THEsvs

IVESR SITYLIBR RESARI

llllllllllllllllllllllllllllllllll lllllllllll

31293 01688 542

 

 

 

 

 

This is to certify that the

dissertation entitled

Energy Utilization Modeling of Animal Draft Power
(EUMDAP) For Kenyan Small-Holder Semi-Arid
Agriculture

presented by

George S N Mungai

has been accepted towards fulﬁllment
of the requirements for

Ph.D. degree in Agricultural Technology
and Systems Management

 

 

%W#W

Major professor

Date May 15, 1998

MSU is an Afﬁrmative Action/Equal Opportunity Institution 0-12771

 

 

LIBRARY

Michigan State
l Unlverslty

 

 

 

PLACE IN RETURN BOX
to remove this checkout from your record.
TO AVOID FINES return on or before date due.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1/98 chlRClDabOuopﬁS—p.“

ENERGY UTILIZATION MODELING OF
ANIMAL DRAFT POWER
(EUMDAP)

FOR KENYAN SMALL-HOLDER SEMI-ARID
AGRICULTURE

By

George S N Mungai

A DISSERTATION

Submitted to
Michigan State University
in partial ﬁilﬁllment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

in
Agricultural Technology and Systems management
Department of Agricultural Engineering

1998

ABSTRACT
ENERGY UTILIZATION MODELING OF
ANIMAL DRAFT POWER
(EUMDAP)
FOR KENYAN SMALL-HOLDER SEMI-ARID
AGRICULTURE
By

George S N Mungai

A computer simulation model developed by this study was used for modeling of
energy utilization of draft animal power by small-holder farmers in semi-arid areas of
Kenya. EUMDAP simulation modeling program can be used to monitor existing DAP-
SYSs or project a proposed system that is not operational yet. The user is provided with
output data that enables decisions to be made about the current level of utilization of
DAP-SYSs. The program can also provide data about the expectations of a proposed
DAP-SYS in terms of energy utilization, the amount of feed required to perform tillage
and other tasks at the farm level as well as the time required to complete tillage
operations in time for planting before the onset of rains in semi-arid lands.

Simulations performed with the model showed that the current rate of DAP-SYS
utilization in days per year can be increased from the existing average of about 60 days
per year to over 100 days per year. The low utilization rate of current DAP-SYSs can be
attributed to the scarcity of alternative implements that can be used after the tillage

operations season is over.

The quality of feed available for draﬁ oxen determines the energy that the animals can
generate for work without losing weight. Since draﬁ oxen in semi-arid areas lose about
20% of their body weight during the dry spell preceding the seed-bed preparation season,
it is imperative that adequate feed (quality and quantity) is provided for tillage work

periods if draft animals are to be maintained in optimum physical working condition.

Legacy passes on. . .

Were my father, deceased at
my tender age of four alive,

he ’d be delighted to know that
his legacy lingers on.

He plowed for family, relatives,
neighbors and friends,
when few were the ox-teams
that tilled the valleys and plains of Ndeiya.
His fame earned him a name in our dialect
that literally means:
father of plows;
for the young and old alike knew him and
called him so.

 

Mungai Ndethi has been gone many a decade,
but in me flows the same desire -
to ease the burden and drudgery
of tillage.

At times I’ve wondered why, and how come
I chose draﬁ animal power to pursue
in studies this far?

Like Elisha, it ’3 my father ’s mantle that fell on, and
the legacy was passed on!

Dedication

Its been a long journey this far we ’ve come;
We 've toiled modestly and honorably;
In tasks great and small, to do the needful
For which we’ve traveled abroad.

Alone I ’d have been worn, tired and discouraged;
But with Anne my constant companion,
We ’ve covered considerable ground

A mate like her its hard to ﬁnd;
One who cares and tends for the family;
She ’s diligently met immediate needs and more,
While her helper was afar.

She ’s worked tirelessly in and out of academia;
Achieving her goals too.

It ’s to her that I dedicate this work;
And to our four daughters:
Catherine, Caroline, Lilian and Pauline
T o whom we pass on the baton.

The horizon is theirs,
T 0 read and to learn where we left
And carry on the pursuit of knowledge
In ﬁelds diverse to heights unparalleled.

Finally and most of all
I give glory and honor to my Lord and Savior
Jesus Chist who brings seemingly
impossible undertakings within reach.
For without Him I could do NOTHING!

For its through Him that
”I can do all things through
Christ Jesus which strengthens me. ”
Philippians 4:13.

ACKNOWLEDGEMENTS

The author wishes to acknowledge the following persons and institutions for their
support during the study period and process of the research work:

Dr. Thomas B. Burkhardt, the major professor, for his personal, professional and
faithful encouragement and advice throughout my period of study.

Dr. Robert Wilkinson for his guidance and material support in the initial stages of the
development of the research idea as well as the encouragement and support during the
ﬁeld data collection as well as serving in my academic committee.

Dr. Richard Bemsten for the advice and time spent during the development of the
instrument for ﬁeld data collection and serving in my academic committee.

Dr. Frederik Derksen for encouragement during the process of formulating the
research idea, support in material acquisition and serving in my academic committee.

Mr. Sven Andersson of Department of Agricultural Engineering, Swedish University
of Agriculture, Uppsala, for providing the expertise required in the development and
building of the data acquisition system without which this exercise would not have
materialized.

The Dutch Aid Agency and the department of Agricultural Engineering, University of
Nairobi for the funding of both the academic program and the ﬁeld data collection

exercise.

vi

TABLE OF CONTENTS

List of Tables ...................................................................
List of Figures ...................................................................

Chapter I INTRODUCTION .............................................
1.1 Background ...................................................
1.2 Draft animal power technology ............................
1.3 Objectives of the study ......................................

Chapter II LITERATURE REVIEW .....................................
2.1 Draft animal power in agriculture .........................
2.2 Draft animal power in Kenya ...............................
2.3 Draft animal power utilization ..............................
2.4 Draft animal power energy output ...........................
2.5 Draft animal power energy utilization models ............
2.6 Data acquisition models .....................................
2.7 Computer simulation models ................................

Chapter III DATA COLLECTION EQUIPMENT ......................
3.1 Introduction ...................................................
3.2 Data acquisition systems hardware .........................
3.3 Data acquisition system software ..........................
3.4 Calibration of data sensors ..................................

Chapter IV FIELD EXPERIMENTAL METHODOLOGY ...........
4.1 Data collection sites .......................................

4.3 Energy expenditure data collection ........................
4.3.1 Data logger programming ............................
4.3.2 Experiments at University of Nairobi farm .........
4.3.3 Experiments in farmers’ ﬁelds .......................

Chapter V FIED EXPERIMENTAL RESULTS .......................

5.1 Socio-economic data analysis ............................
5.1.1 Land holdings ........................................
5.1.2 Farm implements and hand tools ..................
5.1.3 Draﬁ oxen ............................................
5.1.4 Animal feed supply .................................
5.1.5 Farm labor and energy supply ....................
5.1.6 Using two pairs of oxen ............................
5.1.7 Start of plowing .....................................

vii

xiii

HUI—OH

13
14
17
18
21
25
27

'30

3O
3O
36
38

45
47
51
55
56

58
58
58
61
63

65
66
66

5.1.8 Draft animal power utilization ......................
5.1.9 Draft oxen life-span ....................................
5.1.10 DAP renting ﬁ'om others and hiring out ..........
5.2 Energy utilization data analysis ................................
5.2.1 Preliminary data analysis procedure ...............
5.2.2 Experimental results .................................
5.2.2.1 University of Nairobi experiments .......
5.2.2.2 Farmers’ ﬁeld experiments ...............
5.2.3 Use of ﬁeld data for modeling .....................

Chapter VI COMPUTER SIMULATION MODEL
DEVELOPMENT .............................................
6.1 Systems approach and concepts ........................
6.2 Simulation modeling ......................................
6.3 Computer simulation modeling .........................
6.4 EUMDAP computer simulation model

development ................................................
6.4.1 Identiﬁcation of needs ................................
6.4.2 Annual utilization of draft animals ................
6.4.2.1 Physical conditions of draft animals
6.4.2.2 Annual utilization of draft oxen ..........
6.4.2.3 Quality and quantity of feed ............
6.4.2.4 Soil conditions .............................
6.4.2.5 Unreliable rainfall .........................
6.4.3 Formulation of conceptual model .................
6.4.4 Data collection and analysis ......................
6.4.5 Construction of mathematical model ...........
6.4.5.1 Energy needs for oxen ...................
6.4.5.2 Work energy from draft oxen ...........
6.4.5.3 Energy utilization efﬁciency .............
6.4.5.4 Crop yield and returns .....................
6.4.6 Model implementation ............................
6.4.6.1 Land and soil parameters ...............
6.4.6.2 Draft oxen parameters .....................
6.4.6.3 Feed for draft oxen .......................
6.4.6.4 Operational inputs .........................
6.4.6.5 Implement usage ..........................
6.4.6.6 Renting of DAP-SYS ....................
6.4.6.7 Costs of DAP system usage ..............
6.4.6.8 Projected life-span of implements
and oxen ....................................

6.4.8 Simulating DAP-SYS ...............................
6.4.8.1 Scenario one .................................
6.4.8.2 Scenario two .................................

67
69
70
71
73
74
79
80
84

87
87
92
94

95

96

99

100
101
102
103
103
105
108
116
117
120
128
130
131
132
133
133
134
134
134
135

135
135
137
138
141

6.4.8.3 Scenario three ................................. 142

6.4.8.4 Scenario four ................................... 144
6.4.8.5 Scenario ﬁve .................................. 145
6.4.8.6 Scenario six ................................. 146
6.4.8.7 Scenario seven ................................ 148
6.4.8.8 Scenario eight ................................ 149
6.4.9 Application of EUMDAP model ...................... 150
Chapter VII DISCUSSIONS, CONCLUSIONS AND FUTURE
RESEARCH ............................................... 153
7.1 Discussions ................................................. 153
7.1.1 DAS data ............................................. 153
7.1.2 Socio-economic data ............................... 154
7.1.3 Development of DAS .............................. 155
7.1.4 Implementation of DAP-SYS data acquisition
system ................................................. 156
7.1.5 Development and implementation of EUMDAP
simulation model .................................... 157
7.2 Conclusions ............................................... 158
7.3 Limitations of the study and future research ......... 159
Appendix A Empirical values for energy computation .. 161
Appendix B Caloriﬁc equivalent of RER ................. 162
Appendix C RAMLOG EI-9000 data logger ............ 163
Appendix D Kistler accelerometer ......................... 165
Appendix E Time Domain Reﬂectometry instrument .. 166
Appendix F EUMDAP computer program listing ....... 168
Appendix G simulation print-outs .......................... 187
Appendix H Raw data print-out .............................. 193
Appendix I Questionnaire on draft animal power ....... 200
Appendix J Glossary of equations derived for the
simulation model ............................. 211
Bibliography ..................................................... 214

ix

Table 1.1

Table 3.1

Table 4.1

Table 4.2

Table 5.1

Table 5.2

Table 5.3

Table 5.4

Table 5.5

Table 5.6

Table 5.7

Table 5.8

Table 5.9

LIST OF TABLES

Percentage distribution of Kenyan small-holdings

by size and province ............................................. 4

Calibration response equations .................................. 42
Kenya’s agro-climatic zones and rainfall regimes 45
Parameter settings for RAMLOG E19000 logger DAS ..... 52

The t-test results for independent samples
(equality of means) of land size for Machakos and
Mbeere districts .................................................... 6O

Descriptive statistics of the small-holder farms in
Machakos and Mbeere districts .................................. 60

Stem-and-leaf plot of the land size in hectares
(pooled for Machakos and Mbeere districts) .. 62

Stem-and-leaf plot of the tillage land size in hectares
(pooled for Machakos and Mbeere districts) .. 62

Pooled distribution of ox implements in Machakos
and Mbeere districts .............................................. 64

Ownership and usage of draft oxen in Machakos

and Mbeere districts .............................................. 64
Livestock feed supply and sufﬁciency 65
Timing of plowing .............................................. 67

The t-tests for independent samples of daily work
days, ox-life span and annual work days (equality
of means) for Machakos and Mbeere dsistricts .. 68

X

Table 5.10 Stem-and-leaf plot of the annual DAP utilization in
days (pooled for Machakos and Mbeere districts) .. 69

Table 5.11 DAP annual utilization ....................................... 70

Table 5.12 Hiring and renting of draft oxen in Machakos
and Mbeere districts .......................................... 71

Table 5.13 Summary of performance results for University
of Nairobi experiments ...................................... 76

Table 5.14 ANOVA results of the University of
Nairobi experiments ....................................... . 80

Table 5.15 Summary of performance results for experiments
on the farmers’ ﬁelds in Machakos and

Mbeere districts ............................................ 82
Table 5.16 AN OVA results of experiments on farmers’

ﬁelds in Machakos and Mbeere districts . 84
Table 6.1 Energy values of feeds .................................. 120

Table 6.2 Simulation #1 - Tillage with two oxen weighing
250 kg each ............................................. 140

Table 6.3 Simulation #2 - Tillage with two oxen weighing
350 kg each ............................................. 140

Table 6.4 Simulation #3 - Tillage with four oxen weighing
250 kg each ............................................. 143

Table 6.5 Simulation #4 - Tillage with four oxen weighing
350 kg each ............................................ 143

Table 6.6 Simulation #5 —Energy utilization per season with
four oxen weighing 250 kg each working in
tandem ............................................... 147

Table 6.7 Simulation #6 — Energy utilization per season with
four oxen weighing 350 kg each working in
tandem ................................................ 147

Table 6.8 Simulation #7 - Tillage with two oxen weighing

xi

350 kg each using different feeds .................. 149

Table 6.9 Simulation #8 - Tillage with two oxen weighing
350 kg each using different feeds ................. 150

xii

Figure 3.1

Figure 3.2

Figure 3.3

Figure 3.4

Figure 3.5

Figure 5.1

Figure 5.2

Figure 6.1

Figure 6.2

Figure 6.3

Figure 6.4

LIST OF FIGURES

Block diagram of Data Acquisition System (DAS) .......... 33
Schematic view of the animal-implement Data

Acquisition System instrumentation ............................ 34
Depth sensor calibration curve ................................. 39
Horizontal draft calibration curve ............................ 41
Vertical draft calibration curve .............................. 41

PROLEC-TRA graphical output for a Nairobi
university experiment showing heart rate and
three digital channels ......................................... 77

Graphic output for ﬁle “lOWRRl” showing the
horizontal draft force and depth sensor traces ........... 78

Draft Animal Power system identiﬁcation
diagram ....................................................... 89

Step by step simulation model development ............ 97

Energy Utilization Model for Draft Animal Power
(EUMDAP) conceptual model ............................ 106

Algorithm ﬂow chart of the Energy Utilization Model
for Draft Animal Power .................................. 109

xiii

CHAPTER I
INTRODUCTION

1.1 Background

Semi-arid lands of sub-Saharan Africa have been experiencing rapid population
growth, both from natural growth and from in—migration from neighboring high potential
areas. The majority of the farmers in these areas are subsistent small-holders with land
holdings of less than 8 ha (20 acres). In Kenya, arable agricultural production in the semi-
arid lands, commonly referred to as marginal lands, is hinged on the use of draft animal
power (DAP) as the primary and dominant mode of mechanization. The Victory
moldboard plow is used almost exclusively for tillage and weeding operations and
sometimes for hurried plow-planting when rains come before the farmers have prepared
the land for planting. The energy level of the predominantly Zebu draft oxen used in
semi-arid areas is usually perceived by many small-holder farmers as too low for tillage
work when a single pair of oxen is used. Hence most farmers who own more than one
pair of oxen tend to use two or three pairs in tandem, but at low operating efﬁciency in
hours per day per unit draft animal. However, a well managed and maintained pair of
Zebu oxen is capable of carrying out tillage work in suitable soil conditions. In general,
the power from the draft animals is not used to its annual potential because the utilization
level of draft oxen in work days per year is very low.

The amount of food required to meet the needs of developing countries is speedily
increasing, especially in Africa. According to F AO estimates cited by Mohan Raj (Ed.
1992), by the year 2000 sub-Saharan Aﬁ'ica will be producing only 75% of its required

food supply, if present trends of declining food production continue. With this increasing

1

2

food supply, if present trends of declining food production continue. With this increasing
need for food security, new and veritable indigenous crop production methods (especially
tillage systems in the famine prone marginal areas) are required to enhance sustainability
of land and other locally available resources in the face of highly variable and erratic
agro-climatic conditions. This need is made more critical by the limited and often
degraded land resources, hard and poor soil conditions, poor physical condition of draft
animals and the low economic level of the majority of the farming community. Many
small-holder farmers devise various cultural and traditional innovations to improve the
chances of obtaining a harvest particularly due to the unreliability of rainfall in both
quantity and distribution. For example, some farmers in agro-ecological zones III and IV
of Machakos and Mbeere (formerly Embu district) districts hurriedly plow-plant as soon
as the ﬁrst rain showers fall. This strategy enables them to maximize on the limited
moisture available and also take advantage of the softened ground conditions as the draft
animals are usually too weak to plow the hardened soils before the rains set in.

In general, over 80 percent of the rural population in most developing countries
depend on agriculture for their food supply and economic sustenance. The majority of
the farmers are small-holders who operate at a subsistence level. Iftikhar and Kinsey
(Eds. 1984) in their recommendations to International Labor Organization (ILO)
emphasized that development of agricultural production for most sub-Saharan Africa
should be directed to the small-holder farmers who aggregately produce the bulk of the
food and export commodities. In Kenya, the aggregate value of agricultural production
by small-holder farmers outstrips that of the large-scale capital intensive farmers

(Republic of Kenya, Economic Survey 1963, 1970). Small-holders, therefore, play a

3

signiﬁcant role in the economy of the country. Overall, the agricultural industry provides
livelihood to about 90 percent of the total population of about 25 million (1990 census),
employs about 85 percent of the total labor force either directly or indirectly in agro-
industries or agricultural support industries, and generates over one third of the nation’s

gross domestic product- GDP - (Pearson et a1. 1995).

There are about 1.5 million small-holder farmers (Table 1.1) in Kenya of which about
74 percent have less than 3.0 hectares of land while the vast majority (96%) have 8 ha or
less (Republic of Kenya, Statistical Abstracts 1983). The total land area of the country is
about 56.91 million hectares and only about 17 percent is climatically suitable for
cropped agriculture in the high and medium potential lands (Republic of Kenya,
Statistical Abstracts 1978). According to Muchiri and Minto (1977) and Muchiri (1981),
of the 6.82 million hectares cultivated in Kenya, about 84 percent were cultivated by
hand tools, 12 percent were oxen cultivated and only 3.5 percent were cultivated by

tractors.

The predominant technology used by the small-holder farmers for their agricultural
mechanization is either hand tools, draft animal power (DAP) or a combination of both.
Attempts to alleviate the mechanization constraints of small-holder farmers should be
designed to use locally available resources that are socio-economically viable,
employment generating and not labor-displacing. Although rural employment is a major
concern, the potential for increasing agricultural productivity through human power alone
is severely limited by the inadequacy of human beings to develop and sustain
ample power for extended periods of time. A human being is rated at about 0.07 kW

while minimum power requirement for efficient agriculture has been estimated at about

4

0.35 kW per hectare (Kline et a1. 1969). De Vries (1986) estimated that an average
household of 5 people can cultivate 2 to 4 hectares using hand tools. However, the
average size of farming communities’ households in the small-holder is estimated at 4
adults and the land size needed for subsistence food crop production is about 2 hectares
in the medium and high potential agricultural zones. The land size required to sustain a

household in the semi-arid lands is estimated at about 4 hectares.

Table 1.1 Percentage distribution of Kenyan small-holdings by size and province
(except the pastoral and large farms). 1974 - 1975.

 

Provinces (%)

 

 

Acreage Rift

(ha) Central Coast Eastern Nyanza Valley Western Total
below 0.5 6.74 20.37 9.55 15.72 22.99 21.53 13.91
0.5 - 1 10.50 18.46 17.35 28.18 12.11 17.67 17.92
1 - 2 36.96 22.52 33.39 22.03 17.59 17.27 26.99
2 - 3 16.47 11.42 14.82 15.23 14.78 14.68 15.11
3 - 4 11.86 7.68 8.57 6.79 10.11 8.60 8.89
4 - 5 5.92 8.15 9.63 8.13 5.72 4.36 7.22
5 - 8 7.63 7 5.87 5.43 4.09 6.87 10.22 6.50
8 and more 3.92 5.53 1.22 1.83 9.83 5.68 3.47

number of holdings 329,530 69,861 353,431 386,431 89,823 254,618 1,483,422

 

Source: Republic of Kenya Statistical Abstracts (1 978).

The limited productivity of semi-arid lands calls for both intensive and extensive
cultivation of land to offset the risk of climatic uncertainties and meet the food
requirements to feed the increasing populations. The prospects of adopting engine or

tractor power for small-holder agriculture has little or no potential for the majority of

5

subsistence farmers in Kenya. A World Bank study cited by Pingali et a1. (1987)
concluded that most African farming systems will go through stages of gradual
intensiﬁcation and almost invariably DAP will become viable before tractorization
becomes economically feasible. As in most developing countries, the use of engine
power mechanization is limited to large-scale farming which is a small proportion of the
total farm land. Increasing land and labor productivity for the small-holder farmers
requires adoption of suitable, sustainable, appropriate and affordable technologies to
augment the family labor available at the farm level. DAP is already in widespread use
in the semi-arid agricultural zones of Kenya. It is, therefore, imperative and appropriate
to make efforts to increase the efﬁciency of utilizing DAP for both intensive and
extensive cultivation of land. This can be done through better nutritional management of
draft animals and Optimization of the use of their energy output in order to increase
productivity and hence help alleviate the plight of small-holder farmers in the marginal

lands.
1.2 Draft animal power technology

Although DAP was introduced into Kenya about eighty years ago (Mutebwa, 1979;
Starkey, 1990), it has not made an impact in the overall agricultural development partly
due to lack of institutional support. In the 1960’s the mechanization strategy adopted by
most Aﬁican countries was that of leaping from hand tools into the advanced level of
mechanization in an attempt to mimic the contemporary state of development in the
Western world. For over twenty years since Kenya became independent, policy makers
considered DAP technology a retrogressive approach to agricultural development. The

government and other development agencies promoted engine or tractor use through

6
subsidy and tariff reduction on imported machinery and facilitated institutional support

structures for the same. This approach implicitly discouraged the establishment of
suitable institutions for DAP promotion. This lead to the offshoot of a “bimodal” rather
than a “unimodal” mechanization organization structure as described by Ruttan (1990) in
his models of agricultural development. The trend is now changing. Efforts are being
made to pursue DAP as a viable option for mechanizing small-holder agriculture. There
is need for support of both an institutional structure and technical and technological
innovations to effectively utilize the current resources and cultural endowments of the

small-scale semi-arid farming community.

Treating technical development of DAP as endogenous rather than an exogenous
independent factor, along with the technological innovations for crop production, will
make DAP play the role of a catalyst in inducing agricultural productivity as prescribed
by Ruttan and Hayami (1990). It is possible and desirable to do so, given the existing
wide adoption of DAP in semi-arid lands of Kenya. Small-holder farmers are rational
decision-makers who are keen in providing for their own welfare while maintaining the
production capacity of their land by using sustainable production methods. Evidence of
the latter is the soil and water conservation practices that are a common feature especially
in the Machakos district such as bench terraces designed to improve water inﬁltration and
retention.

There is now an increasing realization of DAP’s potential as an intermediate
technological innovation for the small-holder farmers. Kenya’s public institutions,
development agencies and non-government organizations (N GOs) are enhancing their

participation and promotion of DAP usage by the small-holder farmers. The

7

establishment of Kenya Network for Draught Animal Technology (KENDAT) in 1993 is
one example of this trend (Kaurnbutho et al. 1995). The main reasons for this new
development approach are the failure of engine or tractor power to meet agricultural
mechanization needs of small scale farmers and the need to alleviate chronic food
shortages especially in the medium potential regions of the country that always require

government intervention through food rations during drought years.

The soil and climatic conditions of the medium potential lands inhibit land
preparation with draft animals before the onset of rains due to the hard soil conditions
and the poor physical state of the draft animals. The rainfall pattern of the semi-arid
areas is bi-modal and ranges between 612 and 857 mm annually, split almost equally
between the long rains falling in March-April and the short rains falling in October-
November. Evapotranspiration levels are high and in much of the area potential
evapotranspiration is exceeded by rainfall only in the months of April and November
(Republic of Kenya, Statistical Abstracts 1987). During both rainy seasons, the
distribution of rainfall is subject to considerable ﬂuctuation that is unpredictable. The
highly variable long rains fail 30% of the time as compared to 10 to 20% rate of failure of

the short rains (Siberfein and Marilyn, 1989).

Although DAP technology is culturally integrated by the small-holder semi-arid
farming sector, it is not utilized to its optimum potential as observed from the low
acreage plowed by farmers and the existence of few off-fann activities employing draft
animals. Considering DAP as an enabling capital input, it can be used to maximize its
enabling potential productivity capacity in order to optimize the productivity of land and

labor. While labor at the household level is considered an inelastic economic resource,

 

8

land has the potential of both expansion and intensive utilization in the semi-arid

agricultural zones.

Lack of technical guidelines in DAP has led to inefﬁcient utilization of animal power
from both the energy-use standpoint and equipment design and development aspect. This
contention is evident from the fact that farmers design and adopt their own tillage
practices such as plow-planting to meet their mechanization needs. Over the years
government mechanization extension services have placed greater emphasis on large-
scale (capital intensive) mechanization at the expense of DAP mechanization. In order
for farmers and researchers to realize the maximum potential of DAP, there is need for
increased concerted efforts in research at both basic and applied levels. DAP has the
potential of enhancing agricultural productivity, as well as improving the welfare of the
rural communities, if adequately managed and utilized during both the “working” season
and off season periods. It has distinct beneﬁts over the use of hand tools and engine

power for the small-holder farmers, which include:

reduction of human and animal fatigue (drudgery)

increasing agricultural productivity

expansion of area cultivated

reduction or elimination of labor bottlenecks

minimal use of imported inputs (foreign exchange conservation)
renewable energy source at low running cost

improvement of timeliness or increase of the intensity of operations
multi-purpose use of animals

manure for garden fertilization

DAP use in Kenya for over eight decades has remained relatively static. Farmers
have tended to use the same tools, basically a Victory moldboard plow for both primary

tillage (often plow-planting especially when late in land preparation) and inter-row

9

weeding as observed by the researcher during his tenure in mechanization extension
services. There is little evidence of the use of other DAP farm tools, except ox carts for
transportation. Lack of technological innovations has kept the farming community in the
predominantly DAP farming zones in a “static” economic equilibrium limiting them from
making upward advancement of their welfare. Utilization of this power source has also
remained quite low, as farmers do not use their animals in-between the cropping seasons.
One way of getting the most out of DAP in small-holder agricultural production is to
increase its utilization in terms of productive work-days per year, acreage plowed, and
diversiﬁcation of its application at the farm and regional level.

Personal contact with researchers and agricultural extension workers revealed that
there is a perception that if modest technical changes that would not demand high extra
investment, but instead improve the utilization and management practices of DAP are
introduced, farmers would be willing to make changes that have the potential of high
returns. In order to make the best use of DAP, the potential power output from animals
and the draft force demand by the implements must be known or be estimated accurately.
Matching implements to the available animal power would enhance the efﬁciency of
energy use and create less fatigue to the prime mover (animals) and operator. However,
unlike engine power which can be easily estimated from the speciﬁc caloriﬁc value of
petroleum fuels and the predetermined power output of the engines, DAP energy source
does not render itself to convenient methods of prediction. However, the physiological
demands of an animal at work and the mechanical draft requirements of the implements
are two primary variables that can be used for the determination of the level and rate of

energy utilization in DAP systems.

10

The data base on DAP utilization is scant because research in this area has been
limited. Availability of a DAP data base would beneﬁt researchers on DAP farming
systems, rural development project implementers and agricultural mechanization
extension workers who would use the data on DAP for prediction of energy utilization,
power requirement or modeling purposes.

Tillage is considered the most energy demanding cultivation exercise at the farm level.
Almost one-third of the energy needed in the total farm operations is taken up by tillage
(Kitani, 1987). Draft animals derive their work energy from the feed they eat.
Regrettably, they have to be fed throughout the year, despite their being used for only a
fraction of the time. It is therefore important to make the best use of the animals during
the working season in order to efﬁciently use their energy output. Understanding the
animal-implement interaction and the effect of the variables that inﬂuence the
performance of the DAP system as a whole would enhance the utilization of this energy

source through computer simulation modeling.

A computer simulation model, Energy Utilization Model for Draft Animal Power
(EUMDAP), was developed by this research exercise to be used for the improvement of
DAP utilization. The program is capable of simulating various scenarios given varying
ﬁeld conditions, size of animals, implement size and the feed quality and quantity for the
draft animals. Development of computer simulation models requires the use of historical
data or acquisition of primary ﬁeld data for parameter estimation. Limited research has
been conducted in the area of animal-implement systems due to the difﬁculty of
simultaneously monitoring both the mechanical output from animals and the

physiological parameters that affect the DAP system performance. In this study,

11

therefore, ﬁeld data collection was performed to obtain authentic parameter estimators in
Machakos and Mbeere districts in Eastern province of Kenya where DAP is widely
adopted. Simultaneous collection of mechanical and physiological data of the DAP
systems was done using a sophisticated electronic data acquisition system (DAS). A total
of seven physiological and physical parameters were monitored. The physiological data
of the oxen and the equipment performance data were collected for varying soil types and
conditions, varying operational speeds and various teams of animals, as well as varying
depths and widths of tillage. Historical and operational data at the household level were
also gathered through a questionnaire administered by the researcher with a random

sample of small-holder farmers in the two districts.
1.3 Objectives of the study

The primary objective of this research exercise was to develop a computer simulation
model for energy utilization of DAP systems that can be used by researchers and
extension workers in the promotion and improvement of DAP technology in order to
enhance utilization of DAP systems. The main focus of the research project was in the
semi-arid areas of Machakos and Mbeere districts where DAP is predominantly used.
Speciﬁc objectives of the study included the following:

1 Develop and build an electronic data acquisition system for monitoring physiological
and mechanical parameters of an animal-implement system simultaneously.

2 Perform DAP-SYS data collection in the farmers’ ﬁelds with electronic DAS to be
used for validation of the computer simulation model.

3 Obtain socio-economic ﬁeld data from a random sample of farmers in semi-arid lands
for estimating DAP-SYS operational parameters for the computer simulation model.

4 Develop a computer simulation model that would provide estimates of various output
variables for optimum utilization of DAP systems in varying ﬁeld conditions.

12

5 Perform simulations for various DAP system scenarios for different levels of
applications to include:

0 sole use of DAP-SYS by owner-operator
DAP-SYS owner-operator/hiring out

In order to achieve the above objectives, a DAS for collecting and monitoring energy
input and output of a DAP system was designed and built (see description in chapter III).
It was capable of collecting physiological and physical data of the animal-implement
system at work. The data collected with the DAS from the DAP system at work were
monitored, logged and saved with a commercially available data logger (RAMLOG
EI9000') and a portable personal computer.

The physiological variables monitored for the animal were: heart rate, respiration rate,
stepping rate and body temperature. Four mechanical variables: depth of tillage, ground
speed and horizontal and vertical draft force were monitored simultaneously with the

physiological data. The results of the ﬁeld tests are discussed in chapter V.

 

' Trade names are used in this dissertation solely to provide speciﬁc information. Mention of a product
name does not constitute an endorsement of the product by the author to the exclusion of other products not
mentioned.

CHAPTER II
LITERATURE REVIEW
2.1 Draft animal power in agriculture

The use of animals to augment human beings’ limited power supply dates back to
about 3200 BC. when the Summerians ﬁrst used animals for plowing, threshing and
transport (Zeuner, 1963). Though animal power preceded tractor power in the historical
development of agricultural mechanization, literature is scarce on the prediction and
evaluation of this age-old power source. DAP technology has been considered as an
intermediate technology that would phase out for improved, efﬁcient and more complex
technologies. Though this has been the case for the Western civilizations where
constraints imposed by inelastic supply of labor have been offset by technical and
technological advances in agricultural production, DAP still plays a signiﬁcant role in the
agricultural production and transportation sectors of most developing countries.

Draft animals commonly used for tillage, processing, transportation and water
pumping are horses, mules, oxen, buffaloes, camels and donkeys. Oxen, and donkeys are
the most pOpular work animals in sub-Saharan Africa. Donkeys are used mainly as pack
animals and occasionally to pull carts while oxen are used primarily for tillage purposes.
Sims and Ramirez (1989) postulate that the use of animals as a power source is extremely
important. They quote Smith (1981) and Lawrence and Pearson (1985), who estimated
that there are about 200 million draft oxen and buffaloes in the world that produce 85%
of the power used in developing countries’ agriculture. Kemp (1987) documents that
global animal power represents about 35 GW of power source in the countries where

DAP is used.

14

An estimated 15 million draft animals are in regular use in sub-Saharan Africa,
cultivating between 5 and 10% of the land area planted with annual crops (F A0, 1983).
The trend to use DAP in Sub-Saharan Africa is increasing. Starkey (1986) emphasizes
that since there is currently a signiﬁcant interest in animal traction, it is important that the
resources being allocated to DAP systems are effectively utilized. Singh et al. (1989)
concurs with other proponents of DAP by stating that this vast power resource should be
utilized more efﬁciently. An increase in the number of hours DAP systems are used per
year through hiring out and off season usage would be a way of increasing the systems’
utilization level.

Attempts to promote the use of engine power through government tractor hire services
for small-holder agriculture in sub-Saharan Africa have been a technical and economic
disaster (Pingali, et al. 1987). The failure and subsequent disbanding of the government
of Kenya’s National Tractor Hire Service that was started in the early 19608 bears
testimony to this contention. The use of DAP in most developing countries will continue
to play a signiﬁcant role in the foreseeable future for the food production and
transportation in developing countries. Research ﬁndings and development efforts on
DAP technology will contribute in a great measure towards improving the utilization and
effectiveness of this mode of mechanization.

2.2 Draft animal power in Kenya

The role of DAP in agricultural mechanization has a varied history in Aﬁica and other
developing countries. In Kenya animal power was introduced over 80 years ago by white
settlers (Mutebwa, 1979; Starkey, 1990). The ox plow was ﬁrst used in Machakos

district by settlers in 1903 and adopted by local farmers in 1910 (Downing et al. 1988).

15

This innovation was not widely accepted primarily because there was little institutional
support. For over twenty years since Kenya became independent in 1963, policy makers
did not consider DAP as being important in agricultural development. This meant that
the advancement of DAP as a mechanization option was relegated to the enterprising
small-holder farmers and non-govemment development agencies. Despite this apparent
indifference by the public institutions on DAP promotion, Starkey (1988) has
documented that about 700,000 working cattle are employed in agricultural production in

Kenya predominantly in Eastern, Nyanza and Western provinces.

In the recent past, there has been a change in the right direction for DAP proponents.
Kenya’s development plan of 1974-78 enunciated a policy of giving priority to increasing
the productivity of Kenya’s small-scale farmers which included emphasis on animal
powered equipment and other ‘appropriate technologies’. A workshop on “Farm
Equipment Innovations for Agricultural Development and Rural Industrialization” held at
Kabete, Kenya in 1975 (Westley and Johnson, Eds. 1975) accelerated this positive
change of attitude on DAP. Policy makers have realized the critical role that DAP plays
in tillage and transportation for small-holder agricultural producers. The National Food
Policy Paper of 1981 (Republic of Kenya, 1981) speciﬁcally cited promotion of animal

power by stating that:

...the main aim of the policy will be the development and wider usage of more
appropriate technology to increase labor productivity and to reduce the present
emphasis on imported capital-intensive equipment. The availability of agricultural
machinery, particularly that required for land preparation and seeding, will be
increased through programs to supply mechanized ox-drawn and hand equipment.

16

The majority of draft oxen used in Kenya are Zebu (Bos indicus) whose weight
seldom exceeds 350 kg and often drops to 250 kg in the dry season when natural pasture
is deﬁcient of forage. It is believed that the Zebu cattle were introduced into Africa from
India in about 1500 B. C. They are also found in East Asia, Gulf Coast states of the
USA, and Latin America where they are called Brahman cattle. Zebus are better adapted
to the tropics because they have been selected under these conditions for hundreds of
years by both artiﬁcial and natural means. They have adaptive traits like resistance to, or
tolerance of pests and diseases; tolerance of intense sunshine, heat and humidity; and
ability to utilize high-ﬁber content forages (Lasley, 1981; Mukasa-Mugerwa, 1989).

They also mature late but produce little milk.

The Zebu cattle are characterized by a pronounced hump over the shoulder and neck,
horns which usually curve up, and drooping ears. In Kenya, as in most sub-Saharan
African countries, oxen (mature castrated bulls) and steers (young castrated males) are
used for draft work. Female cattle are primarily used as breeding stock and for domestic
milk production and not for draft purposes. Draft cattle are usually trained from the age
of two to four years and work for a period of between ﬁve and ten years before they are

replaced.

Although in some developing countries single animals are used either with head yokes
or breast harnesses, this research project concentrated on a pair of oxen harnessed
together with a neck yoke. The harnessing system most commonly used in Kenya as a
force transmission system linking the animals to their working implements consists of a
yoke and a chain. The chain’s length is used to regulate the vertical position of the line

of pull, increasing or reducing the angle of pull. The ideal line of pull being collinear

17

with the animals’ center of . mass, the implement’s center of gravity and the soil’s force
(Inns, 1990). Two major types of yokes are used in Africa, those that are tied to the
horns of the animal and those that take power from the withers (Starkey, 1989). The
latter version of yokes are often referred to as neck or shoulder yokes. The withers refers
to the part of the back that is over the shoulders, directly above the thoracic vertebra. In
Zebu cattle the withers are immediately in front of the hump. The hump, therefore, helps
to support the yoke which is placed right ahead of it. Kenyan small-holder farmers use

neck yokes exclusively.

Neck yokes are wooden poles used to hitch two animals together side-by-side. In their
simplest form, they are made of a wooden pole with pegs inserted in holes on either side
of the animal’s neck that help to restrain lateral movement. Improvement of neck yokes
has been concerned with increasing the contact area of the yoke on the animal and also

padding the yoke to minimize bodily abrasion and subsequent sores (Oudman, 1994).
2.3 DAP utilization

The overall utilization of draft animals can be measured in terms of the number of hours
the animals work per year. Small-holder farmers in Kenya tend to use their animals mainly
for tillage and to some limited extent for transportation. Land preparation is carried our
during a three-week period in each of the two cropping seasons. Depending on the acreage
under cultivation, draft animals can be used for a maximum of about 180 hours per year
when working for 6 hours per day for 6 ﬁve-day weeks on tillage work for the two
cropping seasons. Srivastava (1987) and Thakur (1989) both document that annual

utilization of oxen in India ranges from 300 to 1500 hours when used for multi-task

l8
purposes. The economics of using DAP in Kenya are currently limited to the immediate
needs of land preparation in order to plant in time.

The draft oxen are rarely used beyond the land preparation period when they are in their
prime physical condition. After the tillage season is over and forage supply is plentiful the
oxen regain their physical condition. Although most farmers have only one type of ox
implement (Victory plow), there are scope and potential for increasing the utilization of
DAP for small-holder farmers by engaging the animals in other on and off-farm tasks.
Muchiri (1984) cited Heyer et al. (1976) who noted that:

As long as the ox equipment in use consists of the mouldboard plough, ox carts
and little else, as at present, the useﬁrlness of oxen is very limited. Only in a few
areas are oxen used at all extensively and while the scope for using oxen is
obviously restricted to areas where the topography is suitable and the availability
of land sufﬁcient, there is certainly scope for more intensive use of oxen in some
areas and there are additional areas that could beneﬁt from the use of oxen at
present.
There are a number of other DAP implements that have been tried and utilized in other sub-
Saharan African countries that could be useful in Kenya’s semi-arid lands. An assortment
of weeding implements, ridgers and mutli-purpose tool-bars have been developed and used
in Tanzania, Zambia and Zimbabwe (Stevens, 1994; Kwiligwa et a1. (1994); Kayumbo,
1994; and Hagmann 1994). The main drawback in the acquisition of single purpose
implements is the high relative price of an implement that is used for a speciﬁc task only.
2.4 DAP energy output

Physical work by hmnans and animals is a complex interactive process involving both

the active and passive elements in a power transmission system. The normal regime for

energy supply to the animal’s muscles is up to the sub-maximal level of metabolism

(aerobic). In draft cattle the likelihood of reaching anaerobic threshold is only when

19

activities involving sudden , intense efforts are done, say pulling a cart out of a ditch
(Pearson and Archibald, 1989) or when the animals are constrained to work beyond their
aerobic limits. Continued work in anaerobic range results in the accumulation of lactic acid
in the muscles and body ﬂuids leading to fatigue which inhibits ﬁrrther glycogen
breakdown (Wilmore et al. 1994).

A draft animal can be considered a ‘machine’ that converts chemical energy to
mechanical energy in the muscle tissue which does the work. An analogous comparison of
draft animals with mechanical chemical energy converters (engines), shows that the basal
energy supply of an animal are utilized for maintenance and growth when not doing any
work. A fossil fuel engine on the other hand needs “basal” energy (idling) for basic
functions such as cooling, battery charging, and overcoming friction.

