Citation
Soil fertility and productivity aspects of alley cropping leucaena leucocephala and cassia siamea under semiarid conditions at Machakos, Kenya

Material Information

Title:
Soil fertility and productivity aspects of alley cropping leucaena leucocephala and cassia siamea under semiarid conditions at Machakos, Kenya
Creator:
Adan, Bashir Jama, 1957-
Publication Date:
Language:
English
Physical Description:
xvi, 267 leaves : ill. ; 29 cm.

Subjects

Subjects / Keywords:
Forest Resources and Conservation thesis Ph. D
Dissertations, Academic -- Forest Resources and Conservation -- UF
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1993.
Bibliography:
Includes bibliographical references (leaves 247-266).
General Note:
Typescript.
General Note:
Vita.
Funding:
Electronic resources created as part of a prototype UF Institutional Repository and Faculty Papers project by the University of Florida.
Statement of Responsibility:
by Bashir Jama Adan.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
The University of Florida George A. Smathers Libraries respect the intellectual property rights of others and do not claim any copyright interest in this item. This item may be protected by copyright but is made available here under a claim of fair use (17 U.S.C. §107) for non-profit research and educational purposes. Users of this work have responsibility for determining copyright status prior to reusing, publishing or reproducing this item for purposes other than what is allowed by fair use or other copyright exemptions. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder. The Smathers Libraries would like to learn more about this item and invite individuals or organizations to contact the RDS coordinator (ufdissertations@uflib.ufl.edu) with any additional information they can provide.
Resource Identifier:
030180528 ( ALEPH )
30335723 ( OCLC )
AJZ0674 ( NOTIS )

Downloads

This item has the following downloads:


Full Text













SOIL FERTILITY AND PRODUCTIVITY ASPECTS OF ALLEY
CROPPING LEUCAENA LEUCOCEPRALA AND CASSIA
SIAMEA UNDER SEMIARID CONDITIONS
AT MACHAKOS, KENYA














BY

BASHIR JAMA ADAN


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


1993

































To my late father, Jama Adan Derow, who passed away in November 1992, while I was in Gainesville writing this dissertation.














ACKNOWLEDGMENTS
The author would like to express his sincere thanks to Dr. P.K.R. Nair for his constant advice, patience, sincerity, and moral support during the course of this project. This research certainly would have been more difficult if it were not for the support and encouragement that Dr. Nair extended on many difficult occasions. The author also would like to thank the other members of the supervisory committee, Drs. H.L. Gholz, C.P.P. Reid, P.E. Hildebrand, and H.L. Popenoe. Special thanks are due to Dr. M.R. Rao of the International Centre for Research in Agroforestry (ICRAF), Nairobi, for his field supervision, support and suggestions regarding the dissertation. In addition, thanks are due to Peter Kurira Wambugu, manager of ICRAF Research Station, Machakos, Kenya, for supervising field activities and data collection, especially when the author was away in 1989-1990.

Appreciation is extended to ICRAF, the Ford Foundation and the Rockefeller Foundation, who have provided financial support for this research; and the National Dryland Research Station (Katumani), Kenya, which has provided laboratory facilities; and to ICRAF once again, for granting leave-ofabsence to the author during the study period.


iii









Special thanks are given to Richard Coe of ICRAF and Steven Linda, Department of Statistics, University of Florida, for their suggestions and assistance in the analysis of the experimental data. Thanks are also due to forestry/agroforestry students Paramu Mafongoya, Todd Johnson, Reinhold Muschler, Paul Campell, Loice Omoro, Mutiah Govindarajan, Don Mee, Mark Follis, Chris Latt, Antonio Brunori, and Anne Todd-Bockerie and to Christine Kelly of agronomy department.

The secretarial services provided by Marie Kimenye of ICRAF during the period of field research are highly appreciated. Furthermore, appreciation is extended to Pam White for her assistance in the preparation of this manuscript.

Finally, the author wishes to express his deepest

appreciation for the support, understanding and sacrifice made by his wife Fatuma (Nimoo), daughters (Aisha and Khadija), father (who passed away in the while the author was writing-up the dissertation--"rahimuallah"), mother, brothers, and sisters.















TABLE OF CONTENTS

Pacre

ACKNOWLEDGMENTS . . . . . ii

LIST OF TABLES . . . . . . . vii

LIST OF FIGURES . . . . . . . . . x

ABSTRACT . . . . . . xv

CHAPTERS

1 INTRODUCTION. . . 1

2 PRODUCTIVITY ASPECTS OF ALLEY CROPPING
WITH LEUCAENA LEUCOCEPHALA AND CASSIA SIAMEA
IN SEMIARID TROPICS AT MACHAKOS, KENYA . . 11

Introduction . . . . . . . . . . . . . 11
Materials and Methods .15
Results . . . . . . . . 30
Discussion . . . . . . 57
Conclusions . . . . . 73

3 SOIL FERTILITY ASPECTS OF ALLEY CROPPING
WITH LEUCAENA LEUCOCEPHALA AND CASSIA
SIAMEA IN SEMIARID TROPICS AT MACHAKOS,
KENYA . . . . . . . . . . . . 75

Introduction . . . . . . . . . . 75
Materials and Methods . . . 78
Results . . . . . . 83
Discussion . . . . . . . . . . 86
Conclusions . . . . . 95

4 DYNAMICS OF SOIL FERTILITY CHANGES: RATES OF
DECOMPOSITION, N-MINERALIZATION AND UPTAKE
BY MAIZE FROM MULCH OF LEUCAENA LEUCOCEPHALA
AND CASSIA SIAMEA UNDER SEMIARID CONDITIONS
AT MACHAKOS, KENYA . . . . . . 97

Introduction . . 97
Materials and Methods . . . . 100
Results . . 110
Discussion . . . . . . . . 132
Conclusions . . . . . . . . . . 145



















5 INTERACTIONS BETWEEN MAIZE AND HEDGEROWS
OF LEUCAENA LEUCOCEPHALA AND CASSIA SIAMEA
IN ALLEY CROPPING SYSTEM IN SEMIARID TROPICS
AT MACHAKOS, KENYA .


Introduction . . . . . . . .
Materials and Methods .
Results . . . . . . . . . . . .
Discussion . . . . . . . .
Conclusions . . . . . . . . . .

6 POTENTIAL FOR FARMER ADOPTION OF ALLEY
CROPPING OF MULTIPURPOSE TREES AND SHRUBS
IN SEMIARID AREAS OF MACHAKOS, KENYA

Introduction . . . . . . . . *
Study area and methodology . .
Results . . . . . . . * . .
Discussion . . . . . . . .
Conclusions . . . . . .

7 SUMMARY AND CONCLUSIONS .

LITERATURE CITED .

BIOGRAPHICAL SKETCH . .


. 147 . 149 . 158 . 183 . 195



* 197

* 197 . 201
204
. 226 9 238

. 241

. 247

* 267


Pace


147















LIST OF TABLES


Page


2-1. Mulch yield of Leucaena leucocephala and
Cassia siamea intercropped hedgerows or
planted in sole stands outside the
cropped area . * . . . . 0.


2-2. Concentration and amounts of nutrients in
Leucaena leucocephala and Cassia siamea
mulch from alley cropped hedgerows .

2-3. Effect of alley cropping and block planting
systems of Leucaena leucocephala and
Cassia siamea on maize grain yield
during six cropping seasons . . .

2-4. Effect of alley cropping and block planting
systems on relative grain yield of maize .


* . 39


2-5. Land equivalent ratios (LER) of alley cropping
Leucaena leucocephala and Cassia siamea compared with the block planting system
with same ratios of tree and crop land
occupancy . * * . . . . . . . . . . . . . 48

2-6. Nitrogen budgets, utilization efficiencies
and yield of maize grown in alley cropping systems of Leucaena leucocephala and Cassia
siamea at a hedgerow spacing of 6.7 m . . . . . . 50

2-7. Addition of phosphorus through mulch and
removal by maize in six seasons of cropping
under Leucaena leucocephala and Cassia
siamea alley cropping . . . . . . . . . . . . . 52

2-8. Fine root density at two soil depths along a
meter-long transect from 6 m apart hedgerows of
Leucaena leucocephala and Cassia siamea
intercropped with maize. . . . . . . 55

3-1. Chemical and physical characteristics of
the top soil (0-20 cm) at the beginning of
the study, October 1987 . . . 79


vii


* . 0










Page
3-2. Chemical characteristics of the dry mulch
of alley cropped hedgerows of Leucaena
leucocephala and Cassia siamea . . . . . . . . 81

3-3. Chemical and physical characteristics of
the top soil (0-20 cm) under alley cropping
systems with hedgerows of Leucaena leucocephala
and Cassia siamea in November, 1991 . . . . . . 84

3-4. Changes in soil Mehlich-l P levels after
three years of alley cropping hedgerows of
Leucaena leucocephala and Cassia siamea with
maize, November 1991 . . . . . . . 85

3-5. Concentration of phosphorus in the mulch
(leaves plus twigs) of alley cropped
hegderows of Leucaena leucocephala and
Cassia siamea at the start (1987) and
end (1991) of the study . . . . 87

3-6. Relative indices of KCl-extract N from
ion-exchange resins incorporated into the
soil of alley cropped hedgerows of Leucaena
leucocephala and Cassia siamea with maize
during the "short-rains" cropping season
of 1991 . . . . . . . . . . . . . . . * * . . . 88

4-1. Chemical and physical characteristics of the
mulch of alley cropped hedgerows Leucaena
leucocephala and Cassia siamea . . . . . . . . 106

4-2. Intercepts, slopes, spline points and
coefficients of determination (r2) of the regression lines for phase one and phase
two of leucaena and cassia mulch
decomposition . . . . 11

4-3. Decomposition parameters of soil incorporated
and surface placed Leucaena leucocephala and
Cassia siamea twigs placed outside the
litter bags, second season of 1991 . . . . . . 118

4-4. Effects of soil incorporated Leucaena
leucocephala and Cassia siamea mulch on
the biomass yield, N-concentration, N-yield and apparent N recovery of maize planted in
pots, 1991 . . . . . . . . . . . . . . . . . 128


viii










Pacre
4-5. Relative indices of cumulative NH4- N, NO3- N
and total N mobilized 16 weeks after soil
incorporation of 2 t ha-I equivalents of Leucaena leucocephala and Cassia siamea
mulch in pots with maize, 1991 . 0 . . . . . . 131

5-1. Nitrogen content of 30 day-old maize shoots
(leaves plus stalks) sampled near (0.45 m) and
away (0.45 mm) from hedgerows of Leucocephala leucocephala and Cassia siamea, fifth season,
1991 . . . . 164

5-2. Relative water content of flag leaves of
maize plants sampled near (0.45 m) and
away (3.5 m) from alley cropped hedgerows
of Leucaena leucocephala and Cassia siamea,
on April 26, 1991 (sixth season) . 165

5-3. Water content of whole maize plants sampled
near (0.45 m) and away (3.5 m) from alley
cropped hedgerows of Leucocephala leucocephala
and Cassia siamea on April 26, sixth season,
1991 . . . . . . . . . . 166

5-4. Relative water content of flag leaves of maize
plants sampled near (0.45 m) and away (3.5 m)
from alley cropped hedgerows of Leucaena
leucocephala and Cassia siamea leucaena and
cassia on May 28, 1991 (sixth season) o . . . . 168

5-5. Relative water content of flag leaves of maize
plants sampled near (0.45 m) and away (3.5 m)
from alley cropped hedgerows of Leucaena
leucocephala and Cassia siamea on January 6,
1992 (seventh season) . . . . . . . . . . . . . 169

5-6. Saturated hydraulic conductivity of soil near
(0.45 m) and away (3.5 m) from alley cropped hedgerows of Leucaena leucocephala and Cassia
siamea, March 1992 (season seven) . . . . . . . 173

5-7. Effects of three nitrogen rates and two
irrigation levels on the biomass yield of
maize after two months growth in pots, 1991 . . 184

6-1. Farm and household characteristics of the 60
farmers interviewed at Machakos, Kenya,
August 1991 . . . . . . . 205










Pacre


6-2. Interests of farmers in alley cropping, farm
size and number of members in a household
at Machakos, Kenya, August 1991 . . . . . . . 207

6-3. Factors which influenced significantly
interests of farmers in alley cropping for
fodder or for mulch and the frequency of
farmers under each factor, Machakos, Kenya,
August 1991 . . . . . . . . . . . . . . 216

6-4. The p-values of dependency of resource
possession, their use or purchase on the presence
of coffee on the farm, Machakos, Kenya,
August 1991 . . . . . . . . . .217

6-5. Estimates of maize yield (with or without
inputs) of farmers with interests in fodder
or mulch aspects of alley cropping, Machakos,
Kenya, August 1991. . . . . . . . . 219

6-6. Functions of tree nurseries possessed by
30% of the 60 farmers, 1991 interviewed,
Machakos, Kenya, August 1991 . . . . . . . . 224

6-7. Reasons given by farmers without nurseries
on their farms, Machakos, Kenya, August 1991 . 225















LIST OF FIGURES


Page
2-1. Seasonal rainfall and the local names
(long-rains and short-rains) of the
cropping seasons during the study . 16

2-2. Mulch yield of Leucaena leucocephala and
Cassia siamea in alley cropping and
block/sole planting systems . . . 33

2-3. Effect of three hedgerow spacings on maize
grain yield, averaged over the yields
under the two species, Leucaena leucocephala
and Cassia siamea . . . 37

2-4. Effect of sole stands of Leucaena leucocephala
and Cassia siamea planted at three land
occupancy on maize yield during six seasons . . 38

2-5. Relationship between rainfall and maize mean
yield of all treatments in each of the six
cropping season . 42

2-6. Maize grain yield under alley cropping and
block planting systems of Leucaena leucocephala
and Cassia siamea expressed as percent of the
sole maize control yield . . . 44

2-7. Maize productivity (Kg m-2) under alley cropping
Leucaena leucocephala and Cassia siamea hedgerows
for six seasons . . . . . . 45

2-8. Productivity of maize (kg mm" of rainfall)
under alley cropping of Leucaena leucocephala and Cassia siamea hedgerows for six cropping
seasons * . . . . . . . . * . . . 46

2-9. Fine root density in the top 40 cm soil profile
of alley cropped hedgerows of Leucaena
leucocephala and Cassia siamea . . . . . . . . . 54

2-10. Effect of three mulch rates of Leucaena
leucoc-:-hala and Cassia siamea on biomass
yield of maize grown in pots, 1991 . . . . . . . 56










Pace


4-1. Monthly rainfall and mean temperature during
the study, 1991 . . . . . . . . . . . . . 101

4-2. Decomposition patterns of the leaves of
Leucaena leucocephala and Cassia siamea . 113

4-3. Decomposition patterns of the twigs of Leucaena
leucocephala and Cassia siamea . . . . . 117

4-4. Decomposition patterns of the mulch (leaves
plus twigs) of Leucaena leucocephala and
Cassia siamea . . . . . . . . . . . . . . . . 120

4-5. Cumulative percent of N mineralized from soil
incorporated mulch of Leucaena leucocephala and
Cassia siamea after 16 weeks . . . . . . . . 123

4-6. Relationship between the patterns of N release
from the mulch of Leucaena leucocephala and
Cassia siamea and biomass yield of maise
from four consecutive crops of maize sown
in pots and harvested at an interval of
four weeks . . . . . . . . . . 126

5-1. Maize yield response to distance from alley
cropped hedgerows of Leucaena leucocephala
and Cassia siamea, 1990 short-rains . . . . . 159

5-2. Maize yield response to distance from alley
cropped hedgerows of Leucaena leucocephala
and Cassia siamea, 1991 long-rains . . . . . . 160

5-3. Maize yield response to distance from alley
cropped hedgerows of Leucaena leucocephala
and Cassia siamea, 1991 short-rains . . . . . 161

5-4. Bulk density of top 5 cm soil profile
near (0.45 m) alley cropped hedgerows of Leucaena leucocephala and Cassia siamea,
1991 short-rains . . . . . . . . . . . . . . . 170

5-5. Infiltration rates of water into the soil
near (0.45 m) and away (3.5 m) from alley cropped hedgerows of Leucaena leucocephala
and Cassia siamea, 1991 short-rains . . . . . 172

5-6. Soil water content near (0.45 m) alley cropped
hedgerows of Leucaena leucocephala and Cassia
siamea determined at the end of 1991
long-rains . . 174


xii










Page


5-7. Soil water content near (0.45 m) alley cropped
hedgerows of Leucaena leucocephala and Cassia
siamea determined at the end of 1991
short-rains . . . . . . . 175

5-8. Difference in gravimetric water content
(%, w/w) between soil sampled at 0-20 cm
depth near (0.45 m) and away (3.5 m) from
alley cropped hedgerows of alley cropped
hedgerows of Leucaena leucocephala and
Cassia siamea in 1991 . . . . . 177

5-9. Monthly average readings of tensiometers placed
in the soil at 15 cm depths near (0.45 m) and
away (3.5 m) from alley cropped hedgerows
of Leucaena leucocephala and Cassia siamea in the long- and short-rain
seasons of 1991 . . . . . . . . . . 178

5-10. Difference in mean monthly temperature of
top soil (10 cm) measured near (0.45 m) and away (3.5 m) from alley cropped hedgerows of
Leucaena leucocephala and in the long- and
short-rain seasons of 1991 . . . . . . . . . . 179

5-11. Count of fauna in 25 x 25 x 25 cm top soil
monoliths taken from near (0.45 m) and away
(3.5 m) from alley cropped hedgerows of Leucaena
leucocephala and Cassia siamea at the end of
1991 short-rain season . . * . . . . * . . . . 181

5-12. Effects of irrigation and fertilization, in
a combination or alone, on maize grain yield
in the field (no alley cropping with hedgerows
of Leucaena leucocephala and Cassia siamea),
1991 . 0. . . . . . . . . . . . . . . . 182


xiii














Abstract of Dissertation Presented to the Graduate School of
the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy


SOIL FERTILITY AND PRODUCTIVITY ASPECTS OF ALLEY
CROPPING LEUCAENA LEUCOCEPHALA AND CASSIA
SIAMEA UNDER SEMIARID CONDITIONS AT MACHAKOS, KENYA

By

Bashir Jama Adan

May 1993

Chairman: Dr. P.K.R. Nair
Major Department: Forest Resources and Conservation

Soil productivity aspects of alley cropping under

semiarid conditions were investigated in a three-year study at Machakos, Kenya. The objectives were to: (1) determine if alley cropping with leguminous trees and shrubs improves soil fertility and productivity, (2) separate the effects of mulch per se from those of in situ presence of the hedges,

(3) assess the extent of tree-crop interaction patterns, and

(4) determine the major factors that influence farmers' adoption of alley cropping. The study consisted of Leucaena leucocephala and Cassia siamea alley cropped with maize (Zea mays); the woody species were also grown in sole-stand blocks outside the cropped field, their prunings applied as mulch to the cropped area in equivalent land areas.


xiv








Difference in maize yield between alley-cropped plots and the plots receiving mulch from the block plantings showed that growing the hedges in situ had a positive effect on maize yield. Yield advantage of alley cropping over the separate crop and hedge planting'systems, expressed as land equivalent ratios, were 36% and 12% for cassia and leucaena, respectively. Leucaena produced more biomass (4 t ha-1 yr'I) than cassia. (2 ha-' yr-'); also leucaena had higher fine-root density than cassia in the top soil (0-40 cm). Therefore, leucaena was more competitive than cassia, and over the years, maize yield declined by 12% under leucaena whereas it increased by 8% under cassia.

For both species, maize nearer the hedgerows yielded

more than those away from them. Soil water content was also higher near the hedgerows. Chemical and physical properties of the top soil (0-20 cm) did not differ significantly between treatments (with or without mulch or alley cropping) or over time for a given treatment. The quantities of mulch produced by the hedgerows were generally low, and they decomposed quickly (half life of four weeks); these could be th'e main reasons for the lack of significant changes in soil properties.

Farmers were generally more interested in the fodder than the mulch aspects of alley cropping. Because of the initial high investments of inputs such as labor and seedlings that required to establish the hedgerows, alley cropping is more likely to be adopted by relatively








resourceful farmers such as coffee growers than by poorer farmers.


xvi















CHAPTER 1
INTRODUCTION


In many parts of the tropics, semiarid lands that have traditionally not been under arable farming are increasingly being brought under rained agriculture. For example, in Kenya where arid and semiarid lands comprise four-fifths of the country's land area, arable farming has been extended to vast areas of dry or marginal lands during recent years. However, there are some formidable obstacles to food production in many of these areas. In addition to rainfall constraints, the soil's productive capacity is inherently low in many areas. In areas where low level subsistence farming has been followed, fertilizer use and other means of nutrient input to soil are seldom practiced. Consequently, low and declining yields of crops are common. For instance, maize yields less than 1 t ha-1, even in seasons of normal rainfall are typical of farms in semiarid areas of Machakos, Kenya (Nadar and Faught, 1984; Mbogo, 1991).

Under traditional shifting cultivation, which has been the predominant agricultural system in many areas of the tropics (both humid and semiarid), long rest or fallow periods in between short cropping periods have traditionally been used to improve fertility of soils exhausted by cultivation (Nye and Greenland, 1960). However, land

1











f allowing results in satisfactory regeneration of soil fertility only when the length of the fallow is adequately long. The length of the fallow-period or rest-period requirement of a soil depends on a number of factors, such as soil type, climatic conditions and land-use practices (Young, 1989). For example, on Alfisols in semiarid zones, it is estimated that fallow periods of one to two years out of every three years are required even with intermediate levels of inputs. With the increasing population, fallow periods have either been reduced drastically or totally abandoned in many places. Thus, keeping the land in fallow as a means of sustaining soil fertility is no longer a feasible management option.

Use of fertilizers is an option considered by many to increase soil fertility and productivity. Chemical fertilizers are, however, seldom used by many resource-poor,

tropical, subsistence f armers under conditions as in Machakos (Ockwell et al., 1991). Fertilizers are expensive

and/or not readily available (Mbogo, 1991). Application of farmyard manure is another widely adopted practice which, in fact, is suggested to be the principal source of nutrients for crops on most farms in Machakos (Probert et al., 1990). However, the quantity of manure available or applied is often too low to sustain continuous crop production at meaningful yield levels (Ikombo, 1984; Ockwell et al., 1991).












Thus, expansion of arable lands, extended f allows

and/or increased fertilizer use are not viable solutions to the problem of declining soil productivity and poor crop yields in most semiarid parts. There is a need, therefore, for alternative low-input technologies that can intensify and/or sustain production from existing croplands. One technology that has been tried with some success in the humid lowland tropics is alley cropping (also called hedgerow intercropping). It consists of growing food crops in alleys between planted hedgerows of multipurpose trees and shrubs (MPTS), usually nitrogen-fixing ones. The hedgerows are periodically cut back at the beginning of, as well as during, cropping. This prevents shading of the associated crop and the hedge prunings can be a source of nutrients to the crop (Kang et al., 1990). By retaining woody perennials in crop production fields on a continuing basis, alley cropping simulates the role of fallow in soil fertility regeneration in shifting cultivation. It also involves, to some extent, the principle of intercropping of legumes and cereals (e.g., maize and beans), which is a traditional farming practice in Kenya as well as many other African countries.

The most distinctive element of agroforestry

practices, and in particular, alley cropping, is the MPTS. The notion of trees as multipurpose plants is based on their productive attributes (for example, food, fodder, fuelwood











and poles) and their protective or service functions in relation, for example, to soil conservation and soil fertility improvement. Leucaena leucocephala Lam. De Wit (Leguminosae, Mimosoideae) and Cassia siamea Lam. (Leguminosae, Caesalpinoideae) are two MPTS widely used in alley cropping studies in both the humid lowlands and the semiarid tropics (Nair, 1987; Kang et al., 1989; Singh et al., 1989). This is probably so because both species are legumes and it is assumed that soil fertility improvement is among their potential benefits. (In subsequent references to the two species, the generic names leucaena and cassia will be used as common names.)

Leucaena had long been used for shade and soil improvement in tree and coffee plantations in Java, Indonesia, as early as 1900 (Dijkman, 1950). Properties that make leucaena suitable for soil fertility improvement are summarized in Young (1989). They include: high biomass production (10-25 t ha"'yr-1, DM), high nitrogen fixation (100-500 kg N ha- yr-1), high levels of nitrogen in leaves (2.5%-4.0%), and, thus, a high rate of nitrogen return in litter and prunings. The main soil limitation is a reduction in its growth on acid soils (Brewbaker, 1987). Also, attack by the psyllid, Heteropsylla cubana, is currently serious in some regions. Like leucaena, cassia has the capacity to grow on poor soils and has been used in alley cropping trials (Yamoah et al., 1986a), although the









5

extent of its soil-improving potential is little known compared to that of leucaena. There is even doubt as to whether it is nitrogen fixing, and established opinion is that it is not.

In the humid lowlands of Nigeria, many trials have shown that crop yields could be improved under alley cropping with MPTS. For instance, an eight-year alley cropping trial conducted by Kang et al. (1990) showed that using Leucaena leucocephala prunings only (in the range of 8-10 t ha"1 yr'-), maize yield could be maintained at a reasonable level of 2 t ha-i, compared to 0.66 t ha- without leucaena. Additions of leucaena prunings also maintained higher soil organic matter than when no prunings were added. Several other studies in the humid lowlands have also given favorable results from alley cropping (e.g., Yamoah et al., 1986a). As a result, there has been considerable interest in alley cropping, and the technology has been tried in a number of other places including the semiarid tropics (SATs). However, the results of alley cropping have not been positive in all places.

Fundamental to the concept of alley cropping is the

proposition that trees improve soils. There is evidence to support that soil fertility and, to some extent, crop yields in semiarid areas could be improved by the presence of trees on farmlands e.g., Faidherbia albida (synonym: Acacia albida) in West Africa (Bon Kongou, 1992) and Prosopis











cineraria in India (Mann and Saxena, 1980). The various mechanisms, demonstrated or hypothesized but not yet fully proven, are reviewed by Sanchez (1987), Young (1989), and Nair (1990). Some of the processes demonstrated or hypothesized by which trees improve soil fertility include

additions of litter, N2-fixation, nutrient uptake from subsoil and protection from erosion. While trees may improve fertility of the soil beneath them and the yield of crops near them, they will also compete with crops for nutrients and moisture if their root distribution is similar to and in the same soil zone as that of annual crops (Huxley et al., 1977; Ewel et al., 1982; Jonsson et al., 1988). For example, most of the fine, feeder roots of many common trees are found within the 20 cm top soil (Comerford et al., 1984). Besides, roots of the tree may be well established by the time the annual crop is sown. It is this potential for effective competition, particularly for soil resources, that has been suggested as the reason for decreased crop yields under alley cropping in some humid lowlands of the tropics (Evenssen, 1989; Szott et al., 1990) and semiarid tropics (Singh et al., 1986; Nair, 1987; Mittal and Singh, 1989).

To minimize the potential for crop yield losses under alley cropping, there should be benefits of soil fertility and crop pe rformance that can only be realized if the hedges and crops are planted together, i.e., benefits that would be












lost if they were grown separately. These include mostly gradual changes in soil organic matter, nutrient content and physical properties such as bulk density, water infiltration and porosity. The presence of the hedges in situ may, through their root turnover and reduction of soil erosion, sustain soil organic matter (Young, 1989). This may be of particular importance in low-input agricultural systems where all above-ground biomass is removed from the field.

If the contribution of the in situ presence of the

hedgerows to soil fertility relative to that of the mulch is negligible, then planting the hedges separately from crops and then transferring their mulch to crop fields would be a more logical approach than growing the two together. Even then, an understanding of the relative roles of the quantity and quality (i.e., decomposition rates) of the mulch produced would be relevant as this may enhance the efficiency of its use.

In the humid tropics, mulch yields from the hedgerows

are high (in the range of 8-10 t ha-1 yr-1) . Also, the rates of decomposition are rapid. This makes nutrients in the mulch rapidly available, for instance, in about 30 days in the case of leucaena (Budelman, 1988; Gichuru and Kang, 1989). So, the mulch, in essence, resembles inorganic fertilizer that provides a readily available supply of nutrients to the crop. Crop response when mulch is removed is therefore observable within a growing season. In the












SATs, unlike in the humidtropics, mulch yields from alley cropped species are lower than in the humid tropics. For

instance, mulch yields of 2 t ha-1 yr-' from leucaena is typical in SATs (Nair, 1987) as opposed to 8-10 t ha-1 yr-1 in lowland humid tropics. As Huxley (1986) postulated, this low mulch yield in SATs may limit the potential of alley cropping to improve fertility and productivity of soils in these regions.

Information on the yields as well as the rates of decomposition of mulch from alley cropped MPTS (which in turn affects the ability of the mulch to provide nutrients to the crops and to contribute to soil organic matter) are relatively scarce in the SATs. Also, the benefits of the in situ presence of hedgerows to soil fertility and productivity are not well understood in either the humid or the semiarid tropics.

In addition to these technical questions, there is

also the issue of social acceptability of the new technology by the target farmers. It is necessary to determine how farmers will adapt and use the technology under conditions unique to their systems of farming. For instance, in the semiarid areas where pruning from hedgerows as well as crop residues are used as fodder (Singh et al., 1986), the in situ presence of hedgerows may be the main avenue for improving soil fertility. Studies of this nature will provide insights into determining the technical feasibility












as well as the management regime of alley cropping to improve the fertility and productivity of soils in the semiarid tropics. These technical and social questions were the prime motivations in undertaking this study. This study had with four main objectives:

1. To determine if alley cropping with leguminous

MPTS substantially improves soil fertility and

productivity in the semiarid conditions of

Machakos, Kenya.

2. To determine or separate effects due to mulch per

se from those due to the in situ presence of the

hedges within the crop field.

3. To assess the patterns of tree-crop interaction

and to determine factors that may be responsible

for the observed interactions.

4. To determine factors that may significantly

influence adoption of the technology by farmers.

Plant productivity aspects of the crop and hedgerows of leucaena and cassia are presented in Chapter 2. Chapter

3 addresses the effect of the alley cropping and mulch quantity and quality on soil chemical, physical and biological properties; trends over time as well as within a cropping season are discussed. Patterns of N availability determined by means of frequent soil sampling and/or in situ mineralization using ion-exchange resins are also discussed in Chapter 3.










10

In Chapter 4, studies on decomposition, mineralization and uptake patterns of leucaena and cassia mulches are presented. Uptake patterns of N and P from the mulch was assessed by means of monthly harvests and determination of nutrient content of potted maize plants. Chapter 5 addresses crop yield profile at the tree-crop interface. For this, the last three of the seven droppings seasons of the study were used. Also presented in Chapter 5 are the results of field and pot studies of water and nutrient status of soil at the tree-crop interface. These studies were conducted to explain the observed interaction between the hedgerows and the crop. Chapter 6 presents results of a survey of 60 farms and discusses factors that may significantly influence adoption of alley cropping in the semiarid tropics.

Finally, Chapter 7 provides a summary of Chapters 1-6. It highlights the major findings of the study and suggests aspects of alley cropping in the semiarid tropics with potential for further research.













CHAPTER 2
PRODUCTIVITY ASPECTS OF ALLEY CROPPING WITH LEUCAENA LEUCOCEPHALA AND CASSIA SIAMEA IN SEMIARID TROPICS AT MACHAKOS, KENYA


Introduction

Alley cropping, also called hedgerow cropping, is an agroforestry technology where annual crops are grown in the alleys between rows of trees/shrubs that are regularly pruned and maintained as low hedges. The prunings from the hedgerows, when applied to the crops in the alleys, are reported to improve or sustain crop yields. This is suggested by many studies on cropping from the humid lowland tropics. For instance, Kang et al. (1985) reported from southern Nigeria that, on an Alfisol, using leucaena prunings only, maize yield could be maintained at a "reasonable" level of 2 t ha-1 as against 0.66 t ha"1 without leucaena prunings and fertilizer. An extensive review by Kang et al. (1990) of other such studies from the humid tropics showed similar results. Several other studies also show that alley cropping can sustain soil nutrient status and minimize organic matter decline in humid to subhumid regions (Lal, 1989; Yamoah et al., 1986b).

Typically, alley cropping studies involve use of leguminous trees/shrubs. This is because nitrogen availability is commonly (though by no means exclusively) a limiting factor in tropical agriculture. -It is also the 11










element which more than any other is under biological regulation in natural ecosystems (Swift and Sanchez, 1984). Nitrogen available to crops in alley cropping may be enhanced through processes of biological fixation of atmospheric N (Roskoski, 1986; Dommerques, 1987) and subsequent release through processes of nodule sloughing (Sanginga et al., 1986) or excretion of nitrogen from the root nodules (Brophy and Heichel, 1989; Wacquant et al., 1989). However, the significance of the latter process remains doubtful (Vallis, 1978).

Prunings from the hedgerows are the main source of nutrients to the crops. This is evident from the significant declines in crop yield that occur when the application of mulch is withdrawn (Kang et al., 1981). The extent of supply of N from the mulch to the crop would depend, among other factors, on the rainfall conditions of the site, the soil, the hedgerow species, and management practices such as spacing and the pruning regime. In the humid lowlands of West Africa, Kang et al. (1981) and Duguma et al. (1988) reported a wide range of N yields of about 150-560 kg N ha-i yr" for alley cropped leucaena. Similar information from the semiarid tropics is scarce. A major exception is the work of Lulandala and Hall (1990) in Tanzania that noted nitrogen returns through frequent pruning of leucaena hedges could be as high as 88 kg ha-' yr-.









13

One of the impacts of the above-mentioned studies has been to stimulate interest and studies on alley cropping in the SATs. In many areas of SATs, soil fertility is low (Steiner et al., 1988) and chemical fertilization alone (i.e., no irrigation) could improve yields (Ludwig, 1987). Fertilizers are, however, expensive for many resource-poor farmers and/or are not readily available.

In many regions, and in particular the SATs, the

pruning are more likely to be used for fodder than as mulch (Singh et al., 1986; Ong et al., 1991). Even if the pruning are used as mulch for the crops, the effects on crop yields may not be significant (Singh et al., 1989). Under these conditions, the N contribution by the hedgerows to the soil may primarily be through below-ground processes, particularly where N fixing species are used. Whether a significant amount of N can accrue through these processes and whether this N is available to the companion crop have not been convincingly established (Ta et al., 1986; Danso et al., 1987; Rao et al., 19B7).

If growing the trees within the crop fields is not beneficial for the crops but the trees can provide other valuable benefits (e.g., fodder), separate planting of trees might be preferable. This would be particularly appropriate 'in the SATs where many studies report significantly reduced crop yields under alley cropping (Singh, et al., 1986; Mittal and Singh, 1989; Nair, 1987). Under these











conditions, the trees could be grown separately from the crops, and their biomass harvested for use as fodder or mulch in a "cut and carry system."

The magnitude of the effects of in situ planting of the hedgerows would be influenced significantly by differences in the rooting patterns and densities of the hedgerow species. Trees with high root densities (expressed as cm cm-3 of soil volume) in the rooting zone of the crops are likely to result in less positive in situ benefits to crops than trees with low root densities. Also, for a given species, between-row spacing of the hedges, or put in other words, the ratio of tree:crop land occupancy, could modify the effects of in situ planting. Because of the increased chances of competition, hedgerows with narrow between-row spacings would be more likely to have less positive in-situ effects on crops than widely spaced hedgerows. Narrowly spaced hedgerows may, on the other hand, produce higher quantities of biomass and provide greater scope for nutrient recycling than widely spaced hedgerows.

From the foregoing, it is apparent that an

understanding of the in situ versus mulch benefits of growing hedgerows in crops is necessary. Additionally, it is important to have an understanding of the rooting densities of species involved. The objective of this study was to separate the effects of mulch from those of in situ planting of the hedgerows on productivity (biomass and crop











yields) in alley cropping systems of Leucaena leucocephala and Cassia siamea in semiarid Machakos, Kenya.



Materials and Methods

Site Description

The study was carried out under rainfed conditions at the Field Station of the International Centre for Research in Agroforestry (ICRAF) located at Machakos, Kenya (latitude 10 N 33' S, longitude of 370 14' E and 1596 m elevation). Mean annual temperature is 190 C and annual rainfall ranges from 400-1100 mm with a mean of 750 mm. This rain falls in two seasons in almost equal amounts. There are two rainfed cropping seasons for the principal agricultural crops. The first rains of the year during March-June are locally referred to as the long-rains; the second rains of the year during October-January are referred to as the short-rains. Of the two, the short-rains are more reliable than the longrains at Machakos. The seasons and the monthly rainfall during the study period are presented in Figure 2-1. Generally, monthly rainfall equals or exceeds evapotranspiration only in April, May and November.

The soil at the study site is texturally well-drained, dark, reddish-brown sandy clay, derived from the basement complex gneiss. According to the Kenya/FAO (Food and Agriculture Organization of the United Nations), the soils are Luvisols, including the ferric and eutric forms (or























300

LR = Long-rains 250 SR - Short-rains


200
E

- 150 '0

100


50




0 - -- -- I-- - - I ---- 4 I - - -- I- --- I--- I -- -1988 1989 1989 1990 1990 1991 1991
SR LR SR LR SR LR SR
Month and cropping season


Figure 2-1. Seasonal rainfall and the local names
(long-rains and short-rains) of the cropping seasons during the study.









17

Kandic Rhodustalfs according to the United States Department of Agriculture Soil Taxonomy). The soils of the research station are in general weakly to moderately leached, with a weakly acid reaction (pH [water] 6.0-6.5), a medium base saturation (50-80%), low to moderate levels of organic matter (top soil organic carbon: 1%-1.5%), nitrogen (N) (0.08%) and phosphorus (P) (<16 ppm). Other nutrients are reportedly present in adequate levels to support crop production (Kibe et al., 1981).

Clay and sand content in the top 15 cm averaged 28% and 67%, respectively, in 1987. Kaolinite is the predominant mineral in the clay fraction. The top soil had an effective cation exchange capacity of 9.0 meq per 100 g of soil at the initiation of the study in October 1987.



Methods


Field study

Experimental layout.--Hedgerows of Leucaena leucocephala and Cassia siamea were established in October 1987. Alley widths of the hedgerows were 4, 5 and 6.7 m. In-row spacing of trees within the hedgerows was 0.5 m. Alley widths used were computed based on field observations (Nair, 1987) that 3-4 year old hedgerows, spaced 4 m apart and pruned frequently to low heights, occupied 25% of total (aboveground) land area. Similarly, the land occupancy ratio by the hedgerows spaced 5 and 6.7 m would be 20% and












15%, respectively. Sole stands of leucaena and cassia occupying 15%, 20% and 25% of crop land were also established for comparison to the hedgerow intercropped plots. In-row spacing of the trees within the block/sole stands of hedgerows was 1.0 m by 1.0 m. Control plots included in the experiment were (a) maize crop which received mulch only in amounts equivalent to what was produced by the hedgerow intercropped plots and, (b) maize crop which received no mulch but fertilizer only at the

rates of 40 kg and 17 kg of N and P ha-' season-'
The experimental design was split-plot, replicated three times. Main plot treatments, randomized within the blocks, were the hedgerow species (i.e., leucaena or cassia). Subplot treatments were the factorial combination of the three ratios of crop to tree land occupancy and the control plots with mulch-only. The fertilizer-only control plot was included in each main plot (i.e., 2 replications per block). This gave 6 replications for the fertilizeronly control. Main plot size was 20 x 50 m, with two meter borders separating sub-plots of 5 x 20 m. Between blocks, a 4-meter wide path was provided. Sole stands of leucaena and cassia hedges which provided mulch for the mulch-only control plots were also included in each block.

'Initially, a control plot with no inputs was

inadvertently left out. The control plot of an adjacent experiment, established about the same time as this study










19

was used instead. Nevertheless, since this control plot was outside the setting of the present study, it was not included in the statistical analyses. Rather, it was used only for relative comparisons.

Root studies.--Root studies were initiated in the last cropping season of the study with the objective of identifying the cause of differences in the effects of the two hedgerows species on maize yield. These studies were conducted on hedgerows outside the main experiment consisting of two blocks of cassia and leucaena hedgerows,

6 m wide, established in 1983. Until 1988i maize and beans were sown in rotation in the alleys between the hedgerows in these plots. The maize crop was fertilized with N and P at the rate of 40 and 18 kg ha-' season-'; the beans were not. In addition, pruning from the hedgerows were applied to the cropped alleys. In 1988, fertilizer application was stopped and maize was sown in both seasons.

For each species, blocks of soil 20 x 20 x 20 cm were removed from six transects 3 meters long from the base of the hedgerows during the 1992 short-rains cropping season. The soil blocks were taken from 0-20 and 20-40 cm soil depths. The samples were brought back to washing sheds in plastic bags. The soil blocks were then washed with a spray of water aided by hand manipulation to separate medium-sized

and fine roots from the 'soil in sieves with mesh sizes of 5 mm and 0.5 mm, respectively. Roots less than 2 mm in











diameter were classified as fine roots, those of 2-5 mm as medium-sized or small roots (Bohm,.1979).

The fine roots of leucaena and cassia were

distinguishable easily from maize roots based on color and texture. Maize had whitish and succulent roots, cassia had black roots, and leucaena had fine, cream-colored roots, which were strong relative to cassia. The majority of the fine roots of the two species (which were planted separately) and the maize roots were sorted in the same way. Some were hard to differentiate and were separated under a magnifier. Dead roots with brittle steel and degenerated epidermal and cortical tissues were not included.

Root length was determined by the line-intercept

method of Tenant (1975) using a tray with grid units of 1.3 cm (half-inch). With these grid units, the total number of intercepts equals the total length of root in centimeters.

Crop and tree rnanagement.--The hedgerows occupied 15%, 20% and 25% of the gross land area. The remainder of the land, viz, 85%, 80% and 75%, was sown to maize at approximately 37,000 plants ha-' with a spacing of 90 cm between rows and 30 cm within rows. Katuinani Composite, a drought-tolerant maize variety was used. The crops and trees were only rain-fed.

Fertilizer was applied at the rate of 40, 18 and 0 kg ha-' season-' of N, P and K, respectively, to maize in the fertilizer-only control, alley cropped maize and maize next











to the block/sole planted hedgerows. The mulch-only plots received no fertilizer.

The hedgerows were pruned 2-3 times in a cropping

season (depending on the rainfall conditions of the season) to 50 cm height aboveground. First pruning was done in November 1988, this being the first of the seven cropping seasons with the treatments applied. At each pruning of the hedgerows, which typically took place 1-2 weeks before the onset of rains, the biomass harvested (hereafter referred to as mulch) was determined. It was then chopped into small pieces, samples taken for dry matter determination and the rest spread on the field. Typically, mulch harvested before the onset of rains was incorporated into the soil during field preparation and planting of the crops. However, mulch from later pruning during the life of the crop was either partially incorporated (in the process of weeding the crops) or generally left on the soil surface. In all crops, the plots were hand tilled and weeded.

At each harvest of maize, which took place

approximately 120 days after sowing, the yields of the grain, cobs and stalks were determined. Samples of each

component were oven-dried at 105*C for 48 hours and dry weight determined. Maize stalks and cobs were not returned to the plots, in line with the farmers, practice in the study region where maize stalks are a valuable livestock feed.









22

In cropping seasons 1-4, yields of all maize rows in a plot were composited at harvest. However, in seasons 5, 6 and 7, maize yields of each row in a plot was separately harvested and recorded. Also, in the last three cropping seasons, sub-samples of the grain, cobs and stover of some selected treatments, as explained later, were analyzed for nitrogen and phosphorus concentrations. Nitrogen was determined by the standard micro-kjeldahl method and P by UV visible spectrophotometry. In all seasons, two rows of maize on all four sides of the plot were left as guard rows.



Pot Study

A pot study was initiated in January 1992 to determine the amounts of mulch that would produce significant effects on maize yields under a given rainfall condition. The treatments included a factorial combination of 4 rates of leucaena and cassia mulch and 2 levels of irrigation. The four mulch rates were applications equivalent to 0, 2, 4 and

8 t ha-1 season-', dry matter. The two irrigation levels were the equivalents of 200 and 400 mm received in the last two cropping seasons of the study. The pots were arranged in a randomized block design with 3 replications in a transparent polythene-roof house. In each block, pots with one level of fertilizer at a rate equivalent to 40 N and 17

P kg ha-1 season-' were also included.











Soil put into the pots (approximately 20 kg) was

collected from three randomly selected positions near (0.45 m) and away (3.5 m) from the alley cropped plot of the 6.7 m wide hedgerows. The amounts of irrigation water applied to the pots were based on the difference between soil water storage at field capacity (15%) and wilting point (7.8%) determined at the initiation of the study. The plots were irrigated twice weekly.

Three seeds of maize were sown in each pot. Three days after germination, the seedlings were thinned to one per pot. Total plant harvest (i.e., shoot plus roots) was

done after four weeks. After oven drying at 65*C for 48 hours, dry mass was determined. on the same day as the harvest, the pots were reason for another four weeks. This repeated sowing and harvesting from the pots was performed for a total period of 16 weeks.



Yield Computation-Crop and Hedge Biomass from Field Study

Biomass and maize yields of the hedgerow planting

systems (expressed in t ha-1, dry weight) were computed to include area occupied by both the hedgerows and the crop. Hence, the yields are expressed on gross as opposed to net area basis or area occupied by crop only. In situ effects of the hedgerows were calculated as the difference in mulch and crop yields of the systems where hedgerows were












intercropped and where they were planted in sole/block stands outside the crop area.

To estimate in situ effects of the intercropped

hedges, the block-planted hedgerows should not affect the adjacent crop. In an on-farm situation, the block-planted trees could be on land not fit for cropping that is at some distance from the crop land. In on-station situations, distant planting may not be feasible, especially under conditions where treatments are randomly located in small plots. Two approaches, polythene sheets vertically inserted into the soil to 50 cm depths (Ong et al., 1991) and trenching (Verinumbe and Okali, 1985) have been used to minimize effects of adjacent rows or blocks of trees on crops. The effectiveness of these approaches depends on the depths of insertion of the polythene sheets into the soil or of trenching. Tree roots can go deep below the polythene sheet or the trench and resurface (Corlett, 1989; Ong et al., 1992).

In the present study, row-by-row harvest of the crop of the widest alley (6.7 m) was used to determine the distance of influence of the block-planted hedgerows on the yield of the adjacent crop. A graphical plot of the yield of the rows against the distance of the rows from the hedgerows was then made. From this plot, the rows of maize influenced by the block-planted hedges were determined and discarded. The yields of the remaining rows were used to










25
calculate the per hectare yield of maize planted next to the block planting system of hedges outside the crops.

In the section on results below, only yields of the

second season to the sixth season crops are discussed. The yield of the first season was excluded because the hedgerows were pruned only once (since they were not yet wellestablished). In the other seasons the hedgerows were pruned 2-3 times. Also, the crop yield of the first season crop was at least twice as much as the yield of the other seasons, ascribed more to the initial site conditions than to the effects of treatments. Detailed presentation and discussion of the biomass and crop yields are limited to the 6.7 m wide alley in the block planting system. This spacing of the alleys and ratio of tree:crop land occupancy offers the most land for cropping. This is of practical significance to farmers.



Trends in Maize Yield over Time

Trends in maize yield over time were analyzed by expressing the yield of each treatment in a season-as a percent of the average yield of all treatments of the.second crop. The patterns of maize yield over time were also examined by expressing the yields of the two maize planting systems as a percent of the yields of the fertilizer-only control treatment.












The productivity of alley cropping compared to the sole planting system of the hedgerows and the crop was assessed by calculating land equivalent ratios (Willey, 1979) based on grain and total biomass yield of grain, stalk and the hedgerows. The block planting and alley cropping, systems with the same ratio of land apportioned to tree and crop (15:85) was used.



Nutrient Budget of Nitrogen and Phosphorus

A simple budget for N and P was computed for selected treatments during the last three cropping seasons of the study. It was only in these seasons that nutrient content of the crop was regularly determined. The N budget was calculated by considering N applied through the mulch or as fertilizer, N removed in crop (grain, cob and stover), and N still on the plot in mulch not decomposed. The P budget was similarly calculated except that P remaining in mulch not decomposed was not computed because sufficient amounts of mulch did not remain for determination of both N and P.

Nitrogen and P removed from the plots were calculated by summing the N and P removed by the crop (i.e.f grain, cobs and stover) for the last three seasons of the study. Since N and P data for the crop were available for only three out of the seven crops regression equations relating N and P removed in crops 5-7 to yields in these seasons were determined. These equations were then used to calculate N












and P removed in crops of seasons 2-4. The regression equations were obtained first by multiplying the N and P concentrations of the crop (grain, cobs, stover) by their respective yields to obtain N and P yields of the crop. The N and P values were then regressed against the crop yields which resulted in the following equations: N(kg) grain yield =0.015 * grain yield - 0.0006; r2 =0.98 N(kg) cob yield =0.0028 * cob yield + 0.0002; r20. 95 N(kg) stover yield = 0.0056 * stover yield + 0.0008; r2=0.91 Similarly, regression equations for P yield in grain, stover and cob yields are as follows:

P(kg) grain yield =0.0018 *grain yield + 0.0002; r2-0.93 P(kg) stover yield =0.0021 *stover yield -0.001; r'=0.92 P(kg) cob yield =0.0005 * cob yield - 0.00002; r2=0.96 Nitrogen remaining in mulch not decomposed was calculated using the N concentration of the soil-incorporated mulch at the end of the study in March 1992. Typically, N remaining in mulch is calculated using the decomposition constant of the mulch (Palm, 1988). However, where the pattern of decomposition of the mulch has more than one phase as was the case in the present study (Chapter 4), use of one decomposition constant for both phases may not be appropriate. Nitrogen remaining was calculated for each season's mulch input into the soil and then summed over all the seasons.












Apparent nitrogen utilization efficiencies were

calculated for crops seasons 2-6 and includes grain, cobs and stover. The efficiencies were calculated by determining the nitrogen yield of the treatments where mulch was applied to the crop (dry matter x nitrogen concentration) less nitrogen yield of the no fertilizer control treatment and expressing this as a percentage of the amount of nitrogen via the mulch or the fertilizer. This method has also been used by others to estimate utilization efficiency of mulch N by crops (Kang et al., 1981; Yamoah et al., 1986b; Palm, 1988).



Statistical Analyses

The effects of the two species and systems of hedge planting on crop and biomass yields were determined using the General Linear Models (GLM) procedure in the Statistical Analysis System (SAS, 1992). Because of the unbalanced design caused by the fertilizer-only treatment (which was applied at only one level as opposed to other treatments which were applied at three levels,), the analysis was performed in two stages. First, the fertilizer-only treatment was omitted and the analysis of split-plot design was performed. Second, depending on what factors were significant in the split-plot analysis of variance (ANOVA) test, comparisons with controls were performed in a subsequent ANOVA that included all treatments. In both











analyses, differences between means of treatments were determined by pairwise contrasts of the Least Squares Means (LSM). A statistically significant difference was declared to exist between means of any pair of treatments when the pvalue was less than or equal to 0.05.

Yield response of the systems over seasons was also analyzed with the GLM/SAS procedure. The option for repeated measures ANOVA was included in the GLM procedure to account for the systematic effects of harvests from the same plots over time. The difference in mean yield of treatments between seasons was determined by pairwise contrasts of the LSM.

The GLM/SAS procedure was also used to analyze the root density data. The option for repeated measures analysis was included to account for the effects of "fixed" distances of root observations from the hedgerows. The significance of differences between means of root densities at various distances from the stumps of the hedgerows was determined by pairwise contrasts of the least squares means.

In the pot studies also, the GLM/SAS procedure was

used to analyze the data. As in the field study, the option for repeated measures was included in the ANOVA to account for the effects of repeated harvests from the same pots. In addition, orthogonal polynomial contrasts for linear, quadratic and cubic effects were included in the GLM procedure to test significance of the relationships between










30

irrigation levels, mulch species and mulch rates. Pairwise comparison of LSM was used to test the significance of differences between treatment means. Dunnet's test, which controls the Type 1 experiment error for comparisons of all treatments against a control, was used to determine significance of differences between the mulch rates and the control pots that had only one level of fertilizer application.


Results


Field Study


Mulch yield

Leucaena, intercropped in hedgerows, produced

significantly higher mulch (+25%) than when the species was planted in sole/block stands (Table 2-1). For cassia, however, the +14% observed difference was not significant.

Mean mulch yield of leucaena, whether it was intercropped or planted in sole/block stands, was significantly higher (+46%) than that of cassia (Table 2-1). Between the three hedgerow spacings of the intercropped system, mulch yield of the 4 m hedgerows was significantly (p = 0.04) higher (on average, 21%) than that of the other two hedgerow spacings. The mulch yields of 5 and 6.7 in-wide hedgerows were not significantly different (p = 0.23).









31


Table 2-1. Mulch yield of Leucaena leucocephala and Cassia
siamea intercropped hedgerows or planted in sole
stands outside the cropped area.


t ha-I yrHedgerow Alley Block
Species Cropping Planting* Mean


Leucaena 4.4" 3.3b 3.9a

Cassia 2.2 a 1.9a 2.ib

Mean 3.3 2.6 3.0


*Block planting in an area equal to that estimated to be occupied by the hedgerows in an alley cropping system. a,'bFor a given species, means followed by the same letter are
not significantly different (p = 0.05).












In the first year, there was no significant difference between mulch yields from the two species, but after the second year, leucaena yields were significantly higher than those of cassia (Fig. 2-2). Also, a significant difference was observed in the response of the two species to rainfall.

While changes in rainfall had a significant effect on the mulch yield of leucaena, no effects were observed for cassia. Mulch yield of leucaena peaked in 1990 when the annual rainfall was highest (964 mm) during the study period. In 1991 when rainfall dropped to only 22% of the 1990 total (Fig. 2-1), mulch yield of leucaena dropped by half. For cassia, there was no significant response (Fig. 2-2). Cassia mulch yield tended to decline over time, although the differences in yield between years were not significant. Although the effect of rainfall on mulch yield was more pronounced than the effect of age of the hedgerows, given the same rainfall, mulch yield of leucaena did increase with age. For instance, with similar seasonal rainfalls total of 330 mrm, mulch yields of leucaena hedgerows were 40% more in the fifth season than in the second season. Similar trends of increase in cassia mulch with age were not detectable.

Where hedgerows were planted as sole stands/blocks outside the crop fields, differences in mulch yields per hectare among the three tree:crop ratios were not
























3.5 Leucaena
Intercropping

3.0 Leucaena
3.0 Sole

D Cassia
2.5 2 .lIntercropping
Cassia
J 2.0- XX Sole
--- 2.0 7 .


1.5 b

75
1.0
C
C

0.5 - x


0
1989 1990 1991
Year






Figure 2-2. Mulch yield of Leucaena leucocephala and Cassia siamea in alley cropping and block/sole planting systems.










34

significant. However, leucaena always yielded significantly more mulch than cassia.

A count of all plants in the three replicates in the last season of cropping showed that cassia had a higher percentage of dead plants than leucaena (using the number of seedlings planted in 1987 as a base). In cassia, up to 5% (t 7% standard error) of the intercropped hedges and 11.6% (t 15% standard error) of the block planted plants were dead. In contrast, no plants were recorded as dead in

either leucaena systems. Cassia's mortality was largely attributed to a commonly observed stem borer, Xylentes capensis (Cassidae) and a wood boring beetle, Xyloposus sp. (Bostrychidae).



Nutrient concentrations and contents of the mulches

In addition to a supply of N and P, the mulch provided substantial quantities of other nutrients to the soil (Table 2-2). While concentrations of the other nutrients remained stable over time in both species, the concentration of P increased considerably. For instance, the levels of P in cassia mulch increased 25% from 1987 to 1991 (compare 0.20 to 0.25%).



Crop yields

The combined analyses of the data over seasons

indicated that there were significant interactions between













Table 2-2.


Concentration and amounts of nutrients in Leucaena leucocephala and Cassia siamea mulch from alley cropped hedgerows.


Nutrient concentration (%) and amounts (kg
ha-' yr"1, in parenthesis)


Species N P K Ca Mg S


Leucaena 3.5a 0.2a 1.9 0.7a 0.3a 0.1a

(136.0) (7.5) (71.7) (26.4) (11.3) (3.8) Cassia 2.9a 0.3" 1.8a 0.7a 0.2a 0.1a

(54.7) (5.7) (33.7) (13.2) (3.8) (1.9)

*Within a column, values of nutrient concentration with different letters are significantly different (p = 0.05). The amounts of all nutrients in leucaena mulch were significantly higher than those in cassia (p = 0.05).












seasons, systems, and hedgerow species on crop yield. Therefore, ANOVA by season was necessary in order to understand the true main effects of species and systems.

The ANOVA by season showed that in four out of the six

(i.e., seasons 2, 4, 5, and 6), alley cropped maize yielded higher with cassia than with leucaena (Table 2-3). A significant difference between the yield of leucaena alley cropped maize and that next to the block planted system of leucaena was observed in only two out of the six seasons (the third and the fourth). For cassia, the difference was also significant in two out of the six seasons (the second and the third). The difference in maize yield under the two systems represents the in situ effects of the hedgerows on crop yield. Averaged over the seasons, the in situ effects of cassia on maize yields were 22% while those of leucaena were 9%.

Except in the second season, differences between the effects of the three hedgerow spacings (i.e., 4, 5, and 6.7 m) on maize yield were not significantly different (Fig. 2-3). In the second season, maize yields under the 6.7 m hedgerows were significantly higher than those of the other two spacings. In the block planting systems, significant differences in maize yields were observed in seasons 2 and 4 only and 4 only (Fig. 2-4).

A comparison of the mean yields of all the treatments (Table 2-3, last column) shows that the fertilized maize



























cc
4-I




N Co


2.5




2.0 1.5


1.0 0.5


R


-


3
Cropping


4m

--W
5mr


6.7m
sea&@*


4
Season


Figure 2-3. Effect of three hedgerow spacings on maize grain yield, averaged over yields under the two species, Leucaena leucocephala and Cassia siamea. Vertical bars indicate standard error of difference of means.









Percent land under tree and crop

F2575


20:80 15:85

$ Is ve


1 2 3 4 5 6
Cropping Season


Figure 2-4. Effect of sole stands of Leucaena leucocephala and Cassia siamea planted at three land occupancy on maize yield during six seasons. Vertical bars indicate standard error of difference of means.


4.0 3.5 4 3.0 .C 2.5


N
* 2.0


1.5 1.0











Table 2-3. Effect of alley cropping and
during six cropping seasons.
row.


block planting systems of maize grain yield Rainfall for each season is included in the last


Maize yield (t ha-) for each season


System 2 3 4 5 6 7 Mean


Alley cropping (leucaena) 2.5a', 3.0ab 2 b 2. 3"b 0.6&b,c 2.3a 2.2ab

Alley cropping (cassia) 2.4 a' 3.9d 2.4b 3.4a 1.4a 2.8c 2.7c

Block planting (leucaena) 2.5' 2.7b' 2.0c 2.6b 0.5c 1.7b 2.0a

Block planting (cassia) 2.3ab 20.bc 2.Oc 2.9"ab 09abc 1.5b 2.1ab

Control (leucaena mulch) 2. 0,,b 2. 6bc 2 2abc 1. 6d 0.8abc 1. 7b 1 . 8b Control (cassia mulch) 1. 9ab 2. Y 2. 1abc 2. 1cd 0. 7abc 1.8b 1.8a

Control (fertilizer) 1.6b 3.1l,b 2.9a'b 3.1l'b 1.3a'b 2.7a 2.5c

Control (absolute)* 2.0 2.9 1.8 2.1 1.1 2.4 2.1

Mean 2.1 2.8 2.1 2.4 0.9 2.0 2.1

Rainfall (mm) 330 441 631 333 214 373 336


a'b'cdWithin a season, values followed by different letters are
(p = 0.05).

*From an adjacent experiment.


significantly different












(control) yielded significantly higher than all the other treatments except the cassia alley cropping system. Lack of major differences between maize yields of the various treatments across the seasons of the study is evident from Table 2-3. In general, maize yield was positively related to the rainfall patterns. However, in the third season, in spite of the relatively high rainfall (631 mm), the average of all treatments was only 2.2 t ha-1 which was similar to the yield obtained in some other seasons with lower rainfall (e.g., seasons 1 and 6 with rainfall of 330 mm and 373 mm, respectively). The low yield with 631 mm rainfall could be ascribed to excessive soil loss and runoff that also caused physical damage to the crop; about half the rainfall was estimated to have been lost in the runoff (data not presented).



Maintenance of maize yield over time

Trends of maize yield over time were analyzed by expressing the yield of each treatment in a season as a percent of the average yield of all treatments of the second crop (Table 2-4). The yield of all treatments generally remained stable over seasons, including that of the control with no inputs. Yield increased or declined mainly in direct response to the season's rainfall, with a linear relationship between rainfall and yield (Fig. 2-5). This














Table 2-4. Effect of alley cropping block planting systems on
relative grain yield of maize.


System/Treatment Season

2 3 4 5 6 7 Mean

Alley cropping
Leucaena 100 124 95 95 25 90 88

Cassia 95 157 95 138 57 114 109

Block planting
Leucaena 119 129 95 124 24 81 95

Cassia 110 138 95 138 43 71 99

Mulch-only
Leucaena 96 124 105 76 38 81 87

Cassia 90 110 100 100 33 86 87

Fertilizer-only 76 148 138 149 62 129 117

Absolute control 96 138 86 100 52 114 98

Mean 98 134 101 115 42 96 98


NOTE: Relative yield was calculated as a ratio of actual to
average yield of all treatments in the second season,
expressed as percent.











3.0





2.5





2.0



C

1.5 CD
N




1.0





0.5 L
200


500


250 300 350 400 450


Rainfall (mm)





Figure 2-5. Relationship between rainfall and maize mean yield of all treatments in each of the six cropping seasons. Numbers on the regression line refer to the cropping seasons.










43
suggests that at this site, soil moisture was more limiting than differences between treatments.

The patterns of maize yield over time were also

examined by expressing the yields of the two maize planting systems as a percent of the yields of the fertilizer-only control treatment (Fig. 2-6). Averaged over seasons, maize yield declined by 12% per season under leucaena alley cropping, with the largest decline in the third year (season 6). In contrast, maize yield under cassia alley cropping increased by 8% per season. In the second year (season 6), when yield under leucaena declined by 45% compared to the control, yield under cassia was 27% higher than the control. The increase in maize yield under cassia, in spite of the hedgerows occupying land, suggested that productivity of maize (kg m-") under cassia alley cropping was higher per unit net (cropped) area than of the sole crop (Fig. 2-7). Most of the reduction in maize yield and in productivity under leucaena, compared to the yield of the sole crop, occurred in the last three seasons of the study. The patterns of productivity per unit of land or rainfall were similar. On average, mean maize productivity per unit of rainfall under cassia alley cropping was +9% while that under leucaena was -15% (Fig. 2-8). The increased maize productivity under cassia alley cropping offsets the 15% loss in maize yield when hedgerows occupied 15% of the croppable land.



























Leucaena Cassia Leucaena Cassia
(intercropped) (intercropped) (sole) (sole)
-% m = 6 9 A . . - ,


Saa. _ 1


1990


1991


Year











Figure 2-6. Maize grain yield under alley cropping and block planting systems of Leucaena leucocephala and Cassia siamea expressed as percent of the sole maize control yield.


140 120 100


80 60


1989

























Leucaena Cassia Control
- - -,, .


0.5


N
E 0.4



-o 0.3



"0.2

N
- 0.1



0


Figure 2-7. Maize productivity (kg m-2) under alley cropping Leucaena leucocephala and Cassia siamea hedgerows for six seasons. Vertical bars indicate standard error of difference of means.


1 2 3 4 5 6
Cropping season






























TI


Leucaena
(intercropping)

Cassia
(intercropping)
- -A- s
Maize (sole)


- �-.~j- - - -~


7
7


2 2 3 4 5 6

Cropping season











Figure 2-8. Productivity of maize (kg mm-' of
rainfall) under alley cropping of Leucaena leucocephala and Cassia siamea hedgerows for six cropping seasons. Vertical bars indicate standard error of difference of means.


0

EB





4











Land equivalent ratios (LERI

Alley cropping leucaena and cassia had LER greater

than 1.0 of maize grain yield or the combined yield of the crop and hedgerow biomnass. Averaged across species and seasons, LER of maize grain yield under alley cropping was

1.24 (or 24% advantage over the maize grain of the sole planting system of the hedges and crop). No particular trend in LER of maize yield across seasons was discernible under both species (Table 2-5). Between the two species, LER of maize grain yield under cassia (1.36) was significantly higher than under leucaena (1.12). This difference was due to the lower crop yield under leucaena compared to cassia.

When averaged across species, the combined biomass of the hedgerows and the crop yield (grain plus stalks) had LER of 1.10 (i.e., an advantage of 10% by alley cropping over the sole system). The LER of maize grain yield under alley cropping was greater than 1.0 because of lower yield of maize yield under the sole system compared to alley cropping in nearly all seasons. The biomass yield of the hedges under alley cropping and sole planting systems were similar when the highest yielding sole planting system was used in LER computation.












Table 2-5.


Land equivalent ratios (LER) of alley cropping Leucaena .eucocephala and Cassia siamea compared with the block planting system with same ratios of tree and crop land occupancy.


Year and Seasons Within a Year
1989 1990 1991

Species 1 2 1 2 1 2 Mean


Leucaena 1.0a 1.09 1.15 .0.87 1.23 1.35 1.12

(1.10) (0.97) (1.15) (1.02) (1.29) (1.26) (1.13)

Cassia 1.04 1.33 1.21 1.17 1.56 1.87 1.36

(0.99) (1.24) (1.00) (0.96) (1.10) (1.15) (1.07)


LER = (Biomass yield of the tree species in alley cropping plus yield of crop in alley cropping) divided by (biomass yield of the tree species in sole system plus yield of crop in sole system). In both the alley cropping and sole planting systems, the percent of land occupied by the hedges and the crop was 15 and 85, respectively. In the sole planting system, the biomass yield of
the hedgerows in the 25:75 tree:crop percent land occupancy ratio was used because this was the most appropriate (i.e., highest yielding) sole planting system.

a LER of maize grain yield only. Number in bracket is the total LER, i.e., the grain and the stalk yield of the maize and the tree biomass which was used as mulch.












Nutrient budgets

Nitrocgen.--Generally, leucaena had significantly

higher mulch yield (Table 2-1) and N concentration (Table 2-2) than cassia. As a consequence, the N yield of leucaena

was significantly higher (136 kg N ha-1 yr-1) than that of cassia (59 kg N ha-1 yr') .

Nitrogen uptake was high and essentially of similar pattern in all treatments, including the absolute control (Table 2-6). It was only the fertilizer-control and the alley cropping system of cassia that had N-yields considerably higher than the rest of the treatments. This is because of the relatively high maize yields of these two treatments (i.e., the fertilizer-only control and the hedgerow intercropped plots of cassia).

Apparent recovery/utilization efficiency of N applied through the mulch or fertilizer was low in all treatments (Table 2-6, columns 7 and 8). The N-recovery values were low primarily because maize and N-yield of the absolute control were almost equal to those of the other treatments with inputs. This suggests that the N needs of the crop were adequately met by native soil N. Averaged across treatments, an apparent N-recovery of 9% summed over the N recoveries through the grain, stover and cobs was observed. Most of the N (67%) was recovered in the grain component of the crop and the rest (33%) in the stover and cobs.










Table 2-6.


Nitrogen budgets, utilization efficiencies and yield of maize grown in alley cropping of Leucaena leucocephala and Cassia siamea, at a hedgerow spacing of 6.7 m for 6 seasons.


N from
mulch Unac- Maize
and/or N-utilization N-remaining count- yield
fertil- N removed during 6 seasons efficiencyb in mulch ed for (t/ha/
izer kq ha'I (% N seaTreatment (kg) Grain cob Stover Total Grain Total kg ha- (%) %) son)


Absolute
control* 0 179 9 89 277 . 2.1

Alley
cropping
(Leucaena) 648 186 9 83 278 1.0 0.2 103 16 41 2.2

Alley
cropping
(Cassia) 417 225 9 93 327 11 12 39 9 13 2.7

Mulch-only
(Leucaena) 408 192 10 87 289 3 3 87 21 8 1.8

ulch-only
(Cassia) 177 190 9 91 290 6 7 33 19 0 1.8

Fertilizer
only 240 222 9 97 328 18 22 - - o 2.5

aFertilizer was applied to the alley cropped plots at the rate of 40 and 17 kg ha- season-' N and P, respectively.
"Sample calculation using maize grain yield of 6 seasons from alley cropping of cassia: N utilization efficiency (11%) = N removed in grain yield (225 kg) less N removed in grain yield of the absolute control (179 kg) divided by total N applied (417 kg) times 100. cBased on N concentration of the mulch at the end of the litter bag decomposition studies of mulch incorporated in October 1991 to March 1992. d100 -(percent of applied N in mulch or fertilizer removed in crop plus N remaining in undecomposed mulch).


*From an adjacent experiment.












Trends in N uptake were similar to those of the crop yields, i.e., the higher the yield of the crop, the higher the N-yield. Because of the lower yields under leucaena alley cropping compared to those under cassia (Table 2-3), apparent N utilization efficiency under leucaena was low. However, in the sole croppings with only mulch application, apparent N utilization efficiency values from the mulch of leucaena and cassia were similar.

N uptake and apparent N utilization efficiency were highest in the fertilized control plot, in both grain as well as stover and cobs. N uptake by the maize in the fertilizer-control treatment was 42% more than the amount of N applied. Similar high N uptake levels exceeding the amounts applied were also observed with cassia alley cropping. For these two treatments, all of the N applied through the fertilizer or mulch could be accounted for.

Phosphorus.--In six seasons, P removed by the crop (i.e., grain, stover and cobs) significantly exceeded the amount of P added through mulch or fertilizer (Table 2-7). Averaged across species, P removed through the crop was 164% more than P applied through the mulch of leucaena and cassia. As a consequence of the lower mulch yield of cassia compared to leucaena, P applied thr ough cassia mulch was significantly lower than P applied through leucaena mulch. Hence, P removed through the crop in plots with cassia was significantly higher (215%) than that from plots with












Table 2-7. Addition of phosphorus through mulch and removal
by. maize in six seasons of cropping under
Leucaena leucocephala and Cassia siamea alley
cropping.


Phosphorus (kg ha"1)


Mulch Source Addition Removal Difference


Leucaena 22.8a 52.1a -29.3a

Cassia 15.8b 49.8a -34.0b

Mean 19.3 52.0 31.7

'Values in a column with different letters are not significantly different (p = 0.05).












leucaena mulch (127%). The interaction between N and P uptake/removal by the crop was not examined in this study. Root Density

Root density differed substantially between species and also between soil depths. In both species, the densities of both fine- and medium-sized roots were significantly higher near than away from the hedgerows (Fig. 2-9). Fine root densities of leucaena at both 0-20 cm and 20-40 cm soil depths were significantly higher at nearly all distances from the hedgerows than that of cassia (Table 2-8). Interaction effects between species and soil depths on root density were not significant. In both species, root density at 20-40 cm depth was significantly higher than at 0-20 cm.


Pot Study

During four harvests, the species effects and the

interaction effects of species and mulch rate on the biomass yield of maize were not significant (Fig. 2-10). However, effects of the mulch rates on maize yield were highly

significant, with the yields increasing direct proportion to mulch rates (Fig. 2-10). Averaged across species and irrigation levels (i.e., the equivalents of 100 and 200 mm of rainfall in two months period), biomass yield of maize

from the 0 and 2 t ha-1 mulch rates were not significantly





















0.12


0.1


0.08
E
C.,
,t 0.06

C

4 0.04
0
0

0.02


0


Leucaena (0-20) U Cassia (0-20) " Leucaena (20-40) Cassia (20-40)


.LUL~. L-fI-g.~ ~lE~b liEu iMEli~hii~iiiniSinIEIE


20 40 60 80 100120140160180200220240260280300


Distance (cm) from tree hedges


Figure 2-9. Fine root density in the top 40 cm soil profile of alley cropped hedgerows of Leucaena leucocephala and Cassia siamea.


ii . 4












Table 2-8. Fine root density at two soil depths along a
meter-long transect from 6 m apart hedgerows of
Leucaena leucocephala and Cassia siamea
intercropped with maize.


Soil depth (cm)

Hedgerow Species 0-20 20-40 Mean


Leucaena 0.41 0.64 0.53a

Cassia 0.23 0.47 0.35b

Mean 0.32a 0.56b 0.44


NOTE: Fine root density = cm root per cm3 of soil. a bFor a given row and column, means followed by the same letters are not significantly different (p = 0.05).

























Leucaena: Y = 0.15X + 16.
2
Cassia: Y = 0.24X + 14.2;r


7 r = 0.85 = 0.84

- 0


Leucaena Cassia
leucocephala siamea =.=m- -. ., -,


10 20 30 40 5
Mulch rates (g pot'1,dry weight)


Figure 2-10. Effects of three mulch rates of
Leucaena leucocephala and Cassia siamea on the biomass yield of maize grown in pots, 1991.












different. The effects of these lower rates were also not significantly different from the effects of the fertilizer only.

The effects of irrigation depended on the harvest. In harvests 1 and 4 (i.e., 4 and 16 weeks after mulch application to the pots, respectively), differences in yield between the two levels of irrigation were not significant. However, in harvests 2 and 3, the differences in maize biomass yield were significant.



Discussion


Mulch Yield of the Hedgerows

on average, leucaena produced 46% more mulch than

cassia. The higher yield of leucaena was observable in both systems of hedge planting. Similar differences in the mulch yields of the species had been noted earlier by Nair (1987) working at the same site. The differences in biomass production between the two species could be ascribed to inherent growth differences as well as their site utilization efficiencies. In the lowland humid tropics, differences in the biomass yields of leucaena and cassia appear to be even more pronounced than in the semiarid tropics (Kang et al., 1981; Kang et al., 1985). Averaged over the three years of study, mulch yield of leucaena was

approximately 4 t ha-' yr-' (dry matter). Similar yields of leucaena have also been reported by Rao et al. (1990) from












alley cropping studies spanning over 4 years from semiarid India.

Between the two systems of planting (i.e.,

intercropped vs block/sole stands), intercropped hedgerows of leucaena yielded significantly higher (+33%) mulch than the block planting on an equal area basis in all but in the first year. In cassia, differences in the mulch yield of the two systems were not significantly different. The higher biomass of the alley cropped plots could be attributed to (a) availability of more space and hence reduced competition for soil water and nutrients (especially when the annual crop was not in the field) and, (b) indirect benefits from husbandry practices of the maize crop, e.g., fertilization, weeding and tillage. The higher mulch yield of the intercropped hedgerows compared to the yields of the block/sole stands indicates the benefits of growing the hedgerows within the crops.

In the block planting system, although the hedges

occupied the same land area as the intercropped hedges, they were confined to limited space and not spread out. Hence, plant-to-plant competition was relatively severe. Typically, the hedges at the borders of the block planting system always grew taller and yielded more biomass than the rows inside the blocks. Enhanced tree-to-tree competition may have contributed to the slightly higher mortality,












especially of cassia, observed in the hedges planted in block/sole stands compared to those intercropped.

The high pruning frequency (4-6 times a year) may also have stressed the cassia hedges more than leucaena, and subsequently made cassia less resistant to pest attack. Another experiment on the site of the present study (P. Kiepe, personal communication) showed that with only two prunings per year, mulch yields similar to those obtained in the present study could be obtained from intercropped hedgerows of cassia. Nevertheless, reduced frequency of pruning may exaggerate crop loss under alley cropping due to enhanced shading.

It is also possible that the frequent pruning and

"export" of mulch from the area of the block planted hedges impoverished the site in terms of nutrient storage, which in turn contributed to the low yields of the block planted hedges.



Crop Yield



Effects of mulch on yield

Because yields of maize in the mulched and unmulched control plots were comparable, it would appear that low levels of mulch obtained from the hedgerows limited maize yields. Significant response in biomass yield of maize to increasing rates of mulch i-n the pot studies also supports












this observation. This response could be ascribed to enhanced availability of nutrients; positive impact of mulch on soil physical properties (e.g., improved aeration and moisture conservation) could be another reason, but this aspect was not examined in this study.

It has been argued that if mulch applied in a season could be increased to 3-4 t ha-', maize yield may be increased significantly even in sites similar to the one of the present study with relatively fertile soils. For instance, Mugendi (1990), working at a site adjacent to the present study, reported from a study spanning 14 seasons that application of leucaena and cassia mulch at rates equivalent to 5 t ha-i consistently gave significantly higher maize yields than applications of 1it ha-1 or those of the control plot with no mulch additions. Moreover, more pronounced effects would be expected in nutrient-poor than in nutrient-rich sites. For instance, Palm (1988) observed that in more nutrient-poor and/or in acid soils, increased rates of mulch application resulted in crop responses similar to those of inorganic nitrogen fertilizers.

Because a species like leucaena has a higher fodder value compared to cassia which has no fodder value (NAS, 1984), in semiarid areas, the prunings from the leucaena hedges are more likely to be removed from the field for use as fodder (Singh et al., 1989). The use of the prunings for











fodder becomes even more appropriate when mulch may not result in significant gains in crop yields. Similar observation was made by Singh et al. (1989) from semiarid India. Furthermore, in nutrient-poor and/or acidic soils, crop yields may not be sustained by mulch even if all the crop residues and hedge mulch remained in the cror fields (Szott et al., 1991).

Although the small quantities of mulch from alley

cropped hedgerows may not improve crop yields, it could be argued that their application may still be necessary to minimize long-term declines in soil fertility and productivity. It is hypothesized that such declines in fertility and productivity could result from loss of nutrients removed in continuous harvests of mulch and tree products (Kimmins, 1977; Hook et al., 1982; Lulandala and Hall, 1988). With or without the removal of hedgerow mulch, initial soil conditions, crop type and their yields would likely influence the onset of yield declines. Use of manure from livestock fed with the pruning may reduce yield- and soil-fertility declines. In addition, in the semiarid tropics where the crop yields are generally low compared to those of the humid tropics, removal of mulch in the range of

1-2 t ha-1 season-' may not show any significant impact on yields and fertility conditions of the soil, at least in the short term. This view is supported by the absence of significant differences in maize yield between plots with










62
mulch and those without for the seasons of this study. From the foregoing, therefore, it is evident that improvements in crop yields through the effects of mulch produced by alley cropped hedgerows are probably limited in the semiarid tropics.



Supply of N and P by the mulch and removal by the crop

Leucaena mulch, when returned to the soil, provided more N than what the crop removed in 6 seasons. On the other hand, cassia supplied only half of the crop N. Most of the N removed was harvested in the grain (67%), an observation in agreement with that of others (Allison, 1973). Based on the N budget analysis, it does appear, therefore, that if all of the mulch N is available to the crop, leucaena could meet the crop's seasonal N requirements.

Besides N, the supply of the other nutrients is

important for the maintenance of high maize yield over time. Removal of the mulch which supplies substantial quantities of nutrients other than N could therefore have significant effects on yield, even when N is provided via fertilizer. This view is supported by observations of Gichuru and Kang (1989) who showed that plots receiving no pruning had, in general, lower maize grain yield regardless of N application compared to those receiving pruning.










63
Soil P appears to be a nutrient that could rapidly be depleted through continuous harvest of mulch and crops from the systems studied. This is particularly so in soils such as the Alfisols of the present study where the native exchangeable soil P levels are low (less than 16 ppm), inputs through the mulch are low, and, yet, outputs through crop harvest exceed inputs significantly. The 7 kg ha- yr-I P supplied through leucaena mulch is similar to P levels in leucaena mulch noted by Lulandala and Hall (1990) from semiarid Tanzania. Most of the P in the crop was removed in the grain and stover, in almost equal percentages. Only about 3% of the total P in the crop was in the cobs. Averaged over species, only 37% of the P needs of the crop was balanced by leucaena and cassia mulch. The shortfall was more pronounced in cassia alley cropping because of cassia's low mulch yield compared to leucaena's.

Based on the P budgets (notwithstanding the

possibility that not all mulch P is available to crop), it does appear that the recycling potential of cassia mulch from the alley cropping system is inadequate to meet the P requirements of the maize crop. The balance between P supply and crop need may be even less favorable under conditions of high crop yield such as in seasons of good rainfall. Kang and Wilson (1987) observed that the P content of leucaena prunings was inadequate for one crop of maize at the humid lowlands of Nigeria. Also, while












mulch-P, particularly that of leucaena, could meet the P requirements of maize grain yield of about 2 t ha"' as suggested by Torres (1983), it is evident that it does not meet the total needs of the crop (i.e., grain, stover and cobs).

The considerable increase (16%) in the P concentration of cassia mulch in 1991 over the levels of 1987 and the absence of similar increase in leucaena mulch suggest differences in the patterns of P uptake by the two species. The cause of this difference is unknown. Some postulations are that cassia is more mycorrhizal than leucaena or supporting more phosphate solubilizing bacteria than leucaena in its rhizosphere.



Apparent utilization efficiency of mulch-N

The apparent low use efficiency of mulch-N can best be explained by the equally high crop and N yield of the sole maize control compared to the yields of mulch treatments. This is another indication of the presence of sufficient native soil N at the study site for the maize crop without additional inputs. Low utilization efficiencies of mulch-N have also been reported by several other workers. For example, Palm (1988) observed mulch-N utilization efficiency of 16% by upland rice from mulches of leguminous shrubs. Yamoah et al. (1986b) noted N-utilization efficiencies of near 20% from the mulch of leguminous shrubs with the











addition of fertilizer. Sanginga et al. (1988) also observed low N-utilization efficiencies from leucaena mulch but they emphasized that the absence or presence of N-fixing bacteria inoculation made a difference. Other workers (Kang et al., 1981; Guevarra, 1976) have also reported low apparent N-utilization efficiencies from mulch of leguminous shrubs.

Factors such as time lags between mulch-N

mineralization and crop N demands (Anderson and Ingram, 1989), immobilization of mulch-N and soil-N by microbes in the process of mulch decomposition, at least in the initial stages (Palm and Sanchez, 1991) and volatilization losses from surface applied mulch (Terman, 1979; Nelson, 1982) could also have contributed to the apparent low mulch-N utilization efficiencies observed. Also, about 22% of the undecomposed mulch-N was still tied up in mulch of both species. This cannot be described as inefficiency, because this N is still in the system and might be utilized by subsequent crops. Admittedly, without more precise determinations such as the use of "5N techniques, it is difficult to assess the fate of mulch-N in soil (Sanginga et al. 1988). Management practices that synchronize periods of peak N release from mulch with periods of peak N demands by the crop may enhance effects of the small quantities of mulch available from the hedgerows.












In situ effects of hedgerows on crop yield

The in situ benefits of growing the hedgerows was reflected in the higher maize yields between the alley cropping system and the block planting system and LER of alley cropping > 1.0. The in situ effects of the hedgerows may result not only from their root and rhizosphere effects but also from their ability to protect the soil against erosion and to serve as windbreaks (Wiesrum, 1984). From semiarid India, Rao et al. (1991) provide evidence for reduction in runoff and soil loss under alley cropping with leucaena.

The in situ benefits of the hedgerows will be in perspective when viewed against the yields of the appropriate sole-crop control. This will determine if, indeed, there has been a positive gain over the nointercropping situation. Such a comparison can be made from the last column of Table 2-3, which shows, for instance, that the maize yield of the sole-maize control (fertilizeronly) was -8% compared to that of cassia alley cropping and +12% compared to the leucaena alley cropping. Because of reduced maize yields under leucaena alley cropping, LER or the yield advantage of leucaena alley cropping was significantly lower than that under cassia. These differences in yield reductions under alley cropping of the two species suggest that leucaena is more competitive with maize than cassia. The reductions in maize yield under












leucaena alley cropping could be more severe than observed in the present study in sites that are both nutrient-poor and dry (Chirwa, 1991).

Leucaena became more competitive than cassia starting in the second year. Similar observations on the onset of leucaena's competition were made also by Rao et al. (1990) working with alley cropping in semiarid India. The significantly higher mulch production as well as the presence of high fine-root densities of leucaena in the top soil compared to that of cassia provides additional clues to leucaena's higher competitiveness. It is possible that a higher pruning frequency than the 4-6 times a year of leucaena in the present study may have reduced its competition with the intercrcropped maize.

The higher maize yield under cassia alley cropping compared with that under leucaena or under the sole crop could be ascribed to significantly more positive crophedgerow interactions observed under cassia. These positive interactions could have mitigated high reductions in maize yields from hedgerows with narrow alleys compared to those with wide alleys. This could explain the lack of significant differences in the three alley widths (spacings) used in the present study. Because more land was under hedges in the narrow alley s than in the wide alleys, maize yield from the narrow alleys was expected to be less than from those of the wide alleys.










68
Maize-yield reduction observed under leucaena was not likely due to competition for light, because the hedges were pruned frequently during the rainy season. Also, it was probably not due to competition for nutrients because the plots were fertilized and leucaena (which was inoculated at the start of the study) may have fixed considerable levels for its own N needs. Instead, the reduction in yield was probably most likely due to competition for water, a suggestion also made by others regarding leucaena alley cropping in semiarid India (Rao et al., 1991). In the sixth season when rainfall was 39% below normal, the significantly lower maize yields under leucaena alley cropping compared to the control provide support to the explanation of water limitation. The argument that competition was more for water and less for nutrients was supported by the results of additional field studies (Chapter 5) involving factorial combination of fertilizer (N and P) and irrigation levels where a positive response in maize yield to N application was not significantly different from that due to the effects of irrigation only. Prior to the initiation of the present study, the site was a savanna woodland cropped to pigeon pea for one-and-one-half years, so that soil N may still have been relatively high. Another evidence for water being more limiting than nutrients is provided by the linear relationship between the maize yield and rainfall over the six seasons.












Trends in maize yield over time

All treatments had similar levels of relative yield maintenance as long as rainfall received in a season was normal, a situation more pronounced under leucaena than under cassia. Much longer studies than three years are required because the effects of the hedgerows are likely to be more pronounced with age. Also, long-term studies would more clearly reflect the high variability in seasonal rainfall. From six-year studies in t he humid lowlands of Nigeria, Lal (1991) observed that total maize yields from alley cropping of leucaena and Gliricidia sepium were about 10% lower than those of the control. In the present study, total maize yield after 6 seasons under alley cropping with leucaena was reduced by 12% while that under cassia was increased by 8%. Therefore, under management conditions similar to the ones of the present study, it does appear that cassia has more potential to sustain maize grain yield than leucaena.

