Citation
Sustaining crop phosphorus nutrition of highly leached oxisols of the Amazon basin of Brazil through use of organic amendments

Material Information

Title:
Sustaining crop phosphorus nutrition of highly leached oxisols of the Amazon basin of Brazil through use of organic amendments
Creator:
Singh, Braj K., 1955-
Publication Date:
Language:
English
Physical Description:
xiv, 150 leaves : ill. ; 29 cm.

Subjects

Subjects / Keywords:
Adsorption ( jstor )
Corn ( jstor )
Crops ( jstor )
Farmers ( jstor )
Incubation ( jstor )
Organic fertilizers ( jstor )
Phosphates ( jstor )
Phosphorus ( jstor )
Soil science ( jstor )
Soils ( jstor )
Dissertations, Academic -- Soil Science -- UF ( lcsh )
Soil Science thesis Ph. D ( lcsh )
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1990.
Bibliography:
Includes bibliographical references (leaves 138-149).
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Braj K. Singh.

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:
024294771 ( ALEPH )
23536005 ( OCLC )

Downloads

This item has the following downloads:


Full Text











SUSTAINING CROP PHOSPHORUS NUTRITION OF HIGHLY LEACHED OXISOLS OF THE AMAZON BASIN OF BRAZIL THROUGH USE OF
ORGANIC AMENDMENTS


















By
BRAJ K. SINGH



















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

1990 FFLORIDA LIBRARY































This work is dedicated to my parents: Shree Mrit B. Singh and Shrimati Naina Devi, and to my brothers: Sri Biswanath Singh and Nawal K. Singh.














ACKNOWLEDGMENTS


The author wishes to express his sincere appreciation to Dr. J. B. Sartain, the chairman of the supervisory committee, for his excellent guidance, assistance, and continued interest in this study. His patience, understanding, and personal friendship throughout the academic and research program has been most invaluable and made the author's stay in the United States a most rewarding and stimulating experience. The financial support provided by him to cover part of research expenses in Brazil is gratefully acknowledged.

Sincere appreciation is extended to Dr. P.E. Hildebrand for providing a farming systems assistantship and taking personal initiatives to work out a joint research venture with the TropSoils Collaborative Research Support Program between Cornell University and the Brazilian Agricultural Research Organization (EMBRAPA). His deep concerns, constructive criticisms, and moral support as the member of supervisory committee helped broaden the author's understanding of farming systems and contributed to the successful completion of this manuscript. The author also thanks the other members of the supervisory committee, Dr.



iii








E.A. Hanlon, Dr. D.A. Graetz, and Dr. K.L. Buhr for their suggestions, support, and editorial comments. A special note of gratitude is extended to Dr. C.K. Hiebsch for substituting Dr. K.L. Buhr as a member of supervisory committee and reviewing the manuscript.

A particular note of appreciation is extended to Dr. Hugh Popenoe, Director of International Programs, for providing primary funding for the author's Ph. D. program. The author is indebted to and gratefully acknowledges the financial and technical support provided by TropSoils and EMBRAPA. Particular gratitude is expressed to Dr. Walter Bowen for his invaluable guidance in design and execution of the research and manuscript review. Special thanks go to Mr. Manoel da Silva Cravo and his extraordinary group of enthusiastic technicians, Raimundo Vitoriano de Oliveira, Agilau de Araujo Rodrigues, Edilza da Silva Richa, Emanoel dos Santos Alencar, Teofanes Moreira de Souza Junior, and Onelia Maria Pereira de Almeida for conducting the routine analysis of plant and soil samples. The author is grateful to Mrs. Eda M. Souza, CEPA-AM for her support in putting the on-farm project together.

A word of gratitude is also extended to Dr. K.D. Sayre, Dr. W. H. Freeman, Dr. C.N. Hittle, Dr. Richard Harwood, and Dr. Douglas Beck for their help in getting the author into the graduate school. Thanks go to S.K. Patel for his help in




iv









representing the author at UF during the author's one year stay in Brazil.

The author is indebted to his beloved wife for her encouragement, support, and patience.













































v














TABLE OF CONTENTS



ACKNOWLEDGEMENTS . ... . . .... . . . . . . . . iii

LIST OF TABLES ...... . ......... .... vii

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

ABSTRACT ................... . . . . .iii

CHAPTER I INTRODUCTION . ............... 1

Statement of Problem . ............. . 1
Goals and Objectives . ............. . 4

CHAPTER II ORGANIC AMENDMENTS AND PHOSPHORUS ADSORPTION ISOTHERMS . ............... 5

Soil Constituents and Phosphorus Adsorption . . . 5 Materials and Methods . ............. 10
Results and Discussion . ............. 14
Conclusion . .................. . 35

CHAPTER III ORGANIC AMENDMENTS AND CROP PHOSPHORUS NUTRITION . . . . . . . . . . . . . . . . . . . . . . 37

Phosphorus Management Strategy . ......... 37 Materials and Methods . .......... . . . 45
Results and Discussion . ............. 52
Conclusion . .................. . 71

CHAPTER IV RECOMMENDATION DOMAINS AND MODIFIED STABILITY ANALYSIS . ................. 72

Introduction . ........... ...... . 72
Materials and Methods . ............. 79
Results and Discussion . ............. 86
Conclusion . .................. . 126

CHAPTER V SUMMARY AND CONCLUSIONS . . . . . . . . . . 129 APPENDIX ECONOMIC ANALYSIS . ........... . 134

REFERENCE LIST . .................. . 138

BIOGRAPHICAL SKETCH . ................. 150

vi















LIST OF TABLES


Table 2-1. Ionic species and concentration of nutrient
solution used with silica to simulate solution chemistry of a Xanthic Hapludox in a phosphorus
release study by organic amendments . ...... 13

Table 2-2: Chemical composition of organic amendments
used in the incubation study. .......... 16

Table 2-3: Selected physical and chemical properties of
Ap horizon of the Xanthic Hapludox used in
adsorption studies. ............... 17

Table 2-4. Release of 0.01 M CaC12 extractable
P (Ag g") following incubation of organic
amendments with soil and silica
as matrix substratum. .............. 23

Table 2-5. Langmuir parameters (k and b) for P
adsorption by soil incubated with different
organic amendments. ............... 29

Table 2-6. Langmuir parameters (k and b) based on net
release of P measured by sequential extraction of
simulated silica matrix with 0.01 M CaCl2. . . . 32

Table 2-7. Difference in the estimated values of
Langmuir parameters (b and k) based on the
estimation of preadsorbed P by simulated silica
matrix technique and sequential extraction. . . . 33

Table 3-1. Factorial arrangements of treatments for the
greenhouse study . ............... . 48

Table 3-2. Description of treatments tested in the
field . . . . . . . . . . . . . . . . . . . ... . . 50

Table 3-3. Selected chemical properties of the fine
fraction (<2 mm) of the Xanthic Hapludox. . . . . 55

Table 3-4. Selected physical properties of the Xanthic
Hapludox. ................... . 55



vii








Table 3-5. Selected chemical properties of the fine
fraction (<2 mm) of the Xanthic Hapludox. . . . . 56

Table 3-6. Analysis of variance for maize herbage
dry weight production per pot in
the glasshouse study. . . . .......... . 57

Table 3-7. Mean changes in soil pH (H20) following
application of selected organic amendments in the
glasshouse study . ................ 60

Table 3-8. Orthogonal contrasts of maize grain yield
under different treatments applied in a 30 cm wide
band at UEPAE research station, Amazonas, Brazil. 62

Table 4-1. Application of N, P, and K in different
treatments tested in on-farm experimentation for maize
and cowpea crops in the municipality of Rio Preto da
Eva, Amazonas, Brazil. . ............. . 82

Table 4-2. Characterization of experimental plots for
maize testings in the municipality of Rio Preto da
Eva, Amazonas, Brazil . . . . . . ............. . 89

Table 4-3. Characteristics of experimental plots for
cowpea trials. ........... . . ...... 90

Table 4-4. Relationship between soil characteristics
with year in crop production in different land
types. . .. ............... ... . 91

Table 4-5. Relationship between soil characteristics
measured after treatment application with grain
yield for maize and cowpea crops in the
municipality of Rio Preto da Eva, Amazonas,
Brazil ..... . ............ . . .... . 92

Table 4-6. Summary of ANOVA for multilocatonal maize
testing in the municipality of Rio preto da Eva,
Amazonas, Brazil. . . . . . . . . . . . . . . . . 97

Table 4-7. Duncan Multiple Range Test (DMRT) for maize
crop in the municipal of Rio preto da Eva,
Amazonas, Brazil. . . . . . . . . . . . . . . . . 97

Table 4-8. Combined Analysis of Variance for maize
grain yield in the municipal of Rio Preto da Eva,
Amazonas, Brazil. . . . . . . . . . . . . . . . . 98

Table 4-9. Environmental index (e) for maize production in
the Municipalilty of Rio Preto da Eva, Amazonas,
Brazil..................... .. 100

viii








Table 4-10. Summary of ANOVA for multilocatonal cowpea
testing. . .............. . . . . . . 110

Table 4-11. Duncan Multiple Range Test (DMRT) for
cowpea trials conducted in the municipal of Rio
Preto da Eva, Amazonas, Brazil. ........ . 110

Table 4-12. Combined Analysis of Variance for
cowpea experiments. .............. . 111

Table 4-13. Technology selection for a given land type
based on different evaluation criteria for maize
cultivation . .................. . 124

Table 4-14. Technology selection for a given land type
based on different evaluation criteria for cowpea
cultivation . .................. 124




































ix














LIST OF FIGURES




Figure 2-1. Acid-base potentiometric titration curves
for the Ap horizon of Xanthic Hapludox with
varying concentration of CaC12 ......... 18

Figure 2-2. Phosphorus adsorption isotherms following
35 d (a) and 150 d (b) of soil incubation with Mucuna aterrima (Mucuna), Chicken Manure (CM),
Aerobically Processed City Waste (PCW), and
Control. ........... . . . ........ . . 21

Figure 2-3. Phosphorus desorption isotherms following
35 d (a), and 150 d (b) of soil incubation with
Mucuna aterrima (Mucuna), Chicken Manure (CM),
Aerobically Processed City Waste (PCW), and
Control. . ... ...... . . . . . . . . 22

Figure 2-4. A schematic representation of molecular
structures of oxides of phosphorus (a), and oxides
of silicon (b). ................ . 26

Figure 2-5. Effect of incubation period and incubation
matrix on P release pattern from different organic
amendments in a laboratory study. . ....... . 27

Figure 2-6. Phosphorus adsorption isotherms for the
Xanthic Hapludox (0.01 M CaCl2) fitted to Langmuir
equation. The lines in the figure represents
fitted equation. ................ 28

Figure 2-7. A comparison of P adsorption by Xanthic
Hapludox based on the estimation of preadsorbed P by sequential extraction of soil or silica matrix
incubated with organic amendments. . ........ 34

Figure 3-1. Geographic location of Amazon basin in
Brazil (a), on-station and farming systems
research (FSR) sites (b), and effective rainfall
during the period of August, 1988 until August
1989 (c) at EMBRAPA station in Manaus, Brazil.. . 47



x








Figure 3-2. Effect of different rates of selected
organic amendments in combination with different
rates of inorganic P on maize dry matter
production at 65 d after planting in a glasshouse
study . . . . . . . . . . . . . . . . . . . . . . 58

Figure 3-3. Relationship between Mehlich-I extractable
soil P (a), and maize leaf tissue
P concentration (b) with maize dry matter yield. 59

Figure 3-4. Rate of inorganic P applied through triple
superphosphate and its effect on maize yield in a
maize-maize rotation. . ........... .. . 64

Figure 3-5. Effect of selected organic amendments,
applied in a quantity equivalent to provide 26.4 kg ha'I of P, on sustaining Mehlich-I extractable
soil P pool in a maize-maize rotation on a Xanthic
Hapludox............... . . . . . . . . 65

Figure 3-6. Effect of selected organic amendments,
applied in combination with inorganic phosphorus
source in a quantity equivalent to provide 8.8 kg
ha"1 of P, on sustaining Mehlich-I extractable
soil P pool in a maize-maize rotation on a Xanthic
Hapludox. . ................ ... . 67

Figure 3-7. Leaching of calcium from surface horizon of
highly leached Xanthic Hapludox as influenced by
different organic amendments. . ......... . 69

Figure 3-8. Relationship between P concentration and
concentration of Zn and Cu in maize leaf tissue.. 70

Figure 4-1. Effective rainfall and suggested shift in
the planting date for maize in Rio Preto da Eva,
Amazonas, Brazil. (a) common practice, (b)
suggested practice. . ......... . .... 83

Figure 4-2. Range of different soil characteristics by
recommendation domains for maize trials. . ..... . 94

Figure 4-3. Range of different soil characteristics by
recommendation domains for cowpea trials. ..... 95

Figure 4-4a. Response of different treatments to
environmental index for maize production, Rio Preto da Eva, Amazonas, Brazil. . .............. . 102

Figure 4-4b. Distribution of confidence intervals for
maize production in poor (e<1.95 mg ha'1), and
good (e>1.95 mg hal1) environments. . . . . . . . 103

xi









Figure 4-5. Relationship of net income/cash cost, net
income/total cost, and net income with
environmental index, in on-farm maize trials from
Rio Preto da Eva, Amazonas, Brazil. ....... 106

Figure 4-6. Distribution of confidence intervals for
net income/cash cost, net income/total cost, and
net income for different treatments used for maize
cultivation. . .......... . . . . . . . . 108

Figure 4-7a. Response of different treatments to
environmental index for cowpea production, Rio
Preto da Eva, Amazonas, Brazil. ......... 115

Figure 4-7b. Distribution of confidence intervals for
cowpea production in poor (e<1.32 mg ha'l), and
good (e>1.32 Mg ha-1) environments. . . ....... . 116

Figure 4-8. Relationship of net income, net income/cash
cost and net income/total cost with environmental index, in on-farm cowpea trials from Rio Preto da
Eva, Amazonas, Brazil. . ............. . 117

Figure 4-9. Distribution of confidence intervals for
net income/cash cost for selected treatments in
poor (e<1.32 Mg ha"') and good (e>1.32 Mg haI )
environments for cowpea cultivation. . ....... 119

Figure 4-10. Distribution of confidence intervals for
net income/total cost for selected treatments in
poor (e<1.32 Mg ha ') and good (e>1.32 Mg ha1')
environments for cowpea cultivation. . ....... 120

Figure 4-11. Distribution of confidence intervals for
net income for selected treatments in poor (e<1.32 Mg ha1) and good (e>1.32 Mg ha') environments for
cowpea cultivation. ............... 121

Figure 4-12. Recommendation domains, based on yield,
and their relationship to land types in the
municipality of Rio Preto da Eva,Amazonas,
Brazil.. ................... .. 125

Figure 4-13. Recommendation domains, based on net
return to cash cost, and their relationship
to land types in the municipality of Rio Preto
da Eva, Amazonas, Brazil.. ............ 126




xii














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

SUSTAINING CROP PHOSPHORUS NUTRITION OF HIGHLY LEACHED
OXISOLS OF THE AMAZON BASIN OF BRAZIL THROUGH USE OF ORGANIC AMENDMENTS

By

Braj K. Singh

August 1990


Chairperson: Dr. Jerry B. Sartain Major Department: Soil Science

Retention of phosphorus by iron and aluminum oxides

plays an important role in determining ultimate availability of P to plants in highly weathered soils of the tropics. Different management strategies have been proposed to overcome this problem. The overall objective of this research was to study the influence of organic amendments (plant origin, manure, and aerobically processed city waste (PCW)) in sustaining P nutrition of Oxisol and to test the performances of selected alternatives in a wide range of production environments.

There was a marked difference in P adsorption and

desorption by soil preincubated with organic amendments. The difference was attributed to incubation period and chemical composition of amendments. Correction for preadsorbed P by


xiii








soil based on simulated silica matrix (SSMT) resulted in a better estimation of the Langmuir adsorption maxima. This technique involved mixing of acid washed sand with a nutrient solution (without P) and inoculation of the sand with microbes and incubation with organic amendments.

In the glasshouse experiment the highest dry matter

(DM) production was obtained with the CM. A strong relationship was observed between DM production and MehlichI extractable P (r2 = 0.88).

In the on-station field experiments, CM applied plots produced more grain in first as well as second crops compared to the plots which had received Canavalia ensiformis, Mucuna aterrima, PCW, and TSP. All organic amendments improved the soil P reserve and reduced Ca leaching indicating that application of organic amendments could lead to sustained crop P nutrition.

In farmers' field studies, three amendments were tested in different land types with maize and cowpea crops. The same amount of P applied from different amendments had different effect on maize and cowpea production. However, the selection of a given technology for a given land type (environment) was dependent on farmers' goals. Based on the criteria of grain production CM was recommended for maize in all environments and for cowpea in poor environments.





xiv














CHAPTER I
INTRODUCTION



Statement of Problem



A leaching environment together with persistent high

rainfall and temperature are the defining conditions for the development of Oxisols, a soil order which occupies over 1.12 billion ha and ranks 5th in worldwide distribution (USDA, 1975). Their similarity as a group stems from the composition of the colloid fraction rich in iron and alumina minerals, and deficient in other nutrients. Sustained production on these soils can be achieved only with adequate application of lime, N, P, K, Mg, Zn, Cu, B, and Mo (Sanchez et al., 1982). Among other nutrients, P deficiency appears to be the most crucial and may show up as early as the second year of cultivation. Large amounts of applied P are required to attain levels of soil solution P which are adequate for high crop yields (Yost et al., 1979).

Rapidly increasing cost of P fertilizer, and limited P supply in the tropics, have led to extensive P management studies on these soils. Fertilizer application techniques (band, broadcast, and band+broadcast) have been widely



1








2

investigated. Immediate and long term effects of such applications have been studied by Kamprath (1967), Yost (1977), and Yost et al. (1981). These studies have suggested that a high initial P application rates can reduce the P fixation capacity (Barrow, 1974), increase the cation exchange capacity (Keng and Uehara, 1974), and increase soil pH.


Advanced Research with Organic Amendments

Along with fertilizer application techniques, attempts are underway to manipulate the mineral surface chemistry and inactivate the high reaction capacity of sesquioxides by application of organic amendments (Larsen et al., 1959; Nagarajah et al., 1970; Yuan, 1980). It is believed that organic amendments can form a coating on the mineral surface (Easterwood and Sartain, 1990), and complex Fe3 and Al3 (Larsen et al., 1959) which will reduce P fixation. However, the role of an organic amendment as a P source has not received much attention. Attempts have been made to understand the factors responsible for P mineralization from organic amendments. Amendment P content (Singh, and Jones, 1977) and decomposition characteristics (Sweeney and Graetz, 1988) are important factors in understanding the P release patterns and predicting residual effects.

Use of P adsorption isotherms in predicting P

requirements of crops and soils has received wide attention








3

(Jones and Benson, 1975; Solis and Torrent, 1989). However, the techniques used to construct P adsorption isotherms in the presence of amendments are similar to those used for soil without amendment. Generally a correction for preadsorbed P is introduced (Mokwunye, 1977; Solis and Torrent, 1989). An assumption is made that adsorption and desorption are equal, which is not true (Sample et al., 1980). Therefore, it leads to unreliable prediction of crop P requirement and residual effect of applied P. Organic and Inorganic P Interaction

A large amount of organic amendment will be required to provide the total P requirement for a crop. An efficient way to use a suboptimal dose of an organic amendment is band application with inorganic P. Such an application reduces P immobilization from P-poor amendments, and eliminates P deficiency in the early stages of crop growth. Much remains to be learned about the interaction of organic amendments with inorganic P. Most laboratory work has not been adequately tested under field conditions. Limited studies conducted in this area indicate that diammonium phosphate and monoammonium phosphate can solubilize organic matter (Bell and Black, 1970; Giordano et al., 1971). As the solubilized organic matter is carried to a new location in the soil, it may reprecipitate, covering soil mineral surfaces which otherwise could have participated in P retention reactions.








4



Transfer of Technology

It is widely believed that future efforts to increase food production must be directed towards the marginal lands of the developing world (Shaner et al., 1982). To achieve this goal, more research on appropriate technology is required with direct farmer involvement in problem identification, research priority identification, and technology evaluation (Harwood, 1979). Such an attempt will lead to location-specific technology tailored to fit farmers' circumstances, and accelerate the process of technology diffusion (Hildebrand, 1983).



Goals and Objectives

The overall goal of this research was to examine the role of organic amendments in sustaining P nutrition of a highly leached Oxisol. The specific objectives were: (i) to devise a technique to measure the P release patterns from a decomposing organic material, (ii) to examine the effect of a combined application of organic and inorganic P in a band on sustaining crop P nutrition, and (iii) to conduct farmers' field trials to validate on-station research results and delineate recommendation domains for selected treatments.














CHAPTER II
ORGANIC AMENDMENTS AND PHOSPHORUS ADSORPTION ISOTHERMS


Soil Constituents and Phosphorus Adsorption


It is widely believed that hydrous oxides of Fe and Al, and calcium carbonate play key roles in P retention. Although controversy exists among researchers about the mechanism of P retention by these compounds (Muljadi et al., 1966; Hingston et al., 1967; Rajan et al., 1974), the phasic nature of P adsorption has been well recognized by several researchers (Bache, 1964; Munns and Fox, 1976). The first phase of adsorption is due to a high energy chemisorption of small amounts of P on the soil surface. The second phase is comprised of a precipitation reaction followed by a low energy sorption of P onto the precipitate.


Phosphorus Adsorption Isotherms

Phosphorus adsorption isotherms have been used

extensively for describing the P adsorption characteristics of various soils and in estimating the P requirement of different crops (Fox and Kamprath, 1970; Jones and Benson, 1975; Mokwunye, 1977; Singh and Jones, 1977; Solis and Torrent, 1989). The isotherm technique involves



5








6

equilibration of a known amount of soil for a limited time in a KC1, NaCl or CaC12 solution containing various amounts of K2HPO4, NaH2PO4 or Ca(H2PO4)2 (Olsen and Watanabe, 1957; Syers et al., 1973; Singh and Jones, 1976). Phosphorus removed from solution is considered to be adsorbed. Plant P requirement is calculated based on the amount of P needed to bring the concentration of supernatant solution to a specified concentration (generally 0.2 Ag mL ).

Lancmuir euation. A number of researchers have

investigated soil P adsorption characteristics using the Langmuir equation (Olsen and Watanabe, 1957; Woodruff and Kamprath, 1965; Gunary, 1970; Borggaard, 1983). A frequently used linear form is:

(c/x/m) = (1/kb) + (c/b)

Where:

x/m = amount of P adsorbed per unit weight of soil

b = the adsorption maxima

c = the equilibrium P concentration, Ag mL1

k = a constant related to the bonding energy of the adsorbent for adsorbate.

For a given uniform population of sites, the value of adsorption maxima can also be evaluated by plotting x/m vs. x/m/c, obtained from a rearranged Langmuir equation commonly referred to as the 'Eadie-Hofstee' plot (Syres et al., 1973). The underlying assumptions in each case are that adsorption sites on the particle surface are uniform, and








7

the maximum adsorption possible corresponds to a complete monomolecular layer. Both of these postulates do not hold for a heterogenous medium like soil (Larsen, 1967).

Deviations from the conventional Langmuir relationship at high equilibrium P concentrations (above 15 Ag mL') have already been reported (Olsen and Watanabe, 1957; Hsu and Rennie, 1962) which have led to the development of an extended form of the Langmuir equation. Gunary (1970) included a square-root term in the Langmuir equation. Holford et al. (1974), Syers et al. (1973), and Holford (1983) used a two-surface Langmuir equation. The two surface equation has not, however, been universally accepted and Posner and Bowden (1980) have discussed the futility of attempting to split isotherms into two or more regions. Similarly, a curve-fit error in estimating the Langmuir adsorption maxima was described by Harter (1984). He claimed the test of linearity was inadequate because plotting concentration against itself reduces data variability and always provides a significant correlation coefficient. He suggested that a better test of the fit is to ascertain whether the adsorption isotherm has the shape of the equation model.

In spite of drawbacks, the Langmuir equation is used

widely to describe P adsorption by soil. The major advantage of the Langmuir equation is that it allows for the calculation of adsorption maxima along with a relative








8

bonding energy term for P sorption (Syere et al., 1973). The values of adsorption maxima and bonding energy can be related to various soil properties which will supply information about the nature of the reaction between soil and fertilizer P, and can aid in the prediction of plant available P (Olsen, 1953).


Correction for Initial Surface Phosphate

Olsen and Watanabe (1957) have demonstrated the effect of correction for initial surface phosphate on the constants derived from the Langmuir isotherm. For Pierry clay and Owyhee silt loam soils the correction increased the adsorption maxima and the bonding energy constant. Ideally, the adsorption would be determined in a system in which the surface is free of the adsorbate ion. Usually this restriction is not feasible. Therefore, a correction for initial P is made by adding the amount of surface P determined by a separate analysis. One of the commonly used methods to measure native adsorbed phosphate is by isotopic exchange (Olsen and Watanabe, 1957; Holford et al., 1974). This method involves shaking a soil sample in an electrolyte to which carrier-free 32P is added. An aliquot is taken after a specified time and 31P is determined. Reddy (1990) has suggested a least squares fit for the determination of native P adsorbed on the soil surface. He suggested the following calculations:








9

S = KC - S

Where:

SI = amount of added P sorbed, Ag g-1

Sp = Y axis intercept-representing native soil P in the adsorbed phase, Ag g

C = P in solution, Ag mL'

K = linear adsorption coefficient (estimated without considering native adsorbed P, Sp), mL g-1

Syers et al. (1973) reported that at lower

concentrations the Eadie-Hofstee plot was more useful for evaluating P sorption than the conventional Langmuir equation, because the plot expanded the low P concentration region.

The methods described above are based on the assumption that all surface-retained P is displacable either by isotopic dilution or sequential extraction techniques in the presence of a weak extracting salt solution. However, evidence suggested that with increasing contact period and temperature, P becames less ready to exchange with isotopically labeled P (Barrow and Shaw, 1975). In addition, adsorption and desorption isotherms are different, and the relationship between the quantity of adsorbed P and concentration is not totally reversible (Sample et al. 1980). Therefore, estimation of surface held native P based on the techniques described above might underestimate the true adsorption maxima and bonding energy values. Thus, any








10

prediction of how the soil P status will change upon cropping must be considered unreliable (Bowman and Olsen, 1985).

The purpose of this study was to examine the effect of different organic amendments on P adsorption and desorption isotherms, and to devise a technique to measure net P release from organic amendments during their decomposition. Such a technique should lead to a direct and accurate estimation of the quantity of P which originated from decomposing organic amendments and adsorbed by soil. This knowledge will aid in the accurate estimation of soil P adsorption maxima and bonding energy.


Materials and Methods

All amendments mucuna ([Mucuna aterrima (Piper and Tracy) Merr], kudzu [Pueraria phaseoloides (Roxb.) Benth], canavalia [Canavalia ensiformis (L.) DC], maize (Zea mays L.), peanut (Arachis hypogaea L.), tephrosia [Tephrosia candida (Roxb.) DC], cowpea (Vigna unguiculata), mixed gramineae (Grass), aerobically processed city waste (PCW), and chicken manure (CM)) were dried at 65 OC for 72 hrs for dry matter determination. Dried plant material was ground in a Wiley mill and passed through a 0.5 mm screen. A weighed (0.2 g) sample of the ground plant material was digested with H2SO4 and H202 for the determination of macro and micro elements. Total nitrogen was determined by the micro-








11

Kjeldahl procedure. Potassium, Ca, Mg, Fe, Zn, Cu, and Mn were determined with an atomic adsorption spectrophotometer, and P was determined colorimetrically by the Murphy and Riley (1962) procedure.



Incubation study

Soil matrix. Soil used for this study was obtained from the Ap horizon of a Xanthic Hapludox (clayey, kaolinitic, isohyperthermic) (EMBRAPA, 1979) which had PZNC (point of zero net charge) at pH of 4.2. Five grams of organic ground material were mixed with 95 g of air-dried soil and incubated in plastic containers for 35, 65, and 150 d at 30 � 30C. The soil in each container was mixed thoroughly every

6 d and was kept at 45 % moisture content. After each incubation period, 2-g duplicate soil samples were extracted with 0.01 M CaCl2 with a soil to solution ratio of 1:10 and

1 h shaking. The cycle was repeated three times. In calculating P desorbed, an allowance was made for the 2 mL of supernatant carried over from each cycle (Fox and Kamprath, 1970).

Adsorption was determined using 2-g air-dried samples, in duplicate, equilibrated for 6 d with 20 mL of 0-60 gg mL1 P solution in 0.01 M CaCl2 in polyethylene tubes at 25 � 20C. To inhibit microbial activities a few drops of toluene were added to each tube. After 6 d the solution P was determined by the ascorbic acid method (Murphy and Riley,








12

1962). Phosphorus removed from solution was considered adsorbed.

Desorption was studied by equilibrating the soil from the adsorption study in 20 mL 0.01 M CaC12 for 6 hrs (Singh and Jones, 1977). Increases in solution P were measured and considered desorbed P. Adsorption and desorption isotherms were constructed and a standard Langmuir equation was fitted to calculate adsorption maxima and bonding energy (Olsen and Watanabe, 1957).

Simulated silica matrix (SSMT). Hydrochloric acid (0.1 M) washed, fine silica was mixed with a nutrient solution containing N, K, Ca, Mg, Zn, Mo, Mn, Fe, Cu, B, S, and Cl (P excluded) (EMBRAPA, 1976) to simulate the nutrient requirement for maize in a solution culture. Solution pH was adjusted to 5.0 with 0.1 M HC1. Table 2-1 highlights the chemical compounds used to make the nutrient solution and the resulting concentrations. The inoculation of silica matrix was carried out with microbes grown on potato dextrose agar.

Five grams of ground organic amendments (the same ones used in incubation study with soil) were mixed with 95 g of silica matrix and incubated for 35, 65, and 150 d at 30 � 30C. The matrix was kept moist and mixed thoroughly every 6 d. After each incubation period, duplicate 2-g samples were extracted with 0.01 M CaCl2 following the same procedure as described for soil. The sum of P detected by sequential








13







Table 2-1. Ionic species and concentration of nutrient
solution used with silica to simulate solution
chemistry of a Xanthic Hapludox in a phosphorus release
study by organic amendments.



Source Species Conc+.
No. Chemical mmol (charge) L

1. Ca(NO3)2.4H20 Ca2+ 2.690
NH4NO3 NH4 , NO3 9.180

2. KC1 K++ 1.829
K2SO4 Cl 0.670 KNO3 S 0.480

3. Mg(N03) 2. 6H20 Mg2+ 0.650

4. FeHEDTA Fe3 0.035

5. MnC12. 4H20 Mn2+ 0.007
H3BO3 B 0.019
ZnSO4.7H20 Zn 0.0018 CuSO4. 5H20 Cu 0.0005 Na2MoO4.2H20 Mo 0.0006

+ pH was adjusted to 5.0 with 0.1M HC1. The reported concentration is considered adequate for maize growth in a solution culture.

+ The final concentration for K is calculated from KC1, K2SO4, and KNO3.








14



extraction with CaC12 was considered total P released from the amendments because of the inability of silica to adsorb P. This value was used to construct a P release curve for each amendment.



Calculation of total P adsorbed. Total adsorbed P was calculated as follows:

TAP = Np + Sp + OAp

where:

TAP = total adsorbed P on soil surface, Ag g

Np = native P present in adsorbed phase Ag gSp = applied inorganic P in adsorbed phase Ag gOAp = phosphorus from organic amendment in adsorbed phase Ag g �

The amount of P released from the organic amendments and adsorbed on the soil surfaces (OAp) was estimated by SSMT. The native exchangeable P on control samples was determined by the method of least squares (Reddy, 1990). For comparison purposes, P adsorbed on the soil surface following incubation with organic amendments was also determined by sequential extraction with 0.01 M CaCl2.








15



Results and Discussion

Organic amendments differed in chemical composition

(Table 2-2). The highest P concentration was observed in CM



(2.77 g 100 g-') and the lowest in PCW (0.16 g 100 g '). The highest C:P ratio was for maize (189.0) followed by grass (180.0).

Soil used was obtained from the Ap horizon of a highly leached oxisol classified as Xanthic Hapludox (clayey kaolinitic, isohyperthermic) (EMBRAPA, 1979). The soil was acidic (pH = 4.5), and low in total P (170 Ag g-1 soil) and Mehlich-I extractable P (3.3 Ag gol soil). Clay content ranged from 78-82% and the Al saturation was 78.4% (Table 23). The point of zero net charge measured by potentiometric titration in CaC12 solution was at pH 4.2 (Fig. 2-1). The native exchangeable P measured by the least squares method was 3.0 g g-' soil.








Table 2-2. Chemical composition of organic amendments used in the incubation
study.


OA+ Ca Mg K C P N C:P C:N Cu Zn Mn
-1 - 1 g 100g __ gg1 9Canav. 2.27 0.17 1.90 39.6 0.31 3.95 128 10 2.5 8.4 174.0 Peanut 1.44 0.27 1.95 38.5 0.33 3.32 116 11 2.5 8.9 70.0 Maize 0.35 0.19 2.25 39.6 0.21 2.93 189 13 5.0 40.1 95.0 Teph. 1.01 0.19 1.75 43.9 0.29 3.31 151 13 2.5 26.1 41.0 Mucuna 1.05 0.15 1.80 42.5 0.28 3.34 152 12 2.5 28.0 103.0 Grass 0.70 0.25 1.60 34.1 0.19 1.43 179 23 5.0 36.9 123.0 CM 7.62 0.51 2.10 17.8 2.77 1.46 6 12 92.5 102.9 399.0 PCW 1.57 0.14 0.35 22.3 0.16 0.95 139 23 155.0 280.8 213.0 Puer. 0.45 0.18 1.55 42.1 0.28 2.52 150 16 2.5 24.3 212.0 Cowpea 1.22 0.16 2.20 41.9 0.31 2.75 134 15 2.5 19.2 98.0

+ Organic amendments; Mucuna aterrima (Mucuna), Pueraria phaseoloides (Puer.), Canavalia ensiformis (Canav.), Zea mays (Maize), Arachis hypoqaea (Peanut), Tephrosia candida (Teph), Vicna unguiculata (Cowpea), mixed gramineae (Grass), aerobically digested city waste (PCW) and chicken manure (CM).














Or








17










Table 2-3: Selected physical and chemical properties of Ap
horizon of the Xanthic Hapludox used in adsorption
studies.


Parameters+ Value Units of Measured Measurement

Bulk density 1.25 g cm3 Clay 815 g kg"
pH H20, KC1 4.5, 4.2
Net Charge 0.22 cmol (-) kg
Oxides Fe, Al 0.075, 0.325 g 100 g
Al Sat. (ECEC) 78.4 %
-1
Total P 170.0 Ag g soil
Mehlich-I P 3.3 A g - soil
Al, Fe (P) 29.0, 1.4 jg g soil







18

10


Depth 0-20 cm


ic 0










C3O







5
6I













10l
o
5 ,
















2 3 4 5 6 7 pH

Figure 2-1. Acid-base potentiometric titration curves for
the Ap horizon of Xanthic Hapludox with varying
concentration of CaC12








19

Adsorption Isotherms

The results of adsorption and desorption studies are presented in Fig. 2-2 and 2-3. Incubation of the soil with amendments for 35 d influenced P adsorption. This was more pronounced at higher equilibrium solution P concentrations (Fig. 2-2a). At a P concentrations >1.9 gg mLI , PCW amended soil adsorbed more P than the unamended soil. A similar result was obtained by Singh and Jones (1976) for organic amendments with low P content and high C:P ratio. On the other hand, soil samples amended with CM and Mucuna adsorbed less P than the unamended soil. As the time of incubation increased to 150 d, all amendments reduced P adsorption compared to the control (Fig. 2-2b). A five-fold reduction in P adsorption was observed for the CM treatment at an equilibrium P concentration of 0.3 Ag g-1 soil. The reduction was perhaps due to net mineralization of P even from P-poor organic amendments with time. Another factor in P mineralization is the C:P ratio. When this ratio remains less than about 200:1, immobilization predominates during the initial stages of decomposition. But, as the decomposition proceeds, this ratio becomes narrower due to the concentration of P in decomposing residue and continuous degradation of C by microorganisms.

As plants deplete soil solution P, the solution must be continuously recharged, if good growth is to be maintained. Recharge occurs when P is desorbed from the soil surface,








20

and this will happen in sufficient quantity only if the soil has a large capacity to sorb and therefore desorb P. Thus, even though the soils with high sesquioxides content require more P to achieve a given level of P in soil solution, they have the compensating value of being able to supply P to the soil solution as it is taken up by plants. The results of desorption studies are presented in Fig. 2-3.

Soil incubated with Mucuna aterrima (Mucuna), PCW, and the control for 35 d did not show any difference in P desorption. However, desorption was higher for all amendments incubated with soil for 150 d (Fig. 2-3b). In fact, CM reduced adsorption and increased desorption at all incubation periods. The observed differences of organic amendments in influencing P desorption characteristics of the soil could be attributed to their decompositional characteristics, P content, C:P and C:N ratios, and concentration of such elements as Zn and Cu. Incubation of soil with low P containing organic materials may not initially influence soil P desorption characteristics.



Phosphorus release from organic amendments

Phosphorus release data for different amendments in

soil and silica substratum are presented in Table 2-4. More P was detected in the silica matrix for all amendments compared to the soil matrix. With increasing incubation time, a decrease in P was observed for maize and grass








21
Adsorption
600
35 Days a) o 500 o- 0
C0
Mucuna CM PCW Control o :"
400 --- ----A --O 4 0 4
300 - 4.
0
200 A /


n0







o

200 150 Days b)

4 0 0 - - ------ o - -- -- - o

100




.100 o................



0.01 0.03 0.1 0.3 1 3 10 30 Solution P (ug m)




Figure 2-2. Phosphorus adsorption isotherms following 35 d
(a) and 150 d (b) of soil incubation with Mucuna aterrima (Mucuna), Chicken Manure (CM), Aerobically Processed City Waste (PCW), and Control.
Waste (PCW), and Control.








22
Desorption
600
35 days a) 500


400 - Mucuna CM PCW Control
-- -- ........ ....... ---- .....

300


200 - o


0 100 o
O

3
"o
0)
0)
*0
0 600


CO500 150 Days b)
(A a 500 Mucuna CM PCW Control A, o
400 A 0 ----A---- A A
0A
400 0

o L
300

o A
200 A
---------.... .----f k
100
0A A
0 0 II R

0.003 0.01 0.03 0.1 0.3 1 3 10

Solution P (ug mC)


Figure 2-3. Phosphorus desorption isotherms following 35 d
(a), and 150 d (b) of soil incubation with Mucuna
aterrima (Mucuna), Chicken Manure (CM), Aerobically
Processed City Waste (PCW), and Control.








23







Table 2-4. Release of 0.01 M CaC12 extractable P (pg g )
following incubation of organic amendments with soil
and silica as matrix substratum.


Org.+ Soil Silica amend.

35 65 150 35 65 150
---------------- Days------------------CM 109a* 204.9a 245.1a 447a 1031a 1380a Mucuna 21b 14.Obc 16.9b 80b 124b 160b Tephrosia 22b 14.9b 19.0b 65b 116b 152b Cowpea 15c 11.5bc 16.2bc 43c 70c 92c Peanut 18bc 10.3cd 8.6bcd 41c 117b 142b Grass 6e 6.7de 3.4d 37c 54c 102c Maize 14cd 12.5bc 3.9d 37c 77c 1llc PCW 3e 2.8e 10.8bcd 34c 62c 64d Canavalia 14cd 11.8bc 15.3bc 31c 83c 148b Kudzu 10d 11.4bc 14.1bcd 23c 71c 94c Control 4e 4.4e 5.2cd Trace Trace 0.1e

CV% 6.3 14.0 8.2 12.3 7.5 12.0

* Means in the same column followed by the same letter are
not significantly different at the 95% level of
probability, as determined by Duncan's Multiple Range
Test.
+ Mucuna aterrima (Mucuna), Pueraria phaseoloides (Kudzu), Canavalia ensiformis (Canav.), Zea mays (Maize), Arachis hypoqaea (Peanut), Tephrosia candida (Tephrosia), Vigna unguiculata (Cowpea), Mixed Gramineae (Grass), Aerobically Processed City Waste (PCW), and Chicken Manure (CM).








24

treatments in the soil matrix. Processed city waste appeared to immobilize soil P as indicated by the P measurement made at 35 d of incubation but at 150 d a net release was observed. All leguminous amendments followed similar P release patterns. The highest P release was observed for CM treatment at 150 d of incubation (245 gg g-1 soil) and the lowest was for maize and grass treatments (3.9, and 3.4 Ag g-' soil, respectively).

When silica was used as the incubation matrix, there was a net release of P from all amendments which indicated that P released from amendments was being adsorbed by soil (soil matrix) and was not all exchangeable in sequential extraction with 0.01 M CaC12. Maize and grass treatments which had shown no net P release at 150 d in the soil matrix released over 100 Mg g-1 silica in the silica matrix. Given the similarity in molecular structure and chemical behavior of Si and P (Fig. 2-4) it is believed that added Si increases water soluble and easily extractable P (Adams, 1980), and Si does not absorb P, because both of them in ionic forms are negatively charged. It is also suggested that silicate and phosphate ions compete for the same adsorption sites on Al(Fe)-oxide surfaces (Mekaru and Uehara, 1972) and they form insoluble precipitates with such common ions as Al, Fe, and Ca. This suggests that all P mineralized from organic amendments remained in the solution and was easily extractable by 0.01 M CaCl2.








25

To understand the P release pattern from different amendments, surface response curves were fitted and are illustrated in Fig. 2-5. Phosphorus release in silica for maize, PCW, and CM followed a logarithmic function and the r2 for PCW, CM, and maize treatments were 0.73, 0.82, and 0.98. A trace amount of P (0.1 Ag g- ) was detected in the control treatment in the silica matrix confirming that the matrix was not contaminated and the P detected came solely from the decomposition of amendments.


Lanqmuir Parameters

Soil Matrix. A standard Langmuir equations were fitted to the data after correction was made for preadsorbed P using sequential extraction techniques (Fig. 2-6). This was followed by calculation of the adsorption maxima and bonding energy for each treatment (Table 2-5). All but Pueraria phaseoloides (Kudzu) increased the Langmuir adsorption maxima at 35 d of incubation compared to the control. The extent of variation among different amendments ranged from 536 to 818 Ag g-1 soil at 35 d of incubation. It is noteworthy that PCW had the highest adsorption maxima (818 Mg g- ). The bonding energy was reduced for all treatments compared to the control. As the time of incubation increased from 35 d to 150 d, the adsorption maxima decreased for all amendments compared to the control treatment. But the








26












a) b)








O Shared 0 * Si eo gP o














Figure 2-4. A schematic representation of molecular
structures of oxides of phosphorus (a), and oxides of
silicon (b).









27
70.0 1,600.0
a) - "- A b)
60.0 1,400.0

,-' PCW 1,200.0- CM
50.0 - .- ,Silica 1,000.0
40.0 ' ,
A Soil = - 0.32 + 0.07X 800.0 Soil = 100.8 + 1.03X
30.0 - r 2 =0.90 ,' r 2 =0.77
Silca = -29.5 + 19.6logx 600.0 ,'
20.0 r2 =0.73 A Silica = - 1716 + 6281ogX 400.0
Soil r2=0.82
10.0 - 200.0

0.0 0.0
0 20 40 60 80 100 120 140 160 20 40 60 80 100 120 140 160


120.0 0.5
C) Maize ,o� d) Control

100.0
0.4

80 Soil = 0.0002X + 0.05 80.0 A
0.3 r 2=0.15 Soil = 40.7 - 7.2logX
60.0
r 2 =0.92 Silica = 0.0004X - 0.02
0.2
. r 2 =0.92 40.0 '
Silica = - 138.5 + 50.3logX

20.0 r 2 =0.98 0.1


0. 0. ,- " I
20 40 60 80 100 120 140 160020 40 60 80 100 120 140 160

Incubation (days)





Figure 2-5. Effect of incubation period and incubation
matrix on P release pattern from different organic
amendments in a laboratory study.








28






0.02
PCW = 0.00123 X + 0.00096 150 Days r 20.94 Grass = 0.00099X + 0.00051
0.015
1... r 2 =0.84 E PCW

0.01 o Tephrosia

y ,, Tephrosia = 0.00085x + 0.00049 , r 2 =0.91 CM
0.005 - C
CM 0.00013X + 0.00003 r =0.76


0 _ - - -0 5 10 15 20 25 P in solution (ug/mL)

















Figure 2-6. Phosphorus adsorption isotherms for the Xanthic
Hapludox (0.01 M CaC12) incubated with organic amendments for 150 d. The lines in the figure
represents fitted Langmuir equation.










Table 2-5. Langmuir parameters (k and b) for P sorption by soil incubated
with different organic amendments.


Incubation Period

35 Days 65 Days 150 Days
Org.
Amend. r2 b+ kl r2 b k r2 b k

Canavalia 0.87** 638 0.384 0.81** 633 0.303 0.89** 528 0.264 Peanut 0.73** 632 0.233 0.87** 522 0.282 0.85** 509 0.229 Maize 0.94** 560 0.785 0.84** 670 0.308 0.86** 673 0.431 Tephrosia 0.69* 768 0.229 0.91** 477 0.291 0.91** 465 0.246 Mucuna 0.90** 536 0.207 0.64* 654 0.087 0.49ns 695 0.120 CM 0.66* 643 0.080 0.67* 618 0.079 0.76** 496 0.103 Kudzu 0.93** 471 0.417 0.94** 577 0.476 0.95** 535 0.599 PCW 0.67* 818 0.224 0.66** 815 0.271 0.94** 517 0.375 Cowpea 0.62* 729 0.146 0.89* 585 0.244 0.91** 543 0.473 Grass 0.74** 765 0.247 0.80* 707 0.331 0.84** 603 0.363 Control 0.94** 535 0.673 0.95** 544 0.941 0.98** 591 1.941

*, ** Significant at the 0.05 and 0.01 levels, respectively.

+ Mucuna aterrima (Mucuna), Pueraria phaseoloides (Puer.), Canavalia ensiformis (Canav.), Zea mays (Maize), Arachis hypoqaea (Peanut), Tephrosia candida (Teph), Vicna unquiculata (Cowpea), mixed gramineae (Grass), aerobically digested city waste (PCW) and chicken manure (CM).

+ b, P adsorption maxima, 1g P g-1

� k, bonding energy, mL g P-1








30

bonding energy did not follow any defined pattern. For the grass treatment (mixture of different grasses) bonding energy value increased from 0.25 to 0.37 mL Ag-1, for Canavalia ensiformis (Canavalia) bonding energy dropped from

0.39 to 0.26 mL Ag-', and for CM it remained at 0.10 mL g-. This finding does not agree with the data obtained by Hundal et al. (1988) where all studied amendments reduced the bonding energy as the incubation period increased from 20 d to 40 d.

The range of P in organic amendments varied from 1.7 to 25.0 g kg-1 on a dry matter basis. There was a high concentration of Zn and Cu in PCW, and Ca in CM. Being polyvalent cations, they have a high affinity for P adsorption through cation bridging (Holford and Mattingly, 1975; Haynes, 1989). The decomposition rate constants of the organic amendments were different and so was the release of different elements as the decomposition proceeded. These factors also may have contributed to the observed differences in P adsorption maxima and bonding energy.

Silica matrix. The value of the total adsorbed P was

adjusted based on the amount of P released from the organic amendments in the silica matrix. The Langmuir equation was fitted to the data, and the value of adsorption maxima and bonding energy were recalculated for selected treatments. The results are presented in Tables 2-6 and 2-7. The recalculated adsorption maxima and bonding energy values








31

were higher than the values obtained based on the P release from the amendments in the soil matrix. At 150 d for the CM treatment, the adsorption maxima were higher by 1170 Mg g-1 soil and the bonding energy was higher by 20 fold. The low adsorption maxima (496 jg g-') and bonding energy (0.10 g mL'1) for soil matrix indicated that the P capacity factor for CM applied plots was lower compared to plots which received other amendments and, therefore, CM plots should have the lowest residual P effect in field testings. But, the result from the Silica Matrix technique indicated otherwise. The high adsorption maxima for CM at 150 d of incubation (1667 Mg g-') and the high bonding energy (2.0 mL Ag-) are indications of sustained P desorption and residual P availability.

Predicted and measured phosphorus adsorption. The data on P adsorption maxima calculated by the Langmuir equation and actual P adsorption based on P release from amendments in soil and silica matrices are presented in Fig. 2-7. For the CM treatment, with increasing incubation time P adsorption maxima increased. Continued adsorption beyond adsorption maxima indicated the presence of a precipitation reaction, or multilayer adsorption. For tephrosia and kudzu, the adsorption maxima decreased with time. A good agreement among adsorption maxima and actual adsorption, calculated based on SSMT, was observed at higher equilibrating solution








Table 2-6. Langmuir parameters (k and b) based on net release of P
measured by sequential extraction of simulated silica matrix with 0.01 M CaC12.


Incubation Period
35 Days 65 Days 150 Days
Org.+
amend. r2 b+ k r2 b k r2 b k

Silica
Tephrosia 0.87** 714 0.42 0.96** 556 0.56 0.97** 588 0.55 CM 0.99** 909 0.61 1.00** 1429 1.17 1.00** 1667 2.00 Kudzu 0.95** 502 0.43 0.97** 625 0.70 0.97** 625 0.89 PCW 0.97** 908 0.22 0.90** 769 0.45 0.97** 588 0.63 Grass 0.94** 769 0.32 0.92** 714 0.47 0.93** 667 0.65


** Significant at P = 0.01 level. ns = not significant

+ Pueraria phaseoloides (Kudzu), Tephrosia candida (Tephrosia), Mixed Gramineae (Grass), Aerobically Processed City Waste (PCW), and Chicken Manure (CM).

+ b, P adsorption maxima, Ag P g� k, bonding energy, mL g P-







Table 2-7. Difference in the estimated values of Langmuir parameters (b and k) based
on the estimation of preadsorbed P by simulated silica matrix technique and
sequential extraction.


Incubation Period

35 Days 65 Days 150 Days

Org.+
amend. b+ k b k b k


Tephrosia -54.1 (14.2) 0.19 (0.01) 77.6 (8.3) 0.27 (0.04) 122.2 (8.4) 0.30 (0.01)
CM 265.1 (35.8) 0.53 (0.02) 810.6 (58.3) 1.09 (0.09) 1170.7 (50.4) 1.90 (0.12)
Kudju 29.2 (5.9) 0.01 (0.00) 48.0 (4.9) 0.22 (0.02) 90.0 (6.8) 0.29 (0.01)
PCW 91.3 (15.9) 0.00 (0.00) -46.8 (15.2) 0.18 (0.01) 71.2 (7.2) 0.25 (0.02)
Grass 4.3 (1.3) 0.07 (0.01) 7.3 (1.4) 0.14 (0.01) 63.7 (6.9) 0.29 (0.02)

+ Pueraria phaseoloides (Kudzu), Tephrosia candida (Tephrosia), Mixed Gramineae (Grass), Aerobically Processed City Waste (PCW), and Chicken Manure (CM). + b, P adsorption maxima, pg P g-'

� k, bonding energy, mL Ag PNumbers in the parenthesis are standard deviations.









34
2,000 2,000
a) CM a b) Tephrosia SM
1,500 a 1,500
b SSM

LAM

1,000- a [ 1,000aa a
b a

500C - 0 500b a
O



0 0
35d 65d 150d 35d 65d 150d
2,000 c - 2,000

c) Kudj d) PW
0 d) PCW

1,500 1,500



1,000- 1,00oo a


a a
b ..b ob . ab b ab




0 0
35d 65d 150d 35d 65d 150d Incubation Time












Figure 2-7. A comparison of P adsorption by Xanthic Hapludox
based on the estimation of preadsorbed P by sequential
extraction of soil or silica matrix incubated with
organic amendments.








35

concentration (60 gg g-1 soil) which indicated the validity of this technique in predicting P requirements of amended soils. In highly weathered soils, extractable P is usually low but the amount of P which the soil can immobilize varies greatly because of the variation in the reactive surfaces. Addition of organic amendments changed the reactive soil surface as indicated by the change in adsorption maxima and bonding energy.



Conclusion

Decomposition of organic amendments influenced P

adsorption and desorption by soil. Amendments with low P content (PCW, maize, and grass) immobilized P during early incubation periods. Higher amounts of P were desorbed from soil samples incubated with CM. Soil incubated with Mucuna, PCW, and the control for 35 d did not show any difference in P desorption. However, desorption was higher for all amendments incubated with soil for 150 d. Incubation of soil with low P-containing organic materials may not influence soil P desorption characteristics initially.

Amendments' P content and C:P ratio did not prove to

be helpful in predicting P mineralization. Amendments with a C:P ratio of 1:139 (PCW) immobilized P during their decomposition while others with similar or even higher C:P ratios had a net P release in soil matrix. All amendments








36

released P in the silica matrix independent of their P content and C:P ratio.

The use of SSMT to measure release of P from

decomposing organic amendments aid in the calculation of adsorption maxima compared to the seqential extraction of amended soil with 0.01 M CaC12 solution. The calculated values were in close agreement with actual adsorption measured at high equilibrium P concentration.














CHAPTER III
ORGANIC AMENDMENTS AND CROP PHOSPHORUS NUTRITION



Phosphorus Management Strategy


In highly leached soils, the minerals with permanent charge have been either severely altered or completely weathered out, so that the surface charge arises from adsorption of potential determining ions such as hydrogen and hydroxyl. The magnitude of the surface charge is expressed by a combined Gouy-Chapman and Nerst equation. This equation provides a theoretical basis for increasing the cation retention capacity of a soil by lowering pHo (value of soil pH at which net surface charge is zero). One way of lowering pHo is to increase the organic matter, phosphorus, or silica content of the soil (Uehara and Gillman, 1981).

Highly weathered soils are also very poor in total and plant available P. Their high P fixing capacity requires high doses of applied P to meet crop demands. The classic work of de Wit (1953) on physical theory of fertilizer placement predicts that when suboptimum quantities of fertilizer are used, restricted placement is desirable (Fox et al., 1986). de Wit based his analysis on nutrient uptake 37








38

by entire root mass, and part of the root mass immersed in nutrient solution and established a relationship between Ur (g of nutrient taken up by the plant when part of the root mass, Xr, was immersed in the solution) and Ub (g of nutrient taken up by the plant when the entire root mass was immersed in the solution).

Ur/Ub = (Xr/Xb)"44

This relationship appears to be independent of crop type and nutrient solution concentration. Under field conditions Ub, Ur, Xr, and Xb take on new meanings.

Ub = uptake rate from broadcast fertilizer

Ur = uptake rate from banded fertilizer

Xr = width of the fertilizer band

Xb = distance between the crop rows

The use of data from P sorption curves concerning soil solution P concentration to predict P uptake and crop yield in combination with P placement analysis offers a way to increase fertilizer use efficiency. Faced with a high P fixing soil and a small quantity of fertilizer, the fertilizer can be used to the best advantage by concentrating it in a band so that the P concentration in the soil solution is identical to the concentration that produces a maximum yield in a broadcast application. Any deviation from this optimum value, either to higher or to lower concentrations, leads to less than an optimum return per unit of fertilizer input.








39

Management of Organic Amendments

Organic amendments can affect the reaction of P in the soil through complexation of polyvalent cations which are major phosphate adsorption sites. It is widely believed that humus in association with cations such as Fe 3, Al and Ca2+ retains significant amounts of P. Appelt et al. (1975) prepared a hydroxy-Al-humic acid complex that adsorbed P because of the creation of new P adsorption sites. They concluded that any increase in organic content of a soil could lead to greater adsorption. A study conducted by Swift and Haynes (1989) also showed that Al-organic matter associations have a significant phosphate adsorption capacity. Indeed, the Al-humate adsorbed amounts of phosphate similar to those commonly reported for Al and Fe oxides (McLaughlin et al., 1981) on a w/w basis. However, simple organic acids, fulvic acids, and humic acids had no effect on P adsorption by volcanic ash-derived soils. For these soils, P was preferentially adsorbed over the organic acids studied (Appelt et al.,1975).

The addition of organic materials to high P fixing soil can decrease, increase, or leave virtually unaltered the P fixation capacity (Yuan, 1980). The reduced fixation is the result of: 1. complexation of Fe, Al, and Ca by organic anions (Larsen et al, 1959), 2. competition of organic anions and P for the same adsorption sites (Nagarajah et al, 1970), 3. development of organic coatings on mineral surface








40

(Easterwood and Sartain, 1987), and 4. reduction in bonding energy of adsorbed anions resulting in low residence time for adsorbed P (Hudal et al, 1988). Increased adsorption may be due to: 1. cation bridging between organic anions, and Fe 3, Al , and Ca2+ leading to formation of new sites for P adsorption (Appelt et al, 1975), and 2. ability of organic ligands to maintain hydroxy-Al, and Fe in a non-crystalline state and thus maintaining a greater surface area (Swift and Haynes, 1989).

Amendments chemical composition. Although organic acids are an integral part of all organic matter by far they are not the only reactive component influencing P adsorption. Singh and Jones (1976) suggested that the P content of organic residues plays an important role in the release or fixation of added P. Similarly, chemical and decompositional characteristics of the organic matter may influence total CO2 evolution (Sweeney and Graetz, 1988). This gas when dissolved in water, forms carbonic acid, which is capable of decomposing certain primary minerals. Elemental ratios such as C/N and C/P are also considered valuable indicators for net mineralization or immobilization of N and P contained in organic amendments.

One of the important studies in this area was

conducted by Blair and Boland (1978) who examined the release of P from white clover residue in high and low P status soils. Their results suggested that the addition of








41

plant material resulted in immobilization of soil P only in the low P soil in the absence of plants. In the high P soil no immobilization of P was observed.

Crop residues have an effect on the nutrient status of the soil. The cumulative effects of increasing quantities of organic residues on available nutrients in soil were studied for 11 years by Larson et al. (1978). They reported that addition of 16 tons/ha of plant residue per year to the soil increased the amount of N, S and P by 37, 45 and 14% respectively, over the control treatment. They also found that the NH4-N production, weak acid soluble P and exchangeable K in the soil were increased as a result of increasing the addition of organic residues.

Solubilization of Mn and Fe in soil were affected by wheat straw and alfalfa amendments (Elliot and Blaylock 1975, Sims, 1986). The release was greater for Mn than Fe and also much higher at 30 kPa than 50 kPa moisture tension. The release of Mn and Fe from the soil column followed the following order: alfalfa > wheat straw > soil alone. They suggested that the potential accumulation of soluble Mn in well drained soil was possible where there was a large quantity of plant residues incorporated into the soil.

Application technique. In the case of organic manures, changes in organic P will to some extent depend on whether the material is left on the surface of the soil or is plowed








42

in. Douglas et al. (1980) reported that the method of placement, composition of residues and loading rates were important factors influencing mineralization or immobilization of N and S. A higher mineralization rate of N from residue incorporated at 4 cm soil depth as compared to the surface was also observed by Brown and Dickey (1970) and Cocharan et al. (1980). Data on P mineralization from organic matter in the literature is scarce. Inorganic Phosphorus Management Strategy

Salinas and Sanchez (1976) have outlined a three-point strategy for P management under limited resource conditions.

1. Use of cheaper sources of P. Two main sources are phosphate rocks (PR) and thermally altered sources, such as basic slags and the Rhenania phosphates. Numerous reports have appeared in the literature regarding the fertilizer value of PR as compared to other sources of P fertilizers, e.g., superphosphate (Khasawneh and Doll, 1978; Hammond et al., 1986; Hernandez and Sartain, 1985). Recently, Hellums et al. (1989) compared the potential agronomic value of some rock phosphate from South America and West Africa and concluded that in addition to P the PRs with medium to high reactivity have a potential Ca supply value. 2. improved soil test interpretations, and 3. improved placement methods.








43

Fertilizer placement. The advantage of banding

phosphate fertilizer is well known. What is generally not understood is the sensitivity of nutrient uptake to band width. de Wit (1953) demonstrated that a soil that is virtually incapable of supporting a crop with 100 kg/ha of broadcast phosphorus will produce nearly 50% of maximum yield with the same amount of fertilizer applied in a narrow band. A more detailed description of de Wit's analysis and some of the assumptions contained in the analysis are further elaborated by van Wijk (1966), Uehara and Gillman (1981), and Fox et al. (1986). In agreement with the theory Fox and Keng (1978) reported a better response from localized P placement as compared with complete incorporation if suboptimal P rates were used, but if quantities of P were sufficient, best results were obtained from incorporating P in the entire soil volume.

Kamprath (1967) found that similar maize yields were obtained by annual banded applications of 22 kg P/ha for seven years as were obtained by an initial P application of 350 kg/ha. Banding, therefore, saved more than half of the P requirement. Applying N and P in knifed bands has been shown to be an effective method of applying N and P to winter wheat (Leikam et al., 1978, 1983). Experiments on maize have shown that dual-placed N and P increases P uptake and maize grain yield more than when P is banded to the side or below the seed (Raun et al., 1987).








44

Band spacing of applied N and P fertilizer affects the concentration of these nutrients in the applied band and the probability of roots contacting the band. Sleight et al. (1984) showed that in high P-fixing soil, increasing the probability that root-fertilizer contact will occur is more important than reducing soil-fertilizer contact during the first week of oat (Avena sativa L.) growth.

There is also a threshold value of soil solution P concentration beyond which the P uptake rate does not increase (Jungk, and Barber, 1974). Anghinoni and Barber (1980) in a P placement experiment reported maize root growth stimulation in the portion of the soil where P was added. Maximum shoot dry weight in their experiment was obtained by placing the fertilizer in 0.25 of the soil volume.



Summary

Application of P in narrow bands results in reduced fixation of applied P and improves crop P nutrition. This belief is based on: (i) localized P is protected to some degree against irreversible adsorption or precipitation reactions with the soil, (ii) localized P may be more readily accessible to seedling roots than P widely distributed in the soil, and (iii) plants can be adequately supplied with P through a few roots which proliferate in the fertilizer band. Further, combined application of organic








45

amendments with inorganic P in a band will reduce direct contact of inorganic P with a large volume of soil.

In this context, the overall objective of this research was to devise a technique to sustain P nutrition in a highly leached Oxisols with the help of organic and inorganic P sources applied in narrow bands. The specific objectives were to: (i) study the effect of different amendments and their rate of application on maize dry matter production and herbage nutrient concentration, (ii) examine the beneficial effect of the combined application of organic amendment with inorganic P in a narrow band as measured by maize grain production and soil nutrient dynamics, and (iii) measure the residual effect of applied treatments in relation to selected soil chemical characteristics.




Materials and Methods

This experiment was conducted at the Empresa Brasileira de Pesquisa Agropecuaria (EMBRAPA) station located 30 km north of Manaus at an elevation of 50 m in the Amazonas state of Brazil (Fig. 3-1). Climate in the Manaus region has been classified in the Koppen nomenclature as afi, tropical, humid and hot (Goes and Ribeiro, 1976). The soil used in the study has been classified as Xanthic Hapludox (clayey kaolinitic, isohyperthermic) (EMBRAPA, 1979).








46

Glasshouse Experiment

A factorial arrangement of three factors to yield 18

treatment combinations was employed in a randomized complete block design (RCBD) with four replications (Table 3-1). Five kg of unlimed soil from the Ap horizon of a Xanthic Hapludox was mixed in a pot with organic amendments (<2 cm long) 7 d prior to maize seeding. Maize (variety BR 5110) was planted and thinned to 2 plants per pot 7 d after planting. The experiment was harvested at 65 d. At harvest, leaves immediately below and opposite to the ear leaf (70% plants had begun to develop ear) were collected for chemical analysis. A 0.2 g sample of the ground leaf tissue was digested with H2SO4 and H202. Potassium, Ca, Mg, Fe, Zn, Cu, Mn were determined with an atomic adsorption spectrophotometer, and P was determined colorimetrically using the Murphy and Riley (1962) procedure. Soil P was extracted with the Mehlich-I extractant P (0.05 M HC1 +

0.0125 M H2SO4, with a soil solution ratio of 1:10, and 5 min shaking time), unbuffered 1 M KCl (1:10 soil:solution ratio) was used for the determination of extractable Al. Aluminum was determined by titrating the extract with 0.1 M NaOH to bromthymol blue endpoint. Soil reaction (pH) was determined in water using a soil:water ratio of 1:2.5 and in

1 M KC1 using the same ratio. Maize dry matter production was recorded and surface response curves were fitted in the case of interaction effects among factors.








47

60 50 AM-o10

Rio Preto da Eva
0 0 FSR
A AZ N Rio Amazonas
Manaus
10 aila 10 EMBRAPA MANAUS


B R A Z IL B r s.2

Rio Negro
a) b) 60 50




Effective Rainfall, mm
400
400 21 23
27 25
300
18
24
200 -25 2

16
100 -
20
3 515

C)
-100

-200
A'88 S O N D J'89 F M A M J J A Months





Figure 3-1. Geographic location of Amazon basin in Brazil
(a), on-station and farming systems research (FSR)
sites (b), and effective rainfall during the period of August 1988 until August 1989 (c) at EMBRAPA station in
Manaus, Brazil.








48





Table 3-1. Factorial arrangements of treatments for the glasshouse study.


Type Amendments+ Rate P Equivalent kg ha"'

Mucuna 1&2 8.8, 17.6 Organic CM 1&2 8.8, 17.6
PCW 1&2 8.8, 17.6

Inorganic TSP 1,2&3 0, 8.8, 17.6 + Mucuna (Mucuna aterrima), CM (Chicken Manure), PCW (Aerobically Processed City Waste) TSP (Triple superphosphate)


Field Experiment

The site was cleared by burning an existing sugarcane crop. Sugarcane stems and rhizomes were taken out of the field. A uniformity trial using maize as an indicator crop was planted for 60 d in order to record the existing variation in the field. Visual field ratings combined with soil and plant chemical analysis results were used to select a uniform area for planting the field experiment. The selected area was divided into 48 plots and soil samples were taken from three consecutive depths (0-15, 15-30, and 30-45 cm) in each plot.

Dolomitic lime (2 Mg ha" ) was applied over the entire area 2 weeks prior to maize planting. Nitrogen and K were

-11
applied at the rate of 200 and 100 kg ha'. Zinc, Mo, and Mn (2 kg ha"I of each) were applied in the band. Half of the N








49

as urea was broadcast over the entire plot and the remaining half was sidedressed at 35 and 70 d after planting. All K was applied basal broadcast. Six furrows each 20 cm wide and 8-10 cm deep were opened in each plot 80 cm apart. The plot size was 10x5 m.

Mucuna aterrima and Canavalia ensiformis grown for 60 d in an adjoining field were harvested and passed through a hay chopper to produce a uniform size of less than 6 cm. All organic amendments were applied in bands to which different rates of P from triple superphosphate (TSP) was added. A soil cover about 2 cm thick was put over the amendments. Twelve treatments were tested in a RCBD with 4 replications. The P release information from the SSMT technique (Chapter II) was used to calculate the amount of a given amendment required to supply P equivalent to 8.8 and 26.4 kg haI.

First maize. Maize variety BR-5110 was planted on December 27, 1988. Plant population was adjusted to approximately 55 x 103 plants ha"1 during the first sidedress at 30 d after planting. Soil samples were collected before planting, during tasseling (65 d after planting) and within a week after the harvest. Leaf samples were collected during tasseling. They were taken from immediately below and opposite to ear leaf. The inner four rows were harvested for grain. Grain production was recorded at 15% moisture level.







Table 3-2. Description of treatments tested in the field



Source Amendment Type Treatment+ P Equivalent
-- kg haI -Mucuna aterrima M60 26.4 M20 8.8
ORGANIC Canavalia ensiformis C60 26.4 C20 8.8
Processed City Waste PCW60 26.4 PCW20 8.8
Chicken Manure CM60 26.4 CM20 8.8

INORGANIC Tiple Superphosphate TSP20 8.8 TSP40 17.6
TSP60 26.4

ORGANIC+INORGANIC Mucuna, Canavalia, M20+TSP20 17.6 PCW, CM, TSP C20+TSP20 17.6 PCW20+TSP20 17.6
CM20+TSP20 17.6

CONTROL Control 0

+ Whenever P from organic and inorganic sources was applied together, organic amendments were applied first in furrows followed by the application of inorganic P.







-Ln
CD








51

Second maize. This trial was conducted to measure the residual effects of the treatments applied during the first maize crop. Maize stover was taken out of the field. Nitrogen was applied in the same manner as for the first crop. No other nutrient applications were made. Maize (variety BR-5110) was dibble planted in the rows and the population was adjusted to 55 x 103 plants ha '. Soil samples were collected during tasseling and after harvest. Again 1020 maize leaves were sampled during tasseling initiation from each plot. Maize grain production was recorded from four inner rows and was adjusted to 15% moisture content.


Chemical Analysis of Soil and Plant Materials

Soil pH was determined in water, 0.01 M CaCl2, and 1.0 M KC1 using a soil:solution ratio of 1:2.5. Total P was determined by wet combustion method and inorganic P fractionation was carried out using Chang and Jackson procedure (1965). The pH of NH4F was adjusted to 8.2. Mehlich-I extractant (0.05 M HC1 + 0.0125 M H2SO4) was used to extract soil P, K, Zn, Cu, and Mo. Phosphorus was determined colorimetrically and K by a flame photometer. Aluminum, Ca, and Mg were extracted with unbuffered 1 M KCl (1:10 soil:solution ratio). Aluminum was determined by titrating the extract with 0.1 M NaOH to bromthymol blue endpoint, and Ca and Mg were determined with an atomic absorption spectrophotometer. Soil apparent density was








52

determined by the method of a measuring cylinder. The cylinder was struck against a rubber pad 10 times from a distance of 10 cm. The final weight of the cylinder was taken and the apparent density was calculated as the following: Apparent Density (g cm"3) = Weight of dry soil at 1050C / Volume of the soil in the cylinder. Particle size distribution was measured by the pipet method.

All organic amendments and maize leaf samples were

dried at 65 OC for 72 hours for dry matter determination. Subsamples of the dried plant material were ground in a Wiley mill to pass a 0.5-mm screen. A 0.2-gm sample of the ground plant material was digested with H2SO4 and H202. Phosphorus was determined calorimetrically, and K, Ca, Mg, and micronutrients were determined with an atomic absorption spectrophotometer.


Results and Discussion

Selected chemical and physical properties of soil used in the glasshouse and field studies are presented in Tables 3-3 through 3-5.

With increasing soil depth, a reduction in Al-P was

observed. But, the Langmuir adsorption maxima increased with the depth (789 Ag g-' soil at 30-45 cm) (Table 3-3). It is interesting to note that with over 80% clay content the soil had a hydraulic conductivity of 25.1 cm h"1 (Table 3-4). The ammonium oxalate extracable Fe and Al were high and so was








53

the Al saturation (>76.0%) calculated from the effective cation exchange capacity (ECEC) (Table 3-5).



Glasshouse Study

An analysis of variance (Table 3-6) presented for maize herbage dry weight yield indicated the presence of a three way interaction among types of organic amendment, rate of amendment, and rate of TSP. Among the three amendments tested, PCW produced the lowest yield when applied alone or in combination with TSP (Fig. 3-2a). It is interesting to note that when PCW was applied at higherr rate (equivalent to 17.6 kg ha'1 of P) it produced less dry matter per pot compared to when this amendment was applied at lower rate (Figure 3-2a). A possible explanation is that the slow rate of release of nutrients from the material resulted in an initial P deficiency for maize seedlings. There was also a high concentration of Zn and Cu in this material (280, 155 Ag g-', respectively) which may have played a role in providing cation bridging between organic and P anions making P less available (Murray and Linder, 1984). By forming organo-metal complexes, humic and fulvic acids as well as simple acids can dissolve or decompose such minerals as feldspar, gibsite, goethite, hematite, and mica (Schnitzer, 1977). Therefore, organic amendments with high Zn and Cu content may have limited potential as a source of P in highly leached soils.








54

The highest dry matter yield per pot was obtained with CM (Figure 3-2b). Combining the application of CM and TSP provided a higher dry matter yield at rate 2 (equivalent of 17.6 kg ha'1 of P) compared to rate 1 at 0, 8.8, and 17.6 kg ha'1 P as TSP.

There was a strong relatioship between Mehlich-I extractable P and dry matter production (r2=0.88), and tissue P concentration and dry matter production (r2=0.79) (Figure 3-3). The relationship was linear for both indicators.

Several researchers have attributed the positive effect of organic amendments in improving P nutrition to a change in soil pH (Sanchez and Uehara, 1980) which improves the plant growth conditions and increases the solubility of native P. Soil pH increased approximately by one unit (4.55.5) in response to the application of PCW (Table 3-7). But this treatment produced the lowest dry matter. Application of Mucuna aterrima also improved the soil pH but the magnitude of improvement was less by 0.4 unit. The control treatment and CM had essentially the same pH, but CM had higher dry matter production.








55





Table 3-3. Selected chemical properties of the fine fraction (<2 mm) of the Xanthic Hapludox.


Depth Org. Chang & Jackson+ Langmui
carbon TP OP Al-P Fe-P Ca-P MI-P maxT.

cm g kg'-- -------------- Ag g -----------------0-15 14.6 200 25.2 51.6 0.3 0.3 2.4 550 15-30 12.2 120 14.5 18.5 2.8 0.2 1.8 620 30-45 8.3 90 6.7 5.2 0.1 0.1 1.2 789

+ pH of NH4F was adjusted to 8.2. The reductant soluble P is not included.
+ 0.01 M CaC12 was used as an electrolyte. OP = Organic phosphorus, TP = Total phosphorus MI-P = Mehlich I extractable P.





Table 3-4. Selected physical properties of the Xanthic
Hapludox.


Depth Bulk Fine fraction Hydraulic
Density Sand Silt Clay+ Conductivity

cm g cm"3 -- kg kg"I of <2 mm -- cm h"'

0-15 1.11 0.14 0.11 0.75 25.1 15-30 1.30 0.11 0.06 0.88 6.2 30-45 1.16 0.09 0.09 0.82 7.3

+ Fine sand fraction is also included.
















Table 3-5. Selected chemical properties of the fine fraction (<2 mm) of the Xanthic
Hapludox.


Depth pH Extractable bases ECEC Charqe+ Oxides+ Al Sat.
H20 KC1 Ca Mg K Al (-) (+) Fe Al (ECEC)

cm -------- cmol (+) kg ------- -- g 10g" -0-15 4.7 4.3 0.26 0.06 0.05 1.20 1.57 1.96 0.77 0.07 0.30 76.4
15-30 4.4 4.1 0.24 0.04 0.02 1.31 1.61 0.60 1.43 0.08 0.35 81.4
30-45 4.3 4.1 0.19 0.05 0.01 1.24 1.49 0.64 1.63 0.09 0.38 83.2


Acid ammonium oxalate extraction.
Acid-base potentiometric titration.












ON1









57
















Table 3-6. Analysis of variance for maize herbage dry weight
production per pot in the glasshouse study.


Source+ DF F Value Pr > F

REP 3 1.4 0.23 ORGANIC 2 970.2 0.00 RATE 1 189.2 0.00 ORGANIC*RATE 2 144.0 0.00 INRATE 2 127.9 0.00 ORGANIC*INRATE 4 17.6 0.00 RATE*INRATE 2 0.1 0.86 ORGANIC*RATE*INRATE 4 8.3 0.00 ERROR 51 CORRECTED TOTAL 71

CV% 13.0

+ REP = Replication, ORGANIC = Organic amendments, RATE = Rate of Organic amendments, INRATE = Rate of Inorganic P.









PCW CM 58
30 120
0
Y = 64.9 + 0.99X
Y = 13.83 + 0.30X ,
25 100 r2 =0.87 Rate2 ..'
r 2=0.76 Rate 1



0 6 . .ta) b)
Rate 1
o Y = 7.84 + 0.04X


10 - ARate2 40 Y =41.62 + 0.55X

A 12
Ar =0.86
5 ' ' 1 20
0 4.4 8.8 13.2 17.6 22.0 0 4.4 8.8 13.2 17.6 22.0 P, kg ha-1
Mucuna
100


r2 =0.88:










Figure 3-2. Effect of different rates of selected organic


amendments in combination with different rates of
Y = 35.1 + 0.97X

inorganic P on maize dry matter production at 650.87

30after planting in a glasshouse study.)
20
0 4.4 8.8 13.2 17.6 22.0

P, kg ha-1



Figure 3-2. Effect of different rates of selected organic
amendments in combination with different rates of
inorganic P on maize dry matter production at 65 d
after planting in a glasshouse study.