The energy derived from feed by animals is referred to as metabolizable energy (ME).
MB is obtained from the digestible energy (DB) of the feed that the animals eat. Gross
energy (GE) of the feed minus the energy lost in the feces constitutes DE. ME and DE are
strongly correlated. DE as a proportion of GE may vary from 0.3 for very mature,
weathered forage to nearly 0.9 for processed, high quality cereal grains. For most forages
and cereal grains, the ratio of ME to DB is about 0.8 but can vary considerably depending
on intake and feed resource (ARC, 1980; NRC 1996). The priority in the energy utilization
by an animal is to meet the net maintenance energy (NEno needs ﬁrst. The remaining
energy is then available as production or growth energy (NEp), for storage of body mass or
for work as in the case of working animals. Under-fed draft animals utilize their storage

energy for work leading to loss of weight.

20

Theoretically it is possible to calculate the amount of feed necessary for an animal in
order to get a certain level of energy output (Crossley and Kilgour, 1983). The gross
energy (GE) value of typical feed stuffs can either be measured in the laboratory using
bomb calorimeter or it can be determined from animal nutrition literature such as National
Agricultural Research (NRC) or Agricultural Research Council (ARC). Mature animals
require energy for bodily maintenance and production (milk, meat) or work whereas
growing animals require extra energy for growth as well.

Growing cattle require about 1.3 to 1.7 kg of total digestible nutrients (TDN)2 above
maintenance in order to gain 0.45 kg per day (cited by Sen, 1966). According to F A0
(1972), the maintenance requirement of working oxen in West Africa was estimated at 2.6
TDN. Lawrence (1985) and Singh (1985) in separate experiments concluded that the
energy used by oxen during a normal working day is equivalent to between 1.67 and 2.5
times the energy they require for the maintenance need when working 5.5 hours per day.
The major constraint to providing this energy where only poor quality feed (containing
about 9MJ of ME/kg of dry matter [DMD is available, is the voluntary dry matter intake
(V DMI-or appetite limit) of the animal.

In determining the performance of draft animals, the question arises as to which
measurable parameters should be used and varied in such a manner as to optimize
performance of the DAP system. The energy available from the animals is a direct output
from the feed the animals consume. The duration of working for the draft animals depends
on the quality and quantity of the feed they eat and also on the amount of energy they

expend per unit time (power developed). The work that the animals are able to perform is

 

2 1 kg TDN = 4.4 Mcal DE (NRC, 1966); 1 Meal = 4.184 M]

21

based on the physiological processes associated with movement and muscular activity
(Pearson, 1985; Macmillan, 1985). While prediction of the energy requirements for DAP
is not directly deterministic, the rate at which animals use oxygen in sub-maximal (aerobic)
metabolism has a linear relationship with the power which animals can develop.

Mueller et al. (1994) working with donkeys showed that oxygen consumption varies
linearly with mass-speciﬁc work rate (Wt/kg). Gottlieb-Vedi et al. (1991) demonstrated
that the volumetric rate of oxygen consumption (V 02 ) and draft of standard bred trotters
had a direct relationship. Kuhlmann et al. (1985) also arrived at similar conclusions when
working with Hereford calves on a treadmill. Although oxygen uptake is the most direct
measurement of energy expenditure in animals (Brody, 1964), it is difﬁcult to monitor
oxygen utilization in ﬁeld conditions. The constraints inherent in the determination of
animal performance are the ﬁnite energy reserves that decline with time and the limitation
of the amount of oxygen intake for oxidation of the energy source (Macmillan, 1985).

Various researchers have shown that the heart rate (HR) can be used to predict the
energy expenditure of animals since V02 is directly related to cardiovascular response of
animals. The aim of any physiological study of draft animals should be to knit together the
observations made in the laboratory with ﬁeld experiences, so that a complete picture can
be constructed (Pearson, 1985). Pearson (1985) further emphasizes that measurements
such as heart rate, respiration rate and body temperature are relatively easily monitored in
animals in the ﬁeld and can be useful in comparative studies of animals at work.

2.5 DAP energy utilization models
Experiments conducted by the Center for Tropical Veterinary Medicine (CTVM) in

Costa Rica showed that there is very little need for extra protein during work for draft

22

animals. However, there was need of more energy feed to avoid loss of weight (Lawrence,
1985). If the net energy required for work is known and the heat increment associated
with work is also known, combining the two would enable computation of the extra
metabolizable energy required for work and hence the feed requirements. Lawrence (1985)
used the following factorial model to compute the extra energy used for work including

energy for walking, carrying loads, pulling loads and walking uphill:

E = AFM + BFL+—I:::+ 9'81HM

 

m

E = extra energy used for work, kl

A = energy used to move 1 kg of body weight 1 m horizontally, J
F = horizontal distance traveled, km

= live-weight, kg

= energy used to move 1 kg of applied load 1 m horizontally, J
= load carried, kg

work done whilst pulling loads, kJ

efﬁciency of doing mechanical work (work done/energy used)
efﬁciency of raising body weight (work done raising body)
vertical distance traveled, m

mungrwg
||

Empirical and experimental data are available for the parameters used in the above
model for various sizes and types of draft animals (Appendix A). The model developed by
Lawrence is useful for estimating energy requirements for performing various ﬁeld tasks.
It is especially useful when the energy demand from draft equipment and the maintenance
energy of draft animals can be accurately estimated. Simulation of the energy needs to
perform farm and off-fann tasks with animal-implement systems without subjecting the

draft animals to body weight loss can be done with this model.

23

Crossley and Kilgour (1983) developed energy-feed relationships that can be used to
compute the quantity and quality of feed required for a pair of animals to perform a

particular task. The relationships used for the calculations are:

DE = 0.45 to 0.85 * GE
ME = 0.8 "‘ DE
MR = 8.3 + (0.091W,) MJ/day
AL = 0.025 * W, kg/day
1 kWh = 3.6 Ml
where:
DE = digestible energy, MJ
GE = gross energy, MJ
ME = metabolizable energy, MJ
W, = live body weight of oxen, kg
MR = maintenance ratio
AL = appetite limit
kWh = kilowatt-hour

These relationships can be used to estimate the amount of feed required per season and
hence the area of land required to maintain the animal per year, provided the yield per unit
area is known. This model together with Lawrence’s model are complementary in the
estimation of draft energy requirements and the amount of feed required to provide this
energy to the animals.

Models based on physiological parameters have been developed and used by various
researchers. Estimates of energy use derived from the heart rates of working animals and
their body weight have been accurately predicted for draft oxen. Richards and Lawrence

(1984) derived a predictive equation for Brahman oxen (Bos indicus) that uses the relative

 

24
heart rate (RHR) and the metabolic weightJ (W,)°‘7’ of the animal to determine energy

requirements of working draft cattle under ﬁeld conditions. The RHR was deﬁned as:

_ WorkingHeartRate

RHR — _
Re stmgHeartRate

 

(2)

The weight of the animals used in their experiments ranged between 510 and 630 kg.
Other researchers have developed similar models that conﬁrm the validity of Richards and
Lawrence’s model in varying ambient conditions and size of cattle. Even though the
animals which they used are larger than those found in Kenya; Sneddon (1986), Mathers et
al. (1985), and Sneddon et al. (1984) have all documented that the model works well and is
independent of ambient temperature and animal size. The regression coefﬁcient of the
model developed by Sneddon et al. (1984), equation (5), compared well with that of

Richards and Lawrence (1984), equation (3) as shown below:

EEW = 24.94RHR - 16.25 (r = 0.91; n = 49; P=0. 001) (3)

HR = 0.0318EE + 46.2 (r = 0.94; rt = 56) (4)

EE, = 25.4mm -17.7 (r = 0.95; n = 54) (5)
where:

EBw = actual energy expenditure per unit metabolic weight, W/(W,)°'7s

HR = heart rate

Extrapolation of the value of EEw when RHR = 1.0 (the energy demand of the oxen at

resting heart rate) in the ﬁrst model was found to be 6.94 W/(W, )°'7’. This ﬁgure derived

 

3 Metabolic weight (W0 75) or Empty Body Weight (EB WO- 75) is the body weight of an animal in
kilograms raised to the three-fourths power ﬂVRC, l 981; NRC, l 996).

25

experimentally by Richards and Lawrence (1984) compares well with 6.62 W/(W,)°'7s
obtained for Japanese cows by Yamamoto et al. (1979).

O’Neill and Kemp (1989) working at the Central Institute of Agricultural Engineering
(CIAE), in Bhopal, India showed that the HR and work rate of Brahman cattle and
buffaloes have a direct relationship. Their model shown below predicts the heart rate for

Brahman cattle from the work rate:

HR = 0.06P + 72 (r = 0.95) (6)
where:
P = power developed by animals, W.
HR = heart rate

Since these models have been tested under both laboratory and ﬁeld conditions, they can
be used with conﬁdence to model the energy requirements of Zebu cattle (Bos indicus).
Subsequently the amount of feed required to provide the net energy (N [3,) that will give the
power needed from the animals can be estimated from the MB of the feed. This would
require that the maintenance energy (N13,) or the basal metabolism rate (BMR), voluntary
activity and heat increment due to feeding and digestion, be accurately estimated too. The
various energy utilization models developed can be used to predict the energy use of Zebu
oxen depending on the ﬁeld conditions and the allometric parameters of the draft animals.
2.6 Data acquisition systems

Electronic data acquisition systems have been developed and used to collect data, or
monitor variables of biological and physical systems in production and research

environments. One of the major advantages in this advent of technological data

acquisition methodology is the speed, accuracy and data storage capacity of the systems.

26

Rugged and miniaturized computer packages are now available for ﬁeld data collection
and storage. Several research institutions have developed tailor-made packages for DAP-
SYS data acquisition primarily to collect physical and physiological data independently.
CEEMAT in France, Silsoe Research Institute (UK) and the Swedish University of
Agricultural Sciences, Uppsala, have developed instrumentation for monitoring DAP
energy utilization. O’Neill (1989) documented the development of a computerized
instrumentation package that has been used for a comprehensive assessment of draft
animal performance. This package has been used in India and Ethiopia as reported by
O’Neill and Kemp (1989). The data collected using the instrumentation package included
mechanical variables to monitor the energy expenditure by the animals and the
physiological responses of the animals at work. The instrumentation was made to work in
such a way that the "animal-implement" system was considered as a single entity composed
of various components in order to observe the interactions that occur between animals and
implements and which may inﬂuence the system's performance. The conclusions drawn
from the experiments conducted by O'Neill and Kemp (1989) indicated that the
instrumentation was reliable.

The ideal method for the determination of energy use is by the collection and analysis of
data on oxygen consumption rates (V02 - ml/kg/hr) and the corresponding rates of carbon
dioxide emission (V C02 - ml/kg/hr). The respiratory exchange ratio - VC02N02 - (RER)
obtained from the ratio of oxygen used versus carbon dioxide produced can be used to
compute the heat increment and hence the energy used as well as the mix of substrates
(types and mixtures of fuel) burned in the process (Wilmore and Costill, 1994). The RER

value varies with the type of fuels being used for energy production (Appendix B). The

 

27

amount of oxygen needed to completely oxidize a molecule of carbohydrate or fat is
proportional to the amount of carbon in that fuel. For example, glucose has six carbon
atoms. During combustion six molecules of oxygen are used to produce 6C02 molecules,
6H20 molecules and 38 ATP molecules". For example, RER for carbohydrates is 1.0, fats
have an RER of 0.71 and proteins 0.69. Oxidation of these fuels yields 5.05, 4.69 and 4.46
kcal per liter of oxygen consumed respectively (Wilmore and Costill, 1994). Instruments
like the Oxymax System by Columbus Instruments, Ohio, USA, are perfected for
measurements of gaseous exchanges in animals but are restricted to laboratory applications
where work situations are simulated with small animals such as rats and rabbits.

According to Pearson (1985) and Rautaray (1987) the most important physiological
responses to work that can be measured on working animals without causing them too
much discomfort are changes in heart rate, breathing rate and temperature. The greatest
stress to the animals as observed by Rautaray (1987) was due to erratic draft force which
elicited a higher heart rate response than a smooth draft force at the same mean power
output.

2.7 Computer simulation models

Simulation modeling in agriculture has grown in parallel with the popularity of
systems analysis. The demand for a better understanding of the real agricultural systems
makes simulation modeling a valuable tool for analyzing existing systems and designing
new ones. Like other farming systems, DAP systems are dynamic in nature and are
based on biological, physical and economic principles. A systems approach to DAP

simulation enables the various components involved in the production of and utilization

 

‘ 6C02 +C, 11,, o, —)6C02 +6H2 0+38A TP. .-. the RER =6VC02 /6V02 =1.0

28

of energy by animals to be monitored and predictions made for varying endogenous and
exogenous factors.

The concept of modeling, as deﬁned by Manetsch and Park (1993), is the construction
of an abstract model as a representation of “the real thing” where the latter is the real
situation or problem that is being studied, and the model is a useful representation of that
situation that can be dealt with intellectually and reasonably in terms of mathematics or
computer programming. Computer simulation models have become useful tools in
modeling agricultural production systems. This research technique is considered cost
effective because near-real systems that mimic real world physical systems can be
designed and constructed on computers. There is no limit of the level, number and
conﬁgurations of manipulations that a comptrter model can be subjected to, hence
providing large possibilities of scenarios. Naylor et al. (1968) describes the purpose of a
simulation model as being to enable the analyst to determine how one or more changes in
aspects of a modeled system may affect other aspects of the system as a whole. System
modeling on computers has been applied in machinery systems management. Various
researchers have used computer simulation for machinery management systems. Tokida
(1992) developed a Tillage Machinery Management Model for Rice Farming Projects in
Kenya. Pingali et al. (1987) noted that appropriate agricultural mechanization should be
developed based on a farming system model. Ismail (1991) also developed a Simulation
Model for Field Crop Production Machinery System. Little work, though, has been done
on computer simulation of DAP systems. Phillip et al. (1986) applied principles of linear

programming in their research on economic implications of animal power in Nigeria.

29

Although much work has been done on DAP in various developing countries,
centralized and comprehensive documentation is scarce. Renewed interest in DAP has
changed the status quo. In the recent past, research agencies and individuals have made
efforts to document DAP literature and other related materials. Notable publications
include a selected animal traction bibliography by Goe and Hailu (1983) and a
comprehensive one by Starkey (1991). Several researchers have collected data that can be
used to estimate coefficients of variables for DAP parameters. Data collection for a
computer simulation program requires that parameters be estimated accurately. Similarly,
the range of variables used should be realistic in order to reﬂect real world scenarios. The
data collection procedures used in this research included a soio-economic data survey
administered through a questionnaire by the researcher and an electronic DAS for the
physiological and mechanical data of draft oxen at work in two districts in Eastern Province

of Kenya.

CHAPTER III

DATA COLLECTION INSTRUMENTS

3.1 Introduction

The data required for the development of the DAP computer simulation model were
obtained through two methods using two types of instruments. Socio-economic baseline
data covering historical aspects, operational factors and resource endowments for a cross-
section of small-holder farmers in two administrative districts of Eastern Province of
Kenya were obtained through a survey questionnaire. The procedure of the baseline data
collection is described in chapter IV. Secondly, a custom-made DAS was used to collect
empirical data of energy utilization of draft animals while performing tillage operations.
This chapter will describe the DAS equipment. Both the empirical and socio-economic
data collected were used to design and implement an energy utilization model code
named Energy Utilization Model for Draft Animal Power (EUMDAP) for DAP systems
in semi-arid lands of Kenya.
3.2 Data acquisition system hardware

Various researchers have used data acquisition systems similar to the one described in
this chapter for collecting physiological and mechanical data that can be used in the
estimation and prediction of energy expenditure of draft animals at work. Draft animal
energy data monitors have recently gained importance as researchers investigate the
nature and manner of energy delivery from draft animals to tillage and other farm
implements. The motivation driving this renewed interest is the need to enhance the

utilization of DAP energy source for improved land and labor productivity especially for

30

31

small-holder agriculture in semi-arid lands of developing countries where the risk of
inadequate feed supply is a chronic issue. The DAS for this research was designed and
made by the researcher in collaboration with the Swedish University of Agricultural
Sciences, Uppsala, Sweden, speciﬁcally to meet the animal draft energy data acquisition
needs for DAP systems of working draft oxen.

Animal power sources derive their energy for work from metabolic fuels stored in the
animals’ muscle and liver (glycogen peripheral pools) as well as in their fat (adipose
tissue). The process of energy conversion from the stored chemical energy to mechanical
energy at the cellular level (catabolism) is a complex one involving the breakdown of the
compound adenosine triphosphate (ATP), the immediate energy source for muscle
contraction, which contains a high energy phosphate bond (Bursztein et al. 1989).
Animals utilize oxidative metabolism (aerobic) for continuous energy supply when
working (Wilmore and Costill, 1994). The process of breaking down ﬁnite ATP and
replenishing it is a continuous one for any muscular activity to take place in an animal.

Energy expenditure is the key factor for quantifying animal work load physiologically.
The most direct method of determining the energy expenditure of animals is to measure
the gaseous exchange (02 and CO2) by direct calorimetry (heat increment) or indirect
calorimetry because the level of energy expenditure and the type of fuel oxidized are
directly proportional and related to the volume of oxygen consumed by the animal and

the volume of carbon dioxide released during the combustion of energy fuels.

The purpose of the DAS developed for this project was to monitor the mechanical
energy produced by oxen at work and the effect that the rate and intensity of work has on

the physiological response of animals at work. Traditionally the force exerted by draft

32

animals whilst pulling loads has been measured using a dynamometer such as a
mechanical spring, a hydraulic cylinder or strain gauge on a mechanical linkage. The
measuring transducer connects the power source (animal) and the load it pulls in such a
way that the reading obtained is proportional to the force exerted by the animals
(Lawrence and Pearson, 1985). Due to limitations of making direct accurate
measurement of energy expenditure by animals in ﬁeld conditions, new methods of
estimating it have been devised. Some physiological changes such as cardiovascular
responses to work can be used to provide information on the rate of energy expenditure
of animals in work load situations. Estimation of energy use through relative heart rate
(RHR) and the metabolic weight, (W,)°'75, of oxen have been used by various researchers
successfully (Richards and Lawrence 1984; Sneddon 1986; Mathers and Sneddon 1984
and O’Neill 1989). This research exercise used similar DAS equipment tailor-made for
the project and utilizing a commercial data logger to obtain both physiological and

mechanical data from DAP systems of oxen at work.

The DAS equipment consisted of four physiological sensors, three mechanical
sensors, 3 data logger, a portable computer, and a battery power source. The electronic
circuitry of the mechanical sensors was housed in a box together with a 7.0 V battery that
provided power for the whole DAS (Figure 3.1). A separate box contained the four
physiological sensors. The layout of the system enabled the data transmission cables
from each box to be fed into the data logger for temporary storage of the data before
dumping it into the portable computer or ﬂoppy diskettes. The system was designed to
operate as an integral part of the animal-implement DAP system with all the data from

the transducers being monitored and recorded simultaneously. Figure 3.2 shows the

33

complete layout of the instrumentation as it was mounted on both the draft oxen and the
tillage implement. When the DAS was collecting data, it was operated as one integral
DAS-DAP system operating as a unit with all the data collected being channeled to the

RAMLOG E19000 data logger. Three electronic circuit cards for the physiological and

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

TILLAGE DEPTH ﬂ
VERTICAL DRAFT >__ B
A
HORIZONTAL DRAFT CIRCUIT T
>—4 BOX T
E
GROUND SPEED JL R
Y
W
HEART RATE RAMLOG E19000
DArALOGGER
BODY TEMPERATURE J1— ._
ALPHA - L To
“5mm" ”“5 l_]—‘l— CIRCUIT NUMERIC C ————t°°m
3mm BOX KEYPAD D “‘5 23”
ADAPTER “
STEP RATE l—ﬂ

 

 

 

P SIOLOGI AL SENSORS

 

Figure 3.1: Block diagram of Data Acquisition System (DAS).

mechanical transducers were built and housed in three separate boxes. The circuits for
the four physiological variables, heart rate, breathing rate, body temperature and
stepping rate, were housed in a plastic box that was secured on the side of the
animal’s body with a leather belt strap during the data collection exercise. This box

was mounted in such a way that the researcher could monitor the analog needle

34
display which monitored the operation of the heart rate sensor during the data collection

process. The box containing the tillage depth and ground speed electronic circuitry and

the battery power pack were contained in a separate box which was mounted on the plow

beam together with the load cell circuit box (junction box). The load cell sensor output

 

Figure 3.2: Schematic view of the animal-implement Data Acquisition System instrumentation.
1: Load cell; 2: Depth-ground speed sensor junction box; 3: Plow beam; 4: Body temperature
sensor; 5: Physiological electronic-junction box; 6: Heart rate sensor; 7: Respiration rate

sensor; 8: Depth-ground speed measuring wheel; 9: Ground speed sensor; 10: Depth sensor:

11: Mechanical electronic-junction box; 12: Ramlog E19000 data logger; 13: Step rate sensor.

cables transmitted data to the logger via the depth-speed wheel circuit box. Two shielded
cables, one for each respective set of sensors, transmitted the data from the
respective junction boxes to the data logger through a 37-pin female-male adapter
connector. The transmission cables were long enough to allow the researcher to walk
along-side the DAS-DAP system some distance away from the plow operator and the
animals as he operated the data logger.

One 25.4 cm (10”) single-bottom Victory moldboard plow was modiﬁed and used

for all of the experiments including those at the farmers’ ﬁelds. Using the same plow for

35

all experiments eliminated equipment variability as a factor. One plow share was used
for all experiments. It did not show obvious signs of wear that would have necessitated
its replacement.

The central feature of the DAS was a 20 channel hand held data logger (RAMLOG
E19000) weighing about 500 gm including batteries. The logger is an interactive
commercial microcomputer developed by A.B.I. Data of Belgium. It operates on a real
time clock with a programmable system software for monitoring, sampling, recording,
and saving data in a 128 KB RAM. It operates on a power supply range of 4.2 to 7 V
provided by either four AA alkaline batteries or a commercially available rechargeable
power pack (PU-1) that has a life span of 30 hours when fully charged. The logger’s
memory capacity is 1.5 x 106 data points when in continuous use with multiple channels
(RAMLOG E19000 User Manual). The micro-computer-user interface is through a full
alphanumeric keyboard that is menu driven.

The signal sensors were connected to the RAMLOG data logger through four types of
interface modules (signal conditioners): voltage/current inputs; counter inputs/digital
event inputs; thermocouple inputs; and temperature/humidity/pressure inputs. The
differential and single ended analog data acquisition sensors had a programmable range
from i 100mV to i 10V. The analog-to-digital converter (ADC) had a capability of 11
bits plus sign, a programmable gain ampliﬁer and software calibration. Technical details
of the logger are provided in Appendix C.

When the logger RAM ﬁlled up, the data were transferred to a portable computer via
an R8232 cable using the built-in data transfer software (FIRMWARE). The portable

computer used for downloading the data was an IBM compatible CityNote with a

 

36
486DX hard disk of 4 MB working memory (RAM) and 170 MB storage ROM with aa

operating speed of 25 MHz.

Seven data sensors were used for the physical and physiological variables of the
animal-implement system. Three mechanical sensors monitored the tillage depth, ground
speed and the draft force (vertical and horizontal components) generated by the
implement. The plow beam of the Victory plow used for the experiments was cut into
two pieces to accommodate the draft force transducers (load cell) which were bolted as an
integral part of the two section-plow beam. A 50.8 cm (20”) bicycle wheel (depth-speed
wheel) was instrumented to measure tillage depth and ground speed. The axle of the
depth-speed wheel was connected to a hinged frame whose vertical section was bolted to
the plow beam in such a way that the wheel rolled freely on the unplowed land and also
the wheel’s axis of rotation moved vertically in response to the changing depth of tillage
because of its hinged joint (Fig. 3.2).

The speed sensor (pulse generator) was incorporated with the axle of the depth-speed
wheel thereby rotating with the wheel to measure the ground speed. It generated pulses
per unit time interval at the rate of 500 pulses per wheel revolution. The pulse generator
was mounted as an integral part of the axle of the wheel. A linear potentiometer was used
for the tillage depth measurement. It was mounted at the hinge joint of the depth—speed
wheel frame to register the vertical displacements of the wheel. The range of depth
wheel’s vertical measurement was from 0 to 29 cm.

3.3 Data acquisition system software

The RAMLOG E19000 data logger can be programmed to work on all 20 channels or

only a few of them, depending on the number of transducers in use. Each of the channels

37
can also be programmed to receive data in accordance with the individual needs of the

sensors and also to meet the needs of the user. The logger has sixteen analog channels
(channels 1 to 16), two 8-bit counters (channels 17 and 18) one 16-bit counter (channel
19) and one 8-bit digital input channel (channel 20). The program used for monitoring,
recording and saving of the data as well as programming of the channels, was a built-in
system software (FIRMWARE) that was stored in the logger’s non-volatile memory. The
data acquisition rate ranges from 10 msec interval (100 data points/sec) for one channel to
100 msec interval (10 data point/sec/channel) for 16 channels. On-site monitoring of the
data is done through 2 lines of a 16 character liquid crystal display (LCD). User
communication is done through a 20-key keypad with triple functions, ﬁrll alphanumeric
with punctuation, arithmetic operators and function keys. Initialization of data collection
is triggered by the “F1’ key and data collection is terminated by “F2” key. This enables
the user to vary the time and or distance covered by a particular “run”. Data sampling
and recording automatically stops when the logger power supply falls below 4.4 V
reserving the remnant power for memory content reading and peripheral transfer or
dumping of the data. The data holding capacity is 77,500 data points, (analog - 12 bits)
for the 128 Kbytes version. The capacity of the logger when using 16 channels at a
sampling rate of 10 data points per second per channel was about 8 minutes.

A complementary menu-driven data processing software (PROLEC-TRA EI9000) that
is stored in the portable computer handles the data transfer to a PC (via RS 232 cable) as
well as subsequent initial data analysis and data saving in various formats. Processed
data can be saved in ASCII (Text, Lotus 123 or Dbase), or graphic versions. For each

sensor, the PROLEC-TRA software calculates the average, maximum and minimum

38

values. These statistics are stored in the volatile memory of the logger together with the
individual data points. During each data collection exercise, before a new ”run” is started
the user is prompted to create a new record name to be used for subsequent data saving in
the same ﬁle.

A maximum of 8 digital and or analog channels can be displayed on the computer
screen at any one time in graphic format. Print-outs of the results are available in
graphic, tabular or mathematical formats for 8 channels at a time.

3.4 Calibration of data sensors

The ground speed and the depth sensors were calibrated by using the data logger to
read the voltage generated by the potentiometer and also the pulses recorded over a
predetermined distance on the ground. The ground speed was later computed on the
basis of the rate at which the pulses were generated. Calibration of each of the
mechanical sensors was done and a calibration response equation computed. Calibrating
the potentiometer involved clamping the wheel assembly on a bench vice, varying the
height of the wheel and taking data logger voltage readings corresponding to the depth.
The R2 value for the calibration curve was 0.999 (Figure 3.3).

Calibration of the speed sensor was done to establish the number of pulses per
revolution and the respective ground displacement. Data logger readings of the
cumulative pulses generated over a ﬁxed distance, time taken and the number of
revolutions over the distance were used for the computation of ground speed.

The energy that the animals develop is translated into a pull force when they are used

for work. The draft force generated and put to useful work constitutes a fraction of the

Depth (cm)

39

total effort that the animals produce. Measuring the pull force and comparing it with the

actual net energy expenditure would provide an estimate of the energy utilization of the

 

   

 

 

 

30
25 _
20 I.
15 .
10
y = 34.376x-72.704

5 R2 = 0.999

0 4 ¢ + i

2.10 2.30 2.50 2.70 2.90 3.10

Logger reading (V)

Figure 3.3: Depth sensor calibration curve

animals at work and also the gross and net energetic efﬁciency. The draft force was the
most rapidly changing variable. It was measured through vertical and horizontal draft
force components of the pull force using an extended octagonal ring type transducer
(dynamometer). The force transducer was integrated into the plow beam (Figure 3.2) in
such away that the horizontal and vertical components of the forces acting on the soil
engaging part of the plow were measured independent of each other. The analog signal
(voltage) from the load cell was digitized and logged at 0.06 sec intervals (approximately
17 Hz). The transducer was calibrated by loading it in both “X” and “Y” axes when the

unit was clamped in a rig speciﬁcally made for the calibration process. Analog voltage

40

readings of each load were recorded and regression analysis done. Calibration of the
transducer showed that the interaction between the vertical and horizontal forces was less
than 1.5% and the observed hysteresis was insigniﬁcant. The R2 value of the calibration
curve for the horizontal and the vertical draft were 1.0 and 0.9998 respectively (Figures
3.4 and 3.5). The calibration response equations and the R2 values for the mechanical
sensors are shown in Table 3.1.

The regulation of the amount of nutrients oxidized, blood ﬂow rate and the oxygen
demand for the generation of energy from food substrates is done by the cardiovascular
system. Gottlieb—Vedi et al. (1991) established that there is a linear relationship between
the heart rate of draft horses and the rate of oxygen uptake. The pumping rate of the
heart of a working animal can, therefore, be used as an indicator of the energy demand
and expenditure of a working animal. Whereas most animals work at sub-maximal levels
of heart rate, the maximum heart rate indicates the physiological endurance limit.
Measurement of heart rate in a laboratory environment can be done with instruments that
have been perfected for that purpose. Richards and Lawrence (1984) used
electrocardiograph (ECG) to record the heart rate of oxen. The instrument measures the
potential difference associated with ventricular depolarization (QRS complex) phase of
the heart beat. In their experiments they found that the heart rate correlation with the
energy expenditure of oxen indicated that heart rate can be used for the prediction of
energy expenditure.

Measurement of heart rate in ﬁeld situations requires a method that would be accurate
and un-intrusive to the animals. A method of detecting the changes in infra-red

absorptance of the ear as blood pulses through it was used successfully by O’Neill (1989)

41

1200

 

1000 A

800 ..

600 4»

Load (N)

400

    

y =11848x - 1056.7

200 . R2=1

 

 

 

0.09 0.11 0.13 0.15 0.17
Logger reading (V)

Figure 3.4: Horizontal draft calibration curve

 

1000 -.

800 .,

600 ..

Load (N)

400 -.

y = 10601x-23.808
200 ._ R2: 0.9993

 

0 l i a: t Y
0.005 0.025 0.045 0.065 0.085 0.105
Lower wading (V)

 

 

Figure 3.5: Vertical draft calibration curve

 

42
and Zerbini et al. (1992). An ear-clip sensor comprising of two arrays of three diodes

was used to sense the QRS complex or peak surge of blood ﬂow. One array emits near
infra-red radiation and the other, on the other side of the ear, is sensitive to it. A signal
conditioning circuit was made to give an electronic square wave coincident with the
maximum absorptance of the blood ﬂow. A similar sensor was fabricated for this
research. Several prototypes of the heart rate sensor were made and tried in the ﬁeld for
sensitivity and accuracy. The ﬁnal version was clamped on the ear of the animal and
secured with medical tape to minimize interference from sunlight. The signal from the
heart rate sensor was channeled through both analog (voltage) and digital circuits.
Obtaining a suitable position for the heart rate sensor to provide transmittal of the signal
to the logger was a considerable challenge. Analysis of the heart rate was done using the

digital signal.

Table 3.1: Calibration response equations

 

 

Channel Variable Calibration Response Equation R2 n
1 Horizontal Draft Draft (N) =11848.1279 ‘ V - 1056.663 0.9999 27
2 Vertical Draft Draft (N) = 10600.575 ‘ V - 23.808 0.9998 27
5 Depth (cm) Depth (em) = 34.3756 " V - 72.7044 0.999 27
19 Distance (m) Distance (m) = 0.004347 ' P - 3.6447 0.9443 13

 

The rate of breathing, although by itself is not an indicator of the energy demand from
working animals, can be used to indicate the physiological response of the intensity of

work. The breathing or respiration rate was determined by sensing the air movement

43

through a tube in one of the animal’s nostrils. Inside the tube was a miniature heater
element. The heater yields differential temperature signals depending on the moisture
content of the air ﬂowing during inhalation and exhalation. The differential temperature
signals are counted in real time and are used for computing the breathing rate.

The body temperature was measured with a standard bead thermistor model made by
Fenwal Electronics USA. This is a small NTC Type 112 K fast response thermistor
measuring 1.09 mm (0.043”) in diameter and is hermetically sealed in glass. It is highly
reliable and has long term stability and had a 20% tolerance at 25°C. It was mounted in a

plastic probe that gave it protection for handling. Its resistance at 25°C is 100,000 9 and

requires a minimum power supply of 0.4 mW/°C.

The step rate sensor used was a commercial accelerometer type 8308A model K-Beam
made by Kistler of Switzerland. It consists of a miniature sensor with built-in electronics
to measure the lowest acceleration down to 0 Hz. The sensor generates an electronic
pulse during peak acceleration which is logged in real time. It is suitable for a wide range
of applications which include measurements in low speed machines and in bio-
mechanics. It has a high sensitivity and high temperature stability. The zero reading of
the signal is set at +2.5 V so that the output voltage is always positive and has a full scale
output of i1 V. It was connected to an A/D converter with 0 - 5 V input. A leather
pouch was made for the sensor which was strapped on the front foreleg of the animal.
The sensitive axis of the accelerometer was mounted parallel to the direction of the

movement. Technical details of the accelerometer are provided in Appendix D.

CHAPTER IV

FIELD EXPERIMENTAL METHODOLOGY

4.1 Data collection sites

The locations for the ﬁeld experiments were chosen in two districts in Eastern
Province. Although DAP is widely used in three Provinces of Kenya (Eastern, Western
and Nyanza), Western and Nyanza Provinces are in the high potential agro-climatic zone
(Table 4.1). The agro-climatic zones of most areas of Eastern Province where DAP is the
dominant mechanization system is zone 111, an area with medium to low crop production
potential that is generally referred to as semi-arid. The climatic conditions of agro-
climatic zone 111 describe the medium and low potential area from the standpoint of
agricultural output. On the other hand zone 11 is considered as high potential area which
does not have similar agroclimatic constraints that limit agricultural production present in
zone 111.

Eastern Province is divided into four administrative Districts (Machakos, Embu, Meru,
and Kitui). Machakos and Embu were in the process of being subdivided into Machakos
and Makueni districts and Embu and Mbeere districts, respectively, during the time the
research was being conducted. The research sites chosen for the ﬁeld work were in the
new Machakos and Mbeere districts. The choice of the two districts was based on
logistical considerations, and also on the fact that they represented the semi-arid agro-
climatic zones. Parts of lower Meru district that borders with Mbeere district have
similar climatic conditions as Mbeere. Most of Kitui district which borders with

Machakos district has the same climatic regime as Machakos district.

44

45

Table 4.1 Kenya’s agro-climatic zones and rainfall regimes

 

 

Agro—ecological Area Annual
Zone Description (ka) % of Total Rainfall (mm)
1 Humid and Semi-humid minor % in Kenya5
11 High Potential 67,850 11.89 857.5 or more
111 Medium Potential .
(Semi-Arid) 31,570 5.54 612.5 to 857.5
IV Low Potential 422,370 74.07 612.5 or less
V and V1 All Other Land 48,430 8.49
Total Land Area 570,220 100.00

 

Each district is divided into four or ﬁve administrative divisions. Research sites selected
for the Machakos district were in three of the fOur divisions - Mwala, Central and
Kalama. Mbeere District is divided into two divisions - Mbeere and Gachoka - which
were covered in the research. In each of the ﬁve divisions, farmer surveys were
performed to obtain socio-economic data of the small-holder peasant farmers which
represented the historical and Operational data base for the computer simulation program.
4.2 Socio-economic data collection

The socio-economic data were collected through the administration of a questionnaire
(Appendix I) with 80 randomly selected farmers in Machakos and Mbeere districts.
Sixteen farmers were randomly sampled from each of the ﬁve administrative divisions of
the two districts and interviewed to obtain the required data base. In selecting the
farmers, the only criteria that had to be met was that the land size of the farmers was not

to exceed 8 ha (20 acres). As noted earlier, the vast majority (96%) of small-holder

 

5 climate governed by altitude, not moisture.

46

farmers in semi-arid areas have 8 or less hectares of land on which they grow their food
crops and also keep livestock.
The data collected included the following general areas of DAP utilization:

1» farm size and acreage tilled
- mechanization resources available
0 oxen utilization
hours and number of days oxen were used
ntunber of days farmer hired or rented oxen
rates of renting and hiring (cost)
tillage practices
timing of tillage operations
post plowing operations
labor availability
family size
hired labor
crop yields and sales
0 current energy constraints and future plans

All interviews were conducted by the researcher personally after the working season of
the 1995 short rains (September through November) and after the working season of the
1996 long rains (March through May). All of the interviews were conducted by the
researcher in order to minimize personal errors or enumerator error, reduce the data
collection costs and maintain consistency of eliciting information from the farmers. The
researcher translated the English questionnaire to the respondents’ local dialects with
which he was familiar in both districts. The researcher understood the two languages
spoken in Mbeere district (Kiembu) and in Machakos district (Kikamba). Likewise the
farmers in the two districts can understand Kikuyu, the mother tongue of the researcher.
There was, therefore, no need to translate the questionnaire into either the national

language (Kiswahili) or the native languages spoken in the two districts.

47

A cross-section of the farmers interviewed were followed up for the DAP energy data
collection with the DAS during the long rain tillage period when they were preparing the
land for planting. Three farmers were selected out of the sixteen in each division. The
DAS energy data collection was performed on a randomized block design pattern.
Collection of the energy data was preferably done during the actual tillage period in order
to obtain data that were pertinent to the farmer’s calendar of cropping events and at the
time that farmers usually prepare their land for planting. Small-holder farmers in the two
districts do not leave any of their open land fallow. It was therefore not possible to get
free parcels of land for carrying out the energy experiments when the tillage season was
over.

4.3 Energy expenditure data collection

The energy expenditure data collection was done at the University of Nairobi farm
(Field Station) and in farmers’ ﬁelds. The procedure for collecting energy data was the
same for both the Nairobi experiments and the experiments in farmers’ ﬁelds. The DAS
equipment and the farmers’ animals were used with the same modiﬁed moldboard plow
to obtain data on physiological response of animals at work and the mechanical energy
requirement of the DAP implement. In each case soil moisture data was obtained as well
as the data on the soil shear strength. Tests of each kind were taken at four sites in the
ﬁeld at random. For each test site three moisture content samples were obtained at three
locations at three depths (5cm, 10cm and 15 cm) in each location for a total of nine
samples. Soil moisture content was obtained with Time Domain Reﬂectometry (TDR)
that measured instantaneous volumetric soil water content in situ (Appendix E). Soil

shear strength readings were also obtained with a hand penetrometer at three places in the

48

ﬁeld. In each location readings were obtained at three depths (5cm, 10cm and 15 cm).
The shear strength values were used to estimate the speciﬁc resistance of the soil for
simulation.

The weights of the animals used for the ﬁeld tests were estimated using body physical
measurements. The two measurements needed to compute the weight of the oxen were
the body length, from the point of the shoulder to the point of the rump-pinbone (length)
and the circumference (heart girth), from a point slightly behind the shoulder blade, down
over the fore-ribs and under the body behind the elbow. The conversion of the
measurements to the weight was done with the following formula adapted from

Ensirninger (1990) and the measurements taken in cm:

h t 'th 2 *1 th
Bodyweight(kg) =( W W620 “3 (7)

 

The above weight conversion formula or variations of it have been used for weight
estimation of cattle by various researchers in East Africa (Young, 1972; and Semenye,
1979).

Estimating the oxen weights using the physical measurements method is considered
accurate enough for the purpose for which the estimated weights are used. The oxen used
at the University of Nairobi experiments were measured both by a weigh bridge and with
the physical measurements in order to ﬁnd out how closely the physical measurements
estimated oxen weights. The data obtained showed that the estimate was within 21:10%

which for the purpose of the experiments performed was acceptable. The University of

49

Nairobi oxen like most of the draft oxen used in Kenya were locally bred Zebu type.
However, the University of Nairobi oxen were well fed and maintained and hence the
estimation of the weight by physical measurement method gave a higher value than the
actual weight. Although estimating the weight of the farmers’ draft oxen with equation
(7) was expected to be within i10%, this method was likely to overestimate the oxen
weights due to the poorer maintenance standards of the farmers’ oxen.

The purpose of the tests at the University farm was to perform a pretest with the DAS
in controlled experiments. From these tests relationships between the oxen’s
physiological parameters and the mechanical energy used to pull implements when
varying the draft force were obtained. Changing the depth and width of tillage varied
draft force. The pretests were necessary to ensure that all the instrumentation was
working well and that the data being collected were reliable. Data obtained from the ﬁeld
tests were analyzed and the results are presented in Chapter V.

The mechanical sensors were relatively easy to work with because they were already
connected to the source of the signal. The physiological transducers were, however,
difﬁcult to handle because they had to be connected to the animals at various positions.
The draft oxen at the University farm were well trained and so working with them was
not difﬁcult. Positioning of the sensors on the animal required patience and practice to
do it well.