Because of the low measured concentrations of

extractable N and P in the soils of the study station (Kibe et al., 1981), fertilizer N and P was applied to the crop each season at the minimum of the recommended rates for farms in the area. However, the general lack of significant differences in maize yields between treatments suggests that N and P were not limiting, again supported by the results of the pot study.












Root Density

The significantly higher yield of leucaena mulch and presence of significantly higher density of its fine roots in the top soil could be the causes of the lower maize yield under alley cropping with leucaena compared with cassia. In a competitive situation, the sharing of limited soil water and nutrient resources by competing plants is proportional to their effective rooting length/density (Bowen, 1985). Generally, most of the fine roots of trees are found in the upper 50 cm of the soil (Fayle, 1975; Deans and Ford, 1983). High root density is an ecological adaptation of plants, particularly those growing on soils with chemical and/or physical limitation in the lower horizons (Bowen, 1973). This is also true of plants in nutrient and water stressed environments (Chapin, 1980). The studies of several workers in both the humid lowland and the semiarid tropics (Kang et al., 1981; Corlett, 1989; Singh et al., 1989; Rao et al., 1991, 1992; Ruhigwa et al., 1992; Toky et al., 1992) corroborate the high root densities of leucaena and cassia found in the present study.

Unfortunately, most of the roots of annual crops and, in particular those of maize, are found in the same soil horizon as those of trees (Jonsson et al., 1988). Therefore, competition for limited resources (e.g., soil water and nutrients) between trees with well established root systems and the seasonal crop with less










71
well-established roots is unavoidable. More often than not, the agronomic crop is the loser in the competitive situation as indicated, in this case, by the low maize yields under leucaena. Studies of other workers from the semiarid tropics also show low maize yields under leucaena alley cropping (Nair, 1987; Singh et al., 1989; Corlett et al., 1989; Rao et al., 1990, 1991; Ong et al., 1992). The present study also shows that all species are not as competitive as leucaena with crops, and that an understanding of the rooting system of species to be used for intercropping is necessary.

Typically, fine root density of trees declines with

distance from the stem (Roberts, 1976; Ford and Deans, 1977; Persson, 1980; Dhyani et al., 1990) which is consistent with the observations of the present study. Along the entire 3 m long trenches from the hedgerows studied, root density of leucaena was significantly higher than that of cassia. The extended high root-density of alley cropped leucaena into the cropped Area and the resultant low yields of intercrops would suggest that potentials of intercropping leucaena without physical separation from crops are limited. Approaches to minimize below-ground competition between hedgerows and crops, at least experimentally, include root, pruning and/or placement of physical barriers such as galvanized iron or polythene sheets (Singh et al., 1989; Rao











et al., Ong et al., 1992), trenching (Verinumbe, 1985) and root pruning (Fernandez, 1990).

With root pruning/barrier techniques, improvements in crop yield compared to a control have been noted, with significantly increased yields at some times (e.g., Corlett, 1989) and not at others (Ong et al., 1992). Root pruning is, however, labor demanding and unless there are significant gains to be foregone by not having the hedgerows in situ, farmers would be most unlikely to adopt it. Besides, roots of the hedgerows can arch back to the surface soil even when physical barriers are placed to depths of one meter below the soil surface and hence render barrier or trenching efforts ineffective (Ong et al., 1992).

Separate planting of hedges in blocks or sole stands on land outside the crops that is not currently cropped, or on land not suitable for cropping or on the periphery of the farm may be options to minimize competition of leucaena with crops. This could be an option particularly for leucaena which has a desirable product (fodder) but which cannot be intercropped because it is competitive with crops. However expected losses in crop yield are proportional to the percentage of land occupied by the hedges. This lost land occurred because the amounts of mulch obtained from the

hedges were low (on average 1-2 t ha-I season-') and had no compensatory effects on crop yield. However, the results of the pot studies suggest that if the levels of mulch from












hedgerows could be increased to application rates of -3-4 t ha-' season-', gains in crop yields could be increased even from the relatively fertile site of the present study.



Conclusions

Alley cropping of Leucaena .eucocephala was

detrimental to maize yield over the six cropping seasons. on average, maize yield under leucaena alley cropping dropped by 12% of the sole crop while that of cassia increased by 8%. Leucaena, on the other hand, produced significantly higher mulch biomass (50% more) than cassia. The higher mulch biomass production and higher density of fine roots in the top 40 cm soil are factors that contribute to leucaena's significantly higher competitiveness with maize than cassia's.

The in situ benefits of growing the hedges on maize

yield were positive for both species but significantly more under cassia compared to those under leucaena. Planting the hedges outside the crops may be a management approach to minimize competition between the crop and the tree hedgerows, especially for a species like leucaena with desirable fodder value. However, this approach produced yield reductions higher than alley cropping with both leucaena and cassia, with the loss nearly equal to the percentage of land apportioned to the hedges.











The magnitude of the mulch produced by leucaena and cassia, 1 and 2 t ha-1 season', respectively, was not sufficient enough to cause significant gains in maize yield at the study site. There are indications that gains in maize yield could be significant if the amounts of mulch

applied could be increased to 4 t ha-' season and beyond. It is unlikely, however, that such increases in mulch yield from the hedges could be made without compromising further crop yields.

At this point, the results of the present study do not indicate sufficient short-term benefits to recommend alley cropping with leucaena under conditions similar to the ones of the present study. However, cassia does have potential if (a) the 8% increase in maize yield outweighs the costs of production and, (b) mortality of cassia after 3-4 years is low enough not to necessitate replanting of the hedgerows. Follow-ilp studies on factors responsible for cassia's positive in situ effects may enhance its potentials for alley cropping under semiarid conditions.














CHAPTER 3
SOIL FERTILITY ASPECTS OF ALLEY CROPPING
LEUCAENA LEUCOCEPHALA AND CASSIA SIAMEA IN
SEMIARID TROPICS AT MACHAKOS, KENYA


Introduction
Agroforestry practices are suggested as low-input systems with potentials to meet the critical needs to improve or sustain food production in the tropics (Swift and Sanchez, 1984). One practice of agroforestry which has been the subject of much interest and study is alley cropping, also called hedgerow intercropping (Kang et al., 1990). It

consists of growing food crops in alleys between planted hedgerows of multipurpose trees and shrubs. The hedgerows are periodically cut back at the beginning of as well as during cropping to prevent shading and to provide mulch to the associated crop.

One of the most important premises of alley cropping is that the addition of organic mulch, especially if it is rich in nutrients, has favorable effects on soil physical and chemical properties, and hence on productivity of crops. However, the ability of the mulch to provide nutrients and to maintain or improve soil properties would depend, in addition to its quality and time of application, on the quantity. For instance, Kang et al. (1990) observed from six-year-long studies at the humid lowland of West Africa that with continuous addition of leucaena pruning (8-10 t ha- yr" , dry 75









76
matter), soil organic matter (SOM) and nutrients were

maintained at higher levels than where the prunings were removed. Similar observations are made by other workers also from the lowland humid tropics (Atta-Krah et al. 1985; Yaxnoah

et al. 1986c) . However, in the SATs, ecological conditions do not allow the high rate of mulch production found in the humid tropics. Hence, signif icant gains in SOM from the addition of mulch may be limited. Based on the effects of mulch per se

there might be no conceivable advantage of alley cropping systems on SOM in the SATs. However, turnover rates of high root mass often associated with agroecosystems with trees (Ewel et al., 1982), such as alley cropping, may contribute significantly to organic matter and nutrient contents of the soil. Moreover, where leaching losses of soil- and appliednutrients may be a problem, the roots could minimize these through a combination of processes including sorption and uptake from deeper layers (Stark and Jordan, 1978; Jaiyebo and Moore, 1964). These processes may result in the concentration of nutrients in the mulch of the hedgerows which could, in turn, improve crop nutrition (Fernandez, 1990). In some

soils, there are indications that the roots may even recycle nutrients from the subsoil which could in turn result in top

soil improvement of pH and exchangeable bases (Lal, 1989). In addition, due to their longevity and to their protective benefits, intercropped hedgerows may minimize runoff and soil erosion loss (Lal, 1989; Rao et al., 1991).










77
Notwithstanding the positive attributes and potentials of alley cropping discussed above, competition for soil

resources (nutrients and water) between the hedgerows and the crops has to be considered since this could result in

significant losses of crop yield under alley cropping (Rao et al., 1991; Nair, 1987; Singh et al., 1989). The losses in yield may, indeed, obscure any gains in soil fertility due to the presence of the hedgerows. To justify mixing of multipurpose tree hedgerows with crops, particularly by resourcepoor farmers, it is important, therefore, to understand the

benefits to soil fertility that would otherwise be lost if the hedgerows were planted separate from the crops. This

knowledge is of particular significance in the semiarid tropics where soil moisture is a major factor limiting crop

production and where soil fertility improvement aspects of alley cropping have not yet been wellunderstood.

This study was undertaken in this scenario with the following objectives: (a) to evaluate the changes in some selected soil chemical and physical properties under alley cropping of maize with Leucaena leucocephala and Cassia siamea over three years and, (b) to separate the effects of mulch per se from the in situ presence of the hedgerow on the selected soil chemical and physical properties.












Materials and Methods


Site Descrintion

The location, rainfall and soil characteristics of the

site are described in Chapter 2. Table 3-1 shows the chemical and physical characteristics of the top soil (0-20 cm) at the beginning of the study (October, 1987).



Treatments and Experimental Layout

The experimental treatments consisted of two systems of Leucaena .eucocephala and Cassia siamea: (a) intercropped hedgerows and (b) planted as blocks/sole stands outside the crop. In the intercropped system, the hedgerows were 6.67 m wide. The hedgerows in both the intercropped and the block/sole planting systems occupied 15% of the total area, and the crop occupied the remainder of the land (85%). Plots

with fertilizer-only and mulch of leucaena or cassia only were included as controls. As an absolute control, no-mulch or fertilizer plots from an adjacent experiment was used. The

experimental design was randomized block design with three replications. The field experiment was conducted for more than 3 years (October 1987 to March 1992).



Crop and Tree Management

In all systems, the soil was hand tilled and maize (Katumani Composite var.) planted twice a year at a stand of












Table 3-1. Chemical and physical
soil (0-20 cm) at the
October 1987.


characteristics of the top beginning of the study,


Treatments

Leucaena Cassia Control
Soil Property plots plots plots Mean


pH (1:2.5 soil:H20) 6.2 6.2 6.2 6.2

Org. C (%) 0.9 1.0 0.8 0.9

Total N (%) 0.1 0.1 0.1 0.1

P (ppm, Mehlich-1) 12.7 12.0 12.7 12.5

CEC (meq/100 g soil) 9.6 9.5 9.0 9.4

Clay (%) 28.7 30.7 26.7 28.7

Bulk density (g/cm3) 1.3 1.3 1.3 1.3


experimental


At the time of soil sampling in 1987, the treatments had not yet been introduced.









80
37,000 plants/ha. Moderate levels of fertilizer (40 and 18 kg of N and P/ha) were added to the maize in the hedgerow intercropped and the block planted systems each season. Weeds were hand removed at least twice in each season. At harvest, maize grain, stalks and cobs were removed from the field as was the practice of the farmers in the area.

The tree hedgerows were pruned back to 50 cm above ground with 2-3 prunings a season. The prunings obtained were chopped into small pieces and applied as mulch to the plots. The chemical characteristics of the mulch are shown in Table 3-2. If tillage and/or weeding schedules shortly preceded periods of mulch application, the mulch generally became incorporated into the soil; otherwise, it remained on the soil surface.

Soil sampling was always carried to 20 cm depths. A composite sample of 10 auger cores was taken for each treatment. The soil was air-dried, crushed to pass a 40mesh sieve, and mixed. Samples were analyzed at the Analytical Laboratory of the Katumani Research Station, Machakos, Kenya by the following methods: total N by the micro-Kjeldahl method (Jackson, 1958), organic carbon by the Walkley-Black procedure (Jackson, 1958), P by the Mehlich-l double acid procedure and cation exchange capacity (CEC) by Atomic Absorption Spectrophotometry methods. Soil pH was determined using a glass electrode pH meter (model Bibby), at soil to water ratio of 1:2.5.









81


Table 3-2. Chemical characteristics of the dry mulch of
alley cropped hedgerows of Leucaena leucocephala
and Cassia siamea.


Poly
N P K Ca Mg C:N Lignin phenol
(%) (%) (%) (%) (%) ratio (%) (%)


Leucaena 3.5 0.2 1.9 0.7 0.3 14 8.1 4.7

Cassia 2.9 0.2 1.8 0.7 0.2 17 8.4 3.7











Relative availability of N03- N and NH44 N was

monitored in the last season. Ion exchange resin bags (Binkley and Matson, 1983; Binkley and Hart, 1989) were used. In each plot/replicate, one bag for anions and one for cations weighing 5 g were placed in the top 10 cm soil profile. After the harvest of the crop, the resins were retrieved and NO.- N and NH4* N extracted with 2 N KCl. Levels of NO,- N and NH4* N from the KCl-extracts were determined using Standard Autoanalyzer methods at the Soil Testing Laboratory of the University of Florida, Gainesville.



Soil Physical Properties

Bulk density (Blake, 1965) is the only physical

property of the soil discussed in this chapter. In the last cropping season, soil was sampled with a semi-wide auger and 5 cm long to a total depth of 20 cm (i.e., 0-5, 5-10, 10-15 and 15-20 cm). Bulk density of each plot/replicate was computed as the average values of samples taken at two positions near the hedgerows (0.45 m) and two in the middle of the alleys (3.5 m away from the hedgerows).



Statistical Analysis

ANOVA procedures were used to determine statistically significant (p = 0.05) differences between the levels of











soil properties monitored under the various treatments and over time. Because the initial (1987) soil levels of percent clay and P differed significantly among plots, these two parameters were used as covariates in the analysis of the final (1991) soil P levels. Also, for the KC1 extractable N, analyses were performed on logarithmically transformed data, because frequencies from a set of 5 resin bags were not normally distributed. When significant effects of treatments were detected, means were separated by pairwise contrasts of the LSM using the GLM procedure in SAS (SAS, 1992). Differences between treatments were declared significant at p = 0.05.


Results



Soil Chemical Properties

With the exception of P, there were no significant

differences among treatments in the levels of the chemical characteristics measured (Table 3-3). Soil pH, organic carbon, total N and CEC in the different treatments were similar. With respect to P, however, there were significant differences among treatments. in general, plots with cassia mulch tended to have a higher level of available soil P. In particular, the levels of P from the plots with application of cassia mulch-only were significantly higher than the levels of all the other treatments (Table 3-4). The











Table 3-3. Chemical and physical characteristics of top soil (0-20 cm) under alley
cropping with hedgerows of Leucaena leucocephala and Cassia siamea hedgerows,
November 1991.


KCI CEC
pH Organic extract (meq/ Bulk
(1:2.5 C Total N (pg/g Mehlich-1 100 g density
System soil:H20 (%) N (%) soil)* P (ppm) soil) (g/cm)


Alley cropping
Leucaena 6.3a 0.83a 0.09a 9.91a 16.7a 8.42a 1.23a
Cassia 6.2a 0.81a 0.09a 8.15" 23.Ob 7.67a 1.46a

Block planting
Leucaena 6.2' 0.94a 0.08a 8.55a 18.2b 7.63a 1.37a
Cassia 6.2a 0.83a 0.09a 9.17a 19.5b 9.26a 1.33a

Controls
Mulch
(leucaena) 6.2a 0.80a 0.11 8.09a 11.38 7.21a 1.31a
Mulch
(cassia) 6.6a 0.93a 0.06a 7.43a 30.0c 9.68a 1.34a

Fertilizer 6.3a 0.72a 0.07a 6.14a 21.2b 7.60a 1.20a

Maize-only** 6.38 0.85a 0.07a 6.77a 11.0' 9.33a 1.44a

a'b'c=dIn a column, values with different letters are significantly different (p = 0.05).

*Indicates average of 3 sampling dates.

**From an adjacent experiment (no fertilizer, no mulch).




Full Text
180
Soil Fauna
Total count of soil fauna was significantly higher
near the hedgerows compared to away from the hedgerows (Fig.
5-11). Soil fauna commonly observed included termites, wood
lice and millipedes; earthworms were absent.
Supplementary Studies
Field study: effect of irrigation and N
fertilizer on maize yield
A significant response to irrigation was observed
(Fig. 5-12). There was no response to increasing rates of N
application. These observations suggest that maize yield at
the site was limited more by soil moisture and less by soil
N.
Pot Study; effects of factorial combination
of irrigation and fertilization on yield of
maize grown on soil collected from near and
away from the hedgerows
No significant 2- or 3-way interaction effects of N
rates, water levels and soil source (i.e., near or away from
the hedgerows) were observed. The effect of soil source
(i.e., near or away from the hedgerows) on maize yield was
not significant. However, main effects of the N rates and
the irrigation levels were significant. Biomass yield of
yield of maize was not significant. Main effects of the N
rates and the irrigation levels were, however, significant.


CHAPTER 1
INTRODUCTION
In many parts of the tropics, semiarid lands that have
traditionally not been under arable farming are increasingly
being brought under rainfed agriculture. For example, in
Kenya where arid and semiarid lands comprise four-fifths of
the country's land area, arable farming has been extended to
vast areas of dry or marginal lands during recent years.
However, there are some formidable obstacles to food
production in many of these areas. In addition to rainfall
constraints, the soil's productive capacity is inherently
low in many areas. In areas where low level subsistence
farming has been followed, fertilizer use and other means of
nutrient input to soil are seldom practiced. Consequently,
low and declining yields of crops are common. For instance,
maize yields less than 1 t ha*1, even in seasons of normal
rainfall are typical of farms in semiarid areas of Machakos,
Kenya (Nadar and Faught, 1984; Mbogo, 1991).
Under traditional shifting cultivation, which has been
the predominant agricultural system in many areas of the
tropics (both humid and semiarid), long rest or fallow
periods in between short cropping periods have traditionally
been used to improve fertility of soils exhausted by
cultivation (Nye and Greenland, 1960). However, land
1


91
noted in their rates of decomposition (particularly on the
soil surface), and no long-term differences in soil organic
C may be expected.
Although there may in fact be (unmeasured) differences
in the quality of the mulch of the two species, the lack of
significant differences observed in organic C of soil under
the mulch of the two species suggests that these differences
were not as important as the quantity of mulch in impacting
on soil organic C. The effects of residues (mainly those of
crops) on soil organic C content are highly related to the
amount and only slightly to the type of residue applied
(Rasmussen and Collins, 1991; Sowden, 1968). Both the
quantity and the quality factors of the mulch are critical
because of the high loss of the C applied as C02.
Typically, laboratory incubation studies estimate that as
much as 60-75% of the C from crop residues could be evolved
as C02 after one year in the soil (Martin and Haider, 1980).
The effects of hedgerow intercropped systems on soil
organic C, if any, appeared to be due more to the effects of
the mulch and less to the in situ presence of the hedgerows.
The effects of only the small quantities of mulch added
(2-4 t ha'1 yr'1) appeared to have been sufficient to sustain
soil organic C after three years of cropping. Evidence for
this is provided by the lack of significant difference
observed in the levels of soil organic C between the


256
Lalf R. 1991. Myths and scientific realities of
agroforestry as a strategy for sustainable management
for soils in the tropics. Advances in Soil Science 15,
91-137.
Landon, J.R., (ed.), 1990. Booker tropical soil manual: a
handbook for tropical soil survey and agricultural land
evaluation in the tropics and sub-tropics. John Wiley
and Sons, New York.
Lightfoot, C., 1987. Indigenous research and on-farm
trials. Agricultural Administration and Extension 24,
79-89.
Lousier, J.D., and Parkinson, D., 1976. Litter
decomposition in a cool temperate deciduous forest.
Canadian Journal of Botany 54, 419-436.
Ludwig, J.A., 1987. Primary productivity in arid lands.
Myths and realities. Journal of Arid Environment 13,
1-7.
Lulandala, L.L.L., and Hall, J.B., 1990. Nutrient removals
in harvesting of Leucaena hedgerows at Mafiga,
Morogoro, Tanzania. Forest Ecology and Management 35,
207-216.
Maher, C., 1937. Soil erosion and land fertility in the
Ukamba (Kitui) reserve. Ministry of Agriculture,
Nairobi, Kenya.
Mann, H.S., and Saxena, S.K., 1980. Khejri (Prosopis
cineraria) in the Indian desert: Its role in
agroforestry. CAZRI Monograph 11. Central Arid Zone
Research Institute, Jodhpur, Rajasthan, India.
Martin, J.P., and Haider, K., 1980. Microbial degradation
and stabilization of 14c-labelled lignin, phenols, and
phenolic polymers in relation to soil humus formation,
pp. 77-140. In: Kirk, T.K., Higuchi, T., and Chang,
H.M. (eds.), Lignin biodegradation; microbiology
chemistry and potential applications. Vol. 2, CRC
Press, West Palm Beach, Florida.


202
exercise, stratification of the farms relative to their
location on the hills (low, middle and top slopes) was felt
necessary. One reason for stratification was that
infrastructure in the area was strongly influenced by slope
factors of the hills; farms on the low and middle slopes
were more accessible than farms on the upper slopes. In
addition, farms on the top slopes and particularly those on
the windward side had more rainfall (on average, 1000 mm a
year) than farms on the low slopes (with about 700 mm
rainfall a year). On the basis of the above stratification,
a list of farmers of the three administrative sublocations
was obtained from the Ministry of Agriculture staff in
charge of each of the three sublocations. However, in the
Kiima-Kimwe sub-location, an extension agent of the Catholic
Diocese of Machakos, helped provide a more comprehensive
list of farmers. This was because of the inadequacy of the
list provided by the agricultural office of this location.
Twenty farms were randomly selected from the list of
farmers of each sublocation. Thus, a total of 60 farmers
for the three locations were selected. Interviews were
conducted on single-day visits in the months of July and
August, 1991. A one-page questionnaire was used as a guide
during the interviews. Interviews and visits at each farm
lasted about an hour. The interview team was made up of the
author and the extension agent of the Catholic Diocese of
Machakos who also helped with the interpretation of the


187
than away from them. Therefore, the availability of
nutrients should be higher near the hedgerows than away from
them (although significant differences were not measurable).
With relatively more soil moisture content near the
hedgerows, uptake of nutrients by the plants near the
hedgerows was probably enhanced. Typically, an increase in
soil moisture results in a corresponding increase in the
uptake of nutrients (Tisdale and Nelson, 1975). Also, the
less compacted soil (low bulk density) near the hedgerows
could have contributed to an increased uptake of nutrients.
It is possible, therefore, that besides enhanced soil
moisture, the presence and/or availability of more nutrients
near the hedgerows was a contributory factor to the higher
maize yield observed near the hedgerows than away from them.
The presence of more roots near the hedgerows could have
enhanced availability of nutrients more near the hedgerows
than away from them not only through root death and
subsequent mineralization but also through the effects of
root exudates in mobilizing nutrients from insoluble
materials.
Soil Water Status Near and Away
from the Hedgerows
Top soil (0-20 cm) gravimetric water content was
generally higher near the hedges than away from them. This
was so during all months of sampling, although a
statistically significant difference was detected in only


5
extent of its soil-improving potential is little known
compared to that of leucaena. There is even doubt as to
whether it is nitrogen fixing, and established opinion is
that it is not.
In the humid lowlands of Nigeria, many trials have
shown that crop yields could be improved under alley
cropping with MPTS. For instance, an eight-year alley
cropping trial conducted by Kang et al. (1990) showed that
using Leucaena leucocephala prunings only (in the range of
8-10 t ha'1 yr'1), maize yield could be maintained at a
reasonable level of 2 t ha'1, compared to 0.66 t ha'1 without
leucaena. Additions of leucaena prunings also maintained
higher soil organic matter than when no prunings were added.
Several other studies in the humid lowlands have also given
favorable results from alley cropping (e.g., Yamoah et al.,
1986a). As a result, there has been considerable interest
in alley cropping, and the technology has been tried in a
number of other places including the semiarid tropics
(SATs). However, the results of alley cropping have not
been positive in all places.
Fundamental to the concept of alley cropping is the
proposition that trees improve soils. There is evidence to
support that soil fertility and, to some extent, crop yields
in semiarid areas could be improved by the presence of trees
on farmlands e.g., Faidherbia albida (synopym: Acacia
albida) in West Africa (Bon Kongou, 1992) and Prosopis


63
Soil P appears to be a nutrient that could rapidly be
depleted through continuous harvest of mulch and crops from
the systems studied. This is particularly so in soils such
as the Alfisols of the present study where the native
exchangeable soil P levels are low (less than 16 ppm),
inputs through the mulch are low, and, yet, outputs through
crop harvest exceed inputs significantly. The 7 kg ha-1 yr"1
P supplied through leucaena mulch is similar to P levels in
leucaena mulch noted by Lulandala and Hall (1990) from
semiarid Tanzania. Most of the P in the crop was removed in
the grain and stover, in almost equal percentages. Only
about 3% of the total P in the crop was in the cobs.
Averaged over species, only 37% of the P needs of the crop
was balanced by leucaena and cassia mulch. The shortfall
was more pronounced in cassia alley cropping because of
cassia's low mulch yield compared to leucaena's.
Based on the P budgets (notwithstanding the
possibility that not all mulch P is available to crop), it
does appear that the recycling potential of cassia mulch
from the alley cropping system is inadequate to meet the P
requirements of the maize crop. The balance between P
supply and crop need may be even less favorable under
conditions of high crop yield such as in seasons of good
rainfall. Kang and Wilson (1987) observed that the P
content of leucaena prunings was inadequate for one crop of
maize at the humid lowlands of Nigeria. Also, while


145
application of a mixture of leaves and twigs could be a
management tool to reduce the rate of decomposition and
hence contribute to soil organic matter build-up.
Conclusions
Decomposition patterns of the mulch of both species
were biphasic, i.e., an initial phase of a rapid rate
followed by a second phase of a lower rate. When soil
incorporated, the decomposition rates of the mulch (leaves
plus twigs) of both leucaena and cassia were similar (about
12% per week in the first phase and 2% per week in the
second phase). However, when placed on the soil surface,
leucaena mulch decomposed more rapidly than cassia mulch in
the first phase of decomposition.
Narrow C:N ratio appears to explain the rapid rates of
decomposition of both species. In the second phase, high
C:N ratio (or lignin) appears to explain best the low rates
of decomposition of both species. In leucaena, however,
polyphenols also seemed to be involved in regulating the
rate of decomposition of the second phase.
Synchrony was not observed between periods of peak
mulch-N release (i.e, four weeks after application of mulch
to the soil) and peak uptake by the maize crop (eight weeks
after sowing). Low apparent recovery of the mulch-N applied
by the crop observed (34% from leucaena and 4% from cassia)
was perhaps a reflection of the poor synchrony between
mulch-N mineralization and uptake by the crop. It is


124
respectively. The difference between these rates of mulch-N
release of the two species was not significant. In the second
phase, the rates of N release were significantly lower than
those of the first phase, 6.3% and 3.2% for leucaena and
cassia, respectively. These rates were not significantly
different. The spline points of the regression lines of the
two phases occurred at 8.1 weeks for leucaena and 8.7 weeks
for cassia after incorporation of mulch into the soil. The
difference in the spline points of two species was also not
significant. The half-life of N loss for leucaena and cassia
in the first phase of decomposition were at 4.1 and 4.4 weeks,
respectively.
In the first season, the cumulative mulch-N mineralized
after 16 weeks was 73.7% and 55.5% of the total for leucaena
and cassia, respectively. In the second season, N
mineralization of both species was similar, viz, 98% for each
species. At week four of the second season, 57.5% of the
leucaena and 58.4% of the total cassia mulch-N was
mineralized. Net N-immobilization by the mulch was detected
between weeks 8 and 16 for both species (Fig. 4-6) in the
second season. At week eight, the mulch-N concentration of
leucaena and cassia mulch was 1.7 and 1.9%, respectively,
suggesting that these concentrations of N may be close to the
critical levels for initial net mineralization of incorporated
legume (or below these levels, N-immobilization by soil
microbes may occur).


245
Improved levels of available soil N may result in
improved nutrition and yield of crop, particularly if the
periods of peak N mineralization from the mulch (four weeks
after application) are matched/synchronized with periods of
peak N uptake by maize (6-8 weeks after sowing). The low
apparent recovery of the mulch-N applied could be a
reflection of the absence of synchrony between the peaks of
mulch-N mineralization and uptake by the crop. In the
field, apparent N recoveries from the mulch of leucaena and
cassia were 3% and 7%, respectively. Studies using 1SN
tracers into the pathways and pools of N transfer from the
mulch may provide insights into the potentials of improving
the utilization efficiency of the limited amounts of mulch
available from the hedgerows. In addition, studies on time
of mulch application and/or the use of mixtures of mulches
of different qualities (i.e., variable rates of
decomposition) may help determine to what extent management
practices can achieve synchrony between mulch N
mineralization and uptake by the crop.
Based on the 8% increase in maize yield under cassia
alley cropping as opposed to 12% decline under leucaena, it
is tempting to recommend cassia alley cropping to farmers.
To put such a recommendation into perspective would,
however, require an understanding of the costs associated
with the 8% increase in maize yield from cassia alley
cropping as well as additional costs of replanting the


resourceful farmers such as coffee growers than by poorer
farmers.
xvi


BIOGRAPHICAL SKETCH
Bashir Jama Adan was born on July 7, 1957 to a nomadic
family on the outskirts of Modogashe, Garissa District,
North Eastern Province, Kenya. Between 1966 to 1972, Bashir
completed his Certificate of Primary Education at Garissa
Primary School. Between 1973 and 1978, Bashir obtained his
Secondary and High School Certificates from Alliance High
School, Kikuyu, Kenya. Bashir attended the University of
Nairobi, Kenya, between 1979 and 1982, where he earned a
Bachelor of Science degree (honors) in agriculture. Bashir
completed his Master of Science degree in agronomy in 1988,
received also from the University of Nairobi. In 1982, he
was employed by the Ministry of Energy, Kenya, as the center
manager for Mtwapa Agroforestry/Energy Center, Coast
Province. In 1982, he assumed the position of research
assistant at the International Centre for Research in
Agroforestry (ICRAF), Nairobi, Kenya. In 1989, Bashir
initiated studies at the University of Florida for a
doctoral degree in Forestry Resources and Conservation (with
a specialization in Agroforestry).
267


137
The presence of high polyphenol content is reported to
influence the rates of decomposition of leguminous residues
with high N but yet with low decomposition rates (Vallis and
Jones, 1973; Weeratna, 1979; Ladd et al., 1981). Other
suggested factors include polyphenol-to-N ratio (Palm and
Sanchez, 1991; Oglesby and Fownes, 1992) and lignin plus
polyphenol to N ratio (Fox et al., 1990). The concentration
of polyphenol in the leaves of leucaena used in the present
study were high relative to that used by others, for
instance, Palm and Sanchez (1991). However, the rates of
decomposition in the first phase did not appear to be
influenced by the concentration of polyphenols in the
leaves. In fact, leucaena leaves decomposed at
significantly higher rate than cassia leaves, in spite of
the latter's lower polyphenol content. These observations
suggest that polyphenols (and their ratio with N) were
either not important, or that the levels and the types of
polyphenols present in leucaena leaves were not rate
limiting. It is also possible that the polyphenols present
in leucaena and cassia leaves were water-soluble and readily
dissolved from decomposing residues before impacting
negatively on the activities of the decomposer community
(Olson and Reiners, 1983; Baldwin et al., 1983). Thus, the
primary factor regulating the rates of decomposition of the
first phase of both species appears to be narrow C:N ratio,
and not polyphenol or lignin content.


CHAPTER 7
SUMMARY AND CONCLUSIONS
Maize yield of alley-cropped plots of leucaena and
cassia was higher than those of plots receiving mulch from
the block planting system. This suggested that the effects
of growing the hedges in situ was positive. The yield
advantage of alley cropping over the separate block planting
system, expressed as land equivalent ratio, was 36% and 12%
for cassia and leucaena, respectively. The amounts of mulch
applied to the alley cropped and block plating systems being
equal, the higher yield of the alley cropping systems could
be ascribed to effects associated with the in situ presence
of the hedgerows. Although not examined in the present
study, the in situ effects of the hedgerows could include
root and rhizosphere activities of the hedgerows (e.g.,
biological N-fixation, mycorrhizal association) and soil
erosion control. There is evidence from studies of others
that the soil erosion control aspects of alley cropped
hedgerows can be significant.
Between the two species, leucaena alley cropping
reduced maize by 12% per season compared to the sole crop
while that under cassia increased by 8% per season.
Leucaena produced more biomass (4 t ha'1 yr'1) than cassia
241


155
from the hedgerows. Soil analyses were conducted at the
Analytical Laboratory of the Katumani Research Station,
Machakos, Kenya. In addition to total N, relative index of
available N was determined with ion-exchange resin bags
buried to 10 cm depth near and away from the hedgerows in
the last season (seventh) of the study. On retrieval of the
ion exchange resin bags at the end of the cropping season, N
trapped by the resins was extracted with 2 M KC1. Ammonium-
N and nitrate-N contents of the KCl-extract were determined
at the Soil Testing Laboratory of the University of Florida,
Gainesville.
Supplementary Studies
To determine the relative influence of soil water or
fertility (N) on the observed tree crop interaction,
supplementary field and pot studies involving factorial
combination of fertilizer N and irrigation were initiated in
the last season of the study.
Field studies: irrigation and fertilization
Treatments to evaluate the response of maize yield to
N fertilization and irrigation, in combination or alone, was
introduced in small plots that were formerly unplanted
strips between the plots of the main experiment. The
treatments of the experiment were: irrigation only,
fertilization only, irrigation plus fertilization and


133
Lang, 1982). Plant structural materials and microbial
products comprise a greater proportion of the recalcitrant
residual mass in the second phase (Cheshire et al., 1974).
The predicted and the observed values of the mulch
remaining not decomposed were regressed well by two-stage
exponential model. This model, which has also been used by
other workers (Howard and Howard, 1974; Hunt, 1977; Lousier
and Parkinson, 1976) has considerable advantages in
providing insights into the process of the decomposition
process over statistical procedures typically used to
analyze decomposition data. Wieder and Lang (1983) provides
a comprehensive critique of the analytical methods used in
examining decomposition data obtained from litter bags.
The general absence of significant differences between
the rates of decomposition of the soil incorporated and
surface placed mulch was unexpected. Other workers have
observed that soil incorporated mulch decomposes faster than
surface placed mulch (Read et al., 1985). Generally, the
rates of decomposition of plant materials are more rapid in
moist wet and warm than in dry conditions (Swift et al.,
1979; Shield and Paul, 1973). Moisture conditions generally
increase with depth from the soil surface. Microbial
activity also increases linearly with increase in soil
water, at least up to 50% of the soil water holding capacity
(Murwira et al., 1990). Extremes of drying and wetting
conditions which cause conditions less favorable for
decomposition (Parr and Papendick, 1978) predominate more on


TABLE OF CONTENTS
Page
ACKNOWLEDGMENTS
LIST OF TABLES. . vii
LIST OF FIGURES x
ABSTRACT
CHAPTERS
1 INTRODUCTION 1
2 PRODUCTIVITY ASPECTS OF ALLEY CROPPING
WITH LEUCAENA LEUCOCEPHALA AND CASSIA SIAMEA
IN SEMIARID TROPICS AT MACHAROS, KENYA . 11
Introduction 11
Materials and Methods 15
Results 30
Discussion 57
Conclusions 73
3 SOIL FERTILITY ASPECTS OF ALLEY CROPPING
WITH LEUCAENA LEUCOCEPHALA AND CASSIA
SIAMEA IN SEMIARID TROPICS AT MACHAROS,
KENYA 75
Introduction 75
Materials and Methods 78
Results 83
Discussion 86
Conclusions 95
4 DYNAMICS OF SOIL FERTILITY CHANGES: RATES OF
DECOMPOSITION, N-MINERALIZATION AND UPTAKE
BY MAIZE FROM MULCH OF LEUCAENA LEUCOCEPHALA
AND CASSIA SIAMEA UNDER SEMIARID CONDITIONS
AT MACHAROS, KENYA 97
Introduction 97
Materials and Methods 100
Results ....... 110
Discussion 132
Conclusions 145
v


169
Table 5-5. Relative water content of flag leaves of maize
plants sampled near (0.45 m) and away (3.5 m)
alley cropped hedgerows of Leucaena leucocephala
and Cassia siamea on January 6, 1991 (seventh
season).
Water content
(%)
Near
Away
Species
(0.45 m)
(3.5 m)
Mean
Leucaena
31.6
o

o
m
30.8
Cassia
31.3
34.8
33.0
Mean
31.4
32.4
31.9
Means in a column and row followed by the same letters are
not significantly different (p = 0.05).


157
was based on amounts of rainfall received in the last two
cropping seasons, 214 mm and 374 mm. Hence, the pots
received only half of a season's rainfall. The quantity of
water applied to each pot was determined as the quantity
required to replenish loss of "available soil water," i.e.,
the difference between field capacity and wilting point.
For the 100 mm rainfall equivalent, the amount of water
required per pot corresponded to the application of 0.2
liters of water a day; for the 200 mm rainfall equivalent,
the level was 0.4 liters a day. The pots were irrigated
twice a week.
Statistical Analysis
ANOVA of split-plot design was used. For the field
studies of the two species involving measurements near and
away from the hedgerows, the option for repeated measures in
the GLM procedure of the SAS (SAS, 1991) was included.
Repeated measures option was necessary in order to adjust
for the fixed (as opposed to random) effects of crop
positions from the hedgerows and/or seasons. Analysis of
the data from the pot studies was similar to that for field
study described above since the soil was collected from
fixed positions near and away from the hedgerows.
In the field study involving irrigation and
fertilization, standard ANOVA of randomized block design was


194
In addition to soil moisture, it is also possible that
crop yields were reduced more under leucaena compared to the
sole crop or the crop under cassia due to competition for
nutrients. A basis for this hypothesis is provided by
studies by Chirwa (1991) in semiarid Zambia, where
reductions in maize yields under leucaena hedgerow
intercropping were observed to be less with the application
of fertilizer and more without fertilizer application.
The lower content of soil water under leucaena
compared to cassia would suggest that leucaena has a higher
root density than cassia. This was, indeed, observed to be
so. Root density of leucaena was higher than that of cassia
at nearly all distances considered (up to 3 m away from the
hedgerows) in the top 40 cm soil profile (Chapter 2).
Studies by Jonsson et al. (1988) also showed that both
leucaena and maize had their highest density of fine roots
in the same soil depth, suggesting that the potentials for
competition are high.
Two implications emerge from the observed differences
in the rooting density and water depletion patterns of the
two species. First is that leucaena would allow less
intercropping than cassia. This is particularly so in
semiarid areas where water, and to a lesser extent, soil
fertility, are major constraints to production. Yield of
crops with significant amount of their roots in the 0-80 cm
soil depth could be adversely affected under leucaena due to


7
lost if they were grown separately. These include mostly
gradual changes in soil organic matter, nutrient content and
physical properties such as bulk density, water infiltration
and porosity. The presence of the hedges in situ may,
through their root turnover and reduction of soil erosion,
sustain soil organic matter (Young, 1989). This may be of
particular importance in low-input agricultural systems
where all above-ground biomass is removed from the field.
If the contribution of the in situ presence of the
hedgerows to soil fertility relative to that of the mulch is
negligible, then planting the hedges separately from crops
and then transferring their mulch to crop fields would be a
more logical approach than growing the two together. Even
then, an understanding of the relative roles of the quantity
and quality (i.e., decomposition rates) of the mulch
produced would be relevant as this may enhance the
efficiency of its use.
In the humid tropics, mulch yields from the hedgerows
are high (in the range of 8-10 t ha"1 yr"1) Also, the rates
of decomposition are rapid. This makes nutrients in the
mulch rapidly available, for instance, in about 30 days in
the case of leucaena (Budelman, 1988; Gichuru and Kang,
1989). So, the mulch, in essence, resembles inorganic
fertilizer that provides a readily available supply of
nutrients to the crop. Crop response when mulch is removed
is therefore observable within a growing season. In the


34
significant. However, leucaena always yielded significantly
more mulch than cassia.
A count of all plants in the three replicates in the
last season of cropping showed that cassia had a higher
percentage of dead plants than leucaena (using the number of
seedlings planted in 1987 as a base). In cassia, up to 5%
( 7% standard error) of the intercropped hedges and 11.6%
( 15% standard error) of the block planted plants were
dead. In contrast, no plants were recorded as dead in
either leucaena systems. Cassia's mortality was largely
attributed to a commonly observed stem borer, Xylentes
capensis (Cassidae) and a wood boring beetle, Xyloposus sp.
(Bostrychidae).
Nutrient concentrations and contents of the mulches
In addition to a supply of N and P, the mulch provided
substantial quantities of other nutrients to the soil (Table
2-2). While concentrations of the other nutrients remained
stable over time in both species, the concentration of P
increased considerably. For instance, the levels of P in
cassia mulch increased 25% from 1987 to 1991 (compare 0.20
to 0.25%).
Crop yields
The combined analyses of the data over seasons
indicated that there were significant interactions between


102
Treatments
In the first season, decomposition patterns of soil
incorporated mulch (leaves plus twigs) of leucaena and
cassia were evaluated. A randomized block design with three
replications was used. On March 23, 1991 (which was the
beginning of the first season), fresh mulch samples of
leucaena and cassia weighing 50 g (20 g dry weight) were put
into litter bags (33 x 33 cm with 5 x 5 m mesh size) and
incorporated into the soil at a depth of 10 cm. The twigs
had diameters ranging from 5-10 mm. At that time, no
rainfall had yet occurred. The mulch was obtained from the
hedgerows at the end of the dry season. Senescence of the
leaves and/or low production of new foliage reduced the
proportion of leaves to twigs to 1:1 (w/w). During the
rainy season, however, the leaf ratio of the mulch would be
expected to be more.
In the second season, mulch of similar composition as
in the first season was soil incorporated on October 24,
1991. Some additional treatments were included in this
season. These were
1. Litter bags of mulch (leaves plus twigs) placed on
the soil surface and soil incorporated;
Soil incorporated and surface placed twigs (i.e.,
excluding the leaves) in litter bags;
2.


163
the hedgerows), the first maize row next to the hedgerows
was 15% and 21% more productive under leucaena and cassia,
respectively, than the next row.
Crop Nutrients (N and P) and
Water Content Measurements
No significant effects were observed due to distance
of crop rows from the hedgerows of both species on N and P
content of the crop grain. Similarly, there was no
significant difference in N and P content of the biomass of
maize sampled before final harvest of the crop in all the
three seasons. For example, in the sixth season with very
low rainfall, biomass N of maize near and away from the
hedgerows was not significantly different (Table 5-1).
At the time of maize biomass sampling in the sixth
season, maize was severely stressed by drought. This
observation was most pronounced away from the hedgerows. It
was then hypothesized that at the time of maize stress,
plants near the hedgerows would have higher water content
than those away from them. However, there were no
significant differences in water content of plants near or
away from the hedgerows (Tables 5-2 and 5-3). Nevertheless,
trends of higher water content in plants near the hedgerows
than those away from them were detectable.
Differences in plant water content were more apparent
between plants under the two hedgerow species than between
positions of plants within a given hedgerow species. For


CHAPTER 4
DYNAMICS OF SOIL FERTILITY CHANGES: RATES OF
DECOMPOSITION, N-MINERALIZATION AND UPTAKE BY
MAIZE FROM MULCH OF LEUCAENA LEUCOCEPHALA
AND CASSIA SIAMEA UNDER SEMIARID CONDITIONS
AT MACHAROS, KENYA
Introduction
The use of leguminous plants as a source of nitrogen
for crops is particularly important in many parts of the
tropics where fertilizer use is often economically not
feasible. Agroforestry practices, and in particular alley
cropping or hedgerow intercropping is a technology that is
often mentioned as having the potential to improve soil
fertility and productivity in the tropics (Kang et al.,
1990) One of the basic principles of the management of
alley cropping is the periodical pruning of the hedgerows to
reduce their shading and competition with the food crops
(Kang et al., 1985). The prunings left as mulch on the
surface between the hedgerows are claimed to improve the
physical, chemical and biological properties of the soil
(Kang et al., 1984). The contribution of the plant residues
to the fertility of the soil will largely depend on the
amount of biomass so applied and on its rates of
decomposition.
The decomposition of plant residues is known to be
primarily affected by the N content of the biomass (Campell,
97


136
any one site, an understanding of the mulch factors that
regulate rates of decomposition is important.
Mulch Quality and the Rates of Decomposition
and N Mineralization
First phase
The rates of decomposition of the two phases appeared
to be regulated by different factors. In the first phase,
the C:N ratio seemed to be the main factor for both species.
Generally, organic materials with high N concentrations or
narrow C:N ratio decompose and mineralize fast unless
constrained by environmental and plant factors such as
polyphenols (Parr and Papendick, 1978; Vallis and Jones,
1973). The C:N ratios of leucaena and cassia mulch, 11:1
and 13:1, respectively, are below the critical ratio of 33:1
that Bartholomew (1965) had mentioned as the limit beyond
which immobilization may occur. Also, the N concentrations
of the leaves of both species were above the critical level
of between 1.5% and 2.5% that is generally considered as the
minimum level before net mineralization occurs (Allison,
1973; Stevenson, 1986; Fox et al., 1990). As a consequence
of narrow C:N ratio, the rates of N mineralization of the
mulch high with more than 50% of the mulch-N in both species
released within the first four weeks after soil
incorporation. The rates of N-mineralization declined
significantly after eight weeks which would perhaps suggest
occurrence of net N-immobilization.


17
Kandic Rhodustalfs according to the United States Department
of Agriculture Soil Taxonomy). The soils of the research
station are in general weakly to moderately leached, with a
weakly acid reaction (pH [water] 6.0-6.5), a medium base
saturation (50-80%), low to moderate levels of organic
matter (top soil organic carbon: 1%-1.5%), nitrogen (N)
(0.08%) and phosphorus (P) (<16 ppm). Other nutrients are
reportedly present in adeguate levels to support crop
production (Kibe et al., 1981).
Clay and sand content in the top 15 cm averaged 28%
and 67%, respectively, in 1987. Kaolinite is the
predominant mineral in the clay fraction. The top soil had
an effective cation exchange capacity of 9.0 meq per 100 g
of soil at the initiation of the study in October 1987.
Methods
Field study
Experimental layout.Hedgerows of Leucaena leuco-
cephala and Cassia siamea were established in October 1987.
Alley widths of the hedgerows were 4, 5 and 6.7 m. In-row
spacing of trees within the hedgerows was 0.5 m. Alley
widths used were computed based on field observations (Nair,
1987) that 3-4 year old hedgerows, spaced 4 m apart and
pruned frequently to low heights, occupied 25% of total
(aboveground) land area. Similarly, the land occupancy
ratio by the hedgerows spaced 5 and 6.7 m would be 20% and


210
interviewed) let their livestock into the crop fields after
harvest.
Weed control
On average, crops were weeded twice in a season.
Labor for weeding was mainly family labor. However, labor
was often hired or exchanged with neighbors. The difference
in the frequency of weeding by the two interest groups
(fodder and mulch) was not significant.
Sixty-eight percent of the farmers interviewed
responded positively to exchange of labor with other members
of the community. Forty-two percent of this exchange labor
was for help with weeding. Only a small percentage, 3% and
4%, respectively, of the exchange labor was for crop
planting and harvesting. Farmers who did not exchange labor
often cited exchange of labor as becoming increasingly more
expensive in terms of providing for food for work done.
In terms of hiring labor, 60% of the farmers
interviewed hired labor to help with weeding and harvesting.
However, labor for harvesting was mainly for picking coffee.
Seventy-eight percent of the fodder-interest group and 50%
of the mulch-interest group hired labor. This difference
was significant.


3.0
2.5
0 L-, ¡ r
0 2 4 6
MULCH
Cassia
(incorporated)
e
Leucaena
(incorporated)
B- .
Cassia
(surface)
Leucaena
(surface)
8 10 12 14
Time (weeks)
-e-
16
18
120


2
fallowing results in satisfactory regeneration of soil
fertility only when the length of the fallow is adequately
long. The length of the fallow-period or rest-period
requirement of a soil depends on a number of factors, such
as soil type, climatic conditions and land-use practices
(Young, 1989). For example, on Alfisols in semiarid zones,
it is estimated that fallow periods of one to two years out
of every three years are required even with intermediate
levels of inputs. With the increasing population, fallow
periods have either been reduced drastically or totally
abandoned in many places. Thus, keeping the land in fallow
as a means of sustaining soil fertility is no longer a
feasible management option.
Use of fertilizers is an option considered by many to
increase soil fertility and productivity. Chemical
fertilizers are, however, seldom used by many resource-poor,
tropical, subsistence farmers under conditions as in
Machakos (Ockwell et al., 1991). Fertilizers are expensive
and/or not readily available (Mbogo, 1991). Application of
farmyard manure is another widely adopted practice which, in
fact, is suggested to be the principal source of nutrients
for crops on most farms in Machakos (Probert et al., 1990).
However, the quantity of manure available or applied is
often too low to sustain continuous crop production at
meaningful yield levels (Ikombo, 1984; Ockwell et al.,
1991) .


188
three months (August, November and December). At soil
profiles below 20 cm, however, no trend was observed of
higher water content near compared to away from the
hedgerows, suggesting that the influence of the hedgerows
was probably confined more to the top soil and less to the
sub-soil.
One explanation for the lack of significant
differences in top soil (0-20 cm) water content near the
hedgerows and away from them could be the higher
transpirational loss associated with the high yield near the
hedgerows. The magnitude of water loss through
transpiration could have been large enough to result in
similar levels of soil water content near the hedgerows.
Away from the hedgerows, the yield of the crop was low,
therefore transpirational loss was equally low. Also,
complete or partial closure of stomates, an adaptive
mechanism of plants under water stress (Slatyer, 1967),
could have minimized losses of water from soil distant from
hedges. If these physiological processes were indeed
operational, their net effect could result in similar levels
of soil water content near and away from the hedgerows.
Also, the lack of significant differences observed in the
relative water content of plants (or of leaf discs) near the
hedgerows and away from them is probably a reflection of the
high and low consumptive use of soil water by the crop near
and away from the hedgerows.


40
(control) yielded significantly higher than all the other
treatments except the cassia alley cropping system. Lack of
major differences between maize yields of the various
treatments across the seasons of the study is evident from
Table 2-3. In general, maize yield was positively related
to the rainfall patterns. However, in the third season, in
spite of the relatively high rainfall (631 mm), the average
of all treatments was only 2.2 t ha'1 which was similar to
the yield obtained in some other seasons with lower rainfall
(e.g., seasons 1 and 6 with rainfall of 330 mm and 373 mm,
respectively). The low yield with 631 mm rainfall could be
ascribed to excessive soil loss and runoff that also caused
physical damage to the crop; about half the rainfall was
estimated to have been lost in the runoff (data not
presented).
Maintenance of maize yield over time
Trends of maize yield over time were analyzed by
expressing the yield of each treatment in a season as a
percent of the average yield of all treatments of the second
crop (Table 2-4). The yield of all treatments generally
remained stable over seasons, including that of the control
with no inputs. Yield increased or declined mainly in
direct response to the season's rainfall, with a linear
relationship between rainfall and yield (Fig. 2-5). This


Table 3-1. Chemical and physical characteristics of the top
soil (0-20 cm) at the beginning of the study,
October 1987.
Soil Property
Leucaena
plots
Treatments
Cassia
plots
Control
plots
Mean
pH (1:2.5 soil:H20)
6.2
6.2
6.2
6.2
Org. C (%)
0.9
1.0
0.8
0.9
Total N (%)
0.1
0.1
0.1
0.1
P (ppm, Mehlich-1)
12.7
12.0
12.7
12.5
CEC (meq/100 g soil)
9.6
9.5
9.0
9.4
Clay (%)
28.7
30.7
26.7
28.7
Bulk density (g/cm3)
1.3
1.3
1.3
1.3
At the time of soil sampling in 1987, the experimental
treatments had not yet been introduced.


182
6
Control (no inputs) Fert. Irrig.
Treatments
Figure 5-12. Effects of irrigation and
in combination or alone, on maize grain yield
(no alley cropping with hedgerows of Leucaena
and Cassia siamea), 1991.
b
Irrig. + Fert.
fertilization,
in the field
leucocephala


Page
3-2. Chemical characteristics of the dry mulch
of alley cropped hedgerows of Leucaena
leucocephala and Cassia siamea 81
3-3. Chemical and physical characteristics of
the top soil (0-20 cm) under alley cropping
systems with hedgerows of Leucaena leucocephala
and Cassia siamea in November, 1991 84
3-4. Changes in soil Mehlich-1 P levels after
three years of alley cropping hedgerows of
Leucaena leucocephala and Cassia siamea with
maize, November 1991 .............. 85
3-5. Concentration of phosphorus in the mulch
(leaves plus twigs) of alley cropped
hegderows of Leucaena leucocephala and
Cassia siamea at the start (1987) and
end (1991) of the study ........ 87
3-6. Relative indices of KCl-extract N from
ion-exchange resins incorporated into the
soil of alley cropped hedgerows of Leucaena
leucocephala and Cassia siamea with maize
during the "short-rains" cropping season
of 1991 88
4-1. Chemical and physical characteristics of the
mulch of alley cropped hedgerows Leucaena
leucocephala and Cassia siamea 106
4-2. Intercepts, slopes, spline points and
coefficients of determination (r2) of the
regression lines for phase one and phase
two of leucaena and cassia mulch
decomposition ................. Ill
4-3. Decomposition parameters of soil incorporated
and surface placed Leucaena leucocephala and
Cassia siamea twigs placed outside the
litter bags, second season of 1991 ...... 118
4-4. Effects of soil incorporated Leucaena
leucocephala and Cassia siamea mulch on
the biomass yield, N-concentration, N-yield
and apparent N recovery of maize planted in
pots, 1991 128
viii


114
of leucaena leaves on the surface or within the soil.
Decomposition rate of leucaena leaves was significantly
higher than surface-placed cassia leaves. Soil-incorporated
and surface-placed leucaena leaves had physically
disappeared by weeks 7.5 and 6.5, respectively. The rate
of decomposition of leucaena leaves, averaged across
placement methods, was 15.4% per week, with a corresponding
half-life of 4.5 weeks.
For cassia, soil-incorporated leaves had significantly
higher rates of phase one decomposition than surface-placed
leaves. The rates of phase one decomposition of cassia
i
leaves within the soil and on the soil surface were 13.1%
and 11.1% per week, respectively. Half-lives corresponding
to these rates were 5.3 and 6.2 weeks. Phase two of the
soil-incorporated cassia leaves started at week 8.5 while
that of the soil-surface-placed leaves started at week 9.2.
The differences in the start of phase two decomposition of
the soil-incorporated or surface-placed cassia leaves were
not significant. Also, the rates of leaf decomposition in
phase two of the soil-surface-placed and soil-incorporated
leaves were not significantly different. On average, the
leaves decomposed at a rate of 2.3% per week in the second
phase. Half-life corresponding to the average rate of
decomposition in the second phase was 30.1 weeks.


82
Relative availability of N03" N and NH4+ N was
monitored in the last season. Ion exchange resin bags
(Binkley and Matson, 1983; Binkley and Hart, 1989) were
used. In each plot/replicate, one bag for anions and one
for cations weighing 5 g were placed in the top 10 cm soil
profile. After the harvest of the crop, the resins were
retrieved and N03" N and NH4+N extracted with 2 N KC1.
Levels of NO," N and NH4* N from the KCl-extracts were
determined using Standard Autoanalyzer methods at the Soil
Testing Laboratory of the University of Florida,
Gainesville.
Soil physical properties
Bulk density (Blake, 1965) is the only physical
property of the soil discussed in this chapter. In the last
cropping season, soil was sampled with a semi-wide auger
and 5 cm long to a total depth of 20 cm (i.e., 0-5, 5-10,
10-15 and 15-20 cm). Bulk density of each plot/replicate
was computed as the average values of samples taken at two
positions near the hedgerows (0.45 m) and two in the middle
of the alleys (3.5 m away from the hedgerows).
Statistical Analysis
ANOVA procedures were used to determine statistically
significant (p = 0.05) differences between the levels of


Maize yield (kg mm*1 of rainfall)
46
Figure 2-8. Productivity of maize (kg nun'1 of
rainfall) under alley cropping of Leucaena leucocephala and
Cassia siamea hedgerows for six cropping seasons. Vertical
bars indicate standard error of difference of means.


107
ha"1) and two levels of irrigation (i.e., the equivalents of
200 and 400 mm of rainfall received in a cropping season of
16 weeks) to potted maize. Each factorial combination of
species, mulch rate and irrigation levels was replicated
thrice in a completely randomized block design. Pots with
one level of fertilizer, the equivalents of 40 kg N ha"1 in
the form of mono-ammonium phosphate and pots with no inputs
(i.e., mulch or fertilizer) were also included as controls.
In this chapter, only the effects of one of the mulch
rates (i.e., 2 t ha"1, averaged across the two irrigation
levels) is discussed to demonstrate the patterns of mulch-N
mineralization and uptake. The 2 t ha'1 mulch rate was
equivalent to the highest amount of mulch obtained from the
hedgerows in a season and applied to the fields (Chapter 2).
Polythene cylinders (30 cm long and 30 cm wide) closed
at the bottom were used as pots. The pots were filled with
the top 20 cm soil collected and composited from positions
near (0.45 m) and distant (3.5 m) from intercropped hedgerow
plots of leucaena and cassia. Soil in each pot weighed 20
kg when dried to 8% moisture content (w/w) with a bulk
density of 1.53 g cm"3.
To obtain an index of N availability from the mulch,
5 gm of ion-exchange resin (Binkley and Matson, 1983;
Binkley and Hart, 1989) sealed in bags made of plastic
stocking were placed in each pot about 5 cm above the bottom


27
and P removed in crops of seasons 2-4. The regression
equations were obtained first by multiplying the N and P
concentrations of the crop (grain, cobs, stover) by their
respective yields to obtain N and P yields of the crop. The
N and P values were then regressed against the crop yields
which resulted in the following equations:
N(kg) grain yield = 0.015 grain yield 0.0006; r2=0.98
N(kg) cob yield = 0.0028 cob yield + 0.0002; r2=0.95
N(kg) stover yield = 0.0056 stover yield + 0.0008; r2*=0.91
Similarly, regression equations for P yield in grain, stover
and cob yields are as follows:
P(kg) grain yield = 0.0018 grain yield + 0.0002; r2=0.93
P(kg) stover yield ** 0.0021 stover yield -0.001; r2=0.92
P(kg) cob yield = 0.0005 cob yield 0.00002; r2=0.96
Nitrogen remaining in mulch not decomposed was calculated
using the N concentration of the soil-incorporated mulch at
the end of the study in March 1992. Typically, N remaining
in mulch is calculated using the decomposition constant of
the mulch (Palm, 1988). However, where the pattern of
decomposition of the mulch has more than one phase as was
the case in the present study (Chapter 4), use of one
decomposition constant for both phases may not be
appropriate. Nitrogen remaining was calculated for each
season's mulch input into the soil and then summed over all
the seasons.


96
With respect to soil P, significant increases were
observed in plots with cassia compared to those with
leucaena at the end of the three years period. Also, plots
with cassia had higher levels of available soil P compared
to the other treatments in the final year of the study. The
higher P level under plots with cassia was attributed to the
considerable increase in the concentration of P in the mulch
of cassia over time. Explanation for the higher
concentration of P in cassia mulch compared to that of than
leucaena is a matter of speculation; some hypothesis are
differences in root to shoot ratios, mycorrhizal
association, microbial population and activity (e.g., P-
solubilizing bacteria).
The limited amount of mulch available from the
hedgerows of leucaena and cassia may have short-term or
immediate benefits on soil fertility, as is evident from the
increases in soil available N and P, but may not have
significant impacts on soil organic carbon. The contribution
of the in situ presence of the hedgerows was not
significant, during the study. Therefore, there may not be
any special advantages, in terms of soil fertility
improvement, from growing the hedgerows within the crop
field.


215
5. Prior knowledge of alley cropping.
Table 6-3 provides frequency distribution of farmers under
each of these factors against the interests of farmers in
alley cropping.
Use of ox-plow and fertilizer by the fodder-interest
group was significantly higher than by the mulch-interest
group. On the other hand, frequency of farmers using mulch
on fields was significantly greater for the mulch-interest
group than the frequency for the fodder-interest group.
The fodder-interest group was exposed more to alley
cropping (45%) than the mulch-interest group (35%), although
the difference was not significant (p = 0.42). This
difference in prior knowledge of the technology did
influence significantly the farmers' proposed use of the
technology. Between the two interest groups, 67% of the
fodder-interest group had coffee and only 35% of the mulch-
interest group had coffee. This difference was significant.
The use of chemical fertilizers, hired labor, ox-
plows, purchased manure, purchase of animal feed, frequency
of weeding, and purchase of food were all observed to be
significantly greater for the coffee than for the noncoffee
farmers. All the other factors considered (e.g., farm size,
number of livestock, number of household members doing off-
farm work, etc.) were not significantly dependent on
possession of coffee. Table 6-4 provides the p-values of


Table 3-3. Chemical and physical characteristics of top soil (0-20 cm) under alley
cropping with hedgerows of Leucaena leucocephala and Cassia siamea hedgerows,
November 1991.
System
PH
(1:2.5
soil:H20
Organic
C
(%)
Total
N (%)
KCl
extract
N (pg/g
soil)*
Mehlich-1
P (ppm)
CEC
(meg/
100 g
soil)
Bulk
density
(g/cm3)
Aliev cropping
Leucaena
6.3a
0.83*
0.09*
9.91*
16.7*
8.42*
1.23*
Cassia
6.2*
0.81*
0.09*
8.15*
23.0b
7.67*
1.46*
Block planting
Leucaena
6.2*
0.94*
0.08
8.55*
18.2b
7.63*
1.37*
Cassia
6.2*
0.83*
0.09*
9.17*
19.5b
9.26*
1.33*
Controls
Mulch
(leucaena)
Mulch
6.2*
0.80*
0.11
8.09*
11.3*
7.21*
1.31*
(cassia)
6.6*
0.93*
0.06*
7.43*
30.0C
9.68*
1.34*
Fertilizer
6.3*
0.72*
0.07
6.14*
21.2b
7.60*
1.20*
Maize-only**
6.3*
0.85*
0.07
6.77*
11.0*
9.33
1.44*
a,b,c,dIn a coiUnm,
values with
different
letters
are signific
antly different (p =
0.05).
Indicates average of 3 sampling dates.
From an adjacent experiment (no fertilizer, no mulch).
oo
t


Root density (cm cm3)
54
0.12
0.1
Leucaena (0-20)
Cassia (0-20)
0.08
0.06
0.04
0.02
0
Leucaena (20-40) ^ Cassia (20-40)
80 100 120 140 160 180 200 220 240 260 280 300
Distance (cm) from tree hedges
Figure 2-9. Fine root density in the top 40 cm soil
profile of alley cropped hedgerows of Leucaena leucocephala
and Cassia siamea.


211
Fallowing of land
Because the farms were generally of small acreage,
only 18% of the farmers interviewed mentioned fallowing as a
practice. The difference in the use of land fallowing
between the fodder and mulch interest groups of alley
cropping was not significant (p = 0.54).
Use of mulch
Use of mulch was common. Typically, mulch materials
were crop residues, lopping from trees on the farms and
prunings from farm fence bushes. Many farms had Lantana
caara and Tithonia diversifolia bushes as live fences.
Prunings from these shrubs provided fodder and firewood when
dry and mulch when green. When fresh, lantana could be
poisonous to cattle. Sixty-seven of the farmers interviewed
responded positively to the use of mulch. Of this number,
60% were those with fodder interests and 85% were those with
mulch interests. This difference between the two groups was
significant.
Intercropping
Forty-five percent of the farmers interviewed
practiced intercropping. Of these, the percentages of
fodder and mulch-interest groups were similar (46%).
Intercrops of maize and beans (both phaseolus and vigna) or
maize and pigeon pea were common. Of the two, maize and


Figure 4-4. Decomposition patterns of the mulch (leaves plus twigs) of
Leucaena leucocephala and Cassia siamea after 16 weeks.
Cassia (incorporated): Y = 2.3 0.22X 0.01 (X-11.3); r2 = 0.96.
Leucaena (incorporated): Y = 2.3 0.23X 0.02 (X 8.0); r2 = 0.95.
Cassia (surface): Y = 2.4 0.25X 0.04 (X 8.5); r2 = 0.97.
Leucaena (surface): Y = 2.2 0.33X 0.09 (X 6.0)*; r2 = 0.98.
Indicates presence of significant difference from the other treatments.


71
well-established roots is unavoidable. More often than not,
the agronomic crop is the loser in the competitive situation
as indicated, in this case, by the low maize yields under
leucaena. Studies of other workers from the semiarid
tropics also show low maize yields under leucaena alley
cropping (Nair, 1987; Singh et al., 1989; Corlett et al.,
1989; Rao et al., 1990, 1991; Ong et al., 1992). The
present study also shows that all species are not as
competitive as leucaena with crops, and that an
understanding of the rooting system of species to be used
for intercropping is necessary.
Typically, fine root density of trees declines with
distance from the stem (Roberts, 1976; Ford and Deans, 1977;
Persson, 1980; Dhyani et al., 1990) which is consistent with
the observations of the present study. Along the entire 3 m
long trenches from the hedgerows studied, root density of
leucaena was significantly higher than that of cassia. The
extended high root-density of alley cropped leucaena into
the cropped s^rea and the resultant low yields of intercrops
would suggest that potentials of intercropping leucaena
without physical separation from crops are limited.
Approaches to minimize below-ground competition between
hedgerows and crops, at least experimentally, include root
pruning and/or placement of physical barriers such as
galvanized iron or polythene sheets (Singh et al., 1989; Rao


259
Ockwell, A.P.f Muhammad, L.f Nguluu, S., Parton, K.A.,
Jones, R.K., and McCown, R.L., 1991. Characteristics
of improved technologies that affect their adoption in
semi-arid tropics of eastern Kenya. Farming Systems
Research-Extension 2 (1), 29-46.
Oglesby, K.A., and Fownes, J.H., 1992. Effects of chemical
composition on nitrogen mineralization from green
manures of seven tropical leguminous trees. Plant and
Soil 143, 127-132.
Olson, R.K, and Reiners, W.A., 1983. Nitrification in a
subalpine balsam fir soils: tests for inhibitory-
factors. Soil Biology and Biochemistry 15, 413-418
Ong, C.K., 1991. Interactions of light, water and nutrients
in agroforestry systems, pp. 107-124. In: M.E. Avery,
M.G.R.Cannell and C.K. Ong (eds.), Biophysical research
for Asian agroforestry. Winrock International, New
Delhi.
Ong, C.K., Corlett, J.E., Singh, R.P., and Black, C.R.,
1991. Above and below ground interactions in
agroforestry systems. Forest Ecology and Management
45, 45-57.
Ong, C.K., Rao, M.R., and Mathuva, M., 1992. Trees and
crops: competition for resources above and below the
ground. Agroforestry Today 4 (2), 4-5. ICRAF,
Nairobi, Kenya.
Palada, M.C., 1989. On-farm research methods for alley
cropping, pp. 84-91. In: Kang, B.T. and Reynolds, L. ,
(eds.), Alley farming in the humid and sub-humid
tropics. IDRC, Ottawa, Canada.
Palm, C.A., 1988. Mulch and nitrogen dynamics in an alley
cropping system in the Peruvian Amazon. Ph.D.
Dissertation. North Carolina State University,
Raleigh.
Palm, C.A., and Sanchez, P.A., 1991. Nitrogen release from
the leaves of some tropical legumes as affected by
their lignin and polyphenolic contents. Soil Biology
and Biochemistry 23 (1), 83-88.
Parker, D.T., 1962. Decomposition in the field of buried
and surface-applied corn stalk residue. Soil Science
Society of America 26, 559-562.


212
pigeon pea intercrop was the more preferred. The beans in a
maize and bean intercrop ripened earlier (in about two
months) than the maize (four months). Early ripening of
beans made their harvest difficult in the midst of the maize
stands. Also, in the process of harvesting beans, damage to
the immature maize crop could be severe. Beans in a maize
intercrop were also observed by 52% of the farmers
interviewed to have competitive effects. On the other hand,
pigeon pea is an perennial crop and its harvest occurs long
after maize is out of the fields.
The presence or absence of intercropping appeared to
be influenced most by farm size. Intercropping was
predominantly a practice on the small farms. Of those
farmers that intercropped, 60% had farm sizes of two or less
hectares. On the other hand, 85% of those farmers who did
not intercrop had farm sizes greater than two hectares.
Use of improved seeds
Farmers bought seeds, in particular those of the
drought tolerant maize varieties (e.g., Katumani Composite).
Sixty-seven percent of the farmers interviewed bought seeds.
Also, the purchase of seeds was mentioned by most farmers to
be the one farm input that required most money. Farmers who
did not buy seeds saved some from their harvests. Of those
purchasing seeds, the percentages of the fodder and the


Page
4-5. Relative indices of cumulative NH4+ N, N03 N
and total N mobilized 16 weeks after soil
incorporation of 2 t ha'1 equivalents of
Leucaena leucocephala and Cassia si amea
mulch in pots with maize, 1991 ........ 131
5-1. Nitrogen content of 30 day-old maize shoots
(leaves plus stalks) sampled near (0.45 m) and
away (0.45 mm) from hedgerows of Leucocephala
leucocephala and Cassia siamea, fifth season,
1991 164
5-2. Relative water content of flag leaves of
maize plants sampled near (0.45 m) and
away (3.5 m) from alley cropped hedgerows
of Leucaena leucocephala and Cassia siamea,
on April 26, 1991 (sixth season) 165
5-3. Water content of whole maize plants sampled
near (0.45 m) and away (3.5 m) from alley
cropped hedgerows of Leucocephala leucocephala
and Cassia siamea on April 2*6, sixth season,
1991 166
5-4. Relative water content of flag leaves of maize
plants sampled near (0.45 m) and away (3.5 m)
from alley cropped hedgerows of Leucaena
leucocephala and Cassia siamea leucaena and
cassia on May 28, 1991 (sixth season) 168
5-5. Relative water content of flag leaves of maize
plants sampled near (0.45 m) and away (3.5 m)
from alley cropped hedgerows of Leucaena
leucocephala and Cassia siamea on January 6,
1992 (seventh season) 169
5-6. Saturated hydraulic conductivity of soil near
(0.45 m) and away (3.5 m) from alley cropped
hedgerows of Leucaena leucocephala and Cassia
siamea, March 1992 (season seven) 173
5-7. Effects of three nitrogen rates and two
irrigation levels on the biomass yield of
maize after two months growth in pots, 1991 . 184
6-1. Farm and household characteristics of the 60
farmers interviewed at Machakos, Kenya,
August 1991 ...... 205
ix


Special thanks are given to Richard Coe of ICRAF and
Steven Linda, Department of Statistics, University of
Florida, for their suggestions and assistance in the
analysis of the experimental data. Thanks are also due to
forestry/agroforestry students Paramu Mafongoya, Todd
Johnson, Reinhold Muschler, Paul Campell, Loice Omoro,
Mutiah Govindarajan, Don Mee, Mark Follis, Chris Latt,
Antonio Brunori, and Anne Todd-Bockerie and to Christine
Kelly of agronomy department.
The secretarial services provided by Marie Kimenye of
ICRAF during the period of field research are highly
appreciated. Furthermore, appreciation is extended to Pam
White for her assistance in the preparation of this
manuscript.
Finally, the author wishes to express his deepest
appreciation for the support, understanding and sacrifice
made by his wife Fatuma (Nimoo), daughters (Aisha and
Khadija), father (who passed away in the while the author
was writing-up the dissertation"rahimuallah"), mother,
brothers, and sisters.
iv


72
et al., Ong et al., 1992), trenching (Verinumbe, 1985) and
root pruning (Fernandez, 1990).
With root pruning/barrier techniques, improvements in
crop yield compared to a control have been noted, with
significantly increased yields at some times (e.g., Corlett,
1989) and not at others (Ong et al., 1992). Root pruning
is, however, labor demanding and unless there are
significant gains to be foregone by not having the hedgerows
in situ, farmers would be most unlikely to adopt it.
Besides, roots of the hedgerows can arch back to the surface
soil even when physical barriers are placed to depths of one
meter below the soil surface and hence render barrier or
trenching efforts ineffective (Ong et al., 1992).
Separate planting of hedges in blocks or sole stands
on land outside the crops that is not currently cropped, or
on land not suitable for cropping or on the periphery of the
farm may be options to minimize competition of leucaena with
crops. This could be an option particularly for leucaena
which has a desirable product (fodder) but which cannot be
intercropped because it is competitive with crops. However,
expected losses in crop yield are proportional to the
percentage of land occupied by the hedges. This lost land
occurred because the amounts of mulch obtained from the
hedges were low (on average 1-2 t ha'1 season*1) and had no
compensatory effects on crop yield. However, the results of
the pot studies suggest that if the levels of mulch from


ACKNOWLEDGMENTS
The author would like to express his sincere thanks to
Dr. P.K'.R. Nair for his constant advice, patience,
sincerity, and moral support during the course of this
project. This research certainly would have been more
difficult if it were not for the support and encouragement
that Dr. Nair extended on many difficult occasions. The
author also would like to thank the other members of the
supervisory committee, Drs. H.L. Gholz, C.P.P. Reid, P.E.
Hildebrand, and H.L. Popenoe. Special thanks are due to Dr.
M.R. Rao of the International Centre for Research in
Agroforestry (ICRAF), Nairobi, for his field supervision,
support and suggestions regarding the dissertation. In
addition, thanks are due to Peter Kurira Wambugu, manager of
ICRAF Research Station, Machakos, Kenya, for supervising
field activities and data collection, especially when the
author was away in 1989-1990.
Appreciation is extended to ICRAF, the Ford Foundation
and the Rockefeller Foundation, who have provided financial
support for this research; and the National Dryland Research
Station (Katumani), Kenya, which has provided laboratory
facilities; and to ICRAF once again, for granting leave-of-
absence to the author during the study period.
iii


Figure 4-2. Decomposition patterns of the leaves of Leucaena leucocephala
and Cassia siamea.
Cassia (incorporated): Y = 2.8 0.26X 0.05 (X-8.5); r2 = 0.94.
Leucaena (incorporated): Y = 2.9 0.31X; r2 = 0.98.
Cassia (surface): Y = 3.0 0.11X 0.05 (X 9.2); r2 =
Leucaena (surface): Y = 3.0 0.31X; r2 = 0.97.
0.91.


12
element which more than any other is under biological
regulation in natural ecosystems (Swift and Sanchez, 1984).
Nitrogen available to crops in alley cropping may be
enhanced through processes of biological fixation of
atmospheric N (Roskoski, 1986; Dommerques, 1987) and
subsequent release through processes of nodule sloughing
(Sanginga et al., 1986) or excretion of nitrogen from the
root nodules (Brophy and Heichel, 1989; Wacquant et al.,
1989). However, the significance of the latter process
remains doubtful (Vallis, 1978).
Prunings from the hedgerows are the main source of
nutrients to the crops. This is evident from the
significant declines in crop yield that occur when the
application of mulch is withdrawn (Kang et al., 1981). The
extent of supply of N from the mulch to the crop would
depend, among other factors, on the rainfall conditions of
the site, the soil, the hedgerow species, and management
practices such as spacing and the pruning regime. In the
humid lowlands of West Africa, Kang et al. (1981) and Duguma
et al. (1988) reported a wide range of N yields of about
150-560 kg N ha'1 yr'1 for alley cropped leucaena. Similar
information from the semiarid tropics is scarce. A major
exception is the work of Lulandala and Hall (1990) in
Tanzania that noted nitrogen returns through frequent
pruning of leucaena hedges could be as high as 88 kg ha'1
yr'1.


Rainfall (mm)
16
1988 1989 1989 1990 1990 1991 1991
SR lr SR LR SR LR SR
Month and cropping season
Figure 2-1. Seasonal rainfall and the local names
(long-rains and short-rains) of the cropping seasons during
the study.


223
grow on the farms. With the exception of Eucalyptus
species, a popular tree among farmers, many of acacia
species and croton are termite resistant. Hence, survival
was high. Eucalyptus, on the other hand, is not resistant
to termite attack. But because Eucalyptus seeds heavily,
farmers plant many seeds and hope that some plants survive.
And, indeed, its survival was highly variable, with an
average of 53% (sd: 38%) across the farmers interviewed.
Availability of planting materials was cited by many
farmers as a big constraint. To ensure their own supply of
seedlings, 30% of the farmers interviewed had at one time or
another a nursery on the farm. The functions of the tree
nurseries possessed are shown in Table 6-6. Farmers
established nurseries because of the nonavailability of
seedlings from other sources. Government and non
governmental nurseries were either too far away, not
dependable or often did not carry species desired. In
addition, without one's own transport, one would not be able
to carry many seedlings up the hills. However, 70% of the
farmers interviewed did not possess a nursery; the reasons
are shown in Table 6-7.
The major problems of raising one's own nursery/-
seedlings are lack of water and availability of seeds. A
number of farmers said they were planting trees to have a
secure source of seeds. Because the area is semiarid, a
sufficient supply of water is a problem. Most of the


77
Notwithstanding the positive attributes and potentials
of alley cropping discussed above, competition for soil
resources (nutrients and water) between the hedgerows and the
crops has to be considered since this could result in
significant losses of crop yield under alley cropping (Rao et
al., 1991; Nair, 1987; Singh et al., 1989). The losses in
yield may, indeed, obscure any gains in soil fertility due to
the presence of the hedgerows. To justify mixing of multi
purpose tree hedgerows with crops, particularly by resource-
poor farmers, it is important, therefore, to understand the
benefits to soil fertility that would otherwise be lost if the
hedgerows were planted separate from the crops. This
knowledge is of particular significance in the semiarid
tropics where soil moisture is a major factor limiting crop
production and where soil fertility improvement aspects of
alley cropping have not yet been well understood.
This study was undertaken in this scenario with the
following objectives: (a) to evaluate the changes in some
selected soil chemical and physical properties under alley
cropping of maize with Leucaena leucocephala and Cassia siamea
over three years and, (b) to separate the effects of mulch per
se from the in situ presence of the hedgerow on the selected
soil chemical and physical properties.


141
of N-mineralization of the mulch in the field are
superimposed on the patterns of N uptake and biomass yield
of maize from the pot study. The rates of N release from
the mulch of both species were highest in the first four
weeks application, with 80-85% of the mulch N being
mineralized. After weeks 8 and 6 for leucaena and cassia,
respectively, the rates of N release were on a declining
trend. This would suggest N immobilization, an observation
made also by Oglesby and Fownes (1992). In field
conditions, synchrony may be difficult to establish because
of the dynamic equilibrium that exists between
mineralization and its control by immobilization and losses
through volatilization, denitrification and leaching (Woomer
and Ingram, 1990).
From the pot studies, maize biomass yield and N uptake
were highest in the period spanning 8-12 weeks after mulch
application. Typically, the period between the onset of
flowering or silking is when N requirements by maize is at
peak (Arnon, 1975) or when daily rate of N accumulation is
highest (Sayre, 1955). For a 120-day maize variety such as
the1 one used in the present study, the onset of flowering or
silking occurs between weeks 6 (Bromfield, 1969; Nadar,
1984) and 8 (Hanway, 1962). In weeks 4 and 16, when N
requirements are low, both biomass yield and N uptake of the
plants were significantly lower than those of weeks 8-12.
Splitapplication of the mulch, timing of application,
and placement method and controls on the quality of the


218
resources possessed or purchased that depended significantly
on farmer's possession of coffee. Even though the use of
fertilizer was significantly greater for coffee than for
noncoffee farmers, maize yields of the two groups were not
significantly (p = 0.35) different (Table 6-5). This
suggests that fertilizer was probably applied mainly to
coffee.
Factors Affecting Farmers' Strategies
to Improve or Maintain Fertility and
Productivity of Their Soils
Most of the strategies discussed above reguired cash
outlay. For example, "Fanya-Juu" terracing was expensive
and a major investment on many farms. Between the two
interest groups, the fodder-interest group evidently had
more purchasing power for farm inputs and household needs
(e.g., fertilizer, ox-plows, labor, feed for livestock, food
for the family, and firewood) than the mulch-interest group.
There were two major sources of cash income in the
area, viz., sale of coffee and off-farm employment. Many of
the families interviewed had one to two of their members
working off-farm. Forty-four percent of the farmers
interviewed had at least one member of the family engaged in
regular off-farm employment. Most of those engaged in off-
farm employment were men since 60% of the farmers found on
the farm during the survey were women.


230
(use of fertilizers, ox-plows, knowledge of the technology),
with the exception of mulch use, also depended significantly
on the presence or absence of coffee trees on the farm.
To invest in production strategies, farmers had to spend
money. Labor had to be hired for soil conservation works,
"Fanya Juu" construction and coffee picking, seeds had to be
purchased, and ox-plows had to be hired. The ability to
spend appeared to depend strongly on the possession of a
cash generating activity on the farm. In the area of the
present study, there were two main cash-generating
activities: off-farm work and sale of coffee. Other minor
sources of revenue included sale of fruits, wood products
(e.g., poles), and sisal baskets. Income from off-farm work
and from trees (fruit and nonfruit trees) was probably
significant to the household but did not differ
significantly among farms. What differed between farms was
income from coffee.
Farmers with coffee invested more in inputs (e.g.,
bought fertilizers, manure, possessed or hired ox-plow,
bought feed for their livestock) than noncoffee farmers. In
addition, coffee farmers hired more labor and weeded more
frequently. This suggests that farmers with coffeea cash
crophad greater purchasing power than those farmers
without coffee.


Figure 4-3. Decomposition patterns of the twigs of Leucaena leucocephala
and Cassia siamea.
Cassia (incorporated): Y = 0.97 0.25X 0.06 (X 6.5); r2 = 0.96.
Leucaena (incorporated): Y = 1.1 0.15X 0.04 (X 7.4)*; r2 = 0.85.
Cassia (surface): Y = 1.1 0.23X 0.06 (X 6.0); r2 = 0.77.
Leucaena (surface): Y 1.3 0.2IX 0.09 (X 7.5); r2 = 0.77.
*Indicates presence of significant difference from the other treatments.


174
Figure 5-6. Soil water near (0.45 m) alley cropped
hedgerows of Leucaena leucocephala and Cassia siamea
determined at the end of 1991 long-rains.


229
decision. On the other hand, some questions raised by the
mulch-interest group, when they visited the research
station, suggested that their initial response was suspect.
For example, on visiting the research station, both interest
groups raised serious questions about potential competition
between the trees and the crops.
Potentials for Adoption of Aliev Cropping
From the positive response of the farmers surveyed,
62% for fodder use and 33% for mulch use, it would appear
that there is potential for adoption of the alley cropping
technology. Only 5% of the farmers interviewed were
indifferent to the technology. Between the two end uses,
alley cropping is more likely to be adopted for fodder than
for mulch. Coffee farmers are more likely to adopt alley
cropping than those without coffee. For example, of the 62%
interested in fodder, 76% were coffee farmers. Of the total
number of farmers interviewed, 57% were farmers with coffee.
Like the various strategies used by farmers to improve soil
fertility and crop production (e.g., terracing, weeding),
alley cropping would require investments (e.g., purchase of
seedlings). Among the five factors identified to determine
farmers' interests in the use of alley cropping (fodder or
mulch), the presence or absence of coffee trees on the farm
was significant. The presence or absence of other factors


87
Table 3-5. Concentration of phosphorus in the mulch (leaves
plus twigs) of alley cropped hedgerows of
Leucaena leucocephala and Cassia siamea at the
start (1987) and end (1991) of the study.
Species
1987
1991
Leucaena leucocephala
o
CM

o
0.19
Cassias siamea
0.20
0.25
Values in a column with
significantly different
the same letters are not
(p = 0.05).


206
household? over 50% of the family members were children
under 18 years of age. Land ownership was family free-hold
title basis.
Of the 60 farmers interviewed, 62% were interested in
fodder and 33% in mulch. Three farmers (5%) had no interest
in alley cropping or were undecided. Farm sizes of the
fodder- and mulch-interest groups were not significantly
different, on average three hectares. Also, farm size did
not influence significantly (p = 0.67) the fodder or mulch
interests of the farmers interviewed (Table 6-2). The labor
requirements of alley cropping is often seen as a bottleneck
to its adoption. However, most of the farmers (88%) did not
see labor for pruning the hedgerows as a problem.
Maize was the main food crop, grown alone or in
combination with legumes (predominantly pigeon pea). All
farmers expressed insufficiency of food supply. Indeed, 67%
of the farmers interviewed bought their staple food (maize)
at one time or another. Between the two groups, 80% of the
mulch-interest group and 57% of the fodder-interest group
farmers bought food. Other food crops included beans and
various vegetables. Cash crops were coffee (mainly in two
locationsKivandini and Kiima Kimwe) and fruit trees
(mango, guava, citrus and avocado).
Farmers kept livestock (cows, goats, sheep and
chicken). In terms of cattle, the average number per farm
was two head (sd: 4), with a range from 0 to 12 head per


195
enhanced competition. Unfortunately, most of the annual
crops have their roots within the top part of the soil
profile. Crops that have either short periods of maturation
(e.g., cowpea) or that are drought tolerant (e.g., millet)
may provide opportunities for intercropping with leucaena
where this may still be necessary. Secondly, through root
turn-over, particularly of the fine roots, leucaena would be
likely to add more to soil organic matter than cassia. This
may improve both physical and chemical properties of the
soil. For instance, hydraulic conductivity of the soil
could be enhanced by macropores opened following the death
of roots (Barley, 1954), and more roots are under leucaena
than under cassia. But these benefits of soil improvements
under alley cropping with leucaena seems to be nullified by
enhanced tree-crop competition for soil moisture under
semiarid conditions.
Conclusions
This study demonstrates that:
1. Although the effects may be small, interaction
between tree hedgerows and crops could be
positive, even in semiarid environments.
2. The magnitude of the interaction effects on the
crop yield is dependent on the hedgerow species,
and more positive under cassia than under
leucaena.


