• TABLE OF CONTENTS
HIDE
 Front Cover
 Preface
 Table of Contents
 1.0 General Introduction and 2.0...
 3.0 Introduction to livestock...
 4.0 Introduction to agroforestry...
 Appendix A: Glossary
 5.0 Appendix A: Glossary
 6.0 Appendix B: Major referenc...
 Back Cover














Title: On-farm experimentation : guidelines for using OFE methodology in crops, livestock and agroforestry experimentation
CITATION THUMBNAILS PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00080815/00001
 Material Information
Title: On-farm experimentation : guidelines for using OFE methodology in crops, livestock and agroforestry experimentation
Physical Description: Book
Language: English
Creator: Stroud, Ann
Publisher: CIMMYT Eastern African Economics Programme
Publication Date: 1985
 Record Information
Bibliographic ID: UF00080815
Volume ID: VID00001
Source Institution: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: oclc - 189864502

Table of Contents
    Front Cover
        Front Cover
    Preface
        Preface
    Table of Contents
        Table of Contents
    1.0 General Introduction and 2.0 Crop experiments
        Page 1
        Page 2
        Page 3
        Page 4
        Page 5
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
        Page 18
        Page 19
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
        Page 28
        Page 29
    3.0 Introduction to livestock experimentation
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
        Page 35
        Page 36
        Page 37
        Page 38
        Page 39
        Page 40
        Page 41
        Page 42
        Page 43
        Page 44
    4.0 Introduction to agroforestry experiments
        Page 45
        Page 46
        Page 47
        Page 48
    Appendix A: Glossary
        Page 49
    5.0 Appendix A: Glossary
        Page 50
        Page 51
        Page 52
        Page 53
        Page 54
        Page 55
        Page 56
    6.0 Appendix B: Major references
        Page 57
        Page 58
        Page 59
    Back Cover
        Page 60
Full Text


PETER E. ILDIEiJAND






CIMMYT




AN OCCASIONAL SERIES OF PAPERS AND NOTES ON METHODOLOGIES AND PROCEDURES
USEFUL IN FARM SYSTEMS RESEARCH AND IN THE ECONOMIC INTERPRETATION OF
AGRICULTURAL EXPERIMENTS.


NUMBER 13.


TEACHING NOTES
NOVEMBER, 1985


ON-FARM


EXPERIMENTATION


GUIDELINES FOR USING OFE METHODOLOGY IN CROPS, LIVESTOCK
AND AGROFORESTRY EXPERIMENTATION


Dr Ann Stroud














CIMMYT EASTERN AFRICAN ECONOMICS PROGRAMME
INTERNATIONAL MAIZE AND WHEAT IMPROVEMENT CENTRE (CIMMYT)
P.O. BOX 25171, NAIROBI, KENYA, TELEPHONES: 592054, 592206, 592151
Telex: 22040 ILRAD Cables: Cencimmyt, Nairobi









PREFACE





The CIMMYT East African Program has been conducting a series of
Workshops in On-Farm Research (OFR/FSP) Methodology in the region.
Prior to the posting of CIMMYT Regional Agronomists to the Nairobi
Office, we sought assistance in conducting these Workshops from
agronomists in the region who are familiar with CIMMYT's approach
to On-Farm Research. Dr Ann Stroud, an agronomist currently working
with REDSO/USAID as Regional Weed Agronomist is one who participated
in many of our Workshops.

During these Workshops we realized that there was no single document
which provides a comprehensive coverage of the materials taught in
our Workshops. Therefore, we requested Dr Ann Stroud to put
together'all materials she has presented in these Workshops into
training notes for easy reference by trainees. This three part series
of teaching notes on On-Farm Experimentation includes:
Number 11:- Concepts and Principles; Number 12:- Evaluation of
On Farm Trials; and Number 13:- Guidelines for using OFE Methodology
in Crops, Livestock and Agroforestry Experimentation. These notes
are meant to.be used by teachers, planners, researchers and
extensionists who have had some previous background in agricultural
research. Although reference is consistently made to 'researchers
in the series, it should be understood that 'extensionists' can also
be 'researchers'. This current version is a preliminary working
draft. It will be updated as we receive comments from the users.
We would therefore request constructive comments, criticisms and
suggestions from the users to update this document.







P. Anandajayasekeram
Regional Training Officer
CIMMYT Economics









TABLE OF CONTENTS


Subject Page Number


1.0 General Introduction.............. ............... .......

2.0 Crop experiments ................. ..... .................. 1
2.1 Soil fertility...................... ..... .......... 1
2.2 Varietal testing..... ................. .. *******. 8
2.3 Pest control .......... .......... ....... ........... 17
2.4 Cultural practices................................ 21

3.0 Introduction to livestock experimentation............... 30
3.1 Differences in ORT compared to OFE with
crop research............... ....................... 30
3.2 System components to consider (ILCA) ............... 30
3.3 Reasons to do ORT with animals...................... 31
3.4 Type and management of experiments................ 31
3.5 A sample of trial design questions.................. 32
3.6 Experimental design considerations............... 33
3.7 Types of measurements............................. 34
3.8 Arrangements with the farmer...................... 34
3.9 Choice of farmers and/or animals................... 36
3.10 Evaluation criteria.................................... 37
3.11 Examples. ....................... .............. .... 37

4.0 Introduction to agroforestry experiments
4.1 What is .Agroforestry .............. ................. 45
4.2 Planning ............................................ 46
4.3 Characteristics of experimental treatments.......... 46
4.4 Implemention ........................................... 46
4.5 Evaluation ...................... .................. 47

5.0 Appendix A:
Glossary.................................................. 49

6.0 Appendix B:
Major References.......... .................................. 57








1.0 GENERAL INTRODUCTION


In the OFR process, the diagnostic phase helps to familiarize the
researcher with the farmer's situation; identifying areas wherd the
researcher can design or test new technologies which will assist the
farmer. In this phase, problems, constraints and missed opportunities
should be identified and prioritized; possible solutions identified,
scrutinized for their feasibility and prioritized for various defined
research domains or target groups. After reviewing the literature,
the researcher can begin to clarify what further information is needed
to result in a useful recommendations) and a more or less detailed
site description. Background information should be collected for the
area.

The researcher now begins the experiment phase where specific objective
need to be set, experiments designed, implemented'and evaluated.
Technology progress from design to testing within the 'research' phase,
to further testing, modification and demonstration within a 'research/
extension' phase, finally to an 'extension' phase. The researcher
should remember that the objective is to test solutions while involving
farmers, the final users: constant exchange with farmers is necessary.
The researchers should remain flexible, continue evaluating farmer
feedback and adjust the objectives accordingly. "Compromises, changes,
reversals of direction or even termination of projects may be
appropriate (but difficult)..."(Rhoades, 1985). The farming systems
perspective (FSSP) embodied in On Farm Experimentation does not aim at
profound transformation of agricultural production systems but at the
development of an institutional framework or way of thinking in order
to facilitate fast adoption of useful improved practices.

2.0 CROPS EXPERIMENTS

2.1 Fertilizer Experiments

Consider fertilizer as coming from inorganic (chemical)
or organic (manure, compost, crop residue) sources.

1. Types of experiments:
a. Fertilizer rates, time or methods of application,
type of fertilizer materials, or any combination of
these plus interactions of these factors with other
management factors.

b. Fertilizer behaviour in the soil movement,
fixation, transformation.

c. Soil test correlation and calibration to be
implemented over a range of soils over years and
across management levels.


d. Residual both long and short term.









e. Innoculum.


2. Considerations:

R.A. Morris (1981) has made a list of the types of
considerations and questions that need to be asked when
embarking on a fertilizer rate experiment. The same
sorts of questions can also be asked for time method of
application:

To determine fertilizer rates for crops ask some of the
following questions:

What is the crop and what is its nutrient requirement
for high yields and for moderate yields?

What is known about patterns of nutrient uptake during
crop growth?

What is the soil and how much of each required nutrient
can it supply?

What are the current fertilizer recommendations for each
soil mapping unit or region? Is it easy to obtain a
soil analysis and are the test data for the soils
reliable? (That is, has the soil test been calibrated
for the soils and crops of the pattern?) Early in the
project it may be important to know only if an element
is deficient. Later the need may be to determine
"optimum" levels.. Ultimately, the need may be to
determine carry-over or residual effects and to compare
practicable methods of improving fertilizer efficiency.

Can the site be stratified on the basis of distinguishing
soil features?

Other questions pertain to the capacity to do research:

How diversified is the area to be studied?

What is the size of the research staff and how familiar
are they with soil fertility research, experimental
designs, and methods of data analysis?

What are the immediate and long-range objectives?

The analysis and interpretation of data a fertilizer
study is not complete until data are properly interpreted
ask the following questions:


. What are current fertilizer and product prices?










Will farmers purchase fertilizers with the benefit of
government-sponsored production loans?

Will fertilizers be subsidized or will a current subsidy
be reduced?

How specific must the fertilizer recommendation be?
And, can a specific recommendation be justified?

The research approach will be affected by background
information such as past fertilizer response experiments,
soil test results, maps, and related information.
The approach will also be affected by the size and experience
of research staff, and the complexity of the site.

In summary, the type and number of trials is determined by:

What is already known
Soil and fertility conditions in the area.

3. Im plementation and evaluate ion:

a. It is important to communicate with farmers as they can
provide detailed information concerning differences in
drainage and soil type within their farm, past management,
crop history, weather, residual effect.

b. It may be difficult to have reproducibility of fertilizer
response over an area due to inherent field differences.
Understand these differences and how they would affect
response. The number of fields sampled should be
increased and several years data collected in order to
broaden the results and inferences.

c. Use local methods of measurement but equate them to
accurate, scientific measures.

d. If the trial is farmer managed, monitor to check on
farmer modification of treatment applications.

e. Management of non-experimental variables such as:
weeding, seeding depth, plant population, etc. will be
important in order to interpret results. If management
is not accurate or possible, data should be collected
to note these variables.

f. Trial maintenance may be affected by higher rates of
fertilizer. For example, weeding may have to be more
frequent.











g. Review data for reliability looking for missing and
erroneous values. Erroneous values should be
investigated: re-calculate, check with visual observa-
tions made. If still biologically plausible, keep the
data, if not, discard it and create a value.

h. It is important to use economic evaluation techniques
such as partial budget, variable costs and net benefits,
marginal rates of return, to have an idea of the risks
associated with Fertilizer use.

4. Trial exami les

a. (Matlon, 1984)

'Farmers' tests conducted in Upper Volta in 1982 had
measured the profitability and risks associated with the
recommended dose of NPK (14:23:15) on cereal varieties.
The analysis did not answer the question of whether the
recommended dose was optimal by financial and economic
criteria and whether the risks were the same at levels
other than the recommended dose. To answer these
questions required data from the tests that would allow
a comparison of yield responses at different fertilizer
levels and the calculation of profit distributions.
Moreover, the profitability of urea in combination with
cotton complex fertilizer had not yet been tested in
Upper Volta under farmers' conditions.

A joint researcher- and farmer-managed trial was set up
with the objectives to:

SEstimate response functions to cotton complex fertilizer
in each of the three agroclimatic zones, and, based on
these results, calculate levels that maximize financial
and economic profitability in the short term;

Calculate the probability distribution of gains and
losses associated with a range of fertilizer doses
applied to local and improved varieties in different
regions;

Measure the profitability of applying urea at a
recommended dose and the probability of losses and gains,
again by variety and region; and

Identify and measure the effects of management factors
(e.g., soil preparation, fertilizer use) and microenviron-
mental factors (e.g. soil type) on returns.