The step rate on the animal’s foreleg sensor was the easiest one to attach. There were

practical problems of connecting the breathing rate sensor, the body temperature sensor

50

and the heart rate sensor. Several attempts were made before placing the temperature
probe into the rectum of the oxen. Medical tape had to be used to maintain the
temperature probe in position. The heart rate sensor comprised of two arrays of three
diodes in an ear clip that was clamped into place over the animal’s car. It was difﬁcult to
ﬁnd a suitable position from where the heart rate reading could be obtained. The analog
heart rate needle indicator on the physiological sensors junction box was used to identify
a suitable position because it indicated the pulses when the signal was received by the
electronic circuit. When a suitable position was found, medical tape was used to secure it
in place as well as to help keep sun light out of the sensors. The heart rate sensor worked
on the principle of absorptance of near infra-red radiation (0.7 to 102 um wavelength)
which is emitted and sent through the ear by three diodes while another set of diodes on
the other side of the ear is sensitive to it. When blood surges through the blood veins of
the ear, the infra-red radiation is absorbed by the blood. The degree of absorption
depends on the volume of blood ﬂowing. The signal conditioning circuit was made to
give a binary pulse read-out (from 0 to 1) when the peak surge of blood was sensed. The
peak surge corresponds to the systolic phase (QRS complex) of the cardiac cycle when
the heart is in the ventricular contraction. The interval between each QRS complex
represented the period of each heart beat.

The breathing rate sensor was housed in a plastic tube that was attached to a muzzle
and then placed into the nostril of the animal with a muzzle holding it in place especially
because the animal tried to remove it with its tongue. It was ensured that the tube was
ﬁtted into the nostril so that the animal breathed through it. The sensor received the

breathing rate signal through the differential temperature of the inhaled and exhaled air.

51
Moist air coming from the animal’s body cooled the miniature heater on the sensor.

Inhaled air being drier caused the temperature of the miniature heater to rise. The
differential temperature was conditioned and read as pulsesby the data logger.

Trials at the farmers’ ﬁelds were conducted to obtain the variability of the energy
expenditure in various farming situations and with different sets of oxen. The same
equipment (plow) was used for both the University of Nairobi experiments and the trials
on farmers’ ﬁelds. The reason for this approach was to eliminate the equipment factor as
a variable. The integration of the draft measurement transducer into the plow beam also
precluded the use of different plows. The Victory plow used for the experiments is
almost exclusively used by DAP farmers in Kenya. The plow’s widespread use is
primarily due to lack of other suitable implements in the market. The Victory plow is a
single bottom moldboard plow ranging in size from 30 to 38 cm (12” to 15”) wide. A
ground wheel that rides ahead of the plow bottom is used for depth adjustment. The
width of plowing is regulated with a U-bolt that is pOsitioned laterally across the front
end of the beam.

4.3.1 Data logger programming

Before the RAMLOG E19000 data logger could be used to collect, record and store
data, it had to be programmed or conﬁgured for the desired mode of operation. A two-
line l6-character alphanumeric liquid crystal display (LCD) provides a display for the
key pad entries as the user programs the logger. Nine of the 20 channels available in the
data logger were used for these experiments. Table 4.1 shows the logger channels and the
transducers that each of them processed. Further information on the logger parameter

settings are available in Appendix C. Three of the four mechanical sensors (horizontal

52

and vertical draft, tillage depth), the battery power, and the body temperature of the
animal were monitored by the analog channels of the data logger with the output of the
reading in volts (V). The ground speed sensor was monitored in pulses (P) by channel 19
which is a 16-bit counter. The heart rate was monitored in both analog (channel 6) and
digital (channel 20) formats. The same process was used for the respiration rate sensor in
channels 7 and 20 while the digital channel measured the step rate only. The digital
channel (20) has eight digital (binary) inputs. The heart rate was required for the
estimation of the energy expenditure of working oxen. Other physiological parameters,

including respiration rate were obtained mainly to monitor the animals’ response to work.

Table 4.2 Parameter settings for RAMLOG E19000 logger DAS

 

Channel Number Function Scale Range

 

Mechanical Sensors:

 

 

1 Horizontal Draft 0.5 V 0 - 4kN
2 Vertical Draft 0.5 V . 0 - 2kN
5 Tillage Depth 5.0 V 0 - 29 cm
19 Ground Speed Pulse 0 - 2 m/sec
9 Battery 10 V
Physiological Sensors:
6, 20 Heart Rate 10 V 40-160 beats/min
7, 20 Respiration Rate 10 V 14 -140 breaths/min
8 Body Temperature 0.5 V 25 - 45° C
20 Stepping Rate Pulse 0 - 100 steps/min

 

The heart rate and respiration rate were monitored in both analog and digital forms in
order to identify the peaks of the traces of the pulse rate and also correlate the trace with
the binary (0, 1) or “low” and “high” positions of the pulse rate. The peaks of the trace

and the digital output were clearly identiﬁable in the graphical outputs of the PROLEC-

 

53

TRA over the time interval of the experiments. The data analysis results covered in
chapter V show graphical outputs in both analog and digital format. A standard mode (or
point-to-point mode) of conﬁguration for the logger was preferred for the energy data
collection. In this mode the data logger sampled each sensor’s signal for a short duration
of time which was determined by the length of the run or the time span in which the data
was to be collected. The number of points visited by the logger for each sensor was
contingent on the rate or frequency of sampling. For instance at a sampling rate of 10
Hz, 1200 data points per channel (sensor) would be visited and recorded in two minutes.
For each channel, the logger calculated and recorded the average, maximum and
minimum values as well as the individual values. In this mode, the measured values were
also displayed cyclically on the LCD for on-site monitoring of the logger readings as the
sampling proceeded. The cumulative data points were saved in the read only memory
(ROM) of the logger at the end of each experiment in a speciﬁc ﬁle name for subsequent
down loading to a computer.

Recording conﬁguration also included setting the date and time, deﬁning the recording
name for each of the ﬁeld trials, the name of each run, the recording interval in centi-
seconds and the record begin option. The appropriate record begin option for the ﬁeld
experiments was through the data logger keypad. Pressing the F1 key when the operator
was ready to collect data started the data collection process. This was done after the
animals and the implements were in a steady state mode. Pressing the F2 key stopped the
data collection. The ability to stop and start data collection any time using the “F 1” and

“F2” keys was convenient especially when the animals stopped or encountered an

54
obstruction (stone or stump) in the ground. At the end of each run the logger prompted

for a new record name to be used for the subsequent saving of the data.

In the standard mode, each channel was conﬁgured individually to meet the respective
sensor’s parameter settings as well as the sampling and recording frequency needs. The
parameter settings for the channels included:

the channel number
function of the sensor
scale or measuring range
units of measurement
scaling factor

record type and
recording interval

The recording interval is a multiple of the sampling frequency, which allows the
recording interval to be varied to suit each signal’s rate of change and the memory
capacity of the logger. Channels that had fast changing sensors should be sampled at a
higher frequency to capture the short interval of change. For these research experiments
the sampling frequency for all the sensors was set at 16.7 Hz (0.06 sec interval) based on
the rate of change of the draft force signal which was the most rapidly changing one. For
all of the sensors, all of the sampled data points were recorded so that 16 records were
obtained per second for each parameter as shown in the printout in Appendix F.

After the initial programming, only a new record name was needed for the subsequent
runs until the logger’s memory was exhausted and the data transferred to the portable
computer or into ﬂoppy diskettes. After the data were transferred, it was necessary to
reprogram the logger for all the channels except the date and time before proceeding with

further data collection.

55

4.3.2 Experiments at University of Nairobi farm

The ﬁeld site for the controlled experiment performed at the University of Nairobi was
the Field Station farm. The ﬁeld was fairly ﬂat and had been under a maize crop during
the preceding short rains season. The design of the experiments was randomized block
design (RBD) consisting of six treatments replicated three times. The six treatments
comprised of three depths (5cm, 10cm and 15cm) and two widths of tillage. Although
attempts were made to maintain the required depth and width of tillage, the moldboard
plow used was not designed for precise settings for depth and width. The actual depth
and width of tillage was roughly estimated by the operator and was later measured at the
end of each “run”. Each replication was treated as a block and all the experiments for

each block were performed and completed in one day. The variables of the treatments

were:
Treatment Deﬁnition
(replication 1)
5NRR1 5 cm depth, narrow width
5WR1 5 cm depth, medium width
10NRR1 10 cm depth, narrow width
lOWRl 10 cm depth, medium width
15NRR1 15 cm depth, narrow width
15WR1 15 cm depth, medium width

At the start of each experiment the following data and information were obtained

for each ﬁeld.

0 soil shear strength
0 soil moisture content
0 weather conditions

At the start of each set of tests, a data recording conﬁguration was programmed for each

56

ﬁeld. This included such parameters as the date and time, ﬁle name to be used for down
loading data, and the name of the “run”. Conﬁguration for each of the channels used was
done by setting the following parameters:

channel number

function

unit of measurement

scaling factor

sampling and recording interval

The test “runs” ranged in length from about 20 m to about 37 m and the completion time
for the test runs ranged from 14.5 to 120.0 seconds. The length of the “runs” was subject
to the speed of operation and the time that the data collection was started relative to the
end of the ﬁeld (see Table 5.10 in chapter V). Data collection was not started until the
DAP-SYS was considered to be in a stable operating condition. This caused a wide
variability of the length of the test “runs”. Data collection was also stopped before the
end of the ﬁeld if the oxen stopped due to obstruction in the ground. However the
sampling and recording rate was ﬁxed at 16.67 Hz. for all the channels.
4.3.3 Experiments in farmers’ ﬁelds

The energy data collection at the farmers’ ﬁelds was performed for 15 of the total
farmers sampled with the survey questionnaire. For each of the farmers’ oxen teams
tested, physical dimensions of the animals were taken for estimating their weights. The
same instrumented Victory moldboard plow was used for all the experiments both at the
University of Nairobi and all off-campus energy data collection.

Experiments were performed in ﬁve administrative divisions. Mbeere district has only

two divisions while Machakos district has four divisions of which three were used for the

57

experiments. Each of the divisions was considered an experimental block consisting of
three draft animal teams (one from each farmer) to be tested. Therefore, the data analysis
for each block (or division) was done separately before it was pooled for the whole
district. Since the same tillage equipment was used for all experiments, the three draft
oxen teams were used as factors for the trials. In each site (farm), the tillage trials were
replicated three times within the same day for each of the three farmers per division. The
experimental plan was randomized block design (RBD) for the divisions with the
experiments done at three different sites. The same operator was used with the tillage
system throughout all the trials. He controlled the implement depth and speed while the
researcher did the data logging. The farmer’s help was, however, required to hitch the
animals and help control the animals while at work with voice commands.

The length of the test “runs” in the farmers’ ﬁelds varied from one farmer to another
depending on the conﬁguration of the ﬁelds, the level of control of the animals and the
starting of the data collection. The shortest ﬁeld “run” was about 14 m while the longest

was about 58 m.

CHAPTER V
FIELD EXPERIMENTAL RESULTS

The ﬁeld experimental energy data collected with the DAS and the socio-economic
survey data obtained from the farmers’ surveys were analyzed using different statistical
packages. Eighty farmers were surveyed for the socio-economic data in ﬁve
administrative divisions of Machakos and Mbeere districts, Eastern Province of Kenya.
Of the 80 farmers surveyed, 48 were from Machakos district and 32 were from Mbeere
district. The socio-economic data were analyzed with the SPSS data analysis package
while the electronic instrumentation data were analyzed with the complementary
PROLEC-TRA software of the RAMLOG-E19000 data logger and the EXCEL
spreadsheet program. The data obtained from the two sources were used for computation
of the energy utilization of DAP in small-holder tillage Operations, and the results were
used for the EUMDAP computer simulation model design and implementation. The
SPSS data analysis procedures used were mainly the, descriptive statistics which were
used for determining the mean, maximum and the expected variability of the energy input
variables and the resources used for the EUMDAP computer simulation model.
5.1 Socio-economic data analysis
5.1.1 Land holdings

The small-holder farmers surveyed were limited to those farmers that had a total
farm land area of 8 ha (20 acres) or less. As indicated earlier in Table 1.1, only about 4%
of small-holder farmers in Kenya have more than 8 ha of land. In Eastern Province
where the research was conducted only 1.22% of the farmers have farm sizes in excess of

8 ha.

58

59

The results of the pooled data showed that ninety-eight per cent of the farmers
surveyed own their land, and less than 2% lease the land that they cultivate. This is an
important factor because most of the farmers can use their land as collateral for obtaining
agricultural development loans for future investments.

The average land size of the pooled data was 4.69 ha (11.6 acres). When the data
were analyzed by t-tests with the independent samples routine of SPSS for the two
districts, the results showed that the null hypothesis that the mean difference of the land
size of the two samples (0.33 ha, P = 0.05) was the same could not be rejected because
the observed signiﬁcance level was more than 5% (34.8%).

The results also showed that the mean difference of zero for the two districts was
included in the 95% conﬁdence interval meaning that the mean land size value obtained
in the two districts (Table 5.1) was the same. This statistic conﬁrms that the average size
of the small-holder farms in the two districts surveyed may be considered the same.
Similarly the land currently used for tillage or land generally referred to as cropped land
for the two districts had a mean difference of 0.046 ha with a signiﬁcance level of 87.21%
(P>5%) and the value zero of the difference was within the 95% conﬁdence interval
showing that for the two population samples used, the size of land currently used for
tillage by farmers had the same mean size in the two districts. The size of the grazing
land was also found to have a signiﬁcant level above 5% (5.8%) with a mean difference
of 0.538 ha. Hence from the data obtained for the two districts, farmers’ land size, the
land under tillage and the land allocated to grazing had the same mean size in the two

districts surveyed.

60

Descriptive statistics obtained from further analysis of the pooled data showed that
about 56.3% of the farmers had between 4 and 6 ha (10 to 15 acres) and only 10% had
more than 6 ha (15 acres). About half of the land-holdings were cultivated for crop
production while the rest was reserved for grazing purposes (Table 5.2). The range of

the individual farm size holdings was from 1.21 ha (3 acres) to 8.00 ha (20 acres).

Table 5.1: The t-test results for independent samples (equality of means) of land
size for Machakos and Mbeere districts.

 

 

Variable Mean Difference d.f. 95% CI for Difference Standard Error 2-Tail Sig.
Land size 0.3324 78 -0.369, 1.034 0.352 0.348
Cropped land 0.046 78 -0.522, 0.614 0.285 0.872
Grazing land 0.5383 74 -0.019, 1.095 0.280 0.058

 

Table 5.2: Descriptive statistics of the small-holder farms in Machakos
and Mbeere districts

 

Descriptive Statistics (ha); n = 80

 

 

Variable Minimum Mean Maximum Std. Dev.
Land size ' 1.21 4.69 3.00 1.54
Cropped land 0.60 2.83 6.07 1.24
Grazing land 0.00 1.79 6.48 1.21

 

The distribution of the land holdings in the two districts shows that the most
frequently occurring land size is between 4.0 to 4.8 ha. (9.9 to 11.9 acres). The mean
land size is included in this range of most frequently occurring land size. The “stem and

leaf” plots of SPSS, which provides frequency distribution, shows that the land size has a

61

normal distribution with land size between 4.0 and 4.8 ha being the most frequent size of
small-holding with a count of 24 out of 80 farms. A modiﬁed print-out of the SPSS
output is shown in Table 5 .3 showing the distribution of the land size and the range in
each cluster of cases. The. stem column represents the integer value of land size while the
leaf digits represent the decimal (tenth) value of land size within the cluster. For
instance, the most frequently occurring land size has 24 cases which consist of eight 4.0
ha, two of 4.6 ha and fourteen of 4.8 ha cases. The tail ends of the histogram has two
cases of 1.2 ha. and 1.6 ha. on the lower end and one case of 8.1 ha on the upper end.
The stem width is a multiplier factor of the stern before adding on the leaf for each of the
cases'.

The cropped land had a similar distribution pattern showing also that the most frequent
size of cropped land is between 2.0 to 2.8 ha. which is about half the size of the land
holdings (Table 5.4).

5.1.2 Farm implements and hand tools

Small-holder farmers in semi-arid areas use hand tools or DAP or a combination of
both for their agricultural mechanization needs. The results of the survey show that the
variety of DAP implements available at the farm level was limited to plows, ox-carts and
cultivators. Although there were about 54% of the farmers that had both a plow and an

ox-cart, the low utilization of DAP after the tillage season is over indicates that ox-carts

are not widely used for transportation purposes. Multipurpose use of DAP is also limited

 

' In general to compute the value of the cases in a stem and leaf histogram, the stem is multiplied by the
stem width and the values added to each leaf one at a time. For a stem width of l, the leaf is a decimal
value (10m) while a stem width of 10 has a leaf of unit values.

62

Table 5.3: Stem-and-leaf plot of the farm size in hectares (pooled for Machakos and
Mbeere districts).

Farm size ha

 

Frequency Stem (ha) Leaf (decimal, 10'h ha)
2.00 I . 26
10.00 2 . 0044444888
12.00 3 . 022222226666
24.00 4 . 000000006688888888888888
14.00 5 . 02244666666666
10.00 6 . 0000000000
7.00 7 . 2222226
1.00 8 . 1

Stem width: 1.00
Each leaf: 1 case(s)
Source: Modified SPSS print-out

Table 5.4: Stem-and-leaf plot of the cropped land size in hectares (pooled for Machakos

and Mbeere districts).
Croppgd land,I ha
Frequency Stern (ha) Leaf (decimal, 10‘h ha)
6.00 0 678888
1 3 .00 1 0222226666666
24.00 2 . 000000444444444468888888
20.00 3 . 22222222222222566666
14.00 4 00000000044888
2.00 5 66

Stem width: 1.00
Each leaf: 1 case(s)
Source: Modified SPSS print-out.

because only 10% of the farmers owned a plow, an ox-cart and a cultivator while 33%
had a single plow only and about 3% had no ox-implement at all (Table 5.5). The
scarcity of cultivators among the farmers explains why more than 50% of the farmers use

a moldboard plow for inter-row weeding of pure stand (single crop) maize

63
plantations. The few farmers that had a cultivator either bought them second hand from

outside the district or arranged to import them directly from outside the country. The
local DAP market does not cater for the distribution of cultivation or weeding
implements. This is primarily because there are no local manufacturers of weeding
implements and no formal importation from outside either. Although farmers indicate
that weeding is a major concern, weeding implements are not in demand especially
because they are single purpose tools whose cost may not be justiﬁed for non-commercial
crop production. However, farmers owned a wide variety of hand tools including
machetes, digging forks, hoes, spades, manure forks, wheel barrows, and axes. The
number and type of hand tools available at the farm level depend on the family size and
the number of adult-equivalent labor in the household. The average farm labor force
available at the farm on a regular basis was three adults.
5.1.3 Draft oxen

The majority of the farmers (68.8%) had only one pair of trained oxen and 27.5% had-
two pairs. Some of the farmers with two pairs of trained oxen did not use the two pairs
simultaneously for tillage, especially when the land was plowed either soon after harvest
or when the rains set in. The main reason for using only one pair of oxen while owning
two pairs of trained ones was because the soils were soft enough for a single pair to
manage at the time the tillage was done. The two pairs of oxen were used alternately to
keep them in condition, otherwise, the second pair was on stand-by in case it was needed
to work in ‘fours’ when the ground conditions demanded higher draft force than one pair
could provide. The expected productive working life span of the working oxen was

reported by about 53% of the farmers to range between 7 and 10 years. Table 5.6

64
summarizes the data on the ownership and usage of draft oxen.

Table 5.5: Pooled distribution of ox implements in Machakos and

 

 

Mbeere districts.
Ox Implement Frequency Percent
None 3 3.8
Plow only 26 32.4
Plow and ox-cart , 43 53.8
Plow, Ox-cart and cultivator 8 10.0

 

Table 5.6: Ownership and usage of draft oxen in Machakos and Mbeere districts.

 

 

 

Response Number of oxen used for tillage
Ownership of Oxen Frequency °/o Frequency %
No oxen owned 3 3.7 3 3.7
One pair 55 68.8 58 . 72.5
Two pairs 22 27.5 19 23.8
5.1.4 Animal feed supply

The primary feed source for the draft oxen is natural grazing and natural hay
(harvested natural grass) either from the farmer’s own land or that of friends (or relatives)
which can be made available free of charge or from rented grazing land. Table 5.7 shows I
that 85% of the respondents indicated that they depend on grazing land as the main
source of feed. However, about 43% of the respondents indicated that they had sufﬁcient
feed supply for their livestock from their own grazing land. Renting grazing land from

neighbors that either have no livestock or have more grazing land than they need is a

65
common practice in the two districts surveyed. About 31% of the respondents depend on

renting grazing land for supplementary livestock feed whereas only 6% indicated that
renting was their main source of animal feed. Commercial feeds are not commonly used
due to lack of availability and their high cost when brought in from the high potential

districts. Less than 4% of the farmers use commercial feeds for their draft oxen.

Table 5.7: Livestock feed supply and sufficiency

 

 

Main Food Frequency % Feed Sufﬁciency Frequency "/o

No Livestock 3 3.75 No Livestock 3 3.75
Grazing Fields 68 85.00 Sufﬁcient 43 53.75
Renting 6 7.50 Rented 3 l 38.75
Commercial feed 3 3.75 Commercial feed 3 3.75

 

5.1.5 Farm labor and energy supply

Seventy-ﬁve percent of the farmers indicated that they did not have sufficient energy
for all their farming needs. When asked what area they would need extra energy for,
38.8% indicated that weeding required more energy than was available at their disposal.
They also indicated that labor was most scarce during the hand weeding and harvesting
periods of the crop production season. Twenty-ﬁve percent indicated that plowing
required more labor than they had available at the farm. Plowing was also considered by
about 95% of the respondents as the most difﬁcult, crucial and most labor demanding
activity at the farm. While labor bottlenecks during the peak weeding and harvesting

period represented the highest energy deﬁciency, food requirements for the draft oxen

66
was cited as the next most energy deﬁcient area. Weeding by hand was also considered

to be the most time demanding activity on the farm followed by harvesting when the crop
is good.
5.1.6 Using two pairs of oxen

There is a tendency of farmers to use more than one pair of oxen for plowing.
Respondents were asked the reason for using two or more pairs of oxen in tandem. Their
response indicated that in most cases the second pair was used just because it was
available and also that using it kept all the draft animals in good working condition.
While 73.8% of the respondents said that they used only one pair of oxen, 18.8%
indicated that they used the second pair because it was already trained and available.
Interestingly, only 2.5% used the second pair of oxen because a single pair of oxen was
inadequate. This implies that the use of a second pair of oxen in tandem is not
necessitated by draft requirements. Most of the farmers that used one pair of oxen
prepared their land after the rains came or soon after harvest when the ground was friable.
5.1. 7 Start of plowing

The timing of plowing for the next cropping season is always critical if the season is to
succeed. Table 5.8 shows that about 55% of the farmers plow soon after harvest, mainly
because they want to bury stubble. The other reasons for plowing soon after harvest are
that the soil conditions after the crop is harvested are friable (low draft requirements), the
animals are still in good physical condition, and there is sufﬁcient feed for them. This
practice was observed mainly with the farmers that receive organic farming extension

services from a non-government organization (N GO) in the two districts.

67

The next most important time of starting tillage work was reported to be at the onset of
the rains. The main reason why many farmers start plowing at the onset of the rains was
because the animals were not strong enough to work before the rains due to the prevailing
hard soil conditions at the land preparation time caused by feed shortage. Justiﬁcation
for this timing was the softened ground conditions because the ground conditions would

be too hard for the weakened oxen prior to the falling of the initial rain showers.

Table 5.8: Timing of plowing

 

 

-Plowing time ﬁequency Percent Rilsons
Soon after harvest 44 55 Bury Stubble
At ﬁrst rain showers 20 20 Soil softened
Just before the rains 16 25 Soils heavy when wet

 

5.1.8 DAP utilization

The amount of time that draft oxen are utilized per'year and the number of hours that
they work per day are measures of the level of utilization of the DAP. Over 70% of the
farmers reported that they used their draft oxen between 5 and 6 hours per day during the
peak plowing season. These results were consistent in both districts. The t-test results
showed that the null hypothesis that the mean difference of hours worked by oxen per
day is zero could not be rejected. The mean daily work hours in Machakos was 5.96
whereas that of Mbeere was 6.03. The mean difference of -0.07 hour was included in the
95% conﬁdence interval for difference and therefore the mean of the daily work hours in

the two sample sites can be considered to be the same.

68

The analysis of the annual utilization of draft oxen showed that the mean annual
utilization of DAP in Machakos was 45.67 days as compared to 40.65 days in Mbeere.
However, the mean difference value of 5.02 days was found to be not signiﬁcant at
P = 0.05. This difference was expected to occur about 23.4% of the time if the mean
annual working days of the two sampled populations were equal (Table 5.9). The pooled
stem and leaf procedure (Table 5.10) shows that most of the farmers (about 70%) use
their oxen for 50 days or less per year with a few on either of the tail ends of the
distribution (3 farmers more than 80 days and 5 farmers less than 20 days per year).

The utilization of oxen drops off after the planting period is over and only picks up
again at the beginning of the next tillage season. Generally draft oxen are utilized for less

than 50 days per agricultural year (October - September). It is notable that this rate of

Table 5.9: "The t-tests for independent samples of daily work days, ox-life span and
annual work days (equality of means) for Machakos and Mbeere districts.

 

 

Variable Mean Diff d.f. 95% CI for Diff Standard Error 2-Tail Sig.
Daily Work Hours -0.0739 77 -0.440, 0.292 0.184 0.688
Ox Life Span 0.3000 76 -0.956, 1.556 0.630 0.636
Annual Work Days 5.0215 77 -3.318, 13.361 4.188 0.234

 

DAP utilization is far short of the potential rate of utilization, especially if DAP can be
used in-between the tillage seasons. The range of the days animals work per year was
wide meaning that there are farmers that use their animals marginally while others
attempt to make the best use of the draft oxen. The mean utilization of draft oxen per

year was found to be 43.7 days. About 30% of the farmers used their oxen for more

69

Table 5.10: Stem-and-leaf plot of the annual DAP utilization in days (pooled for
Machakos and Mbeere districts).

DAP utilization. days
Frequency Stem (x10 days) Leaf (x1 day)

 

2.00 0 . 58

3.00 l . 024

13.00 2 . 0000224455589
13.00 3 . 0000455556668
12.00 4 . 000000055559
17.00 5 . 00000000000002255
11.00 6 . 00000005556

5.00 7 . 00005

2.00 8 . 00

1.00 9 . 0

Stem width: 10

Each leaf: 1 case(s)
Source: Modified SPSS print-out.

than 50 days per year as shown (Table 5.1 1). This low utilization rate is indicative of the
fact that draft oxen are used primarily for tillage operations. This premise was conﬁrmed
by more than 98% of the farmers who reported that their main activity with oxen was
plowing the land.
5.1. 9 Draft oxen life-span

One of the factors that determines the life span of draft oxen is the age at which the
animal are trained to work. This varies from 2 to 4 years. According to the responses the
null hypothesis that the two groups of farmers interviewed had the same mean working
life-span of oxen could not be rejected. While the mean reported oxen working life-span

for Machakos was 8.33 years. that of Mbeere was 8.03 years. The mean difference

70

Table 5.11: DAP annual utilization

 

 

Annual DAP Usage Frequency Percent
Less than 50 days 43 53.75
50 days 13 26.25
More than 50 days 24 30.00

 

of 0.30 was expected to occur 63.6% (P>0.05) of the time if the two population means
were equal. Most of the farmers reported that they use the animals for up to 10 years
before they replace them. It is commonly documented in the literature that the longevity
of working draft oxen is an average of 5 to 6 years as the reasonable working life-span.
However, it is not unusual to find farmers that have used the same oxen pair for 15 years
or more. The data obtained from the ﬁeld had six farmers who had used their current
draft oxen for 15 or more years.

5.1.10 DAP renting from others or hiring out.

Renting out draft oxen or hiring draft oxen from other farmers for income generation
was not a common practice. When farmers borrowed draft oxen from neighbors, friends
or relatives. it was mainly to return a favor and not for income generation. Less than
29% of the farmers (Table 5.12) rented out their draft oxen to other farmers for income
generation. These were primarily the farmers with more than one pair of oxen. The main
limitation against hiring or renting practice is the timeliness of operation. Over 78% of
the respondents indicated that they do not hire other farmers’ oxen. When they are late
for land preparation (and they always try not to be late), they borrow draft oxen from

friends or relatives. Only 6% of the respondents indicated that they rented draft oxen

71

from other farmers because their own oxen could not complete tillage in time for timely
planting (low capacity).

Under the present level of management of draft oxen, most farmers do not seem to
have the extra capacity that could be used for hiring out to other farmers due to the short
duration of land preparation time. Improving the level of management of DAP can make
available extra hours which can be used to earn extra income for the farmers and
facilitate the promotion of mechanization for the farmers without draft oxen.

The number of years that the farmers have used DAP in the two districts had a mean
difference of 11.9 (Machakos 28.3 and Mbeere 16.5). The 2-tail signiﬁcance level for the
number of years farmers have used oxen was small (P<0.0005) and the 95% conﬁdence
interval was 6.813 to 16.920. This result indicated that the farmers in the two districts
have a mean DAP usage period that is signiﬁcantly different. This ﬁnding was
reasonable because the farmers in Machakos district have used DAP for a longer period

than most of the other districts in the country.

Table 5.12: Hiring and renting of draft oxen in Machakos and Mbeere districts.

 

 

Hiring from others . Frequency % Renting to others Frequency °/o

No renting 63 78.80 No renting 32 40.0
Low capacity of own oxen 6 7.56 Return favor 25 31.3
Miscellaneous Reasons 1 l 13.75 Earn Income 23 28.8

 

5.2 Energy utilization data analysis
Data collection with the electronic instrumentation (DAS) was performed through

experiments that were done at the University of Nairobi farm and in the farmers’ ﬁelds in

72
Machakos and Mbeere (former Embu) districts. The DAP-SYS energy utilization data

obtained ﬁ'om these ﬁeld tests provided data for quantifying linkages and parameter
estimates for the EUMDAP computer simulation model.

In general, all DAS experiments performed at the University of Nairobi and in the
farmers’ ﬁelds were performed using the same equipment and procedure. Only the draft
animals, the time of the experiment and the soil type and conditions varied when the
experiments were performed in the two districts with the farmers’ animals at their farms.
The same Operator (an employee of the researcher) handled the implement (Victory plow)
throughout all the experiments in order to maintain consistency in the control of the speed
of operation, depth of tillage and the width covered.

The length and the number of “runs” in each of the ﬁelds depended on the size and
conﬁguration of the ﬁeld. The length of the “runs” ranged between 14 m and about 57 m.
A minimum of three “runs” was done in each of the farmers’ ﬁelds. For each farmer’s
ﬁeld a data ﬁle for saving the data was created and it contained all the individual “runs”
performed at the site. A different record name identiﬁed the name of each “run”. At the
end of each “run” the data record was saved in the same ﬁle that was created for the ﬁeld
site. Initially the data were saved in the data logger’s RAM until the memory was
exhausted. The RAM had sufﬁcient memory to store the data from ﬁve or more runs
depending on the length of each run. The data were then dumped into a ﬂoppy diskette
using an RS232 data transfer cable, and the PROLEC-TRA software. PROLEC-TRA
software is the processing program for the RAMLOG E19000 data logger allowing one to
read and to process the measurements taken by the data logger. The software is a user-

friendly menu driven program.

73

5.2.1 Preliminary data analysis procedure

The number of “records” contained in a ﬁle depended on the length and time of each
of the ﬁeld tests conducted. The naming of the ﬁles and the “records” was such that the
researcher could identify from the ﬁle name the location and the ﬁeld where the
experiments were performed.

The preliminary data analysis software (PROLEC-TRA) was saved in the computer
and used to review and inspect all the data ﬁles in graphical format. Tabular data was
later retrieved in EXCEL for subsequent analysis and computation of the output values in
physical units.

The process of data analysis using the data logger software started with a display on
the computer screen of all records for the ﬁle selected to be analyzed. When a record was
selected for analysis, a graphic presentation of the entire time span of the “run” was
displayed. The graphic display showed all details of the displayed channels including:
date and time of the experiments; name of the ﬁle and “record” name; the units of each
sensor values; and a function-key menu for other options of viewing the data.

One of the options included the choice of the channels to be displayed, printed or
exported; up to a maximum of 8 analog, digital or a combination of both. Viewing of the
average, maximum and minimum values was provided in an overlay window in the
graphics display. Zooming-in on the graphics over a selected “window” of the display
was also an option that conveniently enabled sampling of short spans of data to be viewed
in a magniﬁed form. This was especially useful for the three physiological channels with

both analog and digital data records (heart rate, respiration rate and stepping rate).

74
Channel 20 which had 8 digital inputs was used for the digital data monitoring and

recording.

Various data output options were available for saving or printing the preliminary data:
print-out (graphics, Tables or Math results); ASCII ﬁle saving (Text, Lotus 1-2-3,
Dbase); and saving Graph ﬁles (.PCX) through the PROLEC-TRA program. However,
only the 8 displayed channels were saved in any of the formats chosen which included the
main channels needed for the analysis. The ninth channel was used for the battery power
monitor, and was omitted in the subsequent ﬁles. The data were saved in ASCII and also
in graphics mode for subsequent analysis in EXCEL and for export into word processing
programs. A print-out of one of the raw ﬁeld data ﬁles is shown in Appendix H which
was imported into EXCEL from the PROLEC-TRA software. The print-out shows nine
channels including the battery monitor channel.

5.2.2 Experimental results

The experiments conducted at the University of Nairobi were performed with one pair
of oxen. The purpose of the experiments was to provide preliminary estimates for the
relationships of the relative heart rate (RHR) versus the draft generated by the animals at
work. This was done through varying the depth and width of the tillage implement.
Three depths (5, 10 and 15 cm) were used at two widths of 20 cm and 30 cm. A total of
eighteen test “runs” were performed in three blocks in a randomized block design format.
Each block was done and completed in one day. The data sampling and recording rate
used for all channels was 16.67 Hz (0.06 sec intervals). At this sampling rate the time
taken for the 594 data points recorded for experiment number “5NRR1 ” was about one-

half minute (35 seconds). The ﬁle names represent the depth of tillage (5 cm), narrow

75
width (NR) and replication one (R1). The descriptive statistical results of the speed, draft

force, depth and power of the university of Nairobi experiments are shown in Table 5.13
grouped according to the blocks of the experimental design. The results of the depth data
shown in the table shows a variation of the recorded depth from the depth that was
targeted. Although the implement operator attempted to maintain consistency of the
depth of plowing, this discrepancy was consistent in all experiments. The depth sensor
was properly calibrated and therefore the error was possibly due to the operator not
maintaining the plow at a consistent depth.

The data were obtained in “snap-shots” of the test runs which lasted between about 20
seconds and two minutes - when the draﬁ animals were in a steady working condition.
Data collection was triggered by pressing the “F1” key, while pressing the “F2” key
marked the completion of the data acquisition process. This ability to control the process
of data acquisition was very important, especially when there was need to interrupt the
process when the draft animals either stopped or insMents needed attention.

One of the outputs of the PROLEC-TRA data analysis software for an experiment at
the university of Nairobi (Figure 5.1) shows one “run” that lasted for about 20 seconds.
Two channels (6 and 20) out of the eight possible were chosen for this graphical display.
The analog heart rate output (channel 6) is shown in the upper part of the graph in the
form of a trace which is also represented as voltage on the left hand side vertical axis.
The axis was, however, labeled “P” to represent the pulse rate of the heart. The label on
the right hand side vertical axis represented the raw values of the step rate sensor before

conversion to actual step rate in steps per minute.

76

 

 

 

an on... mu... S... v. .m... .9. ....N ohm. mm... 2.... av... we. v.5. one 95% .
an e. m. mm... ....c. «... «.... mm. ..m. an... M... 3.... a... 3.. ....m ..3 $5.22
mm a3... who mod. n. mm... 3... .m... ..m... cm... 2.... cm... mm. com ....m 9.3...
Nv v.0. 3.6 :6. m. ...... cw... cm... ems cm... 3... 5.... mm. v. . m ..mn 3.2...
Ne w. m. Rd ....0. m. mm... a... 3... ...m.m on... ma... 2... on. ...... one gm
.... Em. mm... ....e. m. cm... ...... .N. .....v ...... ca... m... N... «.mm ....m 9.2m
On Q. m. mad :6. m. N... 3.. and ..N... .N... am... .v... 3.. m. . m can gm.
om o3... 3.6 «On. N. 3.... mm. mm... N . .a n . ... mu... cm... 3.. Wow coo 9.12m .
cm a3... 3.6 3.2 N... 3... 3.... mm... ..m... mm... 3... mm... am. m. . m new 3.3...
on v. m. mmd ....o. m. 2... o... ..m... mm.@ a. ... am... No... On. 5.0m ....m 55.2...
..m v. m. 2.6 2.... m. mm... 3... ...... on... ..m... 5... 2... N... man 3:. 9.3m
..m Em. mud 2.... m. R... o... 3.. end a... .3... m... a... m.~m ..nm 32m
..m V. m. 2.6 2.... 2 mm... 3.. am... «No a. ... Na... 3... 3.. Yum 8.. 3.32
mm v. m. 2.6 ...... m. om... m... .m... mm... ...... an... ...... mm. new ...n 75.22
.... m3... 3.0 8.2 m. Va... 8... ...... 3.6 cm... R... N... a... Yam mg 5.3...
.... a3... 2.6 3.... m. R... a... ...... nmé mm... ...... N. .. em. .....m on... 15.2...
mm $9. «...... 3.». v. ..N... 2.. mm... ..m.m o. ... ca... ...... . ... man ...... 1.5%
on 3.... mad ....c. m. cm... v... mm. and ...... ma... 3... .~.. Qmm vow .....Zm
32$. kc. mm 63.x} R 6§§ 85. $9.. ..qu so: :82 $0.. 58.). Son. 5.02 «E. Z .85.).
85— .83 a ...... ...—2.x”.— .eeaua a: .3. ....m ....m ....m ....m ......— 4.2.x”.—
..8m use”... 3.33 ...ocm ...—:3 :2... 39.0 :32 «3M. .83...— «Eu. 5%.. 33. .8on 9.0.. as...

 

8:08.898 £832 .0 36.03:: 8.. 3.32 maggot“... .3 588.5 um . .m 0.2....

77

 

 

 

 

 

 

 

18
P . , 246
5
0 WA 244
-5 l ' ’ ' 242
—1a 240
sec
BEGIN DATE nun DATE N31 119 1 sea

 

 

05-83-96 85-83-96 6 : HR
15:36:30 15:36:49

 

 

 

 

 

 

Figure 5.1: PROLEC-TRA graphical output for a Nairobi university experiment
showing heart rate trace and three digital channels.

The heart rate, respiration rate and the stepping rate are represented in digital form
(0,1) in the lower half of the graph as l, 2, and 3 respectively. The analog graph of the
heart rate represents the output of the infra-red absorption which registered the high and
low (1, 0) peaks of the blood flow through the ear. It is worth noting that the heart rate
trace in the upper half of the graph does not appear like a typical eletro-cardiograrn
(ECG) trace which clearly deﬁnes the three components of the cardiac cycle (P wave,
QRS complex and T wave). The heart rate sensor picks the systolic or the ventricular
depolarization phase (QRS complex) of the heart beat which is shown as the spike (peak)
of the trace. The equivalent phase of QRS complex in the digital mode is represented by
the number “1” of the binary digits. The raw data of the step rate is represented in the

graph as pulses per minute on the right hand side of the vertical axis.

 

78

A graphic print-out showing two traces of the horizontal draft and depth of tillage
are shown in Figure 5.2. The traces lasted for lasted 52 seconds recording a total of

866 data points. The left hand vertical axis represents the draﬁ force in volts which

 

 

 

 

 

O
8 .
3-
l8

 

SEC
BEGIN BRIE END DRTE KBTEXPT liﬂURBl I 560
10-05-96 10-85-96 1 :H08
12:53:47 12:54:39 5 IDEPIH

 

 

 

 

 

 

Figure 5.2: Graphic output for ﬁle number “10WRR1” showing the horizontal draft
force and depth sensor traces.

were converted to force (N) with the sensor’s transfer function. The right hand
' vertical axis represented the depth of tillage in analogous voltage readings which
were computed into the actual depth in cm using the sensor’s transfer function. The
transfer functions for each mechanical sensors are tabulated in Table 3.1.

Two phases of the draft force are clearly seen in the graph. Both the depth of
tillage and the draft force remained steady. The draft force generated by the
implement was a factor of the depth to which the implement penetrated. In one

section of the trace, the soil resistance was rapidly ﬂuctuating as shown by the rapid

 

79

changes of draft force and depth of tillage. The second section of the ground shows a
more uniform and consistent soil resistance. Both sections of the graph show a more
uniform and consistent soil resistance. In both sections of the graph, the depth of
tillage and the draft force followed the same pattern in which the deeper tillage
resulted in higher draft force.
5.2.2.1 University of Nairobi experiments

The summarized data obtained from the controlled experiments at the university
of Nairobi were analyzed with the ANOVA procedures to obtain statistical validity.
The observed difference between the F values of the blocks for the variables are as
shown in Table 5.14. The mechanical variables measured with the DAS which
included the draft force, depth of tillage, ground speed and the computed power were
all found not to be signiﬁcantly different. The respective probability of obtaining the
F values shown for each of the variables if the means of the sampled populations
were the same was far less than 5% value for which the ANOVA tests were
performed. The largest F value was for the energy variable which was 0.0191.
Therefore, the null hypothesis that the speed, draft force and power variables are the
same were rejected at P<0.05. The P values obtained from SPSS for each of the
variables indicate the probability of obtaining the computed F value if the null
hypothesis of equal means is true. For example the probability of obtaining an F

value of 20.464 for the draft force is 1.70263E-5 if the null hypothesis is true. The

 

80
Table 5.14: AN OVA results of the university of Nairobi experiments

 

 

Variables df. df2 F P value
Draft Force 5 12 20.464" 1.70263E-5
Ground Speed 5 12 16.783" 4.74E-05
Power 5 12 4.2218" 0.019107
Energy expenditure 5 12 3.450" 0.036581

 

” signiﬁcant at P = 0.05 [The critical value for df. = 2, df, = 6 and P (0.05)= 3.11]

results show that the variation of the draft force, ground speed, the energy expenditure
and the power developed by the animals can be attributed to the varying depth and width
of the implement. Hence the physiological response that the draft oxen were subjected to
when working can be attributed to the draft force required by the implement at work in
the prevailing soil conditions and setting of the tillage implement.