59








120
Y = 0.45 + 2.75P a) b) Y =-24.6 + 309.5P
r2=0.88 r 2=0.79
100



S80 A




S60-AA E A AAA





E **
AA
* AA


20
A

20 A A


0 I I . I . I I A I I A I

0 5 10 15 20 25 30 35
0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 Soil P, ug g' (M-1)
% Leaf P







Figure 3-3. Relationship between Mehlich-I extractable
soil P (a), and maize leaf tissue P concentration (b)
with maize dry matter yield.








60

















Table 3-7. Mean canges in soil pH (H20) following
application of selected organic amendments in the
glasshouse study


Amend+. Rate' Obs. Mean pH SD

Mucuna 1 12 5.09 0.22 Mucuna 2 12 5.47 0.20 CM 1 12 4.80 0.14 CM 2 12 4.87 0.26 PCW 1 12 5.48 0.19 PCW 1 12 5.85 0.23 Control 0 4 4.53 0.09 TSP 1 4 4.72 0.27 TSP 2 4 4.70 0.08

+ TSP = Triple superphosphate, CM = Chicken Manure, PCW =
Aerobically processed city waste.
� Rate 1 & 2 are equivalent to 8.8 and 17.6 kg ha-1 of P.








61

Field Study

Rainfall distribution data for both cropping cycles are presented in Figure 3-1c. The distribution is bi-modal with a short dry season. During this period evaporation is greater than precipitation.

Grain Yield. Single degree of freedom orthogonal

contrasts of maize grain production during the first crop (Table 3-8) indicated that one application of PCW equivalent of 26.4 kg P ha' produced only 1.41 Mg ha"1 of maize. This level of production was inferior to the control treatment. Canavalia ensiformis, Mucuna aterrima, and TSP when applied separately to provide 26.4 kg ha' of P produced the same amount of grain during first and second cropping cycles. But the CM treatment produced more grain than other treatments in both the first and second crops. This was expected based on the results of the adsorption study presented in Chapter II. One explanation is that the faster rate of decomposition (0.26 g/100 g d"') and the high P content in CM maintained the soil P concentration at a high level from the beginning of plant growth. In a review paper, Olsen and Barber (1977) concluded that an annual application of manure and superphosphate resulted in an increased level of 0.01M CaC12- and 0.5 M NaHCO3-extractable P. In most studies, manure treated soils tend to support a higher level of soluble P than soil treated with an equivalent amount of superphosphate.









62







Table 3-8. Orthogonal contrasts of maize grain yield under
different treatments applied in a 30 cm wide band
at UEPAE research station, Manaus, Brazil.


Treatment+ Mean Grain Yield
Cropl Crop2 Contrasts Cropl Crop2
-- Mg ha -M60 3.44 1.65 C60 VS TSP60 ns ns C60 3.56 2.07 CM60 VS TSP60 ** ** PCW60 1.41 1.67 PCW VS TSP60 ** ns CM60 4.63 2.37 M60 VS TSP60 ns ns M20+TSP20 3.55 1.41 C20+TSP20 VS TSP40 * ns C20+TSP20 3.62 1.81 M20+TSP20 VS TSP40 ns ns PCW20+TSP20 1.66 1.90 PCW20+TSP20 VS TSP40 ** ns CM20+TSP20 4.10 1.92 CM20+TSP20 VS TSP40 ** ns Control 1.71 1.10
TSP20 2.63 1.04
TSP40 3.12 1.46
TSP60 3.69 1.63

*, ** Significantly different at 0.05, and 0.01 level of probability. ns = not significantly different at 0.05 level of probability.

+ M60, C60, PCW60, and CM60 = Mucuna, Canavalia, Aerobically processed city waste, and Chicken manure applied to provide equivalent of 26.4 kg ha'1 of P.

TSP20, TSP40, and TSP60 = 8.8, 17.6, and 26.4 kg ha1' P from triple superphosphate (TSP).








63

Combined application of 8.8 kg P ha"I from organic amendments and the same amount from TSP giving an application rate of 17.6 kg P ha"' P was compared to 17.6 kg P ha"1 from TSP. Differential results relative to amendment source was observed. For the first crop, the combination of TSP with CM increased yield 30% relative to TSP. Mucuna aterrima did not influence yield while Canavalia ensiformis combined with TSP produced higher yield (3.6 compared to 3.1 Mg ha'1 for TSP). Chicken manure was still the best of all treatments during the second crop. But, PCW performance improved over the first crop. And there was no difference between PCW60 and TSP60, and PCW20+TSP20 combined. Such findings highlight the limitations of incubation studies conducted in the laboratory for a short span of time in predicting long term effects of organic amendments on soil P dynamics.

A response surface plot (Figure 3-4) for TSP indicated a linear response between grain yield and rate of TSP up to 26.4 kg P ha'I (r2 = 0.90). The magnitude of this response was declining in the second crop but the regression analysis did not show any difference with varying rate of TSP.

Change in soil phosphorus status. The change in P

status of the soil over the 240 d cropping cycle for all treatments, except PCW, followed a cubic surface response (Figure 3-5). The data presented are for 0-30 cm depth. All treatments improved soil P status compared to the control.







64







4.5


4.0 Y = 1.82 + 0.03P A r 2=0.90 Crop 1
_ 3.53.0




2.0
0.. Crop 2

1.5

0.1 P
1.0 Y = 1.0e r 2=0.41 ns
0.5
0 8.8 17.6 26.4 P, kg ha1







Figure 3-4. Rate of inorganic P applied through triple
superphosphate and its effect on maize yield in a
maize-maize rotation.








65


Trt. a bi b2 b R2

TSP60 4.79 5.53E-01 -6.30E-03 1.65E-05 0.75 M60 4.77 4.08E-01 -3.49E-03 7.58E-06 0.85 Control 4.05 -6.92E-03 -2.00E-04 7.18E-07 0.95
PCW60 4.07 3.96E-02 -3.08E-04 7.91E-07 0.15ns
C60. 3.62 3.02E-01 -2.56E-03 5.56E-06 0.99


25.0



210- x
M60
T
16.9







A8.8







0.7 1
0 47 93 140 187 233 280

Days





Figure 3-5. Effect of selected organic amendments, applied
in a quantity equivalent to provide 26.4 kg ha1 of P, on sustaining Mehlich-I extractable soil P pool in a
maize-maize rotation on a Xanthic Hapludox. M = Mucuna,
C = Canavalia








66

The actual P values ranged from 0.2 gg g-1 soil (control) to about 8.0 Ag g-1 soil for other amendments measured at the end of 240 d. This finding conflicts with the results obtained by Izza and Indiati (1982) where they studied the effect of various organic farm products on soil available P. They found that incorporation of these materials in high P fixing soil produced no effect on soil available P.

At 65 d TSP60 and M60 treatments had the same soil P status. There was a sharp decrease in soil P with TSP60 treatment as the cropping season progressed compared to M60 and C60. The plots which received PCW maintained low P which could not be described with the aid of any polynomial. When TSP20 was combined with M20 the P status improved compared to TSP40 (Figure 3-6). Application of inorganic P in the proximity of the organic amendment may have reduced the exposure of inorganic P to a larger soil volume leading to reduced fixation. And the decomposition of organic material may have inactivated active soil adsorption sites in the localized band. Increased effectiveness of soil amendments such as lime, in the proximity of organic matter has been reported by Ahmed and Tan (1988). Combined application of C20 with TSP20 was as good as TSP40, and PCW20 applied with TSP20 was inferior not only to TSP40 but to all treatments, except control.








67



Trt. a bI b2 t R2 TSP40 4.97 3.13E-01 -3.66E-03 9.57E-06 0.78 M20+TSP20 4.22 3.48E-01 -3.18E-03 7.14E-06 0.90 PCW20+TSP20 4.41 6.26E-02 -8.02E-04 2.24E-06 0.30ns TSP40 3.77 2.96E-01 -2.69E-03 5.92E-06 0.92


1B.0
x

5.2- a M20+TSP20 C2 TSP

12.3
O C/)

=9.5- TSP40


6.7



3.8 - PCW20+TSP20




0 47 93 140 B7 233 280 Days




Figure 3-6. Effect of selected organic amendments, applied
in combination with inorganic phosphorus source in a
quantity equivalent to provide 8.8 kg ha-' of P, on
sustaining Mehlich-I extractable soil P pool in a
maize-maize rotation on a Xanthic Hapludox.








68

Movement of Ca in the soil profile over the 240 d period indicated that application of TSP may reduce the soilCa pool (Figure 3-7). The level of Ca in the 0-30 cm depth was 2.30 cmol (+) kg-1. This value dropped to 0.50 cmol (+) kg"1 within 240 d. Maize plants from the control treatment plot also had the lowest uptake of Ca. Application of M60 and CM60 improved the Ca status in 0-15 cm. These materials after decomposition released cations which were an integral part of their composition (Larsen et al., 1972). Deep-rooting green manure crops offer the advantage of recycling cations which have been leached to a deeper soil profile. There is considerable evidence in the literature that organic ligands can hold polyvalent cations and prevent them from leaching (Moreno, 1960). Building a cation reserve is the matter of great importance in acid soils where soil chemical and physical conditions favor their rapid depletion through leaching.

High P concentration in leaf tissue reduced Zn and Cu concentration in leaf (Figure 3-8). This can be attributed to the chelation of these elements by fulvic acids (Saar and Weber, 1982). However, there is not enough evidence to conclude that at low leaf P concentration there will be a higher uptake of these metals. It has been shown that soils with high level of P are less likely to exhibit Cu toxicity (Sartain and Street, 1980).








69
0 0 0.5 1.0 1.5 2.0 2.5 0.0 1.0 2.0 3.0 4.0



a) b) so A Days

A A Days
0 120 180 240 0 120 180 240
45 - , a

Control CM60

E

.4
CL
0 0.5 1.0 1.5 2.0 2.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0


c)


30
Days Days
S A --.A-120 180 240
0 120 180 240 120 180 24
45-1" -F

M60 TSP60


60
Ca (M-1), cmol (+) kg '





Figure 3-7. Leaching of calcium from surface horizon of
highly leached Xanthic Hapludox as influenced by
different organic amendments.









25.0 70

Cu
20.0
00

15.0
O ED


10.0 0
O EIIIID OCDT
0 [2JJE D 0O


5.0
PP O O OO 0 0
0. , , , 0 ,

" 0.0

c,)
3 40.0

35.0 0 0
o D
Zn
30.0
O

25.0 - o oo0
000 O0 20.0 - 0 B ao D DO000 0 15.0 o0 o
OE O 00 0 . 0 0a 0 51.0 |0 o oo o o 0


5.0 - " I1 - .0 I 0
OO 0 O O 0 O 0

0.0 0.1 0.2 0.3 0.4 0.5 0.6

% P (leaf)



Figure 3-8. Relationship between leaf tissue P and
concentration of Zn and Cu in maize leaf tissue.








71

Conclusion

In the glasshouse experiment, among the three

amendments tested PCW produced the lowest dry matter per pot when applied alone or in combination with TSP. The highest dry matter production was obtained with the application of CM. A strong relationship was observed between dry matter production and Mehlich-I extractable P (r2 = 0.88).

In the field experiments, CM applied plots produced

more grain in first as well as second crops compared to the plots which had received Canavalia ensiformis, Mucuna aterrima, PCW, and TSP. This finding agreed with the prediction made from the laboratory incubation studies using silica as the matrix substratum to evaluate the decomposition rate of organic amendments. The prediction was made based on P adsorption maxima and binding energy values, and indicated that CM applied plots will produce superior yield compare to the plots which received other amendments.

When CM was applied in combination with TSP higher

grain yield was obtained compared to the same amount of P applied from TSP. All organic amendments improved the soil P reserve and reduced Ca leaching.














CHAPTER IV
RECOMMENDATION DOMAINS AND MODIFIED STABILITY ANALYSIS



Introduction


The problem of non-adoption of new technologies among small farmers in third world countries has been widely recognized, not as the farmers' unwillingness to accept change but rather as the inappropriateness of the technology to the real conditions that exist at the farm level in terms of the social, economic, and natural resource endowment (Shanner, et al., 1982). This chapter explores 1) techniques to understand farmers' circumstances, 2) the concept of recommendation domains (RDs), 3) the change in soil fertility and its impact on crop production, and 4) the use of modified stability analysis as a tool for the selection of appropriate technology for farmers with different production goals.



Research Data Base

An applied researchable problem should be based on farmers' immediate concerns. Such concerns are often documented through formal surveys with a structured questionnaire. Although this technique produces statistics 72








73

on farm input and output, it is expensive and time consuming. In order to make surveying cost-effective, Vander Veen and Mathema (1978) used key informants to gather baseline information. This speeded-up the process but it lacked interactive information.

Working for the farmers' cause requires direct involvement in problem identification. Rapid rural appraisals, often referred to as exploratory or diagnostic surveys, are a simple and relatively quick method of identifying constraints that operate in a defined area (Abalu, et al., 1987). The sondeo (Hildebrand, 1981) is one of the techniques of rapid rural appraisal which combines different disciplines in a team. Its primary purpose is to acquaint technicians with the area in which they are going to work. No questionnaires are used; rather informal interviews with farmers are conducted.



Recommendation Domains (RD)

In order to be cost effective, the research activities have to address the problems of and provide solutions for relatively large numbers of people. It is necessary, therefore, to classify farmers with similar circumstances into recommendation domains: groups of farmers for whom it is possible to make more or less the same recommendation (Perrin et al., 1978; Byerlee et al., 1980). More recently the concept of RDs has been extended to Research Domains and








74

Diffusion Domains (DDs) (Hildebrand, 1986). Many researchers, however, have considered RDs as a synonym for cropping systems (Zandstra, et al., 1981), farming types (Njobvu, 1986), and homogeneous groups (Moussie and Muhitira (1988). The limit set by the definition of Byerlee et al. (1980) has also been broadened to include agroclimatic zones and individual fields, in addition to farms. Criteria for Delineating Recommendation Domains

There is a great debate on the criteria used for

delineation of RDs. Socioeconomic criteria may be just as important as agroclimatic, and agroecological variables (Njobvu, 1986) in delineating domains. If so, the resulting domains are often not amenable to geographical mapping because farmers of different domains may be interspersed in a given area. Moussie and Muhitira (1988) attempted to classify farmers into relatively homogenous groups using cluster and discriminant statistical analysis. This method employed the use of qualitative information obtained from sondeos, and the selection of key variables from in-depth, formal surveys to obtain a discriminant function which helps to identify the most important variables in classifying farmers into RDs. Swinton and Samba (1986) used four agronomic criteria: average annual rainfall, soil fertility, soil texture, and depth to the subterranean water table for defining agricultural technology recommendation domains.








75

Tshabalala and Holland (1986) indicated that the "average" farmer is a myth and the programs designed to help him or her will fail, but used the RD concept in matching the improved technology to the group likely to be interested in taking advantage of it.

Economic criteria are also important in the delineation of RDs (Hildebrand and Poey, 1985). Improved technology often requires more cash and labor investments. Both resources are scarce on small subsistence family farms. Under limited cash availability, and so many competing uses for it, a farmer will consider an option which will give him or her the highest return per dollar invested. In this context, economic evaluation criteria such as net return ha'1, net return/total cost, and net return/cash cost play important roles.

The RD concept is being used in other disciplines (Fattori, 1990) and the use is being considered as an extension tool guiding the effective dissemination of technology appropriate to small farm conditions.


Stability Analysis

The stable performance of crop cultivars over a wide range of environmental conditions is generally regarded as desirable, but there is disagreement both on its definition and on the most appropriate methods for its statistical measurement from yield data trials (Becker, 1981; Hill and








76

Baylor, 1983). The most widely used stability analysis has been linear regression of cultivar yield on an environmental index derived from the mean of all or a subset of cultivars at each location or environment (Eberhart and Russell, 1966). Stability analysis has been used extensively to select genotypes that interact less with the environment and, therefore, are considered stable. Mackenzie, et al. suggested the use of stability analysis in predicting the response of a tested variety under different growing conditions for comparing the worthiness of selected potato clones for further replicated testing before naming and release. A regression coefficient of unity has been considered favorable for selection of stable genotypes. Hildebrand (1990) discussed the drawbacks of the use of the regression coefficient in selecting adaptable, or stable genotypes.

Hill and Baylor (1983) pointed out that perennial crop yields are usually measured on the same plot over a number of years, so problems with stability analysis may be encountered due to a differential change in the yields of the entries as the stand ages. They suggested an alternative orthogonal contrast analysis that partitions the variation over environmental components for each entry into sources due to environmental components (year, site, and management) and all possible interactions between these factors.








77

The reason Eberhart and Russell (1966) used the site mean as an environmental index was because a lack of knowledge of the relationship of macro-environmental differences such as temperature gradients, rainfall distribution, and soil types did not permit the computation of an index which could transform the environment into a continuous variable. However, attempts continue to quantify the production environment. Advancements in using the multivariate approach to group soils in the field based on variations in systematic and random components are forthcoming (Winding and Dress, 1983). Systematic variation is caused by difference in the parent material, relief and biological action as well as soil management practices such as fertilizer application and tillage. Random variation which is called "noise" by Burrough (1983), represents the statistical heterogeneity of the soil.

The need for assessing the factors causing soil

microvariability in the tropics has been stressed by Moorman and Keng (1978). In general, some soil characteristics are mutually correlated with each other (Norris, 1970). Hence, factors causing soil variation which are reflected in one or more of the soil characteristics may be used as criteria for grouping soil.

To analyze the causes of soil variation Kosaki and

Anthony (1989) applied principal component analysis (PCA), which is a mathematical technique used to summarize data and








78

investigate the relationship among variables. Variables employed for PCA in their study included soil pH, organic carbon, available P, exchangeable Ca, exchangeable Mg, exchangeable K, sand, silt, and clay. For the computation of principal components, they used a correlation matrix. However, soil variation is only one factor among many which influences the performance of a technology in a given environment. Much remains to be understood concerning the complicated interactions among agronomic, economic, social, and cultural variables which have a bearing on the performance of a technology. Modified Stability Analysis and Farming Systems

Farming Systems Research/Extension, is considered to be a dynamic, interactive, and problem oriented approach to develop technology for farmers, particularly those with limited resources. The technological base of FSR/E is onstation research but it constitutes only a part of the overall FSR/E program. The main activities are concentrated in farmers' fields with direct farmer involvement in technology evaluation and feedback. The farmers' field trials involve a few treatments, and often without replications. Farmer to farmer variation in management practices for a given experiment is not controlled. Instead, any unusual practice is recorded. Under these circumstances, modified stability analysis (MSA) (Hildebrand, 1984;








79

Upraity, et al., 1985, Fattori, 1990) has been used to select environment-specific production techniques. This technique does not depend on the concept that a regression coefficient of unity is always favorable for the selection of a technology. Adherence to this concept leads to rejection of superior technology for a specific environment in search of a 'stable' technology.



Objectives

The overall goal of this research was to identify

appropriate technology for maize and cowpea production by small farmers. The specific objectives were to: (i) measure the performance of selected treatments in different land types, (ii) study the changes in soil fertility parameters as influenced by the application of organic amendments, and (iii) develop location-specific recommendations.



Materials and Methods

Site Description

Two small farming communities (mean cultivated size = 3 ha/farm) in the municipality of Rio Preto da Eva, located in the state of Amazonas, Brazil were selected for on-farm experimentation. The area is accessible only by small motorboats. In trying to improve living conditions of these marginal farmers, the government of Brazil was just beginning a small watershed management program. The project








80

was intended to bring awareness among farmers to preserve the environment and assist them in improving their agricultural production. The Brazilian national agricultural research institution (EMBRAPA) has a mandate of developing appropriate technology for different farming conditions in this relatively inaccessible area. Developing a Research Base

Secondary information regarding indigenous farming

practices of the area was collected from published sources. A rapid appraisal of the area was conducted with a multidisciplinary team of scientists participating from various research disciplines and state planning and agriculture extension organizations who visited the area on three different occasions to collect and verify information obtained in group discussions or during individual communication with farmers. Farmers' knowledge of indigenous technology, agronomic practices, and land types being used were recorded. An extensive soil sampling program was carried out to understand soil physical and chemical characteristics and relate them to farmers' rationale for assigning a particular cropping pattern to a given land type. Farmers played an active role in technology design, execution and evaluation.








81

Selection of Treatments

Three treatments from previous on-station research were selected for comparison with farmers practices' (FP) for growing maize (Zea mavs L.) and cowpea (Vigna unguiculata) (Table 4-1). Several locations were identified to encompass the different land types within the soil family limit of Clayey, Isohyperthermic, Xanthic Hapludox. A shift in maize planting date was agreed upon so maize maturity would coincide with the beginning of the dry season (Fig. 4-1). For the maize crop, all treatments except FP received 100 kg ha'1 of N from urea half applied broadcast before planting and the other half in two additional split applications. All K (60 kg ha'1) was applied basal broadcast except to the FP plots. Processed city waste, CM, and TSP were applied in 25 cm bands. The maize variety BR-5110 was planted in rows 80 cm apart. No nitrogen was applied to the cowpea crop. The cowpea variety IPEAN V-69 was planted in rows 60 cm apart. Plot size for maize and cowpea varied from 100-200 square meters. The organic amendments were applied in a 20 cm band. Both crops were harvested for grain. For growing maize and cowpea the main differences between FP and the improved practice was that the improved practice received fertilizer. The land preparation and planting methods consisted of clearing the area by slash and burn, followed by manual land preparation and planting with sticks. This is a common FP in the area.








82

The trial was planted in a RCB design with two

replications wherever possible. Some locations had only one replication because of limited area allocated by farmers for the trial.



Table 4-1. Application of N, P, and K in different
treatments tested in on-farm experimentation for maize
and cowpea crops in the municipality of Rio Preto da
Eva, Amazonas, Brazil.


Maize Cowpea
Treatment N P K N P K
--------- kg ha ---------FP+ 0 0 0 0 2+ 0
CM2O+TSP20 100 8.8+8.8 60 0 8.8+8.8 60 PCW20+TSP20 100 8.8+8.8 60 0 8.8+8.8 60 TSP40 100 8.8+8.8 60 0 0+17.6 60

+ No restriction was imposed on the amount of nutrient or type of cultural practice to be used. This treatment varied from farm to farm.
+ the range for P was 0-4 kg ha'. The value of P reported is the mean.


Soil samples were taken before planting, during maize and cowpea flowering, and after harvest. All soil samples were analyzed for pH, Ca, Mg, Al, K, and P. Particle size analysis was conducted only on samples taken before planting. Plant tissue samples were analyzed for P, Zn, Cu, and Mn. Their results are used in Chapter III. The analytical techniques for soil and plant tissue samples were essentially the same as described in Chapters II, and III.








83








Effective Rainfall, mm 400

300

200

100

0


-100a) (b)

-200
A'88 S O N D J'89 F M A M J J A Months










Figure 4-1. Effective rainfall and suggested shift in the
planting date for maize in Rio Preto da Eva, Amazonas,
Brazil. (a) common practice, (b) suggested practice.








84

Statistical Analysis

Combined analysis of variance (CANOVA). A combined

analysis of variance was conducted for replicated trials. A preliminary analysis for homogeneity of variances from the individual ANOVAs was based on the chi-squared test. In the case of a significant chi-squared test, locations with a coefficient of variation >20% were excluded from the CANOVA (Gomez and Gomez, 1985). No attempt was made to employ a missing plot technique to recover lost data or data which violated some assumptions of the ANOVA. It is common to lose part of on-farm experiments to destruction of experimental plants, loss of harvested samples, etc. Data from two locations for maize and one for cowpea were lost due to misunderstanding with farmers about the harvest date. The trials were harvested before the agreed upon date. At three locations (2 for maize and one for cowpea), the crop failed due to late planting. Data from one location for cowpea could not be used for ANOVA because the second harvest yield for all treatments was mixed without recording production separately.

Stepwise reQression. Stepwise regression is a method for either the forward, or backward selection of variables based on an F statistic of R2 significance at the level specified in the model (SAS Institute, Inc., 1985). This technique was used to identify factors responsible for the wide range of variation in maize and cowpea production over








85

locations. For this purpose, yield obtained from each treatment over locations was stepwise regressed on the values of soil pH, ECEC, Mehlich-I extractable-P, and Al saturation. These variables were measured from soil samples taken during tasseling stage for maize and flowering stage for cowpea (Hanway, 1967).

Modified stability analysis. The production

environments were separated based on the average yield of all treatments at each location (Eberhart and Russell, 1966). In this way, environment becomes a continuous, quantifiable variable. Yield for each treatment was related to environment by simple linear regression.

Yij = a + be

where Yij = yield from treatment i at the jth location, and

ej = jth environmental index.

The use of the regression coefficient 'b' in the linear equations with a value near one to select the "stable" treatment over all environments was avoided. Instead superior treatments were identified for groups of environments, or RDs (Hildebrand, 1984).

The distribution of confidence intervals for the treatments within RDs for both crops was calculated as follows:

CI = Y � taS/n"

Where;








86

Y = the mean treatment yield, or other criterion within the RD,

a = the level of confidence,

t, = value from a two-tailed "t" table,

S = [Ex2/(n-1)] or standard deviation of treatment

yield, or of other evaluation criterion within the RD, and

n = no. of locations or environments in each recommendation domain.

The CI test was used to provide an assessment of the risk of low yields and unacceptable levels of other evaluation criteria within each RD.



Results and Discussion

The project area is inhabited by subsistence farmers who practice a slash-and-burn, rotational type of agricultural system and market surplus production and products gathered and hunted in the forest. The two major land types used for crops are: (i) area cleared for the first time by slash-and-burn from primary forest (PF), and

(ii) area cleared from secondary forest which was left in fallow 5-7 years (SF). For this trial, a third type was used, (iii) area considered undesirable for agricultural activities (WL). The chemical and physical properties of these soils are presented in Tables 4-2 and 4-3. All sampled soils were acidic (median pH H20 = 4.5) with very low ECEC




Full Text
Confidence Interval, %
116
Yield, Mg/ha
Figure 4-7b. Distribution of confidence intervals for cowpea
production in poor (e<1.32 mg ha'1), and good (e>1.32
Mg ha'1) environments.


41
plant material resulted in immobilization of soil P only in
the low P soil in the absence of plants. In the high P soil
no immobilization of P was observed.
Crop residues have an effect on the nutrient status of
the soil. The cumulative effects of increasing quantities of
organic residues on available nutrients in soil were studied
for 11 years by Larson et al. (1978). They reported that
addition of 16 tons/ha of plant residue per year to the soil
increased the amount of N, S and P by 37, 45 and 14%
respectively, over the control treatment. They also found
that the NH4-N production, weak acid soluble P and
exchangeable K in the soil were increased as a result of
increasing the addition of organic residues.
Solubilization of Mn and Fe in soil were affected by
wheat straw and alfalfa amendments (Elliot and Blaylock
1975, Sims, 1986) The release was greater for Mn than Fe
and also much higher at 30 kPa than 50 kPa moisture
tension. The release of Mn and Fe from the soil column
followed the following order: alfalfa > wheat straw > soil
alone. They suggested that the potential accumulation of
soluble Mn in well drained soil was possible where there was
a large quantity of plant residues incorporated into the
soil.
Application technique. In the case of organic manures,
changes in organic P will to some extent depend on whether
the material is left on the surface of the soil or is plowed


54
The highest dry matter yield per pot was obtained with
CM (Figure 3-2b). Combining the application of CM and TSP
provided a higher dry matter yield at rate 2 (equivalent of
17.6 kg ha1 of P) compared to rate 1 at 0, 8.8, and 17.6 kg
ha1 P as TSP.
There was a strong relatioship between Mehlich-I
extractable P and dry matter production (r2=0.88), and
tissue P concentration and dry matter production (r2=0.79)
(Figure 3-3). The relationship was linear for both
indicators.
Several researchers have attributed the positive effect
of organic amendments in improving P nutrition to a change
in soil pH (Sanchez and Uehara, 1980) which improves the
plant growth conditions and increases the solubility of
native P. Soil pH increased approximately by one unit (4.5-
5.5) in response to the application of PCW (Table 3-7). But
this treatment produced the lowest dry matter. Application
of Mucuna aterrima also improved the soil pH but the
magnitude of improvement was less by 0.4 unit. The control
treatment and CM had essentially the same pH, but CM had
higher dry matter production.


91
Table 4-4. Relationship between soil characteristics with
year in crop production in different land types.
Land Type^
Soil
Characteristics
Intercept
Slope
r2
Primary
pH
5.25
-0.15
(0.12)
0.14
Forest (PF)
ECEC
3.60
-0.80
(0.38)
0.33*
A1 Sat. (%)
61.1
6.5
(0.6)
0.18
Mehlich-I (P)
10.3
-2.15
(1.29)
0.23
Secondary
pH
4.75
-0.17
(0.09)
0.38*
Forest (SF)
ECEC
2.33
-0.16
(0.15)
0.03
A1 Sat. (%)
81.5
4.0
(1.6)
0.20
Mehlich-I (P)
8.5
-2.60
(0.72)
0.72*
-f No. of observations in PF were 11, and in SF 7.
Numbers in the parenthesis are standard error of slope
estimates.


63
Combined application of 8.8 kg P ha1 from organic
amendments and the same amount from TSP giving an
application rate of 17.6 kg P ha 1 P was compared to 17.6 kg
P ha1 from TSP. Differential results relative to amendment
source was observed. For the first crop, the combination of
TSP with CM increased yield 30% relative to TSP. Mucuna
aterrima did not influence yield while Canavalia ensiformis
combined with TSP produced higher yield (3.6 compared to 3.1
Mg ha'1 for TSP) Chicken manure was still the best of all
treatments during the second crop. But, PCW performance
improved over the first crop. And there was no difference
between PCW60 and TSP60, and PCW20+TSP20 combined. Such
findings highlight the limitations of incubation studies
conducted in the laboratory for a short span of time in
predicting long term effects of organic amendments on soil P
dynamics.
A response surface plot (Figure 3-4) for TSP indicated
a linear response between grain yield and rate of TSP up to
26.4 kg P ha1 (r2 = 0.90). The magnitude of this response
was declining in the second crop but the regression analysis
did not show any difference with varying rate of TSP.
Change in soil phosphorus status. The change in P
status of the soil over the 240 d cropping cycle for all
treatments, except PCW, followed a cubic surface response
(Figure 3-5). The data presented are for 0-30 cm depth. All
treatments improved soil P status compared to the control.


14
extraction with CaCl2 was considered total P released from
the amendments because of the inability of silica to adsorb
P. This value was used to construct a P release curve for
each amendment.
Calculation of total P adsorbed. Total adsorbed P was
calculated as follows:
TAP = N + S + OA
p p p
where:
TAP = total adsorbed P on soil surface, nq g1
Np = native P present in adsorbed phase nq g1
Sp = applied inorganic P in adsorbed phase /ng g"1
0Ap = phosphorus from organic amendment in adsorbed phase
Mg g'1.
The amount of P released from the organic amendments and
adsorbed on the soil surfaces (0Ap) was estimated by SSMT.
The native exchangeable P on control samples was determined
by the method of least squares (Reddy, 1990). For comparison
purposes, P adsorbed on the soil surface following
incubation with organic amendments was also determined by
sequential extraction with 0.01 M CaCl2.


ugg1 (Leaf)
25.0
Cu
20.0

15.0

m
10.0
5.0
0.0
111 rm cu
2dd d
T]
O ~ffi i i i i i i i
L
n
Hi ittA 11 rrn
l
Figure 3-8. Relationship between leaf tissue P and
concentration of Zn and Cu in maize leaf tissue.


CHAPTER IV
RECOMMENDATION DOMAINS AND MODIFIED STABILITY ANALYSIS
Introduction
The problem of non-adoption of new technologies among
small farmers in third world countries has been widely
recognized, not as the farmers' unwillingness to accept
change but rather as the inappropriateness of the technology
to the real conditions that exist at the farm level in terms
of the social, economic, and natural resource endowment
(Shanner, et al., 1982). This chapter explores 1)
techniques to understand farmers' circumstances, 2) the
concept of recommendation domains (RDs), 3) the change in
soil fertility and its impact on crop production, and 4) the
use of modified stability analysis as a tool for the
selection of appropriate technology for farmers with
different production goals.
Research Data Base
An applied researchable problem should be based on
farmers' immediate concerns. Such concerns are often
documented through formal surveys with a structured
questionnaire. Although this technique produces statistics
72


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.
Sartain, Chairman
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 of Doctor of Philosophy.
Peter E. Hildebrand
Professor of Food and Resource
Economics
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.
Donald A. Graetz
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 of Doctor of Philosophy.
-Arfl oulgA
Edward A. Hanlon
Associate Professor of Soil
Science


136
APPENDIX B
ECONOMIC ANALYSIS FOR MAIZE
Yield, income, and cost from the maize
trials, Rio Preto da Eva, Manaus, Brazil. (US$)
Yield Gross Total Net Net Income/
Mg ha"1 Income Cost Income Total Cost
FP
0.15
30
18
12
0.7
0.00
0
12
-12
-1.0
0.00
0
12
-12
-1.0
0.25
50
22
28
1.3
0.15
30
18
12
0.7
2.20
440
100
340
3.4
2.50
500
112
388
3.5
0.20
40
20
20
1.0
PCW20+TSP20
0.15
30
213
-183
-0.9
1.10
220
251
-31
-0.1
0.00
0
207
-207
-1.0
1.10
220
251
-31
-0.1
0.70
140
235
-95
-0.4
1.00
200
247
-47
-0.2
1.40
280
263
16
0.1
0.70
140
235
-95
-0.4
TSP40
1.30
260
150
110
0.7
3.40
680
234
446
1.9
0.15
30
104
-74
-0.7
1.60
320
162
158
1.0
3.40
680
234
446
1.9
4.20
840
266
574
2.2
4.50
900
278
622
2.2
3.50
700
238
462
1.9


CHAPTER II
ORGANIC AMENDMENTS AND PHOSPHORUS ADSORPTION
ISOTHERMS
Soil Constituents and Phosphorus Adsorption
It is widely believed that hydrous oxides of Fe and Al,
and calcium carbonate play key roles in P retention.
Although controversy exists among researchers about the
mechanism of P retention by these compounds (Muljadi et al.,
1966; Hingston et al., 1967; Rajan et al., 1974), the phasic
nature of P adsorption has been well recognized by several
researchers (Bache, 1964; Munns and Fox, 1976). The first
phase of adsorption is due to a high energy chemisorption of
small amounts of P on the soil surface. The second phase is
comprised of a precipitation reaction followed by a low
- >
energy sorption of P onto the precipitate.
Phosphorus Adsorption Isotherms
Phosphorus adsorption isotherms have been used
extensively for describing the P adsorption characteristics
of various soils and in estimating the P requirement of
different crops (Fox and Kamprath, 1970; Jones and Benson,
1975; Mokwunye, 1977; Singh and Jones, 1977; Solis and
Torrent, 1989). The isotherm technique involves
5


45
amendments with inorganic P in a band will reduce direct
contact of inorganic P with a large volume of soil.
In this context, the overall objective of this research
was to devise a technique to sustain P nutrition in a highly
leached Oxisols with the help of organic and inorganic P
sources applied in narrow bands. The specific objectives
were to: (i) study the effect of different amendments and
their rate of application on maize dry matter production and
herbage nutrient concentration, (ii) examine the beneficial
effect of the combined application of organic amendment with
inorganic P in a narrow band as measured by maize grain
production and soil nutrient dynamics, and (iii) measure the
residual effect of applied treatments in relation to
selected soil chemical characteristics.
Materials and Methods
This experiment was conducted at the Empresa Brasileira
de Pesquisa Agropecuaria (EMBRAPA) station located 30 km
north of Manaus at an elevation of 50 m in the Amazonas
state of Brazil (Fig. 3-1). Climate in the Manaus region has
been classified in the Koppen nomenclature as afi, tropical,
humid and hot (Goes and Ribeiro, 1976). The soil used in the
study has been classified as Xanthic Hapludox (clayey
kaolinitic, isohyperthermic) (EMBRAPA, 1979).


60
Table 3-7. Mean canges in soil pH (H20) following
application of selected organic amendments in the
glasshouse study
Amende.
Rate5
Obs.
Mean pH
SD
Mucuna
1
12
5.09
0.22
Mucuna
2
12
5.47
0.20
CM
1
12
4.80
0.14
CM
2
12
4.87
0.26
PCW
1
12
5.48
0.19
PCW
1
12
5.85
0.23
Control
0
4
4.53
0.09
TSP
1
4
4.72
0.27
TSP
2
4
4.70
0.08
TSP = Triple superphosphate, CM = Chicken Manure, PCW =
Aerobically processed city waste.
§ Rate 1 & 2 are equivalent to 8.8 and 17.6 kg ha'1 of P.


39
Management of Organic Amendments
Organic amendments can affect the reaction of P in the
soil through complexation of polyvalent cations which are
major phosphate adsorption sites. It is widely believed that
humus in association with cations such as Fe3+, Al3+, and Ca2+
retains significant amounts of P. Appelt et al. (1975)
prepared a hydroxy-Al-humic acid complex that adsorbed P
because of the creation of new P adsorption sites. They
concluded that any increase in organic content of a soil
could lead to greater adsorption. A study conducted by Swift
and Haynes (1989) also showed that Al-organic matter
associations have a significant phosphate adsorption
capacity. Indeed, the Al-humate adsorbed amounts of
phosphate similar to those commonly reported for Al and Fe
oxides (McLaughlin et al., 1981) on a w/w basis. However,
simple organic acids, fulvic acids, and humic acids had no
effect on P adsorption by volcanic ash-derived soils. For
these soils, P was preferentially adsorbed over the organic
acids studied (Appelt et al.,1975).
The addition of organic materials to high P fixing soil
can decrease, increase, or leave virtually unaltered the P
fixation capacity (Yuan, 1980). The reduced fixation is the
result of: 1. complexation of Fe, Al, and Ca by organic
anions (Larsen et al, 1959), 2. competition of organic
anions and P for the same adsorption sites (Nagarajah et al,
1970), 3. development of organic coatings on mineral surface


13
Table 2-1. Ionic species and concentration of nutrient
solution used with silica to simulate solution
chemistry of a Xanthic Hapludox in a phosphorus release
study by organic amendments.
Source Species Conc^.
No. Chemical mmol (charge) L
1.
Ca(N03)2.4H20
Ca2+
2.690
nh4no3
nh4+, no3
9.180
2.
KC1
K++
1.829
k2so4
Cl'
0.670
kno3
S
0.480
3 .
Mg(N03)2.6H20
Mg2+
0.650
4 .
FeHEDTA
Fe3+
0.035
5.
MnCl2.4H20
Mn2+
0.007
h3bo3
B
0.019
ZnS04.7H20
Zn
0.0018
CuS04.5H20
Cu
0.0005
Na2Mo04.2H20
Mo
0.0006
+ PH was adjusted to 5.0 with 0.1M HC1. The reported
concentration is considered adequate for maize growth in a
solution culture.
4= The final concentration for K is calculated from KC1,
K2S04, and KN03.


Table 2-2. Chemical composition of organic amendments used in the incubation
study.
0A+ Ca Mg K C P N C: P C:N Cu Zn Mn
g lOOg'1 M9 g 1
Canav.
2.27
0.17
1.90
39.6
0.31
3.95
128
10
2.5
8.4
174.0
Peanut
1.44
0.27
1.95
38.5
0.33
3.32
116
11
2.5
8.9
70.0
Maize
0.35
0.19
2.25
39.6
0.21
2.93
189
13
5.0
40.1
95.0
Teph.
1.01
0.19
1.75
43.9
0.29
3.31
151
13
2.5
26.1
41.0
Mucuna
1.05
0.15
1.80
42.5
0.28
3.34
152
12
2.5
28.0
103.0
Grass
0.70
0.25
1.60
34.1
0.19
1.43
179
23
5.0
36.9
123.0
CM
7.62
0.51
2.10
17.8
2.77
1.46
6
12
92.5
102.9
399.0
PCW
1.57
0.14
0.35
22.3
0.16
0.95
139
23
155.0
280.8
213.0
Puer.
0.45
0.18
1.55
42.1
0.28
2.52
150
16
2.5
24.3
212.0
Cowpea
1.22
0.16
2.20
41.9
0.31
2.75
134
15
2.5
19.2
98.0
-f- Organic amendments; Mucuna aterrima (Mucuna) Pueraria phaseoloides (Puer.) ,
Canavalia ensiformis (Canav.), Zea mays (Maize), Arachis hvpoqaea (Peanut), Tephrosia
candida (Teph), Viqna unquiculata (Cowpea), mixed gramineae (Grass), aerobically
digested city waste (PCW) and chicken manure (CM).


19
Adsorption Isotherms
The results of adsorption and desorption studies are
presented in Fig. 2-2 and 2-3. Incubation of the soil with
amendments for 35 d influenced P adsorption. This was more
pronounced at higher equilibrium solution P concentrations
(Fig. 2-2a) At a P concentrations >1.9 /ug mL'1, PCW amended
soil adsorbed more P than the unamended soil. A similar
result was obtained by Singh and Jones (1976) for organic
amendments with low P content and high C:P ratio. On the
other hand, soil samples amended with CM and Mucuna adsorbed
less P than the unamended soil. As the time of incubation
increased to 150 d, all amendments reduced P adsorption
compared to the control (Fig. 2-2b). A five-fold reduction
in P adsorption was observed for the CM treatment at an
equilibrium P concentration of 0.3 jug g1 soil. The
reduction was perhaps due to net mineralization of P even
from P-poor organic amendments with time. Another factor in
P mineralization is the C:P ratio. When this ratio remains
less than about 200:1, immobilization predominates during
the initial stages of decomposition. But, as the
decomposition proceeds, this ratio becomes narrower due to
the concentration of P in decomposing residue and continuous
degradation of C by microorganisms.
As plants deplete soil solution P, the solution must be
continuously recharged, if good growth is to be maintained.
Recharge occurs when P is desorbed from the soil surface,


representing the author at UF during the author's one year
stay in Brazil.
The author is indebted to his beloved wife for her
encouragement, support, and patience.
v


P adsorbed (ug/g soil)
Adsorption
21
Figure 2-2. Phosphorus adsorption isotherms following 35 d
(a) and 150 d (b) of soil incubation with Mucuna aterrima
(Mucuna), Chicken Manure (CM), Aerobically Processed City
Waste (PCW), and Control.


P adsorbed (ug/g soil)
Desorption
22
Solution P (ug mL!)
Figure 2-3. Phosphorus desorption isotherms following 35 d
(a), and 150 d (b) of soil incubation with Mucuna
aterrima (Mucuna), Chicken Manure (CM), Aerobically
Processed City Waste (PCW), and Control.


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.
Kenneth L. Buhr
Assistant Professor of
Agronomy
This dissertation was submitted to the Graduate Faculty
of 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.
AUGUST 1990
'We
Dean,/College of Agriculture
Dean, Graduate School


128
practice. However, FP can not be continued for more than a
year due to adverse effect on soil properties. The aluminum
saturation was increasing each year in PF by 7%, and
Mehlich-I P was decreasing by 2.1 ¡ig g'1 soil annually.
For cowpea, CM20+TSP20 outperformed all treatments in
poor environments (eel. 32 Mg ha1) while in good
environments (RD2) (e>1.32 Mg ha1) TSP40 had clear
advantage over other treatments based on all evaluation
criteria. Farmer practice had the highest return per dollar
compared to other treatments in RD2 (37.7 $/$ ). The
delineation of RDs was not amenable to geographical mapping
because fields of different domains were interspersed in the
area. A change in evaluation criteria lead to change in the
demarcation of RDs.


114
The Cl calculation (Fig. 4-10a-b) indicated that
CM20+TSP20 gave higher, and stable return in poor
environments compared to other treatments. At 90% Cl, FP
produced from $ 0 to 2.2 compared to $ 2.8 to 3.8 for
CM20+TSP20. But FP was the best performer in good
environments and provided a return of $ 4.2 to 5.8 for each
$ invested in total cost.
Net income. The value of net income ha1 from different
environments was regressed against e (Fig. 4-8c). The net
income was higher for TSP40 treatment in good environments.
But CM20+TSP20 was superior to TSP40 in poor environments.
As the environment for cowpea cultivation improved the net
income from FP was also increasing. Processed city waste
treatment (PCW20+TSP20) was always inferior to other
treatments. Confidence intervals were calculated for FP,
TSP40, and CM20+TSP20 (Fig. 4-lla-b).
Based on this criterion FP was inferior to TSP40 in
good environments, and to CM20+TSP20 in poor environments.
At 90% Cl the value of net return for TSP40 fell in the
range of $ 820-1080 ha'1 compared to $ 300 to 750 for FP.


Confidence Coefficient (%)
Net Income/Cash Cost (US $)
Net Income/Cash Cost (US $)
Figure 4-9. Distribution of confidence intervals for net
income/cash cost for selected treatments in poor
(e<1.32 Mg ha 1) and good (e>1.32 Mg ha1) environments
for cowpea cultivation.


42
in. Douglas et al. (1980) reported that the method of
placement, composition of residues and loading rates were
important factors influencing mineralization or
immobilization of N and S. A higher mineralization rate of N
from residue incorporated at 4 cm soil depth as compared to
the surface was also observed by Brown and Dickey (1970)
and Cocharan et al. (1980). Data on P mineralization from
organic matter in the literature is scarce.
Inorganic Phosphorus Management Strategy
Salinas and Sanchez (1976) have outlined a three-point
strategy for P management under limited resource conditions.
1. Use of cheaper sources of P. Two main sources are
phosphate rocks (PR) and thermally altered sources, such as
basic slags and the Rhenania phosphates. Numerous reports
have appeared in the literature regarding the fertilizer
value of PR as compared to other sources of P fertilizers,
e.g., superphosphate (Khasawneh and Doll, 1978; Hammond et
al., 1986; Hernandez and Sartain, 1985). Recently, Heliums
et al. (1989) compared the potential agronomic value of some
rock phosphate from South America and West Africa and
concluded that in addition to P the PRs with medium to high
reactivity have a potential Ca supply value. 2. improved
soil test interpretations, and 3. improved placement
methods.


36
released P in the silica matrix independent of their P
content and C:P ratio.
The use of SSMT to measure release of P from
decomposing organic amendments aid in the calculation of
adsorption maxima compared to the seqential extraction of
amended soil with 0.01 M CaCl2 solution. The calculated
values were in close agreement with actual adsorption
measured at high equilibrium P concentration.


118
Net Income/Cash Cost ($/$)
0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2
Environmental Index (e), Mg/ha
Net Income, $/ha
Environmental Index (e), Mg/ha
0.6
2.2


35
concentration (60 /g g 1 soil) which indicated the validity
of this technique in predicting P requirements of amended
soils. In highly weathered soils, extractable P is usually
low but the amount of P which the soil can immobilize varies
greatly because of the variation in the reactive surfaces.
Addition of organic amendments changed the reactive soil
surface as indicated by the change in adsorption maxima and
bonding energy.
Conclusion
Decomposition of organic amendments influenced P
adsorption and desorption by soil. Amendments with low P
content (PCW, maize, and grass) immobilized P during early
soil samples incubated with CM. Soil incubated with Mucuna,
PCW, and the control for 35 d did not show any difference in
P desorption. However, desorption was higher for all
amendments incubated with soil for 150 d. Incubation of soil
with low P-containing organic materials may not influence
soil P desorption characteristics initially.
Amendments' P content and C:P ratio did not prove to
be helpful in predicting P mineralization. Amendments with a
C:P ratio of 1:139 (PCW) immobilized P during their
decomposition while others with similar or even higher C:P
ratios had a net P release in soil matrix. All amendments


6
equilibration of a known amount of soil for a limited time
in a KC1, NaCl or CaCl2 solution containing various amounts
of K2HP04, NaH2P04 or Ca(H2P04)2 (Olsen and Watanabe, 1957;
Syers et al., 1973; Singh and Jones, 1976). Phosphorus
removed from solution is considered to be adsorbed. Plant P
requirement is calculated based on the amount of P needed to
bring the concentration of supernatant solution to a
specified concentration (generally 0.2 nq mL"1) .
Langmuir equation. A number of researchers have
investigated soil P adsorption characteristics using the
Langmuir equation (Olsen and Watanabe, 1957; Woodruff and
Kamprath, 1965; Gunary, 1970; Borggaard, 1983). A frequently
used linear form is:
(c/x/m) = (1/kb) + (c/b)
Where:
x/m = amount of P adsorbed per unit weight of soil
b = the adsorption maxima
c = the equilibrium P concentration, nq mL"1
k = a constant related to the bonding energy of the
adsorbent for adsorbate.
For a given uniform population of sites, the value of
adsorption maxima can also be evaluated by plotting x/m vs.
x/m/c, obtained from a rearranged Langmuir equation commonly
referred to as the 1Eadie-Hofstee plot (Syres et al.,
1973) The underlying assumptions in each case are that
adsorption sites on the particle surface are uniform, and


76
Baylor, 1983). The most widely used stability analysis has
been linear regression of cultivar yield on an environmental
index derived from the mean of all or a subset of cultivars
at each location or environment (Eberhart and Russell,
1966). Stability analysis has been used extensively to
select genotypes that interact less with the environment
and, therefore, are considered stable. Mackenzie, et al.
suggested the use of stability analysis in predicting the
response of a tested variety under different growing
conditions for comparing the worthiness of selected potato
clones for further replicated testing before naming and
release. A regression coefficient of unity has been
considered favorable for selection of stable genotypes.
Hildebrand (1990) discussed the drawbacks of the use of the
regression coefficient in selecting adaptable, or stable
genotypes.
Hill and Baylor (1983) pointed out that perennial crop
yields are usually measured on the same plot over a number
of years, so problems with stability analysis may be
encountered due to a differential change in the yields of
the entries as the stand ages. They suggested an alternative
orthogonal contrast analysis that partitions the variation
over environmental components for each entry into sources
due to environmental components (year, site, and management)
and all possible interactions between these factors.


15
Results and Discussion
Organic amendments differed in chemical composition
(Table 2-2). The highest P concentration was observed in CM
(2.77 g 100 g'1) and the lowest in PCW (0.16 g 100 g1). The
highest C:P ratio was for maize (189.0) followed by grass
(180.0).
Soil used was obtained from the Ap horizon of a highly
leached oxisol classified as Xanthic Hapludox (clayey
kaolinitic, isohyperthermic) (EMBRAPA, 1979). The soil was
acidic (pH = 4.5), and low in total P (170 /g g'1 soil) and
Mehlich-I extractable P (3.3 iiq g'1 soil). Clay content
ranged from 78-82% and the A1 saturation was 78.4% (Table 2-
3). The point of zero net charge measured by potentiometric
titration in CaCl2 solution was at pH 4.2 (Fig. 2-1). The
native exchangeable P measured by the least sguares method
was 3.0 nq g"1 soil.


126
C O W P E A
Recommendation Domain I
Recommendation Domain II
Figure 4-13. Recommendation domains, based on net return to
cash cost, and their relationship to land types in the
municipality of Rio Preto da Eva, Amazonas, Brazil.


8
bonding energy term for P sorption (Syere et al., 1973). The
values of adsorption maxima and bonding energy can be
related to various soil properties which will supply
information about the nature of the reaction between soil
and fertilizer P, and can aid in the prediction of plant
available P (Olsen, 1953).
Correction for Initial Surface Phosphate
Olsen and Watanabe (1957) have demonstrated the effect
of correction for initial surface phosphate on the constants
derived from the Langmuir isotherm. For Pierry clay and
Owyhee silt loam soils the correction increased the
adsorption maxima and the bonding energy constant. Ideally,
the adsorption would be determined in a system in which the
surface is free of the adsorbate ion. Usually this
restriction is not feasible. Therefore, a correction for
initial P is made by adding the amount of surface P
determined by a separate analysis. One of the commonly used
methods to measure native adsorbed phosphate is by isotopic
exchange (Olsen and Watanabe, 1957; Holford et al., 1974).
This method involves shaking a soil sample in an electrolyte
to which carrier-free 32P is added. An aliquot is taken
after a specified time and 31P is determined. Reddy (1990)
has suggested a least squares fit for the determination of
native P adsorbed on the soil surface. He suggested the
following calculations:


3
(Jones and Benson, 1975; Solis and Torrent, 1989). However,
the techniques used to construct P adsorption isotherms in
the presence of amendments are similar to those used for
soil without amendment. Generally a correction for
preadsorbed P is introduced (Mokwunye, 1977; Solis and
Torrent, 1989). An assumption is made that adsorption and
desorption are equal, which is not true (Sample et al.,
1980). Therefore, it leads to unreliable prediction of crop
P requirement and residual effect of applied P.
Organic and Inorganic P Interaction
A larqe amount of orqanic amendment will be required to
provide the total P requirement for a crop. An efficient way
to use a suboptimal dose of an organic amendment is band
application with inorganic P. Such an application reduces P
immobilization from P-poor amendments, and eliminates P
deficiency in the early stages of crop growth. Much remains
to be learned about the interaction of organic amendments
with inorganic P. Most laboratory work has not been
adequately tested under field conditions. Limited studies
conducted in this area indicate that diammonium phosphate
and monoammonium phosphate can solubilize organic matter
(Bell and Black, 1970; Giordano et al., 1971). As the
solubilized organic matter is carried to a new location in
the soil, it may reprecipitate, covering soil mineral
surfaces which otherwise could have participated in P
retention reactions.


82
The trial was planted in a RCB design with two
replications wherever possible. Some locations had only one
replication because of limited area allocated by farmers for
the trial.
Table 4-1. Application of N, P, and K in different
treatments tested in on-farm experimentation for maize
and cowpea crops in the municipality of Rio Preto da
Eva, Amazonas, Brazil.
Treatment
Maize
Cowoea
N
P
kg
K
ha'1
N
P K
FP+
0
0
0
0
2+
0
CM20+TSP20
100
8.8+8.8
60
0
88+88
60
PCW20+TSP20
100
8.8+8.8
60
0
8.8+8.8
60
TSP40
100
8.8+8.8
60
0
0+17.6
60
-f No restriction was imposed on the amount of nutrient or
type of cultural practice to be used. This treatment
varied from farm to farm.
4= the range for P was 0-4 kg ha1. The value of P reported
is the mean.
Soil samples were taken before planting, during maize
and cowpea flowering, and after harvest. All soil samples
were analyzed for pH, Ca, Mg, Al, K, and P. Particle size
analysis was conducted only on samples taken before
planting. Plant tissue samples were analyzed for P, Zn, Cu,
and Mn. Their results are used in Chapter III. The
analytical technigues for soil and plant tissue samples were
essentially the same as described in Chapters II, and III.


1.85 minimum acceptable level
Yield, Mg/ha
Figure 4-4b. Distribution of confidence intervals for maize
production in poor (e<1.85 mg ha*1) and good (e>1.85
mg ha"1) environments.


96
Maize Experiment
Results of combined analysis of variance
Results of the ANOVA for the maize experiments are
presented in Table 4-6. According to the F-test (P>0.05)
treatments influenced yield at all locations. Treatment
CM20+TSP20, according to DMRT (Table 4-7), was superior to
all tested treatments at all locations (0.65-4.40 Mg ha'1)
and the FP failed or produced the lowest yield (maximum 0.25
Mg ha1) .
A combined analysis of variance (CANOVA) based on the
chi-sguare test at the 5% level of significance (meaning
that variances from all five locations are homogeneous), and
the random effect model (locations were selected randomly),
is presented in Table 4-8. Only the 5 locations with
replications were used in CANOVA. The presence of treatment
x location interaction hindered making a statistically valid
statement for a given treatment for all locations, even
though CM20+TSP20 outperformed other treatments as
indicated by the DMRT test.


25
To understand the P release pattern from different
amendments, surface response curves were fitted and are
illustrated in Fig. 2-5. Phosphorus release in
silica for maize, PCW, and CM followed a logarithmic
function and the r2 for PCW, CM, and maize treatments were
0.73, 0.82, and 0.98. A trace amount of P (0.1 /Lxg g'1) was
detected in the control treatment in the silica matrix
confirming that the matrix was not contaminated and the P
detected came solely from the decomposition of amendments.
Langmuir Parameters
Soil Matrix. A standard Langmuir equations were fitted
to the data after correction was made for preadsorbed P
using sequential extraction techniques (Fig. 2-6). This was
followed by calculation of the adsorption maxima and bonding
energy for each treatment (Table 2-5). All but Pueraria
phaseoloides (Kudzu) increased the Langmuir adsorption
maxima at 35 d of incubation compared to the control. The
extent of variation among different amendments ranged from
536 to 818 nq g'1 soil at 35 d of incubation. It is
noteworthy that PCW had the highest adsorption maxima (818
jug g1) The bonding energy was reduced for all treatments
compared to the control. As the time of incubation increased
from 35 d to 150 d, the adsorption maxima decreased for all
amendments compared to the control treatment. But the


86
Y = the mean treatment yield, or other criterion within
the RD,
a = the level of confidence,
t = value from a two-tailed "t" table,
a
S = [Zx2/(n-l)]% or standard deviation of treatment
yield, or of other evaluation criterion within the RD, and
n = no. of locations or environments in each
recommendation domain.
The Cl test was used to provide an assessment of the risk of
low yields and unacceptable levels of other evaluation
criteria within each RD.
Results and Discussion
The project area is inhabited by subsistence farmers
who practice a slash-and-burn, rotational type of
agricultural system and market surplus production and
products gathered and hunted in the forest. The two major
land types used for crops are: (i) area cleared for the
first time by slash-and-burn from primary forest (PF), and
(ii) area cleared from secondary forest which was left in
fallow 5-7 years (SF). For this trial, a third type was
used, (iii) area considered undesirable for agricultural
activities (WL). The chemical and physical properties of
these soils are presented in Tables 4-2 and 4-3. All sampled
soils were acidic (median pH H20 = 4.5) with very low ECEC


68
Movement of Ca in the soil profile over the 240 d
period indicated that application of TSP may reduce the
soilCa pool (Figure 3-7). The level of Ca in the 0-30 cm
depth was 2.30 cmol (+) kg'1. This value dropped to 0.50
cmol (+) kg'1 within 240 d. Maize plants from the control
treatment plot also had the lowest uptake of Ca. Application
of M60 and CM60 improved the Ca status in 0-15 cm. These
materials after decomposition released cations which were an
integral part of their composition (Larsen et al., 1972).
Deep-rooting green manure crops offer the advantage of
recycling cations which have been leached to a deeper soil
profile. There is considerable evidence in the literature
that organic ligands can hold polyvalent cations and prevent
them from leaching (Moreno, 1960). Building a cation reserve
is the matter of great importance in acid soils where soil
chemical and physical conditions favor their rapid depletion
through leaching.
High P concentration in leaf tissue reduced Zn and Cu
concentration in leaf (Figure 3-8). This can be attributed
to the chelation of these elements by fulvic acids (Saar and
Weber, 1982). However, there is not enough evidence to
conclude that at low leaf P concentration there will be a
higher uptake of these metals. It has been shown that soils
with high level of P are less likely to exhibit Cu toxicity
(Sartain and Street, 1980).


61
Field Study
Rainfall distribution data for both cropping cycles are
presented in Figure 3-lc. The distribution is bi-modal with
a short dry season. During this period evaporation is
greater than precipitation.
Grain Yield. Single degree of freedom orthogonal
contrasts of maize grain production during the first crop
(Table 3-8) indicated that one application of PCW equivalent
of 26.4 kg P ha'1 produced only 1.41 Mg ha"1 of maize. This
level of production was inferior to the control treatment.
Canavalia ensiformis. Mucuna aterrima. and TSP when applied
separately to provide 26.4 kg ha1 of P produced the same
amount of grain during first and second cropping cycles. But
the CM treatment produced more grain than other treatments
in both the first and second crops. This was expected based
on the results of the adsorption study presented in Chapter
II. One explanation is that the faster rate of decomposition
(0.26 g/100 g d'1) and the high P content in CM maintained
the soil P concentration at a high level from the beginning
of plant growth. In a review paper, Olsen and Barber (1977)
concluded that an annual application of manure and
superphosphate resulted in an increased level of 0.01M
CaCl2- and 0.5 M NaHCOj-extractable P. In most studies,
manure treated soils tend to support a higher level of
soluble P than soil treated with an equivalent amount of
superphosphate.


9
s1 = k'c Sp
Where:
S1 = amount of added P sorbed, nq g'1
Sp = Y axis intercept-representing native soil P in the
adsorbed phase, nq g1
C = P in solution, nq mL1
k' = linear adsorption coefficient (estimated without
considering native adsorbed P, Sp) mL g1.
Syers et al. (1973) reported that at lower
concentrations the Eadie-Hofstee plot was more useful for
evaluating P sorption than the conventional Langmuir
equation, because the plot expanded the low P concentration
region.
The methods described above are based on the assumption
that all surface-retained P is displacable either by
isotopic dilution or sequential extraction techniques in the
presence of a weak extracting salt solution. However,
evidence suggested that with increasing contact period and
temperature, P becames less ready to exchange with
isotopically labeled P (Barrow and Shaw, 1975). In addition,
adsorption and desorption isotherms are different, and the
relationship between the quantity of adsorbed P and
concentration is not totally reversible (Sample et al.
1980). Therefore, estimation of surface held native P based
on the techniques described above might underestimate the
true adsorption maxima and bonding energy values. Thus, any


Ill
Table 4-12. Combined Analysis of Variance^ for
cowpea experiments.
Source
DF
EMS
F Value
Pr > F
BLOCK
1
0.07
LOC
2
0.80
1.7
0.20
BLOCK*LOC
2
0.47
TRT
3
2.40
26.6
0.00
TRT*LOC
6
0.09
4.7
0.03
BLOCK*TRT*LOC 9
0.02
-f Chi-square test for homogeneity of variances was
significant. All sites with coefficient of variation > 20%
were excluded from the combined analysis.
Random model. LOC was tested against BL0CK*L0C, TRT against
TRT*LOC, and TRT*L0C against BL0CK*TRT*L0C.


24
treatments in the soil matrix. Processed city waste appeared
to immobilize soil P as indicated by the P measurement made
at 35 d of incubation but at 150 d a net release was
observed. All leguminous amendments followed similar P
release patterns. The highest P release was observed for CM
treatment at 150 d of incubation (245 /ug g1 soil) and the
lowest was for maize and grass treatments (3.9, and 3.4 ng
g'1 soil, respectively) .
When silica was used as the incubation matrix, there
was a net release of P from all amendments which indicated
that P released from amendments was being adsorbed by soil
(soil matrix) and was not all exchangeable in sequential
extraction with 0.01 M CaCl2. Maize and grass treatments
which had shown no net P release at 150 d in the soil matrix
released over 100 /g g'1 silica in the silica matrix. Given
the similarity in molecular structure and chemical behavior
of Si and P (Fig. 2-4) it is believed that added Si
increases water soluble and easily extractable P (Adams,
1980), and Si does not absorb P, because both of them in
ionic forms are negatively charged. It is also suggested
that silicate and phosphate ions compete for the same
adsorption sites on Al(Fe)-oxide surfaces (Mekaru and
Uehara, 1972) and they form insoluble precipitates with such
common ions as Al, Fe, and Ca. This suggests that all P
mineralized from organic amendments remained in the solution
and was easily extractable by 0.01 M CaCl2.


139
Blair, G. J., and O. W. Boland. 1978. The release of
phosphorus from plant material added to soil. Aust. J.
Soil Res. 16:101-111.
Bowman, R.A., and S.R. Olsen. 1985. Assessment of phosphate
buffering capacity: 2. Greenhouse methods. Soil Sci.
140:387-392.
Brown, P. L., and D. D. Dickey. 1970. Losses of wheat straw
residue under simulated conditions. Soil Sci. Soc.
Amer. Proc. 34:118-121.
Burrough, P.A. 1983. Multiscale source of spatial variation
in soil: I. The application of fractal concepts to
nested level of soil variation. J. Soil Sci. 34:577-
597.
Byerlee, D., L. Harrington, and M. Collinson. 1980. Planning
technologies appropriate for farmers: Concepts and
procedures. International Maize and Wheat Improvement
Center (CIMMYT), Mexico.
Byerlee, D., L. Harrington, and D.L. Winkelmann. 1982.
Farming Systems Research: Issues in research strategy
and technology design. Am. J. Agri. Econ. 64:897-904.
Chang, S.C., and M.L. Jackson. 1957. Fractionation of soil
phosphorus. Soil Sci. 84:133-144.
Cochran, V. L., L. F. Elliot, and R. I. Papendick. 1980.
Carbon and nitrogen movement from surface applied wheat
straw. Soil Sci. Soc. Amer. J. 44:978-982.
de Wit, C.T. 1953. A physical theory on placement of
fertilizers. Versl. Landbouwk. Onderzoek. No. 59.4.
Douglas, Jr. C. L., R. R. Allmaras, P. E. Rasmussen, R. E.
Ramig, and N. C. Roger, Jr. 1980. Wheat straw
composition and placement effects on decomposition in
dryland agriculture of the pacific Northwest. Soil Sci.
Soc. Amer. J. 44:833-837.
Easterwood, G.W. and J.B. Sartain. 1990. Organic coatings on
P fertilizers: Influence on plant growth on a Florida
Ultisol (in review). Soil and Crop Sci. Soc. Fla. Proc.
49:
Eberhart, S.A., W.A. Russell. 1966. Stability parameters for
comparing varieties. Crop Sci. 6:36-40.


2
investigated. Immediate and long term effects of such
applications have been studied by Kamprath (1967), Yost
(1977), and Yost et al. (1981). These studies have suggested
that a high initial P application rates can reduce the P
fixation capacity (Barrow, 1974), increase the cation
exchange capacity (Keng and Uehara, 1974), and increase soil
pH.
Advanced Research with Organic Amendments
Along with fertilizer application technigues, attempts
are underway to manipulate the mineral surface chemistry and
inactivate the high reaction capacity of sesguioxides by
application of organic amendments (Larsen et al., 1959;
Nagarajah et al., 1970; Yuan, 1980). It is believed that
*-- ^orm a coating on the mineral surface
1990) and complex Fe3+ and Al3+
(Larsen et al., 1959) which will reduce P fixation. However,
the role of an organic amendment as a P source has not
received much attention. Attempts have been made to
understand the factors responsible for P mineralization from
organic amendments. Amendment P content (Singh, and Jones,
1977) and decomposition characteristics (Sweeney and Graetz,
1988) are important factors in understanding the P release
patterns and predicting residual effects.
Use of P adsorption isotherms in predicting P
requirements of crops and soils has received wide attention


65
Trt.
a
b1
b2
b3
2
RZ
TSP60
4.79
5.53E-01
-6.30E-03
1.65E-05
0.75
M60
4.77
4.08E-01
-3.49E-03
7.58E-06
0.85
Control 4.05
-6.92E-03
-2.00E-04
7.18E-07
0.95
PCW60
4.07
3.96E-02
-3.08E-04
7.91E-07
0.15ns
C60.
3.62
3.02E-01
-2.56E-03
5.56E-06
0.99
Figure 3-5. Effect of selected organic amendments, applied
in a quantity equivalent to provide 26.4 kg ha"1 of P,
on sustaining Mehlich-I extractable soil P pool in a
maize-maize rotation on a Xanthic Hapludox. M = Mucuna,
C = Canavalia


131
An analysis of variance for maize herbage dry weight
from the glasshouse study, indicated the presence of an
interaction among type and rate of amendments. Among three
amendments tested, PCW produced the lowest yield. It is
interesting to note that when PCW was applied at a higher
rate (equivalent of 17.6 kg P ha'1) the result was inferior
compared to the lower rate (equivalent of 8.8 kg P ha'1).
The highest dry matter yield per pot was obtained with CM.
A single degree of orthogonal contrast for maize grain
production during the first crop showed that application of
PCW equivalent to 2 6.4 kg ha"1 of P produced 1.41 Mg ha'1 of
maize. This level of production was inferior to the control
treatment. No significant differences were found among
canavalia, mucuna, and TSP when applied to provide P at the
rate of 26.4 kg ha'1. The CM treatment outyielded all
treatments in first as well as second crops.
Combined application of 8.8 kg P ha'1 from organic
amendments and the same amount from TSP was compared with
17.6 kg P ha'1 from TSP. For the first crop the combination
with CM was superior compared to TSP and an improvement of
30% in yield was recorded.
The change in P status of soil over 240 d of cropping
cycle for all but PCW treatment followed a cubic surface
response curve. All treatments improved P status of soil
compared to the control. As the cropping season progressed,
there was a sharp decrease in soil P with TSP60 compared to


BIOGRAPHICAL SKETCH
Braj K. Singh was born on May 22, 1955, in a small
village in Nepal. B.K. graduated from high school in
Kalaiya, Bara, in 1971. He received his B.S. and M.S. in
agronomy from Peoples' Friendship University in Moscow,
USSR, in 1980. From the same university he earned a Russian-
English interpreter's diploma. He worked as an assistant
lecturer of agronomy, an extension officer, and cropping
systems/farming systems agronomist in Nepal from 1980 to
1986. He was awarded a Fulbright scholarship in 1986 and
later a farming systems assistantship from the University of
Florida. B.K. conducted his dissertation project in
collaboration with TropSoil and EMBRAPA in Manaus, Brazil.
He is married to Zoila Alvarado Singh and has a son,
Alexander Alvarado Singh, and a daughter Carol Alvarado
Singh.
150


46
Glasshouse Experiment
A factorial arrangement of three factors to yield 18
treatment combinations was employed in a randomized complete
block design (RCBD) with four replications (Table 3-1). Five
kg of unlimed soil from the Ap horizon of a Xanthic Hapludox
was mixed in a pot with organic amendments (<2 cm long) 7 d
prior to maize seeding. Maize (variety BR 5110) was planted
and thinned to 2 plants per pot 7 d after planting. The
experiment was harvested at 65 d. At harvest, leaves
immediately below and opposite to the ear leaf (70% plants
had begun to develop ear) were collected for chemical
analysis. A 0.2 g sample of the ground leaf tissue was
digested with H2S04 and H202. Potassium, Ca, Mg, Fe, Zn, Cu,
Mn were determined with an atomic adsorption
spectrophotometer, and P was determined colorimetrically
using the Murphy and Riley (1962) procedure. Soil P was
extracted with the Mehlich-I extractant P (0.05 M HCl +
0.0125 M H2S04, with a soil solution ratio of 1:10, and 5
min shaking time), unbuffered 1 M KC1 (1:10 soil:solution
ratio) was used for the determination of extractable Al.
Aluminum was determined by titrating the extract with 0.1 M
NaOH to bromthymol blue endpoint. Soil reaction (pH) was
determined in water using a soil:water ratio of 1:2.5 and in
1 M KC1 using the same ratio. Maize dry matter production
was recorded and surface response curves were fitted in the
case of interaction effects among factors.


81
Selection of Treatments
Three treatments from previous on-station research were
selected for comparison with farmers practices' (FP) for
growing maize (Zea mays L.) and cowpea (Vigna uncruiculata)
(Table 4-1). Several locations were identified to encompass
the different land types within the soil family limit of
Clayey, Isohyperthermic, Xanthic Hapludox. A shift in maize
planting date was agreed upon so maize maturity would
coincide with the beginning of the dry season (Fig. 4-1).
For the maize crop, all treatments except FP received 100 kg
ha"1 of N from urea half applied broadcast before planting
and the other half in two additional split applications. All
K (60 kg ha"1) was applied basal broadcast except to the FP
plots. Processed city waste, CM, and TSP were applied in 25
cm bands. The maize variety BR-5110 was planted in rows 80
cm apart. No nitrogen was applied to the cowpea crop. The
cowpea variety IPEAN V-69 was planted in rows 60 cm apart.
Plot size for maize and cowpea varied from 100-200 square
meters. The organic amendments were applied in a 20 cm band.
Both crops were harvested for grain. For growing maize and
cowpea the main differences between FP and the improved
practice was that the improved practice received fertilizer.
The land preparation and planting methods consisted of
clearing the area by slash and burn, followed by manual land
preparation and planting with sticks. This is a common FP in
the area.


109
Cowpea Experiment
Results of combined analysis of variance
Cowpea response was evaluated at 13 locations. Eight
were replicated and five were not. The ANOVA and DMRT for
locations are presented in Table 4-10 and 4-11. According to
the F-test, treatments did not influence yield (P>0.05) at
three locations. At location 1, CM20+TSP20 was as good as
TSP40 while CM20+TSP20 did not differ from PCW20+TSP20
(Table 4-12). A CANOVA based on a similar computation as for
maize is presented in Table 4-9. A significant chi-square
test (chi-square = 11.78) suggested the elimination of all
locations with cv > 20%. Therefore, this test was carried
out with only three of the 8 replicated locations. The
treatment x location interaction was found to be significant
(P <0.03). Gomez and Gomez (1985) suggested the partitioning
of treatment x location interaction using either the
homogeneous site approach or the homogeneous treatment
approach. However, by rejecting 10 locations out of 13, it
was felt that the locations were no longer a true
representation of the population.