The trial was designed to combine researcher and farmer
management because the amounts of fertilizer applied had to
be precise, whereas, fertilizer doses in up to 30% of all
cases.,

A fertilizer-response trial combined with a varietal test
seemed to be the most workable. Field assistants would
intervene to apply fertilizer on plots demarcated within
farmers' tests of improved and local cereal varieties, and
all other operations were to be performed by the farmers.

Six fertilizer doses were selected. Included was the
recommended dose (100 kg/ha) of cotton complex fertilizer
with and without urea. The number of treatments/farmer
was limited to four so that errors in reporting would
not be unacceptably large. All farmers received three
treatments (0; 100 kg NPK/ha; 100kg NPK/ha plus 50 kg urea),
and the remaining three treatments were randomly
*distributed, with each farmer receiving one (50 kg NPK/ha;
200 kg NPK/ha or 400 kg NPK/ha). Detailed data on
operations were collected for each of the eight test plots.
Yields were measured by field enumerators, harvesting each
plot completely.

b. (Morris, 1981)

From previous studies in the area, the crop responds to at
least 50 kg N/ha. The responses to phosphorous and
potassium are uncertain. There is interest to test the
response to higher nitrogen rates and for the presence of
phosphorous and potassium dificiencies. There are sufficient
personnel-to conduct a moderately large conventional
experiment plus six small superimposed N-P-K trials and
there is someone on the staff who is familiar with computation
of basis statistical analysis with a pocket calculator.

This example evaluates several treatment combinations in the
range in which fertilizer response may occur. Additionally,
an example of a test for treatment interactions with fields
in analyses of variance (AOV) of the superimposed trial is
given.

Twelve selected treatment combinations for the conventional
experiments are shown in Table 1. For the superimposed
trials a subset of six treatment combinations was chosen
(Table 2). It will permit you to examine the response to
phosphorous, potassium, and the phosphorous-potassium
interaction at 70kg N/ha, and the response to nitrogen
without phosphorous or potassium fertilizer. The AOVs are
also presented in Tables I and 2.











Table 1.


An example of a conventional 12-treatment
nitrogen-phosphorous-potassium fertilizer experiment


Treatment Treatment Grain yield (t/ha) by farm
No. legosiat ion.
N P K R1 R2 R" R- X


3.54
4.05
4.15
3.50
3.90
4.30
3.10
3.91
4.09
3.94
3.92
4.08


4.11
3.83
4.12
4.18
3.E85
3.87
3.68
4.07
4.31
3.87
4.10
4.07


3.47
4. 12
3.59
1. C00
4.17
3.96
4.22
4.07
4.59
4.07
4.07
4.49


3.39
3.16
3.52
4.02
3.94
3.43
3.91
4.16
4.32
4.63
4.03
3.93


3.62
3.79
3.84
3.92
3.97
3.89
3.98
4.11
4.33
3.88
3.03
4.14


___ ANOQVA___
S MS F


BLOCKS
N
P
K
N X P
N X P
N X K
N X P' X K
Error


Yield (t/ha)


50
50
50
50
70
70
70W
70
90'
90
90
90


Source


0.088
0.518
0.266
0.252


3
2
1
1
2
2
1
2
33


K-
K+


6.55
3.36
3.18


0.265
1. 036
0.266
0.252

0.00
0.00
0.00
2.611

P-
3.97
3.85

N'so
N-...


0.0791

P+
3.95
4.06

3.52
3.86
4.23


~i-i-^------~,--------


-----------------------------I-I-~--~











V




N CO N
CM N ow r-
H-C r-N Ll H


M N HO 0







O M11
Sr- c u -
0 0 1 e ,


.a co ..
P4) C4

0




o

a a



p c

4J 0 0 (1 $ 1
0













4-)
0I N 1


0) 0 0 U )n
N tt) ( i
41. H C 4) 4.2 0 r 1















0 0 0.
40
0) u 1 0 0 Co cO

SH. .N. .H .H












m0 0 0 0 0 0 m c
VD 0







'0 H3 C N o



) o o 00 c Co
C o 0 CO r Co CO CO


.0U0
Q) N.4 N 14 C1 Cf) 11
1 C) M cO 3n 1 0 ro. Co q3
4.2 -r O- 'Zr 0t c0 0r -
0 *3


S






00 H 14 C) Co C) r- c1
03 0 I sT H r I N cO N f' 0



X o
S4 r0 C) 1 CH Co C' '
eN cn o H c0' H r 'N



03 c m C; C; CO -
)0

rCi
.0


cci

H) cn m C i C -) rC) C) C)









S 000 000

4-E H N O O O O










On the basis of the AOVs on both conventional experiments and the
superimposed trials, treatment differences occurred. Phosphorous
and potassium did not influence yield but nitrogen did. The
results (yield levels, treatment responses, and-error) are in
approximate agreement between the two types of experiments,
suggesting that similar behaviour toward fertilizer applications
would occur over many fields in the area.

Economic analysis can be applied to the nitrogen response data, but
from casual inspection, it appears that response was rather linear
over the range tested. Under a simple prof i t-max i mixing assumption,
the highest nitrogen rate would, therefore, be more profitable than
the middle rate, provided the cost of an increment of nitrogen is
less than the value of additional yield. With responses similar to
that for nitrogen in this example, input availabilities, cost
constraints, and risk factors should be considered jointly by
agronomists and economists.

2.2 Varietal Testing

1. ]Types of experiments

a. Varietal performance yield variability, seedling
establishment, insect and disease resistance,
competitiveness with weeds, consumer acceptability,
fit into local systems, storability.

b. Varietal interaction with agronomic practices and
environmental factors time of planting, soil
fertility, planting density, types of land preparation
and planting methods, rainfall patterns, soil drainage
factors, local pest control methods.

c. Detailed studies of varietal characteristics -
flowering time, head type, lodging and tillering,
yield components,leaf angle.

2. Considerations

a. Understand the multiple uses of local varieties when
developing evaluation criteria for new introduced varieties.

b. Understand the crop management practices used by
farmers when deciding on treatments or management of
non-experimental variables. What background are you
evaluating the new variety in?

3. Implementation and evaluation

a. A variety may not respond the same in relation to other
varieties under different management levels (=crossover)










Ex:LmpEof a vi3.ri ety crc over: 7/'Y


. /


/
- - A


Mean location yield


Poor site


Key: A =
B =
C =
1:1=


Good site


Good only in poor sites
Good in both
Good only in good sites
Where both varieties perform equally
across all sites.


b. Compare the new variety with the local variety under
local management conditions. If using improvedmanagement,
test the local variety under these conditions for comparison.

c. Use farmer's evaluation criteria as your own. Don't just
concentrate on yield.

d. Measure labor input differences, as this may indicate whether
or not the variety will fit into the system.Complete an economic
evaluation.

e. If higher levels of management are needed for optimum
performance (e.g fertilizer), complete an economic evaluation.

f. Plot size should be large enough, especially when farmer
managed, to give an indication of performance under the farmer's
conditions.. Plots should not be too large imposing unreasonable
burden or risk on the farmer.

g. It is advised, that frequent visits be made to the trials to
monitor and note crucial operations such as planting, fertilizer
application, germination rates, pest attacks, etc. Also note
farmer's activities outside the experiment area.


'7-
I 7'


Variety
yield


-7
-A -


X` / C


I*I~
t


r -r










h. Because variability between sites may be large, a comparison of
mean yields from all sites may not give significant results.
Alternative analysis methods include: t-.-tests of mean differences
with paired observations for each site; post-stratification of
sites according to principle site and management characteristics
(see example below); distribution of yields around the mean
(concentrated or skewed).

i. Analyze results over a range of environmental conditions to see
the extent of yield variability.

j. In statistical analysis, correlation is a useful technique to
compare a given variety's performance in monocrop and in an
intercrop. Check for interactions between genotype and over
years performance. If the variety performs the same in
monocrop and intercrop, then the researcher does not need to
implement separate breeding experiments to determine performance.

k. Seek farmer assessment but understand biases.

4. Trial examples:

a. (Peter Hildebrand, 1984)

i. Treatments:. Local maize + fertilizer
Local maize fertilizer
Improved maize + fertilizer
Improved maize fertilizer

1 replication/site; 14 sites in two villages

ii. Indexed: rainfall, temperature, soil fertility.

i'ii. AOV results:

Source Significance level

Total
Varieties (V) NS
Fertilizer (F) *
VX F
Villages **
Error

iv. Graphed combined results of 14 trials of variety (V) vs
fertilizer (F).

Local V had higher average yield and greater stability
for both F situations.









- Used confidence intervals:


(T t. Sr;)






50





Confidence
coefficient
(%)




100


Note: Narrower confidence
intervals were more
satisfactory


Yield


local variety
.... improved variety

v. Fit each V to each F level using regression. An environmental
index (e) was used to test V response to F. (e) was
determined by judgement of poor vs good maize environments
primarily based on rainfall amount and reliability, and soil
fertility.


I
I
Yield I
I


improved

local


0 1


(e)
With Fertilizer
The V's respond differently


to environment:


Local maize superior in poor environment
Improved maize superior in good environment


-~--L ---^I- I ------ -










vi, Partition farms into 2:

a. If e < 2, then the local variety yielded 1.5t/ha and was
superior whether fertilized or nct (9 farms). This
indicated a poorer environment.

b. If e > 2, then local, unfertilized variety yielded
> 1.5t/ha and improved variety is superior.


I

Yield I
i


I


improved

1ocal


O a I 2 3
9 farms 5 farms


vii. Graph the frequency distribution of confidence coefficients
of yields for:

A. 9 poorer environment farms:

No fertilizer Fertilizer



i /
50 \



ConfidenceI
coefficient ./



100 /
i \


Yield (t/ha)

= local variety
= improved variety

Conclude:
.1. with or without fertilizer local variety is better.
2. without fertilizer local variety is superior










E. 5 better environments:


50




Confidence
coeffici ent n
(%)

100


No fertilizer Fertilizer




\ \


yield (t/ha)


= local variety
.... improved variety

Conclude:
1. With or without fertilizer the improved variety
is better.
2. The yield difference is greater with fertilizer
than without.

viii. Points

a. This method measures response to good or poor
environments; hence, if in another year the data
points might shift but environment "3" is still "3".

b. Problems could come if the range were very narrow
causing extrapolation to be extreme or if usual (e)
is much higher or lower.











(Matlon, 1984)

Table 3. Mean yields (kg/ha) of improved and local sorghums by
position along'the toposequence at two levels of management
in level-5 farmers' tests, Nakomtenga and Nabitenga 198i


Low m an agiem nt Hi gh management
E:35--I 38-3 S CSH5 Local E35-1 338-3 CSH5 Local


F'P ateau_
Mean yield/(kg/ha)
Obervations
UFi2.erQ sl p e
Mean yield (kg/ha)
Standard deviation
Observations
Mid selo.pe
\Mean yield (kg/ha)
Standard deviation
Observations
Lowerslope
Mean yield (kg/ha)
Standard deviation
Observations


0

268
286
8

685
609
17

810
645
4


318
1

305
395
7

311
376
16

516
655
6


144
1

773
377
9

537
374
15

602
313
4


189
1

605
473
12

626
459
24

606
525
7


0

996
668
8
b

1405
763
17

1389
1162
4


185 8133
1 1


1048
693
7

915
362
16

1106
799
6


1256
480
9

1369
583
15

1202
1033
4


Poststratification
sorghum varieties,


analysis (Table 3) of mean yields for two improved
one hybrid, and a local variety suggested that local


varieties and, to a lesser degree, the hybrid CSH 5, were more widely
adaptable than E35-1 but that E35-1 was best adapted to fields on the
lower half of the slope under low-input management and to both mid-and
lower slope fields under high management.