Likewise the energy expenditure of the draft oxen versus the power demand from the
equipment that was monitored and recorded by the DAS shows that there is a relationship
between the power output and the draft force generated. The AN OVA results show a
statistically signiﬁcant outcome from which the null hypothesis that the means of the
energy expenditure of the animals subjected to different energy demand levels have the
same values is rejected.

5. 2. 2.2 Farmers ’ ﬁeld experiments

Out of the eighty farmers that were surveyed for the socio-economic data, 15 of them

had their draft animals tested for the power output and energy consumption at their farms.

The procedure of ﬁeld tests with the farmers’ draft animals was the same as that of the

81

University of Nairobi experiments. For each of the animal teams, six test runs of tillage
were performed with the instrumented plow. Three depths (10, 15 and 20 cm) and two
widths of tillage were used for each farm. Estimates of the width of tillage were done by
the operator of the implement and an accurate measurement of the width done after the
implement had passed. Three measurements of width were taken for each “run”. The
1 position of the preceding furrow was marked and used for measuring the actual width of
tillage for the preceding “run”.

A summary of the mean and standard deviation of four mechanical variables and
corresponding values of the draft animal energy expenditure data is provided in Table
5.15. The energy expenditure calculations were performed with equation (3) of the
model developed by Richards and Lawrence (1984) and which was discussed in section
2.5. According to the model, the power demand of oxen equivalent to relative heart rate
of 1 (at rest) is 6.94 W/Wt0‘75. Hence the computation of the energy expenditure for
work was done without the resting metabolism equivalent. The data were further
analyzed with the one-way ANOVA procedures for each division. The test results
analyzed were carried out in ﬁve administrative divisions three in Machakos and two in
Mbeere. Data from each division consisted of three farm sites where the farmer’s draft
oxen were tested with the DAS instrumentation. Each division was considered as an
experimental unit with three pairs of different teams of draft oxen for testing. From
Table 5.16, the data for the three variables from Kalarna division, the null hypothesis of
equality of the means for draft force, depth of tillage and power was rejected. This means
that the three variables in each of the test sites were signiﬁcantly different due to other

factors and not by chance. Hence. the different sites with different soil types have

82

a... .28. m ..m 0.0....

 

 

: 2.... 3.0 ”0.2 N. . o . ... v0... 2 .N NN.2 N . ... mm... 0.... no... néN ..wv MM.
.. an... 3.0 ”0.2 N. . .N... 00... 2 .m 3.2 2... no... 2... «0.. ..vM ..wm NM.
.. an... 3.0 ”0.2 N. . .... NV... 2.N MN. . .N... 2.... .... an... N...N nN0 m Eh... .M.
on NN... 3.0 «0.2 N. . w . ... ..v... N . .m 3.2 N. ... 0w... N. ... N0... o.Nm ..NN NM.
nm N..... 3.0 «0.2 N.. o... 3... ..mN ”...... mm... N0... ...... ...... N.mN ..N0 NM.
0m No... 3.0 ”0.2 N.. v. ... me... v...” .N.... v. ... an... N... 3... ... . N ...v N 8.8... .M.
3 ...... NNo .. . .0. m. .N... 3... ..vN .0.2 w . ... 0.... .. . ... ov... «.NN ..0m MM.
2. ...... No ....0. m. 2... 0m... ...N woN. N... .M.... .... mm... 0..N ..mm NM.
Nv ...... No ....0. m. N... mm... 0....~ .....0. o.... m... 2... Nm... ..NN ..nm . E... .M.
co.m.>.o «.953.
.v 00... 3.0 w0.2 N.. o... .m... ..m.m ..o.2 ..N... ...... NN... 2.... ...m ..2. MM.
um 00... 3.0 ..0.2 N.. N... .m... .....m .M.... .N... ...»... .N... ..w... voN ..mw NM.
mm 00... 3.0 ”0.2 N.. m . ... m0... we... 0m.2 N... no... N... ..0... .N ....0 m Em... .M.
-- N... No 2.0. n. v... Nm... .o.m 2.2 .N... mm... o... .0... N.NN ....0 MM.
: N... No ....0. m. 2... mm... 3...” NN.... ...... cm... 2... m0... 0.2 ....m NM.
-- N... No .....0. m. o.... .0... omen. oN.... ..N... on... .N... N»... ”.2 ..nm N :5... .M.
mm v»... 3.0 «0.2 N. . .... an... omN m . .v. ..N... ...... w . ... 3.... ....N ..N0 .M.
0m «a... 3.0 «0.2 N.. N... ...... .m.m ...... ...... 2.... 0.... .m... ...V. ..3. NM.
.v v»... 3.0 ..0.2 N.. N... on... ..oN 0N... NN... no... .N... .0... o.Nm o2. . .Em... .M.
:o.m.>.D .EEoU
.... mm... 3.0 w0.2 N.. NN... w... 0.... wow 0.... .o... ...... oN.. «.NN 0.9 MM.
om an... 3.0 ”0.2 N. 0.»... No... and .. . .m 3.... ”0... .... on. 0.... 2... NM.
..m ...... 3.0 ”0.2 N.. N..... N... Ned 2.6 2... .N... .... w... ...NN 00.. m E... .M.
0m 0.... 3.0 ..0.2 N. . NN... v0... o0.N omo .N... o... ...... ..m... 0.... ...... .M.
...N 0.... 2.0 .0.... N. _ N... .N... ..NN N... N... mo... oo... ms... ....N ..0m NM.
Nv 0.... 3.0 ..0.2 N.. . .... E... 3N woo ...... ..o... o..... 05... ”.NN ..om N SE... .M.
-- ..o... N... .. . .0. m. m . ... vs... 2N 00.... v. ... oo... we... ...... ..NN ....m MM.
: ..o... No .. . .0. .M.. .... N0... ovN Nmo o. ... cw... o..... 2.... ...: ..v. . NM.
-- o..o... No ....0. m. o..... 3.... m... o... ...... .o... ...... .w... ..mm ....o . EB... .M.
:o.m.>.n. 9:23.
"mas—«.1052
58>. E. n- RES. n. 50.3.. 8.3. $0.. 53.2 $9.. :52 $9.. ecu—,— 5... cue—2 .E. g .532...
25. .92.... .83 .85 .= ... ....m ....m ....m ....m an... acacia...
:05 «.83 9.3.32.3. 3.5:... 5.92 «a... .33.... «E... .....o: 35. .325 2... ......

 

 

 

$8.5... 0.82.2 ...... 8.5.85. :. 0.2.. £25... 2.. :o 35:...35 5.. 3.28. egg—8.39... ....mEEzm ”m . .w 22......

83

 

 

 

2. h... MNd ....0. M.. MN... 0... MM... MM.» ...... Mo... ...... N... 06M .60 MM.

.... .. . .. MNd 2.0. M.. MM... Mo... NN.N MN.0 ..M... 0w... .... N... ....M 0... NM.

MM .... MN... 2.0. M.. .N... 00... «.N.. N... NV... M... N... o... ..MN ..M.. M FEW... .M.

M... N... MN.@ ....0. M.. 5N... 0.... ..M.N MM... MN... ..M... N... .N.. ...M M2. MM.

3. N... MN.o ....0. M.. ..N... M... ..MN 3.0 M... Ma... .... .N.. M...M ..N... NM.

2. N... MN... ....0. M.. NM... Mo... ..MN vM.v ..M... 0.... 0.... 0N. ...M. ..wo N SE... .M.

.... 2... MNd ....0. M.. ..M... ...... 0... M0.. 0N... .M.. ...... NN.. ﬁNv Maw MM.

...V ...... MNd 2.0. M.. ..M... Ma... M...M ......V. MN... MM... ...... M...— wéM ..M... NM.

Nv ...... MN... 2.0. M.. 0M... .... NEM ..MM. NN... M0... a... M0. 0.3. ....N. . E2... .M.

2073... 8.9.80

.. ...... MN... ....0. M.. NN... ...... ..M.N .M..N ..M... MM... 0.... 0N.. «xv. .2. MM.

.. ...... MN... 2.0. M.. MN... Mm... M..M .50 .N.. M0... ...... MN.. ...MN N.0 NM.

-- ...... MN... 2.0. M.. MM... .... Me... M...0 MN... 3... NN.. on... ...M 30 M E3... .M.

-- ..N. Mn. .. 3.x. .... MM... ...... o0.N ...... MN... 00... . M... NM. . 0..N MoM MM.

-- ..N. Mb... 50.x. v. ..N... 0.... MwN .>.M. ..M... ..M. M . ... 00... ..N ....N NM.

-- «.N.. 2.... 50.2 .... ..N... .M.. ....V NN.M. ..M... 0.... .N... ..0... N.0N ...M N ES... .M.

-- Mo... MN... ....0. M.. ..N... .M... ..MM .0.M. ..M... ..M.. w... ..M... vNM ....0 MM.

-- Mo... MN..9 h . .0. M. . MN... ..M... wMN .0.... ..N... 0.... w . ... NM... de ....M NM.

-- Mo... MNd ....0. M.. ..N... ..N. .0.M 0..» .N... ...... M... 0.... ..MN hNM . ES. .M.

53>... owmxaa

mam—mm:

2.5}. K.... a. ......i. .. eexx... 93M. $9.. 5.95. $9.. 59:. $9.. .392 $9.. .392 «E. = .533...

9.3. 3.2.”. is: :20 .= .z ....m ....m ....m ....w .3: 3:95:25
.95 ...—93 9.3.0.3.... .99.... 5.9.). «a... .939. .39. 5.9.. 35. 099cm .2... CF...

 

 

 

84
varying soil resistance which caused the draft force as well as the power required to

operate the tillage implement to be different. The computed F value was more than the

critical value for all the three variables.

Table 5.16: ANOVA results of farmers’ ﬁeld experiments in Machakos and Mbeere

 

districts.
Draft Force Tillage Depth _ Power
df,/ df./ df,/
Division (If2 F P value dfj F P value df2 F P value

 

Kalama 2/6 206.725" 2.93 E-06 2/6 37.470" 0.00040 2/6 16.146" 0.00384
Central 2/6 0.3069 0.74661 2/6 2.2494 0.18664 2/6 0.0729 0.93042
Mwala 2/6 5.5815" 0.04272 2/6 4.2465 0.07095 2/6 0.6834 0.54024
Gachoka 2/6 4.871 1 0.05536 2/6 25.242 0.00119 2/6 0.5799 0.58847
Mbeere 2/6 1.1435 0.37953 2/6 9.0240 0.01553 2/6 0.2067 0.81881

 

.. signiﬁcant at P = 0.05 [The critical value for df. = 2, df2 = 6 and P (0.05)= 5.14.]

5.2.3 Use of ﬁeld data for modeling

The results of the data collected and analyzed were used as a comparing for the output
of the simulation model. The EUMDAP simulation model was designed to make best
estimates of the utilization of the DAP-SYSs in semi-arid lands of Kenya. The results of
the data was used as guideline of the boundaries in which the DAP-SYSs operate. For
instance the results of the data showed that the land holding size ranged between 1.21 and
8.0 ha per household. The simulations performed with the model were constrained within
these land holding limits in order to represent the real world situation. The data also
showed that about half of the land holdings are under crop and the rest is under natural
forage which is harvested for feeding animals during the drier times of the year. This
information was useful in determining the effect of expanding the cropped land beyond or

reducing it below the average cropped land.

85
The data showed that the majority of farmers (about 70%) owned and used only one

pair of oxen. The use of two animals for tillage only was therefore prevalent in the two
administrative districts. The implication of this finding was that any improvement of the
utilization of draft animals would be based mainly on the extensive use of the available
oxen and not on acquisition of multiple teams. Since only less than 30% of the farmers
train and use two pairs of oxen, simulations with more than one pair of oxen would be
limited to farmers with larger farm sizes especially because of feed requirements.

The data on the availability and use of feed for the draft oxen showed that most of the
farmers (85%) depend on the feed available from their own farms. One of the objectives
of the research was to investigate ways of assisting the farmers to improve their ability to
prepare their land in time for timely planting using their current resource endowments.
Feed as an energy source is an important factor in the management of draft animals for
work. There is little use of commercial feed for livestock in the research areas.
Indigenous feeds (including maize stubble) were used for projecting the energy needs of
the draﬁ oxen for the simulation modeling using EUMDAP.

According to the data, the annual utilization of the draft animals during the cropping
season was limited to only the short period of land preparation for planting and partly for
inter-row weeding of single crop plantations of maize. Simulation of DAP-SYS
utilization was done with the intention of utilizing the draﬁ animals for longer periods
than the current average of about 45 days per year. Draft animal power systems are not
considered a source of generating direct income in semi-arid areas. Using draﬁ animals
between households is culturally acceptable on the basis of friendship and return of

favors. Since over 70% of the farmers indicated that they do not rent other farmers’ oxen,

86
using “renting” in the EUMDAP simulation was an indicator of the extra capacity that the

farmers may have and which may be used for expansion of their own cropped land if so
desired.

The energy expenditure of the actual draft working animals in the farmers’ ﬁelds was
intended to be used for the validation of the simulation outputs obtained from the
EUMDAP simulation modeling. Although the ﬁeld data and the output of simulations
may not be exactly the same, the range and variability that the ﬁeld data provided for
various soil types and conditions was important in comparing the simulation outputs with

real world cases.

CHAPTER VI
COMPUTER SIMULATION MODEL DEVELOPMENT

6.1 Systems approach and concepts

Application of systems science to solve problems of physical, biological or social
nature continues to gain momentum in various disciplines. A “system approach” to the
analysis of problems comprised of various components is used for analyzing systems
with the aid of computers. Computer simulation models of a system are designed,
implemented and used for analysis of a problem. Analysis is deﬁned here as the
determination of model outputs for a given set of inputs and for a given model structure
which approximates the structure of the real world system being studied. Problems
requiring analysis may arise from existing and operational systems or from entirely new
concepts that address a new challenge arising from a felt need. Needs that arise from an
existing system present a management decision-making problem. This type of systems
approach is also referred to as management or control problems. In addressing a new and
previously non-existent need requires that a new system is designed from scratch and
implemented. In the latter case, a set of given system inputs and system outputs are
available or perceived and a system structure that would predict values for the system
desired outputs has to be designed and implemented.

A system is deﬁned as a hierarchical structure with deﬁned boundaries within which a
combination of inter-related components are integrated to perform speciﬁc functions in
order to achieve speciﬁc objectives (Ogata, 1978). There are two main types of systems:

static and dynamic. Static systems operate in steady or equilibrium states with outputs

87

88

being independent of the preceding values of variables. Dynamic systems on the other
hand have memory. The system outputs change with time as system variables interact
with each other. All systems have inputs and outputs and there are two types of inputs.
Exogenous or environmental inputs are factors that are external to the system. Although
this type of input may affect the system, they are not controllable by the analyst. The
second type of inputs are called endogenous or controllable inputs. These are an integral
part of the system and can be altered to determine the system behavior which in turn
determines the outputs obtained.

The objective of a system analyst is to optimize the output of a system. Outputs are
either used as inputs for subsequent system components or they can be used as
performance measures of the system being studied. Since outputs can be both desirable
and undesirable, optimization of a system requires minimizing the undesirable outputs
while maximizing the desired outputs. This is achieved through varying or adjusting the
levels and interactions of the endogenous input variables.

An example of an undesirable output is atmospheric pollution resulting from industrial
production. In the farming environment, undesirable outputs may include chemical
leaching into ground water or a disproportionate increase of labor requirement for hand
weeding and hence the cost of crop production when draﬁ animal power is used for
tillage to expand the acreage of tillage.

The computer simulation model designed and developed by the researcher is derived
from an existing real system of draﬁ animal power (DAP-SYS). The DAP-SYS can be
identiﬁed in terms of design parameters, inputs, outputs, and performance measures as

shown in Figure 6.1.

89

839.0 5080.390. 89.9.9 939.. .98.:< can. ”..0 99:3“.

5.308.. .89 09829:.

 

 

 

 

 

 

992.: :95:
509.09.. b9w0ED
..9:9.9_....9 93.9909...
2:2. :5 w:..:9: m>md<ﬂ 9:..9mw._..wh%90.www.
9:89:. 9.5.... .....9. ..o 9.95 3.6.. :. w:..:9._ m>m-x.<n.
99:95.59:— m>m-m<.. :23... 93.9.90 ...—93 26 w:.:.« . 85.9.5239 8.89:0
3:989:39. 099.. 08929:. 59.999.09.59... m9._9< . ._ 9.9 ”5:.th 99.. ..om
.509. 3.0993 0:9: 08929:. mmdDm<mE > ...9.:9mo ...9 99M 9:... 93......
93.50 99:99:22 99252909999 3.... 95529.: 92529 9985.
9995239.». 29.99: 25 22:295.
.5525 099 h.
.... 9a.. 99
9mg... ...9E.... - Fl .99“ 3%..“
9 958m. :59 .39.: , _ $5-55. :28 .8 new
9 3.8 92. 20 83205 i _ 53% .26.. 32:2 989 9597: 9.9390528
893:9: .2: 99822: _ .
mHDnFDO Guam”... ) -
9.9.. w:.._.: ..<..

:3. 93..

..9..o.. 5099.52.99:

:99. :o .9298.

99.... 099 h.

99.... :39...

95959.95. ..9 99....

m99.>..9m b9:.:9.9>
2.33.2. mDOZmOOxm.

 

 

\I

...ZmEZOM..>Zm

 

90

The design parameters of a DAP-SYS are considered ﬁxed and serve to deﬁne the
structure of the system. They are the basis by which a particular DAP-SYS is identiﬁed
in terms of the geographic location, acreage of the land, size of the implement, size of
oxen, feed quality, and the harnessing technology used among others. The design
parameters, of a particular DAP-SYS determine the ability of the system to efﬁciently
produce the need-fulﬁlling desired outputs.

Examples of the controllable input variables that ﬂow into the DAP-SYS include the
training level of the system operator and animals, feed quality and quantity for the draft
oxen, acreage under crops, DAP-SYS renting time in hours, and system utilization in
hours per day. The DAP-SYS also is subjected to other inputs that are outside the control
of the system. These are the exogenous inputs emanating from the “environment”. Some
of these uncontrollable inputs include the weather, mechanization policy, draﬁ oxen
hiring rate, and the price of such inputs as feed and implements.

The objective of the DAP-SYS is to produce desired outputs while minimizing the
undesired outputs. One of the main desired outputs is an increase in the annual utilization
of the DAP-SYS in days per year. Increase in the cropped acreage, higher net farm
income and timely tillage operations are other desired outputs expected from
optimization of the DAP-SYS. Muchiri and Minto (1977) documented that timely tillage
and hence subsequent timely planting enhances weed control as the crops get a head start.
The crop also takes full advantage of the initial rainfall and therefore the probability of
obtaining a crop and an improved yield is increased.

In the process of meeting the system objectives, some undesired outputs such as

increased need for feed quality and or feed quantity, higher cost of DAP-SYS

9]

maintenance and increased demand for hand weeding are inevitable. These will tend to
reduce the net beneﬁts of the system and should, therefore, be minimized.

The measurement of the desired outputs is determined quantitatively by performance
measures that include extra income generated, labor use efﬁciency, work days per year,
acreage plowed per season, increased crop production (total cropped area and total yield),
and the number of hours the DAP-SYS is rented out.

According to Manetsch and Park (1993), a systems approach to problem solving is a
technique that is needs driven. The design of a system begins with identiﬁcation of a
need or needs from which a problem deﬁnition is formulated and subsequently a system
structure is created. The boundaries of a system determine which components are
included in the system. Once the boundaries are identiﬁed, other components outside the
system are considered to be the environment with which the system interacts.

The DAP-SYS can be considered as a black box which receives both exogenous and
endogenous inputs and which provides outputs, both desirable and undesirable. The
simulation model developed is expected to behave as closely as possible to the real world
DAP-SYS. When the system is operational, the level of outputs is determined through
the system performance measures which are used to evaluate the performance of the
system. When the system inputs of an operational simulation model are varied, the
performance measures are used to determine and evaluate the change of outputs in
response to changes of input type or level.

The design of the computer simulation model of the DAP-SYS involved the
“dissecting” of the system to its bare constituent parts in order to build a “wholistic”

simulation model consisting of the aggregate parts or components with linkages which

92

represent directional ﬂow rates of materials or services into and out of components. The
preceding chapters I through V have dealt with the prerequisites that are necessary for
the design, construction and implementation of a computer simulation model. Literature
review and data collection and analysis established the foundation on which the
analogous simulation model of DAP-SYS can be made, implemented and used for
obtaining information about the performance of the real physical system, without the
rigorous testing and evaluation of the biological (draﬁ animals) and mechanical
(implements) components of the real world system. The beneﬁts of this type of approach
in research cannot be overemphasized. Cost effectiveness notwithstanding, there is no
limit of the possibilities that the simulation model can render to a researcher using a well
designed and comprehensive computer simulation model.
6.2 Simulation modeling

A model is deﬁned as a mathematical or physical system obeying certain speciﬁed
conditions whose behavior is used to understand a physical, biological or social system to
which it is analogous in some way. Simulation usually means that there exists a
computer program or other functional model of a system on which diﬂ‘erent design and
management strategies can be tried. Simulation modeling involves a model that is
capable of performing in an analogous manner akin to an actual physical system.
Analyzing a simulation model enables a system analyst to understand the effects that
changes in system variables impose on the performance of a system without necessarily
building the actual system.

Although systems analysis can be done without computers, the computational power

of computers makes many system analysis tools easier to use. Simulation models are

93

used to display the impacts of alternative decisions upon a number of measures of system
performance which in turn can be used by decision makers to arrive at courses of action
which are acceptable or “good”. Two types of systems are commonly used in solving
problems of physical or non-physical nature. The ﬁrst is called the black box approach
where inputs and outputs can be observed and measured, but the process of
transformation of the inputs and outputs remains unknown or is of less importance to the
user. The second type of a system is called the structural approach which begins with a
careful examination of a system structure and theory to determine the basic system
components and linkages. Both the black box and the structured approach can be used to
simulate the same physical system.

The Energy Utilization Model for Draﬁ Animal Power (EUMDAP) computer
simulation model developed by the researcher in this project incorporated the two
approaches. Essentially the computer simulation model developed was designed to be
used for the modeling of a DAP-SYS operating in semi-arid lands. Energy utilization by
draft animals for tillage purposes was the primary focus in developing the model. The
objective of the exercise was to enhance the utilization of the energy from draft animals
both in terms of the annual days the system is used and also the efﬁciency of using this
power source for crop production. Although the conversion of feed to energy in animals
is well understood, the modeling of energy production and utilization by draft animals
was not done at the cellular level. Modeling the energy ﬂow through the DAP-SYS was
done at the macro level using the already established mathematical relationships of the
feed energy values and the animals’ metabolic rates for both maintenance and work and

the existing models on energy used by animals at work.

94

Computer simulation models have been utilized for the design of new systems or for
the management of existing physical systems. Principles of system analysis are used to
determine how a set of components will behave in response to changes in the value of
inputs or the interactions of inputs. In describing a system, consideration is given to the
entities or components that would be included within the boundaries of the system. A
detailed description of the components of the system and their linkages enables the
design of the system to be done without making too many assumptions.

6.3 Computer simulation modeling

According to Neelamkavil (1987), a model is deﬁned as a simpliﬁed representation of
a system intended to enhance the ability to understand, predict, and possibly control the
behavior of a physical system. Simulation has also been deﬁned by Rockwell (1965) as
meaning to duplicate the essence of a system without actually attaining reality. Hence a
computer simulation model is one that is adapted for simulation on a computer.
Computer simulation models have become useful tools in simulating biological and
physical systems of agricultural production entities. The draﬁ animal power (DAP-SYS)
physical system was represented in abstract terms by the EUMDAP computer simulation
model to achieve speciﬁc objectives. EUMDAP enables monitoring of the physiological
response and interaction of the biological component (draft oxen) with the mechanical
component (implements) of the DAP-SYS in an abstract setting. The purpose of the
EUMDAP computer model was to simulate real DAP-SYSs by using varying rates and
levels of input variables in order to achieve optimum energy utilization of draft animals

in small-holder agricultural production systems in semi-arid areas.

95

Computer simulation models are used as alternatives to analytic solution methods
using mathematical abstractions of the system to relate various functions that together
affect each other in achieving a set of goals or objectives. The design of computer
simulation models is preceded by graphical or conceptual models which provide the
inter-relations between the various components, their linkages and direction of ﬂow of
materials or services.

Naylor et al. (1986) describes the purpose of a simulation model as being to enable the
analyst to determine how one or more changes in aspects of a modeled system may affect
other aspects of the system as a whole. Coefﬁcients of equations used to relate system
components are derived from empirical and historical data. The data used for the
development and construction of the EUMDAP computer simulation model were
collected in the ﬁeld using socio-economic and DAS instruments to obtain ﬁrst hand
baseline and operational data from farmers that predominantly use draft animal power in
Kenya. Historical and secondary data from literature were also used for parameters and
variables that could not be obtained directly from the ﬁeld surveys or energy utilization
experiments.

6.4 EUMDAP computer simulation model development

In a system approach to problem solving, the process of developing a computer

simulation model to solve a problem involves various stages which include:

identiﬁcation of needs

formulation of the problem
formulation of the conceptual model
data collection and analysis
formulation of the mathematical model
model implementation

veriﬁcation of the model

96

o validation of the model

0 sensitivity analysis

0 evaluation of the simulation model
Each of these stages provides information that is used for the subsequent stage of model
development. A checking process is used recursively to ensure that each stage is covered
comprehensively before proceeding to the next one and that the desired structure of the
model is progressively formulated. Figure 6.2 shows the step-by-step stages of the
development of a simulation model. The dotted arrows from the left represent the ﬂow
and accumulation of information between iterations which is processed and used for the
model development. This step-wise development process for EUMDAP computer
simulation model of the DAP-SYS is described in the succeeding sections.
6.4.1 Identification of needs

Identifying the needs of the DAP-SYS used by the small-holder farmers of semi—arid
areas of Kenya was the ﬁrst stage of the process of formulating the computer simulation
model. While other agronomic and socio-economic factors of agricultural production are
important, the simulation model which was developed is hinged on the timing of tillage
operations as critical in ensuring that farmers have a good chance of obtaining a harvest.
The majority of farmers in the study area are largely subsistence growers producing
mainly maize and beans as the primary seasonal crops. Maize, the staple food of Kenya

is grown by virtually all small-holder farmers in the semi-arid lands. The primary

objective of a subsistence farmer is to ensure food production (food security) for his

positive and normative information

97

needs

1

......... a: formulation of the problem Jie———

 

/“ ‘\‘\

 

 

 

-1: complete i: — ‘ 1_\I___’
\\\‘ ‘ x// ’
Y
uuuuuuu ){ formulation of conceptual model
//,/ "\\\N\
1‘; adequate ,
\ /,/

\//

‘1

{data collection and analysis]

 

 

 

——-——>I----),i formulation of mathematical mocﬂl

 

 

 

I model implementation J

 

 

 

veriﬁcation ]

 

 

V,

0“

\

- :‘satisfactory , ' i; i—e
4L,
l validation J

V
. -------- ->{ sensitivity analysis]

 

 

 

 

\

/. \\.\_ N
., :. \ satisfactory; ,... .; ———3

YI/

evaluation of simulation results

 

 

 

V

 

 

 

 

Figure 6.2: Step by step simulation model development

98

household. When maize is grown as a pure stand (single crop), limited post-planting
mechanization like weeding can be done with draft oxen. Draft animals may be used for
the ﬁrst weeding of pure stand crops while the crops are small. Subsequent intra-row
hand weeding is labor intensive. Expansion of cropped land is often constrained to the
acreage that the family labor can handle in manual weeding. The EUMDAP simulation
model implementation factors this limitation in projecting the land expansion
possibilities.

Small-holder farmers in semi-arid areas face various food production challenges.
Foremost among these is that the prevailing soil conditions during the optimum land
preparation period are hard. More often than not farmers wait for the initial rain
showers to soften the ground before they can embark on the tillage process. This delays
land preparation which causes late planting and subsequently late weeding. Without
timely planting and weeding, which are the main factors affecting yield, other
recommended management practices and technological inputs such as improved seed,
fertilizer, plant population and crop protection yield low returns. In semi-arid areas the
loss of yield due to delayed planting is very substantial. According to experiments on
yield of maize in semi-arid areas, up to 500 kg per hectare [5 quintals] loss of yield
occurs per week of late planting or late weeding (Johnston, 1984). It is therefore
imperative that timely planting in order to maximize on the scanty rainfall be given ﬁrst
priority in modeling for tillage with draft animal power systems.

The soils prevalent in the study area include the well-drained dark reddish Luvisols
and heavy clay Vertisols, generally known as black cotton soils. These soils tend to be

hard and impenetrable in the dry season due to “cementing” on the surface for the red

99
soils and crusting for the black soils. On the other hand black soils become heavy and

sticky when very wet but the red soils are manageable when wet. The results of the data
collected indicated that about 55% of the farmers do the ﬁrst plowing soon after harvest
when the soils are neither too dry nor too wet. This post-harvest plowing enables tillage
to take place before the rains and facilitates timely planting. The underlying need for the
small-holder subsistence farmer is to produce sufﬁcient food for the household despite
the climatic and environmental odds operating against crop production. The problems
sited above that hinder food production are factored into the EUMDAP simulation model
in order to arrive at an optimtun combination of resources and timing of operations for
ensuring that the farmer obtains a crop.
6.4.2 Formulation of the problem

The process of formulation of the problem entails the identiﬁcation or deﬁnition of the
problem to be addressed and setting out the objectives to be realized in the simulation
model. As stated earlier, DAP utilization in semi-arid lands of Kenya is bereft with many
constraints. Some of the problems are associated with the harsh climatic and
environmental conditions prevalent in this geographic area. The low productivity of the
land makes the predominantly peasant farming community prone to chronic food
shortages necessitating government food supply interventions during the drought stricken
years. The main and most pressing limitations that the EUMDAP computer simulation
model incorporates include:
poor physical state of the draft oxen during the tillage season.
low annual utilization of draft animal power
inadequate quality and quantity of feed for draft oxen

hard and dry soil conditions
erratic and poorly distributed rainfall pattern

100
6. 4.2.] Physical condition of draft animals

The primary objective of the EUMDAP computer simulation model was to formulate a
tillage system model that would utilize the current technological innovations at the
farmer’s disposal in such a way that land, draft power and labor productivity would be
optimized. One of the main concerns at the outset was the prevalent and chronic loss of
weight of the draft animals both before and during the tillage season. The optimum land
preparation time of the two cropping seasons falls when the grazing ﬁelds have been
depleted of adequate fodder for the livestock. About one third (30%) of the farmers
surveyed harvest and store maize stover for the feed deﬁcient periods. Data collected
from the ﬁeld by the researcher shows that commercial supplementary feeds are not
readily available in the semi-arid areas. It is estimated that, in general, livestock in the
semi-arid areas lose up to about 20% of their body weight by the end of the dry season.
This situation invariably leads to the use of the draft animals at their worst physical
condition. The draft requirements of the single bottom Victory plow commonly used by
farmers does not justify the use of more than two oxen at a time. However, some farmers
train more animals than they would normally need so that they are available to work
together (in fours) during the peak power demand period if the need arises. Field data
collected by the researcher found that although about 30% of all the farmers interviewed
have four draft animals trained for work, only about 25% of all respondents actually use
the two pairs together.

The use of multiple pairs of draft animals does not take full advantage of each
animal’s potential power output. The efficiency of multiple pairs of animals declines

with each additional animal in the team. This loss of energetic efficiency is relative to the

lOl
tractive effort of a single animal. The efﬁciency of utilization per animal declines with

each extra animal used. Goe and McDowell (1980) documented this loss as amounting to
7.5% for two, 15% for three, 22% for four 30% for ﬁve and 37% for six animals. It is
therefore desirable to utilize a single pair of oxen to its maximum potential to avoid the
cumulative loss of power when multiple pairs of oxen are used in tandem. In order to
enhance DAP utilization for farmers with four oxen the EUMDAP model simulates the
use of single pairs intermittently in blocks of two or three hour intervals.
6. 4.2.2 Annual utilization of draft oxen

The number of hours that oxen are utilized per year depends on the variety of activities
in which the animals are engaged. Farmers with more than two trained oxen have a lower
net annual draft oxen utilization rate per animal. The survey data obtained from the ﬁeld
also shows that farmers who use only a plow log relatively fewer hours per year when
compared with farmers who use oxen for other purposes. The total average time that
draft animals are used in the semi-arid areas was found by the survey to be about 58
hours per year. There is a substantial untapped potential for higher utilization of DAP-
SYSs both during the tillage and post tillage periods. While tillage is the most important
and the most energy demanding activity at the farm level, farmers with other implements
such as ox-carts and improvised transport sleds log in more draft oxen hours per year than
those with only a plow.

At the end of the plowing season the grazing ﬁelds have sufﬁcient and nutritive
forage. The oxen soon gain back their condition but unfortunately they are put to little
use except for weeding and miscellaneous transportation like carting water, building

materials, ﬁrewood and other items. The simulation model attempts to maximize the

102

utilization of a single pair of draft oxen during and after the tillage season period. By so
doing the annual utilization and the potential returns that the farmer can expect are
increased. As noted in section 2.3, Indian draft power farmers use their animals almost
three times as much per year as the Kenyan draft animal farmers.

6.4.2.3 Quality and quantity of feed

In general, farmers depend on natural grazing to feed their livestock including draft
oxen. The milk producing cattle receive preferential feed regime (where available). The
natural grazing ﬁelds in semi-arid areas have a very low livestock carrying capacity.
Ideally 2 to 4 ha are required per livestock unit (LU) for adequate feeding (Silberfein,
1989). According to the ﬁndings of the data collected from the research areas, about half
of the land owned by small-holder semi-arid farmers is used for crop production. The
remaining land is barely enough to maintain the farmers’ livestock. Hence the quantity
of fodder available for livestock on natural grazing falls far short of the minimum
required except after the rainy period.

The metabolizable energy (ME) feed value of natural pasture ranges between 4.0 and
8.5 MJ/kg DM (Zerbini et al. 1992). According to guidelines by ARC (1980) the ME
available is attenuated by an efficiency of utilization factor (K...) and also by the dry
matter content of the feed (DMC). Given an appetite limit or voluntary dry matter intake
(VDMI) of 2.5% (ARC, 1980; NRC, 1996) per unit weight (kg) of the animal per day,
draft animals are unable to ingest sufﬁcient low quality feed to meet both maintenance
and work energy needs. The amount of VDMI that a 350 kg ox can ingest per day is 8.75
kg of DM. From this amount of feed the ox would extract between 35.0 MJ and 74.38 .

M] of metabolizable energy (MB) per day from natural grazing pasture. Using models

103
developed by Brody (1964) and used in the EUMDAP simulation modeling (see equation

13 in section 6.4.5.2) the energy for maintenance (Em) for a 350 kg ox would be 23.87
MI per day.
6. 4.2.4 Soil conditions

The prevalent highly drained sandy soils of semi-arid lands and the less common black
cotton soils are both difficult to work when dry. The high clay content of these soils
causes hardening when the soil moisture declines with receding rain. This hardening
increases the draft requirements of the tillage tools. Though timely land preparation is of
prime importance in the semi-arid region, it is often not practical to engage animals in
tillage operations when the soils are still dry. A characteristic shear resistance in these
soils when dry is 50 kN per square meter. A pair of oxen (each weighing 350 kg) would
develop a total force of 0.65 kN based on a 10% rule of thumbl for each animal and
92.5% “tractive efﬁciency” for a two-animal team2 (Goe and McDowell, 1980). For a
25.4 cm (10 in) plow operating in a dry soil at a depth of 10 cm, the draft requirements of
the plow would be about 1.27 kN which is beyond the maximum tractive force generated
by a pair of 350 kg oxen.
6. 4. 2.5 Unreliable rainfall

The minimum required moisture in rain-fed agriculture is 726 mm (30 in) of rainfall
for crops to mature. In Machakos and Mbeere districts, the probability of obtaining more

than 590.] mm in the March-May season is only 10% while that of obtaining 670.5 mm

 

' It is generally considered that oxen develop a pull force equivalent to 10% of their body weight (rule of
thumb). ,

2 If one ox is working, it is expected to develop a pull force approximately 10% of its body weight. When
two oxen work in a team, the pair loses 7.5% of its potential pull force so that the total “tractive efﬁciency”
or pulling ability is reduced to 92.5% of the potential pull force (Goe and McDowell, 1980)

104
or more is only 10% in October-December season. The two districts studied not only

have erratic and poorly distributed rainfall but also the probability of drought is high,
occurring about one in every ten years (Tiffen and Gichuki, 1994).

Some of the limitations of the expansion of land use for the small-holder farmer in the
medium and low potential areas are the low livestock holding capacity, the small acreage
of the land and the poor, and dry soil conditions. When the soils of semi-arid areas are
dry, they tend to crust and harden. In the dry period the soils develop high implement
draft force which is often beyond the capacity of the draft oxen. In addition, the physical
condition of the draft animals is at its lowest at the time of tillage for the next season’s
planting due to inadequate fodder. The climatic conditions of these lands also pose a
moisture supply problem due to poor distribution of rainfall during the crops’ growing
season. The basic need that the farmer has to meet given the economic and
environmental scenario he is in is to produce enough food for his household. The needs
and challenges stated above that the small-holder farmer faces are used for subsequent
process of problem formulation. Better utilization of the DAP system is expected to
provide increased returns for the household in order to improve the provision of food.

The problems that a farmer faces in trying to produce enough food for his family and
preferably have some marketable produce left are technical and technological, as well as
environmental and cultural. The implements available at the household level for the
preparation of land by timely plowing are limited to mainly a moldboard plow which is
incapable of penetrating the hard soils during the optimum land preparation period before
the onset of rains. It is imperative that the land preparation is done in a timely manner to

ensure that the planting is completed before the rains fall. Given the scanty rainfall that is

105
prevalent in the semi-arid areas, the time of planting relative to the onset of rainfall can

determine the success or failure of a cropping season. The problem to be addressed in
attempting to assist the farmer can be deﬁned as:

How can the small-holder draﬁ animal power farmer in the semi-arid area of Kenya
combine the production resources at hand to prepare the land on time in order to: (a)
maximize on the scarce and poorly distributed rain fall and (b) optimize on the
available draft animal power so as to (c) produce enough food for his household?

The simulation model addressed the following areas of the DAP-SYSs of semi-arid

lands:

the potential economic gain in increased utilization of draft animal power
the potential of using only one pair of well trained oxen.

multi-purpose use of draft animal power for increased income generation
hiring draft animal power

land-use expansion

6.4.3 Formulation of conceptual model

The DAP-SYS is comprised of both physical and biological components that work
together as an entity to provide draft force to the soil in tillage operations. The oxen
(prime movers) are the source of power for the system, while the operator(s) serve as the
power controller. The animal, implement and soil interact together with each other to
achieve the objective of accomplishing a task (tillage or transportation) within certain
constraints of time and resources. The EUMDAP conceptual model shown in Figure 6.3
identiﬁes ﬁve components of the system which include:
draft oxen
harness mechanism
implement
operator

land (soil)
These components are interconnected through linkages which are services or materials

ﬂowing at varying ﬂow rates. The draft oxen receive water, feed, management and

106

training as inputs so that they can deliver the required energy to the harnessing
mechanism that in turn delivers draft force to the implements. The ultimate objective is
to manipulate the soil conditions to a suitable tilth for crop growth.

Detailed descriptions of the inputs and linkages are provided in the mathematical
model development in section 6.4.5. The exogenous factors or inputs that affect the
EUMDAP model are mainly institutional and environmental in nature and which are not
inﬂuenced by the system. Four institutional factors that inﬂuence the EUIVHDAP system

were identiﬁed. The government policy on agricultural mechanization can hinder

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Management ’,
Training .
Water '
Control Food :
I I Draft Force . I
. H I- Draft Oxen .
amess . ‘ u
- r Mechanical Energy .
: Draft Force :5 .
' Soil Resistance '
. Implement 3 :

. Control Tillage E
' ' Weeding V .
Wag Management '
'# Operator . Soil Shear Strength Land (Soil) ' .'