58
alley cropping studies spanning over 4 years from semiarid
India.
Between the two systems of planting (i.e.,
intercropped vs block/sole stands), intercropped hedgerows
of leucaena yielded significantly higher (+33%) mulch than
the block planting on an egual area basis in all but in the
first year. In cassia, differences in the mulch yield of
the two systems were not significantly different. The
higher biomass of the alley cropped plots could be
attributed to (a) availability of more space and hence
reduced competition for soil water and nutrients (especially
when the annual crop was not in the field) and, (b) indirect
benefits from husbandry practices of the maize crop, e.g.,
fertilization, weeding and tillage. The higher mulch yield
of the intercropped hedgerows compared to the yields of the
block/sole stands indicates the benefits of growing the
hedgerows within the crops.
In the block planting system, although the hedges
occupied the same land area as the intercropped hedges, they
were confined to limited space and not spread out. Hence,
plant-to-plant competition was relatively severe.
Typically, the hedges at the borders of the block planting
system always grew taller and yielded more biomass than the
rows inside the blocks. Enhanced tree-to-tree competition
may have contributed to the slightly higher mortality,


129
of leucaena. Apparent N recovery is expressed as percentage
difference between N-yield of the maize biomass with mulch or
fertilizer application and the N-yield of the control
treatment, and then divided by the amount of N applied through
fertilizer or mulch.
In the first harvest, apparent N recovery from the mulch
of cassia was low and, in fact, significantly negative. This
was because of the low maize biomass yield from the cassia-
mulched pots compared to the control pots. Compared to the
first harvest, maize biomass as well as the N-yield of the
biomass of all treatments (including the control) was
significantly higher in the second harvest, resulting in lower
apparent N recovery of N applied compared to the first
harvest. This was most pronounced with cassia (0.6%). The
fertilizer treatment had significantly the highest recovery
(7.1%) of N applied.
In the third harvest, no significant differences were
observed in biomass and N-yield between treatments. In this
season, apparent N recovery rates from the mulch and the
fertilizer treatments ranged from 9.2-11.4 %.
In the fourth harvest, apparent N recovery by the maize
biomass from the mulch of leucaena was significantly higher
than that with cassia mulch or with fertilizer. Because of
the relatively lower maize biomass yield from pots with cassia
mulch compared to those of the other treatments in this
harvest, apparent N recovery from cassia mulch was negative


Maize grain yield (kg m'2 )
45
t
/
Figure 2-7. Maize productivity (kg m'2) under alley
cropping Leucaena leucocephala and Cassia siantea hedgerows
for six seasons. Vertical bars indicate standard error of
difference of means.


183
Biomass yield of maize in the pots increased in a
significantly linear manner with increase in N rates (Table
5-7), irrespective of the source of the soil in the pots,
i.e., near or away from the hedgerows, suggesting that the
status of the soil N near the hedgerows and away from them
was similar.
With respect to the two levels of irrigation, biomass
yield of the maize increased significantly in all pots with
increase in water (irrigation) level. Biomass yield of
maize from pots with the equivalents of 200 mm rainfall was
significantly higher than those with the equivalent of 100
mm.
Discussion
Crop Yield Profile and Explanatory Factors
The decline observed in maize grain yield with
distance from the hedgerows of both species has also been
reported by others at the ICRAF Field Station, the site of
the present study (Sang and Hoekstra, 1987; Mungai, 1991;
Ong et al., 1992). The general trend of higher yield near
the hedgerows is also comparable to observations of enhanced
production of grasses near single trees, a typical
observation in many semiarid ecozones (Radwanski and
Wickens, 1967; Belksy et al., 1989; 1992; Wilson et al.,
1990). The increase in yield nearer the hedgerows was most
pronounced in the second and third seasons. In these
f


233
The hedgerows, on the other hand, produced at most 2 t ha'1
season'1 from systems occupying 15%-25% of the land. The
potentials of mulch to improve yields is, therefore, limited
by the quantity of the mulch that the hedgerows can produce.
Theoretically, mulch yield could be increased by putting 50%
of the crop land to trees. It is unlikely, however, that
farmers with two to three hectares land would put 50% of
their land to trees/hedges.
Between the two tree species evaluated on the station,
leucaena was more competitive with crops than cassia.
Averaged over seasons, maize yield increased under cassia by
8% per season, while yields declined by 12% per season under
leucaena. However, the presence of mulch- and fodder-
interest groups suggests two recommendation domains
(research or extension): cassia for farmers with mulch-
interests and leucaena for farmers with fodder-interests.
The characteristics of the two tree species also lend
themselves to meeting the separate needs of the farmers.
Leucaena can be used for both fodder and mulch while cassia
has mulch but no known fodder value. Because leucaena is
competitive with crops, its planting would be outside the
crop fields which, as Arnold (1987) postulated, could
provide productive applications for under-utilized land,
labor or capital.


47
Land equivalent ratios (LER)
Alley cropping leucaena and cassia had LER greater
than 1.0 of maize grain yield or the combined yield of the
crop and hedgerow biomass. Averaged across species and
seasons, LER of maize grain yield under alley cropping was
1.24 (or 24% advantage over the maize grain of the sole
planting system of the hedges and crop). No particular
trend in LER of maize yield across seasons was discernible
under both species (Table 2-5). Between the two species,
LER of maize grain yield under cassia (1.36) was
significantly higher than under leucaena (1.12). This
difference was due to the lower crop yield under leucaena
compared to cassia.
When averaged across species, the combined biomass of
the hedgerows and the crop yield (grain plus stalks) had LER
of 1.10 (i.e., an advantage of 10% by alley cropping over
the sole system). The LER of maize grain yield under alley
cropping was greater than 1.0 because of lower yield of
maize yield under the sole system compared to alley cropping
in nearly all seasons. The biomass yield of the hedges
under alley cropping and sole planting systems were similar
when the highest yielding sole planting system was used in
LER computation.


196
3. Enhanced soil moisture near the hedgerows, more
than improved soil fertility, could explain the
higher maize yield observed near the hedgerows.
The hedgerows, particularly that of leucaena, improved
soil bulk density near them. This, in turn, enhanced the
rates of infiltration and hydraulic conductivity of the
soil. If the observed positive effects of the hedgerows on
the crop and the physical properties of the soil near them
are not site-, species-, and management-specific, then alley
cropping of multipurpose trees, particularly cassia, may
have a role in improving productivity of soils in the
semiarid tropics.


Mulch N mineralized (%)
123
Time (weeks)
Figure 4-5. Cumulative percent of N mineralized from
soil incorporated mulch of Leucaena leucocephala and Cassia
siamea after 16 weeks.


70
Root Density
The significantly higher yield of leucaena mulch and
presence of significantly higher density of its fine roots
in the top soil could be the causes of the lower maize yield
under alley cropping with leucaena compared with cassia. In
a competitive situation, the sharing of limited soil water
and nutrient resources by competing plants is proportional
to their effective rooting length/density (Bowen, 1985).
Generally, most of the fine roots of trees are found in the
upper 50 cm of the soil (Fayle, 1975; Deans and Ford, 1983).
High root density is an ecological adaptation of plants,
particularly those growing on soils with chemical and/or
physical limitation in the lower horizons (Bowen, 1973).
This is also true of plants in nutrient and water stressed
environments (Chapin, 1980). The studies of several workers
in both the humid lowland and the semiarid tropics (Kang et
al., 1981; Corlett, 1989; Singh et al., 1989; Rao et al.,
1991, 1992; Ruhigwa et al., 1992; Toky et al., 1992)
corroborate the high root densities of leucaena and cassia
found in the present study.
Unfortunately, most of the roots of annual crops and,
in particular those of maize, are found in the same soil
horizon as those of trees (Jonsson et al., 1988).
Therefore, competition for limited resources (e.g., soil
water and nutrients) between trees with well established
root systems and the seasonal crop with less


10D
Materials and Methods
Site Description
The study was conducted for two cropping seasons of
1991 at the Field Station of ICRAF located at Machakos,
Kenya (latitude Io 33' S and longitude of 37 14' E and 1596
m elevation; mean annual temperature, 19C). In the first
season, the total rainfall was 214 mm while that of the
second was 373 mm (Fig. 4-1). These monthly totals,
however, do not reveal the considerable variations that
occur within and between seasons. Monthly rainfall equals
or exceeds evapotranspiration only in April, May and
November. The general characteristics of the soils of the
study site, and in particular, that of the top soil (0-20
cm) are described in Chapter 3.
Methods
Field and laboratory studies
Field studies were carried out for both seasons of
1991 in alley cropping plots of leucaena and cassia. Maize
(Katumani Composite cv.) fertilized at the rate of 40 and 18
kg/ha of N and P, respectively, was planted in the alleys
between the hedgerows in both seasons.


LITERATURE CITED
Ahn, J.H., Robertson, B.M., Elliott, R., Gutteridge, R.C.
and Ford, C.W., 1989. Quality assessment of tropical
browse legumes: tannin content and protein degradation.
Animal Feed Science and Technology 27, 147-156.
Alexander, M., 1977. Introduction to soil microbiology.
Second edition. John Wiley and Sons, New York.
Allison, F.E., 1966. The fate of nitrogen applied to soils.
Advances in Agronomy 18, 219-259.
Allison, F.E., 1973. Soil organic matter and its role in
crop production. Developments in Soil Science 3.
Elsevier Scientific Publishing Company, Amsterdam.
Anderson, J.M., and Ingram, J.S.I., 1989. Tropical soil
biology and fertility program: A handbook of methods.
C.A.B International, Wallingford, U.K.
Arnon, I., 1975. Mineral nutrition of maize. International
Potash Institute, Bern, Switzerland, pp. 143-144.
Atta-Krah, A.N., and Francis, P.A., 1987. The role of on-
farm trials in the evaluation of composite
technologies: the case of alley farming in southern
Nigeria. Agroforestry Systems 23, 133-152.
Atta-Krah, A.N., Sumberg, J.E., and Reynolds, L., 1985.
Leguminous fodder trees in the farming systems: an
overview of research at the humid zone program of ILCA
in south-western Nigeria. ILCA, Ibadan, Nigeria.
Ayanaba, A., 1982. The value of mulches in the management
of organic matter in the tropics, pp 97-103. In:
Cerri, C.C., Athi, D., Sadrzcieski, D. (eds.) Regional
Colloquium on Soil Organic Matter Studies. Sao Paulo,
Brazil.
Baldwin, I.T., Olson, R.K., and Reiners, W.A., 1983.
Protein binding phenolics and the inhibition of
nitrification in sub-alpine balsam fir soils. Soil
Biology and Biochemistry 15, 419-423.
247


26
The productivity of alley cropping compared to the
sole planting system of the hedgerows and the crop was
assessed by calculating land equivalent ratios (Willey,
1979) based on grain and total biomass yield of grain, stalk
and the hedgerows. The block planting and alley cropping
systems with the same ratio of land apportioned to tree and
crop (15:85) was used.
Nutrient Budget of Nitrogen and Phosphorus
A simple budget for N and P was computed for selected
treatments during the last three cropping seasons of the
study. It was only in these seasons that nutrient content
of the crop was regularly determined. The N budget was
calculated by considering N applied through the mulch or as
fertilizer, N removed in crop (grain, cob and stover), and N
still on the plot in mulch not decomposed. The P budget was
similarly calculated except that P remaining in mulch not
decomposed was not computed because sufficient amounts of
mulch did not remain for determination of both N and P.
Nitrogen and P removed from the plots were calculated
by summing the N and P removed by the crop (i.e., grain,
cobs and stover) for the last three seasons of the study.
Since N and P data for the crop were available for only
three out of the seven crops, regression equations relating
N and P removed in crops 5-7 to yields in these seasons were
determined. These equations were then used to calculate N


173
Table 5-6. Saturated hydraulic conductivity of soil near
and away from alley cropped hedgerows of
Leucaena leucocephala and Cassia siamea, March
1992 (season seven).
Species
Hvdraulic conductivitv (mm hr'M
Mean
Near
(0.45 m)
Away
(3.5 m)
Leucaena
256.0
174.6b
215.3
Cassia
129.4
152.8
141.1
Mean
192.7
163.7*
178.2
Means in a row followed by the same letters are not
significantly different (p = 0.05).


Abstract of Dissertation Presented to the Graduate School of
the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
SOIL FERTILITY AND PRODUCTIVITY ASPECTS OF ALLEY
CROPPING LEUCAENA LEUCOCEPHALA AND CASSIA
SIAMEA UNDER SEMIARID CONDITIONS AT
MACHAROS, KENYA
By
Bashir Jama Adan
May 1993
Chairman: Dr. P.K.R. Nair
Major Department: Forest Resources and Conservation
Soil productivity aspects of alley cropping under
semiarid conditions were investigated in a three-year study
at Machakos, Kenya. The objectives were to: (1) determine
if alley cropping with leguminous trees and shrubs improves
soil fertility and productivity, (2) separate the effects of
mulch per se from those of in situ presence of the hedges,
(3) assess the extent of tree-crop interaction patterns, and
(4) determine the major factors that influence farmers'
adoption of alley cropping. The study consisted of Leucaena
leucocephala and Cassia siamea alley cropped with maize (Zea
mays); the woody species were also grown in sole-stand
blocks outside the cropped field, their prunings applied as
mulch to the cropped area in equivalent land areas.
xiv


108
of the pot. Separate resin bags were used for anions and
cations.
Three pregerminated seeds of maize were sown to each
pot and thinned to one plant/pot after germination. Four
weeks after sowing, total harvest of both shoots and roots
was made. The pots were then sown for another four
consecutive weeks. At each harvest, the roots were
separated from the soil manually by rinsing and floatation
procedures. The plant materials were then oven dried for 24
h at 65C, weighed, and ground to pass through a 0.2 mm
opening sieve, and analyzed for N concentration.
After the fourth harvest, the resin bags in each pot
were retrieved from the pots and extracted with 100 ml of 2
molar KCl. Ammonium-N and nitrate-N were determined by
standard Autoanalyzer methods at the Soil Testing Laboratory
of the University of Florida, Gainesville.
Statistical analyses
The NLIN procedure for multiphase regression models of
the SAS (SAS, 1992) was used to determine differences
between the rates and patterns of decomposition and N-
mineralization of the mulches. This procedure models the
response variable, dry matter or N remaining in the mulch,
as a continuous function of time.
The single exponential model Yt = Y0 e-kt, where Y0 is
the original amount of material applied and Yt the


219
Table 6-5. Estimates of maize yield (with or without
inputs) of farmers with interests in fodder or
mulch aspects of alley cropping.
Interests
Maize
vield ft ha'1
in alley
cropping
With
inputs
Without
inputs
Mean
Fodder
0.4 (0.2)
0.2 (0.1)
0.3 (0.2)
Mulch
0.3 (0.2)
0.1 (0.1)
0.2 (0.2)
NOTE; Figures in parentheses are the standard deviation of
the mean.


Table 2-6.
Nitrogen budgets, utilization efficiencies and yield of maize grown in alley
cropping of Leucaena leucocephala and Cassia siamea, at a hedgerow spacing of 6.7
m for 6 seasons.
N from
mulch
and/or
fertil
izer
N removed during 6
kg ha*1
seasons
N-utilization
efficiencyb
m
N-remaining
in mulch
Unac
count
ed for
N
Maize
yield
(t/ha/
sea-
Treatment
(kg)
Grain
Cob Stover
Total
Grain
Total
kg ha'1 (%)
(%>
son)
Absolute
control*
0
179
9
89
277
-
-
- -
-
2.1
Alley
cropping
(Leucaena)
648
186
9
83
278
1.0
0.2
103 16
41
2.2
Alley
cropping
(Cassia)
417
225
9
93
327
11
12
39 9
13
2.7
Mulch-only
(Leucaena)
408
192
10
87
289
3
3
87 21
8
1.8
Mulch-only
(Cassia)
177
190
9
91
290
6
7
33 19
0
1.8
Fertilizer
only*
240
222
9
97
328
18
22
-
0
2.5
aFertilizer was applied to the alley cropped plots at the rate of 40 and 17 kg ha-1 season'1 N and
P, respectively.
'Sample calculation using maize grain yield of 6 seasons from alley cropping of cassia: N
utilization efficiency (11%) N removed in grain yield (225 kg) less N removed in grain yield
of the absolute control (179 kg) divided by total N applied (417 kg) times 100.
cBased on N concentration of the mulch at the end of the litter bag decomposition studies of
mulch incorporated in October 1991 to March 1992.
d100 -(percent of applied N in mulch or fertilizer removed in crop plus N remaining in
undecomposed mulch).
*From an adjacent experiment.
U1
o


66
In situ effects of hedgerows on crop yield
The in situ benefits of growing the hedgerows was
reflected in the higher maize yields between the alley
cropping system and the block planting system and LER of
alley cropping > 1.0. The in situ effects of the hedgerows
may result not only from their root and rhizosphere effects
but also from their ability to protect the soil against
erosion and to serve as windbreaks (Wiesrum, 1984). From
semiarid India, Rao et al. (1991) provide evidence for
reduction in runoff and soil loss under alley cropping with
leucaena.
The in situ benefits of the hedgerows will be in
perspective when viewed against the yields of the
appropriate sole-crop control. This will determine if,
indeed, there has been a positive gain over the no
intercropping situation. Such a comparison can be made from
the last column of Table 2-3, which shows, for instance,
that the maize yield of the sole-maize control (fertilizer-
only) was -8% compared to that of cassia alley cropping and
+12% compared to the leucaena alley cropping. Because of
reduced maize yields under leucaena alley cropping, LER or
the yield advantage of leucaena alley cropping was
significantly lower than that under cassia. These
differences in yield reductions under alley cropping of the
two species suggest that leucaena is more competitive with
maize than cassia. The reductions in maize yield under


31
Table 2-1. Mulch yield of Leucaena leucocephala and Cassia
siamea intercropped hedgerows or planted in sole
stands outside the cropped area.
Hedgerow
Species
t ha-1
Alley
Cropping
y r1
Block
Planting*
Mean
Leucaena
4.4
3.3b
3.9s
Cassia
2.2a
1.9*
2. lb
Mean
3.3
2.6
3.0
Block planting in an area equal to that estimated to be
occupied by the hedgerows in an alley cropping system.
*-bFor a given species, means followed by the same letter are
not significantly different (p = 0.05).


216
Table 6-3. Factors which influenced significantly interests
of fanners in alley cropping for fodder or for
mulch and the frequency of farmers under each
factor, Machakos, Kenya, August 1991.
Interests
in alley
cropping
Use
of ox
plow
Use of
crop/trees
mulch at
present
or in the
past
Use of
fertil
izer
Posses
sion of
coffee
trees
on the
farm
Prior
know
ledge
about
alley
cropping
Fodder
27
22
21
37
13
Mulch
7
12
5
20
9


236
readily available planting materials, the station developed
technologies, whatever their potential, may not be adopted
by farmers.
It could be argued that the question of availability
of planting materials should be addressed only after a
viable technology has been developed on the station. The
fallacy with this prevalent attitude is that it would take
years for technologies to reach farmers and have an impact
on their lives. It also denies the researcher valuable
feedback from farmers that could enhance development of a
relevant technology, if for instance, components of the
technology were simultaneously being tried by farmers under
their different farming environments.
Of particular concern is research on some of the
agroforestry practices such as alley cropping that by nature
are long-term and also require a large number of planting
materials. To be effective for soil erosion control, or
mulch production, alley cropping requires a large number of
seedlings. For example, to plant 20% of the land under
hedges spaced, for instance, at 5 meters between rows and
0.5 m between plants in a row would require 800 plants/ha.
Assuming a highly optimistic 50% survival after six months,
one would require at least 1600 seedlings per hectare. If
all the 60 farmers interviewed were to plant 20% of their
land to hedgerows, this would require 96,000 seedlings. One
seedling would cost at least two Kenya shillings (0.1 $);


19
was used instead. Nevertheless, since this control plot was
outside the setting of the present study, it was not
included in the statistical analyses. Rather, it was used
only for relative comparisons.
Root studies.Root studies were initiated in the last
cropping season of the study with the objective of
identifying the cause of differences in the effects of the
two hedgerows species on maize yield. These studies were
conducted on hedgerows outside the main experiment
consisting of two blocks of cassia and leucaena hedgerows,
6 m wide, established in 1983. Until 1988, maize and beans
were sown in rotation in the alleys between the hedgerows in
these plots. The maize crop was fertilized with N and P at
the rate of 40 and 18 kg ha'1 season'1; the beans were not.
In addition, prunings from the hedgerows were applied to the
cropped alleys. In 1988, fertilizer application was stopped
and maize was sown in both seasons.
For each species, blocks of soil 20 x 20 x 20 cm were
removed from six transects 3 meters long from the base of
the hedgerows during the 1992 short-rains cropping season.
The soil blocks were taken from 0-20 and 20-40 cm soil
depths. The samples were brought back to washing sheds in
plastic bags. The soil blocks were then washed with a spray
of water aided by hand manipulation to separate medium-sized
and fine roots from the soil in sieves with mesh sizes of 5
mm and 0.5 mm, respectively. Roots less than 2 mm in


20
diameter were classified as fine roots, those of 2-5 mm as
medium-sized or small roots (Bohm, 1979).
The fine roots of leucaena and cassia were
distinguishable easily from maize roots based on color and
texture. Maize had whitish and succulent roots, cassia had
black roots, and leucaena had fine, cream-colored roots,
which were strong relative to cassia. The majority of the
fine roots of the two species (which were planted
separately) and the maize roots were sorted in the same way.
Some were hard to differentiate and were separated under a
magnifier. Dead roots with brittle steel and degenerated
epidermal and cortical tissues were not included.
Root length was determined by the line-intercept
method of Tenant (1975) using a tray with grid units of 1.3
cm (half-inch). With these grid units, the total number of
intercepts equals the total length of root in centimeters.
Crop and tree management.The hedgerows occupied 15%,
20% and 25% of the gross land area. The remainder of the
land, viz, 85%, 80% and 75%, was sown to maize at
approximately 37,000 plants ha-1 with a spacing of 90 cm
between rows and 30 cm within rows. Katumani Composite, a
drought-tolerant maize variety was used. The crops and
trees were only rain-fed.
Fertilizer was applied at the rate of 40, 18 and 0 kg
ha'1 season'1 of N, P and K, respectively, to maize in the
fertilizer-only control, alley cropped maize and maize next


138
Second phase
Compared to the first phase, it is less clear what
factors regulated the rate of decomposition in the second
phase. Increased C:N ratio (the C probably being in the
form of lignin) and polyphenols could be involved but the
former appears to provide the best explanation. Two
observations favor high C:N (lignin:N in particular) ratio.
The first is the increasing ratio of twigs in the mixture of
twig and leaves over time in the second phase (data not
shown). In spite of their high polyphenol content relative
to the twigs, leucaena leaves had practically all
disappeared at the start of the second phase. Instead, the
twigs with higher lignin content than the leaves were more
dominant than leaves in the second phase. The second
observation was that leucaena twigs with slightly higher
polyphenol concentration than cassia twigs had rates of twig
decomposition not significantly different from those of
cassia.
In the absence of lignin, wide C:N ratio may not be a
rate limiting factor. This is because the filter paper with
the highest C:N ratio had rates of decomposition not
significantly different from the twigs. These observations
support the argument that the form of C, lignin, must be
taken into account along with the C:N ratio in predicting
residue decomposition rates as reported also by Herman et
al., 1977; Parr and Papendick, 1978.


74
The magnitude of the mulch produced by leucaena and
cassia, 1 and 2 t ha*1 season*1, respectively, was not
sufficient enough to cause significant gains in maize yield
at the study site. There are indications that gains in
maize yield could be significant if the amounts of mulch
applied could be increased to 4 t ha*1 season and beyond.
It is unlikely, however, that such increases in mulch yield
from the hedges could be made without compromising further
crop yields.
At this point, the results of the present study do not
indicate sufficient short-term benefits to recommend alley
cropping with leucaena under conditions similar to the ones
of the present study. However, cassia does have potential
if (a) the 8% increase in maize yield outweighs the costs of
production and, (b) mortality of cassia after 3-4 years is
low enough not to necessitate replanting of the hedgerows.
Follow-p studies on factors responsible for cassia's
positive in situ effects may enhance its potentials for
alley cropping under semiarid conditions.


29
analyses, differences between means of treatments were
determined by pairwise contrasts of the Least Squares Means
(LSM). A statistically significant difference was declared
to exist between means of any pair of treatments when the p-
value was less than or equal to 0.05.
Yield response of the systems over seasons was also
analyzed with the GLM/SAS procedure. The option for
repeated measures ANOVA was included in the GLM procedure to
account for the systematic effects of harvests from the same
plots over time. The difference in mean yield of treatments
between seasons was determined by pairwise contrasts of the
LSM.
The GLM/SAS procedure was also used to analyze the
root density data. The option for repeated measures
analysis was included to account for the effects of "fixed"
distances of root observations from the hedgerows. The
significance of differences between means of root densities
at various distances from the stumps of the hedgerows was
determined by pairwise contrasts of the least squares means.
In the pot studies also, the GLM/SAS procedure was
used to analyze the data. As in the field study, the option
for repeated measures was included in the ANOVA to account
for the effects of repeated harvests from the same pots. In
addition, orthogonal polynomial contrasts for linear,
quadratic and cubic effects were included in the GLM
procedure to test significance of the relationships between


184
Table 5-7. Effects of three nitrogen rates and two
irrigation levels on the biomass yield of maize
after ywo
1991.
months growth in
pots, seventh
season,
Nitrooen
rates
(q oot'M
Irrigation
equivalent
(nun)
0
0.64
1.3
Mean
100
32.6
40.8
53.6
42.3a
200
44.5
56.3
67.1
56.0b
Mean
38.6
48.6
60.4
49.2
NOTE: Soil in the pots was collected from near (0.45 m) and
away (3.5 m) from the hedgerows. Soil source did not
significantly affect maize biomass yield. Hence, the
means presented are averaged across soil sources.
*,bMeans in a row and column followed by the same letters are
not significantly different (p = 0.05).


238
have not had ready access to chemicals as have researchers
on the station. Besides, some of the chemicals in the
market (e.g., DDT) are environmentally hazardous.
Typically, farmers use wood ash as a traditional control
measure for termites. However, the success of this practice
was highly variable from farm to farm.
Research into termite control is, indeed, the key to
success of any tree planting activity in this and other
similar areas. At the time of concluding the present study,
termite control studies were missing from the many trials at
the research station. In particular, it would be valuable
to explore the use of nonchemical methods such as wood ash,
their potentials and constraints.
Conclusions
Most farmers were interested in alley cropping. The
interest was perhaps more in the trees than in the alleys
created by the trees. There were two distinct interest
groups: primarily in fodder (62%) and, in mulch (33%).
Interests of farmers in alley cropping depended
significantly on five factors:
1. Presence or absence of coffee,
2. Use of mulch on the crop fields,
3. Use or possession of an ox-plow,
4. Use of fertilizer, and
5. Prior knowledge about the technology.


67
leucaena alley cropping could be more severe than observed
in the present study in sites that are both nutrient-poor
and dry (Chirwa, 1991).
Leucaena became more competitive than cassia starting
in the second year. Similar observations on the onset of
leucaena's competition were made also by Rao et al. (1990)
working with alley cropping in semiarid India. The
significantly higher mulch production as well as the
presence of high fine-root densities of leucaena in the top
soil compared to that of cassia provides additional clues to
leucaena's higher competitiveness. It is possible that a
higher pruning frequency than the 4-6 times a year of
leucaena in the present study may have reduced its
competition with the intercrcropped maize.
The higher maize yield under cassia alley cropping
compared with that under leucaena or under the sole crop
could be ascribed to significantly more positive crop-
hedgerow interactions observed under cassia. These positive
interactions could have mitigated high reductions in maize
yields from hedgerows with narrow alleys compared to those
with wide alleys. This could explain the lack of
significant differences in the three alley widths (spacings)
used in the present study. Because more land was under
hedges in the narrow alleys than in the wide alleys, maize
yield from the narrow alleys was expected to be less than
from those of the wide alleys.


193
permeability (Swift and Sanchez, 1984). Moreover, the
activity of the significantly higher population of fauna
(predominantly termites, beetles and wood lice) near the
hedgerows could have contributed to the lower soil bulk
density observed near the hedgerows compared to away from
them. As observed by Coleman et al. (1991), the role of the
fauna in reducing soil bulk density and enhancing
infiltration rates of water appeared to be important in the
present study. For example, from a subsidiary study, it was
observed that the rates of hydraulic conductivity of the
soil declined significantly in spots near the hedgerows
where a pesticide was applied repeatedly compared to where
the pesticide was not applied (data not presented). These
observations indicated the significant role of the fauna on
the physical properties of the soil.
Differences between Effects of Hedgerow
Species on Crop Yield Profile
The significantly lower yield of maize under leucaena
alleys compared to those under cassia demonstrates that
leucaena is more competitive with crops than cassia, an
observation also made by Nair (1987). Greater depletion of
soil water by leucaena compared to cassia best explained the
lower maize yield under leucaena compared to cassia. A
similar explanation was also made by Singh et al. (1989) who
observed that alley cropping leucaena in semiarid India
reduced crop yields significantly.


94
Mwangi (1989), working at a site adjacent to the present
study, also observed that at rates of mulch application 3-4
times higher than the amounts applied in the present study, P
levels of the top soil nearly doubled after 6 years of
cropping. However, there are biological limitations to the
production of such high mulch yields in SATs. In the absence
of high mulch yields, application of the small amounts of
mulch available may still result in short-term or immediate
benefits (e.g., enhanced availability of top soil N and P).
It may, however, have little effect on the maintenance of soil
organic matter (Stevenson, 1986).
The initial levels of soil organic C and total N were
in the low end of the range typical of Alfisols (Brady,
1990). However, it is worth noting that no significant
changes in either soil organic matter or N were observed
after three years of cropping with an annual removal of
about 60 kg N ha*1 through grain and maize stover. The lack
of significant change in soil organic C and total N was even
observell from the control plots with no additions of N or
mulch. Maize yield from plots with various treatments were
also similar, with differences being more pronounced between
seasons than between treatments. These supporting
observations suggests that it will take more than three
years to detect changes in organic C or that the effects of
the small amounts of mulch added to the soil may have been
sufficient to prevent declines in the initial soil organic
C.


151
with the closest one to the hedgerow being only 45 cm away
and the farthest, 3.5 m away. Moderate levels of fertilizer
(40 and 18 N and P kg ha'1' was applied to the maize every
season.
Measurements
Crop
In the last three seasons (i.e., seasons 5, 6, and 7)
out of a total six of the study, the crop was harvested row-
by-row from the alleys of both species and separated into
grain, stalks and stover. Before final harvest, biomass
yield of each crop was determined twice, at 30 and at 60
days after sowing, near and away from the hedgerows.
At harvest, the N and P content of the crop grain,
stalks, and stover of the row nearest the hedge (referred to
as row one) and the row farthest from the hedge (row four)
were also determined. At first sampling (30 days after
sowing), N and P content of samples from the whole shoot
after milling and passing through a 2 mm sieve was
determined. At the second harvest/sampling, N and P content
of the ear leaf was determined. Standard micro-Kjeldahl
method, modified for plants, was used for N analysis; P was
determined with UV visible Spectrophotometer at 450 nm.
One to two days before the first and second harvest
for dry matter assessment, relative water content of the


37
Figure 2-3. Effect of three hedgerow spacings on
maize grain yield, averaged over yields under the two
species, Leucaena leucocephala and Cassia siamea. Vertical
bars indicate standard error of difference of means.


118
Table 4-3. Decomposition parameters of soil incorporated
and surface placed Leucaena leucocephala and
Cassia
second
siamea
season
twigs outside
of 1991.
the litter bags,
Placement
method/species
Inter
cept
Decomposition
rate (%)
Phase Phase
one one
Spline
point
(weeks)
r2
Soil incoroorated
Cassia
2.2
10.1
6.0
9.3
0.73
Leucaena
2.4
13.1
2.5
9.3
0.89
Soil surface
Cassia
2.2
7.8
1.3
11.4
0.65
Leucaena
2.2
4.8
2.4
8.3
0.78


192
Effects of shade from the hedgerows on soil
temperature may be significant where the orientation of the
hedgerows is north-south, where the pruning frequency of the
hedgerows is low and at sites far from the equator as
observed by Monteith et al. (1991). In the present study,
the hedgerows were east-west oriented, they were frequently
pruned (2 to 3 times in a season) and the site was close to
the equator (Io S 33' E). It is therefore unlikely that the
effects of shade from the hedgerows on soil water content,
particularly during the cropping season, were significant.
During the dry season, however, it is possible that even a
small amount of soil water conserved due to shade effects
from the hedgerows, could have contributed to the enhanced
yield of the crop near the hedgerows in the following
season.
The significantly higher rates of infiltration and
hydraulic conductivity observed near the hedgerows could
have resulted from the lowered soil bulk density observed
near, compared to away, from the hedgerows. Topsoil (0-5 cm
profile) bulk density was significantly reduced near
leucaena hedgerows. Reduction in soil bulk density could
have resulted from the death and turnover of the higher root
density near the hedgerows (Chapter 2). Sloughed-off roots
add to soil organic matter (humus), which in turn, could
reduce soil bulk density. Reduced soil bulk density
improves aeration, stabilizes structure, and increases


48
Table 2-5. Land equivalent ratios (LER) of alley cropping
Leucaena leucocephala and Cassia siamea compared with
the block planting system with same ratios of tree and
crop land occupancy.
Year and Seasons Within a Year
Species
1989
1990
1991
Mean
1
2
1
2
1
2
Leucaena
1.0*
1.09
1.15
0.87
1.23
1.35
1.12
(1.10)
(0.97)
(1.15)
(1.02)
(1.29)
(1.26)
(1.13)
Cassia
1.04
1.33
1.21
1.17
1.56
1.87
1.36
(0.99)
(1.24)
(1.00)
(0.96)
(1.10)
(1.15)
(1.07)
LER = (Biomass yield of the tree species in alley cropping plus
yield of crop in alley cropping) divided by (biomass yield of the
tree species in sole system plus yield of crop in sole system) .
In both the alley cropping and sole planting systems, the percent
of land occupied by the hedges and the crop was 15 and 85,
respectively. In the sole planting system, the biomass yield of
the hedgerows in the 25:75 tree:crop percent land occupancy ratio
was used because this was the most appropriate (i.e., highest
yielding) sole planting system.
*LER of maize grain yield only. Number in bracket is the total
LER, i.e., the grain and the stalk yield of the maize and the tree
biomass which was used as mulch.


73
hedgerows could be increased to application rates of 3-4 t
ha'1 season'1, gains in crop yields could be increased even
from the relatively fertile site of the present study.
Conclusions
Alley cropping of Leucaena leucocephala was
detrimental to maize yield over the six cropping seasons.
On average, maize yield under leucaena alley cropping
dropped by 12% of the sole crop while that of cassia
increased by 8%. Leucaena, on the other hand, produced
significantly higher mulch biomass (50% more) than cassia.
The higher mulch biomass production and higher density of
fine roots in the top 40 cm soil are factors that contribute
to leucaena's significantly higher competitiveness with
maize than cassia's.
The in situ benefits of growing the hedges on maize
yield were positive for both species but significantly more
under cassia compared to those under leucaena. Planting the
hedges outside the crops may be a management approach to
minimize competition between the crop and the tree
hedgerows, especially for a species like leucaena with
desirable fodder value. However, this approach produced
yield reductions higher than alley cropping with both
leucaena and cassia, with the loss nearly equal to the
percentage of land apportioned to the hedges.


190
therefore, increase yield of crops adjacent to the hedgerows
while that of the distant crop is reduced. In addition, the
protective effects of hedgerows could reduce soil loss and
runoff (Rao et al., 1991) and could enhance soil water
content near the hedgerows. It is feasible that the
hedgerows had significant barrier effects since the plots of
the present study were on land with a 3%-5% slope.
Reduced runoff has been suggested by others to explain
an increased yield of crops observed near hedgerows on
sloping lands (Sang and Hoekstra, 1987). Reductions in
runoff under hedgerows, however, could be of minor
importance compared to the competition for soil water
between the roots of trees that are established by the time
the crops are sown (Singh, 1989). A process referred to as
hydraulic lift (Caldwell and Richards, 1989), which involves
absorption of water by deep roots in moist soil, upward
movement through the roots, and release in the upper soil
profile at night, could also be a mechanism by which water
content near the hedgerows was enhanced.
Reduced soil temperature as a result of shade from the
hedgerows could also have contributed to the trends of
higher soil water content near the hedgerows than away from
them. Shade would create differential rates of evaporation,
with higher rates in the open and lower rates near the
hedgerows. This would imply that less water was available
to crops in the open alleys and more to those near the


Maize grain yield (t ha"}
Percent land under tree and crop
38
Figure 2-4. Effect of sole stands of Leucaena
leucocephala and Cassia siamea planted at three land
occupancy on maize yield during six seasons. Vertical bars
indicate standard error of difference of means.


SOIL FERTILITY AND PRODUCTIVITY ASPECTS OF ALLEY
CROPPING LBUCAENA LEUCOCEPHALA AND CASSIA
SIAMEA UNDER SEMIARID CONDITIONS
AT MACHAROS, KENYA
BY
BASHIR JAMA ADAN
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE
UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE
REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
1993

To my late father, Jama Adan Derow, who passed away in
November 1992, while I was in Gainesville writing this
dissertation.

ACKNOWLEDGMENTS
The author would like to express his sincere thanks to
Dr. P.K'.R. Nair for his constant advice, patience,
sincerity, and moral support during the course of this
project. This research certainly would have been more
difficult if it were not for the support and encouragement
that Dr. Nair extended on many difficult occasions. The
author also would like to thank the other members of the
supervisory committee, Drs. H.L. Gholz, C.P.P. Reid, P.E.
Hildebrand, and H.L. Popenoe. Special thanks are due to Dr.
M.R. Rao of the International Centre for Research in
Agroforestry (ICRAF), Nairobi, for his field supervision,
support and suggestions regarding the dissertation. In
addition, thanks are due to Peter Kurira Wambugu, manager of
ICRAF Research Station, Machakos, Kenya, for supervising
field activities and data collection, especially when the
author was away in 1989-1990.
Appreciation is extended to ICRAF, the Ford Foundation
and the Rockefeller Foundation, who have provided financial
support for this research; and the National Dryland Research
Station (Katumani), Kenya, which has provided laboratory
facilities; and to ICRAF once again, for granting leave-of-
absence to the author during the study period.
iii

Special thanks are given to Richard Coe of ICRAF and
Steven Linda, Department of Statistics, University of
Florida, for their suggestions and assistance in the
analysis of the experimental data. Thanks are also due to
forestry/agroforestry students Paramu Mafongoya, Todd
Johnson, Reinhold Muschler, Paul Campell, Loice Omoro,
Mutiah Govindarajan, Don Mee, Mark Follis, Chris Latt,
Antonio Brunori, and Anne Todd-Bockerie and to Christine
Kelly of agronomy department.
The secretarial services provided by Marie Kimenye of
ICRAF during the period of field research are highly
appreciated. Furthermore, appreciation is extended to Pam
White for her assistance in the preparation of this
manuscript.
Finally, the author wishes to express his deepest
appreciation for the support, understanding and sacrifice
made by his wife Fatuma (Nimoo), daughters (Aisha and
Khadija), father (who passed away in the while the author
was writing-up the dissertation"rahimuallah"), mother,
brothers, and sisters.
iv

TABLE OF CONTENTS
Page
ACKNOWLEDGMENTS
LIST OF TABLES. . vii
LIST OF FIGURES x
ABSTRACT
CHAPTERS
1 INTRODUCTION 1
2 PRODUCTIVITY ASPECTS OF ALLEY CROPPING
WITH LEUCAENA LEUCOCEPHALA AND CASSIA SIAMEA
IN SEMIARID TROPICS AT MACHAROS, KENYA . 11
Introduction 11
Materials and Methods 15
Results 30
Discussion 57
Conclusions 73
3 SOIL FERTILITY ASPECTS OF ALLEY CROPPING
WITH LEUCAENA LEUCOCEPHALA AND CASSIA
SIAMEA IN SEMIARID TROPICS AT MACHAROS,
KENYA 75
Introduction 75
Materials and Methods 78
Results 83
Discussion 86
Conclusions 95
4 DYNAMICS OF SOIL FERTILITY CHANGES: RATES OF
DECOMPOSITION, N-MINERALIZATION AND UPTAKE
BY MAIZE FROM MULCH OF LEUCAENA LEUCOCEPHALA
AND CASSIA SIAMEA UNDER SEMIARID CONDITIONS
AT MACHAROS, KENYA 97
Introduction 97
Materials and Methods 100
Results ....... 110
Discussion 132
Conclusions 145
v

Page
5 INTERACTIONS BETWEEN MAIZE AND HEDGEROWS
OF LEUCAENA LEUCOCEPHALA AND CASSIA SI AME A
IN ALLEY CROPPING SYSTEM IN SEMIARID TROPICS
AT MACHAROS, KENYA 147
Introduction 147
Materials and Methods 149
Results .......... 158
Discussion 183
Conclusions 195
6 POTENTIAL FOR FARMER ADOPTION OF ALLEY
CROPPING OF MULTIPURPOSE TREES AND SHRUBS
IN SEMIARID AREAS OF MACHAROS, KENYA ... 197
Introduction 197
Study area and methodology ...... 201
Results ............... 204
Discussion 226
Conclusions ............. 238
7 SUMMARY AND CONCLUSIONS 241
LITERATURE CITED 247
BIOGRAPHICAL SKETCH ................. 267
vi

LIST OF TABLES
Page
2-1. Mulch yield of Leucaena leucocephala and
Cassia siamea intercropped hedgerows or
planted in sole stands outside the
cropped area 31
2-2. Concentration and amounts of nutrients in
Leucaena leucocephala and Cassia siamea
mulch from alley cropped hedgerows ....... 35
2-3. Effect of alley cropping and block planting
systems of Leucaena leucocephala and
Cassia siamea on maize grain yield
during six cropping seasons .... 39
2-4. Effect of alley cropping and block planting
systems on relative grain yield of maize .... 41
2-5. Land equivalent ratios (LER) of alley cropping
Leucaena leucocephala and Cassia siamea
compared with the block planting system
with same ratios of tree and crop land
occupancy .............. 48
2-6. Nitrogen budgets, utilization efficiencies
and yield of maize grown in alley cropping
systems of Leucaena leucocephala and Cassia
siamea at a hedgerow spacing of 6.7 m ..... 50
2-7. Addition of phosphorus through mulch and
removal by maize in six seasons of cropping
under Leucaena leucocephala and Cassia
siamea alley cropping 52
2-8. Fine root density at two soil depths along a
meter-long transect from 6 m apart hedgerows of
Leucaena leucocephala and Cassia siamea
intercropped with maize ...... 55
3-1. Chemical and physical characteristics of
the top soil {0-20 cm) at the beginning of
the study, October 1987 79
Vii

Page
3-2. Chemical characteristics of the dry mulch
of alley cropped hedgerows of Leucaena
leucocephala and Cassia siamea 81
3-3. Chemical and physical characteristics of
the top soil (0-20 cm) under alley cropping
systems with hedgerows of Leucaena leucocephala
and Cassia siamea in November, 1991 84
3-4. Changes in soil Mehlich-1 P levels after
three years of alley cropping hedgerows of
Leucaena leucocephala and Cassia siamea with
maize, November 1991 .............. 85
3-5. Concentration of phosphorus in the mulch
(leaves plus twigs) of alley cropped
hegderows of Leucaena leucocephala and
Cassia siamea at the start (1987) and
end (1991) of the study ........ 87
3-6. Relative indices of KCl-extract N from
ion-exchange resins incorporated into the
soil of alley cropped hedgerows of Leucaena
leucocephala and Cassia siamea with maize
during the "short-rains" cropping season
of 1991 88
4-1. Chemical and physical characteristics of the
mulch of alley cropped hedgerows Leucaena
leucocephala and Cassia siamea 106
4-2. Intercepts, slopes, spline points and
coefficients of determination (r2) of the
regression lines for phase one and phase
two of leucaena and cassia mulch
decomposition ................. Ill
4-3. Decomposition parameters of soil incorporated
and surface placed Leucaena leucocephala and
Cassia siamea twigs placed outside the
litter bags, second season of 1991 ...... 118
4-4. Effects of soil incorporated Leucaena
leucocephala and Cassia siamea mulch on
the biomass yield, N-concentration, N-yield
and apparent N recovery of maize planted in
pots, 1991 128
viii

Page
4-5. Relative indices of cumulative NH4+ N, N03 N
and total N mobilized 16 weeks after soil
incorporation of 2 t ha'1 equivalents of
Leucaena leucocephala and Cassia si amea
mulch in pots with maize, 1991 ........ 131
5-1. Nitrogen content of 30 day-old maize shoots
(leaves plus stalks) sampled near (0.45 m) and
away (0.45 mm) from hedgerows of Leucocephala
leucocephala and Cassia siamea, fifth season,
1991 164
5-2. Relative water content of flag leaves of
maize plants sampled near (0.45 m) and
away (3.5 m) from alley cropped hedgerows
of Leucaena leucocephala and Cassia siamea,
on April 26, 1991 (sixth season) 165
5-3. Water content of whole maize plants sampled
near (0.45 m) and away (3.5 m) from alley
cropped hedgerows of Leucocephala leucocephala
and Cassia siamea on April 2*6, sixth season,
1991 166
5-4. Relative water content of flag leaves of maize
plants sampled near (0.45 m) and away (3.5 m)
from alley cropped hedgerows of Leucaena
leucocephala and Cassia siamea leucaena and
cassia on May 28, 1991 (sixth season) 168
5-5. Relative water content of flag leaves of maize
plants sampled near (0.45 m) and away (3.5 m)
from alley cropped hedgerows of Leucaena
leucocephala and Cassia siamea on January 6,
1992 (seventh season) 169
5-6. Saturated hydraulic conductivity of soil near
(0.45 m) and away (3.5 m) from alley cropped
hedgerows of Leucaena leucocephala and Cassia
siamea, March 1992 (season seven) 173
5-7. Effects of three nitrogen rates and two
irrigation levels on the biomass yield of
maize after two months growth in pots, 1991 . 184
6-1. Farm and household characteristics of the 60
farmers interviewed at Machakos, Kenya,
August 1991 ...... 205
ix

Page
6-2. Interests of fanners in alley cropping, farm
size and number of members in a household
at Machakos, Kenya, August 1991 207
6-3. Factors which influenced significantly
interests of farmers in alley cropping for
fodder or for mulch and the frequency of
farmers under each factor, Machakos, Kenya,
August 1991 216
6-4. The p-values of dependency of resource
possession, their use or purchase on the presence
of coffee on the farm, Machakos, Kenya,
August 1991 217
6-5. Estimates of maize yield (with or without
inputs) of farmers with interests in fodder
or mulch aspects of alley cropping, Machakos,
Kenya, August 1991 219
6-6. Functions of tree nurseries possessed by
30% of the 60 farmers, 1991 interviewed,
Machakos, Kenya, August 1991 . 224
6-7. Reasons given by farmers without nurseries
on their farms, Machakos, Kenya, August 1991 225
x

LIST OF FIGURES
Page
2-1. Seasonal rainfall and the local names
(long-rains and short-rains) of the
cropping seasons during the study 16
2-2. Mulch yield of Leucaena leucocephala and
Cassia siamea in alley cropping and
block/sole planting systems 33
2-3. Effect of three hedgerow spacings on maize
grain yield, averaged over the yields
under the two species, Leucaena leucocephala
and Cassia siamea 37
2-4. Effect of sole stands of Leucaena leucocephala
and Cassia siamea planted at three land
occupancy on maize yield during six seasons . 38
2-5. Relationship between rainfall and maize mean
yield of all treatments in each of the six
cropping season ........ ... 42
2-6. Maize grain yield under alley cropping and
block planting systems of Leucaena leucocephala
and Cassia siamea expressed as percent of the
sole maize control yield 44
2-7. Maize productivity (Kg m'2) under alley cropping
Leucaena leucocephala and Cassia siamea hedgerows
for six seasons 45
2-8. Productivity of maize (kg mm-1 of rainfall)
under alley cropping of Leucaena leucocephala
and Cassia siamea hedgerows for six cropping
seasons 46
2-9. Fine root density in the top 40 cm soil profile
of alley cropped hedgerows of Leucaena
leucocephala and Cassia siamea 54
2-10. Effect of three mulch rates of Leucaena
leucoc-phala and Cassia siamea on biomass
yield of maize grown in pots, 1991 56
xi

Page
4-1. Monthly rainfall and mean temperature during
the study, 1991 101
4-2. Decomposition patterns of the leaves of
Leucaena leucocephala and Cassia siamea. ... 113
4-3. Decomposition patterns of the twigs of Leucaena
leucocephala and Cassia siamea 117
4-4. Decomposition patterns of the mulch (leaves
plus twigs) of Leucaena leucocephala and
Cassia siamea 120
4-5. Cumulative percent of N mineralized from soil
incorporated mulch of Leucaena leucocephala and
Cassia siamea after 16 weeks ......... 123
4-6. Relationship between the patterns of N release
from the mulch of Leucaena leucocephala and
Cassia siamea and biomass yield of maise
from four consecutive crops of maize sown
in pots and harvested at an interval of
four weeks 126
5-1. Maize yield response to distance from alley
cropped hedgerows of Leucaena leucocephala
and Cassia siamea, 1990 short-rains 159
5-2. Maize yield response to distance from alley
cropped hedgerows of Leucaena leucocephala
and Cassia siamea, 1991 long-rains ...... 160
5-3. Maize yield response to distance from alley
cropped hedgerows of Leucaena leucocephala
and Cassia siamea, 1991 short-rains 161
5-4. Bulk density of top 5 cm soil profile
near (0.45 m) alley cropped hedgerows of
Leucaena leucocephala and Cassia siamea,
1991 short-rains 170
5-5. Infiltration rates of water into the soil
near (0.45 m) and away (3.5 m) from alley
cropped hedgerows of Leucaena leucocephala
and Cassia siamea, 1991 short-rains 172
5-6. Soil water content near (0.45 m) alley cropped
hedgerows of Leucaena leucocephala and Cassia
siamea determined at the end of 1991
long-rains ....... 174
xii

Page
5-7. Soil water content near (0.45 m) alley cropped
hedgerows of Leucaena leucocephala and Cassia
siamea determined at the end of 1991
short-rains 175
5-8. Difference in gravimetric water content
(%, w/w) between soil sampled at 0-20 cm
depth near (0.45 m) and away (3.5 m) from
alley cropped hedgerows of alley cropped
hedgerows of Leucaena leucocephala and
Cassia siamea in 1991 177
5-9. Monthly average readings of tensiometers placed
in the soil at 15 cm depths near (0.45 m) and
away (3.5 m) from alley cropped hedgerows
of Leucaena leucocephala and Cassia
siamea in the long- and short-rain
seasons of 1991 . 178
5-10. Difference in mean monthly temperature of
top soil (10 cm) measured near (0.45 m) and
away (3.5 m) from alley cropped hedgerows of
Leucaena leucocephala and in the long- and
short-rain seasons of 1991 179
5-11. Count of fauna in 25 x 25 x 25 cm top soil
monoliths taken from near (0.45 m) and away
(3.5 m) from alley cropped hedgerows of Leucaena
leucocephala and Cassia siamea at the end of
1991 short-rain season 181
5-12. Effects of irrigation and fertilization, in
a combination or alone, on maize grain yield
in the field (no alley cropping with hedgerows
of Leucaena leucocephala and Cassia siamea),
1991 182
xiii

Abstract of Dissertation Presented to the Graduate School of
the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
SOIL FERTILITY AND PRODUCTIVITY ASPECTS OF ALLEY
CROPPING LEUCAENA LEUCOCEPHALA AND CASSIA
SIAMEA UNDER SEMIARID CONDITIONS AT
MACHAROS, KENYA
By
Bashir Jama Adan
May 1993
Chairman: Dr. P.K.R. Nair
Major Department: Forest Resources and Conservation
Soil productivity aspects of alley cropping under
semiarid conditions were investigated in a three-year study
at Machakos, Kenya. The objectives were to: (1) determine
if alley cropping with leguminous trees and shrubs improves
soil fertility and productivity, (2) separate the effects of
mulch per se from those of in situ presence of the hedges,
(3) assess the extent of tree-crop interaction patterns, and
(4) determine the major factors that influence farmers'
adoption of alley cropping. The study consisted of Leucaena
leucocephala and Cassia siamea alley cropped with maize (Zea
mays); the woody species were also grown in sole-stand
blocks outside the cropped field, their prunings applied as
mulch to the cropped area in equivalent land areas.
xiv

Difference in maize yield between alley-cropped plots
and the plots receiving mulch from the block plantings
showed that growing the hedges in situ had a positive effect
on maize yield. Yield advantage of alley cropping over the
separate crop and hedge planting systems, expressed as land
equivalent ratios, were 36% and 12% for cassia and leucaena,
respectively. Leucaena produced more biomass (4 t ha"1 yr"1)
than cassia (2 ha'1 yr"1) ; also leucaena had higher fine-root
density than cassia in the top soil (0-40 cm). Therefore,
leucaena was more competitive than cassia, and over the
years, maize yield declined by 12% under leucaena whereas it
increased by 8% under cassia.
For both species, maize nearer the hedgerows yielded
more than those away from them. Soil water content was also
higher near the hedgerows. Chemical and physical properties
of the top soil (0-20 cm) did not differ significantly
between treatments (with or without mulch or alley cropping)
or over time for a given treatment. The quantities of mulch
produced by the hedgerows were generally low, and they
decomposed quickly (half life of 'four weeks); these could be
the main reasons for the lack of significant changes in soil
properties.
Fanners were generally more interested in the fodder
than the mulch aspects of alley cropping. Because of the
initial high investments of inputs such as labor and
seedlings that required to establish the hedgerows, alley
cropping is more likely to be adopted by relatively
xv

resourceful farmers such as coffee growers than by poorer
farmers.
xvi

CHAPTER 1
INTRODUCTION
In many parts of the tropics, semiarid lands that have
traditionally not been under arable farming are increasingly
being brought under rainfed agriculture. For example, in
Kenya where arid and semiarid lands comprise four-fifths of
the country's land area, arable farming has been extended to
vast areas of dry or marginal lands during recent years.
However, there are some formidable obstacles to food
production in many of these areas. In addition to rainfall
constraints, the soil's productive capacity is inherently
low in many areas. In areas where low level subsistence
farming has been followed, fertilizer use and other means of
nutrient input to soil are seldom practiced. Consequently,
low and declining yields of crops are common. For instance,
maize yields less than 1 t ha*1, even in seasons of normal
rainfall are typical of farms in semiarid areas of Machakos,
Kenya (Nadar and Faught, 1984; Mbogo, 1991).
Under traditional shifting cultivation, which has been
the predominant agricultural system in many areas of the
tropics (both humid and semiarid), long rest or fallow
periods in between short cropping periods have traditionally
been used to improve fertility of soils exhausted by
cultivation (Nye and Greenland, 1960). However, land
1

2
fallowing results in satisfactory regeneration of soil
fertility only when the length of the fallow is adequately
long. The length of the fallow-period or rest-period
requirement of a soil depends on a number of factors, such
as soil type, climatic conditions and land-use practices
(Young, 1989). For example, on Alfisols in semiarid zones,
it is estimated that fallow periods of one to two years out
of every three years are required even with intermediate
levels of inputs. With the increasing population, fallow
periods have either been reduced drastically or totally
abandoned in many places. Thus, keeping the land in fallow
as a means of sustaining soil fertility is no longer a
feasible management option.
Use of fertilizers is an option considered by many to
increase soil fertility and productivity. Chemical
fertilizers are, however, seldom used by many resource-poor,
tropical, subsistence farmers under conditions as in
Machakos (Ockwell et al., 1991). Fertilizers are expensive
and/or not readily available (Mbogo, 1991). Application of
farmyard manure is another widely adopted practice which, in
fact, is suggested to be the principal source of nutrients
for crops on most farms in Machakos (Probert et al., 1990).
However, the quantity of manure available or applied is
often too low to sustain continuous crop production at
meaningful yield levels (Ikombo, 1984; Ockwell et al.,
1991) .

Thus, expansion of arable lands, extended fallows
and/or increased fertilizer use are not viable solutions to
the problem of declining soil productivity and poor crop
yields in most semiarid parts. There is a need, therefore,
for alternative low-input technologies that can intensify
and/or sustain production from existing croplands. One
technology that has been tried with some success in the
humid lowland tropics is alley cropping (also called
hedgerow intercropping). It consists of growing food crops
in alleys between planted hedgerows of multipurpose trees
and shrubs (MPTS), usually nitrogen-fixing ones. The
hedgerows are periodically cut back at the beginning of, as
well as during, cropping. This prevents shading of the
associated crop and the hedge prunings can be a source of
nutrients to the crop (Kang et al., 1990). By retaining
woody perennials in crop production fields on a continuing
basis, alley cropping simulates the role of fallow in soil
fertility regeneration in shifting cultivation. It also
involves, to some extent, the principle of intercropping of
legumes and cereals (e.g., maize and beans), which is a
traditional farming practice in Kenya as well as many other
African countries.
The most distinctive element of agroforestry
practices, and in particular, alley cropping, is the MPTS.
The notion of trees as multipurpose plants is based on their
productive attributes (for example, food, fodder, fuelwood

4
and poles) and their protective or service functions in
relation, for example, to soil conservation and soil
fertility improvement. Leucaena leucocephala Lam. De Wit
(Leguminosae, Mimosoideae) and Cassia siamea Lam.
(Leguminosae, Caesalpinoideae) are two MPTS widely used in
alley cropping studies in both the humid lowlands and the
semiarid tropics (Nair, 1987; Kang et al., 1989; Singh et
al., 1989). This is probably so because both species are
legumes and it is assumed that soil fertility improvement is
among their potential benefits. (In subsequent references
to the two species, the generic neones leucaena and cassia
will be used as common names.)
Leucaena had long been used for shade and soil
improvement in tree and coffee plantations in Java,
Indonesia, as early as 1900 (Dijkman, 1950) Properties
that make leucaena suitable for soil fertility improvement
are summarized in Young (1989). They include: high biomass
production (10-25 t ha'1 yr'1, DM), high nitrogen fixation
(100-500 kg N ha'1 yr'1), high levels of nitrogen in leaves
(2.5%-4.0%), and, thus, a high rate of nitrogen return in
litter and prunings. The main soil limitation is a
reduction in its growth on acid soils (Brewbaker, 1987).
Also, attack by the psyllid, Heteropsylla cubana, is
currently serious in some regions. Like leucaena, cassia
has the capacity to grow on poor soils and has been used in
alley cropping trials (Yamoah et al., 1986a), although the

5
extent of its soil-improving potential is little known
compared to that of leucaena. There is even doubt as to
whether it is nitrogen fixing, and established opinion is
that it is not.
In the humid lowlands of Nigeria, many trials have
shown that crop yields could be improved under alley
cropping with MPTS. For instance, an eight-year alley
cropping trial conducted by Kang et al. (1990) showed that
using Leucaena leucocephala prunings only (in the range of
8-10 t ha'1 yr'1), maize yield could be maintained at a
reasonable level of 2 t ha'1, compared to 0.66 t ha'1 without
leucaena. Additions of leucaena prunings also maintained
higher soil organic matter than when no prunings were added.
Several other studies in the humid lowlands have also given
favorable results from alley cropping (e.g., Yamoah et al.,
1986a). As a result, there has been considerable interest
in alley cropping, and the technology has been tried in a
number of other places including the semiarid tropics
(SATs). However, the results of alley cropping have not
been positive in all places.
Fundamental to the concept of alley cropping is the
proposition that trees improve soils. There is evidence to
support that soil fertility and, to some extent, crop yields
in semiarid areas could be improved by the presence of trees
on farmlands e.g., Faidherbia albida (synopym: Acacia
albida) in West Africa (Bon Kongou, 1992) and Prosopis

6
cineraria in India (Mann and Saxena, 1980). The various
mechanisms, demonstrated or hypothesized but not yet fully
proven, are reviewed by Sanchez (1987), Young (1989), and
Nair (1990). Some of the processes demonstrated or
hypothesized by which trees improve soil fertility include
additions of litter, N2-fixation, nutrient uptake from
subsoil and protection from erosion. While trees may
improve fertility of the soil beneath them and the yield of
crops near them, they will also compete with crops for
nutrients and moisture if their root distribution is similar
to and in the same soil zone as that of annual crops (Huxley
et al., 1977; Ewel et al., 1982; Jonsson et al., 1988). For
example, most of the fine, feeder roots of many common trees
are found within the 20 cm top soil (Comerford et al.,
1984). Besides, roots of the tree may be well established
by the time the annual crop is sown. It is this potential
for effective competition, particularly for soil resources,
that has been suggested as the reason for decreased crop
yields under alley cropping in some humid lowlands of the
tropics (Evenssen, 1989; Szott et al., 1990) and semiarid
tropics (Singh et al., 1986; Nair, 1987; Mittal and Singh,
1989).
To minimize the potential for crop yield losses under
alley cropping, there should be benefits of soil fertility
and crop performance that can only be realized if the hedges
and crops are planted together, i.e., benefits that would be

7
lost if they were grown separately. These include mostly
gradual changes in soil organic matter, nutrient content and
physical properties such as bulk density, water infiltration
and porosity. The presence of the hedges in situ may,
through their root turnover and reduction of soil erosion,
sustain soil organic matter (Young, 1989). This may be of
particular importance in low-input agricultural systems
where all above-ground biomass is removed from the field.
If the contribution of the in situ presence of the
hedgerows to soil fertility relative to that of the mulch is
negligible, then planting the hedges separately from crops
and then transferring their mulch to crop fields would be a
more logical approach than growing the two together. Even
then, an understanding of the relative roles of the quantity
and quality (i.e., decomposition rates) of the mulch
produced would be relevant as this may enhance the
efficiency of its use.
In the humid tropics, mulch yields from the hedgerows
are high (in the range of 8-10 t ha"1 yr"1) Also, the rates
of decomposition are rapid. This makes nutrients in the
mulch rapidly available, for instance, in about 30 days in
the case of leucaena (Budelman, 1988; Gichuru and Kang,
1989). So, the mulch, in essence, resembles inorganic
fertilizer that provides a readily available supply of
nutrients to the crop. Crop response when mulch is removed
is therefore observable within a growing season. In the

8
SATs, unlike in the humid tropics, mulch yields from alley
cropped species are lower than in the humid tropics. For
instance, mulch yields of 2 t ha1 yr*1 from leucaena is
typical in SATs (Nair, 1987) as opposed to 8-10 t ha1 yr1
in lowland humid tropics. As Huxley (1986) postulated, this
low mulch yield in SATs may limit the potential of alley
cropping to improve fertility and productivity of soils in
these regions.
Information on the yields as well as the rates of
decomposition of mulch from alley cropped MPTS (which in
turn affects the ability of the mulch to provide nutrients
to the crops and to contribute to soil organic matter) are
relatively scarce in the SATs. Also, the benefits of the in
situ presence of hedgerows to soil fertility and
productivity are not well understood in either the humid or
the semiarid tropics.
In addition to these technical questions, there is
also the issue of social acceptability of the new technology
by the target farmers. It is necessary to determine how
farmers will adapt and use the technology under conditions
unique to their systems of farming. For instance, in the
semiarid areas where prunings from hedgerows as well as crop
residues are used as fodder (Singh et al., 1986), the in
situ presence of hedgerows may be the main avenue for
improving soil fertility. Studies of this nature will
provide insights into determining the technical feasibility

9
as well as the management regime of alley cropping to
improve the fertility and productivity of soils in the
semiarid tropics. These technical and social questions were
the prime motivations in undertaking this study. This study
had with four main objectives:
1. To determine if alley cropping with leguminous
MPTS substantially improves soil fertility and
productivity in the semiarid conditions of
Machakos, Kenya.
2. To determine or separate effects due to mulch per
se from those due to the in situ presence of the
hedges within the crop field.
3. To assess the patterns of tree-crop interaction
and to determine factors that may be responsible
for the observed interactions.
4. To determine factors that may significantly
influence adoption of the technology by farmers.
Plant productivity aspects of the crop and hedgerows
of leucaena and cassia are presented in Chapter 2. Chapter
3 addresses the effect of the alley cropping and mulch
quantity and quality on soil chemical, physical and
biological properties; trends over time as well as within a
cropping season are discussed. Patterns of N availability
determined by means of frequent soil sampling and/or in situ
mineralization using ion-exchange resins are also discussed
in Chapter 3.

10
In Chapter 4, studies on decomposition, mineralization
and uptake patterns of leucaena and cassia mulches are
presented. Uptake patterns of N and P from the mulch was
assessed by means of monthly harvests and determination of
nutrient content of potted maize plants. Chapter 5
addresses crop yield profile at the tree-crop interface.
For this, the last three of the seven croppings seasons of
the study were used. Also presented in Chapter 5 are the
results of field and pot studies of water and nutrient
status of soil at the tree-crop interface. These studies
were conducted to explain the observed interaction between
the hedgerows and the crop. Chapter 6 presents results of a
survey of 60 farms and discusses factors that may
significantly influence adoption of alley cropping in the
semiarid tropics.
Finally, Chapter 7 provides a summary of Chapters
1-6. It highlights the major findings of the study and
suggests aspects of alley cropping in the semiarid tropics
with potential for further research.

CHAPTER 2
PRODUCTIVITY ASPECTS OF ALLEY CROPPING WITH LEUCAENA
LEUCOCEPHALA AND CASSIA SI AME A IN SEMIARID TROPICS AT
MACHAROS, KENYA
Introduction
Alley cropping, also called hedgerow cropping, is an
agroforestry technology where annual crops are grown in the
alleys between rows of trees/shrubs that are regularly
pruned and maintained as low hedges. The prunings from the
hedgerows, when applied to the crops in the alleys, are
reported to improve or sustain crop yields. This is
suggested by many studies on cropping from the humid lowland
tropics. For instance, Kang et al. (1985) reported from
southern Nigeria that, on an Alfisol, using leucaena
prunings only, maize yield could be maintained at a
"reasonable" level of 2 t ha-1 as against 0.66 t ha*1
without leucaena prunings and fertilizer. An extensive
review by Kang et al. (1990) of other such studies from the
humid tropics showed similar results. Several other studies
also show that alley cropping can sustain soil nutrient
status and minimize organic matter decline in humid to
subhumid regions (Lai, 1989; Yamoah et al., 1986b).
Typically, alley cropping studies involve use of
leguminous trees/shrubs. This is because nitrogen
availability is commonly (though by no means exclusively) a
limiting factor in tropical agriculture. -It is also the
11

12
element which more than any other is under biological
regulation in natural ecosystems (Swift and Sanchez, 1984).
Nitrogen available to crops in alley cropping may be
enhanced through processes of biological fixation of
atmospheric N (Roskoski, 1986; Dommerques, 1987) and
subsequent release through processes of nodule sloughing
(Sanginga et al., 1986) or excretion of nitrogen from the
root nodules (Brophy and Heichel, 1989; Wacquant et al.,
1989). However, the significance of the latter process
remains doubtful (Vallis, 1978).
Prunings from the hedgerows are the main source of
nutrients to the crops. This is evident from the
significant declines in crop yield that occur when the
application of mulch is withdrawn (Kang et al., 1981). The
extent of supply of N from the mulch to the crop would
depend, among other factors, on the rainfall conditions of
the site, the soil, the hedgerow species, and management
practices such as spacing and the pruning regime. In the
humid lowlands of West Africa, Kang et al. (1981) and Duguma
et al. (1988) reported a wide range of N yields of about
150-560 kg N ha'1 yr'1 for alley cropped leucaena. Similar
information from the semiarid tropics is scarce. A major
exception is the work of Lulandala and Hall (1990) in
Tanzania that noted nitrogen returns through frequent
pruning of leucaena hedges could be as high as 88 kg ha'1
yr'1.

13
One of the impacts of the above-mentioned studies has
been to stimulate interest and studies on alley cropping in
the SATs. In many areas of SATs, soil fertility is low
(Steiner et al., 1988) and chemical fertilization alone
(i.e.f no irrigation) could improve yields (Ludwig, 1987).
Fertilizers are, however, expensive for many resource-poor
farmers and/or are not readily available.
In many regions, and in particular the SATs, the
prunings are more likely to be used for fodder than as mulch
(Singh et al., 1986; Ong et al., 1991). Even if the
prunings are used as mulch for the crops, the effects on
crop yields may not be significant (Singh et al., 1989).
Under these conditions, the N contribution by the hedgerows
to the soil may primarily be through below-ground processes,
particularly where N fixing species are used. Whether a
significant amount of N can accrue through these processes
and whether this N is available to the companion crop have
not been convincingly established (Ta et al., 1986; Danso et
al., 1987; Rao et al., 1987).
If growing the trees within the crop fields is not
beneficial for the crops but the trees can provide other
valuable benefits (e.g., fodder), separate planting of trees
might be preferable. This would be particularly appropriate
in the SATs where many studies report significantly reduced
crop yields under alley cropping (Singh, et al., 1986;
Mittal and Singh, 1989; Nair, 1987). Under these

14
conditions, the trees could be grown separately from the
crops, and their biomass harvested for use as fodder or
mulch in a "cut and carry system."
The magnitude of the effects of in situ planting of
the hedgerows would be influenced significantly by
differences in the rooting patterns and densities of the
hedgerow species. Trees with high root densities (expressed
as cm cm*3 of soil volume) in the rooting zone of the crops
are likely to result in less positive in situ benefits to
crops than trees with low root densities. Also, for a given
species, between-row spacing of the hedges, or put in other
words, the ratio of tree:crop land occupancy, could modify
the effects of in situ planting. Because of the increased
chances of competition, hedgerows with narrow between-row
spacings would be more likely to have less positive in-situ
effects on crops than widely spaced hedgerows. Narrowly
spaced hedgerows may, on the other hand, produce higher
quantities of biomass and provide greater scope for nutrient
recycling than widely spaced hedgerows.
From the foregoing, it is apparent that an
understanding of the in situ versus mulch benefits of
growing hedgerows in crops is necessary. Additionally, it
is important to have an understanding of the rooting
densities of species involved. The objective of this study
was to separate the effects of mulch from those of in situ
planting of the hedgerows on productivity (biomass and crop

15
yields) in alley cropping systems of Leucaena leucocephala
and Cassia siamea in semiarid Machakos, Kenya.
Materials and Methods
Site Description
The study was carried out under rainfed conditions at
the Field Station of the International Centre for Research
in Agroforestry (ICRAF) located at Machakos, Kenya (latitude
Io N 33' S, longitude of 37 14' E and 1596 m elevation).
Mean annual temperature is 19 C and annual rainfall ranges
from 400-1100 mm with a mean of 750 mm. This rain falls in
two seasons in almost equal amounts. There are two rainfed
cropping seasons for the principal agricultural crops. The
first rains of the year during March-June are locally
referred to as the long-rains; the second rains of the year
during October-January are referred to as the short-rains.
Of the two, the short-rains are more reliable than the long-
rains at Machakos. The seasons and the monthly rainfall
during the study period are presented in Figure 2-1.
Generally, monthly rainfall equals or exceeds
evapotranspiration only in April, May and November.
The soil at the study site is texturally well-drained,
dark, reddish-brown sandy clay, derived from the basement
complex gneiss. According to the Kenya/FAO (Food and
Agriculture Organization of the United Nations), the soils
are Luvisols, including the ferric and eutric forms (or

Rainfall (mm)
16
1988 1989 1989 1990 1990 1991 1991
SR lr SR LR SR LR SR
Month and cropping season
Figure 2-1. Seasonal rainfall and the local names
(long-rains and short-rains) of the cropping seasons during
the study.

17
Kandic Rhodustalfs according to the United States Department
of Agriculture Soil Taxonomy). The soils of the research
station are in general weakly to moderately leached, with a
weakly acid reaction (pH [water] 6.0-6.5), a medium base
saturation (50-80%), low to moderate levels of organic
matter (top soil organic carbon: 1%-1.5%), nitrogen (N)
(0.08%) and phosphorus (P) (<16 ppm). Other nutrients are
reportedly present in adeguate levels to support crop
production (Kibe et al., 1981).
Clay and sand content in the top 15 cm averaged 28%
and 67%, respectively, in 1987. Kaolinite is the
predominant mineral in the clay fraction. The top soil had
an effective cation exchange capacity of 9.0 meq per 100 g
of soil at the initiation of the study in October 1987.
Methods
Field study
Experimental layout.Hedgerows of Leucaena leuco-
cephala and Cassia siamea were established in October 1987.
Alley widths of the hedgerows were 4, 5 and 6.7 m. In-row
spacing of trees within the hedgerows was 0.5 m. Alley
widths used were computed based on field observations (Nair,
1987) that 3-4 year old hedgerows, spaced 4 m apart and
pruned frequently to low heights, occupied 25% of total
(aboveground) land area. Similarly, the land occupancy
ratio by the hedgerows spaced 5 and 6.7 m would be 20% and

18
15%, respectively. Sole stands of leucaena and cassia
occupying 15%, 20% and 25% of crop land were also
established for comparison to the hedgerow intercropped
plots. In-row spacing of the trees within the block/sole
stands of hedgerows was 1.0 m by 1.0 m. Control plots
included in the experiment were (a) maize crop which
received mulch only in amounts equivalent to what was
produced by the hedgerow intercropped plots and, (b) maize
crop which received no mulch but fertilizer only at the
rates of 40 kg and 17 kg of N and P ha*1 season'1.
The experimental design was split-plot, replicated
three times. Main plot treatments, randomized within the
blocks, were the hedgerow species (i.e., leucaena or
cassia). Subplot treatments were the factorial combination
of the three ratios of crop to tree land occupancy and the
control plots with mulch-only. The fertilizer-only control
plot was included in each main plot (i.e., 2 replications
per block). This gave 6 replications for the fertilizer-
only control. Main plot size was 20 x 50 m, with two meter
borders separating sub-plots of 5 x 20 m. Between blocks, a
4-meter wide path was provided. Sole stands of leucaena and
cassia hedges which provided mulch for the mulch-only
control plots were also included in each block.
' Initially, a control plot with no inputs was
inadvertently left out. The control plot of an adjacent
experiment, established about the same time as this study

19
was used instead. Nevertheless, since this control plot was
outside the setting of the present study, it was not
included in the statistical analyses. Rather, it was used
only for relative comparisons.
Root studies.Root studies were initiated in the last
cropping season of the study with the objective of
identifying the cause of differences in the effects of the
two hedgerows species on maize yield. These studies were
conducted on hedgerows outside the main experiment
consisting of two blocks of cassia and leucaena hedgerows,
6 m wide, established in 1983. Until 1988, maize and beans
were sown in rotation in the alleys between the hedgerows in
these plots. The maize crop was fertilized with N and P at
the rate of 40 and 18 kg ha'1 season'1; the beans were not.
In addition, prunings from the hedgerows were applied to the
cropped alleys. In 1988, fertilizer application was stopped
and maize was sown in both seasons.
For each species, blocks of soil 20 x 20 x 20 cm were
removed from six transects 3 meters long from the base of
the hedgerows during the 1992 short-rains cropping season.
The soil blocks were taken from 0-20 and 20-40 cm soil
depths. The samples were brought back to washing sheds in
plastic bags. The soil blocks were then washed with a spray
of water aided by hand manipulation to separate medium-sized
and fine roots from the soil in sieves with mesh sizes of 5
mm and 0.5 mm, respectively. Roots less than 2 mm in

20
diameter were classified as fine roots, those of 2-5 mm as
medium-sized or small roots (Bohm, 1979).
The fine roots of leucaena and cassia were
distinguishable easily from maize roots based on color and
texture. Maize had whitish and succulent roots, cassia had
black roots, and leucaena had fine, cream-colored roots,
which were strong relative to cassia. The majority of the
fine roots of the two species (which were planted
separately) and the maize roots were sorted in the same way.
Some were hard to differentiate and were separated under a
magnifier. Dead roots with brittle steel and degenerated
epidermal and cortical tissues were not included.
Root length was determined by the line-intercept
method of Tenant (1975) using a tray with grid units of 1.3
cm (half-inch). With these grid units, the total number of
intercepts equals the total length of root in centimeters.
Crop and tree management.The hedgerows occupied 15%,
20% and 25% of the gross land area. The remainder of the
land, viz, 85%, 80% and 75%, was sown to maize at
approximately 37,000 plants ha-1 with a spacing of 90 cm
between rows and 30 cm within rows. Katumani Composite, a
drought-tolerant maize variety was used. The crops and
trees were only rain-fed.
Fertilizer was applied at the rate of 40, 18 and 0 kg
ha'1 season'1 of N, P and K, respectively, to maize in the
fertilizer-only control, alley cropped maize and maize next

21
to the block/sole planted hedgerows. The mulch-only plots
received no fertilizer.
The hedgerows were pruned 2-3 times in a cropping
season (depending on the rainfall conditions of the season)
to 50 cm height aboveground. First pruning was done in
November 1988, this being the first of the seven cropping
seasons with the treatments applied. At each pruning of the
hedgerows, which typically took place 1-2 weeks before the
onset of rains, the biomass harvested (hereafter referred to
as mulch) was determined. It was then chopped into small
pieces, samples taken for dry matter determination and the
rest spread on the field. Typically, mulch harvested before
the onset of rains was incorporated into the soil during
field preparation and planting of the crops. However, mulch
from later pruning during the life of the crop was either
partially incorporated (in the process of weeding the crops)
or generally left on the soil surface. In all crops, the
plots were hand tilled and weeded.
At each harvest of maize, which took place
approximately 120 days after sowing, the yields of the
grain, cobs and stalks were determined. Samples of each
component were oven-dried at 105C for 48 hours and dry
weight determined. Maize stalks and cobs were not returned
to the plots, in line with the farmers' practice in the
study region where maize stalks are a valuable livestock
feed-.

22
In cropping seasons 1-4, yields of all maize rows in a
plot were composited at harvest. However, in seasons 5, 6
and 7, maize yields of each row in a plot was separately
harvested and recorded. Also, in the last three cropping
seasons, sub-samples of the grain, cobs and stover of some
selected treatments, as explained later, were analyzed for
nitrogen and phosphorus concentrations. Nitrogen was
determined by the standard micro-kjeldahl method and P by UV
visible spectrophotometry. In all seasons, two rows of
maize on all four sides of the plot were left as guard rows.
Pot Study
A pot study was initiated in January 1992 to determine
the amounts of mulch that would produce significant effects
on maize yields under a given rainfall condition. The
treatments included a factorial combination of 4 rates of
leucaena and cassia mulch and 2 levels of irrigation. The
four mulch rates were applications equivalent to 0, 2, 4 and
8 t ha*1 season*1, dry matter. The two irrigation levels
were the equivalents of 200 and 400 mm received in the last
two cropping seasons of the study. The pots were arranged
in a randomized block design with 3 replications in a
transparent polythene-roof house. In each block, pots with
one level of fertilizer at a rate equivalent to 40 N and 17
P kg ha'1 season*1 were also included.

23
Soil put into the pots (approximately 20 kg) was
collected from three randomly selected positions near (0.45
m) and away (3.5 m) from the alley cropped plot of the 6.7 m
wide hedgerows. The amounts of irrigation water applied to
the pots were based on the difference between soil water
storage at field capacity (15%) and wilting point (7.8%)
determined at the initiation of the study. The plots were
irrigated twice weekly.
Three seeds of maize were sown in each pot. Three
days after germination, the seedlings were thinned to one
per pot. Total plant harvest (i.e., shoot plus roots) was
done after four weeks. After oven drying at 65C for 48
hours, dry mass was determined. On the same day as the
harvest, the pots were resown for another four weeks. This
repeated sowing and harvesting from the pots was performed
for a total period of 16 weeks.
Yield ComputationCrop and Hedge
Biomass from Field Study
Biomass and maize yields of the hedgerow planting
systems (expressed in t ha1, dry weight) were computed to
include area occupied by both the hedgerows and the crop.
Hence, the yields are expressed on gross as opposed to net
area basis or area occupied by crop only. In situ effects
of the hedgerows were calculated as the difference in mulch
and crop yields of the systems where hedgerows were

24
intercropped and where they were planted in sole/block
stands outside the crop area.
To estimate in situ effects of the intercropped
hedges, the block-planted hedgerows should not affect the
adjacent crop. In an on-farm situation, the block-planted
trees could be on land not fit for cropping that is at some
distance from the crop land. In on-station situations,
distant planting may not be feasible, especially under
conditions where treatments are randomly located in small
plots. Two approaches, polythene sheets vertically inserted
into the soil to 50 cm depths (Ong et al., 1991) and
trenching (Verinumbe and Okali, 1985) have been used to
minimize effects of adjacent rows or blocks of trees on
crops. The effectiveness of these approaches depends on the
depths of insertion of the polythene sheets into the soil or
of trenching. Tree roots can go deep below the polythene
sheet or the trench and resurface (Corlett, 1989; Ong et
al., 1992).
In the present study, row-by-row harvest of the crop
of the widest alley (6.7 m) was used to determine the
distance of influence of the block-planted hedgerows on the
yield of the adjacent crop. A graphical plot of the yield
of the rows against the distance of the rows from the
hedgerows was then made. From this plot, the rows of maize
influenced by the block-planted hedges were determined and
discarded. The yields of the remaining rows were used to

25
calculate the per hectare yield of maize planted next to the
block planting system of hedges outside the crops.
In the section on results below, only yields of the
second season to the sixth season crops are discussed. The
yield of the first season was excluded because the hedgerows
were pruned only once (since they were not yet well-
established). In the other seasons the hedgerows were
pruned 2-3 times. Also, the crop yield of the first season
crop was at least twice as much as the yield of the other
seasons, ascribed more to the initial site conditions than
to the effects of treatments. Detailed presentation and
discussion of the biomass and crop yields are limited to the
6.7 m wide alley in the block planting system. This spacing
of the alleys and ratio of tree:crop land occupancy offers
the most land for cropping. This is of practical
significance to farmers.
Trends in Maize Yield over Time
Trends in maize yield over time were analyzed by
expressing the yield of each treatment in a season as a
percent of the average yield of all treatments of the second
F C:
crop. The patterns of maize yield over time were also
(
examined by expressing the yields of the two maize planting
systems as a percent of the yields of the fertilizer-only
control treatment.

26
The productivity of alley cropping compared to the
sole planting system of the hedgerows and the crop was
assessed by calculating land equivalent ratios (Willey,
1979) based on grain and total biomass yield of grain, stalk
and the hedgerows. The block planting and alley cropping
systems with the same ratio of land apportioned to tree and
crop (15:85) was used.
Nutrient Budget of Nitrogen and Phosphorus
A simple budget for N and P was computed for selected
treatments during the last three cropping seasons of the
study. It was only in these seasons that nutrient content
of the crop was regularly determined. The N budget was
calculated by considering N applied through the mulch or as
fertilizer, N removed in crop (grain, cob and stover), and N
still on the plot in mulch not decomposed. The P budget was
similarly calculated except that P remaining in mulch not
decomposed was not computed because sufficient amounts of
mulch did not remain for determination of both N and P.
Nitrogen and P removed from the plots were calculated
by summing the N and P removed by the crop (i.e., grain,
cobs and stover) for the last three seasons of the study.
Since N and P data for the crop were available for only
three out of the seven crops, regression equations relating
N and P removed in crops 5-7 to yields in these seasons were
determined. These equations were then used to calculate N

27
and P removed in crops of seasons 2-4. The regression
equations were obtained first by multiplying the N and P
concentrations of the crop (grain, cobs, stover) by their
respective yields to obtain N and P yields of the crop. The
N and P values were then regressed against the crop yields
which resulted in the following equations:
N(kg) grain yield = 0.015 grain yield 0.0006; r2=0.98
N(kg) cob yield = 0.0028 cob yield + 0.0002; r2=0.95
N(kg) stover yield = 0.0056 stover yield + 0.0008; r2*=0.91
Similarly, regression equations for P yield in grain, stover
and cob yields are as follows:
P(kg) grain yield = 0.0018 grain yield + 0.0002; r2=0.93
P(kg) stover yield ** 0.0021 stover yield -0.001; r2=0.92
P(kg) cob yield = 0.0005 cob yield 0.00002; r2=0.96
Nitrogen remaining in mulch not decomposed was calculated
using the N concentration of the soil-incorporated mulch at
the end of the study in March 1992. Typically, N remaining
in mulch is calculated using the decomposition constant of
the mulch (Palm, 1988). However, where the pattern of
decomposition of the mulch has more than one phase as was
the case in the present study (Chapter 4), use of one
decomposition constant for both phases may not be
appropriate. Nitrogen remaining was calculated for each
season's mulch input into the soil and then summed over all
the seasons.

28
Apparent nitrogen utilization efficiencies were
calculated for crops seasons 2-6 and includes grain, cobs
and stover. The efficiencies were calculated by determining
the nitrogen yield of the treatments where mulch was applied
to the crop (dry matter x nitrogen concentration) less
nitrogen yield of the no fertilizer control treatment and
expressing this as a percentage of the amount of nitrogen
via the mulch or the fertilizer. This method has also been
used by others to estimate utilization efficiency of mulch N
by crops (Kang et al., 1981; Yamoah et al., 1986b; Palm,
1988).
Statistical Analyses
The effects of the two species and systems of hedge
planting on crop and biomass yields were determined using
the General Linear Models (GLM) procedure in the Statistical
Analysis System (SAS, 1992). Because of the unbalanced
design caused by the fertilizer-only treatment (which was
applied at only one level as opposed to other treatments
which were applied at three levels.), the analysis was
performed in two stages. First, the fertilizer-only
treatment was omitted and the analysis of split-plot design
was performed. Second, depending on what factors were
significant in the split-plot analysis of variance (ANOVA)
test, comparisons with controls were performed in a
subsequent ANOVA that included all treatments. In both

29
analyses, differences between means of treatments were
determined by pairwise contrasts of the Least Squares Means
(LSM). A statistically significant difference was declared
to exist between means of any pair of treatments when the p-
value was less than or equal to 0.05.
Yield response of the systems over seasons was also
analyzed with the GLM/SAS procedure. The option for
repeated measures ANOVA was included in the GLM procedure to
account for the systematic effects of harvests from the same
plots over time. The difference in mean yield of treatments
between seasons was determined by pairwise contrasts of the
LSM.
The GLM/SAS procedure was also used to analyze the
root density data. The option for repeated measures
analysis was included to account for the effects of "fixed"
distances of root observations from the hedgerows. The
significance of differences between means of root densities
at various distances from the stumps of the hedgerows was
determined by pairwise contrasts of the least squares means.
In the pot studies also, the GLM/SAS procedure was
used to analyze the data. As in the field study, the option
for repeated measures was included in the ANOVA to account
for the effects of repeated harvests from the same pots. In
addition, orthogonal polynomial contrasts for linear,
quadratic and cubic effects were included in the GLM
procedure to test significance of the relationships between

30
irrigation levels, mulch species and mulch rates. Pairwise
comparison of LSM was used to test the significance of
differences between treatment means. Dunnet's test, which
controls the Type 1 experiment error for comparisons of all
treatments against a control, was used to determine
significance of differences between the mulch rates and the
control pots that had only one level of fertilizer
application.
Results
Field Study
Mulch yield
Leucaena, intercropped in hedgerows, produced
significantly higher mulch (+25%) than when the species was
planted in sole/block stands (Table 2-1). For cassia,
however, the +14% observed difference was not significant.
Mean mulch yield of leucaena, whether it was
intercropped or planted in sole/block stands, was
significantly higher (+46%) than that of cassia (Table 2-1).
Between the three hedgerow spacings of the intercropped
system, mulch yield of the 4 m hedgerows was significantly
(p = 0.04) higher (on average, 21%) than that of the other
two hedgerow spacings. The mulch yields of 5 and 6.7
m-wide hedgerows were not significantly different
(p = 0.23).
I

31
Table 2-1. Mulch yield of Leucaena leucocephala and Cassia
siamea intercropped hedgerows or planted in sole
stands outside the cropped area.
Hedgerow
Species
t ha-1
Alley
Cropping
y r1
Block
Planting*
Mean
Leucaena
4.4
3.3b
3.9s
Cassia
2.2a
1.9*
2. lb
Mean
3.3
2.6
3.0
Block planting in an area equal to that estimated to be
occupied by the hedgerows in an alley cropping system.
*-bFor a given species, means followed by the same letter are
not significantly different (p = 0.05).