Combining poststratification analysis with data on labour use and factor
returns (for the test varieties and for all other farm-level activities
included in the baseline survey) can elucidate probable adoption patterns
and fit within existing systems. For example, in 1980, an analysis of
yields across field locations showed that E35-1 achieved significantly
(P <0.05) greater yields only on fields where it received larger amounts
of organic refuse that is, fields adjacent to family dwellings.
Baseline data showed that these plots are predominantly sown with maize
and red sorghum, budgets were calculated, and the returns to both land
and labour for E35-1 were compared with those for the alternative crops
sown near the compound. The analysis revealed that, on highly manured
soils, E35-1 was significantly more profitable than local sorghums but
not more profitable than maize. Moreover, because maize is harvested
1 month earlier than E35-1, it serves a critical role in providing
calories before the major cereal harvests." This source of food during
the hunger period would be forgone if E35-1 were substituted for maize.


273
1

1102
553
12

1197
454
24

1150
588
7


'-~-~--"~~1------1~1I -~ -----'-- -~ I' ~------~ ---- ---~----`----I









Also, technical budgets showed that soil preparation and planting of the
shorter-cycle and later-pl.anted E35-1 conflicted with the first weeding
of local sorghum. The conflict would be eliminated if E35-1 were
substituted for local varieties. Thus, the improved variety would
probably be adopted primarily on the most fertile soils, but as a
replacement of : local sorghums .rather than maize.

One such budget analysis (Table 4) showed no consistent or significant
differences between E35-1 and the local variety in returns to either
land or labour and no trend in differences as one moved from low to
high-cost management. Although the low number of observations and the
high variation in data makes conclusions somewhat suspect, the local
variety appears to be at least as responsive as E35-1. For example,
in several management classes, the local variety responded relatively
more to chemical fertilizer than did E35-1-. Also, the rate of return to
incremental costs over the base management class (zero tillage and
no fertilizer) tended to fall with the adoption of higher cost systems.
Nevertheless, the marginal return to total costs in the fully
developed system (traction plowing, chemical fertilizer, and manure)
remained attractive for both varieties at between 140% and 180%.









16







m o
C) bO~i (

Itn Co o -o -
0; p 13 m <4 co 0 co co T n 0 in w r
S co co co cn ii co v N
44C .C c, 14 %0 o c'I Hr

0 r v
-H 0u
HOn) Hj 0 0 C) o o ,
I () (C P Vn '0i 0co (1 0 co co (1r w
O ( Cr 10(1 C(N 1 tn -4 r-q Lo C
O N3
W c
H U

SC O. O LI OC, Or, O V 0

SL 0 .-O)( CJ 0- Z


'-4

M 0a
C g U44 -. C-) (N (N- .7

0( 'ro 0

QC ico l


Q 1O '0 (N 4 0, 01 n u




p rq C Q) co i/-1 .m l a, \
a) t-o00 CE. r-O) (104 WCV ('4IT C(


0 C: *- '0 '0 0 -
0o 0 In COV) c 0 c



0L(Ci (' C C-N (N (n ( >0
C C) H 0. u n -I (1 0D I OD H 1





(c* O'

c o 0 H u C 0 -0 H O
0O 0l Ox( N O

S,-I El 43




0
-41
H( (1( %D )o 4 co (O H cp u






O N0 r- 'D "0 v 4 U
H o ro 0 H 'rs i-q rr c o -( IC






:t '' `0r. CC I O V- H T O U C3
SC r- O0 m m 04 w
0 .0 o 0' ) H 3 QN i C) U
or 0 .0 T o









'A) (3 C14 I c- 4 .o 0 C 0 0
C) Q! Its (n C) I)C tz H
to C 0 n IT 0 n' q -( ol In ^ C C 0
5( co 1 --i r- C4 o r- N
C H 0t- ( N I N

H2 0 H m N o, 0- cu H C
O ".4 OD H H C. 3 Q
c '-1 r, V t

0) 0'C C) CIO
0 m C co VII r o N 1-. q 0 ci :

m I*i "L ( CN "I a)CM 00 C)
0 ) C) LI -4 0C3 -
GI p > w >
0 c m* ,, 4
CO CC CC r H 0 '- r-H- C ( C)






v c- C > (N rN( oc (C
:0 3 C:
H* .- .-. -o C) (C C) e h ~ r 0

9) C) 0-C-C H 0C' C H- 0 0









0C Cl CB (10 ( 0 0
0 v 1i 0 r-






C LI 04 (0 60 0Or) vJ U H 0 0o C) C3
:J O)LI 0 -C HH H w & 4
'4 0 .7HC- o Hrf (NO sj '- (C ). J






w 0 C 1 cP 0 0 C3 C -- .03ia
Hi N g C) 00 a 0( 3 Q 3 c

o CC> BC3
(y n B ri ci0 sA o o) 0(C C n xC 6 )




0 3)0 0s"










2.3 Weled, Insect and Disease Pest control experimimnts

1. Typecs of eor.iments

a. Biological control, pesticide rate-;, timings or frequencies
at certain crop growth stages and methods of application,
variety testing for resistance or tolerance, cultural
practices; or a combination of methods. Cultural practices
may not he directly related to removing the pest but
indirectly affect ng thle :incidence of the pest, e.g. time of
planting and insect presence; or land preparation method and
the effect on weed germination, etc.

b. Study of pest biology behaviour under varying circumstances
including scouting for incidence of various life stages.

c. Study of disease transmission process including vector
biology.

d. Storage pests:
This is not an. all inclusive list of possibilities. An
example of various roles of entomologists are reviewed in
the following table.

2. Implementation:-

a. Consider crop/animal practices concerning rotations, grazing
patterns (weeds) etc., and how this may affect pest incidence
as this may interact strongly or be affected by farm manage-
ment practices. Gather information from the farmer concerning
these aspects as well as environmental interactions (weather,
soil etc.)

b. Consider evenness of distribution and incidence of pest in
study area. The researcher may need to scout in advance to
find areas where the necessary ]evel of pest incidence is
present.

c. Identify current farmer pest control practices. Ascertain why
farmers think they work. Use these as a 'control' comparison
treatment.

d. Trials, by and large, will be managed by researchers when new
risky practices are used (pesticides) especially when screening
these techniques.

e. When screening new tools (for weeding, spraying, etc.) the
researcher may want to have a preliminary test with farmers
(completely farmer managed) in the exploratory experiment
stage to determine a major tool design flawQ.











Division of roles for
systems sites:


entomologists in experiment stations and cropping


Pest control method
researclh-producti on


Ecology and pest
management






Chemical control











Host plant
resistance


Cultural control











Biocontrol


Basic research activity Applied


(technology generation)

Taxonomy
Pest bionomics
Economic threshold
determination
National pest control
recommendations.

Screening (efficiency)
Dosage
Formulation
Method of application
Timing and frequency
Residues
Phytotoxicity
Environmental impact
assessment
Toxicology

Varietal screening
Mode of resistance
Genetics

Seasonal effect
Spacing
Fertilizer
Tillage
Trap crop
Intercropping
Crop residue management
Crop rotation
Crop maturity
(Micro level studies)

Taxonomy
Natural enemy effective-
ness and bionomics
Introduction of exotic
species
Augumentation (mass
release)
Conservation


activity (technology
specif icat ion)
Pest complex determination
Population assessments
Target farmer (behaviour,
resource level, and
managerial capabilities)


Timing and frequency
Method of application
Minimum dosage









Verification of resistance
Deselection of susceptible
varieties

Planting time
Synchronous planting
Crop residue management
Tillage
Removal of alternate hosts
(Macro level studies)






Natural enemy complex
determination
Populations of natural
enemies



Conservation


(Zandstra, et al 1981)


I -- -- -I --


----


--










f. In screening, pesticides try to choose levels that are
compatible at farmer's resources to start with.

g. A greater number of Field sites may need to be sampled to
guard against variability or failure of pest incidence.

h. As a control treatment, a '0' control level may not make any
sense to include lnl.ess it is the farmer's practice.

i. One may need to adjust the layout of weed experiments in order
to cover more uniform areas. By trying to control variation
within each site the researcher can more easily analyze
treatment response.

j. Plot size for insect or disease trials in particular may need
to be large, depending upon the type of trial. In weed manage-
ment work, consider having larger plots for labour measurements
or for cultural practice implementation. Use smaller plots
for screening work.

k. Use local measures for the amount of pesticide applied in the
equipment the farmer would most likely use.

1. Monitoring of insect control trials may take greater manpower.
Consider this in planning a program. Also consider training
needs of technicians in advance. Work out an easy-to-follow
methodology.

m. Insect and disease trials need perhaps longer repetition over
time because of variations in pest populations year to year.

3. Evaluation:
a. Pesticide use:

i. Is the increased yield level achieved by use of
pesticides large enough to make the practice econo-
mically viable for the small farmer?

ii. Are materials being used too hazardous, considering
the conditions and management level of the small
farmers?

iii. Are the support services adequate, if the farmer is to
depend on this as a method of control? For example,
are supplies of materials, equipment, maintenance
requirements and spare parts etc.adequate?

iv. Is there extension assistance for training inexperienced
farmers? For following up usage problems?










v. Is the outlay of cash rather than use of labour
(especially in weed control) problematic? What are the
social trade-offs in terms of possible 'displaced'
labour?

vi. Consider risk of alternative methods.

b. Cultural practice:
i. Is the timing or method suggested compatible with the
farming system? (labour, resources, other activities)

ii. Are the tools (or varieties) available in large enough
quantity to enable farmers to adopt the practice if
they like it?

iii. What are the economic and social ramifications? Returns
to labour?

iv. Consider risk related to methods.

4. Examples:
a. Treatment arrangement: (Zandstra, et al, 1981)
Experimental design for yield-loss assessment and
determination of the optimal insect control recommendation
for transplanted rice. The treatments listed are repeated
a several field sites.


E I 1. Complete protection (high level)
[ I
E Seedbed 0.75 kg a.i monocrotophos/ha sprays
C 7 and 14 DE
E Vegetative 1.0 kg a.i monocrotophos/ha sprays
C 5,15, and 25 DT
Yield Reproductive kg a.i chlorpyrifos/ha sprays,
loss E 35, 45 and 55 DT
assess-[ Ripening 1 kg a.i. Y-BHC/ha sprays (x3) weekly
ment E intervals after flowering

EC 2. Omit seedbed protection (high level)
C I 3. Omit vegetative protection (high level)
C I 4. Omit reproductive protection (high level) _
C[ 5. Omit ripening protection (high level)
C[ 6. Untreated control
7. Recommended practice 1 kg a.i carbofuran G/ha soil
incorporated plus economic thresholds I
8. Alternative practice 1.5 kg a.i diazinon G/ha
broadcast 5 DT plus economic thresholds
9. Low-cost technology Economic thresholds

800 to 1000 m2 field











b. Weed control evaluation (Hook, 1979)

Let us look at this data from the present farmer's point of
view, and the allocation of his and his family's time. Which
treatment will give him greatest return on time invested in
weed control This data is shown in table 2.


Table 2 Rice Yields Compared to Weeding Times, and Returns per
__ Hour Invested in Weeding (3)


Treatment


No weeding
Hand weed
Hand weed once
Single rotary
Rotary weed/hand weed
MCPA/Rotary .weeding
24D/Rotary weeding
Rotary weeding MCPA
Rotary weeding/24D
Dichlobenil/MCPA
Propanil/MCPA
Dichlobenil


Weed weight
_m/m2__
650
0
92
287
10
112
87
157
146
293
147
350


Weeding time
Man hour/ha


616
323
152
368
104
109
143
116
11


Clearly from this standpoint the 'best' treatment (dichlobenil)
leaves 350 gm per sq. m. weeds at harvest. This is low down on
the % control curve. The 'worst' is the clean weeded treatment
which produced the greatest yield because of the large number
of weeding hours. The traditional weed research conclusion from
this experiment would be that none of the treatments out yielded
the control, and frop the chemical companies point of view, no
recommendation would be forthcoming, because yield is tradition-
ally the sole mechanism of comparison, and not yield related to
effort.