84
Statistical Analysis
Combined analysis of variance (CANOVA). A combined
analysis of variance was conducted for replicated trials. A
preliminary analysis for homogeneity of variances from the
individual ANOVAs was based on the chi-squared test. In the
case of a significant chi-squared test, locations with a
coefficient of variation >20% were excluded from the CANOVA
(Gomez and Gomez, 1985). No attempt was made to employ a
missing plot technique to recover lost data or data which
violated some assumptions of the ANOVA. It is common to lose
part of on-farm experiments to destruction of experimental
plants, loss of harvested samples, etc. Data from two
locations for maize and one for cowpea were lost due to
misunderstanding with farmers about the harvest date. The
trials were harvested before the agreed upon date. At three
locations (2 for maize and one for cowpea), the crop failed
due to late planting. Data from one location for cowpea
could not be used for ANOVA because the second harvest yield
for all treatments was mixed without recording production
separately.
Stepwise regression. Stepwise regression is a method
for either the forward, or backward selection of variables
based on an F statistic of R2 significance at the level
specified in the model (SAS Institute, Inc., 1985). This
technique was used to identify factors responsible for the
wide range of variation in maize and cowpea production over


TABLE OF CONTENTS
ACKNOWLEDGEMENTS iii
LIST OF TABLES vii
LIST OF FIGURES X
ABSTRACT xiii
CHAPTER I INTRODUCTION 1
Statement of Problem 1
Goals and Objectives 4
CHAPTER II ORGANIC AMENDMENTS AND PHOSPHORUS
ADSORPTION ISOTHERMS 5
Soil Constituents and Phosphorus Adsorption ... 5
Materials and Methods 10
Results and Discussion 14
Conclusion 35
CHAPTER III ORGANIC AMENDMENTS AND CROP PHOSPHORUS
NUTRITION 37
Phosphorus Management Strategy 37
Materials and Methods 45
Results and Discussion 52
Conclusion 71
CHAPTER IV RECOMMENDATION DOMAINS AND MODIFIED
STABILITY ANALYSIS 72
Introduction 72
Materials and Methods 79
Results and Discussion 86
Conclusion 12 6
CHAPTER V SUMMARY AND CONCLUSIONS 129
APPENDIX ECONOMIC ANALYSIS 134
REFERENCE LIST 138
BIOGRAPHICAL SKETCH 150
vi


REFERENCE LIST
Abalu, G.O.I., N.M. Fisher, and Y. Abdullahi. 1987. Rapid
rural appraisal for generating appropriate technologies
for peasant fanners: Some experiences from Northern
Nigeria. Agri. Systems 25:311-324.
Adams, F. 1980. Interactions of phosphorus with other
elements in soils and in plants, p. 655-680. In M.
Stelly and R. C. Dinauer (ed.) The role of phosphorus
in agriculture. ASA, CSSA, and SSSA, Madison, WI.
Anghinoni, I., and S.A. Barber. 1980. Phosphorus application
rate and distribution in the soil and phosphorus uptake
by corn. Soil Sci. Soc. Am. J. 44:1041-1044.
Appelt, H., N.T. Coleman, and P.F. Pratt. 1975. Interaction
between organic compounds, minerals and ions in
volcanic-ash derived soils. II. Effects of organic
compounds on the adsorption of phosphate. Soil Sci.
Soc. Am. Proc. 39:628-630.
Bache, B.W. 1964. Aluminum and iron phosphate studies
related to soils. II. Reaction between phosphates and
hydrous oxides. J. Soil Sci. 15:110-116.
Barrow, N.J. 1974. The effect of previous additions of
phosphate on phosphate adsorption by soils. Soil Sci.
118:82-89.
Barrow, N.J. 1983. A mechanism model for describing the
sorption and desorption of phosphate by soil. J. Soil
Sci. 32:555-570.
Becker, H.C. 1981. Correlations among some statistical
measures of phenotypic stability. Euphytica 30:835-840.
Bell, L.C., and C.A. Black. 1970. Comparison of methods for
identifying crystalline products produced by
interaction of orthophosphate fertilizers with soils.
Soil Sci. Soc. Am. Proc. 34:579-582.
138


Table 2-5. Langmuir parameters (k and b) for P sorption by soil incubated
with different organic amendments.
Org.+
Amend.
35 Days
65 Days
Incubation Period
150 Days
r2
b+
k§
r2
b
k
r2
b
k
Canavalia
0.87**
638
0.384
0.81**
633
0.303
0.89**
528
0.264
Peanut
0.73**
632
0.233
0.87**
522
0.282
0.85**
509
0.229
Maize
0.94**
560
0.785
0.84**
670
0.308
0.86**
673
0.431
Tephrosia
0.69*
768
0.229
0.91**
477
0.291
0.91**
465
0.246
Mucuna
0.90**
536
0.207
0.64*
654
0.087
0.49ns
695
0.120
CM
0.66*
643
0.080
0.67*
618
0.079
0.76**
496
0.103
Kudzu
0.93**
471
0.417
0.94**
577
0.476
0.95**
535
0.599
PCW
0.67*
818
0.224
0.66**
815
0.271
0.94**
517
0.375
Cowpea
0.62*
729
0.146
0.89*
585
0.244
0.91**
543
0.473
Grass
0.74**
765
0.247
0.80*
707
0.331
0.84**
603
0.363
Control
0.94**
535
0.673
0.95**
544
0.941
0.98**
591
1.941
*, ** Significant at the 0.05 and 0.01 levels, respectively.
-f Mucuna aterrima (Mucuna), Pueraria phaseoloides (Puer.), Canavalia ensiformis
(Canav.), Zea mays (Maize), Arachis hypoqaea (Peanut), Tephrosia candida (Teph),
Vicrna uncruiculata (Cowpea) mixed gramineae (Grass) aerobically digested city waste
(PCW) and chicken manure (CM).
+ b, P adsorption maxima, nq P g'1
§ k, bonding energy, mL g P"1


99
Modified stability analysis
Yield. An environmental index (e) was computed for all
environments (Table 4-9). Location 3 was the poorest
environment (mean e for two replications = 0.20 Mg ha'1). It
had a very high Al saturation (95%), low ECEC (1.35 cmol
( + ) kg'1), very low water pH (3.9), and only a trace of
Mehlich-I extractable P (Table 4-2). Farmers classified this
location as WL, obviously with good reasons. Scattered
occurrence of WL has led many scientists to be skeptical
about permanent agricultural development on highly leached
soils of the tropics (McNeill, 1964; Sioli, 1980). But the
latest advancements with long term fertility experiments
have offered potential for sustained production with proper
management (Sanchez et al., 1982).
The value of e for the best environment (location 7)
was 3.10 Mg ha'1. This location was recently cleared from
primary forest and was in the first year of cultivation.
Soil pH was very favorable (pH H20 = 5.2) with a ECEC of
4.21 cmol (+) charge kg'1 soil, and Al saturation of only
58%.
The MSA indicated that PCW20+TSP20 performed poorly in
all environments compared to the other treatments with
amendments based on maize grain production (Fig. 4-4a).
However, an exponential decrease in production under FP was
observed with decline in environmental guality. The low


1.-6 6n
27
Figure 2-5. Effect of incubation period and incubation
matrix on P release pattern from different organic
amendments in a laboratory study.


88
respectively. This analysis provided evidence that
restoration of soil fertility by letting secondary forest
take over for 5-7 years is unlikely to happen.
Soil Chemical Properties and Yield
A stepwise regression was carried out to understand the
relationship between maize and cowpea yield and soil
chemical properties. Emphasis was placed on how different
amendments influenced Mehlich-I extractable P. All
treatments except FP had received an equal amount of P from
organic and inorganic sources. Table 4-5 presents results
of the stepwise regression on amended soil characteristics.
For FP in maize, each unit increase in A1 saturation
decreased the production by 0.06 Mg ha'1. Application of
PCW20+TSP20 did not improve soil P level, instead an
improvement in soil pH was observed (similar results were
obtained in glasshouse study). Application of TSP40 and
CM20+TSP20 resulted in improved soil P. However, the
magnitude of yield improvement with similar increases in P
levels were different for different treatments. It was
higher for TSP40 treatment by 0.90 Mg ha'1 compared to
CM20+TSP40.
It is widely believed that low soil pH is often
associated with Al and Mn toxicity and Ca deficiency. An
acidic soil reaction can reduce rhizobia growth, nodule


133
increasing at the rate of 7% per year in PF compared to 4%
in SF. The rate of decline in Mehlich-I P in both land types
was in the range of 2.1-2.6/zg"1 soil. A relatively low ECEC
in both land types (2.3-3.6 cmol (+) charge kg1 soil)
indicated that most cations were leached following heavy
rainfall. And those present were being taken up by
vegetation or being washed out at the rate of 0.80 and 0.16
cmol (+) charge kg1 soil every year from PF and SF,
respectively.
The data presented in Chapter IV suggested that the
delineation of RDs was dependent on the technology
evaluation criteria, and was not amenable to geographical
mapping because fields of different domains were
interspersed. A change in evaluation criteria led to a
change in the demarcation of RDs. It was also observed that
a complete restoration of soil fertility by letting
secondary forest take over for 5-7 years was unlikely to
happen in the rain forest.


59
% Leaf P
Figure 3-3. Relationship between Mehlich-I extractable
soil P (a), and maize leaf tissue P concentration (b)
with maize dry matter yield.


Table 3-2. Description of treatments tested in the field
Source
Amendment Type
Treatment^ P Equivalent
kg ha1
Mucuna aterrima
M60
26.4
M2 0
8.8
ORGANIC
Canavalia ensiformis
C60
26.4
C20
8.8
Processed City Waste
PCW 60
26.4
PCW 2 0
8.8
Chicken Manure
CM60
26.4
CM20
8.8
INORGANIC
Tiple Superphosphate
TSP20
8.8
TSP40
17.6
TSP60
26.4
ORGANIC+INORGANIC
Mucuna, Canavalia,
M20+TSP20
17.6
PCW, CM, TSP
C20+TSP20
17.6
PCW20+TSP20
17.6
CM2 0+TSP2 0
17.6
CONTROL
Control
0
f Whenever P from
organic and inorganic sources
was applied
together
amendments were applied first in furrows followed by the application of inorganic P.
U1
o


E.A. Hanlon, Dr. D.A. Graetz, and Dr. K.L. Buhr for their
suggestions, support, and editorial comments. A special note
of gratitude is extended to Dr. C.K. Hiebsch for
substituting Dr. K.L. Buhr as a member of supervisory
committee and reviewing the manuscript.
A particular note of appreciation is extended to Dr.
Hugh Popenoe, Director of International Programs, for
providing primary funding for the author's Ph. D. program.
The author is indebted to and gratefully acknowledges the
financial and technical support provided by TropSoils and
EMBRAPA. Particular gratitude is expressed to Dr. Walter
Bowen for his invaluable guidance in design and execution of
the research and manuscript review. Special thanks go to Mr.
Manoel da Silva Cravo and his extraordinary group of
enthusiastic technicians, Raimundo Vitoriano de Oliveira,
Agilau de Araujo Rodrigues, Edilza da Silva Richa, Emanoel
dos Santos Alencar, Teofanes Moreira de Souza Junior, and
Onelia Maria Pereira de Almeida for conducting the routine
analysis of plant and soil samples. The author is grateful
to Mrs. Eda M. Souza, CEPA-AM for her support in putting the
on-farm project together.
A word of gratitude is also extended to Dr. K.D. Sayre,
Dr. W. H. Freeman, Dr. C.N. Hittle, Dr. Richard Harwood, and
Dr. Douglas Beck for their help in getting the author into
the graduate school. Thanks go to S.K. Patel for his help in
iv


CHAPTER III
ORGANIC AMENDMENTS AND CROP PHOSPHORUS NUTRITION
Phosphorus Management Strategy
In highly leached soils, the minerals with permanent
charge have been either severely altered or completely
weathered out, so that the surface charge arises from
adsorption of potential determining ions such as hydrogen
and hydroxyl. The magnitude of the surface charge is
expressed by a combined Gouy-Chapman and Nerst equation.
This equation provides a theoretical basis for increasing
the cation retention capacity of a soil by lowering pH0
(value of soil pH at which net surface charge is zero). One
way of lowering pH0 is to increase the organic matter,
phosphorus, or silica content of the soil (Uehara and
Gillman, 1981).
Highly weathered soils are also very poor in total and
plant available P. Their high P fixing capacity requires
high doses of applied P to meet crop demands. The classic
work of de Wit (1953) on physical theory of fertilizer
placement predicts that when suboptimum quantities of
fertilizer are used, restricted placement is desirable (Fox
et al., 1986). de Wit based his analysis on nutrient uptake
37


137
Appendix B contd
CM20+TSP20
2.85
570
241
329
1.4
4.40
880
303
577
1.9
0.65
130
153
-23
-0.2
2.80
560
239
321
1.3
3.60
720
271
449
1.7
3.60
720
271
449
1.7
4.00
800
287
513
1.8
4.00
800
287
513
1.8



PAGE 1

SUSTAINING CROP PHOSPHORUS NUTRITION OF HIGHLY LEACHED OXISOLS OF THE AMAZON BASIN OF BRAZIL THROUGH USE OF ORGANIC AMENDMENTS By BRAJ K. SINGH 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 OF FLORIDA LIBRARIES

PAGE 2

This work is dedicated to my parents: Shree Mr it B. Singh and Shrimati Naina Devi, and to my brothers: Sri Biswanath Singh and Nawal K. Singh.

PAGE 3

ACKNOWLEDGMENTS The author wishes to express his sincere appreciation to Dr. J. B. Sartain, the chairman of the supervisory committee, for his excellent guidance, assistance, and continued interest in this study. His patience, understanding, and personal friendship throughout the academic and research program has been most invaluable and made the author's stay in the United States a most rewarding and stimulating experience. The financial support provided by him to cover part of research expenses in Brazil is gratefully acknowledged. Sincere appreciation is extended to Dr. P.E. Hildebrand for providing a farming systems assistantship and taking personal initiatives to work out a joint research venture with the TropSoils Collaborative Research Support Program between Cornell University and the Brazilian Agricultural Research Organization (EMBRAPA) . His deep concerns, constructive criticisms, and moral support as the member of supervisory committee helped broaden the author's understanding of farming systems and contributed to the successful completion of this manuscript. The author also thanks the other members of the supervisory committee, Dr. iii

PAGE 4

E.A. Hanlon, Dr. D. A. Graetz, and Dr. K.L. Buhr for their suggestions, support, and editorial comments. A special note of gratitude is extended to Dr. C.K. Hiebsch for substituting Dr. K.L. Buhr as a member of supervisory committee and reviewing the manuscript. A particular note of appreciation is extended to Dr. Hugh Popenoe, Director of International Programs, for providing primary funding for the author's Ph. D. program. The author is indebted to and gratefully acknowledges the financial and technical support provided by TropSoils and EMBRAPA. Particular gratitude is expressed to Dr. Walter Bowen for his invaluable guidance in design and execution of the research and manuscript review. Special thanks go to Mr. Manoel da Silva Cravo and his extraordinary group of enthusiastic technicians, Raimundo Vitoriano de Oliveira, Agilau de Araujo Rodrigues, Edilza da Silva Richa, Emanoel dos Santos Alencar, Teofanes Moreira de Souza Junior, and Onelia Maria Pereira de Almeida for conducting the routine analysis of plant and soil samples. The author is grateful to Mrs. Eda M. Souza, CEPA-AM for her support in putting the on-farm project together. A word of gratitude is also extended to Dr. K.D. Sayre, Dr. W. H. Freeman, Dr. C.N. Hittle, Dr. Richard Harwood, and Dr. Douglas Beck for their help in getting the author into the graduate school. Thanks go to S.K. Patel for his help in iv

PAGE 5

representing the author at UF during the author's one year stay in Brazil. The author is indebted to his beloved wife for her encouragement, support, and patience.

PAGE 6

TABLE OF CONTENTS ACKNOWLEDGEMENTS iii LIST OF TABLES vii LIST OF FIGURES X ABSTRACT xiii CHAPTER I INTRODUCTION 1 Statement of Problem 1 Goals and Objectives 4 CHAPTER II ORGANIC AMENDMENTS AND PHOSPHORUS ADSORPTION ISOTHERMS 5 Soil Constituents and Phosphorus Adsorption ... 5 Materials and Methods 10 Results and Discussion 14 Conclusion 35 CHAPTER III ORGANIC AMENDMENTS AND CROP PHOSPHORUS NUTRITION 37 Phosphorus Management Strategy 37 Materials and Methods 45 Results and Discussion 52 Conclusion 71 CHAPTER IV RECOMMENDATION DOMAINS AND MODIFIED STABILITY ANALYSIS 72 Introduction 72 Materials and Methods 79 Results and Discussion 86 Conclusion 126 CHAPTER V SUMMARY AND CONCLUSIONS 12 9 APPENDIX ECONOMIC ANALYSIS 134 REFERENCE LIST 138 BIOGRAPHICAL SKETCH 150 vi

PAGE 7

LIST OF TABLES Table 2-1. Ionic species and concentration of nutrient solution used with silica to simulate solution chemistry of a Xanthic Hapludox in a phosphorus release study by organic amendments 13 Table 2-2: Chemical composition of organic amendments used in the incubation study 16 Table 2-3: Selected physical and chemical properties of Ap horizon of the Xanthic Hapludox used in adsorption studies 17 Table 2-4. Release of 0.01 M CaCl 2 extractable P (/ig g" 1 ) following incubation of organic amendments with soil and silica as matrix substratum 23 Table 2-5. Langmuir parameters (k and b) for P adsorption by soil incubated with different organic amendments 29 Table 2-6. Langmuir parameters (k and b) based on net release of P measured by sequential extraction of simulated silica matrix with 0.01 M CaCl 2 . ... 32 Table 2-7. Difference in the estimated values of Langmuir parameters (b and k) based on the estimation of preadsorbed P by simulated silica matrix technique and sequential extraction. ... 33 Table 3-1. Factorial arrangements of treatments for the greenhouse study 48 Table 3-2. Description of treatments tested in the field 50 Table 3-3. Selected chemical properties of the fine fraction (<2 mm) of the Xanthic Hapludox 55 Table 3-4. Selected physical properties of the Xanthic Hapludox 55 vii

PAGE 8

Table 3-5. Selected chemical properties of the fine fraction (<2 mm) of the Xanthic Hapludox 56 Table 3-6. Analysis of variance for maize herbage dry weight production per pot in the glasshouse study 57 Table 3-7. Mean changes in soil pH (H 2 0) following application of selected organic amendments in the glasshouse study 60 Table 3-8. Orthogonal contrasts of maize grain yield under different treatments applied in a 30 cm wide band at UEPAE research station, Amazonas, Brazil. 62 Table 4-1. Application of N, P, and K in different treatments tested in on-farm experimentation for maize and cowpea crops in the municipality of Rio Preto da Eva, Amazonas, Brazil 82 Table 4-2. Characterization of experimental plots for maize testings in the municipality of Rio Preto da Eva, Amazonas, Brazil 89 Table 4-3. Characteristics of experimental plots for cowpea trials 90 Table 4-4. Relationship between soil characteristics with year in crop production in different land types 91 Table 4-5. Relationship between soil characteristics measured after treatment application with grain yield for maize and cowpea crops in the municipality of Rio Preto da Eva, Amazonas, Brazil 92 Table 4-6. Summary of ANOVA for multilocatonal maize testing in the municipality of Rio preto da Eva, Amazonas, Brazil 97 Table 4-7. Duncan Multiple Range Test (DMRT) for maize crop in the municipal of Rio preto da Eva, Amazonas, Brazil 97 Table 4-8. Combined Analysis of Variance for maize grain yield in the municipal of Rio Preto da Eva, Amazonas, Brazil 98 Table 4-9. Environmental index (e) for maize production in the Municipal ilty of Rio Preto da Eva, Amazonas, Brazil 100 viii

PAGE 9

Table 4-10. Summary of ANOVA for multilocatonal cowpea testing 110 Table 4-11. Duncan Multiple Range Test (DMRT) for cowpea trials conducted in the municipal of Rio Preto da Eva, Amazonas, Brazil 110 Table 4-12. Combined Analysis of Variance for cowpea experiments. Ill Table 4-13. Technology selection for a given land type based on different evaluation criteria for maize cultivation 124 Table 4-14. Technology selection for a given land type based on different evaluation criteria for cowpea cultivation 124 ix

PAGE 10

LIST OF FIGURES Figure 2-1. Acid-base potentiometric titration curves for the Ap horizon of Xanthic Hapludox with varying concentration of CaCl 2 . 18 Figure 2-2. Phosphorus adsorption isotherms following 35 d (a) and 150 d (b) of soil incubation with Mucuna aterrima (Mucuna) , Chicken Manure (CM) , Aerobically Processed City Waste (PCW) , and Control 21 Figure 2-3. Phosphorus desorption isotherms following 35 d (a) , and 150 d (b) of soil incubation with Mucuna aterrima (Mucuna) , Chicken Manure (CM) , Aerobically Processed City Waste (PCW) , and Control 22 Figure 2-4. A schematic representation of molecular structures of oxides of phosphorus (a) , and oxides of silicon (b) 26 Figure 2-5. Effect of incubation period and incubation matrix on P release pattern from different organic amendments in a laboratory study 27 Figure 2-6. Phosphorus adsorption isotherms for the Xanthic Hapludox (0.01 M CaCl 2 ) fitted to Langmuir equation. The lines in the figure represents fitted equation 28 Figure 2-7. A comparison of P adsorption by Xanthic Hapludox based on the estimation of preadsorbed P by sequential extraction of soil or silica matrix incubated with organic amendments 34 Figure 3-1. Geographic location of Amazon basin in Brazil (a) , on-station and farming systems research (FSR) sites (b) , and effective rainfall during the period of August, 1988 until August 1989 (c) at EMBRAPA station in Manaus, Brazil.. . 47 x

PAGE 11

Figure 3-2. Effect of different rates of selected organic amendments in combination with different rates of inorganic P on maize dry matter production at 65 d after planting in a glasshouse study 58 Figure 3-3. Relationship between Mehlich-I extractable soil P (a) , and maize leaf tissue P concentration (b) with maize dry matter yield. 59 Figure 3-4. Rate of inorganic P applied through triple superphosphate and its effect on maize yield in a maize-maize rotation 64 Figure 3-5. Effect of selected organic amendments, applied in a quantity equivalent to provide 26.4 kg ha" 1 of P, on sustaining Mehlich-I extractable soil P pool in a maize-maize rotation on a Xanthic Hapludox 65 Figure 3-6. Effect of selected organic amendments, applied in combination with inorganic phosphorus source in a quantity equivalent to provide 8.8 kg ha" 1 of P, on sustaining Mehlich-I extractable soil P pool in a maize-maize rotation on a Xanthic Hapludox 67 Figure 3-7. Leaching of calcium from surface horizon of highly leached Xanthic Hapludox as influenced by different organic amendments 69 Figure 3-8. Relationship between P concentration and concentration of Zn and Cu in maize leaf tissue. . 70 Figure 4-1. Effective rainfall and suggested shift in the planting date for maize in Rio Preto da Eva, Amazonas, Brazil, (a) common practice, (b) suggested practice 83 Figure 4-2. Range of different soil characteristics by recommendation domains for maize trials 94 Figure 4-3. Range of different soil characteristics by recommendation domains for cowpea trials 95 Figure 4-4a. Response of different treatments to environmental index for maize production, Rio Preto da Eva, Amazonas, Brazil 102 Figure 4-4b. Distribution of confidence intervals for maize production in poor (e<1.95 mg ha" 1 ), and good (e>1.95 mg ha" 1 ) environments 103 xi

PAGE 12

Figure 4-5. Relationship of net income/cash cost, net income/total cost, and net income with environmental index, in on-farm maize trials from Rio Preto da Eva, Amazonas, Brazil 106 Figure 4-6. Distribution of confidence intervals for net income/cash cost, net income/total cost, and net income for different treatments used for maize cultivation 108 Figure 4-7a. Response of different treatments to environmental index for cowpea production, Rio Preto da Eva, Amazonas, Brazil 115 Figure 4-7b. Distribution of confidence intervals for cowpea production in poor (e<1.32 mg ha ), and good (e>1.32 Mg ha" 1 ) environments 116 Figure 4-8. Relationship of net income, net income/cash cost and net income/total cost with environmental index, in on-farm cowpea trials from Rio Preto da Eva, Amazonas, Brazil 117 Figure 4-9. Distribution of confidence intervals for net income/cash cost for selected treatments in poor (e<1.32 Mg ha" 1 ) and good (e>1.32 Mg ha" 1 ) environments for cowpea cultivation 119 Figure 4-10. Distribution of confidence intervals for net income/total cost for selected treatments in poor (e<1.32 Mg ha" 1 ) and good (e>1.32 Mg ha" 1 ) environments for cowpea cultivation 120 Figure 4-11. Distribution of confidence intervals for net income for selected treatments in poor (e<1.32 Mg ha" 1 ) and good (e>1.32 Mg ha" 1 ) environments for cowpea cultivation 121 Figure 4-12. Recommendation domains, based on yield, and their relationship to land types in the municipality of Rio Preto da Eva, Amazonas, Brazil 125 Figure 4-13. Recommendation domains, based on net return to cash cost, and their relationship to land types in the municipality of Rio Preto da Eva, Amazonas, Brazil 12 6 xii

PAGE 13

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 SUSTAINING CROP PHOSPHORUS NUTRITION OF HIGHLY LEACHED OXISOLS OF THE AMAZON BASIN OF BRAZIL THROUGH USE OF ORGANIC AMENDMENTS By Braj K. Singh August 1990 Chairperson: Dr. Jerry B. Sartain Major Department: Soil Science Retention of phosphorus by iron and aluminum oxides plays an important role in determining ultimate availability of P to plants in highly weathered soils of the tropics. Different management strategies have been proposed to overcome this problem. The overall objective of this research was to study the influence of organic amendments (plant origin, manure, and aerobically processed city waste (PCW) ) in sustaining P nutrition of Oxisol and to test the performances of selected alternatives in a wide range of production environments. There was a marked difference in P adsorption and desorption by soil preincubated with organic amendments. The difference was attributed to incubation period and chemical composition of amendments. Correction for preadsorbed P by xiii

PAGE 14

soil based on simulated silica matrix (SSMT) resulted in a better estimation of the Langmuir adsorption maxima. This technique involved mixing of acid washed sand with a nutrient solution (without P) and inoculation of the sand with microbes and incubation with organic amendments. In the glasshouse experiment the highest dry matter (DM) production was obtained with the CM. A strong relationship was observed between DM production and MehlichI extractable P (r 2 = 0.88). In the on-station field experiments, CM applied plots produced more grain in first as well as second crops compared to the plots which had received Canavalia ensiformis , Mucuna aterrima , PCW, and TSP. All organic amendments improved the soil P reserve and reduced Ca leaching indicating that application of organic amendments could lead to sustained crop P nutrition. In farmers' field studies, three amendments were tested in different land types with maize and cowpea crops. The same amount of P applied from different amendments had different effect on maize and cowpea production. However, the selection of a given technology for a given land type (environment) was dependent on farmers* goals. Based on the criteria of grain production CM was recommended for maize in all environments and for cowpea in poor environments. xiv

PAGE 15

CHAPTER I INTRODUCTION Statement of Problem A leaching environment together with persistent high rainfall and temperature are the defining conditions for the development of Oxisols, a soil order which occupies over 1.12 billion ha and ranks 5th in worldwide distribution (USDA, 1975) . Their similarity as a group stems from the composition of the colloid fraction rich in iron and alumina minerals, and deficient in other nutrients. Sustained production on these soils can be achieved only with adequate application of lime, N, P, K, Mg, Zn, Cu, B, and Mo (Sanchez et al., 1982). Among other nutrients, P deficiency appears to be the most crucial and may show up as early as the second year of cultivation. Large amounts of applied P are required to attain levels of soil solution P which are adequate for high crop yields (Yost et al., 1979). Rapidly increasing cost of P fertilizer, and limited P supply in the tropics, have led to extensive P management studies on these soils. Fertilizer application techniques (band, broadcast, and band+broadcast) have been widely 1

PAGE 16

2 investigated. Immediate and long term effects of such applications have been studied by Kamprath (1967), Yost (1977), and Yost et al. (1981). These studies have suggested that a high initial P application rates can reduce the P fixation capacity (Barrow, 1974) , increase the cation exchange capacity (Keng and Uehara, 1974) , and increase soil pH. Advanced Research with Organic Amendments Along with fertilizer application techniques, attempts are underway to manipulate the mineral surface chemistry and inactivate the high reaction capacity of sesquioxides by application of organic amendments (Larsen et al., 1959; Nagarajah et al., 1970; Yuan, 1980). It is believed that organic amendments can form a coating on the mineral surface (Easterwood and Sartain, 1990) , and complex Fe 3+ and Al 3+ (Larsen et al., 1959) which will reduce P fixation. However, the role of an organic amendment as a P source has not received much attention. Attempts have been made to understand the factors responsible for P mineralization from organic amendments. Amendment P content (Singh, and Jones, 1977) and decomposition characteristics (Sweeney and Graetz, 1988) are important factors in understanding the P release patterns and predicting residual effects. Use of P adsorption isotherms in predicting P requirements of crops and soils has received wide attention

PAGE 17

(Jones and Benson, 1975; Solis and Torrent, 1989). However, the techniques used to construct P adsorption isotherms in the presence of amendments are similar to those used for soil without amendment. Generally a correction for preadsorbed P is introduced (Mokwunye, 1977; Solis and Torrent, 1989) . An assumption is made that adsorption and desorption are equal, which is not true (Sample et al., 1980) . Therefore, it leads to unreliable prediction of crop P requirement and residual effect of applied P. Organic and Inorganic P Interaction A large amount of organic amendment will be required to provide the total P requirement for a crop. An efficient way to use a suboptimal dose of an organic amendment is band application with inorganic P. Such an application reduces P immobilization from P-poor amendments, and eliminates P deficiency in the early stages of crop growth. Much remains to be learned about the interaction of organic amendments with inorganic P. Most laboratory work has not been adequately tested under field conditions. Limited studies conducted in this area indicate that diammonium phosphate and monoammonium phosphate can solubilize organic matter (Bell and Black, 1970; Giordano et al., 1971). As the solubilized organic matter is carried to a new location in the soil, it may reprecipitate, covering soil mineral surfaces which otherwise could have participated in P retention reactions.

PAGE 18

4 Transfer of Technology It is widely believed that future efforts to increase food production must be directed towards the marginal lands of the developing world (Shaner et al., 1982). To achieve this goal, more research on appropriate technology is required with direct farmer involvement in problem identification, research priority identification, and technology evaluation (Harwood, 1979) . Such an attempt will lead to location-specific technology tailored to fit farmers' circumstances, and accelerate the process of technology diffusion (Hildebrand, 1983) . Goals and Objectives The overall goal of this research was to examine the role of organic amendments in sustaining P nutrition of a highly leached Oxisol. The specific objectives were: (i) to devise a technique to measure the P release patterns from a decomposing organic material, (ii) to examine the effect of a combined application of organic and inorganic P in a band on sustaining crop P nutrition, and (iii) to conduct farmers' field trials to validate on-station research results and delineate recommendation domains for selected treatments .

PAGE 19

CHAPTER II ORGANIC AMENDMENTS AND PHOSPHORUS ADSORPTION ISOTHERMS Soil Constituents and Phosphorus Adsorption It is widely believed that hydrous oxides of Fe and Al, and calcium carbonate play key roles in P retention. Although controversy exists among researchers about the mechanism of P retention by these compounds (Muljadi et al., 1966; Hingston et al . , 1967; Rajan et al., 1974), the phasic nature of P adsorption has been well recognized by several researchers (Bache, 1964; Munns and Fox, 1976). The first phase of adsorption is due to a high energy chemisorption of small amounts of P on the soil surface. The second phase is comprised of a precipitation reaction followed by a low energy sorption of P onto the precipitate. Phosphorus Adsorption Isotherms Phosphorus adsorption isotherms have been used extensively for describing the P adsorption characteristics of various soils and in estimating the P requirement of different crops (Fox and Kamprath, 1970; Jones and Benson, 1975; Mokwunye, 1977; Singh and Jones, 1977; Solis and Torrent, 1989) . The isotherm technique involves 5

PAGE 20

6 equilibration of a known amount of soil for a limited time in a KC1, NaCl or CaCl 2 solution containing various amounts of K 2 HP0 4 , NaH 2 P0 4 or Ca(H 2 P0 4 ) 2 (Olsen and Watanabe, 1957; Syers et al., 1973; Singh and Jones, 1976). Phosphorus removed from solution is considered to be adsorbed. Plant P requirement is calculated based on the amount of P needed to bring the concentration of supernatant solution to a specified concentration (generally 0.2 jtxg mL" 1 ) . Lancrmuir equation . A number of researchers have investigated soil P adsorption characteristics using the Langmuir equation (Olsen and Watanabe, 1957; Woodruff and Kamprath, 1965; Gunary, 1970; Borggaard, 1983). A frequently used linear form is: (c/x/m) = (1/kb) + (c/b) Where : x/m = amount of P adsorbed per unit weight of soil b = the adsorption maxima c = the equilibrium P concentration, ng mL" 1 k = a constant related to the bonding energy of the adsorbent for adsorbate. For a given uniform population of sites, the value of adsorption maxima can also be evaluated by plotting x/m vs. x/m/c, obtained from a rearranged Langmuir equation commonly referred to as the ' Eadie-Hof stee ' plot (Syres et al., 1973) . The underlying assumptions in each case are that adsorption sites on the particle surface are uniform, and

PAGE 21

the maximum adsorption possible corresponds to a complete monomolecular layer. Both of these postulates do not hold for a heterogenous medium like soil (Larsen, 1967) . Deviations from the conventional Langmuir relationship at high equilibrium P concentrations (above 15 /xg ml/ 1 ) have already been reported (Olsen and Watanabe, 1957; Hsu and Rennie, 1962) which have led to the development of an extended form of the Langmuir equation. Gunary (1970) included a square-root term in the Langmuir equation. Holford et al. (1974), Syers et al. (1973), and Holford (1983) used a two-surface Langmuir equation. The two surface equation has not, however, been universally accepted and Posner and Bowden (1980) have discussed the futility of attempting to split isotherms into two or more regions. Similarly, a curve-fit error in estimating the Langmuir adsorption maxima was described by Harter (1984) . He claimed the test of linearity was inadequate because plotting concentration against itself reduces data variability and always provides a significant correlation coefficient. He suggested that a better test of the fit is to ascertain whether the adsorption isotherm has the shape of the equation model . In spite of drawbacks, the Langmuir equation is used widely to describe P adsorption by soil. The major advantage of the Langmuir equation is that it allows for the calculation of adsorption maxima along with a relative

PAGE 22

8 bonding energy term for P sorption (Syere et al., 1973). The values of adsorption maxima and bonding energy can be related to various soil properties which will supply information about the nature of the reaction between soil and fertilizer P, and can aid in the prediction of plant available P (Olsen, 1953) . Correction for Initial Surface Phosphate Olsen and Watanabe (1957) have demonstrated the effect of correction for initial surface phosphate on the constants derived from the Langmuir isotherm. For Pierry clay and Owyhee silt loam soils the correction increased the adsorption maxima and the bonding energy constant. Ideally, the adsorption would be determined in a system in which the surface is free of the adsorbate ion. Usually this restriction is not feasible. Therefore, a correction for initial P is made by adding the amount of surface P determined by a separate analysis. One of the commonly used methods to measure native adsorbed phosphate is by isotopic exchange (Olsen and Watanabe, 1957; Holford et al., 1974). This method involves shaking a soil sample in an electrolyte to which carrier-free 32 P is added. An aliguot is taken after a specified time and 31 P is determined. Reddy (1990) has suggested a least sguares fit for the determination of native P adsorbed on the soil surface. He suggested the following calculations:

PAGE 23

9 S 1 = k'c S Where : S 1 = amount of added P sorbed, nq g' 1 S p = Y axis intercept-representing native soil P in the adsorbed phase, /xg g" C = P in solution, /ug mL" 1 k' = linear adsorption coefficient (estimated without considering native adsorbed P, S p ) , mL g . Syers et al. (1973) reported that at lower concentrations the Eadie-Hof stee plot was more useful for evaluating P sorption than the conventional Langmuir equation, because the plot expanded the low P concentration region. The methods described above are based on the assumption that all surface-retained P is displacable either by isotopic dilution or sequential extraction techniques in the presence of a weak extracting salt solution. However, evidence suggested that with increasing contact period and temperature, P becames less ready to exchange with isotopically labeled P (Barrow and Shaw, 1975) . In addition, adsorption and desorption isotherms are different, and the relationship between the quantity of adsorbed P and concentration is not totally reversible (Sample et al. 1980) . Therefore, estimation of surface held native P based on the techniques described above might underestimate the true adsorption maxima and bonding energy values. Thus, any

PAGE 24

prediction of how the soil P status will change upon cropping must be considered unreliable (Bowman and Olsen, 1985) . The purpose of this study was to examine the effect of different organic amendments on P adsorption and desorption isotherms, and to devise a technigue to measure net P release from organic amendments during their decomposition. Such a technigue should lead to a direct and accurate estimation of the guantity of P which originated from decomposing organic amendments and adsorbed by soil. This knowledge will aid in the accurate estimation of soil P adsorption maxima and bonding energy. Materials and Methods All amendments mucuna ( [ Mucuna aterrima (Piper and Tracy) Merr] , kudzu [ Pueraria phaseoloides (Roxb.) Benth] , canavalia [Canavalia ensiformis (L.) DC], maize (Zea mays L.), peanut ( Arachis hypogaea L. ) , tephrosia [Tephrosia Candida (Roxb.) DC], cowpea (Vigna unguiculata ) , mixed gramineae (Grass) , aerobically processed city waste (PCW) , and chicken manure (CM)) were dried at 65 °C for 72 hrs for dry matter determination. Dried plant material was ground in a Wiley mill and passed through a 0.5 mm screen. A weighed (0.2 g) sample of the ground plant material was digested with H 2 S0 4 and H 2 0 2 for the determination of macro and micro elements. Total nitrogen was determined by the micro-

PAGE 25

11 Kjeldahl procedure. Potassium, Ca, Mg, Fe, Zn, Cu, and Mn were determined with an atomic adsorption spectrophotometer, and P was determined colorimetrically by the Murphy and Riley (1962) procedure. Incubation study Soil matrix . Soil used for this study was obtained from the Ap horizon of a Xanthic Hapludox (clayey, kaolinitic, isohyperthermic) (EMBRAPA, 1979) which had PZNC (point of zero net charge) at pH of 4.2. Five grams of organic ground material were mixed with 95 g of air-dried soil and incubated in plastic containers for 35, 65, and 150 d at 30 ± 3°C. The soil in each container was mixed thoroughly every 6 d and was kept at 45 % moisture content. After each incubation period, 2-g duplicate soil samples were extracted with 0.01 M CaCl 2 with a soil to solution ratio of 1:10 and 1 h shaking. The cycle was repeated three times. In calculating P desorbed, an allowance was made for the 2 mL of supernatant carried over from each cycle (Fox and Kamprath, 1970) . Adsorption was determined using 2-g air-dried samples, in duplicate, eguilibrated for 6 d with 20 mL of 0-60 nq mL" P solution in 0.01 M CaCl 2 in polyethylene tubes at 25 ± 2°C. To inhibit microbial activities a few drops of toluene were added to each tube. After 6 d the solution P was determined by the ascorbic acid method (Murphy and Riley,

PAGE 26

12 1962) . Phosphorus removed from solution was considered adsorbed. Desorption was studied by equilibrating the soil from the adsorption study in 20 mL 0.01 M CaCl 2 for 6 hrs (Singh and Jones, 1977) . Increases in solution P were measured and considered desorbed P. Adsorption and desorption isotherms were constructed and a standard Langmuir equation was fitted to calculate adsorption maxima and bonding energy (Olsen and Watanabe, 1957) . Simulated silica matrix (SSMT) . Hydrochloric acid (0.1 M) washed, fine silica was mixed with a nutrient solution containing N, K, Ca, Mg, Zn, Mo, Mn, Fe, Cu, B, S, and CI (P excluded) (EMBRAPA, 1976) to simulate the nutrient requirement for maize in a solution culture. Solution pH was adjusted to 5.0 with 0.1 M HC1. Table 2-1 highlights the chemical compounds used to make the nutrient solution and the resulting concentrations. The inoculation of silica matrix was carried out with microbes grown on potato dextrose agar. Five grams of ground organic amendments (the same ones used in incubation study with soil) were mixed with 95 g of silica matrix and incubated for 35, 65, and 150 d at 30 ± 3°C. The matrix was kept moist and mixed thoroughly every 6 d. After each incubation period, duplicate 2-g samples were extracted with 0.01 M CaCl 2 following the same procedure as described for soil. The sum of P detected by sequential

PAGE 27

13 Table 2-1. Ionic species and concentration of nutrient solution used with silica to simulate solution chemistry of a Xanthic Hapludox in a phosphorus release study by organic amendments. Source Species Cone 1 ". No. Chemical mmol (charge) L 1. Ca(N0 3 ) 2 .4H 2 0 Ca 2+ 2.690 NH 4 N0 3 NH 4 \ N0 3 " 9. 180 2. KC1 K*+ 1.829 K 2 S0 4 CI" 0. 670 KN0 3 s 0.480 3 . Mg(N0 3 ) 2 .6H 2 0 Mg 2+ 0. 650 4 . FeHEDTA Fe 3+ 0.035 5. MnCl 2 .4H 2 0 Mn 2+ 0.007 H3BO3 B 0.019 ZnS0 4 .7H 2 0 Zn 0.0018 CuS0 4 . 5H 2 0 Cu 0.0005 Na 2 Mo0 4 .2H 2 0 Mo 0.0006 f pH was adjusted to 5.0 with 0.1M HC1 . The reported concentration is considered adequate for maize growth in a solution culture. 4= The final concentration for K is calculated from KC1, K 2 S0 4 , and KN0 3 .

PAGE 28

14 extraction with CaCl 2 was considered total P released from the amendments because of the inability of silica to adsorb P. This value was used to construct a P release curve for each amendment. Calculation of total P adsorbed . Total adsorbed P was calculated as follows: TAP = N + S + 0A„ P P P where : TAP = total adsorbed P on soil surface, jug g" 1 N = native P present in adsorbed phase pg g" 1 S p = applied inorganic P in adsorbed phase /xg g" 1 0A p = phosphorus from organic amendment in adsorbed phase Hg g" 1 . The amount of P released from the organic amendments and adsorbed on the soil surfaces (OA ) was estimated by SSMT. The native exchangeable P on control samples was determined by the method of least squares (Reddy, 1990) . For comparison purposes, P adsorbed on the soil surface following incubation with organic amendments was also determined by sequential extraction with 0.01 M CaCl 2 .

PAGE 29

15 Results and Discussion Organic amendments differed in chemical composition (Table 2-2) . The highest P concentration was observed in CM (2.77 g 100 g" 1 ) and the lowest in PCW (0.16 g 100 g" 1 ) . The highest C:P ratio was for maize (189.0) followed by grass (180.0) . Soil used was obtained from the Ap horizon of a highly leached oxisol classified as Xanthic Hapludox (clayey kaolinitic, isohyperthermic) (EMBRAPA, 1979) . The soil was acidic (pH = 4.5), and low in total P (170 /xg g" 1 soil) and Mehlich-I extractable P (3.3 jug g" 1 soil). Clay content ranged from 78-82% and the Al saturation was 78.4% (Table 23) . The point of zero net charge measured by potentiometric titration in CaCl 2 solution was at pH 4.2 (Fig. 2-1). The native exchangeable P measured by the least sguares method was 3.0 /xg g" 1 soil.

PAGE 30

c 2 CP CP a. p u SB • • u u u * o H CP (0 u ro Eh i— I t— I i— I i— I i— I t— I H H H incMcoH'tfcovoincMin HnH(j>eooih^fflH 00000004000 voinvocninHcocorHcn CPcocPcocM^r-^cMCMH cococo^} , 'i , corHcM*j'' 3 0) • C (0 • Q) (0 •H (/I o u rc o — a) • Eh o O O o o o o o o O Q) •H 42 O H CO CO CP CO CM CO -P 0 CP *f o CM CP rH rH CP u H rH rH CO CM CM a) U) ro (0 CP rH rH O CP CP CO CO CM a) • • • • • • • • • • CO CO o CO CM o CP •H N ' «* CM CM CO O CO CM H o no H CM rH (0 n O QJ (0 IT) IT) o in in O in O in in a) (0 rH w CJ a CM CM in CM CM in CM in CM CM (0 O CP in Q H a > Q) (0 fO a) H in c O H m CM CO CM CO cm cp H CM CP CP O CM rH CO in in CO in CO a) s X u id Q) •rH c N B -rH Q) O «3 rH 3 2 S "ro c ro fO > e ro 0 c •rH e U a) rH N ' X rH ro o (1) ro •rH P CO -p 4: ro ro 0 rH n c • 0 c > •H ro 0 ro 3 c CT X ro C is u u ia ro p Efl c c •H Cr a) a> rH p > o ro C 4-1 Q) •rH e C cup Q) 0) •rH o Eh U -H ro c •H T5 (0 rH ro a> CP (0 TS p rH > •H O ro TS CD c C CP ro ro •rH + u u TS

PAGE 31

Table 2-3: Selected physical and chemical properties of Ap horizon of the Xanthic Hapludox used in adsorption studies. Parameters^ Measured Value Units of Measurement Bulk density Clay pH H 2 0, KC1 Net Charge Oxides Fe, Al Al Sat. (ECEC) Total P Mehlich-I P Al, Fe (P) 1.25 815 4.5, 4.2 0.22 0.075, 0.325 78.4 170.0 3.3 29.0, 1.4 g cm g kg" cmol (-) kg g 100 g" 1 % Hq g" 1 soil jug g" 1 soil Hq q~ soil

PAGE 32

10 U) o E o, CD D) CC 0 JZ o CD O i S CO +^ CD 10 Depth 0-20 cm PZNC = 4.2 CaCfe 2.0M 0.2M 0.02M 0.004M '. e-" v' / / PH 6 Figure 2-1. Acid-base potentiometric titration curves for the Ap horizon of Xanthic Hapludox with varying concentration of CaCl 2 .

PAGE 33

19 Adsorption Isotherms The results of adsorption and desorption studies are presented in Fig. 2-2 and 2-3. Incubation of the soil with amendments for 35 d influenced P adsorption. This was more pronounced at higher equilibrium solution P concentrations (Fig. 2-2a) . At a P concentrations >1.9 /xg mL" 1 , PCW amended soil adsorbed more P than the unamended soil. A similar result was obtained by Singh and Jones (1976) for organic amendments with low P content and high C:P ratio. On the other hand, soil samples amended with CM and Mucuna adsorbed less P than the unamended soil. As the time of incubation increased to 150 d, all amendments reduced P adsorption compared to the control (Fig. 2-2b) . A five-fold reduction in P adsorption was observed for the CM treatment at an equilibrium P concentration of 0.3 jug g" 1 soil. The reduction was perhaps due to net mineralization of P even from P-poor organic amendments with time. Another factor in P mineralization is the C:P ratio. When this ratio remains less than about 200:1, immobilization predominates during the initial stages of decomposition. But, as the decomposition proceeds, this ratio becomes narrower due to the concentration of P in decomposing residue and continuous degradation of C by microorganisms. As plants deplete soil solution P, the solution must be continuously recharged, if good growth is to be maintained. Recharge occurs when P is desorbed from the soil surface,

PAGE 34

20 and this will happen in sufficient quantity only if the soil has a large capacity to sorb and therefore desorb P. Thus, even though the soils with high sesquioxides content require more P to achieve a given level of P in soil solution, they have the compensating value of being able to supply P to the soil solution as it is taken up by plants. The results of desorption studies are presented in Fig. 2-3. Soil incubated with Mucuna aterrima (Mucuna) , PCW, and the control for 35 d did not show any difference in P desorption. However, desorption was higher for all amendments incubated with soil for 150 d (Fig. 2-3b) . In fact, CM reduced adsorption and increased desorption at all incubation periods. The observed differences of organic amendments in influencing P desorption characteristics of the soil could be attributed to their decompositional characteristics, P content, C:P and C:N ratios, and concentration of such elements as Zn and Cu. Incubation of soil with low P containing organic materials may not initially influence soil P desorption characteristics. Phosphorus release from organic amendments Phosphorus release data for different amendments in soil and silica substratum are presented in Table 2-4. More P was detected in the silica matrix for all amendments compared to the soil matrix. With increasing incubation time, a decrease in P was observed for maize and grass

PAGE 35

Figure 2-2. Phosphorus adsorption isotherms following 35 d (a) and 150 d (b) of soil incubation with Mucuna aterrima (Mucuna) , Chicken Manure (CM) , Aerobically Processed City Waste (PCW) , and Control.

PAGE 36

Figure 2-3. Phosphorus desorption isotherms following 35 d (a) , and 150 d (b) of soil incubation with Mucuna aterrima (Mucuna) , Chicken Manure (CM) , Aerobically Processed City Waste (PCW) , and Control.

PAGE 37

Table 2-4. Release of 0.01 M CaCl 2 extractable P (jig g" ) following incubation of organic amendments with soil and silica as matrix substratum. Org . T Soil Silica amend . 35 65 150 35 65 150 Days CM 109a* 204. ,9a 245. , la 447a 1031a 1380a Mucuna 21b 14, , Obc 16. ,9b 80b 124b 160b Tephrosia 22b 14, ,9b 19. , 0b 65b 116b 152b Cowpea 15c 11, , 5bc 16. , 2bc 43c 70c 92c Peanut 18bc 10, , 3cd 8, , 6bcd 41c 117b 142b Grass 6e 6, , 7de 3 , ,4d 37c 54c 102c Maize 14cd 12, , 5bc 3 , ,9d 37c 77c 111c PCW 3e 2, , 8e 10, , 8bcd 34c 62c 64d Canavalia 14cd 11, , 8bc 15, , 3bc 31c 83c 148b Kudzu lOd 11, , 4bc 14. , lbcd 23c 71c 94c Control 4e 4. . 4e 5. , 2cd Trace Trace O.le CV% 6.3 14, , 0 8. ,2 12.3 7.5 12.0 * Means in the same column followed by the same letter are not significantly different at the 95% level of probability, as determined by Duncan's Multiple Range Test. f Mucuna aterrima (Mucuna) , Pueraria phaseoloides (Kudzu) , Canavalia ensif ormis (Canav.), Zea mays (Maize), Arachis h ypoqaea (Peanut) , Tephrosia Candida (Tephrosia) , Vigna unguiculata (Cowpea) , Mixed Gramineae (Grass) , Aerobically Processed City Waste (PCW) , and Chicken Manure (CM) .

PAGE 38

24 treatments in the soil matrix. Processed city waste appeared to immobilize soil P as indicated by the P measurement made at 35 d of incubation but at 150 d a net release was observed. All leguminous amendments followed similar P release patterns. The highest P release was observed for CM treatment at 150 d of incubation (245 /xg g* 1 soil) and the lowest was for maize and grass treatments (3.9, and 3.4 /xg g" 1 soil, respectively) . When silica was used as the incubation matrix, there was a net release of P from all amendments which indicated that P released from amendments was being adsorbed by soil (soil matrix) and was not all exchangeable in sequential extraction with 0.01 M CaCl 2 . Maize and grass treatments which had shown no net P release at 150 d in the soil matrix released over 100 /xg g" 1 silica in the silica matrix. Given the similarity in molecular structure and chemical behavior of Si and P (Fig. 2-4) it is believed that added Si increases water soluble and easily extractable P (Adams, 1980) , and Si does not absorb P, because both of them in ionic forms are negatively charged. It is also suggested that silicate and phosphate ions compete for the same adsorption sites on Al(Fe) -oxide surfaces (Mekaru and Uehara, 1972) and they form insoluble precipitates with such common ions as Al, Fe, and Ca. This suggests that all P mineralized from organic amendments remained in the solution and was easily extractable by 0.01 M CaCl 2 .

PAGE 39

25 To understand the P release pattern from different amendments, surface response curves were fitted and are illustrated in Fig. 2-5. Phosphorus release in silica for maize, PCW, and CM followed a logarithmic function and the r 2 for PCW, CM, and maize treatments were 0.73, 0.82, and 0.98. A trace amount of P (0.1 /xg g" 1 ) was detected in the control treatment in the silica matrix confirming that the matrix was not contaminated and the P detected came solely from the decomposition of amendments. Lancrmuir Parameters Soil Matrix . A standard Langmuir eguations were fitted to the data after correction was made for preadsorbed P using sequential extraction techniques (Fig. 2-6) . This was followed by calculation of the adsorption maxima and bonding energy for each treatment (Table 2-5) . All but Pueraria phaseoloides (Kudzu) increased the Langmuir adsorption maxima at 3 5 d of incubation compared to the control. The extent of variation among different amendments ranged from 536 to 818 nq g" 1 soil at 35 d of incubation. It is noteworthy that PCW had the highest adsorption maxima (818 Mg g" 1 ) • The bonding energy was reduced for all treatments compared to the control. As the time of incubation increased from 35 d to 150 d, the adsorption maxima decreased for all amendments compared to the control treatment. But the

PAGE 40

O Shared 0 • Si 6d o #P @ ° Figure 2-4. A schematic representation of molecular structures of oxides of phosphorus (a) , and oxides silicon (b) .

PAGE 41

70.0 60.0 50.0 40.0 30.0 20.0 10.0 27 3 0.0 a) PCW Silica Soil = 0.32 + 0.07X r 2 =0.90 Silica = -29.5 + 19.6logX r 2 =0.73 Soil 20 40 60 80 100 120 140 160 1 ,600.0 1 ,400.0 1 ,200.0 1,000.0 800.0 600.0 400.0 200.0 0.0 '. b ) CM * A * * * * / _ / Soil = 100.8 + 1.03X / / * * » r 2 =0.77 t A Silica = 1716 + 628logX r 2 =0.82 20 40 60 80 100 120 140 160 120.0 100.0 80.0 60.0 40.0 20.0 0.Q C) Maize * * * * * * * Soil = 40.7 7.2logX t 0 1 * r 2 =0.92 t t Silica = 138.5 + 50.3logX r 2 =0.98 iii 0.5 0.4 0.3 0.2 0.1 20 40 60 80 100 120 140 160 0.0 d) Control Soil = 0.0002X + 0.05 r 2 =0.15 Silica = 0.0004X 0.02 r 2 =0.92 20 40 60 80 100 120 140 160 Incubation (days) Figure 2-5. Effect of incubation period and incubation matrix on P release pattern from different organic amendments in a laboratory study.

PAGE 42

28 Figure 2-6. Phosphorus adsorption isotherms for the Xanthic Hapludox (0.01 M CaCl 2 ) incubated with organic amendments for 150 d. The lines in the figure represents fitted Langmuir equation.

PAGE 43

29 cd p id u •H o c o -H -p & o in o rr, W c +i in rH CD -p (0 Q o IT) 01 > CO Q in VD > <0 Q if) co "3* CFi i— i VD o CO OA in CO CO i i VD CN CO CN o OA VD CN CN CN rH rH in CO CO OA O O o O O o o o o o rH CO CTi CO in in vo in P» CO CO rH CN O rCTi o\ CO rH o ca If) in vo VO in in in VD in X * X tfl X * X X X * * * X X C X X X X * CTi in vo H i o o o O O o o O O o o rH CD CO CN co rH oi VO rH rH rH •H o CO o CTl co l> rCO P co CN CO oo o o CN CN CO CA 0 a) o O o o o o o o O o o tfl t , 1*1 CN o p» CO r~in in [*» m CN p> r* in H t-~ rH CO o VD m VO VO VO in co in r-~ in w rH 0) > * X X X X * 0) X •X X X x X •X •X x X K rH rH r— rH VO OA o in co co co G\ vo vo VD co co OA rH O o o o O o o O O o o o • v_^ CO in Oi o VD CO co CO co CN co r-1 CN fl CN CN CN o CN rH CN VD (Q o o o o o o o o o o o If) O • o CO CN o CO VO co rH CO OA in in n CO vo vo CO H CN VD CO CD VD in in vo CO r» in -P -P * * * X X -X * ftJ * •X X •X X -X * X * CO CJ> o IX) CO CM -P CO o\ x> 0> X) OA VO VD P» OA c o o o o o o o o O o O o CO CO •H H rH tfl rH c0 p 0 fO CO o rT3 > CD U C CD tfl r< • C CO c N si 3 N ft tfl +J CT> i ft-p CD •H Eh o *H T3 Cfi cO CD r»-t TS P 0) -H tfl T3 CD (Tt tu C tT> *rn CO •H i l~l U CO >1 •H rH r-H tfl rH CO 0 CO u U 0 si -H •> 0 X! CD 0 * Eh U u CD CD CO 3 tfl (D O CD tfl CO si a -p 3 — C tfl CO W d) CO O CD O C Q -H > g CO (0 rH •H w tr. rJ •H CO u CD Q) (0 X H •H • ft 5 < B a u 3. 'ft "cO Cn c N CD CD CO*" •rH ft rH g 0 (0 a •rH g a s 0 c X X u CO CO g g >^ tfl CO > CO B c rH B (0 P 0) 0 0) •H i CO •H c >H rH u P CD u (0 •H ft CD CD 0 si C CT 1 P N •rH 0 0 C CO •H D T3 T3 CO c c CO C C CO O > ft £) 0 (0 (0 e c 9" CO & y X! u •H ft + > -H-

PAGE 44

30 bonding energy did not follow any defined pattern. For the grass treatment (mixture of different grasses) bonding energy value increased from 0.25 to 0.37 mL jug" 1 , for Canavalia ensiformis (Canavalia) bonding energy dropped from 0.39 to 0.26 mL Mg \ and for CM it remained at 0.10 mL g~ 1 . This finding does not agree with the data obtained by Hundal et al. (1988) where all studied amendments reduced the bonding energy as the incubation period increased from 20 d to 40 d. The range of P in organic amendments varied from 1.7 to 25.0 g kg" 1 on a dry matter basis. There was a high concentration of Zn and Cu in PCW, and Ca in CM. Being polyvalent cations, they have a high affinity for P adsorption through cation bridging (Holford and Mattingly, 1975; Haynes, 1989). The decomposition rate constants of the organic amendments were different and so was the release of different elements as the decomposition proceeded. These factors also may have contributed to the observed differences in P adsorption maxima and bonding energy. Silica matrix . The value of the total adsorbed P was adjusted based on the amount of P released from the organic amendments in the silica matrix. The Langmuir equation was fitted to the data, and the value of adsorption maxima and bonding energy were recalculated for selected treatments. The results are presented in Tables 2-6 and 2-7. The recalculated adsorption maxima and bonding energy values

PAGE 45

were higher than the values obtained based on the P release from the amendments in the soil matrix. At 150 d for the CM treatment, the adsorption maxima were higher by 1170 /ig g" 1 soil and the bonding energy was higher by 20 fold. The low adsorption maxima (496 fig g" 1 ) and bonding energy (0.10 nq mL" 1 ) for soil matrix indicated that the P capacity factor for CM applied plots was lower compared to plots which received other amendments and, therefore, CM plots should have the lowest residual P effect in field testings. But, the result from the Silica Matrix technique indicated otherwise. The high adsorption maxima for CM at 150 d of incubation (1667 jizg g" 1 ) and the high bonding energy (2.0 mL /ig 1 ) are indications of sustained P desorption and residual P availability. Predicted and measured phosphorus adsorption . The data on P adsorption maxima calculated by the Langmuir equation and actual P adsorption based on P release from amendments in soil and silica matrices are presented in Fig. 2-7. For the CM treatment, with increasing incubation time P adsorption maxima increased. Continued adsorption beyond adsorption maxima indicated the presence of a precipitation reaction, or multilayer adsorption. For tephrosia and kudzu, the adsorption maxima decreased with time. A good agreement among adsorption maxima and actual adsorption, calculated based on SSMT, was observed at higher equilibrating solution

PAGE 46

32 u to u 21 o • o Xj p ft -H 4H o X •H a) ^ in p (0 (0 a) rH a) to ^ o •H P a) "H c w C TS o a) T3 (1) W (0 P rC rH B •H 01 xi i 10 XI VO T3 0) (0 a) a) rH g xt (0 EH T3 0 -H a) ft c o •H P (0 x 0 c m O cn rn in in o co VO VD o CM o o o cn r> > co VO IT) co rro CO VO CM co VO (0 IT) 1X1 in VO X) 0 o •H rH rH •rH * * * * w * * * * * O m a> O cn CTl (M M o rH O o O VO rO in r> IT) H A! O rH o o o cn > Q in ta > (0 Q in ro h— T3 • C en a) n n O in o\ *r in cm cm vo H in vo r> r> ***** ***** VO O > O CM C7\ O CTl CTl CT< O H O O O CM H CO CM CM VO CM CO O O O O O en cm co cn h o o o vo rcn in cn r***** ***** > cn in rCO 0> Cn Cn Cn O O O O O 03 •H CO O U 3 W 43 N W CD H O ft P (0 p c to o •H 4H •H c Cn •H CO * * a) (0 0) B •rH (0 •a 0) 0) M C (0 -rH cn o Jh ft C a) o -H ax: 0) u Eh C (0 •rH c (0 O (0 •H 10 O rl x: o CD Eh =1 CO EJ cn 0) T3 (0 9E u ft 1 p •H u h cn CD W 3 to ft ^ o cn ft Cn (0 0 H X id 6 c o •H p Vh o CO TS (0 ft ft Cn SL e >1 Cn >H CD C CD Cn c •H T3 C o X!

PAGE 47

1-1 >1 71 o « «i (0 Q o in > CO Q in (fl > CO Q in + T3 • C c i O t0 H N H (M N O H O O O O O O O O o o cn in cn m a\ cm cm cm O H O O O • CO CM C\ • o • • • co in vo r> vo (M h O M h NOOHP1 cm o r~ vo Tf (J\ (M H rt o o o o o o o o o o > CTl • ^ • co • in • CO in M* H H id o w n I s O CO ID c^ rh i< 'j* CO I rl (N| O O H o o o o o o o o o o 01 fl H O h i— i in o o o o o o o o CM CO ^ (Tl — • • o\ • n "a* in • in • h r> in in rH H rl CM fl fl in a\ h in vo cm cn I CM CO -H n o ^ 3 uj -r-> m ft T3 S to a) to a) c •H S (0 M O n $ x •H S S -H C (fl a) 0 * U 0 & -H ft£ a) u C to to rs •H c to u u a< f0 •H a) [fl p o Ul u CO x: a o E-t p •rH u 3 T3 n a) T> Ul « Q) 0 o [fl O CD [fl fO x: u fO u •H o ft u 0) CO H to U Ul QJ Ul 3 fO (X ^ 3. fO e •H to e c o •H P ft o tfl fO o< >i CP ft 0) c 0) Cn C •H c o XI I I '"0 fO p Ul 0) M CO Ul -H Ul 0) X! p c 0) p. (0 ft 0) XJ p c H Ul u X5 1 53

PAGE 48

Incubation Time Figure 2-7. A comparison of P adsorption by Xanthic Hapludox based on the estimation of preadsorbed P by sequential extraction of soil or silica matrix incubated with organic amendments.

PAGE 49

35 concentration (60 nq g' 1 soil) which indicated the validity of this technique in predicting P requirements of amended soils. In highly weathered soils, extractable P is usually low but the amount of P which the soil can immobilize varies greatly because of the variation in the reactive surfaces. Addition of organic amendments changed the reactive soil surface as indicated by the change in adsorption maxima and bonding energy. Conclusion Decomposition of organic amendments influenced P adsorption and desorption by soil. Amendments with low P content (PCW, maize, and grass) immobilized P during early incubation periods. Higher amounts of P were desorbed from soil samples incubated with CM. Soil incubated with Mucuna, PCW, and the control for 35 d did not show any difference in P desorption. However, desorption was higher for all amendments incubated with soil for 150 d. Incubation of soil with low P-containing organic materials may not influence soil P desorption characteristics initially. Amendments' P content and C:P ratio did not prove to be helpful in predicting P mineralization. Amendments with a C:P ratio of 1:139 (PCW) immobilized P during their decomposition while others with similar or even higher C:P ratios had a net P release in soil matrix. All amendments

PAGE 50

36 released P in the silica matrix independent of their P content and C:P ratio. The use of SSMT to measure release of P from decomposing organic amendments aid in the calculation of adsorption maxima compared to the segential extraction of amended soil with 0.01 M CaCl 2 solution. The calculated values were in close agreement with actual adsorption measured at high eguilibrium P concentration.

PAGE 51

CHAPTER III ORGANIC AMENDMENTS AND CROP PHOSPHORUS NUTRITION Phosphorus Management Strategy In highly leached soils, the minerals with permanent charge have been either severely altered or completely weathered out, so that the surface charge arises from adsorption of potential determining ions such as hydrogen and hydroxyl . The magnitude of the surface charge is expressed by a combined Gouy-Chapman and Nerst eguation. This eguation provides a theoretical basis for increasing the cation retention capacity of a soil by lowering pH Q (value of soil pH at which net surface charge is zero) . One way of lowering pH 0 is to increase the organic matter, phosphorus, or silica content of the soil (Uehara and Gillman, 1981) . Highly weathered soils are also very poor in total and plant available P. Their high P fixing capacity reguires high doses of applied P to meet crop demands. The classic work of de Wit (1953) on physical theory of fertilizer placement predicts that when suboptimum quantities of fertilizer are used, restricted placement is desirable (Fox et al., 1986). de Wit based his analysis on nutrient uptake 37

PAGE 52

38 by entire root mass, and part of the root mass immersed in nutrient solution and established a relationship between Ur (g of nutrient taken up by the plant when part of the root mass, Xr, was immersed in the solution) and Ub (g of nutrient taken up by the plant when the entire root mass was immersed in the solution) . Ur/Ub = (Xr/Xb) 0 '" This relationship appears to be independent of crop type and nutrient solution concentration. Under field conditions Ub, Ur, Xr, and Xb take on new meanings. Ub = uptake rate from broadcast fertilizer Ur = uptake rate from banded fertilizer Xr = width of the fertilizer band Xb = distance between the crop rows The use of data from P sorption curves concerning soil solution P concentration to predict P uptake and crop yield in combination with P placement analysis offers a way to increase fertilizer use efficiency. Faced with a high P fixing soil and a small quantity of fertilizer, the fertilizer can be used to the best advantage by concentrating it in a band so that the P concentration in the soil solution is identical to the concentration that produces a maximum yield in a broadcast application. Any deviation from this optimum value, either to higher or to lower concentrations, leads to less than an optimum return per unit of fertilizer input.

PAGE 53

39 Management of Organic Amendments Organic amendments can affect the reaction of P in the soil through complexation of polyvalent cations which are major phosphate adsorption sites. It is widely believed that humus in association with cations such as Fe 3+ , Al 3+ , and Ca 2+ retains significant amounts of P. Appelt et al. (1975) prepared a hydroxy-Al-humic acid complex that adsorbed P because of the creation of new P adsorption sites. They concluded that any increase in organic content of a soil could lead to greater adsorption. A study conducted by Swift and Haynes (1989) also showed that Al-organic matter associations have a significant phosphate adsorption capacity. Indeed, the Al-humate adsorbed amounts of phosphate similar to those commonly reported for Al and Fe oxides (McLaughlin et al., 1981) on a w/w basis. However, simple organic acids, fulvic acids, and humic acids had no effect on P adsorption by volcanic ash-derived soils. For these soils, P was preferentially adsorbed over the organic acids studied (Appelt et al.,1975). The addition of organic materials to high P fixing soil can decrease, increase, or leave virtually unaltered the P fixation capacity (Yuan, 1980) . The reduced fixation is the result of: 1. complexation of Fe, Al, and Ca by organic anions (Larsen et al, 1959), 2. competition of organic anions and P for the same adsorption sites (Nagarajah et al, 1970), 3. development of organic coatings on mineral surface

PAGE 54

40 (Easterwood and Sartain, 1987), and 4. reduction in bonding energy of adsorbed anions resulting in low residence time for adsorbed P (Hudal et al, 1988) . Increased adsorption may be due to: 1. cation bridging between organic anions, and Fe 3+ , Al 3+ , and Ca 2+ leading to formation of new sites for P adsorption (Appelt et al, 1975), and 2. ability of organic ligands to maintain hydroxy-Al, and Fe in a non-crystalline state and thus maintaining a greater surface area (Swift and Haynes, 1989) . Amendments chemical composition . Although organic acids are an integral part of all organic matter by far they are not the only reactive component influencing P adsorption. Singh and Jones (1976) suggested that the P content of organic residues plays an important role in the release or fixation of added P. Similarly, chemical and decompositional characteristics of the organic matter may influence total C0 2 evolution (Sweeney and Graetz, 1988). This gas when dissolved in water, forms carbonic acid, which is capable of decomposing certain primary minerals. Elemental ratios such as C/N and C/P are also considered valuable indicators for net mineralization or immobilization of N and P contained in organic amendments. One of the important studies in this area was conducted by Blair and Boland (1978) who examined the release of P from white clover residue in high and low P status soils. Their results suggested that the addition of

PAGE 55

41 plant material resulted in immobilization of soil P only in the low P soil in the absence of plants. In the high P soil no immobilization of P was observed. Crop residues have an effect on the nutrient status of the soil. The cumulative effects of increasing guantities of organic residues on available nutrients in soil were studied for 11 years by Larson et al. (1978) . They reported that addition of 16 tons/ha of plant residue per year to the soil increased the amount of N, S and P by 37, 45 and 14% respectively, over the control treatment. They also found that the NH 4 -N production, weak acid soluble P and exchangeable K in the soil were increased as a result of increasing the addition of organic residues. Solubilization of Mn and Fe in soil were affected by wheat straw and alfalfa amendments (Elliot and Blaylock 1975, Sims, 1986) . The release was greater for Mn than Fe and also much higher at 30 kPa than 50 kPa moisture tension. The release of Mn and Fe from the soil column followed the following order: alfalfa > wheat straw > soil alone. They suggested that the potential accumulation of soluble Mn in well drained soil was possible where there was a large guantity of plant residues incorporated into the soil . Application technigue . In the case of organic manures, changes in organic P will to some extent depend on whether the material is left on the surface of the soil or is plowed

PAGE 56

42 in. Douglas et al. (1980) reported that the method of placement, composition of residues and loading rates were important factors influencing mineralization or immobilization of N and S. A higher mineralization rate of N from residue incorporated at 4 cm soil depth as compared to the surface was also observed by Brown and Dickey (1970) and Cocharan et al. (1980) . Data on P mineralization from organic matter in the literature is scarce. Inorganic Phosphorus Management Strategy Salinas and Sanchez (1976) have outlined a three-point strategy for P management under limited resource conditions. 1. Use of cheaper sources of P. Two main sources are phosphate rocks (PR) and thermally altered sources, such as basic slags and the Rhenania phosphates. Numerous reports have appeared in the literature regarding the fertilizer value of PR as compared to other sources of P fertilizers, e.g., superphosphate (Khasawneh and Doll, 1978; Hammond et al., 1986; Hernandez and Sartain, 1985). Recently, Heliums et al. (1989) compared the potential agronomic value of some rock phosphate from South America and West Africa and concluded that in addition to P the PRs with medium to high reactivity have a potential Ca supply value. 2. improved soil test interpretations, and 3. improved placement methods .

PAGE 57

43 Fertilizer placement . The advantage of banding phosphate fertilizer is well known. What is generally not understood is the sensitivity of nutrient uptake to band width, de Wit (1953) demonstrated that a soil that is virtually incapable of supporting a crop with 100 kg/ha of broadcast phosphorus will produce nearly 50% of maximum yield with the same amount of fertilizer applied in a narrow band. A more detailed description of de Wit's analysis and some of the assumptions contained in the analysis are further elaborated by van Wijk (1966) , Uehara and Gillman (1981), and Fox et al . (1986). In agreement with the theory Fox and Keng (1978) reported a better response from localized P placement as compared with complete incorporation if suboptimal P rates were used, but if quantities of P were sufficient, best results were obtained from incorporating P in the entire soil volume. Kamprath (1967) found that similar maize yields were obtained by annual banded applications of 22 kg P/ha for seven years as were obtained by an initial P application of 350 kg/ha. Banding, therefore, saved more than half of the P requirement. Applying N and P in knifed bands has been shown to be an effective method of applying N and P to winter wheat (Leikam et al., 1978, 1983). Experiments on maize have shown that dual-placed N and P increases P uptake and maize grain yield more than when P is banded to the side or below the seed (Raun et al., 1987).

PAGE 58

44 Band spacing of applied N and P fertilizer affects the concentration of these nutrients in the applied band and the probability of roots contacting the band. Sleight et al. (1984) showed that in high P-fixing soil, increasing the probability that root-fertilizer contact will occur is more important than reducing soil-fertilizer contact during the first week of oat ( Avena sativa L. ) growthThere is also a threshold value of soil solution P concentration beyond which the P uptake rate does not increase (Jungk, and Barber, 1974) . Anghinoni and Barber (1980) in a P placement experiment reported maize root growth stimulation in the portion of the soil where P was added. Maximum shoot dry weight in their experiment was obtained by placing the fertilizer in 0.25 of the soil volume . Summary Application of P in narrow bands results in reduced fixation of applied P and improves crop P nutrition. This belief is based on: (i) localized P is protected to some degree against irreversible adsorption or precipitation reactions with the soil, (ii) localized P may be more readily accessible to seedling roots than P widely distributed in the soil, and (iii) plants can be adeguately supplied with P through a few roots which proliferate in the fertilizer band. Further, combined application of organic

PAGE 59

45 amendments with inorganic P in a band will reduce direct contact of inorganic P with a large volume of soil . In this context, the overall objective of this research was to devise a technigue to sustain P nutrition in a highly leached Oxisols with the help of organic and inorganic P sources applied in narrow bands. The specific objectives were to: (i) study the effect of different amendments and their rate of application on maize dry matter production and herbage nutrient concentration, (ii) examine the beneficial effect of the combined application of organic amendment with inorganic P in a narrow band as measured by maize grain production and soil nutrient dynamics, and (iii) measure the residual effect of applied treatments in relation to selected soil chemical characteristics. Materials and Methods This experiment was conducted at the Empresa Brasileira de Pesguisa Agropecuaria (EMBRAPA) station located 30 km north of Manaus at an elevation of 50 m in the Amazonas state of Brazil (Fig. 3-1) . Climate in the Manaus region has been classified in the Koppen nomenclature as afi, tropical, humid and hot (Goes and Ribeiro, 197 6) . The soil used in the study has been classified as Xanthic Hapludox (clayey kaolinitic, isohyperthermic) (EMBRAPA, 1979) .

PAGE 60

46 Glasshouse Experiment A factorial arrangement of three factors to yield 18 treatment combinations was employed in a randomized complete block design (RCBD) with four replications (Table 3-1) . Five kg of unlimed soil from the Ap horizon of a Xanthic Hapludox was mixed in a pot with organic amendments (<2 cm long) 7 d prior to maize seeding. Maize (variety BR 5110) was planted and thinned to 2 plants per pot 7 d after planting. The experiment was harvested at 65 d. At harvest, leaves immediately below and opposite to the ear leaf (70% plants had begun to develop ear) were collected for chemical analysis. A 0.2 g sample of the ground leaf tissue was digested with H 2 S0 4 and H 2 0 2 . Potassium, Ca, Mg, Fe, Zn, Cu, Mn were determined with an atomic adsorption spectrophotometer, and P was determined colorimetrically using the Murphy and Riley (1962) procedure. Soil P was extracted with the Mehlich-I extractant P (0.05 M HCl + 0.0125 M H 2 S0 4 , with a soil solution ratio of 1:10, and 5 min shaking time), unbuffered 1 M KC1 (1:10 soil : solution ratio) was used for the determination of extractable Al. Aluminum was determined by titrating the extract with 0.1 M NaOH to bromthymol blue endpoint. Soil reaction (pH) was determined in water using a soil: water ratio of 1:2.5 and in 1 M KC1 using the same ratio. Maize dry matter production was recorded and surface response curves were fitted in the case of interaction effects among factors.

PAGE 61

Figure 3-1. Geographic location of Amazon basin in Brazil (a) , on-station and farming systems research (FSR) sites (b) , and effective rainfall during the period of August 1988 until August 1989 (c) at EMBRAPA station in Manaus, Brazil.

PAGE 62

Table 3-1. Factorial arrangements of treatments for the glasshouse study. Type Amendments''' Rate P Equivalent kg ha" 1 Mucuna 1&2 8.8, 17.6 Organic CM 1&2 8.8, 17.6 PCW 1&2 8.8, 17.6 Inorganic TSP 1,2&3 0, 8.8, 17.6 f Mucuna ( Mucuna aterrima ) , CM (Chicken Manure) , PCW (Aerobically Processed City Waste) TSP (Triple superphosphate ) Field Experiment The site was cleared by burning an existing sugarcane crop. Sugarcane stems and rhizomes were taken out of the field. A uniformity trial using maize as an indicator crop was planted for 60 d in order to record the existing variation in the field. Visual field ratings combined with soil and plant chemical analysis results were used to select a uniform area for planting the field experiment. The selected area was divided into 48 plots and soil samples were taken from three consecutive depths (0-15, 15-30, and 30-45 cm) in each plot. Dolomitic lime (2 Mg ha" 1 ) was applied over the entire area 2 weeks prior to maize planting. Nitrogen and K were applied at the rate of 200 and 100 kg ha* 1 . Zinc, Mo, and Mn (2 kg ha" 1 of each) were applied in the band. Half of the N

PAGE 63

49 as urea was broadcast over the entire plot and the remaining half was sidedressed at 35 and 70 d after planting. All K was applied basal broadcast. Six furrows each 20 cm wide and 8-10 cm deep were opened in each plot 80 cm apart. The plot size was 10x5 m. Mucuna aterrima and Canavalia ensiformis grown for 60 d in an adjoining field were harvested and passed through a hay chopper to produce a uniform size of less than 6 cm. All organic amendments were applied in bands to which different rates of P from triple superphosphate (TSP) was added. A soil cover about 2 cm thick was put over the amendments. Twelve treatments were tested in a RCBD with 4 replications. The P release information from the SSMT technique (Chapter II) was used to calculate the amount of a given amendment required to supply P equivalent to 8.8 and 26.4 kg ha" 1 . First maize . Maize variety BR-5110 was planted on December 27, 1988. Plant population was adjusted to approximately 55 x 10 3 plants ha" 1 during the first sidedress at 30 d after planting. Soil samples were collected before planting, during tasseling (65 d after planting) and within a week after the harvest. Leaf samples were collected during tasseling. They were taken from immediately below and opposite to ear leaf. The inner four rows were harvested for grain. Grain production was recorded at 15% moisture level.