32
In the first year, there was no significant difference
between mulch yields from the two species, but after the
second year, leucaena yields were significantly higher than
those of cassia (Fig. 2-2). Also, a significant difference
was observed in the response of the two species to rainfall.
While changes in rainfall had a significant effect on
the mulch yield of leucaena, no effects were observed for
cassia. Mulch yield of leucaena peaked in 1990 when the
annual rainfall was highest (964 mm) during the study
period. In 1991 when rainfall dropped to only 22% of the
1990 total (Fig. 2-1), mulch yield of leucaena dropped by
half. For cassia, there was no significant response (Fig.
2-2). Cassia mulch yield tended to decline over time,
although the differences in yield between years were not
significant. Although the effect of rainfall on mulch yield
was more pronounced than the effect of age of the hedgerows,
given the same rainfall, mulch yield of leucaena did
increase with age. For instance, with similar seasonal
rainfalls total of 330 mm, mulch yields of leucaena
hedgerows were 40% more in the fifth season than in the
second season. Similar trends of increase in cassia mulch
with age were not detectable.
Where hedgerows were planted as sole stands/blocks
outside the crop fields, differences in mulch yields per
hectare among the three tree:crop ratios were not

Mulch yield (t ha _1)
33
3.5
3.0
2.5
a
1989
1990
Year
Leucaena
Intercropping
Leucaena
Sole
Cassia
Intercropping
Cassia
Sole
Figure 2-2. Mulch yield of Leucaena leucocephala and
Cassia siamea in alley cropping and block/sole planting
systems.

34
significant. However, leucaena always yielded significantly
more mulch than cassia.
A count of all plants in the three replicates in the
last season of cropping showed that cassia had a higher
percentage of dead plants than leucaena (using the number of
seedlings planted in 1987 as a base). In cassia, up to 5%
( 7% standard error) of the intercropped hedges and 11.6%
( 15% standard error) of the block planted plants were
dead. In contrast, no plants were recorded as dead in
either leucaena systems. Cassia's mortality was largely
attributed to a commonly observed stem borer, Xylentes
capensis (Cassidae) and a wood boring beetle, Xyloposus sp.
(Bostrychidae).
Nutrient concentrations and contents of the mulches
In addition to a supply of N and P, the mulch provided
substantial quantities of other nutrients to the soil (Table
2-2). While concentrations of the other nutrients remained
stable over time in both species, the concentration of P
increased considerably. For instance, the levels of P in
cassia mulch increased 25% from 1987 to 1991 (compare 0.20
to 0.25%).
Crop yields
The combined analyses of the data over seasons
indicated that there were significant interactions between

35
Table 2-2. Concentration and amounts of nutrients in Leucaena
leucocephala and Cassia siamea mulch from alley
cropped hedgerows.
Species
Nutrient concentration (%) and amounts (kg
ha'1 yr'1, in parenthesis)
N
P
K
Ca
Mg
S
Leucaena
3.5
0.2
1.9
0.7
0.3*
0.1
(136.0)
(7.5)
(71.7)
(26.4)
(11-3)
(3.8)
Cassia
2.9
0.3
1.8
0.7
0.2
0.1
(54.7)
(5.7)
(33.7)
(13.2)
(3.8)
(1.9)
Within a column, values of nutrient concentration with
different letters are significantly different (p = 0.05).
The amounts of all nutrients in leucaena mulch were
significantly higher than those in cassia (p = 0.05).

seasons, systems, and hedgerow species on crop yield.
Therefore, ANOVA by season was necessary in order to
understand the true main effects of species and systems.
The ANOVA by season showed that in four out of the six
(i.e., seasons 2, 4, 5, and 6), alley cropped maize yielded
higher with cassia than with leucaena (Table 2-3). A
significant difference between the yield of leucaena alley
cropped maize and that next to the block planted system of
leucaena was observed in only two out of the six seasons
(the third and the fourth). For cassia, the difference was
also significant in two out of the six seasons (the second
and the third). The difference in maize yield under the two
systems represents the in situ effects of the hedgerows on
crop yield. Averaged over the seasons, the in situ effects
of cassia on maize yields were 22% while those of leucaena
were 9%.
Except in the second season, differences between the
effects of the three hedgerow spacings (i.e., 4, 5, and 6.7
m) on maize yield were not significantly different (Fig.
2-3). In the second season, maize yields under the 6.7 m
hedgerows were significantly higher than those of the other
two spacings. In the block planting systems, significant
differences in maize yields were observed in seasons 2 and 4
only and 4 only (Fig. 2-4).
A comparison of the mean yields of all the treatments
(Table 2-3, last column) shows that the fertilized maize

37
Figure 2-3. Effect of three hedgerow spacings on
maize grain yield, averaged over yields under the two
species, Leucaena leucocephala and Cassia siamea. Vertical
bars indicate standard error of difference of means.

Maize grain yield (t ha"}
Percent land under tree and crop
38
Figure 2-4. Effect of sole stands of Leucaena
leucocephala and Cassia siamea planted at three land
occupancy on maize yield during six seasons. Vertical bars
indicate standard error of difference of means.

Table 2-3. Effect of alley cropping and block planting systems of maize grain yield
during six cropping seasons. Rainfall for each season is included in the last
row.
System
Maize yield (t ha
_1) for each season
Mean
2
3
4
5
6
7
Alley cropping
(leucaena)
2.5a,b
3.0a,b
2.3a'b
2.3a,b
0.6a,b,c
2.3
2.2a,b
Alley cropping
(cassia)
2.4a'b
3.9d
2.4a,b
3.4a
1.4a
2.8C
2.7C
Block planting
(leucaena)
2.5a
2.7b,c
2.0C
2.6b
0.5C
1.7b
2.0a
Block planting
(cassia)
2.3a,b
2.9b,c
2.0C
2.9a,b
0.9a'b'
1.5b
2. la'b
Control (leucaena mulch)
2.0a,b
2.6b,c
2.2a,b'c
1.6d
0.8a,b'c
1.7b
1.8b
Control (cassia
mulch)
1.9a,b
2.3C
2. i-.b,c
2. lc,d
0.7a'b,c
1.8b
1.8a
Control (fertilizer)
1.6b
3. la-b
2.9a,b
3. la'b
1.3a,b
2.7#
2.5C
Control (absolute)*
2.0
2.9
1.8
2.1
1.1
2.4
2.1
Mean
2.1
2.8
2.1
2.4
0.9
2.0
2.1
Rainfall (mm)
330
441
631
333
214
373
336
a'b'c'dwithin a season, values followed by different letters are significantly different
(p = 0.05).
to
From an adjacent experiment.

40
(control) yielded significantly higher than all the other
treatments except the cassia alley cropping system. Lack of
major differences between maize yields of the various
treatments across the seasons of the study is evident from
Table 2-3. In general, maize yield was positively related
to the rainfall patterns. However, in the third season, in
spite of the relatively high rainfall (631 mm), the average
of all treatments was only 2.2 t ha'1 which was similar to
the yield obtained in some other seasons with lower rainfall
(e.g., seasons 1 and 6 with rainfall of 330 mm and 373 mm,
respectively). The low yield with 631 mm rainfall could be
ascribed to excessive soil loss and runoff that also caused
physical damage to the crop; about half the rainfall was
estimated to have been lost in the runoff (data not
presented).
Maintenance of maize yield over time
Trends of maize yield over time were analyzed by
expressing the yield of each treatment in a season as a
percent of the average yield of all treatments of the second
crop (Table 2-4). The yield of all treatments generally
remained stable over seasons, including that of the control
with no inputs. Yield increased or declined mainly in
direct response to the season's rainfall, with a linear
relationship between rainfall and yield (Fig. 2-5). This

41
Table 2-4. Effect of alley cropping block planting systems on
relative grain yield of maize.
System/Treatment
Season
2
3
4
5
6
7
Mean
Aliev cropping
Leucaena
100
124
95
95
25
90
88
Cassia
95
157
95
138
57
114
109
Block planting
Leucaena
119
129
95
124
24
81
95
Cassia
110
138
95
138
43
71
99
Mulch-only
Leucaena
96
124
105
76
38
81
87
Cassia
90
110
100
100
33
86
87
Fertilizer-only
76
148
138
149
62
129
117
Absolute control
96
138
86
100
52
114
98
Mean
98
134
101
115
42
96
98
NOTE: Relative yield was calculated as a ratio of actual to
average yield of all treatments in the second season,
expressed as percent.

Maize grain yield (t ha
42
Figure 2-5. Relationship between rainfall and maize
mean yield of all treatments in each of the six cropping
seasons. Numbers on the regression line refer to the
cropping seasons.

43
suggests that at this site, soil moisture was more limiting
than differences between treatments.
The patterns of maize yield over time were also
examined by expressing the yields of the two maize planting
systems as a percent of the yields of the fertilizer-only
control treatment (Fig. 2-6). Averaged over seasons, maize
yield declined by 12% per season under leucaena alley
cropping, with the largest decline in the third year (season
6). In contrast, maize yield under cassia alley cropping
increased by 8% per season. In the second year (season 6),
when yield under leucaena declined by 45% compared to the
control, yield under cassia was 27% higher than the control.
The increase in maize yield under cassia, in spite of the
hedgerows occupying land, suggested that productivity of
maize (kg m~2) under cassia alley cropping was higher per
unit net (cropped) area than of the sole crop (Fig. 2-7).
Most of the reduction in maize yield and in productivity
under leucaena, compared to the yield of the sole crop,
occurred in the last three seasons of the study. The
patterns of productivity per unit of land or rainfall were
similar. On average, mean maize productivity per unit of
rainfall under cassia alley cropping was +9% while that
under leucaena was -15% (Fig. 2-8). The increased maize
productivity under cassia alley cropping offsets the 15%
loss in maize yield when hedgerows occupied 15% of the
croppable land.

Maize yield as % of sole crop
44
140
120
100
80
60
40
Leucaena
Cassia
Leucaena
Cassia
(intercropped)
(intercropped)
(sole)
(sole)
&
MifM $
IT1
1989
1990
Year
1991
Figure 2-6. Maize grain yield under alley cropping
and block planting systems of Leucaena leucocephala and
Cassia siamea expressed as percent of the sole maize control
yield.

Maize grain yield (kg m'2 )
45
t
/
Figure 2-7. Maize productivity (kg m'2) under alley
cropping Leucaena leucocephala and Cassia siantea hedgerows
for six seasons. Vertical bars indicate standard error of
difference of means.

Maize yield (kg mm*1 of rainfall)
46
Figure 2-8. Productivity of maize (kg nun'1 of
rainfall) under alley cropping of Leucaena leucocephala and
Cassia siamea hedgerows for six cropping seasons. Vertical
bars indicate standard error of difference of means.

47
Land equivalent ratios (LER)
Alley cropping leucaena and cassia had LER greater
than 1.0 of maize grain yield or the combined yield of the
crop and hedgerow biomass. Averaged across species and
seasons, LER of maize grain yield under alley cropping was
1.24 (or 24% advantage over the maize grain of the sole
planting system of the hedges and crop). No particular
trend in LER of maize yield across seasons was discernible
under both species (Table 2-5). Between the two species,
LER of maize grain yield under cassia (1.36) was
significantly higher than under leucaena (1.12). This
difference was due to the lower crop yield under leucaena
compared to cassia.
When averaged across species, the combined biomass of
the hedgerows and the crop yield (grain plus stalks) had LER
of 1.10 (i.e., an advantage of 10% by alley cropping over
the sole system). The LER of maize grain yield under alley
cropping was greater than 1.0 because of lower yield of
maize yield under the sole system compared to alley cropping
in nearly all seasons. The biomass yield of the hedges
under alley cropping and sole planting systems were similar
when the highest yielding sole planting system was used in
LER computation.

48
Table 2-5. Land equivalent ratios (LER) of alley cropping
Leucaena leucocephala and Cassia siamea compared with
the block planting system with same ratios of tree and
crop land occupancy.
Year and Seasons Within a Year
Species
1989
1990
1991
Mean
1
2
1
2
1
2
Leucaena
1.0*
1.09
1.15
0.87
1.23
1.35
1.12
(1.10)
(0.97)
(1.15)
(1.02)
(1.29)
(1.26)
(1.13)
Cassia
1.04
1.33
1.21
1.17
1.56
1.87
1.36
(0.99)
(1.24)
(1.00)
(0.96)
(1.10)
(1.15)
(1.07)
LER = (Biomass yield of the tree species in alley cropping plus
yield of crop in alley cropping) divided by (biomass yield of the
tree species in sole system plus yield of crop in sole system) .
In both the alley cropping and sole planting systems, the percent
of land occupied by the hedges and the crop was 15 and 85,
respectively. In the sole planting system, the biomass yield of
the hedgerows in the 25:75 tree:crop percent land occupancy ratio
was used because this was the most appropriate (i.e., highest
yielding) sole planting system.
*LER of maize grain yield only. Number in bracket is the total
LER, i.e., the grain and the stalk yield of the maize and the tree
biomass which was used as mulch.

49
Nutrient budgets
Nitrogen.Generally, leucaena had significantly
higher mulch yield (Table 2-1) and N concentration (Table
2-2) than cassia. As a consequence, the N yield of leucaena
was significantly higher (136 kg N ha-1 yr*1) than that of
cassia (59 kg N ha*1 yr*1).
Nitrogen uptake was high and essentially of similar
pattern in all treatments, including the absolute control
(Table 2-6). It was only the fertilizer-control and the
alley cropping system of cassia that had N-yields
considerably higher than the rest of the treatments. This
is because of the relatively high maize yields of these two
treatments (i.e., the fertilizer-only control and the
hedgerow intercropped plots of cassia).
Apparent recovery/utilization efficiency of N applied
through the mulch or fertilizer was low in all treatments
(Table 2-6, columns 7 and 8). The N-recovery values were
low primarily because maize and N-yield of the absolute
control were almost equal to those of the other treatments
with inputs. This suggests that the N needs of the crop
were adequately met by native soil N. Averaged across
treatments, an apparent N-recovery of 9% summed over the N
recoveries through the grain, stover and cobs was observed.
Most of the N (67%) was recovered in the grain component of
the crop and the rest (33%) in the stover and cobs.

Table 2-6.
Nitrogen budgets, utilization efficiencies and yield of maize grown in alley
cropping of Leucaena leucocephala and Cassia siamea, at a hedgerow spacing of 6.7
m for 6 seasons.
N from
mulch
and/or
fertil
izer
N removed during 6
kg ha*1
seasons
N-utilization
efficiencyb
m
N-remaining
in mulch
Unac
count
ed for
N
Maize
yield
(t/ha/
sea-
Treatment
(kg)
Grain
Cob Stover
Total
Grain
Total
kg ha'1 (%)
(%>
son)
Absolute
control*
0
179
9
89
277
-
-
- -
-
2.1
Alley
cropping
(Leucaena)
648
186
9
83
278
1.0
0.2
103 16
41
2.2
Alley
cropping
(Cassia)
417
225
9
93
327
11
12
39 9
13
2.7
Mulch-only
(Leucaena)
408
192
10
87
289
3
3
87 21
8
1.8
Mulch-only
(Cassia)
177
190
9
91
290
6
7
33 19
0
1.8
Fertilizer
only*
240
222
9
97
328
18
22
-
0
2.5
aFertilizer was applied to the alley cropped plots at the rate of 40 and 17 kg ha-1 season'1 N and
P, respectively.
'Sample calculation using maize grain yield of 6 seasons from alley cropping of cassia: N
utilization efficiency (11%) N removed in grain yield (225 kg) less N removed in grain yield
of the absolute control (179 kg) divided by total N applied (417 kg) times 100.
cBased on N concentration of the mulch at the end of the litter bag decomposition studies of
mulch incorporated in October 1991 to March 1992.
d100 -(percent of applied N in mulch or fertilizer removed in crop plus N remaining in
undecomposed mulch).
*From an adjacent experiment.
U1
o

51
Trends in N uptake were similar to those of the crop
yields, i.e., the higher the yield of the crop, the higher
the N-yield. Because of the lower yields under leucaena
alley cropping compared to those under cassia (Table 2-3),
apparent N utilization efficiency under leucaena was low.
However, in the sole croppings with only mulch application,
apparent N utilization efficiency values from the mulch of
leucaena and cassia were similar.
N uptake and apparent N utilization efficiency were
highest in the fertilized control plot, in both grain as
well as stover and cobs. N uptake by the maize in the
fertilizer-control treatment was 42% more than the amount of
N applied. Similar high N uptake levels exceeding the
amounts applied were also observed with cassia alley
cropping. For these two treatments, all of the N applied
through the fertilizer or mulch could be accounted for.
Phosphorus.In six seasons, P removed by the crop
(i.e., grain, stover and cobs) significantly exceeded the
amount of P added through mulch or fertilizer (Table 2-7).
Averaged across species, P removed through the crop was 164%
more than P applied through the mulch of leucaena and
cassia. As a consequence of the lower mulch yield of cassia
compared to leucaena, P applied through cassia mulch was
significantly lower than P applied through leucaena mulch.
Hence, P removed through the crop in plots with cassia was
significantly higher (215%) than that from plots with
/

52
Table 2-7. Addition of phosphorus through mulch and removal
by maize in six seasons of cropping under
Leucaena leucocephala and Cassia siamea alley
cropping.
Mulch Source
Phosphorus
Addition
(kg ha"1)
Removal
Difference
Leucaena
22.8
52.1*
-29.3*
Cassia
15.8b
49.8
-34.0b
Mean
19.3
52.0
31.7
Values in a column with different letters are not
significantly different (p = 0.05).

53
leucaena mulch (127%). The interaction between N and P
uptake/removal by the crop was not examined in this study.
Root Density
Root density differed substantially between species
and also between soil depths. In both species, the
densities of both fine- and medium-sized roots were
significantly higher near than away from the hedgerows (Fig.
2-9). Fine root densities of leucaena at both 0-20 cm and
20-40 cm soil depths were significantly higher at nearly all
distances from the hedgerows than that of cassia (Table
2-8). Interaction effects between species and soil depths
on root density were not significant. In both species, root
density at 20-40 cm depth was significantly higher than at
0-20 cm.
Pot Study
During four harvests, the species effects and the
interaction effects of species and mulch rate on the biomass
yield of maize were not significant (Fig. 2-10). However,
effects of the mulch rates on maize yield were highly
significant, with the yields increasing direct proportion to
mulch rates (Fig. 2-10). Averaged across species and
irrigation levels (i.e., the eguivalents of 100 and 200 mm
of rainfall in two months period), biomass yield of maize
from the 0 and 2 t ha"1 mulch rates were not significantly

Root density (cm cm3)
54
0.12
0.1
Leucaena (0-20)
Cassia (0-20)
0.08
0.06
0.04
0.02
0
Leucaena (20-40) ^ Cassia (20-40)
80 100 120 140 160 180 200 220 240 260 280 300
Distance (cm) from tree hedges
Figure 2-9. Fine root density in the top 40 cm soil
profile of alley cropped hedgerows of Leucaena leucocephala
and Cassia siamea.

55
Table 2-8. Fine root density at two soil depths along a
meter-long transect from 6 m apart hedgerows of
Leucaena leucocephala and Cassia siamea
intercropped with maize.
Hedgerow Species
Soil
0-20
deDth (cml
20-40
Mean
Leucaena
0.41
0.64
0.53*
Cassia
0.23
0.47
0.35b
Mean
0.32
0.56b
0.44
NOTE: Fine root density = cm root per cm3 of soil.
*,bFor a given row and column, means followed by the same
letters are not significantly different (p = 0.05).

Biomass yield of maize (g pot'1)
56
Figure 2-10. Effects of three mulch rates of
Leucaena leucocephala and Cassia siamea on the biomass yield
of maize grown in pots, 1991.

57
different. The effects of these lower rates were also not
significantly different from the effects of the fertilizer
only.
The effects of irrigation depended on the harvest. In
harvests 1 and 4 (i.e., 4 and 16 weeks after mulch
application to the pots, respectively), differences in yield
between the two levels of irrigation were not significant.
However, in harvests 2 and 3, the differences in maize
biomass yield were significant.
Discussion
Mulch Yield of the Hedgerows
On average, leucaena produced 46% more mulch than
cassia. The higher yield of leucaena was observable in both
systems of hedge planting. Similar differences in the mulch
yields of the species had been noted earlier by Nair (1987)
working at the same site. The differences in biomass
production between the two species could be ascribed to
inherent growth differences as well as their site
utilization efficiencies. In the lowland humid tropics,
differences in the biomass yields of leucaena and cassia
appear to be even more pronounced than in the semiarid
tropics (Kang et al., 1981; Kang et al., 1985). Averaged
over the three years of study, mulch yield of leucaena was
approximately 4 t ha*1 yr'1 (dry matter) Similar yields of
leucaena have also been reported by Rao et al. (1990) from

58
alley cropping studies spanning over 4 years from semiarid
India.
Between the two systems of planting (i.e.,
intercropped vs block/sole stands), intercropped hedgerows
of leucaena yielded significantly higher (+33%) mulch than
the block planting on an egual area basis in all but in the
first year. In cassia, differences in the mulch yield of
the two systems were not significantly different. The
higher biomass of the alley cropped plots could be
attributed to (a) availability of more space and hence
reduced competition for soil water and nutrients (especially
when the annual crop was not in the field) and, (b) indirect
benefits from husbandry practices of the maize crop, e.g.,
fertilization, weeding and tillage. The higher mulch yield
of the intercropped hedgerows compared to the yields of the
block/sole stands indicates the benefits of growing the
hedgerows within the crops.
In the block planting system, although the hedges
occupied the same land area as the intercropped hedges, they
were confined to limited space and not spread out. Hence,
plant-to-plant competition was relatively severe.
Typically, the hedges at the borders of the block planting
system always grew taller and yielded more biomass than the
rows inside the blocks. Enhanced tree-to-tree competition
may have contributed to the slightly higher mortality,

59
especially of cassia, observed in the hedges planted in
block/sole stands compared to those intercropped.
The high pruning frequency (4-6 times a year) may also
have stressed the cassia hedges more than leucaena, and
subsequently made cassia less resistant to pest attack.
Another experiment on the site of the present study (P.
Kiepe, personal communication) showed that with only two
prunings per year, mulch yields similar to those obtained in
the present study could be obtained from intercropped
hedgerows of cassia. Nevertheless, reduced frequency of
pruning may exaggerate crop loss under alley cropping due to
enhanced shading.
It is also possible that the frequent pruning and
"export" of mulch from the area of the block planted hedges
impoverished the site in terms of nutrient storage, which in
turn contributed to the low yields of the block planted
hedges.
Crop Yield
Effects of mulch on yield
Because yields of maize in the mulched and unmulched
control plots were comparable, it would appear that low
levels of mulch obtained from the hedgerows limited maize
yields. Significant response in biomass yield of maize to
increasing rates of mulch in the pot studies also supports

60
this observation. This response could be ascribed to
enhanced availability of nutrients; positive impact of mulch
on soil physical properties (e.g., improved aeration and
moisture conservation) could be another reason, but this
aspect was not examined in this study.
It has been argued that if mulch applied in a season
could be increased to 3-4 t ha*1, maize yield may be
increased significantly even in sites similar to the one of
the present study with relatively fertile soils. For
instance, Mugendi (1990), working at a site adjacent to the
present study, reported from a study spanning 14 seasons
that application of leucaena and cassia mulch at rates
equivalent to 5 t ha"1 consistently gave significantly
higher maize yields than applications of 1 t ha*1 or those
of the control plot with no mulch additions. Moreover,
more pronounced effects would be expected in nutrient-poor
than in nutrient-rich sites. For instance, Palm (1988)
observed that in more nutrient-poor and/or in acid soils,
increased rates of mulch application resulted in crop
responses similar to those of inorganic nitrogen
fertilizers.
Because a species like leucaena has a higher fodder
value compared to cassia which has no fodder value (NAS,
1984), in semiarid areas, the prunings from the leucaena
hedges are more likely to be removed from the field for use
as fodder (Singh et al., 1989). The use of the prunings for

61
fodder becomes even more appropriate when mulch may not
result in significant gains in crop yields. Similar
observation was made by Singh et al. (1989) from semiarid
India. Furthermore, in nutrient-poor and/or acidic soils,
crop yields may not be sustained by mulch even if all the
crop residues and hedge mulch remained in the crof fields
(Szott et al., 1991).
Although the small quantities of mulch from alley
cropped hedgerows may not improve crop yields, it could be
argued that their application may still be necessary to
minimize long-term declines in soil fertility and
productivity. It is hypothesized that such declines in
fertility and productivity could result from loss of
nutrients removed in continuous harvests of mulch and tree
products (Kimmins, 1977; Hook et al., 1982; Lulandala and
Hall, 1988). With or without the removal of hedgerow mulch,
initial soil conditions, crop type and their yields would
likely influence the onset of yield declines. Use of manure
from livestock fed with the prunings may reduce yield- and
soil-fertility declines. In addition, in the semiarid
tropics where the crop yields are generally low compared to
those of the humid tropics, removal of mulch in the range of
1-2 t ha"1 season"1 may not show any significant impact on
yields and fertility conditions of the soil, at least in the
short term. This view is supported by the absence of
significant differences in maize yield between plots with

62
mulch and those without for the seasons of this study. From
the foregoing, therefore, it is evident that improvements in
crop yields through the effects of mulch produced by alley
cropped hedgerows are probably limited in the semiarid
tropics.
Supply of N and P by the mulch
and removal by the crop
Leucaena mulch, when returned to the soil, provided
more N than what the crop removed in 6 seasons. On the
other hand, cassia supplied only half of the crop N. Most
of the N removed was harvested in the grain (67%), an
observation in agreement with that of others (Allison,
1973). Based on the N budget analysis, it does appear,
therefore, that if all of the mulch N is available to the
crop, leucaena could meet the crop's seasonal N
requirements.
Besides N, the supply of the other nutrients is
important for the maintenance of high maize yield over time.
Removal of the mulch which supplies substantial quantities
of nutrients other than N could therefore have significant
effects on yield, even when N is provided via fertilizer.
This view is supported by observations of Gichuru and Kang
(1989) who showed that plots receiving no prunings had, in
general, lower maize grain yield regardless of N application
compared to those receiving prunings.

63
Soil P appears to be a nutrient that could rapidly be
depleted through continuous harvest of mulch and crops from
the systems studied. This is particularly so in soils such
as the Alfisols of the present study where the native
exchangeable soil P levels are low (less than 16 ppm),
inputs through the mulch are low, and, yet, outputs through
crop harvest exceed inputs significantly. The 7 kg ha-1 yr"1
P supplied through leucaena mulch is similar to P levels in
leucaena mulch noted by Lulandala and Hall (1990) from
semiarid Tanzania. Most of the P in the crop was removed in
the grain and stover, in almost equal percentages. Only
about 3% of the total P in the crop was in the cobs.
Averaged over species, only 37% of the P needs of the crop
was balanced by leucaena and cassia mulch. The shortfall
was more pronounced in cassia alley cropping because of
cassia's low mulch yield compared to leucaena's.
Based on the P budgets (notwithstanding the
possibility that not all mulch P is available to crop), it
does appear that the recycling potential of cassia mulch
from the alley cropping system is inadequate to meet the P
requirements of the maize crop. The balance between P
supply and crop need may be even less favorable under
conditions of high crop yield such as in seasons of good
rainfall. Kang and Wilson (1987) observed that the P
content of leucaena prunings was inadequate for one crop of
maize at the humid lowlands of Nigeria. Also, while

64
mulch-P, particularly that of leucaena, could meet the P
requirements of maize grain yield of about 2 t ha'1 as
suggested by Torres (1983), it is evident that it does not
meet the total needs of the crop (i.e., grain, stover and
cobs).
The considerable increase (16%) in the P concentration
of cassia mulch in 1991 over the levels of 1987 and the
absence of similar increase in leucaena mulch suggest
differences in the patterns of P uptake by the two species.
The cause of this difference is unknown. Some postulations
are that cassia is more mycorrhizal than leucaena or
supporting more phosphate solubilizing bacteria than
leucaena in its rhizosphere.
Apparent utilization efficiency of mulch-N
The apparent low use efficiency of mulch-N can best be
explained by the equally high crop and N yield of the sole
maize control compared to the yields of mulch treatments.
This is another indication of the presence of sufficient
native soil N at the study site for the maize crop without
additional inputs. Low utilization efficiencies of mulch-N
have also been reported by several other workers. For
example, Palm (1988) observed mulch-N utilization efficiency
of 16% by upland rice from mulches of leguminous shrubs.
Yamoah et al. (1986b) noted N-utilization efficiencies of
near 20% from the mulch of leguminous shrubs with the

65
addition of fertilizer. Sanginga et al. (1988) also
observed low N-utilization efficiencies from leucaena mulch
but they emphasized that the absence or presence of N-fixing
bacteria inoculation made a difference. Other workers (Kang
et al., 1981; Guevarra, 1976) have also reported low
apparent N-utilization efficiencies from mulch of leguminous
shrubs.
Factors such as time lags between mulch-N
mineralization and crop N demands (Anderson and Ingram,
1989), immobilization of mulch-N and soil-N by microbes in
the process of mulch decomposition, at least in the initial
stages (Palm and Sanchez, 1991) and volatilization losses
from surface applied mulch (Terman, 1979; Nelson, 1982)
could also have contributed to the apparent low mulch-N
utilization efficiencies observed. Also, about 22% of the
undecomposed mulch-N was still tied up in mulch of both
species. This cannot be described as inefficiency, because
this N is still in the system and might be utilized by
subsequent crops. Admittedly, without more precise
determinations such as the use of 15N techniques, it is
difficult to assess the fate of mulch-N in soil (Sanginga et
al. 1988). Management practices that synchronize periods of
peak N release from mulch with periods of peak N demands by
the crop may enhance effects of the small quantities of
mulch available from the hedgerows.

66
In situ effects of hedgerows on crop yield
The in situ benefits of growing the hedgerows was
reflected in the higher maize yields between the alley
cropping system and the block planting system and LER of
alley cropping > 1.0. The in situ effects of the hedgerows
may result not only from their root and rhizosphere effects
but also from their ability to protect the soil against
erosion and to serve as windbreaks (Wiesrum, 1984). From
semiarid India, Rao et al. (1991) provide evidence for
reduction in runoff and soil loss under alley cropping with
leucaena.
The in situ benefits of the hedgerows will be in
perspective when viewed against the yields of the
appropriate sole-crop control. This will determine if,
indeed, there has been a positive gain over the no
intercropping situation. Such a comparison can be made from
the last column of Table 2-3, which shows, for instance,
that the maize yield of the sole-maize control (fertilizer-
only) was -8% compared to that of cassia alley cropping and
+12% compared to the leucaena alley cropping. Because of
reduced maize yields under leucaena alley cropping, LER or
the yield advantage of leucaena alley cropping was
significantly lower than that under cassia. These
differences in yield reductions under alley cropping of the
two species suggest that leucaena is more competitive with
maize than cassia. The reductions in maize yield under

67
leucaena alley cropping could be more severe than observed
in the present study in sites that are both nutrient-poor
and dry (Chirwa, 1991).
Leucaena became more competitive than cassia starting
in the second year. Similar observations on the onset of
leucaena's competition were made also by Rao et al. (1990)
working with alley cropping in semiarid India. The
significantly higher mulch production as well as the
presence of high fine-root densities of leucaena in the top
soil compared to that of cassia provides additional clues to
leucaena's higher competitiveness. It is possible that a
higher pruning frequency than the 4-6 times a year of
leucaena in the present study may have reduced its
competition with the intercrcropped maize.
The higher maize yield under cassia alley cropping
compared with that under leucaena or under the sole crop
could be ascribed to significantly more positive crop-
hedgerow interactions observed under cassia. These positive
interactions could have mitigated high reductions in maize
yields from hedgerows with narrow alleys compared to those
with wide alleys. This could explain the lack of
significant differences in the three alley widths (spacings)
used in the present study. Because more land was under
hedges in the narrow alleys than in the wide alleys, maize
yield from the narrow alleys was expected to be less than
from those of the wide alleys.

68
Maize-yield reduction observed under leucaena was not
likely due to competition for light, because the hedges were
pruned frequently during the rainy season. Also, it was
probably not due to competition for nutrients because the
plots were fertilized and leucaena (which was inoculated at
the start of the study) may have fixed considerable levels
for its own N needs. Instead, the reduction in yield was
probably most likely due to competition for water, a
suggestion also made by others regarding leucaena alley
cropping in semiarid India (Rao et al., 1991). In the sixth
season when rainfall was 39% below normal, the significantly
lower maize yields under leucaena alley cropping compared to
the control provide support to the explanation of water
limitation. The argument that competition was more for
water and less for nutrients was supported by the results of
additional field studies (Chapter 5) involving factorial
combination of fertilizer (N and P) and irrigation levels
where a positive response in maize yield to N application
was not significantly different from that due to the effects
of irrigation only. Prior to the initiation of the present
study, the site was a savanna woodland cropped to pigeon pea
for one-and-one-half years, so that soil N may still have
been relatively high. Another evidence for water being more
limiting than nutrients is provided by the linear
relationship between the maize yield and rainfall over the
six seasons.

69
Trends in maize yield over time
All treatments had similar levels of relative yield
maintenance as long as rainfall received in a season was
normal, a situation more pronounced under leucaena than
under cassia. Much longer studies than three years are
required because the effects of the hedgerows are likely to
be more pronounced with age. Also, long-term studies would
more clearly reflect the high variability in seasonal
rainfall. From six-year studies in the humid lowlands of
Nigeria, Lai (1991) observed that total maize yields from
alley cropping of leucaena and Gliricidia sepium were about
10% lower than those of the control. In the present study,
total maize yield after 6 seasons under alley cropping with
leucaena was reduced by 12% while that under cassia was
increased by 8%. Therefore, under management conditions
similar to the ones of the present study, it does appear
that cassia has more potential to sustain maize grain yield
than leucaena.
Because of the low measured concentrations of
extractable N and P in the soils of the study station (Kibe
et al., 1981), fertilizer N and P was applied to the crop
each season at the minimum of the recommended rates for
farms in the area. However, the general lack of significant
differences in maize yields between treatments suggests that
N and P were not limiting, again supported by the results of
the pot study.

70
Root Density
The significantly higher yield of leucaena mulch and
presence of significantly higher density of its fine roots
in the top soil could be the causes of the lower maize yield
under alley cropping with leucaena compared with cassia. In
a competitive situation, the sharing of limited soil water
and nutrient resources by competing plants is proportional
to their effective rooting length/density (Bowen, 1985).
Generally, most of the fine roots of trees are found in the
upper 50 cm of the soil (Fayle, 1975; Deans and Ford, 1983).
High root density is an ecological adaptation of plants,
particularly those growing on soils with chemical and/or
physical limitation in the lower horizons (Bowen, 1973).
This is also true of plants in nutrient and water stressed
environments (Chapin, 1980). The studies of several workers
in both the humid lowland and the semiarid tropics (Kang et
al., 1981; Corlett, 1989; Singh et al., 1989; Rao et al.,
1991, 1992; Ruhigwa et al., 1992; Toky et al., 1992)
corroborate the high root densities of leucaena and cassia
found in the present study.
Unfortunately, most of the roots of annual crops and,
in particular those of maize, are found in the same soil
horizon as those of trees (Jonsson et al., 1988).
Therefore, competition for limited resources (e.g., soil
water and nutrients) between trees with well established
root systems and the seasonal crop with less

71
well-established roots is unavoidable. More often than not,
the agronomic crop is the loser in the competitive situation
as indicated, in this case, by the low maize yields under
leucaena. Studies of other workers from the semiarid
tropics also show low maize yields under leucaena alley
cropping (Nair, 1987; Singh et al., 1989; Corlett et al.,
1989; Rao et al., 1990, 1991; Ong et al., 1992). The
present study also shows that all species are not as
competitive as leucaena with crops, and that an
understanding of the rooting system of species to be used
for intercropping is necessary.
Typically, fine root density of trees declines with
distance from the stem (Roberts, 1976; Ford and Deans, 1977;
Persson, 1980; Dhyani et al., 1990) which is consistent with
the observations of the present study. Along the entire 3 m
long trenches from the hedgerows studied, root density of
leucaena was significantly higher than that of cassia. The
extended high root-density of alley cropped leucaena into
the cropped s^rea and the resultant low yields of intercrops
would suggest that potentials of intercropping leucaena
without physical separation from crops are limited.
Approaches to minimize below-ground competition between
hedgerows and crops, at least experimentally, include root
pruning and/or placement of physical barriers such as
galvanized iron or polythene sheets (Singh et al., 1989; Rao

72
et al., Ong et al., 1992), trenching (Verinumbe, 1985) and
root pruning (Fernandez, 1990).
With root pruning/barrier techniques, improvements in
crop yield compared to a control have been noted, with
significantly increased yields at some times (e.g., Corlett,
1989) and not at others (Ong et al., 1992). Root pruning
is, however, labor demanding and unless there are
significant gains to be foregone by not having the hedgerows
in situ, farmers would be most unlikely to adopt it.
Besides, roots of the hedgerows can arch back to the surface
soil even when physical barriers are placed to depths of one
meter below the soil surface and hence render barrier or
trenching efforts ineffective (Ong et al., 1992).
Separate planting of hedges in blocks or sole stands
on land outside the crops that is not currently cropped, or
on land not suitable for cropping or on the periphery of the
farm may be options to minimize competition of leucaena with
crops. This could be an option particularly for leucaena
which has a desirable product (fodder) but which cannot be
intercropped because it is competitive with crops. However,
expected losses in crop yield are proportional to the
percentage of land occupied by the hedges. This lost land
occurred because the amounts of mulch obtained from the
hedges were low (on average 1-2 t ha'1 season*1) and had no
compensatory effects on crop yield. However, the results of
the pot studies suggest that if the levels of mulch from

73
hedgerows could be increased to application rates of 3-4 t
ha'1 season'1, gains in crop yields could be increased even
from the relatively fertile site of the present study.
Conclusions
Alley cropping of Leucaena leucocephala was
detrimental to maize yield over the six cropping seasons.
On average, maize yield under leucaena alley cropping
dropped by 12% of the sole crop while that of cassia
increased by 8%. Leucaena, on the other hand, produced
significantly higher mulch biomass (50% more) than cassia.
The higher mulch biomass production and higher density of
fine roots in the top 40 cm soil are factors that contribute
to leucaena's significantly higher competitiveness with
maize than cassia's.
The in situ benefits of growing the hedges on maize
yield were positive for both species but significantly more
under cassia compared to those under leucaena. Planting the
hedges outside the crops may be a management approach to
minimize competition between the crop and the tree
hedgerows, especially for a species like leucaena with
desirable fodder value. However, this approach produced
yield reductions higher than alley cropping with both
leucaena and cassia, with the loss nearly equal to the
percentage of land apportioned to the hedges.

74
The magnitude of the mulch produced by leucaena and
cassia, 1 and 2 t ha*1 season*1, respectively, was not
sufficient enough to cause significant gains in maize yield
at the study site. There are indications that gains in
maize yield could be significant if the amounts of mulch
applied could be increased to 4 t ha*1 season and beyond.
It is unlikely, however, that such increases in mulch yield
from the hedges could be made without compromising further
crop yields.
At this point, the results of the present study do not
indicate sufficient short-term benefits to recommend alley
cropping with leucaena under conditions similar to the ones
of the present study. However, cassia does have potential
if (a) the 8% increase in maize yield outweighs the costs of
production and, (b) mortality of cassia after 3-4 years is
low enough not to necessitate replanting of the hedgerows.
Follow-p studies on factors responsible for cassia's
positive in situ effects may enhance its potentials for
alley cropping under semiarid conditions.

CHAPTER 3
SOIL FERTILITY ASPECTS OF ALLEY CROPPING
LEUCAENA LEUCOCEPHALA AND CASSIA SIAMEA IN
SEMIARID TROPICS AT MACHAROS, KENYA
Introduction
Agroforestry practices are suggested as low-input
systems with potentials to meet the critical needs to improve
or sustain food production in the tropics (Swift and Sanchez,
1984). One practice of agroforestry which has been the
subject of much interest and study is alley cropping, also
called hedgerow intercropping (Kang et al., 1990). It
consists of growing food crops in alleys between planted
hedgerows of multipurpose trees and shrubs. The hedgerows are
periodically cut back at the beginning of as well as during
cropping to prevent shading and to provide mulch to the
associated crop. ^
One of the most important premises of alley cropping is
that the addition of organic mulch, especially if it is rich
in nutrients, has favorable effects on soil physical and
chemical properties, and hence on productivity of crops.
However, the ability of the mulch to provide nutrients and to
maintain or improve soil properties would depend, in addition
to its quality and time of application, on the quantity. For
instance, Kang et al. (1990) observed from six-year-long
studies at the humid lowland of West Africa that with
continuous addition of leucaena pruning (8-10 t ha'1 yr*1, dry
75

76
matter), soil organic matter (SOM) and nutrients were
maintained at higher levels than where the prunings were
removed. Similar observations are made by other workers also
from the lowland humid tropics (Atta-Krah et al. 1985; Yamoah
et al. 1986c). However, in the SATs, ecological conditions do
not allow the high rate of mulch production found in the humid
tropics. Hence, significant gains in SOM from the addition of
mulch may be limited. Based on the effects of mulch per se
there might be no conceivable advantage of alley cropping
systems on SOM in the SATs. However, turnover rates of high
root mass often associated with agroecosystems with trees
(Ewel et al., 1982), such as alley cropping, may contribute
significantly to organic matter and nutrient contents of the
soil. Moreover, where leaching losses of soil- and applied-
nutrients may be a problem, the roots could minimize these
through a combination of processes including sorption and
uptake from deeper layers (Stark and Jordan, 1978; Jaiyebo and
Moore, 1964). These processes may result in the concentration
of nutrients in the mulch of the hedgerows which could, in
turn, improve crop nutrition (Fernandez, 1990). In some
soils, there are indications that the roots may even recycle
nutrients from the subsoil which could in turn result in top
soil improvement of pH and exchangeable bases (Lai, 1989). In
addition, due to their longevity and to their protective
benefits, intercropped hedgerows may minimize runoff and soil
erosion loss (Lai, 1989; Rao et al., 1991).

77
Notwithstanding the positive attributes and potentials
of alley cropping discussed above, competition for soil
resources (nutrients and water) between the hedgerows and the
crops has to be considered since this could result in
significant losses of crop yield under alley cropping (Rao et
al., 1991; Nair, 1987; Singh et al., 1989). The losses in
yield may, indeed, obscure any gains in soil fertility due to
the presence of the hedgerows. To justify mixing of multi
purpose tree hedgerows with crops, particularly by resource-
poor farmers, it is important, therefore, to understand the
benefits to soil fertility that would otherwise be lost if the
hedgerows were planted separate from the crops. This
knowledge is of particular significance in the semiarid
tropics where soil moisture is a major factor limiting crop
production and where soil fertility improvement aspects of
alley cropping have not yet been well understood.
This study was undertaken in this scenario with the
following objectives: (a) to evaluate the changes in some
selected soil chemical and physical properties under alley
cropping of maize with Leucaena leucocephala and Cassia siamea
over three years and, (b) to separate the effects of mulch per
se from the in situ presence of the hedgerow on the selected
soil chemical and physical properties.

78
Materials and Methods
Site Description
The location, rainfall and soil characteristics of the
site are described in Chapter 2. Table 3-1 shows the chemical
and physical characteristics of the top soil (0-20 cm) at the
beginning of the study (October, 1987).
Treatments and Experimental Layout
The experimental treatments consisted of two systems of
Leucaena leucocephala and Cassia siameai (a) intercropped
hedgerows and (b) planted as blocks/sole stands outside the
crop. In the intercropped system, the hedgerows were 6.67 m
wide. The hedgerows in both the intercropped and the
block/sole planting systems occupied 15% of the total area,
and the crop occupied the remainder of the land (85%). Plots
with fertilizer-only and mulch of leucaena or cassia only were
included as controls. As an absolute control, no-mulch or
fertilizer plots from an adjacent experiment was used. The
experimental design was randomized block design with three
replications. The field experiment was conducted for more
than 3 years (October 1987 to March 1992).
Crop and Tree Management
In all systems, the soil was hand tilled and maize
(Katumani Composite var.) planted twice a year at a stand of

Table 3-1. Chemical and physical characteristics of the top
soil (0-20 cm) at the beginning of the study,
October 1987.
Soil Property
Leucaena
plots
Treatments
Cassia
plots
Control
plots
Mean
pH (1:2.5 soil:H20)
6.2
6.2
6.2
6.2
Org. C (%)
0.9
1.0
0.8
0.9
Total N (%)
0.1
0.1
0.1
0.1
P (ppm, Mehlich-1)
12.7
12.0
12.7
12.5
CEC (meq/100 g soil)
9.6
9.5
9.0
9.4
Clay (%)
28.7
30.7
26.7
28.7
Bulk density (g/cm3)
1.3
1.3
1.3
1.3
At the time of soil sampling in 1987, the experimental
treatments had not yet been introduced.

80
37,000 plants/ha. Moderate levels of fertilizer (40 and 18
kg of N and P/ha) were added to the maize in the hedgerow
intercropped and the block planted systems each season.
Weeds were hand removed at least twice in each season. At
harvest, maize grain, stalks and cobs were removed from the
field as was the practice of the farmers in the area.
The tree hedgerows were pruned back to 50 cm above
ground with 2-3 prunings a season. The prunings obtained
were chopped into small pieces and applied as mulch to the
plots. The chemical characteristics of the mulch are shown
in Table 3-2. If tillage and/or weeding schedules shortly
preceded periods of mulch application, the mulch generally
became incorporated into the soil; otherwise, it remained on
the soil surface.
Soil sampling was always carried to 20 cm depths. A
composite sample of 10 auger cores was taken for each
treatment. The soil was air-dried, crushed to pass a 40-
mesh sieve, and mixed. Samples were analyzed at the
Analytical Laboratory of the Katumani Research Station,
Machakos, Kenya by the following methods: total N by the
micro-Kjeldahl method (Jackson, 1958), organic carbon by the
Walkley-Black procedure (Jackson, 1958), P by the Mehlich-1
double acid procedure and cation exchange capacity (CEC) by
Atomic Absorption Spectrophotometry methods. Soil pH was
determined using a glass electrode pH meter (model Bibby),
at soil to water ratio of 1:2.5.

81
Table 3-2. Chemical characteristics of the dry mulch of
alley cropped hedgerows of Leucaena leucocephala
and Cassia siamea.
N
(%)
P
(%)
K
(%)
Ca
(%)
Mg
(%)
C:N
ratio
Lignin
(%)
Poly
phenol
(%)
Leucaena
3.5
0.2
1.9
0.7
0.3
14
8.1
4.7
Cassia
2.9
0.2
1.8
0.7
0.2
17
8.4
3.7

82
Relative availability of N03" N and NH4+ N was
monitored in the last season. Ion exchange resin bags
(Binkley and Matson, 1983; Binkley and Hart, 1989) were
used. In each plot/replicate, one bag for anions and one
for cations weighing 5 g were placed in the top 10 cm soil
profile. After the harvest of the crop, the resins were
retrieved and N03" N and NH4+N extracted with 2 N KC1.
Levels of NO," N and NH4* N from the KCl-extracts were
determined using Standard Autoanalyzer methods at the Soil
Testing Laboratory of the University of Florida,
Gainesville.
Soil physical properties
Bulk density (Blake, 1965) is the only physical
property of the soil discussed in this chapter. In the last
cropping season, soil was sampled with a semi-wide auger
and 5 cm long to a total depth of 20 cm (i.e., 0-5, 5-10,
10-15 and 15-20 cm). Bulk density of each plot/replicate
was computed as the average values of samples taken at two
positions near the hedgerows (0.45 m) and two in the middle
of the alleys (3.5 m away from the hedgerows).
Statistical Analysis
ANOVA procedures were used to determine statistically
significant (p = 0.05) differences between the levels of

83
soil properties monitored under the various treatments and
over time. Because the initial (1987) soil levels of
percent clay and P differed significantly among plots, these
two parameters were used as covariates in the analysis of
the final (1991) soil P levels. Also, for the KC1
extractable N, analyses were performed on logarithmically
transformed data, because frequencies from a set of 5 resin
bags were not normally distributed. When significant
effects of treatments were detected, means were separated by
pairwise contrasts of the LSM using the GLM procedure in SAS
(SAS, 1992). Differences between treatments were declared
significant at p 0.05.
Results
Soil Chemical Properties
With the exception of P, there were no significant
differences among treatments in the levels of the chemical
characteristics measured (Table 3-3). Soil pH, organic
carbon, total N and CEC in the different treatments were
similar. With respect to P, however, there were significant
differences among treatments. In general, plots with cassia
mulch tended to have a higher level of available soil P. In
particular, the levels of P from the plots with application
of cassia mulch-only were significantly higher than the
levels of all the other treatments (Table 3-4). The

Table 3-3. Chemical and physical characteristics of top soil (0-20 cm) under alley
cropping with hedgerows of Leucaena leucocephala and Cassia siamea hedgerows,
November 1991.
System
PH
(1:2.5
soil:H20
Organic
C
(%)
Total
N (%)
KCl
extract
N (pg/g
soil)*
Mehlich-1
P (ppm)
CEC
(meg/
100 g
soil)
Bulk
density
(g/cm3)
Aliev cropping
Leucaena
6.3a
0.83*
0.09*
9.91*
16.7*
8.42*
1.23*
Cassia
6.2*
0.81*
0.09*
8.15*
23.0b
7.67*
1.46*
Block planting
Leucaena
6.2*
0.94*
0.08
8.55*
18.2b
7.63*
1.37*
Cassia
6.2*
0.83*
0.09*
9.17*
19.5b
9.26*
1.33*
Controls
Mulch
(leucaena)
Mulch
6.2*
0.80*
0.11
8.09*
11.3*
7.21*
1.31*
(cassia)
6.6*
0.93*
0.06*
7.43*
30.0C
9.68*
1.34*
Fertilizer
6.3*
0.72*
0.07
6.14*
21.2b
7.60*
1.20*
Maize-only**
6.3*
0.85*
0.07
6.77*
11.0*
9.33
1.44*
a,b,c,dIn a coiUnm,
values with
different
letters
are signific
antly different (p =
0.05).
Indicates average of 3 sampling dates.
From an adjacent experiment (no fertilizer, no mulch).
oo
t

85
Table 3-4. Changes in soil Mehlich-1 P levels after three
years of alley cropping hedgerows of Leucaena
leucocephala and Cassia siamea hedgerows with
maize, November 1991.
System*
Change in soil
Mehlich-1 P
(ppm)
Aliev intercroDoina
Leucaena leucocephala
4.0a
Cassia siamea
11.0b
Controls
Mulch (Leucaena leucocephala)
-1.3a
Mulch (Cassia siamea)
17.3a
Fertilizer-only
CD

in
cr
Maize-only
-1.7*
*The values of the block planting system were similar.
a,bValues within a column with different letters are
significantly different (p = 0.05).

86
difference between the cassia mulch-only plot and the
similarly treated leucaena plot was, for instance, 62%.
This difference was observed even when the initial (1987)
soil P and percent clay content were used as covariates for
the final (1991) P measurements.
The higher P in plots with cassia mulch could be
attributed to increased concentrations of P observed in
cassia mulch over the years (Table 3-5). Cassia mulch had
24% higher P content than leucaena in 1991. In contrast,
plots with leucaena mulch had significantly higher KCl
extractable N03' N and NH4+ N than the plots with cassia
mulch or the absolute control (Table 3-6). The levels of
relative indices of N03" N and NH4+ N in the plots with
leucaena mulch were, however, not significantly different
from those of plots with fertilizer.
Soil Bulk Densities
No significant difference was observed in soil bulk
density of the different treatments (Table 3-2). Also, no
significant change was observed in the bulk density of any
treatment over the three years of study.
Discussion
Soil Chemical Properties
Despite the differences in the rates and, to some
extent, in the quality of leucaena and cassia mulch

87
Table 3-5. Concentration of phosphorus in the mulch (leaves
plus twigs) of alley cropped hedgerows of
Leucaena leucocephala and Cassia siamea at the
start (1987) and end (1991) of the study.
Species
1987
1991
Leucaena leucocephala
o
CM

o
0.19
Cassias siamea
0.20
0.25
Values in a column with
significantly different
the same letters are not
(p = 0.05).

88
Table 3-6. Relative indices of KCl-extract nitrogen from
ion-exchange resins incorporated into the soil
of alley cropped
leucocephala and
the "short-rains
hedgerows of Leucaena
Cassia siamea with maize
" cropping season of 1991
during

KC1 extract-N (;jg/g
resin)
System
nh4+ n
no3- n
Total
Aliev croDDina
Leucaena leucocephala
29.5a
24.8b
54.3b
Cassia siamea
15.2C
15.3C
30.5d
Controls
Fertilizer
24.3b
36.9a
61.2a
Mulch (leucaena)
22.4b
39.0a
61.4a
Mulch (cassia)
18.2C
25.7b
43.9a
Maize-only
21.4b
20.8b,c
42.2a
a'b'c'dValUes within a column with different
significantly different (p = 0.05).
letters are

89
applied, no significant treatment-related differences were
found in soil organic C and N. Changes in organic C and N
after three years of cropping were also not significant.
Insufficient quantities of mulch with high rates of
decomposition could be an explanation for the lack of
significant differences. The former explanation appears
more plausible than the latter as Kang et al., (1985) also
reported significant increases in SOM as well as
exchangeable K, Ca and Mg levels in both the top- and sub
soil from the application of high rates of leucaena mulch
(7 t ha'1 yr'1). The mulch decomposition rates in the study
of Kang et al. and the present one were similar. Also, pot
studies involving additions of high rates of leucaena and
cassia mulch (i.e., 4 and 8 t ha'1 season'1) to soil from
plots of the present study resulted in significant increases
in maize biomass yield. This was probably due to the
effects of increased soil organic matter and supply of
nutrients from the mulch. In the pot studies, rates of
equivalent mulch application lower than 4 t ha'1 yr'1 had no
significant effects on maize biomass yield. Other workers
have also reported increases in soil organic carbon content
from the additions of high rates of leucaena and cassia
mulch (Yamoah et al., 1986c). However, as studies by Palm
(1988) suggest, differences between sites, soils and mulch
species could significantly impact the effects of high rates
of mulch application on soil organic C.

90
Generally, though, high rates of mulch application (in
the range of 10-20 t ha'1 yr'1 dry matter) are effective in
increasing soil organic C, as concluded by Huxley (1986)
from a comprehensive review of the literature. The
quantities of mulch obtained from the hedgerows in the
present study were only about 2 and 4 t ha'1 yr'1 from cassia
and leucaena, respectively (Chapter 2). Hence, the
limitation to increasing soil organic C and N in the soil of
the field plots of the present study was most probably the
limited amounts of mulch applied. Under such conditions, it
may take more time to detect changes in soil organic C and
total N because the additions and possible changes are small
relative to the soil store (Powlson et al., 1987).
Although productivity may decline due to insufficient
availability of nitrogen (Ayanaba, 1982), slowly decomposing
mulches, high in lignin, polyphenol content or high C/N
ratios, typically, have more potential to increase soil
organic carbon than fast decomposing ones (DeHaan, 1977;
Sivapalan, 1982). The characteristics of the mulch of the
two species were similar (Table 3-1); however, their rates
of decomposition on the soil surface differed slightly, with
cassia decomposing more slowly than leucaena (Chapter 4).
Thus, the lack of significant differences observed in soil
organic C between soils under cassia and leucaena mulch
could be due to a lack of differences in the mulch quality
of the two species. The mulch of both species may have been
of relatively high quality, despite the slight differences

91
noted in their rates of decomposition (particularly on the
soil surface), and no long-term differences in soil organic
C may be expected.
Although there may in fact be (unmeasured) differences
in the quality of the mulch of the two species, the lack of
significant differences observed in organic C of soil under
the mulch of the two species suggests that these differences
were not as important as the quantity of mulch in impacting
on soil organic C. The effects of residues (mainly those of
crops) on soil organic C content are highly related to the
amount and only slightly to the type of residue applied
(Rasmussen and Collins, 1991; Sowden, 1968). Both the
quantity and the quality factors of the mulch are critical
because of the high loss of the C applied as C02.
Typically, laboratory incubation studies estimate that as
much as 60-75% of the C from crop residues could be evolved
as C02 after one year in the soil (Martin and Haider, 1980).
The effects of hedgerow intercropped systems on soil
organic C, if any, appeared to be due more to the effects of
the mulch and less to the in situ presence of the hedgerows.
The effects of only the small quantities of mulch added
(2-4 t ha'1 yr'1) appeared to have been sufficient to sustain
soil organic C after three years of cropping. Evidence for
this is provided by the lack of significant difference
observed in the levels of soil organic C between the

92
hedgerow intercropped and the block planted system of both
species. The levels of mulch produced by the hedgerows in the
present study were within the range of plant residues
requirements estimated by Young (1989) to maintain top soil
organic matter in semiarid and subhumid ecozones of the
tropics.
The contribution of the root system of alley cropped
hedgerows to soil organic matter, and thus to soil
fertility, has been little studied. In addition to the
contribution of above ground litter, turnover of high root
biomass, which is generally higher in agroecosystems
involving trees than in other land use systems (Ewel et al.,
1982), is a mechanism by which trees are hypothesized to
improve soil fertility. It is also hypothesized that,
through their deep roots, trees such as leucaena and cassia
are able to absorb nutrients from soil depths that crops
cannot reach, and thus, improve soil fertility through
litter fall and root turnover (Nair, 1984). While the roots
of the intercropped hedgerows may contribute to organic C
levels of the soil over time, this was not observed in the
short-duration of the present study (3-4 years). It is
implicit from the above that, especially for a species such
as leucaena which is highly competitive, the presence of the
hedgerows in the crops may not be necessary for purposes of
maintaining soil organic matter/soil fertility.
Although there were no significant differences in soil
organic C and total-N between treatments, the results of the
KCl extractions indicated that plots with leucaena (whether

93
intercropped or received mulch only) had relatively higher
available N than those with cassia (Table 3-6). This could
be attributed to the higher yield of leucaena mulch which
also had higher concentration of N and higher rate of
decomposition (particularly the leaf fraction) than cassia
mulch. It is also possible that apart from the release of
its own N upon decomposition, some of the higher effects of
the mulch of leucaena on the relative index of available N
was to stimulate more mineralization of the native soil N.
This could be through the so-called "priming" effect of
added N on native soil N (Hauck and Bremner, 1976; Jenkinson
et al., 1985) .
It does appear that, besides N, the small quantity of
mulch applied (particularly that of cassia) had significant
effects on the availability of top soil P. This was most
pronounced where only mulch of cassia was applied. Similar
observations have been made by Yamoah et al. (1986c) from
the humid lowland tropics of Nigeria. The increase in soil
P under cassia plots could be ascribed to enhanced P
contribution from the mulch. Over the three years, the
concentration of P in cassia increased considerably while
that of leucaena remained unchanged. Explanation for the
higher concentration of P in cassia mulch compared to that
of leucaena is a matter of speculation; some hypothesis are
differences in root to shoot ratios, mycorrhizal
association, microbial population and activity (e.g., P-
solubilizing bacteria).

94
Mwangi (1989), working at a site adjacent to the present
study, also observed that at rates of mulch application 3-4
times higher than the amounts applied in the present study, P
levels of the top soil nearly doubled after 6 years of
cropping. However, there are biological limitations to the
production of such high mulch yields in SATs. In the absence
of high mulch yields, application of the small amounts of
mulch available may still result in short-term or immediate
benefits (e.g., enhanced availability of top soil N and P).
It may, however, have little effect on the maintenance of soil
organic matter (Stevenson, 1986).
The initial levels of soil organic C and total N were
in the low end of the range typical of Alfisols (Brady,
1990). However, it is worth noting that no significant
changes in either soil organic matter or N were observed
after three years of cropping with an annual removal of
about 60 kg N ha*1 through grain and maize stover. The lack
of significant change in soil organic C and total N was even
observell from the control plots with no additions of N or
mulch. Maize yield from plots with various treatments were
also similar, with differences being more pronounced between
seasons than between treatments. These supporting
observations suggests that it will take more than three
years to detect changes in organic C or that the effects of
the small amounts of mulch added to the soil may have been
sufficient to prevent declines in the initial soil organic
C.

95
Soil Physical Properties
The lack of significant change in soil bulk density
could, in turn, be ascribed to lack of change in soil
organic matter. Like many other soil properties, soil bulk
density is influenced by changes in soil organic matter
(Allison, 1973). Several workers have noted significant
improvements in soil bulk density in the humid lowlands of
Nigeria where mulch yields as high as 8-10 t ha"1 yr"1 (dry
matter) are obtained from intercropped hedgerows,
particularly leucaena in studies spanning four years (Wilson
and Kang, 1982; Yamoah et al., 1986c).
Conclusions
No significant changes were observed in soil organic
C, KCl-extractable N, CEC and bulk density after three
years. This was so with or without alley cropping of
Leucaena leucocephala and Cassia siamea. The low amounts of
mulch obtained from the hedgerows (coupled with the
relatively high rate of decomposition of the mulch) appeared
to be the main limitation to significant improvements in the
levels of the soil chemical and physical properties
monitored. Although the changes in organic C and total-N
over time were not significant, the small amounts of mulch
(particularly those of leucaena) resulted in significantly
higher relative index of N03" N and NH4+ N availability
compared to the plots with no additions of mulch.

96
With respect to soil P, significant increases were
observed in plots with cassia compared to those with
leucaena at the end of the three years period. Also, plots
with cassia had higher levels of available soil P compared
to the other treatments in the final year of the study. The
higher P level under plots with cassia was attributed to the
considerable increase in the concentration of P in the mulch
of cassia over time. Explanation for the higher
concentration of P in cassia mulch compared to that of than
leucaena is a matter of speculation; some hypothesis are
differences in root to shoot ratios, mycorrhizal
association, microbial population and activity (e.g., P-
solubilizing bacteria).
The limited amount of mulch available from the
hedgerows of leucaena and cassia may have short-term or
immediate benefits on soil fertility, as is evident from the
increases in soil available N and P, but may not have
significant impacts on soil organic carbon. The contribution
of the in situ presence of the hedgerows was not
significant, during the study. Therefore, there may not be
any special advantages, in terms of soil fertility
improvement, from growing the hedgerows within the crop
field.

CHAPTER 4
DYNAMICS OF SOIL FERTILITY CHANGES: RATES OF
DECOMPOSITION, N-MINERALIZATION AND UPTAKE BY
MAIZE FROM MULCH OF LEUCAENA LEUCOCEPHALA
AND CASSIA SIAMEA UNDER SEMIARID CONDITIONS
AT MACHAROS, KENYA
Introduction
The use of leguminous plants as a source of nitrogen
for crops is particularly important in many parts of the
tropics where fertilizer use is often economically not
feasible. Agroforestry practices, and in particular alley
cropping or hedgerow intercropping is a technology that is
often mentioned as having the potential to improve soil
fertility and productivity in the tropics (Kang et al.,
1990) One of the basic principles of the management of
alley cropping is the periodical pruning of the hedgerows to
reduce their shading and competition with the food crops
(Kang et al., 1985). The prunings left as mulch on the
surface between the hedgerows are claimed to improve the
physical, chemical and biological properties of the soil
(Kang et al., 1984). The contribution of the plant residues
to the fertility of the soil will largely depend on the
amount of biomass so applied and on its rates of
decomposition.
The decomposition of plant residues is known to be
primarily affected by the N content of the biomass (Campell,
97

98
1983) and the organic components, such as lignin (Alexander,
1977, Herman et al., 1977), polyphenols (Vallis and Jones,
1973; Palm and Sanchez, 1991) and soluble carbohydrates
(Cheshire et al., 1974; Wieder and Lang, 1982), and by
environmental and management factors (Wilson et al., 1986).
The rates at which mulches decompose have important
implications on soil organic matter and N dynamics. Slowly
decomposing materials may increase soil organic matter, at
least in temperate conditions (Dehaan, 1976; Silvapalan,
1982; Kelly and Stevenson, 1987) but crop yields may decline
due to insufficient availability of N. On the other hand,
fast decomposing materials may provide a short-term increase
in N but may have little effect on the maintenance of soil
organic matter content (Vine, 1953; Stevenson, 1986). Even
from fast decomposing mulches, the efficiency of nutrients-
use, in particular of N, by crops is low compared to
inorganic fertilizers (Fox et al., 1990; Gutteridge, 1992).
Most field studies of decomposition of hedgerow
prunings or mulches have been made in the humid tropics
(Yamoah et al., 1986a; Swift et al. 1981; Palm and Sanchez,
1991). Similar studies in the semiarid tropics are scarce.
Although some field studies on decomposition of hedgerow
prunings have been conducted in the semiarid tropics
(Mugendi, 1990), they have not considered characterization
of plant and environmental factors that regulate the
process. There could be similarities in the patterns and
rates of decomposition of prunings in the humid and semiarid

99
tropics, but equally, there could be differences caused by
sheer differences in climate. Moreover, differences in the
quality of the mulch (e.g., ratio of leaves to twigs as well
as nutrients, polyphenols and lignin content) and management
practices (e.g., soil incorporation or surface placement)
could cause differences in the patterns and rates of
decomposition between sites. An understanding of the effect
of these factors on mulch decomposition could provide
guidelines for increasing the efficiency of N-use by crop
from mulches and to sustain soil fertility in the semiarid
tropics.
The objectives of this study were three-fold:
1. To evaluate the effects of soil placement methods
(i.e., incorporation vs surface placement) on the
rates of decomposition of Leucaena leucocephala
and Cassia siamea mulch under field conditions.
2. To determine the impacts of the mulch quality
(i.e., nutrient content, percent lignin and total
soluble polyphenols) on the rates of decomposition
and N mineralization.
To evaluate the potentials for synchronizing mulch
N-mineralization and uptake by maize.
3.

10D
Materials and Methods
Site Description
The study was conducted for two cropping seasons of
1991 at the Field Station of ICRAF located at Machakos,
Kenya (latitude Io 33' S and longitude of 37 14' E and 1596
m elevation; mean annual temperature, 19C). In the first
season, the total rainfall was 214 mm while that of the
second was 373 mm (Fig. 4-1). These monthly totals,
however, do not reveal the considerable variations that
occur within and between seasons. Monthly rainfall equals
or exceeds evapotranspiration only in April, May and
November. The general characteristics of the soils of the
study site, and in particular, that of the top soil (0-20
cm) are described in Chapter 3.
Methods
Field and laboratory studies
Field studies were carried out for both seasons of
1991 in alley cropping plots of leucaena and cassia. Maize
(Katumani Composite cv.) fertilized at the rate of 40 and 18
kg/ha of N and P, respectively, was planted in the alleys
between the hedgerows in both seasons.

Rainfall (mm)
101
Month
Rainfall
Temperature (C)
Figure 4-1. Monthly rainfall and mean temperature
during the study, 1991.
Temperature (C)

102
Treatments
In the first season, decomposition patterns of soil
incorporated mulch (leaves plus twigs) of leucaena and
cassia were evaluated. A randomized block design with three
replications was used. On March 23, 1991 (which was the
beginning of the first season), fresh mulch samples of
leucaena and cassia weighing 50 g (20 g dry weight) were put
into litter bags (33 x 33 cm with 5 x 5 m mesh size) and
incorporated into the soil at a depth of 10 cm. The twigs
had diameters ranging from 5-10 mm. At that time, no
rainfall had yet occurred. The mulch was obtained from the
hedgerows at the end of the dry season. Senescence of the
leaves and/or low production of new foliage reduced the
proportion of leaves to twigs to 1:1 (w/w). During the
rainy season, however, the leaf ratio of the mulch would be
expected to be more.
In the second season, mulch of similar composition as
in the first season was soil incorporated on October 24,
1991. Some additional treatments were included in this
season. These were
1. Litter bags of mulch (leaves plus twigs) placed on
the soil surface and soil incorporated;
Soil incorporated and surface placed twigs (i.e.,
excluding the leaves) in litter bags;
2.

103
3. Bundles of twigs tied with a nylon string (no
litter bags) incorporated into the soil and placed
on the surface, and
4 Soil incorporated filter paper in litter bags.
The split-plot design had the species as the main-plots and
the placement methods of the mulch (i.e., soil incorporated
or surface applied) as the subplot treatments.
The purpose of the nylon tied twigs treatment was to
determine to what extent the litter bags influenced the
rates of decomposition. Whatman No. 2 filter paper, chopped
into small pieces, which is pure cellulose with practically
no N, was used to determine the importance of C:N ratio of
the mulch relative to the organic compounds (e.g., lignin
and polyphenols) in influencing rates of decomposition.
In both seasons, mulch similar in composition to the
soil incorporated or surface-applied was analyzed in the
laboratory to determine dry matter and nutrient contents.
The litter bags and nylon-tied twigs were retrieved from the
field every two weeks for a total period of 18 weeks.
Sufficient numbers of bags were used to allow the removal of
a bag after every two weeks. Once retrieved, the bags were
wrapped in a plastic sheet to minimize losses during
transportation from the field to the laboratory. In the
laboratory, soil and debris adhering to the mulch were
removed carefully with a toothbrush and a soft paint brush.

104
The mulch was then sorted into leaves and twigs. Dry weight
of each component was recorded after oven drying at 65C for
48 hours.
Ash-free dry weight of the mulch retrieved from the
soil was obtained following combustion of the mulch to ashes
in a muffle furnace at 550C for 3 hours. Fresh mulch was
also ashed to determine initial ash content. The following
formula after Cochran (1991) was used to calculate ash-free
dry weight and percent N: Corrected dry weight = Dry weight
-increase in total residual ash over the ash content of the
fresh mulch.
To determine the contents of nutrients, the mulch was
oven-dried at 65C for 24 h before weighing. It was then
ground to pass through a 0.2 mm screen, thoroughly mixed,
and sub-sampled for total-N analysis (three replicates),
according to micro-Kjedahl acid digestion procedure
(Jackson, 1958) modified for plant materials. Plant
materials of known N-concentrations from the archives of the
National Dryland Research Station Laboratory, Machakos,
Kenya, were included in the analyses to check the accuracy
of the analytical procedures. In the first season,
determination of N concentration was done only at the start
and end of the study. In the second, however, N
concentration of the mulch was determined at each time it
was retrieved from the soil. Carbon (C) content of the
mulch remaining undecomposed after every two weeks was not

105
determined but assumed to be 50% of the ash-free dry weight
of the mulch. From the contents of C and N in the mulch,
the ratio of C:N was determined.
Acid digestible lignin and ash concentrations were
determined by the Van Soest method (Van Soest, 1968). Total
soluble polyphenols were determined at the Natural Resources
Institute, United Kingdom, using a revised Folin-Denis
method (King and Heath, 1967). This involved extraction of
the leaves, twigs and combined leaves plus twigs in 50%
aqueous methanol for two hours in water bath (80C). The
materials had previously been air dried for two days and
stored under room temperature before analysis. Extraction
and determination of polyphenol were carried out in
duplicates for each sample, using tannic acid supplied by
Sigma Chemicals as a standard. Total soluble polyphenols
are expressed as percent of tannic acid equivalents. The
chemical and physical characteristics of the mulch used are
shown in Table 4-1.
Pot studies
Pot studies were initiated in January 1992, to
determine interactive effects of mulch quality of the two
species and the seasonal rainfall on the patterns of N-
mineralization and uptake by maize. It consisted of the
application of a factorial combination of four mulch rates
of leucaena and cassia (the equivalents of 0, 2, 4, 8 t

Table 4-1.
Chemical
Leucaena
and physical properties
leucocephala and Cassia
of the
si antea
mulch of alley
cropped hedgerows of
Leucaena
Cassia
Property
Leaves
Twigs
Average
Leaves
Twigs
Average
Nitrogen (%)
4.4
2.6
3.5
4.0
1.7
2.9
Polyphenols
(%)
6.9
2.4
4.7
5.5
1.9
3.7
Lignin (%)
5.4
10.8
8.1
6.8
9.9
8.4
Ash (%)
-
-
7.1
-
-
5.8
C:N ratio
11
19
14
13
29
17
NOTE: It was assumed that 50% of the mulch material was carbon.

107
ha"1) and two levels of irrigation (i.e., the equivalents of
200 and 400 mm of rainfall received in a cropping season of
16 weeks) to potted maize. Each factorial combination of
species, mulch rate and irrigation levels was replicated
thrice in a completely randomized block design. Pots with
one level of fertilizer, the equivalents of 40 kg N ha"1 in
the form of mono-ammonium phosphate and pots with no inputs
(i.e., mulch or fertilizer) were also included as controls.
In this chapter, only the effects of one of the mulch
rates (i.e., 2 t ha"1, averaged across the two irrigation
levels) is discussed to demonstrate the patterns of mulch-N
mineralization and uptake. The 2 t ha'1 mulch rate was
equivalent to the highest amount of mulch obtained from the
hedgerows in a season and applied to the fields (Chapter 2).
Polythene cylinders (30 cm long and 30 cm wide) closed
at the bottom were used as pots. The pots were filled with
the top 20 cm soil collected and composited from positions
near (0.45 m) and distant (3.5 m) from intercropped hedgerow
plots of leucaena and cassia. Soil in each pot weighed 20
kg when dried to 8% moisture content (w/w) with a bulk
density of 1.53 g cm"3.
To obtain an index of N availability from the mulch,
5 gm of ion-exchange resin (Binkley and Matson, 1983;
Binkley and Hart, 1989) sealed in bags made of plastic
stocking were placed in each pot about 5 cm above the bottom

108
of the pot. Separate resin bags were used for anions and
cations.
Three pregerminated seeds of maize were sown to each
pot and thinned to one plant/pot after germination. Four
weeks after sowing, total harvest of both shoots and roots
was made. The pots were then sown for another four
consecutive weeks. At each harvest, the roots were
separated from the soil manually by rinsing and floatation
procedures. The plant materials were then oven dried for 24
h at 65C, weighed, and ground to pass through a 0.2 mm
opening sieve, and analyzed for N concentration.
After the fourth harvest, the resin bags in each pot
were retrieved from the pots and extracted with 100 ml of 2
molar KCl. Ammonium-N and nitrate-N were determined by
standard Autoanalyzer methods at the Soil Testing Laboratory
of the University of Florida, Gainesville.
Statistical analyses
The NLIN procedure for multiphase regression models of
the SAS (SAS, 1992) was used to determine differences
between the rates and patterns of decomposition and N-
mineralization of the mulches. This procedure models the
response variable, dry matter or N remaining in the mulch,
as a continuous function of time.
The single exponential model Yt = Y0 e-kt, where Y0 is
the original amount of material applied and Yt the

109
proportion of the initial dry matter or nitrogen remaining
after a period of time t (in weeks) was first fitted. A
plot of time against the natural logarithm of this first-
order exponential model (which linearizes the model) was
made for each replicate. The logarithmic plots revealed
that the patterns of dry matter or N loss were biphasic and
a single negative exponential model could not provide the
best fit.
Regression parameters (i.e., the intercept and slope
of each phase) and the spline (meeting) points of the
regression lines of the two phases were then generated for
each replicate of the two species. ANOVA test was then
performed on each of the regression parameters to determine
significance of differences between the regression
parameters of the two species.
For the pot experiment, the GLM procedure of the SAS
(SAS, 1992) was used to test for differences between the
effects of the four treatments (i.e., leucaena mulch, cassia
mulch, fertilizer control and the absolute control) on maize
biomass yield of the pots. The option for repeated measures
ANOVA was included in the GLM procedure to account for the
systematic effects of repeated harvests of biomass from the
same pots over time.
When significant effects of treatments were detected,
means were separated by pairwise contrasts of the LSM

110
procedure. Differences between treatments were declared
significant at p = 0.05.
Results
Field Studies: Patterns of Mulch
Decomposition and N-mineralization
In general, the patterns of decomposition of leucaena
and cassia mulch (i.e., leaves plus twigs), the leaves and
twigs separate, as well as the filter paper were biphasic.
The parameters of the two-phase regression lines fitted
(i.e., the intercept, the slopes, the spline point of the
two phases and the coefficients of determinations) of the
mulch material are presented in Table 4-2. With the
exception of leucaena leaves, all other materials had an
initial phase of rapid decline rate followed by a second
phase of comparatively lower rate. Leucaena leaves had only
one phase of decomposition.
Effects of Placement Method:
Soil Surface or Incorporation
Leaves
Significant interactive effects were observed between
species and soil placement method on the rate of
decomposition of the leaves (Fig. 4-2). No significant
difference was detected between the rates of decomposition

Table 4-2. Intercepts, slopes, spline points and coefficients of determination (r2) of
the regression lines for phase one and phase two of Leucaena and Cassia mulch
decomposition.
Decomposition rate (%)
Spline
Intercept Phase one Phase two point
First season
Mulch (incorporated)
Cassia 2.2
Leucaena 2.0
Second season
Mulch (incorporated)
Cassia 2.3
Leucaena 2.3
Mulch (surface)
Cassia 2.3
Leucaena 2.0
Twigs (incorporated)
Cassia 1.0
Leucaena 1.1
Leaves (surface)
Cassia 3.0
Leucaena 2.9
Leaves (incorporated)
Cassia 2.8
Leucaena 3.0
Twigs (surface)
Cassia 1.5
Leucaena 1.3
Filter paper (incorporated) 1.8
12.1
0.1
9.1
0.82
13.8
*
*
0.84
11.0
2.0
9.3
0.95
12.5
1.9
7.0
0.97
13.3
2.4
7.9
0.97
16.5
3.2
7.3
0.98
12.3
3.0
8.4
0.96
7.5
2.0
7.4
0.85
11.1
2.4
9.2
0.91
15.5
*
*
0.97
13.1
2.3
8.5
0.94
15.5
*
k
0.98
11.5
3.2
7.5
0.91
10.5
4.6
8.4
0.77
10.5
5.2
11.1
0.81
NOTE: Mulch refers to leaves plus twigs.
*
Refers to absence of phase two in the
decomposition pattern of the species.

Figure 4-2. Decomposition patterns of the leaves of Leucaena leucocephala
and Cassia siamea.
Cassia (incorporated): Y = 2.8 0.26X 0.05 (X-8.5); r2 = 0.94.
Leucaena (incorporated): Y = 2.9 0.31X; r2 = 0.98.
Cassia (surface): Y = 3.0 0.11X 0.05 (X 9.2); r2 =
Leucaena (surface): Y = 3.0 0.31X; r2 = 0.97.
0.91.

Carbon (log g) remaining
3.5
>

114
of leucaena leaves on the surface or within the soil.
Decomposition rate of leucaena leaves was significantly
higher than surface-placed cassia leaves. Soil-incorporated
and surface-placed leucaena leaves had physically
disappeared by weeks 7.5 and 6.5, respectively. The rate
of decomposition of leucaena leaves, averaged across
placement methods, was 15.4% per week, with a corresponding
half-life of 4.5 weeks.
For cassia, soil-incorporated leaves had significantly
higher rates of phase one decomposition than surface-placed
leaves. The rates of phase one decomposition of cassia
i
leaves within the soil and on the soil surface were 13.1%
and 11.1% per week, respectively. Half-lives corresponding
to these rates were 5.3 and 6.2 weeks. Phase two of the
soil-incorporated cassia leaves started at week 8.5 while
that of the soil-surface-placed leaves started at week 9.2.
The differences in the start of phase two decomposition of
the soil-incorporated or surface-placed cassia leaves were
not significant. Also, the rates of leaf decomposition in
phase two of the soil-surface-placed and soil-incorporated
leaves were not significantly different. On average, the
leaves decomposed at a rate of 2.3% per week in the second
phase. Half-life corresponding to the average rate of
decomposition in the second phase was 30.1 weeks.

115
Twigs
No significant effects of species or placement methods
on the rate of twig decomposition of both phase one and two
were observed in either season (Fig. 4-3). However, the
ratio of leucaena to cassia twigs remaining over time
provides interesting contrasts between the two species (data
not shown). Leucaena had a higher ratio of twigs remaining
not decomposed than cassia in both seasons and placement
methods. This was most pronounced in the first season when
an average ratio of leucaena to cassia twigs of 1.6:1 was
observed at the end of the study. This would also suggest
that leucaena twigs were more resistant to decomposition
than cassia.
Rates of decomposition of twigs placed outside the
litter bags (Table 4-3) were not significantly different
from those inside the litter bags, regardless of placement
methods.
Mulch (leaves plus twigs)
On the soil surface, the rate of decomposition of
leucaena mulch in the first phase was significantly higher
than that of cassia (Fig. 4-4). Leucaena mulch placed on
the soil surface also had a significantly higher rate of
phase one decomposition than the soil-incorporated mulch in

Figure 4-3. Decomposition patterns of the twigs of Leucaena leucocephala
and Cassia siamea.
Cassia (incorporated): Y = 0.97 0.25X 0.06 (X 6.5); r2 = 0.96.
Leucaena (incorporated): Y = 1.1 0.15X 0.04 (X 7.4)*; r2 = 0.85.
Cassia (surface): Y = 1.1 0.23X 0.06 (X 6.0); r2 = 0.77.
Leucaena (surface): Y 1.3 0.2IX 0.09 (X 7.5); r2 = 0.77.
*Indicates presence of significant difference from the other treatments.

Carbon {log g) remaining
2.0
117

118
Table 4-3. Decomposition parameters of soil incorporated
and surface placed Leucaena leucocephala and
Cassia
second
siamea
season
twigs outside
of 1991.
the litter bags,
Placement
method/species
Inter
cept
Decomposition
rate (%)
Phase Phase
one one
Spline
point
(weeks)
r2
Soil incoroorated
Cassia
2.2
10.1
6.0
9.3
0.73
Leucaena
2.4
13.1
2.5
9.3
0.89
Soil surface
Cassia
2.2
7.8
1.3
11.4
0.65
Leucaena
2.2
4.8
2.4
8.3
0.78

Figure 4-4. Decomposition patterns of the mulch (leaves plus twigs) of
Leucaena leucocephala and Cassia siamea after 16 weeks.
Cassia (incorporated): Y = 2.3 0.22X 0.01 (X-11.3); r2 = 0.96.
Leucaena (incorporated): Y = 2.3 0.23X 0.02 (X 8.0); r2 = 0.95.
Cassia (surface): Y = 2.4 0.25X 0.04 (X 8.5); r2 = 0.97.
Leucaena (surface): Y = 2.2 0.33X 0.09 (X 6.0)*; r2 = 0.98.
Indicates presence of significant difference from the other treatments.

3.0
2.5
0 L-, ¡ r
0 2 4 6
MULCH
Cassia
(incorporated)
e
Leucaena
(incorporated)
B- .
Cassia
(surface)
Leucaena
(surface)
8 10 12 14
Time (weeks)
-e-
16
18
120

121
the second season. Soil-surface-placed leucaena mulch
decomposed at the rate of 17% per week in phase one, with a
half-life of 4.1 weeks. On the other hand, cassia mulch on
the soil surface had a phase one decomposition rate of 12.5%
per week. The half-life corresponding to the rate of
decomposition of the cassia mulch was 5.5 weeks. In the first
season, only soil incorporation was carried out, hence,
comparison of effects of placement methods was not possible.
Within the soil, the rates of phase one or phase two
decomposition of the two species were not significantly
different in the first season or the second season.
Interaction effects between species and placement methods on
the rates of decomposition in the first phase were also not
significant in both seasons. On average, both species
decomposed at the rate of 11.0% and 11.5% per week in phase
one in the first and second season, respectively. The half-
life corresponding to these rates of decomposition were 6.3
and 6.1 weeks in the first and second season, respectively.
In the second phase, the effects of species, placement
methods and season were not significant. Averaged across
species and placement methods, the rates of decomposition in
phase two were 2.7% per week in the first season and 2.1% per
week in the second season. Half-lives corresponding to these
rates were 25.2 weeks and 33.2 weeks in the first and second
seasons, respectively.
As expected, twigs generally decomposed at significantly
lower rates in phase one than leaves. However, the ratio of

122
leaves to twigs remaining over time (data not presented) shows
interesting contrasts that suggest some exceptions to the
generalization. In both seasons, the ratio of leaves to twigs
of cassia mulch remaining over time was consistently higher
than that of leucaena. This was true for both soil-
incorporated and surface-placed mulch. As a result, the rates
of cassia leaf decomposition were often only marginally higher
than those of the twigs. In leucaena, on the other hand, the
ratio of leaves to twigs declined rapidly over time. This is
consistent with the observations of the significantly higher
rate of decomposition of leucaena leaves compared to twigs.
Filter paper decomposition
Soil incorporated filter paper had both phases one and
two rates of decomposition. The rates were similar to those
of the mulch or the twigs incorporated (Table 4-2). Soil
surface placement was not included as a treatment for the
filter papers.
Trends of nitrogen loss from the mulch
The patterns of N loss from the mulch of leucaena and
cassia were similar to those of the dry matter or carbon loss
discussed above. In both species, an initial phase of rapid
rate was observed which declined significantly with time (Fig.
4-5). The rates of N loss of leucaena and cassia mulch in the
first phase of decomposition were 15.9% and 16.9% per week,

Mulch N mineralized (%)
123
Time (weeks)
Figure 4-5. Cumulative percent of N mineralized from
soil incorporated mulch of Leucaena leucocephala and Cassia
siamea after 16 weeks.