2.4 Cultural Practice Experiments:

1. Types of trials:
a. Plant population trials include: between row and in row
spacing combinations, planting patterns, thinning, etc.
perhaps comparing different varieties; planting dates.

b. Intercropping trials including various cropping patterns,
populations, time of planting of various species on a given
area in relation to other species.


Yield
Kg/ha
Hour
1611
5201
4122
3917
4473
4174
4140
3414
3218
3642
3664
3098


Rice/
man


5.8
7.7
15.6
7.8
24.7
23.1
12.6
13.8
19.2
19.4
28. 1


- -1--~1-1-'











c. Tillage or planting operations including primary and
secondary operations: land preparation, seed bed preparation
and planting methods, secondary cultivations for hilling,
ridging, weeding, creating a soil mulch, transplanting, plant
establishment problems. Consider timing, methods and tools.

d. Maintenance operations including mulching, pruning, topping
or stripping, staking, culling, etc. method and timing.

e. Water management including irrigation management or water
harvesting (rain management) operations method and timing.

Note: In reality, experiments often include more than one of
these factors.

2. Implementation:
a. Understand what the farmer is presently doing and why.
Understand yield variability and inherent risk in the system
(see Fig. 1 & 2)

b. Map labor distribution and consider the effect a change in
actual practices may have on this factor in the treatment
choosing stage.

c. When choosing treatments, think of the implication of the
proposed change on other activities. For example, by
increasing plant density, can weeding be effectively carried
out.

d. Farmers, when assisting in experimental management, may nebd
a "training course" in advance to fully understand how and
what they are supposed to do.

e. It is very important to have the farmer involved in the
management of these types of experiments at an early stage.
This is due to important treatment and management inter-
actions.

f. Watch out for treatments that don't make sense, e.g. weeding
at a given time when there are no weeds; applying fertilizer
when a poor plant stand may not justify the application.

g. Define intercropping experiments carefully to avoid compli-
cated and too many treatments when working with farmers.
A superimposed trial can be very useful in this type of
experiment.

h. Replication over several sites (multi-locational) for these
types of trials is useful in helping to expand the inference
of the trial.










i. Plot sizes will tend to be larger when trying to get a
realistic measurement of labor inputs, ease of carrying
out a given procedure, etc.

j. Data collection concerning farmer management and environmental
factors may need to be more detailed and well-planned in
advance. Field workers may need training in this aspect. This
information is often necessary to interpret variable results.
Because farmers are often involved in the treatment or non-
experimental variables management in this type of trial, the
possibility of misunderstanding or changing the experiment is
greater. More detailed data collection safeguards against
having unexplained results.

3. Evaluation:

a. Labor demands and timing of operations is crucial to determine
when comparing practices. This will help determine system
compatability and farmer acceptance. Often, a new cultural
method is under trial, not to increase yield, but to help
re-distribute labor to diminish labor-bottleneck problems.

b. Farmer management and environmental factors play a large
role in determining the results of these types of trials.
Evaluation procedures must take this into account.

c. Consider the risk related to a given cultural practice when
evaluating treatments.

4. Examles:

a. The following figures, tables and graphs were extracted to
illustrate comparison of traditional and recommended
technology of both a maize-bean intercrop and maize sole-
cropped. Figures demonstrate variability of performance in
different management classes; tables illustrate economic
comparisons; the graphs illustrate the certainty of
performance of the two technologies.

Note: The performance of the recommended technology is not
always best or most consistent.












BOTSWANA: Clive Lightfoot, FS Newsletter No. 6(1981)


Figure I Annual Variation in Sorghum

I


Yield
Kg/ha


Productivity


LIVESTOCK AND CROP SURVEYS'


Figure 2 Distribution of Total Farm Production



:i











0




.1*
.r( ..

C f -
c


o
0, 8
n
1
N


.. 0

0n -


Example of -yield variability and inherent risk in system









Testing of Technology with small Farmers in Jinotega, Nicaragua
by Icaza, Lagamann, CATIE, 1982.


Figure 1: Distribution of Gross Returns of a

Traditional
Technology


Freque-
ncy
(in %)


50:

40:

36:

20:

10!


1 2 3
CLASSES


50:
I
401

30

201


4 5-


1= <5000 C/ha
2= 5000-8000
3= 8001-11000 C/ha
Figure 2: Distribution of Maize Yields in
i I


Freque-
ncy
(in %)


40:

30:
I2
201
10
10:


40

30!

201


1 2 3 4 5
CLASSES
Figure 3: Distribution of Maize Yields in Los
i


Freque-


40:


30:

20!


ncy
(in %) 10!


1 2 3
CLASSES
1= <1000 kg/ha
2= 1000-1499 kg/ha


4 45


40

30:

20

10


3= 1500-1000
4= 2000-2500


Maize-Bean Package in Suni


Recommended
Technology


1 2 3 4 5
CLASSES
4=11001-14000 C/ha
5= >14000 C/ha


Sisle


Robles


CLASSES


CLASSES
kg/ha 5=>2500 kg/ha
kg/ha


--I


i, ~ ~ .


cl--~-----


I I











Table 6: Comparison of gross between the farmers' and
the recommended technologies gin the Sisle area.


Farmers' technology
(n=8)


Recommended technology
(n=8)


Gross margin (C$/ha)


Average
Min.- max.
C.V. (in %)


2.479
443 4746
54.3


2.638.
53 7093
102.6


Gross marain/man-day


Average
Min. max.
C.V. (in %.)


50
6 135
70.1


46
9 172
105.5


Table 7: Comparison of gross margins between the farmers' and
the recommended technologies in the Loss Robles area.


Farmers' technology
(n=10)


Recommended technology
(n=13)


Gross martin (C$/ha)


Average
Min. max.
C.V. (in %)


3850.
807 7526
53.7


2304
493 4673
62.5


Gross marain/man-dav


Average
Min. max.
C.V. (in %.)


51
20 75
41.3


31
7 62
61.8


--*-----II---------~---I-----'I


,,,~I -- I--I I-


--


-- ----------


- ---------------------- -- --- --










Figure 5: Cumulative Distribution of Gross Margin/ha in the Sisle Area.


Cumul. ati ve
Frequency
(in % )


100

90 1

80

70 "
A I
*O J
60 /
/
50 1 /

40 /

30


A /
20 /

10
-----O-----


1000 2000 3000 4000 5000 6000
GROSS MARGIN (IN CORDOSAS/ha)


Cumulative
Frequency
(in %)


100

90

80

70

60.

50

40

30

20 1

10


/
-I


0 1000 20C

Recommended Tec


0 300 4000 5000 6000 7000
GROSS MARGIN (IN CORDOBAS/ha)
:hnology ..". Farmer's Technology


7000


-- ----------------------------


,











b. The following table illustrates a trial completed at 20 sites
(Francis and San ders_1978)_


i. Bean cultivars, densities and designs
wiLthL tree cro-_ inp s p. ys tes in CAT.
Trial Planting date No.of Bean


replica-
t ions
5
3
4
4

4
4
4
4
4
3
4
4
4
2
4
3
4
4
4


'RCB = randomized complete block,
SSP = split-split plot.


culti var


P259
P259
P259
P259
P259
P6
P6
P259
P589
P589
P589
P526
P589
P589
P589
P589
P589
P589
P589
P364


in 20 field trials


Bean
density
(1I000/ ha)
.44
250
300
160
320
320
160
320
160
160
160
160
160
160
160.
160
160
160
160
160


Experiment
design'

RCB
SP
SP
RCB
RCB
RCB
SSP
RCB
RCB
SSP
SP
SSP
SSP
SP
Lattice
SP
SP
SP
SSP
SP


SP = split plot,


ii. The reduction costs for three croonina systems at


Cost item

Land preparation
Seed: bean
maize
Planting
Irrigation
Fertilizer"
Weed control
Insect and
disease control
Harvest
Support system4
Fixed costs
Total costs


System
Maize
1,500

300
200
500
1,300
900

700
6,600

6,000
18,000


(100)


(200)
(300)
(200)
(900)


5, OOC
1,100

300
500C
500C
300


(100) 1,60C
(6000) 5,700
10, OOC
(7800) 31,00
(7800) 31,000


Bean
(5000)


(300)
(300)
(100)
(300)

(200)
S(5000)
(5000)

(16,200)


Bean-Maize
1,500 (100)
1,100
300
500 (500)
500 (300)
1,300 (200)
300 (300)

2,300 (300)
8,200 (7500)


6 aQ00
22,000


(9200)


7501
7506
7507
7509
7510
7513
7514
7515
7516
7517
7518
7525
7602
7605
7606
7614
7617
7618
7619
7624


Feb
May
Jun
May
May
Aug
Aug
Aug
Nov
Nov
Nov
Dec
Jan
Apr
Jul
Aug
Aug
Aug
Aug
Oct


1975
1975
1975
1975
1975
1975
1975
1975
1975
1975
1975
1975
1976
1976
1976
1976
1976
1976
1976
1976


CIATi


--. 1 ---1-1..---1- 11----- ------ --------


ii-~- iCIAT I-


II- -~-- I- ----^I----


I- --'-~-~-- -


--- ---~ II- -------- --










1 Costs in Columbian pesos/ha, approximate conversation
CP$30 = US$1; labour cost in parentheses following each figure.

2 Land preparation by.hand was necessary to preserve the support
system in the field.

s Nitrogen (urea) and complete fertilizer were applied to all
systems, but only charged to costs of maize and associated
cropping; only the cost of complete fertilizer.was charged
to monocrop beans.

4 Support system costs for wild cane stakes (useful life of
three seasons) include labour costs for installing the system.

s Costs include interest and depreciation of fixed capital.


iii. Net Income from three cropping systems at several bean/maize
price ratios

US$/ha Monoculture
beans'
601 2,000

50

40
Net income
(CP$1,000 301 1,000
/ha) Beans + maize2
201

10
Monoculture
a maize:
o1 -- --- ---
O F

7 2 3 4 5 6 7 8
-10o Price ratio (beans/maize)

-20



'Monoculture bean yields 3,000 kg/ha; costs CP$31,000/ha.

"Bean yields 1,200 kg/ha + maize yields 5,000 kg/ha; costs
CP$22,000/ha.

"Monoculture maize yields 5,000 kg/ha; costs CP$18,000/ha.










3.0 LIVESTOCK EXPERIMENTATION -"In-herd or On-range trials"(ORT)

3.1 Differences in ORT compared to OFE with crop research:

1. Animals interact with crops. Crops represent 'by-product'
as feed.

2. Most animals have a long production cycle; therefore,
need to know more about the farm or ranching system
before one embarks on ORT.

3. Animals don't always produce a final, measurable product-
work, for example.

4. There is more emphasis on the farmer than on the site
due to management factor having an over-riding influence.

5. It may be more important to understand motivation, interest
and social consequences for ORT, eg. ownership complexities
affecting management of animals, food and water.

6. It may be more difficult to have more than one replicate
on a single farm; therefore, there may be a need
to have a large number of farmers. In this case, animal
equals site.

7. ORT have longer commitments and more risk involved so
may require more in the way of complementary services
or greater compensation.

8. Animal systems are mobile or nomadic in character vs.
stationary in crops.

9. Asychronous production aspect may make ORT more
difficult to implement and monitor.

10. Animal nutrition information is difficult to get in ORT,

11. Time demands on the farmer may be greater for participation.

12. Local animals may be used on a research station as an
intermediate measure to complete isolation from the
field or complete involvement.