Q . - . . P 1 ............................................... . . . . ...... '
Training
[ Veterinary Services; Oxen Training
. Jua Kali
Regir Services I
Equimm‘ 09"“ Government Policy } "°"'°"°"

 

 

(INSTITUTIONAL ENVIRONMENT)

Fig. 6.3 Energy Utilization Model for Draft Animal Power (EUMDAP) conceptual model

107
or facilitate the promotion of draft “animal power in semi-arid areas. As

mentioned elsewhere, it was not until about 1975,. that the government of Kenya
explicitly enunciated a policy statement that considered draft animal power as a viable
mechanization option for small-holder farmers. Both the promotion of DAP-SYS and the
availability of alternative equipment for DAP-SYSs depend on the level of mechanization
that the government supports. Implementation of government policy is channeled
through agricultural extension services and the results of the research services provided
by both the public and private sectors. The fourth exogenous input considered is the
availability of artisans for the maintenance and repair of tillage implements.

For each of the system components a set of input and output variables was identiﬁed.
Some inevitable constraints limited a comprehensive inclusion of all possible exogenous
and endogenous variables that would affect the EUMDAP system. However, inputs that
had direct inﬂuence on the performance of the EUMDAP system were included to the
extent that data were available for parameter and variable determination. In designing the
model of the DAP-SYS, a careful consideration of the boundaries of the system was
taken into account. Inputs that affect the system from outside the system boundaries are
considered environmental inputs. An example of such an input is the agricultural
extension ofﬁcer and the government policy on small-holder agricultural mechanization.

An adequate conceptual model determines how well the ﬁnal system model represents
the actual or envisaged physical system. Determination of the required data and the
form, type and method of obtaining them depends on how comprehensive and detailed
the conceptual model is. Similarly, the validity and relevance of the mathematical

relationships developed and used in the abstract modeling depend on the linkages shown

l08

in the conceptual model. The linkages between the components are the paths of the ﬂow
of material and services between the system components.

The EUMDAP computer simulation modeling program is interactive with the user. At
the start of the program parameter and variable value entries are required from the user as
shown in Figure 6.4a through Figure 6.4f of the algorithm or logic ﬂow of the computer
simulation model. The user is required to select from pre-determined parameter settings
and variables for some of the input data such as the type of soil, implement types, and
depth and width of operation of tillage. The listing of the QBasic computer code is
provided in Appendix F.

6.4.4 Data collection and analysis

The process of obtaining DAP utilization data was done through the administration of
a questionnaire with which socio-economic data from a cross-section of small-holder
farmers in Mbeere and Machakos districts were obtained. The data collected through the
questionnaire provided parameter estimators for the computer simulation model.
Historical data were obtained where possible, otherwise best estimates were obtained
from farmers on variables such as the life span of working oxen. The size of the holdings
and time spent for tillage and other operation with draft oxen, as well as the labor used
for weeding with and without oxen, was obtained from farmers through the questionnaire.
Details of the procedure used and the data collected are covered in Chapter IV and the
results of the data analysis are summarized in Chapter V.

The energy demand by the implement and the energy available from the animals

were obtained through an electronic data collection system tailor-made for the research

109

 

input
farm size

 

 

 

, " in ut
,/ cr ped
land area

 

 

 

V ' .
, 1. before rains ,

 

 

 

 

 

 

 

_,,v’ "\
" 53139 x--. 2 at onset of /
Staff! 0 ;- <———— rains
». . V . tillage —- 3.soon after
‘ ~\\ /,/’ harvest
-x/séiw- ' 1,1- say flay /
\ - \> .. . y oam
~\\ fartrynpseorl // k— 3. clsgnloam /
. 4. blac cotton/
,// b ' input
, num er \\ /
Of draft oxen /,,,-4-—— number of /./
~\ I// . oxen ,1
input ,
ages of /
draft oxen /
ix.

 

(1,.)

Figure 6.4a: Algorithm ﬂow chart of the Energy Utilization Model for Draft Animal
Power.

110

   
 
   
     
  
     

_ select . l. > 3 years
servaceftrme With 2. 2 to 3 years
ra oxe 3. < 1 year

 

 

 

   
 
 
   
   

1. natural

   

select
feed for draft
oxen

select ,
feed quantity
type

pasture
2. cut grass-hay
3. maize stover

     

   

1. ﬁeld pasture

2. harvested
forage

Figure 6.4b: Algorithm ﬂow chart of the Energy Utilization Model for Draft Animal

Power (cont ’d).

111

input
% of days 12/
work/wee

 

 

 

in‘put
# of work
hours/da

 

   

l. acreage
2. distance per
week

 

 

 

 

 

N
input
DAP, labor and ,
equipment costs

 

   

 

 
  

. select
implement
options

input
age of
m/board plow

Figure 6.4c: Algorithm ﬂow chart of the Energy Utilization Model for Draft Animal
Power (cont ’d).

 
 
     

1. m/board plow
onlé
2. plow cart
3. ox-cart only

 

se cc
de thof
o ow1n 1

selec
width of
. lowrn 1

selec

default speed 0
. lowrng

 

 
 
  

112

 

l. shallow (.lm)
. medium (.2 m)
3. deep (.25 m)

 

 

1. narrow (.25 m)

2. medium (.3 m)
3. wide (.35 m)

input
preferred speed
of operation

 

 

 

# input
lweeciays per
2. distance/day

Figure 6.4d: Algorithm ﬂow chart of the Energy Utilization Model for Draft Animal

Power (cont ’d).

113

 

 

 

 

 

 

 

 

 

 

 

 

 

 

h 1d I/#i1"p"’
«:1: fé’n’sepSn ,pi::‘><——, ' , 5 per ,
nee/cls/ . 2. dIstance/day/
/ //hirin ’ 1. acreage ,
/ out D \E Y 2. distance per/
\ /’ week /
Nl/ ‘
.l‘
,. input
DAP, labor and
equipment costs
l
Outputs
user input variables DAP eneng expenditure
theoretrcal ﬁeld total DA energy needs
effective ﬁeld gross energy
annual utilization net energy
Rental capacity absolute energy
""386 time tillage capacity
draft force transport capacity
i ﬁeld efﬁCIency Impe ment energy
I
.v.

Figure 6.4e: Algorithm ﬂow chart of the Energy Utilization Model for Draft Animal
Power (cont ’d).

114

I; I; 1)

\ .
\. -

l

 

 

 

 

 

 

.. r“ print results \Z‘:~————' to
out//’ printer
3

."\
/ g/ \3
\\

\\

S "‘ ..

z"

\ 7’ save result

7"\ /

\\ _/,/-
\ / /’/ USC\‘\\‘
_. default ﬁlename )—

\-
\\

x/ ’
1' create /

/ " and input /

 

 

 

 

 

-—“‘L——\ ,2 ﬁlename /
I“ end i l /
output

results saved
In azyilename

 

 

 

( end .>

Figure 6.4f: Algorithm ﬂow chart of the Energy Utilization Model for Draft Animal
Power (cont ’d).

115

project as described in sections 3.2 and 3.3. Four physiological parameters of the oxen

at work were monitored with the farmers’ animals at work. Using models developed
by Richards and Lawrence (1984), estimation of the energy produced by working oxen
was estimated from the heart rate of the animals at work including heart rate, stepping
rate, breathing rate and body temperature. By simultaneous measurement of the
mechanical energy output of the animal to the implement, it was possible to relate the
energy provided by the animals to the speciﬁc draft force developed. Mechanical energy
parameters monitored for the DAP-SYS included horizontal and vertical draft force,
speed of working and depth of tillage as well as the width of tillage. A Victory plow was
used for the whole exercise of electronic data collection. Detailed description of the data
acquisition system is covered in Chapter III.

Soil and other environmental factors that inﬂuence the quality and rate of tillage work
were measured in each ﬁeld. These included the soil shear strength and the soil moisture
content. From the contacts made with the farmers, it was possible to draw conclusions
about the level of support that the extension ofﬁcers give in the promotion of small-
holder agricultural mechanization in both districts. This was especially noteworthy with
respect to the training of the farmers’ animals to adopt a better system of control of
animals at work. In Mbeere district, a program of training both the farmers and animals
was being promoted by a non—government organization using the agricultural extension
officers to implement it. The use of a single pair of well trained draft oxen was a
common feature in Mbeere district. There was no comparable program that helped
farmers with DAP mechanization needs in Machakos district. The major distinction

between the DAP management in Mbeere and Machakos districts is the handling of the

116

animals. While farmers in Machakos use voice commands and prodding methods to get
the animals to work, in Mbeere farmers have started adopting the nose-ring-and-reins
control method which the government has been promoting. This method enables the oxen
operator to make turns at the end of the ﬁeld by simply pulling the rope (reins) of the
right-hand ox.

There are limitations of measuring the power output ﬁ'om animals since it is not
deterministic like the power from fossil-fueled machines. The output from animals varies
in the short term (less than a day) due to both endogenous and exogenous factors. It was,
therefore, necessary to ensure that the process of data collection for each farmer’s location
was completed in one working day. so that each farm was treated as a block for data
analysis purposes.

6.4.5 Construction of mathematical model

Mathematical relationships were used in the EUMDAP simulation program to
represent the linkages between various components. Models of DAP energy utilization
that have been used by other researchers were utilized for this purpose for some variables
while new parameter values and relevant data from the research area were used for other
variables. Parameter estimators used in other DAP research and animal nutrition
literature provided a good basis for the development of the mathematical models. The
EUMDAP computer simulation model puts emphasis on the energy production from the
animals and the subsequent utilization of this energy by the implements during tillage and
transportation.

Draft animals are isothermal energy converters with a maximal energetic gross

efﬁciency of 25% that is comparable to that of heat engines (Brody, 1964). However as

II7

a source of energy, animals consume energy continuously as basal metabolism. The
amount of energy they can produce for work depends on the quality and quantity of feed
they take and also the energy stored in the body.
6.4.5.1 Energy needs for draft oxen

Animals provide the energy to do work by converting chemical energy stored in their
bodies to mechanical energy. Unlike mechanical chemical energy converters (for
example petroleum-based engines) which transform chemical fossil energy to mechanical
energy at a predetermined rate, the type of feed that animals eat determines how much
energy they can derive from it for maintenance and work. The energy so converted is in
turn provided to the animals through feed ingested and digested to convert the gross
energy (GE) to digestible energy (DE) and eventually to metabolizable energy (ME)
available to exert movement of muscles when combustion takes place in the muscle cells.
Gross energy also referred to as “heat of combustion” is the energy released as heat when
an organic substance is completely oxidized to carbon dioxide and water. The
relationships between the GE and the ME are well established.

The ME of draft animal feeds was used to estimate the quantity of feed required by
draft animals for maintenance and work. MB is deﬁned by Harris (1966) as the feed
intake gross energy minus fecal energy, minus energy in the gaseous products of

digestion, minus urinary energy as represented in equation (8).

ME =GE-FE—GPD-UE (8)

where:
ME = metabolizable energy, MJ/ kg DM
GE = gross energy, MJ/kg DM

118

FE = fecal energy, MJ/kg DM
GPD = combustible gases produced in the digestive tract
UE = gross energy of urine, MJ/kg DM

The value of ME of feeds are documented in publications such as Agricultural Research
Council (ARC, 1980) and National Research Council (NRC, 1996) and other animal
nutrition publications. Natural grazing is the most common source of feed for draft
animals in semi-arid areas. The ME energy value of natural forage is considered to be
between 4.0 and 8.5 MI per kg of dry matter (DM). In contrast, high energy feeds used
for dairy cattle such as groundnut cake have a ME value of about 14.2 M] per kg DM.

Due to the low quality of feed value available to draft animals and the intake
constraint due to the appetite limit of the animal, draft animals fed on natural grazing
hardly eat enough feed for maintenance and work especially during the dry spells
between the two cropping seasons of semi-arid areas. Farmers salvage maize stover and
harvest grass (natural hay) to provide feed for the animals for the dry periods of the year.
Maize stover and harvested hay provide a more efﬁcient energy source for animals
because the animals are stall fed minimizing energy cost of walking. Commercial high
energy value feeds are usually not easily available in semi-arid areas because of low
demand but the prices are also prohibitive. While the energy availability from local feeds
is a limiting factor, the appetite limit or the VDMI of animals also affects the energy
intake of working animals. Generally VDMI is estimated at 2.5% of the animal’s live
weight (kg) per day (ARC, 1980). VDMI is also referred to as the animal’s appetite

limit. If the energy value of the feed is low, the animal may not obtain sufﬁcient

metabolizable energy for both maintenance and work. Hence most work animals lose

119

weight during the working season and only regain it during off season when there is
sufﬁcient feed with high nutritive value.

Calculation of the daily feed intake by the animals is based on the metabolizable
energy (MB) of the feed, weight of the animal, appetite limit and the dry matter content
of the feed. Equation (9) developed by ARC (1980) was used to compute the maximum

metabolizable energy (ME...) intake for draft oxen per day:

ME," = F4 * ME (9)
where:

ME... = maximum metabolizable energy intake per day, MJ

ME = feed energy value, MJ per kg DM

Fq = feed quantity per day, (0.025/W., kg of DM per day)

The maximum animal energy intake is further attenuated by the efﬁciency of utilization
of metabolizable energy providing the net energy for maintenance and work (NEW). The
efﬁciency of utilization factor (K...) and the net energy for maintenance and work are
computed as shown in equation (10) and (11) also developed by ARC (1980). Equation

(12) computes the total energy need per animal per day for an animal at rest.

K," =0.019*ME+0.53 (10)

NE... = ME... " K... (11)

where:
K... = efficiency of utilization, decimal
NEmw = net energy available for maintenance and work per day, MI
ME... = maximum metabolizable energy intake per day, MJ

120

The energy available for maintenance and work translates to about 60% of ME... whereas
the rest is lost. Some of the “loss” is through the energy used for useless incidental
motions associated with work. According to Brody (1964), most of the “lost” energy is
expended for overcoming the internal resistance. Although the resistance of the body
colloids is energetically wasteful it is biologically useﬁll.

Each animal’s NE...w is computed separately and the total of all the animals in the team
is the total DAP-SYS energy needs. The gross quantity of feed required to provide this
energy for maintenance and work is computed using the gross energy (GE) feed values
and the corresponding digestible energy (DE) and metabolizable energy (ME) values as
well as the dry matter content percentage. The simulations performed with the program
used three different types of feeds commonly available in the semi-arid areas. The feed

values and their corresponding ME and DMc values are shown in Table 6.1.

Table 6.1 Energy values of feeds

 

 

Dry Matter Metabolizable
Content [DM.] Energy [ME]
Feed Type (%) MJ/kg DM
Cut Grass (Natural Hay) 85 9.0
Maize Stover 90 7.5
Pasture 75 11.0

 

6. 4. 5.2 Work energy from draft oxen
The energy that working animals need to exact a draft force to pull the implement is

obtained from the feed that they eat. Work animms should be fed enough feed for their

121

own maintenance and extra for doing work. Loss of body weight results when either the
feed source has insufﬁcient ME value or the maximum ME (ME...) intake is inadequate
to meet both maintenance and work energy requirements. The energy cost of
maintenance is the net energy required to keep an animal in a “steady” energetic state.
About 85% of the maintenance energy is expended through basal energy metabolism. By
deﬁnition basal energy metabolism, also called post-absorptive metabolism or standard
metabolism, is the heat production during complete rest of an animal in a thermo-neutral
environment in a post-absorptive condition; it is the resting energy metabolism in a
thermo-neutral environment uncomplicated by the heat increment of feeding (Brody,
1964).

The energy required for maintenance or basal metabolic rate (BMR) is a function of
the animal’s body weight and age. The BMR energy needs of cattle decreases
exponentially with age at a rate of 3 per cent per year (NRC, 1996). On the other hand
the energy for work depends on both the body weight, the type of work, and the rate of
doing work (speed of operation). Equation (12) developed by Brody (1964) was used in
the simulation modeling for the calculation of the maintenance energy needs of draft

OXCI‘I.

70.5(W, )0'73 "' 4.184 * 60'0"“
1000

3BMR = (12)

 

where:
BMR = basal metabolic rate per day, MJ

 

3 The original equation by Brody (1964) for basal metabolism of mature animals of different species, from
mice to elephants was in C per kg (W.)‘”3 per day. l C (large calorie) = 4.184 .1.

122

W. = weight of animal, kg
Age = age of the animal, yr
Equation (13) was used for the calculation of the net energy for work which is the

energy needed above the maintenance energy level.

NE, = N15,”, - BMR (13)

where:
NEw = net energy for work per day, MI

The total net energy needed by working animals is an aggregate of the basal or
maintenance energy, energy needed for walking, carrying or pulling loads, raising loads
and body weight up a slope as shown in equation (14) that was used by Lawrence and
Stibbards (1990). This is the energy cost above the energy the animal uses when standing

still or lying down.

9.81hW,

E, = Em +Ede, +Ech, +W+
1000b

 

(14)

where:

toral energy used per day, kJ

maintenance energy per day, kJ

energy cost of walking, 2.0J/m kg weight

distance traveled horizontally, km

live weight, kg

energy cost of carrying, J/m per kg — used only for back packs
work done per day, kJ

vertical height traveled, m

efﬁciency of doing work raising body weight, 0.35 for oxen

eeaeaerrw

123

Generally for cattle the extra energy needed for walking above energy required for lying
down is 9% and also the energy expenditure of walking decreases with the speed of
walking (Brody, 1964 and Lawrence and Stibbards, 1990). Equation (15) by Brody

(1964) is a model of the effect of speed on the overall cost of energy expenditure in cattle.

' Y = 44e‘0-268‘ + 39.7 (15)

where:
Y = kcal per 45.45 kg weight per 1.6 km
S = speed of walking, km

Adapted from Brody (1964)

In order to compute the work that animals perform per day as provided for in equation
(14), various ﬁeld and operational parameters were estimated or computed from data
obtained from ﬁeld experiments and historical data.

Equation (16) was used for estimating the theoretical ﬁeld capacity (TF C) for tillage
operations. It is expected that the user of the EUMDAP program will provide operational
data to enable the program to simulate the DAP system’s performance. The input data is
covered in the model implementation section (section 6.4.6).

Computation of implement ﬁeld performance and the energy demand by tillage
implements were done using equations 16 through 21 that were developed from ﬁeld

performance equations in the ASAE Yearbook (1995).

 

' Conversion of this equation to SI units has been done in the modeling after the value of Y is calculated to
maintain the original linearity of the equation.

124

S*W
TFC=_1_0_ (16)

where:
TF C = theoretical ﬁeld capacity, ha/hr

S = speed of operation, km per hr

W = implement width, In
The total theoretical ﬁeld time (T.) for the ﬁeld operations (17) was obtained by the
inverse of equation (16) while the effective ﬁeld time (T.) was computed from equation
(18) which took into consideration the percentage of the machine width that was used (k).
The computation of the ﬁeld efﬁciency (FE) was estimated with equation (19) which
takes into consideration the total time lost (TI...) per hectare due to interruptions such as
implement adjustments, turning at the headlands and lunch breaks. For DAP systems, it
was estimated that the total time lost was about 8% of the total theoretical field time per
hectare. This estimate was based on the assumptions that both the operator and the draft

animals need to take a break for about ten minutes every one hour of work as observed

during the ﬁeld data collection.

 

T =__
' TFC - (’7)
100*?"
T = ’ 18
8 k ( )
where:

T. = theoretical ﬁeld time, hr/ha
T. = effective ﬁeld time, hr/ha
k = percentage of implement width used

125

 

FE = 19
T; - T... ‘ ’
where:
FE = ﬁeld efﬁciency, decimal
T1... = total time lost, hr/ha

The effective ﬁeld capacity (EF C) in ha/hr was estimated by equation (20) which
provides the actual total work done per unit hour. The equation was derived from the

TF C and the efficiency of ﬁeld work.

* 1K
EFC = s_in(/)_§_ (20)

The draft developed by the soil (speciﬁc soil resistance) depends on the amount of
moisture in the soil (moisture content), and the type of soil. The soil types covered in the
two districts included silty clay, clay loam, sandy loam and black cotton soil. The
estimated soil shear strength during the three possible periods of time of tillage were
estimated using a hand penetrometer.

The computation of the total energy used for the ﬁeld work per day required
determination of the number of runs that the DAP-SYS does in completing a day’s work.
The average ﬁeld length for the data collected in the farmers’ ﬁeld was 28.8 m with a
standard deviation of 11.9 m. The minimum and maximum lengths recorded were 12.8
m and 58.3 m respectively. A computer random generator was used to determine the

ﬁeld length using the effective ﬁeld capacity. This calculation also determined the ﬁeld

126

width from the area plowed. The actual working width of the tillage implement was
determined by the implement size and the ﬁeld efﬁciency factor (k). The computation of
the total work done per day in M] was done with equation (21) based on the speciﬁc soil

resistance (kN/mz), the hours worked per day, the EFC, and the actual width of tillage.

 

It 1! t 4' It
Em: = EFC 10000 SSR hr W D (21)
1000 * W
where:
Em... = total work done per day, MI
EF C = effective ﬁeld capacity, ha/hr
SSR = speciﬁc soil resistance, kN/m2
D = depth of tillage, m
hr = hours worked per day, hrs
W = width of tillage, m

The following set of equations were used for the computation of the power (or rate of

working) and the implement energy (IE) of the tillage implement.

P = D I- S (22)
where:
P = power, kW
D = draft, kN

S = speed, m/sec

The draft force in equation (22) was based on the estimated soil resistance calculated
using the depth and width of tillage and the speciﬁc soil resistance and the width and
depth of tillage. The speed of operation is controlled by the draft oxen operator and

depends on the level of experience of both the draft oxen and the operator. Adequate and

127

consistent control of the DAP-SYS depends on the type of commands that the operator
uses to drive the animals.

Implement energy is a measure of the expected energy demand from the prime mover
(draft oxen) per hectare. This was computed using equation (23). The capacity or the
power required from the animals to provide the expected energy depends on the size of
the animal (weight), the type, quality and quantity of feed available for them and the
appetite limit of the animals. Equation (23) was based on the expected speed of operation

of the DAP-SYS as well as the unit soil resistance.

15.51— (23)
EF C
where:
IE implement energy, kW-hr/ha

effective ﬁeld capacity, ha/hr
implement power demand, kW

EFC

The theoretical total ﬁeld energy required to complete the tillage work was

computed as a factor of the acreage to be plowed and the implement energy as shown in

equation (24).
TH; = Lnd * IE (24)
where: _
Tn; = theoretical ﬁeld energy, kW-hrs
Lnd = the acreage for plowing, ha

IE implement energy, kW-hrs/ha

The projected tillage time to complete the ﬁeld work also referred to as the actual time

in days to ﬁnish ﬁeld work was computed using equation (25). The rate of energy

128

consumption for work in animals at work was estimated by Abiye Astatke et al. (1986) as
3.85 MJ per hour. The DAP-SYS energetic capacity (hours/day) was calculated from the
net energy available for work per day (NEW) and the energy production capacity (hours

per day) or tillage capacity of the oxen for work per day as shown in equation (26).

w EFC * Hr (25)

where:
TDFw = actual time to ﬁnish ﬁeld work, days
Lnd = land to be plowed, ha
EF C = effective ﬁeld capacity, ha/hr
Hr = hours available for work per day, hr

N15w (26)

DAP =
c 3.85

 

where:

DAPc = draft animal power working capacity, hrs/day

NEw = net energy for work, MJ/day
6. 4. 5.3 Energy utilization efﬁciency

The efﬁciency of utilization of energy by working oxen was determined for three

levels based on Brody (1964). The gross energy efﬁciency was computed as the ratio of
the work done per day and the total energy expenditure as shown in equation (27). The
net energy efﬁciency of the draft animals was calculated as the ratio of the work done per

day and the energy expenditure above the metabolic energy expenditure per day as shown

in equation (28). Thirdly the absolute energy efﬁciency was computed as the ratio of

129

work done per day and the energy expenditure above the energy expenditure of walking
without a load as shown in equation (29). The net efﬁciency has a better estimation of

the actual energy needed for work animals because it accounts for the energy needed

above the basal or maintenance metabolism.

GrsEf=—f%’—*IOO (27)

MW

where:
GrsEf = gross efﬁciency, %
Emu. = total work done per day, MJ
TNEmw = team total energy expenditure per day, MI

(28)

 

NetEf = E'°'°’ *100
TNEm - TBMR

where:
NetEf = net efﬁciency, %
Emu. = total work done per day, MJ
= total energy expenditure per day, MJ

TBMR = team total basal metabolic rate per day, MJ

AbsEf = EM * 100 (29)
TNEM - En Wlk

where:
AbsEf = absolute efﬁciency, %

Emu. - total work done per day, MJ
EnWlk = energy for walking per day, MJ

130

The total work done per day (Emu ) was derived from equation (21). The energy for
walking, however, was based on Lawrence and Stibbard’s (1985) model which uses an

estimate of 2 I per meter per km as the energy cost of walking for cattle.

The distance walked per day while plowing was computed using equation (30).

 

DSlDy = EFC * TDapCap . 100000 * RDN‘M, (3 O)
Wdth * RDNd',
where:
DstDy = distance walked per day, m
TDapCap = total draft animal power capacity, hr/day
Wdth = width of tillage, m
RDNds. = randomly generated length of ﬁeld, In

6. 4. 5.4 Crop yield and returns

The survey carried out with the farmers showed that the average farmer allocates
about 70% of their cropped land to the production of maize and about 30% of the land to
the production of beans. The socio-economic data collected also included the data for
yield of the main staple food crops (maize and beans). Estimates of the yield of maize
and beans were obtained from the ﬁeld extension officers. The average yield and the
standard deviation of maize and beans were used for estimating the expected yields for
the simulation model. The average values used for these estimates were 15 bags (90 kg)
per hectare for maize with a standard deviation of 3 bags. The average yield used for
beans was 6 bags (90 kg) per hectare and a standard deviation of 1.5 bags. For the
simulation of the staple crop yields, a computer random number generator was used to

obtain estimates using the average and standard deviation values.

131

The farmers’ main concern is to provide sufﬁcient food for his household. Hence, a
reasonable quantity of the staple produce is kept for home use and the extra disposed of in
the market. According to the data obtained from the farmers, an average of 10 bags of
maize and 2 bags of beans are set aside for home consumption.

The income generated ﬁom the farm was estimated from the sale of the excess
produce above the household food requirements. The price of maize and beans at the
time of the research was estimated at KSh. 1,200 and KSh. 4.000 per 90 kg bag
respectively. Computation of the gross net income per season was done for the costs of
the DAP-SYS and the input costs of the crops are listed in section 6.4.6.7.

The majority of the farmers breed their own oxen which they train for draft purposes at
the age of about 2 to 4 years. Farmers who buy draft oxen acquire them when they are
young enough for training. The oxen resale value after they have been used for about ten
years is always higher than the price originally paid for them. Hence, the acquisition cost
and the salvage value of the oxen were not used in the computation of the seasonal
variable expenses of DAP-SYS.

6.4.6 Model implementation

The implementation of the simulation model was done by writing a program in the
computer language that is most widely used in Kenya. The computer language selected
was QBASIC. However, after the program was completed veriﬁed and validated, it was
compiled by QBASIC so that it could be used as a stand-alone program. Since the

computer simulation program is geared for use by both extension ofﬁcers and research

132

workers in DAP in government institutions and NGOs, the choice of the language was an

important consideration.

The EUMDAP program was tested and run for various scenarios in order to determine
that it was computing as expected. This process also helped to determine whether there
were other details or data omitted from the programming. The user of the program is
expected to provide basic information that would be used to create a built-in DAP
utilization scenario.

6. 4. 6.1 Land and soil parameters
The information required from the user about the land and soil parameters is as

follows:

1. The size of the farm (land) in acres. Although ofﬁcial land adjudication
measurements are in hectares, in most parts of Kenya land measurement is best
understood in acres. The computer converts the land size to hectares for subsequent
computation purposes.

2. The size of the cropped land that is to be prepared for crop growth in acres.

3. The time when the tillage is expected to be performed. A choice of three possibilities

is provided for selection.
3.1 As soon as the rains come
3.2 Soon after harvest
3.3 Before the rains come.

4. The type of soil prevalent on the farm. There are mainly four types of soils in

the areas covered by the research. The user is prompted to select one of the

four:

133

4.1 Silty clay
4.2 Sandy loam
4.3 Clay loam
4.4 Black cotton
6. 4. 6.2 Draft oxen parameters
The following parameters and variables are required from the user.
1. Number of draft oxen used (choice of 2 or 4).
2. Age of each oxen
3. The experience of the current animals in terms of the period of time the animals
have been in use. Three categories of experience for each ox were considered:
3.1 Less than one year
3.2 Between 2 and 3 years

3.3 More than three years

4. The weight of each draft animal (kg) if known. Dimensions of the animals was
given as an option if the weight was not known. The heart girth and the length of the
animal are used to estimate the weight if the actual weight of the oxen was unknown .
6. 4. 6.3 Feed for draft oxen
Three feed types and quantity are considered for the simulation model. These are:
1. Maize stover
2. Cut grass or natural hay

3. Natural pasture

134

6. 4. 6.4 Operational inputs
Operational inputs include:
1. Number of days draft oxen are used for tillage per week.
2. Number of hours that draft oxen work per day
3. Time (hours) used for transportation for the household.
4. Renting of DAP-SYS for tillage and or transportation work.
6. 4. 6.5 Implement usage
1. The combination of implements used by the household. Three choices are
offered by the program:
1.1 Moldboard plow only
1.2 Moldboard plow and ox-cart
1.3 Ox-cart only.
2. The age of plow and or ox-cart.
3. Size of implement. (cm for plow and tonnage for ox-cart).
4. Depth of tillage preferred (cm).
5. Width of tillage preferred (cm).
6. Speed of operation.
6.1 Default speed (0.75 m/s or 2.7 km per hour)
6.2 Choice between 0.5 and 1.5 m/s
6. 4. 6.6 Renting out DAP-SYS
1. Acreage plowed on rental basis.
2. Distances and loads carried in renting out ox-carts.

3. Rental rate cost for plowing and transportation.

135

6. 4. 6. 7 Costs of DAP System Usage
The user provides the following costs of DAP system operation per season.

1. Feeding cost of oxen

2. DAP veterinary costs

3. DAP operator labor cost

4. Implement maintenance cost

5. Cost of planting per acre

6. Seeds and fertilizer cost per acre

7. Planting cost per acre

8. Weeding cost per acre

9. Harvesting costs per acre
6. 4. 6.8 Projected life-spans of implement and oxen

1. Expected working life span of oxen

2. Life span of plow

3. Life span of ox cart
6.4.7 Model verification and validation

In order to establish that the calculations and estimates that were performed by the

model were accurate and within reasonable limits, each of the equations used for the
model implementation was veriﬁed with a hand calculator. The equations that were
either derived for the model or modiﬁed from existing ones were tested for consistency of
units as shown in Appendix J. This process of veriﬁcation of the equations used enabled
correction of any errors discovered in the formulation of the mathematical relationships

of the model.

136

On the other hand validation of the model involved ensuring that the output values
obtained from simulations were consistent and reasonable for the DAP-SYSs simulated.
Although not all equations used were validated, the experience of the researcher and the
ﬁeld data obtained from farmers both conﬁrmed that the outputs obtained were within
reasonable values. A few outputs were selected and used for validation purposes.

The effective ﬁeld capacity (EFC) obtained from simulation for a pair of oxen
weighing 250 kg was 0.06 ha per hour when the animals were walking at about 2.6 km
per hour and working for 6.4 hours per day. At this working rate, it was expected that the
oxen would take about 2.6 working days to complete plowing 1 ha of land. The data
collected from the farmers indicated that farmers needed between 2 and 3 days to plow 3
hectare of land when the soil conditions were not too hard. From the researcher’s
experience in working with DAP-SYSs, the EF C obtained was within the expected rate of
work in semi-arid lands.

The expected speed of walking of the DAP-SYS is between 0.5 and 1.5 m/sec (1.8 and
5.4 km per hr). The data obtained from the farmers ﬁeld showed that the average rate of
walking was between 0.55 and 1.30 m/sec. (1.98 and 4.68 km per hr) as shown in Table
5.15. A walking speed of about 0.72 m/sec (2.6 km per hr) for DAP-SYS was therefore
within a comfortable range.

The power supplied from a pair of oxen weighing 250 kg each was estimated by the
model as 0.32 kW and the power from a pair of oxen weighing 350 kg was estimated as
0.45 kW. The ﬁeld data obtained from the farmers showed an average power rating of
draft oxen that ranged between 0.42 to 1.2 kW. This range was for oxen that varied in

size from about 280 to about 400 kg live-weight. The variability of the power supply

137

from working oxen depends on a number of factors including training level and the
ability of the operator to control the animals at work.

The simulated total time required to complete tillage operations ranged between 2.9
and 12.1 days with a pair of oxen weighing 250 kg and plowing between 1.2 and 4.7 ha
of land. Many farmers reported that they spent between 4 and 15 days to complete the
tillage operations depending on the size of the cropped land.

6.4.8 Simulating DAP-SYSs

The validity of the EUMDAP simulation model depends on the accuracy with
which the model can predict or estimate the required parameter and variable values. The
EUMDAP simulation model was tested by simulating various hypothetical scenarios.
The real world DAP-SYSs in the area where the data was collected for the development
of the modeling program were considered to be operating with land-holdings that were
limited to a maximum of 8.1 ha (20 acres). Simulation runs were, therefore, performed
for land sizes of 8.1 ha or less. The size of the oxen used by farmers range in size from
about 200 kg to about 400 kg. Although training of oxen can start at the age of about two
years, oxen weighing 200 kg are not powerful enough to be used for tillage. Farmers use
them for tillage work as part of the training process. The simulations were performed for
draft oxen weighing between 250 kg and 350 kg.

The preceding section (6.4.6) enumerated all the input variables that a user is
expected to provide for the simulation in order for the model to provide the required
outputs. After the entry of the inputs the user has the option of viewing the parameter
values used (inputs) and the outputs on the computer screen (monitor), printing them

directly or saving the data in the computer or on a ﬂoppy disk.

138

The format of data input and output print-out expected by a user from a simulation run
is shown in Appendix G. Several simulations were done by varying the inputs that the
user is expected to provide and also the choices provided in the program in order to
evaluate and observe the performance of the model. In the simulations, the size of the
cropped land was varied from 1.17 to 4.70 ha. The ﬁeld data collection showed that
about 58% of the total farm land is used for arable cropping while the rest is used for
grazing purposes and also for the homestead.
6.4.8.1 Scenario one

The majority of farmers use a single pair of oxen for tillage. The ﬁrst simulation was
performed with a single pair of oxen weighing 250 kg each and aged 3 years. At the
beginning of the plowing season the oxen are usually in poor physical condition and
weigh about 250 kg. This size represents the lower range of draft oxen used by farmers.

The following parameters were kept constant for this simulation:

1. Soil type - Silty clay

2. Feed type - Maize stover

3. Depth of tillage - 10 cm

4. Width of tillage - 25 cm

5. Speed of tillage - 2.6 km per hour

The size of the land plowed was varied from 1.17 ha to 4.70 ha. The program computed
and provided the user with the following categories of outputs which varied with the size
of land plowed (see Appendix G):

. Draft oxen energy supply

. Implement energy demand

139

. Feed energy supply

. DAP system performance measures

. Random estimates of produce used in simulation

. Income variables
The results of the simulation shown in Table 6.2 indicate that the pair of oxen was
capable of providing'0.32 kW of power while the implement energy demand was 0.28
kW. For the range of land sizes used for cropping the time needed for tillage was
between 2.9 and 12.1 days for plowing 1.17 and 4.70 ha respectively. The energy needed
to accomplish the tillage was between 5.38 and 22.26 kW-hrs when the animals were
working at 2.6 km per hour at an effective rate (EFC) of tillage 0.062 ha per hour. The
outputs from this simulation show the time required and the feed demand for a pair of
oxen when used for plowing different sizes of land. The user of the DAP-SYS is
provided with the performance information as well as the expected gross net returns
based on the costs of running the DAP system and the crop production inputs provided by
the user as shown in section 6.4.6.7. The gross net income represents the value of the
disposable produce after the farmer’s domestic home consumption needs are met. The
user can also decide whether the DAP-SYS can be used for other tillage operations for
the season based on the time required to complete the tillage work for the season.

The amount of feed required for tillage per day (above maintenance needs) was
computed as about 3.6 kg of maize stover per day. The energy developed by the oxen for
work was 24.8 M] per day whereas the total energy for maintenance and work was 57.3
M] per day. These energy values were constant for all iterations of this scenario. The

feed ingested by the oxen was about 5.1 kg of maize stover per day which provided a

 

140

 

.592 2n n are... 9.33 92 ”5.99. 9.“ n .53 3.. .98.. ”9%... 8.2 u 28 95803 use. 99 I 98% 9.529

 

9.9.3 m... 8.... ...Nv New . .. mm. . m and am... we... 2....
3:... 0.0 5.9. 00.. «.3 ... 3.». «W: ..N... 9.... 99..”
3&3. ﬁn 5.9. 9.9m New . .. 8.2 an... ..N... we... amd
999... a... 8.9. m.m~ New... 3.... mm... am... we... and
«No... ...M 3.9. .... New... .0... am... ..N... 3.9 3..
c 9.. 8.9. P... NE... and mm... am... we... 5...
.8. .99 ES: 9.. 9. 8-3.. 292-3. 3. >3 2
9899:. 98. .. 9:90 ....93 09.. 9.99.. 39:“... 9.3.05qu 08.890 .....nsm 0:9.
.92 9mg... .83 .9. 099 “. 9999: 8989.98. 3.8.”. 3.9:”. .9393. 099995
39.5 099 h. 898.92 .89... 8989.98. 8989.98. m<D

 

...999 w. omm w:...m.9>> :9x9 93. ....3 :9999m 99.. :9..9N....= 3.3m. .. m... :9..9.:8.m ”m9 9...:H

9.905). w.vm .I. 39:9 ...93 .92 9.90%.. 0.9. u ....93 .9. 099“. $90.9... 9.9 .I. 98.. 3.9093 ...—<89. 0N u 099% 08.9.5

 

33.9 ..N. 99.0 ..N: .0099 Nm. .N an... mm... mm... an...
3%? 9.9. 90.9 We... ”.09” an... am... am... mm... 20m
93.90 N... 90.0 9.9m «.09.. 8.2 mm... ..N... mm... mm...
acém m9 90.0 m.m~ ”.09” 3.... an... ..N... Na... and
2.9... ...v 90.0 .... «.03 .0.” an... an... and a...
9 9N 99.0 ...o. ”.09” wm.m mm... mm... mm... ....
am: 2.9 9.52 we. 9.. 8.-.... «£2-39. >9. 3.. a.
9899:. 98. ... 9:99 ...—93 09.8.93. 39:... 9.3.053”. 08.890 .....95 08...
.92 93...... 9093 .9. 099”. 8:9: 8989.98. 3.8m 3.9:”. .939. 099.95
9.995 099 n. 898.92 .89... 8989.98. 8989.98. A.48.

 

4.999 3. 0mm w:...w.93 :998 93. .....3 :99999 .99 5099...... $.95“. - .... :9..9.:8.m ”N.0 9.99...

141

maximum of 85.1 M] of energy per day as shown in simulation #1 in Appendix G.

In running the simulations it is assumed that the draft animals use only the energy
available for work in performing the ﬁeld work and that maintenance energy or body
reserves are not used for work. Hence, the oxen are not expected to lose weight when
used for draft purposes since simulated outputs utilize only the energy available for work
to perform the ﬁeld operations. However, in the farming conditions, farmers use their
oxen without considering that the animals are not well nourished for the work they do.
This causes the oxen to lose body condition as they utilize body reserves for work.
6.4.8.2 Scenario two

The majority of farmers’ draﬁ oxen do not exceed 350 kg of live weight. Table 6.3
shows scenario two which was done with oxen weighing 350 kg for comparison with the
smaller oxen of scenario one. The same working conditions used in scenario one were
simulated for the larger oxen so that only the oxen size was changed. The simulation
showed that the larger oxen performed the same task of tillage in about 40% less time in
total days of tillage at a working capacity of 10.06 hours per day. Although in reality
farmers do not use their oxen for more than seven hours per day, this simulation showed
that the larger oxen had the capacity to work for about 10 hours per day due to the higher
available net work energy from the animals. The net energy for work for the larger oxen
was 38.7 MJ/day while the smaller oxen had 24.8 MJ/day.

The amount of work done per day (MI/day) was also estimated at about 40% more
than the work done by the smaller oxen. The amount of feed required by the larger oxen
for maintenance needs was estimated at about 30% more than the feed used by oxen

weighing 250 kg (Table 6.3). The total implement energy demand for the two simulations

142

was constant at between 5.38 and 21.52 kW-hrs for 1.17 ha and 4.70 ha respectively.

The two pairs of oxen in simulations 1 and 2 were subjected to the same amount of
work and at the same speed of operation. The feed required to perform the work was,
therefore, the same for the two teams (ranging from 10.7 to 42.7 kg). The amount of feed
required for maintenance feed varied as it depends on the size of the animals. The
feeding cost also reduced the net gross returns obtained from the larger ox team.
6.4.8.3 Scenario three

Another simulation was performed with two pairs of oxen that were used alternately
during the working day. The results of the simulation performed with four oxen
weighing 250 kg each are shown in Table 6.4. Each pair of oxen provided 0.325 kW of
power and a total of 0.65 kW for the two pairs. The working capacity of the oxen was
6.04 hours per day per pair and each pair of oxen in this simulation was expected to work
for about 6 hours per day. The total tillage work per day is, therefore, doubled when
compared with a single pair of oxen. The total work done per day is increased from 6.40
to 12.09 MJ. However, the feed required for maintaining the two pairs of oxen increased
substantially from 896.8 for a single pair to 1793.6 kg of maize stover for two pairs of
oxen per season.