PAGE 64

u 'S'CO^l'CO'tfCO^i'CO CO VO VOCOVOCOVOCOVOCO N N N N o o VO CM O O OOOO[2!2V0CN VOCNVOCNUUgg oo r» vo H CM o o o CN VO (U Pn Pn CO CO CO Eh Eh Eh u H 55 < CD « o u H < o K o 55 VO VO VO VO rr~rH H rl H O O CN CM O4 cu co co + + o o CM CN g u o (M O CN CO Pa Eh CO + Eh O + CM O 5 CM U g cu u o p c o u CO CD H ast hat 0 (d Eh ^ CD U C r4 (0 04 CD rd a) C CO -P -P (C T3 g (0 Eh C (0 •H 01 U 0 rH CO B CO CO ro CO 0) -g" c >
PAGE 65

Second maize . This trial was conducted to measure the residual effects of the treatments applied during the first maize crop. Maize stover was taken out of the field. Nitrogen was applied in the same manner as for the first crop. No other nutrient applications were made. Maize (variety BR-5110) was dibble planted in the rows and the population was adjusted to 55 x 10 3 plants ha" 1 . Soil samples were collected during tassel ing and after harvest. Again 1020 maize leaves were sampled during tasseling initiation from each plot. Maize grain production was recorded from four inner rows and was adjusted to 15% moisture content. Chemical Analysis of Soil and Plant Materials Soil pH was determined in water, 0.01 M CaCl 2 , and 1.0 M KC1 using a soil : solution ratio of 1:2.5. Total P was determined by wet combustion method and inorganic P fractionation was carried out using Chang and Jackson procedure (1965). The pH of NH 4 F was adjusted to 8.2. Mehlich-I extractant (0.05 M HC1 + 0.0125 M H 2 S0 4 ) was used to extract soil P, K, Zn, Cu, and Mo. Phosphorus was determined colorimetrically and K by a flame photometer. Aluminum, Ca, and Mg were extracted with unbuffered 1 M KC1 (1:10 soil : solution ratio). Aluminum was determined by titrating the extract with 0.1 M NaOH to bromthymol blue endpoint, and Ca and Mg were determined with an atomic absorption spectrophotometer. Soil apparent density was

PAGE 66

52 determined by the method of a measuring cylinder. The cylinder was struck against a rubber pad 10 times from a distance of 10 cm. The final weight of the cylinder was taken and the apparent density was calculated as the following: Apparent Density (g cm" 3 ) = Weight of dry soil at 105°C / Volume of the soil in the cylinder. Particle size distribution was measured by the pipet method. All organic amendments and maize leaf samples were dried at 65 °C for 72 hours for dry matter determination. Subsamples of the dried plant material were ground in a Wiley mill to pass a 0.5-mm screen. A 0.2-gm sample of the ground plant material was digested with H 2 S0 4 and H 2 0 2 . Phosphorus was determined calorimetrically , and K, Ca, Mg, and micronutrients were determined with an atomic absorption spectrophotometer . Results and Discussion Selected chemical and physical properties of soil used in the glasshouse and field studies are presented in Tables 3-3 through 3-5. With increasing soil depth, a reduction in Al-P was observed. But, the Langmuir adsorption maxima increased with the depth (789 /ng g" 1 soil at 30-45 cm) (Table 3-3) . It is interesting to note that with over 80% clay content the soil had a hydraulic conductivity of 25.1 cm h" 1 (Table 3-4). The ammonium oxalate extracable Fe and Al were high and so was

PAGE 67

the Al saturation (>76.0%) calculated from the effective cation exchange capacity (ECEC) (Table 3-5) . Glasshouse Study An analysis of variance (Table 3-6) presented for maize herbage dry weight yield indicated the presence of a three way interaction among types of organic amendment, rate of amendment, and rate of TSP. Among the three amendments tested, PCW produced the lowest yield when applied alone or in combination with TSP (Fig. 3-2a) . It is interesting to note that when PCW was applied at higherr rate (equivalent to 17.6 kg ha" 1 of P) it produced less dry matter per pot compared to when this amendment was applied at lower rate (Figure 3-2a) . A possible explanation is that the slow rate of release of nutrients from the material resulted in an initial P deficiency for maize seedlings. There was also a high concentration of Zn and Cu in this material (280, 155 /ng g" 1 , respectively) which may have played a role in providing cation bridging between organic and P anions making P less available (Murray and Linder, 1984) . By forming organo-metal complexes, humic and fulvic acids as well as simple acids can dissolve or decompose such minerals as feldspar, gibsite, goethite, hematite, and mica (Schnitzer, 1977) . Therefore, organic amendments with high Zn and Cu content may have limited potential as a source of P in highly leached soils.

PAGE 68

The highest dry matter yield per pot was obtained with CM (Figure 3-2b) . Combining the application of CM and TSP provided a higher dry matter yield at rate 2 (equivalent of 17.6 kg ha" 1 of P) compared to rate 1 at 0, 8.8, and 17.6 kg ha" 1 P as TSP. There was a strong relatioship between Mehlich-I extractable P and dry matter production (r 2 =0.88), and tissue P concentration and dry matter production (r 2 =0.79) (Figure 3-3) . The relationship was linear for both indicators. Several researchers have attributed the positive effect of organic amendments in improving P nutrition to a change in soil pH (Sanchez and Uehara, 1980) which improves the plant growth conditions and increases the solubility of native P. Soil pH increased approximately by one unit (4.55.5) in response to the application of PCW (Table 3-7). But this treatment produced the lowest dry matter. Application of Mucuna aterrima also improved the soil pH but the magnitude of improvement was less by 0.4 unit. The control treatment and CM had essentially the same pH, but CM had higher dry matter production.

PAGE 69

55 Table 3-3. Selected chemical properties of the fine fraction (<2 mm) of the Xanthic Hapludox. Depth Org. Chang & Jackson ' Langmuir carbon TP OP Al-P Fe-P Ca-P MI-P max''". cm g kg" 1 g"1 Hq 0-15 14.6 200 25.2 51. 6 0.3 0.3 2.4 550 15-30 12.2 120 14.5 18.5 2.8 0.2 1.8 620 30-45 8.3 90 6.7 5.2 0.1 0.1 1.2 789 f pH of NH 4 F was adjusted to 8.2. The reductant soluble P is not included. 4= 0.01 M CaCl 2 was used as an electrolyte. OP = Organic phosphorus, TP = Total phosphorus MI-P = Mehlich I extractable P. Table 3-4. Selected physical properties of the Xanthic Hapludox. Depth Bulk Fine fraction Hydraulic Density Sand Silt Clay'" Conductivity cm g cm -3 — kg kg" 1 of <2 mm — cm h" 1 0-15 1. 11 0. 14 0. 11 0.75 25.1 15-30 1.30 0.11 0.06 0.88 6.2 30-45 1.16 0.09 0.09 0.82 7.3 f Fine sand fraction is also included.

PAGE 70

56 P o (0 w w u w n * CM • • • 1 rH CO I G> o ID co o m rn CO o rH • o • o • o Cn loo Cft l o o o • * o o o r» m rn r» l o H rH j VD O <* ! cn
PAGE 71

57 Table 3-6. Analysis of variance for maize herbage dry weight production per pot in the glasshouse study. Source"'" DF F Value Pr > I REP 3 1.4 0.23 ORGANIC 2 970.2 0.00 RATE 1 189.2 0. 00 ORGAN I C * RATE 2 144 . 0 0.00 INRATE 2 127.9 0.00 ORGANIC* INRATE 4 17.6 0.00 RATE* INRATE 2 0.1 0.86 ORGAN I C *RATE * I NRATE 4 8.3 0. 00 ERROR 51 CORRECTED TOTAL 71 CV% 13.0 f REP = Replication, ORGANIC = Organic amendments, RATE = Rate of Organic amendments, INRATE = Rate of Inorganic P.

PAGE 72

PCW 30 25 O20 Q. CO c 15n 10 Y = 13.83 + 0.30X r2 =0.76 Ra tel a) / Y = 7.84 + 0.04X r 2 =0. A 5ns A Rate 2 A A A 1 1 1 l 1 120 100 20 CM Y = 64.9 + 0.99X r 2 =0.87 Rate 2 A A 58 A .•"'A Y = 41.62 + 0.55X r2 =0.86 4.4 6.8 13.2 17.6 22.0 0 4.4 P, kg ha -1 Mucuna 100 90 80 O 70 CL 60 50 40 30 20 Y = 62.5 + 0.63X Rate 2^ r 2 =0.88 A A A i y ' Ratel / Y = 35.1 + 0.97X r * =0.87 ' c) 1,1, 4.4 8.8 13.2 17.6 22.0 P, kg ha -1 8.8 13.2 17.6 22.0 Figure 3-2. Effect of different rates of selected organic amendments in combination with different rates of inorganic P on maize dry matter production at 65 d after planting in a glasshouse study.

PAGE 73

59 Y = 0.45 + 2.75P a) r 2 =0.88 _E_J_ b) vY = -24.6 + 309.5P r 2 =0.79 0 5 10 15 20 25 30 35 Soil P, ug g 1 (M-l) 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 % Leaf P Figure 3-3. Relationship between Mehlich-I extractable soil P (a) , and maize leaf tissue P concentration (b) with maize dry matter yield.

PAGE 74

60 Table 3-7. Mean canges in soil pH (H 2 0) following application of selected organic amendments in the glasshouse study Amend"'" . Rate 5 Obs. Mean pH SD Mucuna 1 12 5.09 0.22 Mucuna 2 12 5.47 0.20 CM 1 12 4.80 0. 14 CM 2 12 4.87 0.26 PCW 1 12 5.48 0. 19 PCW 1 12 5.85 0.23 Control 0 4 4.53 0. 09 TSP 1 4 4.72 0.27 TSP 2 4 4.70 0.08 4= TSP = Triple superphosphate, CM = Chicken Manure, PCW = Aerobically processed city waste. § Rate 1 & 2 are eguivalent to 8.8 and 17.6 kg ha" 1 of P.

PAGE 75

61 Field Study Rainfall distribution data for both cropping cycles are presented in Figure 3-lc. The distribution is bi-modal with a short dry season. During this period evaporation is greater than precipitation. Grain Yield . Single degree of freedom orthogonal contrasts of maize grain production during the first crop (Table 3-8) indicated that one application of PCW equivalent of 26.4 kg P ha' 1 produced only 1.41 Mg ha of maize. This level of production was inferior to the control treatment. Canavalia ensiformis . Mucuna aterrima , and TSP when applied separately to provide 26.4 kg ha* 1 of P produced the same amount of grain during first and second cropping cycles. But the CM treatment produced more grain than other treatments in both the first and second crops. This was expected based on the results of the adsorption study presented in Chapter II. One explanation is that the faster rate of decomposition (0.26 g/100 g d" 1 ) and the high P content in CM maintained the soil P concentration at a high level from the beginning of plant growth. In a review paper, Olsen and Barber (1977) concluded that an annual application of manure and superphosphate resulted in an increased level of 0.01M CaCl 2 and 0.5 M NaHC0 3 -extractable P. In most studies, manure treated soils tend to support a higher level of soluble P than soil treated with an equivalent amount of superphosphate .

PAGE 76

62 Table 3-8. Orthogonal contrasts of maize grain yield under different treatments applied in a 30 cm wide band at UEPAE research station, Manaus, Brazil. Treatment"'" Mean Grain Yield Cropl — Mg Crop2 ha" 1 -Contrasts Cropl Crop2 M60 3.44 1. 65 C60 VS TSP60 ns ns C60 3.56 2.07 CM60 VS TSP60 ** ** PCW60 1.41 1.67 PCW VS TSP60 ** ns CM60 4.63 2.37 M60 VS TSP60 ns ns M20+TSP20 3.55 1.41 C2 0+TSP2 0 VS TSP4 0 * ns C20+TSP20 3.62 1.81 M20+TSP20 VS TSP40 ns ns PCW20+TSP20 1.66 1.90 PCW20+TSP20 VS TSP40 ** ns CM20+TSP20 4.10 1.92 CM20+TSP20 VS TSP40 ** ns Control 1.71 1. 10 TSP20 2.63 1.04 TSP40 3.12 1.46 TSP60 3.69 1.63 *, ** Significantly different at 0.05, and 0.01 level of probability, ns = not significantly different at 0.05 level of probability. f M60, C60, PCW60, and CM60 = Mucuna, Canavalia, Aerobically processed city waste, and Chicken manure applied to provide equivalent of 26.4 kg ha" 1 of P. TSP20, TSP40, and TSP60 = 8.8, 17.6, and 26.4 kg ha' 1 P from triple superphosphate (TSP) .

PAGE 77

Combined application of 8.8 kg P ha from organic amendments and the same amount from TSP giving an application rate of 17.6 kg P ha' 1 P was compared to 17.6 kg P ha" 1 from TSP. Differential results relative to amendment source was observed. For the first crop, the combination of TSP with CM increased yield 30% relative to TSP. Mucuna aterrima did not influence yield while Canavalia ensiformis combined with TSP produced higher yield (3.6 compared to 3.1 Mg ha" 1 for TSP) . Chicken manure was still the best of all treatments during the second crop. But, PCW performance improved over the first crop. And there was no difference between PCW 60 and TSP60, and PCW2 0+TSP20 combined. Such findings highlight the limitations of incubation studies conducted in the laboratory for a short span of time in predicting long term effects of organic amendments on soil P dynamics . A response surface plot (Figure 3-4) for TSP indicated a linear response between grain yield and rate of TSP up to 26.4 kg P ha" 1 (r 2 = 0.90). The magnitude of this response was declining in the second crop but the regression analysis did not show any difference with varying rate of TSP. Change in soil phosphorus status . The change in P status of the soil over the 240 d cropping cycle for all treatments, except PCW, followed a cubic surface response (Figure 3-5) . The data presented are for 0-30 cm depth. All treatments improved soil P status compared to the control.

PAGE 78

Figure 3-4. Rate of inorganic P applied through triple superphosphate and its effect on maize yield in a maize-maize rotation.

PAGE 79

Trt. a *>1 b 2 *>3 R Z TSP60 4.79 5. 53E-01 -6. 30E-03 1.65E-05 0.75 M60 4.77 4. 08E-01 -3. 49E-03 7.58E-06 0.85 Control 4.05 -6. 92E-03 -2. 00E-04 7. 18E-07 0.95 PCW60 4.07 3 . 96E-02 -3. 08E-04 7.91E-07 0. 15ns C60. 3.62 3 . 02E-01 -2. 56E-03 5.56E-06 0.99 Figure 3-5. Effect of selected organic amendments, applied in a guantity eguivalent to provide 26.4 kg ha" 1 of P, on sustaining Mehlich-I extractable soil P pool in a maize-maize rotation on a Xanthic Hapludox. M = Mucuna, C = Canavalia

PAGE 80

66 The actual P values ranged from 0.2 ng g" 1 soil (control) to about 8.0 /xg g 1 soil for other amendments measured at the end of 240 d. This finding conflicts with the results obtained by Izza and Indiati (1982) where they studied the effect of various organic farm products on soil available P. They found that incorporation of these materials in high P fixing soil produced no effect on soil available P. At 65 d TSP60 and M60 treatments had the same soil P status. There was a sharp decrease in soil P with TSP60 treatment as the cropping season progressed compared to M60 and C60. The plots which received PCW maintained low P which could not be described with the aid of any polynomial. When TSP20 was combined with M20 the P status improved compared to TSP40 (Figure 3-6) . Application of inorganic P in the proximity of the organic amendment may have reduced the exposure of inorganic P to a larger soil volume leading to reduced fixation. And the decomposition of organic material may have inactivated active soil adsorption sites in the localized band. Increased effectiveness of soil amendments such as lime, in the proximity of organic matter has been reported by Ahmed and Tan (1988) . Combined application of C20 with TSP20 was as good as TSP4 0, and PCW20 applied with TSP20 was inferior not only to TSP40 but to all treatments, except control .

PAGE 81

67 Trt. a *>1 b 2 b 3 R 2 I TSP40 4.97 3. 13E-01 -3 . 66E-03 9.57E-06 0. 78 M20+TSP20 4.22 3. 48E-01 -3 . 18E-03 7. 14E-06 0. 90 PCW20+TSP20 4.41 6. 26E-02 -8 . 02E-04 2.24E-06 0. 30ns TSP40 3.77 2 . 96E-01 -2 . 69E-03 5.92E-06 0. 92 233 280 Days Figure 3-6. Effect of selected organic amendments, applied in combination with inorganic phosphorus source in a guantity eguivalent to provide 8.8 kg ha" 1 of P, on sustaining Mehlich-I extractable soil P pool in a maize-maize rotation on a Xanthic Hapludox.

PAGE 82

68 Movement of Ca in the soil profile over the 240 d period indicated that application of TSP may reduce the soilCa pool (Figure 3-7) . The level of Ca in the 0-30 cm depth was 2.30 cmol (+) kg" 1 . This value dropped to 0.50 cmol (+) kg" 1 within 240 d. Maize plants from the control treatment plot also had the lowest uptake of Ca. Application of M60 and CM60 improved the Ca status in 0-15 cm. These materials after decomposition released cations which were an integral part of their composition (Larsen et al., 1972). Deep-rooting green manure crops offer the advantage of recycling cations which have been leached to a deeper soil profile. There is considerable evidence in the literature that organic ligands can hold polyvalent cations and prevent them from leaching (Moreno, 1960) . Building a cation reserve is the matter of great importance in acid soils where soil chemical and physical conditions favor their rapid depletion through leaching. High P concentration in leaf tissue reduced Zn and Cu concentration in leaf (Figure 3-8) . This can be attributed to the chelation of these elements by fulvic acids (Saar and Weber, 1982) . However, there is not enough evidence to conclude that at low leaf P concentration there will be a higher uptake of these metals. It has been shown that soils with high level of P are less likely to exhibit Cu toxicity (Sartain and Street, 1980) .

PAGE 83

69 10 0.5 1.0 1.5 2.0 2.5 o.O 1.0 2.0 3.0 4.0 15 I 1 1 1 ' 4 ' 1 , , , , r-» , r^T 45 60 A 0 120 180 240 M60 0.0 0.5 1.0 1.5 2.0 2.5 3.0 // d) if) Days • -»-A A I 1 f \ * t -li 0 120 180 240 TSP60 Ca (M-I).cmol (+) kg' 1 Figure 3-7. Leaching of calcium from surface horizon of highly leached Xanthic Hapludox as influenced by different organic amendments.

PAGE 84

25.0 Cu b) 20.0 15.0 10.0 5.0 0.0 D) 3 40.0 cd ii 1 1 1 1 1 1 1 m 2 D 0 ° <= rg ffiffrffi i i i rm * m a an infflifciB-g n Q 1 1 inn i i 1 1 1 mXDrfl 1 1 1 1 1 I CD D 35.0 30.0 25.0 20.0 15.0 10.0 5.0 0.0. 0.0 Zn DP a g d g q o an" a On B D D Ha-, n Orfl B am o inn 0 D Dm § Dn I ODD QDg DC amSi a an D did nanD Q an _i_ j_ _L_ 0.1 0.2 0.3 % P (leaf) 0.4 0.5 0.6 Figure 3-8. Relationship between leaf tissue P and concentration of Zn and Cu in maize leaf tissue.

PAGE 85

71 Conclusion In the glasshouse experiment, among the three amendments tested PCW produced the lowest dry matter per pot when applied alone or in combination with TSP. The highest dry matter production was obtained with the application of CM. A strong relationship was observed between dry matter production and Mehlich-I extractable P (r 2 = 0.88). In the field experiments, CM applied plots produced more grain in first as well as second crops compared to the plots which had received Canavalia ensiformis , Mucuna aterrima , PCW, and TSP. This finding agreed with the prediction made from the laboratory incubation studies using silica as the matrix substratum to evaluate the decomposition rate of organic amendments. The prediction was made based on P adsorption maxima and binding energy values, and indicated that CM applied plots will produce superior yield compare to the plots which received other amendments. When CM was applied in combination with TSP higher grain yield was obtained compared to the same amount of P applied from TSP. All organic amendments improved the soil P reserve and reduced Ca leaching.

PAGE 86

CHAPTER IV RECOMMENDATION DOMAINS AND MODIFIED STABILITY ANALYSIS Introduction The problem of non-adoption of new technologies among small farmers in third world countries has been widely recognized, not as the farmers' unwillingness to accept change but rather as the inappropriateness of the technology to the real conditions that exist at the farm level in terms of the social, economic, and natural resource endowment (Shanner, et al., 1982). This chapter explores 1) techniques to understand farmers' circumstances, 2) the concept of recommendation domains (RDs) , 3) the change in soil fertility and its impact on crop production, and 4) the use of modified stability analysis as a tool for the selection of appropriate technology for farmers with different production goals. Research Data Base An applied researchable problem should be based on farmers' immediate concerns. Such concerns are often documented through formal surveys with a structured questionnaire. Although this technique produces statistics 72

PAGE 87

73 on farm input and output, it is expensive and time consuming. In order to make surveying cost-effective, Vander Veen and Mathema (1978) used key informants to gather baseline information. This speeded-up the process but it lacked interactive information. Working for the farmers' cause reguires direct involvement in problem identification. Rapid rural appraisals, often referred to as exploratory or diagnostic surveys, are a simple and relatively guick method of identifying constraints that operate in a defined area (Abalu, et al., 1987). The sondeo (Hildebrand, 1981) is one of the technigues of rapid rural appraisal which combines different disciplines in a team. Its primary purpose is to acguaint technicians with the area in which they are going to work. No guestionnaires are used; rather informal interviews with farmers are conducted. Recommendation Domains (RD) In order to be cost effective, the research activities have to address the problems of and provide solutions for relatively large numbers of people. It is necessary, therefore, to classify farmers with similar circumstances into recommendation domains: groups of farmers for whom it is possible to make more or less the same recommendation (Perrin et al., 1978; Byerlee et al., 1980). More recently the concept of RDs has been extended to Research Domains and

PAGE 88

74 Diffusion Domains (DDs) (Hildebrand, 1986) . Many researchers, however, have considered RDs as a synonym for cropping systems (Zandstra, et al., 1981), farming types (Njobvu, 1986) , and homogeneous groups (Moussie and Muhitira (1988). The limit set by the definition of Byerlee et al. (1980) has also been broadened to include agroclimatic zones and individual fields, in addition to farms. Criteria for Delineating Recommendation Domains There is a great debate on the criteria used for delineation of RDs. Socioeconomic criteria may be just as important as agroclimatic, and agroecological variables (Njobvu, 1986) in delineating domains. If so, the resulting domains are often not amenable to geographical mapping because farmers of different domains may be interspersed in a given area. Moussie and Muhitira (1988) attempted to classify farmers into relatively homogenous groups using cluster and discriminant statistical analysis. This method employed the use of qualitative information obtained from sondeos, and the selection of key variables from in-depth, formal surveys to obtain a discriminant function which helps to identify the most important variables in classifying farmers into RDs. Swinton and Samba (198 6) used four agronomic criteria: average annual rainfall, soil fertility, soil texture, and depth to the subterranean water table for defining agricultural technology recommendation domains.

PAGE 89

75 Tshabalala and Holland (1986) indicated that the "average" farmer is a myth and the programs designed to help him or her will fail, but used the RD concept in matching the improved technology to the group likely to be interested in taking advantage of it. Economic criteria are also important in the delineation of RDs (Hildebrand and Poey, 1985) . Improved technology often reguires more cash and labor investments. Both resources are scarce on small subsistence family farms. Under limited cash availability, and so many competing uses for it, a farmer will consider an option which will give him or her the highest return per dollar invested. In this context, economic evaluation criteria such as net return ha" 1 , net return/total cost, and net return/cash cost play important roles. The RD concept is being used in other disciplines (Fattori, 1990) and the use is being considered as an extension tool guiding the effective dissemination of technology appropriate to small farm conditions. Stability Analysis The stable performance of crop cultivars over a wide range of environmental conditions is generally regarded as desirable, but there is disagreement both on its definition and on the most appropriate methods for its statistical measurement from yield data trials (Becker, 1981; Hill and

PAGE 90

76 Baylor, 1983). The most widely used stability analysis has been linear regression of cultivar yield on an environmental index derived from the mean of all or a subset of cultivars at each location or environment (Eberhart and Russell, 1966) . Stability analysis has been used extensively to select genotypes that interact less with the environment and, therefore, are considered stable. Mackenzie, et al. suggested the use of stability analysis in predicting the response of a tested variety under different growing conditions for comparing the worthiness of selected potato clones for further replicated testing before naming and release. A regression coefficient of unity has been considered favorable for selection of stable genotypes. Hildebrand (1990) discussed the drawbacks of the use of the regression coefficient in selecting adaptable, or stable genotypes . Hill and Baylor (1983) pointed out that perennial crop yields are usually measured on the same plot over a number of years, so problems with stability analysis may be encountered due to a differential change in the yields of the entries as the stand ages. They suggested an alternative orthogonal contrast analysis that partitions the variation over environmental components for each entry into sources due to environmental components (year, site, and management) and all possible interactions between these factors.

PAGE 91

77 The reason Eberhart and Russell (1966) used the site mean as an environmental index was because a lack of knowledge of the relationship of macro-environmental differences such as temperature gradients, rainfall distribution, and soil types did not permit the computation of an index which could transform the environment into a continuous variable. However, attempts continue to guantify the production environment. Advancements in using the multivariate approach to group soils in the field based on variations in systematic and random components are forthcoming (Winding and Dress, 1983) . Systematic variation is caused by difference in the parent material, relief and biological action as well as soil management practices such as fertilizer application and tillage. Random variation which is called "noise" by Burrough (1983) , represents the statistical heterogeneity of the soil. The need for assessing the factors causing soil microvariability in the tropics has been stressed by Moorman and Keng (1978) . In general, some soil characteristics are mutually correlated with each other (Norris, 1970) . Hence, factors causing soil variation which are reflected in one or more of the soil characteristics may be used as criteria for grouping soil. To analyze the causes of soil variation Kosaki and Anthony (1989) applied principal component analysis (PCA) , which is a mathematical technigue used to summarize data and

PAGE 92

investigate the relationship among variables. Variables employed for PCA in their study included soil pH, organic carbon, available P, exchangeable Ca, exchangeable Mg, exchangeable K, sand, silt, and clay. For the computation of principal components, they used a correlation matrix. However, soil variation is only one factor among many which influences the performance of a technology in a given environment. Much remains to be understood concerning the complicated interactions among agronomic, economic, social, and cultural variables which have a bearing on the performance of a technology. Modified Stability Analysis and Farming Systems Farming Systems Research/Extension, is considered to be a dynamic, interactive, and problem oriented approach to develop technology for farmers, particularly those with limited resources. The technological base of FSR/E is onstation research but it constitutes only a part of the overall FSR/E program. The main activities are concentrated in farmers* fields with direct farmer involvement in technology evaluation and feedback. The farmers' field trials involve a few treatments, and often without replications. Farmer to farmer variation in management practices for a given experiment is not controlled. Instead, any unusual practice is recorded. Under these circumstances, modified stability analysis (MSA) (Hildebrand, 1984;

PAGE 93

Upraity, et al., 1985, Fattori, 1990) has been used to select environment-specific production techniques. This technique does not depend on the concept that a regression coefficient of unity is always favorable for the selection of a technology. Adherence to this concept leads to rejection of superior technology for a specific environment in search of a 'stable' technology. Objectives The overall goal of this research was to identify appropriate technology for maize and cowpea production by small farmers. The specific objectives were to: (i) measure the performance of selected treatments in different land types, (ii) study the changes in soil fertility parameters as influenced by the application of organic amendments, and (iii) develop location-specific recommendations. Materials and Methods Site Description Two small farming communities (mean cultivated size = 3 ha/farm) in the municipality of Rio Preto da Eva, located in the state of Amazonas, Brazil were selected for on-farm experimentation. The area is accessible only by small motorboats. In trying to improve living conditions of these marginal farmers, the government of Brazil was just beginning a small watershed management program. The project

PAGE 94

80 was intended to bring awareness among farmers to preserve the environment and assist them in improving their agricultural production. The Brazilian national agricultural research institution (EMBRAPA) has a mandate of developing appropriate technology for different farming conditions in this relatively inaccessible area. Developing a Research Base Secondary information regarding indigenous farming practices of the area was collected from published sources. A rapid appraisal of the area was conducted with a multidisciplinary team of scientists participating from various research disciplines and state planning and agriculture extension organizations who visited the area on three different occasions to collect and verify information obtained in group discussions or during individual communication with farmers. Farmers* knowledge of indigenous technology, agronomic practices, and land types being used were recorded. An extensive soil sampling program was carried out to understand soil physical and chemical characteristics and relate them to farmers 1 rationale for assigning a particular cropping pattern to a given land type. Farmers played an active role in technology design, execution and evaluation.

PAGE 95

81 Selection of Treatments Three treatments from previous on-station research were selected for comparison with farmers practices' (FP) for growing maize (Zea mays L. ) and cowpea (Vigna unguiculata ) (Table 4-1) . Several locations were identified to encompass the different land types within the soil family limit of Clayey, Isohyperthermic, Xanthic Hapludox. A shift in maize planting date was agreed upon so maize maturity would coincide with the beginning of the dry season (Fig. 4-1) . For the maize crop, all treatments except FP received 100 kg ha" 1 of N from urea half applied broadcast before planting and the other half in two additional split applications. All K (60 kg ha" 1 ) was applied basal broadcast except to the FP plots. Processed city waste, CM, and TSP were applied in 25 cm bands. The maize variety BR-5110 was planted in rows 80 cm apart. No nitrogen was applied to the cowpea crop. The cowpea variety IPEAN V-69 was planted in rows 60 cm apart. Plot size for maize and cowpea varied from 100-200 square meters. The organic amendments were applied in a 20 cm band. Both crops were harvested for grain. For growing maize and cowpea the main differences between FP and the improved practice was that the improved practice received fertilizer. The land preparation and planting methods consisted of clearing the area by slash and burn, followed by manual land preparation and planting with sticks. This is a common FP in the area.

PAGE 96

82 The trial was planted in a RCB design with two replications wherever possible. Some locations had only one replication because of limited area allocated by farmers for the trial. Table 4-1. Application of N, P, and K in different treatments tested in on-farm experimentation for maize and cowpea crops in the municipality of Rio Preto da Eva, Amazonas, Brazil. Treatment Maize Cowpea N P K kg ha" 1 N P K FPf 0 0 0 0 2+ 0 CM20+TSP20 100 8.8+8. 8 60 0 8.8+8.8 60 PCW20+TSP20 100 8.8+8. 8 60 0 8.8+8.8 60 TSP40 100 8.8+8. 8 60 0 0+17.6 60 f No restriction was imposed on the amount of nutrient or type of cultural practice to be used. This treatment varied from farm to farm. 4= the range for P was 0-4 kg ha" 1 . The value of P reported is the mean. Soil samples were taken before planting, during maize and cowpea flowering, and after harvest. All soil samples were analyzed for pH, Ca, Mg, Al, K, and P. Particle size analysis was conducted only on samples taken before planting. Plant tissue samples were analyzed for P, Zn, Cu, and Mn. Their results are used in Chapter III. The analytical technigues for soil and plant tissue samples were essentially the same as described in Chapters II, and III.

PAGE 97

83 Effective Rainfall, mm 400 A'88 SON D J'89 F M Months Figure 4-1. Effective rainfall and suggested shift in the planting date for maize in Rio Preto da Eva, Amazonas, Brazil, (a) common practice, (b) suggested practice.

PAGE 98

84 Statistical Analysis Combined analysis of variance (CANOVA) . A combined analysis of variance was conducted for replicated trials. A preliminary analysis for homogeneity of variances from the individual ANOVAs was based on the chi-squared test. In the case of a significant chi-squared test, locations with a coefficient of variation >20% were excluded from the CANOVA (Gomez and Gomez, 1985). No attempt was made to employ a missing plot technique to recover lost data or data which violated some assumptions of the ANOVA. It is common to lose part of on-farm experiments to destruction of experimental plants, loss of harvested samples, etc. Data from two locations for maize and one for cowpea were lost due to misunderstanding with farmers about the harvest date. The trials were harvested before the agreed upon date. At three locations (2 for maize and one for cowpea) , the crop failed due to late planting. Data from one location for cowpea could not be used for ANOVA because the second harvest yield for all treatments was mixed without recording production separately. Stepwise regression . Stepwise regression is a method for either the forward, or backward selection of variables based on an F statistic of R 2 significance at the level specified in the model (SAS Institute, Inc., 1985). This technique was used to identify factors responsible for the wide range of variation in maize and cowpea production over

PAGE 99

85 locations. For this purpose, yield obtained from each treatment over locations was stepwise regressed on the values of soil pH, ECEC, Mehlich-I extractable-P, and Al saturation. These variables were measured from soil samples taken during tasseling stage for maize and flowering stage for cowpea (Hanway, 1967) . Modified stability analysis . The production environments were separated based on the average yield of all treatments at each location (Eberhart and Russell, 1966) . In this way, environment becomes a continuous, quantifiable variable. Yield for each treatment was related to environment by simple linear regression. Yjj = a + be where Y^ = yield from treatment i at the j th location, and ej = j th environmental index. The use of the regression coefficient 'b' in the linear equations with a value near one to select the "stable" treatment over all environments was avoided. Instead superior treatments were identified for groups of environments, or RDs (Hildebrand, 1984) . The distribution of confidence intervals for the treatments within RDs for both crops was calculated as follows: CI Y ± t a S/n* Where ;

PAGE 100

86 Y = the mean treatment yield, or other criterion within the RD, a = the level of confidence, t = value from a two-tailed "t" table, a S = [Sx 2 / (n-1) ]* or standard deviation of treatment yield, or of other evaluation criterion within the RD, and n = no. of locations or environments in each recommendation domain. The CI test was used to provide an assessment of the risk of low yields and unacceptable levels of other evaluation criteria within each RD. Results and Discussion The project area is inhabited by subsistence farmers who practice a slash-and-burn, rotational type of agricultural system and market surplus production and products gathered and hunted in the forest. The two major land types used for crops are: (i) area cleared for the first time by slash-and-burn from primary forest (PF) , and (ii) area cleared from secondary forest which was left in fallow 5-7 years (SF) . For this trial, a third type was used, (iii) area considered undesirable for agricultural activities (WL) . The chemical and physical properties of these soils are presented in Tables 4-2 and 4-3. All sampled soils were acidic (median pH H 2 0 = 4.5) with very low ECEC

PAGE 101

87 (1.35-4.21 cmol (+) charge kg soil), high Al saturation (60-90%) , and low Mehlich-I extractable p (0-12.0 /ig g" 1 ) • Change in Soil Fertility with Time Initial soil characteristics were grouped by land types (PF and SF) . Within each group, soil pH, Al saturation, ECEC, and Mehlich-I P were regressed with the number of years the land was in production. The objective was to see the magnitude of change in soil fertility parameters after deforestation had taken place and the land was used for agricultural purposes. The results are presented in Table 44. The initial values (intercept) of soil pH, Mehlich-I extractable P, and ECEC were higher in PF compared to SF. Al saturation was higher in SF (81.5 % compared to 61.1% for PF) . The rate of decline in pH in PF was 0.15 unit per year which was similar to decline in pH in SF. Aluminum saturation was increasing at the rate of 7% per year in PF compared to 4% in SF. The rate of decline in Mehlich-I P in both land types was in the range of 2.1-2.6 jug g" 1 soil. A relatively low ECEC in both land types (2.3-3.6 cmol (+) charge kg" 1 soil) indicated that most cations were leached following heavy rainfall. And those present were being taken up by vegetation or being washed out at the rate of 0.80 and 0.16 cmol (+) charge kg" 1 soil every year from PF and SF,

PAGE 102

88 respectively. This analysis provided evidence that restoration of soil fertility by letting secondary forest take over for 5-7 years is unlikely to happen. Soil Chemical Properties and Yield A stepwise regression was carried out to understand the relationship between maize and cowpea yield and soil chemical properties. Emphasis was placed on how different amendments influenced Mehlich-I extractable P. All treatments except FP had received an egual amount of P from organic and inorganic sources. Table 4-5 presents results of the stepwise regression on amended soil characteristics. For FP in maize, each unit increase in Al saturation decreased the production by 0.06 Mg ha" 1 . Application of PCW20+TSP20 did not improve soil P level, instead an improvement in soil pH was observed (similar results were obtained in glasshouse study) . Application of TSP40 and CM2 0+TSP20 resulted in improved soil P. However, the magnitude of yield improvement with similar increases in P levels were different for different treatments. It was higher for TSP40 treatment by 0.90 Mg ha" 1 compared to CM20+TSP40. It is widely believed that low soil pH is often associated with Al and Mn toxicity and Ca deficiency. An acidic soil reaction can reduce rhizobia growth, nodule

PAGE 103

rd o 0) o p id CO (0 u -H n x a, +> rH •H CO >1 rH u c > rd XJ CO >1 p H DO c 0 o CP CP o o CP CP o o CP B u CP (1) H Ul tO O • • • • r-» vo in m rH t-cn o 00 W H U> O in vo r-» oo h in cp. vo in CM ^ CM CM "3" ^ PI N CM M (N rH VO in VO in in C\ rH ID CP. CN in co r> co oo in oi to o n in * in ^^. id OHO CM CM Pn tJ CO CO 5 H PI CO H O t H CJ rH r-l O CO 0 CO u u O fO MH Q) >iT >i rd rH x: rd 1 b in tprH •H rH r-l 0 a 0 „ CO X 6 0)7 0 -P T3 CP r-l C X! 0) o tj xj E C 0 C + Cd -H O rH Xi rH O x: £ •h B CO > o re -P c ^ rH CO u CO CD >i u II -p >1 O •iH x> i c -P 0) CO P 0 •H CP 3 CO 0 rH C rH U d) •h rd 0 •rH CO P XS X3 4-1 rH O Dh d) X CO CD X! <4-l CD 0 X3 -P XI -P rH C a E •H O rH O O "H rd rd rd rH 0) CD •p P rH rH CP 0 X O 0 c a) 01 •H 4-1 rd rd T3
PAGE 104

-+ H cd « U -0 o -0 0 c c U 0) o .J OJ o a 0 >i Cp a. CP O O i— I cp CP o o CP I B u Cp o CO • ••••• CO CM rH rH rH CM CM <* CO CTi O H in in in in in co h r> CM > O CO "tf CM CO H H CO H CM VO 0^1 CO VO CO ^ (D H O CO • • • Q • • • r** cm co (0 vo cm "d* Sh -p h co ct> in "S* in cm <* cm ^ r» co •<* «a* rCO VT> CO CO CTi CTi in CO 'S' CTi CM co in co h r-» cn rC -P ^ co VO r> H in H CO in VO o in in rH W CM CO H CM CO CO CM CO CO CM CM in a> >i >H -P >i O -H CO H t-» in rH m CTi VO o CO Xt <*H U (0 CO CM O p« in vo o O ro in rH vo in co VD in VO rin VO E >h (0 •H (0 0 -P Tl C 0) co +J 0 CP 0 o CP o CP CO in CM H cp CM P» W 0 • c u CO CM o CM CM H o CM rH CO H M a) >i «j 0 x: H H H H H H H H H H rH rH H 4H -HO 0) rH x co T3 0 P rH ra (0 f0 VH -H CD 0) O -P -p r-i 0 X [iH fx, fc| &L| pL| |Jh |X| pL| b b b [L| M 0 0 Qj * rH in CM CO rH CTi VO -H W Xi d u Q) 04 W XI w X CM O r» o ID CM in o O o in CO CTi CO VO VO in CM i u (0 E rH U U Cu O
PAGE 105

Table 4-4. Relationship between soil characteristics with year in crop production in different land types. Land Type+ Soil Intercept Slope r 2 Characteristics Primary pH 5.25 -0. 15 (0. 12) 0. 14 Forest (PF) ECEC 3.60 -0. 80 (0. 38) 0. 33* Al Sat. (%) 61. 1 6. 5 (0. 6) 0. 18 MehlichI (P) 10.3 -2. 15 (1. 29) 0. 23 Secondary PH 4.75 -0. 17 (0. 09) 0. 38* Forest (SF) ECEC 2.33 -0. 16 (0. 15) 0. 03 Al Sat. (%) 81.5 4. 0 (1. 6) 0. 20 MehlichI (P) 8.5 -2. 60 (0. 72) 0. 72** -fNo. of observations in PF were 11, and in SF 7. Numbers in the parenthesis are standard error of slope estimates.

PAGE 106

92 Table 4-5. Relationship between soil characteristics measured after treatment application with grain yield for maize and cowpea crops in the municipality of Rio Preto da Eva, Amazonas, Brazil'. Partial Treatments r 2 a b Maize FP (Al Sat. ) 0.79** 6. 39 -0. 06 (0. 04+) PCW20+TSP20 (PH) 0.35* -3 . 03 0. 81 (0. 14) CM20+TSP20 (P) 0.42* -0. 52 0. 34 (0. 06) TSP40 (P) 0.77** 0. 42 0. 34 (0. 07) COWPEA FP (pH) 0.70** -1. 69 0. 97 (0. ID PCW2 0+TSP2 0 (P) 0.59* 0. 21 0. 14 (0. 03) TSP40 (P) 0.38* 1. 03 0. 10 (0. 02) CM20+TSP20 (All) X X X X f Soil characteristics were measured 65 d after treatment application for maize, and 45 d for cowpea. =f The numbers in the parenthesis are the standard errors of b estimates. Variables inside the parenthesis are soil characteristics . The values reported for r 2 , a, and b are for those variables which had a strong relationship with crop production. x None of the variables had a strong relationship with crop production.

PAGE 107

93 initiation, and impair nodule function (Keyser and Munns, 1979) . For cowpea, each unit increase in soil pH increased production by 0.97 Mg ha' 1 for FP. The effect of P from different amendments on cowpea yield varied. The intercept values of 0.21 and 1.03 Mg ha" 1 for PCW20+TSP20, and TSP40 along with 0.14 and 0.10 slope values indicated that same amount of P from these two amendments had a markedly different effect on cowpea yield. Yield variation was attributed to differences in N, Ca, Zn, and Cu content in amendments and high C:P and N:P ratios, which are explained in Chapter II and III. Land Types and Recommendation Domains Soil characteristics for maize and cowpea trials for different RDs are presented in Fig. 4-2 and 4-3. For maize, e<1.85 Mg ha" 1 was considered RD1 (poor environments), and e>1.85 Mg ha" 1 RD2 (good environments). For cowpea e<1.32 Mg ha" 1 was considered RD1, and e>1.32 Mg ha" 1 , RD2 . For both crops in RD1, soil characteristics such as pH, Mehlich-I extractable P, and ECEC values were lower than RD2 , and Al saturation was higher. Locations with favorable soil characteristics for crop production (RD2) were in first or second year of cultivation after clearing from PF, and first year of cultivation after clearing from SF.

PAGE 108

Soil pH 94 4.0 4.5 5.0 ECEC 5.5 6.0 cmol (+) charge/kg Soil 1.0 2.0 3.0 Al Sat. (%) 4.0 5.0 90 80 70 60 50 Mehlich-I (P) ug/g Soil 4.0 8.0 12.0 16.0 e<1 .85 Mg/ha e>1 .85 Mg/ha Figure 4-2. Range of different soil characteristics by recommendation domains for maize trials.

PAGE 109

Soil pH 95 4.0 4.5 5.0 5.5 6.0 ECEC cmol (+) charge/kg Soil 1.0 2.0 3.0 4.0 5.0 Al Sat. (%) 90 Mehlich-I (P) ug/g Soil 80 70 60 50 4.0 8.0 12.0 16.0 e<1.32 Mg/ha e>1.32 Mg/ha Figure 4-3. Range of different soil characteristics by recommendation domains for cowpea trials.

PAGE 110

96 Maize Experiment Results of combined analysis of variance Results of the ANOVA for the maize experiments are presented in Table 4-6. According to the F-test (P>0.05) treatments influenced yield at all locations. Treatment CM20+TSP20, according to DMRT (Table 4-7), was superior to all tested treatments at all locations (0.65-4.40 Mg ha" 1 ) and the FP failed or produced the lowest yield (maximum 0.25 Mg ha' 1 ) . A combined analysis of variance (CANOVA) based on the chi-sguare test at the 5% level of significance (meaning that variances from all five locations are homogeneous) , and the random effect model (locations were selected randomly), is presented in Table 4-8. Only the 5 locations with replications were used in CANOVA. The presence of treatment x location interaction hindered making a statistically valid statement for a given treatment for all locations, even though CM20+TSP20 outperformed other treatments as indicated by the DMRT test.

PAGE 111

Table 4-6. Summary of ANOVA for multilocatonal maize testing in the municipality of Rio preto da Eva, Amazonas, Brazil. Source df EMS LI L2 L3 L4 L5 Block 1 0. 010ns 0 . 080ns 0. 020ns 1. 361* 0. 151ns Trt 3 3. 270** 8 .218** 0. 190** 2.271* 6.418* Error 3 0. 008 0 .067 0.010 0. 101 0.411 Total 7 *, ** significantly different at 0.05 and 0.01 level of probability, ns = not significant. Table 4-7. Duncan Multiple Range Test (DMRT) for maize crop in the municipal of Rio preto da Eva, Amazonas, Brazil .'. Trt LI L2 L3 L4 L5 CM20+TSP20 2.85a+ 4.40a 0. 65a 2.80a 3.60a TSP40 1.3 0b 3.40b 0. 15b 1.60b 3.40b PCW20+TSP20 0.15c 1. 10c 0.01b l.lObc 0.70b FP 0.15c 0. Old 0.01b 0.25c 0.15b CV% 8.0 11.6 50.0 22.1 32.7 f Locations with single replications have been dropped from the ANOVA. =}= Means in the same column followed by the same letter are not significantly different at the 95% level of probability as determined by Duncan Multiple Range Test.

PAGE 112

98 Table 4-8. Combined Analysis of Variance for maize grain yield in the municipal of Rio Preto da Eva, Amazonas, Brazil . Source DF EMS F Value Pr > BLOCK 1 0.21 LOC 4 5.04 14.2 0. 02 BLOCK* LOC 4 0.35 TRT 3 15.81 13.9 0.02 TRT*LOC 12 1.13 9.5 0.01 BLOCK*TRT*LOC 15 0.11 f Chi-square test for homogeneity of variances was not significant (P<0.05). All 5 replicated locations were included in the combined analysis. Random model. LOC was tested against BLOCK*LOC, TRT against TRT* LOC, and TRT* LOC against BLOCK*TRT*LOC .

PAGE 113

99 Modified stability analysis Yield . An environmental index (e) was computed for all environments (Table 4-9) . Location 3 was the poorest environment (mean e for two replications = 0.20 Mg ha" 1 ). It had a very high Al saturation (95%), low ECEC (1.35 cmol ( + ) kg" 1 ), very low water pH (3.9), and only a trace of Mehlich-I extractable P (Table 4-2) . Farmers classified this location as WL, obviously with good reasons. Scattered occurrence of WL has led many scientists to be skeptical about permanent agricultural development on highly leached soils of the tropics (McNeill, 1964; Sioli, 1980). But the latest advancements with long term fertility experiments have offered potential for sustained production with proper management (Sanchez et al., 1982). The value of e for the best environment (location 7) was 3.10 Mg ha' 1 . This location was recently cleared from primary forest and was in the first year of cultivation. Soil pH was very favorable (pH H 2 0 = 5.2) with a ECEC of 4.21 cmol ( + ) charge kg" 1 soil, and Al saturation of only 58%. The MSA indicated that PCW20+TSP20 performed poorly in all environments compared to the other treatments with amendments based on maize grain production (Fig. 4-4a) . However, an exponential decrease in production under FP was observed with decline in environmental guality. The low

PAGE 114

CO H o o o o o o fX3 0 X n I s in o H ft 0 rH • • • • • • •H rH £1 O O CO •* CM CM 0 •H rh o o o o O O c O * IT) «* If) O rH rH 2 O rH • • • • • • s rH XI CM H CO CO a) vo H O O O O IT) If) O * CM O CM VO > rr\ — i U Hi rH XI CM rH CO CM CM c •H CM o o o o CM H If) CO CTl CO B rH • • • • • VO 0 IT) X) O O CO CM rH -H o • P 0 H O O O O O H O rH X cm cn o co H rH • • • • • 0 XI O O CO '* CM u CM o o o o IT) Qt CO O CM CO CO Q) rH • • • • • -* X! O H CM CO H • N 0 rH •H 0 H o o o o CM (0 rH ^ rH ^ CM ^ O rH • • • • • X) O O rH CM H o CM O O O O if) O O H If) rH rH • • • • O — • XI o o o o O CM V rH CO • (0 -H 0 O £ N 0 H o o o o If) rH ^ O O CM CO CM rH rH • • « • £ 03 XI o o o o O — CM o o o o CO ^ 01 O O VO > CO a) (0 rH • • • • • CO — c CM XI CM CM o 0 • X N 0 H o o o o CO CM Q> (0 O CM CM H rH *C E rH • • * • • C 5 X! o H n ^ CM •H CM o o o o If) r-l (0 X CM CM CO rH rH (0 > H rH • • • • • H -P w 0 XI O O H CM rH • c 0 H a) «3 J H o o o o rg T3 H H CM CTl o c rH • • • • 0 0 xt O O rH CM H U 4-> •H u CM O ft O CU CM a) 0 w w a< M rH o Eh CO • -H + En >i >1 cn »J o o + XI X! CM t O & Oj CM 0 cu u w s aT Q) Eh 04 Eh U rH

PAGE 115

101 production with the PCW20+TSP20 treatment could be due to the immobilization of the applied inorganic P in close proximity to the PCW. Low P content, high loading rate, high metal concentration, high C:N, and C:P ratios were apperently responsible for low P release, and high P fixation. The performance of CM2 0+TSP2 0 was better than other tretaments in all environments (Fig. 4-4a) . The production level of 1.85 Mg ha" 1 was identified as a minimum acceptable level. Based on the minimum acceptable production level of 1.85 Mg ha" 1 all environments were divided into two RDs. Within both RDs the probability of yield falling below 1.85 Mg ha" 1 for each treatment was examined with the aid of a confidence interval (CI) test (Fig. 4-4b) . Thus, even with CM20+TSP20 there is 17% chance of yield less than 1.85 Mg ha" 1 in SF2 and WL (RD1) and well over 50% with TSP40. Using these two treatments in PF1, 2 and SF1 (RD2) , there is virtually no probability of yield less than 1.85 Mg ha" 1 . On the other hand FP only reaches 1.85 Mg ha' 1 in PF1 which falls in RD2.

PAGE 116

102 Figure 4-4a. Response of different treatments to environmental index for maize production, Rio Preto da Eva, Amazonas, Brazil.

PAGE 117

40 50 60 70 80 90 1 100 I -1.0 O c © 40 8 50 60 70 80 90 100 0.0 CM20+TSP20 I \ f \ TSP40 \ \ \ j i 103 1.0 y 2.0 3.0 4.0 1 .85 minimum acceptable level A i — i i ' i ' : I I ;i i ;i i •i i i i A i \ • i / Mi \ / • i \ i : . i \ / : . i \ /' \ \ / / \ V /' A -2.0 -1.0 0.0 2.0 3.0 4.0 5.0 6.0 ,0 f 1 .85 minimum acceptable level Yield, Mg/ha Figure 4-4b. Distribution of confidence intervals for maize production in poor (e<1.85 mg ha" 1 ), and good (e>1.85 mg ha" 1 ) environments.

PAGE 118

104 Net income/cash cost . The net income/cash cost ratio is an important criterion for selection of improved technology which requires additional cash input. The information presented in Appendix A indicated that FP had the highest return over invested cash. One dollar invested gave a return of US $ 32.3 at a production level of 3.10 Mg ha' 1 in the best environments. The relationship between e and net return/cash cost ratio followed an exponential function (Fig. 4-5a) . The return over invested cash for other treatments was less than $ 6.3 in all environments. Processed city waste treatment ( PCW2 0+TSP2 0) had a negative return at all production levels. The distribution of confidence intervals for net income/cash cost is presented in Fig. 4-6a. The mean net return from a dollar invested in CM20+TSP20 was $ 3.2 and ranged from $ 2.4 to 4.2. It was observed that 99 % of the time a farmer will receive higher than 2.4 $/$ investment. The average return from a dollar invested in FP was $ 18. However there is 18 % chance that FP will provide a return of less than $ 2.4. Net income/total cost . The net income/total cost ratio indicated that return over total cost was low for all treatments. However, as the production environments improved (e> 1.85 Mg ha" 1 ) the net return to total cost was within the range of $ 1.8 to 3.5 (Fig. 4-5b) . Chicken manure (CM20+TSP20) provided stable return which was within the

PAGE 119

105 range of $ 1.8 to 2.1 per dollar spent and had no probability of a loss (Fig. 4-6b) . Net income . An analysis of net income ha" 1 under four tested treatments (Appendix A, and Fig. 4-5c) indicated that TSP40, and CM20+TSP20 were superior to FP, and PCW20+TSP20 in all environments. Net income from TSP40, and CM20+TSP20 at their highest production levels were $ 622 and 513, respectively compared to $ 388 and 16.5 for FP, and PCW20+TSP20. A confidence interval was calculated for FP and CM20+TSP20 treatments based on net income (Fig. 4 -6c) . Two treatments not considered for CI analysis were PCW20+TSP20, TSP40. Processed city waste ( PCW2 0+TSP2 0 ) had the lowest net income, and TSP40 had agronomic performance similar to that of CM20+TSP20. The confidence interval test indicated that 99 % of the time CM20+TSP20 will produced a net income greater than $ 400 compared to a negative net income to an income of $ 380 for FP practice (Fig. 4-6c) . The narrow range of net income with CM20+TSP20 treatment compared to FP indicated that CM20+TSP20 provided less risk than FP, and had a high net income ha" 1 .

PAGE 120

106 Figure 4-5. Relationship of net income/cash cost, net income/total cost, and net income with environmental index, in on-farm maize trials from Rio Preto da Eva, Amazonas, Brazil. Treatment Intercept b SE Y Estimate Net Income/Cash Cost FP| 0.1 1.85 0.42 PCW20+TSP2 0 -1.0 0.30 0.21 TSP40 1.5 2.71 0.56 CM20+TSP20 0.5 1.42 0.80 Net Income/Total Cost Fpf 0.2 0.83 0.50 PCW20+TSP20 -1.0 0.30 0.21 TSP40 -0.5 1.00 0.33 CM20+TSP20 0.3 0.61 0.38 Net Income ha" 1 FPf 1.0 1.8 0.8 PCW20+TSP20 -217.0 71.3 44.3 TSP40 -142.9 261.0 55.0 CM20+TSP20 60.7 177.5 101.1 f Exponential relationship SE = Standard Error

PAGE 121

Net Income/Cash Cost ($) FP PCW20+TSP20 TSP40 CM20+TSP20 _B_ .-•©••• — . *— Net Income/Total Cost ($/$) 4.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 Environmental Index (e), Mg/ha 0.0 0.5 1.0 1.5 2.0 Net Income, $/ha 800 600 400 200 -200 -400 • • O CM20+TSP20 * ~W~, , ' o TSP40 ---A .-T.A PCW20+TSP20 J I L 0 0.5 1 1.5 2 2.5 3 3.5 Environmental Index (e), Mg/ha Values are expressed in US $

PAGE 122

108 Figure 4-6. Distribution of confidence intervals for net income/cash cost, net income/total cost, and net income for different treatments used for maize cultivation.

PAGE 123

109 Cowpea Experiment Results of combined analysis of variance Cowpea response was evaluated at 13 locations. Eight were replicated and five were not. The ANOVA and DMRT for locations are presented in Table 4-10 and 4-11. According to the F-test, treatments did not influence yield (P>0.05) at three locations. At location 1, CM20+TSP20 was as good as TSP40 while CM20+TSP20 did not differ from PCW20+TSP20 (Table 4-12) . A CANOVA based on a similar computation as for maize is presented in Table 4-9. A significant chi-sguare test (chi-sguare = 11.78) suggested the elimination of all locations with cv > 20%. Therefore, this test was carried out with only three of the 8 replicated locations. The treatment x location interaction was found to be significant (P <0.03). Gomez and Gomez (1985) suggested the partitioning of treatment x location interaction using either the homogeneous site approach or the homogeneous treatment approach. However, by rejecting 10 locations out of 13, it was felt that the locations were no longer a true representation of the population.

PAGE 124

110 3 CO • o H 1 Q 0) a) 0 H M X) (0 o Eh co co IVO in co CM 10 * c * VO H H CO CM O • • • OHO w co c c o co oo o <* H • • • o o o 01 01 C C o «* in H [-» H • • • o o o n c * O CO CM O O H • • • O CM O (0 c * O VO If) h m o • • • o o o co in VO r"~ in co co h • • • o o o * * * * in h o in ^ o • • • o o o 10 c * co co a\ OHO • • • OHO H CO CO X u o o o -P u H U U 03 Eh W to p o Eh 1 1 n ii fl) H Ah rn UJ Jj (-« »-« e •H | 1 VJ uj i r— t i \ » . _j *rn M \J r\ M i-i *3 M vJ o N 1 n o ll rH n v_/ 04 01 4-1 ,—1 Pi •H r 1 ti rH ni 1) HM ni rrl r-| n , j > / — V Q_J n r v. V»J M (— iH UJ I f~i U I Eh 01 f — \ V — » q n 1 1 Hi rrt fl) w fn Id U_| c (|_( (0 (0 H a: fi l-H W cii o r-( Jj , | r 1 Q, CI) l 1 hH •H }^ r*-t jj Q. •f-r Hi /Tt ,_! o 13 0 •H 4-1 •H C c ro M-i tn 0 0 -H c DQ 0 H a * 10 I -* 0) H SX f0 Eh -H O •H c co rJ rJ v£> in CO rJ CM (0 rC a) -H >i C -H c0 r-l o rH P u Eh (C (C3 XI rQ in o o o vo co CM H • • • • H H O O cd (0 (0 co in in o in vo h > VO • • • • CM CM H H (0 (0 (0 (0 in in o in co co in h • • • • H H O O ccj cO Xt rQ in o o in o h in H • • * • CM CM O O fO in CM CM rQ O (0 rQ O O VD CM H H (0 o C • c0 >i« CM H +J i P M XI (0 (0 0 T3 in •HMO) • HOC VO a-p -h H a) +j e m a) u CM CM > h a) a> 4-) H 0) 0) C (0 •H CO CO CO a) o xj +J Hp CO P a) T3 0) >iH XI -rH 0) > ft O ftH O H r-l O rQ CO X) o u ft 4-1 o c c a) g 0) 3 rQ H o a) o > (0 0) rC e (0 CO CO c o a) •H £! -P P id o c o -H 03 C CO a) &H S CD > m 0*
PAGE 125

Ill Table 4-12. Combined Analysis of Variance 1 " for cowpea experiments. Source DF EMS F Value Pr > F BLOCK 1 0.07 LOC 2 0.80 1.7 0.20 BLOCK* LOC 2 0.47 TRT 3 2.40 26.6 0.00 TRT*LOC 6 0.09 4.7 0.03 BLOCK*TRT*LOC 9 0.02 -f Chi-square test for homogeneity of variances was significant. All sites with coefficient of variation > 20% were excluded from the combined analysis. Random model. LOC was tested against BLOCK*LOC, TRT against TRT* LOC, and TRT* LOC against BL0CK*TRT*L0C.

PAGE 126

112 Modified stability analysis Yield . For yield, it was observed that the CM20+TSP20 treatment was stable (slope = 0.03, compared to 1.13 for TSP40) over all environments (e values 0.5-2.0 mg ha 1 ). Treatment CM2 0+TSP20 had a clear advantage over TSP40 in e < 1.32 Mg ha" 1 . But in e > 1.32 Mg ha" 1 TSP40 outyielded CM20+TSP20 (Fig. 4-7a) . The reason for such 'cross-over' interaction was believed to be the additional input of N from CM. Excess N in the better environments caused cowpea plants to grow taller and become susceptible to lodging. In poor environments this phenomenon had no significant bearing on plant growth due to inherent N deficiency in the soil. The point of 'cross-over' interaction (1.32 Mg ha 1 ) was taken as a reference point for the delineation of RDs. All locations having e values < 1.32 (PF3; SF1, 2, 3; WL) were grouped into RD1 (poor environment) , and locations having e values > 1.32 (PF 1, 2) were clustered into RD2 (good environment). The confidence limit test (Fig. 4.7b) indicated that CM20+TSP20 was a stable performer in both RDs. The confidence limits for the CM20+TSP20 treatment fall within 1.4 to 2.1 Mg ha" 1 in RD1, and 1.8 to 1.9 Mg ha" 1 in RD2 compared to 1.0 to 1.8 Mg ha" 1 in RD1 and 2.1 to 2.8 Mg ha" 1 in RD2 for TSP40. The 'cross-over' interaction along with confidence limit test indicated that CM20+TSP20 treatment should be recommended in RD1 and TSP4 0 in RD2 .

PAGE 127

113 Net income/cash cost . The results of MSA for net income/cash cost showed that FP had the highest return over invested cash in environments with e value >1.32 Mg ha (Fig. 4-8a) . A return of as high as 37.7 $/$ cash investment was obtained with FP compared to less than $ 10 for other treatments in all environments. The confidence interval calculation is presented in Fig. 4-9a-b. This test suggested that in poor environments CM20+TSP2 0 provided a mean return of $ 6.0 on each $ invested. This mean fell within the range of $ 4.0 to 8.0. Farmer practice had a wider CI. Therefore, the return over invested cash in CM20+TSP20 in poor environments was better and less risky compared to other treatments. However, in good environments FP was superior to other treatments. Net income/total cost . The point where FP, TSP40, and CM20+TSP20 crossed was at an e value of 1.32 Mg ha . Below this value the net return/total cost ratio was equal for TSP40, and CM20+TSP20 treatments (within $ 2.8-3.0) (Fig. 48b) and they were higher than FP. Processed city waste ( PCW2 0+TSP2 0 ) gave the lowest net return/total cost ratio in all environments which suggests that a farmer will lose money by applying PCW20+TSP2 0 for cowpea production in these environments. In good environments performance of all treatments improved. However, FP gave the highest return over total cost.

PAGE 128

114 The CI calculation (Fig. 4-10a-b) indicated that CM20+TSP2 0 gave higher, and stable return in poor environments compared to other treatments. At 90% CI, FP produced from $ 0 to 2.2 compared to $ 2.8 to 3.8 for CM20+TSP20. But FP was the best performer in good environments and provided a return of $ 4.2 to 5.8 for each $ invested in total cost. Net income . The value of net income ha" 1 from different environments was regressed against e (Fig. 4-8c) . The net income was higher for TSP40 treatment in good environments. But CM20+TSP20 was superior to TSP4 0 in poor environments. As the environment for cowpea cultivation improved the net income from FP was also increasing. Processed city waste treatment ( PCW2 0+TSP2 0 ) was always inferior to other treatments. Confidence intervals were calculated for FP, TSP40, and CM20+TSP20 (Fig. 4-lla-b) . Based on this criterion FP was inferior to TSP40 in good environments, and to CM20+TSP20 in poor environments. At 90% CI the value of net return for TSP40 fell in the range of $ 820-1080 ha" 1 compared to $ 300 to 750 for FP.

PAGE 129

115 Figure 4-7a. Response of different treatments to environmental index for cowpea production, Rio Preto da Eva, Amazonas, Brazil.

PAGE 130

116 0.0 0.5 1.0 1.5 2.0 2.5 3.0 Yield, Mg/ha Figure 4-7b. Distribution of confidence intervals for cowpea production in poor (e<1.32 mg ha" 1 ), and good (e>1.32 Mg ha" 1 ) environments.

PAGE 131

117 Figure 4-8. Relationship of net income, net income/cash cost and net income/total cost with environmental index, in on-farm cowpea trials from Rio Preto da Eva, Amazonas, Brazil . Treatment Intercept b SE Y Estimate Net Income/Cash Cost FP -20.8 24.8 8.36 PCW20+TSP2 0 -3.0 3.2 2.87 TSP40 5.7 0.6 0.12 CM20+TSP20 11.3 -2.3 2.14 Net Income/Total Cost FP -2.5 4.3 5.9 PCW20+TSP20 -2.0 2.1 1.8 TSP40 1.8 1.2 1.2 CM20+TSP20 3.0 0.2 0.1 Net Income ha" 1 FP -521.0 620.4 226.6 PCW20+TSP20 -558.0 597.3 238.4 TSP40 20.7 481.1 114.5 CM20+TSP20 604.1 60.7 23.8 SE = Standard Error

PAGE 132

118 Net Income/Cash Cost ($/$) Net Income/Total Cost ($/$) 0 6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 Environmental Index (e), Mg/ha Net Income, $/ha 1.200.0 TSP40 1.000.0 c 800.0 * % 6000 ** CM20+1M20 jC »' A o 400.0 O FP ^\ J 2000 Si PCW20+TSP20 AS* * 00 n A •200 0 A -400 0 i i Cowpea i i i 1 . 0.6 0.8 1 0 1.2 1.4 1.6 1 8 2.0 2.2 Environmental Index (e), Mg/ha

PAGE 133

40.0 50.0 * 60.0 I 'o 70.0 o 8 o | o O 8 c T3 c o O 100.0 0.0 5.0 10.0 15.0 20.0 25.0 Net Income/Cash Cost (US $) 30.0 35.0 40.0 Figure 4-9. Distribution of confidence intervals for net income/cash cost for selected treatments in poor (e<1.32 Mg ha" 1 ) and good (e>1.32 Mg ha" 1 ) environments for cowpea cultivation.

PAGE 134

120 Figure 4-10. Distribution of confidence intervals for net income/total cost for selected treatments in poor (e<1.32 Mg ha" 1 ) and good (e>1.32 Mg ha" 1 ) environments for cowpea cultivation.

PAGE 135

121 I i 0 o o c © 1 o O 40.0 50.0 £ 60.0 70.0 80.0 90.0 100.0 1 1 1 1 1 1 r e<1.32 Mg/ha I i U i r Mi s TSP40 •; CM20+TSP20 A FP { )i % 1 1 1 1 \ V \ \ K -200.0 o.O 200.0 400.0 600. 0 Net Income (US$/ha) 800.0 1 ,000.0 40.0 50.0 60.0 c o 1 O o o c 5 T3 C O 90.0 100 70.0 80.0 200.0 400.0 600.0 800.0 Net Income (US$/ha) 1 ,000.0 1,200.0 Distribution of confidence intervals for net Figure 4-11. income for selected treatments in poor (e<1.32 Mg ha" ) and good (e>1.32 Mg ha" 1 ) environments for cowpea cultivation.

PAGE 136

122 Technology Selection for Different Reco mmendation Domains The concept of RDs is very beneficial in selecting the nitch for a technology in the overall research domain. However, the usefulness of the concept in field conditions will depend on selecting technology evaluation criteria based on the farmers' needs and aspirations, and using the criteria to divide the research area into RDs. More often there are more than one criteria. A summary of technology selection for a given land type based on different evaluation criteria is presented in Table 4-13 and 14. Maize The selection of appropriate technology is the function of land type and the farmer's goal. If the goal is to obtain highest grain production in PF1, CM20+TSP2 0 should be recommended. In the same land type, if the goal is to obtain highest return to limited cash FP, should be suggested for use. However, one should be aware of the fact that FP is not sustainable in long run and within a year or two the productivity will decline to a level which will no longer offer highest return to invested cash. In SF2, CM20+TSP2 0 out performed all treatments based on all evaluation criteria under consideration. In WL, CM2 0+TSP20 can be recommended based on the yield criteria, but economically none of the tested treatments was sound.

PAGE 137

123 Cowpea For cowpea, all land types were divided into two recommendation domains. In RD2 (PF1 and PF2) , TSP40 should be recommended based on yield and net income ha' 1 . Depending on the evaluation criteria of return to cash cost and return to total cost FP has advantage over other treatments. This practice should not be continued for more than a year if high return to invested cash is expected. In RD1 (PF3, SF1, SF2, SF3 , and WL) , CM20+TSP20 is the best option based on all evaluation criteria. Redefining Recommendation Domains A map of the area including location, land type, year in crop production, and RDs based on the yield criterion is presented in Fig 4-12. The resulting domains are not amenable to geographical mapping because farmers of different domains were interspersed in the area. The use of different evaluation criteria can also lead to regrouping of environments from one RD to another. For example, the use of return to cash cost indicator moved one production environment for maize and two for cowpea from RD1 to RD2 (Fig. 4-13) .

PAGE 138

124 Table 4-13. Technology selection for a given land type based on different evaluation criteria for maize cultivation. Land Type Evaluation Criteria Yield Mg/ha Return to Cash Cost $/s Return to Total Cost $/$ Net Income $ PFl SFl CM CM FP FP FP FP CM CM PF2 SF2 CM CM CM CM CM CM CM CM WL CM NR NR NR Table 4-14. Technology selection for a given land type based on different evaluation criteria for cowpea cultivation. Land Type Evaluation Criteria Yield Return to Return to Net Mg/ha Cash Cost Total Cost Income $/$ $/$ $ PFl TSP FP FP TSP PF2 TSP FP FP TSP PF3 CM CM CM CM SFl CM CM CM CM SF2 CM CM CM CM SF3 CM CM CM CM WL CM CM CM CM

PAGE 139

12 C O W P E A 1.9/3 0.9/9 2.0/10 1.2/5 1-4/4 ^ccm H PF(3) 2.0/7 SF(3)M pF(2) < PF(1) SF(2)H 0.8/8 PF(2) SF(1) 1 .8/1 1 M SF(2) 1.5/1 SF(3) 1.0/2 0.8/13 Recommendation Domain I Recommendation Domain II 1.1/1 SF(2) I 2.8/6 PF(1) 3.1/7 1-4/4 ^ PF (1) R I O 0.2/3 2.2/2 1.9/5 WL PF(1) Maize 2.1/8 Recommendation Domain I Recommendation Domain II Figure 4-12. Recommendation domains, based on yield, and their relationship to land types in the municipality Rio Preto da Eva, Amazonas, Brazil.

PAGE 140

126 1.9/3 0.9/9 ^ SF(3)"* PF(2) RIO C O W P E A 1.4/4 2.0/10 1.2/5 PFfl) 6& PF(3) SF(2)M 0.8/8 1.5/1 SF(3) 0.8/13 1.0/2 Recommendation Domain I Recommendation Domain II 1.1/1 SF(2) I 2.8/6 PF(1) 2.2/2 PF(1) 2.1/8 3.1/7 1-4/4 XX PF(1) R I O 1.9/5 Maize Recommendation Domain I Recommendation Domain II Figure 4-13. Recommendation domains, based on net return to cash cost, and their relationship to land types in the municipality of Rio Preto da Eva, Amazonas, Brazil.

PAGE 141

127 Conclusion The presence of environment-by-treatment interaction masked the mean treatment differences measured by CANOVA. To detect treatment effects, this analysis suggested grouping of similar environments and a repeat of the experiment. The slope estimates in regression analysis were useful in predicting rates of decline in pH, ECEC, and Mehlich-I extaractable P with the year of cultivation. A reduction of 2.1 and 2.6 /xg g* 1 soil in Mehlich-I extractable P was obtained annually in PF, and SF. Although an equivalent of 8.8 kg ha" 1 P was applied from different amendments, their effect on maize and cowpea production varied. Phosphorus from TSP was used more efficiently, although this treatment reduced ECEC, and soil pH. Modified stability analysis was conducted for maize and cowpea using yield, net income/cash cost, net income/total cost, and net income ha" 1 as technology evaluation criteria. Different evaluation criteria led to different conclusions regarding recommendations of appropriate technology for a given environment. These criteria also influenced selection of production environments for grouping into RDs. Based on the evaluation criterion of maximizing maize production CM20+TSP20 was recommended in RD2 (good environment) But there was a clear advantage of FP in RD2 in terms of net return to cash cost. If a farmer's goal is to maximize income from his scarce cash, he ought to use farmers*

PAGE 142

128 practice. However, FP can not be continued for more than a year due to adverse effect on soil properties. The aluminum saturation was increasing each year in PF by 7%, and Mehlich-I P was decreasing by 2.1 \iq g' 1 soil annually. For cowpea, CM20+TSP20 outperformed all treatments in poor environments (e<1.32 Mg ha" ) while in good environments (RD2) (e>1.32 Mg ha" 1 ) TSP40 had clear advantage over other treatments based on all evaluation criteria. Farmer practice had the highest return per dollar compared to other treatments in RD2 (37.7 $/$ ). The delineation of RDs was not amenable to geographical mapping because fields of different domains were interspersed in the area. A change in evaluation criteria lead to change in the demarcation of RDs.

PAGE 143

CHAPTER V SUMMARY AND CONCLUSIONS The objective of this chapter is to summarize the work presented in the preceding four chapters. The overall goal of this research was to examine the role of organic amendments in sustaining P nutrition of highly leached Oxisol and to analyse the performance of selected treatments in farmers' fields. To achieve this goal, laboratory incubation studies, glasshouse studies, and on-station and off-station field studies were conducted. Chapter II presents the findings on P adsorption and desorption by soil as influenced by organic amendments. This chapter also investigates a new technigue for measuring nutrient release from a decomposing material. Incubation of soil with amendments for 35 d had a marked influence on total P adsorption by soil. Soil amended with PCW adsorbed more P than unamended soil. However, with an increase in incubation time, all amendments reduced P fixation. Phosphorus desorption values for soil incubated for 35 d were equal to that of the control, but soil incubated for more than 35 d showed increased P desorption. These findings suggest that the rate of decomposition and P 129

PAGE 144

130 content of the amendments played a key role in influencing adsorption and desorption characteristics of the soil. The results of the decomposition study in a soil matrix indicated that maize and grass treatments showed P immobilization. Processed city waste immobilized soil P at 35 and 65 d, but at 150 d a net release was observed. All legumes followed a similar P release patterns. However, When silica was used as a matrix there was a net release of P from all amendments independent of P content and C:P ratio. The data on P adsorption maxima calculated by the Langmuir equation for CM, based on P release from amendments in soil and silica matrices, indicated that with increasing incubation time, P adsorption maxima increased. Continued adsorption beyond adsorption maxima demonstrated the presence of a precipitation reaction, or multilayer adsorption. For tephrosia and kudzu adsorption maxima decreased with time. A good agreement between adsorption maxima and actual adsorption of P based on SSMT was observed at the higher equilibrating solution P concentration (60 /xg g" 1 soil) which indicated the validity of the SSMT technique in predicting P requirements of amended soils. Chapter III examines the results of glasshouse and onstation field studies on the effect of a suboptimal dose of organic amendments in sustaining crop P nutrition and improving the efficiency of inorganic P.

PAGE 145

131 An analysis of variance for maize herbage dry weight from the glasshouse study, indicated the presence of an interaction among type and rate of amendments. Among three amendments tested, PCW produced the lowest yield. It is interesting to note that when PCW was applied at a higher rate (eguivalent of 17.6 kg P ha" 1 ) the result was inferior compared to the lower rate (equivalent of 8 . 8 kg P ha 1 ) . The highest dry matter yield per pot was obtained with CM. A single degree of orthogonal contrast for maize grain production during the first crop showed that application of PCW equivalent to 2 6.4 kg ha" 1 of P produced 1.41 Mg ha" 1 of maize. This level of production was inferior to the control treatment. No significant differences were found among canavalia, mucuna, and TSP when applied to provide P at the rate of 26.4 kg ha" 1 . The CM treatment outyielded all treatments in first as well as second crops. Combined application of 8.8 kg P ha" 1 from organic amendments and the same amount from TSP was compared with 17.6 kg P ha" 1 from TSP. For the first crop the combination with CM was superior compared to TSP and an improvement of 30% in yield was recorded. The change in P status of soil over 240 d of cropping cycle for all but PCW treatment followed a cubic surface response curve. All treatments improved P status of soil compared to the control. As the cropping season progressed, there was a sharp decrease in soil P with TSP60 compared to

PAGE 146

M60 and C60 (Cajnavalia ensiformes) . Processed city waste (8.8 kg ha " 1 ) applied with TSP20 was inferior to all treatments except the control. The results obtained from glasshouse and on-station field studies suggested that organic amendments can provide sustained crop P nutrition compared to inorganic sources. However, the improved efficiency of inorganic P when applied with an organic amendment will depend on chemical characteristics of the organic amendments. In Chapter IV the results of farmers' field trials are presented. Different criteria were used for the selection of appropriate technology for farmers and delineation of RDs. Based on the evaluation criterion of maximizing production CM20+TSP20 was recommended in good environments (e>1.85 Mg ha" 1 ) for maize. For cowpea, CM20+TSP20 outperformed all treatments in poor environments (e<1.32 Mg ha" 1 ) while in good environments (e>1.32 Mg ha" 1 ) TSP40 had clear advantage. Farmer practice had the highest return over cash cost compared to other treatments for both crops. The result of CANOVA for maize and cowpea grain production provided inconclusive results because of the presence of environment x treatment interaction. The slope estimates of the regression analysis provided valuable information on changes taking place in soil fertility following deforestation. The rate of decline in pH in PF and SF was 0.15 unit per year. Aluminum saturation was

PAGE 147

133 increasing at the rate of 7% per year in PF compared to 4% in SF. The rate of decline in Mehlich-I P in both land types was in the range of 2.1-2.6/xg" 1 soil. A relatively low ECEC in both land types (2.3-3.6 cmol ( + ) charge kg" 1 soil) indicated that most cations were leached following heavy rainfall. And those present were being taken up by vegetation or being washed out at the rate of 0.80 and 0.16 cmol ( + ) charge kg" 1 soil every year from PF and SF, respectively. The data presented in Chapter IV suggested that the delineation of RDs was dependent on the technology evaluation criteria, and was not amenable to geographical mapping because fields of different domains were interspersed. A change in evaluation criteria led to a change in the demarcation of RDs. It was also observed that a complete restoration of soil fertility by letting secondary forest take over for 5-7 years was unlikely to happen in the rain forest.