124
respectively. The difference between these rates of mulch-N
release of the two species was not significant. In the second
phase, the rates of N release were significantly lower than
those of the first phase, 6.3% and 3.2% for leucaena and
cassia, respectively. These rates were not significantly
different. The spline points of the regression lines of the
two phases occurred at 8.1 weeks for leucaena and 8.7 weeks
for cassia after incorporation of mulch into the soil. The
difference in the spline points of two species was also not
significant. The half-life of N loss for leucaena and cassia
in the first phase of decomposition were at 4.1 and 4.4 weeks,
respectively.
In the first season, the cumulative mulch-N mineralized
after 16 weeks was 73.7% and 55.5% of the total for leucaena
and cassia, respectively. In the second season, N
mineralization of both species was similar, viz, 98% for each
species. At week four of the second season, 57.5% of the
leucaena and 58.4% of the total cassia mulch-N was
mineralized. Net N-immobilization by the mulch was detected
between weeks 8 and 16 for both species (Fig. 4-6) in the
second season. At week eight, the mulch-N concentration of
leucaena and cassia mulch was 1.7 and 1.9%, respectively,
suggesting that these concentrations of N may be close to the
critical levels for initial net mineralization of incorporated
legume (or below these levels, N-immobilization by soil
microbes may occur).

Figure 4-6. Relationship between the patterns of N release from the mulch
of Leucaena leucocephala and Cassia siamea and biomass yield of maize from four
consecutive crops of maize sown in pots and harvested at an interval of four
weeks.

Percent N remaining in mulch
100
0 4 8 12 16
Time (weeks) after application of mulch
NJ
CT\
Biomass yield (g/pot)

127
Pot Studies; Maize Biomass Yield and
Synchrony of Hulch-N Release
Biomass yield, N yield, apparent N recovery
and relative index of cumulative soil-N
In the first harvest, leucaena-mulch treatment gave
significantly higher yield of maize biomass than the cassia
mulch, fertilizer or control treatments (Table 4-4). In the
second and third harvests, differences between the effects
of treatments on maize biomass yield were not significant^.
In the fourth harvest, the effects of leucaena mulch on
maize biomass yield were significantly higher than the other
treatments. Among the harvests, biomass yields of maize
were low in the first and fourth harvests and significantly
higher in the second and third harvests for all the
treatments.
There was no significant difference in N-concentration
of the maize biomass between treatments in any given
harvest. However, among the harvests, significant
differences between treatments on the biomass N-yield (i.e.,
N concentration x biomass yield) were observed. In the
first harvest, N-yield of maize biomass with leucaena mulch
was significantly higher than that of other treatments.
This was true also of the percent N recovery from the mulch

Table 4-4. Effects of soil incorporated Leucaena leucocephala and Cassia siamea mulch
on the biomass yield, N-concentration, N-yield and N recovery of maize
planted in pots, 1991.
Time after
Maize
N concen-
N-yield
Apparent
mulch*/treatment
biomass
tration
of the
N
application
Treatment/
yield
of biomass
biomass
recovery
(weeks)**
mulch
(g/pot)
(%)
(g/pot)
(%)
4
Leucaena
4.2
2.9a
0.12a
11.2a
Cassia
2.2
3.1a
0.07b
-5.3C
Fertilizer
2.9
2.9a
0.09b
3.6b
Control
2.7
2.8a
0.08b
-
8
Leucaena
5.5
3.0a
0.17a
4.6b
Cassia
5.9
2.6b,c
0.15a
0.6
Fertilizer
5.5
3.0a,c
0.17a
7.1a
Control
5.9
2.6b,c
0.15a
-
12
Leucaena
5.4
2.6a
0.14a
9.8a
Cassia
5.0
2.7a
0.14a
11.4a
Fertilizer
5.4
2.3b
0.13a
9.2a
Control
4.6
2.2b
0.10a
-
16
Leucaena
4.0
2.3a
0.13a
8.7a
Cassia
2.1
2.0a
0.12a
-2.8C
Fertilizer
3.2
2.1a
0.06b
2.5b
Control
2.5
2.2a
0.06b

a'b'cWithin a week and a column, values with different letters are not significantly
different (p = 0.05).
*Mulch applied was the equivalent of 2 t/ha, dry weight.
**Maize was sown and harvested after 4 weeks repeatedly for a total duration of 16 weeks.
128

129
of leucaena. Apparent N recovery is expressed as percentage
difference between N-yield of the maize biomass with mulch or
fertilizer application and the N-yield of the control
treatment, and then divided by the amount of N applied through
fertilizer or mulch.
In the first harvest, apparent N recovery from the mulch
of cassia was low and, in fact, significantly negative. This
was because of the low maize biomass yield from the cassia-
mulched pots compared to the control pots. Compared to the
first harvest, maize biomass as well as the N-yield of the
biomass of all treatments (including the control) was
significantly higher in the second harvest, resulting in lower
apparent N recovery of N applied compared to the first
harvest. This was most pronounced with cassia (0.6%). The
fertilizer treatment had significantly the highest recovery
(7.1%) of N applied.
In the third harvest, no significant differences were
observed in biomass and N-yield between treatments. In this
season, apparent N recovery rates from the mulch and the
fertilizer treatments ranged from 9.2-11.4 %.
In the fourth harvest, apparent N recovery by the maize
biomass from the mulch of leucaena was significantly higher
than that with cassia mulch or with fertilizer. Because of
the relatively lower maize biomass yield from pots with cassia
mulch compared to those of the other treatments in this
harvest, apparent N recovery from cassia mulch was negative

130
and significantly lower than that from the mulch of leucaena
and fertilizer.
Relative Indices of N-availabilitv from the
Mulches; N-accumulation bv Ion-exchanae Resins
After 16 weeks, the index of total-N mobilized from
leucaena mulch was significantly higher than that with cassia
(Table 4-5). Soil with leucaena mulch had significantly
higher levels of nitrate-N index than that with cassia.
However, the differences in the indices of ammonium-N
accumulation between the soil with leucaena andcassia mulch
were not significant. The mulch of both species resulted in
generally higher indices of N-availability than the control
plots. The indices of both total-N and ammonium-N mobilized
were the highest for the soil of the fertilized pots after 16
weeks and lowest in the control pots.
Synchrony between N uptake bv maize
and N mineralization of the mulch
Synchrony (or lack thereof) between the periods of peak
of N release from the mulch of cassia and leucaena with peak
biomass and N-yield of the maize was determined (Fig. 4-6), by
superimposing the N release curves of the soil-incorporated
leucaena and cassia mulch in the second season field
decomposition studies, on the biomass yield of maize from the
pot studies (Table 4-4) for each of the four weekly harvests.

131
Table 4-5. Relative indices of cumulative NH4+ N, N03~ N and
total N mobilized after soil incorporation of 2 t
ha*1 equivalents of Leucaena leucocephala and
Cassia siamea mulch in pots with maize, second
season of 1991.
Treatment
nh4- n
(yg/g
resin)
no3* n
(pg/g
resin)
Total N
mineral
ized
(pg/g
resin)
Leucaena
66.0*
106.9*
172.9*
Cassia
63.8*
39.9*
103.7b
Fertilizer
124.6b
101.9*
226.5C
Control
35.8C
29.9b
65.7d
*,b'C'dMeans in column followed by different letters are not
significantly different (p = 0.05).

132
Averaged across species, 58% of the mulch N was
mineralized in the first four weeks. In contrast, maize
biomass yield and N uptake were highest during weeks 8-12
after mulch or fertilizer application. Hence, the peaks of
N release from the mulch and uptake by the maize were not
matched or synchronized. Although the N release patterns of
the two species were similar, the differential response of
maize biomass yield to the mulch of the two species in weeks
4 and 16 is worth noting. At both times, the biomass yield
of maize was significantly more with leucaena than with
cassia mulch.
Discussion
Patterns of Mulch Decomposition: Effects
of Quality and Soil placement Methods
The chemical properties of the mulch of the two
species including the relatively high polyphenol of leucaena
leaves are withip the range reported by others (Ahn et al.,
1989; Fox et al., 1990; Palm and Sanchez, 1991; Oglesby and
Fownes, 1992). The patterns of biphase mulch decomposition
consisting of an initial phase of rapid dry matter loss
followed by a second phase of lower rate has also been noted
by other workers working with different materials and in
different ecozones (e.g., Parker, 1962; Van der Kruijs et
al., 1984; Swift, 1986; Kumar and Deepu, 1992; and several
others). The first phase reflects the rapid loss of easily
decomposable materials (carbohydrates) and the second phase,
the slow loss of more recalcitrant materials (Wieder and

133
Lang, 1982). Plant structural materials and microbial
products comprise a greater proportion of the recalcitrant
residual mass in the second phase (Cheshire et al., 1974).
The predicted and the observed values of the mulch
remaining not decomposed were regressed well by two-stage
exponential model. This model, which has also been used by
other workers (Howard and Howard, 1974; Hunt, 1977; Lousier
and Parkinson, 1976) has considerable advantages in
providing insights into the process of the decomposition
process over statistical procedures typically used to
analyze decomposition data. Wieder and Lang (1983) provides
a comprehensive critique of the analytical methods used in
examining decomposition data obtained from litter bags.
The general absence of significant differences between
the rates of decomposition of the soil incorporated and
surface placed mulch was unexpected. Other workers have
observed that soil incorporated mulch decomposes faster than
surface placed mulch (Read et al., 1985). Generally, the
rates of decomposition of plant materials are more rapid in
moist wet and warm than in dry conditions (Swift et al.,
1979; Shield and Paul, 1973). Moisture conditions generally
increase with depth from the soil surface. Microbial
activity also increases linearly with increase in soil
water, at least up to 50% of the soil water holding capacity
(Murwira et al., 1990). Extremes of drying and wetting
conditions which cause conditions less favorable for
decomposition (Parr and Papendick, 1978) predominate more on

134
the surface than inside the soil. Besides moisture, surface
applied residue may also remain deficient in N for a longer
period than the incorporated residue (Summerell and Burgess,
1989; Cochran, 1990).
In spite of all the favorable conditions for
decomposition within the soil, the lack of significant
difference between the soil incorporated and surface placed
mulch suggest that conditions within and outside the soil
were not drastically different. The litter bags were
incorporated to only 10 cm depth. Also, the litter bags
could have created more humid conditions for the mulch in
the bags compared to the mulch outside the bags (Jensen,
1974). Moreover, the higher soil temperatures on the
surface could also account for the unexpectedly high rate of
decomposition on the surface. Under moist conditions, rates
of decomposition of litter on soil surface correlate well
with air temperature (Williams and Gray, 1974; Witkamp,
1966; Comanor and Staffeldt, 1978; Gholz et al., 1985).
The observation that the rates of mulch decomposition
between the two seasons were similar, in spite of the
significant differences in precipitation, would suggest that
soil incorporation of mulches could be an approach to
enhance the rates of mulch decomposition and availability of
nutrients to crops. Also the effects of wind blowing away
mulch from the fields, particularly before the crop stand
develops sufficient cover could be minimized. However, soil
incorporation is labor demanding. As a management practice

135
(and not one done within the schedules of the plowing and
weeding fields), it has to be justified by proportionate
gains in crop yield. Besides, crop response to soil-
incorporated or surface-applied mulch may be inconsistent
between seasons even at the same site as is demonstrated by
the studies of Kang et al. (1981) and Read et al. (1985),
which were both conducted at the International Institute of
Tropical Agriculture (IITA), Ibadan, Nigeria.
Field studies that compare the rates of decomposition
of leucaena and cassia mulches at the same site are scarce.
The rates of decomposition observed in both seasons of the
present study were remarkably similar to those reported by
Budelman (1988) for soil-surface-placed leucaena mulch in
the humid lowland tropics of West Africa. Several other
studies have also reported rates of decomposition of
leucaena similar to those of the present study (Kang and
Duguma, 1985; Weeratna, 1979). Working at IITA, Yamoah et
al. (1986a) noted, however, rates of decomposition of cassia
mulch much lower than those of the present study: in 120
days, 85% of the cassia dry matter had disappeared.
However, in laboratory incubation studies, leucaena has been
observed to decompose faster than cassia (Nyamai, 1992).
Because the rates of decomposition of mulch materials are
influenced by many factors such as site, management ,(e.g.,
soil placement methods) and intrinsic properties of the
mulch, differences as well as similarities between and among
studies are commonly reported and are not surprising. At

136
any one site, an understanding of the mulch factors that
regulate rates of decomposition is important.
Mulch Quality and the Rates of Decomposition
and N Mineralization
First phase
The rates of decomposition of the two phases appeared
to be regulated by different factors. In the first phase,
the C:N ratio seemed to be the main factor for both species.
Generally, organic materials with high N concentrations or
narrow C:N ratio decompose and mineralize fast unless
constrained by environmental and plant factors such as
polyphenols (Parr and Papendick, 1978; Vallis and Jones,
1973). The C:N ratios of leucaena and cassia mulch, 11:1
and 13:1, respectively, are below the critical ratio of 33:1
that Bartholomew (1965) had mentioned as the limit beyond
which immobilization may occur. Also, the N concentrations
of the leaves of both species were above the critical level
of between 1.5% and 2.5% that is generally considered as the
minimum level before net mineralization occurs (Allison,
1973; Stevenson, 1986; Fox et al., 1990). As a consequence
of narrow C:N ratio, the rates of N mineralization of the
mulch high with more than 50% of the mulch-N in both species
released within the first four weeks after soil
incorporation. The rates of N-mineralization declined
significantly after eight weeks which would perhaps suggest
occurrence of net N-immobilization.

137
The presence of high polyphenol content is reported to
influence the rates of decomposition of leguminous residues
with high N but yet with low decomposition rates (Vallis and
Jones, 1973; Weeratna, 1979; Ladd et al., 1981). Other
suggested factors include polyphenol-to-N ratio (Palm and
Sanchez, 1991; Oglesby and Fownes, 1992) and lignin plus
polyphenol to N ratio (Fox et al., 1990). The concentration
of polyphenol in the leaves of leucaena used in the present
study were high relative to that used by others, for
instance, Palm and Sanchez (1991). However, the rates of
decomposition in the first phase did not appear to be
influenced by the concentration of polyphenols in the
leaves. In fact, leucaena leaves decomposed at
significantly higher rate than cassia leaves, in spite of
the latter's lower polyphenol content. These observations
suggest that polyphenols (and their ratio with N) were
either not important, or that the levels and the types of
polyphenols present in leucaena leaves were not rate
limiting. It is also possible that the polyphenols present
in leucaena and cassia leaves were water-soluble and readily
dissolved from decomposing residues before impacting
negatively on the activities of the decomposer community
(Olson and Reiners, 1983; Baldwin et al., 1983). Thus, the
primary factor regulating the rates of decomposition of the
first phase of both species appears to be narrow C:N ratio,
and not polyphenol or lignin content.

138
Second phase
Compared to the first phase, it is less clear what
factors regulated the rate of decomposition in the second
phase. Increased C:N ratio (the C probably being in the
form of lignin) and polyphenols could be involved but the
former appears to provide the best explanation. Two
observations favor high C:N (lignin:N in particular) ratio.
The first is the increasing ratio of twigs in the mixture of
twig and leaves over time in the second phase (data not
shown). In spite of their high polyphenol content relative
to the twigs, leucaena leaves had practically all
disappeared at the start of the second phase. Instead, the
twigs with higher lignin content than the leaves were more
dominant than leaves in the second phase. The second
observation was that leucaena twigs with slightly higher
polyphenol concentration than cassia twigs had rates of twig
decomposition not significantly different from those of
cassia.
In the absence of lignin, wide C:N ratio may not be a
rate limiting factor. This is because the filter paper with
the highest C:N ratio had rates of decomposition not
significantly different from the twigs. These observations
support the argument that the form of C, lignin, must be
taken into account along with the C:N ratio in predicting
residue decomposition rates as reported also by Herman et
al., 1977; Parr and Papendick, 1978.

139
Although high lignin:N ratio appears to be an
explanation for the low rate in the second phase compared to
the first of decomposition, an examination of the lignin:N
ratios of the mulch at the start of phase two suggested
either some differences between the two species and/or the
presence of another factor influencing the rate of
decomposition. Mellilo et al. (1982) observed lignin:N
ratio to provide the best prediction of N release from
decomposing plant materials, although several workers have
not found similar relationship (Schlesinger and Hasey, 1981;
McClaugherty et al., 1985; Taylor et al., 1989). At the
start of phase two, the lignin:N ratio of the leucaena mulch
was 6.5:1 while that of cassia was 20:1. The lignin:N ratio
of cassia of 20:1 is within the upper range of the lignin:N
ratio of the plant materials that Mellilo et al. (1982)
used. Therefore, high C:N ratios or more specifically, high
lignin:N ratio, may be the primary factor regulating the
rate of cassia mulch mineralization in phase two.
For leucaena, the lignin:N ratio of 6.5:1 is below the
wide range of plant lignin:N ratios that were studied by
Mellilo et al. (1982) used. Therefore, wide C:N ratio (or
lignin:N ratio) per se may not be a good predictor for
decomposition of leucaena mulch in the second phase. The
polyphenol levels of leucaena twigs were within the range
observed by other workers to limit the rates of
decomposition of some leguminous materials (Palm and
Sanchez, 1991; Oglesby and Fownes, 1992). Given that

140
leucaena twigs had slightly more lignin and polyphenol than
cassia, the combined lignin and polyphenol content (Fox et
al., 1990) could explain the lack of significant differences
in the rates of decomposition and N-mineralization of the
mulch (leaves plus twigs) from the two species. This
explanation is based on results from the leaves of only two
species and may not extend to other species whose C:N ratio,
lignin, polyphenol concentration and age are outside the
range considered.
For materials with high N concentration (i.e., narrow
C:N ratio) such as leucaena leaves used in this study or
leaves and small twigs of alfalfa (Medicago sativa L.) the
second phase of decomposition may not even exist or be
undetectable (Fox et al., 1990). Since the phases as well
as their rates of decomposition have important implications
on the availability of nutrients from the mulch to the crop,
it is important to firmly establish the relative importance
of each factor involved. This my be achieved through a
study involving a wide range of species in which both the
quantity and the quality of the chemical characteristics are
continually monitored.
Implications of Mulch-N Mineralization
Patterns on Uptake; Potentials for Synchrony
Synchrony is said to be achieved when the peaks of N
release of N from the mulch and demand or uptake by the
crop are matched. In the present study, the peaks were
not synchronized as is demonstrated when the patterns

141
of N-mineralization of the mulch in the field are
superimposed on the patterns of N uptake and biomass yield
of maize from the pot study. The rates of N release from
the mulch of both species were highest in the first four
weeks application, with 80-85% of the mulch N being
mineralized. After weeks 8 and 6 for leucaena and cassia,
respectively, the rates of N release were on a declining
trend. This would suggest N immobilization, an observation
made also by Oglesby and Fownes (1992). In field
conditions, synchrony may be difficult to establish because
of the dynamic equilibrium that exists between
mineralization and its control by immobilization and losses
through volatilization, denitrification and leaching (Woomer
and Ingram, 1990).
From the pot studies, maize biomass yield and N uptake
were highest in the period spanning 8-12 weeks after mulch
application. Typically, the period between the onset of
flowering or silking is when N requirements by maize is at
peak (Arnon, 1975) or when daily rate of N accumulation is
highest (Sayre, 1955). For a 120-day maize variety such as
the1 one used in the present study, the onset of flowering or
silking occurs between weeks 6 (Bromfield, 1969; Nadar,
1984) and 8 (Hanway, 1962). In weeks 4 and 16, when N
requirements are low, both biomass yield and N uptake of the
plants were significantly lower than those of weeks 8-12.
Splitapplication of the mulch, timing of application,
and placement method and controls on the quality of the

142
residues have been suggested as management practices that
may enhance synchrony between the peaks of supply of
nutrients from decomposing organic residues and demand by
the crop (Read et al., 1985; Yamoah et al., 1986a; Woomer
and Ingram, 1990; Palm and Sanchez, 1991). Implicit in
these suggestions is that prospects of attaining synchrony
are greatest when there is diversity of mulch resources and
when management options are relatively flexible.
Application of the mulch four weeks after sowing the
crop rather than at sowing may under the conditions of this
study have achieved synchrony between the peaks of N release
from the mulch and uptake by the crop. Under the conditions
of the present study, the second pruning of the hedgerows
was generally carried out 4-6 weeks after sowing the crop.
Therefore some mulch was available for application four
weeks after sowing the crop. The timing of pruning would,
however, depend on rainfall conditions of the season,
hedgerow species and the rate of regrowth after pruning. At
the site of the present study, leucaena could yield
substantial amounts of mulch in the second pruning but
cassia could not. However, mulch yield of both species was
highest in the first pruning of the hedgerows at the time of
sowing the crops. While storage of mulch from the first
pruning and its application at later date could be a
suggestion, it may not be a practical proposition to
resource-poor farmers.

143
Although the periods of peak release of N from the
mulch (four weeks after application) and uptake by the crop
(eight weeks after sowing) did not coincide, both the
biomass and N yield of the crop was highest eight weeks
after mulch application. This indicates that N mineralized
earlier from the mulch was available to the crop and not
lost as such. Even after 16 weeks post application, crop
response to mulch-N still persisted. These observations are
consistent with those of others that the residual effects of
mulches on crops, even of leucaena leaves that decompose
rapidly, could be significant (Read et al., 1985). This
observation also suggests that the ability of the soil to
retain plant-derived N can be strong compared with the
ability of the crop to take it up, currently or
subsequently, and different loss mechanism to remove it, as
studies by Muller and Sundman (1988) using 1SN have
demonstrated.
The higher apparent N recovery by the crop from
leucaena than from cassia mulch (a total of 34% against 4%
across the four harvests for leucaena and cassia,
respectively) could be due the to the higher N concentration
(20% more) and rate of decomposition of leucaena leaves.
Similar observation has been made by others working with
mulch materials of high and low N concentrations (Fox et
al., 1990). The N recovery rates are, however, low but
comparable to those reported by other workers for mulches of
leguminous trees (Read et al., 1985; Gutteridge, 1992; Fox

14
et al., 1990). The low apparent recoveries of N added may
be due to immobilization by soil microbes, especially if th<
N added through the mulch is low (Allison, 1966). This may
explain the negative apparent N recovery from cassia mulch
in week 8. Management approaches that synchronize the peaki
of mulch N mineralization and N uptake by the crop could
enhance apparent recoveries of N by the crop from the littl(
amounts of mulch applied.
Implications of Decomposition Patterns of
Leucaena and Cassia on Soil N and Organic
Matter Build-up
Rapidly decomposing mulches such as leucaena leaves ir
this study or the leaves of Gliricidia sepium (Yamoah et
al., 1986a; Palm and Sanchez, 1991) may provide sufficient 1
for the crop. When applied at high rates and for a long
period, for instance, 8-10 t ha'1 yr'1 dry matter for 6
years, leucaena could also increase soil organic matter
(Kang et al., 1985). Materials with low decomposition rates
(either due to high C:N ratio, lignin or polyphenol
concentration) such as cassia may not provide enough N for
plant growth in the short term, especially when applied in
small amounts, but they may contribute to the build-up of
organic-N and humus in the soil (Yamoah et al., 1986a).
In addition to the many benefits of humus on soil
chemical and physical properties (Swift and Sanchez, 1984),
increase in soil organic matter may provide a low but
continual supply of N in the long term. For leucaena mulch,

145
application of a mixture of leaves and twigs could be a
management tool to reduce the rate of decomposition and
hence contribute to soil organic matter build-up.
Conclusions
Decomposition patterns of the mulch of both species
were biphasic, i.e., an initial phase of a rapid rate
followed by a second phase of a lower rate. When soil
incorporated, the decomposition rates of the mulch (leaves
plus twigs) of both leucaena and cassia were similar (about
12% per week in the first phase and 2% per week in the
second phase). However, when placed on the soil surface,
leucaena mulch decomposed more rapidly than cassia mulch in
the first phase of decomposition.
Narrow C:N ratio appears to explain the rapid rates of
decomposition of both species. In the second phase, high
C:N ratio (or lignin) appears to explain best the low rates
of decomposition of both species. In leucaena, however,
polyphenols also seemed to be involved in regulating the
rate of decomposition of the second phase.
Synchrony was not observed between periods of peak
mulch-N release (i.e, four weeks after application of mulch
to the soil) and peak uptake by the maize crop (eight weeks
after sowing). Low apparent recovery of the mulch-N applied
by the crop observed (34% from leucaena and 4% from cassia)
was perhaps a reflection of the poor synchrony between
mulch-N mineralization and uptake by the crop. It is

146
concluded that application of the mulch four weeks after the
onset of the rains may help achieve synchrony.

CHAPTER 5
INTERACTIONS BETWEEN MAIZE AND HEDGEROWS OF
LEUCAENA LEUCOCEPHALA AND CASSIA SIAMEA
IN ALLEY CROPPING SYSTEM IN SEMIARID
TROPICS AT MACHAROS, KENYA
Introduction
The promising results of soil fertility improvements
under alley cropping (hedgerow intercropping) in the humid
tropical lowlands (Kang et al., 1990) have caused much
interest in this technology elsewhere. This has
particularly been so in the semiarid tropics where fertility
of the soils are often low and inputs are limited (Steiner
et al., 1988). Studies on alley cropping in the semiarid
tropics, however, indicate significant reductions in crop
yields (Rao et al., 1991; Singh et al., 1989; Nair, 1987).
Below ground competition, particularly between the tree and
the crop is often suggested to be the main limitation to
hedgerow intercropping in the semiarid tropics (Singh et
al., 1989; Ong, et al. 1991; Monteith et al., 1991).
Typically, reductions in crop yields are highest near
the tree hedgerow and they diminish with distance from the
hedgerows (Ong et al. 1991; Duguma, et al., 1988).
Observations of yield increases nearer the hedgerows are also
not uncommon (Sang and Hoekstra, 1987; Huxley et al., 1989;
Ong et al., 1992; Dzowella, 1992). Also, yields of crops may
remain unaffected by the distance from the hedgerows
147

148
(Lai, 1989). These inconsistencies of crop performances in
relation to the tree hedgerows suggest some difficulties in
predicting the pattern of interaction between the tree and
the crop. Differences in seasonal rainfall could further
complicate any predicted pattern of tree-crop interaction.
More studies are, therefore, needed to identify factors that
may predict well the pattern of interaction between the tree
hedgerows and the crop in a given state.
Besides site differences, the hedgerow and crop
species used, their management (e.g., height and frequency
of pruning), and the absence or presence of fertilizers are
some of the factors that could alter the pattern of
interaction between the tree hedgerow and the crop. Studies
on alley cropping in semiarid tropics have focused mainly on
leucaena (Rao et al., 1991; Lulandala and Hall, 1990; Ong
et al., 1992). There is need to include more species in
hedgerow intercropping in semiarid areas. Where other
species have been included (Nair, 1987), crop yields were
lowest under leucaena intercrop. Also, there is a general
paucity of information on the relative importance of soil
resources (e.g., soil fertility and water content) on the
potential interaction between the crop and the hedgerows.
In addition to the inclusion of more species, it is
also necessary to characterize the profile of crop yield,
soil chemical, and physical properties (e.g., N, P and water
content) in relation to the position of the hedgerows. Such
studies will determine not only the nature of interaction

149
(positive or negative), but also identify possible causes of
the observed interaction, and, in turn, may provide insights
into determining the potentials for as well as management of
alley cropping, particularly in the semiarid tropics.
In consideration of these, this study was undertaken
with the following objectives:
1. To evaluate the yield profile of maize
intercropped with hedgerows of Leucaena
leucocephala and Cassia siama in relation to the
distance of the crop from the hedgerows.
2. To assess the relative impacts of soil water and
soil fertility (nitrogen) on the observed profile
of maize yield.
Materials and Methods
Site Description
The study was carried out under rainfed conditions at
the Field Station of ICRAF, located at Machakos, Kenya
(latitude 1 33' S and longitude of 37 14' E and 1596 m
elevation). The patterns of rainfall and soil
characteristics of the site are described in Chapter 2.

150
Methods
Treatment and layout
Hedgerow intercropped plots of Leucaena leucocephala
and Cassia siamear 6.67 m wide were established in October
1987 in randomized block design with three replication. The
hedgerows were part of a larger experiment involving
comparison of hedgerow intercropped verses block planting
systems (see previous chapters). Unlike the other two
hedgerow spacings (i.e, 4 and 5m), the 6.67 m wide
hedgerows were used for the tree-crop interaction study
because they provided the widest separation of hedgerows.
This was necessary in order to determine the distance of
influence of the hedgerows on crop yield.
Tree and crop management
The trees were planted in October 1987 from 6 month
old seedlings raised in pots. After one year of growth in
the field, the trees were pruned back into hedges of 50 cm
height above ground. This height of pruning was maintained
in all cropping seasons with 2-3 prunings in a season. The
prunings were applied as mulch (with or without
incorporation) to the alleys in between the hedgerow. Maize
was sown in the alleys twice a year.
The maize in the alley was spaced 90 cm between row
and 30 cm with rows. There were seven rows in the alleys

151
with the closest one to the hedgerow being only 45 cm away
and the farthest, 3.5 m away. Moderate levels of fertilizer
(40 and 18 N and P kg ha'1' was applied to the maize every
season.
Measurements
Crop
In the last three seasons (i.e., seasons 5, 6, and 7)
out of a total six of the study, the crop was harvested row-
by-row from the alleys of both species and separated into
grain, stalks and stover. Before final harvest, biomass
yield of each crop was determined twice, at 30 and at 60
days after sowing, near and away from the hedgerows.
At harvest, the N and P content of the crop grain,
stalks, and stover of the row nearest the hedge (referred to
as row one) and the row farthest from the hedge (row four)
were also determined. At first sampling (30 days after
sowing), N and P content of samples from the whole shoot
after milling and passing through a 2 mm sieve was
determined. At the second harvest/sampling, N and P content
of the ear leaf was determined. Standard micro-Kjeldahl
method, modified for plants, was used for N analysis; P was
determined with UV visible Spectrophotometer at 450 nm.
One to two days before the first and second harvest
for dry matter assessment, relative water content of the

152
samples was determined by the leaf disc procedure
(Weatherly, 1950). This is a measure of the actual water
content of a tissue with reference to the content of water
in the tissue if it were fully hydrated (Bennett, 1990).
Measurement of relative water content by this procedure
requires obtaining three measurements made on the same
sample: fresh weight, turgid weight, and dry weight. Discs
were cut with a corer of 10 mm diameter from the ear leaves
of maize plants growing near and away from the hedgerows.
Fresh weights of the discs were determined immediately.
Turgid weight was determined after the discs had been
immersed in distilled water for 3 hours in darkness. Dry
weight of the discs was determined after oven drying the
discs for 48 hours at 105C. Percent relative water content
(RWC) was then expressed as:
%RWC = (fresh wt dry wt/turgid wt dry wt) x 100.
In addition to the leaf discs, whole plant water
content was determined in the first and second sampling of
each season as the difference between fresh and dry weight
of the whole plants (shoots) harvested (Turner, 1981).
Whole plants near and away from the hedgerows were
harvested, fresh weight recorded immediately and dry weights
after oven drying for 48 hours at 105C.

153
Soil properties
Bulk density of top soil (0-20 cm depth), water
content and water potential, infiltration rate, hydraulic
conductivity, total-N, P, available N and fauna population
were monitored at crop rows 1 and 4 of the intercropped
hedgerows. Bulk density was determined with a 5 cm auger to
a total depth of 20 cm (i.e, 0-5, 5-10, 10-15 and 15-20 cm)
during the last season of cropping. Two soil cores were
taken at each depth and the average taken as the
representative bulk density. Gravimetric soil water content
was determined to 80 cm depth at intervals of 20 cm depths.
In the fifth season, soil sampling was carried out only on
February 15, 1991 (i.e., end of fifth season). In the sixth
and seventh seasons, soil sampling was carried out at the
end of each month including the months of no cropping.
Fresh weights of the soil were determined in the field and
dry weights in the laboratory after oven drying for 48 hours
at 105C. The difference between fresh and dry weight,
expressed as percent, was taken to represent the gravimetric
water content.
Because of the destructive nature of the gravimetric
sampling method as well as the need to increase the
frequency of monitoring soil water content, tensiometers
(type: Irrometer; Irrometer Co., California) were placed to
15 cm depth at the beginning of the sixth season at crop

154
rows 1 and 4 of the intercropped hedgerows. At the time of
tensiometer placement, ordinary thermometers were also
placed at rows 1 and 4 to 10 cm depth.
After harvest of the sixth and seventh season crops,
infiltration rate of water into the soil near and away from
the hedgerows was determined with the double ring
infiltrometer (Landon, 1984). Infiltration rate was
determined twice at each position (i.e., near and away from
the hedgerows) within a plot. At end of the seventh
cropping season, saturated hydraulic conductivity of the
soil was also determined in the field near and away from the
hedgerows using a disc permeameter (CSIRO, 1988).
Hydraulic conductivity was determined at two spots near and
away from the hedgerows. Also, at the end of cropping in
the seventh season, observations were made on the
populations of soil fauna present near and away from the
hedgerows. This involved taking soil monoliths of 25 cm. x
25 cm x 25 cm and hand sorting for macro-invertebrates (body
lengths greater than 2 mm) (Anderson and Ingram, 1989).
Soil monoliths for the fauna counts were taken from two
spots near and away from the hedgerows.
Analyses of total N was by the micro-Kjeldahl method
(Jackson, 1958), organic carbon by the Walkley-Black
procedure (Jackson, 1958), and P by the Mehlich-1 double
acid procedure. Soil samples for N, P and C determination
were collected from the top 20 cm soil profile near and away

155
from the hedgerows. Soil analyses were conducted at the
Analytical Laboratory of the Katumani Research Station,
Machakos, Kenya. In addition to total N, relative index of
available N was determined with ion-exchange resin bags
buried to 10 cm depth near and away from the hedgerows in
the last season (seventh) of the study. On retrieval of the
ion exchange resin bags at the end of the cropping season, N
trapped by the resins was extracted with 2 M KC1. Ammonium-
N and nitrate-N contents of the KCl-extract were determined
at the Soil Testing Laboratory of the University of Florida,
Gainesville.
Supplementary Studies
To determine the relative influence of soil water or
fertility (N) on the observed tree crop interaction,
supplementary field and pot studies involving factorial
combination of fertilizer N and irrigation were initiated in
the last season of the study.
Field studies: irrigation and fertilization
Treatments to evaluate the response of maize yield to
N fertilization and irrigation, in combination or alone, was
introduced in small plots that were formerly unplanted
strips between the plots of the main experiment. The
treatments of the experiment were: irrigation only,
fertilization only, irrigation plus fertilization and

156
control (no irrigation or fertilization). Each treatment
was replicated thrice in plots of 4 m x 8 m. Amounts of
irrigation water applied was the difference between pan
evaporation losses and rainfall. The plots were irrigated
once a week until harvest. Fertilizer was applied at the
rate of 100 kg N ha-1 at the time of sowing the crop.
Pot Studies: hertilization and
irrigation of soil collected from
near and away from the hedgerows
A factorial combination of three levels of N at
equivalents of 0, 40 and 100 kg N ha'1 season'1, and two
levels of irrigation were chosen as treatments for the pot
studies. Soil was collected from the top 40 cm profile near
(row 1) and away from the hedgerows (row 4) of leucaena and
cassia. The soil was contained in 30 x 30 x 30 cm pots
(polythene bags). Each pot contained 20 kg of soil. Each
of the treatments mentioned above was replicated three times
in randomized block design. Maize (Katumani Composite
variety) was grown in the pots for a total duration of two
months. Total biomass (shoot plus roots) of maize plants in
the pots was harvested and dry matter determined after oven
drying at 105C for 48 hours.
Irrigation levels applied to the pots were the
equivalents of 100 and 200 mm rainfall in a duration of two
months. This period is half a typical cropping season of
four months. The choice of the 100 mm and 200 mm rainfall

157
was based on amounts of rainfall received in the last two
cropping seasons, 214 mm and 374 mm. Hence, the pots
received only half of a season's rainfall. The quantity of
water applied to each pot was determined as the quantity
required to replenish loss of "available soil water," i.e.,
the difference between field capacity and wilting point.
For the 100 mm rainfall equivalent, the amount of water
required per pot corresponded to the application of 0.2
liters of water a day; for the 200 mm rainfall equivalent,
the level was 0.4 liters a day. The pots were irrigated
twice a week.
Statistical Analysis
ANOVA of split-plot design was used. For the field
studies of the two species involving measurements near and
away from the hedgerows, the option for repeated measures in
the GLM procedure of the SAS (SAS, 1991) was included.
Repeated measures option was necessary in order to adjust
for the fixed (as opposed to random) effects of crop
positions from the hedgerows and/or seasons. Analysis of
the data from the pot studies was similar to that for field
study described above since the soil was collected from
fixed positions near and away from the hedgerows.
In the field study involving irrigation and
fertilization, standard ANOVA of randomized block design was

158
performed. In all analyses (i.e., field as well as pot
studies), pairwise comparison of Least Square Means were
used to test significance of differences between treatment
means (p 0.05).
Results
Crop Yield Profiles
The row-by-row maize grain yield profiles of the three
seasons of the study are shown in Figures 5-1 to 5-3. An
overall ANOVA of the combined data of the three seasons
revealed significant interaction effects of season x tree
species x row distance from the hedgerow on yield. Hence,
separate analysis for each season was done to determine the
main effects of species and row-spacing on grain yield.
In season five, only main effects of species was
observed (Fig. 5-1). Maize yield of any row in association
with cassia was significantly higher than the row yields
with leucaena. Effects due to distance of crop rows from
the hedgerows were significant. Averaged across rows, maize
yield under cassia was 28% higher than the yield of the rows
under leucaena. In this season, mean yield of the rows of
the sole-crop (fertilized) was, on average, 26% higher than
the yield under leucaena and 11% lower than yield under
cassia.

159
Nearest hedgerow Farthest from hedgerow
Figure 5-1. Maize yield response to distance from
alley cropped hedgerows of Leucaena leucocephala and Cassia
siamea, 1990 short-rains. Vertical bars indicate standard
error of difference of means.

Maize yield (kg/5 m long row)
160
Figure 5-2. Maize yield response to distance from
hedgerows of Leucaena leucocephala and Cassia siamea, 1991
long-rains. Vertical bars indicate standard error of
difference of means.

Maize yield (kg/5 m long row)
j
161
Figure 5-3. Maize yield response to distance from
alley cropped hedgerows of Leucaena leucocephala and Cassia
siamea, 1991 short-rains. Vertical bars indicate standard
error of difference of means.

162
In season 6, significant main effects of both species
and row distance from the hedgerows were observed (Fig. 5-
2). Interaction effects were not significant. As in the
first season, yield of rows under cassia was significantly
higher than those under leucaena. On average, the
difference was 49%. Compared to the sole crop (fertilized),
yield under cassia, averaged over all the rows, was 5% lower
while that under leucaena was 51% lower. With respect to
the effects of row distance from the hedgerows, grain yield
declined significantly with distance from the hedgerows of
both species.
In season 7, no significant effects of interaction
between species and row distance of crops from the hedgerows
was observed. Main effects of species and distance of the
maize rows from hedgerows were significant. Maize yield
declined in a significantly linear manner with distance from
the hedgerows of both species (Fig. 5-3). On average, maize
yield under cassia was 2% and 18% higher than under sole
crop and leucaena hedgerows, respectively. Therefore, the
maize yield under leucaena was 16% lower than that of the
sole crop.
Averaged over all the three seasons, the first row of
maize next to the hedgerows was 7% and 12% more productive
than the next row in leucaena and cassia, respectively.
When averaged over the sixth and seventh seasons (when
significant trends of yield declines with distance from

163
the hedgerows), the first maize row next to the hedgerows
was 15% and 21% more productive under leucaena and cassia,
respectively, than the next row.
Crop Nutrients (N and P) and
Water Content Measurements
No significant effects were observed due to distance
of crop rows from the hedgerows of both species on N and P
content of the crop grain. Similarly, there was no
significant difference in N and P content of the biomass of
maize sampled before final harvest of the crop in all the
three seasons. For example, in the sixth season with very
low rainfall, biomass N of maize near and away from the
hedgerows was not significantly different (Table 5-1).
At the time of maize biomass sampling in the sixth
season, maize was severely stressed by drought. This
observation was most pronounced away from the hedgerows. It
was then hypothesized that at the time of maize stress,
plants near the hedgerows would have higher water content
than those away from them. However, there were no
significant differences in water content of plants near or
away from the hedgerows (Tables 5-2 and 5-3). Nevertheless,
trends of higher water content in plants near the hedgerows
than those away from them were detectable.
Differences in plant water content were more apparent
between plants under the two hedgerow species than between
positions of plants within a given hedgerow species. For

164
Table 5-1. Nitrogen content of 30 day-old maize shoots
(leaves plus stalks) sampled near (0.45 m) and
away (3.5 m) from alley cropped hedgerows of
Leucaena leucocephala and Cassia siamea, fifth
season, 1991.
N content
(%)
Species
Near
(0.45 m)
Away
(3.5 m)
Mean
Leucaena
3.3
.3.3
3.3a
Cassia
3.1
3.2
3.2a
Mean
3.2
3.3a
3.3
Means in a column
not significantly
and row followed
different (p = 0
by the same
.05) .
letters are

165
Table 5-2. Relative water content of flag leaves of maize
plants sampled near (0.45 m) and away (3.5 m)
from alley cropped hedgerows of Leucaena
leucocephala and Cassia siamea on April 26, 1991
(sixth season).
Water content
m
Near
Away
Species
(0.45 m)
(3.5 m)
Mean
Leucaena
40.2
.34.1
37.2a
Cassia
48.7
46.8
47.8b
Mean
44.4a
40.5a
42.5
a,bMeans in a column and row followed by the same letters are
not significantly different (p = 0.05).

166
Table 5-3. Water content of whole maize plants sampled near
(0.45 m) and away (3.5 m) from alley cropped
hedgerows of Leucaena leucocephala and Cassia
siamea hedgerows, sixth season, 1991.
Water content
(%)
Near
Away
Species
(0.45 m)
(3.5 m)
Mean
Leucaena
81.0
79.3
80.2a
Cassia
86.6
83.0
84.5a
Mean
83.5
81.2
82.3
aMeans in a column and row followed by the same letters are
not significantly different (p *= 0.05).

167
instance, water content of the discs (or discs cut from the
leaves of maize plants) under leucaena was significantly
lower during the dry season than those under cassia (Table
5-4). In the seventh season, when rainfall conditions were
normal, no significant difference in plant water content
near and away from the hedgerows or between species was
observed (Table 5-5).
N content of maize grain from the two hedgerow species
treatments was not significantly different (data not
presented), although a difference of 19% was detected in the
N concentration of the sixth season grain harvest from the
two treatments.
Soil Physical and Chemical Properties
Of the various soil physical and chemical properties
monitored, significant differences due to the effects of the
hedgerow species or of distance of crop rows from the
hedgerows were detected in: (a) bulk density, (b) rates of
water infiltration, (c) hydraulic conductivity (saturated)
of the soil, (d) gravimetric water content, (e) soil
temperature, and (f) soil fauna. Bulk density of top soil
(0-5 cm) near the hedgerows was significantly lower than
away from them (Fig. 5-4). This was so only under leucaena,
not cassia. Also, below 5 cm depth, difference in soil bulk
density near and away from the hedgerows was not
significant.

168
Table 5-4. Relative water content of flag leaves of maize
plants sampled near and away from Leucaena
leucocephala and Cassia siamea hedgerows on May
28, 1991 (sixth season).
Water
content (%}
Species
Near
(0.45 m)
Away
(3.5 m)
Mean
Leucaena
28.3a
45.3b
38.8
Cassia
58.4a
55.4b
56.9
Mean
43.4
50.4
46.9
a,bMeans in a row followed by the same letters are not
significantly different (p = 0.05).

169
Table 5-5. Relative water content of flag leaves of maize
plants sampled near (0.45 m) and away (3.5 m)
alley cropped hedgerows of Leucaena leucocephala
and Cassia siamea on January 6, 1991 (seventh
season).
Water content
(%)
Near
Away
Species
(0.45 m)
(3.5 m)
Mean
Leucaena
31.6
o

o
m
30.8
Cassia
31.3
34.8
33.0
Mean
31.4
32.4
31.9
Means in a column and row followed by the same letters are
not significantly different (p = 0.05).

Bulk density (g cm'
170
2.0
Leucaena
Species
I I Near
153 Away
Cassia
Figure 5-4. Bulk density of top 5 cm soil profile
near (0.45 m) alley cropped hedgerows of Leucaena
leucocephala and Cassia siamea, 1991 short-rains.

171
For infiltration rates, main effects of species were
not significant (Fig. 5-5). The rates of infiltration of
water near the hedgerows of both species were significantly
higher than away from them.
Significant interaction effects were observed between
tree species x distance of crop rows from the hedgerows on
hydraulic conductivity of the soil (Table 5-6). The rates
of hydraulic conductivity near the hedgerows of leucaena
were significantly higher than away from them; in cassia,
the differences were not significantly different.
With respect to gravimetric water content of the soil,
only the main effects of species were significant in any
given date of measurement. Effects of the distance of crop
rows from the hedgerows on soil water content were not
significant. At all dates of measurements, soil water under
cassia was consistently higher than under leucaena. This
was particularly so at soil depths below 20 cm (Figs. 5-6
and 5-7) and was most pronounced in the sixth season when
rainfall was below normal. In the fifth and the seventh
seasons when rainfall was normal, no significant difference
was observed in soil water content under the two species.
At the end of the seventh season, soil water content at all
depths was significantly higher under cassia except at 0-20
cm depth.
In the top 0-20 cm soil depth, gravimetric water
content was generally higher near the hedgerows than away

Infiltration rate (cm/min)
172
Species
Figure 5-5. Infiltration rates of water into soil
near (0.45 m) and away (3.5 m) from alley cropped hedgerows
of Leucaena leucocephala and Cassia siamea, 1991 short-
rains .

173
Table 5-6. Saturated hydraulic conductivity of soil near
and away from alley cropped hedgerows of
Leucaena leucocephala and Cassia siamea, March
1992 (season seven).
Species
Hvdraulic conductivitv (mm hr'M
Mean
Near
(0.45 m)
Away
(3.5 m)
Leucaena
256.0
174.6b
215.3
Cassia
129.4
152.8
141.1
Mean
192.7
163.7*
178.2
Means in a row followed by the same letters are not
significantly different (p = 0.05).

174
Figure 5-6. Soil water near (0.45 m) alley cropped
hedgerows of Leucaena leucocephala and Cassia siamea
determined at the end of 1991 long-rains.

175
v
Water Content (%)
Figure 5-7. Soil water near (0.45 m) alley cropped
hedgerows of Leucaena leucocephala and Cassia siamea
determined at the end of 1991 short-rains.

176
from them (Fig. 5-8). Differences between the effects of
species was not significant. In spite of the trends of
generally higher water content near the hedgerows than away
from them, significant difference was, however, noted in
only August, November and December.
With respect to the tensiometer readings (Fig. 5-9),
no significant differences were observed between the effects
of species or position of the crop row relative to the
hedgerow on soil water potential at 0-15 cm depth. This was
so whether or not daily readings or monthly averages were
considered. As expected, however, soil moisture potentials
were higher (negative) during the dry season (e.g., August)
and less negative during the rainy season (e.g., May and
December).
During the dry season (February and June-September)
soil temperature away from the hedgerows was always
significantly higher than that near the hedgerows
(Fig. 5-10). This was so at both 10:00 hrs and at 14:00
hrs. In the dry season, the difference between soil
temperature near the hedgerows and away from them could be
as high as 3C. In the rainy season (e.g., May and October
to December), the differences were not significant.
Differences between the effects of species on soil
temperature near and away from the hedgerows were not
significant.

177
Month
Figure 5-8. Difference in gravimetric water content
(%, w/w) between soil sampled at 0-20 cm depth near (0.45 m)
and away (3.5 m) from alley cropped hedgerows of Leucaena
leucocephala and Cassia siamea in the long- and short-rain
seasons of 1991.

178
Figure 5-9. Monthly average reading of tensiometers
placed in the soil at 15 cm depths near (0.45 m) and away
(3.5 m) from alley cropped hedgerows of Leucaena
leucocephala and Cassia siamea in the long- and short-rain
seasons of 1991.

179
o
o__
0)
k_
3
4*
CD
W.
k.
0)
CO
03
CL
E
z
0)
*->
CO
=J
c
'o
CO
E
c
>
CO
a>
2
cn
<
c
CD
sz
U
Figure 5-10. Difference in mean monthly temperature
of top soil (10 cm) measured near (0.45 m) and away (3.5 m)
from alley cropped hedgerows of Leucaena leucocephala in the
long- and short-rain seasons of 1991. Vertical bars
indicate standard error of difference of means.

180
Soil Fauna
Total count of soil fauna was significantly higher
near the hedgerows compared to away from the hedgerows (Fig.
5-11). Soil fauna commonly observed included termites, wood
lice and millipedes; earthworms were absent.
Supplementary Studies
Field study: effect of irrigation and N
fertilizer on maize yield
A significant response to irrigation was observed
(Fig. 5-12). There was no response to increasing rates of N
application. These observations suggest that maize yield at
the site was limited more by soil moisture and less by soil
N.
Pot Study; effects of factorial combination
of irrigation and fertilization on yield of
maize grown on soil collected from near and
away from the hedgerows
No significant 2- or 3-way interaction effects of N
rates, water levels and soil source (i.e., near or away from
the hedgerows) were observed. The effect of soil source
(i.e., near or away from the hedgerows) on maize yield was
not significant. However, main effects of the N rates and
the irrigation levels were significant. Biomass yield of
yield of maize was not significant. Main effects of the N
rates and the irrigation levels were, however, significant.

181
Species
Figure 5-11. Counts of fauna in 25 x 25 x 25 cm top
soil monoliths taken near (0.45 m) and away (3.5 m) from
alley cropped hedgerows of Leucaena leucocephala and Cassia
siamea at the end of 1991 short-rain season.

182
6
Control (no inputs) Fert. Irrig.
Treatments
Figure 5-12. Effects of irrigation and
in combination or alone, on maize grain yield
(no alley cropping with hedgerows of Leucaena
and Cassia siamea), 1991.
b
Irrig. + Fert.
fertilization,
in the field
leucocephala

183
Biomass yield of maize in the pots increased in a
significantly linear manner with increase in N rates (Table
5-7), irrespective of the source of the soil in the pots,
i.e., near or away from the hedgerows, suggesting that the
status of the soil N near the hedgerows and away from them
was similar.
With respect to the two levels of irrigation, biomass
yield of the maize increased significantly in all pots with
increase in water (irrigation) level. Biomass yield of
maize from pots with the equivalents of 200 mm rainfall was
significantly higher than those with the equivalent of 100
mm.
Discussion
Crop Yield Profile and Explanatory Factors
The decline observed in maize grain yield with
distance from the hedgerows of both species has also been
reported by others at the ICRAF Field Station, the site of
the present study (Sang and Hoekstra, 1987; Mungai, 1991;
Ong et al., 1992). The general trend of higher yield near
the hedgerows is also comparable to observations of enhanced
production of grasses near single trees, a typical
observation in many semiarid ecozones (Radwanski and
Wickens, 1967; Belksy et al., 1989; 1992; Wilson et al.,
1990). The increase in yield nearer the hedgerows was most
pronounced in the second and third seasons. In these
f

184
Table 5-7. Effects of three nitrogen rates and two
irrigation levels on the biomass yield of maize
after ywo
1991.
months growth in
pots, seventh
season,
Nitrooen
rates
(q oot'M
Irrigation
equivalent
(nun)
0
0.64
1.3
Mean
100
32.6
40.8
53.6
42.3a
200
44.5
56.3
67.1
56.0b
Mean
38.6
48.6
60.4
49.2
NOTE: Soil in the pots was collected from near (0.45 m) and
away (3.5 m) from the hedgerows. Soil source did not
significantly affect maize biomass yield. Hence, the
means presented are averaged across soil sources.
*,bMeans in a row and column followed by the same letters are
not significantly different (p = 0.05).

185
seasons yield declined in a significantly linear trend with
increasing distance from the hedgerows. It was particularly
interesting to observe higher yield nearer the hedgerows in
the second season when rainfall was below normal (only 214
mm compared to the normal 350 mm) .
In previous studies on the reasons for microsite
enrichment by trees (Tiedemann and Klemmedson, 1977; Patten,
1978; Kellman, 1979; Joffre, 1988; Bellksy et al., 1989;
Coleman et al, 1991), higher concentrations of nutrients,
improved soil physical properties, higher microbial biomass
and densities of nematodes, lower temperatures, lower
evapotranspiration rates and higher soil moisture levels are
some of the factors attributed to the higher production of
forage under the canopy of single trees. All or a
combination of some of these factors could equally explain
the higher maize biomass yield near the hedgerows. However,
enhanced soil moisture levels appeared to be a particularly
strong factor than enhanced soil fertility (i.e.,
concentration of nutrients). Two lines of evidence support
this argument. First is the lack of significant response
to fertilizer N application in terms of maize yield whether
grown in pots with soil collected from near or away from the
hedgerows, suggesting that soil N was not a limiting factor
for maize yield. Total N or KCl extract-N contents of the
soil near and away from the hedgerows were not significantly
different. Secondly, the field irrigation studies showed,

186
as in the pot studies, significant responses mainly to
irrigation, and only little to N-fertilization. For
instance, plots with irrigation-only (no fertilizer) had
yields of maize not significantly different from those with
a combination of irrigation and fertilizer. Also yield of
plots without irrigation but with fertilizer-only and those
without fertilizer (i.e., the absolute control) were not
significantly different. Thus, soil moisture rather was the
more serious limiting factor for maize production than soil
fertility.
The hedgerows could have differentially influenced
microclimatological conditions near and away from them. For
instance, atmospheric vapor pressure deficit (VPD) could
have been reduced more near the hedgerows than away from
them. However, this difference could be small as Monteith
(1991) observed from alley cropping studies from semiarid
India. Since dry matter production: water use ratio is
inversely proportional to VPD (Monteith, 1990), it is
possible that reduced VPD near the hedgerows, resulting from
shelter and lower soil temperatures observed near the
hedgerows, could have contributed to the higher yield of
maize near the hedgerows than away from them.
The higher yield of maize near the hedgerows would
also suggest that total amount of nutrients taken up (i.e,
concentration x biomass yield) was higher near the hedgerows

187
than away from them. Therefore, the availability of
nutrients should be higher near the hedgerows than away from
them (although significant differences were not measurable).
With relatively more soil moisture content near the
hedgerows, uptake of nutrients by the plants near the
hedgerows was probably enhanced. Typically, an increase in
soil moisture results in a corresponding increase in the
uptake of nutrients (Tisdale and Nelson, 1975). Also, the
less compacted soil (low bulk density) near the hedgerows
could have contributed to an increased uptake of nutrients.
It is possible, therefore, that besides enhanced soil
moisture, the presence and/or availability of more nutrients
near the hedgerows was a contributory factor to the higher
maize yield observed near the hedgerows than away from them.
The presence of more roots near the hedgerows could have
enhanced availability of nutrients more near the hedgerows
than away from them not only through root death and
subsequent mineralization but also through the effects of
root exudates in mobilizing nutrients from insoluble
materials.
Soil Water Status Near and Away
from the Hedgerows
Top soil (0-20 cm) gravimetric water content was
generally higher near the hedges than away from them. This
was so during all months of sampling, although a
statistically significant difference was detected in only

188
three months (August, November and December). At soil
profiles below 20 cm, however, no trend was observed of
higher water content near compared to away from the
hedgerows, suggesting that the influence of the hedgerows
was probably confined more to the top soil and less to the
sub-soil.
One explanation for the lack of significant
differences in top soil (0-20 cm) water content near the
hedgerows and away from them could be the higher
transpirational loss associated with the high yield near the
hedgerows. The magnitude of water loss through
transpiration could have been large enough to result in
similar levels of soil water content near the hedgerows.
Away from the hedgerows, the yield of the crop was low,
therefore transpirational loss was equally low. Also,
complete or partial closure of stomates, an adaptive
mechanism of plants under water stress (Slatyer, 1967),
could have minimized losses of water from soil distant from
hedges. If these physiological processes were indeed
operational, their net effect could result in similar levels
of soil water content near and away from the hedgerows.
Also, the lack of significant differences observed in the
relative water content of plants (or of leaf discs) near the
hedgerows and away from them is probably a reflection of the
high and low consumptive use of soil water by the crop near
and away from the hedgerows.

189
While trends (at times significant) of higher water
content near the hedgerows were detected with the
gravimetric method, similar trends were not observed with
the tensiometers placed at 15 cm depths. Tensiometers
measure water potentials of the soil at or close to the
point of placement of the porous cup tip. They do not
measure or include the water content of the soil above the
depth of placement. On the other hand, the gravimetric
method integrates the whole depth of the soil considered.
The differences in the mode of operation of the two methods
could explain the apparent discrepancy in the results
obtained with the two methods. The differences also
suggested that gravimetric methods would be recommendable
over tensiometer methods for soils with limited water
content. This is because the gravimetric method integrates
more volume of soil and hence is sensitive enough to detect
small quantities of soil water rather than tensiometers.
Water status near the hedgerows could have been
enhanced in several ways. High infiltration and hydraulic
conductivity were probably the principal processes. Other
studies also show that the hedgerows could physically
intercept as much as 20% of the season's rainfall (Monteith
et al., 1991), especially where the hedgerows are planted
perpendicular to the direction of the prevailing winds.
Under conditions of moisture stress, rainfall intercepted by
the hedgerows could increase yield of the hedgerows and,

190
therefore, increase yield of crops adjacent to the hedgerows
while that of the distant crop is reduced. In addition, the
protective effects of hedgerows could reduce soil loss and
runoff (Rao et al., 1991) and could enhance soil water
content near the hedgerows. It is feasible that the
hedgerows had significant barrier effects since the plots of
the present study were on land with a 3%-5% slope.
Reduced runoff has been suggested by others to explain
an increased yield of crops observed near hedgerows on
sloping lands (Sang and Hoekstra, 1987). Reductions in
runoff under hedgerows, however, could be of minor
importance compared to the competition for soil water
between the roots of trees that are established by the time
the crops are sown (Singh, 1989). A process referred to as
hydraulic lift (Caldwell and Richards, 1989), which involves
absorption of water by deep roots in moist soil, upward
movement through the roots, and release in the upper soil
profile at night, could also be a mechanism by which water
content near the hedgerows was enhanced.
Reduced soil temperature as a result of shade from the
hedgerows could also have contributed to the trends of
higher soil water content near the hedgerows than away from
them. Shade would create differential rates of evaporation,
with higher rates in the open and lower rates near the
hedgerows. This would imply that less water was available
to crops in the open alleys and more to those near the

191
hedgerows. Difference in soil temperature near and away
from the hedgerows was significant at the end of the
cropping period (February, March, June and July) and during
the dry season (August and September). At this time, the
hedges are typically not cut back since the crop was either
mature or was not present in the field. Hence, considerable
shade could build up near the hedgerows that, at times, grew
up to as high as 2 m. During the cropping season, however,
there was no significant difference in soil temperature near
and away from the hedgerows because the hedgerows were cut
back 2-3 times in a season.
Also, cloudy conditions prevailing during the rainy
season could have minimized differences in soil temperature
both near the hedgerows and away from them. So, although
effects of reduced temperature (and therefore reduced
respirational losses) near the hedgerows have been suggested
to explain the higher crop yield near the hedgerows (Mungai,
1991), it is unlikely to have been a major factor in the
present study. The basis for this assertion is that there
were no significant reductions in soil temperature when the
crop was present in the field. A similar conclusion is made
by Corlett et al. (1989) who observed no significant
differences in the canopy temperature of sorghum alley-
cropped with leucaena and sorghum grown in the open at
semiarid conditions at Hyderabad, India.

192
Effects of shade from the hedgerows on soil
temperature may be significant where the orientation of the
hedgerows is north-south, where the pruning frequency of the
hedgerows is low and at sites far from the equator as
observed by Monteith et al. (1991). In the present study,
the hedgerows were east-west oriented, they were frequently
pruned (2 to 3 times in a season) and the site was close to
the equator (Io S 33' E). It is therefore unlikely that the
effects of shade from the hedgerows on soil water content,
particularly during the cropping season, were significant.
During the dry season, however, it is possible that even a
small amount of soil water conserved due to shade effects
from the hedgerows, could have contributed to the enhanced
yield of the crop near the hedgerows in the following
season.
The significantly higher rates of infiltration and
hydraulic conductivity observed near the hedgerows could
have resulted from the lowered soil bulk density observed
near, compared to away, from the hedgerows. Topsoil (0-5 cm
profile) bulk density was significantly reduced near
leucaena hedgerows. Reduction in soil bulk density could
have resulted from the death and turnover of the higher root
density near the hedgerows (Chapter 2). Sloughed-off roots
add to soil organic matter (humus), which in turn, could
reduce soil bulk density. Reduced soil bulk density
improves aeration, stabilizes structure, and increases

193
permeability (Swift and Sanchez, 1984). Moreover, the
activity of the significantly higher population of fauna
(predominantly termites, beetles and wood lice) near the
hedgerows could have contributed to the lower soil bulk
density observed near the hedgerows compared to away from
them. As observed by Coleman et al. (1991), the role of the
fauna in reducing soil bulk density and enhancing
infiltration rates of water appeared to be important in the
present study. For example, from a subsidiary study, it was
observed that the rates of hydraulic conductivity of the
soil declined significantly in spots near the hedgerows
where a pesticide was applied repeatedly compared to where
the pesticide was not applied (data not presented). These
observations indicated the significant role of the fauna on
the physical properties of the soil.
Differences between Effects of Hedgerow
Species on Crop Yield Profile
The significantly lower yield of maize under leucaena
alleys compared to those under cassia demonstrates that
leucaena is more competitive with crops than cassia, an
observation also made by Nair (1987). Greater depletion of
soil water by leucaena compared to cassia best explained the
lower maize yield under leucaena compared to cassia. A
similar explanation was also made by Singh et al. (1989) who
observed that alley cropping leucaena in semiarid India
reduced crop yields significantly.

194
In addition to soil moisture, it is also possible that
crop yields were reduced more under leucaena compared to the
sole crop or the crop under cassia due to competition for
nutrients. A basis for this hypothesis is provided by
studies by Chirwa (1991) in semiarid Zambia, where
reductions in maize yields under leucaena hedgerow
intercropping were observed to be less with the application
of fertilizer and more without fertilizer application.
The lower content of soil water under leucaena
compared to cassia would suggest that leucaena has a higher
root density than cassia. This was, indeed, observed to be
so. Root density of leucaena was higher than that of cassia
at nearly all distances considered (up to 3 m away from the
hedgerows) in the top 40 cm soil profile (Chapter 2).
Studies by Jonsson et al. (1988) also showed that both
leucaena and maize had their highest density of fine roots
in the same soil depth, suggesting that the potentials for
competition are high.
Two implications emerge from the observed differences
in the rooting density and water depletion patterns of the
two species. First is that leucaena would allow less
intercropping than cassia. This is particularly so in
semiarid areas where water, and to a lesser extent, soil
fertility, are major constraints to production. Yield of
crops with significant amount of their roots in the 0-80 cm
soil depth could be adversely affected under leucaena due to

195
enhanced competition. Unfortunately, most of the annual
crops have their roots within the top part of the soil
profile. Crops that have either short periods of maturation
(e.g., cowpea) or that are drought tolerant (e.g., millet)
may provide opportunities for intercropping with leucaena
where this may still be necessary. Secondly, through root
turn-over, particularly of the fine roots, leucaena would be
likely to add more to soil organic matter than cassia. This
may improve both physical and chemical properties of the
soil. For instance, hydraulic conductivity of the soil
could be enhanced by macropores opened following the death
of roots (Barley, 1954), and more roots are under leucaena
than under cassia. But these benefits of soil improvements
under alley cropping with leucaena seems to be nullified by
enhanced tree-crop competition for soil moisture under
semiarid conditions.
Conclusions
This study demonstrates that:
1. Although the effects may be small, interaction
between tree hedgerows and crops could be
positive, even in semiarid environments.
2. The magnitude of the interaction effects on the
crop yield is dependent on the hedgerow species,
and more positive under cassia than under
leucaena.

196
3. Enhanced soil moisture near the hedgerows, more
than improved soil fertility, could explain the
higher maize yield observed near the hedgerows.
The hedgerows, particularly that of leucaena, improved
soil bulk density near them. This, in turn, enhanced the
rates of infiltration and hydraulic conductivity of the
soil. If the observed positive effects of the hedgerows on
the crop and the physical properties of the soil near them
are not site-, species-, and management-specific, then alley
cropping of multipurpose trees, particularly cassia, may
have a role in improving productivity of soils in the
semiarid tropics.

CHAPTER 6
POTENTIAL FOR FARMER ADOPTION OF ALLEY CROPPING
OF MULTIPURPOSE TREES AND SHRUBS IN SEMIARID
AREAS OF MACHAROS, KENYA
Introduction
Agroforestry is often suggested as a major practical
land management alternative for the maintenance of soil
fertility and productivity of lands in the tropics (Nair,
1989), and, in particular, the marginally productive ones,
including the semiarid tropics (Rocheleau et al., 1989).
Among the agroforestry technologies that have received
tremendous research and development effort is alley cropping
(also called hedgerow intercropping). Alley cropping
consists of growing food crops in alleys between planted
hedgerows of multipurpose trees and shrubs, usually
nitrogen-fixing ones. The hedgerows are periodically cut
back at the beginning as well as during cropping to prevent
shading and to provide mulch to the associated crop (Kang et
al., 1990).
Results of alley cropping (at least on-station) in the
humid tropics with respect to soil fertility and crop
performance have generally been positive (Kang et al.,
1990). Compared to the humid lowland tropics, there are
relatively few studies on alley cropping in the semiarid
tropics. Rainfall and soil conditions aside, the impacts of
197

198
alley cropping trees on the yields of crops at any given
site are likely to be influenced by the tree species used.
For instance, Sang and Hoekstra (1987) reported improved
crop yields under cassia in semiarid Kenya; in contrast,
Singh et al. (1989) reported significant yield losses under
leucaena in semiarid India.
In a comprehensive review of agroforestry, Nair (1990)
concluded that alley cropping as an agroforestry practice
has some advantages, but some inputs are required. While
there are certainly many biophysical issues of the
technology in both semiarid and humid tropics still at
stake, perhaps most are on aspects related to the socio
economics of the technology. Understanding of the socio
economics is the key to the adoption of a technology
anywhere. Without this understanding, the technology may
never leave the researcher's desk, as is typical of many on-
station studies (Hildebrand, 1982). For instance, alley
cropping research has been going on at the International
Institute of Tropical Agriculture (IITA), Ibadan, Nigeria,
for at least 10 years (Kang et al., 1990). Yet, outside
IITA research stations, there is little evidence of adoption
of the technology by farmers for purposes of mulch
production on crop fields (Palada, 1989).
It is hypothesized the more simple the technology, the
more rapidly farmers can become proficient with it and adopt
it (Wake et al., 1988; Hildebrand, 1990). However,
agroforestry technologies (e.g., alley cropping) are by

199
nature complex, (at least two components are involved, a
crop and a tree/shrub) and, therefore, may be more difficult
to adopt than systems involving a single component (tree or
crop). Simplicity or complexity of the technology
notwithstanding, adoption of alley cropping may be higher
where the trees provide multiple and valuable products,
e.g., fodder. This is a particularly important
consideration in the semiarid tropics where livestock feed
is scarce during the dry season, and trees often provide
valuable fodder (Singh et al., 1986).
Understanding of the existing farming systems of the
area, production priorities and constraints of the farmers
have long been recommended as an absolutely necessary input
into the agenda and process of on-station technology
development (Spedding and Brockington, 1976). Where these
recommendations have been followed, especially through the
iterative Farming Systems Research and Extension
methodology, there is evidence to suggest that the
potentials for farmer acceptance and adoption of
agricultural technologies are high (Hildebrand, 1981). On-
farm trials, either researcher-managed and/or researcher-
farmer managed (Hildebrand and Poey, 1985; Atta-Krah and
Francis, 1987) are among some of the research approaches
used to enhance farmer participation and evaluation of
technologies being developed. Although certainly better
than the total reliance on on-station trials, the on-farm

200
trials also have limitations. For instance, loss of farmer
participation is sometimes cited as a major problem of on-
farm trials (Lightfoot, 1987).
Because of the long-term nature of agroforestry
practices, strategies different from those of systems
involving only annual crops may be necessary to enhance
adoption. For example, conducting farm surveys specific to
a given agroforestry practice (e.g., alley cropping) may
help capture details of the farming systems of an area that,
when included in the development of the technology on-
station, could enhance its potential for adoption by
farmers.
Being cognizant of the above issues, as well as
fulfilling a need to rationalize an on-going on-station
experiment on soil fertility and productivity aspects of
alley cropping Leucaena leucocephala and Cassia siamea at
the Field Station of the International Center for Research
in Agroforestry (ICRAF), Machakos, Kenya, I conducted an on-
farm survey. Sixty farmers in the neighborhood of the Field
Station were interviewed in July-August 1991. The overall
objective of the study was to assess the potential for
farmer adoption of alley cropping of multipurpose trees and
shrubs.

The specific objectives were: (a) to identify current
strategies employed by farmers to improve the fertility and
productivity of their soils, (b) to determine factors

201
significantly affecting the above and, (c) to identify the
potential uses and constraints of farmers in alley cropping
multipurpose trees and shrubs.
Study Area and Methodology
The study was conducted on farms that lie on the hills
immediately south of Machakos town, Eastern Province, Kenya.
The farthest farm from Machakos town was about 20 km. The
ecology of the area is semiarid with a mean annual rainfall
of about 600 to 700 mm, falling in two seasons: long-rains
(March to July) and short-rains (September to October).
Besides meeting the criteria of being semiarid, the area of
the study was close to the Field Station of ICRAF where
alley cropping research was going on. Thus, as one might
either expect or assume, farmers in this area would be
better informed about and potentially more adoptive of the
technology than distant farmers. ICRAF Field Station is 6
km west of Machakos town.
Three administrative sublocations were selected for
sampling: Kiima-Kimwe, Muvuti and Kivandini. All three
sublocations were constituents of Muvuti location, Central
Division, Machakos District, Kenya. The area is
topographically hilly and the farms were scattered on the
hill slopes.
Based on the outcome of a three day rapid rural
appraisal (Chambers, 1985) or Sondeo (Hildebrand, 1981)

202
exercise, stratification of the farms relative to their
location on the hills (low, middle and top slopes) was felt
necessary. One reason for stratification was that
infrastructure in the area was strongly influenced by slope
factors of the hills; farms on the low and middle slopes
were more accessible than farms on the upper slopes. In
addition, farms on the top slopes and particularly those on
the windward side had more rainfall (on average, 1000 mm a
year) than farms on the low slopes (with about 700 mm
rainfall a year). On the basis of the above stratification,
a list of farmers of the three administrative sublocations
was obtained from the Ministry of Agriculture staff in
charge of each of the three sublocations. However, in the
Kiima-Kimwe sub-location, an extension agent of the Catholic
Diocese of Machakos, helped provide a more comprehensive
list of farmers. This was because of the inadequacy of the
list provided by the agricultural office of this location.
Twenty farms were randomly selected from the list of
farmers of each sublocation. Thus, a total of 60 farmers
for the three locations were selected. Interviews were
conducted on single-day visits in the months of July and
August, 1991. A one-page questionnaire was used as a guide
during the interviews. Interviews and visits at each farm
lasted about an hour. The interview team was made up of the
author and the extension agent of the Catholic Diocese of
Machakos who also helped with the interpretation of the

203
Akamba language. The area is predominantly populated by the
Akamba people. Interpretation was not always necessary
because a large number of the farmers spoke either Swahili
and/or English which the author also speaks. On average,
five farmers were interviewed in a day.
At the conclusion of the survey, of the 60 farmers, a
second stratification or regrouping of the farms was carried
out. The farmers were grouped according to their identified
interests in the use of the pruning from the alley cropped
hedges. Basically, there were two distinct use-interests:
fodder or mulch (soil fertility). It is recognized that the
fodder-interest group could also have mulch as their
secondary interest in the use of the pruning from the
hedgerows and vice versa for the mulch group. Hereafter,
the two groups will be referred to as fodder-interest and
mulch-interest groups. On the basis of these two major use-
interests of the alley cropping technology, three farmers of
each interest group were randomly selected from the low,
middle and upper slopes of each of the three sublocations.
Thus, a subsample of 18 farmers out of the initial sample of
60 farmers was generated. The purpose of this second
stratification or subgrouping was two-fold: (a) to get more
in-depth information of the farming practices and the other
objectives of the study, and (b) to obtain a more informed
evaluation of the technology from the farmers. This sub
group of 18 farmers were taken to ICRAF Field Station over a

204
three day duration. At the station, the fanners were given
a guided tour of the trials where pruning were used as mulch
or as fodder. During the tour, questions and concerns
raised by the fanners were recorded.
Statistical Analysis
Frequency tables of responses to questions were
generated using the SAS (SAS, 1987). Means and their
standard deviations (sd) of various variables are presented.
Dependency or independence of fanner interests in the use of
the pruning of the alley cropping technology (i.e., fodder
or mulch) on any of the variables considered in the survey
was determined using the Chi Square test at the 80%
confidence level.
Results
Farm Household Characteristics and
Farmer Interests in use of Aliev Cropping
Farm size of the farmers interviewed varied widely,
ranging from 0.4 hectares to 17.4 hectares (Table 6-1).
Seventy-seven percent of the farms were in the 0-4 ha size
range; of these, 46% had two or less hectares. Farm size
with the largest frequency distribution (15%) was three
hectares. The sizes of farms can therefore be classified as
small scale. Yet, they support a large number of
dependents. On average, there were 14 family members per

205
Table 6-1
Farm and household characteristics of the 60
farmers interviewed at Machakos, Kenya, August
1991.
Farm
size
range
(ha)
Cumu
lative
distri
bution
(%)
Average
number of
members
in a
house
hold
Average
number of
children
in a
household
Household
members
that were
children
(<18 yrs)
(%)
0-4
77
13 (4)
7 (4)
54
5-8
17
16 (3)
8 (3)
50
9-17
.4 6
14 (6)
7 (5)
50
NOTE:
Figures in
parentheses are
the standard
deviations
of the mean.

206
household? over 50% of the family members were children
under 18 years of age. Land ownership was family free-hold
title basis.
Of the 60 farmers interviewed, 62% were interested in
fodder and 33% in mulch. Three farmers (5%) had no interest
in alley cropping or were undecided. Farm sizes of the
fodder- and mulch-interest groups were not significantly
different, on average three hectares. Also, farm size did
not influence significantly (p = 0.67) the fodder or mulch
interests of the farmers interviewed (Table 6-2). The labor
requirements of alley cropping is often seen as a bottleneck
to its adoption. However, most of the farmers (88%) did not
see labor for pruning the hedgerows as a problem.
Maize was the main food crop, grown alone or in
combination with legumes (predominantly pigeon pea). All
farmers expressed insufficiency of food supply. Indeed, 67%
of the farmers interviewed bought their staple food (maize)
at one time or another. Between the two groups, 80% of the
mulch-interest group and 57% of the fodder-interest group
farmers bought food. Other food crops included beans and
various vegetables. Cash crops were coffee (mainly in two
locationsKivandini and Kiima Kimwe) and fruit trees
(mango, guava, citrus and avocado).
Farmers kept livestock (cows, goats, sheep and
chicken). In terms of cattle, the average number per farm
was two head (sd: 4), with a range from 0 to 12 head per

207
Table 6-2.
Interests of farmers in alley
size and number of members in
Machakos, Kenya, August 1991.
cropping, farm
a household at
Interests
Total
number of
farmers
inter-
Farm
Number
of members
in alley
viewed
size
in a
cropping
(%)
(ha)
household
Fodder
62
2.8 (1.5)
14 (3)
Mulch
33
3.1 (2.1)
13 (6)
Undecided
5
2.5 (1.3)
12 (4)
NOTE: Figures in parentheses are the standard deviations
of the means.

208
farm and mode of 0 head (35%). The fodder-interest group
had, on average, 30% more cattle than the mulch-interest
group. However, the number of livestock kept by both groups
(fodder- and mulch-interest group) were not significantly
different. Although 35% of the farmers interviewed did not
have cattle, some had immediate plans for buying them.
In terms of goats, the average holding was nine (sd:
8), with a range from 0 to 30. Twenty percent of the
farmers interviewed did not have goats. Differences in goat
possession of the fodder and mulch-interest groups were not
significant. However, the fodder-interest group had, on
average, 25% more goats than the mulch-interest group. For
sheep, the number possessed ranged from 0-13 with an average
of four (sd:4) sheep per farm. Eighty-three percent of the
farmers did not have sheep.
Farmers kept more chickens than other livestock.
Ninety-three percent of the farmers interviewed kept
chickens probably because of their relatively lower feed
requirements compared to other livestock. The range of
chickens kept ranged from 0 to 2000 birds. The average,
when the single farmer with 2000 birds was excluded as an
outlier, was 21 birds (sd: 21); the mode 20 birds (18.3% of
the farmers interviewed). Seven percent of the farmers
interviewed had no chickens. The difference in chicken
possession of the fodder and mulch-interest groups was not
significant (p = 0.50).

209
Strategies of Farmers to Improve
Fertility and Productivity
Nine practices were identified as the main strategies.
They will be discussed individually, but not in order of
importance.
Use of livestock manure
This was the major source of nutrient input for the
crops on most farms. All farmers interviewed used livestock
manure on their crop fields. Manure was applied on the crop
rows and not broadcasted on the whole field. Row
application of manure economized use of this scarce
resource. The main source of manure was their own farm.
Between the fodder and mulch-interest groups, the percentage
of each group that bought manure was similar, approximately
%
39%. ;
Manure from cattle owned was scarce, given that each
farm had, on average, only two cattle. Occasionally,
farmers with and without livestock purchased manure.
Thirty-five percent of the farmers interviewed did not have
cows and 37% bought manure at one time or another. Often
manure was bought from a few farmers with large broiler
houses within the area of the present study or from cattle
ranches 40-50 km away. Lack of feed and loss of livestock
during previous droughts were the reasons given for the low
number or absence of cattle on farms. Crop residues were
the major source of feed and farmers (92% of those

210
interviewed) let their livestock into the crop fields after
harvest.
Weed control
On average, crops were weeded twice in a season.
Labor for weeding was mainly family labor. However, labor
was often hired or exchanged with neighbors. The difference
in the frequency of weeding by the two interest groups
(fodder and mulch) was not significant.
Sixty-eight percent of the farmers interviewed
responded positively to exchange of labor with other members
of the community. Forty-two percent of this exchange labor
was for help with weeding. Only a small percentage, 3% and
4%, respectively, of the exchange labor was for crop
planting and harvesting. Farmers who did not exchange labor
often cited exchange of labor as becoming increasingly more
expensive in terms of providing for food for work done.
In terms of hiring labor, 60% of the farmers
interviewed hired labor to help with weeding and harvesting.
However, labor for harvesting was mainly for picking coffee.
Seventy-eight percent of the fodder-interest group and 50%
of the mulch-interest group hired labor. This difference
was significant.

211
Fallowing of land
Because the farms were generally of small acreage,
only 18% of the farmers interviewed mentioned fallowing as a
practice. The difference in the use of land fallowing
between the fodder and mulch interest groups of alley
cropping was not significant (p = 0.54).
Use of mulch
Use of mulch was common. Typically, mulch materials
were crop residues, lopping from trees on the farms and
prunings from farm fence bushes. Many farms had Lantana
caara and Tithonia diversifolia bushes as live fences.
Prunings from these shrubs provided fodder and firewood when
dry and mulch when green. When fresh, lantana could be
poisonous to cattle. Sixty-seven of the farmers interviewed
responded positively to the use of mulch. Of this number,
60% were those with fodder interests and 85% were those with
mulch interests. This difference between the two groups was
significant.
Intercropping
Forty-five percent of the farmers interviewed
practiced intercropping. Of these, the percentages of
fodder and mulch-interest groups were similar (46%).
Intercrops of maize and beans (both phaseolus and vigna) or
maize and pigeon pea were common. Of the two, maize and

212
pigeon pea intercrop was the more preferred. The beans in a
maize and bean intercrop ripened earlier (in about two
months) than the maize (four months). Early ripening of
beans made their harvest difficult in the midst of the maize
stands. Also, in the process of harvesting beans, damage to
the immature maize crop could be severe. Beans in a maize
intercrop were also observed by 52% of the farmers
interviewed to have competitive effects. On the other hand,
pigeon pea is an perennial crop and its harvest occurs long
after maize is out of the fields.
The presence or absence of intercropping appeared to
be influenced most by farm size. Intercropping was
predominantly a practice on the small farms. Of those
farmers that intercropped, 60% had farm sizes of two or less
hectares. On the other hand, 85% of those farmers who did
not intercrop had farm sizes greater than two hectares.
Use of improved seeds
Farmers bought seeds, in particular those of the
drought tolerant maize varieties (e.g., Katumani Composite).
Sixty-seven percent of the farmers interviewed bought seeds.
Also, the purchase of seeds was mentioned by most farmers to
be the one farm input that required most money. Farmers who
did not buy seeds saved some from their harvests. Of those
purchasing seeds, the percentages of the fodder and the

213
mulch-interest groups were similar (65%). The purchase of
seeds was not influenced significantly by farm size.
Use of chemical fertilizers
From the sample survey of 60 farmers, 65% used
chemical fertilizers on their crops at the time of the
survey or in prior seasons. Farmers with coffee crops
responded more positively to use of fertilizers than farmers
without coffee. Between the two groups, 76% of the fodder-
interest group and 55% of the mulch-interest group used
fertilizers. This difference in fertilizer use of the two
groups was significant. However, only 3% of the farmers
interviewed mentioned chemical fertilizers as their most
expensive farm input, suggesting that use of chemical
fertilizers was not a major practice. Also, chemical
fertilizers, unlike seeds, were not regularly purchased.
For farmers with coffee crops, fertilizer was obtained at a
subsidized cost from the coffee factories.
Use of ox-plow
A significant number (67%) of the farmers interviewed
used ox-plows for land preparation and/or for weeding. At
times, those who did not own the oxen owned only a plow, and
vice versa. This often necessitated an exchange of the ox-
plow parts between farmers. Possession of both the oxen and
the plow were expensive for individual farmers. Some

214
fanners had only the plow because their oxen died for some
reason (e.g., during a major drought in 1987). In some
isolated cases, farmers had sold their oxen to meet some
urgent needs for cash. Between the two interest groups, 72%
and 35%, respectively, of the fodder- and mulch-interest ox-
plowed their fields. This difference was significant.
Use of soil erosion control measures
The common measure of soil erosion control was
terracing with soil mounds thrown on the upper side of cut
off trenches. This practice is locally called "Fanya Juu."
It is an expensive exercise to undertake initially, but it
is also an effective erosion control measure that requires
little maintenance thereafter. A significant proportion
(70%) of the farms surveyed were terraced well with the
"Fanya Juu." Differences in the percentage of terraced
farms among the fodder- and mulch-interest groups were not
significant.
Factors that Influenced Interests
of Farmers in Aliev Cropping
Five factors significantly influenced the proposed use
of the alley cropped trees for fodder or for mulch:
1. Possession of coffee trees,
2. Possession or use of ox-plow(s),
3. Use of mulch,
4. Use of fertilizer and,

215
5. Prior knowledge of alley cropping.
Table 6-3 provides frequency distribution of farmers under
each of these factors against the interests of farmers in
alley cropping.
Use of ox-plow and fertilizer by the fodder-interest
group was significantly higher than by the mulch-interest
group. On the other hand, frequency of farmers using mulch
on fields was significantly greater for the mulch-interest
group than the frequency for the fodder-interest group.
The fodder-interest group was exposed more to alley
cropping (45%) than the mulch-interest group (35%), although
the difference was not significant (p = 0.42). This
difference in prior knowledge of the technology did
influence significantly the farmers' proposed use of the
technology. Between the two interest groups, 67% of the
fodder-interest group had coffee and only 35% of the mulch-
interest group had coffee. This difference was significant.
The use of chemical fertilizers, hired labor, ox-
plows, purchased manure, purchase of animal feed, frequency
of weeding, and purchase of food were all observed to be
significantly greater for the coffee than for the noncoffee
farmers. All the other factors considered (e.g., farm size,
number of livestock, number of household members doing off-
farm work, etc.) were not significantly dependent on
possession of coffee. Table 6-4 provides the p-values of

216
Table 6-3. Factors which influenced significantly interests
of fanners in alley cropping for fodder or for
mulch and the frequency of farmers under each
factor, Machakos, Kenya, August 1991.
Interests
in alley
cropping
Use
of ox
plow
Use of
crop/trees
mulch at
present
or in the
past
Use of
fertil
izer
Posses
sion of
coffee
trees
on the
farm
Prior
know
ledge
about
alley
cropping
Fodder
27
22
21
37
13
Mulch
7
12
5
20
9

217
Table 6-4,. The p-values of dependency of resource
possession, their use or purchase on the
presence of coffee on the farm, Machakos, Kenya,
August 1991.
Resource
p-values of dependency
of resource
use/possession
on the presence
of coffee on
the farm
Purchase of food
0.07
Purchase of chemical fertilizer
0.04
Hire of ox-plow
0.01
Purchase of manure
0.17
Purchase of animal feed
0.13
Hire of labor
0.06
Frequency of weeding
0.04

218
resources possessed or purchased that depended significantly
on farmer's possession of coffee. Even though the use of
fertilizer was significantly greater for coffee than for
noncoffee farmers, maize yields of the two groups were not
significantly (p = 0.35) different (Table 6-5). This
suggests that fertilizer was probably applied mainly to
coffee.
Factors Affecting Farmers' Strategies
to Improve or Maintain Fertility and
Productivity of Their Soils
Most of the strategies discussed above reguired cash
outlay. For example, "Fanya-Juu" terracing was expensive
and a major investment on many farms. Between the two
interest groups, the fodder-interest group evidently had
more purchasing power for farm inputs and household needs
(e.g., fertilizer, ox-plows, labor, feed for livestock, food
for the family, and firewood) than the mulch-interest group.
There were two major sources of cash income in the
area, viz., sale of coffee and off-farm employment. Many of
the families interviewed had one to two of their members
working off-farm. Forty-four percent of the farmers
interviewed had at least one member of the family engaged in
regular off-farm employment. Most of those engaged in off-
farm employment were men since 60% of the farmers found on
the farm during the survey were women.