The main advantage in using two pairs of oxen alternately is when the time for tillage
is a constraint and the farmers wish to complete the work in time for planting. This is
often the case in the semi-arid areas where the time available for seedbed preparation and
subsequent planting is critical.

The feed requirement for maintenance of the oxen in both simulations is a major

component of the cost of maintaining the oxen. The three simulations performed show

 

143

.93.: 2.: u .9......» .33 .oz ”.339. 9.: n .33 a. 98: as»... 8.2 u as. 95.33 .85. 9.9 u 98% 2529

 

. ....Nm 9... .9.». 9.8 v.»..».~ mm. .m »m... »N... .9... .....v
9. N... ...M .9.». m9» v.»cm.m 9.9.». . »9... »m... .9... 99.»
999..» ».N .9.». 9.9m ...»..m... 8.2 »m... »m... .9... 9N.m
.»9.9. Gm .9.». m.mm 9.8»... v»... »m... »N... .9... »md
«.9.... 9.. .9.». .... v.»..».~ .9.» »m... »N... .9... »»..
.. N. .9.». 9...... v.»u..m.m »».m »9..~ »N... .9... 9...
.99. as: .99.... 9. 9. cf... .85... 3. 3. 2.
9899:. 98.... 9:90 9.993 92.993. 3.80 989.899.”. .89890 9.995 .89..
.97. 999:... 4.93 99.. 999.... 99:9: 8989.98. 3.80 99.8.. 9939.. 999995
9995 999". 898.92 .98... 8989.98. 8989.98. 9<0

 

...999 m. on» m:...w.93 :998 89.. ....3 :99999 .99 8999...... 399:0 - v... 99.8.98.» ”9.9 9.99...

.932 99.9.. n .955 .33 .oz 8.29. 9.9 u .33 s9 .93... ”.992. 2 .9. u 25. 95.33 59.5. 9.9 u 98% 9.529

 

.9»...9 N9 9...N. EN». 9.99.... mm..~ »9... »N... 99... 9:.
9.9.99 9.9 9...N. 9.9.. 9.99.... 9m». »9... »N... 99... 99.»
.89....V m... 9...~. 9.9m 9.999.. 8.2 »99 »N... 99... 9m. 9
N»».m~ 1m 9...N. m.m~ 9.99.... v»... »9... »N... 99... »9.N
m9».9 9N 9...m. ..9. 9699.. .9.» »m... »m... 99... »»..
.. 9.. 9...N. W... 9M8. 9.9.9 »m... »N... 99... 9...
.99. 939 5...). 9. 9. 3.-.... 98.-.... 3. 3. a.
9899:. 98. .. 9:90 ...93 999.999.. 999:... 998.899.”. 89.890 9.99.6 .89..
.97. 999...... .83 8.. 999... 9988 8989.98. 39:0 3.9:... .9399 99999.0
999... 999... 898.92 .98 ... 8989.98. 8989.98. 9.0

 

...999 9. 93 98.3.9? :998 8.9.. ....3 :99999 .99 8999...... 9999:... - m... :9..9.:8.m ”9.9 9.99...

144

that the feed used for work is the same for each scenario. This is expected because the
total work done in tillage is the same and the total demand for the “fuel energy’ should be
the same. The major difference between the single pair and the two pair teams is the time
required to complete the work.

The net gross returns decreased slightly as compared to the single pair of ox teams due
to the extra quantity of feed required for the second pair of oxen.
6.4.8.4 Scenario four

Using four oxen that weigh 350 kg each alternately further reduced the time required
to complete the tillage operations (Table 6.5). Each pair developed about 0.455 kW of
power. The total daily power available per day was 0.91 kW. The two pairs performed
18.91 M] of work per day. The maintenance feed requirements increased to over 2,300
kg for the season. Since the work done in tillage remained the same the amount of feed
required for the work was the same but the work could be completed in about 30% less
days.

The simulation indicates that the two pairs were able to do a total of 19 hours of work
per day. This may not be feasible because each team would have to work for a period of
over 8 hours. However, farmers with sufﬁcient labor and other resources can use
multiple teams of oxen simultaneously and achieve the ﬁeld time simulated. The
researcher observed that some farmers with multiple teams of oxen engaged two or more
teams at the same time and thereby increase their BF C.

The simulations with two pairs of oxen available have the potential of increasing the
ﬁeld time and also the actual acreage covered per day. The major drawback and often the

major bottleneck is the extra feed required for the maintenance of the oxen. This is

145

especially the case when the animals are used for owner-operator work only and no
renting out for income generation.

The larger draft oxen have more work capacity but use more feed per day for work
over a shorter period of time. As Tables 6.4 and 6.5 show, the smaller oxen need about
6.7 kg whereas the larger oxen need 10.6 kg of feed for work per day. The simulations of
larger two pair teams show that the draft animals have both extra energy to work without
straining themselves in the soil conditions simulated. They also have extra capacity
because they are able to complete tillage in fewer days. This is achieved by increasing
the ﬁeld time due to the higher energy level and also utilizing “blocks” of time in which
different pairs can work either alternately or simultaneously. This extra capacity can be
utilized for generating income for the household through renting the oxen out for tillage
and other activities. The quantity of feed required for the work done remained the same
for the two types of DAP systems (simulations 1 through 4) because the computation is
based on the total work done (hectares).

The gross net income from disposable produce ranged between zero for the smallest
parcel of land tilled to KSh. 60,851 for 4.70 ha of land when using the smaller pairs of
oxen and between zero and KSh. 52,711 for the larger team of ox teams. The larger ox
teams required a larger amount of feed for maintenance. As in the preceding simulations,
the quantity of feed used for work was the same for the two different oxen sizes because
the amount of work performed was the same for each pair.
6.4.8.5 Scenario ﬁve

In some cases farmers tend to use two pairs of oxen in tandem particularly when

working in heavier soils like black cotton soil (see table 6.8). When four oxen weighing

 

 

146

250 kg were used together in tandem the power generated by both oxen dropped from
0.65 kW to 0.55 kW. This was due to a “tractive efﬁciency” power loss factor of about
22% for a team of four oxen working together. This loss was not considered signiﬁcant
especially because the implement power demand was only 0.28 kW for the working
conditions simulated. In heavy soils such as wet black cotton soils, the extra power from
a team of four oxen can be useful. Farmers with soil conditions that demand extra
“tractive force” from oxen usually use a team of two pairs in tandem (Table 6.6).
However, this is only necessary when the tillage is done outside the ideal seed
preparation period when the soil conditions are difﬁcult.

Theoretically, the two pairs of oxen can do about 12.15 hours of work per day. Since
the oxen have extra capacity to work beyond the demand from the implement, it is
possible to use them for two blocks of time without sustaining loss in body weight.

The gross net income returns for simulation 5 and 6 were computed at the same input
costs as in simulations 3 and 4. The returns were the same for respective ox teams as
shown in Tables 6.4, 6.5, 6.6 and 6.7. The only factor that was varied in the income
computation was the quantity of feed required for each team. Other input costs including
DAP-SYS costs were held constant.
6.4.8.6 Scenario six

Four oxen weighing 350 kg each working in tandem experience a loss of power ﬁ'om
0.91 kW for a team of two oxen to 0.76 kW for a team of four oxen (Table 6.7). The
total feed requirements and the total energy used for work remain the same like for two
pairs of oxen used alternately.

Using four draft oxen in tandem does not provide much improvement in the overall

 

 

_‘;..¢L

 

147

 

.35.). 9...: u .998 .83 .oz 8.9... 9..: n .33 a. .98. 99.2. 8... u 2... 99...... use. 9.9 u 98% 2529

 

. .9.~9 9.9 .9.». 999 9.»9m.~ N9..N »9.9 »~.9 99.9 99.9
9.~.99 9.9 .9.». 9.99 9.»9m.m 9N». »99 »~.9 99.9 99..”
999. . m ».m .9.». 9.9m 9.»9m.m 99.9 . »99 »m.9 99.9 9~.m
_»9.9. N...” .9.». 0mm 9.»9m.~ 9»... »99 »N.9 99.9 »m.~
999.. 9.. .9.». ..9. 9.»9m.m .9.» »99 »m.9 99.9 »»..
9 N. .9.». 9.9. 9.t».9m.m »m.m »99 »~.9 99.9 9...
.9. ...: .9...: 9. 9. 9.-.... .....-3. 3. 3. ...
9899:. 98. .. 9:90 ...93 99.9.99... 999:0 33.999990 999890 9.9999 999..
.92 999...... ...93 .9. 999 .... 99:9: 8989.98. 999:0 99.990 .9399 99999.0
9.9.0 999 ... 998.92 .98... 8989.98. 8989.98. 9<0

 

899:9. :. 89.83
...999 9.. 99» 8.99.93 :99: .99.. ......» 999999 .99 99.89....9 99.990 - 9... 928.985 “9.9 9.99..

9999.3. 3.99 .1. 93999 ...93 .92 999.». 9.9 u .83 .9. 999... .999)... 2.2 u 9:... 8.9.83 ".58.. 9..... N 9999. 98.9.0

 

.m».99 9.9 99.9. 9.99 9.999.. mm. . m »n .9 »m.9 99.9 99.9
9.9.99 929 99.9. 9.9m 99.99.. 9m». »99 »m.9 99.9 99.9
999.99 m .9 999. 9.9m 9.99.. 99.2 »9 .9 »~.9 m m .9 99.9
«»».mm 9.9. 99.9. 0mm 9.99.. 9»... »99 »m.9 99.9 »m.~
m9».9 9.9 99.9. ..9. 9.999.. .9.» »m.9 »m.9 99.9 »».. .
9 9.. 99.9. 9.9. 9.m99.. »m.n »n.9 »~.9 99.9 9...
.8. a... 9.2.2 9. 9. 9.-.... .....-3. 3. 3. a.
9899:. 98. ... 9:90 .993 999999.. 99.80 9989:9990 9:9890 9895 9:90
.92 999...... ...93 .9. 999 ... 99:9: 8989.98. 99.990 99.990 .9399 99999.0
999.0 999 ... 998.92 .98 .. 8989.98. 8989.98. 9<0

 

. 899:9. :. 99.9.9.»
..999 w. 93 9:93.93 9999 .99.. ......» 999999 .99 :9..99....9 99.990 - 9... 9999.985 .99 9.99...

I48

performance of the DAP-SYS except in heavier soils. Mechanization extension services
recommend the use of only one pair of oxen at a time. When four oxen were used
together in the two simulations (scenario five and six) the power generated was about 2.7
times more than the power demanded by the implement for the larger oxen and about 1.9
times for the smaller oxen. This excess power could only be effectively used if a larger
implement was used.
6.4.8.7 Scenario seven

One of the major concerns for the farmers is the high draft force demand in heavier
soils when wet in the case of black cotton soils and when the clay soils are dry. A
simulation was performed for varying soil types and using oxen weighing 350 kg each as
shown in Table 6.8.

g The two heavier soils (clay loam and black cotton soils) generated draft power of 0.62
and 0.66 kW respectively while the team of oxen could only pull the plow at a maximum
rate of 0.45 kW. The two oxen were only capable of generating sufficient power to plow
in the lighter soils (silty clay and sandy loam). In the real world scenario, the researcher
observed that farmers often subject their draft oxen to loads that are beyond their pulling
capacity. The result is poor performance and low working efficiency as the animals are
coerced to pull beyond their limit. The. stress that the animals are subjected to leads to
weight loss and bruises from the harnesses as the operators force them to work.

In simulation 6, two pairs of oxen weighing 350 kg working in tandem developed
0.76 kW of power which could be used for the higher energy demand of the clay loam
and black cotton soils that require 0.62 kW and 0.66 kW respectively. A team of two

pairs of oxen weighing 250 kg and working in tandem is only capable of providing a

 

 

149

Table 6.8: Simulation #7 - Tillage with two oxen weighing 350 kg each in different soils.

 

 

DAP Implement lmplement Total

Power Energy Energy Implement Work Net

Supply Demand Expenditure Energy Done Efficiency
Soil Type kW kW kW-hr/ha kW-hrs MJ/day %
Silty clay 0.45 0.28 4.58 21.52 10.01 27
Sandy loam 0.45 0.36 5.73 26.90 12.51 34
Clay loam 0.45 0.62 10.02 47.08 21.89 60
Black cotton 0.45 0.66 10.60 49.77 23.15 63

 

Ground speed = 2.6 km/hour; Working time = 10.06 hours/day; Net work energy = BBSWJ/day

maximum of 0.55 kW which is insufﬁcient for the heavier soils depicted in Table 6.8.
6.4.8.8 Scenario eight

The type of feed and the quality of feed the animals are fed determines how much
energy they can generate for both maintenance and work. Three different feed types
were incorporated in the model based on the feeds available in semi-arid areas. A
simulation of the feed type with a team of two oxen weighing 350 kg and plowing 4.7 ha
at 2.6 km per hour generated the data shown in Table 6.9. The quantity of feed required
per animal for each type of feed depends on the quality and the digestibility of the feed.
Although draft animals are often used when the natural pasture is scarce, the feed value
of natural pasture is higher than that of stored maize stover or harvested grass during the
period aﬁer the rains have come.

The draft animals require a lesser quantity of the feed with a higher metabolizable
energy value. The animals fed with better feed value have a higher work capacity per

day because they generate more energy for work per day.

 

 

150

Table 6.9: Simulation #8 - Tillage with two oxen weighing 350 kg each using different

 

 

feeds.
Maximum Net DAP DAP Total Feed Feed
Metabolizable Energy Power Work Work Required Required
Energy for work Supply Done Capacity Per Season for Work

Feed Type MJ/day MJ/day kW MJ/day hrs/day kg kg
Natural
pasture 107.90 34.94 0.45 9.03 9.08 1,320 47
Cut grass
(hay) 100.41 25.60 0.45 6.62 6.65 1,4 I 9 51
Maize
stover 101.21 23.68 0.45 6.12 6.15 1,408 50

 

Ground speed = 2.6 km/hr

The net energy available for work depends on the quality of the feed, its digestibility
and also the quantity that the draft oxen can eat per day (maximum metabolizable energy
per day). The net energy for work is a direct relationship with the weight of the oxen.
High quality feed is not readily available in semi-arid lands of Kenya due to high cost.
Farmers preserve cut grass from rented or their own grazing ﬁelds for the dry periods of
the year when the natural feed is scarce. This source of feed is considered important in
making feed available in the lean periods of the season. Farmers who cannot afford
supplemental feed are disadvantaged during the drier spells as their animals depreciate
due to insufﬁcient feed and are in poor physical state when the tillage season starts.

6.4.9 Application of EUMDAP model

The EUMDAP simulation model has been developed to enable estimation of energy
requirements and utilization for DAP-SYSs in the semi-arid lands of Kenya. Since the
use of draft power in the marginal lands continues to be the main mechanization option,

farmers using the system need an improvement in the management and utilization of this

 

 

 

 

 

151

power source.

The results obtained from simulations of the program are comparable with the real
world scenario. The main application of the simulation results for real farming situations
will be in the extension of DAP-SYS in terms of the utilization capacity of draﬁ oxen.
Currently the utilization of DAP-SYSs is about 60 days per year on average. The
potential can be improved to about 180 days per year if the DAP-SYSs are used during
the period after tillage work is over through transportation and other off-farm tasks.
Increased use of DAP-SYSs in income generating activities has the potential of raising
the farmers’ economic level.

The model shows that the larger draft oxen have the capacity to work more hours per
day, enabling farmers to complete tillage operations sooner. The model can be used to
estimate and enable the extension workers to provide advice to farmers on the size of
draﬁ animals and the quantity of feed needed to be used for the ﬁeld operations.

The program generates the quantity of feed required for performing tillage and other
ﬁeld work based on the type of feed at the farmers’ disposal. A combination of the
working capacity and the feed required as well as the size of draft oxen required would
also facilitate the choice of the size of implement that can be used in order to perform the
tillage work in time for timely planting.

The potential for generating extra income from renting the draft oxen has not been
utilized sufficiently in the semi-arid areas. There are farmers that have two pairs of oxen
that are under-utilized. The response received from the questionnaire indicated that

farmers with two pairs of oxen only use them as stand-by in case they are needed.

 

 

 

 

 

152

For the farmers that prepare their land soon after harvest, there is sufﬁcient capacity to
generate more work among other farmers on rental basis since land plowed after harvest
needs less time and energy at regular tillage time. Land plowed after harvest provides the
farmer with two distinct advantages. The initial tillage after harvest makes the ground
workable at the beginning of the tillage season because the ground has already been
broken. Secondly, the stubble incorporated during the tillage provides humus enriching
the nutritive value for the next crop. Water inﬁltration and timely land preparation give
the crops an edge over weeds which compete for moisture and nutrients.

Post-harvest tillage also allows the use of the draft oxen when they are in prime
physical condition. Feed is readily available and so the draft animals do not lose their
weight during tillage. Consequently, during the ﬁnal land preparation at the beginning of
the season, the draft force demand by the soil is signiﬁcantly less in land that had post-
harvest tillage than the land being freshly plowed. It is expected that draﬁ oxen’s energy
utilization efﬁciency is enhanced and less feed is required to complete tillage on time for

the next cropping season.

 

 

 

 

 

CHAPTER VII
DISCUSSIONS, CONCLUSIONS AND
FUTURE RESEARCH
7.1 Discussions

The use of DAP-SYSs in semi-arid areas of Kenya will continue to fulﬁll the
mechanization needs of small-holder farmers for many years. Since attempts to promote
capital intensive mechanization strategies for small-holders have not succeeded, it is
desirable to continue to promote DAP in the areas where it is currently used. Contacts
made with farmers and mechanization extension workers by the researcher during the
period of this study and earlier indicate that there is need for concerted efforts in
enhancing the use of DAP for small-holder agriculture. Other developing countries like
India that use DAP-SYSs utilize this power source for longer periods per year than the
Kenyan farmers do. One of the reasons for this low utilization capacity is lack of a
variety of implements that can be used to utilize DAP after tillage operations are over.
There are various activities that DAP can be utilized for including transportation of water,
ﬁrewood and farm produce as well as soil and water conservation practices.

DAP farmers are creative in developing cultural tillage practices that enable them to
operate their limited tillage implements for a variety of purposes including weeding and
implementation of soil conservation measures like terraces and dam construction. Since
the oxen are idle most of the calendar year, the government and other non-govemment

development agencies can encourage the use of DAP for soil conservation work. One

farmer interviewed by the researcher uses his oxen for dam construction and repair

153

 

 

l54

after the tillage season is over. Semi-arid areas have chronic water shortage problems
which need to be addressed through promotion of water conservation measures.
7.1.1 DAS data

The data collected with the questionnaire and those collected with the electronic
DAS instrumentation were used for the development and implementation of the
EUMDAP computer simulation model. The tillage depth and width were used for the
computation of the draft force that the implement generated for each of the tests
performed in the farmers’ ﬁelds. The draft force measured was later used to validate the
simulation outputs during the testing of the program as discussed in section 6.4.8.

The mechanical data were also used for the computation of the power demand
from the oxen based on the ground speed and the draft force measured by the
instrumentation. The power demand as computed from the actual ﬁeld data collected and
summarized in Table 5.16 provided data for comparison with the estimated power
demand of simulated scenarios for validating the model.

The physiological data obtained with the DAS were needed to monitor the draft
oxen’s response to work. The primary variable for this purpose was the heart rate which
was incorporated in the model developed by Richards and Lawrence (1984), equation (3)
for the computation of the energy expenditure for work.

The other three variables, body temperature, respiration rate and the stepping rate
. were not utilized in the model. The body temperature did not vary within the time spans
that the measurements were taken. The longest period of measuring the data in the ﬁeld
was about 120 seconds. The stepping rate was computed to estimate the rate of walking

of the oxen as compared with the actual ground speed.

155

7.1.2 Socio-economic data

The socio-economic and DAP Operational system data that were obtained from
the farmers in the two districts enabled the development of the mathematical models of
the computer simulation model. The relationships of the components of the real world
DAP system were used to formulate the linkages of the computer model. For instance the
data obtained for the land size and the amount of land used for cropping were used to
compute the probable land size of the farms used for the simulations.

Other data that included the actual yield from the farms and the sale price of the
farm produce were also used to make estimates of the expected returns (net gross income)
to the farmer. It was noted that only a fraction of the produce was marketed. The main
concern for any farmer is to provide adequate food for their household. Hence, certain
quantities of both maize and beans (the staple foods) are retained for family use. This
was taken into consideration in the computation of the gross net returns.

Though most of the farmers had only one pair of oxen, those farmers with four
trained oxen had excess DAP working capacity that could have been utilized for rental
purposes to bring more income into the household. In particular, the low annual usage of
DAP means that there is untapped capacity which can be used for other non-farm
activities aﬁer the tillage season is over. There is potential for improving the utilization
of DAP-SYS in semi-arid agriculture of Kenya through extensive and intensive use of the
currently available energy source and resources at the farrners’ disposal.

7.1.3 Development of DAS
The DAS developed in collaboration with other researchers was a good prototype

that performed as well as expected. The mechanical sensors had been perfected through

 

156

earlier research done by researchers at Swedish University of Agricultural Sciences,
Uppsala, Sweden. The mechanical sensors worked very well but the physiological
sensors had to be handled carefully and checked regularly to ensure that data was being
received by the logger as required.

Although there were some practical difficulties with the physiological data
sensors, the data obtained enabled the researcher to make conclusive inferences about the
energy that the animals expend and the amount of work they do. Simultaneous
collection of mechanical and physiological data is the most suitable way of relating the
energetic capacity of draft animals and their actual work performance.

7.1.4 Implementation of DAP-SYS data acquisition system

The data obtained with the use of electronic DAS were useful in providing the
required estimates for the energy used for tillage and other activities.‘ The mechanical
data provided valuable estimates of the draft requirements in different soils and using
different size of draft animals. The speed of operation and the actual amount of work
done per unit time (EF C) were used for the modeling program to maintain real ﬁeld
situations in making simulations.

The design and the construction of a tailor-made electronic DAS for mechanical
and physiological data of DAP-SYSs was successful. This instrumentation package and
variations of it are currently being used for research in Kenya by the University of
Nairobi. There is need for improving the way the physiological sensors are connected to
the animal particularly the heart rate sensor, the temperature probe and the breathing rate

sensor. The heart rate was the most important physiological parameter monitored. Since

 

 

 

157

getting the best position in order to obtain a reading was very challenging, it will be
necessary to improve the sensing mechanism of the heart rate sensor.
7.1.5 Development and implementation of EUMDAP simulation model

A computer simulation model was developed and tested. The EUMDAP
simulation model performed well and was demonstrated through simulations of various
scenarios of DAP-SYSs.

This version of the program has the capacity of being improved for greater
versatility and use by a wider range of users and not restricted to semi-arid draft animal
power. The Energy Utilization Model for Draft Animal Power provided useful data on
the simulation of working draﬁ animal power systems in various ﬁeld scenarios and with
different sizes of draft oxen. The program is DOS (disk operating system) based and
operates as a stand-alone program which makes it easier to apply and be used with
minimum computer skills and expertise. In order to have the program provide the user
with the required simulations, the program requests operational and resources data from
the user. These data include the size of the farm, the resources available from the farmer,
the potential feed and the operational variables that include the time the animals are used
for tillage and transportation. The costs involved in DAP-SYS management and the cost
of inputs for the crop production are also requested for in order to estimate the expected
returns to the farmer.

The user has the option of saving the data in a ﬂoppy disk or in a computer.
Alternatively, print-outs of the input and output data are provided at the end of the
simulation. The user is furnished with DAP-SYS performance data including the

potential capacity for using, the draft animals to log more working time (days per season).

 

 

158

7.2 Conclusions

This research project set out several objectives which were fulﬁlled in varying
degrees of success. The ﬁrst objective was to develop, design and build an electronic
DAS that would collect simultaneously both physiological and mechanical data from
working DAP-SYSs. This exercise used the expertise from several sources including
the work done at the Silsoe Research Institute, Bedford, England and the on-going

project work at the Swedish Agricultural Engineering Department in Uppsala,

 

Sweden. A working prototype of DAS was built to project speciﬁcations and was
successfully used to obtain useful data used for validating the EUMDAP simulation
model developed. Improvements of the working model will be a continuous process
as the need for this type of applied and basic research is gaining momentum.

The data collected from the farmers’ ﬁeld with both the DAS and the questionnaire
was useful in developing and validating of the EUMDAP computer simulation
program. The computer program was tested for validity using the ﬁeld data obtained
with both the DAS and the questionnaire. The program will be utilized for simulations
of DAP-SYSs in the semi-arid areas of Kenya for increasing utilization of DAP as a
power source. The physiological data obtained was only useful for the energy
utilization estimates using the heart rate of the working oxen. The other three
physiological variables were not used. However, the mechanical data obtained
provided a basis on which the simulated output of the EUMDAP program was
compared for validity in the semi-arid areas setting.

Baseline data collection was done for a total of 80 farmers in ﬁve administrative

divisions in two Districts in semi-arid Kenya. The objective was to obtain sufficient data

 

 

159

that would be used for the parameter and variable estimates to be used for the computer
simulation model. The data obtained were useful in the somputer model development
which was implemented successfully and used for simulations that monitored the energy
utilization of draﬁ oxen for semi-arid areas.

The ultimate objective was to develop and implement a working computer simulation
model for DAP-SYS energy utilization. The model (code-named EUMDAP) was
successfully developed and implemented. It is available for use in extension and research
services in DAP mechanization in semi-arid areas with the potential for expanding for
other areas where DAP utilization is prevalent. In the process of implementing the
model, various scenarios were simulated that included sole use of DAP by owner-
operator and owner-operator/hiring out to other farmers.

7.3 Limitations of the study and future research

The research work done for this study was limited to obtaining authentic data for
validating a computer simulation program that was developed. The study did not address
many issues that farmers grapple with for their mechanization needs. Some of the issues
include the need for suitable implements that would enable the farmers to engage the
DAP-SYSs before the onset of rains in order to prepare their land in time for timely
planting. Due to constraints of time, adequate data for weeding operations and
transportation could not be collected. Weeding is a major consideration when farmers
consider expanding their cropped land size. Transportation is also an important factor in
expanding the use of DAP after tillage operations are done. The instrumentation used for

the data collection did not include measurements for transportation energy use.

 

 

 

160
Since the government of Kenya has realized the importance of DAP-SYS in small-

holder agriculture, it is time that the opportunity was utilized to advance the cause of
DAP mechanization systems. This research exercise was unique and paves the way for
similar work in the future. Primary data in DAP-SYS is hard to come by as the
researcher experienced in the process of obtaining authentic data for Kenayan DAP-SYS.
Although other researchers have performed similar type of data collection on DAP-SYS,
this was the ﬁrst time that such experiments were conducted in Kenya. Building on the
experience gained from this exercise, it is recommended that the following be undertaken
in order to strengthen and augment the current eforts in DAP-SYS research:

0 perform more extensive research with wider spectrum of farmers in the semi-arid
regions of the country.

0 conduct more controlled experiments on the feeding regime of draft oxen in order to
establish the energy relationships of current local feeds and energy demand of draft
oxen.

o carry out similar experiments in other high potential provinces of Kenya where DAP
is used. .

o perform further research on the harnessing mechanism and the effect of the same on
transmission of draft animal power implements.

0 develop a national data base for DAP-SYS and make it available to researchers in this
area.

0 the DAS can be used for a variety of research undertakings including monitoring of
better harnessing systems to improve the power delivery to implements.

 

APPENDIX A

 

 

161

APPENDIX A

Appendix A provides empirical values for factors used for the computation of extra
energy used by draft oxen at work. These factors were obtained from various sources as
shown in the comments column. The numerical values of the factors include the standard
deviation and the actual value used for the energy data computation. Application of the
factors in the simulation model was done in equation (1) as shown in section 2.5. The

four factors included:

M M
A Joules per kg live weight per m traveled

B Joules per kg carried per m traveled

C Ratio of work done raising body weight per energy used.

Empirical Values for factors used to calculate the extra energy consumption of
draught animals for work. Adapted from Lawrence (1985)

 

Factor and units Numerical Type of Comments
value i S. E. animal

A: Joules per kg live 2.09 i 0.062 Brahman Walking speed range

 

 

 

weight per m traveled cattle 0.4 - 1.6 m/sec

“ 2.0 Cattle
B: Joules per kg carried 2.60 i 0.19 Cattle Load placed on saddle
per m traveled over animal’s shoulders
C: Ratio work done 0.298 $0.006 Brahman Data for animals in
pulling/energy used Cattle double and single yokes

+ a few for single
animals with collars

 

D: Ratio work done 0.35 Cattle A.R.C. (1980)
raising body
weight/energy used

 

 

APPENDIX B

 

Appendix B provides the caloriﬁc equivalent of the respiratory exchange ratios (RER)
and the relative percentages of kcal of carbohydrates and fats. The energy release from
the food substrates depends on the relative combinations of carbohydrates and fats. The
values of RER ranges from 0.71 to 1.0. The lower end of the RER scale (0.71) represents
the energy released when 100% fats are burned in oxygen releasing 4.69 kcal of energy
per liter of oxygen consumed whereas the RER ratio of 1.0 represents the energy released
when 100% of carbohydrates are burned in oxygen releasing 5.05 kcal of energy per liter

of oxygen consumed.

Caloriﬁc Equivalent of the Respiratory Exchange Ratio and % kcal from CH0 and

162

APPENDIX B

Fats. Adapted from Wilmore and Cost!!! (1 994)

 

 

 

 

Energy % kcal
Respiratory
Exchange kcal . L"Oz Carbohydrates Fats
Ratio
0.71 4.69 0 100
0.75 4.74 15.6 84.4
0.80 4.80 33.4 66.6
0.85 4.86 50.7 49.3
0.90 4.92 67.5 32.5
0.95 4.99 84.0 16.0
1.00 5 .05 100.0 0

 

 

 

 

 

APPENDIX C

'1

 

 

163

APPENDIX C

Appendix C gives technical details for the RAMLOG EI-9000 data logger. These data
were obtained from the operation manual of the logger and serves as a guide in
determining the number and type of channels that the logger uses for data collection.
Operational parameters including the range of the power supply and the memory capacity

of the mini-computer of the logger are provided for quick reference.

RAMLOG EI-9000 Data Logger Technical Details

Inputs: 16 analog single ended inputs or 8 differential inputs.
2 counters
8 digital inputs

Memory: 128 kb RAM in standard version.
77,500 measurements (analog - 12 bits) for 128 kb version.

Analog inputs: Individually programmable scales for every input.
Scales: i100mV, i200mV, i500mV,i1V,i2V,i5V,i1 0V.
Resolution i200 points.
Calibration error: 0.1% F S
Over-voltage protection: max i200 V DC
Maximum common mode voltage (differential inputs): i35V DC
Input impedance: > IOOMQ
Input numbers: single ended - channel 1 to 16
differential - channels 1 to 8

Counter inputs: Choice - two 8 bits counters (standard) or one 16 bits counter
(option)
Maximum frequency - 1 MHz
Counter range - 255 (8 bits) or 65,535 (16 bits)
Signal 'ITL/CMOS level, Vmax = 5V
Input numbers - channels 17 and 18 (8 bits), channel 19 (16 bits)

Digital inputs: TTL/CMOS level, Vmax = 5V
Input number - channel 20 binary values stored over 8 bits (10 to
C17

Real time clock: Maximum drift 2 sec/day at 20 °C.

 

 

 

Sampling frequency:

Programmable range:

Display:

Power supply:

Battery econmy:

Connections:

Temperature range:
Humidity:

Case:

Dimensions:

Weight:

164

100 Hz on 1 channel
10 Hz on 16 channels.

i100 mV to i10 V

LCD 2 lines of 16 characters.
Engineering units before, during and after recording.

Working range - 4.2 V to 7 V

4 alkaline batteries type LR6 (AA)

Option 4 NiCd accumulators

External power supply connector 6.7 V DC 200 mA

“STANDBY” current drain - 50 uA
Auto switch to “STANDBY” mode if:
0 > 30 seconds of keyboard inctivity
0 between measurements in recording mode.

Communication - Dsub 9 pins connector

Measure - Dsub 37 pins connector

External sensor power supply - Dsub 37 pins connector (pin 4)
- 10°C to + 60°C

0 to 95% RH no condensation.

Plastic ABS

180 x 100 x 45 mm

500 grams (batteries included)

 

 

Kistler Accelerometer

165

APPENDIX D

The accelerometer used for the measurement of the step rate of the working animals

was a commercially available sensor manufactured by Kistler. Appendix D provides the

technical details of the sensor. These details include the operational range as well as the

sensitivity and the type electrical supply and the current outputs.

Type:

Range:
Sensitivity(i5%):
Frequency range (i5%):
Mass without cable:

Operating temperature range:

Linearity:
Supply current:
Output:
Voltage (full scale)

Bias voltage (15%):

Impedance:

Kistler Accelerometer

8308A K-BEAM
g 21:10
mV/g 100
Hz 0 - 150
g 10
°C -20 to 85
% FSD <°i2
mV 9 - 20
V :1
VDC 2.
Q 10

 

 

 

166

Appendix E

The soil moisture content was monitored with an electronic sensor that measured the

soil moisture content in situ. Appendix B provides the technical details of the Time

Domain Reﬂectometry

instrument. The technical details include the range of operation,

the power supply, memory capacity and the electronic partiCulars of the instrumentation.

Model:

Measuring range:
Measuring accuracy:
Operating temperature:

Power supply:

Connecting ports:

Memory:

Time Domain Reﬂectometry Instrument

6050Xl Trace System

0 - 100% Volumetric moisture content
i 2% or better

0 - 45 °C

2 each 6 amp-hr, sealed, gelled electrolyte batteries

Supply recharge time, 9 hours.

Auxiliary power input 18-24 volts AC or DC amp, for battery
recharge or independent operation.

External battery input: 12 volts for independent operation.

BNC port for wave guide connection.

RS 232 serial port for data transfer.

Multiplex port - 15 pin D-SUB for sequence
Switch access for unattended logging of multiple sites using
external equipment.

Power port - 8 pin DIN

Standard 256 kb memory board with storage for at least 166
graphs/6000 readings.
Optional 256 kb memory board additions to maximum of 1 Mb
storage.
Automatic data tagging of reading time/date/reading number plus
use deﬁned label.
Autologging capability with reading interval range from one day
to one per minute.

. :r-lb

 

 

167

Electronic particulars: Measuring pulse rise time less than 200 picoseconds.
Sampling resolution, 10 picoseconds.

Graphic display - 128 x 256 dot super twist matrix, back-lighted
LCD.

Battery charge status indicator light (3)
Circuit breaker protection.

Hardware - 8 slot card cage construction, 4 system boards, 4
optional slots.

 

APPENDIX F

168

APPENDIX F

EUMDAP Computer Program Listing

COLOR 7, 1, 1

CLS

LOCATE 9, 2

LOCATE 10, 8

PRINT

PRINT " ENERGY UTILIZATION MODEL "
PRINT " FOR"

PRINT " DRAFT ANIMAL POWER-SYSTEMS"
PRINT

PRINT" (EUMDAP-SYS)"

LOCATE I9, 24

PRINT "PRESS ANY KEY TO CONTINUE";

1 a3 = INKEYS: IF a3 = "" THEN I

CLS

LOCATE 10, 10

PRINT " This program simulates the utilization of draft"
PRINT " animal power (DAP) system in "
PRINT " small-holder agriculture of"

PRINT " Semi-Arid areas of Kenya."
LOCATE 20, 24

PRINT "PRESS ANY KEY TO CONTINUE";
2 a3 = INKEYS: IF a3 = "" THEN 2

CLS

LOCATE 10, 10

PRINT " Three DAP Systems are used for the simulations:"
PRINT

PRINT " 1. Sole Owner-operator of DAP System."
PRINT " 2. Owner-operator plus renting out of DAP System. "
PRINT" 3. None DAP System owners renting oxen."
PRINT

LOCATE I9, 24

PRINT "PRESS ANY KEY TO CONTINUE";

3 a8 = INKEYS: IF a8 = "" THEN 3

CLS

LOCATE 10, 5

PRINT " The following categories of information are required from the user"
PRINT " in order to compute the utilization data of the DAP systemz"
PRINT

PRINT " 1. Land and soil parameters 4. Draft oxen usage data"
PRINT " 2. Ox feed quality and quantity 5. Operational parameters "
PRINT " 3. Implement type and size"

PRINT

PRINT " After each entry press the <ENTER> key."

PRINT

LOCATE 20, 14

PRINT " PRESS ANY KEY TO CONTINUE";

4 a3 = INKEYS: IF a5 = "" THEN 4

 

 

I69

 

' LAND AND SOIL PARAMETERS
CLS
100
PRINT
LOCATE 12, 15
INPUT " What is the size of your farm in acres? ", Land
CLS
LOCATE 10, 10
1F Land > 20 OR Land < 0 THEN GOTO 10
INPUT " How much of this land is under crop (acres)? ", Crand
IF Crand > Land THEN
10
BEEP

 

CLS

LOCATE 8, 5

PRINT "OUT OF RANGE: Cropped land is more than available land."
LOCATE 9, 5

PRINT " Or Land should be 20 acres or less."

PRINT " Try again."

PRINT

GOTO 100

END IF

Land! = Land! l2.47

Crand! = Crand! / 2.47 'acres to hectares

200 CLS

LOCATE IO, 10

PRINT " When do you start tillage? Choose from below (I to 3)"
PRINT

PRINT " 1. As soon as the rains come"
PRINT " 2. Soon after harvest"
PRINT " 3. Before the rains"
INPUT " Tillage Start = ", TiISt$
IF TilSt$ < "I " OR TilStS >= "4" THEN
BEEP
PRINT
PRINT" Try again. Enter 1 to 3"
GOTO 200
END IF
CLS

300 LOCATE 10, 10
PRINT " What type of soil do you have? Choose from 1 to 4."

PRINT " 1. Silty Clay"
PRINT " 2. Clay Loam"
PRINT" 3. Sandy Loam"
PRINT " 4. Black Cotton"
INPUT " Soil Type =", SolTyS
PRINT
IF SolTyS < "1" OR SolTy$ >= "5" THEN

BEEP

LOCATE 8, 20

PRINT "Try again, Enter 1 to 4."
GOTO 300

 

 

I70

END IF

IF TilSt$ = "I" AND SoITyS = "1" THEN SoilStr = 15
IF TilStS = "I " AND SolTy$ = "2" THEN SoilStr = 20
IF TiIStS = "I " AND SolTyS = "3" THEN SoilStr = 30
IF TilSt$ = "I " AND SoITyS = "4" THEN SoilStr = 40

IF TilStS = "2" AND SolTyS = "l " THEN SoilStr = 20
IF TilSt$ = "2" AND SolTyS = "2" THEN SoilStr = 25
IF TilStS = "2" AND SolTyS = "3" THEN SoilStr = 35
IF TilStS = "2" AND SolTyS = "4" THEN SoilStr = 45

IF TilStS = "3" AND SoITyS = "1" THEN SoilStr = 25
IF TilStS = "3" AND SolTyS = "2" THEN SoilStr = 35
IF TiIStS = "3" AND SolTyS = "3" THEN SoilStr = 40
IF TilStS = "3" AND SolTyS = "4" THEN SoilStr = 50

' DRAFT OXEN
CLS
400 LOCATE 10, IO
INPUT " How many trained oxen do you have (2 or 4)? ", Oxen
IF Oxen = 2 OR Oxen = 4 THEN GOTO 500
IF Oxen < 2 OR Oxen > 4 OR Oxen = 3 THEN
BEEP
CLS
LOCATE 8, 22
PRINT " TRY AGAIN"
GOTO 400
END IF

 

 

500
LOCATE 10, 10
PRINT "How many years have you used the current oxen (I to 3 years)"
PRINT
INPUT " Level of training =". OxTr
IF OxTr <= 0 OR OxTr >= 4 THEN
BEEP
CLS
LOCATE 8, 20
PRINT "Try again. Enter between I and 3 years"
GOTO 500
END IF

CLS

600

LOCATE 10, 6

INPUT " Do you have the weights (kg) of the oxen (Y or N)"; WtOx$
1F WtOxS = "Y" OR WtOxS = "y" THEN GOTO 610
IF WtOxS = "N" OR WtOxS = "n" THEN GOTO 620
BEEP
CLS
LOCATE 8, 50
PRINT "Enter Y or N "
GOTO 600

610 CLS

 

 

 

171

IF Oxen = 2 THEN GOTO 611
IF Oxen = 4 THEN GOTO 612
611
LOCATE 12, 10
PRINT
PRINT " Enter weight (200 to 400 kg) and age of each oxz"
PRINT
INPUT " Weight of 0x1 =", WtOxI
INPUT " Weight of 0x2 =", WtOx2
PRINT
IF WtOxI < 200 OR WtOxl > 400 OR WtOx2 < 200 OR WtOx2 > 400 THEN
CLS
BEEP
LOCATE 9, 7
PRINT "Ox weight must be more than 200 kg and less than 400 kg"
PRINT
PRINT " TRY AGAIN"
GOTO 61 I
END IF

613
INPUT " Age of 0x1 =", AgerI
INPUT " Age of 0x2 =", AgerZ
IF Agerl <= OxTr OR Ager2 <= OxTr THEN
BEEP
CLS
LOCATE 9, 10
PRINT "Age of oxen MUST be more than years of training (Training level)"
GOTO 613
END IF
GOTO 700

‘weight data for 4 oxen

612

LOCATE 10, 10

PRINT " Enter weight (kg) and age of each oxz"
INPUT " Weight of 0x1 =", WtOxl
INPUT " Weight of 0x2 =", WtOx2
INPUT " Weight of 0x3 =", WtOx3
INPUT " Weight of 0x4 =", WtOx4

IF WtOxl < 200 OR WtOxI > 400 OR WtOx2 < 200 OR WtOx2 > 400 THEN
BEEP

CLS
LOCATE 7, 10
PRINT "Ox weight must be more than 200 and less than 400"
PRINT " TRY AGAIN "
SLEEP 3
GOTO 612
END IF
IF WtOx3 < 200 OR WtOx3 > 400 OR WtOx4 < 200 OR WtOx4 > 400 THEN
BEEP
CLS
LOCATE 7, 10
PRINT "Ox weight must be more than 200 and less than 400"

 

172

PRINT " TRY AGAIN"

SLEEP 3

GOTO 612
END IF
614
INPUT " Age of 0x1 =", Agerl
INPUT " Age of 0x2 =", Age0x2
INPUT " Age of 0x3 =", Ager3
INPUT " Age of 0x4 =", Ager4

IF Agerl <= OxTr OR Ager2 <= OxTr OR Age0x3 <= OxTr OR Age0x4 <= OxTr THEN
BEEP
LOCATE 9, 10
PRINT "Age of oxen MUST be more than years of training (Training level)"
SLEEP 2
GOTO 614
END IF
GOTO 700

620 CLS
IF Oxen = 2 THEN GOTO 615
IF Oxen = 4 THEN GOTO 616

'measurements of 2 oxen

615

LOCATE 10, 8

PRINT " Enter the measurements of the 2 oxen in cm:"

PRINT

PRINT " (Heartgirth should be AT LEAST 125 cm)"

PRINT" (Length should be AT LEAST 145 cm)"

LOCATE 14, 10

INPUT "0x1 Heartgirth and length (HeartGirth, Length) =", HrtGrOxI, LntOxI
LOCATE 16, 10

INPUT "OX2 Heartgirth and length (HeartGirth, Length) =", HrtGrOx2, LntOx2
WtOxl = (HrtGrOxl /2.54) A 2 * (LntOxl /2.54) / 660

WtOx2 = (HrtGrOx2 / 2.54) A 2 " (LntOx2 / 2.54) / 660

GOTO 613

'measurements of 4 oxen

6 I 6

LOCATE 12, 10

PRINT "Enter the measurements of the 4 oxen in cm:"

PRINT " Heartgirth should be at least 125 cm"

PRINT " Length should be at least 145 cm"

INPUT "0x1 Heartgirth and length (HeartGirth, Length) =", HrtGrOxl, LntOxI
INPUT "0x2 Heartgirth and length (HeartGirth, Length) =", HrtGrOx2, LntOx2
INPUT "0x3 Heartgirth and length (HeartGirth, Length) =", HrtGrOx3, Lnt0x3
INPUT "0x4 Heartgirth and length (HeartGirth, Length) =", HrtGrOx4, Lnt0x4
WtOxI =(HrtGrOx1 /2.54) A 2 "' (LntOxl / 2.54) / 660

WtOx2 = (HrtGrOx2 / 2.54) A 2 " (LntOx2 / 2.54) / 660

WtOx3 = (HrtGrOx3 / 2.54) A 2 "‘ (LntOx3 /2.54) / 660

WtOx4 = (HrtGrOx4 / 2.54) A 2 “ (Lnt0x4 / 2.54) / 660

GOTO 614

 

 

I73

FEED FOR THE OXEN .................

 

700 CLS
LOCATE 10, 10
PRINT " What is the Primary feed for the draft animals?"
PRINT " Choose from I to 3"
PRINT " 1. Maize stover"
PRINT " 2. Cut grass (Hay)"
PRINT" 3. Natural pasture"
INPUT " Primary Feed =", PrdeS
IF PrdeS <= "0" OR Prde$ > "3" THEN
BEEP
PRINT" Try again. Enter 1, 2, or 3"
GOTO 700
END IF

IF Prde$ = "1" THEN ME = 7.5
IF Pde8 = "1" THEN DMc = .9
IF PrdeS = "2" THEN GE = 91

1F PrdeS = "2" THEN DMc = .85
IF PrdeS = "3" THEN GE = 11
IF PrdeS = "3" THEN DMc = .75

OPERATIONAL INPUTS

 

 

800 CLS
LOCATE 10, 10
PRINT " How many days do you use oxen for tillage per week?"
PRINT
INPUT " Number of days per week? =", Nde%
IF Nde% <= 0 OR Nde% > 7 THEN

BEEP

LOCATE 13, 15

PRINT" Out of Range. Try again."

INPUT " Days worked per week (I to 7)? =", Nde%
END IF
CLS

LOCATE 12, 10

PRINT "
PRINT
INPUT "

How many hours do oxen work on tillage per day?"

Number of hours per day =", HrDy!

IMPLEMENT USAGE

 

 

CLS

900 LOCATE 10, 10

PRINT "
PRINT "
PRINT "
PRINT "
PRINT "
INPUT "

Which of the following implements do you use?"
Choose one combination:"
1. Moldboard plow only"
2. Moldboard plow and Ox-Cart "
3. Ox-Cart only"
Type of implements ="; Impl$

IF Impl$ = "1" OR ImplS = "2" THEN GOTO 910
IF ImplS = "3" THEN GOTO 920
IF ImplS <= "0" OR lmpl$ > "3" THEN

 

 

I74

BEEP

CLS

GOTO 900
END IF

910 CLS
LOCATE 10, 14
PRINT "Select depth of tillage:"

PRINT " 1. Shallow (.10 m)"
PRINT " 2. Medium ( .15 m)"
PRINT " 3. Deep (.20 m)"
INPUT" Depth of tillage =", Dpth

IFDpth= 1 THEN Dpth=.l
1F Dpth =2 THEN Dpth = .15
IF Dpth = 3 THEN Dpth =2

IF Dpth < 0 OR Dpth > 3 THEN GOTO 1010 ELSE 1100
1010

BEEP

LOCATE 10, 14

PRINT "Depth is out of range!"

I 100 CLS

LOCATE 10, 14

PRINT "Select width of tillage:"

PRINT" 1. Narrow (.25 m)"
PRINT " 2. Medium (.30 m)"
PRINT " 3. Wide (.33 m)"

INPUT " Width of tillage =", Wdth

IF Wdth = I THEN Wdth = .25
IF Wdth = 2 THEN Wdth = .3
IF Wdth = 3 THEN Wdth = .33

IF Wdth < 0 OR Wdth > 3 THEN GOTO 1120 ELSE 1200
1120 BEEP
LOCATE 10, 14
PRINT "Width is out of range!"
GOTO 1100
IF lmpIS = "2" THEN GOTO 920
920

1200 CLS
LOCATE 10, I4
INPUT "Use default tillage operation speed (Y or N)"; DefSpdS
IF Def‘SpdS = "Y" OR DefSpdS = "y" THEN
Spd = .75
GOTO 1300
END IF
lF DefSpdS = "N" OR DefSpdS = "n" THEN GOTO 1220
BEEP
PRINT "Enter Y or N "
GOTO 1200
1220
CLS
LOCATE 10, I4

 

 

175

INPUT "Speed of operation (0.5 to 1.5 m/s) =", Spd
IF Spd < .5 OR Spd > 1.5 THEN

BEEP

CLS

LOCATE 8, 10

PRINT "Try again"

GOTO 1220
END IF

1300 CLS
LOCATE 10, 10
INPUT "Days per week of household transportation with DAP =", TrDys

INPUT " Number of weeks of transportation per season =", WksTr

CLS

LOCATE 10, 10

INPUT "Distance (km) per day of household transportation =", Ter

INPUT " Pay load per trip of transportation (tons) = ", PyLd

DstT = WksTr * TrDys "‘ TrKrn * 1000 'Distance per season m

'HseLoad = DstT * PyLd “ 9.81 "' 1000/ 1000 'Annual transport energy, MJ

IF OxTr = 1 THEN Spd = Spd “ .95
IF OxTr = 2 THEN Spd = Spd “ .85
IF OxTr = 3 THEN Spd = Spd "‘ .8

 

' COSTS of DAP SYSTEM USAGE-----—-----
CLS

leCst = 3500

PICstSn = leCst - (.l "‘ leCst)/ 10 “ 2 '10% salvage value after 10 years
OthCst = 4000

OthCstSn = OthCst - (.I "‘ OthCst)/ 10 " 2 '

LOCATE II, 10

INPUT "Feeding cost of each oxen per season =", F dngtOx
ngCstOx = F dngtOx * Oxen

LOCATE 12, 10

INPUT "DAP veterinary costs per draft oxen per season = ", VetCst
VetCst = VetCst "' Oxen

LOCATE I3, 8

INPUT "DAP operator cost per season (KSh) = ", DapOpCst

CLS

LOCATE 8, 8

INPUT "Implement maintenance cost (Ksh) per season = ", ImpCst
LOCATE 9, 8

INPUT "Hand Weeding cost per acre (Ksh)=", deCst

deCst = (deCst / 2.47) ‘ Crand 'weeding cost per ha
LOCATE 10, 8

SdFtCst = 2500

SdFtCst = (SdFtCst / 2.47) "‘ Crand

LOCATE 12, 8

INPUT "Planting cost per acre per season (Ksh)=", PIngt

Plngt = (Plngt / 2.47) "' Crand

LOCATE I4, 8
INPUT "Harvesting cost per acre per season=", HvstCst
HvstCst = (HvstCst / 2.47) "' Crand 'cost per ha

'Seasonal returns

SanCst = VetCst + DapOpCst + ImpCst + deCst + SdFtCst + Plngt + HvstCst

 

176

SanCst = leCstSn + OthCstSn + ngCstOx 'plow dep; oxcart dep; Oxenfeeding
TtCst = SanCst + SanCst

' INCOME - RETURNS ------

 

ManYld = 15
SthzYId = 3
ManYld = 6
StdBnYld = 1.5
MzYId = (ManYld + (RND * SthzYId))
BnYld = (ManYld + (RND(O) * StdBnYld))
IF TiIStS = "3" THEN
MzYld = MzYld * .9
BnYld = BnYId * .9
END IF
IF TilStS = "2" THEN
MzYld = MzYld '90 kg bags
BnYld = BnYld
END IF
IF TiIStS = "1" THEN
MzYId = MzYld " .8 '20% reduction in yield
BnYld = BnYld " .8
END IF
TMzYld = Crand * .7 * MzYld '90 kg bags of maize on 75% of land
TBnYld = Crand * .3 “ BnYId '90 kg bags of beans
Mle = TMzYld - 10
BnSl = TBnYld - 2
CrpInc = (MzSI "' 1800) + (BnSl "‘ 5000) 'gross income per season
HirInc = HirAc * 600 ‘income from rental plowing
TotInc = CrpInc + HirInc 'gross total income
GrsNthn = TotInc - TtCst

'------------HIRING OUT AND RENTING IN OF DAP-SYSTEM
CLS
8000
LOCATE 12, 10
INPUT "Is DAP hired out to other farmers (Y or N) ="; HirOutS
IF HirOutS = "Y" OR HirOutS = " " THEN GOTO 8600
IF HirOut$ = "N" OR HirOut$ = "n" THEN GOTO 5000
CLS
BEEP
LOCATE 8, 10
PRINT " Enter Y or N "
GOTO 8000

8600 CLS

LOCATE 11, 10

PRINT "Enter 'zero (0)' if no hiring out done"

CLS

LOCATE 12, 10

INPUT "Acreage (acres) plowed for hire per season =", HirAc
HirAc = HirAc / 2.47

CLS
LOCATE 10, 10
PRINT

 

I77

INPUT" Hired out transportation distance (km per week) on hire =", Hier
HrDstT = Hier * 1000

5000 CLS

LOCATE 10, 10

INPUT "Do you hire DAP System from others (Y or N)? ", Hir$
IF HirS = "Y" OR HirS = "y" THEN GOTO 5100

IF Hir$ = "N" OR Hir$ = "n" THEN GOTO 6100

BEEP

GOTO 5000

5100 CLS

LOCATE 10, 10

INPUT "How much land do you plow with hired oxen (acres)? =", HirAcPl
HirAcPl = HirAcPl / 2.47

 

6100

' 1. ENERGY DEMAND FROM OXEN

CLS

TDpth = Dpth ‘ 100

TWdth = Wdth "‘ 100

'999

TCrand = Crand + HirAc

k = 90 'k = % of implement width used

TFC = Spd "' Wdth " 3.6/ 10 'TFC = theoretical ﬁeld capacity (ha/hr)
Tt = 1 / TFC 'Tt = Theoretical ﬁeld time (hr/ha)
Te = 100 “ Tt / k 'Te = Effective ﬁeld time (hr/ha)
Tlos = .08 "' Tt 'Tlos = total time lost (8% of Tt)
FE = Tt / (Te - Tlos) 'FE = Field efficiency (dec)

EFC = Spd * Wdth "‘ FE * 3.6/ 10 'EFC = Effective ﬁeld capacity (ha/hr)"
ATimTil = TCrand / (EF C " HrDy) 'TimTil = Time to complete tillage (days)
Drft = Wdth ‘ Dpth "' SoilStr 'kN

Pwr = Drtt " Spd 'kW = (kN "‘ m/s) /1000

ImplEn = Pwr / EFC 'ImpEn = speciﬁc energy in kW-hr/ha

TrsEn = HseLoad * 9.81 / 1000 ’transport energy per year, MJ

'---------2. DAP-SYS ENERGY SUPPLY & PERFORMANCE

IF Oxen = 2 THEN GOTO 1400
IF Oxen = 4 THEN GOTO 1500

1400 'Two oxen teams
I'Energy from Oxl

Fql = .025 " WtOxI 'Fq = Voluntary dry matter intake (kg of Dry Matter per day)
'IF PrdeS = "3" AND TilStS = "1" OR TilStS = "3" THEN
'ME=GE"' .45‘ .8" .85
' ELSE
' IF PrdeS = "1" OR Prde$ = "2" AND TiIStS = "2" THEN
' ME=GE"'.45*.8
' END IF
'END IF
MEMI = Fql " ME " DMc 'Max Metabolizable energy per day; DMc=Dry matter content (dec)
Km = .019 ‘ ME + .53 'Km = efficiency of utilization (ARC standard)
NEme = MEMI " Km 'NEmw = net energy for maintenance and work per day
BMRI = (70.5 "' WtOxI A .75 "' 4.184 "' EXP(-.03 * Age0x1))/ 1000 'Basal Metabolic rate MJ/day

 

 

178

NEW] = NEmwl - BMRI 'NEw = net energy for work, MJ/day
'Energy from 0x2
Fq2 = .025 * WtOx2 'Fq = Voluntary dry matter intake (kg of Dry Matter per day)

'IF PrdeS = "3" AND TilSt$ = "1" OR TilStS = "3" THEN
'ME=GE"‘.45" .8“ .85
' ELSE
' IF Prde$ = "1" OR PrdeS = "2" AND TiIStS = "2" THEN
' ME=GE"' .45" .8
‘ END IF
'END IF
MEm2 = Fq2 " ME "' DMc 'Max Metabolizable energy per day; DMc=Dry matter content (dec)
Km = .019 " ME + .53 'Km = efﬁciency of utilization (ARC standard)
NEmw2 = MEm2 ‘ Km 'NEmw = net energy for maintenance and work per day
BMR2 = (70.5 " WtOx2 A .75 * 4.184 " EXP(-.03 * Ager2))/ 1000 'Basal Metabolic rate MJ/day
NEw2 = NEmw2 - BMR2 'NEw = net energy for work, MJ/day
TBMR = BMR] + BMR2

'total energy of two oxen

TNEmw = NEme + NEmw2 ‘total net energy for team/day: MJ/day
TMEm = MEMI + MEm2 'max ME/day

TNEw = NEW] + NEw2 'net energy for work/day

TNEw = TNEw

TDapCap = TNEw / 3.85 'capacity in hrs of work per day at 3.85 MJ/hr

TATimTil = TCrand / (EFC "‘ TDapCap) 'actual time to ﬁnish work: days
’Wks = TATimTil / Nde%
Owar =((WtOx1 + WtOx2) ‘ 9.81 * .I * .925 " Spd)/ 1000'power kW from two oxen

Etot = EF C * 10000 “ SoilStr "‘ Wdth "' Dpth "‘ TDapCap / (Wdth " 1000) 'total work done per day MJ
GrsEf = (Etot / TNEmw) "' 100 'gross efﬁciency %

NetEf = (Etot / (TNEmw - TBMR)) “ 100 'net efﬁciency %

DstDy = EFC “ TDapCap "‘ 10000 / Wdth 'distance traveled per day working, m
Oth = WtOxl + WtOx2

EnWlk = Oth * 2! * DstDy/ 1000000 'energy for walking for both oxen, MJ
AbsEf = (Etot / (TNEmw - EnWlk)) " 100 'absolute efﬁciency %

FdMn = TBMR "' 180/ (DMc * ME) 'seasonal maintenance feed needs kg

Fde = TNEw " TATimTil / (DMc "' ME)

'Trst = HseTrLd / (DMc "‘ ME)

Grstqu = FdMn + Fde + Trst

GrsEand = Grstqu “ DMc " ME

PothT = GrsEand / TNEmw 'potential work time, days

GOTO 3600
1500 'F our oxen team
'energy from Oxl

Fql = .025 " WtOxI 'Fq = Voluntary dry matter intake (kg of Dry Matter per day)
'IF Prde$ = "3" AND TilSt$ = "1" OR TilStS = "3" THEN
'ME=GE"' .45" .8“ .85
'ELSE ,
' IF Prde$ = "I " OR PrdeS = "2" AND TilSt$ = "2" THEN
' ME=GE" .45‘ .8

,_ -7 .mu—u‘ ...a-

 

 

I79

‘ END IF
'END IF
MEMI = Fql " ME "‘ DMc 'Max Metabolizable energy per day; DMc=Dry matter content (dec)
Km = .019 * ME + .53 'Km = efficiency of utilization (ARC standard)
NEmwl = MEMI * Km 'NEmw = net energy for maintenance and work per day
BMRI = (70.5 " WtOxl A .75 “' 4.184 * EXP(-.03 " Agerl))/ 1000 'Basal Metabolic rate MJ/day
NEwl = NEme - BMRI 'NEw = net energy for work, MJ/day

'Energy from 0x2

Fq2 = .025 "' WtOx2 'Fq = Voluntary dry matter intake (kg of Dry Matter per day)
'IF PrdeS = "3" AND TilSt$ = "1" OR TilSt$ = "3" THEN
'ME=GE* .45 "‘ .8* .85
‘ ELSE
' IF PrdeS = "I " OR Prde$ = "2" AND TilStS = "2" THEN
' ME=GE* .45" .8
' END IF
'END IF
MEm2 = Fq2 * ME * DMc 'Max Metabolizable energy per day; DMc=Dry matter content (dec)
Km = .019 " ME + .53 'Km = efﬁciency of utilization (ARC standard)
NEmw2 = MEm2 "' Km 'NEmw = net energy for maintenance and work per day
BMR2 = (70.5 * WtOx2 A .75 "' 4.184 "‘ EXP(-.03 "‘ Age0x2))/ 1000 'Basal Metabolic rate MJ/day
NEw2 = NEmw2 - BMR2 'NEw = net energy for work, MJ/day

'computation from one half team

TtNEmw = NEmwl + NEmw2 'total MJ/day for team I

TtMEm = MEMl + MEm2 'max ME, MJ/day

TtNEw = NEW] + NEw2 'net work energy MJ/day

TtDapCap = TtNEw / 3.85 'number of hours per day

TtOwar =((WtOx1 + WtOx2) " 9.81 * .1 "' .925 '- Spd)/ 1000'power karom two oxen

'energy from 0x3

Fq3 = .025 “ WtOx3 'Fq = Voluntary dry matter intake (kg of Dry Matter per day)
'IF Prde$ = "3" AND TilStS = "1" OR TilSt$ = "3" THEN
'ME=GE" .45“ .8“ .85
' ELSE
' IF PrdeS = "I " OR PrdeS = "2" AND TiIStS = "2" THEN
' ME=GE"‘ .45" .8
' END IF
'END IF
MEm3 = Fq3 "‘ ME "‘ DMc 'Max Metabolizable energy per day; DMc=Dry matter content (dec)
Km = .019 "‘ ME + .53 'Km = efﬁciency of utilization (ARC standard)
NEmw3 = MEm3 "' Km 'NEmw = net energy for maintenance and work per day
BMR3 = (70.5 "‘ WtOx3 A .75 “ 4.184 "‘ EXP(-.03 "' Ager3))/ 1000 'Basal Metabolic rate MJ/day
NEw3 = NEmw3 - BMR3 'NEw = net energy for work, MJ/day

'Energy from 0x4

Fq4 = .025 * WtOx4 'Fq = Voluntary dry matter intake (kg of Dry Matter per day)
'IF Prde$ = "3" AND TilSt$ = "1" OR TilSt$ = "3" THEN
'ME=GE* .45"‘ .8‘ .85
' ELSE ME = GE * .45 "' .8
'END IF
MEm4 = Fq4 “ ME "' DMc 'Max Metabolizable energy per day; DMc=Dry matter content (dec)

 

 

180

Km = .019 * ME + .53 'Km = efﬁciency of utilization (ARC standard)

NEmw4 = MEm4 "‘ Km 'NEmw = net energy for maintenance and work per day

BMR4 = (70.5 "‘ WtOx4 A .75 " 4.184 * EXP(-.03 "‘ Ager4))/ 1000 'Basal Metabolic rate MJ/day
NEw4 = NEmw4 - BMR4 'NEw = net energy for work, MJ/day

'calculation for second pair of oxen

T2NEmw = NEmw3 + NEmw4 'total MJ/day for team 1

T2MEm = MEm3 + MEm4 'max ME, MJ/day

T2NEw = NEw3 + NEw4 'net work energy MJ/day

T2DapCap = T2NEw / 3.85

TZOwar = ((WtOx3 + WtOx4) * 9.81 * .1 “ .925 "‘ Spd)/ 1000'power kW from two oxen

'computation of totals

TBMR = BMR] + BMR2 + BMR3 + BMR4
TNEmw = TtNEmw + T2NEmw

TMEm = TtMEm + T2MEm

TNEw = TtNEw + TYNEw

TNEw = TNEw “ EXP(.08 * Spd)
TDapCap = TtDapCap + T2DapCap
TATimTil = TCrand / (EFC ‘* TDapCap)
Wks = TATimTil / Nde%

Owar = TtOwar + 120me

Fthn =(BMR1 + BMR2) * 180/ (DMc * ME) 'seasonal maintenance feed needs kg
F2dMn = (BMR3 + BMR4) * 180/ (DMc " ME)
FdMn = Fthn + F2dMn

Ftde = TtNEw " TATimTil / (DMc "' ME) 'per season
F2de = T2NEw * TATirnTil / (DMc " ME)

Fde = Ftde + F2de

'TrsEn = HseLoad "‘ 9.81 / 1000 'transport energy per year, MJ
Trst = TrsEn / (DMc "' ME)

Grstqu = FdMn + Fde

GrsEand = Grstqu * DMc * ME

3600 'general for both

Etot = EFC "' 10000 "‘ SoilStr * Wdth * Dpth " TDapCap / (Wdth "' 1000) 'total work done per day MJ
GrsEf = (Etot / TNEmw) ‘ 100 'gross efﬁciency %

NetEf = (Etot / (TNEmw - TBMR)) * 100 'net efﬁciency %

DstDy = EFC * TDapCap * 10000 / Wdth 'distance traveled per day working, m

'0th = WtOxl + WtOx2 + WtOx3 + WtOx4

EnWlk = Oth "' 2! "' DstDy/ 1000000 'energy for walking for both oxen, MJ

AbsEf = (Etot / (TNEmw - EnWlk)) " 100 'absolute efﬁciency %

'probability oif working days

IF TiIStS = "1" THEN SnTlDys = 10
IF TilStS = "2" THEN SnTlDys = 25
IF TilSt$ = "3" THEN SnTlDys = 20

PothT = GrsEand / TNEmw
ExCap = PothT - SnTlDys 'extra work potential in days
TFEnD = TCrand "‘ ImplEn 'total ﬁeld implement energy demand

 

181

‘ 4. PREPARATION OF DATA FOR PRINTOUT OF OUTPUT

 

Spdd = Spd "‘ 3.6
DMcc = DMc * 100
Dpthh = Dpth * 100
Wdthh = Wdth * 100
IF TilStS = "1" THEN TilSt$ = "'as soon as rains come’"

IF TilStS = "2" THEN TilSt$ = "'soon after harvest'"

IF TilStS = "3" THEN TilStS = ’"before the rains'"

IF SolTy$ = "I " THEN SolTy$ = "'Silty clay."'

IF SolTy$ = "2" THEN SolTyS = "'Clay loam."'

1F SolTyS = "3" THEN SolTy$ = "'Sandy loam."'

IF SolTyS = "4" THEN SolTyS = "'Black cotton.”

IF PrdeS = "1" THEN PrdeS = "'maize stover'"

IF PrdeS = "2" THEN Pde8 = "'cut grass (hay)"'

IF PrdeS = "3" THEN Prde$ = "'natural pasture'"

IF Impl$ = "1" THEN Impl$ = "'moldboard plow.'"

1F Impl$ = "2" THEN Impl$ = "'moldboard plow and ox cart."'

IF Impl$ = "3" THEN lmplS = "'0x cart only.”

2650 CLS

LOCATE 10, 12

INPUT "Do you want to print out the outputs on a printer (Y/N)"; PrtOut$
IF PrtOut$ = "Y" OR PrtOut$ = "y" GOTO 2700

IF PrtOut$ = "N" OR PrtOut$ = "n" THEN GOTO 2710

BEEP
GOTO 2650
2710

' CREATING NEW FILE NAME

 

 

LOCATE 10, 10

PRINT " Creating a ﬁle for saving the results"
PRINT

LOCATE 12, 8

PRINT "The default ﬁle name is 'A:\z\l .dat'"
FileName$ = "A:\z\1.dat"

4000
LOCATE l4, 8
INPUT " Use default ﬁle name? (Y/N):", DeFNmeS

IF DeFNmeS = "Y" OR DeFNmeS = "y" GOTO 4200
IF DeFNme$ = "N" OR DeFNmeS = "n" THEN GOTO 4100
BEEP

CLS

GOTO 4000

4100

CLS

LOCATE 10, I4

INPUT "Enter New File Name: ", FileNameS
LOCATE 13, 10

PRINT "File Name ="; FileName$

'OPEN F ileName$ FOR OUTPUT AS #1

999 ' SAVING THE RESULTS IN A FILE
4200

 

 

 

182

PRINT " Please wait"
PRINT " Saving Results to the File"
PRINT #1,

.'111LJ111LELLILIIIIJL
ITIUTTUIIITI I ‘II

PRINT#I,

ILLJELILLIIIIIIIIJIIIIILHIlIIILJILJILLIIJIJI11L;1i;4..L"
TTTrI IITITT‘TTII’IITIITTIVYITIITITWIITIITYIIITTYIII

PRINT#I,"
PRINT#I,"
PRINT#I,"
PRINT#I,
PRINT#I,"
PRINT#I,

"inrnirrtnmrur J IJIII
IFTITIIIT T ITIT

PRINT #1,
PRINT #1,"
PRINT #1,"

PRINT #1, ""

PRINT#I,"
PRINT#I,

Energy Utilization Model for Draft Animal Power"
for Small-holder Agriculture"
(E U M D A P) "

George S N Mungai "

LLIIIILLILIILAJIIIILILIILJnllL
IT ITI I] III I

ILJLLJEI
I TYTTVT

rLlLLmJJ'
ITrTITrr

Herebelow are the results of the DAP-SYS simulation performed based on the",
inputs provided."

IN P U T S PROVIDED"

PRINT #1, USING "1. The land size of the farm is #.## ha out of which"; Landl;
PRINT #1 , " the cropped land for each"

PRINT #1, USING " season is #.## ha. with a soil type of“; Cand;

PRINT #1, " "; SolTy$

PRINT #1 , "2. The number of oxen used by system is"; Oxen

IF Oxen = 4 GOTO 992

PRINT #1, " with the following weightsz"

PRINT #1, USING " Oxl is ###.# kg."; WtOxl
PRINT #1, USING " 0x2 is ###.# kg."; WtOx2
PRINT #1, " Their ages are:"

PRINT #1, " age of 0x1 is "; Agerl;

PRINT #1, " years"

PRINT #1, " age of 0x2 is "; Ager2;

PRINT #1, " years"

GOTO 993

992

PRINT #1, USING " Oxl is ###.# kg"; WtOxl

PRINT #1, USING " 0x2 is ###.# kg"; WtOx2
PRINT #1, USING " 0x3 is ###.# kg"; WtOx3

PRINT #1, USING " 0x4 is ###.# kg"; WtOx4

PRINT #1, " Their ages are:"

PRINT #1, " age of 0x1 is "; AgerI;

PRINT #1, " years"

PRINT #1, " age of 0x2 is "; Ager2;

PRINT #1, " years"

PRINT #1 , " age of 0x3 is "; Age0x3;

PRINT #1, " years"

PRINT # I , " age of 0x4 is "; Ager4;

PRINT #1, " years"

993

PRINT #1, "3. The type of feed available was"; PrdeS

PRINT #1, "4. The implement(s) used were: "; Irnpl$

PRINT #1, "5. Tillage was started "; TilStS;

PRINT #1, " and the plowing was done at a"

PRINT #1, " depth of"; Dpthh; "cm. and";

PRINT # I, " a width of"; Wdthh; "cm."

PRINT #1, "6. The speed of operation was"; Spdd; "km per hour"
PRINT #1, "7. The oxen were used for"; Nde%; "days per week";

4
I

183

PRINT #1, " and for"; HrDyl; "hours per day."

PRINT #1, ""

PRINT#I," OUTPUTS"

PRINT #1,

PRINT #1, "1. Draft oxen energy supply"

PRINT #1,

PRINT #1, ; USING " 1.1 Maximum metabolizable energy (MI/day): #,###.## "; TMEm
PRINT #1, USING " 1.2 Total net energy (maintenance and work-MJ/day): #,###.## "; TNEmw
PRINT #1, USING " 1.3 Net energy for work (MI/day): ###.## "; TNEw

PRINT #1, USING " 1.4 DAP-SYS power supply (kW): #.## "; Owar

PRINT #1, USING " 1.5 DAP-SYS work capacity (hours/day): ###.## "; TDapCap
PRINT #1,

PRINT #1, "2. Implement energy demand"

PRINT #1,

PRINT #1, USING " 2.1 Implement power demand (kW): #.## "; Pwr

PRINT #1, USING " 2.2 Implement energy expenditure (kW-hr/ha): ###.## "; ImplEn
PRINT #1, USING " 2.3 Total ﬁeld implement energy demand (kW-hrs): ##.## "; TFEnD
PRINT #1,

PRINT #1, "3. Feed energy supply"

PRINT #1 ,

PRINT #1, USING " 3.1 Gross total season's maintenance feed (kg): #,###.## "; FdMn
PRINT #1, USING " 3.2 Total feed required for tillage work (kg): #,###.## "; Fde

PRINT #1, USING " 3.3 Gross feed required per season (kg): ##,###.## "; Grstqu
PRINT #1,

PRINT #1, "4. DAP System performance measures"

PRINT #1,

PRINT #1, USING " 4.] Effective ﬁeld capacity (ha/hr): #.### "; EFC

PRINT #1, USING " 4.2 Total daily energy expenditure (MJ): #,###.## "; TNEmw
PRINT #1, USING " 4.3 Actual time to ﬁnish work (days): #,###.## "; TATimTil
PRINT #1, USING " 4.4 Potential work time (days): #,###.## "; PothT
PRINT #1, USING " 4.5 Total daily tillage work done (MI): ##.##"; Etot

PRINT #1, USING " 4.6 Gross eﬁciency of DAP-SYS (%): ##1##"; GrsEf
PRINT #1, USING " 4.7 Net efﬁciency of DAP-SYS (%): ##.##"; NetEf

PRINT #1, USING " 4.8 Excess potential DAP-SYS work capacity (days): #,###.##"; ExCap
PRINT #1,
PRINT #1, "5. Random Estimates of produce used in simulation"

PRINT #1 ,

PRINT #1, USING " 5.1 Maize yield per hectare (90 kg bags): ##.##"; MzYld

PRINT #1, USING " 5.2 Total maize yield of farm (90 kg bags): ###.##"; TMzYld
PRINT #1, USING " 5.3 Quantity of maize sold (90 kg bags): ###.## "; Mle

PRINT #1, USING " 5.4 Beans yield per hectare (90 kg bags): ##.##"; BnYld

PRINT #1, USING " 5.5 Total beans yield of farm (90 kg bags): ###.##"; TBnYld
PRINT #1, USING " 5.6 Quantity of beans sold (90 kg bags): ###.## "; BnSl

PRINT # I ,

PRINT #1, "6. Income sources"

PRINT #1,

PRINT #1, USING " 6.1 Income from sale of crops (KSh): ###,###.##"; CrpInc
PRINT #1, USING " 6.2 Income from rental of DAP-SYS (KSh): ###,###.##"; HirInc
PRINT #1, USING " 6.3 Gross net income for season (KSh): ###,###.##"; GrsNthn
888

CLS

PRINT "Etot"; Etot

PRINT "TDapCap"; TDapCap
PRINT "TCrand"; TCrand
PRINT "ATimTil"; ATimTil
PRINT "MEm 1 "; MEMI

 

 

184

PRINT "Km"; Km

PRINT "BMR! "; BMRI
PRINT "BMR2"; BMR2
PRINT "NEmw l "; NEmwl
PRINT "EFC"; EFC

PRINT "Hrdy"; HrDy

PRINT "TBMR"; TBMR
PRINT "TNEmw"; TNEmw
PRINT "Grstqu"; Grstqu
PRINT "GrsEand"; GrsEand

 

CLOSE #1
END

2700
LPRINT

"l LLIALJILIIJIJILLLILILLIIJIJEIIIIIJEJAIILIIILJLLJI 11
YTIIITIITTIITWIIITIYIITTIIIITTIITTFITITTIWTIIIITTI

PRINT
LPRINT " Energy Utilization Model for Draft Animal Power"

LPRINT " for Small-holder Agriculture"

LPRINT " (E U M D A P) "

LPRINT

LPRINT " George S N Mungai "
LPRINT

LPRINT

LPRINT " Herebelow are the results of the DAP-SYS simulation perfomed;"
LPRINT " based on the inputs provided."

LPRINT ""

LPRINT" INPUTS PROVIDED"

LPRINT

LPRINT USING "I. The land size of the farm is #.## ha out of which"; Landl;
LPRINT " the cropped land for each"

LPRINT USING " season is #.## ha. with a soil type of"; Cand;

LPRINT " "; SolTyS

LPRINT "2. The number of oxen used by system is"; Oxen

IF Oxen = 4 GOTO 994

LPRINT " with the following weights:"

LPRINT USING " Oxl is ###.# kg."; WtOxl

LPRINT USING " 0x2 is ###.# kg"; WtOx2

LPRINT " 'Iheir ages are:"

LPRINT " age of 0x1 is "; Agerl;

LPRINT " years"

LPRINT " age of 0x2 is "; AgerZ;

LPRINT " years"

GOTO 995

994

LPRINT USING " Oxl is ###.# kg"; WtOxl

LPRINT USING " Ox2 is ###.# kg"; WtOx2

LPRINT USING " 0x3 is ###.# kg"; WtOx3

LPRINT USING " 0x4 is ###.# kg"; WtOx4

LPRINT " Their ages are:"

LPRINT " age of 0x1 is "; AgerI;

LPRINT " years"

LPRINT " age of 0x2 is "; Ager2;

LPRINT " years"

 

111 I1
TII I

14 ILmrrrurrrrrnkrtrnriri 11111 errrrhmnm"
TTI 111 rrrT ITT 11111 T rrrrwr

 

 

 

185

LPRINT " age of 0x3 is "; Ager3;

LPRINT " years"

LPRINT " age of 0x4 is "; Ager4;

LPRINT " years"

995

LPRINT "3. The type of feed available was"; PrdeS

LPRINT "4. The implement(s) used were: "; Impl$

LPRINT "5. Tillage was started "; TilStS;

LPRINT " and the plowing was done at a"

LPRINT " depth of'; Dpthh; "cm. and";

LPRINT " a width of“; Wdthh; "cm."

LPRINT "6. The speed of operation was"; Spdd; "km per hour"
LPRINT "7. The oxen were used for"; Nde%; "days per week";
LPRINT " and for"; HrDy!; "hours per day."

LPRINT ""

LPRINT" O U T P U T S"

LPRINT

LPRINT "1. Draft oxen energy supply"

LPRINT

LPRINT USING " 1.1 Maximum metabolizable energy (MI/day): #,###.## "; TMEm
LPRINT USING " 1.2 Total net energy (maintenance and work-MJ/day): #,###.## "; TNEmw
LPRINT USING " 1.3 Net energy for work (MI/day): ###.## "; TNEw
LPRINT USING " 1.4 DAP-SYS power supply (kW): #.## "; Owar
LPRINT USING " 1.5 DAP-SYS work capacity (hours/day): ###.## "; TDapCap
LPRINT

LPRINT "2. Implement energy demand"

LPRINT

LPRINT USING " 2.1 Implement power demand (kW): #.## "; Pwr

LPRINT USING " 2.2 Implement energy expenditure (kW-hr/ha): ###.## "; ImplEn
LPRINT USING " 2.3 Total ﬁeld implement energy demand (kW-hrs): ##.## "; TFEnD
LPRINT

LPRINT "3. Feed energy supply"

LPRINT

LPRINT USING " 3.1 Gross total season's maintenance feed (kg): #,###.## "; FdMn
LPRINT USING " 3.2 Total season's feed required for work (kg): #,###.## "; Fde

LPRINT USING " 3.3 Gross feed required per season (kg): ##,###.## "; Grstqu
LPRINT

LPRINT "4. DAP System performance measures"

LPRINT

LPRINT USING " 4.1 Effective ﬁeld capacity (ha/hr): #.## "; EFC

LPRINT USING " 4.2 Total daily energy expenditure (MJ): #,###.## "; TNEmw
LPRINT USING " 4.3 Actual time to ﬁnish work (days): ##.## "; TATimTil
LPRINT USING " 4.4 Potential work time (days): #,###.## "; PothT
LPRINT USING " 4.5 Total daily tillage work done (MJ): ##.##"; Etot

LPRINT USING " 4.6 Gross eﬁciency of DAP-SYS (%): ##.##"; GrsEf
LPRINT USING " 4.7 Net efﬁciency of DAP-SYS (%): ##.##"; NetEf
LPRINT USING " 4.8 Gross net return per season (KSh): ##,###.##"; GrsNthn
LPRINT USING " 4.9 Excess DAP-SYS work capacity (days): #,###.##"; ExCap
LPRINT #1,

LPRINT #1, "5. Random Estimates of produce used in simulation"

LPRINT #1,

LPRINT #1, USING " 5.1 Maize yield per hectare (90 kg bags): ##.##"; MzYld
LPRINT #1, USING " 5.2 Total maize yield of farm (90 kg bags): ###.##"; TMzYld
LPRINT #1, USING " 5.3 Quantity of maize sold (90 kg bags): ###.## "; MzSI
LPRINT #1, USING " 5.4 Beans yield per hectare (90 kg bags): ##.##"; BnYId

LPRINT #1, USING " 5.5 Total beans yield of farm (90 kg bags): ###.##"; TBnYId

 

 

186

LPRINT #1, USING " 5.6 Quantity of beans sold (90 kg bags):
LPRINT #1,

LPRINT #1, "6. Income sources"

LPRINT #1,

LPRINT #1, USING " 6.1 Income from sale of crops (KSh):

LPRINT #1, USING " 6.2 Income from rental of DAP-SYS (KSh):

LPRINT #1, USING " 6.3 Gross net income for season (KSh):

END

###.## "; BnSl

###,###.##"; CrpInc
###,###.##"; HirInc
###,###.##"; GrsNthn

 

 

 

 

APPENDIX C

187

APPENDIX C

Appendix G shows three print-outs of simulations ran with the EUMDAP computer
simulation program. This print-out is the same as the user would obtain after simulations
are carried out. The data input by the user and the output values of the simulation are
shown in the print-outs. The simulations were performed for two oxen weighing 250 kg
each and plowing between 1.17 to 3.29 ha of land with silty clay soil type at a speed of

2.6 km per hour.

Simulation #1:

 

+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

Energy Utilization Model for Draft Animal Power
for Small-holder Agriculture
(E U M D A P)

George S N Mungai
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

Herebelow are the results of the DAP-SYS simulation performed
based on the inputs provided.

I N P U T S P R O V I D E D

l. The land size of the farm is 5.67 ha out of which the cropped land
for each season is 3.29 ha. with a soil type of 'Silty clay.‘
2. The number of oxen used by system is 2
with the following weights:
Oxl is 250.0 kg.
0x2 is 250.0 kg.
Their ages are:
age of 0x1 is 3 years
age of 0x2 is 3 years
3. The type of feed available was'maize stover'
The implement(s) used were: 'moldboard plow.‘
5. Tillage was started 'as soon as rains come' and the plowing was done
at a depth of 10 cm. and a width of 25 cm.
The speed of operation was 2.565 km per hour
7. The oxen were used for 5 days per week and for 6 hours per day.