PAGE 148

134 APPENDIX A ECONOMIC ANALYSIS FOR COWPEA Yield, income, and cost from the cowpea trials, Rio Preto da Eva, Amazonas, Brazil. (US $) Yield Gross Total Net Net Income/ Mg ha" 1 Income Cost Income Total Cost FP 0.70 350 67 283 4.2 0.50 250 55 195 3.5 1.45 725 112 613 5.5 0 . 60 300 61 239 3.9 0. 15 75 34 41 1.2 0. 15 75 34 41 1.2 1.70 850 127 723 5.7 0. 10 50 31 19 0.6 0.20 100 37 63 1.7 2.20 1100 157 943 6.0 1.20 600 97 503 5.2 1.50 750 115 635 5.5 0.00 0 25 -25 -1.0 PCW2 0+TSP20 0.90 450 239 211 0.9 0.65 325 224 101 0.5 1.95 975 302 673 2.2 1.20 600 257 343 1.3 0.50 250 215 35 0.2 0.50 250 215 35 0.2 1.65 825 284 541 1.9 0.20 100 197 -97 -0.5 0.40 200 209 -9 -0.0 1.90 950 299 651 2.2 1.50 750 275 475 1.7 1.80 900 293 607 2.1 0.00 0 185 -185 -1.0

PAGE 149

135 Appendix A contd. . . TSP40 O 1 ft 1 150 230 920 4 . 0 i in 550 158 392 2.5 1250 242 1008 4.2 800 188 612 3.3 2 10 1050 218 832 3.8 1.35 675 173 502 2.9 2.65 1325 251 1074 4.3 1.30 650 170 480 2.8 1.20 600 164 436 2.7 4.2 2.60 1300 248 1052 2.20 1100 224 876 3.9 2.10 1050 218 832 3.8 1.30 650 170 480 2 . 8 CM2 0+TSP20 1.80 900 211 689 3.3 1.50 750 193 557 2.9 1.90 950 217 733 3.4 2.25 1125 238 887 3.7 2.05 1025 226 799 3.5 1.35 675 184 491 2.7 2.15 1075 232 843 3.6 1.65 825 202 623 3.1 1.70 850 205 645 3.1 1.40 700 187 513 2.7 1.90 950 217 733 3.4 1.70 850 205 645 3.1 2.00 1000 223 777 3.5

PAGE 150

APPENDIX B ECONOMIC ANALYSIS FOR MAIZE Yield, income, and cost from the maize trials, Rio Preto da Eva, Manaus, Brazil. (US$) Yield Gross Total Net Net Income/ Mg ha" 1 Income Cost Income Total Cost FP 0.15 30 18 12 0.7 0.00 0 12 -12 -1.0 0.00 0 12 -12 -1.0 0.25 50 22 28 1.3 0. 15 30 18 12 0.7 2.20 440 100 340 3.4 2.50 500 112 388 3.5 0.20 40 20 20 1.0 PCW2 0+TSP2 0 0. 15 30 213 -183 -0.9 1. 10 220 251 -31 -0.1 0.00 0 207 -207 -1.0 1. 10 220 251 -31 -0.1 0.70 140 235 -95 -0.4 1.00 200 247 -47 -0.2 1.40 280 263 16 0.1 0.70 140 235 -95 -0.4 TSP4 0 1.30 260 150 110 0.7 3.40 680 234 446 1.9 0. 15 30 104 -74 -0.7 1.60 320 162 158 1.0 3.40 680 234 446 1.9 4.20 840 266 574 2.2 4.50 900 278 622 2 . 2 3.50 700 238 462 1.9

PAGE 151

137 Appendix B contd CM2 0+TSP2 0 2.85 570 241 329 1.4 4.40 880 303 577 1.9 0.65 130 153 -23 -0.2 2.80 560 239 321 1.3 3.60 720 271 449 1.7 3.60 720 271 449 1.7 4.00 800 287 513 1.8 4.00 800 287 513 1.8

PAGE 152

REFERENCE LIST Abalu, G.O.I. , N.M. Fisher, and Y. Abdullahi. 1987. Rapid rural appraisal for generating appropriate technologies for peasant farmers: Some experiences from Northern Nigeria. Agri. Systems 25:311-324. Adams, F. 1980. Interactions of phosphorus with other elements in soils and in plants, p. 655-680. In M. Stelly and R. C. Dinauer (ed.) The role of phosphorus in agriculture. ASA, CSSA, and SSSA, Madison, WI. Anghinoni, I., and S.A. Barber. 1980. Phosphorus application rate and distribution in the soil and phosphorus uptake by corn. Soil Sci. Soc. Am. J. 44:1041-1044. Appelt, H., N.T. Coleman, and P.F. Pratt. 1975. Interaction between organic compounds, minerals and ions in volcanic-ash derived soils. II. Effects of organic compounds on the adsorption of phosphate. Soil Sci. Soc. Am. Proc. 39:628-630. Bache, B.W. 1964. Aluminum and iron phosphate studies related to soils. II. Reaction between phosphates and hydrous oxides. J. Soil Sci. 15:110-116. Barrow, N.J. 1974. The effect of previous additions of phosphate on phosphate adsorption by soils. Soil Sci. 118:82-89. Barrow, N.J. 1983. A mechanism model for describing the sorption and desorption of phosphate by soil. J. Soil Sci. 32:555-570. Becker, H.C. 1981. Correlations among some statistical measures of phenotypic stability. Euphytica 30:835-840. Bell, L.C., and C.A. Black. 1970. Comparison of methods for identifying crystalline products produced by interaction of orthophosphate fertilizers with soils. Soil Sci. Soc. Am. Proc. 34:579-582. 138

PAGE 153

139 Blair, G. J., and O. W. Boland. 1978. The release of phosphorus from plant material added to soil. Aust. J. Soil Res. 16:101-111. Bowman, R.A., and S.R. Olsen. 1985. Assessment of phosphate buffering capacity: 2. Greenhouse methods. Soil Sci. 140: 387-392 . Brown, P. L. , and D. D. Dickey. 1970. Losses of wheat straw residue under simulated conditions. Soil Sci. Soc. Amer. Proc. 34:118-121. Burrough, P. A. 1983. Multiscale source of spatial variation in soil: I. The application of fractal concepts to nested level of soil variation. J. Soil Sci. 34:577597. Byerlee, D. , L. Harrington, and M. Collinson. 1980. Planning technologies appropriate for farmers: Concepts and procedures. International Maize and Wheat Improvement Center (CIMMYT) , Mexico. Byerlee, D. , L. Harrington, and D.L. Winkelmann. 1982. Farming Systems Research: Issues in research strategy and technology design. Am. J. Agri. Econ. 64:897-904. Chang, S.C., and M.L. Jackson. 1957. Fractionation of soil phosphorus. Soil Sci. 84:133-144. Cochran, V. L. , L. F. Elliot, and R. I. Papendick. 1980. Carbon and nitrogen movement from surface applied wheat straw. Soil Sci. Soc. Amer. J. 44:978-982. de Wit, C.T. 1953. A physical theory on placement of fertilizers. Versl. Landbouwk. Onderzoek. No. 59.4. Douglas, Jr. C. L. , R. R. Allmaras, P. E. Rasmussen, R. E. Ramig, and N. C. Roger, Jr. 1980. Wheat straw composition and placement effects on decomposition in dryland agriculture of the pacific Northwest. Soil Sci. Soc. Amer. J. 44:833-837. Easterwood, G.W. and J.B. Sartain. 1990. Organic coatings on P fertilizers: Influence on plant growth on a Florida Ultisol (in review). Soil and Crop Sci. Soc. Fla. Proc. 49: Eberhart, S.A., W.A. Russell. 1966. Stability parameters for comparing varieties. Crop Sci. 6:36-40.

PAGE 154

140 Elliot, L. F., and J. W. Blaylock. 1975. Effects of wheat straw and alfalfa amendments on solubilization of manganese and iron in soil. Soil Sci. 120:205-211. Empresa Brasileira de Pesquisa Agropequaria. 1976. Composisao da solucao nutritiva para milho. Sete Lagoas, CNPMS, MG, Brazil. Empresa Brasileira de Pesquisa Agropequaria. 1979. Guia de excursao XVII Cogreso Braziliero de sciencia do solo, Manaus, EMBRAPA. Servicio Nacional de Levantamento e Conservacao do solos, Rio de Janeiro, Brazil. Empresa Brasileira de Pesquisa Agropequaria. 1984. Boletim agrometeorologico no. 6. EMBRAPA, UEPAE de Manaus, Manaus, Brazil. Evans, C.E., and E.J. Kamprath. 1970. Lime response as related to percent Al saturation, solution Al, and organic matter content. Soil Sci. Soc. Am. Proc. 34:893-896. Fattori, T.R., F.B. Mather, P.E. Hildebrand. 1990. Methodology for partitioning poultry producers into recommendation domains. Agri. Systems. 32:197-205. Fox, R.L., and E.J. Kamprath. 1970. Phosphate sorption isotherms for evaluating the phosphate requirements of soils. Soil Sci. Am. Proc. 34:902-906. Fox, R.L., and B.T. Kang. 1978. Influence of phosphorus fertilizer placement and fertilization rate of maize nutrition. Soil Sci. 125:34-40. Fox, R.L., W.M.H. Saunders, S.S.S. Rajan. 1986. Phosphorus nutrition of pasture species: Phosphorus requirement and root saturation values. Soil Sci. Soc. Am. J. 50:142-148. Francis, C.A. , and P.E. Hildebrand. 1989. Farming systems research/ extension and the concepts of sustainability p. 1-8. Ninth annual farming systems symposium, Univ. Arkansas, Fayetville, AR. 8-11 Oct., 1989. Gilman, G.P. 1981. Effects of pH and ionic strength on the cation exchange capacity of soils with variable charge Aust. J. Soil Res. 19:94-96. Giordano, P.M., J.J. Mortvedt, and D.A. Mays. 1975. Effect of municipal waste on crop yields and uptake of heavy metals. J. Environ. Qual. 3:394-399.

PAGE 155

141 Gomez, K.A., and A. A. Gomez. 1985. Statistical procedures for agricultural research. John Wiley and Sons, New York, USA. Gunary, D. 1970. A new adsorption isotherm for phosphate in soil. J. Soil Sci. 21:72-77. Hammond, L.L., S.H. Chien, and A.U. Mokwunye. 1986. Agronomic value of unacidulated and partially acidulated phosphate rocks indigenous to tropics. Adv. Agron. 40:89-140. Harter, R.D., and G. Smith, 1981. Langmuir eguation and alternate methods of studying "adsorption" reactions in soils, p. 167-182. In R.H. Dowdy, et al. (ed) Chemistry in the soil environment. Spec. publ. no. 40. Am. Soc. of Agron., Madison, WI . Harter, R.D. 1984. Curve-fit errors in Langmuir adsorption maxima. Soil Sci. Soc. Am. J. 48:749-752. Harwood, R. 1979. Small farm development: Understanding and improving farming systems in the humid tropics. Westview Press, Boulder, Colo. Haynes, R.J. 1982. Effects of liming on phosphate availability in acid soils. Plant Soil 68:289-308. Heliums, D.T., S.H. Chien, and J.T. Touchton. 1989. Potential agronomic value of calcium in some phosphate rocks from South America and West Africa. Soil Sci. Soc. Am. J. 53:459-462. Hildebrand, P.E. 1981. Combining disciplines in rapid appraisal: The 'sondeo' approach. Agricultural Administration 8:423-432. Hildebrand, P.E. 1984. Modified stability analysis of farmer-managed on-farm trials. Agron J. 76:271-274. Hildebrand. P.E. 1986. The Sondeo: A team rapid survey approach, p. 93-98. In P. E. Hildebrand (ed) Perspective on farming systems research and extension. Lynne Reiner publishers, Boulder, Colorado. Hildebrand, P.E. 1990. Modified stability analysis and onfarm research to breed specific adaptability for ecological diversity. Paper presented at the symposium on Genotype-by-Environment Interaction and Plant Breeding. LSU, Baton Rouge.

PAGE 156

142 Hill, R.R., Jr. and J.E. Baylor 1983. Genotype x environment interaction analysis in alfalfa. Crop sci. 23:811-815. Hingston, F.J., R.J. Atkinson, A.M. Po^ner 1967. Specific adsorption of anions. Nature 215.1459 1461. Holford, I.C.R., and G.E.G. Mattingly. 1974. The high and low-energy phosphate absorbing surfaces in calcareous soils. J. Soil Sci. 26:407-417. Holford, I.C.R., Wedderbern, R.W.M. , and Mattingly, G.E.G. 1974 A Langmuir two-surface equation as a model tor phosphate adsorption by soils. J. Soil Sci. 25:242-255. Hsu P.H., and D.A. Rennie. Reactions of phosphate in ' aluminum systems: II. Precipitation of phosphate by exchangeable aluminum on a cation regin. Can. J. Soil Sci. 42:210-221. Hundal, H.S., C.R. Biswas, and A.C. Vig. 1988. Phosphorus sorption characteristics of flooded soil amended with green manure. Trop. Agric. 65:185-187. Izza, C, and R. Indiati. 1982. Effect of farm organic residues added to the soil on phosphorus sorption. Crops and Soils. 43:78-90 Jones, J. P. and J. A. Benson. 1975. Phosphate sorption isotherms for fertilizer needs of sweet corn ( Zea mays) grown on a high P fixing soil. Comm. Soil Sci. Plant Anal. 6:465-477. Jungk, A., and S.A. Barber, 1974. Phosphate uptake rate of corn as related to the proportion of roots exposed to phosphate. Agron. J. 66:554-557. Kamprath, E.J. 1967. Residual effects of large application of phosphorus on high fixing soils. Agron. J. 59:25-27. Kamprath, E.J. 1970. Exchangeable aluminum as a criterion for liming leached mineral soils. Soil Sci. Soc. Am. Proc. 34:252-254. Keng, J., and G. Uehara. 1974. Chemistry, mineralogy , and taxonomy of Oxisols and Ultisols. Soil and Crop Sci. Soc. Fla, Proc. 33:119-126. Khasawneh, F.E., and E.C. Doll. 1978. The use of phosphate rock for direct application. Adv. Agron. 30:159-206.

PAGE 157

143 Kosaki T. and S.R.J. Anthony. 1989. Multivariate approach to grouping soils in small fields. I. extraction of factors causing soil variation by principal component analysis. Soil Sci. Plant Nutr. 35:469-477. Kyuma, K. , and T. Tulaphitak. 1985. Changes in soil fertility and tilth under shifting cultivation. I. general description of soil and effect of burning on the soil characteristics. Soil Sci. Plant Nutr. 31:227238. Larsen, S. 1967. Soil phosphorus. Adv. Agron. 19:151-210. Larsen, J.E., G.F. Warren, and R. Langston. 1959. Effect of iron, aluminum and humic acid on phosphorus fixation by organic soils. Soil Sci. Soc. Proc. 438-440 Larsen, W. E., R. F. Holt, and C. W. Carlson. 1978. Residues for soil conservation, p. 1-17. In W. R. Oswald (ed.). Crop residue management systems. Amer. Soc. Agron. J. 64:204-208. Leal, J.R., and A.C.X. Velloso. 1973. Dessorcao de fosfato absorbido em latosolos sob vegetacao de cerrado. II. Reversibilidade da isoterma de adsorbcao de fosfato em relacao ao pH da solucao em eguilibrio. Pesq. Agropec. Bras. (Ser. Agron). 8:89-92. Leikam, D.R. , R.E. Lamond, P.J. Gallagher, and L.S. Murphy. 1978. Improving N-P application. Agrichem. Age 22(3) :6. Leikam, D.R, L.S. Murphy, D.E. Kissel, D.A. Whitney, and H.C. Moser. 1983. Effects of N and P application method and nitrogen source on winter wheat grain yield and leaf tissue phosphorus. Soil Sci. Soc. Am. J. 47:530535. Lin, C.S., and G. Butler. 1990. Cluster analyses for analyzing two-way classification data. Agron. J. 82:344-349. McNeill, M. 1964. America. Science 211:86-86 Mead, J. A. 1981. A comparison of the Langmuir, Freundlich, and Temkin equations to describe phosphate sorption properties of soils. Aust. J. Soil Res. 19:333.342. Meda, L. , and G. F. Cerofolini. 1989. Physical chemistry of, in and on silicon. Springer series in material science v. 8. Springer-Verlag Heidelberg, Germany.

PAGE 158

144 Mekaru, T. . and G. Uehara. 1972. Anion adsorption in ferruginous tropical soils. Soil Sci. Soc. Am. Proc. 36:296-300. Mendez, J., and E.J. Kamprath. 1978. Liming of Latosols and the effect on P response. Soil Sci. Soc. Am. J. 41:8688. Mokwunye, U. 1977. Phosphorus fertilizers in Nigerian savana soils. I-Use of phosphorus sorption isotherms to estimate the phosphorus requirement of maize at Samaru. Trop. Agric. 54:265-270. Moormann, F.R., and B.T. Kang. 1978. Microvariability of soils in the tropics and its agronomic implications with special reference to West Africa, p 29-43. In Diversity of the soil in the tropics. ASA, Madison, WI. Morais, F.I., A.L. Page, and L.J. Lund. 1976. The effect of pH, salt concentration, and nature of electrolytes on the charge characteristics of Brazilian tropical soils. Soil Sci. Soc. Am. 40:521-527. Moreno, E.C., W.L. Lindsay, and G. Osborn. 1960. Reactions of dicalcium phosphate dihydrate in soil. Soil Sci. 90:58-68. Moussie, M., and C. Muhitira. 1988. Classification of farmers into recommendation domains. Proceedings of Farming Systems Research/Extension Symposium. University of Arkansas, p. 241-250 Muljadi, D., A.M. Posner, and J. P. Quirk. 1966. The mechanism of phosphate adsorption by kaolinite, gibsite, and pseudoboehmite. Part II. The location of adsorption sites. J. Soil Sci. 17:230-237. Munns, D.N., and R.L. Fox. 1976. The slow reaction which continues after phosphate adsorption: Kinetics and equilibrium in some tropical soils. Soil Sci. Soc. Am. J. 40:46-51. Murphy, J., and J. P. Riley. 1962. A modified single solution method for the determination of phosphate in natural waters. Anal. Chem. Acta. 27:31-36. Murray, K. , and P.W. Linder. 1984. Fulvic acids: Structures and metal binding. II. Predominant metal binding sites. J. Soil Sci. 33:217-222.

PAGE 159

145 Nagarajah, S., A.M Posner, and J.P Quirk. 1970. Competitive adsorption of phosphate with polyglacturonate and other organic anions on kaolinite and oxide surfaces. Nature 228:83-84. Nair P.S., T.J. Logen, A.N. Sharpley, L.E. Sommers, M.A. Tabatabai, and T.L. Yuan. 1984. Interlaboratory comparison of a standardized phosphorus adsorption procedure. J. Environ. Qual. 13:591-595. Njobvu, C.A. 1986. Factors influencing recommendation domain boundaries of the farming system and levels of agricultural development in Lusaka province, Zambia. Proceedings of Farming Systems Research/Extension Symposium. Kansas State University, pp. 254-259. Norris, J.M. 1970. Mutivariate methods in the study of soils. Soils Fertil. 33:13-18. Olsen, S.R. 1953. Inorganic phosphorus in alkaline and calcareous soils. Agronomy 4:89-122. Olsen, S.R and Watanabe, 1957. A method to determine a phosphorus adsorption maximum of soils as measured by the Langmuir isotherm. Soil Sci. Soc. Am. Proc. 21:144.149. Olsen, S.R., and S.A. Barber. 1977. The effect of waste application on soil phosphorus and potassium, p. 197215. In L.F Elliott and F.J. Estivenson (ed) Soils for management of organic waste and waste waters. Am. Soc. Agron. Madison, WI. Parr, J. F. , and R. I. Papendick. 1978. Factors affecting the decomposition of crop residues by microorganisms, p. 101-129. In W. R. Oschwald (ed.). Crop residues management systems. ASA Special pub. No. 31, Amer. Soc. Agron., Madison, WI. Perrott, K.W. 1978. The influence of organic matter extracted from humified clover on the properties of amorphous aluminosilicates . II. Phosphate retention. Aust. J. Soil Res. 16:341-346. Posner, A.M., and J.W. Bowden. 1980. Adsorption isotherms: Should they be split? J. Soil Sci. 31:1-10. Rajan, S.S.S., K. W. Perrott, and W.M.H. Saunders. 1974. Identification of phosphate reactive sites of hydrous alumina from proton consumption during phosphate adsorption at constant pH values. J. Soil Sci. 25:438447

PAGE 160

146 Rajan, S.S.S., and R.L. Fox. 1975. P hos P^ te . ^rp^rj by soils. II. Reaction in tropical acid soils. Soil Sci. Soc. Am. Proc. 39:446-451. Raun, W.R., D.H. Sander, and R.A. Olson. 1987. Phosphorus fertilizer carriers and their placement for minimum till corn under sprinkler irrigation. Soil Sci. Soc. Am. J. 51:1055-1062. Reddy K.R. 1990. Phosphorus retention capacity of stream sediments and associated wetlands. Final report submitted to the South Florida Water Management District. West Palm Beach, PI. Reeve, N.G., and M.E. Summer. 1970. Effects of aluminum toxicity and phosphorus fixation on crop growth on Oxisol from Natal. Soil Sci. Soc. Am. Proc. 34:263-267. Salinas, J.G., and P. A. Sanchez. 1976. Soil-plant relationships affecting varieties and species differences in tolerance to low available soil phosphorus. Ciencia e Cultura (Brazil) 28 (2 ): 156-168 . Sample, E.C., R.J. Soper, and G.J. Racz. 1980. Reactions of phosphate fertilizers in soils, p. 263-310. In M. Stelly (ed.) The role of phosphorus in agriculture. ASA, CSSA, and SSSA, Madison, WI . Sanchez, P. A. , and S.W. Buol. 1975. Soils of the tropics and the world food crisis. Science 188:598-603. Sanchez, P. A. , D.E. Bandy, J.H. Villachica, and J.J. Nicholaides. 1982. Amazon basin soils: Management for continuous crop production. Science 216:821-827. Sartain, J.B., and J.J. Street. 1980. Systems for supplying micronutrients. Florida Fertilizer and Lime Conference Proc. 10:1-21. SAS Institute, Inc. 1985. SAS User's guide: Statistics. SAS Institute, Inc, Cary, NC. Schnitzer, M. , and h. Kodama. 1977. Reactions of minerals with soil humic substances, p. 741-770. In J.B. Dixon (ed.) Minerals in soil environments. SSSA, Madison, WI. Shanner, W.W., P.F. Phillip, and W.R. Schmehl. 1982. Farming systems research and development: Guidelines for developing countries. Westview Press, Boulder, Colorado.

PAGE 161

147 Singh, B. B., and J. P. Jones. 1976. Phosphorus sorption and 9 Resorption characteristics of soil as affected by organic residues. Soil Sci. Soc. Amer. J. 40:389-394. Singh, B.B., and J. P. Jones. 1977. Phosphorus sorption isotherm for evaluating phosphorus requirements of lettuce at five temperature regimes. Plant Soil 46:3144. Sioli, H. 1980. Amazonia. p257-268. In F. Barbira-Scazzochio (ed.) Land, People, and Planning in Contemporary Amazonia. Cambridge University Press, Cambridge. Sleight, D.M., D.H. Sander, and G.A. Peterson. 1984. Effect of fertilizer phosphorus placement on the availability of phosphorus. Soil Sci. Soc. Am. J. 48:336-340. Smyth, T.J., and M. Cravo. 1990. Critical phosphorus levels for a corn and cowpea in a Brazilian Amazon Oxisol. Agron. J. 82:309-313. Smyth T.J., and M. Cravo. 1990. Phosphorus management for continuous corn-cowpea production in a Brazilian Amazon Oxisol. Agon. J. 82:305-309. Solis, P., and J. Torrent. 1989. Phosphate sorption by calcareous Vertisol and Inceptisol of Spain. Soil Sci. Soc. Am. J. 53:456-459. Sweeney, D.W., and D.A. Graetz . 1988. Chemical and decompositional characteristics of anaerobic digester effluent applied to soil. J. Environ. Qual. 17:309-313. Swift, R.S. and R.J. Haynes. 1989. The effects of pH and drying on adsorption of phosphate by aluminum-organic matter associations. J. Soil Science. 40:773-781. Swinton, S.M., and L.A. Samba. 1986. Defining agricultural recommendation domains in South-Central Niger. Proceedings of Farming Systems Research/Extension Symposium. Kansas State University, pp. 318-331. Syers, J.K., M.G. Browman, G.W. Smillie, and R.B. Corey. 1973. Phosphate sorption by soils evaluated by the Langmuir adsorption equation. Soil Sci. Soc. Am. Proc. 37:358-363. Tate, K.R., and B.K.G. Theng. 1980. Organic and its interactions with inorganic soil constituents, p. 225-

PAGE 162

148 249. in Theng, B.K.G. (ed.) soils with variable charge. New Zealand Soc. Soil Sci. Tshabalala M. and D. Holland. 1986. Recommendation domains arid tne design of on-farm trials research and extension in Lesotho. Proceedings of Farming Systems Research/Extension Symposium. Kansas State University, pp. 345-355. Tullv, R.C., and A.M. Alberti. 1985. Recommendation domains reconsidered. 1986. Proceedings of Farming Systems Research/Extension Symposium. Kansas State University, pp. 236-253. Uehara, G. and G.P. Gillman. 1980. Charge characteristics of soils with variable and permanent charge minerals: I. Theory. Soil Sci. Soc. Am. J. 44:252-255. Upraity, V.N., K.D. Joshi, and B.K. Singh. 1984. Variety adoption: A function of agronomic and socio-economic variables. CSP, Nepal. USDA, Soil Survey Staff. 1975. Soil Taxonomy, USDA Agriculture Handbook 43 6, Washington D.C. Vander Veen, M. , and S.B. Mathema, 1979. Key Informant Survey result for Lele, Lalitpur. HMG, Nepal. Van Raij, B. and M. Peech. 1972. Electrochemical properties of some Oxisols and Alfisols of the tropics. Soil Sci. Soc. Am. Proc. 36:587-593. van Wijk, W.R. 1966. Introduction, the physical method, p. 1-16. In W.R. van Wijk (ed) Physics of plant environment. North Holand Publishing Co., Amsterdam. Wilding, L.P., N.E. Smeck, and L.R. Drees. 1977. Silica in soils: Quartz, crystobalite, tridymite and opal. p. 471-552. In J.B. Dixon and S.B. Weed (ed.). Minerals in soil environments. SSSA, Madison, WI. Woodruff, J.R. and E.J. Kamprath. 1965. Phosphorus adsorption maximum as measured by the Langmuir isotherm and its relationship to phosphorus availability. Soil Sci. Soc. Proc. 29:148-150. Yost, R.S., 1977. Effect of rate and placement on availability and residual value of P in an Oxisol of Central Brazil. Ph.D. Diss. N.C. State University. Yost, R.S., E.J. Kamprath, E. Lobato, and G.C. Naderman, Jr. 1979. Phosphorus response of corn on an Oxisol as

PAGE 163

149 influenced by rates and placement. Soil Sci. Soc. Am. J. 43:338-343. Yost, R.S., E.J. Kamprath, G.C. Naderman, and E. Lobato. 1981. Residual effects of phosphorus applications on a high phosphorus adsorbing oxisol of central Brazil. Soil Sci. Soc. Am. J. 45:540-543. Yuan T.L. 1980. Adsorption of phosphate and water'ex^ractable soil organic material by synthetic aluminum silicates and acid soils. Soil Sci. Soc. Am. J. 44:951955. Yuan, T.L., and D.E. Lucas. 1982. Retention of phosphorus by sandy soils as evaluated by adsorption isotherms. Soil and Crop Sci. Soc. Fla. Proc. 41:195-201. Zandstra, H.G., E.C. Price, J. A. Litsinger, and R.A. Morris. 1981. A methodology for on-farm cropping systems research. IRRI, Manila, Philippines.

PAGE 164

BIOGRAPHICAL SKETCH Braj K. Singh was born on May 22, 1955, in a small village in Nepal. B.K. graduated from high school in Kalaiya, Bara, in 1971. He received his B.S. and M.S. in agronomy from Peoples' Friendship University in Moscow, USSR, in 1980. From the same university he earned a RussianEnglish interpreter's diploma. He worked as an assistant lecturer of agronomy, an extension officer, and cropping systems/ farming systems agronomist in Nepal from 1980 to 1986. He was awarded a Fulbright scholarship in 1986 and later a farming systems assistantship from the University of Florida. B.K. conducted his dissertation project in collaboration with TropSoil and EMBRAPA in Manaus, Brazil. He is married to Zoila Alvarado Singh and has a son, Alexander Alvarado Singh, and a daughter Carol Alvarado Singh. 150

PAGE 165

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. Sartain, Chairman 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 of Doctor of Philosophy. Professor of Food and Resource Economics 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. Donald A. Graetz 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 of Doctor of Philosophy. Edward A. Hanlon Associate Professor of Soil Science

PAGE 166

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. Kenneth L. Buhr Assistant Professor of Agronomy This dissertation was submitted to the Graduate Faculty of 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. AUGUST 1990 Dean, /College of Agriculture Dean, Graduate School


79
Upraity, et al., 1985, Fattori, 1990) has been used to
select environment-specific production techniques. This
technique does not depend on the concept that a regression
coefficient of unity is always favorable for the selection
of a technology. Adherence to this concept leads to
rejection of superior technology for a specific environment
in search of a 'stable' technology.
Objectives
The overall goal of this research was to identify
appropriate technology for maize and cowpea production by
small farmers. The specific objectives were to: (i) measure
the performance of selected treatments in different land
types, (ii) study the changes in soil fertility parameters
as influenced by the application of organic amendments, and
(iii) develop location-specific recommendations.
Materials and Methods
Site Description
Two small farming communities (mean cultivated size = 3
ha/farm) in the municipality of Rio Preto da Eva, located in
the state of Amazonas, Brazil were selected for on-farm
experimentation. The area is accessible only by small
motorboats. In trying to improve living conditions of these
marginal farmers, the government of Brazil was just
beginning a small watershed management program. The project


SUSTAINING CROP PHOSPHORUS NUTRITION OF HIGHLY LEACHED
OXISOLS OF THE AMAZON BASIN OF BRAZIL THROUGH USE OF
ORGANIC AMENDMENTS
By
BRAJ K. SINGH
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
1990
tJUMWERSlTY OF FLORIDA LIBRARIES


40
(Easterwood and Sartain, 1987), and 4. reduction in bonding
energy of adsorbed anions resulting in low residence time
for adsorbed P (Hudal et al, 1988). Increased adsorption may
be due to: 1. cation bridging between organic anions, and
Fe3+, Al3+, and Ca2+ leading to formation of new sites for P
adsorption (Appelt et al, 1975), and 2. ability of organic
ligands to maintain hydroxy-Al, and Fe in a non-crystalline
state and thus maintaining a greater surface area (Swift and
Haynes, 1989).
Amendments chemical composition. Although organic acids
are an integral part of all organic matter by far they are
not the only reactive component influencing P adsorption.
Singh and Jones (1976) suggested that the P content of
organic residues plays an important role in the release or
fixation of added P. Similarly, chemical and decompositional
characteristics of the organic matter may influence total
C02 evolution (Sweeney and Graetz, 1988). This gas when
dissolved in water, forms carbonic acid, which is capable of
decomposing certain primary minerals. Elemental ratios such
as C/N and C/P are also considered valuable indicators for
net mineralization or immobilization of N and P contained in
organic amendments.
One of the important studies in this area was
conducted by Blair and Boland (1978) who examined the
release of P from white clover residue in high and low P
status soils. Their results suggested that the addition of


124
Table 4-13. Technology selection for a given land type based
on different evaluation criteria for maize cultivation.
Land Type Evaluation Criteria
Yield
Mg/ha
Return to
Cash Cost
$/$
Return
Total
$/$
i to
Cost
Net
Income
$
PF1
CM
FP
FP
CM
SF1
CM
FP
FP
CM
PF2
CM
CM
CM
CM
SF2
CM
CM
CM
CM
WL
CM
NR
NR
NR
Table 4-14.
Technology selection
for a
given
land type
on different evaluation criteria for cowpea
cultivation.
Land Type Evaluation Criteria
Yield
Mg/ha
Return to
Cash Cost
$/$
Return to
Total Cost
$/$
Net
Income
$
PF1
TSP
FP
FP
TSP
PF2
TSP
FP
FP
TSP
PF3
CM
CM
CM
CM
SF1
CM
CM
CM
CM
SF2
CM
CM
CM
CM
SF3
CM
CM
CM
CM
WL
CM
CM
CM
CM


LIST OF TABLES
Table 2-1. Ionic species and concentration of nutrient
solution used with silica to simulate solution
chemistry of a Xanthic Hapludox in a phosphorus
release study by organic amendments 13
Table 2-2: Chemical composition of organic amendments
used in the incubation study 16
Table 2-3: Selected physical and chemical properties of
Ap horizon of the Xanthic Hapludox used in
adsorption studies 17
Table 2-4. Release of 0.01 M CaCl2 extractable
P (M9 g following incubation of organic
amendments with soil and silica
as matrix substratum 23
Table 2-5. Langmuir parameters (k and b) for P
adsorption by soil incubated with different
organic amendments 29
Table 2-6. Langmuir parameters (k and b) based on net
release of P measured by sequential extraction of
simulated silica matrix with 0.01 M CaCl2. ... 32
Table 2-7. Difference in the estimated values of
Langmuir parameters (b and k) based on the
estimation of preadsorbed P by simulated silica
matrix technique and sequential extraction. ... 33
Table 3-1. Factorial arrangements of treatments for the
greenhouse study 48
Table 3-2. Description of treatments tested in the
field 50
Table 3-3. Selected chemical properties of the fine
fraction (<2 mm) of the Xanthic Hapludox 55
Table 3-4. Selected physical properties of the Xanthic
Hapludox 55
vii


soil based on simulated silica matrix (SSMT) resulted in a
better estimation of the Langmuir adsorption maxima. This
technique involved mixing of acid washed sand with a
nutrient solution (without P) and inoculation of the sand
with microbes and incubation with organic amendments.
In the glasshouse experiment the highest dry matter
(DM) production was obtained with the CM. A strong
relationship was observed between DM production and Mehlich-
I extractable P (r2 = 0.88).
In the on-station field experiments, CM applied plots
produced more grain in first as well as second crops
compared to the plots which had received Canavalia
ensiformis. Mucuna aterrima, PCW, and TSP. All organic
amendments improved the soil P reserve and reduced Ca
leaching indicating that application of organic amendments
could lead to sustained crop P nutrition.
In farmers' field studies, three amendments were tested
in different land types with maize and cowpea crops. The
same amount of P applied from different amendments had
different effect on maize and cowpea production. However,
the selection of a given technology for a given land type
(environment) was dependent on farmers' goals. Based on the
criteria of grain production CM was recommended for maize in
all environments and for cowpea in poor environments.
xiv


104
Net income/cash cost. The net income/cash cost ratio is
an important criterion for selection of improved technology
which requires additional cash input. The information
presented in Appendix A indicated that FP had the highest
return over invested cash. One dollar invested gave a return
of US $ 32.3 at a production level of 3.10 Mg ha'1 in the
best environments. The relationship between e and net
return/cash cost ratio followed an exponential function
(Fig. 4-5a). The return over invested cash for other
treatments was less than $ 6.3 in all environments.
Processed city waste treatment ( PCW20+TSP20) had a negative
return at all production levels. The distribution of
confidence intervals for net income/cash cost is presented
in Fig. 4-6a. The mean net return from a dollar invested in
CM20+TSP20 was $ 3.2 and ranged from $ 2.4 to 4.2. It was
observed that 99 % of the time a farmer will receive higher
than 2.4 $/$ investment. The average return from a dollar
invested in FP was $ 18. However there is 18 % chance that
FP will provide a return of less than $ 2.4.
Net income/total cost. The net income/total cost ratio
indicated that return over total cost was low for all
treatments. However, as the production environments improved
(e> 1.85 Mg ha1) the net return to total cost was within
the range of $ 1.8 to 3.5 (Fig. 4-5b). Chicken manure
(CM20+TSP20) provided stable return which was within the


4
Transfer of Technology
It is widely believed that future efforts to increase
food production must be directed towards the marginal lands
of the developing world (Shaner et al., 1982). To achieve
this goal, more research on appropriate technology is
required with direct farmer involvement in problem
identification, research priority identification, and
technology evaluation (Harwood, 1979). Such an attempt will
lead to location-specific technology tailored to fit
farmers' circumstances, and accelerate the process of
technology diffusion (Hildebrand, 1983).
Goals and Objectives
The overall goal of this research was to examine the
role of organic amendments in sustaining P nutrition of a
highly leached Oxisol. The specific objectives were: (i) to
devise a technique to measure the P release patterns from a
decomposing organic material, (ii) to examine the effect of
a combined application of organic and inorganic P in a band
on sustaining crop P nutrition, and (iii) to conduct
farmers' field trials to validate on-station research
results and delineate recommendation domains for selected
treatments.


Table 2-7. Difference in the estimated values of Langmuir parameters (b and k) based
on the estimation of preadsorbed P by simulated silica matrix technique and
sequential extraction.
Incubation Period
35 Days
rg.+
amend. bT ks
65 Days 150 Days
b kb
k
Tephrosia -54.1 (14.2) 0.19 (0.01) 77.6
CM 265.1 (35.8) 0.53 (0.02) 810.6
Kudju 29.2 (5.9) 0.01 (0.00) 48.0
PCW 91.3 (15.9) 0.00 (0.00) -46.8
Grass 4.3 (1.3) 0.07 (0.01) 7.3
(8.3) 0.27 (0.04) 122.2 (8.4) 0.30 (0.01)
(58.3) 1.09 (0.09) 1170.7 (50.4) 1.90 (0.12)
(4.9) 0.22 (0.02) 90.0 (6.8) 0.29 (0.01)
(15.2) 0.18 (0.01) 71.2 (7.2) 0.25 (0.02)
(1.4) 0.14 (0.01) 63.7 (6.9) 0.29 (0.02)
-f Pueraria phaseoloides (Kudzu), Tephrosia candida (Tephrosia), Mixed Gramineae
(Grass), Aerobically Processed City Waste (PCW), and Chicken Manure (CM).
=f= b, P adsorption maxima, fig P g1
§ k, bonding energy, mL /ig P1
Numbers in the parenthesis are standard deviations.
CO
u>


140
Elliot, L. F., and J. W. Blaylock. 1975. Effects of wheat
straw and alfalfa amendments on solubilization of
manganese and iron in soil. Soil Sci. 120:205-211.
Empresa Brasileira de Pesquisa Agropequaria. 1976.
Composisao da solucao nutritiva para milho. Sete
Lagoas, CNPMS, MG, Brazil.
Empresa Brasileira de Pesquisa Agropequaria. 1979. Guia de
excursao XVII Cogreso Braziliero de sciencia do solo,
Manaus, EMBRAPA. Servicio Nacional de Levantamento e
Conservacao do solos, Rio de Janeiro, Brazil.
Empresa Brasileira de Pesquisa Agropequaria. 1984. Boletim
agrometeorologico no. 6. EMBRAPA, UEPAE de Manaus,
Manaus, Brazil.
Evans, C.E., and E.J. Kamprath. 1970. Lime response as
related to percent A1 saturation, solution Al, and
organic matter content. Soil Sci. Soc. Am. Proc.
34:893-896.
Fattori, T.R., F.B. Mather, P.E. Hildebrand. 1990.
Methodology for partitioning poultry producers into
recommendation domains. Agri. Systems. 32:197-205.
Fox, R.L., and E.J. Kamprath. 1970. Phosphate sorption
isotherms for evaluating the phosphate requirements of
soils. Soil Sci. Am. Proc. 34:902-906.
Fox, R.L., and B.T. Kang. 1978. Influence of phosphorus
fertilizer placement and fertilization rate of maize
nutrition. Soil Sci. 125:34-40.
Fox, R.L., W.M.H. Saunders, S.S.S. Rajan. 1986. Phosphorus
nutrition of pasture species: Phosphorus requirement
and root saturation values. Soil Sci. Soc. Am. J.
50:142-148.
Francis, C.A., and P.E. Hildebrand. 1989. Farming systems
research/ extension and the concepts of sustainability,
p. 1-8. Ninth annual farming systems symposium, Univ.
Arkansas, Fayetville, AR. 8-11 Oct., 1989.
Gilman, G.P. 1981. Effects of pH and ionic strength on the
cation exchange capacity of soils with variable charge.
Aust. J. Soil Res. 19:94-96.
Giordano, P.M., J.J. Mortvedt, and D.A. Mays. 1975. Effect
of municipal waste on crop yields and uptake of heavy
metals. J. Environ. Qual. 3:394-399.


30
bonding energy did not follow any defined pattern. For the
grass treatment (mixture of different grasses) bonding
energy value increased from 0.25 to 0.37 mL ng'1, for
Canavalia ensiformis (Canavalia) bonding energy dropped from
0.39 to 0.26 mL nq1, and for CM it remained at 0.10 mL g'1.
This finding does not agree with the data obtained by Hundal
et al. (1988) where all studied amendments reduced the
bonding energy as the incubation period increased from 20 d
to 40 d.
The range of P in organic amendments varied from 1.7 to
25.0 g kg'1 on a dry matter basis. There was a high
concentration of Zn and Cu in PCW, and Ca in CM. Being
polyvalent cations, they have a high affinity for P
adsorption through cation bridging (Holford and Mattingly,
1975; Haynes, 1989). The decomposition rate constants of the
organic amendments were different and so was the release of
different elements as the decomposition proceeded. These
factors also may have contributed to the observed
differences in P adsorption maxima and bonding energy.
Silica matrix. The value of the total adsorbed P was
adjusted based on the amount of P released from the organic
amendments in the silica matrix. The Langmuir equation was
fitted to the data, and the value of adsorption maxima and
bonding energy were recalculated for selected treatments.
The results are presented in Tables 2-6 and 2-7. The
recalculated adsorption maxima and bonding energy values


18
pH
Figure 2-1. Acid-base potentioinetric titration curves for
the Ap horizon of Xanthic Hapludox with varying
concentration of CaCl2.


122
Technology Selection for Different Recommendation Domains
The concept of RDs is very beneficial in selecting the
nitch for a technology in the overall research domain.
However, the usefulness of the concept in field conditions
will depend on selecting technology evaluation criteria
based on the farmers' needs and aspirations, and using the
criteria to divide the research area into RDs. More often
there are more than one criteria. A summary of technology
selection for a given land type based on different
evaluation criteria is presented in Table 4-13 and 14.
Maize
The selection of appropriate technology is the function
of land type and the farmer's goal. If the goal is to obtain
highest grain production in PF1, CM20+TSP20 should be
recommended. In the same land type, if the goal is to obtain
highest return to limited cash FP, should be suggested for
use. However, one should be aware of the fact that FP is not
sustainable in long run and within a year or two the
productivity will decline to a level which will no longer
offer highest return to invested cash. In SF2, CM20+TSP20
out performed all treatments based on all evaluation
criteria under consideration. In WL, CM20+TSP20 can be
recommended based on the yield criteria, but economically
none of the tested treatments was sound.


143
Kosaki T., and S.R.J. Anthony. 1989. Multivariate approach
to grouping soils in small fields. I. extraction of
factors causing soil variation by principal component
analysis. Soil Sci. Plant Nutr. 35:469-477.
Kyuma, K., and T. Tulaphitak. 1985. Changes in soil
fertility and tilth under shifting cultivation. I.
general description of soil and effect of burning on
the soil characteristics. Soil Sci. Plant Nutr. 31:227-
238.
Larsen, S. 1967. Soil phosphorus. Adv. Agron. 19:151-210.
Larsen, J.E., G.F. Warren, and R. Langston. 1959. Effect of
iron, aluminum and humic acid on phosphorus fixation by
organic soils. Soil Sci. Soc. Proc. 438-440
Larsen, W. E., R. F. Holt, and C. W. Carlson. 1978. Residues
for soil conservation, p. 1-17. In W. R. Oswald (ed.).
Crop residue management systems. Amer. Soc. Agron. J.
64:204-208.
Leal, J.R., and A.C.X. Velloso. 1973. Dessorcao de fosfato
absorbido em latosolos sob vegetacao de cerrado. II.
Reversibilidade da isoterma de adsorbcao de fosfato em
relacao ao pH da solucao em equilibrio. Pesq. Agropec.
Bras. (Ser. Agron). 8:89-92.
Leikam, D.R., R.E. Lamond, P.J. Gallagher, and L.S. Murphy.
1978. Improving N-P application. Agrichem. Age 22(3):6.
Leikam, D.R, L.S. Murphy, D.E. Kissel, D.A. Whitney, and
H.C. Moser. 1983. Effects of N and P application method
and nitrogen source on winter wheat grain yield and
leaf tissue phosphorus. Soil Sci. Soc. Am. J. 47:530-
535.
Lin, C.S., and G. Butler. 1990. Cluster analyses for
analyzing two-way classification data. Agron. J.
82:344-349.
McNeill, M. 1964. America. Science 211:86-86
Mead, J.A. 1981. A comparison of the Langmuir, Freundlich,
and Temkin equations to describe phosphate sorption
properties of soils. Aust. J. Soil Res. 19:333.342.
Meda, L., and G. F. Cerofolini. 1989. Physical chemistry of,
in and on silicon. Springer series in material science
v. 8. Springer-Verlag Heidelberg, Germany.


75
Tshabalala and Holland (1986) indicated that the "average"
fanner is a myth and the programs designed to help him or
her will fail, but used the RD concept in matching the
improved technology to the group likely to be interested in
taking advantage of it.
Economic criteria are also important in the delineation
of RDs (Hildebrand and Poey, 1985). Improved technology
often requires more cash and labor investments. Both
resources are scarce on small subsistence family farms.
Under limited cash availability, and so many competing uses
for it, a farmer will consider an option which will give him
or her the highest return per dollar invested. In this
context, economic evaluation criteria such as net return
ha'1, net return/total cost, and net return/cash cost play
important roles.
The RD concept is being used in other disciplines
(Fattori, 1990) and the use is being considered as an
extension tool guiding the effective dissemination of
technology appropriate to small farm conditions.
Stability Analysis
The stable performance of crop cultivars over a wide
range of environmental conditions is generally regarded as
desirable, but there is disagreement both on its definition
and on the most appropriate methods for its statistical
measurement from yield data trials (Becker, 1981; Hill and


47
A88 S O N D J89 F M A M J J A
Months
Figure 3-1. Geographic location of Amazon basin in Brazil
(a), on-station and farming systems research (FSR)
sites (b), and effective rainfall during the period of
August 1988 until August 1989 (c) at EMBRAPA station in
Manaus, Brazil.


Cofidence Coefficient (%)
108
Net Income/Cash Cost
Net Income ($/ha)
Figure 4-6. Distribution of confidence intervals for net
income/cash cost, net income/total cost, and net income
for different treatments used for maize cultivation.


87
(1.35-4.21 cmol (+) charge kg'1 soil), high Al saturation
(60-90%), and low Mehlich-I extractable
p (0-12.0 /xg g"1)
Change in Soil Fertility with Time
Initial soil characteristics were grouped by land types
(PF and SF). Within each group, soil pH, Al saturation,
ECEC, and Mehlich-I P were regressed with the number of
years the land was in production. The objective was to see
the magnitude of change in soil fertility parameters after
deforestation had taken place and the land was used for
agricultural purposes. The results are presented in Table 4-
4. The initial values (intercept) of soil pH, Mehlich-I
extractable P, and ECEC were higher in PF compared to SF. Al
saturation was higher in SF (81.5 % compared to 61.1% for
PF) .
The rate of decline in pH in PF was 0.15 unit per year
which was similar to decline in pH in SF. Aluminum
saturation was increasing at the rate of 7% per year in PF
compared to 4% in SF. The rate of decline in Mehlich-I P in
both land types was in the range of 2.1-2.6 nq g'1 soil. A
relatively low ECEC in both land types (2.3-3.6 cmol (+)
charge kg1 soil) indicated that most cations were leached
following heavy rainfall. And those present were being taken
up by vegetation or being washed out at the rate of 0.80 and
0.16 cmol (+) charge kg'1 soil every year from PF and SF,


Table 3-5. Selected chemical properties of the fine
fraction (<2 mm) of the Xanthic Hapludox 56
Table 3-6. Analysis of variance for maize herbage
dry weight production per pot in
the glasshouse study 57
Table 3-7. Mean changes in soil pH (H20) following
application of selected organic amendments in the
glasshouse study 60
Table 3-8. Orthogonal contrasts of maize grain yield
under different treatments applied in a 30 cm wide
band at UEPAE research station, Amazonas, Brazil. 62
Table 4-1. Application of N, P, and K in different
treatments tested in on-farm experimentation for maize
and cowpea crops in the municipality of Rio Preto da
Eva, Amazonas, Brazil 82
Table 4-2. Characterization of experimental plots for
maize testings in the municipality of Rio Preto da
Eva, Amazonas, Brazil 89
Table 4-3. Characteristics of experimental plots for
cowpea trials 90
Table 4-4. Relationship between soil characteristics
with year in crop production in different land
types 91
Table 4-5. Relationship between soil characteristics
measured after treatment application with grain
yield for maize and cowpea crops in the
municipality of Rio Preto da Eva, Amazonas,
Brazil 92
Table 4-6. Summary of ANOVA for multilocatonal maize
testing in the municipality of Rio preto da Eva,
Amazonas, Brazil 97
Table 4-7. Duncan Multiple Range Test (DMRT) for maize
crop in the municipal of Rio preto da Eva,
Amazonas, Brazil 97
Table 4-8. Combined Analysis of Variance for maize
grain yield in the municipal of Rio Preto da Eva,
Amazonas, Brazil 98
Table 4-9. Environmental index (e) for maize production in
the Municipality of Rio Preto da Eva, Amazonas,
Brazil 100
viii


Yield, Mg ha"1
102
WL
SF 2 PF 2
PF 1
SF 1
PF 1
Environmental Index (e), Mg ha 1
Figure 4-4a. Response of different treatments to
environmental index for maize production, Rio Preto da
Eva, Amazonas, Brazil.


94
O 4.0 8.0 12.0 16.0
e<1.85 Mg/ha e>1.85Mg/ha
Figure 4-2. Range of different soil characteristics by
recommendation domains for maize trials.


98
Table 4-8. Combined Analysis of Variance for maize grain
yield in the municipal of Rio Preto da Eva, Amazonas,
Brazil.
Source
DF
EMS
F Value
Pr >
BLOCK
1
0.21
LOC
4
5.04
14.2
0.02
BLOCK*LOC
4
0.35
TRT
3
15.81
13.9
0.02
TRT*LOC
12
1.13
9.5
0.01
BLOCK*TRT*LOC
15
0.11
f Chi-square test for homogeneity of variances was not
significant (P<0.05). All 5 replicated locations were
included in the combined analysis.
Random model. LOC was tested against BLOCK*LOC, TRT against
TRT*LOC, and TRT*LOC against BLOCK*TRT*LOC.


74
Diffusion Domains (DDs) (Hildebrand, 1986). Many
researchers, however, have considered RDs as a synonym for
cropping systems (Zandstra, et al., 1981), farming types
(Njobvu, 1986), and homogeneous groups (Moussie and Muhitira
(1988). The limit set by the definition of Byerlee et al.
(1980) has also been broadened to include agroclimatic zones
and individual fields, in addition to farms.
Criteria for Delineating Recommendation Domains
There is a great debate on the criteria used for
delineation of RDs. Socioeconomic criteria may be just as
important as agroclimatic, and agroecological variables
(Njobvu, 1986) in delineating domains. If so, the resulting
domains are often not amenable to geographical mapping
because farmers of different domains may be interspersed in
a given area. Moussie and Muhitira (1988) attempted to
classify farmers into relatively homogenous groups using
cluster and discriminant statistical analysis. This method
employed the use of qualitative information obtained from
sondeos, and the selection of key variables from in-depth,
formal surveys to obtain a discriminant function which helps
to identify the most important variables in classifying
farmers into RDs. Swinton and Samba (1986) used four
agronomic criteria: average annual rainfall, soil fertility,
soil texture, and depth to the subterranean water table for
defining agricultural technology recommendation domains.


113
Net income/cash cost. The results of MSA for net
income/cash cost showed that FP had the highest return over
invested cash in environments with e value >1.32 Mg ha 1
(Fig. 4-8a). A return of as high as 37.7 $/$ cash investment
was obtained with FP compared to less than $ 10 for other
treatments in all environments. The confidence interval
calculation is presented in Fig. 4-9a-b. This test suggested
that in poor environments CM20+TSP20 provided a mean return
of $ 6.0 on each $ invested. This mean fell within the range
of $ 4.0 to 8.0. Farmer practice had a wider Cl. Therefore,
the return over invested cash in CM20+TSP20 in poor
environments was better and less risky compared to other
treatments. However, in good environments FP was superior to
other treatments.
Net income/total cost. The point where FP, TSP40, and
CM20+TSP20 crossed was at an e value of 1.32 Mg ha1. Below
this value the net return/total cost ratio was equal for
TSP40, and CM20+TSP20 treatments (within $ 2.8-3.0) (Fig. 4-
8b) and they were higher than FP. Processed city waste
(PCW20+TSP20) gave the lowest net return/total cost ratio in
all environments which suggests that a farmer will lose
money by applying PCW20+TSP20 for cowpea production in these
environments. In good environments performance of all
treatments improved. However, FP gave the highest return
over total cost.


125
C O W P E A
Recommendation Domain I
Recommendation Domain II
Figure 4-12. Recommendation domains, based on yield, and
their relationship to land types in the municipality of
Rio Preto da Eva,Amazonas, Brazil.


11
Kjeldahl procedure. Potassium, Ca, Mg, Fe, Zn, Cu, and Mn
were determined with an atomic adsorption spectrophotometer,
and P was determined colorimetrically by the Murphy and
Riley (1962) procedure.
Incubation study
Soil matrix. Soil used for this study was obtained from
the Ap horizon of a Xanthic Hapludox (clayey, kaolinitic,
isohyperthermic) (EMBRAPA, 1979) which had PZNC (point of
zero net charge) at pH of 4.2. Five grams of organic ground
material were mixed with 95 g of air-dried soil and
incubated in plastic containers for 35, 65, and 150 d at 30
3C. The soil in each container was mixed thoroughly every
6 d and was kept at 45 % moisture content. After each
incubation period, 2-g duplicate soil samples were extracted
with 0.01 M CaCl2 with a soil to solution ratio of 1:10 and
1 h shaking. The cycle was repeated three times. In
calculating P desorbed, an allowance was made for the 2 mL
of supernatant carried over from each cycle (Fox and
Kamprath, 1970).
Adsorption was determined using 2-g air-dried samples,
in duplicate, eguilibrated for 6 d with 20 mL of 0-60 nq
mL 1 P solution in 0.01 M CaCl2 in polyethylene tubes at 25
2C. To inhibit microbial activities a few drops of toluene
were added to each tube. After 6 d the solution P was
determined by the ascorbic acid method (Murphy and Riley,


77
The reason Eberhart and Russell (1966) used the site
mean as an environmental index was because a lack of
knowledge of the relationship of macro-environmental
differences such as temperature gradients, rainfall
distribution, and soil types did not permit the computation
of an index which could transform the environment into a
continuous variable. However, attempts continue to guantify
the production environment. Advancements in using the
multivariate approach to group soils in the field based on
variations in systematic and random components are
forthcoming (Winding and Dress, 1983). Systematic variation
is caused by difference in the parent material, relief and
biological action as well as soil management practices such
as fertilizer application and tillage. Random variation
which is called "noise" by Burrough (1983), represents the
statistical heterogeneity of the soil.
The need for assessing the factors causing soil
microvariability in the tropics has been stressed by Moorman
and Keng (1978). In general, some soil characteristics are
mutually correlated with each other (Norris, 1970). Hence,
factors causing soil variation which are reflected in one or
more of the soil characteristics may be used as criteria for
grouping soil.
To analyze the causes of soil variation Kosaki and
Anthony (1989) applied principal component analysis (PCA),
which is a mathematical technigue used to summarize data and


Table 3-5. Selected chemical properties of the fine fraction (<2 mm) of the Xanthic
Hapludox.
Depth
OH
Extractable bases ECEC
Charcre"^
Oxides^
Al Sat
h2o
KC1
Ca
Mg
K Al
(-)
( + )
Fe
Al
(ECEC
cm
0-15
4.7
4.3
0.26
0.06
cmol (+)
0.05 1.20
kg'1
1.57
1.96
0.77
g
0.07
100g
0.30
1
76.4
15-30
4.4
4.1
0.24
0.04
0.02 1.31
1.61
0.60
1.43
0.08
0.35
81.4
30-45
4.3
4.1
0.19
0.05
0.01 1.24
1.49
0.64
1.63
0.09
0.38
83.2
Acid ammonium oxalate extraction.
Acid-base potentiometric titration.
U1
CTi


149
influenced by rates and placement. Soil Sci. Soc. Am.
J. 43:338-343.
Yost, R.S., E.J. Kamprath, G.C. Naderman, and E. Lobato.
1981. Residual effects of phosphorus applications on a
high phosphorus adsorbing oxisol of central Brazil.
Soil Sci. Soc. Am. J. 45:540-543.
Yuan, T.L. 1980. Adsorption of phosphate and water-
extractable soil organic material by synthetic aluminum
silicates and acid soils. Soil Sci. Soc. Am. J. 44:951-
955.
Yuan, T.L., and D.E. Lucas. 1982. Retention of phosphorus by
sandy soils as evaluated by adsorption isotherms. Soil
and Crop Sci. Soc. Fla. Proc. 41:195-201.
Zandstra, H.G., E.C. Price, J.A. Litsinger, and R.A. Morris.
1981. A methodology for on-farm cropping systems
research. IRRI, Manila, Philippines.


71
Conclusion
In the glasshouse experiment, among the three
amendments tested PCW produced the lowest dry matter per pot
when applied alone or in combination with TSP. The highest
dry matter production was obtained with the application of
CM. A strong relationship was observed between dry matter
production and Mehlich-I extractable P (r2 = 0.88).
In the field experiments, CM applied plots produced
more grain in first as well as second crops compared to the
plots which had received Canavalia ensiformis. Mucuna
aterrima. PCW, and TSP. This finding agreed with the
prediction made from the laboratory incubation studies using
silica as the matrix substratum to evaluate the
decomposition rate of organic amendments. The prediction was
made based on P adsorption maxima and binding energy values,
and indicated that CM applied plots will produce superior
yield compare to the plots which received other amendments.
When CM was applied in combination with TSP higher
grain yield was obtained compared to the same amount of P
applied from TSP. All organic amendments improved the soil P
reserve and reduced Ca leaching.


Table 4-2. Characterization of experimental plots for maize
testings in the municipality of Rio Preto da Eva, Amazonas, Brazil.
General^ Physical Chemical
(e)
Loc.
Land
Type
Year in
Prod.
Density
Sand
Clay
Silt
PH
ECEC+
A1
Sat.
P(M-I)
g cm3
g
H
O
O
iQ
i

g
I00g'1
Mg g'1
3.10
7
PF
1
1.20
68.6
25.5
5.9
5.2
4.21
58.3
7.4
2.75
6
PF
1
1.04
82.7
14.2
3.1
5.1
3.45
69.1
7.1
2.23
2
SF
1
1.07
60.7
31.8
7.5
4.6
2.29
91.7
4.5
2.10
8
PF
1
1.37
75.0
21.0
3.9
4.5
2.26
79.2
6.8
1.96
5
PF
2
1.15
56.5
35.3
8.2
4.6
2.45
80.0
5.0
1.43
4
SF
2
1.04
80.7
14.8
4.5
4.1
3.12
94.8
2.8
1.11
1
SF
2
1.14
47.4
39.3
13.3
4.2
1.99
90.7
2.0
0.20
3
WL
4
1.05
59.6
32.2
8.1
3.9
1.35
94.8
trace
f PF = Area cleared for the first time by slash and burn from primary forest.
SF = Area cleared from the secondary forest which was left for 5-7 years
for building up of soil fertility, (e) = Environmental Index, Mg ha'1
4= ECEC = Effective cation exchange capacity (cmol+ charge kg'1 soil)
§ Mehlich I extractable phosphorus
00


130
content of the amendments played a key role in influencing
adsorption and desorption characteristics of the soil.
The results of the decomposition study in a soil matrix
indicated that maize and grass treatments showed P
immobilization. Processed city waste immobilized soil P at
35 and 65 d, but at 150 d a net release was observed. All
legumes followed a similar P release patterns. However, When
silica was used as a matrix there was a net release of P
from all amendments independent of P content and C:P ratio.
The data on P adsorption maxima calculated by the
Langmuir equation for CM, based on P release from amendments
in soil and silica matrices, indicated that with increasing
incubation time, P adsorption maxima increased. Continued
adsorption beyond adsorption maxima demonstrated the
presence of a precipitation reaction, or multilayer
adsorption. For tephrosia and kudzu adsorption maxima
decreased with time. A good agreement between adsorption
maxima and actual adsorption of P based on SSMT was observed
at the higher equilibrating solution P concentration (60 ¡iq
g'1 soil) which indicated the validity of the SSMT technique
in predicting P requirements of amended soils.
Chapter III examines the results of glasshouse and on-
station field studies on the effect of a suboptimal dose of
organic amendments in sustaining crop P nutrition and
improving the efficiency of inorganic P.


xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID EL1EYNUPZ_DEKP7L INGEST_TIME 2015-02-19T21:03:46Z PACKAGE AA00028751_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES


31
were higher than the values obtained based on the P release
from the amendments in the soil matrix. At 150 d for the CM
treatment, the adsorption maxima were higher by 1170 /j.g g 1
soil and the bonding energy was higher by 20 fold. The low
adsorption maxima (496 /g g1) and bonding energy (0.10 /Ltg
mL1) for soil matrix indicated that the P capacity factor
for CM applied plots was lower compared to plots which
received other amendments and, therefore, CM plots should
have the lowest residual P effect in field testings. But,
the result from the Silica Matrix technique indicated
otherwise. The high adsorption maxima for CM at 150 d of
incubation (1667 /g g1) and the high bonding energy (2.0 mL
/xg"1) are indications of sustained P desorption and residual
P availability.
Predicted and measured phosphorus adsorption. The data
on P adsorption maxima calculated by the Langmuir equation
and actual P adsorption based on P release from amendments
in soil and silica matrices are presented in Fig. 2-7. For
the CM treatment, with increasing incubation time P
adsorption maxima increased. Continued adsorption beyond
adsorption maxima indicated the presence of a precipitation
reaction, or multilayer adsorption. For tephrosia and kudzu,
the adsorption maxima decreased with time. A good agreement
among adsorption maxima and actual adsorption, calculated
based on SSMT, was observed at higher equilibrating solution


105
range of $ 1.8 to 2.1 per dollar spent and had no
probability of a loss (Fig. 4-6b).
Net income. An analysis of net income ha'1 under four
tested treatments (Appendix A, and Fig. 4-5c) indicated that
TSP40, and CM20+TSP20 were superior to FP, and PCW20+TSP20
in all environments. Net income from TSP40, and CM20+TSP20
at their highest production levels were $ 622 and 513,
respectively compared to $ 388 and 16.5 for FP, and
PCW20+TSP20. A confidence interval was calculated for FP and
CM20+TSP20 treatments based on net income (Fig. 4-6c). Two
treatments not considered for Cl analysis were PCW20+TSP20,
TSP40. Processed city waste (PCW20+TSP20) had the lowest net
income, and TSP40 had agronomic performance similar to that
of CM20+TSP20. The confidence interval test indicated that
99 % of the time CM20+TSP20 will produced a net income
greater than $ 400 compared to a negative net income to an
income of $ 380 for FP practice (Fig. 4-6c). The narrow
range of net income with CM20+TSP20 treatment compared to FP
indicated that CM20+TSP20 provided less risk than FP, and
had a high net income ha'1.


83
Effective Rainfall, mm
400
300
200
100
o
-100
-200
A88 S O N D J'89 F M A M J J A
Months
Figure 4-1. Effective rainfall and suggested shift in the
planting date for maize in Rio Preto da Eva, Amazonas,
Brazil, (a) common practice, (b) suggested practice.


141
Gomez, K.A., and A.A. Gomez. 1985. Statistical procedures
for agricultural research. John Wiley and Sons, New
York, USA.
Gunary, D. 1970. A new adsorption isotherm for phosphate in
soil. J. Soil Sci. 21:72-77.
Hammond, L.L., S.H. Chien, and A.U. Mokwunye. 1986.
Agronomic value of unacidulated and partially
acidulated phosphate rocks indigenous to tropics. Adv.
Agron. 40:89-140.
Harter, R.D., and G. Smith, 1981. Langmuir equation and
alternate methods of studying "adsorption" reactions in
soils, p.167-182. In R.H. Dowdy, et al. (ed) Chemistry
in the soil environment. Spec. publ. no. 40. Am. Soc.
of Agron., Madison, WI.
Harter, R.D. 1984. Curve-fit errors in Langmuir adsorption
maxima. Soil Sci. Soc. Am. J. 48:749-752.
Harwood, R. 1979. Small farm development: Understanding and
improving farming systems in the humid tropics.
Westview Press, Boulder, Colo.
Haynes, R.J. 1982. Effects of liming on phosphate
availability in acid soils. Plant Soil 68:289-308.
Heliums, D.T., S.H. Chien, and J.T. Touchton. 1989.
Potential agronomic value of calcium in some phosphate
rocks from South America and West Africa. Soil Sci.
Soc. Am. J. 53:459-462.
Hildebrand, P.E. 1981. Combining disciplines in rapid
appraisal: The 'sondeo' approach. Agricultural
Administration 8:423-432.
Hildebrand, P.E. 1984. Modified stability analysis of
farmer-managed on-farm trials. Agron J. 76:271-274.
Hildebrand. P.E. 1986. The Sondeo: A team rapid survey
approach, p. 93-98. In P. E. Hildebrand (ed)
Perspective on farming systems research and extension.
Lynne Reiner publishers, Boulder, Colorado.
Hildebrand, P.E. 1990. Modified stability analysis and on-
farm research to breed specific adaptability for
ecological diversity. Paper presented at the symposium
on Genotype-by-Environment Interaction and Plant
Breeding. LSU, Baton Rouge.


51
Second maize. This trial was conducted to measure the
residual effects of the treatments applied during the first
maize crop. Maize stover was taken out of the field.
Nitrogen was applied in the same manner as for the first
crop. No other nutrient applications were made. Maize
(variety BR-5110) was dibble planted in the rows and the
population was adjusted to 55 x 103 plants ha1. Soil samples
were collected during tasseling and after harvest. Again 10-
20 maize leaves were sampled during tasseling initiation
from each plot. Maize grain production was recorded from
four inner rows and was adjusted to 15% moisture content.
Chemical Analysis of Soil and Plant Materials
Soil pH was determined in water, 0.01 M CaCl2, and 1.0
M KC1 using a soil:solution ratio of 1:2.5. Total P was
determined by wet combustion method and inorganic P
fractionation was carried out using Chang and Jackson
procedure (1965). The pH of NH^F was adjusted to 8.2.
Mehlich-I extractant (0.05 M HC1 + 0.0125 M H2SO 4) was used
to extract soil P, K, Zn, Cu, and Mo. Phosphorus was
determined colorimetrically and K by a flame photometer.
Aluminum, Ca, and Mg were extracted with unbuffered 1 M KC1
(1:10 soil:solution ratio). Aluminum was determined by
titrating the extract with 0.1 M NaOH to bromthymol blue
endpoint, and Ca and Mg were determined with an atomic
absorption spectrophotometer. Soil apparent density was


106
Figure 4-5. Relationship of net income/cash cost, net
income/total cost, and net income with environmental
index, in on-farm maize trials from Rio Preto da Eva,
Amazonas, Brazil.
Treatment
Intercept
b SE Y Esti
Net Income/Cash
Cost
FPf
0.1
1.85
0.42
PCW20+TSP20
-1.0
0.30
0.21
TSP40
1.5
2.71
0.56
CM20+TSP20
0.5
1.42
0.80
Net
Income/Total
Cost
FP-f
0.2
0.83
0.50
PCW20+TSP20
-1.0
0.30
0.21
TSP40
-0.5
1.00
0.33
CM20+TSP20
0.3
0.61
0.38
Net Income ha"
1
FPf
1.0
1.8
0.8
PCW20+TSP20
-217.0
71.3
44.3
TSP40
-142.9
261.0
55.0
CM20+TSP20
60.7
177.5
101.1
f Exponential relationship
SE = Standard Error


38
by entire root mass, and part of the root mass immersed in
nutrient solution and established a relationship between Ur
(g of nutrient taken up by the plant when part of the root
mass, Xr, was immersed in the solution) and Ub (g of
nutrient taken up by the plant when the entire root mass was
immersed in the solution).
Ur/Ub = (Xr/Xb)0-44
This relationship appears to be independent of crop type and
nutrient solution concentration. Under field conditions Ub,
Ur, Xr, and Xb take on new meanings.
Ub = uptake rate from broadcast fertilizer
Ur = uptake rate from banded fertilizer
Xr = width of the fertilizer band
Xb = distance between the crop rows
The use of data from P sorption curves concerning soil
solution P concentration to predict P uptake and crop yield
in combination with P placement analysis offers a way to
increase fertilizer use efficiency. Faced with a high P
fixing soil and a small guantity of fertilizer, the
fertilizer can be used to the best advantage by
concentrating it in a band so that the P concentration in
the soil solution is identical to the concentration that
produces a maximum yield in a broadcast application. Any
deviation from this optimum value, either to higher or to
lower concentrations, leads to less than an optimum return
per unit of fertilizer input.


Table 4-3. Characteristics of experimental plots for cowpea trials.
General^
Physical
Chemical
(e)
Loc
Land
Year
in Density
Sand
Clay
Silt
pH ECEC' Al
P(M-I)
Type
Prod.
Sat.
g cm"3
g I00g"1
g lOOg
'1 Mg
g1
1.92
7
PF
1
1.30
63.8
28.8
7.4
5.2
3.24
59.8
8.0
1.80
10
PF
1
1.29
52.1
37.1
10.8
5.4
2.21
61.2
12.9
1.67
3
PF
2
1.00
80.7
16.2
3.1
5.3
1.25
80.1
5.0
1.60
12
PF
2
1.29
67.7
27.6
4.7
4.9
1.91
63.2
7.0
1.55
11
PF
2
1.28
45.7
41.9
12.4
5.0
1.72
64.0
10.6
1.42
4
PF
2
1.15
56.5
35.3
8.2
5.1
2.31
86.6
4.0
1.25
1
SF
1
1.07
60.7
31.8
7.5
4.7
1.35
74.1
7.6
1.10
5
PF
3
1.22
70.1
23.5
6.4
4.9
0.99
82.3
2.3
1.00
2
SF
2
1.11
57.5
35.3
7.2
4.6
2.34
94.9
3.4
1.00
8
SF
3
1.39
60.9
36.1
3.0
4.3
1.83
87.5
trace
0.97
13
SF
2
1.42
74.6
20.7
4.7
4.3
1.94
83.4
6.1
0.95
9
SF
3
1.17
45.0
45.4
9.6
4.3
1.20
94.5
2.0
0.73
6
WL
2
1.41
71.3
25.9
2.9
4.6
1.66
94.2
4.8
-f PF = Area cleared for the first time by slash and burn from primary forest.
SF = Area cleared from the secondary forest which was left for 5-7 years for
building up of soil fertility.
ECEC = Effective cation exchange capacity (cmol+ charge kg'1 soil)
§ Mehlich I extractable phosphorus
O


Figure 3-2. Effect of different rates of selected
organic amendments in combination with different
rates of inorganic P on maize dry matter
production at 65 d after planting in a glasshouse
study 58
Figure 3-3. Relationship between Mehlich-I extractable
soil P (a), and maize leaf tissue
P concentration (b) with maize dry matter yield. 59
Figure 3-4. Rate of inorganic P applied through triple
superphosphate and its effect on maize yield in a
maize-maize rotation 64
Figure 3-5. Effect of selected organic amendments,
applied in a quantity equivalent to provide 26.4
kg ha'1 of P, on sustaining Mehlich-I extractable
soil P pool in a maize-maize rotation on a Xanthic
Hapludox 65
Figure 3-6. Effect of selected organic amendments,
applied in combination with inorganic phosphorus
source in a quantity equivalent to provide 8.8 kg
ha1 of P, on sustaining Mehlich-I extractable
soil P pool in a maize-maize rotation on a Xanthic
Hapludox 67
Figure 3-7. Leaching of calcium from surface horizon of
highly leached Xanthic Hapludox as influenced by
different organic amendments 69
Figure 3-8. Relationship between P concentration and
concentration of Zn and Cu in maize leaf tissue.. 70
Figure 4-1. Effective rainfall and suggested shift in
the plnting date for maize in Rio Preto da Eva,
Amazonas, Brazil, (a) common practice, (b)
suggested practice 83
Figure 4-2. Range of different soil characteristics by
recommendation domains for maize trials 94
Figure 4-3. Range of different soil characteristics by
recommendation domains for cowpea trials 95
Figure 4-4a. Response of different treatments to
environmental index for maize production, Rio Preto da
Eva, Amazonas, Brazil 102
Figure 4-4b. Distribution of confidence intervals for
maize production in poor (e<1.95 mg ha1), and
good (e>1.95 mg ha1) environments 103
xi


7
the maximum adsorption possible corresponds to a complete
monomolecular layer. Both of these postulates do not hold
for a heterogenous medium like soil (Larsen, 1967).
Deviations from the conventional Langmuir relationship
at high equilibrium P concentrations (above 15 nq mL 1) have
already been reported (Olsen and Watanabe, 1957; Hsu and
Rennie, 1962) which have led to the development of an
extended form of the Langmuir equation. Gunary (1970)
included a square-root term in the Langmuir equation.
Holford et al. (1974), Syers et al. (1973), and Holford
(1983) used a two-surface Langmuir equation. The two surface
equation has not, however, been universally accepted and
Posner and Bowden (1980) have discussed the futility of
attempting to split isotherms into two or more regions.
Similarly, a curve-fit error in estimating the Langmuir
adsorption maxima was described by Harter (1984). He claimed
the test of linearity was inadequate because plotting
concentration against itself reduces data variability and
always provides a significant correlation coefficient. He
suggested that a better test of the fit is to ascertain
whether the adsorption isotherm has the shape of the
equation model.
In spite of drawbacks, the Langmuir equation is used
widely to describe P adsorption by soil. The major advantage
of the Langmuir equation is that it allows for the
calculation of adsorption maxima along with a relative


CHAPTER V
SUMMARY AND CONCLUSIONS
The objective of this chapter is to summarize the work
presented in the preceding four chapters. The overall goal
of this research was to examine the role of organic
amendments in sustaining P nutrition of highly leached
Oxisol and to analyse the performance of selected treatments
in farmers' fields. To achieve this goal, laboratory
incubation studies, glasshouse studies, and on-station and
off-station field studies were conducted.
Chapter II presents the findings on P adsorption and
desorption by soil as influenced by organic amendments. This
chapter also investigates a new technique for measuring
nutrient release from a decomposing material.
Incubation of soil with amendments for 35 d had a
marked influence on total P adsorption by soil. Soil amended
with PCW adsorbed more P than unamended soil. However, with
an increase in incubation time, all amendments reduced P
fixation. Phosphorus desorption values for soil incubated
for 35 d were equal to that of the control, but soil
incubated for more than 35 d showed increased P desorption.
These findings suggest that the rate of decomposition and P
129


112
Modified stability analysis
Yield. For yield, it was observed that the CM20+TSP20
treatment was stable (slope = 0.03, compared to 1.13 for
TSP40) over all environments (e values 0.5-2.0 mg ha1).
Treatment CM20+TSP20 had a clear advantage over TSP40 in e <
1.32 Mg ha'1. But in e > 1.32 Mg ha'1 TSP40 outyielded
CM20+TSP20 (Fig. 4-7a). The reason for such 'cross-over'
interaction was believed to be the additional input of N
from CM. Excess N in the better environments caused cowpea
plants to grow taller and become susceptible to lodging. In
poor environments this phenomenon had no significant bearing
on plant growth due to inherent N deficiency in the soil.
The point of 'cross-over' interaction (1.32 Mg ha'1)
was taken as a reference point for the delineation of RDs.
All locations having e values < 1.32 (PF3; SF1, 2, 3; WL)
were grouped into RD1 (poor environment), and locations
having e values > 1.32 (PF 1, 2) were clustered into RD2
(good environment). The confidence limit test (Fig. 4.7b)
indicated that CM20+TSP20 was a stable performer in both
RDs. The confidence limits for the CM20+TSP20 treatment fall
within 1.4 to 2.1 Mg ha'1 in RD1, and 1.8 to 1.9 Mg ha1 in
RD2 compared to 1.0 to 1.8 Mg ha1 in RD1 and 2.1 to 2.8 Mg
ha1 in RD2 for TSP40. The 'cross-over' interaction along
with confidence limit test indicated that CM20+TSP20
treatment should be recommended in RD1 and TSP40 in RD2.