219
Table 6-5. Estimates of maize yield (with or without
inputs) of farmers with interests in fodder or
mulch aspects of alley cropping.
Interests
Maize
vield ft ha'1
in alley
cropping
With
inputs
Without
inputs
Mean
Fodder
0.4 (0.2)
0.2 (0.1)
0.3 (0.2)
Mulch
0.3 (0.2)
0.1 (0.1)
0.2 (0.2)
NOTE; Figures in parentheses are the standard deviation of
the mean.

220
Coffee was the major cash crop of the area. Fifty-
seven of the farmers interviewed had coffee. Among the
coffee farmers, 76% were interested in the fodder aspects of
alley cropping. The difference in the frequency of farmers
without coffee but interested in alley cropping for fodder
or for mulch was not significant.
In addition to off-farm employment, other sources of
cash were sale of livestock (e.g., poultry), baskets (mainly
made of sisal), fruits, charcoal, and other tree products.
Trees on the Farm: Functions and
Constraints to their Planting
Predominant species
Both indigenous and exotic species (fruit and non
fruit) were present on the farms. On an average farm size
(about two hectares), 10%-20% of the land could be under
trees (both fruits and nonfruit trees). Typically, only
fast growing exotic species good for pole and firewood
(e.g., Eucalyptus camaldulensis and Grevillea robusta) were
planted. Indigenous or naturalized species (e.g., Acacia
nilotica, Croton megalocarpus and some Commiphora spp.) were
not planted but left to grow from volunteer seedlings and
protected from damage by animals and termites. These
species were found mainly on the lower slopes of the hills.
On the cooler upper slopes, Acacia mearnsii (wattle tree)

221
was common. Also, both cassia and leucaena were present,
but only on a few farms on the lower and middle slopes.
Fruit trees, e.g., mango, citrus, avocado, guava, and
loquat were raised on the farm, on land cropped and not
cropped. With the exception of mangoes and citrus, most of
the fruit trees were planted near the homestead. Citrus was
generally planted inside the "Fanya Juu" or soil
conservation trenches. The trenches acted as macro
catchments for water and provided more water to the citrus
plants than planting in the open.
Uses of tree products
Trees provided many products and services. Firewood
and building materials were two uses often mentioned by the
farmers. Because firewood was scarce, trees with good
firewood properties and with fast growth rates (e.g.,
Eucalyptus) were in high demand. With the exception of a
few farmers that bordered government land reserves, most did
not have access to firewood beyond their farms. Of the
farmers interviewed, 45% bought firewood. Between the two
farmer interest groups in alley cropping, 57% and 25% of the
fodder- and mulch-interest groups, respectively, bought
firewood. The difference between the two groups was
significant.
Like firewood, fodder was also scarce. Forty-five
percent of the farmers interviewed bought livestock feed.

222
Trees and shrubs also provided fodder, particularly during
the dry season. For example, the dry leaves of Lantana
camara and Tithonia diversifolia hedges were used as fodder.
Some farmers with leucaena trees for shade near the
homestead occasionally lopped the trees for fodder.
If available, fruit trees were number one choice of
farmers in favor of all other trees. Fruit trees provide
food and cash to the family. Also, some fruit trees such as
mangoes were lopped for firewood. Besides mangoes which
were scattered on the cropland, nonfruit trees (e.g., Acacia
spp. ) left deliberately on cropped land provided shade
during field activities (plowing, planting, and harvesting).
Constraints to tree planting
Ability to establish easily from seed or seedling,
particularly under arid and competitive condition from weeds
and other plants was an important consideration of farmers
in the choice of the tree to plant. Equally important was
the ability of trees to withstand termite attack. A
significant number (86%) of the farmers interviewed
mentioned termites as problems in establishing trees.
Wood ash was predominantly the traditional termite
control measure used. Some farmers said wood ash controls
termites but many also said it does work or works only
temporarily. The seriousness of the termite problem was
reflected in the choice of species planted or allowed to

223
grow on the farms. With the exception of Eucalyptus
species, a popular tree among farmers, many of acacia
species and croton are termite resistant. Hence, survival
was high. Eucalyptus, on the other hand, is not resistant
to termite attack. But because Eucalyptus seeds heavily,
farmers plant many seeds and hope that some plants survive.
And, indeed, its survival was highly variable, with an
average of 53% (sd: 38%) across the farmers interviewed.
Availability of planting materials was cited by many
farmers as a big constraint. To ensure their own supply of
seedlings, 30% of the farmers interviewed had at one time or
another a nursery on the farm. The functions of the tree
nurseries possessed are shown in Table 6-6. Farmers
established nurseries because of the nonavailability of
seedlings from other sources. Government and non
governmental nurseries were either too far away, not
dependable or often did not carry species desired. In
addition, without one's own transport, one would not be able
to carry many seedlings up the hills. However, 70% of the
farmers interviewed did not possess a nursery; the reasons
are shown in Table 6-7.
The major problems of raising one's own nursery/-
seedlings are lack of water and availability of seeds. A
number of farmers said they were planting trees to have a
secure source of seeds. Because the area is semiarid, a
sufficient supply of water is a problem. Most of the

224
Table 6-6.
Functions of tree nurseries possessed by 30% of
the 60 farmers interviewed, Machakos, Kenya,
August 1991.
Function of
nursery
Distribution of farmers
with tree nurseries
(%)
Fruit, Firewood, Timber
11.7
Fruit
18.3
Total
30.0

225
Table 6-7. Reasons given by farmers without nurseries on
their farms, Machakos, Kenya, August 1991.
Reasons
Distribution of respondents
(%)
Water
26.7
Seeds
13.3
Not decided
11.7
Lack of knowledge
3.3
Planning
1.7
Pests (e.g., chickens)
1.7
Termites
5.0
Labor
6.7
Total
70.1

226
fanners with nurseries were close to some water source,
e.g., a seasonal river.
Discussion
Strategies of Farmers to Enhance
Soil Fertility and Crop Production
Production of sufficient food for the family was
probably the major goal for every farmer. This is based on
the observation that households were large, averaging 14
family members, 50% of which were children. In spite of the
various strategies employed by the farmers to improve or
sustain food production (intercropping, weeding,
fertilizing, etc.), crop yields were low. This was more so
without inputs (i.e., over 50% reduction of that with
inputs). This is perhaps a reflection of the low fertility
status of the soils and insufficient rainfall. Since the
survey was conducted at the end of a season with poor rains,
the contribution of soil fertility to the low yields may not
explain the dramatically reduced yields. Under seasons of
normal rainfall, yields are more likely to be higher with or
without inputs compared to those observed during the survey.
However, estimates of on-farm maize yields less than
1 t ha'1 are typical even in seasons of normal rainfall
(Nadar and Faught, 1984; Mbogo, 1991).
To increase food production, farmers with limited land
were not renting more land to produce food. Instead, they

227
were buying food. Fanners with large land holdings were
also buying food and their crop yields were not
significantly different from those with small holdings of
land. These observations suggest that food production was
constrained perhaps less by land availability than by cash
to buy inputs (fertilizer, seeds, manure, ox-plow) or by
labor. These observations are consistent with studies of
others from the region of the present study (Myers, 1982;
Harsh and Mbatha, 1978).
Given that 76% of the farms were two hectares or less,
the interest of the farmers with such small holdings in
putting some land under alley cropping is surprising. In
addition to supporting the view that land was not
constraining crop production, farmers' positive response to
alley cropping also suggests that farmers saw the practice
as a means to increase production of food or fodder from
their small farms. Farmers probably saw alley cropping to
be similar to intercropping of food crops, which was
predominantly a practice of farmers with small land holdings
(two hectares or less). In addition to increasing
production per unit area, intercropping is done for many
other reasons, such as higher returns to management,
benefits of diversity, even if less productive (thus,
reducing risk), reduction of incidence of diseases, control
of weeds, spread of labor demands, and satisfaction of

228
dietary requirements. In some cases, it can also reduce
need for fertilizer.
The significantly higher interest in alley cropping
for fodder (62%) rather than for mulch (33%) was expected
from farmers in an arid area where feeding livestock is a
problem. However, the large number of farmers (33%) who
have livestock but who want to use the technology for mulch
rather than for fodder raises some speculation. First, it
is possible that the mulch-interest group was not as well
informed as the fodder-interest group and that their
responses could change to fodder as they acquire more
knowledge. This change is feasible because the response of
the farmers to fodder or mulch use was observed to depend
significantly on prior knowledge about alley cropping.
Coffee farmers, who knew most about the technology,
responded positively to fodder use. Secondly, if one
assumes that the response was not due to lack of knowledge,
then it would appear that there was a real interest in alley
cropping for purposes of mulch. It is possible that some
farmers were well informed about the technology and,
therefore, made a rational decision. For example, coffee
farmers were more informed about the technology than the
noncoffee farmers, and most coffee farmers responded
positively to fodder-interest. However, the considerable
number of coffee farmers (24%) who responded positively to
mulch-interests would suggest that they made a rational

229
decision. On the other hand, some questions raised by the
mulch-interest group, when they visited the research
station, suggested that their initial response was suspect.
For example, on visiting the research station, both interest
groups raised serious questions about potential competition
between the trees and the crops.
Potentials for Adoption of Aliev Cropping
From the positive response of the farmers surveyed,
62% for fodder use and 33% for mulch use, it would appear
that there is potential for adoption of the alley cropping
technology. Only 5% of the farmers interviewed were
indifferent to the technology. Between the two end uses,
alley cropping is more likely to be adopted for fodder than
for mulch. Coffee farmers are more likely to adopt alley
cropping than those without coffee. For example, of the 62%
interested in fodder, 76% were coffee farmers. Of the total
number of farmers interviewed, 57% were farmers with coffee.
Like the various strategies used by farmers to improve soil
fertility and crop production (e.g., terracing, weeding),
alley cropping would require investments (e.g., purchase of
seedlings). Among the five factors identified to determine
farmers' interests in the use of alley cropping (fodder or
mulch), the presence or absence of coffee trees on the farm
was significant. The presence or absence of other factors

230
(use of fertilizers, ox-plows, knowledge of the technology),
with the exception of mulch use, also depended significantly
on the presence or absence of coffee trees on the farm.
To invest in production strategies, farmers had to spend
money. Labor had to be hired for soil conservation works,
"Fanya Juu" construction and coffee picking, seeds had to be
purchased, and ox-plows had to be hired. The ability to
spend appeared to depend strongly on the possession of a
cash generating activity on the farm. In the area of the
present study, there were two main cash-generating
activities: off-farm work and sale of coffee. Other minor
sources of revenue included sale of fruits, wood products
(e.g., poles), and sisal baskets. Income from off-farm work
and from trees (fruit and nonfruit trees) was probably
significant to the household but did not differ
significantly among farms. What differed between farms was
income from coffee.
Farmers with coffee invested more in inputs (e.g.,
bought fertilizers, manure, possessed or hired ox-plow,
bought feed for their livestock) than noncoffee farmers. In
addition, coffee farmers hired more labor and weeded more
frequently. This suggests that farmers with coffeea cash
crophad greater purchasing power than those farmers
without coffee.

231
The labor requirement is always a factor which rural
people take into consideration when deciding whether or not
to adopt a new practice (Hoskin, 1987). Alley cropping is a
labor intensive practice and the costs of production
increase considerably if additional labor has to be hired
(Hoekstra, 1987). It is suggested that alley cropping is
economically attractive under conditions of severe cash
constraint where hired labor is available at relatively low
cost (Raintree and Turray, 1980; Verinumbe et al., 1984;
andf Sumberg et al., 1987). In the area of the study,
relatively low cost labor was available probably due to
declining farm size and productivity. Arnold (1987)
hypothesized that both of these factors result in an
increased labor pool seeking off-farm employment which may,
in turn, influence the adoption of agroforestry practices.
Under these conditions, alley cropping could be economically
viable for farmers, such as those producing coffee, who have
the ability to undertake the initial investments required by
the technology and, then, to hire the labor necessary to
manage it.
It could be argued that coffee farmers showed a higher
interest in alley cropping for fodder possibly because most
of their land was under coffee production rather than food
or feed crop production. However, farm size did not
influence significantly the interests of farmers in the use
of alley cropping for fodder or for mulch. Coffee farmers

232
could have seen the technology as a means to increase
productivity on their small holdings. Planting trees for
fodder could reduce the coffee farmers' need for purchasing
livestock feed off the farm. Farmers with coffee bought
more feed for their livestock than those farmers without
coffee. By purchasing less from the outside, farmers with
coffee will have even more revenue to invest in other
activities (e.g., buying an ox-plow that is now probably
being rented, milk cows, etc.). The presence of animals on
the farm may in turn minimize the current purchase of
manure. The accrued savings could alternatively be used to
purchase manure, to increase food production, or to pay for
the education of the children. These projections are in
line with the view that one of the economic benefits of
agroforestry is to create capital stocks to meet
intermittent costs and unforeseen contingencies (Arnold,
1987) .
Cash generating activities are absolutely necessary
for farmers to buy farm inputs, food, and invest in cash
generating activities (e.g., fruit trees, vegetables, and
poultry keeping). Without inputs, production is low. The
low crop yields (0.3 t ha-1) even if increased ten-fold "by
some miraculous technology" are unlikely to sustain the food
needs of the large household members. From the on-station
studies, a significant response to mulch was obtained only
when it was applied at a rate greater than 4 t ha'1 season'1.

233
The hedgerows, on the other hand, produced at most 2 t ha'1
season'1 from systems occupying 15%-25% of the land. The
potentials of mulch to improve yields is, therefore, limited
by the quantity of the mulch that the hedgerows can produce.
Theoretically, mulch yield could be increased by putting 50%
of the crop land to trees. It is unlikely, however, that
farmers with two to three hectares land would put 50% of
their land to trees/hedges.
Between the two tree species evaluated on the station,
leucaena was more competitive with crops than cassia.
Averaged over seasons, maize yield increased under cassia by
8% per season, while yields declined by 12% per season under
leucaena. However, the presence of mulch- and fodder-
interest groups suggests two recommendation domains
(research or extension): cassia for farmers with mulch-
interests and leucaena for farmers with fodder-interests.
The characteristics of the two tree species also lend
themselves to meeting the separate needs of the farmers.
Leucaena can be used for both fodder and mulch while cassia
has mulch but no known fodder value. Because leucaena is
competitive with crops, its planting would be outside the
crop fields which, as Arnold (1987) postulated, could
provide productive applications for under-utilized land,
labor or capital.

234
Emerging Themes
One of the themes emerging from this study is that
farmers invested in soil fertility improvement and
productivity. Soil conservation, weeding, possession, or
use of an ox-plow could all be considered as investments.
These investments require cash and so will alley cropping
for establishment and maintenance of the hedgerows.
Farmers with coffee are more likely to acquire the
inputs to improve or sustain food production. Coffee was
<3
used as collateral to secure loans for education of the
children. Coffee is just an example of how a cash
generating activity can influence farmers' potential to
invest and adopt new technologies. It could be replaced by
any other enterprise. In fact, because of low prices of
coffee at the time of the present study, farmers were
replacing their coffee with fruit trees, vegetables, and
poultry, as alternative cash generating activities. One
could also visualize trees planted for production and sale
of fodder as an alternative technology. Planting and sale
of fruits, poles from trees (predominantly Eucalyptus but
also Grevillea), and firewood was common.
The changes from coffee to other enterprises as cash
generating activities are, indeed, positive. Coffee prices
have oscillated widely and farmers have not been able to
intercrop coffee (which they would like to do) because of
strict government protection of coffee. Sale of products

235
from noncoffee enterprises is, however, unlikely to have
significant impact as a cash generating means unless they
receive similar support to that of coffee in terms of
research, extension, and marketing. Poor marketing and over
production can lead to losses. For instance, there have
been occasions, when mango and guava trees were in season,
fruit was rotting in the fields. Only farmers who could
afford to market their fruits outside the local market would
survive under such conditions of glut and poor marketing
opportunities. Fodder from hedgerows, if grown for sale,
could' have marketing problems over time similar to those of
fruits. This, however, is difficult to speculate about
since there have hardly been any fodder products from the
farms presented for sale at the markets.
Implications of Availability of
Trees and the Problems of Termites
on the Adoption of Alley Cropping
Seeds and seedlings
The question of availability of planting materials
(seeds or seedlings) was a major bottleneck to the farmers'
efforts to plant trees. To enhance availability of planting
materials, some farmers maintained their own nursery in
spite of the presence of various constraints. The question
of availability of planting materials is one that deserves
more discussion since it has implications on both the
research and development aspects of alley cropping. Without

236
readily available planting materials, the station developed
technologies, whatever their potential, may not be adopted
by farmers.
It could be argued that the question of availability
of planting materials should be addressed only after a
viable technology has been developed on the station. The
fallacy with this prevalent attitude is that it would take
years for technologies to reach farmers and have an impact
on their lives. It also denies the researcher valuable
feedback from farmers that could enhance development of a
relevant technology, if for instance, components of the
technology were simultaneously being tried by farmers under
their different farming environments.
Of particular concern is research on some of the
agroforestry practices such as alley cropping that by nature
are long-term and also require a large number of planting
materials. To be effective for soil erosion control, or
mulch production, alley cropping requires a large number of
seedlings. For example, to plant 20% of the land under
hedges spaced, for instance, at 5 meters between rows and
0.5 m between plants in a row would require 800 plants/ha.
Assuming a highly optimistic 50% survival after six months,
one would require at least 1600 seedlings per hectare. If
all the 60 farmers interviewed were to plant 20% of their
land to hedgerows, this would require 96,000 seedlings. One
seedling would cost at least two Kenya shillings (0.1 $);

237
96,000 seedlings would require $9,600 or $160 per farm. By
local standards, this is a lot of money and not readily
available to resource-poor farmers. In addition, there is
no infrastructure in the area to supply or sustain such a
supply of seedlings. Seed-based technology and not
seedlings would probably be a cost-effective means of
meeting potential demands for such amounts of planting
materials.
Unfortunately, all alley cropping studies at the
station were based on the use of transplanted seedlings with
a total lack of any research on direct planting of seeds in
the field. This is probably so because the uniform
establishment of trees under the conditions of the study was
easier from seedlings than from seeds. Uniform establish
ment of trees is an important consideration under conditions
of experimentation. An equally important question about
availability raised by the farmers was care and management
of the trees after planting. These included prevalent dry
conditions, browse by animals, weeds and termites. Of most
concern was damage of trees by termites.
Termites
Most of the farmers interviewed (86%) mentioned
termites as a major problem to raising trees. Termites were
also a major problem in the research station but were
overcome with termite control chemicals. However, farmers

238
have not had ready access to chemicals as have researchers
on the station. Besides, some of the chemicals in the
market (e.g., DDT) are environmentally hazardous.
Typically, farmers use wood ash as a traditional control
measure for termites. However, the success of this practice
was highly variable from farm to farm.
Research into termite control is, indeed, the key to
success of any tree planting activity in this and other
similar areas. At the time of concluding the present study,
termite control studies were missing from the many trials at
the research station. In particular, it would be valuable
to explore the use of nonchemical methods such as wood ash,
their potentials and constraints.
Conclusions
Most farmers were interested in alley cropping. The
interest was perhaps more in the trees than in the alleys
created by the trees. There were two distinct interest
groups: primarily in fodder (62%) and, in mulch (33%).
Interests of farmers in alley cropping depended
significantly on five factors:
1. Presence or absence of coffee,
2. Use of mulch on the crop fields,
3. Use or possession of an ox-plow,
4. Use of fertilizer, and
5. Prior knowledge about the technology.

239
With the exception of mulch use, the presence or absence of
the other factors depended significantly on the presence or
absence of coffee on the farm. For instance, 67% of the
fodder-interest group produced coffee and 76% of all coffee
producers were interested in the fodder production from
alley cropping. It did appear that coffee provided the
capital required to purchase inputs to raise or sustain
sufficient food production.
Farmers used many methods to maintain fertility and
productivity of their soils. The major ones were soil
conservation measures, use of livestock manure, use of
fertilizer, purchase of improved seeds, weed control, and
intercropping. Most of these measures required cash inputs.
Alley cropping would also require cash inputs for
establishment and maintenance of the hedgerows. Thus, alley
cropping would likely be limited to farmers with capital
(e.g., coffee farmers).
Although the farm holdings were small, with 76% of the
farmers sampled having less than three hectares, farm size
appeared unlikely to constrain adoption of the technology.
The major constraints identified were the capital investment
required to plant a large number of seedlings, availability
of the trees, termite problems, and fears of decreased crop
yields, due to competition from the trees.
Based on the results from the on-station studies,
which showed that leucaena was competitive with crops while

240
cassia was not, it is suggested that alley cropping cassia
could be targeted to all farmers, and specifically, to
farmers with mulch-interest. On the other hand, leucaena
could be targeted mainly to farmers with fodder-interest.
However, the research suggested that leucaena be planted
separately from the crops. Notwithstanding the above
recommendation, cassia alley cropping has to be viewed
against the magnitude of gain in crop yield (8% over the
control), problems of pests, and mortality associated with
cassia (which may make replanting necessary after 3-4
years), and the costs and availability of the large number
of seedlings required.
Ninety-two percent of the farmers interviewed let
their livestock into the fields after the harvest of crops.
Therefore, whether alley cropped or planted separately from
crops, the trees/hedges are likely to be damaged by
livestock grazing in the crop fields.

CHAPTER 7
SUMMARY AND CONCLUSIONS
Maize yield of alley-cropped plots of leucaena and
cassia was higher than those of plots receiving mulch from
the block planting system. This suggested that the effects
of growing the hedges in situ was positive. The yield
advantage of alley cropping over the separate block planting
system, expressed as land equivalent ratio, was 36% and 12%
for cassia and leucaena, respectively. The amounts of mulch
applied to the alley cropped and block plating systems being
equal, the higher yield of the alley cropping systems could
be ascribed to effects associated with the in situ presence
of the hedgerows. Although not examined in the present
study, the in situ effects of the hedgerows could include
root and rhizosphere activities of the hedgerows (e.g.,
biological N-fixation, mycorrhizal association) and soil
erosion control. There is evidence from studies of others
that the soil erosion control aspects of alley cropped
hedgerows can be significant.
Between the two species, leucaena alley cropping
reduced maize by 12% per season compared to the sole crop
while that under cassia increased by 8% per season.
Leucaena produced more biomass (4 t ha'1 yr'1) than cassia
241

242
(2 t ha'1 yr'1) ; also leucaena had higher fine root density
than cassia in the top soil (0-40 cm). Because of these,
leucaena was more competitive than cassia with maize. The
difference between the effects of the two species indicates
how performance of alley cropped hedgerows depends on the
tree species used, and suggests the need to include more
than one species in alley cropping studies. It also
suggests that management of the hedgerows may have to be
different even in the same experiment. It is possible that
leucaena would have been less competitive than it was in the
present study if the pruning frequency was more than 4-6
times a year. The use of more drought tolerant hedgerow
species and annual crops than leucaena, cassia and maize
could also have changed the results of the present study.
The reduced yield of maize under alley cropping
leucaena was due more to competition for water, and less for
nutrients since the plots were fertilized with both N and P.
Competition for light was probably negligible given that the
hedgerows were regularly (at least twice in a season) pruned
low (to 50 cm heights above ground). The presence and
effects, if any, of allelopathic chemicals was not studied.
Interestingly, maize yield nearer the hedgerows yielded
more than those away from them. The higher yield near the
hedgerows was attributed to higher water content near the
hedgerows than away from them. This resulted from lowered
soil bulk density and hence higher rates of infiltration

243
near the hedgerows than away from them. It is possible that
other factors such as mycorrhizal association, root
turnover, and root exudates, and microbial activity could
have contributed to the higher yield near the hedgerows than
away from them. These factors were not examined in the
present study.
Planting the hedgerows outside the crops may also be a
management approach to avoid competition between the crop
and the tree hedgerows. Such a practice would particularly
be appropriate for a species like leucaena that has fodder
value but whose intercropping is detrimental to crop yields.
However, separate planting of the hedges from the crops (on
land that would otherwise be cropped) resulted in greater
reduction of maize yield than did alley cropping of both
species. The yield loss in the separate hedge and crop
planting system was nearly equal to the percentage of land
allotted to the hedges. This was so because the mulch
produced from the separate hedge and crop planting system
did not compensate (in terms of crop yield) for land lost
from cropping.
The amounts of mulch produced by cassia and leucaena
averaged 2 and 4 t ha"1 yr"1, respectively. The yields of
the block planted hedgerows outside the crop field were 14%
and 25% lower those of the alley cropped hedgerows of
leucaena and cassia, respectively. In the humid tropics
where mulch yields of alley cropped leucaena are reportedly

244
as high as 8 to 10 t ha"1 yr"1) significant improvements in
both soil fertility and crop yields have been noted. The
low biomass yield of the hedgerows that also decomposed fast
(half life of four weeks) was not sufficient to effect
significant improvements in soil fertility (organic C, total
N and CEC) and maize yields. Maize yields of plots with or
without mulch were similar. If the amounts of mulch applied
in a season could be increased to 4 t ha"1 or more, there
were indications from pot studies that increases in maize
yield could be significant. To obtain such levels of mulch
would, however, require that the density of the hedgerows be
increased to 2 to 4 times that of the present study. This
could result in further loss in crop yields due to reduction
of area available for cropping and possible competition from
the hedgerows.
The small amounts of mulch produced by the hedgerows
may have had significant effects on soil fertility and
productivity if the present studies were long-term and/or
the site was nutrient-poor. The present study was conducted
for only three years and on land not previously cropped for
more than 2 years. Also, moderate levels of chemical
fertilizer (40 and 18 kg N and P, respectively) were applied
to the plots each season. The higher relative indices of
available (KCl-extract) N observed in plots that were
mulched compared to those that were not may be important to
the productivity of N deficient soils.

245
Improved levels of available soil N may result in
improved nutrition and yield of crop, particularly if the
periods of peak N mineralization from the mulch (four weeks
after application) are matched/synchronized with periods of
peak N uptake by maize (6-8 weeks after sowing). The low
apparent recovery of the mulch-N applied could be a
reflection of the absence of synchrony between the peaks of
mulch-N mineralization and uptake by the crop. In the
field, apparent N recoveries from the mulch of leucaena and
cassia were 3% and 7%, respectively. Studies using 1SN
tracers into the pathways and pools of N transfer from the
mulch may provide insights into the potentials of improving
the utilization efficiency of the limited amounts of mulch
available from the hedgerows. In addition, studies on time
of mulch application and/or the use of mixtures of mulches
of different qualities (i.e., variable rates of
decomposition) may help determine to what extent management
practices can achieve synchrony between mulch N
mineralization and uptake by the crop.
Based on the 8% increase in maize yield under cassia
alley cropping as opposed to 12% decline under leucaena, it
is tempting to recommend cassia alley cropping to farmers.
To put such a recommendation into perspective would,
however, require an understanding of the costs associated
with the 8% increase in maize yield from cassia alley
cropping as well as additional costs of replanting the

246
hedges due to mortality of cassia with time. Up to 5% and
11% of the initial stands of the intercropped and block
planted hedgerows of cassia were dead 3-4 years after
planting. If it is economically positive and problems of
mortality are minimal or controllable, cassia alley cropping
would be recommendable on all farms, and particularly where
there are interests in mulch. It could also be recommended
for soil erosion control on all farms.
Most of the farmers were interested in the fodder than
the mulch aspects of alley cropping. Leucaena has fodder
value, cassia does not. But leucaena alley cropping is
favorable to crop yields. Therefore, planting of leucaena
separately from crops, perhaps on lands not suitable for
cropping, would be recommendable. Because of the initial
high investments required to establish the hedgerows, alley
cropping or block planting systems of hedgerows is more
likely to be adopted by relatively resourceful farmers such
as coffee growers than by poorer farmers. Termites are
major problems to on-farm tree planting. Therefore,
research into termite control methods that can be afforded
by farmers and that are also not environmentally hazardous
are necessary.

LITERATURE CITED
Ahn, J.H., Robertson, B.M., Elliott, R., Gutteridge, R.C.
and Ford, C.W., 1989. Quality assessment of tropical
browse legumes: tannin content and protein degradation.
Animal Feed Science and Technology 27, 147-156.
Alexander, M., 1977. Introduction to soil microbiology.
Second edition. John Wiley and Sons, New York.
Allison, F.E., 1966. The fate of nitrogen applied to soils.
Advances in Agronomy 18, 219-259.
Allison, F.E., 1973. Soil organic matter and its role in
crop production. Developments in Soil Science 3.
Elsevier Scientific Publishing Company, Amsterdam.
Anderson, J.M., and Ingram, J.S.I., 1989. Tropical soil
biology and fertility program: A handbook of methods.
C.A.B International, Wallingford, U.K.
Arnon, I., 1975. Mineral nutrition of maize. International
Potash Institute, Bern, Switzerland, pp. 143-144.
Atta-Krah, A.N., and Francis, P.A., 1987. The role of on-
farm trials in the evaluation of composite
technologies: the case of alley farming in southern
Nigeria. Agroforestry Systems 23, 133-152.
Atta-Krah, A.N., Sumberg, J.E., and Reynolds, L., 1985.
Leguminous fodder trees in the farming systems: an
overview of research at the humid zone program of ILCA
in south-western Nigeria. ILCA, Ibadan, Nigeria.
Ayanaba, A., 1982. The value of mulches in the management
of organic matter in the tropics, pp 97-103. In:
Cerri, C.C., Athi, D., Sadrzcieski, D. (eds.) Regional
Colloquium on Soil Organic Matter Studies. Sao Paulo,
Brazil.
Baldwin, I.T., Olson, R.K., and Reiners, W.A., 1983.
Protein binding phenolics and the inhibition of
nitrification in sub-alpine balsam fir soils. Soil
Biology and Biochemistry 15, 419-423.
247

248
Barley, K.P., 1954. Effects of root growth and decay on
the permeability of synthetic sandy loam. Soil Science
78, 205-210.
Bartholomew, W.V., 1965. Mineralization and immobilization
of nitrogen in the decomposition of plant and animal
residues. In: Bartholomew, W.V. and Clark, F.E.
(eds.) Soil Nitrogen. American Society of Agronomy,
Madison.
Belsky, A.J., Amundson, R.G., Dixbury, J.M., Riha, R.J.,
Ali, A.R., and Mwonga, S.M., 1989. The effects of
trees on their physical, chemical and biological
environments in a semiarid savannah in Kenya. Journal
of Applied Ecology, 26, 1005-1024.
Belsky, A.J., 1992. Effects of trees on nutritional quality
of understorey gramineous forag in tropical savannas.
Tropical Grasslands 26, 12-20
Bennet, J.M., 1990. Problems associated with measuring
plant water status. HortScience 25, 1551-1554.
Berg, B., 1986. Nutrient release from litter and humus in
coniferous forest soilsa mini review. Scandinavian
Journal of Forestry Research 1, 359-369.
Binkley, D., and Hart, S.C., 1989. The components of
nitrogen availability assessment in forest soils.
Advances in Soil Science 10, 57-112.
Binkley, D., and Matson, P., 1983. Ion exchange resin bag
method for assessing forest soil nitrogen availability.
Soil Science Society of America Journal 47, 1050-1052.
Blake, G.R. 1965. Bulk density, pp. 374-390. In: Black,
C.A., (ed.), Methods of soil analysis. American
Society of Agronomy.
Bohm, W., 1979. Methods of studying root systems.
Springer-Verlag, Berlin.
Bonkoungou, E.G., 1992. Sociocultural and economic
functions of Acacia albida, pp. 1-8. In: Vandenbeldt,
R.J. (ed.), Faitherbia albida in the West African
semiarid tropics. Proceedings of a workshop held at
Niamey, Niger, 22-26 April.
Bowen, G.D., 1985. Roots as components of tree
productivity, pp. 303-315. In: Cannell, M.G.R., and
Jackson, J.E. (eds.), Attributes of trees as crop
plants. Institute of Terrestrial Ecology, Huntingdon,
U.K.

249
Brady, N.C., 1990. The nature and properties of soil. 10th
Edition. Macmillan Publishing Co., New York.
Brewbaker, J.L., 1987. Leucaena: a multipurpose tree genus
for tropical agroforestry, pp 289-323. In: Steppler,
H.A. and Nair, P.K. (eds.), Agroforestry: a decade of
development. ICRAF, Nairobi, Kenya.
Bromfield, A.R., 1969. Uptake of phosphorus and other
nutrients by maize in Western Nigeria. Experimental
Agriculture 5, 91-100.
Brophy, L.S., and Heichel, G.H., 1989. Nitrogen release
from roots of alfalfa and soybean grown in sand
culture. Plant and Soil 16, 77-84.
Budelman, A., 1988. The decomposition of the leaf mulches
of Leucaena leucocephala, Gliricidia sepium and
Flemingia macrophylla under humid tropical conditions.
Agroforestry Systems 7, 33-45.
Caldwell, M.M., and Richards, J.H., 1989. Hydraulic lift:
water efflux from upper roots improves effectiveness of
water uptake by deep roots. Oecologia 79, 1-5.
Campbell, C.A., 1983. Soil organic carbon, nitrogen and
fertility, pp. 173-291. In: Schnitzer, M. and Khan,
S.U. (eds.), Soil organic matter. Developments in
Soil Science, No. 8, second impression. Elsevier
Science Publishers, Amsterdam.
Chambers, R., 1985. Shortcut methods of gathering social
information for rural development projects. In:
Cernea, M.M., (ed.), Putting people first
Sociological variables in rural development. Oxford
University Press, New York.
Chapin, F. S., 1980. The mineral nutrition of wild plants.
Annual Review of Ecology and Systematics 11, 233-260.
Cheshire, M. V., Mundie, C.M., and Shepherd, H., 1974.
Transformations of sugars when rye hemicellulose
labelled with 14C decompose in soil. Journal of Soil
Science 25, 90-95.
Chirwa, P.W., 1991. Soil moisture changes and maize
productivity under alley cropping with leucaena and
Flemingia hedgerows in semi-arid conditions in Lusaka,
Zambia. M.S. Thesis, University of Florida,
Gainesville.
Cochran, V.L., 1991. Decomposition of barley straw in a
sub-arctic soil in the field. Biology and Fertility of
Soils 10, 227-232.

250
Coleman, D.C., Edwards, A.L., Belksy, A.J., and Mwonga, A.,
1991. The distribution and abundance of soil nematodes
in East African Savannas. Biology and Fertility of
Soils 12, 67-72.
Comanor, P.L., and Staffeldt, E.E., 1978. Decomposition of
plant litter in two western North American deserts,
pp. 31-49. In: West, N.E., and Skujins, J.J., (eds.),
Nitrogen in desert ecosystems. US/IBP series #9.
Dowden, Hutchinson and Ross, Inc. Publishers.
Comerford, N.B., Kidder, G., and Mollitor, A.V., 1984.
Importance of subsoil fertility to forest plant
nutrition. In: Stone, E.L., (ed.), Forest soils and
treatment impacts. Proceedings of the Sixth North
American Forest Soils Conference, Univ. of Tennessee,
Knoxville.
Corlett, J.E., 1989. Leucaena/millet alley cropping in
India: Microclimate and productivity. Ph.D.
Dissertation, University of Nottingham, U.K.
Corlett, J.E., Ong, C.K., and Black, C.R., 1989.
Microclimate modification in intercropping and alley
cropping systems, pp. 419-430. In: Reifsnyder, W.S.,
and Darnhoffer, T.O., (eds.), Meteorology and
agroforestry. ICRAF, Nairobi, Kenya.
CSIRO, 1988. CSIRO disc permeameter instruction manual.
Commonwealth Scientific Industrial Research
Organization, Canberra, Australia.
Danso, S.K.A., Zapata, F., and Hardarson, G., 1987.
Nitrogen fixation in fava beans as affected by plant
population density in sole or intercropped systems with
barley. Soil Biology and Biochemistry 19, 411-415.
Deans, J., and Ford, E.D., 1983. Modelling root structure
and stability. Plant and Soil 71, 189-196.
DeHaan, S., 1976. Humus, its formation, its relation with
the mineral part of the soil, and its significance on
soil for soil productivity, pp. 21-30. In: Soil
organic matter studies. Volume 1. International Atomic
Energy Agency, Vienna, Austria.
Dhyani, S.K., Narain, P. and Singh, R.K., 1990. Studies on
root distribution of five multipurpose tree species in
Doon Valley, India. Agroforestry Systems 12, 149-161.
Dijkman, M.J., 1950. Leucaenaa promising soil erosion
control plant. Economic Botany 4, 337-47.

251
Dommerques, Y.R., 1987. The role of biological nitrogen
fixation in agroforestry, pp. 245-271. In: Steppler,
H.A., and Nair, P.K.R., (eds.), Agroforestry: a decade
of development, ICRAF, Nairobi, Kenya.
Duguma, B., Kang, B.T., and Okali, D.U.U., 1988. Effect of
pruning intensity of three woody leguminous species in
alley cropping with maize and cowpea on an alfisol.
Agroforestry Systems 6, 19-35.
Dzowella, B., 1992. AFRENA (Agroforestry Research Networks
for Africa). Annual Report, Zimbabwe.
Ericksen, F.I., and Whitney, A.S., 1981. Effects of light
intensity on growth of some tropical forage species: I.
Interaction of light intensity and nitrogen
fertilization on six forage grasses. Agronomy Journal
73, 427-433.
Evenssen, C.L.I., 1989. Alley cropping and green manuring
for upland crop production in West Sumatra. Ph.D.
dissertation. University of Hawaii, Hilo.
Ewel, J., Benedict, F., Berish, C., Brown, B., Gliessman,
S., Amador, M., Bermudez, R., Martinez, A., Maranda,
R., and Price, N., 1982. Leaf area, light
transmission, roots and leaf damage in nine tropical
plant communities. Agroecosystems 7, 305-326.
Fayle, D.C.F., 1975. Distribution of radial growth during
the development of red pine root systems. Canadian
Journal of Forest Research 4, 28-38.
Felker, P., 1978. State-of-the-art: Acacia albida as a
complementary intercrop with annual crops. Univ. of
California, Riverside.
Fernandez, E.C.M., 1990. Alley cropping on acid soils.
Ph.D. dissertation, North Carolina State University,
Raleigh.
Ford, E.D., and Deans, J.D., 1977. Growth of Sitka Spruce
plantation: spatial distribution and seasonal
fluctuations of lengths, weights and carbohydrate
concentration of fine roots. Plant and Soil 47, 463-
485.
Fox, R.H., Myers, R.J.K. and Vallis, I., 1990. The nitrogen
mineralization rate of legume residues in soil as
influenced by their polyphenol, lignin, and nitrogen
contents. Plant and Soil 129, 251-259.

252
Gholz, H.L., Perry, C.S., Cropper, W.P. Jr., and Hendry,
L.C., 1985. Litter fall, decomposition and nitrogen
and phosphorus dynamics in a chronosequence of slash
pine (Pinus elliottii) plantations. Forest Science 31,
463-478.
Gichuru, M.P., and Kang, B.T., 1989. Calliandra calothyrsus
(Meissn.) in an alley cropping system with sequentially
cropped maize and cowpea in southwestern Nigeria.
Agroforestry Systems 9 (3), 191-204.
Guevarra, A.B., 1976. Management of Leucaena leucocephala
for maximum yield and nitrogen contribution to
intercropped corn. Ph.D. dissertation, University of
Hawaii, Honolulu.
Gutteridge, R.C., 1992. Evaluation of the leaf of a range
of tree legumes as a source of nitrogen for crop
growth. Experimental Agriculture 28, 195-202.
Hanway, J.J., 1962. Corn growth and composition in relation
to soil fertility. II. Uptake of N, P and K and their
distribution in different plant parts during the
growing season. Agronomy Journal 54, 217-22.
Harsh, C.T., and Mbatha, B., 1978. The economics of small
holder agriculture in two areas of the semi-arid lands
of Kenya. Report No. 3. Kenya marginal semi-arid
lands pre-investment study. Government of Kenya/USAID,
Nairobi, Kenya.
Hauck R.D., and Bremner, J.M., 1976. Use of tracers for
soil for soil nitrogen research. Advances in Agronomy
28, 19-266
Haynes, R.J., 1986. The decomposition process:
mineralization, immobilization, humus formation and
degradation, pp. 52-176. In: Haynes, R.J. (ed.),
Mineral nitrogen in the plant-soil system, Academic
Press, Orlando, FL.
Herman, W.A., McGill, W.B., and Dormarr, J.F., 1977.
Effects of initial chemical composition on
decomposition of roots of three grass species.
Canadian Journal of Soil Science 57, 205-215.
Hildebrand, P.E., 1981. Combining disciplines in rapid
appraisal: the sondeo approach. Agricultural
Administration 8, 423-32.

253
Hildebrand, P.E., 1982. Farming systems research: issues
in research strategy and technology design: discussion.
American Journal of Agricultural Economics 64 (5), 905-
906.
Hildebrand, P.E., 1990. Farming Systems research
extension, pp. 131-143. In: Jones, J.G.W and Street,
P.R. (eds.), Systems theory applied to agriculture and
the food chain. Elsevier Applied Science, Amsterdam.
Hildebrand, P.E., and Poey, F., 1985. On-farm agronomic
trials in farming systems research and extension.
Lynne Reinner Publishers, Inc., Boulder, Colorado.
Hoekstra, D., 1985. The use of economics in diagnosis and
design of agroforestry systems. Working Paper 29,
ICRAF, Nairobi, Kenya.
Hook, R.I.V., Johnson, D.W., West, D.C. and Mann, L.K.,
1982. Environmental effects of harvesting forests for
energy. Forest Ecology and Management 4, 79-94.
Hoskins, M.W., 1987. Agroforestry and the social milieu,
pp. 192-203. In: Steppler, H. A. and Nair, P.K.R.
(eds.), Agroforestry: a decade of development.
ICRAF, Nairobi, Kenya.
Howard, P.J.A., and Howard, D.M., 1974. Microbial
decomposition of tree and tree and shrub leaf litter.
1. Weight loss and chemical composition of decomposing
litter. Oikos 25, 341-352.
Hunt, H.W., 1977. A simulation model for decomposition in
grasslands. Ecology 58, 469-484.
Huxley, P.A., 1986. Rationalizing research on hedgerow
intercropping: an overview. ICRAF Working Paper No.
40. ICRAF, Nairobi, Kenya.
Huxley, P.A., Darnhoffer, T., Pinney, A., Akunda, E., and
Gatama, D., 1989. The tree/crop interface: a project
designed to generate experimental methodology.
Agroforestry Abstracts 2 (4), 127-145.
Huxley, P.A., Patel, R.Z., Kebaara, A.M., and Mitchell,
H.W. 1977. Tracer studies with P32 on the
distribution of functional roots of Arabica coffee in
Kenya. Annals of Applied Biology 77, 159-180.
Ikombo, B. M., 1984. Effects of farm yard manure and
fertilizers on maize in semi-arid areas of Eastern
Kenya. East Africa Agricultural and Forestry Journal
44, 266-74.

254
Jackson, M.L., 1958. Soil chemical analysis. Prentice-
Hall, Englewood, New Jersey.
Jaiyebo, E.O., and Moore, E.O., 1964. Soil fertility and
nutrient storage in different soil vegetation systems
in a tropical rain forest environment. Tropical
Agriculture (Trinidad) 41, 129-130.
Jenkinson, D.S., Fox, R.H., and Rayner, J.H., 1985.
Interactions between fertilizer nitrogen and soil
nitrogenthe so-called "priming" effect. Journal of
Soil Science 36, 425-444.
Jensen, V., 1974. Decomposition of angiosperm tree leaf
litter, pp. 69-104. In: Dickinson, C.H., Pugh, G.J.F.
(eds.), Biology of plant decomposition. Vol. 1.
Academic Press, New York.
Joffre, R., Vacher, J., De Los Llanos, C., and Long, G. ,
1988. The dehesa: an agrosilvopastoral system of the
Mediterranean region with special reference to the
Sierra Morena area of Spain. Agroforestry Systems 6,
71-96.
Jones, C.A., 1985. C4 grasses and cereals. Wiley and Sons,
New York.
Jonsson, K., Fidjeland, L., Maghembe, J.A., and Hogberg, P.,
1988. The vertical distribution of fine roots of five
tree species and maize in Morogoro, Tanzania.
Agroforestry Systems 6, 63-69.
Kang, B.T., and Duguma, B., 1985. Nitrogen movement in
alley cropping systems. In: Kang, B.T., and van den
Heide, J. (eds.), Nitrogen in farming systems in the
humid and the sub-humid tropics. Institute of Soil
Fertility, Harn, The Netherlands.
Kang, B. T., Grimme, H. and Lawson, T.L., 1985. Alley
cropping sequentially cropped maize and cow pea with
leucaena on a sandy soil in southern Nigeria. Plant
and Soil 85, 267-277.
Kang, B.T., Reynolds, L. and Atta-Krah, A.N., 1990. Alley
farming. Advances in Agronomy 43, 315-359.

255
Kang, B.T., and Wilson, G.F., 1987. Agroforestry: a decade
of development, pp 227-243. In: Steppler, H.A., and
Nair, P.K.R. (eds.), International Centre for Research
in Agroforestry, Nairobi, Kenya.
Kang, B.T., Wilson, G.F. and Lawson, T.L., 1984. Alley
cropping: a stable alternative to shifting
cultivation. International Institute of Tropical
Agriculture, Ibadan, Nigeria, pp. 22.
Kang, B.T., Wilson, G.F., and Sipkens, L., 1981. Alley
cropping maize (Zea mays) and Leucaena (Leucaena
leucocephala) in Southern Nigeria. Plant and Soil 63,
165-179.
Reliman, M., 1979. Soil enrichment by neotropical savanna
trees. Journal of Ecology 67, 565-577.
Kelly, K.R., and Stevenson, F.J., 1987. Effects of carbon
source on immobilization and chemical distribution of
fertilizer nitrogen in soil. Soil Science Society of
America Journal 51, 946-951.
Kibe, J.M., Ochung, H., and Macharia, D.N., 1981. Soils and
vegetation of the of the ICRAF experimental farm,
Machakos District. Detailed Soil Survey No. D23, Kenya
Soil Survey, Nairobi.
Kimmins, J.P., 1977. Evaluation of the consequences for
future tree productivity of loss of nutrients on whole
tree harvesting. Forest Ecology and Management 1, 169-
71.
King, H.G.C., and Heath, G.W., 1967. The chemical analysis
of small samples of leaf material and the relationship
between the disappearance and composition of leaves.
Pedobiologia 7, 192-197.
Kumar, M.B., and Deepu, J.D., 1992. Litter production and
decomposition dynamics in moist deciduous forests of
the Western Ghats in Peninsular India. Forest Ecology
and Management 50, 181-201
Ladd, J.N., Oades, J.M. and Amato, M., 1981. Distribution
and recovery of nitrogen from legume residues
decomposing in soil sown to wheat in the field. Soil
Biology and Biochemistry 13, 251-256.
Lai, R., 1989. Agroforestry systems and soil surface
management of a tropical alfisol: 1. Soil moisture and
crop yield. Agroforestry Systems 8, 7-29.

256
Lalf R. 1991. Myths and scientific realities of
agroforestry as a strategy for sustainable management
for soils in the tropics. Advances in Soil Science 15,
91-137.
Landon, J.R., (ed.), 1990. Booker tropical soil manual: a
handbook for tropical soil survey and agricultural land
evaluation in the tropics and sub-tropics. John Wiley
and Sons, New York.
Lightfoot, C., 1987. Indigenous research and on-farm
trials. Agricultural Administration and Extension 24,
79-89.
Lousier, J.D., and Parkinson, D., 1976. Litter
decomposition in a cool temperate deciduous forest.
Canadian Journal of Botany 54, 419-436.
Ludwig, J.A., 1987. Primary productivity in arid lands.
Myths and realities. Journal of Arid Environment 13,
1-7.
Lulandala, L.L.L., and Hall, J.B., 1990. Nutrient removals
in harvesting of Leucaena hedgerows at Mafiga,
Morogoro, Tanzania. Forest Ecology and Management 35,
207-216.
Maher, C., 1937. Soil erosion and land fertility in the
Ukamba (Kitui) reserve. Ministry of Agriculture,
Nairobi, Kenya.
Mann, H.S., and Saxena, S.K., 1980. Khejri (Prosopis
cineraria) in the Indian desert: Its role in
agroforestry. CAZRI Monograph 11. Central Arid Zone
Research Institute, Jodhpur, Rajasthan, India.
Martin, J.P., and Haider, K., 1980. Microbial degradation
and stabilization of 14c-labelled lignin, phenols, and
phenolic polymers in relation to soil humus formation,
pp. 77-140. In: Kirk, T.K., Higuchi, T., and Chang,
H.M. (eds.), Lignin biodegradation; microbiology
chemistry and potential applications. Vol. 2, CRC
Press, West Palm Beach, Florida.

257
Martin, J.P., and Stott, D.E., 1983. pp.1-75. In: Skujins,
I.J., (ed.), Uses of microbiological processes in arid
lands for desertification control and increased
productivity. Proc. 6th Annual International Symposium
on Environment and Biogeochemistry. UNEP, Paris.
Mbogo, S.G., 1991. Production profilecrop production,pp.
10-19. In: Tifien, M., (ed.), Environmental change
and dryland management in Machakos District, Kenya,
1930-1990. Ministry of Reclamation and Development of
Arid, Semi-arid areas and Wastelands, Nairobi.
McClaugherty, C.A., Pastor, J., Aber, J.D., Melillo, J.M.,
1985. Forest litter decomposition in relation to soil
nitrogen dynamics and litter quality. Ecology 63, 621-
626.
Melillo, J.M., Aber J.D., and Muratore, J.F., 1982.
Nitrogen and lignin control of hardwood leaf litter
decomposition dynamics. Ecology 63, 621-626.
Mittal, S.P., and Singh, P., 1989. Intercropping field
crops between rows of Leucaena leucocephala under
rainfed conditions in northern India. Agroforestry
Systems 8, 165-172.
I
Monteith, J.L., 1990. Conservative behavior in the response
of crops to water and light, pp. 3-14. In: Rabbinge,
R. (ed.), Theoretical production ecology: hindsight
and perspectives. Pudoc, Wageningen, The Netherlands.
Monteith, J.L., Ong, C.K., and Corlett, J.E., 1991.
Microclimatic interactions in agroforestry systems.
Forest Ecology and Management 45, 31-44.
Mugendi, D.N., 1990. Plant nutrient aspects of mulch
incorporation in alley cropping trials of semi-arid
Machakos, Kenya. M.S. Thesis. University of Nairobi,
Kenya.
Mugendi, D.N., 1991. Effect of green manure and hedgerow
intercropping on crop production: The DARP Experience.
Paper Presented at the Dryland Agroforestry Research
Project (DARP) workshop held at Machakos, September.
Muller, M.M., and Sundman, V., 1988. The fate of nitrogen
(15N) released from different plant materials during
decomposition under field conditions. Plant and Soil
105, 133-139.
Mungai, D.N., 1991. Microclimatic aspects of hedgerow
intercropping in semi-arid Machakos, Kenya. Ph.D.
Dissertation. University of Nairobi, Kenya.

258
Murwira, H.K., Kirchmann, H., and Swift, M.J., 1990. The
effects of moisture on the decomposition of cattle
manure. Plant and Soil 122, 197-199.
Mwangi, P., 1989. Soil plant nutrient relationship as
affected by Leucaena leucocephala, Cassia siamea and
Terminalia brownii mulch. M.S. Thesis, University of
Nairobi, Kenya.
Myers, L.R., 1982. A sociological approach to farming
systems in Kenya. Cornell International Agriculture,
Mimeo, Ithaca, N.Y.
Nadar, H.M., and Faught, W.A., 1984. Maize yield response
to different levels of nitrogen and phosphorus
fertilizer application. East Africa Agricultural and
Forestry Journal 44, 147-156.
Nair, P.K.R., 1984. Soil productivity aspects of
agroforestry. ICRAF. Nairobi, Kenya.
Nair, P.K.R., 1987. ICRAF Field Station at Machakos. A
demonstration and training site for agroforestry
technologies. Agroforestry Systems 5, 383-393.
Nair, P.K.R., 1989. The role of trees in soil productivity
and protection, pp. 567-589. In: P.K.R. Nair (ed.),
Agroforestry systems in the tropics. Kluwer Academic
Publishers, Dordrecht, Netherlands.
Nair, P.K.R., 1990. The prospects of agroforestry in the
tropics. World Bank Technical Paper no. 131.
NAS, 1984. Leucaena: promising forage and tree crop for
the tropics. Second edition. National Academy of
Sciences, Washington, DC., 100 pp.
Nelson, D.W., 1982. Gaseous losses of nitrogen other than
through denitrification, pp. 327-363. In: Stevenson,
F.J. (ed.), Nitrogen in agricultural soils. American
Society of Agronomy, Madison, Wisconsin.
Nyamai, D.O., 1992. Investigations on decomposition of
foliage of woody species using a perfusion method.
Plant and Soil 139, 239-245.
Nye, P.H., and Greenland, D.J., 1960. The soil under
shifting cultivation. Technical Communication 51.
Harpenden, UK. Commonwealth Bureau of Soils, 144 pp.

259
Ockwell, A.P.f Muhammad, L.f Nguluu, S., Parton, K.A.,
Jones, R.K., and McCown, R.L., 1991. Characteristics
of improved technologies that affect their adoption in
semi-arid tropics of eastern Kenya. Farming Systems
Research-Extension 2 (1), 29-46.
Oglesby, K.A., and Fownes, J.H., 1992. Effects of chemical
composition on nitrogen mineralization from green
manures of seven tropical leguminous trees. Plant and
Soil 143, 127-132.
Olson, R.K, and Reiners, W.A., 1983. Nitrification in a
subalpine balsam fir soils: tests for inhibitory-
factors. Soil Biology and Biochemistry 15, 413-418
Ong, C.K., 1991. Interactions of light, water and nutrients
in agroforestry systems, pp. 107-124. In: M.E. Avery,
M.G.R.Cannell and C.K. Ong (eds.), Biophysical research
for Asian agroforestry. Winrock International, New
Delhi.
Ong, C.K., Corlett, J.E., Singh, R.P., and Black, C.R.,
1991. Above and below ground interactions in
agroforestry systems. Forest Ecology and Management
45, 45-57.
Ong, C.K., Rao, M.R., and Mathuva, M., 1992. Trees and
crops: competition for resources above and below the
ground. Agroforestry Today 4 (2), 4-5. ICRAF,
Nairobi, Kenya.
Palada, M.C., 1989. On-farm research methods for alley
cropping, pp. 84-91. In: Kang, B.T. and Reynolds, L. ,
(eds.), Alley farming in the humid and sub-humid
tropics. IDRC, Ottawa, Canada.
Palm, C.A., 1988. Mulch and nitrogen dynamics in an alley
cropping system in the Peruvian Amazon. Ph.D.
Dissertation. North Carolina State University,
Raleigh.
Palm, C.A., and Sanchez, P.A., 1991. Nitrogen release from
the leaves of some tropical legumes as affected by
their lignin and polyphenolic contents. Soil Biology
and Biochemistry 23 (1), 83-88.
Parker, D.T., 1962. Decomposition in the field of buried
and surface-applied corn stalk residue. Soil Science
Society of America 26, 559-562.

260
Parr, J.F., and Papendick, R.I., 1978. Factors affecting
the decomposition of crop residues by microorganisms,
pp. 101-129. In: Oschwald, W.R. (ed.), Crop residue
management systems. American Society of Agronomy,
Madison, Wisconsin.
Patten, D.T., 1978. Productivity and production efficiency
of an upper Sonoran Desert ephemeral community.
American Journal of Botany 65, 891-895.
Persson, H., 1980. Spatial distribution of fine-root
growth, mortality and decomposition in a young Scots
pine stand in central Sweden. Oikos 4, 77-87.
Powlson, D.S., Brookes, P.C., and Christensen, B.T., 1987.
Measurement of soil microbial biomass provides an early
indication of the changes in total soil organic matter
due to straw incorporation. Soil Biology and
Biochemistry 19, 159-164.
Probert, M.E., Okalebo, J.R. and Simpson, J.R., 1990. The
role of farm yard manure for improving soil fertility.
KARI-ACIAR-CSIRO Symposium on a search for strategies
for sustainable dryland cropping in semi-arid eastern
Kenya. Nairobi, 10-11 September.
Radwanski, S.A., and Wickens, G.E., 1967. The ecology of
Acacia albida on mantle soils in Zalingei, Jabel Marra,
Sudan. Journal of Applied Ecology 4, 569-578.
Raintree, J.B., and Turray, F., 1980. Linear programming
model of an experimental leucaena-rice alley cropping
system. IITA Research Briefs 1, 5-7.
Rao, M.R., Muraya, P., and Huxley, P.A., 1992. Observations
on some tree root systems grown in agroforestry
intercrop situations including their graphical
presentation. ICRAF, Nairobi.
Rao, M.R. Ong, C.K.,, Pathak, P., and Sharma, M.M. 1991.
Productivity of annual cropping and agroforestry on a
shallow alfisol in semi-arid India. Agroforestry
Systems 15, 51-63.
Rao, M.R., Reg, T.J., and Willey, R.W., 1987. Response of
cereals to nitrogen in sole cropping and intercropping
with different legumes. Plant and Soil 101, 167-177.
Rao, M.R., Sharma, M.M., and Ong, C.K., 1990. A study of
the potential of hedgerow intercropping in semi-arid
India using two-way systematic design. Agroforestry
Systems 11, 243-258.

261
Rasmussen, P.E., and Collins, H.P., 1991. Long-term impacts
of tillage, fertilizer and crop residue on soil organic
matter in temperate semi-arid regions. Advances in
Agronomy 45, 93-134.
Read, M.D., Kang, B.T., and Wilson, G.F., 1985. Use of
Leucaena leucocephala (Lam. de Wit) leaves as nitrogen
source for crop production. Fertilizer Research 8,
107-116.
Roberts, J., 1976. A study of the.root distribution and
growth in a Pinus silvestris L. (Scots pine) plantation
in East Anglia. Plant and Soil 44, 607-621.
Rocheleau, D.E., Weber, F. and Field-Juma, A., 1988.
Agroforestry in dryland Africa: participatory
planningprocess and methods. ICRAF, Nairobi, Kenya.
Roskoski, J.P., 1986. Nitrogen fixation in leguminous
trees. Ceiba 27, 227-229.
Ruhigwa, B.A., Gichuru, M.P., Mambani, B., and Tariah, N.M.,
1992. Root distribution of Acioa barteri, Alchornea
cordi folia, Cassia siamea, and Gmelina arbrea in an
acid ultisol. Agroforestry Systems 19, 67-78.
Sanchez, P.A., 1987. Soil productivity and sustainability
in agroforestry systems. In: Steppler, H.A., and
Nair, P.K.R. (eds.), Agroforestry: A decade of
development. ICRAF, Nairobi, Kenya.
Sang, F.K., and Hoekstra, D.A., 1987. On-station hedgerow
results of the first two seasons of the operational
phase. Research Report No. 4, Dryland Agroforestry
Research Project (DARP), Machakos, Kenya. KARI/ICRAF,
Nairobi.
Sanginga, N., Gwaze, D., and Swift, M.J., 1991. Nutrient
requirements of exotic tree species in Zimbabwe. Plant
and Soil 132, 197-205.
Sanginga, N., Mulongoy, K., and Ayanaba, A., 1986.
Inoculation of Leucaena leucocephala (Lam.) de Wit with
Rhizobium and its contribution to a subsequent maize
crop. Biological Agriculture and Horticulture 3, 347-
352.
SAS, 1992. Statistical analysis system for linear models.
Third edition, SAS Institute Inc., N.C.
Sayre, I.D., 1955. Mineral nutrition of corn, pp. 293-314.
In: Spraque, G.F. (ed.), Corn and corn improvement.
Academic Press, New York.

262
Schlesinger, W.H., and Hasey, M.M., 1981. Decomposition of
Chaparral shrub foliage: losses of organic and
inorganic constituents from deciduous and evergreen
leaves. Ecology 62, 762-774.
Shield, J.A., and Paul, E.A., 1973. Decomposition of Re
labelled plant material under field conditions.
Canadian Journal of Soil Science 53, 297-306.
Sivapalan, K., 1982. Humification of polyphenol-rich plant
residues. Soil Biology and Biochemistry 14, 309-310.
Sivapalan, K., Fernando, V. and Thenabadu, M.W., 1985. N-
mineralization in polyphenol-rich plant residues and
their effect on nitrification of applied ammonium
sulfate. Soil Biology and Biochemistry 17 (4), 547-
551.
Singh, R.P., Van den Beldt, Hocking, R., and Korwar, G.R.,
1986. Alley croppping in semi-arid regions of India,
pp. 109-122. In: Kang, B.T. and Reynolds, L. (eds.),
Alley farming in the humid and sub-humid tropics.
Proceeding of an international workshop held at the
International Institute of Tropical Agriculture (IITA),
Ibadan, Nigeria, March, 10-14
Singh, R.P., Ong, C.K., and Saharan, N., 1989. Above and
below-ground interactions in alley cropping in semi-
arid India. Agroforestry Systems 9 (3), 259-274.
Slatyer, R.O., 1967. Plant water relations. Academic
Press, New York.
Sowden, F.J., 1968. Soil organic matter and crop yields as
influenced by the frequency of summer fallowing.
Canadian Journal of Soil Science 48, 331-339.
Spedding, C.R.W., and Brockington, N.R., 1978.
Experimentation in agricultural systems. Agricultural
Systems 1, 47-56.
Stark, N.M., and Jordan, C.F., 1978. Nutrient retention by
the root mat of an Amazonian vegetation. Tropical
Ecology 59, 434-437.
Steiner, J.L., Day, J.C., Papendick, R.I., Meyer, R.E., and
Bertrand, A.R., 1988. Improving and sustaining
productivity in dryland regions of developing
countries. Advances in Soil Science 8, 79-122.
Stevenson, F.J., 1986. Cycles of soil carbon, nitrogen,
phosphorus, sulfur, micronutrients. John Wiley, New
York.

263
Sumberg, J.E., Maclntire, J., Okali, C. and Atta-Krah, A.N.,
1987. Economic analysis of alley farming with small
ruminants.. ILCA Bulletin 28, 2-6.
Summerell, B.A., and Burgess, L.W., 1989. Decomposition and
chemical composition of cereal straw. Soil Biology and
Biochemistry 21, 551-559
Swift, M.J., 1986. Enhancement of tropical soil fertility:
the role of biological research, pp. 17-27. In:
International Soil Reference and Information Center,
Wageningen, The Netherlands, Annual Report for 1985.
Swift, M.J., Russel-Smith, A., and Perfect, T.J., 1981.
Decomposition and mineral-nutrient dynamics of plant
litter in a regenerating bush-fallow in sub-humid
tropical Africa. Journal of Ecology 69, 981-995.
Swift, M.J., and Sanchez, P.A., 1984. Biological management
of tropical soil fertility for sustained productivity.
Nature and Resources 20(4), 2-10.
Swift, M.J., Heal, O.W., and Anderson, J.M., 1979.
Decomposition in terrestrial ecosystems. Studies in
ecology, Vol. 5. University of California Press,
Berkeley.
Szott, L.T., Fernandez, E.C.M., and Sanchez, P.A., 1990.
Soil-plant interactions in agroforestry systems. In:
Proceedings of the World Conference on Agroforestry
held at Edinburgh, 25-29 July, 1989. Department of
Forestry and Natural Resources, University of
Edinburgh, Scotland.
Szott, L.T., Palm, C.A., and Sanchez, P.A., 1991.
Agroforestry in acid soils of the humid tropics.
Advances in Agronomy 45, 275-301.
Ta, T.C., McDonald, F.D.H., and Faris, M.A., 1986.
Excretion of assimilated nitrogen from N2 fixed by
nodulated roots of alfalfa (Medicago sativa L.).
Canadian Journal of Botany 64, 2063-2067.
Taylor, B.R., Parkinson, D., and Parsons, W.E.J., 1989.
Nitrogen and lignin content as predictors of litter
decay rates: A microsm test. Ecology 70 (1), 97-104.
Tenant, D., 1975. A test of modified line intersect method
of estimating root length. Journal of Ecology 63, 995-
1001.

Teman, G.L., 1979. Volatilization losses of nitrogen as
ammonia from surface applied fertilizers, organic
amendments and crop residues. Advances in Agronomy 31,
189-223.
Tiedemann, A.R., and Klemmedson, I.O., 1977. Effect of
mesquite trees on vegetation and soils in the desert
grassland. Journal of Range Management 30, 361-367.
Tisdale, S.L., and Nelson, W.L., 1975. Soil fertility and
fertilizers. Macmillan, New York.
Toky, O.P., and Bisht, R.P., 1992. Observations on the
rooting patterns of some agroforestry trees in an arid
region of north-western India. Agroforestry Systems
18, 245-263.
Torres, F., 1983. Potential contribution of leucaena
hedgerows intercropped with maize to the production of
organic nitrogen and fuelwood in the lowland tropics.
Agroforestry Systems 1, 323-345.
Turner, N.C., 1981. Techniques and experimental approaches
for measurements of plant water status. Plant and Soil
58, 339-366
Vallis, I., 1978. Nitrogen relationship in grass/legume
mixtures, pp. 190-201. In: Wilson, J.R. (ed.), Plant
relations in pastures. CSIRO, Australia.
Vallis I., and Jones, R.J., 1973. Net mineralization of
nitrogen in leaves and leaf litter of Desmodium
intortum and Phaseolus atropurpureus mixed with soil.
Soil Biology and Biochemistry 5, 391-398.
Van der Kruijs, A.C.B.M., Van der Heide, J. and Kang, B.T.,
1984. Observations on decomposition rates of leaves of
several shrubs and tree species applied as mulch under
humid tropical conditions. In: Nitrogen management in
tropical soils. Proc. of International Conference held
at IITA, Ibadan, Nigeria.
Van Soest, P.J., 1968. Determination of lignin and
cellulose in acid-detergent fiber with permanganate.
Journal of the Association of Official Agricultural
Chemists 51, 780-785.
Verinumbe, I., Knipscheer, H.C., and Enabor, E.E., 1984.
The economic potential of leguminous tree crops in
zero-tillage cropping in Nigeria: A linear programming
model. Agroforestry Systems 2, 129-138.

265
Verinumbe, I., and Okali, D.U.U., 1985. The influence of
coppiced teak (Tectona granis, L.F.) regrowth and
roots on intercropped maize (Zea mays). Agroforestry
Systems 3, 381-386.
Vine, H., 1953. Experiments on the maintenance of soil
fertility at Ibadan, Nigeria, 1922-51. Empire Journal
of Experimental Agriculture 21, 65-85.
Wacquant, J.P., Oukinder, M., and Jacquard, P., 1989.
Evidence for a periodic excretion of nitrogen by roots
of grass-legume associations. Plant and Soil 116, 57-
68.
Wake, J.L., Kiker, C.F,. and Hildebrand, P.E., 1988.
Systematic learning of agricultural technologies.
Agricultural Systems 1, 47-56.
Weatherley, P.E., 1950. Studies in the water relations of
the cotton plant. 1. The field measurement of water
deficits in leaves. New Phytologist 49, 81-97.
Weeraratna, C.S., 1979. Pattern of nitrogen release during
decomposition of some green manures in a tropical
alluvial soil. Plant and Soil 53, 287-294.
Wieder, R.K., and Lang, G.E., 1982. A critique of the
analytical methods used in examining decomposition data
obtained from litter bags. Ecology 63(6), 1636-1642.
Wiersum, K.F., 1984. Surface erosion under various tropical
agroforestry systems, pp. 231-239. In: O'Loughlin,
C.L. and Pierce, A.J. (eds), Symposium on effects of
forest land use on erosion and slope stability, East-
West Center, Honolulu, Hawaii.
Willey, R.W., 1979. Intercroppingits importance and
research needs. Part 1. Competition and yield
advantages. Field Crops Abstracts 32 (1), 1-10.
Williams, S.T., and Gray, T.R.G., 1974. Decomposition of
litter on the soil surface. pp. 611-632. In:
Dickinson, C.H., Pugh, G.J.F., (eds), Biology of plant
decomposition. Vol. 1. Academic Press, New York.
Wilson, G.F., and Kang, B.T., 1982. Developing stable and
productive biological cropping systems for the humid
tropics, pp.193-204. In: Stonehouse, B., (ed.),
Biological husbandry. Butterworth, London.

266
Wilson, G.F., Kang, B.T., and Mulongoy, K., 1986. Alley
cropping trees as sources of green-manure and mulch in
the tropics. Biological Agriculture and Horticulture
3, 251-267.
Wilson, J.R., Hill, K., Cameron, D.M., and Shelton, H.M.,
1990. The growth of Paspalum notatum under the shade
of a Eucalyptus granis plantation canopy or in full
sun. Tropical Grasslands 20, 134-143.
Witkamp, M., 1966. Decomposition of leaf litter in relation
to environment, microflora, and microbial respiration.
Ecology 47, 194-202.
Woomer, P.L., and Ingram, J.S.I., (eds.), 1990. The biology
and fertility of tropical soils. Report of the
Tropical Soil Biology and Fertility Program (TSBF).
UNESCO-ROSTA, Nairobi, Kenya.
Yamoah, C.F., Agboola, A.A., and Mulongoy, K., 1986a.
Decomposition, nitrogen release and weed control by
pruning of selected alley cropping shrubs.
Agroforestry Systems 4(3), 239-246.
Yamoah, C.F., Agboola, A.A., and Wilson, G.F., 1986b.
Nutrient contribution and maize performance in alley
cropping systems. Agroforestry Systems 4, 247-254.
Yamoah, C.F., Agboola, A.A., Wilson, G.F., and Mulongoy, K.,
1986c. Soil properties as affected by the use of
leguminous shrubs for alley cropping with maize.
Agriculture, Ecosystems and Environment 18, 167-177.
Young, A., 1989. Agroforestry for soil conservation. CAB
International, Wallingford, U.K. and ICRAF, Nairobi.

BIOGRAPHICAL SKETCH
Bashir Jama Adan was born on July 7, 1957 to a nomadic
family on the outskirts of Modogashe, Garissa District,
North Eastern Province, Kenya. Between 1966 to 1972, Bashir
completed his Certificate of Primary Education at Garissa
Primary School. Between 1973 and 1978, Bashir obtained his
Secondary and High School Certificates from Alliance High
School, Kikuyu, Kenya. Bashir attended the University of
Nairobi, Kenya, between 1979 and 1982, where he earned a
Bachelor of Science degree (honors) in agriculture. Bashir
completed his Master of Science degree in agronomy in 1988,
received also from the University of Nairobi. In 1982, he
was employed by the Ministry of Energy, Kenya, as the center
manager for Mtwapa Agroforestry/Energy Center, Coast
Province. In 1982, he assumed the position of research
assistant at the International Centre for Research in
Agroforestry (ICRAF), Nairobi, Kenya. In 1989, Bashir
initiated studies at the University of Florida for a
doctoral degree in Forestry Resources and Conservation (with
a specialization in Agroforestry).
267

I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality, as
a dissertation for the degree of Doctor of Philosophy.
Q )
P.K.R. Nair, Chair
Professor of Forest Resources
and Conservation
I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality, as
a dissertation for the degree of Doctor of Philosophy.
H.L. Gholz /
Professor of Forest Resources
and Conservation
I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality, as
a dissertation for the degree of Doctor of Philosophy.
C.P.P. Reid
Professor of Forest Resources
and Conservation
I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality, as
a dissertation for the degree of Doctor of Philosophy.
i&L. Pd'penoe /
Professor of Soil Science

I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality, as
a dissertation for the degree
P.E. Hildebrand
Professor of Food and Resource
Economics
This dissertation was submitted to the Graduate
Faculty of the School of Forest Resources and Conservation
in the College of Agriculture and to the Graduate School and
was accepted as partial fulfillment of the requirements for
the degree of Doctor of Philosophy.
May 1993
Conservation
Dean, Graduate School



200
trials also have limitations. For instance, loss of farmer
participation is sometimes cited as a major problem of on-
farm trials (Lightfoot, 1987).
Because of the long-term nature of agroforestry
practices, strategies different from those of systems
involving only annual crops may be necessary to enhance
adoption. For example, conducting farm surveys specific to
a given agroforestry practice (e.g., alley cropping) may
help capture details of the farming systems of an area that,
when included in the development of the technology on-
station, could enhance its potential for adoption by
farmers.
Being cognizant of the above issues, as well as
fulfilling a need to rationalize an on-going on-station
experiment on soil fertility and productivity aspects of
alley cropping Leucaena leucocephala and Cassia siamea at
the Field Station of the International Center for Research
in Agroforestry (ICRAF), Machakos, Kenya, I conducted an on-
farm survey. Sixty farmers in the neighborhood of the Field
Station were interviewed in July-August 1991. The overall
objective of the study was to assess the potential for
farmer adoption of alley cropping of multipurpose trees and
shrubs.

The specific objectives were: (a) to identify current
strategies employed by farmers to improve the fertility and
productivity of their soils, (b) to determine factors


Maize grain yield (t ha
42
Figure 2-5. Relationship between rainfall and maize
mean yield of all treatments in each of the six cropping
seasons. Numbers on the regression line refer to the
cropping seasons.


132
Averaged across species, 58% of the mulch N was
mineralized in the first four weeks. In contrast, maize
biomass yield and N uptake were highest during weeks 8-12
after mulch or fertilizer application. Hence, the peaks of
N release from the mulch and uptake by the maize were not
matched or synchronized. Although the N release patterns of
the two species were similar, the differential response of
maize biomass yield to the mulch of the two species in weeks
4 and 16 is worth noting. At both times, the biomass yield
of maize was significantly more with leucaena than with
cassia mulch.
Discussion
Patterns of Mulch Decomposition: Effects
of Quality and Soil placement Methods
The chemical properties of the mulch of the two
species including the relatively high polyphenol of leucaena
leaves are withip the range reported by others (Ahn et al.,
1989; Fox et al., 1990; Palm and Sanchez, 1991; Oglesby and
Fownes, 1992). The patterns of biphase mulch decomposition
consisting of an initial phase of rapid dry matter loss
followed by a second phase of lower rate has also been noted
by other workers working with different materials and in
different ecozones (e.g., Parker, 1962; Van der Kruijs et
al., 1984; Swift, 1986; Kumar and Deepu, 1992; and several
others). The first phase reflects the rapid loss of easily
decomposable materials (carbohydrates) and the second phase,
the slow loss of more recalcitrant materials (Wieder and


68
Maize-yield reduction observed under leucaena was not
likely due to competition for light, because the hedges were
pruned frequently during the rainy season. Also, it was
probably not due to competition for nutrients because the
plots were fertilized and leucaena (which was inoculated at
the start of the study) may have fixed considerable levels
for its own N needs. Instead, the reduction in yield was
probably most likely due to competition for water, a
suggestion also made by others regarding leucaena alley
cropping in semiarid India (Rao et al., 1991). In the sixth
season when rainfall was 39% below normal, the significantly
lower maize yields under leucaena alley cropping compared to
the control provide support to the explanation of water
limitation. The argument that competition was more for
water and less for nutrients was supported by the results of
additional field studies (Chapter 5) involving factorial
combination of fertilizer (N and P) and irrigation levels
where a positive response in maize yield to N application
was not significantly different from that due to the effects
of irrigation only. Prior to the initiation of the present
study, the site was a savanna woodland cropped to pigeon pea
for one-and-one-half years, so that soil N may still have
been relatively high. Another evidence for water being more
limiting than nutrients is provided by the linear
relationship between the maize yield and rainfall over the
six seasons.


95
Soil Physical Properties
The lack of significant change in soil bulk density
could, in turn, be ascribed to lack of change in soil
organic matter. Like many other soil properties, soil bulk
density is influenced by changes in soil organic matter
(Allison, 1973). Several workers have noted significant
improvements in soil bulk density in the humid lowlands of
Nigeria where mulch yields as high as 8-10 t ha"1 yr"1 (dry
matter) are obtained from intercropped hedgerows,
particularly leucaena in studies spanning four years (Wilson
and Kang, 1982; Yamoah et al., 1986c).
Conclusions
No significant changes were observed in soil organic
C, KCl-extractable N, CEC and bulk density after three
years. This was so with or without alley cropping of
Leucaena leucocephala and Cassia siamea. The low amounts of
mulch obtained from the hedgerows (coupled with the
relatively high rate of decomposition of the mulch) appeared
to be the main limitation to significant improvements in the
levels of the soil chemical and physical properties
monitored. Although the changes in organic C and total-N
over time were not significant, the small amounts of mulch
(particularly those of leucaena) resulted in significantly
higher relative index of N03" N and NH4+ N availability
compared to the plots with no additions of mulch.


86
difference between the cassia mulch-only plot and the
similarly treated leucaena plot was, for instance, 62%.
This difference was observed even when the initial (1987)
soil P and percent clay content were used as covariates for
the final (1991) P measurements.
The higher P in plots with cassia mulch could be
attributed to increased concentrations of P observed in
cassia mulch over the years (Table 3-5). Cassia mulch had
24% higher P content than leucaena in 1991. In contrast,
plots with leucaena mulch had significantly higher KCl
extractable N03' N and NH4+ N than the plots with cassia
mulch or the absolute control (Table 3-6). The levels of
relative indices of N03" N and NH4+ N in the plots with
leucaena mulch were, however, not significantly different
from those of plots with fertilizer.
Soil Bulk Densities
No significant difference was observed in soil bulk
density of the different treatments (Table 3-2). Also, no
significant change was observed in the bulk density of any
treatment over the three years of study.
Discussion
Soil Chemical Properties
Despite the differences in the rates and, to some
extent, in the quality of leucaena and cassia mulch


59
especially of cassia, observed in the hedges planted in
block/sole stands compared to those intercropped.
The high pruning frequency (4-6 times a year) may also
have stressed the cassia hedges more than leucaena, and
subsequently made cassia less resistant to pest attack.
Another experiment on the site of the present study (P.
Kiepe, personal communication) showed that with only two
prunings per year, mulch yields similar to those obtained in
the present study could be obtained from intercropped
hedgerows of cassia. Nevertheless, reduced frequency of
pruning may exaggerate crop loss under alley cropping due to
enhanced shading.
It is also possible that the frequent pruning and
"export" of mulch from the area of the block planted hedges
impoverished the site in terms of nutrient storage, which in
turn contributed to the low yields of the block planted
hedges.
Crop Yield
Effects of mulch on yield
Because yields of maize in the mulched and unmulched
control plots were comparable, it would appear that low
levels of mulch obtained from the hedgerows limited maize
yields. Significant response in biomass yield of maize to
increasing rates of mulch in the pot studies also supports


266
Wilson, G.F., Kang, B.T., and Mulongoy, K., 1986. Alley
cropping trees as sources of green-manure and mulch in
the tropics. Biological Agriculture and Horticulture
3, 251-267.
Wilson, J.R., Hill, K., Cameron, D.M., and Shelton, H.M.,
1990. The growth of Paspalum notatum under the shade
of a Eucalyptus granis plantation canopy or in full
sun. Tropical Grasslands 20, 134-143.
Witkamp, M., 1966. Decomposition of leaf litter in relation
to environment, microflora, and microbial respiration.
Ecology 47, 194-202.
Woomer, P.L., and Ingram, J.S.I., (eds.), 1990. The biology
and fertility of tropical soils. Report of the
Tropical Soil Biology and Fertility Program (TSBF).
UNESCO-ROSTA, Nairobi, Kenya.
Yamoah, C.F., Agboola, A.A., and Mulongoy, K., 1986a.
Decomposition, nitrogen release and weed control by
pruning of selected alley cropping shrubs.
Agroforestry Systems 4(3), 239-246.
Yamoah, C.F., Agboola, A.A., and Wilson, G.F., 1986b.
Nutrient contribution and maize performance in alley
cropping systems. Agroforestry Systems 4, 247-254.
Yamoah, C.F., Agboola, A.A., Wilson, G.F., and Mulongoy, K.,
1986c. Soil properties as affected by the use of
leguminous shrubs for alley cropping with maize.
Agriculture, Ecosystems and Environment 18, 167-177.
Young, A., 1989. Agroforestry for soil conservation. CAB
International, Wallingford, U.K. and ICRAF, Nairobi.


Carbon {log g) remaining
2.0
117


166
Table 5-3. Water content of whole maize plants sampled near
(0.45 m) and away (3.5 m) from alley cropped
hedgerows of Leucaena leucocephala and Cassia
siamea hedgerows, sixth season, 1991.
Water content
(%)
Near
Away
Species
(0.45 m)
(3.5 m)
Mean
Leucaena
81.0
79.3
80.2a
Cassia
86.6
83.0
84.5a
Mean
83.5
81.2
82.3
aMeans in a column and row followed by the same letters are
not significantly different (p *= 0.05).


249
Brady, N.C., 1990. The nature and properties of soil. 10th
Edition. Macmillan Publishing Co., New York.
Brewbaker, J.L., 1987. Leucaena: a multipurpose tree genus
for tropical agroforestry, pp 289-323. In: Steppler,
H.A. and Nair, P.K. (eds.), Agroforestry: a decade of
development. ICRAF, Nairobi, Kenya.
Bromfield, A.R., 1969. Uptake of phosphorus and other
nutrients by maize in Western Nigeria. Experimental
Agriculture 5, 91-100.
Brophy, L.S., and Heichel, G.H., 1989. Nitrogen release
from roots of alfalfa and soybean grown in sand
culture. Plant and Soil 16, 77-84.
Budelman, A., 1988. The decomposition of the leaf mulches
of Leucaena leucocephala, Gliricidia sepium and
Flemingia macrophylla under humid tropical conditions.
Agroforestry Systems 7, 33-45.
Caldwell, M.M., and Richards, J.H., 1989. Hydraulic lift:
water efflux from upper roots improves effectiveness of
water uptake by deep roots. Oecologia 79, 1-5.
Campbell, C.A., 1983. Soil organic carbon, nitrogen and
fertility, pp. 173-291. In: Schnitzer, M. and Khan,
S.U. (eds.), Soil organic matter. Developments in
Soil Science, No. 8, second impression. Elsevier
Science Publishers, Amsterdam.
Chambers, R., 1985. Shortcut methods of gathering social
information for rural development projects. In:
Cernea, M.M., (ed.), Putting people first
Sociological variables in rural development. Oxford
University Press, New York.
Chapin, F. S., 1980. The mineral nutrition of wild plants.
Annual Review of Ecology and Systematics 11, 233-260.
Cheshire, M. V., Mundie, C.M., and Shepherd, H., 1974.
Transformations of sugars when rye hemicellulose
labelled with 14C decompose in soil. Journal of Soil
Science 25, 90-95.
Chirwa, P.W., 1991. Soil moisture changes and maize
productivity under alley cropping with leucaena and
Flemingia hedgerows in semi-arid conditions in Lusaka,
Zambia. M.S. Thesis, University of Florida,
Gainesville.
Cochran, V.L., 1991. Decomposition of barley straw in a
sub-arctic soil in the field. Biology and Fertility of
Soils 10, 227-232.


25
calculate the per hectare yield of maize planted next to the
block planting system of hedges outside the crops.
In the section on results below, only yields of the
second season to the sixth season crops are discussed. The
yield of the first season was excluded because the hedgerows
were pruned only once (since they were not yet well-
established). In the other seasons the hedgerows were
pruned 2-3 times. Also, the crop yield of the first season
crop was at least twice as much as the yield of the other
seasons, ascribed more to the initial site conditions than
to the effects of treatments. Detailed presentation and
discussion of the biomass and crop yields are limited to the
6.7 m wide alley in the block planting system. This spacing
of the alleys and ratio of tree:crop land occupancy offers
the most land for cropping. This is of practical
significance to farmers.
Trends in Maize Yield over Time
Trends in maize yield over time were analyzed by
expressing the yield of each treatment in a season as a
percent of the average yield of all treatments of the second
F C:
crop. The patterns of maize yield over time were also
(
examined by expressing the yields of the two maize planting
systems as a percent of the yields of the fertilizer-only
control treatment.


Page
4-1. Monthly rainfall and mean temperature during
the study, 1991 101
4-2. Decomposition patterns of the leaves of
Leucaena leucocephala and Cassia siamea. ... 113
4-3. Decomposition patterns of the twigs of Leucaena
leucocephala and Cassia siamea 117
4-4. Decomposition patterns of the mulch (leaves
plus twigs) of Leucaena leucocephala and
Cassia siamea 120
4-5. Cumulative percent of N mineralized from soil
incorporated mulch of Leucaena leucocephala and
Cassia siamea after 16 weeks ......... 123
4-6. Relationship between the patterns of N release
from the mulch of Leucaena leucocephala and
Cassia siamea and biomass yield of maise
from four consecutive crops of maize sown
in pots and harvested at an interval of
four weeks 126
5-1. Maize yield response to distance from alley
cropped hedgerows of Leucaena leucocephala
and Cassia siamea, 1990 short-rains 159
5-2. Maize yield response to distance from alley
cropped hedgerows of Leucaena leucocephala
and Cassia siamea, 1991 long-rains ...... 160
5-3. Maize yield response to distance from alley
cropped hedgerows of Leucaena leucocephala
and Cassia siamea, 1991 short-rains 161
5-4. Bulk density of top 5 cm soil profile
near (0.45 m) alley cropped hedgerows of
Leucaena leucocephala and Cassia siamea,
1991 short-rains 170
5-5. Infiltration rates of water into the soil
near (0.45 m) and away (3.5 m) from alley
cropped hedgerows of Leucaena leucocephala
and Cassia siamea, 1991 short-rains 172
5-6. Soil water content near (0.45 m) alley cropped
hedgerows of Leucaena leucocephala and Cassia
siamea determined at the end of 1991
long-rains ....... 174
xii


93
intercropped or received mulch only) had relatively higher
available N than those with cassia (Table 3-6). This could
be attributed to the higher yield of leucaena mulch which
also had higher concentration of N and higher rate of
decomposition (particularly the leaf fraction) than cassia
mulch. It is also possible that apart from the release of
its own N upon decomposition, some of the higher effects of
the mulch of leucaena on the relative index of available N
was to stimulate more mineralization of the native soil N.
This could be through the so-called "priming" effect of
added N on native soil N (Hauck and Bremner, 1976; Jenkinson
et al., 1985) .
It does appear that, besides N, the small quantity of
mulch applied (particularly that of cassia) had significant
effects on the availability of top soil P. This was most
pronounced where only mulch of cassia was applied. Similar
observations have been made by Yamoah et al. (1986c) from
the humid lowland tropics of Nigeria. The increase in soil
P under cassia plots could be ascribed to enhanced P
contribution from the mulch. Over the three years, the
concentration of P in cassia increased considerably while
that of leucaena remained unchanged. Explanation for the
higher concentration of P in cassia mulch compared to that
of leucaena is a matter of speculation; some hypothesis are
differences in root to shoot ratios, mycorrhizal
association, microbial population and activity (e.g., P-
solubilizing bacteria).