.13

O\

O U T P U T S

1. Draft oxen energy supply

 

 

188

Maximum metabolizable energy (MJ/day):

Net energy for work (MJ/day):
DAP-SYS power supply (kW):
DAP-SYS work capacity (hours/day):

I—‘l—‘t—‘l—‘H
UlbLONH

Implement energy demand

2.1 Implement power demand (kW):
2.2 Implement energy expenditure (kW-hr/ha):
2.3 Total field implement energy demand (kW-hrs):

Feed energy supply
Gross total season's maintenance feed (kg):

Total feed required for tillage work (kg):
Gross feed required per season (kg):

LAJUJUJ
DUMP

DAP System performance measures

Effective field capacity (ha/hr):
Total daily energy expenditure (MJ):
Actual time to finish work (days):
Potential work time (days):

Total daily tillage work done (MJ):
Gross efficiency of DAP-SYS (%):

Net efficiency of DAP-SYS (%):

AAA-5.5.5.545
ooxrmmbwmw

Random Estimates of produce used in simulation

Maize yield per hectare (90 kg bags):
Total maize yield of farm (90 kg bags):
Quantity of maize sold (90 kg bags):
Beans yield per hectare (90 kg bags):
Total beans yield of farm (90 kg bags):
Quantity of beans sold (90 kg bags):

U'IUIU'IUTU'IUI
OWUTIbWNr-J

Income sources

6.1 Income from sale of crops (KSh):
6.2 Income from rental of DAP-SYS (KSh):
6.3 Gross net income for season (KSh):

Total net energy (maintenance and work-MJ/day):

Excess potential DAP-SYS work capacity (days):

85.05
57.29
24.77
0.32
6.43

0.28
4.58
15.07

896.8
29.9
926.7

0.062
57.29
8.2
110.05
6.40
11
27
100.1

LA) U" U1 N LA) l-’
o e o H H w
U" (D \I

bubb-

55,546
0
40,720

 

 

189

Simulation #2:

 

+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

Energy Utilization Model for Draft Animal Power
for Small-holder Agriculture
(E U M D A P)

George S N Mungai
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

Herebelow are the results of the DAP-SYS simulation performed
based on the inputs provided.

I N P U T S P R O V I D E D

l. The land size of the farm is 4.45 ha out of which the cropped land
for each season is 2.58 ha. with a soil type of 'Silty clay.‘
2. The number of oxen used by system is 2
with the following weights:
0x1 is 250.0 kg.
0x2 is 250.0 kg.
Their ages are:
age of 0x1 is 3 years
age of 0x2 is 3 years
3. The type of feed available was'maize stover'
4. The implement(s) used were: 'moldboard plow.‘
5. Tillage was started 'as soon as rains come' and the plowing was done
at a depth of 10 cm. and a width of 25 cm.
6. The speed of operation was 2.565 km per hour
7. The oxen were used for 5 days per week and for 6 hours per day.

0 U T P U T S

1. Draft oxen energy supply

1.1 Maximum metabolizable energy (MJ/day): 85.05
1.2 Total net energy (maintenance and work—MJ/day): 57.29
1.3 Net energy for work (MJ/day): 24.77
1.4 DAP-SYS power supply (kW): 0.32
1.5 DAP-SYS work capacity (hours/day): 6.43

2. Implement energy demand

2.1 Implement power demand (kW): 0.28
2.2 Implement energy expenditure (kW-hr/ha): 4.58
2.3 Total field implement energy demand (kW-hrs): 11.84

3. Feed energy supply

3.1 Gross total season's maintenance feed (kg): 896.8
3.2 Total feed required for tillage work (kg): 23.5
3.3 Gross feed required per season (kg): 920.3

4. DAP System performance measures

bbbbbbbb
ooqoxmwaI-J

190

Effective field capacity (ha/hr):
Total daily energy expenditure (MJ):
Actual time to finish work (days):
Potential work time (days):

Total daily tillage work done (MJ):
Gross efficiency of DAP-SYS (%):

Net efficiency of DAP-SYS (%):

Excess potential DAP-SYS work capacity (days):

Random Estimates of produce used in simulation

UIU‘IU'IUIUTU'I

ONUTALUNH

Maize yield per hectare (90 kg bags):
Total maize yield of farm (90 kg bags):
Quantity of maize sold (90 kg bags):
Beans yield per hectare (90 kg bags):
Total beans yield of farm (90 kg bags):
Quantity of beans sold (90 kg bags):

Income sources

6.1 Income from sale of crops (KSh):
6.2 Income from rental of DAP-SYS (KSh):
6.3 Gross net income for season (KSh):

0.062
57.29
6.5
109.29
6.40
11

27
99.3

113-bu.)
CDCD\.I

NubU'll-‘NH

38,541
0
24,969

 

 

I91

Simulation #3:

 

+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

Energy Utilization Model for Draft Animal Power
for Small- holder Agriculture
(EUMDAP)

7 - George S N Mungai
+++++++++++++++++++++++++++++++++++++++++++++t4+4++++++++++++++++++++++

Herebelow are the results of the DAP-SYS simulation performed
based - . ii---
on the inputs provided.

IAN P U T s PROVIDED

1. The land size of the farm is 2.02 ha out of which the cropped land
for each season is 1.17 ha. with a soil type of 'Silty clay.‘
2. The number of oxen used by system is 2
with the following weights:
0x1 is 250.0 kg.
0x2 is 250.0 kg.
Their ages are:
age of 0x1 is 3 years
age of 0x2 is 3 years
3. The type of feed available was'maize stover'
4. The implement(s) used were: 'moldboard plow.I
5. Tillage was started 'as soon as rains come' and the plowing was done
at a depth of 10 cm. and a width of 25 cm.
6. The speed of operation was 2. 565 km per hour
7. The oxen were used for 5 days per week and for 6 hours per day.

0 U T P U T S

1. Draft oxen energy supply

1.1 Maximum metabolizable energy (MJ/day): 85.05
1.2 Total net energy (maintenance and work-MJ/day): 57.29
1.3 Net energy for work (MJ/day): 24.77
1.4 DAP-SYS power supply (kW): 0.32
1.5 DAP-SYS work capacity (hours/day): 6.43

2. Implement energy demand

2.1 Implement power demand (kW): 0.28
2.2 Implement energy expenditure (kW—hr/ha): 4.58
2.3 Total field implement energy demand (kW-hrs): 5.38

3. Feed energy supply

3.1 Gross total season's maintenance feed (kg): 896.8
3.2 Total feed required for tillage work (kg): 10.7
3.3 Gross feed required per season (kg): 907.5

4. DAP System performance measures

 

 

 

bbbbbbbb
Queen-e-

(DQONUTbUJNl-J

192

Effective field capacity (ha/hr):
Total daily energy expenditure (MJ):
Actual time to finish work (days):
Potential work time (days):

Total daily tillage work done (MJ):
Gross efficiency of DAP-SYS (%):

Net efficiency of DAP-SYS (%):

Excess potential DAP-SYS work capacity (days):

Random Estimates of produce used in simulation

U'IUIU'IUTUIUI

ONU'IAle—J

Maize yield per hectare (90 kg bags):
Total maize yield of farm (90 kg bags):
Quantity of maize sold (90 kg bags):
Beans yield per hectare (90 kg bags):
Total beans yield of farm (90 kg bags):
Quantity of beans sold (90 kg bags):

Income sources

6.1 Income from sale of crops (KSh):
6.2 Income from rental of DAP-SYS (KSh):
6.3 Gross net income for season (KSh):

0.062
57.29
2.9
107.77
6.40
11

27
97.8

 

 

APPENDIX H

193

APPENDIX H

Raw data printout of File #: Musau2

One of the ﬁles containing raw ﬁeld data is provided with slight modiﬁcations to
indicate the eight channels used for data collection and the real time lapse. Also included
is a count of the data points and statistical calculations for each channel values at the end
of the ﬁle. The ﬁle printed out was File number Musau2 for one of the farmers in

Kalama division.

Horizontal Vertical Heart Respiration Body Ground Rate
Data Time Draft Draft Rate Rate T

Points Csec

 

I94

 

 

 

 

I95

 

 

I96

 

I97

 

198

 

I99

Horizontal Vertical Heart Respiration Body Ground Rate
Draft Draft Rate Rate T

1 1
0.1
0.1

1

1

Maximum

 

 

 

APPENDIX I

 

200

APPENDIX I

Questionnaire on Draft Animal Power

(DAP) Energy Utilization
in Small-Holder Agriculture
in Kenya

Information about the respondent

Respondent’s name

 

 

I

Province

 

District Division
Location Sub-location

 

 

 

 

 

Information about the enumerator

Enumerator’s name

Date of interview

Place of interview

[eg Respondent ’s house, farm, field]

Time interview began am/pm [check one]
Time interview ended am/pm [check one]
Comments (What was the respondent doing on your arrival?)

 

 

 

 

 

 

 

Terms of the interview

This questionnaire will be administered on a voluntary basis. The respondents will not be coerced to
provide any of the information being asked for. The data collected will be used to develop programs for
improving Draft Animal Power for small-holder farmers. Respondenm will be clearly explained this
purpose which is intended to help them in their mechanization needs. The researcher takes full
responsibility on the data received and will not release the same to a third party without the respondenm ’
permission. ‘

 

 

The respondent '5 name, address and other personalﬂamily data will be used only for identifying the source
of the data. T he family structure data, household income/expenditure data will be used with confidentiality.

 

Farm Map
Make a sketch of all the ﬁelds that were worked on during the "short ” and “long " rains last year.

INSTRUCTIONS:

( To enumerator)

First draw the farmers house and reference marks like a major road, mountain, river. Secondly (a)
draw all the fields that the family has. (b) include all the ﬁelds that the family rented or borrowed
during the past year. Thirdly number all the ﬁelds. F ourthly write the sig of each field in the sketch.
Then list the principal crops planted during short and long rains in the space provided. F iﬁh if the
family has other land located elsewhere, ask the farmer its location and make notes about it.

 

 

 

 

 

 

(For enumerator to the farmer)

201 '

We would like to make simple sketches of your farm showing where the various ﬁelds you worked on last
year were located for each of the two seasons.

Farm Sketch:
North

 

202

Field
Short Rains: Number Main Crops

 

 

 

 

 

Long Rains: Number Main Crops

 

 

 

 

5.
Field

Other Land: Number Location Terms (eg Rented, Borrowed)
1.

 

 

 

 

 

 

 

2.
3.
4

 

 

CHECK LIST:

1) Did you include the farmer’s house and major reference marks?

2) Did you include all the ﬁelds?

3) Did you number all the ﬁelds.

4) Did you indicate all the principal crops grown in short and long rains?

5) Did you make notes on the location of ﬁelds away from the “homestead”?

 

 

III .-
A

 

203

Block 1
I. Resources Inventory
[1.1] What is the size of your farm? Acres
[1.2] How many parcels of land did you work on (a) last season? (short rains)
[1.3] (b) the preceding season (long rains)

Table 1: Field sizes and characteristics.
own easy to
(Season) I rented (Acres) 2 rolling 2 medium soft
2 borrowed 3 steep 3 hard
3 communal 4 4 hard
P

 

I well cleared
2 trees/stumps
cleared

3

204

Income and Draft Animal Power Costs

Production and Sales
We would like to ask some questions related to the crops grown and sold from the use of draft oxen.
Table 2: Short Rains Season

Grown Area Total Production Sales Price

 

Name A UN

 

Table 3: Rains Season
rown
Field Name A
I

 

1.9 Totals

Block 2

Prime Mover/Equipment.
Hand Tools: How many of the following hand tools do you have?

 

 

 

 

 

 

 

Number Other hand tools:
[2.” Hand hoe Name Number
[2.2] Digging fork [2.5]
[2.3] Machetes [2.6]
[2.4] Axe [2.7]

 

Draft Animal Power

[2.8] Do you own trained draft oxen? [0 = no; I = yes]
[2.9] If yes, how many?

How many of the following DAP implements do you own?

 

 

 

 

 

 

 

Number Other DAP implements
[2.10] Moldboard plow Name Number
[2.11] Harrow [2.14]
[2.12] Planter/Seeder [2.15]
[2.13] Trailer/Cart [2.16]

 

 

[2.17] What is the main activity for which you use the draft oxen?
Enumerator: Check one main activity.
a) Plowing
b) Weeding
c) Haulage
[2.18] When do you start your land preparation for the following season?

 

 

205

Enumerator: Check % response.

a) Soon after harvest

b) Just before rains come

c) After rains come

d) Other (specify)
[2.19] Why do you start at that time?

 

 

 

[2.20] Do you own a tractor? [ 0 = no; I = yes]
[2.21] If yes, what is the size of the tractor? HP.
Block 3.

Draft Animal Power Operation

[3.1] How many oxen do you use for plowing?

[3.2] If you use more than two, why?
Enumerator: Check items that apply.
a) Oxen too weak
b) Ground too hard
c) Training other team
d) All oxen available
e) Other reason (Speciﬁz)
[3.3] How many persons operate the oxen team?
[3.4] Who is the principal ox-team operator?
a) Self
b) Wife/Husband
c) Employee
d) Son/daughter

 

 

 

 

Table 4: Use of draft oxen

Field Number

Activities
to

 

[3.9] What other work did your oxen do on your farm in the two seasons?

 

 

 

 

 

 

List the types of work:
Short Rains Season Time Used

Type of Work Days Hours/day
3) _
b) _
C) __ __
Long Rains Season Time Used

Type of Work Days Hours/day
21) ——
b) _

C)

 

 

 

 

 

206

[3.10] Approximately how many days per season do you use the oxen for aﬂ work? days
[3.11] About how many days do you use the oxen in-between the seasons? . days
[3.12] List three main activities that you use the oxen for in-between the seasons?

a)

b)

C)
[3.13] What are the two most ENERGY or POWER demanding activities on your farm?

a)

b)

 

[3.14] During the farming season what activity takes up the greatest amount of time? (Enumerator:
not necessarily using oxen)

 

 

 

Block 4
II. Utilization of Draft Oxen
Hiring Out Oxen '
[4.1] Did you hire out your oxen to other farmers last season? [0 = no; I = yes] '4
[4.2] If yes, give the two most important reasons why you hired out your oxen.
8)
b)

 

[4.3] If you hired oxen out for plowing:
Fill in the hiring details for each season below:

Ksh/acre of Ksh/acre

 

[4.4] If you hired oxen out for haulage:
Fill in the hiring out details for each season below:

 

 

207

[4.5] If you hired out oxen for other activities:

Rains

activities ow Num Hours Num
of Ksh/hr of

 

Block 5.

III. Feed Sources & Usage

[5.1] What was your primary feed source?
a) Communal grazing

 

 

 

 

b) Own grazing ﬁeld

c) Bought feed

d) Any other (Specrjy)
[5.2] Did your oxen get enough feed from your farm? [0 = no; 1 = yes]
[5.3] What was the acreage of your grazing land? acres
[5.4] If answer to Q 5.2 is No did you use extra feed? [0 = no; 1 = yes]
[5.5] If yes to Q 5.4, which extra feed did you use?

Extra feed used Source of feed [2 = bought; 3 = free]

a) Hay

b) Silage (green fodder)

c) Farm waste
d) Concentrates
e) Any other (Specijfv)

 

Block 6.

IV. Renting of Draft Animals

Renting oxen from other farmers.
[6.1] Did you hire oxen from other farmers? [0 = no; I = yes]
Enumerator: If answer is l_V_o_ go to question 6.1].
[6.2] Which season did you hire other farmer’s oxen? [2 = short rains; 3 = long rains]
[6.3] For what activity did you hire the oxen?
a) Plowing
b) Haulage
c) Any other (Specify)
[6.4] If hired for plowing, what acreage were they used for? acres.
[6.5] If hired for haulage, how many hours were they used? hours.
[6.6] If hired for other activity, how many hours were they used for? hours.
[6.7] How was the cost computed? Choose one.
a) Per acre
b) Per hour
c) Per day
[6.8] What was the cost of plowing per acre? Ksh. per acre
[6.9] What was the cost of haulage per hour? Ksh. per hour.
[6.10] What was the unit charge for other activity? Ksh per

 

 

Farmers that did not rent oxen from other farmers.

 

[6.11] Why did you not hire oxen from other farmers?
a) Had own capacity
b) Not available
c) Expensive
d) Poor workmanship
e) Available too late
f) Any other reason (Specify)
[6.12] What other options do you have for tillage?
a) Hire tractors
b) Family labor
c) Hired labor
(1) Friends
[6.13] Did you borrow oxen from friends or relatives? [0 = no; 1 = yes]
[6.14] Why did you borrow oxen from friends or relatives?
a) They offered other services
b) In return for favor given earlier.
c) Other reasons (Speciﬁz)

 

 

 

Block 7

7. Maintenance Costs of Draft Oxen
[7.1] In which year did you start using draft oxen? 19
[7.2] For how many years have you been using oxen ? years.
[7.3] How did you acquire the initial draft oxen?
Enumerator: Choose one from respondent.
a) Purchased
b) Inherited from parents
c) Other source (Specrjy)
[7.4] If purchased the oxen, when did you do so? 19
[7.5] What was the price per animal at that time? Ksh.
[7.6] Who trained the oxen team for you?
Enumerator: Choose one that applies.
a) Self
b) Friends
c) Employee
[7.7] For how long do you expect to use a team of oxen before replacement? years.
[7.8] How much would you expect each ox to fetch in the market currently? Ksh.
[7.9] What method would you use to replace the oxen?
a) Buy trained ones.
b) Buy and train
c) Select from old stock and train
d) How much would a trained ox cost? Ksh.
[7.10] If you choose to buy, how much would you expect to pay? Ksh.

 

 

 

 

 

 

 

 

 

 

 

[7.11] Did you incur any veterinary costs on the animals last year? [0 = no; I = yes]
[7.12] If yes, how much? Ksh.
[7.13] Were there any other maintenance costs on your animals? (Speciﬁz).
Other maintenance costs How much

a) Ksh.

b) Ksh.

c) Ksh.

d) Ksh.

 

[7.14] When did you acquire the oxen implements that you have and at what price?

Implements Year acquired Price paid
a) Moldboard plow 19 Ksh.

209

 

b) Harrow l9 Ksh. __
c) Planter/seeder I9 Ksh. __
d) Trailer/cart l9 Ksh. __
Block 8
Future Plans
[8.1] Are you meeting all your farm energy needs? [0 = no; I = yes]
[8.2] What are the IE major energy constraints areas on your farm?
a) Plowing
b) Weeding
c) Haulage
d) Harvesting
e) Any other (Specrﬁr)

 

[8.3] What would be the best tl_1_r§ options that you would consider to improve your energy needs?
a) Hire tractor
b) Buy more oxen
c) Grow more feed
(I) Buy quality feed
e) Hire more labor
f) Buy ox-cart
g) Hire oxen

Block 9

Credit
[9.1] Have you ever obtained a loan for mechanization needs? [0 = no; 1 = yes]
[9.2] If yes, which year did you get one? 19
[9.3] How much was the loan? Ksh
[9.4] If m to question 9.1, would you consider taking a loan to meet any of your energy needs?
[0=no;1=yes]
[9.5] What would be your best option for a loan?
a) Agricultural Finance Corporation (AFC)
b) Bank
c) Friends
(I) Relatives
f) Village money-lender
g) Other (Speciﬁ’)
[9.6] What would you use as security for the loan?
a) Title deed
b) Daft oxen

c) Other property (Speciﬁ’)

Hill

 

 

V. STRUCTURE OF THE FAMILY

 

[10.1] How many people currently live in your house (home)?

 

210
Table 5: Family Structure
Relation Gender highest level Does
1 head 2 spouse 1 M of schooling he/she
3 child 4 cousin 2 F completed work on
5 aunt/uncle the farm?
6 employee 0 = no
Name 1 = yes

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

\OOOQQUI-bUJN—‘m:

 

[10.2] What is the current wage rate per day for the following tasks?

 

a) Weeding Ksh
b) Harvesting Ksh.
c) Plowing with oxen Ksh.

d) Any other task (Specify) Ksh.

[10.3] What is the draft animal operator’s wage rate? Ksh. per month.

 

APPENDIX J

211

 

 

 

 

m: «a I m: « c
35:6 I I l I . H .N
2:: Eu: «Ecotm 5: 06—. g .m ... 2: o3. 2.5. m.
3.5 W. 1 ml; lot: 1 ..N
2:: Eva E63885. é a 3 ~ 9% 2.54. 2
case seas ﬂ 5 H.332 r R- 1 ﬁg 1B! n PE
20: 30:285. :62: 3 E 5- a: x: ... .w 2% EH< 2
35:6 .... ..
:53 c8 .335 62 .26 cod :2 E mos—g =< NEE I M? H M2 3350 m—
ass—e 2 a 2.133. _ as ... «E
SE £3565 38m zen can .85. E 83 .258 EEwtO $.89». 32.? 2.1302 96o: 62m .2
33% to;
use 8:38:85 ... ... as
.8 38:0 82 26% :32? mm 2;?» .958“. cozuou a a 5v. VN ... N: H M2 and: Dm< _ _
assess ace ..
c2235: xwuocm Ev. Co 55335 mmd + m: ... mad H M Sac: om< o—
QmEqEv 3.55 s ...
033:3806 482 Ems; boon do .x. On a E m: ... K H ..N: 8:: 0m< a
9:5 3:5 .«o :23sz :235m 0225 ~on 5
a cozmscm

 

.6on coca—25m m<QEDm 05 Sm Bicep 32:33 .8 538—0

_. X—DZm—A—md.

212

 

 

   

 

 

 

 

 

 

 

hat you 5: 3.5» bum» mm. m

E 2:. a 9825 HI I n IEII anQ

3688 $5 9‘ 5: £ 3&2 .53th cm
its maumaqﬂmm ..

$me 20c A3 H .363 = ..II km:

E 33 32352: B ... 5‘ 3 IN 32th mm
g
93%: 3‘ H .3 I 59‘
35:0 Soc .88. .5 I k3 m: * “EA III EH 336D vm
3‘ 5‘ URN

aﬁséxv 335 ﬁIIT: n MIV I u E

25:53:: 050on 5* .3 I x»3 R 33.80 8
00m
$3 5:8 E Q3
«80 25829:— E m. * Q N R 33.50 mm
35:5 ESE 3%. 5H3 Wm _ ﬂQV a 3...

335 20c Eek 24 3:5 ".582 3 3 gunman .zwm .Q . k .2 _. Hi I m 33.5 F.

€35 $033 8V .. ﬂEHI Imml R; n mummy I2 H DRE
Soc o>toobm a: E :62: .5 a: W ... x: ... m. 9% 2.5. cm

s - 9

255003 _H.=\HI ”NH—H A83 I. H MK

@5350 20E 3 rm o3. 2H< 2
8:5 35: mo :23:th :ozaacm 8.3m :8. E
u nouns—um

 

213

 

 

   

 

 

 

 

 

A33 03.: T : c3
.8 25. .52 3s 3 @XEBESSIEEQ 323a
52.5 5% “on ﬁ 5 #3 HSQVS :PJ «3: I a
x33 :28 8535 28$: "382 s 25.2 . auumua . Em I S“ Q 32.5 on
«E 2m I EVE
ﬁo\ov OO— * 3:: H \QMQV
xocuBEo 230%? new: 33:0 .3 0:5— W Avg: >85 mm
«E: I $5
09— * 3:: H \Wumz
gov 35650 32 new: 335 we 0:3— .m $3: 395 mm
3::
SE
00— ... ~IaII§ H \thG
ax; 55650 880 new: 335 .3 0:3. m 3.3: buohm mm
82:. mo cozgton cogscm 350m 2on E

5:33.—

 

214

Bibliography

Agricultural Research Council. 1980. The nutrient requirements of ruminant livestock.
Commonwealth Agricultural Bureaux, Slough.

Brody, Samuel. 1964. Bioenergetics and growth with special reference to the efﬁciency
complex in domestic animals. Hafner Publishing Company Inc. New York.

Campbell, J. K. 1990. Dibble sticks, donkeys, and diesels. Machines in production.
International Rice Research Institute. Po Box 933 Manila, Phillipines.

Crossley, Peter and John Kilgour. 1983. Small farm mechanization in developing
countries. John Wiley & Sons, Chichester, U. K.

Downing, T. E., D. N. Mungai, and H. R. Muturi. 1988. Drought climatology and
development of the climatic scenarios. In: The impact of climatic variations on
agriculture: Vol 2: Assesments in semi-arid regions (Eds. M. L. Parry, T. R.
Carter and N. T. Konjin) pp. 149 - 174. Kluwer Academic Publishers, Dordrecht,
Netherlands.

Ensminger, ME. 1989. Animal Science (Animal Agricultural Series). Interstate
Publishers Inc. Dansville, Ilinois.

FAO, 1983. Production yearbook, Rome, Italy

FAO. 1972. The employment of draft animals in Africa. Food and Agricultural
Organization of United Nations. Rome, Italy.

F A0, 1972. Production yearbook, Rome, Italy

Goe, Michael R. and Robert E. McDowell. 1980. Animal traction: guidelines for
utilization. Departmentof Animal Science. Cornell University, Ithaca New York.
14853

Gottlieb-Vedi, M. B. Essen-Gustavbsson and S. G. B. Persson. 1991. Draught load and
speed compared by submaximal tests on a treadmill. Equine Exercise Physiology 3: 92-
96, 1991

Hagmann, Jurgen. 1994. Development of an animal-drawn disc ridger for a tied ridging
system of conservation tillage. In: Improving Animal Traction Technology.
Proceedings of the ﬁrst workshop of the Animal Traction Network for Eastern and
Southern Africa (ATNESA) held 18-23 January 1992, Lusaka, Zambia. Technical
Center for Agricultural and Rural Cooperation (ACP-EU Lorne Convention)
Wageningen, The Netherlands.

215

Harris, Lorin E. 1966. Biological energy interrelationships and glossary of energy terms.
Publication 1411 National Academy of Sciences. National Research Council,
Washington, D. C. ‘

Heyer, Judith and J. K. Waweru. 1976. The development of the small farm areas. In:
Judith Heyer, J. K. Maitha and W. M. Senga (eds), Agricultural Dvelopment in Kenya:
An economic assesment, Oxford University Press, Nairobi.

Iftikhar, A. and Bill Kinsey (Eds) 1984. Farm equipment innovations in Eastern and
Central Southern Africa. A study prepared for the ILO Ofﬁce within the framework of
the World Employment Programme. Gower. Biddles Ltd, Guilford and King ’s Lynn.
GB.

Inns, Frank. 1990. The mechanics of animal-draught cultivation implements. Agricultural
Engineer, Spring 1990. Pp 13-17 .

Ismail, Wan Ishak Wan. 1991. Simulation model for ﬁeld crop production machinery
system. Unpublished Ph D dissertation. Michigan State University, East Lansing,
Michigan.

Johnston, Bruce F. 1984. Farm equipment innovations in Eastern Africa: policy
considerations. In: Farm Equipment Innovations in Eastern and Central Southern
A ﬁica. Iftikhar Ahmed and Bill H. Kinsey (Eds). International Labor Organization
1984. Gower.

Kaumbutho, P. G., A. Anyangu, P. M. Owende and J. Mbuthia. 1995. Meeting the
challenges to draught animal technologies in Kenya. Proceedings of a workshop
held at the Kenya Commercial Bank Institute of Banking and Finance, Ngong
Kenya. March 27 - 31, 1995.

Kayumbo, A. K. 1994. Techniques and implements for weeding with draft animals in
Tanzania. In: Improving Animal Traction Technology. Proceedings of the ﬁrst
workshop of the Animal Traction Network for Eastern and Southern Africa
(ATNESA) held 18-23 January 1992, Lusaka, Zambia. Technical Center for
Agricultural and Rural Cooperation (ACP-EU Lorne Convention) Wageningen, The
Netherlands.

Kemp, D. C. 1987. Draught animal power—some recent and current work. World animal
review, 63 Jul-Sept; 7-14.

Kline, C. K., D. A. G. Green, R. L. Donahune and B. A. Stout. 1969. Agricultural
mechanization in equatorial Aﬁ'ica Institute of international agriculture research.
Report No. 6. Michigan State University, East Lansing, Michigan.

Kuhlmann, W. D. D. S. Hodgson and M. R. Fedde. 1985. Respiratory, cardiovascular,

216

and metabolic adjustments to exercise in the Hereford calf. Journal Applied Physiology.
58(4): 1273-1280, 1985

Kwiligwa, E. M., R. M. Shetto and D. J. Reeds. 1994. The use of animal-drawn
cultivators for maize production in the southern highland of Tanzania. In: Improving
Animal Traction Technology. Proceedings of the ﬁrst workshop of the Animal
Traction Network for Eastern and Southern Africa (ATNESA) held 18-23 January
1992, Lusaka, Zambia. Technical Center for Agricultural and Rural Cooperation
(ACP-EU Lome Convention) Wageningen, The Netherlands.

Lawrence, Peter R. 1985. A review of nutrient requirements of draft oxen. In: Draft
Animal Power for Production. ACIAR Proceedings Series # 10. Proceedings of an
International Workshop Held at James Cook University, Townsville, Queensalnd,
Australia. July 10-16 1985. J. W. Copland (Ed.)

Lawrence, P. R. and R. J. Stibbards. 1990. The energy costs of walking, carrying and
pulling loads on ﬂat surfaces by Brahman cattle and swamp buffalo. Animal
Production, 1990. 50: 29-39.

Lawrence, P. R. and A. R. Pearson. 1985. Factors affecting the measurement of draught
force, work output and power of oxen. Journal Agricultural Science, Cambridge,
(1985), 105, 703-714.

Macmillan, R. H. 1985. Engineering problems in measuring of draft animal
performance. In: Draft Animal Power for Production. ACIAR Proceedings Series #
10. Proceedings of an International Workshop Held at James Cook University,
Townsville, Queensland, Australia. July 10-16, 1985. J. W. Copland (Ed).

Manetsch, T. J. and G. L. Park. 1993. The system approach as a problem solving
methodology. In: Systems Analysis and Simulation with Applications to Economic
and Social Systems. Department of Electrical Engineering, Michigan State University.

Mathers, J. C. and Jennifer C. Sneddon. 1985. Effect of ambient temperature on the
energy costs of activity by tropical cattle. Centre for Tropical Veterinary Medicine,
University of Edinburgh, Easter Bush, Roslin, Midlothian EH25 9RG

Mohan Raj, D. R. (Ed) 1992. Sustainable food production in sub-Saharan Africa.
International Institute of Tropical Agriculture, Pragat Art Printers, Hyderabad, India

McIntire, John. 1990. Evolution of farming systems and what it means for animal traction
research in Africa. In: The Role of Draft Animal Technology in Rural Development.
Proceedings of the International Seminar held at Edinburgh Scotland 2-12 April
1990. Pudoc Scientiﬁc Publishers, Wageningen, 1992.

Muchiri, G. 1984. Farm equipment innovations for smallholders in semi-arid Kenya. In:
Farm equipment innovations in Eastern and Central Southern Africa. A study

217

prepared for the ILO Oﬂice within the ﬁamework of the World Employment
Programme. Iﬁikhar, A. and Bill Kinsey (Eds) 1984 Gower. Biddles Ltd, Guilford
and King 's Lynn. GB.
Muchiri, G. 1981. Farm equipment innovations for small-holder semi-arid agriculture
(FEDSSA). A system approach with special reference to semi-arid Kenya.
Unpublished Ph D manuscript, University of Nairobi.

Muchiri, G. and S. D. Minto. 1977. An approach to research and development in
agricultural mechanization in developing countries. In: Report of the Rural
Technology Meet for East, Central and Southern Aﬁ'ica, Arusha, Tanzania.
Commonwealth Secretariat, London

Mueller, P. J. M. T. Jones. R. E. Rawson, P. J. Van Soest and H. F. Hinttz. 1994. Effect
of increasing work rate on metabolic responses of the donkey (Equus asinus). Journal
Applied Physiology. 77(3): 1431-1438, 1994.

Mukasa-Mugerwa, E. 1989. A review of reproductive performance of female Bos indicus
(Zebu) cattle. ILCA Monograph No. I 6 ILCA, Addis Ababa, Ethiopia.

Mutebwa, A. B. 1979. Determination of mechanization levels with effect to land size and
labor in semi-arid areas of Kenya. M Sc. Thesis. Department of Agricultural
Economics, University of Nairobi, Nairobi, Kenya.

Naylor, T. H., J. L. Balintfy, D. S. Burdick and K. Chu. 1986. Computer simulation
techniques. John Wiley & Sons, Inc. New York.

National Research Council. 1996. Nutrient requirements of beef cattle. Seventh Revised
Edition, 1996. National Academy Press. Washington, D. C. 1996.

National Research Council. 1981. Nutritional energetics of domestic animals & glossary
of energy terms. National Academy Press. Washington, D. C. 1981.

Neelamkavil, Francis. 1987. Computer simulation and modelling. John Wiley & Sons,
Chichester, U. K.

Ogata, K. 1978. Systems Dynamics. Prentice-Hall, Englewood Cliffs, New Jersey.

O’Neill, D. H. 1989. Instrumentation for the measurement of draught animal
performance. ILCA Bulletin. 33, April 1989.

O’Neill, D. H. and D. C. Kemp. 1989. A comparison of work outputs of draught oxen.
Journal of Agricultural Engineering Research. (1989). 43, 33-44.

Oudman, L. 1994. Work on animal power harness technology in Kenya. In: Improving
Animal Traction Technology. Proceedings of the ﬁrst workshop of the Animal
Traction Network for Eastern and Southern Africa (ATNESA) held 18-23 January

218

1992, Lusaka, Zambia. Technical Center for Agricultural and Rural Cooperation
(ACP-EU Lorne Convention) Wageningen, The Netherlands.

Pearson, Anne R. and R. F. Archibald. 1989. Biochemical and hematological changes
associated with short periods of work in draught animals. Animal Production. 1989.
48:3 75 -384.

Pearson, Anne R. 1985. Physiological changes associated with work: some lessons from
the horse. In: Draft Animal Power for Production. ACIAR Proceedings Series # 10.
Proceedings of an International Workshop Held at James Cook University,
Townsville, Queensland, Australia. July 10-16, 1985. J. W. Copland (Ed).

Pearson, 8., Eric Monke, Gem Argwings-Kodhek, Francisco Avillez, Mulinge
Mukumbu, Stefano Pagiola, Daniel Sellen, and Alex Winter-Nelson. 1995.
Agricultural Policy in Kenya. Cornell University Press, Ithaca.

Phillip, D. O. A., G. O. I. Ababu and S. A. Ingawa. 1986. Economic implications of
animal power at the small scale level in the savanna zone of northern Nigeria: a linear
programming simulation of farmer circumstances. In: Animal Power in Farming
Systems. The Proceeding of the Second Animal Traction Networkshop Held
September 19-25, 1986. Freetwon, Sierra Leone.

Pingali, P., Yves Bigot, and Hans P. Binswanger. 1987. Agricultural Mechanization
and the Evolution of F arming Systems in Sub-saharan Aﬁica.

Rautaray, S. K. 1987. Work output of bullocks during ﬁeld operations on black soils. In:
N. S. L. Srivastava and T. P. tha (Eds), Utilization and Economics of Draught
Power. Proceedings of the National Seminar on Animal Energy Utilization, Bhopal.
India. pp 69-87.

Republic of Kenya. 1970. Economic Survey I 970. Central Bureau of Statistics.
Nairobi, Kenya.

Republic of Kenya. 1963. Economic Survey 1963. Central Bureau of Statistics.
Nairobi, Kenya.

Republic of Kenya. 1981. Sessional paper on National Food Policy. (4).

Republic of Kenya. 1987. Statistical Abstracts I98 7, Central Bureau of Statistics,
Nairobi, Kenya.

Republic of Kenya. 1983. Statistical Abstracts 1983, Central Bureau of Statistics,
Nairobi, Kenya.

Republic of Kenya. 1978. Statistical Abstracts 1978, Central Bureau of Statistics,
Nairobi, Kenya.

219

Richards, J. I . and P. R. Lawrence. 1984. The estimation of energy expenditure from
heart rate measurements in working oxen and buffalo. Journal of Agricultural
Sciences, Cambridge. 1986. 102, 711-717

Rockwell, Thomas H. 1965. Use of simulation methodology for solution of operational
system problems. In: Systems Engineering in Agriculture. Paper presented at a
meeting of the American Association for the advancement of science at Berkey,
California. ASAE Publication. PROC-567, St. Joseph, Michigan, 49085.

Ruttan, V. W. and Hayami, Y. 1990. Induced innovation model of agricultural
development. In: Agricultural development in the third world. The John Hopkins
University Press, Baltimore.

Ruttan, V. W. 1990. Models of agricultural development. In: Agricultural development
in the third world. The John Hopkins University Press, Baltimore

Semenye, P. P. 1979. The estimation of liveweight of Zebu cattle from heart girth
measurements. Working Document. International Livestock Center for Africa,
Nairobi. Spetember 1979.

Singh, Gajendra, S. C. L. Premi and V. M. Salogkhe. 1989. Perfomance of bullocks
under different load and climatic conditions. Agricultural Mechanization in Asia,
Africa and Latin America. Vol. 20 No. 4 Autumn I 989. Pp I 5-22.

Silberfein, Marilyn. 1989. Rural change in Machakos, Kenya University Press of
America, Inc. 4720 Boston Way Lanham, MD 20706

Sims, B. G. and A. A. Ramirez.l989. Draft oxen energy expenditure: A Mexican case
study. Agricultural Mechanization in Asia, Aﬁica and Latin America. Vol. 20 No. 2
1989.

Sneddon, Jennifer C. 1986. The effect of treadmill slope on the relationship between
heart rate and energy expenditure in cattle walking on treadmills. Journal of
Agricultural Sciences, Cmbridge. 1986. 106,433-436.

Sneddon, Jennifer C., J. C. Mathers and Claire J. Thomsion. 1984. Prediction of energy
expenditure from heart rate measurements in exercising cattle. Centre for Tropical
Veterinay Medicine, University of Edinburgh, Easter Bush, Roslin, Midlothian EH25
9RG.

Starkey, Paul. 1990. Change in animal traction in Africa and Asia: Implications for
development. In: The role of draﬁ animal technology in rural development.
Proceedings of the international Seminar, 2-12 April I 990, Edinburgh, Scotland.
Pudoc Scientific Publishers, Wageningen, 1992

Starkey, Paul. 1988. Animal traction directory. Vieweg & Sohn Braunchweig/Wiesbaden.

 

220

Starkey, Paul. 1986. The introduction, intensiﬁcation and diversiﬁcation of the use of
animal power in West African farming systems: implications at farm level. In: Animal
power in farming systems. The proceedings of the second West Aﬁica Animal
Traction Networkshop, September 19-25, 1986, Freetown, Sierra Leone.

Stevens, Piet A. 1994. Improving animal-powered tillage systems and weeding technology.
In: Improving Animal Traction Technology. Proceedings of the ﬁrst workshop of the
Animal Traction Network for Eastern and Southern Aﬁica (ATNESA) held 18-23
January 1992, Lusaka, Zambia. Technical Center for Agricultural and Rural
Cooperation (ACP-EU Lome Convention) Wageningen, The Netherlands.

Thakur, T. H., B. Singh and M. P. Singh. 1989. Effects of load on physiological repsonse
of buffaloes in rotary mode. Agricultural Mechanization in Asia, Aﬁica and Latin
America. Vol. 20 No. 4 Autumn 1989. pp 15-22. Check the pages for this article!!

Tiffen, Mary., Montore. M and F. Gichuki. 1994. Population growth and environment
recovery. Policy lessons from Kenya. International Institute for Environment and
Development. Gatekeeper Series No. 45

Tokida, K. 1992. Tillage machinery models for rice farming projects in Kenya.
Unpublished Ph D dissertation. Michigan State University. East Lansing, Michigan.

de Vries, H. C. P. 1986. The combi-plow—an animal implement system in Kenya.
Am Vol. I 7N0. I I 986 pp 27-30.

Westley, S., and B. F. Johnston. Eds. 1975. Proceedings of a workshop on farm
equipment innovations for agricultural development and rural industrialization.
Occasional Paper No. 16, Institute of Development Studies, University of Nairobi,
Nairobi, Kenya

Wilmore, Jack H. and David L. Costill. 1994. Physiology of sport exercise. Human
Kinetics. P 0 Box 5076, Champaign, IL 61825-5076

Yamarnoto, S., J .A. McLean and A. J. Downie. 1979. Estimation of heat production from
heart rate measurements in cattle. British Journal of Nutrition 42, 507-513.

Young, D. L. 1972. The estimation of liveweight from heart girth within speciﬁed age/sex
groups of Kenyan range cattle. East African Agricultural and Forestry Journal.
38: 193-200.

Zerbini, E., T. Gemeda, D. H. O’Neill, P. J. Howell and R. C. Schroter. 1992.
Relationship between cardio-respiratory parameters and draught work output in F1

crossbred dairy cows under ﬁeld conditions. Animal Production. 1992. 55:1-10.

Zeuner, F. E. 1963. A history of domestic animals. Harper and Row. New York.

 

 

Pl GQN

S
WWW
122

S)

TATE UNIV. LIBRARIES
WWWWWWWWWW
301 885422

65

ICHI
WW
:3