3.2 System components to consider (ILCA):

1. Livestock demography, reproduction cycles, growth, health

2. Pasture primary production, cover, browse available









3. Agronomy crop mixtures and patterns, yield, disease,
soils, etc.

4. Examples of system constraints:

a. Mali poor reproduction in sheep and goats, high early
mortality, abortion in Twareg goats.

b. Kenya Masai- delayed age at 1st pasturition, long
pasturition intervals, high proportion
intervals, high proportion of unproductive
stock, recent reproduction problem.

c. Sudan dry season feed shortage.

3.3 Reasons to do ORT with animals:

1. Follow-up and evaluate farmer's management practices in
forage production, identify available inputs, record
timings of activities, yields under farmer conditions.

2. Monitor utilization of crops by animals

3. Estimate animals' feed intake in normal setting.

4. Assess farmer's reaction to the approach and willingness to
adopt.

5. Get feed-back for experimental designs and new approaches.

3.4 Type and management of experiments:

1. TyPes
a. In Sudan, a nutritional deficiency was identified as a
problem during the dry season. Researchers needed
specific diagnostic studies to investigate the following
items to clarify the problem:

key nutritional factors
time limitations on the problem
class of animals affected
impact on herd production
cause and effect relationships (interactions)

b. 'Sentinel' herd -This idea is similar to a 'model' farm
in that the local animal herd is managed by a paid
worker. It does not simulate a true socio-economic
background but helps overcome logistical problems
and may help to reduce time frame of experimentation.


Generally, experiment types are similar to those done










with crops.


c. A food crop management/livestock feed interaction was
studied in Ecuador. The maize food crop was managed
to optimize'returns both for human and livestock
production. The crop thinnings, leaf strippings and
late season weeds were fed to livestock. A reduction
in human food yield was realized as a trade-off.
Researchers studied time and amount of thinning, time
of leaf stripping and weeding and maize varieties to
determine the best management considering the farmer's
objectives.


2. Management:

Management ranges from producer managed (PM) to research
(RM), similar to OFE.

a. In Sudan, a trial was conducted by:
RM = only used producers animals and land. The advantage
was lack of risk for the producer as long as he
compensated for loss.

PM = Researcher recommends a practices) and leaves it to
the producer to manage and adjust the recommendations.
The researcher monitors the adjustment.

b. Joint PM/RM use farmer's site, herd and some labor.
Researcher oversees the experiment implementation because
of trial complexity.

c. In Sudan, researchers did a PM (Producer Managed) feeding
study. They had problems interpreting the results because
they didn't have enough base-line data on the biological
potential of feeding hay to traditional cattle. Therefore,
they needed to have done RM trials first.


3.5 A sample of trial design questions: (in reference to dry season
supplementation)

1. Are producers presently exploiting the productive potential
of their livestock, and how effectively is the resource base
being utilized?

2. Is there potential for improvement through supplemental
feeding?

3. What level of output should be realistically achieved with
supplemental feeds?










4. What should be the aim of supplemental feeding to enhance
survival, maintenance or production?

5. What levels of inputs and options are currently available
in terms of cost and abundance?

6. What level of input is necessary to achieve the objective
of supplementation?

7. How would the proposed use of available inputs conflict
with existing uses and other activities?

8. What is the nature, time-frame and cost of the research
effort required for testing and evaluation?

9. What proportion of the target system is likely to be
affected?

Evaluate the above in terms of priorities and objectives of the
producers; the dynamic and interacting components of the system;
the producers ability and willingness to adopt; trade-offs with
the new technology.

3.6 Experimental design considerations:

1. If a small number of observations per farm are used,
performance of individual animals can have a large effect
on farm averages and on variability between farms.

2. Trade-offs to consider: If you limit number of treatments,
increase.number of farms or increase sample size per
treatment then you may have to reduce secondary data
information collection inorder to handle increased number.
A researcher cannot do this; however, at the expense of
understanding the results.

Or extend duration of trials which will increase the number
of observations and the reliability of the results. This
would help diminish the problem of within farm variability.

3. If the sample is too small, statistical analysis will be
difficult.

4. lt is often difficult to have replication at a given site
because plot size is usually large. Use one replicate per
site. This also encourages farmer understanding and
observation.


5. Two experiment examples:









a. 2 seasons x 3 treatments (feeding regimes) x 10 animals

b. Divide each flock into 3. Each one grazes 2 experimen-
tal forage crops, and the third is the control under
current farmer practice.

3.7 Types of measurements

1. Forage yield and parameters (nutrition, digestibility,
palatability, preference, etc.)

2. Animal performance data milk yield and production, weight,
(multiple measurements necessary), reproductive data,
mortality, disease.

3. Nutrition data 'input' and 'output' vs growth use of
fistulation of local animals for monitoring purposes.

4. Feed management grazing

5. Socio-economic considerations ownership, who manages,
communal aspects, markets, prices, compensation or trade.


3.8 Arrangements with the farmer

Similar considerations in farmer arrangements should be
made compared to OFE, with the following exceptions:

1. Farmer compensation may have to be more generous or
complicated.

a. An example from Egypt:
= basal payment of 100 EGP/mo (1 USD=.24 EGP) to
head of family
= frequent gifts of sugar, flour, clothes, tea
= rental fee for each studied animal (1 EGP/mo/head)
= salary to shepherd (30 EGP/mo)
= beginning of each experiment 15-30 EGP to head
of family
= salaries to assistants as needed (collect forage
species, etc.)
= veterinary care

b. From Sudan:
= additional land
= more frequent waterings
= free supply of inputs not commonly affordable
= technical advice in exchange for strong commitment.

2. Farmer participation problems that can be encountered:











a. Farmers fear loss of production and earning if feeding
treatments are unsuccessful.

b. Farmers fear extra costs imposed.

c. Extra labor related to treatments may be difficult
to find.

d. Diets may be interrupted due to lack of feed resources.

e. Lack of time for farmers to accurately record data.

3. Examples

a. (Sudan) Operator supplied feed and water troughs,
balances, feed storage area, separate place for experi-
mental animals, labor. In this case, the operator was
an owner of a feedlot. The researcher selected the
animals, monitored the treatments and results. No written
agreement was used as this would have offended the
operator.

b. (Nigeria) Researchers found that if veterinary care
(vaccinations) was given free, it created suspicion;
therefore, they decided to charge a nominal fee.

c. (Kenya) Farmer agreement was dependent on the
intervention and farmer's resources required. Initially
researchers supplied goats and if they got good coopera-
tion, the farmer could keep the goats after a one year.
period; however, the farmers were not allowed to sell
the animals. They were given free inputs (vaccinations)
and pay for labor. Bucks were distributed free but the
researchers maintained ownership.

d. Farmers were told about the experiment but not which
treatment they had. This was hoped to help eliminate
bias.









3.9 Choice of farmers and/or animals

Factors to consider when choosing farmers is illustrated by the
following examples:

1. (Egypt) Researchers were introduced by colleagues from
the area; thus were viewed as acceptable by the community.
The researchers selected a Shikh and 5 sons to cooperate.
This family was of high standing in the bedoin community and
customarily everyone follows this most influential person.
The researchers felt this would be a good way to enter the
community and gain the confidence of the community
concerning their interventions.

2. Choice of farmer may depend on the farmer's ability to
provide the needed number of a certain type, age or sex
of animals.

3. (Sudan).- Researchers chose non-lactating female animals for
an experiment because they knew that people sold milk during
the dry season (from lactating animals). Researchers did not
want to disrupt this cash resource.

4. (Sudan): Eliciting farmer trust and confidence: Farmers must
trust that you won't hurt their animals; otherwise, their
cooperation is difficult to get. In this case, researchers
decided to manage their own herd of local animals.
They initially had a problem of attaining research animals
because of owner's reluctance to sell the type needed.
Researchers then paid a worker to manage and move the animals
around as realistically as possible. They designated total
authority to this manager. Eventually, other animal owners
saw that the experimentation was not detrimental and their
confidence was slowly attained. Eventually, researchers were
getting requests from producers to involve their animals in
the study.

5. (Syria): In an experimental program Locations were chosen
to represent different ecological, farming and livestock.

Management systems (3 sites)

= Other criteria: soil type; communal vs limited access
grazing; rainfall.

= Site and farmer choice: considered trade-off between
representativeness within stratification and accessibility.

Cooperative Criteria:
one field/farmer
one resident, responsible adult










no exceptional farmers
S *no closely related farmers
*own at least 21 breeding ewes (grazing trial)
(6/treatment and 3 reserves)
suitable soil type
previous cereal crop
typical cultivation and fertilization history
accessible by vehicle
minimum size 1.2 ha to 0.2 ha grazing trial fields
not too close together
safe from accidental grazing
(* would make exceptions, others absolute)

3.10 Evaluation criteria:

1. Must keep ownership complexities and their relation to
management factors in mind.

2. Remember multiple objectives of livestock keepers food,
sale, intermediate goods (draught, manure, milk, security).
Weighting of these is difficult as animals can provide
several of these at a time.

3. Researcher vs. farmer criteria: Example researcher
looks at shortening the calving interval whereas the
farmer looks at increasing hardiness. Try to match
objectives.

4. Farmer feedback:
Reactions of interest may be confounded: to having a
trial on their land with their animals and to the tested
technology itself. Must design questionnaires carefully
to sort this out.

5. Example (Indonesia): Researchers originally chose farmers
on the basis of land ownership (main stratification) and then
by animal ownership (sub-stratification). When researchers
did analysis of data, they ranked farms first by mortality of
young, then into five production classes (both of which are
management related).

3.11 Examples:

1. Nigeria:
Objective: To study the yields of sorghum and millet
and stover consumption by livestock.
Tested the crops (sorghum and millet) in the farmer's
field to see what amount was left and available to cattle.
Found a large variation in yield due to planting time
differences.
Identified a strong correlation between yield of grain









and dry-matter stover amount. Could use grain yield for
predicting stover yield which would otherwise be difficult
to ascertain by sampling stover alone. Also could predict
stover quality from the grain yield.
Analyzed leaves and stalks separately because they
differed in content and animal preferred eating choice.
Use of the farmer's field gave an insight into farmer's
practices which affected the amount of feed available.

2. Alley farming program (ILCA) Nigeria:

Alley farming is where multipurpose trees are established
in rows with space in-between for annual crops. The trees
are used as fodder for small ruminants. Grazing is managed
between the rows during the off-season.

Site selection Used extension agent's knowledge of the
area and an aerial survey to locate promising villages.
At the village level, they held farmer meetings where
researchers used colorful posters to introduce the idea
and create interest. Interested farmers were investigated
to see if their management levbl and the site matched
objectives. Researchers picked 60 farmers.

Implementation There was a large, initial farmer
involvement. Farmers were given lessons prior to the trial
establishment and were taken to view a pilot on-farm experi-
ment to see the trees established. They were given the
seeds and were allowed to experiment as they please. The
researchers and assistants monitored what each farmer did.
The objective was to see if alley farming was practicable.
When monitoring, they noted such information as the growth
and establishment, crops chosen for intercropping, better
vs poor crop combinations and other management factors.

3. The following Figures and Tables represent interesting
information concerning a research program reported by:
Tully, D; E.F. Thomson, R. Jaubert, T.L. Nordblom. On-farm
trials in Northwestern Syria: Testing the Feasibility of
Animal Forage Legumes as.Grazing and Conserved Feed. Farming
Systems Program ICARDA, P.O. Box 5466, Aleppo, Syria.









FIGURE 1:


Supplementary feed


Grazing areas'


Direct grazing of
barley


Barley stubbles
after harvest


Crop residues from
irrigated areas


Movement of sheep
to wetter areas


LIVESTOCK FEEDING CYCLE AT BUEDA


i a I a a
a


a I I


a-I


Nov


Barley
production
calendar


crop sown! --->


a a a a


Dec Jan Feb Mar
A


Tiller --->


a a
a a


7] a


Apr May Jun Jul Aug Sep Oct

------------
Anth ---> Stubble !Land a
!cultivated!