131
Table 4-5. Relative indices of cumulative NH4+ N, N03~ N and
total N mobilized after soil incorporation of 2 t
ha*1 equivalents of Leucaena leucocephala and
Cassia siamea mulch in pots with maize, second
season of 1991.
Treatment
nh4- n
(yg/g
resin)
no3* n
(pg/g
resin)
Total N
mineral
ized
(pg/g
resin)
Leucaena
66.0*
106.9*
172.9*
Cassia
63.8*
39.9*
103.7b
Fertilizer
124.6b
101.9*
226.5C
Control
35.8C
29.9b
65.7d
*,b'C'dMeans in column followed by different letters are not
significantly different (p = 0.05).


242
(2 t ha'1 yr'1) ; also leucaena had higher fine root density
than cassia in the top soil (0-40 cm). Because of these,
leucaena was more competitive than cassia with maize. The
difference between the effects of the two species indicates
how performance of alley cropped hedgerows depends on the
tree species used, and suggests the need to include more
than one species in alley cropping studies. It also
suggests that management of the hedgerows may have to be
different even in the same experiment. It is possible that
leucaena would have been less competitive than it was in the
present study if the pruning frequency was more than 4-6
times a year. The use of more drought tolerant hedgerow
species and annual crops than leucaena, cassia and maize
could also have changed the results of the present study.
The reduced yield of maize under alley cropping
leucaena was due more to competition for water, and less for
nutrients since the plots were fertilized with both N and P.
Competition for light was probably negligible given that the
hedgerows were regularly (at least twice in a season) pruned
low (to 50 cm heights above ground). The presence and
effects, if any, of allelopathic chemicals was not studied.
Interestingly, maize yield nearer the hedgerows yielded
more than those away from them. The higher yield near the
hedgerows was attributed to higher water content near the
hedgerows than away from them. This resulted from lowered
soil bulk density and hence higher rates of infiltration


246
hedges due to mortality of cassia with time. Up to 5% and
11% of the initial stands of the intercropped and block
planted hedgerows of cassia were dead 3-4 years after
planting. If it is economically positive and problems of
mortality are minimal or controllable, cassia alley cropping
would be recommendable on all farms, and particularly where
there are interests in mulch. It could also be recommended
for soil erosion control on all farms.
Most of the farmers were interested in the fodder than
the mulch aspects of alley cropping. Leucaena has fodder
value, cassia does not. But leucaena alley cropping is
favorable to crop yields. Therefore, planting of leucaena
separately from crops, perhaps on lands not suitable for
cropping, would be recommendable. Because of the initial
high investments required to establish the hedgerows, alley
cropping or block planting systems of hedgerows is more
likely to be adopted by relatively resourceful farmers such
as coffee growers than by poorer farmers. Termites are
major problems to on-farm tree planting. Therefore,
research into termite control methods that can be afforded
by farmers and that are also not environmentally hazardous
are necessary.


53
leucaena mulch (127%). The interaction between N and P
uptake/removal by the crop was not examined in this study.
Root Density
Root density differed substantially between species
and also between soil depths. In both species, the
densities of both fine- and medium-sized roots were
significantly higher near than away from the hedgerows (Fig.
2-9). Fine root densities of leucaena at both 0-20 cm and
20-40 cm soil depths were significantly higher at nearly all
distances from the hedgerows than that of cassia (Table
2-8). Interaction effects between species and soil depths
on root density were not significant. In both species, root
density at 20-40 cm depth was significantly higher than at
0-20 cm.
Pot Study
During four harvests, the species effects and the
interaction effects of species and mulch rate on the biomass
yield of maize were not significant (Fig. 2-10). However,
effects of the mulch rates on maize yield were highly
significant, with the yields increasing direct proportion to
mulch rates (Fig. 2-10). Averaged across species and
irrigation levels (i.e., the eguivalents of 100 and 200 mm
of rainfall in two months period), biomass yield of maize
from the 0 and 2 t ha"1 mulch rates were not significantly


235
from noncoffee enterprises is, however, unlikely to have
significant impact as a cash generating means unless they
receive similar support to that of coffee in terms of
research, extension, and marketing. Poor marketing and over
production can lead to losses. For instance, there have
been occasions, when mango and guava trees were in season,
fruit was rotting in the fields. Only farmers who could
afford to market their fruits outside the local market would
survive under such conditions of glut and poor marketing
opportunities. Fodder from hedgerows, if grown for sale,
could' have marketing problems over time similar to those of
fruits. This, however, is difficult to speculate about
since there have hardly been any fodder products from the
farms presented for sale at the markets.
Implications of Availability of
Trees and the Problems of Termites
on the Adoption of Alley Cropping
Seeds and seedlings
The question of availability of planting materials
(seeds or seedlings) was a major bottleneck to the farmers'
efforts to plant trees. To enhance availability of planting
materials, some farmers maintained their own nursery in
spite of the presence of various constraints. The question
of availability of planting materials is one that deserves
more discussion since it has implications on both the
research and development aspects of alley cropping. Without


78
Materials and Methods
Site Description
The location, rainfall and soil characteristics of the
site are described in Chapter 2. Table 3-1 shows the chemical
and physical characteristics of the top soil (0-20 cm) at the
beginning of the study (October, 1987).
Treatments and Experimental Layout
The experimental treatments consisted of two systems of
Leucaena leucocephala and Cassia siameai (a) intercropped
hedgerows and (b) planted as blocks/sole stands outside the
crop. In the intercropped system, the hedgerows were 6.67 m
wide. The hedgerows in both the intercropped and the
block/sole planting systems occupied 15% of the total area,
and the crop occupied the remainder of the land (85%). Plots
with fertilizer-only and mulch of leucaena or cassia only were
included as controls. As an absolute control, no-mulch or
fertilizer plots from an adjacent experiment was used. The
experimental design was randomized block design with three
replications. The field experiment was conducted for more
than 3 years (October 1987 to March 1992).
Crop and Tree Management
In all systems, the soil was hand tilled and maize
(Katumani Composite var.) planted twice a year at a stand of


Maize yield (kg/5 m long row)
j
161
Figure 5-3. Maize yield response to distance from
alley cropped hedgerows of Leucaena leucocephala and Cassia
siamea, 1991 short-rains. Vertical bars indicate standard
error of difference of means.


189
While trends (at times significant) of higher water
content near the hedgerows were detected with the
gravimetric method, similar trends were not observed with
the tensiometers placed at 15 cm depths. Tensiometers
measure water potentials of the soil at or close to the
point of placement of the porous cup tip. They do not
measure or include the water content of the soil above the
depth of placement. On the other hand, the gravimetric
method integrates the whole depth of the soil considered.
The differences in the mode of operation of the two methods
could explain the apparent discrepancy in the results
obtained with the two methods. The differences also
suggested that gravimetric methods would be recommendable
over tensiometer methods for soils with limited water
content. This is because the gravimetric method integrates
more volume of soil and hence is sensitive enough to detect
small quantities of soil water rather than tensiometers.
Water status near the hedgerows could have been
enhanced in several ways. High infiltration and hydraulic
conductivity were probably the principal processes. Other
studies also show that the hedgerows could physically
intercept as much as 20% of the season's rainfall (Monteith
et al., 1991), especially where the hedgerows are planted
perpendicular to the direction of the prevailing winds.
Under conditions of moisture stress, rainfall intercepted by
the hedgerows could increase yield of the hedgerows and,


130
and significantly lower than that from the mulch of leucaena
and fertilizer.
Relative Indices of N-availabilitv from the
Mulches; N-accumulation bv Ion-exchanae Resins
After 16 weeks, the index of total-N mobilized from
leucaena mulch was significantly higher than that with cassia
(Table 4-5). Soil with leucaena mulch had significantly
higher levels of nitrate-N index than that with cassia.
However, the differences in the indices of ammonium-N
accumulation between the soil with leucaena andcassia mulch
were not significant. The mulch of both species resulted in
generally higher indices of N-availability than the control
plots. The indices of both total-N and ammonium-N mobilized
were the highest for the soil of the fertilized pots after 16
weeks and lowest in the control pots.
Synchrony between N uptake bv maize
and N mineralization of the mulch
Synchrony (or lack thereof) between the periods of peak
of N release from the mulch of cassia and leucaena with peak
biomass and N-yield of the maize was determined (Fig. 4-6), by
superimposing the N release curves of the soil-incorporated
leucaena and cassia mulch in the second season field
decomposition studies, on the biomass yield of maize from the
pot studies (Table 4-4) for each of the four weekly harvests.


Maize yield (kg/5 m long row)
160
Figure 5-2. Maize yield response to distance from
hedgerows of Leucaena leucocephala and Cassia siamea, 1991
long-rains. Vertical bars indicate standard error of
difference of means.


177
Month
Figure 5-8. Difference in gravimetric water content
(%, w/w) between soil sampled at 0-20 cm depth near (0.45 m)
and away (3.5 m) from alley cropped hedgerows of Leucaena
leucocephala and Cassia siamea in the long- and short-rain
seasons of 1991.


110
procedure. Differences between treatments were declared
significant at p = 0.05.
Results
Field Studies: Patterns of Mulch
Decomposition and N-mineralization
In general, the patterns of decomposition of leucaena
and cassia mulch (i.e., leaves plus twigs), the leaves and
twigs separate, as well as the filter paper were biphasic.
The parameters of the two-phase regression lines fitted
(i.e., the intercept, the slopes, the spline point of the
two phases and the coefficients of determinations) of the
mulch material are presented in Table 4-2. With the
exception of leucaena leaves, all other materials had an
initial phase of rapid decline rate followed by a second
phase of comparatively lower rate. Leucaena leaves had only
one phase of decomposition.
Effects of Placement Method:
Soil Surface or Incorporation
Leaves
Significant interactive effects were observed between
species and soil placement method on the rate of
decomposition of the leaves (Fig. 4-2). No significant
difference was detected between the rates of decomposition


Rainfall (mm)
101
Month
Rainfall
Temperature (C)
Figure 4-1. Monthly rainfall and mean temperature
during the study, 1991.
Temperature (C)


85
Table 3-4. Changes in soil Mehlich-1 P levels after three
years of alley cropping hedgerows of Leucaena
leucocephala and Cassia siamea hedgerows with
maize, November 1991.
System*
Change in soil
Mehlich-1 P
(ppm)
Aliev intercroDoina
Leucaena leucocephala
4.0a
Cassia siamea
11.0b
Controls
Mulch (Leucaena leucocephala)
-1.3a
Mulch (Cassia siamea)
17.3a
Fertilizer-only
CD

in
cr
Maize-only
-1.7*
*The values of the block planting system were similar.
a,bValues within a column with different letters are
significantly different (p = 0.05).


224
Table 6-6.
Functions of tree nurseries possessed by 30% of
the 60 farmers interviewed, Machakos, Kenya,
August 1991.
Function of
nursery
Distribution of farmers
with tree nurseries
(%)
Fruit, Firewood, Timber
11.7
Fruit
18.3
Total
30.0


140
leucaena twigs had slightly more lignin and polyphenol than
cassia, the combined lignin and polyphenol content (Fox et
al., 1990) could explain the lack of significant differences
in the rates of decomposition and N-mineralization of the
mulch (leaves plus twigs) from the two species. This
explanation is based on results from the leaves of only two
species and may not extend to other species whose C:N ratio,
lignin, polyphenol concentration and age are outside the
range considered.
For materials with high N concentration (i.e., narrow
C:N ratio) such as leucaena leaves used in this study or
leaves and small twigs of alfalfa (Medicago sativa L.) the
second phase of decomposition may not even exist or be
undetectable (Fox et al., 1990). Since the phases as well
as their rates of decomposition have important implications
on the availability of nutrients from the mulch to the crop,
it is important to firmly establish the relative importance
of each factor involved. This my be achieved through a
study involving a wide range of species in which both the
quantity and the quality of the chemical characteristics are
continually monitored.
Implications of Mulch-N Mineralization
Patterns on Uptake; Potentials for Synchrony
Synchrony is said to be achieved when the peaks of N
release of N from the mulch and demand or uptake by the
crop are matched. In the present study, the peaks were
not synchronized as is demonstrated when the patterns


165
Table 5-2. Relative water content of flag leaves of maize
plants sampled near (0.45 m) and away (3.5 m)
from alley cropped hedgerows of Leucaena
leucocephala and Cassia siamea on April 26, 1991
(sixth season).
Water content
m
Near
Away
Species
(0.45 m)
(3.5 m)
Mean
Leucaena
40.2
.34.1
37.2a
Cassia
48.7
46.8
47.8b
Mean
44.4a
40.5a
42.5
a,bMeans in a column and row followed by the same letters are
not significantly different (p = 0.05).


seasons, systems, and hedgerow species on crop yield.
Therefore, ANOVA by season was necessary in order to
understand the true main effects of species and systems.
The ANOVA by season showed that in four out of the six
(i.e., seasons 2, 4, 5, and 6), alley cropped maize yielded
higher with cassia than with leucaena (Table 2-3). A
significant difference between the yield of leucaena alley
cropped maize and that next to the block planted system of
leucaena was observed in only two out of the six seasons
(the third and the fourth). For cassia, the difference was
also significant in two out of the six seasons (the second
and the third). The difference in maize yield under the two
systems represents the in situ effects of the hedgerows on
crop yield. Averaged over the seasons, the in situ effects
of cassia on maize yields were 22% while those of leucaena
were 9%.
Except in the second season, differences between the
effects of the three hedgerow spacings (i.e., 4, 5, and 6.7
m) on maize yield were not significantly different (Fig.
2-3). In the second season, maize yields under the 6.7 m
hedgerows were significantly higher than those of the other
two spacings. In the block planting systems, significant
differences in maize yields were observed in seasons 2 and 4
only and 4 only (Fig. 2-4).
A comparison of the mean yields of all the treatments
(Table 2-3, last column) shows that the fertilized maize


168
Table 5-4. Relative water content of flag leaves of maize
plants sampled near and away from Leucaena
leucocephala and Cassia siamea hedgerows on May
28, 1991 (sixth season).
Water
content (%}
Species
Near
(0.45 m)
Away
(3.5 m)
Mean
Leucaena
28.3a
45.3b
38.8
Cassia
58.4a
55.4b
56.9
Mean
43.4
50.4
46.9
a,bMeans in a row followed by the same letters are not
significantly different (p = 0.05).


240
cassia was not, it is suggested that alley cropping cassia
could be targeted to all farmers, and specifically, to
farmers with mulch-interest. On the other hand, leucaena
could be targeted mainly to farmers with fodder-interest.
However, the research suggested that leucaena be planted
separately from the crops. Notwithstanding the above
recommendation, cassia alley cropping has to be viewed
against the magnitude of gain in crop yield (8% over the
control), problems of pests, and mortality associated with
cassia (which may make replanting necessary after 3-4
years), and the costs and availability of the large number
of seedlings required.
Ninety-two percent of the farmers interviewed let
their livestock into the fields after the harvest of crops.
Therefore, whether alley cropped or planted separately from
crops, the trees/hedges are likely to be damaged by
livestock grazing in the crop fields.


98
1983) and the organic components, such as lignin (Alexander,
1977, Herman et al., 1977), polyphenols (Vallis and Jones,
1973; Palm and Sanchez, 1991) and soluble carbohydrates
(Cheshire et al., 1974; Wieder and Lang, 1982), and by
environmental and management factors (Wilson et al., 1986).
The rates at which mulches decompose have important
implications on soil organic matter and N dynamics. Slowly
decomposing materials may increase soil organic matter, at
least in temperate conditions (Dehaan, 1976; Silvapalan,
1982; Kelly and Stevenson, 1987) but crop yields may decline
due to insufficient availability of N. On the other hand,
fast decomposing materials may provide a short-term increase
in N but may have little effect on the maintenance of soil
organic matter content (Vine, 1953; Stevenson, 1986). Even
from fast decomposing mulches, the efficiency of nutrients-
use, in particular of N, by crops is low compared to
inorganic fertilizers (Fox et al., 1990; Gutteridge, 1992).
Most field studies of decomposition of hedgerow
prunings or mulches have been made in the humid tropics
(Yamoah et al., 1986a; Swift et al. 1981; Palm and Sanchez,
1991). Similar studies in the semiarid tropics are scarce.
Although some field studies on decomposition of hedgerow
prunings have been conducted in the semiarid tropics
(Mugendi, 1990), they have not considered characterization
of plant and environmental factors that regulate the
process. There could be similarities in the patterns and
rates of decomposition of prunings in the humid and semiarid


205
Table 6-1
Farm and household characteristics of the 60
farmers interviewed at Machakos, Kenya, August
1991.
Farm
size
range
(ha)
Cumu
lative
distri
bution
(%)
Average
number of
members
in a
house
hold
Average
number of
children
in a
household
Household
members
that were
children
(<18 yrs)
(%)
0-4
77
13 (4)
7 (4)
54
5-8
17
16 (3)
8 (3)
50
9-17
.4 6
14 (6)
7 (5)
50
NOTE:
Figures in
parentheses are
the standard
deviations
of the mean.


253
Hildebrand, P.E., 1982. Farming systems research: issues
in research strategy and technology design: discussion.
American Journal of Agricultural Economics 64 (5), 905-
906.
Hildebrand, P.E., 1990. Farming Systems research
extension, pp. 131-143. In: Jones, J.G.W and Street,
P.R. (eds.), Systems theory applied to agriculture and
the food chain. Elsevier Applied Science, Amsterdam.
Hildebrand, P.E., and Poey, F., 1985. On-farm agronomic
trials in farming systems research and extension.
Lynne Reinner Publishers, Inc., Boulder, Colorado.
Hoekstra, D., 1985. The use of economics in diagnosis and
design of agroforestry systems. Working Paper 29,
ICRAF, Nairobi, Kenya.
Hook, R.I.V., Johnson, D.W., West, D.C. and Mann, L.K.,
1982. Environmental effects of harvesting forests for
energy. Forest Ecology and Management 4, 79-94.
Hoskins, M.W., 1987. Agroforestry and the social milieu,
pp. 192-203. In: Steppler, H. A. and Nair, P.K.R.
(eds.), Agroforestry: a decade of development.
ICRAF, Nairobi, Kenya.
Howard, P.J.A., and Howard, D.M., 1974. Microbial
decomposition of tree and tree and shrub leaf litter.
1. Weight loss and chemical composition of decomposing
litter. Oikos 25, 341-352.
Hunt, H.W., 1977. A simulation model for decomposition in
grasslands. Ecology 58, 469-484.
Huxley, P.A., 1986. Rationalizing research on hedgerow
intercropping: an overview. ICRAF Working Paper No.
40. ICRAF, Nairobi, Kenya.
Huxley, P.A., Darnhoffer, T., Pinney, A., Akunda, E., and
Gatama, D., 1989. The tree/crop interface: a project
designed to generate experimental methodology.
Agroforestry Abstracts 2 (4), 127-145.
Huxley, P.A., Patel, R.Z., Kebaara, A.M., and Mitchell,
H.W. 1977. Tracer studies with P32 on the
distribution of functional roots of Arabica coffee in
Kenya. Annals of Applied Biology 77, 159-180.
Ikombo, B. M., 1984. Effects of farm yard manure and
fertilizers on maize in semi-arid areas of Eastern
Kenya. East Africa Agricultural and Forestry Journal
44, 266-74.


201
significantly affecting the above and, (c) to identify the
potential uses and constraints of farmers in alley cropping
multipurpose trees and shrubs.
Study Area and Methodology
The study was conducted on farms that lie on the hills
immediately south of Machakos town, Eastern Province, Kenya.
The farthest farm from Machakos town was about 20 km. The
ecology of the area is semiarid with a mean annual rainfall
of about 600 to 700 mm, falling in two seasons: long-rains
(March to July) and short-rains (September to October).
Besides meeting the criteria of being semiarid, the area of
the study was close to the Field Station of ICRAF where
alley cropping research was going on. Thus, as one might
either expect or assume, farmers in this area would be
better informed about and potentially more adoptive of the
technology than distant farmers. ICRAF Field Station is 6
km west of Machakos town.
Three administrative sublocations were selected for
sampling: Kiima-Kimwe, Muvuti and Kivandini. All three
sublocations were constituents of Muvuti location, Central
Division, Machakos District, Kenya. The area is
topographically hilly and the farms were scattered on the
hill slopes.
Based on the outcome of a three day rapid rural
appraisal (Chambers, 1985) or Sondeo (Hildebrand, 1981)


35
Table 2-2. Concentration and amounts of nutrients in Leucaena
leucocephala and Cassia siamea mulch from alley
cropped hedgerows.
Species
Nutrient concentration (%) and amounts (kg
ha'1 yr'1, in parenthesis)
N
P
K
Ca
Mg
S
Leucaena
3.5
0.2
1.9
0.7
0.3*
0.1
(136.0)
(7.5)
(71.7)
(26.4)
(11-3)
(3.8)
Cassia
2.9
0.3
1.8
0.7
0.2
0.1
(54.7)
(5.7)
(33.7)
(13.2)
(3.8)
(1.9)
Within a column, values of nutrient concentration with
different letters are significantly different (p = 0.05).
The amounts of all nutrients in leucaena mulch were
significantly higher than those in cassia (p = 0.05).


239
With the exception of mulch use, the presence or absence of
the other factors depended significantly on the presence or
absence of coffee on the farm. For instance, 67% of the
fodder-interest group produced coffee and 76% of all coffee
producers were interested in the fodder production from
alley cropping. It did appear that coffee provided the
capital required to purchase inputs to raise or sustain
sufficient food production.
Farmers used many methods to maintain fertility and
productivity of their soils. The major ones were soil
conservation measures, use of livestock manure, use of
fertilizer, purchase of improved seeds, weed control, and
intercropping. Most of these measures required cash inputs.
Alley cropping would also require cash inputs for
establishment and maintenance of the hedgerows. Thus, alley
cropping would likely be limited to farmers with capital
(e.g., coffee farmers).
Although the farm holdings were small, with 76% of the
farmers sampled having less than three hectares, farm size
appeared unlikely to constrain adoption of the technology.
The major constraints identified were the capital investment
required to plant a large number of seedlings, availability
of the trees, termite problems, and fears of decreased crop
yields, due to competition from the trees.
Based on the results from the on-station studies,
which showed that leucaena was competitive with crops while


179
o
o__
0)
k_
3
4*
CD
W.
k.
0)
CO
03
CL
E
z
0)
*->
CO
=J
c
'o
CO
E
c
>
CO
a>
2
cn
<
c
CD
sz
U
Figure 5-10. Difference in mean monthly temperature
of top soil (10 cm) measured near (0.45 m) and away (3.5 m)
from alley cropped hedgerows of Leucaena leucocephala in the
long- and short-rain seasons of 1991. Vertical bars
indicate standard error of difference of means.


Teman, G.L., 1979. Volatilization losses of nitrogen as
ammonia from surface applied fertilizers, organic
amendments and crop residues. Advances in Agronomy 31,
189-223.
Tiedemann, A.R., and Klemmedson, I.O., 1977. Effect of
mesquite trees on vegetation and soils in the desert
grassland. Journal of Range Management 30, 361-367.
Tisdale, S.L., and Nelson, W.L., 1975. Soil fertility and
fertilizers. Macmillan, New York.
Toky, O.P., and Bisht, R.P., 1992. Observations on the
rooting patterns of some agroforestry trees in an arid
region of north-western India. Agroforestry Systems
18, 245-263.
Torres, F., 1983. Potential contribution of leucaena
hedgerows intercropped with maize to the production of
organic nitrogen and fuelwood in the lowland tropics.
Agroforestry Systems 1, 323-345.
Turner, N.C., 1981. Techniques and experimental approaches
for measurements of plant water status. Plant and Soil
58, 339-366
Vallis, I., 1978. Nitrogen relationship in grass/legume
mixtures, pp. 190-201. In: Wilson, J.R. (ed.), Plant
relations in pastures. CSIRO, Australia.
Vallis I., and Jones, R.J., 1973. Net mineralization of
nitrogen in leaves and leaf litter of Desmodium
intortum and Phaseolus atropurpureus mixed with soil.
Soil Biology and Biochemistry 5, 391-398.
Van der Kruijs, A.C.B.M., Van der Heide, J. and Kang, B.T.,
1984. Observations on decomposition rates of leaves of
several shrubs and tree species applied as mulch under
humid tropical conditions. In: Nitrogen management in
tropical soils. Proc. of International Conference held
at IITA, Ibadan, Nigeria.
Van Soest, P.J., 1968. Determination of lignin and
cellulose in acid-detergent fiber with permanganate.
Journal of the Association of Official Agricultural
Chemists 51, 780-785.
Verinumbe, I., Knipscheer, H.C., and Enabor, E.E., 1984.
The economic potential of leguminous tree crops in
zero-tillage cropping in Nigeria: A linear programming
model. Agroforestry Systems 2, 129-138.


127
Pot Studies; Maize Biomass Yield and
Synchrony of Hulch-N Release
Biomass yield, N yield, apparent N recovery
and relative index of cumulative soil-N
In the first harvest, leucaena-mulch treatment gave
significantly higher yield of maize biomass than the cassia
mulch, fertilizer or control treatments (Table 4-4). In the
second and third harvests, differences between the effects
of treatments on maize biomass yield were not significant^.
In the fourth harvest, the effects of leucaena mulch on
maize biomass yield were significantly higher than the other
treatments. Among the harvests, biomass yields of maize
were low in the first and fourth harvests and significantly
higher in the second and third harvests for all the
treatments.
There was no significant difference in N-concentration
of the maize biomass between treatments in any given
harvest. However, among the harvests, significant
differences between treatments on the biomass N-yield (i.e.,
N concentration x biomass yield) were observed. In the
first harvest, N-yield of maize biomass with leucaena mulch
was significantly higher than that of other treatments.
This was true also of the percent N recovery from the mulch


167
instance, water content of the discs (or discs cut from the
leaves of maize plants) under leucaena was significantly
lower during the dry season than those under cassia (Table
5-4). In the seventh season, when rainfall conditions were
normal, no significant difference in plant water content
near and away from the hedgerows or between species was
observed (Table 5-5).
N content of maize grain from the two hedgerow species
treatments was not significantly different (data not
presented), although a difference of 19% was detected in the
N concentration of the sixth season grain harvest from the
two treatments.
Soil Physical and Chemical Properties
Of the various soil physical and chemical properties
monitored, significant differences due to the effects of the
hedgerow species or of distance of crop rows from the
hedgerows were detected in: (a) bulk density, (b) rates of
water infiltration, (c) hydraulic conductivity (saturated)
of the soil, (d) gravimetric water content, (e) soil
temperature, and (f) soil fauna. Bulk density of top soil
(0-5 cm) near the hedgerows was significantly lower than
away from them (Fig. 5-4). This was so only under leucaena,
not cassia. Also, below 5 cm depth, difference in soil bulk
density near and away from the hedgerows was not
significant.


CHAPTER 6
POTENTIAL FOR FARMER ADOPTION OF ALLEY CROPPING
OF MULTIPURPOSE TREES AND SHRUBS IN SEMIARID
AREAS OF MACHAROS, KENYA
Introduction
Agroforestry is often suggested as a major practical
land management alternative for the maintenance of soil
fertility and productivity of lands in the tropics (Nair,
1989), and, in particular, the marginally productive ones,
including the semiarid tropics (Rocheleau et al., 1989).
Among the agroforestry technologies that have received
tremendous research and development effort is alley cropping
(also called hedgerow intercropping). Alley cropping
consists of growing food crops in alleys between planted
hedgerows of multipurpose trees and shrubs, usually
nitrogen-fixing ones. The hedgerows are periodically cut
back at the beginning as well as during cropping to prevent
shading and to provide mulch to the associated crop (Kang et
al., 1990).
Results of alley cropping (at least on-station) in the
humid tropics with respect to soil fertility and crop
performance have generally been positive (Kang et al.,
1990). Compared to the humid lowland tropics, there are
relatively few studies on alley cropping in the semiarid
tropics. Rainfall and soil conditions aside, the impacts of
197


Table 2-3. Effect of alley cropping and block planting systems of maize grain yield
during six cropping seasons. Rainfall for each season is included in the last
row.
System
Maize yield (t ha
_1) for each season
Mean
2
3
4
5
6
7
Alley cropping
(leucaena)
2.5a,b
3.0a,b
2.3a'b
2.3a,b
0.6a,b,c
2.3
2.2a,b
Alley cropping
(cassia)
2.4a'b
3.9d
2.4a,b
3.4a
1.4a
2.8C
2.7C
Block planting
(leucaena)
2.5a
2.7b,c
2.0C
2.6b
0.5C
1.7b
2.0a
Block planting
(cassia)
2.3a,b
2.9b,c
2.0C
2.9a,b
0.9a'b'
1.5b
2. la'b
Control (leucaena mulch)
2.0a,b
2.6b,c
2.2a,b'c
1.6d
0.8a,b'c
1.7b
1.8b
Control (cassia
mulch)
1.9a,b
2.3C
2. i-.b,c
2. lc,d
0.7a'b,c
1.8b
1.8a
Control (fertilizer)
1.6b
3. la-b
2.9a,b
3. la'b
1.3a,b
2.7#
2.5C
Control (absolute)*
2.0
2.9
1.8
2.1
1.1
2.4
2.1
Mean
2.1
2.8
2.1
2.4
0.9
2.0
2.1
Rainfall (mm)
330
441
631
333
214
373
336
a'b'c'dwithin a season, values followed by different letters are significantly different
(p = 0.05).
to
From an adjacent experiment.


51
Trends in N uptake were similar to those of the crop
yields, i.e., the higher the yield of the crop, the higher
the N-yield. Because of the lower yields under leucaena
alley cropping compared to those under cassia (Table 2-3),
apparent N utilization efficiency under leucaena was low.
However, in the sole croppings with only mulch application,
apparent N utilization efficiency values from the mulch of
leucaena and cassia were similar.
N uptake and apparent N utilization efficiency were
highest in the fertilized control plot, in both grain as
well as stover and cobs. N uptake by the maize in the
fertilizer-control treatment was 42% more than the amount of
N applied. Similar high N uptake levels exceeding the
amounts applied were also observed with cassia alley
cropping. For these two treatments, all of the N applied
through the fertilizer or mulch could be accounted for.
Phosphorus.In six seasons, P removed by the crop
(i.e., grain, stover and cobs) significantly exceeded the
amount of P added through mulch or fertilizer (Table 2-7).
Averaged across species, P removed through the crop was 164%
more than P applied through the mulch of leucaena and
cassia. As a consequence of the lower mulch yield of cassia
compared to leucaena, P applied through cassia mulch was
significantly lower than P applied through leucaena mulch.
Hence, P removed through the crop in plots with cassia was
significantly higher (215%) than that from plots with
/


220
Coffee was the major cash crop of the area. Fifty-
seven of the farmers interviewed had coffee. Among the
coffee farmers, 76% were interested in the fodder aspects of
alley cropping. The difference in the frequency of farmers
without coffee but interested in alley cropping for fodder
or for mulch was not significant.
In addition to off-farm employment, other sources of
cash were sale of livestock (e.g., poultry), baskets (mainly
made of sisal), fruits, charcoal, and other tree products.
Trees on the Farm: Functions and
Constraints to their Planting
Predominant species
Both indigenous and exotic species (fruit and non
fruit) were present on the farms. On an average farm size
(about two hectares), 10%-20% of the land could be under
trees (both fruits and nonfruit trees). Typically, only
fast growing exotic species good for pole and firewood
(e.g., Eucalyptus camaldulensis and Grevillea robusta) were
planted. Indigenous or naturalized species (e.g., Acacia
nilotica, Croton megalocarpus and some Commiphora spp.) were
not planted but left to grow from volunteer seedlings and
protected from damage by animals and termites. These
species were found mainly on the lower slopes of the hills.
On the cooler upper slopes, Acacia mearnsii (wattle tree)


204
three day duration. At the station, the fanners were given
a guided tour of the trials where pruning were used as mulch
or as fodder. During the tour, questions and concerns
raised by the fanners were recorded.
Statistical Analysis
Frequency tables of responses to questions were
generated using the SAS (SAS, 1987). Means and their
standard deviations (sd) of various variables are presented.
Dependency or independence of fanner interests in the use of
the pruning of the alley cropping technology (i.e., fodder
or mulch) on any of the variables considered in the survey
was determined using the Chi Square test at the 80%
confidence level.
Results
Farm Household Characteristics and
Farmer Interests in use of Aliev Cropping
Farm size of the farmers interviewed varied widely,
ranging from 0.4 hectares to 17.4 hectares (Table 6-1).
Seventy-seven percent of the farms were in the 0-4 ha size
range; of these, 46% had two or less hectares. Farm size
with the largest frequency distribution (15%) was three
hectares. The sizes of farms can therefore be classified as
small scale. Yet, they support a large number of
dependents. On average, there were 14 family members per


255
Kang, B.T., and Wilson, G.F., 1987. Agroforestry: a decade
of development, pp 227-243. In: Steppler, H.A., and
Nair, P.K.R. (eds.), International Centre for Research
in Agroforestry, Nairobi, Kenya.
Kang, B.T., Wilson, G.F. and Lawson, T.L., 1984. Alley
cropping: a stable alternative to shifting
cultivation. International Institute of Tropical
Agriculture, Ibadan, Nigeria, pp. 22.
Kang, B.T., Wilson, G.F., and Sipkens, L., 1981. Alley
cropping maize (Zea mays) and Leucaena (Leucaena
leucocephala) in Southern Nigeria. Plant and Soil 63,
165-179.
Reliman, M., 1979. Soil enrichment by neotropical savanna
trees. Journal of Ecology 67, 565-577.
Kelly, K.R., and Stevenson, F.J., 1987. Effects of carbon
source on immobilization and chemical distribution of
fertilizer nitrogen in soil. Soil Science Society of
America Journal 51, 946-951.
Kibe, J.M., Ochung, H., and Macharia, D.N., 1981. Soils and
vegetation of the of the ICRAF experimental farm,
Machakos District. Detailed Soil Survey No. D23, Kenya
Soil Survey, Nairobi.
Kimmins, J.P., 1977. Evaluation of the consequences for
future tree productivity of loss of nutrients on whole
tree harvesting. Forest Ecology and Management 1, 169-
71.
King, H.G.C., and Heath, G.W., 1967. The chemical analysis
of small samples of leaf material and the relationship
between the disappearance and composition of leaves.
Pedobiologia 7, 192-197.
Kumar, M.B., and Deepu, J.D., 1992. Litter production and
decomposition dynamics in moist deciduous forests of
the Western Ghats in Peninsular India. Forest Ecology
and Management 50, 181-201
Ladd, J.N., Oades, J.M. and Amato, M., 1981. Distribution
and recovery of nitrogen from legume residues
decomposing in soil sown to wheat in the field. Soil
Biology and Biochemistry 13, 251-256.
Lai, R., 1989. Agroforestry systems and soil surface
management of a tropical alfisol: 1. Soil moisture and
crop yield. Agroforestry Systems 8, 7-29.


21
to the block/sole planted hedgerows. The mulch-only plots
received no fertilizer.
The hedgerows were pruned 2-3 times in a cropping
season (depending on the rainfall conditions of the season)
to 50 cm height aboveground. First pruning was done in
November 1988, this being the first of the seven cropping
seasons with the treatments applied. At each pruning of the
hedgerows, which typically took place 1-2 weeks before the
onset of rains, the biomass harvested (hereafter referred to
as mulch) was determined. It was then chopped into small
pieces, samples taken for dry matter determination and the
rest spread on the field. Typically, mulch harvested before
the onset of rains was incorporated into the soil during
field preparation and planting of the crops. However, mulch
from later pruning during the life of the crop was either
partially incorporated (in the process of weeding the crops)
or generally left on the soil surface. In all crops, the
plots were hand tilled and weeded.
At each harvest of maize, which took place
approximately 120 days after sowing, the yields of the
grain, cobs and stalks were determined. Samples of each
component were oven-dried at 105C for 48 hours and dry
weight determined. Maize stalks and cobs were not returned
to the plots, in line with the farmers' practice in the
study region where maize stalks are a valuable livestock
feed-.


149
(positive or negative), but also identify possible causes of
the observed interaction, and, in turn, may provide insights
into determining the potentials for as well as management of
alley cropping, particularly in the semiarid tropics.
In consideration of these, this study was undertaken
with the following objectives:
1. To evaluate the yield profile of maize
intercropped with hedgerows of Leucaena
leucocephala and Cassia siama in relation to the
distance of the crop from the hedgerows.
2. To assess the relative impacts of soil water and
soil fertility (nitrogen) on the observed profile
of maize yield.
Materials and Methods
Site Description
The study was carried out under rainfed conditions at
the Field Station of ICRAF, located at Machakos, Kenya
(latitude 1 33' S and longitude of 37 14' E and 1596 m
elevation). The patterns of rainfall and soil
characteristics of the site are described in Chapter 2.


55
Table 2-8. Fine root density at two soil depths along a
meter-long transect from 6 m apart hedgerows of
Leucaena leucocephala and Cassia siamea
intercropped with maize.
Hedgerow Species
Soil
0-20
deDth (cml
20-40
Mean
Leucaena
0.41
0.64
0.53*
Cassia
0.23
0.47
0.35b
Mean
0.32
0.56b
0.44
NOTE: Fine root density = cm root per cm3 of soil.
*,bFor a given row and column, means followed by the same
letters are not significantly different (p = 0.05).


14
et al., 1990). The low apparent recoveries of N added may
be due to immobilization by soil microbes, especially if th<
N added through the mulch is low (Allison, 1966). This may
explain the negative apparent N recovery from cassia mulch
in week 8. Management approaches that synchronize the peaki
of mulch N mineralization and N uptake by the crop could
enhance apparent recoveries of N by the crop from the littl(
amounts of mulch applied.
Implications of Decomposition Patterns of
Leucaena and Cassia on Soil N and Organic
Matter Build-up
Rapidly decomposing mulches such as leucaena leaves ir
this study or the leaves of Gliricidia sepium (Yamoah et
al., 1986a; Palm and Sanchez, 1991) may provide sufficient 1
for the crop. When applied at high rates and for a long
period, for instance, 8-10 t ha'1 yr'1 dry matter for 6
years, leucaena could also increase soil organic matter
(Kang et al., 1985). Materials with low decomposition rates
(either due to high C:N ratio, lignin or polyphenol
concentration) such as cassia may not provide enough N for
plant growth in the short term, especially when applied in
small amounts, but they may contribute to the build-up of
organic-N and humus in the soil (Yamoah et al., 1986a).
In addition to the many benefits of humus on soil
chemical and physical properties (Swift and Sanchez, 1984),
increase in soil organic matter may provide a low but
continual supply of N in the long term. For leucaena mulch,


237
96,000 seedlings would require $9,600 or $160 per farm. By
local standards, this is a lot of money and not readily
available to resource-poor farmers. In addition, there is
no infrastructure in the area to supply or sustain such a
supply of seedlings. Seed-based technology and not
seedlings would probably be a cost-effective means of
meeting potential demands for such amounts of planting
materials.
Unfortunately, all alley cropping studies at the
station were based on the use of transplanted seedlings with
a total lack of any research on direct planting of seeds in
the field. This is probably so because the uniform
establishment of trees under the conditions of the study was
easier from seedlings than from seeds. Uniform establish
ment of trees is an important consideration under conditions
of experimentation. An equally important question about
availability raised by the farmers was care and management
of the trees after planting. These included prevalent dry
conditions, browse by animals, weeds and termites. Of most
concern was damage of trees by termites.
Termites
Most of the farmers interviewed (86%) mentioned
termites as a major problem to raising trees. Termites were
also a major problem in the research station but were
overcome with termite control chemicals. However, farmers


To my late father, Jama Adan Derow, who passed away in
November 1992, while I was in Gainesville writing this
dissertation.


28
Apparent nitrogen utilization efficiencies were
calculated for crops seasons 2-6 and includes grain, cobs
and stover. The efficiencies were calculated by determining
the nitrogen yield of the treatments where mulch was applied
to the crop (dry matter x nitrogen concentration) less
nitrogen yield of the no fertilizer control treatment and
expressing this as a percentage of the amount of nitrogen
via the mulch or the fertilizer. This method has also been
used by others to estimate utilization efficiency of mulch N
by crops (Kang et al., 1981; Yamoah et al., 1986b; Palm,
1988).
Statistical Analyses
The effects of the two species and systems of hedge
planting on crop and biomass yields were determined using
the General Linear Models (GLM) procedure in the Statistical
Analysis System (SAS, 1992). Because of the unbalanced
design caused by the fertilizer-only treatment (which was
applied at only one level as opposed to other treatments
which were applied at three levels.), the analysis was
performed in two stages. First, the fertilizer-only
treatment was omitted and the analysis of split-plot design
was performed. Second, depending on what factors were
significant in the split-plot analysis of variance (ANOVA)
test, comparisons with controls were performed in a
subsequent ANOVA that included all treatments. In both


152
samples was determined by the leaf disc procedure
(Weatherly, 1950). This is a measure of the actual water
content of a tissue with reference to the content of water
in the tissue if it were fully hydrated (Bennett, 1990).
Measurement of relative water content by this procedure
requires obtaining three measurements made on the same
sample: fresh weight, turgid weight, and dry weight. Discs
were cut with a corer of 10 mm diameter from the ear leaves
of maize plants growing near and away from the hedgerows.
Fresh weights of the discs were determined immediately.
Turgid weight was determined after the discs had been
immersed in distilled water for 3 hours in darkness. Dry
weight of the discs was determined after oven drying the
discs for 48 hours at 105C. Percent relative water content
(RWC) was then expressed as:
%RWC = (fresh wt dry wt/turgid wt dry wt) x 100.
In addition to the leaf discs, whole plant water
content was determined in the first and second sampling of
each season as the difference between fresh and dry weight
of the whole plants (shoots) harvested (Turner, 1981).
Whole plants near and away from the hedgerows were
harvested, fresh weight recorded immediately and dry weights
after oven drying for 48 hours at 105C.


103
3. Bundles of twigs tied with a nylon string (no
litter bags) incorporated into the soil and placed
on the surface, and
4 Soil incorporated filter paper in litter bags.
The split-plot design had the species as the main-plots and
the placement methods of the mulch (i.e., soil incorporated
or surface applied) as the subplot treatments.
The purpose of the nylon tied twigs treatment was to
determine to what extent the litter bags influenced the
rates of decomposition. Whatman No. 2 filter paper, chopped
into small pieces, which is pure cellulose with practically
no N, was used to determine the importance of C:N ratio of
the mulch relative to the organic compounds (e.g., lignin
and polyphenols) in influencing rates of decomposition.
In both seasons, mulch similar in composition to the
soil incorporated or surface-applied was analyzed in the
laboratory to determine dry matter and nutrient contents.
The litter bags and nylon-tied twigs were retrieved from the
field every two weeks for a total period of 18 weeks.
Sufficient numbers of bags were used to allow the removal of
a bag after every two weeks. Once retrieved, the bags were
wrapped in a plastic sheet to minimize losses during
transportation from the field to the laboratory. In the
laboratory, soil and debris adhering to the mulch were
removed carefully with a toothbrush and a soft paint brush.


61
fodder becomes even more appropriate when mulch may not
result in significant gains in crop yields. Similar
observation was made by Singh et al. (1989) from semiarid
India. Furthermore, in nutrient-poor and/or acidic soils,
crop yields may not be sustained by mulch even if all the
crop residues and hedge mulch remained in the crof fields
(Szott et al., 1991).
Although the small quantities of mulch from alley
cropped hedgerows may not improve crop yields, it could be
argued that their application may still be necessary to
minimize long-term declines in soil fertility and
productivity. It is hypothesized that such declines in
fertility and productivity could result from loss of
nutrients removed in continuous harvests of mulch and tree
products (Kimmins, 1977; Hook et al., 1982; Lulandala and
Hall, 1988). With or without the removal of hedgerow mulch,
initial soil conditions, crop type and their yields would
likely influence the onset of yield declines. Use of manure
from livestock fed with the prunings may reduce yield- and
soil-fertility declines. In addition, in the semiarid
tropics where the crop yields are generally low compared to
those of the humid tropics, removal of mulch in the range of
1-2 t ha"1 season"1 may not show any significant impact on
yields and fertility conditions of the soil, at least in the
short term. This view is supported by the absence of
significant differences in maize yield between plots with


CHAPTER 5
INTERACTIONS BETWEEN MAIZE AND HEDGEROWS OF
LEUCAENA LEUCOCEPHALA AND CASSIA SIAMEA
IN ALLEY CROPPING SYSTEM IN SEMIARID
TROPICS AT MACHAROS, KENYA
Introduction
The promising results of soil fertility improvements
under alley cropping (hedgerow intercropping) in the humid
tropical lowlands (Kang et al., 1990) have caused much
interest in this technology elsewhere. This has
particularly been so in the semiarid tropics where fertility
of the soils are often low and inputs are limited (Steiner
et al., 1988). Studies on alley cropping in the semiarid
tropics, however, indicate significant reductions in crop
yields (Rao et al., 1991; Singh et al., 1989; Nair, 1987).
Below ground competition, particularly between the tree and
the crop is often suggested to be the main limitation to
hedgerow intercropping in the semiarid tropics (Singh et
al., 1989; Ong, et al. 1991; Monteith et al., 1991).
Typically, reductions in crop yields are highest near
the tree hedgerow and they diminish with distance from the
hedgerows (Ong et al. 1991; Duguma, et al., 1988).
Observations of yield increases nearer the hedgerows are also
not uncommon (Sang and Hoekstra, 1987; Huxley et al., 1989;
Ong et al., 1992; Dzowella, 1992). Also, yields of crops may
remain unaffected by the distance from the hedgerows
147


65
addition of fertilizer. Sanginga et al. (1988) also
observed low N-utilization efficiencies from leucaena mulch
but they emphasized that the absence or presence of N-fixing
bacteria inoculation made a difference. Other workers (Kang
et al., 1981; Guevarra, 1976) have also reported low
apparent N-utilization efficiencies from mulch of leguminous
shrubs.
Factors such as time lags between mulch-N
mineralization and crop N demands (Anderson and Ingram,
1989), immobilization of mulch-N and soil-N by microbes in
the process of mulch decomposition, at least in the initial
stages (Palm and Sanchez, 1991) and volatilization losses
from surface applied mulch (Terman, 1979; Nelson, 1982)
could also have contributed to the apparent low mulch-N
utilization efficiencies observed. Also, about 22% of the
undecomposed mulch-N was still tied up in mulch of both
species. This cannot be described as inefficiency, because
this N is still in the system and might be utilized by
subsequent crops. Admittedly, without more precise
determinations such as the use of 15N techniques, it is
difficult to assess the fate of mulch-N in soil (Sanginga et
al. 1988). Management practices that synchronize periods of
peak N release from mulch with periods of peak N demands by
the crop may enhance effects of the small quantities of
mulch available from the hedgerows.


18
15%, respectively. Sole stands of leucaena and cassia
occupying 15%, 20% and 25% of crop land were also
established for comparison to the hedgerow intercropped
plots. In-row spacing of the trees within the block/sole
stands of hedgerows was 1.0 m by 1.0 m. Control plots
included in the experiment were (a) maize crop which
received mulch only in amounts equivalent to what was
produced by the hedgerow intercropped plots and, (b) maize
crop which received no mulch but fertilizer only at the
rates of 40 kg and 17 kg of N and P ha*1 season'1.
The experimental design was split-plot, replicated
three times. Main plot treatments, randomized within the
blocks, were the hedgerow species (i.e., leucaena or
cassia). Subplot treatments were the factorial combination
of the three ratios of crop to tree land occupancy and the
control plots with mulch-only. The fertilizer-only control
plot was included in each main plot (i.e., 2 replications
per block). This gave 6 replications for the fertilizer-
only control. Main plot size was 20 x 50 m, with two meter
borders separating sub-plots of 5 x 20 m. Between blocks, a
4-meter wide path was provided. Sole stands of leucaena and
cassia hedges which provided mulch for the mulch-only
control plots were also included in each block.
' Initially, a control plot with no inputs was
inadvertently left out. The control plot of an adjacent
experiment, established about the same time as this study


13
One of the impacts of the above-mentioned studies has
been to stimulate interest and studies on alley cropping in
the SATs. In many areas of SATs, soil fertility is low
(Steiner et al., 1988) and chemical fertilization alone
(i.e.f no irrigation) could improve yields (Ludwig, 1987).
Fertilizers are, however, expensive for many resource-poor
farmers and/or are not readily available.
In many regions, and in particular the SATs, the
prunings are more likely to be used for fodder than as mulch
(Singh et al., 1986; Ong et al., 1991). Even if the
prunings are used as mulch for the crops, the effects on
crop yields may not be significant (Singh et al., 1989).
Under these conditions, the N contribution by the hedgerows
to the soil may primarily be through below-ground processes,
particularly where N fixing species are used. Whether a
significant amount of N can accrue through these processes
and whether this N is available to the companion crop have
not been convincingly established (Ta et al., 1986; Danso et
al., 1987; Rao et al., 1987).
If growing the trees within the crop fields is not
beneficial for the crops but the trees can provide other
valuable benefits (e.g., fodder), separate planting of trees
might be preferable. This would be particularly appropriate
in the SATs where many studies report significantly reduced
crop yields under alley cropping (Singh, et al., 1986;
Mittal and Singh, 1989; Nair, 1987). Under these


263
Sumberg, J.E., Maclntire, J., Okali, C. and Atta-Krah, A.N.,
1987. Economic analysis of alley farming with small
ruminants.. ILCA Bulletin 28, 2-6.
Summerell, B.A., and Burgess, L.W., 1989. Decomposition and
chemical composition of cereal straw. Soil Biology and
Biochemistry 21, 551-559
Swift, M.J., 1986. Enhancement of tropical soil fertility:
the role of biological research, pp. 17-27. In:
International Soil Reference and Information Center,
Wageningen, The Netherlands, Annual Report for 1985.
Swift, M.J., Russel-Smith, A., and Perfect, T.J., 1981.
Decomposition and mineral-nutrient dynamics of plant
litter in a regenerating bush-fallow in sub-humid
tropical Africa. Journal of Ecology 69, 981-995.
Swift, M.J., and Sanchez, P.A., 1984. Biological management
of tropical soil fertility for sustained productivity.
Nature and Resources 20(4), 2-10.
Swift, M.J., Heal, O.W., and Anderson, J.M., 1979.
Decomposition in terrestrial ecosystems. Studies in
ecology, Vol. 5. University of California Press,
Berkeley.
Szott, L.T., Fernandez, E.C.M., and Sanchez, P.A., 1990.
Soil-plant interactions in agroforestry systems. In:
Proceedings of the World Conference on Agroforestry
held at Edinburgh, 25-29 July, 1989. Department of
Forestry and Natural Resources, University of
Edinburgh, Scotland.
Szott, L.T., Palm, C.A., and Sanchez, P.A., 1991.
Agroforestry in acid soils of the humid tropics.
Advances in Agronomy 45, 275-301.
Ta, T.C., McDonald, F.D.H., and Faris, M.A., 1986.
Excretion of assimilated nitrogen from N2 fixed by
nodulated roots of alfalfa (Medicago sativa L.).
Canadian Journal of Botany 64, 2063-2067.
Taylor, B.R., Parkinson, D., and Parsons, W.E.J., 1989.
Nitrogen and lignin content as predictors of litter
decay rates: A microsm test. Ecology 70 (1), 97-104.
Tenant, D., 1975. A test of modified line intersect method
of estimating root length. Journal of Ecology 63, 995-
1001.


154
rows 1 and 4 of the intercropped hedgerows. At the time of
tensiometer placement, ordinary thermometers were also
placed at rows 1 and 4 to 10 cm depth.
After harvest of the sixth and seventh season crops,
infiltration rate of water into the soil near and away from
the hedgerows was determined with the double ring
infiltrometer (Landon, 1984). Infiltration rate was
determined twice at each position (i.e., near and away from
the hedgerows) within a plot. At end of the seventh
cropping season, saturated hydraulic conductivity of the
soil was also determined in the field near and away from the
hedgerows using a disc permeameter (CSIRO, 1988).
Hydraulic conductivity was determined at two spots near and
away from the hedgerows. Also, at the end of cropping in
the seventh season, observations were made on the
populations of soil fauna present near and away from the
hedgerows. This involved taking soil monoliths of 25 cm. x
25 cm x 25 cm and hand sorting for macro-invertebrates (body
lengths greater than 2 mm) (Anderson and Ingram, 1989).
Soil monoliths for the fauna counts were taken from two
spots near and away from the hedgerows.
Analyses of total N was by the micro-Kjeldahl method
(Jackson, 1958), organic carbon by the Walkley-Black
procedure (Jackson, 1958), and P by the Mehlich-1 double
acid procedure. Soil samples for N, P and C determination
were collected from the top 20 cm soil profile near and away


LIST OF TABLES
Page
2-1. Mulch yield of Leucaena leucocephala and
Cassia siamea intercropped hedgerows or
planted in sole stands outside the
cropped area 31
2-2. Concentration and amounts of nutrients in
Leucaena leucocephala and Cassia siamea
mulch from alley cropped hedgerows ....... 35
2-3. Effect of alley cropping and block planting
systems of Leucaena leucocephala and
Cassia siamea on maize grain yield
during six cropping seasons .... 39
2-4. Effect of alley cropping and block planting
systems on relative grain yield of maize .... 41
2-5. Land equivalent ratios (LER) of alley cropping
Leucaena leucocephala and Cassia siamea
compared with the block planting system
with same ratios of tree and crop land
occupancy .............. 48
2-6. Nitrogen budgets, utilization efficiencies
and yield of maize grown in alley cropping
systems of Leucaena leucocephala and Cassia
siamea at a hedgerow spacing of 6.7 m ..... 50
2-7. Addition of phosphorus through mulch and
removal by maize in six seasons of cropping
under Leucaena leucocephala and Cassia
siamea alley cropping 52
2-8. Fine root density at two soil depths along a
meter-long transect from 6 m apart hedgerows of
Leucaena leucocephala and Cassia siamea
intercropped with maize ...... 55
3-1. Chemical and physical characteristics of
the top soil {0-20 cm) at the beginning of
the study, October 1987 79
Vii


158
performed. In all analyses (i.e., field as well as pot
studies), pairwise comparison of Least Square Means were
used to test significance of differences between treatment
means (p 0.05).
Results
Crop Yield Profiles
The row-by-row maize grain yield profiles of the three
seasons of the study are shown in Figures 5-1 to 5-3. An
overall ANOVA of the combined data of the three seasons
revealed significant interaction effects of season x tree
species x row distance from the hedgerow on yield. Hence,
separate analysis for each season was done to determine the
main effects of species and row-spacing on grain yield.
In season five, only main effects of species was
observed (Fig. 5-1). Maize yield of any row in association
with cassia was significantly higher than the row yields
with leucaena. Effects due to distance of crop rows from
the hedgerows were significant. Averaged across rows, maize
yield under cassia was 28% higher than the yield of the rows
under leucaena. In this season, mean yield of the rows of
the sole-crop (fertilized) was, on average, 26% higher than
the yield under leucaena and 11% lower than yield under
cassia.


185
seasons yield declined in a significantly linear trend with
increasing distance from the hedgerows. It was particularly
interesting to observe higher yield nearer the hedgerows in
the second season when rainfall was below normal (only 214
mm compared to the normal 350 mm) .
In previous studies on the reasons for microsite
enrichment by trees (Tiedemann and Klemmedson, 1977; Patten,
1978; Kellman, 1979; Joffre, 1988; Bellksy et al., 1989;
Coleman et al, 1991), higher concentrations of nutrients,
improved soil physical properties, higher microbial biomass
and densities of nematodes, lower temperatures, lower
evapotranspiration rates and higher soil moisture levels are
some of the factors attributed to the higher production of
forage under the canopy of single trees. All or a
combination of some of these factors could equally explain
the higher maize biomass yield near the hedgerows. However,
enhanced soil moisture levels appeared to be a particularly
strong factor than enhanced soil fertility (i.e.,
concentration of nutrients). Two lines of evidence support
this argument. First is the lack of significant response
to fertilizer N application in terms of maize yield whether
grown in pots with soil collected from near or away from the
hedgerows, suggesting that soil N was not a limiting factor
for maize yield. Total N or KCl extract-N contents of the
soil near and away from the hedgerows were not significantly
different. Secondly, the field irrigation studies showed,


150
Methods
Treatment and layout
Hedgerow intercropped plots of Leucaena leucocephala
and Cassia siamear 6.67 m wide were established in October
1987 in randomized block design with three replication. The
hedgerows were part of a larger experiment involving
comparison of hedgerow intercropped verses block planting
systems (see previous chapters). Unlike the other two
hedgerow spacings (i.e, 4 and 5m), the 6.67 m wide
hedgerows were used for the tree-crop interaction study
because they provided the widest separation of hedgerows.
This was necessary in order to determine the distance of
influence of the hedgerows on crop yield.
Tree and crop management
The trees were planted in October 1987 from 6 month
old seedlings raised in pots. After one year of growth in
the field, the trees were pruned back into hedges of 50 cm
height above ground. This height of pruning was maintained
in all cropping seasons with 2-3 prunings in a season. The
prunings were applied as mulch (with or without
incorporation) to the alleys in between the hedgerow. Maize
was sown in the alleys twice a year.
The maize in the alley was spaced 90 cm between row
and 30 cm with rows. There were seven rows in the alleys


143
Although the periods of peak release of N from the
mulch (four weeks after application) and uptake by the crop
(eight weeks after sowing) did not coincide, both the
biomass and N yield of the crop was highest eight weeks
after mulch application. This indicates that N mineralized
earlier from the mulch was available to the crop and not
lost as such. Even after 16 weeks post application, crop
response to mulch-N still persisted. These observations are
consistent with those of others that the residual effects of
mulches on crops, even of leucaena leaves that decompose
rapidly, could be significant (Read et al., 1985). This
observation also suggests that the ability of the soil to
retain plant-derived N can be strong compared with the
ability of the crop to take it up, currently or
subsequently, and different loss mechanism to remove it, as
studies by Muller and Sundman (1988) using 1SN have
demonstrated.
The higher apparent N recovery by the crop from
leucaena than from cassia mulch (a total of 34% against 4%
across the four harvests for leucaena and cassia,
respectively) could be due the to the higher N concentration
(20% more) and rate of decomposition of leucaena leaves.
Similar observation has been made by others working with
mulch materials of high and low N concentrations (Fox et
al., 1990). The N recovery rates are, however, low but
comparable to those reported by other workers for mulches of
leguminous trees (Read et al., 1985; Gutteridge, 1992; Fox


153
Soil properties
Bulk density of top soil (0-20 cm depth), water
content and water potential, infiltration rate, hydraulic
conductivity, total-N, P, available N and fauna population
were monitored at crop rows 1 and 4 of the intercropped
hedgerows. Bulk density was determined with a 5 cm auger to
a total depth of 20 cm (i.e, 0-5, 5-10, 10-15 and 15-20 cm)
during the last season of cropping. Two soil cores were
taken at each depth and the average taken as the
representative bulk density. Gravimetric soil water content
was determined to 80 cm depth at intervals of 20 cm depths.
In the fifth season, soil sampling was carried out only on
February 15, 1991 (i.e., end of fifth season). In the sixth
and seventh seasons, soil sampling was carried out at the
end of each month including the months of no cropping.
Fresh weights of the soil were determined in the field and
dry weights in the laboratory after oven drying for 48 hours
at 105C. The difference between fresh and dry weight,
expressed as percent, was taken to represent the gravimetric
water content.
Because of the destructive nature of the gravimetric
sampling method as well as the need to increase the
frequency of monitoring soil water content, tensiometers
(type: Irrometer; Irrometer Co., California) were placed to
15 cm depth at the beginning of the sixth season at crop


243
near the hedgerows than away from them. It is possible that
other factors such as mycorrhizal association, root
turnover, and root exudates, and microbial activity could
have contributed to the higher yield near the hedgerows than
away from them. These factors were not examined in the
present study.
Planting the hedgerows outside the crops may also be a
management approach to avoid competition between the crop
and the tree hedgerows. Such a practice would particularly
be appropriate for a species like leucaena that has fodder
value but whose intercropping is detrimental to crop yields.
However, separate planting of the hedges from the crops (on
land that would otherwise be cropped) resulted in greater
reduction of maize yield than did alley cropping of both
species. The yield loss in the separate hedge and crop
planting system was nearly equal to the percentage of land
allotted to the hedges. This was so because the mulch
produced from the separate hedge and crop planting system
did not compensate (in terms of crop yield) for land lost
from cropping.
The amounts of mulch produced by cassia and leucaena
averaged 2 and 4 t ha"1 yr"1, respectively. The yields of
the block planted hedgerows outside the crop field were 14%
and 25% lower those of the alley cropped hedgerows of
leucaena and cassia, respectively. In the humid tropics
where mulch yields of alley cropped leucaena are reportedly


207
Table 6-2.
Interests of farmers in alley
size and number of members in
Machakos, Kenya, August 1991.
cropping, farm
a household at
Interests
Total
number of
farmers
inter-
Farm
Number
of members
in alley
viewed
size
in a
cropping
(%)
(ha)
household
Fodder
62
2.8 (1.5)
14 (3)
Mulch
33
3.1 (2.1)
13 (6)
Undecided
5
2.5 (1.3)
12 (4)
NOTE: Figures in parentheses are the standard deviations
of the means.


Percent N remaining in mulch
100
0 4 8 12 16
Time (weeks) after application of mulch
NJ
CT\
Biomass yield (g/pot)


Thus, expansion of arable lands, extended fallows
and/or increased fertilizer use are not viable solutions to
the problem of declining soil productivity and poor crop
yields in most semiarid parts. There is a need, therefore,
for alternative low-input technologies that can intensify
and/or sustain production from existing croplands. One
technology that has been tried with some success in the
humid lowland tropics is alley cropping (also called
hedgerow intercropping). It consists of growing food crops
in alleys between planted hedgerows of multipurpose trees
and shrubs (MPTS), usually nitrogen-fixing ones. The
hedgerows are periodically cut back at the beginning of, as
well as during, cropping. This prevents shading of the
associated crop and the hedge prunings can be a source of
nutrients to the crop (Kang et al., 1990). By retaining
woody perennials in crop production fields on a continuing
basis, alley cropping simulates the role of fallow in soil
fertility regeneration in shifting cultivation. It also
involves, to some extent, the principle of intercropping of
legumes and cereals (e.g., maize and beans), which is a
traditional farming practice in Kenya as well as many other
African countries.
The most distinctive element of agroforestry
practices, and in particular, alley cropping, is the MPTS.
The notion of trees as multipurpose plants is based on their
productive attributes (for example, food, fodder, fuelwood


134
the surface than inside the soil. Besides moisture, surface
applied residue may also remain deficient in N for a longer
period than the incorporated residue (Summerell and Burgess,
1989; Cochran, 1990).
In spite of all the favorable conditions for
decomposition within the soil, the lack of significant
difference between the soil incorporated and surface placed
mulch suggest that conditions within and outside the soil
were not drastically different. The litter bags were
incorporated to only 10 cm depth. Also, the litter bags
could have created more humid conditions for the mulch in
the bags compared to the mulch outside the bags (Jensen,
1974). Moreover, the higher soil temperatures on the
surface could also account for the unexpectedly high rate of
decomposition on the surface. Under moist conditions, rates
of decomposition of litter on soil surface correlate well
with air temperature (Williams and Gray, 1974; Witkamp,
1966; Comanor and Staffeldt, 1978; Gholz et al., 1985).
The observation that the rates of mulch decomposition
between the two seasons were similar, in spite of the
significant differences in precipitation, would suggest that
soil incorporation of mulches could be an approach to
enhance the rates of mulch decomposition and availability of
nutrients to crops. Also the effects of wind blowing away
mulch from the fields, particularly before the crop stand
develops sufficient cover could be minimized. However, soil
incorporation is labor demanding. As a management practice


148
(Lai, 1989). These inconsistencies of crop performances in
relation to the tree hedgerows suggest some difficulties in
predicting the pattern of interaction between the tree and
the crop. Differences in seasonal rainfall could further
complicate any predicted pattern of tree-crop interaction.
More studies are, therefore, needed to identify factors that
may predict well the pattern of interaction between the tree
hedgerows and the crop in a given state.
Besides site differences, the hedgerow and crop
species used, their management (e.g., height and frequency
of pruning), and the absence or presence of fertilizers are
some of the factors that could alter the pattern of
interaction between the tree hedgerow and the crop. Studies
on alley cropping in semiarid tropics have focused mainly on
leucaena (Rao et al., 1991; Lulandala and Hall, 1990; Ong
et al., 1992). There is need to include more species in
hedgerow intercropping in semiarid areas. Where other
species have been included (Nair, 1987), crop yields were
lowest under leucaena intercrop. Also, there is a general
paucity of information on the relative importance of soil
resources (e.g., soil fertility and water content) on the
potential interaction between the crop and the hedgerows.
In addition to the inclusion of more species, it is
also necessary to characterize the profile of crop yield,
soil chemical, and physical properties (e.g., N, P and water
content) in relation to the position of the hedgerows. Such
studies will determine not only the nature of interaction


64
mulch-P, particularly that of leucaena, could meet the P
requirements of maize grain yield of about 2 t ha'1 as
suggested by Torres (1983), it is evident that it does not
meet the total needs of the crop (i.e., grain, stover and
cobs).
The considerable increase (16%) in the P concentration
of cassia mulch in 1991 over the levels of 1987 and the
absence of similar increase in leucaena mulch suggest
differences in the patterns of P uptake by the two species.
The cause of this difference is unknown. Some postulations
are that cassia is more mycorrhizal than leucaena or
supporting more phosphate solubilizing bacteria than
leucaena in its rhizosphere.
Apparent utilization efficiency of mulch-N
The apparent low use efficiency of mulch-N can best be
explained by the equally high crop and N yield of the sole
maize control compared to the yields of mulch treatments.
This is another indication of the presence of sufficient
native soil N at the study site for the maize crop without
additional inputs. Low utilization efficiencies of mulch-N
have also been reported by several other workers. For
example, Palm (1988) observed mulch-N utilization efficiency
of 16% by upland rice from mulches of leguminous shrubs.
Yamoah et al. (1986b) noted N-utilization efficiencies of
near 20% from the mulch of leguminous shrubs with the


24
intercropped and where they were planted in sole/block
stands outside the crop area.
To estimate in situ effects of the intercropped
hedges, the block-planted hedgerows should not affect the
adjacent crop. In an on-farm situation, the block-planted
trees could be on land not fit for cropping that is at some
distance from the crop land. In on-station situations,
distant planting may not be feasible, especially under
conditions where treatments are randomly located in small
plots. Two approaches, polythene sheets vertically inserted
into the soil to 50 cm depths (Ong et al., 1991) and
trenching (Verinumbe and Okali, 1985) have been used to
minimize effects of adjacent rows or blocks of trees on
crops. The effectiveness of these approaches depends on the
depths of insertion of the polythene sheets into the soil or
of trenching. Tree roots can go deep below the polythene
sheet or the trench and resurface (Corlett, 1989; Ong et
al., 1992).
In the present study, row-by-row harvest of the crop
of the widest alley (6.7 m) was used to determine the
distance of influence of the block-planted hedgerows on the
yield of the adjacent crop. A graphical plot of the yield
of the rows against the distance of the rows from the
hedgerows was then made. From this plot, the rows of maize
influenced by the block-planted hedges were determined and
discarded. The yields of the remaining rows were used to


99
tropics, but equally, there could be differences caused by
sheer differences in climate. Moreover, differences in the
quality of the mulch (e.g., ratio of leaves to twigs as well
as nutrients, polyphenols and lignin content) and management
practices (e.g., soil incorporation or surface placement)
could cause differences in the patterns and rates of
decomposition between sites. An understanding of the effect
of these factors on mulch decomposition could provide
guidelines for increasing the efficiency of N-use by crop
from mulches and to sustain soil fertility in the semiarid
tropics.
The objectives of this study were three-fold:
1. To evaluate the effects of soil placement methods
(i.e., incorporation vs surface placement) on the
rates of decomposition of Leucaena leucocephala
and Cassia siamea mulch under field conditions.
2. To determine the impacts of the mulch quality
(i.e., nutrient content, percent lignin and total
soluble polyphenols) on the rates of decomposition
and N mineralization.
To evaluate the potentials for synchronizing mulch
N-mineralization and uptake by maize.
3.


Maize yield as % of sole crop
44
140
120
100
80
60
40
Leucaena
Cassia
Leucaena
Cassia
(intercropped)
(intercropped)
(sole)
(sole)
&
MifM $
IT1
1989
1990
Year
1991
Figure 2-6. Maize grain yield under alley cropping
and block planting systems of Leucaena leucocephala and
Cassia siamea expressed as percent of the sole maize control
yield.


52
Table 2-7. Addition of phosphorus through mulch and removal
by maize in six seasons of cropping under
Leucaena leucocephala and Cassia siamea alley
cropping.
Mulch Source
Phosphorus
Addition
(kg ha"1)
Removal
Difference
Leucaena
22.8
52.1*
-29.3*
Cassia
15.8b
49.8
-34.0b
Mean
19.3
52.0
31.7
Values in a column with different letters are not
significantly different (p = 0.05).


104
The mulch was then sorted into leaves and twigs. Dry weight
of each component was recorded after oven drying at 65C for
48 hours.
Ash-free dry weight of the mulch retrieved from the
soil was obtained following combustion of the mulch to ashes
in a muffle furnace at 550C for 3 hours. Fresh mulch was
also ashed to determine initial ash content. The following
formula after Cochran (1991) was used to calculate ash-free
dry weight and percent N: Corrected dry weight = Dry weight
-increase in total residual ash over the ash content of the
fresh mulch.
To determine the contents of nutrients, the mulch was
oven-dried at 65C for 24 h before weighing. It was then
ground to pass through a 0.2 mm screen, thoroughly mixed,
and sub-sampled for total-N analysis (three replicates),
according to micro-Kjedahl acid digestion procedure
(Jackson, 1958) modified for plant materials. Plant
materials of known N-concentrations from the archives of the
National Dryland Research Station Laboratory, Machakos,
Kenya, were included in the analyses to check the accuracy
of the analytical procedures. In the first season,
determination of N concentration was done only at the start
and end of the study. In the second, however, N
concentration of the mulch was determined at each time it
was retrieved from the soil. Carbon (C) content of the
mulch remaining undecomposed after every two weeks was not


10
In Chapter 4, studies on decomposition, mineralization
and uptake patterns of leucaena and cassia mulches are
presented. Uptake patterns of N and P from the mulch was
assessed by means of monthly harvests and determination of
nutrient content of potted maize plants. Chapter 5
addresses crop yield profile at the tree-crop interface.
For this, the last three of the seven croppings seasons of
the study were used. Also presented in Chapter 5 are the
results of field and pot studies of water and nutrient
status of soil at the tree-crop interface. These studies
were conducted to explain the observed interaction between
the hedgerows and the crop. Chapter 6 presents results of a
survey of 60 farms and discusses factors that may
significantly influence adoption of alley cropping in the
semiarid tropics.
Finally, Chapter 7 provides a summary of Chapters
1-6. It highlights the major findings of the study and
suggests aspects of alley cropping in the semiarid tropics
with potential for further research.


23
Soil put into the pots (approximately 20 kg) was
collected from three randomly selected positions near (0.45
m) and away (3.5 m) from the alley cropped plot of the 6.7 m
wide hedgerows. The amounts of irrigation water applied to
the pots were based on the difference between soil water
storage at field capacity (15%) and wilting point (7.8%)
determined at the initiation of the study. The plots were
irrigated twice weekly.
Three seeds of maize were sown in each pot. Three
days after germination, the seedlings were thinned to one
per pot. Total plant harvest (i.e., shoot plus roots) was
done after four weeks. After oven drying at 65C for 48
hours, dry mass was determined. On the same day as the
harvest, the pots were resown for another four weeks. This
repeated sowing and harvesting from the pots was performed
for a total period of 16 weeks.
Yield ComputationCrop and Hedge
Biomass from Field Study
Biomass and maize yields of the hedgerow planting
systems (expressed in t ha1, dry weight) were computed to
include area occupied by both the hedgerows and the crop.
Hence, the yields are expressed on gross as opposed to net
area basis or area occupied by crop only. In situ effects
of the hedgerows were calculated as the difference in mulch
and crop yields of the systems where hedgerows were


43
suggests that at this site, soil moisture was more limiting
than differences between treatments.
The patterns of maize yield over time were also
examined by expressing the yields of the two maize planting
systems as a percent of the yields of the fertilizer-only
control treatment (Fig. 2-6). Averaged over seasons, maize
yield declined by 12% per season under leucaena alley
cropping, with the largest decline in the third year (season
6). In contrast, maize yield under cassia alley cropping
increased by 8% per season. In the second year (season 6),
when yield under leucaena declined by 45% compared to the
control, yield under cassia was 27% higher than the control.
The increase in maize yield under cassia, in spite of the
hedgerows occupying land, suggested that productivity of
maize (kg m~2) under cassia alley cropping was higher per
unit net (cropped) area than of the sole crop (Fig. 2-7).
Most of the reduction in maize yield and in productivity
under leucaena, compared to the yield of the sole crop,
occurred in the last three seasons of the study. The
patterns of productivity per unit of land or rainfall were
similar. On average, mean maize productivity per unit of
rainfall under cassia alley cropping was +9% while that
under leucaena was -15% (Fig. 2-8). The increased maize
productivity under cassia alley cropping offsets the 15%
loss in maize yield when hedgerows occupied 15% of the
croppable land.


CHAPTER 2
PRODUCTIVITY ASPECTS OF ALLEY CROPPING WITH LEUCAENA
LEUCOCEPHALA AND CASSIA SI AME A IN SEMIARID TROPICS AT
MACHAROS, KENYA
Introduction
Alley cropping, also called hedgerow cropping, is an
agroforestry technology where annual crops are grown in the
alleys between rows of trees/shrubs that are regularly
pruned and maintained as low hedges. The prunings from the
hedgerows, when applied to the crops in the alleys, are
reported to improve or sustain crop yields. This is
suggested by many studies on cropping from the humid lowland
tropics. For instance, Kang et al. (1985) reported from
southern Nigeria that, on an Alfisol, using leucaena
prunings only, maize yield could be maintained at a
"reasonable" level of 2 t ha-1 as against 0.66 t ha*1
without leucaena prunings and fertilizer. An extensive
review by Kang et al. (1990) of other such studies from the
humid tropics showed similar results. Several other studies
also show that alley cropping can sustain soil nutrient
status and minimize organic matter decline in humid to
subhumid regions (Lai, 1989; Yamoah et al., 1986b).
Typically, alley cropping studies involve use of
leguminous trees/shrubs. This is because nitrogen
availability is commonly (though by no means exclusively) a
limiting factor in tropical agriculture. -It is also the
11


Page
5 INTERACTIONS BETWEEN MAIZE AND HEDGEROWS
OF LEUCAENA LEUCOCEPHALA AND CASSIA SI AME A
IN ALLEY CROPPING SYSTEM IN SEMIARID TROPICS
AT MACHAROS, KENYA 147
Introduction 147
Materials and Methods 149
Results .......... 158
Discussion 183
Conclusions 195
6 POTENTIAL FOR FARMER ADOPTION OF ALLEY
CROPPING OF MULTIPURPOSE TREES AND SHRUBS
IN SEMIARID AREAS OF MACHAROS, KENYA ... 197
Introduction 197
Study area and methodology ...... 201
Results ............... 204
Discussion 226
Conclusions ............. 238
7 SUMMARY AND CONCLUSIONS 241
LITERATURE CITED 247
BIOGRAPHICAL SKETCH ................. 267
vi


257
Martin, J.P., and Stott, D.E., 1983. pp.1-75. In: Skujins,
I.J., (ed.), Uses of microbiological processes in arid
lands for desertification control and increased
productivity. Proc. 6th Annual International Symposium
on Environment and Biogeochemistry. UNEP, Paris.
Mbogo, S.G., 1991. Production profilecrop production,pp.
10-19. In: Tifien, M., (ed.), Environmental change
and dryland management in Machakos District, Kenya,
1930-1990. Ministry of Reclamation and Development of
Arid, Semi-arid areas and Wastelands, Nairobi.
McClaugherty, C.A., Pastor, J., Aber, J.D., Melillo, J.M.,
1985. Forest litter decomposition in relation to soil
nitrogen dynamics and litter quality. Ecology 63, 621-
626.
Melillo, J.M., Aber J.D., and Muratore, J.F., 1982.
Nitrogen and lignin control of hardwood leaf litter
decomposition dynamics. Ecology 63, 621-626.
Mittal, S.P., and Singh, P., 1989. Intercropping field
crops between rows of Leucaena leucocephala under
rainfed conditions in northern India. Agroforestry
Systems 8, 165-172.
I
Monteith, J.L., 1990. Conservative behavior in the response
of crops to water and light, pp. 3-14. In: Rabbinge,
R. (ed.), Theoretical production ecology: hindsight
and perspectives. Pudoc, Wageningen, The Netherlands.
Monteith, J.L., Ong, C.K., and Corlett, J.E., 1991.
Microclimatic interactions in agroforestry systems.
Forest Ecology and Management 45, 31-44.
Mugendi, D.N., 1990. Plant nutrient aspects of mulch
incorporation in alley cropping trials of semi-arid
Machakos, Kenya. M.S. Thesis. University of Nairobi,
Kenya.
Mugendi, D.N., 1991. Effect of green manure and hedgerow
intercropping on crop production: The DARP Experience.
Paper Presented at the Dryland Agroforestry Research
Project (DARP) workshop held at Machakos, September.
Muller, M.M., and Sundman, V., 1988. The fate of nitrogen
(15N) released from different plant materials during
decomposition under field conditions. Plant and Soil
105, 133-139.
Mungai, D.N., 1991. Microclimatic aspects of hedgerow
intercropping in semi-arid Machakos, Kenya. Ph.D.
Dissertation. University of Nairobi, Kenya.


171
For infiltration rates, main effects of species were
not significant (Fig. 5-5). The rates of infiltration of
water near the hedgerows of both species were significantly
higher than away from them.
Significant interaction effects were observed between
tree species x distance of crop rows from the hedgerows on
hydraulic conductivity of the soil (Table 5-6). The rates
of hydraulic conductivity near the hedgerows of leucaena
were significantly higher than away from them; in cassia,
the differences were not significantly different.
With respect to gravimetric water content of the soil,
only the main effects of species were significant in any
given date of measurement. Effects of the distance of crop
rows from the hedgerows on soil water content were not
significant. At all dates of measurements, soil water under
cassia was consistently higher than under leucaena. This
was particularly so at soil depths below 20 cm (Figs. 5-6
and 5-7) and was most pronounced in the sixth season when
rainfall was below normal. In the fifth and the seventh
seasons when rainfall was normal, no significant difference
was observed in soil water content under the two species.
At the end of the seventh season, soil water content at all
depths was significantly higher under cassia except at 0-20
cm depth.
In the top 0-20 cm soil depth, gravimetric water
content was generally higher near the hedgerows than away


15
yields) in alley cropping systems of Leucaena leucocephala
and Cassia siamea in semiarid Machakos, Kenya.
Materials and Methods
Site Description
The study was carried out under rainfed conditions at
the Field Station of the International Centre for Research
in Agroforestry (ICRAF) located at Machakos, Kenya (latitude
Io N 33' S, longitude of 37 14' E and 1596 m elevation).
Mean annual temperature is 19 C and annual rainfall ranges
from 400-1100 mm with a mean of 750 mm. This rain falls in
two seasons in almost equal amounts. There are two rainfed
cropping seasons for the principal agricultural crops. The
first rains of the year during March-June are locally
referred to as the long-rains; the second rains of the year
during October-January are referred to as the short-rains.
Of the two, the short-rains are more reliable than the long-
rains at Machakos. The seasons and the monthly rainfall
during the study period are presented in Figure 2-1.
Generally, monthly rainfall equals or exceeds
evapotranspiration only in April, May and November.
The soil at the study site is texturally well-drained,
dark, reddish-brown sandy clay, derived from the basement
complex gneiss. According to the Kenya/FAO (Food and
Agriculture Organization of the United Nations), the soils
are Luvisols, including the ferric and eutric forms (or


88
Table 3-6. Relative indices of KCl-extract nitrogen from
ion-exchange resins incorporated into the soil
of alley cropped
leucocephala and
the "short-rains
hedgerows of Leucaena
Cassia siamea with maize
" cropping season of 1991
during

KC1 extract-N (;jg/g
resin)
System
nh4+ n
no3- n
Total
Aliev croDDina
Leucaena leucocephala
29.5a
24.8b
54.3b
Cassia siamea
15.2C
15.3C
30.5d
Controls
Fertilizer
24.3b
36.9a
61.2a
Mulch (leucaena)
22.4b
39.0a
61.4a
Mulch (cassia)
18.2C
25.7b
43.9a
Maize-only
21.4b
20.8b,c
42.2a
a'b'c'dValUes within a column with different
significantly different (p = 0.05).
letters are


6
cineraria in India (Mann and Saxena, 1980). The various
mechanisms, demonstrated or hypothesized but not yet fully
proven, are reviewed by Sanchez (1987), Young (1989), and
Nair (1990). Some of the processes demonstrated or
hypothesized by which trees improve soil fertility include
additions of litter, N2-fixation, nutrient uptake from
subsoil and protection from erosion. While trees may
improve fertility of the soil beneath them and the yield of
crops near them, they will also compete with crops for
nutrients and moisture if their root distribution is similar
to and in the same soil zone as that of annual crops (Huxley
et al., 1977; Ewel et al., 1982; Jonsson et al., 1988). For
example, most of the fine, feeder roots of many common trees
are found within the 20 cm top soil (Comerford et al.,
1984). Besides, roots of the tree may be well established
by the time the annual crop is sown. It is this potential
for effective competition, particularly for soil resources,
that has been suggested as the reason for decreased crop
yields under alley cropping in some humid lowlands of the
tropics (Evenssen, 1989; Szott et al., 1990) and semiarid
tropics (Singh et al., 1986; Nair, 1987; Mittal and Singh,
1989).
To minimize the potential for crop yield losses under
alley cropping, there should be benefits of soil fertility
and crop performance that can only be realized if the hedges
and crops are planted together, i.e., benefits that would be


226
fanners with nurseries were close to some water source,
e.g., a seasonal river.
Discussion
Strategies of Farmers to Enhance
Soil Fertility and Crop Production
Production of sufficient food for the family was
probably the major goal for every farmer. This is based on
the observation that households were large, averaging 14
family members, 50% of which were children. In spite of the
various strategies employed by the farmers to improve or
sustain food production (intercropping, weeding,
fertilizing, etc.), crop yields were low. This was more so
without inputs (i.e., over 50% reduction of that with
inputs). This is perhaps a reflection of the low fertility
status of the soils and insufficient rainfall. Since the
survey was conducted at the end of a season with poor rains,
the contribution of soil fertility to the low yields may not
explain the dramatically reduced yields. Under seasons of
normal rainfall, yields are more likely to be higher with or
without inputs compared to those observed during the survey.
However, estimates of on-farm maize yields less than
1 t ha'1 are typical even in seasons of normal rainfall
(Nadar and Faught, 1984; Mbogo, 1991).
To increase food production, farmers with limited land
were not renting more land to produce food. Instead, they


41
Table 2-4. Effect of alley cropping block planting systems on
relative grain yield of maize.
System/Treatment
Season
2
3
4
5
6
7
Mean
Aliev cropping
Leucaena
100
124
95
95
25
90
88
Cassia
95
157
95
138
57
114
109
Block planting
Leucaena
119
129
95
124
24
81
95
Cassia
110
138
95
138
43
71
99
Mulch-only
Leucaena
96
124
105
76
38
81
87
Cassia
90
110
100
100
33
86
87
Fertilizer-only
76
148
138
149
62
129
117
Absolute control
96
138
86
100
52
114
98
Mean
98
134
101
115
42
96
98
NOTE: Relative yield was calculated as a ratio of actual to
average yield of all treatments in the second season,
expressed as percent.


265
Verinumbe, I., and Okali, D.U.U., 1985. The influence of
coppiced teak (Tectona granis, L.F.) regrowth and
roots on intercropped maize (Zea mays). Agroforestry
Systems 3, 381-386.
Vine, H., 1953. Experiments on the maintenance of soil
fertility at Ibadan, Nigeria, 1922-51. Empire Journal
of Experimental Agriculture 21, 65-85.
Wacquant, J.P., Oukinder, M., and Jacquard, P., 1989.
Evidence for a periodic excretion of nitrogen by roots
of grass-legume associations. Plant and Soil 116, 57-
68.
Wake, J.L., Kiker, C.F,. and Hildebrand, P.E., 1988.
Systematic learning of agricultural technologies.
Agricultural Systems 1, 47-56.
Weatherley, P.E., 1950. Studies in the water relations of
the cotton plant. 1. The field measurement of water
deficits in leaves. New Phytologist 49, 81-97.
Weeraratna, C.S., 1979. Pattern of nitrogen release during
decomposition of some green manures in a tropical
alluvial soil. Plant and Soil 53, 287-294.
Wieder, R.K., and Lang, G.E., 1982. A critique of the
analytical methods used in examining decomposition data
obtained from litter bags. Ecology 63(6), 1636-1642.
Wiersum, K.F., 1984. Surface erosion under various tropical
agroforestry systems, pp. 231-239. In: O'Loughlin,
C.L. and Pierce, A.J. (eds), Symposium on effects of
forest land use on erosion and slope stability, East-
West Center, Honolulu, Hawaii.
Willey, R.W., 1979. Intercroppingits importance and
research needs. Part 1. Competition and yield
advantages. Field Crops Abstracts 32 (1), 1-10.
Williams, S.T., and Gray, T.R.G., 1974. Decomposition of
litter on the soil surface. pp. 611-632. In:
Dickinson, C.H., Pugh, G.J.F., (eds), Biology of plant
decomposition. Vol. 1. Academic Press, New York.
Wilson, G.F., and Kang, B.T., 1982. Developing stable and
productive biological cropping systems for the humid
tropics, pp.193-204. In: Stonehouse, B., (ed.),
Biological husbandry. Butterworth, London.