Depth, cm
69
0.0 1.0 2.0 3.0 4.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0
Figure 3-7. Leaching of calcium from surface horizon of
highly leached Xanthic Hapludox as influenced by
different organic amendments.


73
on farm input and output, it is expensive and time
consuming. In order to make surveying cost-effective, Vander
Veen and Mathema (1978) used key informants to gather
baseline information. This speeded-up the process but it
lacked interactive information.
Working for the farmers' cause requires direct
involvement in problem identification. Rapid rural
appraisals, often referred to as exploratory or diagnostic
surveys, are a simple and relatively quick method of
identifying constraints that operate in a defined area
(Abalu, et al., 1987). The sondeo (Hildebrand, 1981) is one
of the techniques of rapid rural appraisal which combines
different disciplines in a team. Its primary purpose is to
acquaint technicians with the area in which they are going
to work. No questionnaires are used; rather informal
interviews with farmers are conducted.
Recommendation Domains (RD)
In order to be cost effective, the research activities
have to address the problems of and provide solutions for
relatively large numbers of people. It is necessary,
therefore, to classify farmers with similar circumstances
into recommendation domains: groups of farmers for whom it
is possible to make more or less the same recommendation
(Perrin et al., 1978; Byerlee et al., 1980). More recently
the concept of RDs has been extended to Research Domains and


142
Hill, R.R., Jr. and J.E. Baylor. 1983. Genotype x
environment interaction analysis in alfalfa. Crop Sci.
23:811-815.
Hingston, F.J., R.J. Atkinson, A.M. Posner, and J.P. Quirk.
1967. Specific adsorption of anions. Nature 215:1459-
1461.
Holford, I.C.R., and G.E.G. Mattingly. 1974. The high and
low-energy phosphate absorbing surfaces in calcareous
soils. J. Soil Sci. 26:407-417.
Holford, I.C.R., Wedderbern, R.W.M., and Mattingly, G.E.G.
1974. A Langmuir two-surface equation as a model for
phosphate adsorption by soils. J. Soil Sci. 25:242-255.
Hsu, P.H., and D.A. Rennie. Reactions of phosphate in
aluminum systems: II. Precipitation of phosphate by
exchangeable aluminum on a cation regin. Can. J. Soil
Sci. 42:210-221.
Hunda1, H.S., C.R. Biswas, and A.C. Vig. 1988. Phosphorus
sorption characteristics of flooded soil amended with
green manure. Trop. Agrie. 65:185-187.
Izza, C., and R. Indiati. 1982. Effect of farm organic
residues added to the soil on phosphorus sorption.
Crops and Soils. 43:78-90
Jones, J.P. and J.A. Benson. 1975. Phosphate sorption
isotherms for fertilizer needs of sweet corn (Zea mays)
grown on a high P fixing soil. Comm. Soil Sci. Plant
Anal. 6:465-477.
Jungk, A., and S.A. Barber, 1974. Phosphate uptake rate of
corn as related to the proportion of roots exposed to
phosphate. Agron. J. 66:554-557.
Kamprath, E.J. 1967. Residual effects of large application
of phosphorus on high fixing soils. Agron. J. 59:25-27.
Kamprath, E.J. 1970. Exchangeable aluminum as a criterion
for liming leached mineral soils. Soil Sci. Soc. Am.
Proc. 34:252-254.
Keng, J., and G. Uehara. 1974. Chemistry, mineralogy and
taxonomy of Oxisols and Ultisols. Soil and Crop Sci.
Soc. Fla, Proc. 33:119-126.
Khasawneh, F.E., and E.C. Doll. 1978. The use of phosphate
rock for direct application. Adv. Agron. 30:159-206.


l d 6
PCW
CM
58
Mucuna
Figure 3-2. Effect of different rates of selected organic
amendments in combination with different rates of
inorganic P on maize dry matter production at 65 d
after planting in a glasshouse study.


123
Cowpea
For cowpea, all land types were divided into two
recommendation domains. In RD2 (PF1 and PF2), TSP40 should
be recommended based on yield and net income ha 1. Depending
on the evaluation criteria of return to cash cost and return
to total cost FP has advantage over other treatments. This
practice should not be continued for more than a year if
high return to invested cash is expected. In RD1 (PF3, SF1,
SF2, SF3, and WL), CM20+TSP20 is the best option based on
all evaluation criteria.
Redefining Recommendation Domains
A map of the area including location, land type, year
in crop production, and RDs based on the yield criterion is
presented in Fig 4-12. The resulting domains are not
amenable to geographical mapping because farmers of
different domains were interspersed in the area. The use of
different evaluation criteria can also lead to regrouping of
environments from one RD to another. For example, the use of
return to cash cost indicator moved one production
environment for maize and two for cowpea from RD1 to RD2
(Fig.4-13) .


148
249. In Theng, B.K.G. (ed.) soils with variable charge.
New Zealand Soc. Soil Sci.
Tshabalala, M., and D. Holland. 1986. Recommendation domains
and the design of on-farm trials research and extension
in Lesotho. Proceedings of Farming Systems
Research/Extension Symposium. Kansas State University,
pp.345-355.
Tully, R.C., and A.M. Alberti. 1985. Recommendation domains
reconsidered. 1986. Proceedings of Farming Systems
Research/Extension Symposium. Kansas State University,
pp. 236-253.
Uehara, G. and G.P. Gillman. 1980. Charge characteristics of
soils with variable and permanent charge minerals: I.
Theory. Soil Sci. Soc. Am. J. 44:252-255.
Upraity, V.N., K.D. Joshi, and B.K. Singh. 1984. Variety
adoption: A function of agronomic and socio-economic
variables. CSP, Nepal.
USDA, Soil Survey Staff. 1975. Soil Taxonomy, USDA
Agriculture Handbook 436, Washington D.C.
Vander Veen, M., and S.B. Mathema, 1979. Key Informant
Survey result for Lele, Lalitpur. HMG, Nepal.
Van Raij, B. and M. Peech. 1972. Electrochemical properties
of some Oxisols and Alfisols of the tropics. Soil Sci.
Soc. Am. Proc. 36:587-593.
van Wijk, W.R. 1966. Introduction, the physical method, p.
1-16. In W.R. van Wijk (ed) Physics of plant
environment. North Holand Publishing Co., Amsterdam.
Wilding, L.P., N.E. Smeck, and L.R. Drees. 1977. Silica in
soils: Quartz, crystobalite, tridymite and opal. p.
471-552. In J.B. Dixon and S.B. Weed (ed.). Minerals in
soil environments. SSSA, Madison, WI.
Woodruff, J.R. and E.J. Kamprath. 1965. Phosphorus
adsorption maximum as measured by the Langmuir isotherm
and its relationship to phosphorus availability. Soil
Sci. Soc. Proc. 29:148-150.
Yost, R.S., 1977. Effect of rate and placement on
availability and residual value of P in an Oxisol of
Central Brazil. Ph.D. Diss. N.C. State University.
Yost, R.S., E.J. Kamprath, E. Lobato, and G.C. Naderman, Jr.
1979. Phosphorus response of corn on an Oxisol as


66
The actual P values ranged from 0.2 fig g1 soil (control) to
about 8.0 ¡ig g'1 soil for other amendments measured at the
end of 240 d. This finding conflicts with the results
obtained by Izza and Indiati (1982) where they studied the
effect of various organic farm products on soil available P.
They found that incorporation of these materials in high P
fixing soil produced no effect on soil available P.
At 65 d TSP60 and M60 treatments had the same soil P
status. There was a sharp decrease in soil P with TSP60
treatment as the cropping season progressed compared to M60
and C60. The plots which received PCW maintained low P which
could not be described with the aid of any polynomial. When
TSP20 was combined with M20 the P status improved compared
to TSP40 (Figure 3-6). Application of inorganic P in the
proximity of the organic amendment may have reduced the
exposure of inorganic P to a larger soil volume leading to
reduced fixation. And the decomposition of organic material
may have inactivated active soil adsorption sites in the
localized band. Increased effectiveness of soil amendments
such as lime, in the proximity of organic matter has been
reported by Ahmed and Tan (1988). Combined application of
C20 with TSP20 was as good as TSP40, and PCW20 applied with
TSP20 was inferior not only to TSP40 but to all treatments,
except control.


93
initiation, and impair nodule function (Keyser and Munns,
1979). For cowpea, each unit increase in soil pH increased
production by 0.97 Mg ha1 for FP. The effect of P from
different amendments on cowpea yield varied. The intercept
values of 0.21 and 1.03 Mg ha'1 for PCW20+TSP20, and TSP40
along with 0.14 and 0.10 slope values indicated that same
amount of P from these two amendments had a markedly
different effect on cowpea yield. Yield variation was
attributed to differences in N, Ca, Zn, and Cu content in
amendments and high C:P and N:P ratios, which are explained
in Chapter II and III.
Land Types and Recommendation Domains
Soil characteristics for maize and cowpea trials for
different RDs are presented in Fig. 4-2 and 4-3. For maize,
e<1.85 Mg ha1 was considered RD1 (poor environments), and
e>1.85 Mg ha'1 RD2 (good environments). For cowpea e<1.32 Mg
ha1 was considered RD1, and e>1.32 Mg ha1, RD2. For both
crops in RD1, soil characteristics such as pH, Mehlich-I
extractable P, and ECEC values were lower than RD2, and A1
saturation was higher. Locations with favorable soil
characteristics for crop production (RD2) were in first or
second year of cultivation after clearing from PF, and first
year of cultivation after clearing from SF.


LU/X/O
28
Figure 2-6. Phosphorus adsorption isotherms for the Xanthic
Hapludox (0.01 M CaCl2) incubated with organic
amendments for 150 d. The lines in the figure
represents fitted Langmuir equation.


26
0 #P
@ 0
Figure 2-4. A schematic representation of molecular
structures of oxides of phosphorus (a), and oxides of
silicon (b).


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
SUSTAINING CROP PHOSPHORUS NUTRITION OF HIGHLY LEACHED
OXISOLS OF THE AMAZON BASIN OF BRAZIL THROUGH USE OF
ORGANIC AMENDMENTS
By
Braj K. Singh
August 1990
Chairperson: Dr. Jerry B. Sartain
Major Department: Soil Science
Retention of phosphorus by iron and aluminum oxides
plays an important role in determining ultimate availability
of P to plants in highly weathered soils of the tropics.
Different management strategies have been proposed to
overcome this problem. The overall objective of this
research was to study the influence of organic amendments
(plant origin, manure, and aerobically processed city waste
(PCW)) in sustaining P nutrition of Oxisol and to test the
performances of selected alternatives in a wide range of
production environments.
There was a marked difference in P adsorption and
desorption by soil preincubated with organic amendments. The
difference was attributed to incubation period and chemical
composition of amendments. Correction for preadsorbed P by
xiii


85
locations. For this purpose, yield obtained from each
treatment over locations was stepwise regressed on the
values of soil pH, ECEC, Mehlich-I extractable-P, and A1
saturation. These variables were measured from soil samples
taken during tasseling stage for maize and flowering stage
for cowpea (Hanway, 1967).
Modified stability analysis. The production
environments were separated based on the average yield of
all treatments at each location (Eberhart and Russell,
1966). In this way, environment becomes a continuous,
quantifiable variable. Yield for each treatment was related
to environment by simple linear regression.
Yjj = a + be
where = yield from treatment i at the jth
location, and
ej = jth environmental index.
The use of the regression coefficient 'b' in the linear
equations with a value near one to select the "stable"
treatment over all environments was avoided. Instead
superior treatments were identified for groups of
environments, or RDs (Hildebrand, 1984).
The distribution of confidence intervals for the
treatments within RDs for both crops was calculated as
follows:
Cl = Y taS/n,/!
Where;


CHAPTER I
INTRODUCTION
Statement of Problem
A leaching environment together with persistent high
rainfall and temperature are the defining conditions for the
development of Oxisols, a soil order which occupies over
1.12 billion ha and ranks 5th in worldwide distribution
(USDA, 1975). Their similarity as a group stems from the
composition of the colloid fraction rich in iron and alumina
minerals, and deficient in other nutrients. Sustained
production on these soils can be achieved only with adequate
application of lime, N, P, K, Mg, Zn, Cu, B, and Mo (Sanchez
et al., 1982). Among other nutrients, P deficiency appears
to be the most crucial and may show up as early as the
second year of cultivation. Large amounts of applied P are
required to attain levels of soil solution P which are
adequate for high crop yields (Yost et al., 1979).
Rapidly increasing cost of P fertilizer, and limited P
supply in the tropics, have led to extensive P management
studies on these soils. Fertilizer application techniques
(band, broadcast, and band+broadcast) have been widely
1


20
and this will happen in sufficient quantity only if the soil
has a large capacity to sorb and therefore desorb P. Thus,
even though the soils with high sesquioxides content require
more P to achieve a given level of P in soil solution, they
have the compensating value of being able to supply P to the
soil solution as it is taken up by plants. The results of
desorption studies are presented in Fig. 2-3.
Soil incubated with Mucuna aterrima (Mucuna), PCW, and
the control for 35 d did not show any difference in P
desorption. However, desorption was higher for all
amendments incubated with soil for 150 d (Fig. 2-3b). In
fact, CM reduced adsorption and increased desorption at all
incubation periods. The observed differences of organic
amendments in influencing P desorption characteristics of
the soil could be attributed to their decompositional
characteristics, P content, C:P and C:N ratios, and
concentration of such elements as Zn and Cu. Incubation of
soil with low P containing organic materials may not
initially influence soil P desorption characteristics.
Phosphorus release from organic amendments
Phosphorus release data for different amendments in
soil and silica substratum are presented in Table 2-4. More
P was detected in the silica matrix for all amendments
compared to the soil matrix. With increasing incubation
time, a decrease in P was observed for maize and grass


95
O 4.0 8.0 12.0 16.0
e<1.32 Mg/ha e>1.32 Mg/ha
Figure 4-3. Range of different soil characteristics by
recommendation domains for cowpea trials.


52
determined by the method of a measuring cylinder. The
cylinder was struck against a rubber pad 10 times from a
distance of 10 cm. The final weight of the cylinder was
taken and the apparent density was calculated as the
following: Apparent Density (g cm3) = Weight of dry soil at
105C / Volume of the soil in the cylinder. Particle size
distribution was measured by the pipet method.
All organic amendments and maize leaf samples were
dried at 65 C for 72 hours for dry matter determination.
Subsamples of the dried plant material were ground in a
Wiley mill to pass a 0.5-mm screen. A 0.2-gm sample of the
ground plant material was digested with H2S04 and H202.
Phosphorus was determined calorimetrically, and K, Ca, Mg,
and micronutrients were determined with an atomic absorption
spectrophotometer.
Results and Discussion
Selected chemical and physical properties of soil used
in the glasshouse and field studies are presented in Tables
3-3 through 3-5.
With increasing soil depth, a reduction in Al-P was
observed. But, the Langmuir adsorption maxima increased with
the depth (789 /zg g'1 soil at 30-45 cm) (Table 3-3) It is
interesting to note that with over 80% clay content the soil
had a hydraulic conductivity of 25.1 cm h'1 (Table 3-4). The
ammonium oxalate extracable Fe and A1 were high and so was


49
as urea was broadcast over the entire plot and the remaining
half was sidedressed at 35 and 70 d after planting. All K
was applied basal broadcast. Six furrows each 20 cm wide and
8-10 cm deep were opened in each plot 80 cm apart. The plot
size was 10x5 m.
Mucuna aterrima and Canavalia ensiformis grown for 60 d
in an adjoining field were harvested and passed through a
hay chopper to produce a uniform size of less than 6 cm. All
organic amendments were applied in bands to which different
rates of P from triple superphosphate (TSP) was added. A
soil cover about 2 cm thick was put over the amendments.
Twelve treatments were tested in a RCBD with 4 replications.
The P release information from the SSMT technique (Chapter
II) was used to calculate the amount of a given amendment
required to supply P equivalent to 8.8 and 26.4 kg ha'1.
First maize. Maize variety BR-5110 was planted on
December 27, 1988. Plant population was adjusted to
approximately 55 x 103 plants ha"1 during the first sidedress
at 30 d after planting. Soil samples were collected before
planting, during tasseling (65 d after planting) and within
a week after the harvest. Leaf samples were collected during
tasseling. They were taken from immediately below and
opposite to ear leaf. The inner four rows were harvested for
grain. Grain production was recorded at 15% moisture level.


This work is dedicated to my parents: Shree Mrit B.
Singh and Shrimati Naina Devi, and to my brothers: Sri
Biswanath Singh and Nawal K. Singh.


Table 2-6. Langmuir parameters (k and b) based on net release of P
measured by sequential extraction of simulated silica matrix with 0.01 M CaCl
Org.^
amend.
35 Days
65
Incubation
Davs
Period
150
Davs
r2
b+
k§
r2
b
k
r2
b
k
Tephrosia
0.87**
714
0.42
0.96**
556
0.56
Silica
0.97** 588
0.55
CM
0.99**
909
0.61
1.00**
1429
1.17
1.00**
1667
2.00
Kudzu
0.95**
502
0.43
0.97**
625
0.70
0.97**
625
0.89
PCW
0.97**
908
0.22
0.90**
769
0.45
0.97**
588
0.63
Grass
0.94**
769
0.32
0.92**
714
0.47
0.93**
667
0.65
** Significant at P = 0.01 level, ns = not significant
-f Pueraria phaseoloides (Kudzu), Tephrosia candida (Tephrosia), Mixed Gramineae
(Grass), Aerobically Processed City Waste (PCW), and Chicken Manure (CM).
4= b, P adsorption maxima, jug P g'1
§ k, bonding energy, mL /g P1


48
Table 3-1. Factorial arrangements of treatments for the
glasshouse study.
Type
Amendments^
Rate
P Equivalent
kg ha'1
Mucuna
1&2
8.8, 17.6
Organic
CM
1&2
8.8, 17.6
PCW
1&2
8.8, 17.6
Inorganic
TSP
1,2&3

rH
CO

00
o
f Mucuna (Mucuna aterrima), CM (Chicken Manure), PCW
(Aerobically Processed City Waste) TSP (Triple
superphosphate)
Field Experiment
The site was cleared by burning an existing sugarcane
crop. Sugarcane stems and rhizomes were taken out of the
field. A uniformity trial using maize as an indicator crop
was planted for 60 d in order to record the existing
variation in the field. Visual field ratings combined with
soil and plant chemical analysis results were used to select
a uniform area for planting the field experiment. The
selected area was divided into 48 plots and soil samples
were taken from three consecutive depths (0-15, 15-30, and
30-45 cm) in each plot.
Dolomitic lime (2 Mg ha1) was applied over the entire
area 2 weeks prior to maize planting. Nitrogen and K were
applied at the rate of 200 and 100 kg ha'1. Zinc, Mo, and Mn
(2 kg ha1 of each) were applied in the band. Half of the N


97
Table 4-6. Summary of ANOVA for multilocatonal maize
testing in the municipality of Rio preto da Eva,
Amazonas, Brazil.
Source df EMS
LI
L2
L3
L4
L5
Block
1
0.010ns
0.080ns
0.020ns
1.361*
0.151ns
Trt
3
3.270**
8.218**
0.190**
2.271*
6.418*
Error
3
0.008
0.067
0.010
0.101
0.411
Total 7
*, ** significantly different at 0.05 and 0.01 level of
probability, ns = not significant.
Table 4-7. Duncan Multiple Range Test (DMRT) for maize crop
in the municipal of Rio preto da Eva, Amazonas,
Brazil .t.
Trt
LI
L2
L3
L4
L5
CM20+TSP20
2.85a+
4.40a
0.65a
2.80a
3.60a
TSP40
1.3 0b
3.40b
0.15b
1.60b
3.40b
PCW20+TSP20
0.15c
1.10c
0.01b
1.lObc
0.70b
FP
0.15c
0. Old
0.01b
0.25c
0.15b
CV%
8.0
11.6
50.0
22.1
32.7
f Locations with single replications have been dropped from
the ANOVA.
4= Means in the same column followed by the same letter are
not significantly different at the 95% level of probability
as determined by Duncan Multiple Range Test.


55
Table 3-3. Selected chemical properties of the fine
fraction (<2 mm) of the Xanthic Hapludox.
Depth Org.
carbon
TP
OP
Chana
& Jackson^
Langmuir
MI-P max'''.
Al-P
Fe-P
Ca-P
a'1
cm
g kg
M9
9
0-15
14.6
200
25.2
51.6
0.3
0.3
2.4
550
15-30
12.2
120
14.5
18.5
2.8
0.2
1.8
620
30-45
8.3
90
6.7
5.2
0.1
0.1
1.2
789
-J- pH of NH4F was adjusted to 8.2. The reductant soluble P is
not included.
4= 0.01 M CaCl2 was used as an electrolyte.
OP = Organic phosphorus, TP = Total phosphorus
MI-P = Mehlich I extractable P.
Table 3-4. Selected physical properties of the Xanthic
Hapludox.
Depth Bulk Fine fraction Hydraulic
Density Sand Silt Clay' Conductivity
cm
g cm'3
kg kg'1
of <2
mm
cm h'1
0-15
1.11
0.14
0.11
0.75
25.1
15-30
1.30
0.11
0.06
0.88
6.2
30-45
1.16
0.09
0.09
0.82
7.3
-f Fine sand fraction is also included.


Yield, Mg ha -1
115
WL
SF 3 PF 3
SF 2 SF 1
PF 2
PF 1
Environmental Index (e), Mg ha 1
Figure 4-7a. Response of different treatments to
environmental index for cowpea production, Rio Preto da
Eva, Amazonas, Brazil.


44
Band spacing of applied N and P fertilizer affects the
concentration of these nutrients in the applied band and the
probability of roots contacting the band. Sleight et al.
(1984) showed that in high P-fixing soil, increasing the
probability that root-fertilizer contact will occur is more
important than reducing soil-fertilizer contact during the
first week of oat (Avena sativa L.) growth.
There is also a threshold value of soil solution P
concentration beyond which the P uptake rate does not
increase (Jungk, and Barber, 1974). Anghinoni and Barber
(1980) in a P placement experiment reported maize root
growth stimulation in the portion of the soil where P was
added. Maximum shoot dry weight in their experiment was
obtained by placing the fertilizer in 0.25 of the soil
volume.
Summary
Application of P in narrow bands results in reduced
fixation of applied P and improves crop P nutrition. This
belief is based on: (i) localized P is protected to some
degree against irreversible adsorption or precipitation
reactions with the soil, (ii) localized P may be more
readily accessible to seedling roots than P widely
distributed in the soil, and (iii) plants can be adeguately
supplied with P through a few roots which proliferate in the
fertilizer band. Further, combined application of organic


Table 4-9. Environmental index (e) (Mg ha'1) for maize production in the Municipal
of Rio Preto da Eva, Amazonas, Brazil.
TRT
Loci
blkl blk2
Loc2
blkl blk2
loc3
blkl
blk2
loc4
blkl blk2
loc5
blkl blk2
loc6
blkl
loc7
blkl
loc8
blkl
FP
0.10
0.20
0.00
0.00
0.00
0.00
0.10
0.40
0.20
0.10
2.20
2.50
0.20
PCW20+TSP20
0.10
0.20
1.20
1.00
0.00
0.00
0.40
1.80
0.90
0.50
1.00
1.40
0.70
TSP4 0
1.20
1.40
3.20
3.60
0.20
0.10
1.20
2.00
3.00
3.80
4.20
4.50
3.50
CM20+TSP20
2.90
2.80
4.10
4.70
0.80
0.50
2.40
3.20
4.30
2.90
3.60
4.00
4.00
(e) by rep
1.07
1.15
2.13
2.33
0.25
0.15
1.02
1.85
2.10
1.82
2.75
3.10
2.10
(e) by loc
1.11
2
.23
0.
20
1.43
1
.96
2.75
3.10
2.10
100


23
Table 2-4. Release of 0.01 M CaCl2 extractable P (/xg g" )
following incubation of organic amendments with soil
and silica as matrix substratum.
Org.+ Soil Silica
amend.
35
65
150
35
65
150
CM
109a*
204.9a
245.1a
447a
1031a
1380a
Mucuna
21b
14.Obc
16.9b
80b
124b
160b
Tephrosia
22b
14.9b
19.0b
65b
116b
152b
Cowpea
15c
11.5bc
16.2bc
43c
70c
92c
Peanut
18bc
10.3cd
8.6bcd
41c
117b
142b
Grass
6e
6.7de
3.4d
37c
54c
102c
Maize
14cd
12.5bc
3.9d
37c
77c
111c
PCW
3e
2 86
10.8bcd
34c
62c
64d
Canavalia
14cd
11.8bc
15.3bc
31c
83c
148b
Kudzu
lOd
11.4bc
14.lbcd
23c
71c
94c
Control
4e
4.4e
5.2cd
Trace
Trace
0. li
CV%
6.3
14.0
8.2
12.3
7.5
12.0
* Means in the same column followed by the same letter are
not significantly different at the 95% level of
probability, as determined by Duncan's Multiple Range
Test.
f Mucuna aterrima (Mucuna), Pueraria phaseoloides (Kudzu),
Canavalia ensiformis (Canav.), Zea mays (Maize), Arachis
hvpogaea (Peanut), Tephrosia candida (Tephrosia), Viona
unguiculata (Cowpea), Mixed Gramineae (Grass), Aerobically
Processed City Waste (PCW), and Chicken Manure (CM).


146
Rajan, S.S.S., and R.L. Fox. 1975. Phosphate adsorption by
soils. II. Reaction in tropical acid soils. Soil Sci.
Soc. Am. Proc. 39:446-451.
Raun, W.R., D.H. Sander, and R.A. Olson. 1987. Phosphorus
fertilizer carriers and their placement for minimum
till corn under sprinkler irrigation. Soil Sci. Soc.
Am. J. 51:1055-1062.
Reddy, K.R. 1990. Phosphorus retention capacity of stream
sediments and associated wetlands. Final report
submitted to the South Florida Water Management
District. West Palm Beach, FI.
Reeve, N.G., and M.E. Summer. 1970. Effects of aluminum
toxicity and phosphorus fixation on crop growth on
Oxisol from Natal. Soil Sci. Soc. Am. Proc. 34:263-267.
Salinas, J.G., and P.A. Sanchez. 1976. Soil-plant
relationships affecting varieties and species
differences in tolerance to low available soil
phosphorus. Ciencia e Cultura (Brazil) 28(2):156-168.
Sample, E.C., R.J. Soper, and G.J. Racz. 1980. Reactions of
phosphate fertilizers in soils, p. 263-310. In M.
Stelly (ed.) The role of phosphorus in agriculture.
ASA, CSSA, and SSSA, Madison, WI.
Sanchez, P.A., and S.W. Buol. 1975. Soils of the tropics and
the world food crisis. Science 188:598-603.
Sanchez, P.A., D.E. Bandy, J.H. Villachica, and J.J.
Nicholaides. 1982. Amazon basin soils: Management for
continuous crop production. Science 216:821-827.
Sartain, J.B., and J.J. Street. 1980. Systems for supplying
micronutrients. Florida Fertilizer and Lime Conference
Proc. 10:1-21.
SAS Institute, Inc. 1985. SAS User's guide: Statistics. SAS
Institute, Inc, Cary, NC.
Schnitzer, M., and h. Kodama. 1977. Reactions of minerals
with soil humic substances, p. 741-770. In J.B. Dixon
(ed.) Minerals in soil environments. SSSA, Madison, WI.
Shanner, W.W., P.F. Phillip, and W.R. Schmehl. 1982. Farming
systems research and development: Guidelines for
developing countries. Westview Press, Boulder,
Colorado.


78
investigate the relationship among variables. Variables
employed for PCA in their study included soil pH, organic
carbon, available P, exchangeable Ca, exchangeable Mg,
exchangeable K, sand, silt, and clay. For the computation of
principal components, they used a correlation matrix.
However, soil variation is only one factor among many which
influences the performance of a technology in a given
environment. Much remains to be understood concerning the
complicated interactions among agronomic, economic, social,
and cultural variables which have a bearing on the
performance of a technology.
Modified Stability Analysis and Farming Systems
Farming Systems Research/Extension, is considered to be
a dynamic, interactive, and problem oriented approach to
develop technology for farmers, particularly those with
limited resources. The technological base of FSR/E is on-
station research but it constitutes only a part of the
overall FSR/E program. The main activities are concentrated
in farmers' fields with direct farmer involvement in
technology evaluation and feedback. The farmers' field
trials involve a few treatments, and often without
replications. Farmer to farmer variation in management
practices for a given experiment is not controlled. Instead,
any unusual practice is recorded. Under these circumstances,
modified stability analysis (MSA) (Hildebrand, 1984;


10
prediction of how the soil P status will change upon
cropping must be considered unreliable (Bowman and Olsen,
1985).
The purpose of this study was to examine the effect of
different organic amendments on P adsorption and desorption
isotherms, and to devise a technique to measure net P
release from organic amendments during their decomposition.
Such a technique should lead to a direct and accurate
estimation of the quantity of P which originated from
decomposing organic amendments and adsorbed by soil. This
knowledge will aid in the accurate estimation of soil P
adsorption maxima and bonding energy.
Materials and Methods
All amendments mucuna ([Mucuna aterrima (Piper and
Tracy) Merr], kudzu [Pueraria phaseoloides (Roxb.) Benth],
canavalia [Canavalia ensiformis (L.) DC], maize (Zea mays
L.), peanut (Arachis hvpoqaea L.), tephrosia fTephrosia
candida (Roxb.) DC], cowpea (Vigna unquiculatal. mixed
gramineae (Grass), aerobically processed city waste (PCW),
and chicken manure (CM)) were dried at 65 C for 72 hrs for
dry matter determination. Dried plant material was ground in
a Wiley mill and passed through a 0.5 mm screen. A weighed
(0.2 g) sample of the ground plant material was digested
with H2S04 and H202 for the determination of macro and micro
elements. Total nitrogen was determined by the micro-


Adsorbed P ug/g Soil
2,000
1,500
1,000
500
35d 65d 150d
2,000
1,500
1,000
500
c) Kudju
b a b
2,000
1,500
1,000
500
34
b) Tephrosia
m sm
r~i ssm
LAM
35d 65d 150d
2,000
1,500
1,000
500
d) PCW
35d 65d 150d
35d 65d 150d
Incubation Time
Figure 2-7. A comparison of P adsorption by Xanthic Hapludox
based on the estimation of preadsorbed P by sequential
extraction of soil or silica matrix incubated with
organic amendments.


120
Net Income/Total Cost (US $)
Figure 4-10. Distribution of confidence intervals for net
income/total cost for selected treatments in poor
(e<1.32 Mg ha'1) and good (e>1.32 Mg ha'1) environments
for cowpea cultivation.


Yield, Mg ha
64
0 8.8 17.6 26.4
P, kg ha'1
Figure 3-4. Rate of inorganic P applied through triple
superphosphate and its effect on maize yield in a
maize-maize rotation.


101
production with the PCW20+TSP20 treatment could be due to
the immobilization of the applied inorganic P in close
proximity to the PCW. Low P content, high loading rate, high
metal concentration, high C:N, and C:P ratios were
apperently responsible for low P release, and high P
fixation. The performance of CM20+TSP20 was better than
other tretaments in all environments (Fig. 4-4a). The
production level of 1.85 Mg ha'1 was identified as a minimum
acceptable level. Based on the minimum acceptable production
level of 1.85 Mg ha'1 all environments were divided into two
RDs. Within both RDs the probability of yield falling below
1.85 Mg ha1 for each treatment was examined with the aid of
a confidence interval (Cl) test (Fig. 4-4b). Thus, even with
CM20+TSP20 there is 17% chance of yield less than 1.85 Mg
ha"1 in SF2 and WL (RD1) and well over 50% with TSP4 0. Using
these two treatments in PF1, 2 and SF1 (RD2), there is
virtually no probability of yield less than 1.85 Mg ha1.
On the other hand FP only reaches 1.85 Mg ha'1 in PF1 which
falls in RD2.


SUSTAINING CROP PHOSPHORUS NUTRITION OF HIGHLY LEACHED
OXISOLS OF THE AMAZON BASIN OF BRAZIL THROUGH USE OF
ORGANIC AMENDMENTS
By
BRAJ K. SINGH
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
1990
tJUMWERSlTY OF FLORIDA LIBRARIES

This work is dedicated to my parents: Shree Mrit B.
Singh and Shrimati Naina Devi, and to my brothers: Sri
Biswanath Singh and Nawal K. Singh.

ACKNOWLE DGMENTS
The author wishes to express his sincere appreciation
to Dr. J. B. Sartain, the chairman of the supervisory
committee, for his excellent guidance, assistance, and
continued interest in this study. His patience,
understanding, and personal friendship throughout the
academic and research program has been most invaluable and
made the author's stay in the United States a most rewarding
and stimulating experience. The financial support provided
by him to cover part of research expenses in Brazil is
gratefully acknowledged.
Sincere appreciation is extended to Dr. P.E. Hildebrand
for providing a farming systems assistantship and taking
personal initiatives to work out a joint research venture
with the TropSoils Collaborative Research Support Program
between Cornell University and the Brazilian Agricultural
Research Organization (EMBRAPA). His deep concerns,
constructive criticisms, and moral support as the member of
supervisory committee helped broaden the author's
understanding of farming systems and contributed to the
successful completion of this manuscript. The author also
thanks the other members of the supervisory committee, Dr.
iii

E.A. Hanlon, Dr. D.A. Graetz, and Dr. K.L. Buhr for their
suggestions, support, and editorial comments. A special note
of gratitude is extended to Dr. C.K. Hiebsch for
substituting Dr. K.L. Buhr as a member of supervisory
committee and reviewing the manuscript.
A particular note of appreciation is extended to Dr.
Hugh Popenoe, Director of International Programs, for
providing primary funding for the author's Ph. D. program.
The author is indebted to and gratefully acknowledges the
financial and technical support provided by TropSoils and
EMBRAPA. Particular gratitude is expressed to Dr. Walter
Bowen for his invaluable guidance in design and execution of
the research and manuscript review. Special thanks go to Mr.
Manoel da Silva Cravo and his extraordinary group of
enthusiastic technicians, Raimundo Vitoriano de Oliveira,
Agilau de Araujo Rodrigues, Edilza da Silva Richa, Emanoel
dos Santos Alencar, Teofanes Moreira de Souza Junior, and
Onelia Maria Pereira de Almeida for conducting the routine
analysis of plant and soil samples. The author is grateful
to Mrs. Eda M. Souza, CEPA-AM for her support in putting the
on-farm project together.
A word of gratitude is also extended to Dr. K.D. Sayre,
Dr. W. H. Freeman, Dr. C.N. Hittle, Dr. Richard Harwood, and
Dr. Douglas Beck for their help in getting the author into
the graduate school. Thanks go to S.K. Patel for his help in
iv

representing the author at UF during the author's one year
stay in Brazil.
The author is indebted to his beloved wife for her
encouragement, support, and patience.
v

TABLE OF CONTENTS
ACKNOWLEDGEMENTS iii
LIST OF TABLES vii
LIST OF FIGURES X
ABSTRACT xiii
CHAPTER I INTRODUCTION 1
Statement of Problem 1
Goals and Objectives 4
CHAPTER II ORGANIC AMENDMENTS AND PHOSPHORUS
ADSORPTION ISOTHERMS 5
Soil Constituents and Phosphorus Adsorption ... 5
Materials and Methods 10
Results and Discussion 14
Conclusion 35
CHAPTER III ORGANIC AMENDMENTS AND CROP PHOSPHORUS
NUTRITION 37
Phosphorus Management Strategy 37
Materials and Methods 45
Results and Discussion 52
Conclusion 71
CHAPTER IV RECOMMENDATION DOMAINS AND MODIFIED
STABILITY ANALYSIS 72
Introduction 72
Materials and Methods 79
Results and Discussion 86
Conclusion 12 6
CHAPTER V SUMMARY AND CONCLUSIONS 129
APPENDIX ECONOMIC ANALYSIS 134
REFERENCE LIST 138
BIOGRAPHICAL SKETCH 150
vi

LIST OF TABLES
Table 2-1. Ionic species and concentration of nutrient
solution used with silica to simulate solution
chemistry of a Xanthic Hapludox in a phosphorus
release study by organic amendments 13
Table 2-2: Chemical composition of organic amendments
used in the incubation study 16
Table 2-3: Selected physical and chemical properties of
Ap horizon of the Xanthic Hapludox used in
adsorption studies 17
Table 2-4. Release of 0.01 M CaCl2 extractable
P (M9 g following incubation of organic
amendments with soil and silica
as matrix substratum 23
Table 2-5. Langmuir parameters (k and b) for P
adsorption by soil incubated with different
organic amendments 29
Table 2-6. Langmuir parameters (k and b) based on net
release of P measured by sequential extraction of
simulated silica matrix with 0.01 M CaCl2. ... 32
Table 2-7. Difference in the estimated values of
Langmuir parameters (b and k) based on the
estimation of preadsorbed P by simulated silica
matrix technique and sequential extraction. ... 33
Table 3-1. Factorial arrangements of treatments for the
greenhouse study 48
Table 3-2. Description of treatments tested in the
field 50
Table 3-3. Selected chemical properties of the fine
fraction (<2 mm) of the Xanthic Hapludox 55
Table 3-4. Selected physical properties of the Xanthic
Hapludox 55
vii

Table 3-5. Selected chemical properties of the fine
fraction (<2 mm) of the Xanthic Hapludox 56
Table 3-6. Analysis of variance for maize herbage
dry weight production per pot in
the glasshouse study 57
Table 3-7. Mean changes in soil pH (H20) following
application of selected organic amendments in the
glasshouse study 60
Table 3-8. Orthogonal contrasts of maize grain yield
under different treatments applied in a 30 cm wide
band at UEPAE research station, Amazonas, Brazil. 62
Table 4-1. Application of N, P, and K in different
treatments tested in on-farm experimentation for maize
and cowpea crops in the municipality of Rio Preto da
Eva, Amazonas, Brazil 82
Table 4-2. Characterization of experimental plots for
maize testings in the municipality of Rio Preto da
Eva, Amazonas, Brazil 89
Table 4-3. Characteristics of experimental plots for
cowpea trials 90
Table 4-4. Relationship between soil characteristics
with year in crop production in different land
types 91
Table 4-5. Relationship between soil characteristics
measured after treatment application with grain
yield for maize and cowpea crops in the
municipality of Rio Preto da Eva, Amazonas,
Brazil 92
Table 4-6. Summary of ANOVA for multilocatonal maize
testing in the municipality of Rio preto da Eva,
Amazonas, Brazil 97
Table 4-7. Duncan Multiple Range Test (DMRT) for maize
crop in the municipal of Rio preto da Eva,
Amazonas, Brazil 97
Table 4-8. Combined Analysis of Variance for maize
grain yield in the municipal of Rio Preto da Eva,
Amazonas, Brazil 98
Table 4-9. Environmental index (e) for maize production in
the Municipality of Rio Preto da Eva, Amazonas,
Brazil 100
viii

Table 4-10. Summary of ANOVA for multilocatonal cowpea
testing 110
Table 4-11. Duncan Multiple Range Test (DMRT) for
cowpea trials conducted in the municipal of Rio
Preto da Eva, Amazonas, Brazil 110
Table 4-12. Combined Analysis of Variance for
cowpea experiments. Ill
Table 4-13. Technology selection for a given land type
based on different evaluation criteria for maize
cultivation 124
Table 4-14. Technology selection for a given land type
based on different evaluation criteria for cowpea
cultivation 124
ix

LIST OF FIGURES
Figure 2-1. Acid-base potentiometric titration curves
for the Ap horizon of Xanthic Hapludox with
varying concentration of CaCl2. 18
Figure 2-2. Phosphorus adsorption isotherms following
35 d (a) and 150 d (b) of soil incubation with
Mucuna aterrima (Mucuna), Chicken Manure (CM),
Aerobically Processed City Waste (PCW), and
Control. 21
Figure 2-3. Phosphorus desorption isotherms following
35 d (a), and 150 d (b) of soil incubation with
Mucuna aterrima (Mucuna), Chicken Manure (CM),
Aerobically Processed City Waste (PCW), and
Control 22
Figure 2-4. A schematic representation of molecular
structures of oxides of phosphorus (a), and oxides
of silicon (b) 26
Figure 2-5. Effect of incubation period and incubation
matrix on P release pattern from different organic
amendments in a laboratory study 27
Figure 2-6. Phosphorus adsorption isotherms for the
Xanthic Hapludox (0.01 M CaCl2) fitted to Langmuir
equation. The lines in the figure represents
fitted equation. 28
Figure 2-7. A comparison of P adsorption by Xanthic
Hapludox based on the estimation of preadsorbed P
by sequential extraction of soil or silica matrix
incubated with organic amendments 34
Figure 3-1. Geographic location of Amazon basin in
Brazil (a), on-station and farming systems
research (FSR) sites (b), and effective rainfall
during the period of August, 1988 until August
1989 (c) at EMBRAPA station in Manaus, Brazil.. 47
x

Figure 3-2. Effect of different rates of selected
organic amendments in combination with different
rates of inorganic P on maize dry matter
production at 65 d after planting in a glasshouse
study 58
Figure 3-3. Relationship between Mehlich-I extractable
soil P (a), and maize leaf tissue
P concentration (b) with maize dry matter yield. 59
Figure 3-4. Rate of inorganic P applied through triple
superphosphate and its effect on maize yield in a
maize-maize rotation 64
Figure 3-5. Effect of selected organic amendments,
applied in a quantity equivalent to provide 26.4
kg ha'1 of P, on sustaining Mehlich-I extractable
soil P pool in a maize-maize rotation on a Xanthic
Hapludox 65
Figure 3-6. Effect of selected organic amendments,
applied in combination with inorganic phosphorus
source in a quantity equivalent to provide 8.8 kg
ha1 of P, on sustaining Mehlich-I extractable
soil P pool in a maize-maize rotation on a Xanthic
Hapludox 67
Figure 3-7. Leaching of calcium from surface horizon of
highly leached Xanthic Hapludox as influenced by
different organic amendments 69
Figure 3-8. Relationship between P concentration and
concentration of Zn and Cu in maize leaf tissue.. 70
Figure 4-1. Effective rainfall and suggested shift in
the plnting date for maize in Rio Preto da Eva,
Amazonas, Brazil, (a) common practice, (b)
suggested practice 83
Figure 4-2. Range of different soil characteristics by
recommendation domains for maize trials 94
Figure 4-3. Range of different soil characteristics by
recommendation domains for cowpea trials 95
Figure 4-4a. Response of different treatments to
environmental index for maize production, Rio Preto da
Eva, Amazonas, Brazil 102
Figure 4-4b. Distribution of confidence intervals for
maize production in poor (e<1.95 mg ha1), and
good (e>1.95 mg ha1) environments 103
xi

106
Figure 4-5. Relationship of net income/cash cost, net
income/total cost, and net income with
environmental index, in on-farm maize trials from
Rio Preto da Eva, Amazonas, Brazil
Figure 4-6. Distribution of confidence intervals for
net income/cash cost, net income/total cost, and
net income for different treatments used for maize
cultivation 108
Figure 4-7a. Response of different treatments to
environmental index for cowpea production, Rio
Preto da Eva, Amazonas, Brazil 115
Figure 4-7b. Distribution of confidence intervals for
cowpea production in poor (e<1.32 mg ha'1), and
good (e>1.32 Mg ha'1) environments 116
Figure 4-8. Relationship of net income, net income/cash
cost and net income/total cost with environmental
index, in on-farm cowpea trials from Rio Preto da
Eva, Amazonas, Brazil 117
Figure 4-9. Distribution of confidence intervals for
net income/cash cost for selected treatments in
poor (e<1.32 Mg ha'1) and good (e>1.32 Mg ha'1)
environments for cowpea cultivation 119
Figure 4-10. Distribution of confidence intervals for
net income/total cost for selected treatments in
poor (e<1.32 Mg ha1) and good (e>1.32 Mg ha'1)
environments for cowpea cultivation 120
Figure 4-11. Distribution of confidence intervals for
net income for selected treatments in poor (e<1.32
Mg ha'1) and good (e>1.32 Mg ha'1) environments for
cowpea cultivation 121
Figure 4-12. Recommendation domains, based on yield,
and their relationship to land types in the
municipality of Rio Preto da Eva,Amazonas,
Brazil 125
Figure 4-13. Recommendation domains, based on net
return to cash cost, and their relationship
to land types in the municipality of Rio Preto
da Eva, Amazonas, Brazil 12 6
xii

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
SUSTAINING CROP PHOSPHORUS NUTRITION OF HIGHLY LEACHED
OXISOLS OF THE AMAZON BASIN OF BRAZIL THROUGH USE OF
ORGANIC AMENDMENTS
By
Braj K. Singh
August 1990
Chairperson: Dr. Jerry B. Sartain
Major Department: Soil Science
Retention of phosphorus by iron and aluminum oxides
plays an important role in determining ultimate availability
of P to plants in highly weathered soils of the tropics.
Different management strategies have been proposed to
overcome this problem. The overall objective of this
research was to study the influence of organic amendments
(plant origin, manure, and aerobically processed city waste
(PCW)) in sustaining P nutrition of Oxisol and to test the
performances of selected alternatives in a wide range of
production environments.
There was a marked difference in P adsorption and
desorption by soil preincubated with organic amendments. The
difference was attributed to incubation period and chemical
composition of amendments. Correction for preadsorbed P by
xiii

soil based on simulated silica matrix (SSMT) resulted in a
better estimation of the Langmuir adsorption maxima. This
technique involved mixing of acid washed sand with a
nutrient solution (without P) and inoculation of the sand
with microbes and incubation with organic amendments.
In the glasshouse experiment the highest dry matter
(DM) production was obtained with the CM. A strong
relationship was observed between DM production and Mehlich-
I extractable P (r2 = 0.88).
In the on-station field experiments, CM applied plots
produced more grain in first as well as second crops
compared to the plots which had received Canavalia
ensiformis. Mucuna aterrima, PCW, and TSP. All organic
amendments improved the soil P reserve and reduced Ca
leaching indicating that application of organic amendments
could lead to sustained crop P nutrition.
In farmers' field studies, three amendments were tested
in different land types with maize and cowpea crops. The
same amount of P applied from different amendments had
different effect on maize and cowpea production. However,
the selection of a given technology for a given land type
(environment) was dependent on farmers' goals. Based on the
criteria of grain production CM was recommended for maize in
all environments and for cowpea in poor environments.
xiv

CHAPTER I
INTRODUCTION
Statement of Problem
A leaching environment together with persistent high
rainfall and temperature are the defining conditions for the
development of Oxisols, a soil order which occupies over
1.12 billion ha and ranks 5th in worldwide distribution
(USDA, 1975). Their similarity as a group stems from the
composition of the colloid fraction rich in iron and alumina
minerals, and deficient in other nutrients. Sustained
production on these soils can be achieved only with adequate
application of lime, N, P, K, Mg, Zn, Cu, B, and Mo (Sanchez
et al., 1982). Among other nutrients, P deficiency appears
to be the most crucial and may show up as early as the
second year of cultivation. Large amounts of applied P are
required to attain levels of soil solution P which are
adequate for high crop yields (Yost et al., 1979).
Rapidly increasing cost of P fertilizer, and limited P
supply in the tropics, have led to extensive P management
studies on these soils. Fertilizer application techniques
(band, broadcast, and band+broadcast) have been widely
1

2
investigated. Immediate and long term effects of such
applications have been studied by Kamprath (1967), Yost
(1977), and Yost et al. (1981). These studies have suggested
that a high initial P application rates can reduce the P
fixation capacity (Barrow, 1974), increase the cation
exchange capacity (Keng and Uehara, 1974), and increase soil
pH.
Advanced Research with Organic Amendments
Along with fertilizer application technigues, attempts
are underway to manipulate the mineral surface chemistry and
inactivate the high reaction capacity of sesguioxides by
application of organic amendments (Larsen et al., 1959;
Nagarajah et al., 1970; Yuan, 1980). It is believed that
*-- ^orm a coating on the mineral surface
1990) and complex Fe3+ and Al3+
(Larsen et al., 1959) which will reduce P fixation. However,
the role of an organic amendment as a P source has not
received much attention. Attempts have been made to
understand the factors responsible for P mineralization from
organic amendments. Amendment P content (Singh, and Jones,
1977) and decomposition characteristics (Sweeney and Graetz,
1988) are important factors in understanding the P release
patterns and predicting residual effects.
Use of P adsorption isotherms in predicting P
requirements of crops and soils has received wide attention

3
(Jones and Benson, 1975; Solis and Torrent, 1989). However,
the techniques used to construct P adsorption isotherms in
the presence of amendments are similar to those used for
soil without amendment. Generally a correction for
preadsorbed P is introduced (Mokwunye, 1977; Solis and
Torrent, 1989). An assumption is made that adsorption and
desorption are equal, which is not true (Sample et al.,
1980). Therefore, it leads to unreliable prediction of crop
P requirement and residual effect of applied P.
Organic and Inorganic P Interaction
A larqe amount of orqanic amendment will be required to
provide the total P requirement for a crop. An efficient way
to use a suboptimal dose of an organic amendment is band
application with inorganic P. Such an application reduces P
immobilization from P-poor amendments, and eliminates P
deficiency in the early stages of crop growth. Much remains
to be learned about the interaction of organic amendments
with inorganic P. Most laboratory work has not been
adequately tested under field conditions. Limited studies
conducted in this area indicate that diammonium phosphate
and monoammonium phosphate can solubilize organic matter
(Bell and Black, 1970; Giordano et al., 1971). As the
solubilized organic matter is carried to a new location in
the soil, it may reprecipitate, covering soil mineral
surfaces which otherwise could have participated in P
retention reactions.

4
Transfer of Technology
It is widely believed that future efforts to increase
food production must be directed towards the marginal lands
of the developing world (Shaner et al., 1982). To achieve
this goal, more research on appropriate technology is
required with direct farmer involvement in problem
identification, research priority identification, and
technology evaluation (Harwood, 1979). Such an attempt will
lead to location-specific technology tailored to fit
farmers' circumstances, and accelerate the process of
technology diffusion (Hildebrand, 1983).
Goals and Objectives
The overall goal of this research was to examine the
role of organic amendments in sustaining P nutrition of a
highly leached Oxisol. The specific objectives were: (i) to
devise a technique to measure the P release patterns from a
decomposing organic material, (ii) to examine the effect of
a combined application of organic and inorganic P in a band
on sustaining crop P nutrition, and (iii) to conduct
farmers' field trials to validate on-station research
results and delineate recommendation domains for selected
treatments.

CHAPTER II
ORGANIC AMENDMENTS AND PHOSPHORUS ADSORPTION
ISOTHERMS
Soil Constituents and Phosphorus Adsorption
It is widely believed that hydrous oxides of Fe and Al,
and calcium carbonate play key roles in P retention.
Although controversy exists among researchers about the
mechanism of P retention by these compounds (Muljadi et al.,
1966; Hingston et al., 1967; Rajan et al., 1974), the phasic
nature of P adsorption has been well recognized by several
researchers (Bache, 1964; Munns and Fox, 1976). The first
phase of adsorption is due to a high energy chemisorption of
small amounts of P on the soil surface. The second phase is
comprised of a precipitation reaction followed by a low
- >
energy sorption of P onto the precipitate.
Phosphorus Adsorption Isotherms
Phosphorus adsorption isotherms have been used
extensively for describing the P adsorption characteristics
of various soils and in estimating the P requirement of
different crops (Fox and Kamprath, 1970; Jones and Benson,
1975; Mokwunye, 1977; Singh and Jones, 1977; Solis and
Torrent, 1989). The isotherm technique involves
5

6
equilibration of a known amount of soil for a limited time
in a KC1, NaCl or CaCl2 solution containing various amounts
of K2HP04, NaH2P04 or Ca(H2P04)2 (Olsen and Watanabe, 1957;
Syers et al., 1973; Singh and Jones, 1976). Phosphorus
removed from solution is considered to be adsorbed. Plant P
requirement is calculated based on the amount of P needed to
bring the concentration of supernatant solution to a
specified concentration (generally 0.2 nq mL"1) .
Langmuir equation. A number of researchers have
investigated soil P adsorption characteristics using the
Langmuir equation (Olsen and Watanabe, 1957; Woodruff and
Kamprath, 1965; Gunary, 1970; Borggaard, 1983). A frequently
used linear form is:
(c/x/m) = (1/kb) + (c/b)
Where:
x/m = amount of P adsorbed per unit weight of soil
b = the adsorption maxima
c = the equilibrium P concentration, nq mL"1
k = a constant related to the bonding energy of the
adsorbent for adsorbate.
For a given uniform population of sites, the value of
adsorption maxima can also be evaluated by plotting x/m vs.
x/m/c, obtained from a rearranged Langmuir equation commonly
referred to as the 1Eadie-Hofstee plot (Syres et al.,
1973) The underlying assumptions in each case are that
adsorption sites on the particle surface are uniform, and

7
the maximum adsorption possible corresponds to a complete
monomolecular layer. Both of these postulates do not hold
for a heterogenous medium like soil (Larsen, 1967).
Deviations from the conventional Langmuir relationship
at high equilibrium P concentrations (above 15 nq mL 1) have
already been reported (Olsen and Watanabe, 1957; Hsu and
Rennie, 1962) which have led to the development of an
extended form of the Langmuir equation. Gunary (1970)
included a square-root term in the Langmuir equation.
Holford et al. (1974), Syers et al. (1973), and Holford
(1983) used a two-surface Langmuir equation. The two surface
equation has not, however, been universally accepted and
Posner and Bowden (1980) have discussed the futility of
attempting to split isotherms into two or more regions.
Similarly, a curve-fit error in estimating the Langmuir
adsorption maxima was described by Harter (1984). He claimed
the test of linearity was inadequate because plotting
concentration against itself reduces data variability and
always provides a significant correlation coefficient. He
suggested that a better test of the fit is to ascertain
whether the adsorption isotherm has the shape of the
equation model.
In spite of drawbacks, the Langmuir equation is used
widely to describe P adsorption by soil. The major advantage
of the Langmuir equation is that it allows for the
calculation of adsorption maxima along with a relative

8
bonding energy term for P sorption (Syere et al., 1973). The
values of adsorption maxima and bonding energy can be
related to various soil properties which will supply
information about the nature of the reaction between soil
and fertilizer P, and can aid in the prediction of plant
available P (Olsen, 1953).
Correction for Initial Surface Phosphate
Olsen and Watanabe (1957) have demonstrated the effect
of correction for initial surface phosphate on the constants
derived from the Langmuir isotherm. For Pierry clay and
Owyhee silt loam soils the correction increased the
adsorption maxima and the bonding energy constant. Ideally,
the adsorption would be determined in a system in which the
surface is free of the adsorbate ion. Usually this
restriction is not feasible. Therefore, a correction for
initial P is made by adding the amount of surface P
determined by a separate analysis. One of the commonly used
methods to measure native adsorbed phosphate is by isotopic
exchange (Olsen and Watanabe, 1957; Holford et al., 1974).
This method involves shaking a soil sample in an electrolyte
to which carrier-free 32P is added. An aliquot is taken
after a specified time and 31P is determined. Reddy (1990)
has suggested a least squares fit for the determination of
native P adsorbed on the soil surface. He suggested the
following calculations:

9
s1 = k'c Sp
Where:
S1 = amount of added P sorbed, nq g'1
Sp = Y axis intercept-representing native soil P in the
adsorbed phase, nq g1
C = P in solution, nq mL1
k' = linear adsorption coefficient (estimated without
considering native adsorbed P, Sp) mL g1.
Syers et al. (1973) reported that at lower
concentrations the Eadie-Hofstee plot was more useful for
evaluating P sorption than the conventional Langmuir
equation, because the plot expanded the low P concentration
region.
The methods described above are based on the assumption
that all surface-retained P is displacable either by
isotopic dilution or sequential extraction techniques in the
presence of a weak extracting salt solution. However,
evidence suggested that with increasing contact period and
temperature, P becames less ready to exchange with
isotopically labeled P (Barrow and Shaw, 1975). In addition,
adsorption and desorption isotherms are different, and the
relationship between the quantity of adsorbed P and
concentration is not totally reversible (Sample et al.
1980). Therefore, estimation of surface held native P based
on the techniques described above might underestimate the
true adsorption maxima and bonding energy values. Thus, any

10
prediction of how the soil P status will change upon
cropping must be considered unreliable (Bowman and Olsen,
1985).
The purpose of this study was to examine the effect of
different organic amendments on P adsorption and desorption
isotherms, and to devise a technique to measure net P
release from organic amendments during their decomposition.
Such a technique should lead to a direct and accurate
estimation of the quantity of P which originated from
decomposing organic amendments and adsorbed by soil. This
knowledge will aid in the accurate estimation of soil P
adsorption maxima and bonding energy.
Materials and Methods
All amendments mucuna ([Mucuna aterrima (Piper and
Tracy) Merr], kudzu [Pueraria phaseoloides (Roxb.) Benth],
canavalia [Canavalia ensiformis (L.) DC], maize (Zea mays
L.), peanut (Arachis hvpoqaea L.), tephrosia fTephrosia
candida (Roxb.) DC], cowpea (Vigna unquiculatal. mixed
gramineae (Grass), aerobically processed city waste (PCW),
and chicken manure (CM)) were dried at 65 C for 72 hrs for
dry matter determination. Dried plant material was ground in
a Wiley mill and passed through a 0.5 mm screen. A weighed
(0.2 g) sample of the ground plant material was digested
with H2S04 and H202 for the determination of macro and micro
elements. Total nitrogen was determined by the micro-

11
Kjeldahl procedure. Potassium, Ca, Mg, Fe, Zn, Cu, and Mn
were determined with an atomic adsorption spectrophotometer,
and P was determined colorimetrically by the Murphy and
Riley (1962) procedure.
Incubation study
Soil matrix. Soil used for this study was obtained from
the Ap horizon of a Xanthic Hapludox (clayey, kaolinitic,
isohyperthermic) (EMBRAPA, 1979) which had PZNC (point of
zero net charge) at pH of 4.2. Five grams of organic ground
material were mixed with 95 g of air-dried soil and
incubated in plastic containers for 35, 65, and 150 d at 30
3C. The soil in each container was mixed thoroughly every
6 d and was kept at 45 % moisture content. After each
incubation period, 2-g duplicate soil samples were extracted
with 0.01 M CaCl2 with a soil to solution ratio of 1:10 and
1 h shaking. The cycle was repeated three times. In
calculating P desorbed, an allowance was made for the 2 mL
of supernatant carried over from each cycle (Fox and
Kamprath, 1970).
Adsorption was determined using 2-g air-dried samples,
in duplicate, eguilibrated for 6 d with 20 mL of 0-60 nq
mL 1 P solution in 0.01 M CaCl2 in polyethylene tubes at 25
2C. To inhibit microbial activities a few drops of toluene
were added to each tube. After 6 d the solution P was
determined by the ascorbic acid method (Murphy and Riley,

12
1962). Phosphorus removed from solution was considered
adsorbed.
Desorption was studied by equilibrating the soil from
the adsorption study in 20 mL 0.01 M CaCl2 for 6 hrs (Singh
and Jones, 1977). Increases in solution P were measured and
considered desorbed P. Adsorption and desorption isotherms
were constructed and a standard Langmuir equation was fitted
to calculate adsorption maxima and bonding energy (Olsen and
Watanabe, 1957).
Simulated silica matrix (SSMT). Hydrochloric acid
(0.1 M) washed, fine silica was mixed with a nutrient
solution containing N, K, Ca, Mg, Zn, Mo, Mn, Fe, Cu, B, S,
and Cl (P excluded) (EMBRAPA, 1976) to simulate the nutrient
requirement for maize in a solution culture. Solution pH was
adjusted to 5.0 with 0.1 M HC1. Table 2-1 highlights the
chemical compounds used to make the nutrient solution and
the resulting concentrations. The inoculation of silica
matrix was carried out with microbes grown on potato
dextrose agar.
Five grams of ground organic amendments (the same ones
used in incubation study with soil) were mixed with 95 g of
silica matrix and incubated for 35, 65, and 150 d at 30
3C. The matrix was kept moist and mixed thoroughly every 6
d. After each incubation period, duplicate 2-g samples were
extracted with 0.01 M CaCl2 following the same procedure as
described for soil. The sum of P detected by sequential

13
Table 2-1. Ionic species and concentration of nutrient
solution used with silica to simulate solution
chemistry of a Xanthic Hapludox in a phosphorus release
study by organic amendments.
Source Species Conc^.
No. Chemical mmol (charge) L
1.
Ca(N03)2.4H20
Ca2+
2.690
nh4no3
nh4+, no3
9.180
2.
KC1
K++
1.829
k2so4
Cl'
0.670
kno3
S
0.480
3 .
Mg(N03)2.6H20
Mg2+
0.650
4 .
FeHEDTA
Fe3+
0.035
5.
MnCl2.4H20
Mn2+
0.007
h3bo3
B
0.019
ZnS04.7H20
Zn
0.0018
CuS04.5H20
Cu
0.0005
Na2Mo04.2H20
Mo
0.0006
+ PH was adjusted to 5.0 with 0.1M HC1. The reported
concentration is considered adequate for maize growth in a
solution culture.
4= The final concentration for K is calculated from KC1,
K2S04, and KN03.

14
extraction with CaCl2 was considered total P released from
the amendments because of the inability of silica to adsorb
P. This value was used to construct a P release curve for
each amendment.
Calculation of total P adsorbed. Total adsorbed P was
calculated as follows:
TAP = N + S + OA
p p p
where:
TAP = total adsorbed P on soil surface, nq g1
Np = native P present in adsorbed phase nq g1
Sp = applied inorganic P in adsorbed phase /ng g"1
0Ap = phosphorus from organic amendment in adsorbed phase
Mg g'1.
The amount of P released from the organic amendments and
adsorbed on the soil surfaces (0Ap) was estimated by SSMT.
The native exchangeable P on control samples was determined
by the method of least squares (Reddy, 1990). For comparison
purposes, P adsorbed on the soil surface following
incubation with organic amendments was also determined by
sequential extraction with 0.01 M CaCl2.

15
Results and Discussion
Organic amendments differed in chemical composition
(Table 2-2). The highest P concentration was observed in CM
(2.77 g 100 g'1) and the lowest in PCW (0.16 g 100 g1). The
highest C:P ratio was for maize (189.0) followed by grass
(180.0).
Soil used was obtained from the Ap horizon of a highly
leached oxisol classified as Xanthic Hapludox (clayey
kaolinitic, isohyperthermic) (EMBRAPA, 1979). The soil was
acidic (pH = 4.5), and low in total P (170 /g g'1 soil) and
Mehlich-I extractable P (3.3 iiq g'1 soil). Clay content
ranged from 78-82% and the A1 saturation was 78.4% (Table 2-
3). The point of zero net charge measured by potentiometric
titration in CaCl2 solution was at pH 4.2 (Fig. 2-1). The
native exchangeable P measured by the least sguares method
was 3.0 nq g"1 soil.

Table 2-2. Chemical composition of organic amendments used in the incubation
study.
0A+ Ca Mg K C P N C: P C:N Cu Zn Mn
g lOOg'1 M9 g 1
Canav.
2.27
0.17
1.90
39.6
0.31
3.95
128
10
2.5
8.4
174.0
Peanut
1.44
0.27
1.95
38.5
0.33
3.32
116
11
2.5
8.9
70.0
Maize
0.35
0.19
2.25
39.6
0.21
2.93
189
13
5.0
40.1
95.0
Teph.
1.01
0.19
1.75
43.9
0.29
3.31
151
13
2.5
26.1
41.0
Mucuna
1.05
0.15
1.80
42.5
0.28
3.34
152
12
2.5
28.0
103.0
Grass
0.70
0.25
1.60
34.1
0.19
1.43
179
23
5.0
36.9
123.0
CM
7.62
0.51
2.10
17.8
2.77
1.46
6
12
92.5
102.9
399.0
PCW
1.57
0.14
0.35
22.3
0.16
0.95
139
23
155.0
280.8
213.0
Puer.
0.45
0.18
1.55
42.1
0.28
2.52
150
16
2.5
24.3
212.0
Cowpea
1.22
0.16
2.20
41.9
0.31
2.75
134
15
2.5
19.2
98.0
-f- Organic amendments; Mucuna aterrima (Mucuna) Pueraria phaseoloides (Puer.) ,
Canavalia ensiformis (Canav.), Zea mays (Maize), Arachis hvpoqaea (Peanut), Tephrosia
candida (Teph), Viqna unquiculata (Cowpea), mixed gramineae (Grass), aerobically
digested city waste (PCW) and chicken manure (CM).

17
Table 2-3: Selected physical and chemical properties of Ap
horizon of the Xanthic Hapludox used in adsorption
studies.
Parameters^
Value
Units of
Measured
Measurement
Bulk density
1.25
g cm"3
g kg"1
Clay
815
pH H20, KCl
4.5, 4.2
->. kg "1
g
Net Charge
0.22
cmol (
Oxides Fe, A1
0.075, 0.325
g 100
A1 Sat. (ECEC)
78.4
%
Mg g'1
Total P
170.0
soil
g"1 soil
Mehlich-I P
3.3
Mg.
Mg g
Al, Fe (P)
29.0, 1.4
soil

18
pH
Figure 2-1. Acid-base potentioinetric titration curves for
the Ap horizon of Xanthic Hapludox with varying
concentration of CaCl2.

19
Adsorption Isotherms
The results of adsorption and desorption studies are
presented in Fig. 2-2 and 2-3. Incubation of the soil with
amendments for 35 d influenced P adsorption. This was more
pronounced at higher equilibrium solution P concentrations
(Fig. 2-2a) At a P concentrations >1.9 /ug mL'1, PCW amended
soil adsorbed more P than the unamended soil. A similar
result was obtained by Singh and Jones (1976) for organic
amendments with low P content and high C:P ratio. On the
other hand, soil samples amended with CM and Mucuna adsorbed
less P than the unamended soil. As the time of incubation
increased to 150 d, all amendments reduced P adsorption
compared to the control (Fig. 2-2b). A five-fold reduction
in P adsorption was observed for the CM treatment at an
equilibrium P concentration of 0.3 jug g1 soil. The
reduction was perhaps due to net mineralization of P even
from P-poor organic amendments with time. Another factor in
P mineralization is the C:P ratio. When this ratio remains
less than about 200:1, immobilization predominates during
the initial stages of decomposition. But, as the
decomposition proceeds, this ratio becomes narrower due to
the concentration of P in decomposing residue and continuous
degradation of C by microorganisms.
As plants deplete soil solution P, the solution must be
continuously recharged, if good growth is to be maintained.
Recharge occurs when P is desorbed from the soil surface,

20
and this will happen in sufficient quantity only if the soil
has a large capacity to sorb and therefore desorb P. Thus,
even though the soils with high sesquioxides content require
more P to achieve a given level of P in soil solution, they
have the compensating value of being able to supply P to the
soil solution as it is taken up by plants. The results of
desorption studies are presented in Fig. 2-3.
Soil incubated with Mucuna aterrima (Mucuna), PCW, and
the control for 35 d did not show any difference in P
desorption. However, desorption was higher for all
amendments incubated with soil for 150 d (Fig. 2-3b). In
fact, CM reduced adsorption and increased desorption at all
incubation periods. The observed differences of organic
amendments in influencing P desorption characteristics of
the soil could be attributed to their decompositional
characteristics, P content, C:P and C:N ratios, and
concentration of such elements as Zn and Cu. Incubation of
soil with low P containing organic materials may not
initially influence soil P desorption characteristics.
Phosphorus release from organic amendments
Phosphorus release data for different amendments in
soil and silica substratum are presented in Table 2-4. More
P was detected in the silica matrix for all amendments
compared to the soil matrix. With increasing incubation
time, a decrease in P was observed for maize and grass

P adsorbed (ug/g soil)
Adsorption
21
Figure 2-2. Phosphorus adsorption isotherms following 35 d
(a) and 150 d (b) of soil incubation with Mucuna aterrima
(Mucuna), Chicken Manure (CM), Aerobically Processed City
Waste (PCW), and Control.

P adsorbed (ug/g soil)
Desorption
22
Solution P (ug mL!)
Figure 2-3. Phosphorus desorption isotherms following 35 d
(a), and 150 d (b) of soil incubation with Mucuna
aterrima (Mucuna), Chicken Manure (CM), Aerobically
Processed City Waste (PCW), and Control.

23
Table 2-4. Release of 0.01 M CaCl2 extractable P (/xg g" )
following incubation of organic amendments with soil
and silica as matrix substratum.
Org.+ Soil Silica
amend.
35
65
150
35
65
150
CM
109a*
204.9a
245.1a
447a
1031a
1380a
Mucuna
21b
14.Obc
16.9b
80b
124b
160b
Tephrosia
22b
14.9b
19.0b
65b
116b
152b
Cowpea
15c
11.5bc
16.2bc
43c
70c
92c
Peanut
18bc
10.3cd
8.6bcd
41c
117b
142b
Grass
6e
6.7de
3.4d
37c
54c
102c
Maize
14cd
12.5bc
3.9d
37c
77c
111c
PCW
3e
2 86
10.8bcd
34c
62c
64d
Canavalia
14cd
11.8bc
15.3bc
31c
83c
148b
Kudzu
lOd
11.4bc
14.lbcd
23c
71c
94c
Control
4e
4.4e
5.2cd
Trace
Trace
0. li
CV%
6.3
14.0
8.2
12.3
7.5
12.0
* Means in the same column followed by the same letter are
not significantly different at the 95% level of
probability, as determined by Duncan's Multiple Range
Test.
f Mucuna aterrima (Mucuna), Pueraria phaseoloides (Kudzu),
Canavalia ensiformis (Canav.), Zea mays (Maize), Arachis
hvpogaea (Peanut), Tephrosia candida (Tephrosia), Viona
unguiculata (Cowpea), Mixed Gramineae (Grass), Aerobically
Processed City Waste (PCW), and Chicken Manure (CM).

24
treatments in the soil matrix. Processed city waste appeared
to immobilize soil P as indicated by the P measurement made
at 35 d of incubation but at 150 d a net release was
observed. All leguminous amendments followed similar P
release patterns. The highest P release was observed for CM
treatment at 150 d of incubation (245 /ug g1 soil) and the
lowest was for maize and grass treatments (3.9, and 3.4 ng
g'1 soil, respectively) .
When silica was used as the incubation matrix, there
was a net release of P from all amendments which indicated
that P released from amendments was being adsorbed by soil
(soil matrix) and was not all exchangeable in sequential
extraction with 0.01 M CaCl2. Maize and grass treatments
which had shown no net P release at 150 d in the soil matrix
released over 100 /g g'1 silica in the silica matrix. Given
the similarity in molecular structure and chemical behavior
of Si and P (Fig. 2-4) it is believed that added Si
increases water soluble and easily extractable P (Adams,
1980), and Si does not absorb P, because both of them in
ionic forms are negatively charged. It is also suggested
that silicate and phosphate ions compete for the same
adsorption sites on Al(Fe)-oxide surfaces (Mekaru and
Uehara, 1972) and they form insoluble precipitates with such
common ions as Al, Fe, and Ca. This suggests that all P
mineralized from organic amendments remained in the solution
and was easily extractable by 0.01 M CaCl2.

25
To understand the P release pattern from different
amendments, surface response curves were fitted and are
illustrated in Fig. 2-5. Phosphorus release in
silica for maize, PCW, and CM followed a logarithmic
function and the r2 for PCW, CM, and maize treatments were
0.73, 0.82, and 0.98. A trace amount of P (0.1 /Lxg g'1) was
detected in the control treatment in the silica matrix
confirming that the matrix was not contaminated and the P
detected came solely from the decomposition of amendments.
Langmuir Parameters
Soil Matrix. A standard Langmuir equations were fitted
to the data after correction was made for preadsorbed P
using sequential extraction techniques (Fig. 2-6). This was
followed by calculation of the adsorption maxima and bonding
energy for each treatment (Table 2-5). All but Pueraria
phaseoloides (Kudzu) increased the Langmuir adsorption
maxima at 35 d of incubation compared to the control. The
extent of variation among different amendments ranged from
536 to 818 nq g'1 soil at 35 d of incubation. It is
noteworthy that PCW had the highest adsorption maxima (818
jug g1) The bonding energy was reduced for all treatments
compared to the control. As the time of incubation increased
from 35 d to 150 d, the adsorption maxima decreased for all
amendments compared to the control treatment. But the

26
0 #P
@ 0
Figure 2-4. A schematic representation of molecular
structures of oxides of phosphorus (a), and oxides of
silicon (b).

1.-6 6n
27
Figure 2-5. Effect of incubation period and incubation
matrix on P release pattern from different organic
amendments in a laboratory study.

LU/X/O
28
Figure 2-6. Phosphorus adsorption isotherms for the Xanthic
Hapludox (0.01 M CaCl2) incubated with organic
amendments for 150 d. The lines in the figure
represents fitted Langmuir equation.