---- i --- _____ --Har
Graze Hay Harvest


Production
calendar
for forage
legumes


Rain/month (mm)


ILand a
Cultivated!


crop sown
------------
60

40 _

20

0 1__

Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct


1 Village based flocks utilize local marginal grazing areas.
Transhumant flocks move to steppe land in spring for grazing.
Source: Jaubert and Thomson 1985. Solid lines represent typical patterns;
dotted lines represent observations in the dry 1983/84 season.


---


Nov


c=











FIGURE 3: BASIC PLOT LAYOUT, COMPLEX ROTATION TRIAL

----------------------------------40 m------------------------------------



-P +P//// /////////// I
-P / / +P/ / / / / / / / / / / / /
----------------------------------LATIHRUS-----------------------------
\ \ \ \ \ \ \ \ \ \ \ \ \ \ -P \ \I\/\/\/ +P/\/\/\/\//\/\/\/\/\/\/\/\
\ \ \ \ \ \ \ \ \ \ \ \ \ \ +carbofuran \/\/\/\/\/\/\/\/\/\/\/\/\/\
\ \ \ \ \ \ \ \ \ \ \ \ \ \ \ +carbofuran \/\/\/\/\/\/\/\/\/\/\/\/\/\
\ \ \ \ \ \ \ \ \ \ \ \ \ \-P\ \ 1/\/\/\/+P\/\/\/\/\/\/\/\/\/\/\/\/\I
--------- ------------------------VETCH ------------------------------
-P / / / /+P / / / / / / / / / / / /


-P / / / / P/ / / / / / / / / / / / /
-P i / +P/
-------------------------------LENTILS ----- ---------------------- m
\ \ \ \ \ \ \ \ \ \ \ \ \ \ -P \ \I/\\/\/+p/\/\/\\/\/\/\/\/\/\/\/\//\I
\ \ \ \ \ \ \ \ \ \ \ \ \ \ \+carbofu.rran/\/l\/l//\/\/\/\/\/\/l\/\/\/\ !
\ \ \ \ \ \ \ \ \ \ \ \ \ \ \+carbofuran\/\/\/\/\/\/\/\\/\/\/\/\/\/\
\ \ \ \ \ \ \ \ \ \ \ \ \ \-P \ \I\/\/\/ +P\/\/\/\/\/\/\/\/\//\/\/\/\
------------------------------- FALLOW-------------------------------
-P / / / / P/ / / / / / / / / / / / /I


-P / / / /+P/ / / / / / / / / / / / /
---------------------------------HBARLEY--N----------------------------
\ \ \ \ \ \ \ \ \ \ \ \ \ \-P \ \!/\/\/\ +P /\/\/\/\/\/\/\/\/\/\\
\ \ \ \ \ \ \ \ \ \ \ \ \ \ \+carbofuran \/\/\/\/\/\/\/\/\/\/\/1/\/\
\ \ \ \ \ \ \ \ \ \ \ \ \ \+carbofuran \/\/\/\/\/\/\/\/\/\/\/\/\/\/\
\ \ \ \ \ \ \ \ \ \ \ \ \ -P \ \ I\/\/\/+P \/\/\/\/\/\/\/\/\/\/\/\/\
--------------------------------BARLEY-+N----------------------------
-P !\/\/\/ +P\/\/\/\/\/\/\/\/\/\/\/\/\
1/\/\/ +P\/\/\/\/\/\/\/\/\/\/\/\/\


SECOND YEAR PHOSPHATE TREATMENTS


= +p = -P = -p = +P =
B _I__ 1______













FIGURE 4: TYPICAL NESTED FLOT LAYOUT (HARVEST TRIAL)


_______ _....._____ I, /






I-- --------------------------(
ROTATION NO 1


ROTATION NO 2
I----------------------------
ROTATION NO 3
-._-------------- .----------- -.-

ROTATION NO 4
----------- -------------- -- .
ROTATION NO 5

S ------------------------
I ROTATION NO 6
- I--- --------------- i


'I


S- - - 200 m -- -



FIGURE 5: BASIC PLOT LAYOUT, SIMPLE ROTATION TRIAL

' / I-I / /I7I I 7 -- / I I, V77 TCI I 7 7 7-7-7-I7i / I- I 1-7
, / / / / / / / / / / / / +P / / / / / +P / / / / / / / / / / / / /
--------------------------------VETCH------------------------------------
I -P -P
II
I^ ^ ^ _ ___ __ .- -- .. S_...


---------------------- -------------------------------------------
--------------------------------FALLOW---------------------------------
I---- - -- ------------------I--= == -~==
/ I I 1 I/ / / I / I / /I / / / I//I / / / / / / / / / / / I

1/ / / / / / / / / / / / / +P/ / / / / /+P / / / / / / / / / / / /
-----------------------------LATHYRUS-----------------------------


-P -P


---------------------------------------------
------------------------------ FALLOW-----------------------------------


:I / / I 7 7 7 / / / / / / I I / / I /7 / / / 7I~~77
I/ / / / / / / / / / / / / +P / / / / / +P / / / / / / / / / / /
S-------------------------------PEAS----


-P -P
i i


I


I !


60 m

I I


I
I
S

I a
I

I I


----- ---


------~~~~-~











TABLE 1: DETAILS OF TRIALS DESIGNS

Complex rotation trials:
Treatments compared:
6 rotations: Lathyrus, vetch, lentils, barley
+ N, fallow, followed by barley
+/- P20z (46 kg/ha)

+/- Carbofuran (30 kg/ha)
Treatments controlled:
Legume seed rate (150 kg/ha)
Barley seed rate (100 kg/ha)
Uniform innoculation
Trial size: 0.2 ha (24 plots x 80 m")
Replication: 4 replicates in each of 3 locations

Simple rotation trials:
Treatments compared:
4 rotations: Lathyrus, vetch, peas and fallow followed
by barley +/- Pa20 (50 kg/ha) on crops only

Treatments controlled:
Legume seed rate (180 kg/ha)
No innoculation
Trial size: 0.35 ha (7 plots x 500 m")
Replication: 5 replicates in one location (plus grazing
trial sites)

Grazing trials:
Treatments compared:
Grazing of vetch, lathyrus and commununal rangeland
+/- P200(50 kg/ha) on crops; DM yield only

Treatments controlled:
Legume seed rate: (180 kg/ha)
No innoculation
Trial size: 2 ha (4 plots x 0.5 ha) plus small fallow strip
Replication: 2 replicates in one location and six replicates
in another

Harvest trial:
Treatment compared:
2 species vetch and lathyrus
Treatments controlled:
Seed rate 150 kg/ha
46 kg/ha phosphate fertilizer
Uniform inoculation
Trial size: 1 ha (2 plots x 0.5 ha)
Replication: 4 replicates in each of 3 locations


























o el


I n

H I
U I
I
I








SIO
0 01
C I
E I
I
SI
I













t I








U
1




I
H 01
N I


X
MIN






0


























N 0
0 K


t r0

94 -1


.*


Cl 0 -j


Cl (**J I-




0 t-




Cl '0 -f \


C. C14 M'


en




0


z c
Z0 N
HOO


0
s U

HO
H
.5.

& -i


Cl











.4 P
OH ON 01-
NO H

N N


NH ( N
G C-4 o 4.a
d3 d" o r
(d W5, Of.
ui t-i &<

I -a
\o


Cl c( o


e
rc

C:
12
* 0
U 0
w2 t
I

.1:










TABLE 3: INPUT COSTS OF ON-FARM


GRAZING TRIALS

(2 HA X 8 FARMS)

AMOUNT APPROX COST"


SIMPLE ROTATION

(0.6 HA X 5 FARMS)I

AMOUNT APPROX COST


HARVEST AND COMPLEX
ROTATION TRIALS
(1.2 HA X 12 FARMS)

AMOUNT APPROX COST


BASIC INPUTS

CULTIVATION 1

SEED 2

FERTILIZER 1
(TSP)

LABOR FOR
PLANTING AND
FERTILIZING

ADDED RESEARCH
INPUTS

FENCING MATERIALS"

MICROPLOT CAGES"

VEHICLES
(@$O.10/KM)

INCENTIVES IN KIND

TOTAL MATERIALS
COSTS


6 HA

880 KG

760 KG


$492

$1662

$451


3 HA

420 KG

96 KG


$207


6.4 KM

160


$ 98

$215

$ 25


14.4 HA

2050 KG

1350 KG


$ 46


$8200

$4100


9100 KM $910

$1538


$17560


$205


1.3 KM $1670


1820 KM $182


7280 KM $728


$2236


$2886


1. Includes 0.35 ha of forage crops and 0.25 ha of barley (second year
of trial).
2. Calculated at 3.9 Syrian pounds = US $1.
3. Full costs. Fencing and microplots may be reused for approximately
five years.


TRIALS









4.0 INTRODUCTION TO AGROFORESTRY EXPERIMENTATION

4.1 What is Agroforestry?

"Agroforestry is a collective name for land use systems and
technologies where woody perennials (trees, shrubs, palms,
bamboo) are deliberately used on the same land management
unit as agricultural crops and/or animals, either in some
form of spatial arrangement or temporal sequence. In agro-
forestry systems, there are both ecological and economic
inter-actions between the different components" (Lundgren
and Raintree, 1983)

Thus, the key characteristics of any agroforestry system are:

-Generally, involves two or more species of pl-ants (or plants
and animals), of which at least one is woody.

Always has two or more outputs.

Usually has a cycle in excess of one year.

Is more complex ecologically (in structure and function)
and economically than a monocropping system.

"The promise of agroforestry lies in the fact that it has the
potential to address some of the key ecological and socio-economic
problems of land use e.g. soil fertility decline, erosion, food
and fuelwood scarcity. The aim is to maximise positive
interaction between the wood species, crops and/or animals so as
to achieve, at the very least, a more sustainable and diversified
production from the land (Lundgren, 1982). It should hence be
possible to achieve present requirements of food and wood while
conserving the soil resources required for future production
(King and Chandler, 1978). Under certain ecological and socio-
economic conditions, it is feasible that a higher total
production, than is possible with other forms of land use, can be
obtained.

It is important to stress that agroforestry is not a panacea for
all land use problems. In each case, it will also be necessary
to consider the opportunity cost of other land use system
(Hoekstra 1983). Thus, an agroforestry system/practice is
justifiable if it makes efficient use of available resources
(land, labour, capital) with respect to sustainable production."
(ICRAF, 1984)

One major difference between traditional agricultural research
and agroforestry is that agroforestry has a very limited number
of tried and tested technologies.










4.2 Planning:

There is a need for a multidisciplinary team because of the
nature of the subject: agronomist, forester, social scientist,
economist A land resource map/survey is advised when planning
the 'research program.

4.3 Characteristics of experimental treatments:

1. Tree arrangements are difficult to duplicate over sites.

2. Management of non-experimental variables is often impossible.

3. Trees may have a long life cycle; therefore, the hazard of
losses over time is greater due to the tree being tree cut
or altered.

4. Production is not sychronized and varies between seasons
(similar to livestock).

5. There are several outputs such as fuel, building materials,
mulch, etc., many of which are renewable. How does the
researcher measure output?

6. Relatively large experimental areas are needed, depending
on the nature of the species.

7. Sociological considerations are necessary, e.g. taboo or
religious connections with trees and others.

8. It may be difficult to get a yield estimate.

.9. Changing markets overtime makes long term economics
difficult to analyse.

10. Land use changes over time especially where there is a high
population density. Because of the long term nature of these
experiments, objectives may change over time.