251
Dommerques, Y.R., 1987. The role of biological nitrogen
fixation in agroforestry, pp. 245-271. In: Steppler,
H.A., and Nair, P.K.R., (eds.), Agroforestry: a decade
of development, ICRAF, Nairobi, Kenya.
Duguma, B., Kang, B.T., and Okali, D.U.U., 1988. Effect of
pruning intensity of three woody leguminous species in
alley cropping with maize and cowpea on an alfisol.
Agroforestry Systems 6, 19-35.
Dzowella, B., 1992. AFRENA (Agroforestry Research Networks
for Africa). Annual Report, Zimbabwe.
Ericksen, F.I., and Whitney, A.S., 1981. Effects of light
intensity on growth of some tropical forage species: I.
Interaction of light intensity and nitrogen
fertilization on six forage grasses. Agronomy Journal
73, 427-433.
Evenssen, C.L.I., 1989. Alley cropping and green manuring
for upland crop production in West Sumatra. Ph.D.
dissertation. University of Hawaii, Hilo.
Ewel, J., Benedict, F., Berish, C., Brown, B., Gliessman,
S., Amador, M., Bermudez, R., Martinez, A., Maranda,
R., and Price, N., 1982. Leaf area, light
transmission, roots and leaf damage in nine tropical
plant communities. Agroecosystems 7, 305-326.
Fayle, D.C.F., 1975. Distribution of radial growth during
the development of red pine root systems. Canadian
Journal of Forest Research 4, 28-38.
Felker, P., 1978. State-of-the-art: Acacia albida as a
complementary intercrop with annual crops. Univ. of
California, Riverside.
Fernandez, E.C.M., 1990. Alley cropping on acid soils.
Ph.D. dissertation, North Carolina State University,
Raleigh.
Ford, E.D., and Deans, J.D., 1977. Growth of Sitka Spruce
plantation: spatial distribution and seasonal
fluctuations of lengths, weights and carbohydrate
concentration of fine roots. Plant and Soil 47, 463-
485.
Fox, R.H., Myers, R.J.K. and Vallis, I., 1990. The nitrogen
mineralization rate of legume residues in soil as
influenced by their polyphenol, lignin, and nitrogen
contents. Plant and Soil 129, 251-259.


221
was common. Also, both cassia and leucaena were present,
but only on a few farms on the lower and middle slopes.
Fruit trees, e.g., mango, citrus, avocado, guava, and
loquat were raised on the farm, on land cropped and not
cropped. With the exception of mangoes and citrus, most of
the fruit trees were planted near the homestead. Citrus was
generally planted inside the "Fanya Juu" or soil
conservation trenches. The trenches acted as macro
catchments for water and provided more water to the citrus
plants than planting in the open.
Uses of tree products
Trees provided many products and services. Firewood
and building materials were two uses often mentioned by the
farmers. Because firewood was scarce, trees with good
firewood properties and with fast growth rates (e.g.,
Eucalyptus) were in high demand. With the exception of a
few farmers that bordered government land reserves, most did
not have access to firewood beyond their farms. Of the
farmers interviewed, 45% bought firewood. Between the two
farmer interest groups in alley cropping, 57% and 25% of the
fodder- and mulch-interest groups, respectively, bought
firewood. The difference between the two groups was
significant.
Like firewood, fodder was also scarce. Forty-five
percent of the farmers interviewed bought livestock feed.


227
were buying food. Fanners with large land holdings were
also buying food and their crop yields were not
significantly different from those with small holdings of
land. These observations suggest that food production was
constrained perhaps less by land availability than by cash
to buy inputs (fertilizer, seeds, manure, ox-plow) or by
labor. These observations are consistent with studies of
others from the region of the present study (Myers, 1982;
Harsh and Mbatha, 1978).
Given that 76% of the farms were two hectares or less,
the interest of the farmers with such small holdings in
putting some land under alley cropping is surprising. In
addition to supporting the view that land was not
constraining crop production, farmers' positive response to
alley cropping also suggests that farmers saw the practice
as a means to increase production of food or fodder from
their small farms. Farmers probably saw alley cropping to
be similar to intercropping of food crops, which was
predominantly a practice of farmers with small land holdings
(two hectares or less). In addition to increasing
production per unit area, intercropping is done for many
other reasons, such as higher returns to management,
benefits of diversity, even if less productive (thus,
reducing risk), reduction of incidence of diseases, control
of weeds, spread of labor demands, and satisfaction of


135
(and not one done within the schedules of the plowing and
weeding fields), it has to be justified by proportionate
gains in crop yield. Besides, crop response to soil-
incorporated or surface-applied mulch may be inconsistent
between seasons even at the same site as is demonstrated by
the studies of Kang et al. (1981) and Read et al. (1985),
which were both conducted at the International Institute of
Tropical Agriculture (IITA), Ibadan, Nigeria.
Field studies that compare the rates of decomposition
of leucaena and cassia mulches at the same site are scarce.
The rates of decomposition observed in both seasons of the
present study were remarkably similar to those reported by
Budelman (1988) for soil-surface-placed leucaena mulch in
the humid lowland tropics of West Africa. Several other
studies have also reported rates of decomposition of
leucaena similar to those of the present study (Kang and
Duguma, 1985; Weeratna, 1979). Working at IITA, Yamoah et
al. (1986a) noted, however, rates of decomposition of cassia
mulch much lower than those of the present study: in 120
days, 85% of the cassia dry matter had disappeared.
However, in laboratory incubation studies, leucaena has been
observed to decompose faster than cassia (Nyamai, 1992).
Because the rates of decomposition of mulch materials are
influenced by many factors such as site, management ,(e.g.,
soil placement methods) and intrinsic properties of the
mulch, differences as well as similarities between and among
studies are commonly reported and are not surprising. At


254
Jackson, M.L., 1958. Soil chemical analysis. Prentice-
Hall, Englewood, New Jersey.
Jaiyebo, E.O., and Moore, E.O., 1964. Soil fertility and
nutrient storage in different soil vegetation systems
in a tropical rain forest environment. Tropical
Agriculture (Trinidad) 41, 129-130.
Jenkinson, D.S., Fox, R.H., and Rayner, J.H., 1985.
Interactions between fertilizer nitrogen and soil
nitrogenthe so-called "priming" effect. Journal of
Soil Science 36, 425-444.
Jensen, V., 1974. Decomposition of angiosperm tree leaf
litter, pp. 69-104. In: Dickinson, C.H., Pugh, G.J.F.
(eds.), Biology of plant decomposition. Vol. 1.
Academic Press, New York.
Joffre, R., Vacher, J., De Los Llanos, C., and Long, G. ,
1988. The dehesa: an agrosilvopastoral system of the
Mediterranean region with special reference to the
Sierra Morena area of Spain. Agroforestry Systems 6,
71-96.
Jones, C.A., 1985. C4 grasses and cereals. Wiley and Sons,
New York.
Jonsson, K., Fidjeland, L., Maghembe, J.A., and Hogberg, P.,
1988. The vertical distribution of fine roots of five
tree species and maize in Morogoro, Tanzania.
Agroforestry Systems 6, 63-69.
Kang, B.T., and Duguma, B., 1985. Nitrogen movement in
alley cropping systems. In: Kang, B.T., and van den
Heide, J. (eds.), Nitrogen in farming systems in the
humid and the sub-humid tropics. Institute of Soil
Fertility, Harn, The Netherlands.
Kang, B. T., Grimme, H. and Lawson, T.L., 1985. Alley
cropping sequentially cropped maize and cow pea with
leucaena on a sandy soil in southern Nigeria. Plant
and Soil 85, 267-277.
Kang, B.T., Reynolds, L. and Atta-Krah, A.N., 1990. Alley
farming. Advances in Agronomy 43, 315-359.


60
this observation. This response could be ascribed to
enhanced availability of nutrients; positive impact of mulch
on soil physical properties (e.g., improved aeration and
moisture conservation) could be another reason, but this
aspect was not examined in this study.
It has been argued that if mulch applied in a season
could be increased to 3-4 t ha*1, maize yield may be
increased significantly even in sites similar to the one of
the present study with relatively fertile soils. For
instance, Mugendi (1990), working at a site adjacent to the
present study, reported from a study spanning 14 seasons
that application of leucaena and cassia mulch at rates
equivalent to 5 t ha"1 consistently gave significantly
higher maize yields than applications of 1 t ha*1 or those
of the control plot with no mulch additions. Moreover,
more pronounced effects would be expected in nutrient-poor
than in nutrient-rich sites. For instance, Palm (1988)
observed that in more nutrient-poor and/or in acid soils,
increased rates of mulch application resulted in crop
responses similar to those of inorganic nitrogen
fertilizers.
Because a species like leucaena has a higher fodder
value compared to cassia which has no fodder value (NAS,
1984), in semiarid areas, the prunings from the leucaena
hedges are more likely to be removed from the field for use
as fodder (Singh et al., 1989). The use of the prunings for


156
control (no irrigation or fertilization). Each treatment
was replicated thrice in plots of 4 m x 8 m. Amounts of
irrigation water applied was the difference between pan
evaporation losses and rainfall. The plots were irrigated
once a week until harvest. Fertilizer was applied at the
rate of 100 kg N ha-1 at the time of sowing the crop.
Pot Studies: hertilization and
irrigation of soil collected from
near and away from the hedgerows
A factorial combination of three levels of N at
equivalents of 0, 40 and 100 kg N ha'1 season'1, and two
levels of irrigation were chosen as treatments for the pot
studies. Soil was collected from the top 40 cm profile near
(row 1) and away from the hedgerows (row 4) of leucaena and
cassia. The soil was contained in 30 x 30 x 30 cm pots
(polythene bags). Each pot contained 20 kg of soil. Each
of the treatments mentioned above was replicated three times
in randomized block design. Maize (Katumani Composite
variety) was grown in the pots for a total duration of two
months. Total biomass (shoot plus roots) of maize plants in
the pots was harvested and dry matter determined after oven
drying at 105C for 48 hours.
Irrigation levels applied to the pots were the
equivalents of 100 and 200 mm rainfall in a duration of two
months. This period is half a typical cropping season of
four months. The choice of the 100 mm and 200 mm rainfall


Infiltration rate (cm/min)
172
Species
Figure 5-5. Infiltration rates of water into soil
near (0.45 m) and away (3.5 m) from alley cropped hedgerows
of Leucaena leucocephala and Cassia siamea, 1991 short-
rains .


208
farm and mode of 0 head (35%). The fodder-interest group
had, on average, 30% more cattle than the mulch-interest
group. However, the number of livestock kept by both groups
(fodder- and mulch-interest group) were not significantly
different. Although 35% of the farmers interviewed did not
have cattle, some had immediate plans for buying them.
In terms of goats, the average holding was nine (sd:
8), with a range from 0 to 30. Twenty percent of the
farmers interviewed did not have goats. Differences in goat
possession of the fodder and mulch-interest groups were not
significant. However, the fodder-interest group had, on
average, 25% more goats than the mulch-interest group. For
sheep, the number possessed ranged from 0-13 with an average
of four (sd:4) sheep per farm. Eighty-three percent of the
farmers did not have sheep.
Farmers kept more chickens than other livestock.
Ninety-three percent of the farmers interviewed kept
chickens probably because of their relatively lower feed
requirements compared to other livestock. The range of
chickens kept ranged from 0 to 2000 birds. The average,
when the single farmer with 2000 birds was excluded as an
outlier, was 21 birds (sd: 21); the mode 20 birds (18.3% of
the farmers interviewed). Seven percent of the farmers
interviewed had no chickens. The difference in chicken
possession of the fodder and mulch-interest groups was not
significant (p = 0.50).


105
determined but assumed to be 50% of the ash-free dry weight
of the mulch. From the contents of C and N in the mulch,
the ratio of C:N was determined.
Acid digestible lignin and ash concentrations were
determined by the Van Soest method (Van Soest, 1968). Total
soluble polyphenols were determined at the Natural Resources
Institute, United Kingdom, using a revised Folin-Denis
method (King and Heath, 1967). This involved extraction of
the leaves, twigs and combined leaves plus twigs in 50%
aqueous methanol for two hours in water bath (80C). The
materials had previously been air dried for two days and
stored under room temperature before analysis. Extraction
and determination of polyphenol were carried out in
duplicates for each sample, using tannic acid supplied by
Sigma Chemicals as a standard. Total soluble polyphenols
are expressed as percent of tannic acid equivalents. The
chemical and physical characteristics of the mulch used are
shown in Table 4-1.
Pot studies
Pot studies were initiated in January 1992, to
determine interactive effects of mulch quality of the two
species and the seasonal rainfall on the patterns of N-
mineralization and uptake by maize. It consisted of the
application of a factorial combination of four mulch rates
of leucaena and cassia (the equivalents of 0, 2, 4, 8 t


214
fanners had only the plow because their oxen died for some
reason (e.g., during a major drought in 1987). In some
isolated cases, farmers had sold their oxen to meet some
urgent needs for cash. Between the two interest groups, 72%
and 35%, respectively, of the fodder- and mulch-interest ox-
plowed their fields. This difference was significant.
Use of soil erosion control measures
The common measure of soil erosion control was
terracing with soil mounds thrown on the upper side of cut
off trenches. This practice is locally called "Fanya Juu."
It is an expensive exercise to undertake initially, but it
is also an effective erosion control measure that requires
little maintenance thereafter. A significant proportion
(70%) of the farms surveyed were terraced well with the
"Fanya Juu." Differences in the percentage of terraced
farms among the fodder- and mulch-interest groups were not
significant.
Factors that Influenced Interests
of Farmers in Aliev Cropping
Five factors significantly influenced the proposed use
of the alley cropped trees for fodder or for mulch:
1. Possession of coffee trees,
2. Possession or use of ox-plow(s),
3. Use of mulch,
4. Use of fertilizer and,


244
as high as 8 to 10 t ha"1 yr"1) significant improvements in
both soil fertility and crop yields have been noted. The
low biomass yield of the hedgerows that also decomposed fast
(half life of four weeks) was not sufficient to effect
significant improvements in soil fertility (organic C, total
N and CEC) and maize yields. Maize yields of plots with or
without mulch were similar. If the amounts of mulch applied
in a season could be increased to 4 t ha"1 or more, there
were indications from pot studies that increases in maize
yield could be significant. To obtain such levels of mulch
would, however, require that the density of the hedgerows be
increased to 2 to 4 times that of the present study. This
could result in further loss in crop yields due to reduction
of area available for cropping and possible competition from
the hedgerows.
The small amounts of mulch produced by the hedgerows
may have had significant effects on soil fertility and
productivity if the present studies were long-term and/or
the site was nutrient-poor. The present study was conducted
for only three years and on land not previously cropped for
more than 2 years. Also, moderate levels of chemical
fertilizer (40 and 18 kg N and P, respectively) were applied
to the plots each season. The higher relative indices of
available (KCl-extract) N observed in plots that were
mulched compared to those that were not may be important to
the productivity of N deficient soils.


176
from them (Fig. 5-8). Differences between the effects of
species was not significant. In spite of the trends of
generally higher water content near the hedgerows than away
from them, significant difference was, however, noted in
only August, November and December.
With respect to the tensiometer readings (Fig. 5-9),
no significant differences were observed between the effects
of species or position of the crop row relative to the
hedgerow on soil water potential at 0-15 cm depth. This was
so whether or not daily readings or monthly averages were
considered. As expected, however, soil moisture potentials
were higher (negative) during the dry season (e.g., August)
and less negative during the rainy season (e.g., May and
December).
During the dry season (February and June-September)
soil temperature away from the hedgerows was always
significantly higher than that near the hedgerows
(Fig. 5-10). This was so at both 10:00 hrs and at 14:00
hrs. In the dry season, the difference between soil
temperature near the hedgerows and away from them could be
as high as 3C. In the rainy season (e.g., May and October
to December), the differences were not significant.
Differences between the effects of species on soil
temperature near and away from the hedgerows were not
significant.


162
In season 6, significant main effects of both species
and row distance from the hedgerows were observed (Fig. 5-
2). Interaction effects were not significant. As in the
first season, yield of rows under cassia was significantly
higher than those under leucaena. On average, the
difference was 49%. Compared to the sole crop (fertilized),
yield under cassia, averaged over all the rows, was 5% lower
while that under leucaena was 51% lower. With respect to
the effects of row distance from the hedgerows, grain yield
declined significantly with distance from the hedgerows of
both species.
In season 7, no significant effects of interaction
between species and row distance of crops from the hedgerows
was observed. Main effects of species and distance of the
maize rows from hedgerows were significant. Maize yield
declined in a significantly linear manner with distance from
the hedgerows of both species (Fig. 5-3). On average, maize
yield under cassia was 2% and 18% higher than under sole
crop and leucaena hedgerows, respectively. Therefore, the
maize yield under leucaena was 16% lower than that of the
sole crop.
Averaged over all the three seasons, the first row of
maize next to the hedgerows was 7% and 12% more productive
than the next row in leucaena and cassia, respectively.
When averaged over the sixth and seventh seasons (when
significant trends of yield declines with distance from


Carbon (log g) remaining
3.5
>


Figure 4-6. Relationship between the patterns of N release from the mulch
of Leucaena leucocephala and Cassia siamea and biomass yield of maize from four
consecutive crops of maize sown in pots and harvested at an interval of four
weeks.


228
dietary requirements. In some cases, it can also reduce
need for fertilizer.
The significantly higher interest in alley cropping
for fodder (62%) rather than for mulch (33%) was expected
from farmers in an arid area where feeding livestock is a
problem. However, the large number of farmers (33%) who
have livestock but who want to use the technology for mulch
rather than for fodder raises some speculation. First, it
is possible that the mulch-interest group was not as well
informed as the fodder-interest group and that their
responses could change to fodder as they acquire more
knowledge. This change is feasible because the response of
the farmers to fodder or mulch use was observed to depend
significantly on prior knowledge about alley cropping.
Coffee farmers, who knew most about the technology,
responded positively to fodder use. Secondly, if one
assumes that the response was not due to lack of knowledge,
then it would appear that there was a real interest in alley
cropping for purposes of mulch. It is possible that some
farmers were well informed about the technology and,
therefore, made a rational decision. For example, coffee
farmers were more informed about the technology than the
noncoffee farmers, and most coffee farmers responded
positively to fodder-interest. However, the considerable
number of coffee farmers (24%) who responded positively to
mulch-interests would suggest that they made a rational


250
Coleman, D.C., Edwards, A.L., Belksy, A.J., and Mwonga, A.,
1991. The distribution and abundance of soil nematodes
in East African Savannas. Biology and Fertility of
Soils 12, 67-72.
Comanor, P.L., and Staffeldt, E.E., 1978. Decomposition of
plant litter in two western North American deserts,
pp. 31-49. In: West, N.E., and Skujins, J.J., (eds.),
Nitrogen in desert ecosystems. US/IBP series #9.
Dowden, Hutchinson and Ross, Inc. Publishers.
Comerford, N.B., Kidder, G., and Mollitor, A.V., 1984.
Importance of subsoil fertility to forest plant
nutrition. In: Stone, E.L., (ed.), Forest soils and
treatment impacts. Proceedings of the Sixth North
American Forest Soils Conference, Univ. of Tennessee,
Knoxville.
Corlett, J.E., 1989. Leucaena/millet alley cropping in
India: Microclimate and productivity. Ph.D.
Dissertation, University of Nottingham, U.K.
Corlett, J.E., Ong, C.K., and Black, C.R., 1989.
Microclimate modification in intercropping and alley
cropping systems, pp. 419-430. In: Reifsnyder, W.S.,
and Darnhoffer, T.O., (eds.), Meteorology and
agroforestry. ICRAF, Nairobi, Kenya.
CSIRO, 1988. CSIRO disc permeameter instruction manual.
Commonwealth Scientific Industrial Research
Organization, Canberra, Australia.
Danso, S.K.A., Zapata, F., and Hardarson, G., 1987.
Nitrogen fixation in fava beans as affected by plant
population density in sole or intercropped systems with
barley. Soil Biology and Biochemistry 19, 411-415.
Deans, J., and Ford, E.D., 1983. Modelling root structure
and stability. Plant and Soil 71, 189-196.
DeHaan, S., 1976. Humus, its formation, its relation with
the mineral part of the soil, and its significance on
soil for soil productivity, pp. 21-30. In: Soil
organic matter studies. Volume 1. International Atomic
Energy Agency, Vienna, Austria.
Dhyani, S.K., Narain, P. and Singh, R.K., 1990. Studies on
root distribution of five multipurpose tree species in
Doon Valley, India. Agroforestry Systems 12, 149-161.
Dijkman, M.J., 1950. Leucaenaa promising soil erosion
control plant. Economic Botany 4, 337-47.


122
leaves to twigs remaining over time (data not presented) shows
interesting contrasts that suggest some exceptions to the
generalization. In both seasons, the ratio of leaves to twigs
of cassia mulch remaining over time was consistently higher
than that of leucaena. This was true for both soil-
incorporated and surface-placed mulch. As a result, the rates
of cassia leaf decomposition were often only marginally higher
than those of the twigs. In leucaena, on the other hand, the
ratio of leaves to twigs declined rapidly over time. This is
consistent with the observations of the significantly higher
rate of decomposition of leucaena leaves compared to twigs.
Filter paper decomposition
Soil incorporated filter paper had both phases one and
two rates of decomposition. The rates were similar to those
of the mulch or the twigs incorporated (Table 4-2). Soil
surface placement was not included as a treatment for the
filter papers.
Trends of nitrogen loss from the mulch
The patterns of N loss from the mulch of leucaena and
cassia were similar to those of the dry matter or carbon loss
discussed above. In both species, an initial phase of rapid
rate was observed which declined significantly with time (Fig.
4-5). The rates of N loss of leucaena and cassia mulch in the
first phase of decomposition were 15.9% and 16.9% per week,


92
hedgerow intercropped and the block planted system of both
species. The levels of mulch produced by the hedgerows in the
present study were within the range of plant residues
requirements estimated by Young (1989) to maintain top soil
organic matter in semiarid and subhumid ecozones of the
tropics.
The contribution of the root system of alley cropped
hedgerows to soil organic matter, and thus to soil
fertility, has been little studied. In addition to the
contribution of above ground litter, turnover of high root
biomass, which is generally higher in agroecosystems
involving trees than in other land use systems (Ewel et al.,
1982), is a mechanism by which trees are hypothesized to
improve soil fertility. It is also hypothesized that,
through their deep roots, trees such as leucaena and cassia
are able to absorb nutrients from soil depths that crops
cannot reach, and thus, improve soil fertility through
litter fall and root turnover (Nair, 1984). While the roots
of the intercropped hedgerows may contribute to organic C
levels of the soil over time, this was not observed in the
short-duration of the present study (3-4 years). It is
implicit from the above that, especially for a species such
as leucaena which is highly competitive, the presence of the
hedgerows in the crops may not be necessary for purposes of
maintaining soil organic matter/soil fertility.
Although there were no significant differences in soil
organic C and total-N between treatments, the results of the
KCl extractions indicated that plots with leucaena (whether


62
mulch and those without for the seasons of this study. From
the foregoing, therefore, it is evident that improvements in
crop yields through the effects of mulch produced by alley
cropped hedgerows are probably limited in the semiarid
tropics.
Supply of N and P by the mulch
and removal by the crop
Leucaena mulch, when returned to the soil, provided
more N than what the crop removed in 6 seasons. On the
other hand, cassia supplied only half of the crop N. Most
of the N removed was harvested in the grain (67%), an
observation in agreement with that of others (Allison,
1973). Based on the N budget analysis, it does appear,
therefore, that if all of the mulch N is available to the
crop, leucaena could meet the crop's seasonal N
requirements.
Besides N, the supply of the other nutrients is
important for the maintenance of high maize yield over time.
Removal of the mulch which supplies substantial quantities
of nutrients other than N could therefore have significant
effects on yield, even when N is provided via fertilizer.
This view is supported by observations of Gichuru and Kang
(1989) who showed that plots receiving no prunings had, in
general, lower maize grain yield regardless of N application
compared to those receiving prunings.


164
Table 5-1. Nitrogen content of 30 day-old maize shoots
(leaves plus stalks) sampled near (0.45 m) and
away (3.5 m) from alley cropped hedgerows of
Leucaena leucocephala and Cassia siamea, fifth
season, 1991.
N content
(%)
Species
Near
(0.45 m)
Away
(3.5 m)
Mean
Leucaena
3.3
.3.3
3.3a
Cassia
3.1
3.2
3.2a
Mean
3.2
3.3a
3.3
Means in a column
not significantly
and row followed
different (p = 0
by the same
.05) .
letters are


49
Nutrient budgets
Nitrogen.Generally, leucaena had significantly
higher mulch yield (Table 2-1) and N concentration (Table
2-2) than cassia. As a consequence, the N yield of leucaena
was significantly higher (136 kg N ha-1 yr*1) than that of
cassia (59 kg N ha*1 yr*1).
Nitrogen uptake was high and essentially of similar
pattern in all treatments, including the absolute control
(Table 2-6). It was only the fertilizer-control and the
alley cropping system of cassia that had N-yields
considerably higher than the rest of the treatments. This
is because of the relatively high maize yields of these two
treatments (i.e., the fertilizer-only control and the
hedgerow intercropped plots of cassia).
Apparent recovery/utilization efficiency of N applied
through the mulch or fertilizer was low in all treatments
(Table 2-6, columns 7 and 8). The N-recovery values were
low primarily because maize and N-yield of the absolute
control were almost equal to those of the other treatments
with inputs. This suggests that the N needs of the crop
were adequately met by native soil N. Averaged across
treatments, an apparent N-recovery of 9% summed over the N
recoveries through the grain, stover and cobs was observed.
Most of the N (67%) was recovered in the grain component of
the crop and the rest (33%) in the stover and cobs.


199
nature complex, (at least two components are involved, a
crop and a tree/shrub) and, therefore, may be more difficult
to adopt than systems involving a single component (tree or
crop). Simplicity or complexity of the technology
notwithstanding, adoption of alley cropping may be higher
where the trees provide multiple and valuable products,
e.g., fodder. This is a particularly important
consideration in the semiarid tropics where livestock feed
is scarce during the dry season, and trees often provide
valuable fodder (Singh et al., 1986).
Understanding of the existing farming systems of the
area, production priorities and constraints of the farmers
have long been recommended as an absolutely necessary input
into the agenda and process of on-station technology
development (Spedding and Brockington, 1976). Where these
recommendations have been followed, especially through the
iterative Farming Systems Research and Extension
methodology, there is evidence to suggest that the
potentials for farmer acceptance and adoption of
agricultural technologies are high (Hildebrand, 1981). On-
farm trials, either researcher-managed and/or researcher-
farmer managed (Hildebrand and Poey, 1985; Atta-Krah and
Francis, 1987) are among some of the research approaches
used to enhance farmer participation and evaluation of
technologies being developed. Although certainly better
than the total reliance on on-station trials, the on-farm


217
Table 6-4,. The p-values of dependency of resource
possession, their use or purchase on the
presence of coffee on the farm, Machakos, Kenya,
August 1991.
Resource
p-values of dependency
of resource
use/possession
on the presence
of coffee on
the farm
Purchase of food
0.07
Purchase of chemical fertilizer
0.04
Hire of ox-plow
0.01
Purchase of manure
0.17
Purchase of animal feed
0.13
Hire of labor
0.06
Frequency of weeding
0.04


109
proportion of the initial dry matter or nitrogen remaining
after a period of time t (in weeks) was first fitted. A
plot of time against the natural logarithm of this first-
order exponential model (which linearizes the model) was
made for each replicate. The logarithmic plots revealed
that the patterns of dry matter or N loss were biphasic and
a single negative exponential model could not provide the
best fit.
Regression parameters (i.e., the intercept and slope
of each phase) and the spline (meeting) points of the
regression lines of the two phases were then generated for
each replicate of the two species. ANOVA test was then
performed on each of the regression parameters to determine
significance of differences between the regression
parameters of the two species.
For the pot experiment, the GLM procedure of the SAS
(SAS, 1992) was used to test for differences between the
effects of the four treatments (i.e., leucaena mulch, cassia
mulch, fertilizer control and the absolute control) on maize
biomass yield of the pots. The option for repeated measures
ANOVA was included in the GLM procedure to account for the
systematic effects of repeated harvests of biomass from the
same pots over time.
When significant effects of treatments were detected,
means were separated by pairwise contrasts of the LSM


30
irrigation levels, mulch species and mulch rates. Pairwise
comparison of LSM was used to test the significance of
differences between treatment means. Dunnet's test, which
controls the Type 1 experiment error for comparisons of all
treatments against a control, was used to determine
significance of differences between the mulch rates and the
control pots that had only one level of fertilizer
application.
Results
Field Study
Mulch yield
Leucaena, intercropped in hedgerows, produced
significantly higher mulch (+25%) than when the species was
planted in sole/block stands (Table 2-1). For cassia,
however, the +14% observed difference was not significant.
Mean mulch yield of leucaena, whether it was
intercropped or planted in sole/block stands, was
significantly higher (+46%) than that of cassia (Table 2-1).
Between the three hedgerow spacings of the intercropped
system, mulch yield of the 4 m hedgerows was significantly
(p = 0.04) higher (on average, 21%) than that of the other
two hedgerow spacings. The mulch yields of 5 and 6.7
m-wide hedgerows were not significantly different
(p = 0.23).
I


181
Species
Figure 5-11. Counts of fauna in 25 x 25 x 25 cm top
soil monoliths taken near (0.45 m) and away (3.5 m) from
alley cropped hedgerows of Leucaena leucocephala and Cassia
siamea at the end of 1991 short-rain season.


248
Barley, K.P., 1954. Effects of root growth and decay on
the permeability of synthetic sandy loam. Soil Science
78, 205-210.
Bartholomew, W.V., 1965. Mineralization and immobilization
of nitrogen in the decomposition of plant and animal
residues. In: Bartholomew, W.V. and Clark, F.E.
(eds.) Soil Nitrogen. American Society of Agronomy,
Madison.
Belsky, A.J., Amundson, R.G., Dixbury, J.M., Riha, R.J.,
Ali, A.R., and Mwonga, S.M., 1989. The effects of
trees on their physical, chemical and biological
environments in a semiarid savannah in Kenya. Journal
of Applied Ecology, 26, 1005-1024.
Belsky, A.J., 1992. Effects of trees on nutritional quality
of understorey gramineous forag in tropical savannas.
Tropical Grasslands 26, 12-20
Bennet, J.M., 1990. Problems associated with measuring
plant water status. HortScience 25, 1551-1554.
Berg, B., 1986. Nutrient release from litter and humus in
coniferous forest soilsa mini review. Scandinavian
Journal of Forestry Research 1, 359-369.
Binkley, D., and Hart, S.C., 1989. The components of
nitrogen availability assessment in forest soils.
Advances in Soil Science 10, 57-112.
Binkley, D., and Matson, P., 1983. Ion exchange resin bag
method for assessing forest soil nitrogen availability.
Soil Science Society of America Journal 47, 1050-1052.
Blake, G.R. 1965. Bulk density, pp. 374-390. In: Black,
C.A., (ed.), Methods of soil analysis. American
Society of Agronomy.
Bohm, W., 1979. Methods of studying root systems.
Springer-Verlag, Berlin.
Bonkoungou, E.G., 1992. Sociocultural and economic
functions of Acacia albida, pp. 1-8. In: Vandenbeldt,
R.J. (ed.), Faitherbia albida in the West African
semiarid tropics. Proceedings of a workshop held at
Niamey, Niger, 22-26 April.
Bowen, G.D., 1985. Roots as components of tree
productivity, pp. 303-315. In: Cannell, M.G.R., and
Jackson, J.E. (eds.), Attributes of trees as crop
plants. Institute of Terrestrial Ecology, Huntingdon,
U.K.


Page
6-2. Interests of fanners in alley cropping, farm
size and number of members in a household
at Machakos, Kenya, August 1991 207
6-3. Factors which influenced significantly
interests of farmers in alley cropping for
fodder or for mulch and the frequency of
farmers under each factor, Machakos, Kenya,
August 1991 216
6-4. The p-values of dependency of resource
possession, their use or purchase on the presence
of coffee on the farm, Machakos, Kenya,
August 1991 217
6-5. Estimates of maize yield (with or without
inputs) of farmers with interests in fodder
or mulch aspects of alley cropping, Machakos,
Kenya, August 1991 219
6-6. Functions of tree nurseries possessed by
30% of the 60 farmers, 1991 interviewed,
Machakos, Kenya, August 1991 . 224
6-7. Reasons given by farmers without nurseries
on their farms, Machakos, Kenya, August 1991 225
x


8
SATs, unlike in the humid tropics, mulch yields from alley
cropped species are lower than in the humid tropics. For
instance, mulch yields of 2 t ha1 yr*1 from leucaena is
typical in SATs (Nair, 1987) as opposed to 8-10 t ha1 yr1
in lowland humid tropics. As Huxley (1986) postulated, this
low mulch yield in SATs may limit the potential of alley
cropping to improve fertility and productivity of soils in
these regions.
Information on the yields as well as the rates of
decomposition of mulch from alley cropped MPTS (which in
turn affects the ability of the mulch to provide nutrients
to the crops and to contribute to soil organic matter) are
relatively scarce in the SATs. Also, the benefits of the in
situ presence of hedgerows to soil fertility and
productivity are not well understood in either the humid or
the semiarid tropics.
In addition to these technical questions, there is
also the issue of social acceptability of the new technology
by the target farmers. It is necessary to determine how
farmers will adapt and use the technology under conditions
unique to their systems of farming. For instance, in the
semiarid areas where prunings from hedgerows as well as crop
residues are used as fodder (Singh et al., 1986), the in
situ presence of hedgerows may be the main avenue for
improving soil fertility. Studies of this nature will
provide insights into determining the technical feasibility


231
The labor requirement is always a factor which rural
people take into consideration when deciding whether or not
to adopt a new practice (Hoskin, 1987). Alley cropping is a
labor intensive practice and the costs of production
increase considerably if additional labor has to be hired
(Hoekstra, 1987). It is suggested that alley cropping is
economically attractive under conditions of severe cash
constraint where hired labor is available at relatively low
cost (Raintree and Turray, 1980; Verinumbe et al., 1984;
andf Sumberg et al., 1987). In the area of the study,
relatively low cost labor was available probably due to
declining farm size and productivity. Arnold (1987)
hypothesized that both of these factors result in an
increased labor pool seeking off-farm employment which may,
in turn, influence the adoption of agroforestry practices.
Under these conditions, alley cropping could be economically
viable for farmers, such as those producing coffee, who have
the ability to undertake the initial investments required by
the technology and, then, to hire the labor necessary to
manage it.
It could be argued that coffee farmers showed a higher
interest in alley cropping for fodder possibly because most
of their land was under coffee production rather than food
or feed crop production. However, farm size did not
influence significantly the interests of farmers in the use
of alley cropping for fodder or for mulch. Coffee farmers


14
conditions, the trees could be grown separately from the
crops, and their biomass harvested for use as fodder or
mulch in a "cut and carry system."
The magnitude of the effects of in situ planting of
the hedgerows would be influenced significantly by
differences in the rooting patterns and densities of the
hedgerow species. Trees with high root densities (expressed
as cm cm*3 of soil volume) in the rooting zone of the crops
are likely to result in less positive in situ benefits to
crops than trees with low root densities. Also, for a given
species, between-row spacing of the hedges, or put in other
words, the ratio of tree:crop land occupancy, could modify
the effects of in situ planting. Because of the increased
chances of competition, hedgerows with narrow between-row
spacings would be more likely to have less positive in-situ
effects on crops than widely spaced hedgerows. Narrowly
spaced hedgerows may, on the other hand, produce higher
quantities of biomass and provide greater scope for nutrient
recycling than widely spaced hedgerows.
From the foregoing, it is apparent that an
understanding of the in situ versus mulch benefits of
growing hedgerows in crops is necessary. Additionally, it
is important to have an understanding of the rooting
densities of species involved. The objective of this study
was to separate the effects of mulch from those of in situ
planting of the hedgerows on productivity (biomass and crop


234
Emerging Themes
One of the themes emerging from this study is that
farmers invested in soil fertility improvement and
productivity. Soil conservation, weeding, possession, or
use of an ox-plow could all be considered as investments.
These investments require cash and so will alley cropping
for establishment and maintenance of the hedgerows.
Farmers with coffee are more likely to acquire the
inputs to improve or sustain food production. Coffee was
<3
used as collateral to secure loans for education of the
children. Coffee is just an example of how a cash
generating activity can influence farmers' potential to
invest and adopt new technologies. It could be replaced by
any other enterprise. In fact, because of low prices of
coffee at the time of the present study, farmers were
replacing their coffee with fruit trees, vegetables, and
poultry, as alternative cash generating activities. One
could also visualize trees planted for production and sale
of fodder as an alternative technology. Planting and sale
of fruits, poles from trees (predominantly Eucalyptus but
also Grevillea), and firewood was common.
The changes from coffee to other enterprises as cash
generating activities are, indeed, positive. Coffee prices
have oscillated widely and farmers have not been able to
intercrop coffee (which they would like to do) because of
strict government protection of coffee. Sale of products


LIST OF FIGURES
Page
2-1. Seasonal rainfall and the local names
(long-rains and short-rains) of the
cropping seasons during the study 16
2-2. Mulch yield of Leucaena leucocephala and
Cassia siamea in alley cropping and
block/sole planting systems 33
2-3. Effect of three hedgerow spacings on maize
grain yield, averaged over the yields
under the two species, Leucaena leucocephala
and Cassia siamea 37
2-4. Effect of sole stands of Leucaena leucocephala
and Cassia siamea planted at three land
occupancy on maize yield during six seasons . 38
2-5. Relationship between rainfall and maize mean
yield of all treatments in each of the six
cropping season ........ ... 42
2-6. Maize grain yield under alley cropping and
block planting systems of Leucaena leucocephala
and Cassia siamea expressed as percent of the
sole maize control yield 44
2-7. Maize productivity (Kg m'2) under alley cropping
Leucaena leucocephala and Cassia siamea hedgerows
for six seasons 45
2-8. Productivity of maize (kg mm-1 of rainfall)
under alley cropping of Leucaena leucocephala
and Cassia siamea hedgerows for six cropping
seasons 46
2-9. Fine root density in the top 40 cm soil profile
of alley cropped hedgerows of Leucaena
leucocephala and Cassia siamea 54
2-10. Effect of three mulch rates of Leucaena
leucoc-phala and Cassia siamea on biomass
yield of maize grown in pots, 1991 56
xi


90
Generally, though, high rates of mulch application (in
the range of 10-20 t ha'1 yr'1 dry matter) are effective in
increasing soil organic C, as concluded by Huxley (1986)
from a comprehensive review of the literature. The
quantities of mulch obtained from the hedgerows in the
present study were only about 2 and 4 t ha'1 yr'1 from cassia
and leucaena, respectively (Chapter 2). Hence, the
limitation to increasing soil organic C and N in the soil of
the field plots of the present study was most probably the
limited amounts of mulch applied. Under such conditions, it
may take more time to detect changes in soil organic C and
total N because the additions and possible changes are small
relative to the soil store (Powlson et al., 1987).
Although productivity may decline due to insufficient
availability of nitrogen (Ayanaba, 1982), slowly decomposing
mulches, high in lignin, polyphenol content or high C/N
ratios, typically, have more potential to increase soil
organic carbon than fast decomposing ones (DeHaan, 1977;
Sivapalan, 1982). The characteristics of the mulch of the
two species were similar (Table 3-1); however, their rates
of decomposition on the soil surface differed slightly, with
cassia decomposing more slowly than leucaena (Chapter 4).
Thus, the lack of significant differences observed in soil
organic C between soils under cassia and leucaena mulch
could be due to a lack of differences in the mulch quality
of the two species. The mulch of both species may have been
of relatively high quality, despite the slight differences


89
applied, no significant treatment-related differences were
found in soil organic C and N. Changes in organic C and N
after three years of cropping were also not significant.
Insufficient quantities of mulch with high rates of
decomposition could be an explanation for the lack of
significant differences. The former explanation appears
more plausible than the latter as Kang et al., (1985) also
reported significant increases in SOM as well as
exchangeable K, Ca and Mg levels in both the top- and sub
soil from the application of high rates of leucaena mulch
(7 t ha'1 yr'1). The mulch decomposition rates in the study
of Kang et al. and the present one were similar. Also, pot
studies involving additions of high rates of leucaena and
cassia mulch (i.e., 4 and 8 t ha'1 season'1) to soil from
plots of the present study resulted in significant increases
in maize biomass yield. This was probably due to the
effects of increased soil organic matter and supply of
nutrients from the mulch. In the pot studies, rates of
equivalent mulch application lower than 4 t ha'1 yr'1 had no
significant effects on maize biomass yield. Other workers
have also reported increases in soil organic carbon content
from the additions of high rates of leucaena and cassia
mulch (Yamoah et al., 1986c). However, as studies by Palm
(1988) suggest, differences between sites, soils and mulch
species could significantly impact the effects of high rates
of mulch application on soil organic C.


178
Figure 5-9. Monthly average reading of tensiometers
placed in the soil at 15 cm depths near (0.45 m) and away
(3.5 m) from alley cropped hedgerows of Leucaena
leucocephala and Cassia siamea in the long- and short-rain
seasons of 1991.


115
Twigs
No significant effects of species or placement methods
on the rate of twig decomposition of both phase one and two
were observed in either season (Fig. 4-3). However, the
ratio of leucaena to cassia twigs remaining over time
provides interesting contrasts between the two species (data
not shown). Leucaena had a higher ratio of twigs remaining
not decomposed than cassia in both seasons and placement
methods. This was most pronounced in the first season when
an average ratio of leucaena to cassia twigs of 1.6:1 was
observed at the end of the study. This would also suggest
that leucaena twigs were more resistant to decomposition
than cassia.
Rates of decomposition of twigs placed outside the
litter bags (Table 4-3) were not significantly different
from those inside the litter bags, regardless of placement
methods.
Mulch (leaves plus twigs)
On the soil surface, the rate of decomposition of
leucaena mulch in the first phase was significantly higher
than that of cassia (Fig. 4-4). Leucaena mulch placed on
the soil surface also had a significantly higher rate of
phase one decomposition than the soil-incorporated mulch in


191
hedgerows. Difference in soil temperature near and away
from the hedgerows was significant at the end of the
cropping period (February, March, June and July) and during
the dry season (August and September). At this time, the
hedges are typically not cut back since the crop was either
mature or was not present in the field. Hence, considerable
shade could build up near the hedgerows that, at times, grew
up to as high as 2 m. During the cropping season, however,
there was no significant difference in soil temperature near
and away from the hedgerows because the hedgerows were cut
back 2-3 times in a season.
Also, cloudy conditions prevailing during the rainy
season could have minimized differences in soil temperature
both near the hedgerows and away from them. So, although
effects of reduced temperature (and therefore reduced
respirational losses) near the hedgerows have been suggested
to explain the higher crop yield near the hedgerows (Mungai,
1991), it is unlikely to have been a major factor in the
present study. The basis for this assertion is that there
were no significant reductions in soil temperature when the
crop was present in the field. A similar conclusion is made
by Corlett et al. (1989) who observed no significant
differences in the canopy temperature of sorghum alley-
cropped with leucaena and sorghum grown in the open at
semiarid conditions at Hyderabad, India.


Table 4-2. Intercepts, slopes, spline points and coefficients of determination (r2) of
the regression lines for phase one and phase two of Leucaena and Cassia mulch
decomposition.
Decomposition rate (%)
Spline
Intercept Phase one Phase two point
First season
Mulch (incorporated)
Cassia 2.2
Leucaena 2.0
Second season
Mulch (incorporated)
Cassia 2.3
Leucaena 2.3
Mulch (surface)
Cassia 2.3
Leucaena 2.0
Twigs (incorporated)
Cassia 1.0
Leucaena 1.1
Leaves (surface)
Cassia 3.0
Leucaena 2.9
Leaves (incorporated)
Cassia 2.8
Leucaena 3.0
Twigs (surface)
Cassia 1.5
Leucaena 1.3
Filter paper (incorporated) 1.8
12.1
0.1
9.1
0.82
13.8
*
*
0.84
11.0
2.0
9.3
0.95
12.5
1.9
7.0
0.97
13.3
2.4
7.9
0.97
16.5
3.2
7.3
0.98
12.3
3.0
8.4
0.96
7.5
2.0
7.4
0.85
11.1
2.4
9.2
0.91
15.5
*
*
0.97
13.1
2.3
8.5
0.94
15.5
*
k
0.98
11.5
3.2
7.5
0.91
10.5
4.6
8.4
0.77
10.5
5.2
11.1
0.81
NOTE: Mulch refers to leaves plus twigs.
*
Refers to absence of phase two in the
decomposition pattern of the species.


159
Nearest hedgerow Farthest from hedgerow
Figure 5-1. Maize yield response to distance from
alley cropped hedgerows of Leucaena leucocephala and Cassia
siamea, 1990 short-rains. Vertical bars indicate standard
error of difference of means.


139
Although high lignin:N ratio appears to be an
explanation for the low rate in the second phase compared to
the first of decomposition, an examination of the lignin:N
ratios of the mulch at the start of phase two suggested
either some differences between the two species and/or the
presence of another factor influencing the rate of
decomposition. Mellilo et al. (1982) observed lignin:N
ratio to provide the best prediction of N release from
decomposing plant materials, although several workers have
not found similar relationship (Schlesinger and Hasey, 1981;
McClaugherty et al., 1985; Taylor et al., 1989). At the
start of phase two, the lignin:N ratio of the leucaena mulch
was 6.5:1 while that of cassia was 20:1. The lignin:N ratio
of cassia of 20:1 is within the upper range of the lignin:N
ratio of the plant materials that Mellilo et al. (1982)
used. Therefore, high C:N ratios or more specifically, high
lignin:N ratio, may be the primary factor regulating the
rate of cassia mulch mineralization in phase two.
For leucaena, the lignin:N ratio of 6.5:1 is below the
wide range of plant lignin:N ratios that were studied by
Mellilo et al. (1982) used. Therefore, wide C:N ratio (or
lignin:N ratio) per se may not be a good predictor for
decomposition of leucaena mulch in the second phase. The
polyphenol levels of leucaena twigs were within the range
observed by other workers to limit the rates of
decomposition of some leguminous materials (Palm and
Sanchez, 1991; Oglesby and Fownes, 1992). Given that


261
Rasmussen, P.E., and Collins, H.P., 1991. Long-term impacts
of tillage, fertilizer and crop residue on soil organic
matter in temperate semi-arid regions. Advances in
Agronomy 45, 93-134.
Read, M.D., Kang, B.T., and Wilson, G.F., 1985. Use of
Leucaena leucocephala (Lam. de Wit) leaves as nitrogen
source for crop production. Fertilizer Research 8,
107-116.
Roberts, J., 1976. A study of the.root distribution and
growth in a Pinus silvestris L. (Scots pine) plantation
in East Anglia. Plant and Soil 44, 607-621.
Rocheleau, D.E., Weber, F. and Field-Juma, A., 1988.
Agroforestry in dryland Africa: participatory
planningprocess and methods. ICRAF, Nairobi, Kenya.
Roskoski, J.P., 1986. Nitrogen fixation in leguminous
trees. Ceiba 27, 227-229.
Ruhigwa, B.A., Gichuru, M.P., Mambani, B., and Tariah, N.M.,
1992. Root distribution of Acioa barteri, Alchornea
cordi folia, Cassia siamea, and Gmelina arbrea in an
acid ultisol. Agroforestry Systems 19, 67-78.
Sanchez, P.A., 1987. Soil productivity and sustainability
in agroforestry systems. In: Steppler, H.A., and
Nair, P.K.R. (eds.), Agroforestry: A decade of
development. ICRAF, Nairobi, Kenya.
Sang, F.K., and Hoekstra, D.A., 1987. On-station hedgerow
results of the first two seasons of the operational
phase. Research Report No. 4, Dryland Agroforestry
Research Project (DARP), Machakos, Kenya. KARI/ICRAF,
Nairobi.
Sanginga, N., Gwaze, D., and Swift, M.J., 1991. Nutrient
requirements of exotic tree species in Zimbabwe. Plant
and Soil 132, 197-205.
Sanginga, N., Mulongoy, K., and Ayanaba, A., 1986.
Inoculation of Leucaena leucocephala (Lam.) de Wit with
Rhizobium and its contribution to a subsequent maize
crop. Biological Agriculture and Horticulture 3, 347-
352.
SAS, 1992. Statistical analysis system for linear models.
Third edition, SAS Institute Inc., N.C.
Sayre, I.D., 1955. Mineral nutrition of corn, pp. 293-314.
In: Spraque, G.F. (ed.), Corn and corn improvement.
Academic Press, New York.


121
the second season. Soil-surface-placed leucaena mulch
decomposed at the rate of 17% per week in phase one, with a
half-life of 4.1 weeks. On the other hand, cassia mulch on
the soil surface had a phase one decomposition rate of 12.5%
per week. The half-life corresponding to the rate of
decomposition of the cassia mulch was 5.5 weeks. In the first
season, only soil incorporation was carried out, hence,
comparison of effects of placement methods was not possible.
Within the soil, the rates of phase one or phase two
decomposition of the two species were not significantly
different in the first season or the second season.
Interaction effects between species and placement methods on
the rates of decomposition in the first phase were also not
significant in both seasons. On average, both species
decomposed at the rate of 11.0% and 11.5% per week in phase
one in the first and second season, respectively. The half-
life corresponding to these rates of decomposition were 6.3
and 6.1 weeks in the first and second season, respectively.
In the second phase, the effects of species, placement
methods and season were not significant. Averaged across
species and placement methods, the rates of decomposition in
phase two were 2.7% per week in the first season and 2.1% per
week in the second season. Half-lives corresponding to these
rates were 25.2 weeks and 33.2 weeks in the first and second
seasons, respectively.
As expected, twigs generally decomposed at significantly
lower rates in phase one than leaves. However, the ratio of


260
Parr, J.F., and Papendick, R.I., 1978. Factors affecting
the decomposition of crop residues by microorganisms,
pp. 101-129. In: Oschwald, W.R. (ed.), Crop residue
management systems. American Society of Agronomy,
Madison, Wisconsin.
Patten, D.T., 1978. Productivity and production efficiency
of an upper Sonoran Desert ephemeral community.
American Journal of Botany 65, 891-895.
Persson, H., 1980. Spatial distribution of fine-root
growth, mortality and decomposition in a young Scots
pine stand in central Sweden. Oikos 4, 77-87.
Powlson, D.S., Brookes, P.C., and Christensen, B.T., 1987.
Measurement of soil microbial biomass provides an early
indication of the changes in total soil organic matter
due to straw incorporation. Soil Biology and
Biochemistry 19, 159-164.
Probert, M.E., Okalebo, J.R. and Simpson, J.R., 1990. The
role of farm yard manure for improving soil fertility.
KARI-ACIAR-CSIRO Symposium on a search for strategies
for sustainable dryland cropping in semi-arid eastern
Kenya. Nairobi, 10-11 September.
Radwanski, S.A., and Wickens, G.E., 1967. The ecology of
Acacia albida on mantle soils in Zalingei, Jabel Marra,
Sudan. Journal of Applied Ecology 4, 569-578.
Raintree, J.B., and Turray, F., 1980. Linear programming
model of an experimental leucaena-rice alley cropping
system. IITA Research Briefs 1, 5-7.
Rao, M.R., Muraya, P., and Huxley, P.A., 1992. Observations
on some tree root systems grown in agroforestry
intercrop situations including their graphical
presentation. ICRAF, Nairobi.
Rao, M.R. Ong, C.K.,, Pathak, P., and Sharma, M.M. 1991.
Productivity of annual cropping and agroforestry on a
shallow alfisol in semi-arid India. Agroforestry
Systems 15, 51-63.
Rao, M.R., Reg, T.J., and Willey, R.W., 1987. Response of
cereals to nitrogen in sole cropping and intercropping
with different legumes. Plant and Soil 101, 167-177.
Rao, M.R., Sharma, M.M., and Ong, C.K., 1990. A study of
the potential of hedgerow intercropping in semi-arid
India using two-way systematic design. Agroforestry
Systems 11, 243-258.


69
Trends in maize yield over time
All treatments had similar levels of relative yield
maintenance as long as rainfall received in a season was
normal, a situation more pronounced under leucaena than
under cassia. Much longer studies than three years are
required because the effects of the hedgerows are likely to
be more pronounced with age. Also, long-term studies would
more clearly reflect the high variability in seasonal
rainfall. From six-year studies in the humid lowlands of
Nigeria, Lai (1991) observed that total maize yields from
alley cropping of leucaena and Gliricidia sepium were about
10% lower than those of the control. In the present study,
total maize yield after 6 seasons under alley cropping with
leucaena was reduced by 12% while that under cassia was
increased by 8%. Therefore, under management conditions
similar to the ones of the present study, it does appear
that cassia has more potential to sustain maize grain yield
than leucaena.
Because of the low measured concentrations of
extractable N and P in the soils of the study station (Kibe
et al., 1981), fertilizer N and P was applied to the crop
each season at the minimum of the recommended rates for
farms in the area. However, the general lack of significant
differences in maize yields between treatments suggests that
N and P were not limiting, again supported by the results of
the pot study.


262
Schlesinger, W.H., and Hasey, M.M., 1981. Decomposition of
Chaparral shrub foliage: losses of organic and
inorganic constituents from deciduous and evergreen
leaves. Ecology 62, 762-774.
Shield, J.A., and Paul, E.A., 1973. Decomposition of Re
labelled plant material under field conditions.
Canadian Journal of Soil Science 53, 297-306.
Sivapalan, K., 1982. Humification of polyphenol-rich plant
residues. Soil Biology and Biochemistry 14, 309-310.
Sivapalan, K., Fernando, V. and Thenabadu, M.W., 1985. N-
mineralization in polyphenol-rich plant residues and
their effect on nitrification of applied ammonium
sulfate. Soil Biology and Biochemistry 17 (4), 547-
551.
Singh, R.P., Van den Beldt, Hocking, R., and Korwar, G.R.,
1986. Alley croppping in semi-arid regions of India,
pp. 109-122. In: Kang, B.T. and Reynolds, L. (eds.),
Alley farming in the humid and sub-humid tropics.
Proceeding of an international workshop held at the
International Institute of Tropical Agriculture (IITA),
Ibadan, Nigeria, March, 10-14
Singh, R.P., Ong, C.K., and Saharan, N., 1989. Above and
below-ground interactions in alley cropping in semi-
arid India. Agroforestry Systems 9 (3), 259-274.
Slatyer, R.O., 1967. Plant water relations. Academic
Press, New York.
Sowden, F.J., 1968. Soil organic matter and crop yields as
influenced by the frequency of summer fallowing.
Canadian Journal of Soil Science 48, 331-339.
Spedding, C.R.W., and Brockington, N.R., 1978.
Experimentation in agricultural systems. Agricultural
Systems 1, 47-56.
Stark, N.M., and Jordan, C.F., 1978. Nutrient retention by
the root mat of an Amazonian vegetation. Tropical
Ecology 59, 434-437.
Steiner, J.L., Day, J.C., Papendick, R.I., Meyer, R.E., and
Bertrand, A.R., 1988. Improving and sustaining
productivity in dryland regions of developing
countries. Advances in Soil Science 8, 79-122.
Stevenson, F.J., 1986. Cycles of soil carbon, nitrogen,
phosphorus, sulfur, micronutrients. John Wiley, New
York.


Bulk density (g cm'
170
2.0
Leucaena
Species
I I Near
153 Away
Cassia
Figure 5-4. Bulk density of top 5 cm soil profile
near (0.45 m) alley cropped hedgerows of Leucaena
leucocephala and Cassia siamea, 1991 short-rains.


81
Table 3-2. Chemical characteristics of the dry mulch of
alley cropped hedgerows of Leucaena leucocephala
and Cassia siamea.
N
(%)
P
(%)
K
(%)
Ca
(%)
Mg
(%)
C:N
ratio
Lignin
(%)
Poly
phenol
(%)
Leucaena
3.5
0.2
1.9
0.7
0.3
14
8.1
4.7
Cassia
2.9
0.2
1.8
0.7
0.2
17
8.4
3.7


32
In the first year, there was no significant difference
between mulch yields from the two species, but after the
second year, leucaena yields were significantly higher than
those of cassia (Fig. 2-2). Also, a significant difference
was observed in the response of the two species to rainfall.
While changes in rainfall had a significant effect on
the mulch yield of leucaena, no effects were observed for
cassia. Mulch yield of leucaena peaked in 1990 when the
annual rainfall was highest (964 mm) during the study
period. In 1991 when rainfall dropped to only 22% of the
1990 total (Fig. 2-1), mulch yield of leucaena dropped by
half. For cassia, there was no significant response (Fig.
2-2). Cassia mulch yield tended to decline over time,
although the differences in yield between years were not
significant. Although the effect of rainfall on mulch yield
was more pronounced than the effect of age of the hedgerows,
given the same rainfall, mulch yield of leucaena did
increase with age. For instance, with similar seasonal
rainfalls total of 330 mm, mulch yields of leucaena
hedgerows were 40% more in the fifth season than in the
second season. Similar trends of increase in cassia mulch
with age were not detectable.
Where hedgerows were planted as sole stands/blocks
outside the crop fields, differences in mulch yields per
hectare among the three tree:crop ratios were not


198
alley cropping trees on the yields of crops at any given
site are likely to be influenced by the tree species used.
For instance, Sang and Hoekstra (1987) reported improved
crop yields under cassia in semiarid Kenya; in contrast,
Singh et al. (1989) reported significant yield losses under
leucaena in semiarid India.
In a comprehensive review of agroforestry, Nair (1990)
concluded that alley cropping as an agroforestry practice
has some advantages, but some inputs are required. While
there are certainly many biophysical issues of the
technology in both semiarid and humid tropics still at
stake, perhaps most are on aspects related to the socio
economics of the technology. Understanding of the socio
economics is the key to the adoption of a technology
anywhere. Without this understanding, the technology may
never leave the researcher's desk, as is typical of many on-
station studies (Hildebrand, 1982). For instance, alley
cropping research has been going on at the International
Institute of Tropical Agriculture (IITA), Ibadan, Nigeria,
for at least 10 years (Kang et al., 1990). Yet, outside
IITA research stations, there is little evidence of adoption
of the technology by farmers for purposes of mulch
production on crop fields (Palada, 1989).
It is hypothesized the more simple the technology, the
more rapidly farmers can become proficient with it and adopt
it (Wake et al., 1988; Hildebrand, 1990). However,
agroforestry technologies (e.g., alley cropping) are by


4
and poles) and their protective or service functions in
relation, for example, to soil conservation and soil
fertility improvement. Leucaena leucocephala Lam. De Wit
(Leguminosae, Mimosoideae) and Cassia siamea Lam.
(Leguminosae, Caesalpinoideae) are two MPTS widely used in
alley cropping studies in both the humid lowlands and the
semiarid tropics (Nair, 1987; Kang et al., 1989; Singh et
al., 1989). This is probably so because both species are
legumes and it is assumed that soil fertility improvement is
among their potential benefits. (In subsequent references
to the two species, the generic neones leucaena and cassia
will be used as common names.)
Leucaena had long been used for shade and soil
improvement in tree and coffee plantations in Java,
Indonesia, as early as 1900 (Dijkman, 1950) Properties
that make leucaena suitable for soil fertility improvement
are summarized in Young (1989). They include: high biomass
production (10-25 t ha'1 yr'1, DM), high nitrogen fixation
(100-500 kg N ha'1 yr'1), high levels of nitrogen in leaves
(2.5%-4.0%), and, thus, a high rate of nitrogen return in
litter and prunings. The main soil limitation is a
reduction in its growth on acid soils (Brewbaker, 1987).
Also, attack by the psyllid, Heteropsylla cubana, is
currently serious in some regions. Like leucaena, cassia
has the capacity to grow on poor soils and has been used in
alley cropping trials (Yamoah et al., 1986a), although the


209
Strategies of Farmers to Improve
Fertility and Productivity
Nine practices were identified as the main strategies.
They will be discussed individually, but not in order of
importance.
Use of livestock manure
This was the major source of nutrient input for the
crops on most farms. All farmers interviewed used livestock
manure on their crop fields. Manure was applied on the crop
rows and not broadcasted on the whole field. Row
application of manure economized use of this scarce
resource. The main source of manure was their own farm.
Between the fodder and mulch-interest groups, the percentage
of each group that bought manure was similar, approximately
%
39%. ;
Manure from cattle owned was scarce, given that each
farm had, on average, only two cattle. Occasionally,
farmers with and without livestock purchased manure.
Thirty-five percent of the farmers interviewed did not have
cows and 37% bought manure at one time or another. Often
manure was bought from a few farmers with large broiler
houses within the area of the present study or from cattle
ranches 40-50 km away. Lack of feed and loss of livestock
during previous droughts were the reasons given for the low
number or absence of cattle on farms. Crop residues were
the major source of feed and farmers (92% of those


I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality, as
a dissertation for the degree
P.E. Hildebrand
Professor of Food and Resource
Economics
This dissertation was submitted to the Graduate
Faculty of the School of Forest Resources and Conservation
in the College of Agriculture and to the Graduate School and
was accepted as partial fulfillment of the requirements for
the degree of Doctor of Philosophy.
May 1993
Conservation
Dean, Graduate School


SOIL FERTILITY AND PRODUCTIVITY ASPECTS OF ALLEY
CROPPING LBUCAENA LEUCOCEPHALA AND CASSIA
SIAMEA UNDER SEMIARID CONDITIONS
AT MACHAROS, KENYA
BY
BASHIR JAMA ADAN
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE
UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE
REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
1993


203
Akamba language. The area is predominantly populated by the
Akamba people. Interpretation was not always necessary
because a large number of the farmers spoke either Swahili
and/or English which the author also speaks. On average,
five farmers were interviewed in a day.
At the conclusion of the survey, of the 60 farmers, a
second stratification or regrouping of the farms was carried
out. The farmers were grouped according to their identified
interests in the use of the pruning from the alley cropped
hedges. Basically, there were two distinct use-interests:
fodder or mulch (soil fertility). It is recognized that the
fodder-interest group could also have mulch as their
secondary interest in the use of the pruning from the
hedgerows and vice versa for the mulch group. Hereafter,
the two groups will be referred to as fodder-interest and
mulch-interest groups. On the basis of these two major use-
interests of the alley cropping technology, three farmers of
each interest group were randomly selected from the low,
middle and upper slopes of each of the three sublocations.
Thus, a subsample of 18 farmers out of the initial sample of
60 farmers was generated. The purpose of this second
stratification or subgrouping was two-fold: (a) to get more
in-depth information of the farming practices and the other
objectives of the study, and (b) to obtain a more informed
evaluation of the technology from the farmers. This sub
group of 18 farmers were taken to ICRAF Field Station over a


222
Trees and shrubs also provided fodder, particularly during
the dry season. For example, the dry leaves of Lantana
camara and Tithonia diversifolia hedges were used as fodder.
Some farmers with leucaena trees for shade near the
homestead occasionally lopped the trees for fodder.
If available, fruit trees were number one choice of
farmers in favor of all other trees. Fruit trees provide
food and cash to the family. Also, some fruit trees such as
mangoes were lopped for firewood. Besides mangoes which
were scattered on the cropland, nonfruit trees (e.g., Acacia
spp. ) left deliberately on cropped land provided shade
during field activities (plowing, planting, and harvesting).
Constraints to tree planting
Ability to establish easily from seed or seedling,
particularly under arid and competitive condition from weeds
and other plants was an important consideration of farmers
in the choice of the tree to plant. Equally important was
the ability of trees to withstand termite attack. A
significant number (86%) of the farmers interviewed
mentioned termites as problems in establishing trees.
Wood ash was predominantly the traditional termite
control measure used. Some farmers said wood ash controls
termites but many also said it does work or works only
temporarily. The seriousness of the termite problem was
reflected in the choice of species planted or allowed to


Mulch yield (t ha _1)
33
3.5
3.0
2.5
a
1989
1990
Year
Leucaena
Intercropping
Leucaena
Sole
Cassia
Intercropping
Cassia
Sole
Figure 2-2. Mulch yield of Leucaena leucocephala and
Cassia siamea in alley cropping and block/sole planting
systems.


142
residues have been suggested as management practices that
may enhance synchrony between the peaks of supply of
nutrients from decomposing organic residues and demand by
the crop (Read et al., 1985; Yamoah et al., 1986a; Woomer
and Ingram, 1990; Palm and Sanchez, 1991). Implicit in
these suggestions is that prospects of attaining synchrony
are greatest when there is diversity of mulch resources and
when management options are relatively flexible.
Application of the mulch four weeks after sowing the
crop rather than at sowing may under the conditions of this
study have achieved synchrony between the peaks of N release
from the mulch and uptake by the crop. Under the conditions
of the present study, the second pruning of the hedgerows
was generally carried out 4-6 weeks after sowing the crop.
Therefore some mulch was available for application four
weeks after sowing the crop. The timing of pruning would,
however, depend on rainfall conditions of the season,
hedgerow species and the rate of regrowth after pruning. At
the site of the present study, leucaena could yield
substantial amounts of mulch in the second pruning but
cassia could not. However, mulch yield of both species was
highest in the first pruning of the hedgerows at the time of
sowing the crops. While storage of mulch from the first
pruning and its application at later date could be a
suggestion, it may not be a practical proposition to
resource-poor farmers.


186
as in the pot studies, significant responses mainly to
irrigation, and only little to N-fertilization. For
instance, plots with irrigation-only (no fertilizer) had
yields of maize not significantly different from those with
a combination of irrigation and fertilizer. Also yield of
plots without irrigation but with fertilizer-only and those
without fertilizer (i.e., the absolute control) were not
significantly different. Thus, soil moisture rather was the
more serious limiting factor for maize production than soil
fertility.
The hedgerows could have differentially influenced
microclimatological conditions near and away from them. For
instance, atmospheric vapor pressure deficit (VPD) could
have been reduced more near the hedgerows than away from
them. However, this difference could be small as Monteith
(1991) observed from alley cropping studies from semiarid
India. Since dry matter production: water use ratio is
inversely proportional to VPD (Monteith, 1990), it is
possible that reduced VPD near the hedgerows, resulting from
shelter and lower soil temperatures observed near the
hedgerows, could have contributed to the higher yield of
maize near the hedgerows than away from them.
The higher yield of maize near the hedgerows would
also suggest that total amount of nutrients taken up (i.e,
concentration x biomass yield) was higher near the hedgerows


146
concluded that application of the mulch four weeks after the
onset of the rains may help achieve synchrony.


57
different. The effects of these lower rates were also not
significantly different from the effects of the fertilizer
only.
The effects of irrigation depended on the harvest. In
harvests 1 and 4 (i.e., 4 and 16 weeks after mulch
application to the pots, respectively), differences in yield
between the two levels of irrigation were not significant.
However, in harvests 2 and 3, the differences in maize
biomass yield were significant.
Discussion
Mulch Yield of the Hedgerows
On average, leucaena produced 46% more mulch than
cassia. The higher yield of leucaena was observable in both
systems of hedge planting. Similar differences in the mulch
yields of the species had been noted earlier by Nair (1987)
working at the same site. The differences in biomass
production between the two species could be ascribed to
inherent growth differences as well as their site
utilization efficiencies. In the lowland humid tropics,
differences in the biomass yields of leucaena and cassia
appear to be even more pronounced than in the semiarid
tropics (Kang et al., 1981; Kang et al., 1985). Averaged
over the three years of study, mulch yield of leucaena was
approximately 4 t ha*1 yr'1 (dry matter) Similar yields of
leucaena have also been reported by Rao et al. (1990) from


213
mulch-interest groups were similar (65%). The purchase of
seeds was not influenced significantly by farm size.
Use of chemical fertilizers
From the sample survey of 60 farmers, 65% used
chemical fertilizers on their crops at the time of the
survey or in prior seasons. Farmers with coffee crops
responded more positively to use of fertilizers than farmers
without coffee. Between the two groups, 76% of the fodder-
interest group and 55% of the mulch-interest group used
fertilizers. This difference in fertilizer use of the two
groups was significant. However, only 3% of the farmers
interviewed mentioned chemical fertilizers as their most
expensive farm input, suggesting that use of chemical
fertilizers was not a major practice. Also, chemical
fertilizers, unlike seeds, were not regularly purchased.
For farmers with coffee crops, fertilizer was obtained at a
subsidized cost from the coffee factories.
Use of ox-plow
A significant number (67%) of the farmers interviewed
used ox-plows for land preparation and/or for weeding. At
times, those who did not own the oxen owned only a plow, and
vice versa. This often necessitated an exchange of the ox-
plow parts between farmers. Possession of both the oxen and
the plow were expensive for individual farmers. Some


225
Table 6-7. Reasons given by farmers without nurseries on
their farms, Machakos, Kenya, August 1991.
Reasons
Distribution of respondents
(%)
Water
26.7
Seeds
13.3
Not decided
11.7
Lack of knowledge
3.3
Planning
1.7
Pests (e.g., chickens)
1.7
Termites
5.0
Labor
6.7
Total
70.1


76
matter), soil organic matter (SOM) and nutrients were
maintained at higher levels than where the prunings were
removed. Similar observations are made by other workers also
from the lowland humid tropics (Atta-Krah et al. 1985; Yamoah
et al. 1986c). However, in the SATs, ecological conditions do
not allow the high rate of mulch production found in the humid
tropics. Hence, significant gains in SOM from the addition of
mulch may be limited. Based on the effects of mulch per se
there might be no conceivable advantage of alley cropping
systems on SOM in the SATs. However, turnover rates of high
root mass often associated with agroecosystems with trees
(Ewel et al., 1982), such as alley cropping, may contribute
significantly to organic matter and nutrient contents of the
soil. Moreover, where leaching losses of soil- and applied-
nutrients may be a problem, the roots could minimize these
through a combination of processes including sorption and
uptake from deeper layers (Stark and Jordan, 1978; Jaiyebo and
Moore, 1964). These processes may result in the concentration
of nutrients in the mulch of the hedgerows which could, in
turn, improve crop nutrition (Fernandez, 1990). In some
soils, there are indications that the roots may even recycle
nutrients from the subsoil which could in turn result in top
soil improvement of pH and exchangeable bases (Lai, 1989). In
addition, due to their longevity and to their protective
benefits, intercropped hedgerows may minimize runoff and soil
erosion loss (Lai, 1989; Rao et al., 1991).


22
In cropping seasons 1-4, yields of all maize rows in a
plot were composited at harvest. However, in seasons 5, 6
and 7, maize yields of each row in a plot was separately
harvested and recorded. Also, in the last three cropping
seasons, sub-samples of the grain, cobs and stover of some
selected treatments, as explained later, were analyzed for
nitrogen and phosphorus concentrations. Nitrogen was
determined by the standard micro-kjeldahl method and P by UV
visible spectrophotometry. In all seasons, two rows of
maize on all four sides of the plot were left as guard rows.
Pot Study
A pot study was initiated in January 1992 to determine
the amounts of mulch that would produce significant effects
on maize yields under a given rainfall condition. The
treatments included a factorial combination of 4 rates of
leucaena and cassia mulch and 2 levels of irrigation. The
four mulch rates were applications equivalent to 0, 2, 4 and
8 t ha*1 season*1, dry matter. The two irrigation levels
were the equivalents of 200 and 400 mm received in the last
two cropping seasons of the study. The pots were arranged
in a randomized block design with 3 replications in a
transparent polythene-roof house. In each block, pots with
one level of fertilizer at a rate equivalent to 40 N and 17
P kg ha'1 season*1 were also included.


Table 4-1.
Chemical
Leucaena
and physical properties
leucocephala and Cassia
of the
si antea
mulch of alley
cropped hedgerows of
Leucaena
Cassia
Property
Leaves
Twigs
Average
Leaves
Twigs
Average
Nitrogen (%)
4.4
2.6
3.5
4.0
1.7
2.9
Polyphenols
(%)
6.9
2.4
4.7
5.5
1.9
3.7
Lignin (%)
5.4
10.8
8.1
6.8
9.9
8.4
Ash (%)
-
-
7.1
-
-
5.8
C:N ratio
11
19
14
13
29
17
NOTE: It was assumed that 50% of the mulch material was carbon.


80
37,000 plants/ha. Moderate levels of fertilizer (40 and 18
kg of N and P/ha) were added to the maize in the hedgerow
intercropped and the block planted systems each season.
Weeds were hand removed at least twice in each season. At
harvest, maize grain, stalks and cobs were removed from the
field as was the practice of the farmers in the area.
The tree hedgerows were pruned back to 50 cm above
ground with 2-3 prunings a season. The prunings obtained
were chopped into small pieces and applied as mulch to the
plots. The chemical characteristics of the mulch are shown
in Table 3-2. If tillage and/or weeding schedules shortly
preceded periods of mulch application, the mulch generally
became incorporated into the soil; otherwise, it remained on
the soil surface.
Soil sampling was always carried to 20 cm depths. A
composite sample of 10 auger cores was taken for each
treatment. The soil was air-dried, crushed to pass a 40-
mesh sieve, and mixed. Samples were analyzed at the
Analytical Laboratory of the Katumani Research Station,
Machakos, Kenya by the following methods: total N by the
micro-Kjeldahl method (Jackson, 1958), organic carbon by the
Walkley-Black procedure (Jackson, 1958), P by the Mehlich-1
double acid procedure and cation exchange capacity (CEC) by
Atomic Absorption Spectrophotometry methods. Soil pH was
determined using a glass electrode pH meter (model Bibby),
at soil to water ratio of 1:2.5.


Page
5-7. Soil water content near (0.45 m) alley cropped
hedgerows of Leucaena leucocephala and Cassia
siamea determined at the end of 1991
short-rains 175
5-8. Difference in gravimetric water content
(%, w/w) between soil sampled at 0-20 cm
depth near (0.45 m) and away (3.5 m) from
alley cropped hedgerows of alley cropped
hedgerows of Leucaena leucocephala and
Cassia siamea in 1991 177
5-9. Monthly average readings of tensiometers placed
in the soil at 15 cm depths near (0.45 m) and
away (3.5 m) from alley cropped hedgerows
of Leucaena leucocephala and Cassia
siamea in the long- and short-rain
seasons of 1991 . 178
5-10. Difference in mean monthly temperature of
top soil (10 cm) measured near (0.45 m) and
away (3.5 m) from alley cropped hedgerows of
Leucaena leucocephala and in the long- and
short-rain seasons of 1991 179
5-11. Count of fauna in 25 x 25 x 25 cm top soil
monoliths taken from near (0.45 m) and away
(3.5 m) from alley cropped hedgerows of Leucaena
leucocephala and Cassia siamea at the end of
1991 short-rain season 181
5-12. Effects of irrigation and fertilization, in
a combination or alone, on maize grain yield
in the field (no alley cropping with hedgerows
of Leucaena leucocephala and Cassia siamea),
1991 182
xiii


252
Gholz, H.L., Perry, C.S., Cropper, W.P. Jr., and Hendry,
L.C., 1985. Litter fall, decomposition and nitrogen
and phosphorus dynamics in a chronosequence of slash
pine (Pinus elliottii) plantations. Forest Science 31,
463-478.
Gichuru, M.P., and Kang, B.T., 1989. Calliandra calothyrsus
(Meissn.) in an alley cropping system with sequentially
cropped maize and cowpea in southwestern Nigeria.
Agroforestry Systems 9 (3), 191-204.
Guevarra, A.B., 1976. Management of Leucaena leucocephala
for maximum yield and nitrogen contribution to
intercropped corn. Ph.D. dissertation, University of
Hawaii, Honolulu.
Gutteridge, R.C., 1992. Evaluation of the leaf of a range
of tree legumes as a source of nitrogen for crop
growth. Experimental Agriculture 28, 195-202.
Hanway, J.J., 1962. Corn growth and composition in relation
to soil fertility. II. Uptake of N, P and K and their
distribution in different plant parts during the
growing season. Agronomy Journal 54, 217-22.
Harsh, C.T., and Mbatha, B., 1978. The economics of small
holder agriculture in two areas of the semi-arid lands
of Kenya. Report No. 3. Kenya marginal semi-arid
lands pre-investment study. Government of Kenya/USAID,
Nairobi, Kenya.
Hauck R.D., and Bremner, J.M., 1976. Use of tracers for
soil for soil nitrogen research. Advances in Agronomy
28, 19-266
Haynes, R.J., 1986. The decomposition process:
mineralization, immobilization, humus formation and
degradation, pp. 52-176. In: Haynes, R.J. (ed.),
Mineral nitrogen in the plant-soil system, Academic
Press, Orlando, FL.
Herman, W.A., McGill, W.B., and Dormarr, J.F., 1977.
Effects of initial chemical composition on
decomposition of roots of three grass species.
Canadian Journal of Soil Science 57, 205-215.
Hildebrand, P.E., 1981. Combining disciplines in rapid
appraisal: the sondeo approach. Agricultural
Administration 8, 423-32.


9
as well as the management regime of alley cropping to
improve the fertility and productivity of soils in the
semiarid tropics. These technical and social questions were
the prime motivations in undertaking this study. This study
had with four main objectives:
1. To determine if alley cropping with leguminous
MPTS substantially improves soil fertility and
productivity in the semiarid conditions of
Machakos, Kenya.
2. To determine or separate effects due to mulch per
se from those due to the in situ presence of the
hedges within the crop field.
3. To assess the patterns of tree-crop interaction
and to determine factors that may be responsible
for the observed interactions.
4. To determine factors that may significantly
influence adoption of the technology by farmers.
Plant productivity aspects of the crop and hedgerows
of leucaena and cassia are presented in Chapter 2. Chapter
3 addresses the effect of the alley cropping and mulch
quantity and quality on soil chemical, physical and
biological properties; trends over time as well as within a
cropping season are discussed. Patterns of N availability
determined by means of frequent soil sampling and/or in situ
mineralization using ion-exchange resins are also discussed
in Chapter 3.


Table 4-4. Effects of soil incorporated Leucaena leucocephala and Cassia siamea mulch
on the biomass yield, N-concentration, N-yield and N recovery of maize
planted in pots, 1991.
Time after
Maize
N concen-
N-yield
Apparent
mulch*/treatment
biomass
tration
of the
N
application
Treatment/
yield
of biomass
biomass
recovery
(weeks)**
mulch
(g/pot)
(%)
(g/pot)
(%)
4
Leucaena
4.2
2.9a
0.12a
11.2a
Cassia
2.2
3.1a
0.07b
-5.3C
Fertilizer
2.9
2.9a
0.09b
3.6b
Control
2.7
2.8a
0.08b
-
8
Leucaena
5.5
3.0a
0.17a
4.6b
Cassia
5.9
2.6b,c
0.15a
0.6
Fertilizer
5.5
3.0a,c
0.17a
7.1a
Control
5.9
2.6b,c
0.15a
-
12
Leucaena
5.4
2.6a
0.14a
9.8a
Cassia
5.0
2.7a
0.14a
11.4a
Fertilizer
5.4
2.3b
0.13a
9.2a
Control
4.6
2.2b
0.10a
-
16
Leucaena
4.0
2.3a
0.13a
8.7a
Cassia
2.1
2.0a
0.12a
-2.8C
Fertilizer
3.2
2.1a
0.06b
2.5b
Control
2.5
2.2a
0.06b

a'b'cWithin a week and a column, values with different letters are not significantly
different (p = 0.05).
*Mulch applied was the equivalent of 2 t/ha, dry weight.
**Maize was sown and harvested after 4 weeks repeatedly for a total duration of 16 weeks.
128


83
soil properties monitored under the various treatments and
over time. Because the initial (1987) soil levels of
percent clay and P differed significantly among plots, these
two parameters were used as covariates in the analysis of
the final (1991) soil P levels. Also, for the KC1
extractable N, analyses were performed on logarithmically
transformed data, because frequencies from a set of 5 resin
bags were not normally distributed. When significant
effects of treatments were detected, means were separated by
pairwise contrasts of the LSM using the GLM procedure in SAS
(SAS, 1992). Differences between treatments were declared
significant at p 0.05.
Results
Soil Chemical Properties
With the exception of P, there were no significant
differences among treatments in the levels of the chemical
characteristics measured (Table 3-3). Soil pH, organic
carbon, total N and CEC in the different treatments were
similar. With respect to P, however, there were significant
differences among treatments. In general, plots with cassia
mulch tended to have a higher level of available soil P. In
particular, the levels of P from the plots with application
of cassia mulch-only were significantly higher than the
levels of all the other treatments (Table 3-4). The


I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality, as
a dissertation for the degree of Doctor of Philosophy.
Q )
P.K.R. Nair, Chair
Professor of Forest Resources
and Conservation
I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality, as
a dissertation for the degree of Doctor of Philosophy.
H.L. Gholz /
Professor of Forest Resources
and Conservation
I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality, as
a dissertation for the degree of Doctor of Philosophy.
C.P.P. Reid
Professor of Forest Resources
and Conservation
I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality, as
a dissertation for the degree of Doctor of Philosophy.
i&L. Pd'penoe /
Professor of Soil Science


175
v
Water Content (%)
Figure 5-7. Soil water near (0.45 m) alley cropped
hedgerows of Leucaena leucocephala and Cassia siamea
determined at the end of 1991 short-rains.


Difference in maize yield between alley-cropped plots
and the plots receiving mulch from the block plantings
showed that growing the hedges in situ had a positive effect
on maize yield. Yield advantage of alley cropping over the
separate crop and hedge planting systems, expressed as land
equivalent ratios, were 36% and 12% for cassia and leucaena,
respectively. Leucaena produced more biomass (4 t ha"1 yr"1)
than cassia (2 ha'1 yr"1) ; also leucaena had higher fine-root
density than cassia in the top soil (0-40 cm). Therefore,
leucaena was more competitive than cassia, and over the
years, maize yield declined by 12% under leucaena whereas it
increased by 8% under cassia.
For both species, maize nearer the hedgerows yielded
more than those away from them. Soil water content was also
higher near the hedgerows. Chemical and physical properties
of the top soil (0-20 cm) did not differ significantly
between treatments (with or without mulch or alley cropping)
or over time for a given treatment. The quantities of mulch
produced by the hedgerows were generally low, and they
decomposed quickly (half life of 'four weeks); these could be
the main reasons for the lack of significant changes in soil
properties.
Fanners were generally more interested in the fodder
than the mulch aspects of alley cropping. Because of the
initial high investments of inputs such as labor and
seedlings that required to establish the hedgerows, alley
cropping is more likely to be adopted by relatively
xv


CHAPTER 3
SOIL FERTILITY ASPECTS OF ALLEY CROPPING
LEUCAENA LEUCOCEPHALA AND CASSIA SIAMEA IN
SEMIARID TROPICS AT MACHAROS, KENYA
Introduction
Agroforestry practices are suggested as low-input
systems with potentials to meet the critical needs to improve
or sustain food production in the tropics (Swift and Sanchez,
1984). One practice of agroforestry which has been the
subject of much interest and study is alley cropping, also
called hedgerow intercropping (Kang et al., 1990). It
consists of growing food crops in alleys between planted
hedgerows of multipurpose trees and shrubs. The hedgerows are
periodically cut back at the beginning of as well as during
cropping to prevent shading and to provide mulch to the
associated crop. ^
One of the most important premises of alley cropping is
that the addition of organic mulch, especially if it is rich
in nutrients, has favorable effects on soil physical and
chemical properties, and hence on productivity of crops.
However, the ability of the mulch to provide nutrients and to
maintain or improve soil properties would depend, in addition
to its quality and time of application, on the quantity. For
instance, Kang et al. (1990) observed from six-year-long
studies at the humid lowland of West Africa that with
continuous addition of leucaena pruning (8-10 t ha'1 yr*1, dry
75


258
Murwira, H.K., Kirchmann, H., and Swift, M.J., 1990. The
effects of moisture on the decomposition of cattle
manure. Plant and Soil 122, 197-199.
Mwangi, P., 1989. Soil plant nutrient relationship as
affected by Leucaena leucocephala, Cassia siamea and
Terminalia brownii mulch. M.S. Thesis, University of
Nairobi, Kenya.
Myers, L.R., 1982. A sociological approach to farming
systems in Kenya. Cornell International Agriculture,
Mimeo, Ithaca, N.Y.
Nadar, H.M., and Faught, W.A., 1984. Maize yield response
to different levels of nitrogen and phosphorus
fertilizer application. East Africa Agricultural and
Forestry Journal 44, 147-156.
Nair, P.K.R., 1984. Soil productivity aspects of
agroforestry. ICRAF. Nairobi, Kenya.
Nair, P.K.R., 1987. ICRAF Field Station at Machakos. A
demonstration and training site for agroforestry
technologies. Agroforestry Systems 5, 383-393.
Nair, P.K.R., 1989. The role of trees in soil productivity
and protection, pp. 567-589. In: P.K.R. Nair (ed.),
Agroforestry systems in the tropics. Kluwer Academic
Publishers, Dordrecht, Netherlands.
Nair, P.K.R., 1990. The prospects of agroforestry in the
tropics. World Bank Technical Paper no. 131.
NAS, 1984. Leucaena: promising forage and tree crop for
the tropics. Second edition. National Academy of
Sciences, Washington, DC., 100 pp.
Nelson, D.W., 1982. Gaseous losses of nitrogen other than
through denitrification, pp. 327-363. In: Stevenson,
F.J. (ed.), Nitrogen in agricultural soils. American
Society of Agronomy, Madison, Wisconsin.
Nyamai, D.O., 1992. Investigations on decomposition of
foliage of woody species using a perfusion method.
Plant and Soil 139, 239-245.
Nye, P.H., and Greenland, D.J., 1960. The soil under
shifting cultivation. Technical Communication 51.
Harpenden, UK. Commonwealth Bureau of Soils, 144 pp.


Biomass yield of maize (g pot'1)
56
Figure 2-10. Effects of three mulch rates of
Leucaena leucocephala and Cassia siamea on the biomass yield
of maize grown in pots, 1991.


232
could have seen the technology as a means to increase
productivity on their small holdings. Planting trees for
fodder could reduce the coffee farmers' need for purchasing
livestock feed off the farm. Farmers with coffee bought
more feed for their livestock than those farmers without
coffee. By purchasing less from the outside, farmers with
coffee will have even more revenue to invest in other
activities (e.g., buying an ox-plow that is now probably
being rented, milk cows, etc.). The presence of animals on
the farm may in turn minimize the current purchase of
manure. The accrued savings could alternatively be used to
purchase manure, to increase food production, or to pay for
the education of the children. These projections are in
line with the view that one of the economic benefits of
agroforestry is to create capital stocks to meet
intermittent costs and unforeseen contingencies (Arnold,
1987) .
Cash generating activities are absolutely necessary
for farmers to buy farm inputs, food, and invest in cash
generating activities (e.g., fruit trees, vegetables, and
poultry keeping). Without inputs, production is low. The
low crop yields (0.3 t ha-1) even if increased ten-fold "by
some miraculous technology" are unlikely to sustain the food
needs of the large household members. From the on-station
studies, a significant response to mulch was obtained only
when it was applied at a rate greater than 4 t ha'1 season'1.