Table 2-5. Langmuir parameters (k and b) for P sorption by soil incubated
with different organic amendments.
Org.+
Amend.
35 Days
65 Days
Incubation Period
150 Days
r2
b+
k§
r2
b
k
r2
b
k
Canavalia
0.87**
638
0.384
0.81**
633
0.303
0.89**
528
0.264
Peanut
0.73**
632
0.233
0.87**
522
0.282
0.85**
509
0.229
Maize
0.94**
560
0.785
0.84**
670
0.308
0.86**
673
0.431
Tephrosia
0.69*
768
0.229
0.91**
477
0.291
0.91**
465
0.246
Mucuna
0.90**
536
0.207
0.64*
654
0.087
0.49ns
695
0.120
CM
0.66*
643
0.080
0.67*
618
0.079
0.76**
496
0.103
Kudzu
0.93**
471
0.417
0.94**
577
0.476
0.95**
535
0.599
PCW
0.67*
818
0.224
0.66**
815
0.271
0.94**
517
0.375
Cowpea
0.62*
729
0.146
0.89*
585
0.244
0.91**
543
0.473
Grass
0.74**
765
0.247
0.80*
707
0.331
0.84**
603
0.363
Control
0.94**
535
0.673
0.95**
544
0.941
0.98**
591
1.941
*, ** Significant at the 0.05 and 0.01 levels, respectively.
-f Mucuna aterrima (Mucuna), Pueraria phaseoloides (Puer.), Canavalia ensiformis
(Canav.), Zea mays (Maize), Arachis hypoqaea (Peanut), Tephrosia candida (Teph),
Vicrna uncruiculata (Cowpea) mixed gramineae (Grass) aerobically digested city waste
(PCW) and chicken manure (CM).
+ b, P adsorption maxima, nq P g'1
§ k, bonding energy, mL g P"1

30
bonding energy did not follow any defined pattern. For the
grass treatment (mixture of different grasses) bonding
energy value increased from 0.25 to 0.37 mL ng'1, for
Canavalia ensiformis (Canavalia) bonding energy dropped from
0.39 to 0.26 mL nq1, and for CM it remained at 0.10 mL g'1.
This finding does not agree with the data obtained by Hundal
et al. (1988) where all studied amendments reduced the
bonding energy as the incubation period increased from 20 d
to 40 d.
The range of P in organic amendments varied from 1.7 to
25.0 g kg'1 on a dry matter basis. There was a high
concentration of Zn and Cu in PCW, and Ca in CM. Being
polyvalent cations, they have a high affinity for P
adsorption through cation bridging (Holford and Mattingly,
1975; Haynes, 1989). The decomposition rate constants of the
organic amendments were different and so was the release of
different elements as the decomposition proceeded. These
factors also may have contributed to the observed
differences in P adsorption maxima and bonding energy.
Silica matrix. The value of the total adsorbed P was
adjusted based on the amount of P released from the organic
amendments in the silica matrix. The Langmuir equation was
fitted to the data, and the value of adsorption maxima and
bonding energy were recalculated for selected treatments.
The results are presented in Tables 2-6 and 2-7. The
recalculated adsorption maxima and bonding energy values

31
were higher than the values obtained based on the P release
from the amendments in the soil matrix. At 150 d for the CM
treatment, the adsorption maxima were higher by 1170 /j.g g 1
soil and the bonding energy was higher by 20 fold. The low
adsorption maxima (496 /g g1) and bonding energy (0.10 /Ltg
mL1) for soil matrix indicated that the P capacity factor
for CM applied plots was lower compared to plots which
received other amendments and, therefore, CM plots should
have the lowest residual P effect in field testings. But,
the result from the Silica Matrix technique indicated
otherwise. The high adsorption maxima for CM at 150 d of
incubation (1667 /g g1) and the high bonding energy (2.0 mL
/xg"1) are indications of sustained P desorption and residual
P availability.
Predicted and measured phosphorus adsorption. The data
on P adsorption maxima calculated by the Langmuir equation
and actual P adsorption based on P release from amendments
in soil and silica matrices are presented in Fig. 2-7. For
the CM treatment, with increasing incubation time P
adsorption maxima increased. Continued adsorption beyond
adsorption maxima indicated the presence of a precipitation
reaction, or multilayer adsorption. For tephrosia and kudzu,
the adsorption maxima decreased with time. A good agreement
among adsorption maxima and actual adsorption, calculated
based on SSMT, was observed at higher equilibrating solution

Table 2-6. Langmuir parameters (k and b) based on net release of P
measured by sequential extraction of simulated silica matrix with 0.01 M CaCl
Org.^
amend.
35 Days
65
Incubation
Davs
Period
150
Davs
r2
b+
k§
r2
b
k
r2
b
k
Tephrosia
0.87**
714
0.42
0.96**
556
0.56
Silica
0.97** 588
0.55
CM
0.99**
909
0.61
1.00**
1429
1.17
1.00**
1667
2.00
Kudzu
0.95**
502
0.43
0.97**
625
0.70
0.97**
625
0.89
PCW
0.97**
908
0.22
0.90**
769
0.45
0.97**
588
0.63
Grass
0.94**
769
0.32
0.92**
714
0.47
0.93**
667
0.65
** Significant at P = 0.01 level, ns = not significant
-f Pueraria phaseoloides (Kudzu), Tephrosia candida (Tephrosia), Mixed Gramineae
(Grass), Aerobically Processed City Waste (PCW), and Chicken Manure (CM).
4= b, P adsorption maxima, jug P g'1
§ k, bonding energy, mL /g P1

Table 2-7. Difference in the estimated values of Langmuir parameters (b and k) based
on the estimation of preadsorbed P by simulated silica matrix technique and
sequential extraction.
Incubation Period
35 Days
rg.+
amend. bT ks
65 Days 150 Days
b kb
k
Tephrosia -54.1 (14.2) 0.19 (0.01) 77.6
CM 265.1 (35.8) 0.53 (0.02) 810.6
Kudju 29.2 (5.9) 0.01 (0.00) 48.0
PCW 91.3 (15.9) 0.00 (0.00) -46.8
Grass 4.3 (1.3) 0.07 (0.01) 7.3
(8.3) 0.27 (0.04) 122.2 (8.4) 0.30 (0.01)
(58.3) 1.09 (0.09) 1170.7 (50.4) 1.90 (0.12)
(4.9) 0.22 (0.02) 90.0 (6.8) 0.29 (0.01)
(15.2) 0.18 (0.01) 71.2 (7.2) 0.25 (0.02)
(1.4) 0.14 (0.01) 63.7 (6.9) 0.29 (0.02)
-f Pueraria phaseoloides (Kudzu), Tephrosia candida (Tephrosia), Mixed Gramineae
(Grass), Aerobically Processed City Waste (PCW), and Chicken Manure (CM).
=f= b, P adsorption maxima, fig P g1
§ k, bonding energy, mL /ig P1
Numbers in the parenthesis are standard deviations.
CO
u>

Adsorbed P ug/g Soil
2,000
1,500
1,000
500
35d 65d 150d
2,000
1,500
1,000
500
c) Kudju
b a b
2,000
1,500
1,000
500
34
b) Tephrosia
m sm
r~i ssm
LAM
35d 65d 150d
2,000
1,500
1,000
500
d) PCW
35d 65d 150d
35d 65d 150d
Incubation Time
Figure 2-7. A comparison of P adsorption by Xanthic Hapludox
based on the estimation of preadsorbed P by sequential
extraction of soil or silica matrix incubated with
organic amendments.

35
concentration (60 /g g 1 soil) which indicated the validity
of this technique in predicting P requirements of amended
soils. In highly weathered soils, extractable P is usually
low but the amount of P which the soil can immobilize varies
greatly because of the variation in the reactive surfaces.
Addition of organic amendments changed the reactive soil
surface as indicated by the change in adsorption maxima and
bonding energy.
Conclusion
Decomposition of organic amendments influenced P
adsorption and desorption by soil. Amendments with low P
content (PCW, maize, and grass) immobilized P during early
soil samples incubated with CM. Soil incubated with Mucuna,
PCW, and the control for 35 d did not show any difference in
P desorption. However, desorption was higher for all
amendments incubated with soil for 150 d. Incubation of soil
with low P-containing organic materials may not influence
soil P desorption characteristics initially.
Amendments' P content and C:P ratio did not prove to
be helpful in predicting P mineralization. Amendments with a
C:P ratio of 1:139 (PCW) immobilized P during their
decomposition while others with similar or even higher C:P
ratios had a net P release in soil matrix. All amendments

36
released P in the silica matrix independent of their P
content and C:P ratio.
The use of SSMT to measure release of P from
decomposing organic amendments aid in the calculation of
adsorption maxima compared to the seqential extraction of
amended soil with 0.01 M CaCl2 solution. The calculated
values were in close agreement with actual adsorption
measured at high equilibrium P concentration.

CHAPTER III
ORGANIC AMENDMENTS AND CROP PHOSPHORUS NUTRITION
Phosphorus Management Strategy
In highly leached soils, the minerals with permanent
charge have been either severely altered or completely
weathered out, so that the surface charge arises from
adsorption of potential determining ions such as hydrogen
and hydroxyl. The magnitude of the surface charge is
expressed by a combined Gouy-Chapman and Nerst equation.
This equation provides a theoretical basis for increasing
the cation retention capacity of a soil by lowering pH0
(value of soil pH at which net surface charge is zero). One
way of lowering pH0 is to increase the organic matter,
phosphorus, or silica content of the soil (Uehara and
Gillman, 1981).
Highly weathered soils are also very poor in total and
plant available P. Their high P fixing capacity requires
high doses of applied P to meet crop demands. The classic
work of de Wit (1953) on physical theory of fertilizer
placement predicts that when suboptimum quantities of
fertilizer are used, restricted placement is desirable (Fox
et al., 1986). de Wit based his analysis on nutrient uptake
37

38
by entire root mass, and part of the root mass immersed in
nutrient solution and established a relationship between Ur
(g of nutrient taken up by the plant when part of the root
mass, Xr, was immersed in the solution) and Ub (g of
nutrient taken up by the plant when the entire root mass was
immersed in the solution).
Ur/Ub = (Xr/Xb)0-44
This relationship appears to be independent of crop type and
nutrient solution concentration. Under field conditions Ub,
Ur, Xr, and Xb take on new meanings.
Ub = uptake rate from broadcast fertilizer
Ur = uptake rate from banded fertilizer
Xr = width of the fertilizer band
Xb = distance between the crop rows
The use of data from P sorption curves concerning soil
solution P concentration to predict P uptake and crop yield
in combination with P placement analysis offers a way to
increase fertilizer use efficiency. Faced with a high P
fixing soil and a small guantity of fertilizer, the
fertilizer can be used to the best advantage by
concentrating it in a band so that the P concentration in
the soil solution is identical to the concentration that
produces a maximum yield in a broadcast application. Any
deviation from this optimum value, either to higher or to
lower concentrations, leads to less than an optimum return
per unit of fertilizer input.

39
Management of Organic Amendments
Organic amendments can affect the reaction of P in the
soil through complexation of polyvalent cations which are
major phosphate adsorption sites. It is widely believed that
humus in association with cations such as Fe3+, Al3+, and Ca2+
retains significant amounts of P. Appelt et al. (1975)
prepared a hydroxy-Al-humic acid complex that adsorbed P
because of the creation of new P adsorption sites. They
concluded that any increase in organic content of a soil
could lead to greater adsorption. A study conducted by Swift
and Haynes (1989) also showed that Al-organic matter
associations have a significant phosphate adsorption
capacity. Indeed, the Al-humate adsorbed amounts of
phosphate similar to those commonly reported for Al and Fe
oxides (McLaughlin et al., 1981) on a w/w basis. However,
simple organic acids, fulvic acids, and humic acids had no
effect on P adsorption by volcanic ash-derived soils. For
these soils, P was preferentially adsorbed over the organic
acids studied (Appelt et al.,1975).
The addition of organic materials to high P fixing soil
can decrease, increase, or leave virtually unaltered the P
fixation capacity (Yuan, 1980). The reduced fixation is the
result of: 1. complexation of Fe, Al, and Ca by organic
anions (Larsen et al, 1959), 2. competition of organic
anions and P for the same adsorption sites (Nagarajah et al,
1970), 3. development of organic coatings on mineral surface

40
(Easterwood and Sartain, 1987), and 4. reduction in bonding
energy of adsorbed anions resulting in low residence time
for adsorbed P (Hudal et al, 1988). Increased adsorption may
be due to: 1. cation bridging between organic anions, and
Fe3+, Al3+, and Ca2+ leading to formation of new sites for P
adsorption (Appelt et al, 1975), and 2. ability of organic
ligands to maintain hydroxy-Al, and Fe in a non-crystalline
state and thus maintaining a greater surface area (Swift and
Haynes, 1989).
Amendments chemical composition. Although organic acids
are an integral part of all organic matter by far they are
not the only reactive component influencing P adsorption.
Singh and Jones (1976) suggested that the P content of
organic residues plays an important role in the release or
fixation of added P. Similarly, chemical and decompositional
characteristics of the organic matter may influence total
C02 evolution (Sweeney and Graetz, 1988). This gas when
dissolved in water, forms carbonic acid, which is capable of
decomposing certain primary minerals. Elemental ratios such
as C/N and C/P are also considered valuable indicators for
net mineralization or immobilization of N and P contained in
organic amendments.
One of the important studies in this area was
conducted by Blair and Boland (1978) who examined the
release of P from white clover residue in high and low P
status soils. Their results suggested that the addition of

41
plant material resulted in immobilization of soil P only in
the low P soil in the absence of plants. In the high P soil
no immobilization of P was observed.
Crop residues have an effect on the nutrient status of
the soil. The cumulative effects of increasing quantities of
organic residues on available nutrients in soil were studied
for 11 years by Larson et al. (1978). They reported that
addition of 16 tons/ha of plant residue per year to the soil
increased the amount of N, S and P by 37, 45 and 14%
respectively, over the control treatment. They also found
that the NH4-N production, weak acid soluble P and
exchangeable K in the soil were increased as a result of
increasing the addition of organic residues.
Solubilization of Mn and Fe in soil were affected by
wheat straw and alfalfa amendments (Elliot and Blaylock
1975, Sims, 1986) The release was greater for Mn than Fe
and also much higher at 30 kPa than 50 kPa moisture
tension. The release of Mn and Fe from the soil column
followed the following order: alfalfa > wheat straw > soil
alone. They suggested that the potential accumulation of
soluble Mn in well drained soil was possible where there was
a large quantity of plant residues incorporated into the
soil.
Application technique. In the case of organic manures,
changes in organic P will to some extent depend on whether
the material is left on the surface of the soil or is plowed

42
in. Douglas et al. (1980) reported that the method of
placement, composition of residues and loading rates were
important factors influencing mineralization or
immobilization of N and S. A higher mineralization rate of N
from residue incorporated at 4 cm soil depth as compared to
the surface was also observed by Brown and Dickey (1970)
and Cocharan et al. (1980). Data on P mineralization from
organic matter in the literature is scarce.
Inorganic Phosphorus Management Strategy
Salinas and Sanchez (1976) have outlined a three-point
strategy for P management under limited resource conditions.
1. Use of cheaper sources of P. Two main sources are
phosphate rocks (PR) and thermally altered sources, such as
basic slags and the Rhenania phosphates. Numerous reports
have appeared in the literature regarding the fertilizer
value of PR as compared to other sources of P fertilizers,
e.g., superphosphate (Khasawneh and Doll, 1978; Hammond et
al., 1986; Hernandez and Sartain, 1985). Recently, Heliums
et al. (1989) compared the potential agronomic value of some
rock phosphate from South America and West Africa and
concluded that in addition to P the PRs with medium to high
reactivity have a potential Ca supply value. 2. improved
soil test interpretations, and 3. improved placement
methods.

43
Fertilizer placement. The advantage of banding
phosphate fertilizer is well known. What is generally not
understood is the sensitivity of nutrient uptake to band
width, de Wit (1953) demonstrated that a soil that is
virtually incapable of supporting a crop with 100 kg/ha of
broadcast phosphorus will produce nearly 50% of maximum
yield with the same amount of fertilizer applied in a narrow
band. A more detailed description of de Wit's analysis and
some of the assumptions contained in the analysis are
further elaborated by van Wijk (1966), Uehara and Gillman
(1981), and Fox et al. (1986). In agreement with the theory
Fox and Keng (1978) reported a better response from
localized P placement as compared with complete
incorporation if suboptimal P rates were used, but if
quantities of P were sufficient, best results were obtained
from incorporating P in the entire soil volume.
Kamprath (1967) found that similar maize yields were
obtained by annual banded applications of 22 kg P/ha for
seven years as were obtained by an initial P application of
350 kg/ha. Banding, therefore, saved more than half of the P
requirement. Applying N and P in knifed bands has been shown
to be an effective method of applying N and P to winter
wheat (Leikam et al., 1978, 1983). Experiments on maize have
shown that dual-placed N and P increases P uptake and maize
grain yield more than when P is banded to the side or below
the seed (Raun et al., 1987).

44
Band spacing of applied N and P fertilizer affects the
concentration of these nutrients in the applied band and the
probability of roots contacting the band. Sleight et al.
(1984) showed that in high P-fixing soil, increasing the
probability that root-fertilizer contact will occur is more
important than reducing soil-fertilizer contact during the
first week of oat (Avena sativa L.) growth.
There is also a threshold value of soil solution P
concentration beyond which the P uptake rate does not
increase (Jungk, and Barber, 1974). Anghinoni and Barber
(1980) in a P placement experiment reported maize root
growth stimulation in the portion of the soil where P was
added. Maximum shoot dry weight in their experiment was
obtained by placing the fertilizer in 0.25 of the soil
volume.
Summary
Application of P in narrow bands results in reduced
fixation of applied P and improves crop P nutrition. This
belief is based on: (i) localized P is protected to some
degree against irreversible adsorption or precipitation
reactions with the soil, (ii) localized P may be more
readily accessible to seedling roots than P widely
distributed in the soil, and (iii) plants can be adeguately
supplied with P through a few roots which proliferate in the
fertilizer band. Further, combined application of organic

45
amendments with inorganic P in a band will reduce direct
contact of inorganic P with a large volume of soil.
In this context, the overall objective of this research
was to devise a technique to sustain P nutrition in a highly
leached Oxisols with the help of organic and inorganic P
sources applied in narrow bands. The specific objectives
were to: (i) study the effect of different amendments and
their rate of application on maize dry matter production and
herbage nutrient concentration, (ii) examine the beneficial
effect of the combined application of organic amendment with
inorganic P in a narrow band as measured by maize grain
production and soil nutrient dynamics, and (iii) measure the
residual effect of applied treatments in relation to
selected soil chemical characteristics.
Materials and Methods
This experiment was conducted at the Empresa Brasileira
de Pesquisa Agropecuaria (EMBRAPA) station located 30 km
north of Manaus at an elevation of 50 m in the Amazonas
state of Brazil (Fig. 3-1). Climate in the Manaus region has
been classified in the Koppen nomenclature as afi, tropical,
humid and hot (Goes and Ribeiro, 1976). The soil used in the
study has been classified as Xanthic Hapludox (clayey
kaolinitic, isohyperthermic) (EMBRAPA, 1979).

46
Glasshouse Experiment
A factorial arrangement of three factors to yield 18
treatment combinations was employed in a randomized complete
block design (RCBD) with four replications (Table 3-1). Five
kg of unlimed soil from the Ap horizon of a Xanthic Hapludox
was mixed in a pot with organic amendments (<2 cm long) 7 d
prior to maize seeding. Maize (variety BR 5110) was planted
and thinned to 2 plants per pot 7 d after planting. The
experiment was harvested at 65 d. At harvest, leaves
immediately below and opposite to the ear leaf (70% plants
had begun to develop ear) were collected for chemical
analysis. A 0.2 g sample of the ground leaf tissue was
digested with H2S04 and H202. Potassium, Ca, Mg, Fe, Zn, Cu,
Mn were determined with an atomic adsorption
spectrophotometer, and P was determined colorimetrically
using the Murphy and Riley (1962) procedure. Soil P was
extracted with the Mehlich-I extractant P (0.05 M HCl +
0.0125 M H2S04, with a soil solution ratio of 1:10, and 5
min shaking time), unbuffered 1 M KC1 (1:10 soil:solution
ratio) was used for the determination of extractable Al.
Aluminum was determined by titrating the extract with 0.1 M
NaOH to bromthymol blue endpoint. Soil reaction (pH) was
determined in water using a soil:water ratio of 1:2.5 and in
1 M KC1 using the same ratio. Maize dry matter production
was recorded and surface response curves were fitted in the
case of interaction effects among factors.

47
A88 S O N D J89 F M A M J J A
Months
Figure 3-1. Geographic location of Amazon basin in Brazil
(a), on-station and farming systems research (FSR)
sites (b), and effective rainfall during the period of
August 1988 until August 1989 (c) at EMBRAPA station in
Manaus, Brazil.

48
Table 3-1. Factorial arrangements of treatments for the
glasshouse study.
Type
Amendments^
Rate
P Equivalent
kg ha'1
Mucuna
1&2
8.8, 17.6
Organic
CM
1&2
8.8, 17.6
PCW
1&2
8.8, 17.6
Inorganic
TSP
1,2&3

rH
CO

00
o
f Mucuna (Mucuna aterrima), CM (Chicken Manure), PCW
(Aerobically Processed City Waste) TSP (Triple
superphosphate)
Field Experiment
The site was cleared by burning an existing sugarcane
crop. Sugarcane stems and rhizomes were taken out of the
field. A uniformity trial using maize as an indicator crop
was planted for 60 d in order to record the existing
variation in the field. Visual field ratings combined with
soil and plant chemical analysis results were used to select
a uniform area for planting the field experiment. The
selected area was divided into 48 plots and soil samples
were taken from three consecutive depths (0-15, 15-30, and
30-45 cm) in each plot.
Dolomitic lime (2 Mg ha1) was applied over the entire
area 2 weeks prior to maize planting. Nitrogen and K were
applied at the rate of 200 and 100 kg ha'1. Zinc, Mo, and Mn
(2 kg ha1 of each) were applied in the band. Half of the N

49
as urea was broadcast over the entire plot and the remaining
half was sidedressed at 35 and 70 d after planting. All K
was applied basal broadcast. Six furrows each 20 cm wide and
8-10 cm deep were opened in each plot 80 cm apart. The plot
size was 10x5 m.
Mucuna aterrima and Canavalia ensiformis grown for 60 d
in an adjoining field were harvested and passed through a
hay chopper to produce a uniform size of less than 6 cm. All
organic amendments were applied in bands to which different
rates of P from triple superphosphate (TSP) was added. A
soil cover about 2 cm thick was put over the amendments.
Twelve treatments were tested in a RCBD with 4 replications.
The P release information from the SSMT technique (Chapter
II) was used to calculate the amount of a given amendment
required to supply P equivalent to 8.8 and 26.4 kg ha'1.
First maize. Maize variety BR-5110 was planted on
December 27, 1988. Plant population was adjusted to
approximately 55 x 103 plants ha"1 during the first sidedress
at 30 d after planting. Soil samples were collected before
planting, during tasseling (65 d after planting) and within
a week after the harvest. Leaf samples were collected during
tasseling. They were taken from immediately below and
opposite to ear leaf. The inner four rows were harvested for
grain. Grain production was recorded at 15% moisture level.

Table 3-2. Description of treatments tested in the field
Source
Amendment Type
Treatment^ P Equivalent
kg ha1
Mucuna aterrima
M60
26.4
M2 0
8.8
ORGANIC
Canavalia ensiformis
C60
26.4
C20
8.8
Processed City Waste
PCW 60
26.4
PCW 2 0
8.8
Chicken Manure
CM60
26.4
CM20
8.8
INORGANIC
Tiple Superphosphate
TSP20
8.8
TSP40
17.6
TSP60
26.4
ORGANIC+INORGANIC
Mucuna, Canavalia,
M20+TSP20
17.6
PCW, CM, TSP
C20+TSP20
17.6
PCW20+TSP20
17.6
CM2 0+TSP2 0
17.6
CONTROL
Control
0
f Whenever P from
organic and inorganic sources
was applied
together
amendments were applied first in furrows followed by the application of inorganic P.
U1
o

51
Second maize. This trial was conducted to measure the
residual effects of the treatments applied during the first
maize crop. Maize stover was taken out of the field.
Nitrogen was applied in the same manner as for the first
crop. No other nutrient applications were made. Maize
(variety BR-5110) was dibble planted in the rows and the
population was adjusted to 55 x 103 plants ha1. Soil samples
were collected during tasseling and after harvest. Again 10-
20 maize leaves were sampled during tasseling initiation
from each plot. Maize grain production was recorded from
four inner rows and was adjusted to 15% moisture content.
Chemical Analysis of Soil and Plant Materials
Soil pH was determined in water, 0.01 M CaCl2, and 1.0
M KC1 using a soil:solution ratio of 1:2.5. Total P was
determined by wet combustion method and inorganic P
fractionation was carried out using Chang and Jackson
procedure (1965). The pH of NH^F was adjusted to 8.2.
Mehlich-I extractant (0.05 M HC1 + 0.0125 M H2SO 4) was used
to extract soil P, K, Zn, Cu, and Mo. Phosphorus was
determined colorimetrically and K by a flame photometer.
Aluminum, Ca, and Mg were extracted with unbuffered 1 M KC1
(1:10 soil:solution ratio). Aluminum was determined by
titrating the extract with 0.1 M NaOH to bromthymol blue
endpoint, and Ca and Mg were determined with an atomic
absorption spectrophotometer. Soil apparent density was

52
determined by the method of a measuring cylinder. The
cylinder was struck against a rubber pad 10 times from a
distance of 10 cm. The final weight of the cylinder was
taken and the apparent density was calculated as the
following: Apparent Density (g cm3) = Weight of dry soil at
105C / Volume of the soil in the cylinder. Particle size
distribution was measured by the pipet method.
All organic amendments and maize leaf samples were
dried at 65 C for 72 hours for dry matter determination.
Subsamples of the dried plant material were ground in a
Wiley mill to pass a 0.5-mm screen. A 0.2-gm sample of the
ground plant material was digested with H2S04 and H202.
Phosphorus was determined calorimetrically, and K, Ca, Mg,
and micronutrients were determined with an atomic absorption
spectrophotometer.
Results and Discussion
Selected chemical and physical properties of soil used
in the glasshouse and field studies are presented in Tables
3-3 through 3-5.
With increasing soil depth, a reduction in Al-P was
observed. But, the Langmuir adsorption maxima increased with
the depth (789 /zg g'1 soil at 30-45 cm) (Table 3-3) It is
interesting to note that with over 80% clay content the soil
had a hydraulic conductivity of 25.1 cm h'1 (Table 3-4). The
ammonium oxalate extracable Fe and A1 were high and so was

53
the Al saturation (>76.0%) calculated from the effective
cation exchange capacity (ECEC) (Table 3-5).
Glasshouse Study
An analysis of variance (Table 3-6) presented for maize
herbage dry weight yield indicated the presence of a three
way interaction among types of organic amendment, rate of
amendment, and rate of TSP. Among the three amendments
tested, PCW produced the lowest yield when applied alone or
in combination with TSP (Fig. 3-2a). It is interesting to
note that when PCW was applied at higherr rate (equivalent
to 17.6 kg ha1 of P) it produced less dry matter per pot
compared to when this amendment was applied at lower rate
(Figure 3-2a). A possible explanation is that the slow rate
of release of nutrients from the material resulted in an
initial P deficiency for maize seedlings. There was also a
high concentration of Zn and Cu in this material (280, 155
/xg g1, respectively) which may have played a role in
providing cation bridging between organic and P anions
making P less available (Murray and Linder, 1984). By
forming organo-metal complexes, humic and fulvic acids as
well as simple acids can dissolve or decompose such minerals
as feldspar, gibsite, goethite, hematite, and mica
(Schnitzer, 1977). Therefore, organic amendments with high
Zn and Cu content may have limited potential as a source of
P in highly leached soils.

54
The highest dry matter yield per pot was obtained with
CM (Figure 3-2b). Combining the application of CM and TSP
provided a higher dry matter yield at rate 2 (equivalent of
17.6 kg ha1 of P) compared to rate 1 at 0, 8.8, and 17.6 kg
ha1 P as TSP.
There was a strong relatioship between Mehlich-I
extractable P and dry matter production (r2=0.88), and
tissue P concentration and dry matter production (r2=0.79)
(Figure 3-3). The relationship was linear for both
indicators.
Several researchers have attributed the positive effect
of organic amendments in improving P nutrition to a change
in soil pH (Sanchez and Uehara, 1980) which improves the
plant growth conditions and increases the solubility of
native P. Soil pH increased approximately by one unit (4.5-
5.5) in response to the application of PCW (Table 3-7). But
this treatment produced the lowest dry matter. Application
of Mucuna aterrima also improved the soil pH but the
magnitude of improvement was less by 0.4 unit. The control
treatment and CM had essentially the same pH, but CM had
higher dry matter production.

55
Table 3-3. Selected chemical properties of the fine
fraction (<2 mm) of the Xanthic Hapludox.
Depth Org.
carbon
TP
OP
Chana
& Jackson^
Langmuir
MI-P max'''.
Al-P
Fe-P
Ca-P
a'1
cm
g kg
M9
9
0-15
14.6
200
25.2
51.6
0.3
0.3
2.4
550
15-30
12.2
120
14.5
18.5
2.8
0.2
1.8
620
30-45
8.3
90
6.7
5.2
0.1
0.1
1.2
789
-J- pH of NH4F was adjusted to 8.2. The reductant soluble P is
not included.
4= 0.01 M CaCl2 was used as an electrolyte.
OP = Organic phosphorus, TP = Total phosphorus
MI-P = Mehlich I extractable P.
Table 3-4. Selected physical properties of the Xanthic
Hapludox.
Depth Bulk Fine fraction Hydraulic
Density Sand Silt Clay' Conductivity
cm
g cm'3
kg kg'1
of <2
mm
cm h'1
0-15
1.11
0.14
0.11
0.75
25.1
15-30
1.30
0.11
0.06
0.88
6.2
30-45
1.16
0.09
0.09
0.82
7.3
-f Fine sand fraction is also included.

Table 3-5. Selected chemical properties of the fine fraction (<2 mm) of the Xanthic
Hapludox.
Depth
OH
Extractable bases ECEC
Charcre"^
Oxides^
Al Sat
h2o
KC1
Ca
Mg
K Al
(-)
( + )
Fe
Al
(ECEC
cm
0-15
4.7
4.3
0.26
0.06
cmol (+)
0.05 1.20
kg'1
1.57
1.96
0.77
g
0.07
100g
0.30
1
76.4
15-30
4.4
4.1
0.24
0.04
0.02 1.31
1.61
0.60
1.43
0.08
0.35
81.4
30-45
4.3
4.1
0.19
0.05
0.01 1.24
1.49
0.64
1.63
0.09
0.38
83.2
Acid ammonium oxalate extraction.
Acid-base potentiometric titration.
U1
CTi

57
Table 3-6. Analysis of variance for maize herbage dry weight
production per pot in the glasshouse study.
Source^
DF
F Value
Pr > F
REP
3
1.4
0.23
ORGANIC
2
970.2
0.00
RATE
1
189.2
0.00
ORGANIC*RATE
2
144.0
0.00
INRATE
2
127.9
0.00
ORGANIC*INRATE
4
17.6
0.00
RATE*INRATE
2
0.1
0.86
ORGANIC*RATE*INRATE
4
8.3
0.00
ERROR
51
CORRECTED TOTAL
71
CV%
13.0
f REP = Replication, ORGANIC = Organic amendments, RATE =
Rate of Organic amendments, INRATE = Rate of Inorganic P.

l d 6
PCW
CM
58
Mucuna
Figure 3-2. Effect of different rates of selected organic
amendments in combination with different rates of
inorganic P on maize dry matter production at 65 d
after planting in a glasshouse study.

59
% Leaf P
Figure 3-3. Relationship between Mehlich-I extractable
soil P (a), and maize leaf tissue P concentration (b)
with maize dry matter yield.

60
Table 3-7. Mean canges in soil pH (H20) following
application of selected organic amendments in the
glasshouse study
Amende.
Rate5
Obs.
Mean pH
SD
Mucuna
1
12
5.09
0.22
Mucuna
2
12
5.47
0.20
CM
1
12
4.80
0.14
CM
2
12
4.87
0.26
PCW
1
12
5.48
0.19
PCW
1
12
5.85
0.23
Control
0
4
4.53
0.09
TSP
1
4
4.72
0.27
TSP
2
4
4.70
0.08
TSP = Triple superphosphate, CM = Chicken Manure, PCW =
Aerobically processed city waste.
§ Rate 1 & 2 are equivalent to 8.8 and 17.6 kg ha'1 of P.

61
Field Study
Rainfall distribution data for both cropping cycles are
presented in Figure 3-lc. The distribution is bi-modal with
a short dry season. During this period evaporation is
greater than precipitation.
Grain Yield. Single degree of freedom orthogonal
contrasts of maize grain production during the first crop
(Table 3-8) indicated that one application of PCW equivalent
of 26.4 kg P ha'1 produced only 1.41 Mg ha"1 of maize. This
level of production was inferior to the control treatment.
Canavalia ensiformis. Mucuna aterrima. and TSP when applied
separately to provide 26.4 kg ha1 of P produced the same
amount of grain during first and second cropping cycles. But
the CM treatment produced more grain than other treatments
in both the first and second crops. This was expected based
on the results of the adsorption study presented in Chapter
II. One explanation is that the faster rate of decomposition
(0.26 g/100 g d'1) and the high P content in CM maintained
the soil P concentration at a high level from the beginning
of plant growth. In a review paper, Olsen and Barber (1977)
concluded that an annual application of manure and
superphosphate resulted in an increased level of 0.01M
CaCl2- and 0.5 M NaHCOj-extractable P. In most studies,
manure treated soils tend to support a higher level of
soluble P than soil treated with an equivalent amount of
superphosphate.

62
Table 3-8. Orthogonal contrasts of maize grain yield under
different treatments applied in a 30 cm wide band
at UEPAE research station, Manaus, Brazil.
Treatment^
Mean Grain Yield
Cropl Crop2 Contrasts Cropl
- Mg ha'1 --
Crop2
M60
3.44
1.65
C60 VS TSP60
ns
ns
C60
3.56
2.07
CM60 VS TSP60
**
**
PCW60
1.41
1.67
PCW VS TSP60
"k:k
ns
CM60
4.63
2.37
M60 VS TSP60
ns
ns
M20+TSP20
3.55
1.41
C20+TSP20 VS TSP40
*
ns
C20+TSP20
3.62
1.81
M20+TSP20 VS TSP40
ns
ns
PCW20+TSP20
i 1.66
1.90
PCW20+TSP20 VS TSP40
**
ns
CM20+TSP20
4.10
1.92
CM20+TSP20 VS TSP40
k k
ns
Control
1.71
1.10
TSP2 0
2.63
1.04
TSP40
3.12
1.46
TSP60
3.69
1.63
*, ** Significantly different at 0.05, and 0.01 level of
probability, ns = not significantly different at 0.05 level
of probability.
f M60, C60, PCW60, and CM60 = Mucuna, Canavalia, Aerobically
processed city waste, and Chicken manure applied to provide
equivalent of 26.4 kg ha'1 of P.
TSP20, TSP4 0, and TSP60 = 8.8, 17.6, and 26.4 kg ha'1 P from
triple superphosphate (TSP).

63
Combined application of 8.8 kg P ha1 from organic
amendments and the same amount from TSP giving an
application rate of 17.6 kg P ha 1 P was compared to 17.6 kg
P ha1 from TSP. Differential results relative to amendment
source was observed. For the first crop, the combination of
TSP with CM increased yield 30% relative to TSP. Mucuna
aterrima did not influence yield while Canavalia ensiformis
combined with TSP produced higher yield (3.6 compared to 3.1
Mg ha'1 for TSP) Chicken manure was still the best of all
treatments during the second crop. But, PCW performance
improved over the first crop. And there was no difference
between PCW60 and TSP60, and PCW20+TSP20 combined. Such
findings highlight the limitations of incubation studies
conducted in the laboratory for a short span of time in
predicting long term effects of organic amendments on soil P
dynamics.
A response surface plot (Figure 3-4) for TSP indicated
a linear response between grain yield and rate of TSP up to
26.4 kg P ha1 (r2 = 0.90). The magnitude of this response
was declining in the second crop but the regression analysis
did not show any difference with varying rate of TSP.
Change in soil phosphorus status. The change in P
status of the soil over the 240 d cropping cycle for all
treatments, except PCW, followed a cubic surface response
(Figure 3-5). The data presented are for 0-30 cm depth. All
treatments improved soil P status compared to the control.

Yield, Mg ha
64
0 8.8 17.6 26.4
P, kg ha'1
Figure 3-4. Rate of inorganic P applied through triple
superphosphate and its effect on maize yield in a
maize-maize rotation.

65
Trt.
a
b1
b2
b3
2
RZ
TSP60
4.79
5.53E-01
-6.30E-03
1.65E-05
0.75
M60
4.77
4.08E-01
-3.49E-03
7.58E-06
0.85
Control 4.05
-6.92E-03
-2.00E-04
7.18E-07
0.95
PCW60
4.07
3.96E-02
-3.08E-04
7.91E-07
0.15ns
C60.
3.62
3.02E-01
-2.56E-03
5.56E-06
0.99
Figure 3-5. Effect of selected organic amendments, applied
in a quantity equivalent to provide 26.4 kg ha"1 of P,
on sustaining Mehlich-I extractable soil P pool in a
maize-maize rotation on a Xanthic Hapludox. M = Mucuna,
C = Canavalia

66
The actual P values ranged from 0.2 fig g1 soil (control) to
about 8.0 ¡ig g'1 soil for other amendments measured at the
end of 240 d. This finding conflicts with the results
obtained by Izza and Indiati (1982) where they studied the
effect of various organic farm products on soil available P.
They found that incorporation of these materials in high P
fixing soil produced no effect on soil available P.
At 65 d TSP60 and M60 treatments had the same soil P
status. There was a sharp decrease in soil P with TSP60
treatment as the cropping season progressed compared to M60
and C60. The plots which received PCW maintained low P which
could not be described with the aid of any polynomial. When
TSP20 was combined with M20 the P status improved compared
to TSP40 (Figure 3-6). Application of inorganic P in the
proximity of the organic amendment may have reduced the
exposure of inorganic P to a larger soil volume leading to
reduced fixation. And the decomposition of organic material
may have inactivated active soil adsorption sites in the
localized band. Increased effectiveness of soil amendments
such as lime, in the proximity of organic matter has been
reported by Ahmed and Tan (1988). Combined application of
C20 with TSP20 was as good as TSP40, and PCW20 applied with
TSP20 was inferior not only to TSP40 but to all treatments,
except control.

67
Trt.
a
bl
b2
, _2
b3 R
TSP40
4.97
3.13E-01
-3.66E-03
9.57E-06 0.78
M20+TSP20
4.22
3.48E-01
-3.18E-03
7.14E-06 0.90
PCW20+TSP20
4.41
6.26E-02
-8.02E-04
2.24E-06 0.30ns
TSP40
3.77
2.96E-01
-2.69E-03
5.92E-06 0.92
Figure 3-6. Effect of selected organic amendments, applied
in combination with inorganic phosphorus source in a
quantity eguivalent to provide 8.8 kg ha1 of P, on
sustaining Mehlich-I extractable soil P pool in a
maize-maize rotation on a Xanthic Hapludox.

68
Movement of Ca in the soil profile over the 240 d
period indicated that application of TSP may reduce the
soilCa pool (Figure 3-7). The level of Ca in the 0-30 cm
depth was 2.30 cmol (+) kg'1. This value dropped to 0.50
cmol (+) kg'1 within 240 d. Maize plants from the control
treatment plot also had the lowest uptake of Ca. Application
of M60 and CM60 improved the Ca status in 0-15 cm. These
materials after decomposition released cations which were an
integral part of their composition (Larsen et al., 1972).
Deep-rooting green manure crops offer the advantage of
recycling cations which have been leached to a deeper soil
profile. There is considerable evidence in the literature
that organic ligands can hold polyvalent cations and prevent
them from leaching (Moreno, 1960). Building a cation reserve
is the matter of great importance in acid soils where soil
chemical and physical conditions favor their rapid depletion
through leaching.
High P concentration in leaf tissue reduced Zn and Cu
concentration in leaf (Figure 3-8). This can be attributed
to the chelation of these elements by fulvic acids (Saar and
Weber, 1982). However, there is not enough evidence to
conclude that at low leaf P concentration there will be a
higher uptake of these metals. It has been shown that soils
with high level of P are less likely to exhibit Cu toxicity
(Sartain and Street, 1980).

Depth, cm
69
0.0 1.0 2.0 3.0 4.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0
Figure 3-7. Leaching of calcium from surface horizon of
highly leached Xanthic Hapludox as influenced by
different organic amendments.

ugg1 (Leaf)
25.0
Cu
20.0

15.0

m
10.0
5.0
0.0
111 rm cu
2dd d
T]
O ~ffi i i i i i i i
L
n
Hi ittA 11 rrn
l
Figure 3-8. Relationship between leaf tissue P and
concentration of Zn and Cu in maize leaf tissue.

71
Conclusion
In the glasshouse experiment, among the three
amendments tested PCW produced the lowest dry matter per pot
when applied alone or in combination with TSP. The highest
dry matter production was obtained with the application of
CM. A strong relationship was observed between dry matter
production and Mehlich-I extractable P (r2 = 0.88).
In the field experiments, CM applied plots produced
more grain in first as well as second crops compared to the
plots which had received Canavalia ensiformis. Mucuna
aterrima. PCW, and TSP. This finding agreed with the
prediction made from the laboratory incubation studies using
silica as the matrix substratum to evaluate the
decomposition rate of organic amendments. The prediction was
made based on P adsorption maxima and binding energy values,
and indicated that CM applied plots will produce superior
yield compare to the plots which received other amendments.
When CM was applied in combination with TSP higher
grain yield was obtained compared to the same amount of P
applied from TSP. All organic amendments improved the soil P
reserve and reduced Ca leaching.

CHAPTER IV
RECOMMENDATION DOMAINS AND MODIFIED STABILITY ANALYSIS
Introduction
The problem of non-adoption of new technologies among
small farmers in third world countries has been widely
recognized, not as the farmers' unwillingness to accept
change but rather as the inappropriateness of the technology
to the real conditions that exist at the farm level in terms
of the social, economic, and natural resource endowment
(Shanner, et al., 1982). This chapter explores 1)
techniques to understand farmers' circumstances, 2) the
concept of recommendation domains (RDs), 3) the change in
soil fertility and its impact on crop production, and 4) the
use of modified stability analysis as a tool for the
selection of appropriate technology for farmers with
different production goals.
Research Data Base
An applied researchable problem should be based on
farmers' immediate concerns. Such concerns are often
documented through formal surveys with a structured
questionnaire. Although this technique produces statistics
72

73
on farm input and output, it is expensive and time
consuming. In order to make surveying cost-effective, Vander
Veen and Mathema (1978) used key informants to gather
baseline information. This speeded-up the process but it
lacked interactive information.
Working for the farmers' cause requires direct
involvement in problem identification. Rapid rural
appraisals, often referred to as exploratory or diagnostic
surveys, are a simple and relatively quick method of
identifying constraints that operate in a defined area
(Abalu, et al., 1987). The sondeo (Hildebrand, 1981) is one
of the techniques of rapid rural appraisal which combines
different disciplines in a team. Its primary purpose is to
acquaint technicians with the area in which they are going
to work. No questionnaires are used; rather informal
interviews with farmers are conducted.
Recommendation Domains (RD)
In order to be cost effective, the research activities
have to address the problems of and provide solutions for
relatively large numbers of people. It is necessary,
therefore, to classify farmers with similar circumstances
into recommendation domains: groups of farmers for whom it
is possible to make more or less the same recommendation
(Perrin et al., 1978; Byerlee et al., 1980). More recently
the concept of RDs has been extended to Research Domains and

74
Diffusion Domains (DDs) (Hildebrand, 1986). Many
researchers, however, have considered RDs as a synonym for
cropping systems (Zandstra, et al., 1981), farming types
(Njobvu, 1986), and homogeneous groups (Moussie and Muhitira
(1988). The limit set by the definition of Byerlee et al.
(1980) has also been broadened to include agroclimatic zones
and individual fields, in addition to farms.
Criteria for Delineating Recommendation Domains
There is a great debate on the criteria used for
delineation of RDs. Socioeconomic criteria may be just as
important as agroclimatic, and agroecological variables
(Njobvu, 1986) in delineating domains. If so, the resulting
domains are often not amenable to geographical mapping
because farmers of different domains may be interspersed in
a given area. Moussie and Muhitira (1988) attempted to
classify farmers into relatively homogenous groups using
cluster and discriminant statistical analysis. This method
employed the use of qualitative information obtained from
sondeos, and the selection of key variables from in-depth,
formal surveys to obtain a discriminant function which helps
to identify the most important variables in classifying
farmers into RDs. Swinton and Samba (1986) used four
agronomic criteria: average annual rainfall, soil fertility,
soil texture, and depth to the subterranean water table for
defining agricultural technology recommendation domains.

75
Tshabalala and Holland (1986) indicated that the "average"
fanner is a myth and the programs designed to help him or
her will fail, but used the RD concept in matching the
improved technology to the group likely to be interested in
taking advantage of it.
Economic criteria are also important in the delineation
of RDs (Hildebrand and Poey, 1985). Improved technology
often requires more cash and labor investments. Both
resources are scarce on small subsistence family farms.
Under limited cash availability, and so many competing uses
for it, a farmer will consider an option which will give him
or her the highest return per dollar invested. In this
context, economic evaluation criteria such as net return
ha'1, net return/total cost, and net return/cash cost play
important roles.
The RD concept is being used in other disciplines
(Fattori, 1990) and the use is being considered as an
extension tool guiding the effective dissemination of
technology appropriate to small farm conditions.
Stability Analysis
The stable performance of crop cultivars over a wide
range of environmental conditions is generally regarded as
desirable, but there is disagreement both on its definition
and on the most appropriate methods for its statistical
measurement from yield data trials (Becker, 1981; Hill and

76
Baylor, 1983). The most widely used stability analysis has
been linear regression of cultivar yield on an environmental
index derived from the mean of all or a subset of cultivars
at each location or environment (Eberhart and Russell,
1966). Stability analysis has been used extensively to
select genotypes that interact less with the environment
and, therefore, are considered stable. Mackenzie, et al.
suggested the use of stability analysis in predicting the
response of a tested variety under different growing
conditions for comparing the worthiness of selected potato
clones for further replicated testing before naming and
release. A regression coefficient of unity has been
considered favorable for selection of stable genotypes.
Hildebrand (1990) discussed the drawbacks of the use of the
regression coefficient in selecting adaptable, or stable
genotypes.
Hill and Baylor (1983) pointed out that perennial crop
yields are usually measured on the same plot over a number
of years, so problems with stability analysis may be
encountered due to a differential change in the yields of
the entries as the stand ages. They suggested an alternative
orthogonal contrast analysis that partitions the variation
over environmental components for each entry into sources
due to environmental components (year, site, and management)
and all possible interactions between these factors.

77
The reason Eberhart and Russell (1966) used the site
mean as an environmental index was because a lack of
knowledge of the relationship of macro-environmental
differences such as temperature gradients, rainfall
distribution, and soil types did not permit the computation
of an index which could transform the environment into a
continuous variable. However, attempts continue to guantify
the production environment. Advancements in using the
multivariate approach to group soils in the field based on
variations in systematic and random components are
forthcoming (Winding and Dress, 1983). Systematic variation
is caused by difference in the parent material, relief and
biological action as well as soil management practices such
as fertilizer application and tillage. Random variation
which is called "noise" by Burrough (1983), represents the
statistical heterogeneity of the soil.
The need for assessing the factors causing soil
microvariability in the tropics has been stressed by Moorman
and Keng (1978). In general, some soil characteristics are
mutually correlated with each other (Norris, 1970). Hence,
factors causing soil variation which are reflected in one or
more of the soil characteristics may be used as criteria for
grouping soil.
To analyze the causes of soil variation Kosaki and
Anthony (1989) applied principal component analysis (PCA),
which is a mathematical technigue used to summarize data and

78
investigate the relationship among variables. Variables
employed for PCA in their study included soil pH, organic
carbon, available P, exchangeable Ca, exchangeable Mg,
exchangeable K, sand, silt, and clay. For the computation of
principal components, they used a correlation matrix.
However, soil variation is only one factor among many which
influences the performance of a technology in a given
environment. Much remains to be understood concerning the
complicated interactions among agronomic, economic, social,
and cultural variables which have a bearing on the
performance of a technology.
Modified Stability Analysis and Farming Systems
Farming Systems Research/Extension, is considered to be
a dynamic, interactive, and problem oriented approach to
develop technology for farmers, particularly those with
limited resources. The technological base of FSR/E is on-
station research but it constitutes only a part of the
overall FSR/E program. The main activities are concentrated
in farmers' fields with direct farmer involvement in
technology evaluation and feedback. The farmers' field
trials involve a few treatments, and often without
replications. Farmer to farmer variation in management
practices for a given experiment is not controlled. Instead,
any unusual practice is recorded. Under these circumstances,
modified stability analysis (MSA) (Hildebrand, 1984;

79
Upraity, et al., 1985, Fattori, 1990) has been used to
select environment-specific production techniques. This
technique does not depend on the concept that a regression
coefficient of unity is always favorable for the selection
of a technology. Adherence to this concept leads to
rejection of superior technology for a specific environment
in search of a 'stable' technology.
Objectives
The overall goal of this research was to identify
appropriate technology for maize and cowpea production by
small farmers. The specific objectives were to: (i) measure
the performance of selected treatments in different land
types, (ii) study the changes in soil fertility parameters
as influenced by the application of organic amendments, and
(iii) develop location-specific recommendations.
Materials and Methods
Site Description
Two small farming communities (mean cultivated size = 3
ha/farm) in the municipality of Rio Preto da Eva, located in
the state of Amazonas, Brazil were selected for on-farm
experimentation. The area is accessible only by small
motorboats. In trying to improve living conditions of these
marginal farmers, the government of Brazil was just
beginning a small watershed management program. The project

80
was intended to bring awareness among farmers to preserve
the environment and assist them in improving their
agricultural production. The Brazilian national agricultural
research institution (EMBRAPA) has a mandate of developing
appropriate technology for different farming conditions in
this relatively inaccessible area.
Developing a Research Base
Secondary information regarding indigenous farming
practices of the area was collected from published sources.
A rapid appraisal of the area was conducted with a
multidisciplinary team of scientists participating from
various research disciplines and state planning and
agriculture extension organizations who visited the area on
three different occasions to collect and verify information
obtained in group discussions or during individual
communication with farmers. Farmers' knowledge of indigenous
technology, agronomic practices, and land types being used
were recorded. An extensive soil sampling program was
carried out to understand soil physical and chemical
characteristics and relate them to farmers' rationale for
assigning a particular cropping pattern to a given land
type. Farmers played an active role in technology design,
execution and evaluation.

81
Selection of Treatments
Three treatments from previous on-station research were
selected for comparison with farmers practices' (FP) for
growing maize (Zea mays L.) and cowpea (Vigna uncruiculata)
(Table 4-1). Several locations were identified to encompass
the different land types within the soil family limit of
Clayey, Isohyperthermic, Xanthic Hapludox. A shift in maize
planting date was agreed upon so maize maturity would
coincide with the beginning of the dry season (Fig. 4-1).
For the maize crop, all treatments except FP received 100 kg
ha"1 of N from urea half applied broadcast before planting
and the other half in two additional split applications. All
K (60 kg ha"1) was applied basal broadcast except to the FP
plots. Processed city waste, CM, and TSP were applied in 25
cm bands. The maize variety BR-5110 was planted in rows 80
cm apart. No nitrogen was applied to the cowpea crop. The
cowpea variety IPEAN V-69 was planted in rows 60 cm apart.
Plot size for maize and cowpea varied from 100-200 square
meters. The organic amendments were applied in a 20 cm band.
Both crops were harvested for grain. For growing maize and
cowpea the main differences between FP and the improved
practice was that the improved practice received fertilizer.
The land preparation and planting methods consisted of
clearing the area by slash and burn, followed by manual land
preparation and planting with sticks. This is a common FP in
the area.

82
The trial was planted in a RCB design with two
replications wherever possible. Some locations had only one
replication because of limited area allocated by farmers for
the trial.
Table 4-1. Application of N, P, and K in different
treatments tested in on-farm experimentation for maize
and cowpea crops in the municipality of Rio Preto da
Eva, Amazonas, Brazil.
Treatment
Maize
Cowoea
N
P
kg
K
ha'1
N
P K
FP+
0
0
0
0
2+
0
CM20+TSP20
100
8.8+8.8
60
0
88+88
60
PCW20+TSP20
100
8.8+8.8
60
0
8.8+8.8
60
TSP40
100
8.8+8.8
60
0
0+17.6
60
-f No restriction was imposed on the amount of nutrient or
type of cultural practice to be used. This treatment
varied from farm to farm.
4= the range for P was 0-4 kg ha1. The value of P reported
is the mean.
Soil samples were taken before planting, during maize
and cowpea flowering, and after harvest. All soil samples
were analyzed for pH, Ca, Mg, Al, K, and P. Particle size
analysis was conducted only on samples taken before
planting. Plant tissue samples were analyzed for P, Zn, Cu,
and Mn. Their results are used in Chapter III. The
analytical technigues for soil and plant tissue samples were
essentially the same as described in Chapters II, and III.

83
Effective Rainfall, mm
400
300
200
100
o
-100
-200
A88 S O N D J'89 F M A M J J A
Months
Figure 4-1. Effective rainfall and suggested shift in the
planting date for maize in Rio Preto da Eva, Amazonas,
Brazil, (a) common practice, (b) suggested practice.

84
Statistical Analysis
Combined analysis of variance (CANOVA). A combined
analysis of variance was conducted for replicated trials. A
preliminary analysis for homogeneity of variances from the
individual ANOVAs was based on the chi-squared test. In the
case of a significant chi-squared test, locations with a
coefficient of variation >20% were excluded from the CANOVA
(Gomez and Gomez, 1985). No attempt was made to employ a
missing plot technique to recover lost data or data which
violated some assumptions of the ANOVA. It is common to lose
part of on-farm experiments to destruction of experimental
plants, loss of harvested samples, etc. Data from two
locations for maize and one for cowpea were lost due to
misunderstanding with farmers about the harvest date. The
trials were harvested before the agreed upon date. At three
locations (2 for maize and one for cowpea), the crop failed
due to late planting. Data from one location for cowpea
could not be used for ANOVA because the second harvest yield
for all treatments was mixed without recording production
separately.
Stepwise regression. Stepwise regression is a method
for either the forward, or backward selection of variables
based on an F statistic of R2 significance at the level
specified in the model (SAS Institute, Inc., 1985). This
technique was used to identify factors responsible for the
wide range of variation in maize and cowpea production over

85
locations. For this purpose, yield obtained from each
treatment over locations was stepwise regressed on the
values of soil pH, ECEC, Mehlich-I extractable-P, and A1
saturation. These variables were measured from soil samples
taken during tasseling stage for maize and flowering stage
for cowpea (Hanway, 1967).
Modified stability analysis. The production
environments were separated based on the average yield of
all treatments at each location (Eberhart and Russell,
1966). In this way, environment becomes a continuous,
quantifiable variable. Yield for each treatment was related
to environment by simple linear regression.
Yjj = a + be
where = yield from treatment i at the jth
location, and
ej = jth environmental index.
The use of the regression coefficient 'b' in the linear
equations with a value near one to select the "stable"
treatment over all environments was avoided. Instead
superior treatments were identified for groups of
environments, or RDs (Hildebrand, 1984).
The distribution of confidence intervals for the
treatments within RDs for both crops was calculated as
follows:
Cl = Y taS/n,/!
Where;

86
Y = the mean treatment yield, or other criterion within
the RD,
a = the level of confidence,
t = value from a two-tailed "t" table,
a
S = [Zx2/(n-l)]% or standard deviation of treatment
yield, or of other evaluation criterion within the RD, and
n = no. of locations or environments in each
recommendation domain.
The Cl test was used to provide an assessment of the risk of
low yields and unacceptable levels of other evaluation
criteria within each RD.
Results and Discussion
The project area is inhabited by subsistence farmers
who practice a slash-and-burn, rotational type of
agricultural system and market surplus production and
products gathered and hunted in the forest. The two major
land types used for crops are: (i) area cleared for the
first time by slash-and-burn from primary forest (PF), and
(ii) area cleared from secondary forest which was left in
fallow 5-7 years (SF). For this trial, a third type was
used, (iii) area considered undesirable for agricultural
activities (WL). The chemical and physical properties of
these soils are presented in Tables 4-2 and 4-3. All sampled
soils were acidic (median pH H20 = 4.5) with very low ECEC

87
(1.35-4.21 cmol (+) charge kg'1 soil), high Al saturation
(60-90%), and low Mehlich-I extractable
p (0-12.0 /xg g"1)
Change in Soil Fertility with Time
Initial soil characteristics were grouped by land types
(PF and SF). Within each group, soil pH, Al saturation,
ECEC, and Mehlich-I P were regressed with the number of
years the land was in production. The objective was to see
the magnitude of change in soil fertility parameters after
deforestation had taken place and the land was used for
agricultural purposes. The results are presented in Table 4-
4. The initial values (intercept) of soil pH, Mehlich-I
extractable P, and ECEC were higher in PF compared to SF. Al
saturation was higher in SF (81.5 % compared to 61.1% for
PF) .
The rate of decline in pH in PF was 0.15 unit per year
which was similar to decline in pH in SF. Aluminum
saturation was increasing at the rate of 7% per year in PF
compared to 4% in SF. The rate of decline in Mehlich-I P in
both land types was in the range of 2.1-2.6 nq g'1 soil. A
relatively low ECEC in both land types (2.3-3.6 cmol (+)
charge kg1 soil) indicated that most cations were leached
following heavy rainfall. And those present were being taken
up by vegetation or being washed out at the rate of 0.80 and
0.16 cmol (+) charge kg'1 soil every year from PF and SF,

88
respectively. This analysis provided evidence that
restoration of soil fertility by letting secondary forest
take over for 5-7 years is unlikely to happen.
Soil Chemical Properties and Yield
A stepwise regression was carried out to understand the
relationship between maize and cowpea yield and soil
chemical properties. Emphasis was placed on how different
amendments influenced Mehlich-I extractable P. All
treatments except FP had received an equal amount of P from
organic and inorganic sources. Table 4-5 presents results
of the stepwise regression on amended soil characteristics.
For FP in maize, each unit increase in A1 saturation
decreased the production by 0.06 Mg ha'1. Application of
PCW20+TSP20 did not improve soil P level, instead an
improvement in soil pH was observed (similar results were
obtained in glasshouse study). Application of TSP40 and
CM20+TSP20 resulted in improved soil P. However, the
magnitude of yield improvement with similar increases in P
levels were different for different treatments. It was
higher for TSP40 treatment by 0.90 Mg ha'1 compared to
CM20+TSP40.
It is widely believed that low soil pH is often
associated with Al and Mn toxicity and Ca deficiency. An
acidic soil reaction can reduce rhizobia growth, nodule

Table 4-2. Characterization of experimental plots for maize
testings in the municipality of Rio Preto da Eva, Amazonas, Brazil.
General^ Physical Chemical
(e)
Loc.
Land
Type
Year in
Prod.
Density
Sand
Clay
Silt
PH
ECEC+
A1
Sat.
P(M-I)
g cm3
g
H
O
O
iQ
i

g
I00g'1
Mg g'1
3.10
7
PF
1
1.20
68.6
25.5
5.9
5.2
4.21
58.3
7.4
2.75
6
PF
1
1.04
82.7
14.2
3.1
5.1
3.45
69.1
7.1
2.23
2
SF
1
1.07
60.7
31.8
7.5
4.6
2.29
91.7
4.5
2.10
8
PF
1
1.37
75.0
21.0
3.9
4.5
2.26
79.2
6.8
1.96
5
PF
2
1.15
56.5
35.3
8.2
4.6
2.45
80.0
5.0
1.43
4
SF
2
1.04
80.7
14.8
4.5
4.1
3.12
94.8
2.8
1.11
1
SF
2
1.14
47.4
39.3
13.3
4.2
1.99
90.7
2.0
0.20
3
WL
4
1.05
59.6
32.2
8.1
3.9
1.35
94.8
trace
f PF = Area cleared for the first time by slash and burn from primary forest.
SF = Area cleared from the secondary forest which was left for 5-7 years
for building up of soil fertility, (e) = Environmental Index, Mg ha'1
4= ECEC = Effective cation exchange capacity (cmol+ charge kg'1 soil)
§ Mehlich I extractable phosphorus
00

Table 4-3. Characteristics of experimental plots for cowpea trials.
General^
Physical
Chemical
(e)
Loc
Land
Year
in Density
Sand
Clay
Silt
pH ECEC' Al
P(M-I)
Type
Prod.
Sat.
g cm"3
g I00g"1
g lOOg
'1 Mg
g1
1.92
7
PF
1
1.30
63.8
28.8
7.4
5.2
3.24
59.8
8.0
1.80
10
PF
1
1.29
52.1
37.1
10.8
5.4
2.21
61.2
12.9
1.67
3
PF
2
1.00
80.7
16.2
3.1
5.3
1.25
80.1
5.0
1.60
12
PF
2
1.29
67.7
27.6
4.7
4.9
1.91
63.2
7.0
1.55
11
PF
2
1.28
45.7
41.9
12.4
5.0
1.72
64.0
10.6
1.42
4
PF
2
1.15
56.5
35.3
8.2
5.1
2.31
86.6
4.0
1.25
1
SF
1
1.07
60.7
31.8
7.5
4.7
1.35
74.1
7.6
1.10
5
PF
3
1.22
70.1
23.5
6.4
4.9
0.99
82.3
2.3
1.00
2
SF
2
1.11
57.5
35.3
7.2
4.6
2.34
94.9
3.4
1.00
8
SF
3
1.39
60.9
36.1
3.0
4.3
1.83
87.5
trace
0.97
13
SF
2
1.42
74.6
20.7
4.7
4.3
1.94
83.4
6.1
0.95
9
SF
3
1.17
45.0
45.4
9.6
4.3
1.20
94.5
2.0
0.73
6
WL
2
1.41
71.3
25.9
2.9
4.6
1.66
94.2
4.8
-f PF = Area cleared for the first time by slash and burn from primary forest.
SF = Area cleared from the secondary forest which was left for 5-7 years for
building up of soil fertility.
ECEC = Effective cation exchange capacity (cmol+ charge kg'1 soil)
§ Mehlich I extractable phosphorus
O

91
Table 4-4. Relationship between soil characteristics with
year in crop production in different land types.
Land Type^
Soil
Characteristics
Intercept
Slope
r2
Primary
pH
5.25
-0.15
(0.12)
0.14
Forest (PF)
ECEC
3.60
-0.80
(0.38)
0.33*
A1 Sat. (%)
61.1
6.5
(0.6)
0.18
Mehlich-I (P)
10.3
-2.15
(1.29)
0.23
Secondary
pH
4.75
-0.17
(0.09)
0.38*
Forest (SF)
ECEC
2.33
-0.16
(0.15)
0.03
A1 Sat. (%)
81.5
4.0
(1.6)
0.20
Mehlich-I (P)
8.5
-2.60
(0.72)
0.72*
-f No. of observations in PF were 11, and in SF 7.
Numbers in the parenthesis are standard error of slope
estimates.

92
Table 4-5. Relationship between soil characteristics
measured after treatment application with grain yield
for maize and cowpea crops in the municipality of Rio
Preto da Eva, Amazonas, Brazil'.
Partial
p
Treatments r a b
Maize
FP (A1 Sat.)
0.79**
6.39
-0.06
(0.04+)
PCW20+TSP20 (pH)
0.35*
-3.03
0.81
(0.14)
CM20+TSP20 (P)
0.42*
-0.52
0.34
(0.06)
TSP40 (P)
0.77**
0.42
0.34
(0.07)
COWPEA
FP (pH)
0.70**
-1.69
0.97
(0.11)
PCW20+TSP20 (P)
0.59*
0.21
0.14
(0.03)
TSP40 (P)
0.38*
1.03
0.10
(0.02)
CM20+TSP20 (All)
X
X
X
X
-(- Soil characteristics were measured 65 d after treatment
application for maize, and 45 d for cowpea.
The numbers in the parenthesis are the standard errors of
b estimates. Variables inside the parenthesis are soil
characteristics.
The values reported for r2, a, and b are for those
variables which had a strong relationship with crop
production.
x None of the variables had a strong relationship with crop
production.

93
initiation, and impair nodule function (Keyser and Munns,
1979). For cowpea, each unit increase in soil pH increased
production by 0.97 Mg ha1 for FP. The effect of P from
different amendments on cowpea yield varied. The intercept
values of 0.21 and 1.03 Mg ha'1 for PCW20+TSP20, and TSP40
along with 0.14 and 0.10 slope values indicated that same
amount of P from these two amendments had a markedly
different effect on cowpea yield. Yield variation was
attributed to differences in N, Ca, Zn, and Cu content in
amendments and high C:P and N:P ratios, which are explained
in Chapter II and III.
Land Types and Recommendation Domains
Soil characteristics for maize and cowpea trials for
different RDs are presented in Fig. 4-2 and 4-3. For maize,
e<1.85 Mg ha1 was considered RD1 (poor environments), and
e>1.85 Mg ha'1 RD2 (good environments). For cowpea e<1.32 Mg
ha1 was considered RD1, and e>1.32 Mg ha1, RD2. For both
crops in RD1, soil characteristics such as pH, Mehlich-I
extractable P, and ECEC values were lower than RD2, and A1
saturation was higher. Locations with favorable soil
characteristics for crop production (RD2) were in first or
second year of cultivation after clearing from PF, and first
year of cultivation after clearing from SF.

94
O 4.0 8.0 12.0 16.0
e<1.85 Mg/ha e>1.85Mg/ha
Figure 4-2. Range of different soil characteristics by
recommendation domains for maize trials.

95
O 4.0 8.0 12.0 16.0
e<1.32 Mg/ha e>1.32 Mg/ha
Figure 4-3. Range of different soil characteristics by
recommendation domains for cowpea trials.

96
Maize Experiment
Results of combined analysis of variance
Results of the ANOVA for the maize experiments are
presented in Table 4-6. According to the F-test (P>0.05)
treatments influenced yield at all locations. Treatment
CM20+TSP20, according to DMRT (Table 4-7), was superior to
all tested treatments at all locations (0.65-4.40 Mg ha'1)
and the FP failed or produced the lowest yield (maximum 0.25
Mg ha1) .
A combined analysis of variance (CANOVA) based on the
chi-sguare test at the 5% level of significance (meaning
that variances from all five locations are homogeneous), and
the random effect model (locations were selected randomly),
is presented in Table 4-8. Only the 5 locations with
replications were used in CANOVA. The presence of treatment
x location interaction hindered making a statistically valid
statement for a given treatment for all locations, even
though CM20+TSP20 outperformed other treatments as
indicated by the DMRT test.

97
Table 4-6. Summary of ANOVA for multilocatonal maize
testing in the municipality of Rio preto da Eva,
Amazonas, Brazil.
Source df EMS
LI
L2
L3
L4
L5
Block
1
0.010ns
0.080ns
0.020ns
1.361*
0.151ns
Trt
3
3.270**
8.218**
0.190**
2.271*
6.418*
Error
3
0.008
0.067
0.010
0.101
0.411
Total 7
*, ** significantly different at 0.05 and 0.01 level of
probability, ns = not significant.
Table 4-7. Duncan Multiple Range Test (DMRT) for maize crop
in the municipal of Rio preto da Eva, Amazonas,
Brazil .t.
Trt
LI
L2
L3
L4
L5
CM20+TSP20
2.85a+
4.40a
0.65a
2.80a
3.60a
TSP40
1.3 0b
3.40b
0.15b
1.60b
3.40b
PCW20+TSP20
0.15c
1.10c
0.01b
1.lObc
0.70b
FP
0.15c
0. Old
0.01b
0.25c
0.15b
CV%
8.0
11.6
50.0
22.1
32.7
f Locations with single replications have been dropped from
the ANOVA.
4= Means in the same column followed by the same letter are
not significantly different at the 95% level of probability
as determined by Duncan Multiple Range Test.

98
Table 4-8. Combined Analysis of Variance for maize grain
yield in the municipal of Rio Preto da Eva, Amazonas,
Brazil.
Source
DF
EMS
F Value
Pr >
BLOCK
1
0.21
LOC
4
5.04
14.2
0.02
BLOCK*LOC
4
0.35
TRT
3
15.81
13.9
0.02
TRT*LOC
12
1.13
9.5
0.01
BLOCK*TRT*LOC
15
0.11
f Chi-square test for homogeneity of variances was not
significant (P<0.05). All 5 replicated locations were
included in the combined analysis.
Random model. LOC was tested against BLOCK*LOC, TRT against
TRT*LOC, and TRT*LOC against BLOCK*TRT*LOC.

99
Modified stability analysis
Yield. An environmental index (e) was computed for all
environments (Table 4-9). Location 3 was the poorest
environment (mean e for two replications = 0.20 Mg ha'1). It
had a very high Al saturation (95%), low ECEC (1.35 cmol
( + ) kg'1), very low water pH (3.9), and only a trace of
Mehlich-I extractable P (Table 4-2). Farmers classified this
location as WL, obviously with good reasons. Scattered
occurrence of WL has led many scientists to be skeptical
about permanent agricultural development on highly leached
soils of the tropics (McNeill, 1964; Sioli, 1980). But the
latest advancements with long term fertility experiments
have offered potential for sustained production with proper
management (Sanchez et al., 1982).
The value of e for the best environment (location 7)
was 3.10 Mg ha'1. This location was recently cleared from
primary forest and was in the first year of cultivation.
Soil pH was very favorable (pH H20 = 5.2) with a ECEC of
4.21 cmol (+) charge kg'1 soil, and Al saturation of only
58%.
The MSA indicated that PCW20+TSP20 performed poorly in
all environments compared to the other treatments with
amendments based on maize grain production (Fig. 4-4a).
However, an exponential decrease in production under FP was
observed with decline in environmental guality. The low

Table 4-9. Environmental index (e) (Mg ha'1) for maize production in the Municipal
of Rio Preto da Eva, Amazonas, Brazil.
TRT
Loci
blkl blk2
Loc2
blkl blk2
loc3
blkl
blk2
loc4
blkl blk2
loc5
blkl blk2
loc6
blkl
loc7
blkl
loc8
blkl
FP
0.10
0.20
0.00
0.00
0.00
0.00
0.10
0.40
0.20
0.10
2.20
2.50
0.20
PCW20+TSP20
0.10
0.20
1.20
1.00
0.00
0.00
0.40
1.80
0.90
0.50
1.00
1.40
0.70
TSP4 0
1.20
1.40
3.20
3.60
0.20
0.10
1.20
2.00
3.00
3.80
4.20
4.50
3.50
CM20+TSP20
2.90
2.80
4.10
4.70
0.80
0.50
2.40
3.20
4.30
2.90
3.60
4.00
4.00
(e) by rep
1.07
1.15
2.13
2.33
0.25
0.15
1.02
1.85
2.10
1.82
2.75
3.10
2.10
(e) by loc
1.11
2
.23
0.
20
1.43
1
.96
2.75
3.10
2.10
100

101
production with the PCW20+TSP20 treatment could be due to
the immobilization of the applied inorganic P in close
proximity to the PCW. Low P content, high loading rate, high
metal concentration, high C:N, and C:P ratios were
apperently responsible for low P release, and high P
fixation. The performance of CM20+TSP20 was better than
other tretaments in all environments (Fig. 4-4a). The
production level of 1.85 Mg ha'1 was identified as a minimum
acceptable level. Based on the minimum acceptable production
level of 1.85 Mg ha'1 all environments were divided into two
RDs. Within both RDs the probability of yield falling below
1.85 Mg ha1 for each treatment was examined with the aid of
a confidence interval (Cl) test (Fig. 4-4b). Thus, even with
CM20+TSP20 there is 17% chance of yield less than 1.85 Mg
ha"1 in SF2 and WL (RD1) and well over 50% with TSP4 0. Using
these two treatments in PF1, 2 and SF1 (RD2), there is
virtually no probability of yield less than 1.85 Mg ha1.
On the other hand FP only reaches 1.85 Mg ha'1 in PF1 which
falls in RD2.

Yield, Mg ha"1
102
WL
SF 2 PF 2
PF 1
SF 1
PF 1
Environmental Index (e), Mg ha 1
Figure 4-4a. Response of different treatments to
environmental index for maize production, Rio Preto da
Eva, Amazonas, Brazil.

1.85 minimum acceptable level
Yield, Mg/ha
Figure 4-4b. Distribution of confidence intervals for maize
production in poor (e<1.85 mg ha*1) and good (e>1.85
mg ha"1) environments.

104
Net income/cash cost. The net income/cash cost ratio is
an important criterion for selection of improved technology
which requires additional cash input. The information
presented in Appendix A indicated that FP had the highest
return over invested cash. One dollar invested gave a return
of US $ 32.3 at a production level of 3.10 Mg ha'1 in the
best environments. The relationship between e and net
return/cash cost ratio followed an exponential function
(Fig. 4-5a). The return over invested cash for other
treatments was less than $ 6.3 in all environments.
Processed city waste treatment ( PCW20+TSP20) had a negative
return at all production levels. The distribution of
confidence intervals for net income/cash cost is presented
in Fig. 4-6a. The mean net return from a dollar invested in
CM20+TSP20 was $ 3.2 and ranged from $ 2.4 to 4.2. It was
observed that 99 % of the time a farmer will receive higher
than 2.4 $/$ investment. The average return from a dollar
invested in FP was $ 18. However there is 18 % chance that
FP will provide a return of less than $ 2.4.
Net income/total cost. The net income/total cost ratio
indicated that return over total cost was low for all
treatments. However, as the production environments improved
(e> 1.85 Mg ha1) the net return to total cost was within
the range of $ 1.8 to 3.5 (Fig. 4-5b). Chicken manure
(CM20+TSP20) provided stable return which was within the

105
range of $ 1.8 to 2.1 per dollar spent and had no
probability of a loss (Fig. 4-6b).
Net income. An analysis of net income ha'1 under four
tested treatments (Appendix A, and Fig. 4-5c) indicated that
TSP40, and CM20+TSP20 were superior to FP, and PCW20+TSP20
in all environments. Net income from TSP40, and CM20+TSP20
at their highest production levels were $ 622 and 513,
respectively compared to $ 388 and 16.5 for FP, and
PCW20+TSP20. A confidence interval was calculated for FP and
CM20+TSP20 treatments based on net income (Fig. 4-6c). Two
treatments not considered for Cl analysis were PCW20+TSP20,
TSP40. Processed city waste (PCW20+TSP20) had the lowest net
income, and TSP40 had agronomic performance similar to that
of CM20+TSP20. The confidence interval test indicated that
99 % of the time CM20+TSP20 will produced a net income
greater than $ 400 compared to a negative net income to an
income of $ 380 for FP practice (Fig. 4-6c). The narrow
range of net income with CM20+TSP20 treatment compared to FP
indicated that CM20+TSP20 provided less risk than FP, and
had a high net income ha'1.

106
Figure 4-5. Relationship of net income/cash cost, net
income/total cost, and net income with environmental
index, in on-farm maize trials from Rio Preto da Eva,
Amazonas, Brazil.
Treatment
Intercept
b SE Y Esti
Net Income/Cash
Cost
FPf
0.1
1.85
0.42
PCW20+TSP20
-1.0
0.30
0.21
TSP40
1.5
2.71
0.56
CM20+TSP20
0.5
1.42
0.80
Net
Income/Total
Cost
FP-f
0.2
0.83
0.50
PCW20+TSP20
-1.0
0.30
0.21
TSP40
-0.5
1.00
0.33
CM20+TSP20
0.3
0.61
0.38
Net Income ha"
1
FPf
1.0
1.8
0.8
PCW20+TSP20
-217.0
71.3
44.3
TSP40
-142.9
261.0
55.0
CM20+TSP20
60.7
177.5
101.1
f Exponential relationship
SE = Standard Error

Net Income/Cash Cost ($)
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5
Environmental Index (e), Mg/ha
Environmental Index (e), Mg/ha
Values are expressed in US $

Cofidence Coefficient (%)
108
Net Income/Cash Cost
Net Income ($/ha)
Figure 4-6. Distribution of confidence intervals for net
income/cash cost, net income/total cost, and net income
for different treatments used for maize cultivation.

109
Cowpea Experiment
Results of combined analysis of variance
Cowpea response was evaluated at 13 locations. Eight
were replicated and five were not. The ANOVA and DMRT for
locations are presented in Table 4-10 and 4-11. According to
the F-test, treatments did not influence yield (P>0.05) at
three locations. At location 1, CM20+TSP20 was as good as
TSP40 while CM20+TSP20 did not differ from PCW20+TSP20
(Table 4-12). A CANOVA based on a similar computation as for
maize is presented in Table 4-9. A significant chi-square
test (chi-square = 11.78) suggested the elimination of all
locations with cv > 20%. Therefore, this test was carried
out with only three of the 8 replicated locations. The
treatment x location interaction was found to be significant
(P <0.03). Gomez and Gomez (1985) suggested the partitioning
of treatment x location interaction using either the
homogeneous site approach or the homogeneous treatment
approach. However, by rejecting 10 locations out of 13, it
was felt that the locations were no longer a true
representation of the population.

Table 4-10. Summary of ANOVA for multilocatonal cowpea testing.
Source
Df
LI
L2
EMS
L3
L4
L5
L6
L7
L8
Block
1
0.08ns
0.55**
0.36ns
0.10ns
0.00ns
0.10ns
0.00ns
0.36ns
Trt
3
1.13*
0.41**
0.37ns
0.96*
2.08*
0.74ns
0.43ns
1.21**
Error
3
0.09
0.00
0.15
0.05
0.12
0.15
0.18
0.01
Total
7
*, ** F-test significantly different at 0.05 and 0.01 level of probability, ns = not
significant.
Table 4-11. Duncan Multiple Range Test (DMRT) for cowpea trials conducted in the
municipal of Rio Preto da Eva, Amazonas.
Trt
LI
L2
L3
L4
L5
L6
L7
L8
1
(aiTciin yield, ng na
CM2 0+TSP2 0
1.80ab
1.50a
1.90a
2.25a
2.05a
1.35a
2.65a
1.65a
TSP40
2.30a
1.10b
2.50a
1.60ab
2.10a
1.35a
2.15a
1.30a
PCW20+TSP20
0.90bc
0.65c
1.95a
1.20bc
0.50b
0.50a
1.70a
0.2 0b
FP
0.70c
0.50c
1.45a
0.60c
0.15b
0.15a
1.65a
0.10b
CV%
21.4
7.2
20.3
16.5
30.0
46.4
21.0
14
tFive locations have been dropped due to single replicates.
Means in the same column followed by the same letter are not significantly
different at the 95% level of probability as determined by Duncan Multiple Range
Test.
o

Ill
Table 4-12. Combined Analysis of Variance^ for
cowpea experiments.
Source
DF
EMS
F Value
Pr > F
BLOCK
1
0.07
LOC
2
0.80
1.7
0.20
BLOCK*LOC
2
0.47
TRT
3
2.40
26.6
0.00
TRT*LOC
6
0.09
4.7
0.03
BLOCK*TRT*LOC 9
0.02
-f Chi-square test for homogeneity of variances was
significant. All sites with coefficient of variation > 20%
were excluded from the combined analysis.
Random model. LOC was tested against BL0CK*L0C, TRT against
TRT*LOC, and TRT*L0C against BL0CK*TRT*L0C.

112
Modified stability analysis
Yield. For yield, it was observed that the CM20+TSP20
treatment was stable (slope = 0.03, compared to 1.13 for
TSP40) over all environments (e values 0.5-2.0 mg ha1).
Treatment CM20+TSP20 had a clear advantage over TSP40 in e <
1.32 Mg ha'1. But in e > 1.32 Mg ha'1 TSP40 outyielded
CM20+TSP20 (Fig. 4-7a). The reason for such 'cross-over'
interaction was believed to be the additional input of N
from CM. Excess N in the better environments caused cowpea
plants to grow taller and become susceptible to lodging. In
poor environments this phenomenon had no significant bearing
on plant growth due to inherent N deficiency in the soil.
The point of 'cross-over' interaction (1.32 Mg ha'1)
was taken as a reference point for the delineation of RDs.
All locations having e values < 1.32 (PF3; SF1, 2, 3; WL)
were grouped into RD1 (poor environment), and locations
having e values > 1.32 (PF 1, 2) were clustered into RD2
(good environment). The confidence limit test (Fig. 4.7b)
indicated that CM20+TSP20 was a stable performer in both
RDs. The confidence limits for the CM20+TSP20 treatment fall
within 1.4 to 2.1 Mg ha'1 in RD1, and 1.8 to 1.9 Mg ha1 in
RD2 compared to 1.0 to 1.8 Mg ha1 in RD1 and 2.1 to 2.8 Mg
ha1 in RD2 for TSP40. The 'cross-over' interaction along
with confidence limit test indicated that CM20+TSP20
treatment should be recommended in RD1 and TSP40 in RD2.

113
Net income/cash cost. The results of MSA for net
income/cash cost showed that FP had the highest return over
invested cash in environments with e value >1.32 Mg ha 1
(Fig. 4-8a). A return of as high as 37.7 $/$ cash investment
was obtained with FP compared to less than $ 10 for other
treatments in all environments. The confidence interval
calculation is presented in Fig. 4-9a-b. This test suggested
that in poor environments CM20+TSP20 provided a mean return
of $ 6.0 on each $ invested. This mean fell within the range
of $ 4.0 to 8.0. Farmer practice had a wider Cl. Therefore,
the return over invested cash in CM20+TSP20 in poor
environments was better and less risky compared to other
treatments. However, in good environments FP was superior to
other treatments.
Net income/total cost. The point where FP, TSP40, and
CM20+TSP20 crossed was at an e value of 1.32 Mg ha1. Below
this value the net return/total cost ratio was equal for
TSP40, and CM20+TSP20 treatments (within $ 2.8-3.0) (Fig. 4-
8b) and they were higher than FP. Processed city waste
(PCW20+TSP20) gave the lowest net return/total cost ratio in
all environments which suggests that a farmer will lose
money by applying PCW20+TSP20 for cowpea production in these
environments. In good environments performance of all
treatments improved. However, FP gave the highest return
over total cost.

114
The Cl calculation (Fig. 4-10a-b) indicated that
CM20+TSP20 gave higher, and stable return in poor
environments compared to other treatments. At 90% Cl, FP
produced from $ 0 to 2.2 compared to $ 2.8 to 3.8 for
CM20+TSP20. But FP was the best performer in good
environments and provided a return of $ 4.2 to 5.8 for each
$ invested in total cost.
Net income. The value of net income ha1 from different
environments was regressed against e (Fig. 4-8c). The net
income was higher for TSP40 treatment in good environments.
But CM20+TSP20 was superior to TSP40 in poor environments.
As the environment for cowpea cultivation improved the net
income from FP was also increasing. Processed city waste
treatment (PCW20+TSP20) was always inferior to other
treatments. Confidence intervals were calculated for FP,
TSP40, and CM20+TSP20 (Fig. 4-lla-b).
Based on this criterion FP was inferior to TSP40 in
good environments, and to CM20+TSP20 in poor environments.
At 90% Cl the value of net return for TSP40 fell in the
range of $ 820-1080 ha'1 compared to $ 300 to 750 for FP.

Yield, Mg ha -1
115
WL
SF 3 PF 3
SF 2 SF 1
PF 2
PF 1
Environmental Index (e), Mg ha 1
Figure 4-7a. Response of different treatments to
environmental index for cowpea production, Rio Preto da
Eva, Amazonas, Brazil.

Confidence Interval, %
116
Yield, Mg/ha
Figure 4-7b. Distribution of confidence intervals for cowpea
production in poor (e<1.32 mg ha'1), and good (e>1.32
Mg ha'1) environments.

117
Figure 4-8. Relationship of net income, net income/cash cost
and net income/total cost with environmental index, in
on-farm cowpea trials from Rio Preto da Eva, Amazonas,
Brazil.
Treatment
Intercept
b
SE Y Estimate
Net
Income/Cash Cost
FP
-20.8
24.8
8.36
PCW20+TSP20
-3.0
3.2
2.87
TSP40
5.7
0.6
0.12
CM20+TSP20
11.3
-2.3
2.14
Net
Income/Total
Cost
FP
-2.5
4.3
5.9
PCW20+TSP20
-2.0
2.1
1.8
TSP40
1.8
1.2
1.2
CM20+TSP20
3.0
0.2
0.1
Net Income ha1
FP
-521.0
620.4
226.6
PCW20+TSP20
-558.0
597.3
238.4
TSP40
20.7
481.1
114.5
CM20+TSP2 0
604.1
60.7
23.8
SE = Standard Error

118
Net Income/Cash Cost ($/$)
0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2
Environmental Index (e), Mg/ha
Net Income, $/ha
Environmental Index (e), Mg/ha
0.6
2.2

Confidence Coefficient (%)
Net Income/Cash Cost (US $)
Net Income/Cash Cost (US $)
Figure 4-9. Distribution of confidence intervals for net
income/cash cost for selected treatments in poor
(e<1.32 Mg ha 1) and good (e>1.32 Mg ha1) environments
for cowpea cultivation.

120
Net Income/Total Cost (US $)
Figure 4-10. Distribution of confidence intervals for net
income/total cost for selected treatments in poor
(e<1.32 Mg ha'1) and good (e>1.32 Mg ha'1) environments
for cowpea cultivation.

121
Net Income (US$/ha)
Net Income (US$/ha)
Figure 4-11. Distribution of confidence intervals for net
income for selected treatments in poor (e<1.32 Mg ha'1)
and good (e>1.32 Mg ha'1) environments for cowpea
cultivation.

122
Technology Selection for Different Recommendation Domains
The concept of RDs is very beneficial in selecting the
nitch for a technology in the overall research domain.
However, the usefulness of the concept in field conditions
will depend on selecting technology evaluation criteria
based on the farmers' needs and aspirations, and using the
criteria to divide the research area into RDs. More often
there are more than one criteria. A summary of technology
selection for a given land type based on different
evaluation criteria is presented in Table 4-13 and 14.
Maize
The selection of appropriate technology is the function
of land type and the farmer's goal. If the goal is to obtain
highest grain production in PF1, CM20+TSP20 should be
recommended. In the same land type, if the goal is to obtain
highest return to limited cash FP, should be suggested for
use. However, one should be aware of the fact that FP is not
sustainable in long run and within a year or two the
productivity will decline to a level which will no longer
offer highest return to invested cash. In SF2, CM20+TSP20
out performed all treatments based on all evaluation
criteria under consideration. In WL, CM20+TSP20 can be
recommended based on the yield criteria, but economically
none of the tested treatments was sound.

123
Cowpea
For cowpea, all land types were divided into two
recommendation domains. In RD2 (PF1 and PF2), TSP40 should
be recommended based on yield and net income ha 1. Depending
on the evaluation criteria of return to cash cost and return
to total cost FP has advantage over other treatments. This
practice should not be continued for more than a year if
high return to invested cash is expected. In RD1 (PF3, SF1,
SF2, SF3, and WL), CM20+TSP20 is the best option based on
all evaluation criteria.
Redefining Recommendation Domains
A map of the area including location, land type, year
in crop production, and RDs based on the yield criterion is
presented in Fig 4-12. The resulting domains are not
amenable to geographical mapping because farmers of
different domains were interspersed in the area. The use of
different evaluation criteria can also lead to regrouping of
environments from one RD to another. For example, the use of
return to cash cost indicator moved one production
environment for maize and two for cowpea from RD1 to RD2
(Fig.4-13) .

124
Table 4-13. Technology selection for a given land type based
on different evaluation criteria for maize cultivation.
Land Type Evaluation Criteria
Yield
Mg/ha
Return to
Cash Cost
$/$
Return
Total
$/$
i to
Cost
Net
Income
$
PF1
CM
FP
FP
CM
SF1
CM
FP
FP
CM
PF2
CM
CM
CM
CM
SF2
CM
CM
CM
CM
WL
CM
NR
NR
NR
Table 4-14.
Technology selection
for a
given
land type
on different evaluation criteria for cowpea
cultivation.
Land Type Evaluation Criteria
Yield
Mg/ha
Return to
Cash Cost
$/$
Return to
Total Cost
$/$
Net
Income
$
PF1
TSP
FP
FP
TSP
PF2
TSP
FP
FP
TSP
PF3
CM
CM
CM
CM
SF1
CM
CM
CM
CM
SF2
CM
CM
CM
CM
SF3
CM
CM
CM
CM
WL
CM
CM
CM
CM

125
C O W P E A
Recommendation Domain I
Recommendation Domain II
Figure 4-12. Recommendation domains, based on yield, and
their relationship to land types in the municipality of
Rio Preto da Eva,Amazonas, Brazil.

126
C O W P E A
Recommendation Domain I
Recommendation Domain II
Figure 4-13. Recommendation domains, based on net return to
cash cost, and their relationship to land types in the
municipality of Rio Preto da Eva, Amazonas, Brazil.

127
Conclusion
The presence of environment-by-treatment interaction
masked the mean treatment differences measured by CANOVA. To
detect treatment effects, this analysis suggested grouping
of similar environments and a repeat of the experiment. The
slope estimates in regression analysis were useful in
predicting rates of decline in pH, ECEC, and Mehlich-I
extaractable P with the year of cultivation. A reduction of
2.1 and 2.6 g g'1 soil in Mehlich-I extractable P was
obtained annually in PF, and SF. Although an equivalent of
8.8 kg ha1 P was applied from different amendments, their
effect on maize and cowpea production varied. Phosphorus
from TSP was used more efficiently, although this treatment
reduced ECEC, and soil pH.
Modified stability analysis was conducted for maize and
cowpea using yield, net income/cash cost, net income/total
cost, and net income ha1 as technology evaluation criteria.
Different evaluation criteria led to different conclusions
regarding recommendations of appropriate technology for a
given environment. These criteria also influenced selection
of production environments for grouping into RDs. Based on
the evaluation criterion of maximizing maize production
CM20+TSP20 was recommended in RD2 (good environment) But
there was a clear advantage of FP in RD2 in terms of net
return to cash cost. If a farmer's goal is to maximize
income from his scarce cash, he ought to use farmers'

128
practice. However, FP can not be continued for more than a
year due to adverse effect on soil properties. The aluminum
saturation was increasing each year in PF by 7%, and
Mehlich-I P was decreasing by 2.1 ¡ig g'1 soil annually.
For cowpea, CM20+TSP20 outperformed all treatments in
poor environments (eel. 32 Mg ha1) while in good
environments (RD2) (e>1.32 Mg ha1) TSP40 had clear
advantage over other treatments based on all evaluation
criteria. Farmer practice had the highest return per dollar
compared to other treatments in RD2 (37.7 $/$ ). The
delineation of RDs was not amenable to geographical mapping
because fields of different domains were interspersed in the
area. A change in evaluation criteria lead to change in the
demarcation of RDs.

CHAPTER V
SUMMARY AND CONCLUSIONS
The objective of this chapter is to summarize the work
presented in the preceding four chapters. The overall goal
of this research was to examine the role of organic
amendments in sustaining P nutrition of highly leached
Oxisol and to analyse the performance of selected treatments
in farmers' fields. To achieve this goal, laboratory
incubation studies, glasshouse studies, and on-station and
off-station field studies were conducted.
Chapter II presents the findings on P adsorption and
desorption by soil as influenced by organic amendments. This
chapter also investigates a new technique for measuring
nutrient release from a decomposing material.
Incubation of soil with amendments for 35 d had a
marked influence on total P adsorption by soil. Soil amended
with PCW adsorbed more P than unamended soil. However, with
an increase in incubation time, all amendments reduced P
fixation. Phosphorus desorption values for soil incubated
for 35 d were equal to that of the control, but soil
incubated for more than 35 d showed increased P desorption.
These findings suggest that the rate of decomposition and P
129

130
content of the amendments played a key role in influencing
adsorption and desorption characteristics of the soil.
The results of the decomposition study in a soil matrix
indicated that maize and grass treatments showed P
immobilization. Processed city waste immobilized soil P at
35 and 65 d, but at 150 d a net release was observed. All
legumes followed a similar P release patterns. However, When
silica was used as a matrix there was a net release of P
from all amendments independent of P content and C:P ratio.
The data on P adsorption maxima calculated by the
Langmuir equation for CM, based on P release from amendments
in soil and silica matrices, indicated that with increasing
incubation time, P adsorption maxima increased. Continued
adsorption beyond adsorption maxima demonstrated the
presence of a precipitation reaction, or multilayer
adsorption. For tephrosia and kudzu adsorption maxima
decreased with time. A good agreement between adsorption
maxima and actual adsorption of P based on SSMT was observed
at the higher equilibrating solution P concentration (60 ¡iq
g'1 soil) which indicated the validity of the SSMT technique
in predicting P requirements of amended soils.
Chapter III examines the results of glasshouse and on-
station field studies on the effect of a suboptimal dose of
organic amendments in sustaining crop P nutrition and
improving the efficiency of inorganic P.

131
An analysis of variance for maize herbage dry weight
from the glasshouse study, indicated the presence of an
interaction among type and rate of amendments. Among three
amendments tested, PCW produced the lowest yield. It is
interesting to note that when PCW was applied at a higher
rate (equivalent of 17.6 kg P ha'1) the result was inferior
compared to the lower rate (equivalent of 8.8 kg P ha'1).
The highest dry matter yield per pot was obtained with CM.
A single degree of orthogonal contrast for maize grain
production during the first crop showed that application of
PCW equivalent to 2 6.4 kg ha"1 of P produced 1.41 Mg ha'1 of
maize. This level of production was inferior to the control
treatment. No significant differences were found among
canavalia, mucuna, and TSP when applied to provide P at the
rate of 26.4 kg ha'1. The CM treatment outyielded all
treatments in first as well as second crops.
Combined application of 8.8 kg P ha'1 from organic
amendments and the same amount from TSP was compared with
17.6 kg P ha'1 from TSP. For the first crop the combination
with CM was superior compared to TSP and an improvement of
30% in yield was recorded.
The change in P status of soil over 240 d of cropping
cycle for all but PCW treatment followed a cubic surface
response curve. All treatments improved P status of soil
compared to the control. As the cropping season progressed,
there was a sharp decrease in soil P with TSP60 compared to

132
M60 and C60 (Canavalia ensiformes). Processed city waste
(8.8 kg ha '1) applied with TSP20 was inferior to all
treatments except the control.
The results obtained from glasshouse and on-station
field studies suggested that organic amendments can provide
sustained crop P nutrition compared to inorganic sources.
However, the improved efficiency of inorganic P when applied
with an organic amendment will depend on chemical
characteristics of the organic amendments.
In Chapter IV the results of farmers' field trials are
presented. Different criteria were used for the selection of
appropriate technology for farmers and delineation of RDs.
Based on the evaluation criterion of maximizing production
CM20+TSP20 was recommended in good environments (e>1.85 Mg
ha1) for maize. For cowpea, CM2 0+TSP2 0 outperformed all
treatments in poor environments (e<1.32 Mg ha1) while in
good environments (e>1.32 Mg ha1) TSP40 had clear
advantage. Farmer practice had the highest return over cash
cost compared to other treatments for both crops.
The result of CANOVA for maize and cowpea grain
production provided inconclusive results because of the
presence of environment x treatment interaction. The slope
estimates of the regression analysis provided valuable
information on changes taking place in soil fertility
following deforestation. The rate of decline in pH in PF and
SF was 0.15 unit per year. Aluminum saturation was

133
increasing at the rate of 7% per year in PF compared to 4%
in SF. The rate of decline in Mehlich-I P in both land types
was in the range of 2.1-2.6/zg"1 soil. A relatively low ECEC
in both land types (2.3-3.6 cmol (+) charge kg1 soil)
indicated that most cations were leached following heavy
rainfall. And those present were being taken up by
vegetation or being washed out at the rate of 0.80 and 0.16
cmol (+) charge kg1 soil every year from PF and SF,
respectively.
The data presented in Chapter IV suggested that the
delineation of RDs was dependent on the technology
evaluation criteria, and was not amenable to geographical
mapping because fields of different domains were
interspersed. A change in evaluation criteria led to a
change in the demarcation of RDs. It was also observed that
a complete restoration of soil fertility by letting
secondary forest take over for 5-7 years was unlikely to
happen in the rain forest.

134
APPENDIX A
ECONOMIC ANALYSIS FOR COWPEA
Yield, income, and cost from the cowpea trials, Rio Preto da
Eva, Amazonas, Brazil. (US $)
Yield Gross Total Net Net Income/
Mg ha 1 Income Cost Income Total Cost
FP
0.70
350
67
283
4.2
0.50
250
55
195
3.5
1.45
725
112
613
5.5
0.60
300
61
239
3.9
0.15
75
34
41
1.2
0.15
75
34
41
1.2
1.70
850
127
723
5.7
0.10
50
31
19
0.6
0.20
100
37
63
1.7
2.20
1100
157
943
6.0
1.20
600
97
503
5.2
1.50
750
115
635
5.5
0.00
0
25
-25
-1.0
PCW20+TSP20
0.90
450
239
211
0.9
0.65
325
224
101
0.5
1.95
975
302
673
2.2
1.20
600
257
343
1.3
0.50
250
215
35
0.2
0.50
250
215
35
0.2
1.65
825
284
541
1.9
0.20
100
197
-97
-0.5
0.40
200
209
-9
-0.0
1.90
950
299
651
2.2
1.50
750
275
475
1.7
1.80
900
293
607
2.1
0.00
0
185
-185
-1.0

Appendix A contd
TSP40
2.30
1150
230
920
4.0
1.10
550
158
392
2.5
2.50
1250
242
1008
4.2
1.60
800
188
612
3.3
2.10
1050
218
832
3.8
1.35
675
173
502
2.9
2.65
1325
251
1074
4.3
1.30
650
170
480
2.8
1.20
600
164
436
2.7
2.60
1300
248
1052
4.2
2.20
1100
224
876
3.9
2.10
1050
218
832
3.8
1.30
650
170
480
2.8
CM2 0+TSP2 0
1.80
900
211
689
3.3
1.50
750
193
557
2.9
1.90
950
217
733
3.4
2.25
1125
238
887
3.7
2.05
1025
226
799
3.5
1.35
675
184
491
2.7
2.15
1075
232
843
3.6
1.65
825
202
623
3.1
1.70
850
205
645
3.1
1.40
700
187
513
2.7
1.90
950
217
733
3.4
1.70
850
205
645
3.1
2.00
1000
223
777
3.5

136
APPENDIX B
ECONOMIC ANALYSIS FOR MAIZE
Yield, income, and cost from the maize
trials, Rio Preto da Eva, Manaus, Brazil. (US$)
Yield Gross Total Net Net Income/
Mg ha"1 Income Cost Income Total Cost
FP
0.15
30
18
12
0.7
0.00
0
12
-12
-1.0
0.00
0
12
-12
-1.0
0.25
50
22
28
1.3
0.15
30
18
12
0.7
2.20
440
100
340
3.4
2.50
500
112
388
3.5
0.20
40
20
20
1.0
PCW20+TSP20
0.15
30
213
-183
-0.9
1.10
220
251
-31
-0.1
0.00
0
207
-207
-1.0
1.10
220
251
-31
-0.1
0.70
140
235
-95
-0.4
1.00
200
247
-47
-0.2
1.40
280
263
16
0.1
0.70
140
235
-95
-0.4
TSP40
1.30
260
150
110
0.7
3.40
680
234
446
1.9
0.15
30
104
-74
-0.7
1.60
320
162
158
1.0
3.40
680
234
446
1.9
4.20
840
266
574
2.2
4.50
900
278
622
2.2
3.50
700
238
462
1.9

137
Appendix B contd
CM20+TSP20
2.85
570
241
329
1.4
4.40
880
303
577
1.9
0.65
130
153
-23
-0.2
2.80
560
239
321
1.3
3.60
720
271
449
1.7
3.60
720
271
449
1.7
4.00
800
287
513
1.8
4.00
800
287
513
1.8

REFERENCE LIST
Abalu, G.O.I., N.M. Fisher, and Y. Abdullahi. 1987. Rapid
rural appraisal for generating appropriate technologies
for peasant fanners: Some experiences from Northern
Nigeria. Agri. Systems 25:311-324.
Adams, F. 1980. Interactions of phosphorus with other
elements in soils and in plants, p. 655-680. In M.
Stelly and R. C. Dinauer (ed.) The role of phosphorus
in agriculture. ASA, CSSA, and SSSA, Madison, WI.
Anghinoni, I., and S.A. Barber. 1980. Phosphorus application
rate and distribution in the soil and phosphorus uptake
by corn. Soil Sci. Soc. Am. J. 44:1041-1044.
Appelt, H., N.T. Coleman, and P.F. Pratt. 1975. Interaction
between organic compounds, minerals and ions in
volcanic-ash derived soils. II. Effects of organic
compounds on the adsorption of phosphate. Soil Sci.
Soc. Am. Proc. 39:628-630.
Bache, B.W. 1964. Aluminum and iron phosphate studies
related to soils. II. Reaction between phosphates and
hydrous oxides. J. Soil Sci. 15:110-116.
Barrow, N.J. 1974. The effect of previous additions of
phosphate on phosphate adsorption by soils. Soil Sci.
118:82-89.
Barrow, N.J. 1983. A mechanism model for describing the
sorption and desorption of phosphate by soil. J. Soil
Sci. 32:555-570.
Becker, H.C. 1981. Correlations among some statistical
measures of phenotypic stability. Euphytica 30:835-840.
Bell, L.C., and C.A. Black. 1970. Comparison of methods for
identifying crystalline products produced by
interaction of orthophosphate fertilizers with soils.
Soil Sci. Soc. Am. Proc. 34:579-582.
138

139
Blair, G. J., and O. W. Boland. 1978. The release of
phosphorus from plant material added to soil. Aust. J.
Soil Res. 16:101-111.
Bowman, R.A., and S.R. Olsen. 1985. Assessment of phosphate
buffering capacity: 2. Greenhouse methods. Soil Sci.
140:387-392.
Brown, P. L., and D. D. Dickey. 1970. Losses of wheat straw
residue under simulated conditions. Soil Sci. Soc.
Amer. Proc. 34:118-121.
Burrough, P.A. 1983. Multiscale source of spatial variation
in soil: I. The application of fractal concepts to
nested level of soil variation. J. Soil Sci. 34:577-
597.
Byerlee, D., L. Harrington, and M. Collinson. 1980. Planning
technologies appropriate for farmers: Concepts and
procedures. International Maize and Wheat Improvement
Center (CIMMYT), Mexico.
Byerlee, D., L. Harrington, and D.L. Winkelmann. 1982.
Farming Systems Research: Issues in research strategy
and technology design. Am. J. Agri. Econ. 64:897-904.
Chang, S.C., and M.L. Jackson. 1957. Fractionation of soil
phosphorus. Soil Sci. 84:133-144.
Cochran, V. L., L. F. Elliot, and R. I. Papendick. 1980.
Carbon and nitrogen movement from surface applied wheat
straw. Soil Sci. Soc. Amer. J. 44:978-982.
de Wit, C.T. 1953. A physical theory on placement of
fertilizers. Versl. Landbouwk. Onderzoek. No. 59.4.
Douglas, Jr. C. L., R. R. Allmaras, P. E. Rasmussen, R. E.
Ramig, and N. C. Roger, Jr. 1980. Wheat straw
composition and placement effects on decomposition in
dryland agriculture of the pacific Northwest. Soil Sci.
Soc. Amer. J. 44:833-837.
Easterwood, G.W. and J.B. Sartain. 1990. Organic coatings on
P fertilizers: Influence on plant growth on a Florida
Ultisol (in review). Soil and Crop Sci. Soc. Fla. Proc.
49:
Eberhart, S.A., W.A. Russell. 1966. Stability parameters for
comparing varieties. Crop Sci. 6:36-40.

140
Elliot, L. F., and J. W. Blaylock. 1975. Effects of wheat
straw and alfalfa amendments on solubilization of
manganese and iron in soil. Soil Sci. 120:205-211.
Empresa Brasileira de Pesquisa Agropequaria. 1976.
Composisao da solucao nutritiva para milho. Sete
Lagoas, CNPMS, MG, Brazil.
Empresa Brasileira de Pesquisa Agropequaria. 1979. Guia de
excursao XVII Cogreso Braziliero de sciencia do solo,
Manaus, EMBRAPA. Servicio Nacional de Levantamento e
Conservacao do solos, Rio de Janeiro, Brazil.
Empresa Brasileira de Pesquisa Agropequaria. 1984. Boletim
agrometeorologico no. 6. EMBRAPA, UEPAE de Manaus,
Manaus, Brazil.
Evans, C.E., and E.J. Kamprath. 1970. Lime response as
related to percent A1 saturation, solution Al, and
organic matter content. Soil Sci. Soc. Am. Proc.
34:893-896.
Fattori, T.R., F.B. Mather, P.E. Hildebrand. 1990.
Methodology for partitioning poultry producers into
recommendation domains. Agri. Systems. 32:197-205.
Fox, R.L., and E.J. Kamprath. 1970. Phosphate sorption
isotherms for evaluating the phosphate requirements of
soils. Soil Sci. Am. Proc. 34:902-906.
Fox, R.L., and B.T. Kang. 1978. Influence of phosphorus
fertilizer placement and fertilization rate of maize
nutrition. Soil Sci. 125:34-40.
Fox, R.L., W.M.H. Saunders, S.S.S. Rajan. 1986. Phosphorus
nutrition of pasture species: Phosphorus requirement
and root saturation values. Soil Sci. Soc. Am. J.
50:142-148.
Francis, C.A., and P.E. Hildebrand. 1989. Farming systems
research/ extension and the concepts of sustainability,
p. 1-8. Ninth annual farming systems symposium, Univ.
Arkansas, Fayetville, AR. 8-11 Oct., 1989.
Gilman, G.P. 1981. Effects of pH and ionic strength on the
cation exchange capacity of soils with variable charge.
Aust. J. Soil Res. 19:94-96.
Giordano, P.M., J.J. Mortvedt, and D.A. Mays. 1975. Effect
of municipal waste on crop yields and uptake of heavy
metals. J. Environ. Qual. 3:394-399.

141
Gomez, K.A., and A.A. Gomez. 1985. Statistical procedures
for agricultural research. John Wiley and Sons, New
York, USA.
Gunary, D. 1970. A new adsorption isotherm for phosphate in
soil. J. Soil Sci. 21:72-77.
Hammond, L.L., S.H. Chien, and A.U. Mokwunye. 1986.
Agronomic value of unacidulated and partially
acidulated phosphate rocks indigenous to tropics. Adv.
Agron. 40:89-140.
Harter, R.D., and G. Smith, 1981. Langmuir equation and
alternate methods of studying "adsorption" reactions in
soils, p.167-182. In R.H. Dowdy, et al. (ed) Chemistry
in the soil environment. Spec. publ. no. 40. Am. Soc.
of Agron., Madison, WI.
Harter, R.D. 1984. Curve-fit errors in Langmuir adsorption
maxima. Soil Sci. Soc. Am. J. 48:749-752.
Harwood, R. 1979. Small farm development: Understanding and
improving farming systems in the humid tropics.
Westview Press, Boulder, Colo.
Haynes, R.J. 1982. Effects of liming on phosphate
availability in acid soils. Plant Soil 68:289-308.
Heliums, D.T., S.H. Chien, and J.T. Touchton. 1989.
Potential agronomic value of calcium in some phosphate
rocks from South America and West Africa. Soil Sci.
Soc. Am. J. 53:459-462.
Hildebrand, P.E. 1981. Combining disciplines in rapid
appraisal: The 'sondeo' approach. Agricultural
Administration 8:423-432.
Hildebrand, P.E. 1984. Modified stability analysis of
farmer-managed on-farm trials. Agron J. 76:271-274.
Hildebrand. P.E. 1986. The Sondeo: A team rapid survey
approach, p. 93-98. In P. E. Hildebrand (ed)
Perspective on farming systems research and extension.
Lynne Reiner publishers, Boulder, Colorado.
Hildebrand, P.E. 1990. Modified stability analysis and on-
farm research to breed specific adaptability for
ecological diversity. Paper presented at the symposium
on Genotype-by-Environment Interaction and Plant
Breeding. LSU, Baton Rouge.

142
Hill, R.R., Jr. and J.E. Baylor. 1983. Genotype x
environment interaction analysis in alfalfa. Crop Sci.
23:811-815.
Hingston, F.J., R.J. Atkinson, A.M. Posner, and J.P. Quirk.
1967. Specific adsorption of anions. Nature 215:1459-
1461.
Holford, I.C.R., and G.E.G. Mattingly. 1974. The high and
low-energy phosphate absorbing surfaces in calcareous
soils. J. Soil Sci. 26:407-417.
Holford, I.C.R., Wedderbern, R.W.M., and Mattingly, G.E.G.
1974. A Langmuir two-surface equation as a model for
phosphate adsorption by soils. J. Soil Sci. 25:242-255.
Hsu, P.H., and D.A. Rennie. Reactions of phosphate in
aluminum systems: II. Precipitation of phosphate by
exchangeable aluminum on a cation regin. Can. J. Soil
Sci. 42:210-221.
Hunda1, H.S., C.R. Biswas, and A.C. Vig. 1988. Phosphorus
sorption characteristics of flooded soil amended with
green manure. Trop. Agrie. 65:185-187.
Izza, C., and R. Indiati. 1982. Effect of farm organic
residues added to the soil on phosphorus sorption.
Crops and Soils. 43:78-90
Jones, J.P. and J.A. Benson. 1975. Phosphate sorption
isotherms for fertilizer needs of sweet corn (Zea mays)
grown on a high P fixing soil. Comm. Soil Sci. Plant
Anal. 6:465-477.
Jungk, A., and S.A. Barber, 1974. Phosphate uptake rate of
corn as related to the proportion of roots exposed to
phosphate. Agron. J. 66:554-557.
Kamprath, E.J. 1967. Residual effects of large application
of phosphorus on high fixing soils. Agron. J. 59:25-27.
Kamprath, E.J. 1970. Exchangeable aluminum as a criterion
for liming leached mineral soils. Soil Sci. Soc. Am.
Proc. 34:252-254.
Keng, J., and G. Uehara. 1974. Chemistry, mineralogy and
taxonomy of Oxisols and Ultisols. Soil and Crop Sci.
Soc. Fla, Proc. 33:119-126.
Khasawneh, F.E., and E.C. Doll. 1978. The use of phosphate
rock for direct application. Adv. Agron. 30:159-206.

143
Kosaki T., and S.R.J. Anthony. 1989. Multivariate approach
to grouping soils in small fields. I. extraction of
factors causing soil variation by principal component
analysis. Soil Sci. Plant Nutr. 35:469-477.
Kyuma, K., and T. Tulaphitak. 1985. Changes in soil
fertility and tilth under shifting cultivation. I.
general description of soil and effect of burning on
the soil characteristics. Soil Sci. Plant Nutr. 31:227-
238.
Larsen, S. 1967. Soil phosphorus. Adv. Agron. 19:151-210.
Larsen, J.E., G.F. Warren, and R. Langston. 1959. Effect of
iron, aluminum and humic acid on phosphorus fixation by
organic soils. Soil Sci. Soc. Proc. 438-440
Larsen, W. E., R. F. Holt, and C. W. Carlson. 1978. Residues
for soil conservation, p. 1-17. In W. R. Oswald (ed.).
Crop residue management systems. Amer. Soc. Agron. J.
64:204-208.
Leal, J.R., and A.C.X. Velloso. 1973. Dessorcao de fosfato
absorbido em latosolos sob vegetacao de cerrado. II.
Reversibilidade da isoterma de adsorbcao de fosfato em
relacao ao pH da solucao em equilibrio. Pesq. Agropec.
Bras. (Ser. Agron). 8:89-92.
Leikam, D.R., R.E. Lamond, P.J. Gallagher, and L.S. Murphy.
1978. Improving N-P application. Agrichem. Age 22(3):6.
Leikam, D.R, L.S. Murphy, D.E. Kissel, D.A. Whitney, and
H.C. Moser. 1983. Effects of N and P application method
and nitrogen source on winter wheat grain yield and
leaf tissue phosphorus. Soil Sci. Soc. Am. J. 47:530-
535.
Lin, C.S., and G. Butler. 1990. Cluster analyses for
analyzing two-way classification data. Agron. J.
82:344-349.
McNeill, M. 1964. America. Science 211:86-86
Mead, J.A. 1981. A comparison of the Langmuir, Freundlich,
and Temkin equations to describe phosphate sorption
properties of soils. Aust. J. Soil Res. 19:333.342.
Meda, L., and G. F. Cerofolini. 1989. Physical chemistry of,
in and on silicon. Springer series in material science
v. 8. Springer-Verlag Heidelberg, Germany.

144
Mekaru, T., and G. Uehara. 1972. Anion adsorption in
ferruginous tropical soils. Soil Sci. Soc. Am. Proc.
36:296-300.
Mendez, J., and E.J. Kamprath. 1978. Liming of Latosols and
the effect on P response. Soil Sci. Soc. Am. J. 41:86-
88.
Mokwunye, U. 1977. Phosphorus fertilizers in Nigerian savana
soils. I-Use of phosphorus sorption isotherms to
estimate the phosphorus requirement of maize at Samaru.
Trop. Agrie. 54:265-270.
Moormann, F.R., and B.T. Kang. 1978. Microvariability of
soils in the tropics and its agronomic implications
with special reference to West Africa, p 29-43. In
Diversity of the soil in the tropics. ASA, Madison, WI.
Moris, F.I., A.L. Page, and L.J. Lund. 1976. The effect of
pH, salt concentration, and nature of electrolytes on
the charge characteristics of Brazilian tropical soils.
Soil Sci. Soc. Am. 40:521-527.
Moreno, E.C., W.L. Lindsay, and G. Osborn. 1960. Reactions
of dicalcium phosphate dihydrate in soil. Soil Sci.
90:58-68.
Moussie, M., and C. Muhitira. 1988. Classification of
farmers into recommendation domains. Proceedings of
Farming Systems Research/Extension Symposium.
University of Arkansas, p.241-250
Muljadi, D., A.M. Posner, and J.P. Quirk. 1966. The
mechanism of phosphate adsorption by kaolinite,
gibsite, and pseudoboehmite. Part II. The location of
adsorption sites. J. Soil Sci. 17:230-237.
Munns, D.N., and R.L. Fox. 1976. The slow reaction which
continues after phosphate adsorption: Kinetics and
equilibrium in some tropical soils. Soil Sci. Soc. Am.
J. 40:46-51.
Murphy, J., and J.P. Riley. 1962. A modified single solution
method for the determination of phosphate in natural
waters. Anal. Chem. Acta. 27:31-36.
Murray, K., and P.W. Linder. 1984. Fulvic acids: Structures
and metal binding. II. Predominant metal binding sites.
J. Soil Sci. 33:217-222.

145
Nagarajah, S., A.M Posner, and J.P Quirk. 1970. Competitive
adsorption of phosphate with polyglacturonate and other
organic anions on kaolinite and oxide surfaces. Nature
228:83-84.
Nair, P.S., T.J. Logen, A.N. Sharpley, L.E. Sommers, M.A.
Tabatabai, and T.L. Yuan. 1984. Interlaboratory
comparison of a standardized phosphorus adsorption
procedure. J. Environ. Qual. 13:591-595.
Njobvu, C.A. 1986. Factors influencing recommendation domain
boundaries of the farming system and levels of
agricultural development in Lusaka province, Zambia.
Proceedings of Farming Systems Research/Extension
Symposium. Kansas State University, pp. 254-259.
Norris, J.M. 1970. Mutivariate methods in the study of
soils. Soils Frtil. 33:13-18.
Olsen, S.R. 1953. Inorganic phosphorus in alkaline and
calcareous soils. Agronomy 4:89-122.
Olsen, S.R and Watanabe, 1957. A method to determine a
phosphorus adsorption maximum of soils as measured by
the Langmuir isotherm. Soil Sci. Soc. Am. Proc.
21:144.149.
Olsen, S.R., and S.A. Barber. 1977. The effect of waste
application on soil phosphorus and potassium, p. 197-
215. In L.F Elliott and F.J. Estivenson (ed) Soils for
management of organic waste and waste waters. Am. Soc.
Agron. Madison, WI.
Parr, J. F., and R. I. Papendick. 1978. Factors affecting
the decomposition of crop residues by microorganisms,
p. 101-129. In W. R. Oschwald (ed.). Crop residues
management systems. ASA Special pub. No. 31, Amer. Soc.
Agron., Madison, WI.
Perrott, K.W. 1978. The influence of organic matter
extracted from humified clover on the properties of
amorphous aluminosilicates. II. Phosphate retention.
Aust. J. Soil Res. 16:341-346.
Posner, A.M., and J.W. Bowden. 1980. Adsorption isotherms:
Should they be split? J. Soil Sci. 31:1-10.
Rajan, S.S.S., K. W. Perrott, and W.M.H. Saunders. 1974.
Identification of phosphate reactive sites of hydrous
alumina from proton consumption during phosphate
adsorption at constant pH values. J. Soil Sci. 25:438-
447

146
Rajan, S.S.S., and R.L. Fox. 1975. Phosphate adsorption by
soils. II. Reaction in tropical acid soils. Soil Sci.
Soc. Am. Proc. 39:446-451.
Raun, W.R., D.H. Sander, and R.A. Olson. 1987. Phosphorus
fertilizer carriers and their placement for minimum
till corn under sprinkler irrigation. Soil Sci. Soc.
Am. J. 51:1055-1062.
Reddy, K.R. 1990. Phosphorus retention capacity of stream
sediments and associated wetlands. Final report
submitted to the South Florida Water Management
District. West Palm Beach, FI.
Reeve, N.G., and M.E. Summer. 1970. Effects of aluminum
toxicity and phosphorus fixation on crop growth on
Oxisol from Natal. Soil Sci. Soc. Am. Proc. 34:263-267.
Salinas, J.G., and P.A. Sanchez. 1976. Soil-plant
relationships affecting varieties and species
differences in tolerance to low available soil
phosphorus. Ciencia e Cultura (Brazil) 28(2):156-168.
Sample, E.C., R.J. Soper, and G.J. Racz. 1980. Reactions of
phosphate fertilizers in soils, p. 263-310. In M.
Stelly (ed.) The role of phosphorus in agriculture.
ASA, CSSA, and SSSA, Madison, WI.
Sanchez, P.A., and S.W. Buol. 1975. Soils of the tropics and
the world food crisis. Science 188:598-603.
Sanchez, P.A., D.E. Bandy, J.H. Villachica, and J.J.
Nicholaides. 1982. Amazon basin soils: Management for
continuous crop production. Science 216:821-827.
Sartain, J.B., and J.J. Street. 1980. Systems for supplying
micronutrients. Florida Fertilizer and Lime Conference
Proc. 10:1-21.
SAS Institute, Inc. 1985. SAS User's guide: Statistics. SAS
Institute, Inc, Cary, NC.
Schnitzer, M., and h. Kodama. 1977. Reactions of minerals
with soil humic substances, p. 741-770. In J.B. Dixon
(ed.) Minerals in soil environments. SSSA, Madison, WI.
Shanner, W.W., P.F. Phillip, and W.R. Schmehl. 1982. Farming
systems research and development: Guidelines for
developing countries. Westview Press, Boulder,
Colorado.

147
Singh, B. B., and J. P. Jones. 1976. Phosphorus sorption and
desorption characteristics of soil as affected by
organic residues. Soil Sci. Soc. Amer. J. 40:389-394.
Singh, B.B., and J.P. Jones. 1977. Phosphorus sorption
isotherm for evaluating phosphorus requirements of
lettuce at five temperature regimes. Plant Soil 46:31-
44.
Sioli, H. 1980. Amazonia. p257-268. In F. Barbira-Scazzochio
(ed.) Land, People, and Planning in Contemporary
Amazonia. Cambridge University Press, Cambridge.
Sleight, D.M., D.H. Sander, and G.A. Peterson. 1984. Effect
of fertilizer phosphorus placement on the availability
of phosphorus. Soil Sci. Soc. Am. J. 48:336-340.
Smyth, T.J., and M. Cravo. 1990. Critical phosphorus levels
for a corn and cowpea in a Brazilian Amazon Oxisol.
Agron. J. 82:309-313.
Smyth, T.J., and M. Cravo. 1990. Phosphorus management for
continuous corn-cowpea production in a Brazilian Amazon
Oxisol. Agon. J. 82:305-309.
Solis, P., and J. Torrent. 1989. Phosphate sorption by
calcareous Vertisol and Inceptisol of Spain. Soil Sci.
Soc. Am. J. 53:456-459.
Sweeney, D.W., and D.A. Graetz. 1988. Chemical and
decompositional characteristics of anaerobic digester
effluent applied to soil. J. Environ. Qual. 17:309-313.
Swift, R.S. and R.J. Haynes. 1989. The effects of pH and
drying on adsorption of phosphate by aluminum-organic
matter associations. J. Soil Science. 40:773-781.
Swinton, S.M., and L.A. Samba. 1986. Defining agricultural
recommendation domains in South-Central Niger.
Proceedings of Farming Systems Research/Extension
Symposium. Kansas State University, pp.318-331.
Syers, J.K., M.G. Browman, G.W. Smillie, and R.B. Corey.
1973. Phosphate sorption by soils evaluated by the
Langmuir adsorption equation. Soil Sci. Soc. Am. Proc.
37:358-363.
Tate, K.R., and B.K.G. Theng. 1980. Organic and its
interactions with inorganic soil constituents, p. 225-

148
249. In Theng, B.K.G. (ed.) soils with variable charge.
New Zealand Soc. Soil Sci.
Tshabalala, M., and D. Holland. 1986. Recommendation domains
and the design of on-farm trials research and extension
in Lesotho. Proceedings of Farming Systems
Research/Extension Symposium. Kansas State University,
pp.345-355.
Tully, R.C., and A.M. Alberti. 1985. Recommendation domains
reconsidered. 1986. Proceedings of Farming Systems
Research/Extension Symposium. Kansas State University,
pp. 236-253.
Uehara, G. and G.P. Gillman. 1980. Charge characteristics of
soils with variable and permanent charge minerals: I.
Theory. Soil Sci. Soc. Am. J. 44:252-255.
Upraity, V.N., K.D. Joshi, and B.K. Singh. 1984. Variety
adoption: A function of agronomic and socio-economic
variables. CSP, Nepal.
USDA, Soil Survey Staff. 1975. Soil Taxonomy, USDA
Agriculture Handbook 436, Washington D.C.
Vander Veen, M., and S.B. Mathema, 1979. Key Informant
Survey result for Lele, Lalitpur. HMG, Nepal.
Van Raij, B. and M. Peech. 1972. Electrochemical properties
of some Oxisols and Alfisols of the tropics. Soil Sci.
Soc. Am. Proc. 36:587-593.
van Wijk, W.R. 1966. Introduction, the physical method, p.
1-16. In W.R. van Wijk (ed) Physics of plant
environment. North Holand Publishing Co., Amsterdam.
Wilding, L.P., N.E. Smeck, and L.R. Drees. 1977. Silica in
soils: Quartz, crystobalite, tridymite and opal. p.
471-552. In J.B. Dixon and S.B. Weed (ed.). Minerals in
soil environments. SSSA, Madison, WI.
Woodruff, J.R. and E.J. Kamprath. 1965. Phosphorus
adsorption maximum as measured by the Langmuir isotherm
and its relationship to phosphorus availability. Soil
Sci. Soc. Proc. 29:148-150.
Yost, R.S., 1977. Effect of rate and placement on
availability and residual value of P in an Oxisol of
Central Brazil. Ph.D. Diss. N.C. State University.
Yost, R.S., E.J. Kamprath, E. Lobato, and G.C. Naderman, Jr.
1979. Phosphorus response of corn on an Oxisol as

149
influenced by rates and placement. Soil Sci. Soc. Am.
J. 43:338-343.
Yost, R.S., E.J. Kamprath, G.C. Naderman, and E. Lobato.
1981. Residual effects of phosphorus applications on a
high phosphorus adsorbing oxisol of central Brazil.
Soil Sci. Soc. Am. J. 45:540-543.
Yuan, T.L. 1980. Adsorption of phosphate and water-
extractable soil organic material by synthetic aluminum
silicates and acid soils. Soil Sci. Soc. Am. J. 44:951-
955.
Yuan, T.L., and D.E. Lucas. 1982. Retention of phosphorus by
sandy soils as evaluated by adsorption isotherms. Soil
and Crop Sci. Soc. Fla. Proc. 41:195-201.
Zandstra, H.G., E.C. Price, J.A. Litsinger, and R.A. Morris.
1981. A methodology for on-farm cropping systems
research. IRRI, Manila, Philippines.

BIOGRAPHICAL SKETCH
Braj K. Singh was born on May 22, 1955, in a small
village in Nepal. B.K. graduated from high school in
Kalaiya, Bara, in 1971. He received his B.S. and M.S. in
agronomy from Peoples' Friendship University in Moscow,
USSR, in 1980. From the same university he earned a Russian-
English interpreter's diploma. He worked as an assistant
lecturer of agronomy, an extension officer, and cropping
systems/farming systems agronomist in Nepal from 1980 to
1986. He was awarded a Fulbright scholarship in 1986 and
later a farming systems assistantship from the University of
Florida. B.K. conducted his dissertation project in
collaboration with TropSoil and EMBRAPA in Manaus, Brazil.
He is married to Zoila Alvarado Singh and has a son,
Alexander Alvarado Singh, and a daughter Carol Alvarado
Singh.
150

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.
Sartain, Chairman
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 of Doctor of Philosophy.
Peter E. Hildebrand
Professor of Food and Resource
Economics
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.
Donald A. Graetz
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 of Doctor of Philosophy.
-Arfl oulgA
Edward A. Hanlon
Associate 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 of Doctor of Philosophy.
Kenneth L. Buhr
Assistant Professor of
Agronomy
This dissertation was submitted to the Graduate Faculty
of 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.
AUGUST 1990
'We
Dean,/College of Agriculture
Dean, Graduate School



Table 4-10. Summary of ANOVA for multilocatonal cowpea testing.
Source
Df
LI
L2
EMS
L3
L4
L5
L6
L7
L8
Block
1
0.08ns
0.55**
0.36ns
0.10ns
0.00ns
0.10ns
0.00ns
0.36ns
Trt
3
1.13*
0.41**
0.37ns
0.96*
2.08*
0.74ns
0.43ns
1.21**
Error
3
0.09
0.00
0.15
0.05
0.12
0.15
0.18
0.01
Total
7
*, ** F-test significantly different at 0.05 and 0.01 level of probability, ns = not
significant.
Table 4-11. Duncan Multiple Range Test (DMRT) for cowpea trials conducted in the
municipal of Rio Preto da Eva, Amazonas.
Trt
LI
L2
L3
L4
L5
L6
L7
L8
1
(aiTciin yield, ng na
CM2 0+TSP2 0
1.80ab
1.50a
1.90a
2.25a
2.05a
1.35a
2.65a
1.65a
TSP40
2.30a
1.10b
2.50a
1.60ab
2.10a
1.35a
2.15a
1.30a
PCW20+TSP20
0.90bc
0.65c
1.95a
1.20bc
0.50b
0.50a
1.70a
0.2 0b
FP
0.70c
0.50c
1.45a
0.60c
0.15b
0.15a
1.65a
0.10b
CV%
21.4
7.2
20.3
16.5
30.0
46.4
21.0
14
tFive locations have been dropped due to single replicates.
Means in the same column followed by the same letter are not significantly
different at the 95% level of probability as determined by Duncan Multiple Range
Test.
o


ACKNOWLE DGMENTS
The author wishes to express his sincere appreciation
to Dr. J. B. Sartain, the chairman of the supervisory
committee, for his excellent guidance, assistance, and
continued interest in this study. His patience,
understanding, and personal friendship throughout the
academic and research program has been most invaluable and
made the author's stay in the United States a most rewarding
and stimulating experience. The financial support provided
by him to cover part of research expenses in Brazil is
gratefully acknowledged.
Sincere appreciation is extended to Dr. P.E. Hildebrand
for providing a farming systems assistantship and taking
personal initiatives to work out a joint research venture
with the TropSoils Collaborative Research Support Program
between Cornell University and the Brazilian Agricultural
Research Organization (EMBRAPA). His deep concerns,
constructive criticisms, and moral support as the member of
supervisory committee helped broaden the author's
understanding of farming systems and contributed to the
successful completion of this manuscript. The author also
thanks the other members of the supervisory committee, Dr.
iii


145
Nagarajah, S., A.M Posner, and J.P Quirk. 1970. Competitive
adsorption of phosphate with polyglacturonate and other
organic anions on kaolinite and oxide surfaces. Nature
228:83-84.
Nair, P.S., T.J. Logen, A.N. Sharpley, L.E. Sommers, M.A.
Tabatabai, and T.L. Yuan. 1984. Interlaboratory
comparison of a standardized phosphorus adsorption
procedure. J. Environ. Qual. 13:591-595.
Njobvu, C.A. 1986. Factors influencing recommendation domain
boundaries of the farming system and levels of
agricultural development in Lusaka province, Zambia.
Proceedings of Farming Systems Research/Extension
Symposium. Kansas State University, pp. 254-259.
Norris, J.M. 1970. Mutivariate methods in the study of
soils. Soils Frtil. 33:13-18.
Olsen, S.R. 1953. Inorganic phosphorus in alkaline and
calcareous soils. Agronomy 4:89-122.
Olsen, S.R and Watanabe, 1957. A method to determine a
phosphorus adsorption maximum of soils as measured by
the Langmuir isotherm. Soil Sci. Soc. Am. Proc.
21:144.149.
Olsen, S.R., and S.A. Barber. 1977. The effect of waste
application on soil phosphorus and potassium, p. 197-
215. In L.F Elliott and F.J. Estivenson (ed) Soils for
management of organic waste and waste waters. Am. Soc.
Agron. Madison, WI.
Parr, J. F., and R. I. Papendick. 1978. Factors affecting
the decomposition of crop residues by microorganisms,
p. 101-129. In W. R. Oschwald (ed.). Crop residues
management systems. ASA Special pub. No. 31, Amer. Soc.
Agron., Madison, WI.
Perrott, K.W. 1978. The influence of organic matter
extracted from humified clover on the properties of
amorphous aluminosilicates. II. Phosphate retention.
Aust. J. Soil Res. 16:341-346.
Posner, A.M., and J.W. Bowden. 1980. Adsorption isotherms:
Should they be split? J. Soil Sci. 31:1-10.
Rajan, S.S.S., K. W. Perrott, and W.M.H. Saunders. 1974.
Identification of phosphate reactive sites of hydrous
alumina from proton consumption during phosphate
adsorption at constant pH values. J. Soil Sci. 25:438-
447


43
Fertilizer placement. The advantage of banding
phosphate fertilizer is well known. What is generally not
understood is the sensitivity of nutrient uptake to band
width, de Wit (1953) demonstrated that a soil that is
virtually incapable of supporting a crop with 100 kg/ha of
broadcast phosphorus will produce nearly 50% of maximum
yield with the same amount of fertilizer applied in a narrow
band. A more detailed description of de Wit's analysis and
some of the assumptions contained in the analysis are
further elaborated by van Wijk (1966), Uehara and Gillman
(1981), and Fox et al. (1986). In agreement with the theory
Fox and Keng (1978) reported a better response from
localized P placement as compared with complete
incorporation if suboptimal P rates were used, but if
quantities of P were sufficient, best results were obtained
from incorporating P in the entire soil volume.
Kamprath (1967) found that similar maize yields were
obtained by annual banded applications of 22 kg P/ha for
seven years as were obtained by an initial P application of
350 kg/ha. Banding, therefore, saved more than half of the P
requirement. Applying N and P in knifed bands has been shown
to be an effective method of applying N and P to winter
wheat (Leikam et al., 1978, 1983). Experiments on maize have
shown that dual-placed N and P increases P uptake and maize
grain yield more than when P is banded to the side or below
the seed (Raun et al., 1987).


132
M60 and C60 (Canavalia ensiformes). Processed city waste
(8.8 kg ha '1) applied with TSP20 was inferior to all
treatments except the control.
The results obtained from glasshouse and on-station
field studies suggested that organic amendments can provide
sustained crop P nutrition compared to inorganic sources.
However, the improved efficiency of inorganic P when applied
with an organic amendment will depend on chemical
characteristics of the organic amendments.
In Chapter IV the results of farmers' field trials are
presented. Different criteria were used for the selection of
appropriate technology for farmers and delineation of RDs.
Based on the evaluation criterion of maximizing production
CM20+TSP20 was recommended in good environments (e>1.85 Mg
ha1) for maize. For cowpea, CM2 0+TSP2 0 outperformed all
treatments in poor environments (e<1.32 Mg ha1) while in
good environments (e>1.32 Mg ha1) TSP40 had clear
advantage. Farmer practice had the highest return over cash
cost compared to other treatments for both crops.
The result of CANOVA for maize and cowpea grain
production provided inconclusive results because of the
presence of environment x treatment interaction. The slope
estimates of the regression analysis provided valuable
information on changes taking place in soil fertility
following deforestation. The rate of decline in pH in PF and
SF was 0.15 unit per year. Aluminum saturation was


117
Figure 4-8. Relationship of net income, net income/cash cost
and net income/total cost with environmental index, in
on-farm cowpea trials from Rio Preto da Eva, Amazonas,
Brazil.
Treatment
Intercept
b
SE Y Estimate
Net
Income/Cash Cost
FP
-20.8
24.8
8.36
PCW20+TSP20
-3.0
3.2
2.87
TSP40
5.7
0.6
0.12
CM20+TSP20
11.3
-2.3
2.14
Net
Income/Total
Cost
FP
-2.5
4.3
5.9
PCW20+TSP20
-2.0
2.1
1.8
TSP40
1.8
1.2
1.2
CM20+TSP20
3.0
0.2
0.1
Net Income ha1
FP
-521.0
620.4
226.6
PCW20+TSP20
-558.0
597.3
238.4
TSP40
20.7
481.1
114.5
CM20+TSP2 0
604.1
60.7
23.8
SE = Standard Error


62
Table 3-8. Orthogonal contrasts of maize grain yield under
different treatments applied in a 30 cm wide band
at UEPAE research station, Manaus, Brazil.
Treatment^
Mean Grain Yield
Cropl Crop2 Contrasts Cropl
- Mg ha'1 --
Crop2
M60
3.44
1.65
C60 VS TSP60
ns
ns
C60
3.56
2.07
CM60 VS TSP60
**
**
PCW60
1.41
1.67
PCW VS TSP60
"k:k
ns
CM60
4.63
2.37
M60 VS TSP60
ns
ns
M20+TSP20
3.55
1.41
C20+TSP20 VS TSP40
*
ns
C20+TSP20
3.62
1.81
M20+TSP20 VS TSP40
ns
ns
PCW20+TSP20
i 1.66
1.90
PCW20+TSP20 VS TSP40
**
ns
CM20+TSP20
4.10
1.92
CM20+TSP20 VS TSP40
k k
ns
Control
1.71
1.10
TSP2 0
2.63
1.04
TSP40
3.12
1.46
TSP60
3.69
1.63
*, ** Significantly different at 0.05, and 0.01 level of
probability, ns = not significantly different at 0.05 level
of probability.
f M60, C60, PCW60, and CM60 = Mucuna, Canavalia, Aerobically
processed city waste, and Chicken manure applied to provide
equivalent of 26.4 kg ha'1 of P.
TSP20, TSP4 0, and TSP60 = 8.8, 17.6, and 26.4 kg ha'1 P from
triple superphosphate (TSP).


92
Table 4-5. Relationship between soil characteristics
measured after treatment application with grain yield
for maize and cowpea crops in the municipality of Rio
Preto da Eva, Amazonas, Brazil'.
Partial
p
Treatments r a b
Maize
FP (A1 Sat.)
0.79**
6.39
-0.06
(0.04+)
PCW20+TSP20 (pH)
0.35*
-3.03
0.81
(0.14)
CM20+TSP20 (P)
0.42*
-0.52
0.34
(0.06)
TSP40 (P)
0.77**
0.42
0.34
(0.07)
COWPEA
FP (pH)
0.70**
-1.69
0.97
(0.11)
PCW20+TSP20 (P)
0.59*
0.21
0.14
(0.03)
TSP40 (P)
0.38*
1.03
0.10
(0.02)
CM20+TSP20 (All)
X
X
X
X
-(- Soil characteristics were measured 65 d after treatment
application for maize, and 45 d for cowpea.
The numbers in the parenthesis are the standard errors of
b estimates. Variables inside the parenthesis are soil
characteristics.
The values reported for r2, a, and b are for those
variables which had a strong relationship with crop
production.
x None of the variables had a strong relationship with crop
production.


147
Singh, B. B., and J. P. Jones. 1976. Phosphorus sorption and
desorption characteristics of soil as affected by
organic residues. Soil Sci. Soc. Amer. J. 40:389-394.
Singh, B.B., and J.P. Jones. 1977. Phosphorus sorption
isotherm for evaluating phosphorus requirements of
lettuce at five temperature regimes. Plant Soil 46:31-
44.
Sioli, H. 1980. Amazonia. p257-268. In F. Barbira-Scazzochio
(ed.) Land, People, and Planning in Contemporary
Amazonia. Cambridge University Press, Cambridge.
Sleight, D.M., D.H. Sander, and G.A. Peterson. 1984. Effect
of fertilizer phosphorus placement on the availability
of phosphorus. Soil Sci. Soc. Am. J. 48:336-340.
Smyth, T.J., and M. Cravo. 1990. Critical phosphorus levels
for a corn and cowpea in a Brazilian Amazon Oxisol.
Agron. J. 82:309-313.
Smyth, T.J., and M. Cravo. 1990. Phosphorus management for
continuous corn-cowpea production in a Brazilian Amazon
Oxisol. Agon. J. 82:305-309.
Solis, P., and J. Torrent. 1989. Phosphate sorption by
calcareous Vertisol and Inceptisol of Spain. Soil Sci.
Soc. Am. J. 53:456-459.
Sweeney, D.W., and D.A. Graetz. 1988. Chemical and
decompositional characteristics of anaerobic digester
effluent applied to soil. J. Environ. Qual. 17:309-313.
Swift, R.S. and R.J. Haynes. 1989. The effects of pH and
drying on adsorption of phosphate by aluminum-organic
matter associations. J. Soil Science. 40:773-781.
Swinton, S.M., and L.A. Samba. 1986. Defining agricultural
recommendation domains in South-Central Niger.
Proceedings of Farming Systems Research/Extension
Symposium. Kansas State University, pp.318-331.
Syers, J.K., M.G. Browman, G.W. Smillie, and R.B. Corey.
1973. Phosphate sorption by soils evaluated by the
Langmuir adsorption equation. Soil Sci. Soc. Am. Proc.
37:358-363.
Tate, K.R., and B.K.G. Theng. 1980. Organic and its
interactions with inorganic soil constituents, p. 225-


134
APPENDIX A
ECONOMIC ANALYSIS FOR COWPEA
Yield, income, and cost from the cowpea trials, Rio Preto da
Eva, Amazonas, Brazil. (US $)
Yield Gross Total Net Net Income/
Mg ha 1 Income Cost Income Total Cost
FP
0.70
350
67
283
4.2
0.50
250
55
195
3.5
1.45
725
112
613
5.5
0.60
300
61
239
3.9
0.15
75
34
41
1.2
0.15
75
34
41
1.2
1.70
850
127
723
5.7
0.10
50
31
19
0.6
0.20
100
37
63
1.7
2.20
1100
157
943
6.0
1.20
600
97
503
5.2
1.50
750
115
635
5.5
0.00
0
25
-25
-1.0
PCW20+TSP20
0.90
450
239
211
0.9
0.65
325
224
101
0.5
1.95
975
302
673
2.2
1.20
600
257
343
1.3
0.50
250
215
35
0.2
0.50
250
215
35
0.2
1.65
825
284
541
1.9
0.20
100
197
-97
-0.5
0.40
200
209
-9
-0.0
1.90
950
299
651
2.2
1.50
750
275
475
1.7
1.80
900
293
607
2.1
0.00
0
185
-185
-1.0


144
Mekaru, T., and G. Uehara. 1972. Anion adsorption in
ferruginous tropical soils. Soil Sci. Soc. Am. Proc.
36:296-300.
Mendez, J., and E.J. Kamprath. 1978. Liming of Latosols and
the effect on P response. Soil Sci. Soc. Am. J. 41:86-
88.
Mokwunye, U. 1977. Phosphorus fertilizers in Nigerian savana
soils. I-Use of phosphorus sorption isotherms to
estimate the phosphorus requirement of maize at Samaru.
Trop. Agrie. 54:265-270.
Moormann, F.R., and B.T. Kang. 1978. Microvariability of
soils in the tropics and its agronomic implications
with special reference to West Africa, p 29-43. In
Diversity of the soil in the tropics. ASA, Madison, WI.
Moris, F.I., A.L. Page, and L.J. Lund. 1976. The effect of
pH, salt concentration, and nature of electrolytes on
the charge characteristics of Brazilian tropical soils.
Soil Sci. Soc. Am. 40:521-527.
Moreno, E.C., W.L. Lindsay, and G. Osborn. 1960. Reactions
of dicalcium phosphate dihydrate in soil. Soil Sci.
90:58-68.
Moussie, M., and C. Muhitira. 1988. Classification of
farmers into recommendation domains. Proceedings of
Farming Systems Research/Extension Symposium.
University of Arkansas, p.241-250
Muljadi, D., A.M. Posner, and J.P. Quirk. 1966. The
mechanism of phosphate adsorption by kaolinite,
gibsite, and pseudoboehmite. Part II. The location of
adsorption sites. J. Soil Sci. 17:230-237.
Munns, D.N., and R.L. Fox. 1976. The slow reaction which
continues after phosphate adsorption: Kinetics and
equilibrium in some tropical soils. Soil Sci. Soc. Am.
J. 40:46-51.
Murphy, J., and J.P. Riley. 1962. A modified single solution
method for the determination of phosphate in natural
waters. Anal. Chem. Acta. 27:31-36.
Murray, K., and P.W. Linder. 1984. Fulvic acids: Structures
and metal binding. II. Predominant metal binding sites.
J. Soil Sci. 33:217-222.


Net Income/Cash Cost ($)
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5
Environmental Index (e), Mg/ha
Environmental Index (e), Mg/ha
Values are expressed in US $


57
Table 3-6. Analysis of variance for maize herbage dry weight
production per pot in the glasshouse study.
Source^
DF
F Value
Pr > F
REP
3
1.4
0.23
ORGANIC
2
970.2
0.00
RATE
1
189.2
0.00
ORGANIC*RATE
2
144.0
0.00
INRATE
2
127.9
0.00
ORGANIC*INRATE
4
17.6
0.00
RATE*INRATE
2
0.1
0.86
ORGANIC*RATE*INRATE
4
8.3
0.00
ERROR
51
CORRECTED TOTAL
71
CV%
13.0
f REP = Replication, ORGANIC = Organic amendments, RATE =
Rate of Organic amendments, INRATE = Rate of Inorganic P.


17
Table 2-3: Selected physical and chemical properties of Ap
horizon of the Xanthic Hapludox used in adsorption
studies.
Parameters^
Value
Units of
Measured
Measurement
Bulk density
1.25
g cm"3
g kg"1
Clay
815
pH H20, KCl
4.5, 4.2
->. kg "1
g
Net Charge
0.22
cmol (
Oxides Fe, A1
0.075, 0.325
g 100
A1 Sat. (ECEC)
78.4
%
Mg g'1
Total P
170.0
soil
g"1 soil
Mehlich-I P
3.3
Mg.
Mg g
Al, Fe (P)
29.0, 1.4
soil


12
1962). Phosphorus removed from solution was considered
adsorbed.
Desorption was studied by equilibrating the soil from
the adsorption study in 20 mL 0.01 M CaCl2 for 6 hrs (Singh
and Jones, 1977). Increases in solution P were measured and
considered desorbed P. Adsorption and desorption isotherms
were constructed and a standard Langmuir equation was fitted
to calculate adsorption maxima and bonding energy (Olsen and
Watanabe, 1957).
Simulated silica matrix (SSMT). Hydrochloric acid
(0.1 M) washed, fine silica was mixed with a nutrient
solution containing N, K, Ca, Mg, Zn, Mo, Mn, Fe, Cu, B, S,
and Cl (P excluded) (EMBRAPA, 1976) to simulate the nutrient
requirement for maize in a solution culture. Solution pH was
adjusted to 5.0 with 0.1 M HC1. Table 2-1 highlights the
chemical compounds used to make the nutrient solution and
the resulting concentrations. The inoculation of silica
matrix was carried out with microbes grown on potato
dextrose agar.
Five grams of ground organic amendments (the same ones
used in incubation study with soil) were mixed with 95 g of
silica matrix and incubated for 35, 65, and 150 d at 30
3C. The matrix was kept moist and mixed thoroughly every 6
d. After each incubation period, duplicate 2-g samples were
extracted with 0.01 M CaCl2 following the same procedure as
described for soil. The sum of P detected by sequential


127
Conclusion
The presence of environment-by-treatment interaction
masked the mean treatment differences measured by CANOVA. To
detect treatment effects, this analysis suggested grouping
of similar environments and a repeat of the experiment. The
slope estimates in regression analysis were useful in
predicting rates of decline in pH, ECEC, and Mehlich-I
extaractable P with the year of cultivation. A reduction of
2.1 and 2.6 g g'1 soil in Mehlich-I extractable P was
obtained annually in PF, and SF. Although an equivalent of
8.8 kg ha1 P was applied from different amendments, their
effect on maize and cowpea production varied. Phosphorus
from TSP was used more efficiently, although this treatment
reduced ECEC, and soil pH.
Modified stability analysis was conducted for maize and
cowpea using yield, net income/cash cost, net income/total
cost, and net income ha1 as technology evaluation criteria.
Different evaluation criteria led to different conclusions
regarding recommendations of appropriate technology for a
given environment. These criteria also influenced selection
of production environments for grouping into RDs. Based on
the evaluation criterion of maximizing maize production
CM20+TSP20 was recommended in RD2 (good environment) But
there was a clear advantage of FP in RD2 in terms of net
return to cash cost. If a farmer's goal is to maximize
income from his scarce cash, he ought to use farmers'


Appendix A contd
TSP40
2.30
1150
230
920
4.0
1.10
550
158
392
2.5
2.50
1250
242
1008
4.2
1.60
800
188
612
3.3
2.10
1050
218
832
3.8
1.35
675
173
502
2.9
2.65
1325
251
1074
4.3
1.30
650
170
480
2.8
1.20
600
164
436
2.7
2.60
1300
248
1052
4.2
2.20
1100
224
876
3.9
2.10
1050
218
832
3.8
1.30
650
170
480
2.8
CM2 0+TSP2 0
1.80
900
211
689
3.3
1.50
750
193
557
2.9
1.90
950
217
733
3.4
2.25
1125
238
887
3.7
2.05
1025
226
799
3.5
1.35
675
184
491
2.7
2.15
1075
232
843
3.6
1.65
825
202
623
3.1
1.70
850
205
645
3.1
1.40
700
187
513
2.7
1.90
950
217
733
3.4
1.70
850
205
645
3.1
2.00
1000
223
777
3.5


Table 4-10. Summary of ANOVA for multilocatonal cowpea
testing 110
Table 4-11. Duncan Multiple Range Test (DMRT) for
cowpea trials conducted in the municipal of Rio
Preto da Eva, Amazonas, Brazil 110
Table 4-12. Combined Analysis of Variance for
cowpea experiments. Ill
Table 4-13. Technology selection for a given land type
based on different evaluation criteria for maize
cultivation 124
Table 4-14. Technology selection for a given land type
based on different evaluation criteria for cowpea
cultivation 124
ix


80
was intended to bring awareness among farmers to preserve
the environment and assist them in improving their
agricultural production. The Brazilian national agricultural
research institution (EMBRAPA) has a mandate of developing
appropriate technology for different farming conditions in
this relatively inaccessible area.
Developing a Research Base
Secondary information regarding indigenous farming
practices of the area was collected from published sources.
A rapid appraisal of the area was conducted with a
multidisciplinary team of scientists participating from
various research disciplines and state planning and
agriculture extension organizations who visited the area on
three different occasions to collect and verify information
obtained in group discussions or during individual
communication with farmers. Farmers' knowledge of indigenous
technology, agronomic practices, and land types being used
were recorded. An extensive soil sampling program was
carried out to understand soil physical and chemical
characteristics and relate them to farmers' rationale for
assigning a particular cropping pattern to a given land
type. Farmers played an active role in technology design,
execution and evaluation.


106
Figure 4-5. Relationship of net income/cash cost, net
income/total cost, and net income with
environmental index, in on-farm maize trials from
Rio Preto da Eva, Amazonas, Brazil
Figure 4-6. Distribution of confidence intervals for
net income/cash cost, net income/total cost, and
net income for different treatments used for maize
cultivation 108
Figure 4-7a. Response of different treatments to
environmental index for cowpea production, Rio
Preto da Eva, Amazonas, Brazil 115
Figure 4-7b. Distribution of confidence intervals for
cowpea production in poor (e<1.32 mg ha'1), and
good (e>1.32 Mg ha'1) environments 116
Figure 4-8. Relationship of net income, net income/cash
cost and net income/total cost with environmental
index, in on-farm cowpea trials from Rio Preto da
Eva, Amazonas, Brazil 117
Figure 4-9. Distribution of confidence intervals for
net income/cash cost for selected treatments in
poor (e<1.32 Mg ha'1) and good (e>1.32 Mg ha'1)
environments for cowpea cultivation 119
Figure 4-10. Distribution of confidence intervals for
net income/total cost for selected treatments in
poor (e<1.32 Mg ha1) and good (e>1.32 Mg ha'1)
environments for cowpea cultivation 120
Figure 4-11. Distribution of confidence intervals for
net income for selected treatments in poor (e<1.32
Mg ha'1) and good (e>1.32 Mg ha'1) environments for
cowpea cultivation 121
Figure 4-12. Recommendation domains, based on yield,
and their relationship to land types in the
municipality of Rio Preto da Eva,Amazonas,
Brazil 125
Figure 4-13. Recommendation domains, based on net
return to cash cost, and their relationship
to land types in the municipality of Rio Preto
da Eva, Amazonas, Brazil 12 6
xii


121
Net Income (US$/ha)
Net Income (US$/ha)
Figure 4-11. Distribution of confidence intervals for net
income for selected treatments in poor (e<1.32 Mg ha'1)
and good (e>1.32 Mg ha'1) environments for cowpea
cultivation.


67
Trt.
a
bl
b2
, _2
b3 R
TSP40
4.97
3.13E-01
-3.66E-03
9.57E-06 0.78
M20+TSP20
4.22
3.48E-01
-3.18E-03
7.14E-06 0.90
PCW20+TSP20
4.41
6.26E-02
-8.02E-04
2.24E-06 0.30ns
TSP40
3.77
2.96E-01
-2.69E-03
5.92E-06 0.92
Figure 3-6. Effect of selected organic amendments, applied
in combination with inorganic phosphorus source in a
quantity eguivalent to provide 8.8 kg ha1 of P, on
sustaining Mehlich-I extractable soil P pool in a
maize-maize rotation on a Xanthic Hapludox.


LIST OF FIGURES
Figure 2-1. Acid-base potentiometric titration curves
for the Ap horizon of Xanthic Hapludox with
varying concentration of CaCl2. 18
Figure 2-2. Phosphorus adsorption isotherms following
35 d (a) and 150 d (b) of soil incubation with
Mucuna aterrima (Mucuna), Chicken Manure (CM),
Aerobically Processed City Waste (PCW), and
Control. 21
Figure 2-3. Phosphorus desorption isotherms following
35 d (a), and 150 d (b) of soil incubation with
Mucuna aterrima (Mucuna), Chicken Manure (CM),
Aerobically Processed City Waste (PCW), and
Control 22
Figure 2-4. A schematic representation of molecular
structures of oxides of phosphorus (a), and oxides
of silicon (b) 26
Figure 2-5. Effect of incubation period and incubation
matrix on P release pattern from different organic
amendments in a laboratory study 27
Figure 2-6. Phosphorus adsorption isotherms for the
Xanthic Hapludox (0.01 M CaCl2) fitted to Langmuir
equation. The lines in the figure represents
fitted equation. 28
Figure 2-7. A comparison of P adsorption by Xanthic
Hapludox based on the estimation of preadsorbed P
by sequential extraction of soil or silica matrix
incubated with organic amendments 34
Figure 3-1. Geographic location of Amazon basin in
Brazil (a), on-station and farming systems
research (FSR) sites (b), and effective rainfall
during the period of August, 1988 until August
1989 (c) at EMBRAPA station in Manaus, Brazil.. 47
x


53
the Al saturation (>76.0%) calculated from the effective
cation exchange capacity (ECEC) (Table 3-5).
Glasshouse Study
An analysis of variance (Table 3-6) presented for maize
herbage dry weight yield indicated the presence of a three
way interaction among types of organic amendment, rate of
amendment, and rate of TSP. Among the three amendments
tested, PCW produced the lowest yield when applied alone or
in combination with TSP (Fig. 3-2a). It is interesting to
note that when PCW was applied at higherr rate (equivalent
to 17.6 kg ha1 of P) it produced less dry matter per pot
compared to when this amendment was applied at lower rate
(Figure 3-2a). A possible explanation is that the slow rate
of release of nutrients from the material resulted in an
initial P deficiency for maize seedlings. There was also a
high concentration of Zn and Cu in this material (280, 155
/xg g1, respectively) which may have played a role in
providing cation bridging between organic and P anions
making P less available (Murray and Linder, 1984). By
forming organo-metal complexes, humic and fulvic acids as
well as simple acids can dissolve or decompose such minerals
as feldspar, gibsite, goethite, hematite, and mica
(Schnitzer, 1977). Therefore, organic amendments with high
Zn and Cu content may have limited potential as a source of
P in highly leached soils.