4.4 Implementation:

1. Know at what stage to take what measurements and how often
to record.

2. Design tools for easy, fast measurement.

3. Time implementation operations with the farmer's schedule.

4. Collect economic data over time.

5. Plan to collect data over time.











4.5 Evaluation

1. Time frame may be longer with agroforestry; look for
consistency in performance over years.

2. Use non-parametric statistics. There is a lack of specified
probability distribution which weakens statistical inferences.

3. Useful criteria include: sustainability of yields, stability
of animal preference if used as feed source: time saving and
labor use; conservation of wood lands; crop yield trade-offs
with increased fuel or animal forage.

4. Error of rejecting something acceptable is high using
traditional statistical methods; therefore, these methods
should be used with others, (See Evaluation Methods in OFE,
Number 12).


5. Farmer's assessment is very valuable.


(See Table 1)











TABLE 1 COMPARISON OF CHARACTERISTICS OF CROPS. LIVESTOCK AND TREES
AND IMPLICATIONS FOR ON FARM TESTING IN AGROFORESTRY PRODU-
CTION SYSTEMS (Modified from Bernsten et al. 1983)


Situati ons


Factor
Component
arrangement



Life cycle


Production
phases


Outputs


Nonmarket
inputs/
outputs

Experimen-
tal unit

Local trad-
itions/
customs


Crops
Generally
standar-
dised


Generally
less than
4 months

All units
synchro-
nized

Only
grain/
tuber and
residue


Few


Small,
divisible


Some
social/


ritual uses


Li vestoc k
Mobil e/
stall fed



Generally
over 1 year


with respect to:


Trees/shrubs
Generally hap-
hazard



Nearly always
more than
1 year


Units seldom Units seldom
synchronized synchronized


Multiple
outputs:
meat, hides
mi 1 k ,man Lre,
power


Many


Multiple, out-
puts: firewood,
fodder, fruit,
timber, poles


Many


Large, non-
divisible


Various
taboos


Large,
divisible


Often complex
owner/use
rights


Im locations
Difficult to measure
and control non-
experimental
variables

Increases costs,
likelihood of losing
experimental units

Difficult to find
comparable units


Difficult to
measure/ value
treatment effect


Difficult to value
input/output


Increases cost,
risk to cooperator

Limits treatments


Management Relatively High
variability low


High


Few


Observation Many
units


Relatively Relatively
homoge- homogenous
nous (domestica-
(domesti- ted)
cated)


Crop
residue
feed


Manure for
crops


Very heteroge-
neous (largely
wild)


Fodder for
animals green
manure for
crops shelter
for animals
and crops.


Difficult to
isolate treatment
effect

Large statistical
variability

Large statistical
variability


Increased cost of
more complex
statistical design


Genetic
makeup


Inter-
action


--


--


-- -- -- -I I- -I-


--









APPENDIX A

GLOSSARY

Agroclimatic Classification the grouping of different physical areas
within a country, a region, or the world into broadly homogeneous
zones based on climatic and edaphic factors.

Agroclimatic distortions When transferring technology which success-
fully works in one situation to another situation, distortion of the
technology may occur due to differences in climate and/or soils.

Agroecological Zone a major area of land that is broadly homogenous
in climatic and edaphic factors, but not necessarily contiguous,
where a specific crop exhibits roughly the same biological expression.

Agroforestry a system of land use wherein annual crops are grown,
mostly in intensive mixed or intercropping methods, under the perennial
forest trees or fruit-cum-timber trees. Also known as agrisilviculture
and forest gardening.

Alley cropping the arrangement of several rows of annual crops
between a row of trees or a perennial crop (pigeon peas).

Arable land refers to land under temporary crops, meadows for mowing or
pasture, land under market and kitchen gardens (including cultivation
under glass), and land temporarily fallow or lying idle.

Benchmark-survey a systematic survey study aimed at collecting data
e.g. existing crops, varieties, yields, socioeconomic constraints,
before a project begins. Data collected depict the existing
picture of the survey areas with regard to selected parameters
and can be used to evaluate the results of the project.

Biological determinants of cropping systems the biological factors
such as crop species, varieties, weeds, insect pests and diseases,
which determine the crop configuration and performance of a
cropping pattern at a given site.

'Bottom-up' Information or research which starts with an understanding
of the existing situation on the ground before attempting to assess
what changes might be useful. 'Diagnosis as a basis for perscription'.

Commodity research the focusing of research on individual crops in
considerable detail.

Component technology the cultural techniques used in the management of
a crop or cropping pattern. Component technologies include variety,
planting method, tilage operations, fertilizer and water management,
pest management, harvesting, etc.










Constraints research Research that aims to identify and rank factors
such as disease, weeds, labor shortage etc. which are'limiting
products on.

Cropping index number of crops grown/year on a given area of
land x: 100.

Crop intensification the concept, approach, method, and process of
growing more crops per year by increasing cropping intensity.

Cropping intensity total cropped area divided by net area available for
cultivation multiplied by 100.

Cropping intensity index (CII) (Menagay E1975]) a time-weighted
land-use index that evaluates the fraction of the total hectare-
months available to the farmer that are used for crop production.

Cropping pattern the yearly sequence and spatial arrangement -of .crops
on a given land area.

Cropping pattern design the crop configuration or sequencing done on
paper for year-round land utilization at a given area considering
physical, biological, and socioeconomic factors prevailing at that
area.

Cropping pattern testing the growing of a designed cropping pattern
at a given site and evaluating biological stability, agronomic
productivity, and economic profitability.

Cropping system the crop production activity of a farm. It comprises
all cropping patterns grown on the farm and their interaction with
farm resources, other household enterprises and the physical,
biological, technological, and socioeconomic factors or environments.

Cropping systems research the research activities, mainly in farmers'
fields, that focus on the understanding of farmers' existing cropping
systems; design, testing, and development of new improved cropping
patterns and component technologies for selected environments to
efficiently utilize available farm resources.

'Cropping systems research site a contiguous area or several selected
areas representing one or more land types in production environments
that occur over an extensive area, where cropping systems scientists
conduct on-farm research trials with cooperating farmers.

Crop rotation The practice of following the crop located on a
particular site with a different crop the following season.

Cultural practices crop husbandry practices including land preparation,
seed selection, weed control, fertilizer and insecticide application,
water control in the field, etc.










Determinants of cropping patterns environmental factors that influence
the performance of cropping patterns and are not readily modifiable by
changes in cultural techniques of crop production.

Double cropping growing of two crops in sequence in a year on a piece
of land by seeding or transplanting one after the harvest of the other.

Dryland farming cropping systems of farmers in the arid and semiarid
zones.

Environmental complex a union of sites when cropping pattern or crop/
animal determinants are the same.

Environmental factors factors over which farmers have little direct
control, including the physical, biological, and socioeconomic aspects
of their setting.

Extrapolation area adaptation domain of a cropping pattern composed of
land types to which the cropping pattern is adapted.

Factor returns Economics recognizes four factors contributing to all
production processes: land, labor, capital and management. Each of
these factors contributes; therefore, there is a return to each of
these contributions embodied in an output.

Fallow when a crop is not grown on a field.

Farm enterprise an individual crop or animal production function within
a farming system which is the smallest unit for which resource use and
cost-return analysis is normally carried out.

Farmer environment the physical, biological, economic, and socio-
"cultural conditions under which the farmer operates his farming
systems.

Farmer feedback The assessment by the farmer of technology or
methodology being tested or demonstrated by extension or research.

Farming system a unique and reasonably stable arrangement of farming
enterprises that a household manages according to well-defined
practices in response to the physical, biological, and socioeconomic
environments and in accordance with the household's goals, preferences,
and resources. These factors combine to influence output and production
methods. More commonality will be found within the system than between
systems. The farming system is part of larger systems and can be
divided into subsystems.

Farming systems research and development (FSRD) an approach to
agricultural research and development that 1) views the whole farm
as a system, and 2) focuses on the interdependencies among the









components under the control of farm household members and how these
components interact with the physical, biological, and socioeconomic
factors not under the household's control. The approach involves
selecting target areas and farmers, identifying problems and
opportunities, designing and executing on-farm research, and
evaluating and implementing the results. In the process, opportunities
for improving public policies and support systems affecting the target
farmers are also considered.

Horizontal revolution in agriculture increased land use by expanding
cultivated land area through the utilization of' fallow and marginal
lands and reclaiming culturable waste lands, thereby increasing
land-use intensity.

Infrastructure the supportive features of and economy often provided
by government, but sometimes provided by private industry, such as
transportation, electricity, water, communications and governmental
organizations.

Intercropping growing two or more crops simultaneously in alternative
rows in the same field.

Interplanting all types of seeding or planting a crop into a growing
stand. It is used especially for annual crops under stands of
perennial crops.

Land equivalent ratio (LER) the land area needed under monoculture to
produce the same amount of crop yields as from 1 ha of intercropping
or mixed cropping. LER is computed:

Maize yield in intercropping Peanut yield in intercropping
(2.5 t/ha) + (1.2 t/ha) = 1.50
LER = Maize yield in monoculture Peanut yield in monoculture
(3.0 t/ha) (1.8 t/ha)

Land type a union of locations within which values of cropping pattern
determinants are the same.

Land use patterns alternative ways to utilize available land resources
over time for agricultural production.

Land utilization index (LUI) The number of days which crops occupy
the land during the year, divided by 365.

Maximum cropping the highest possible production per unit area per
unit time without considering cost of production or net return.

Mixed cropping growing two or more crops simultaneously in the same
field without rows.

Mixed farming systems farming systems with integrated crops, livestock,


52









and other possible household enterprises.


Mixed intercropping growing two or more crops simultaneously
intermingled in the same plot with no distinct row arrangement.

Mixed-row cropping growing two or more crops simultaneously in the
sample plot intermingled within a distinct row arrangement.

Monitoring study making systematic observations through well-designed
procedures on a crop, cropping pattern, farm or experimental trial to
relate resultant effects with observed factors or causes.

Monoculture growing only one crop on the land in a given crop season.

Multidisciplinary approach an approach in which several disciplines
become involved in a project or program with common general objectives.

Multilocation testing the testing of cropping systems technologies
generated at an on-farm research site at other locations within the
large target area to delineate the extrapolation zone as well as to
finally verify technology performance before wide-scale diffusion.

Multiple cropping the practice of growing more than one crop on the
same land in one year. It involves several alternative patterns of
crop arrangement in space and time such as mixed cropping, inter-
cropping, relay-cropping, sequential cropping, double cropping, triple
cropping, etc.

Multiple cropping index (MCI) the sum of areas planted to different
crops harvested during the year, divided by the total cultivated area.

Non-inhibiting non-treatment variables Experiments compare treatments,
but in the management of experiments there are many other practices
which are common to all treatments. These are non-treatment or
non-eLxpgrimental variables usually held constant during classical
experimentation, but rarely constant over a group of farms. Classical
experimentation usually fixes the constant level of these variables
high enough to prevent these management practices from inhibiting the
responses of the treatments.

On-farm research and development agronomic and socioeconomic studies
conducted on the farms with farmers' active participation. The goals
are to develop improved cropping system technologies and to devise
ways to combine these technologies with farmers' knowledge and skills
to efficiently utilize the available farm resources.

Optimized The best possible solution.

Perennial crops drops occupying land for more than 30 months not
including legumes and grasses in permanent pastures.










Physical determinants the important attributes of climate, water, and
land such as rainfall, topography, and hydrology that influence
configuration and performance of cropping patterns.

Pilot production program a small-scale (100 500 ha) production
program to determine the support needed in the large-scale diffusion
of recommended technologies as well as to clearly specify the tasks
and interrelationships of different institutions involved in supporting
the farmers. It also allows a final evaluation of the recommended
cropping technology, the cost of its extension to the farmers, and the
expected benefit.

Plot a contiguous area of land planted in a homogeneous manner during
a defined period, normally 1 year.

Plot plan a diagrammatic representation of the spatial and temporal
combination of crops on a plot during 1 year.

Preproduction evaluation higher level on-farm cropping system
activities consisting of multilocation pattern testing and pilot
production program to delineate the final production program area,
verify technology performance, determine institutional support
requirements, and to help in structuring large-scale production
programs.

Pre-screening Ex ante evaluation of the likely technical and economic
suitability of (treatments) possible solutions which are being
considered for OFE.

Ratoon cropping the cultivation of an additional crop from the
regrowth of stubbles of a main crop after its harvest, thereby
avoiding replanting such as in sugarcane, sorghum, and rice.

Reductionist research The method commonly used in classical
agricultural research whereby all variables except treatments are
held constant to allow precise measurement of reactions to treat-
ments. The method 'reduces' or 'abstracts' out of the real world
to be able to'handle the measurements with precision.

Relay cropping growing two or more crops in a sequence, planting the
succeeding crop after the flowering, but before the harvesting of the
preceding crop.

Research-managed trials experiments done in farmers' fields, but
managed by researchers to attain higher degree of experimental
precision while still getting the effect of some variables existing
on the farms.

Sequential cropping growing two or more crops in a sequence, planting
the succeeding crop after the harvesting of the previous one.








Shifting cultivation a method of cultivation in which several crop
years are followed by several fallow years with the land not under
management during the fallow. The shifting cultivation may involve
shifts around a permanent homestead or village site, or the entire
living area may shift location as the fields for cultivation are moved.

Site description description of an on-farm experiment site with
respect to physical and socio-economic environments and existing
farming systems.

Site selection selecting a contiguous area pr several areas
representing one or more production environments that occur over
an extensive target area, to conduct on-farm experiments to develop
improved farming systems technologies for the target area.

Slash-and-burn system a kind of shifting cultivation in high-rainfall
areas where the cropping period is followed by a fallow period during
which bush or tree growth occurs. For the next cycle of cropping,
the bush or tree growth is again cleared by cutting and burning.

Socioeconomic determinants factors such as marketing facilities, land
tenure system, and credit, which influence the cropping systems of
a given area.

Sole cropping growing one crop alone in pure stand, either as a single
crop or as a sequence of'single crops within the year.

Sondeo An informal survey approach to understanding local farming
systems.

Strip cropping growing two or more crops simultaneously in alternative
plots arranged in strips that can be independently cultivated.

Subsistence farmers farmers who produce mainly to meet family needs and
have little capacity to purchase production inputs or foods.

Superimposed trial a small set of experimental treatments superimposed
on farmers' production plot.

Systems approach studying a system as an entity made up of all its
components and their interrelationships, together with relationships
between the system and its environment. Such study may disturb the
real system itself (e.g. via farmer-managed trials or by comparison
pre-adoption studies of new technology), but more generally is done via
models (e.g. experiments, researcher or farmer-managed on-farm trials
oar both], unit farms, linear programming and other mathematical
simulations) which to varying degrees simulate the real system.

Target area large priority development area in a country for which
improved farming systems technologies are developed through on-farm
research conducted at sites in that area.










Target group/ recommendation domain Almost the same meaning: a group
of farmers for whom the same research effort will be relevant
because they are operating the same farming system.

Technical relationships The physical (eg. yield) output resulting front
a given level of one or more inputs (eg. fertilizer).

Testing of cropping systems the process of evaluating designed croppir
patterns and associated component technologies using selected criteria

'Top down' The imposition or prescription of changes thought to be goc
without any understanding of the local situation. 'Prescription without
diagnosis'.

*Turnaround time period between the harvesting of the preceding crop
and planting of the succeeding crop in a specific field.

Upland crops crops grown under aerobic soil conditions such as
wheat, maize, peanut, beans, etc.

Vertical revolution in agriculture Maximizing production per unit
land are per unit of time using intensive farming practices, high
production inputs, and improved management practices.

Whole-farm analysis A methodology designed to search for optimal
solutions through incorporation of farmers' objectives, farming
systems, and resources to arrive at improved cropping and livestock
patterns and management practices for overall farming systems
performance.

Whole-farm approach.- An essential characteristic of farming systems
research and development in which teams look at a whole farm to
identify problems, opportunities, and interrelationships, to design
and conduct experiments, and to evaluate results.





57



APPENDIX B:

MAJOR REFERENCES USED

Avila, Marcelino, 1984. Monitoring and Evaluating the Effectiveness of
On-Farm Research Trials. FSR Team Leader, RSS, Harare, Zimbabwe.

Cameroon National Root Crops Improvement Program, 1984. Annual Tech.
Report (IITA) IRA B.P 13, Njombe, Cameroon.

Christensen, P. 1981. Observations on the major classifications of
field trials used in FSR. SAFGRAD. Farming Systems Unit, B.P 1783.
Ouagadougou, Upper Volta.

Chambers, R. and B.P. Ghildyal. 1984. Agric. Res. for Resource Poor
Farmers: The Farmer First-And-Last Model. Paper presented at:
National Agricultural Res. Workshop on Nat. Agric. Res. Mgmt at
Nat. Academy of Agric. Res. Mgmt, Rajendrangus, Hyderabad
(July 10-13, 1984).

Davis, J.H.C.; M.C. Amezquita; J.E. Munoz. Border effects and optimum
plot sizes for climbing beans and maize in association and monoculture
CIAT.

Dear, K.B.G and R Mead, 1983. Statistics in Intercropping Technical
Report No.l. The Use of Bivariate Analysis Techniques for the
Presentation, Analysis and Interpretation of Data. Dept. of Applied
Statistics. Univ. of Reading.

Farming Systems Support Project. 1985. Agronomic Experimental Design and
Analysis. preliminary draft edition. University of Florida. Gainesvill
FL, USA

Fernandez, ICRAF, personal communication

Finney, D.J. An Introduction to Statistical Science in Aqriculture,
4th ed. Blackwell Scientific Publications, Oxford, U.K. 1972.

Fisher, Ronald A. The Design of Exgeriments. 5 edition. Hafner
Publishing Co., N.Y c.

Flinn, J.C. 1979. Agroeconomic Considerations in Cassava Intercropping
Research. In: Intercropping with Cassava. Proceedings of Workshop
held at Trivandrum India (1978) IDRC.

Francis, C. A. and J. H. Saunders. 1978. Economic Analysis of Bean and
Maize Systems: Monoculture versus Associated Cropping. Field
Crops Res. 1: 319-335.









Fresco, Louise. Approaches to the Study of Farming and Cropping
.Systems. Paper prepared for FSSP Trainers Workshop at ISU June, 1984.
(Res. fellow at Wageningen).

Hildebrand, P. 1981. Combining Disciplines in Rapid Appraisal:
The Sondceo Approach. Ag. Admin. 8: 423-432.

Hil debrand, P. 1984. Modified Stability Analysis of Farmer Managed
On-Farm Trials. Ag. J. 76: 241-274.

Hildebrand, P.E. and F. Poey. 1985. .On-Farm Aronomic Trials in
Farmin_ Systems Research/Extension. Lynne. Rienner Pub. Inc.
Boulder Co.

Hook, R. J. 1970s. Weed Control Experiments and Farmers Choice.
Dow Chemical Export A. G., P.O. Box 49470, Nairobi, Kenya.

Hogue, M.Z. 1984. CraopinSys tems in Asia. On-Farm Research and
Manaqemn.t, IRRI, P.O. Box 933, Manila, Philippines.

ICARDA/IDRC 1985. Regional Workshop on Research Methodology for
Livestock On-Farm Trials. 25-28 March 1985 at ICARDA Aleppo, Syria
(papers presented used as reference sources).

Icaza J. & Lagemann, J. 1982. Testing of Technology with Small Farmers
in Jinotega, Nicaragua. Trop. Ag. Res. &\Train. Center CATIE.

Kirkby, R.A 1981. The Study of Agronomic Practices and Maize Varieties
.Appropriate to the Circumstances of Small Farmers in Highland Ecuador,
Ph.D Thesis, Cornell University, Ithaca, N.Y.

Kirkby, R. A. and T. Cornick, 1981. Interaction of Crops and Livestock
Production in the Generation of Technology in Sloped Areas. Cornell
University, Ithaca NY USA

Lightfoot, Clive, 1981. Experiences in Designing and Evaluating
Technical Packages for Small Farmers, the Case of Botswana.
Pub. F.S Newsletter No.6 CIMMYT Nairobi.

Little, T. M. and F. J. Hills. Statistical Methods in Agricultural
Research. 2nd edition. U. of California, Davis CA 95616 USA. c. 1976.

Lowe, Carl. Lecture notes, Plant Breeding 212. Cornel University
Ithaca, N.Y.

Matlon, P.; R. Cantrell; D. King; M. Benoit-Cattin. Coming Full Circle:
Farmer's Participation in the Develoment of Technology, IDRC -189e,
Box 8500, Ottawa, Canada KIG 3149, @ 1984.

Milne, G. 1947. A Soil Reconnaissaince Journey Through Parts of
Tanganyika Territory. J. Ecol. 35: 192-265.










Morris, R. A. and F. M. Rumbaoa. 1979. Rainfall Recurrence Analysis
for Extrapolating Rice-based Cropping Patterns. paper presented
at WMO/IRRI Symposium/Planning Meeting on Agrometeorology
of the Rice Crop. IRRI. Los Banos, Phil.

Munro, J.M. 1968. Controls in a Series of Field Experiments. Cotton
Growing Review 45(1):42-53.

Norman, D.W. 1976. Developing Mixed Cropping Systems Relevant to the
Farmers' Environment. Symposium on Intercropping in Semi-Arid Areas.
Univ. of Dar-es-Salaam, Morogoro, Tanzania 1976.

Norman, D.W. 1973. Methodology and Problems of Farm Management
Investigations: Experiences from N. Nigeria. African Rural Employ-
ment Paper Nlo.9. Dept. of Ag. Econ., Michigan State Univ.,
East Lansing, Michigan.

Powell, J.M. 1985. Yields of Sorghum and Millet and Stover Consumption
by Livestock in the Sub-humid Zone of Nigeria. Trop. Ag. 62(1): 77-81.

Raintree, J.B. 1983. Strategies for Enhancing the Adoptability of
Agroforestry Innovations. In: Aqroforestr_ Systems 1:173-187.
Martinus Nijhoff/Dr. W. Junk Publ. The Hague.

Rhoades, R.E. 1982. The Art of the Informal Agric. Survey. Soc. Sci.
Dept. Training Document, 1981. CIP. Aptdo 5969, Lima, Peru.

Rhoades, R. E. 1982. Understanding Small Farmers: Sociocultural
Perspectives on Experimental Farm Trials. CIP. Aptdo 5969,
Lima, Peru

Saeed, M, Francis C.A, Rajewshi J.F. 1984. Maturity effects on
Genotype x environment interactions in grain sorghum. Agron. J.
76(1): 55-58.

Snedecor, G.W. and W.G. Cochran. Statistical Methods. 6th edition,
Iowa State University Press, Ames, Iowa, USA @ 1967.

Btewart, A.B. 1947. I. Planning and Conduct of Fertilizer Expts.
Memoranda on Colonial Fert. Expts. London. His Magesty's
Stationery Office. Colonial No. 214. UK.

Stoop, W.A.. 1982. Some Aspects and Examples of On-Farm Experiments.
Workshop "On-Farm Experimentation", IITA Ibadan, Nigeria, June 1982.

Zandstra, H.G., E.C. Price, J.A. Litsinger, R.A. Morris. A Methodology
for On-Farm Crogpingq Sstems Research. IRRI, Los Banos, Laguna,
Philippines. @ 1981.







University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs