§;\.n..a¢.....u:

“£4

 

 

-"~¢a 1o

uh.

u
~
w
..
. .
.
.
u
—
.

:lih's

.nhi.‘

5. .423 $481,:

 

mummlil’l‘filllmlllillllllmuml

3 1293 01772 1105

This is to certify that the

dissertation entitled

IMPACT OF LEGUME AND FERTILIZER NITROGEN
ON SMALLHOLDER MAIZE (Zea mays L.) CROPPING SYSTEMS
IN NORTHERN ZIMBABWE

presented by

PETER J ERANYAMA

has been accepted towards fulfillment
of the requirements for

Ph.D. degree in CSS

 

tau/3M

Major professor

Date {/36/73

MS U is an Affirmative Action/Equal Opportunity Institution 0- 12771

 

LIBRARY

Michigan State
University

PLACE IN RETURN BOX to remove this checkout from your record.
To AVOID FINE return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ma mus-m4

IMPACT OF LEGUME AND FERTILIZER NITROGEN ON SMALLHOLDER
MAIZE (Zea mays L.) CROPPIN G SYSTEMS IN NORTHERN ZIMBABWE

By

PETER JERANY AMA

A DISSERTATION

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

DOCTOR OF PHILOSOPHY

Department of Crop and Soil Sciences

1998

ABSTRACT

IMPACT or LEGUME AND FERTILIZER NITROGEN ON SMALLHOLDER
MAIZE (Zea mays L.) CROPPING SYSTEMS IN NORTHERN ZIMBABWE

By

Peter Jeranyama

Growing maize (Zea mays L.) in rotation or intercropped with legumes may
maintain soil fertility and prevent yield declines associated with smallholder cropping
systems of Zimbabwe. This research was conducted in Zimbabwe on a Typic
Kandiustalf to (i) evaluate the impact of relay-intercropping a food legume (cowpea;
Vigna unguiculata L.) and a tropical forage legume (sunnhemp; Crotalaria juncea L.)
into maize, and (ii) assess the effects of a systematic rotation of maize with groundnut
(Arachis hwogaea L.) on maize yields and economic returns.

Relay intercropping legumes into maize fertilized at 60 kg N ha‘1 did not result
in yield reductions of the companion maize crop. However, relay-intercropping
legumes into maize fertilized at 120 kg N ha" was associated with yield declines of
20-34% for a companion maize crop. In the subsequent year, maize grain yields were
increased by 20% following maize-legume intercrops relative to continuous maize
when no fertilizer was used. Maize grain yield increases following maize and legume
intercrops were sufficient to pay legume seed outlays in the intercropping year. The

research suggests that intercropped annual herbaceous legumes when integrated with

small amounts of inorganic N fertilizer offers a strategy to meet the N needs on
smallholder farms of Zimbabwe.

Maize grain yields were 0.1-2.2 Mg ha’1 higher following groundnut than
following maize. Fertilizer needs of maize following groundnut were reduced by up to
72 kg N ha’1 compared to continuous maize. However, these results were sensitive to
rainfall distribution. A marginal benefit cost analysis Showed that continuous maize at
92 kg N ha‘1 optimized marginal benefits when compared to rotation in a scenario
where family labor had an opportunity cost. The low groundnut yields and little yield
improvements for maize following groundnut on-farm, and the high labor costs
associated with groundnut made the rotation less profitable than continuous maize,
especially when maize was grown with some fertilizer. The results for groundnut-
maize rotation underline the need for research to (i) increase the yield of groundnut on
smallholder farms and (ii) reduce the associated labor costs in producing groundnut

without adding much to cash costs.

ACKNOWLEDGEMENT

The task of producing a dissertation of this quality requires much help and
encouragement. I thank Dr. Oran B. Hesterman, who served as my major professor for
most of my stay at Michigan State University and Dr. Stephen R Waddington of
CIMMYT-Zimbabwe served as my local advisor. I relied on their wise counsel and
mentorship. Dr. Richard R Harwood served as major professor in the last phase of my
studies at a time I thought mission was impossible. I am indebted to Dr. Richard H.
Bernsten, who taught me economic analysis and Dr. Richard W. Ward, who served on
my guidance committee. Special thanks to Drs. Boyd Ellis, Eunice Foster and
Lawrence Copeland who took a keen interest in my progress.

I could not have done this research without the sacrifices of the sweet heart of
my dreams-wife Letina and son Bongani "Bobo" Tinotenda. They let me take a 2-yr
leave from family while I pursued this research. Letina and Bobo, I cannot thank you
enough.

This research was funded by the Rockefeller Foundation and the WK. Kellogg
Foundation. Professional companionship of Drs. Anil Shrestha, Irvine Mariga, Daniel
Rasse and Mr. Johannes Karigwindi is acknowledged. Milton Kamwendo, the talk
shows at Msasa Park were medicines to my soul. Finally ’thanks be unto the Lord
Jesus Christ who is able to do exceedingly above all that I ask or think, according to

the power that works in me’.

iv

PREFACE

Chapters 1 and 2 of this dissertation are written in the style required for
publication in the Agronomy Journal. Chapter 3 is written in the style required for

publication in the Agricultural Systems.

TABLE OF CONTENTS

LIST OF TABLES
LIST OF FIGURES
CHAPTER ONE: RELAY-INTERCROPPED LEGUMES
lNTO MAIZE (Zea may L.) SYSTEM IN ZIMBABWE
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS AND DISCUSSION
Legume herbage biomass
Legume N yield
Relay-intercropped maize yields
Maize grain N uptake
Subsequent maize response to legume and fertilizer N
Fertilizer replacement values
CONCLUSIONS
REFERENCES
CHAPTER TWO: MAIZE(Zea mays L.) YIELD AND NITROGEN
UPTAKE FOLLOWING GROUNDNUT (Arachis hypogaea L.) ON
SMALLHOLDER FARMS IN NORTHERN ZIIVIBABWE
ABSTRACT
INTRODUCTION

MATERIALS AND METHODS

vi

PAGE
viii

12
12
13
14
15
16
18
31
32

37

37

38

41

RESULTS AND DISCUSSION
Groundnut yields
Maize yields following groundnut
Fertilizer replacement value

CONCLUSION

REFERENCES

CHAPTER THREE: ECONONIIC ANALYSIS OF SMALLHOLDER
MAIZE CROPPING SYSTEMS IN NORTHERN ZIMBABWE

ABSTRACT
INTRODUCTION
ECONOMIC ANALYSIS AND ASSUMPTIONS
ECONOMIC ASSESSMENT
Continuous maize versus maize and groundnut rotation
Maize-legume intercropping
Second year following maize-legume intercrops

CONCLUSION

REFERENCES

vii

PAGE
44
44
44
48
63
64

66

66
68
71
74
74
78
79
92

93

LIST OF TABLES

TABLE TITLE PAGE

1.1 Trophic soil properties at the beginning of study 20
at Domboshava.

1.2 Effect of fertilizer N on legume herbage biomass (DM) 21
and N yield at 75 DAP in 1996.

1.3 Effect of fertilizer N on legume herbage biomass (DM) 22
and N yield at 75 DAP in 1997.

1.4 Fitted Regression Equations for Maize grain yield (GR), 23
grain N content (GRN) and total above-ground biomass (TDM)
in Maize-Legume Intercrop systems as a function of fertilizer
N applied (x).

1.5 Regression equations for maize grain yield (GR), grain N 24

content (GRN), total above-ground biomass (TDM) and
total N in above-ground biomass (TN) as a function of
fertilizer N applied (x) in the subsequent.

2.1 Trophic soil properties at Chinyika and Chiduku in Zimbabwe 50
at the beginning of the study.

2.2 Groundnut kernel, haulm and plowdown N yield at three 51
locations in Zimbabwe.

2.3 Effect of cropping system on unfertilized maize grain yield 52
at three locations in northern Zimbabwe.

2.4 Effect of cropping system on unfertilized second maize crop 53
following groundnut at Domboshava, Zimbabwe.

2.5 Fitted Regression Equations for maize grain yield (GR), grain 54

N content (GRN) and total above-ground biomass (TDM) in two
maize cropping systems as function of fertilizer N applied (x).

viii

2.6

3.1

3.2

3.3

3.4

3.5

3.6

3.7

3.8

3.9

3.10

A31]

Fertilizer replacement values of a maize and groundnut
rotation cropping system at Domboshava.

Partial budget for the first maize crop in a continuous maize
cropping pattern and groundnut crop in a groundnut-maize
rotation in 1994/95 at Domboshava, Zimbabwe.

Partial budget for the second maize crop in a continuous maize
cropping pattern and maize following the groundnut crop in a
groundnut-maize rotation in 1995/96 at Domboshava, Zimbabwe.

Partial budget for the third maize crop in a continuous maize
cropping pattern and second maize crop following groundnut in a
groundnut-maize rotation in l996/97 at Domboshava, Zimbabwe.

Summary of net benefits associated with two fertilizer rates
over three seasons (1994-1997) in a continuous maize cropping
pattern and groundnut-maize rotation at Domboshava, Zimbabwe.

Partial budget for the first maize crop in a continuous maize
cropping pattern and groundnut crop in a groundnut-maize
rotation in 1995/96 at Chinyika, Zimbabwe.

Partial budget for the second maize crop in a continuous maize
cropping pattern and the maize crop following groundnut in a
groundnut-maize rotation in l996/97 at Chinyika, Zimbabwe.

Summary of net benefits associated with two fertilizer rates over
two seasons (1995-1997) at Chinyika, Zimbabwe.

Partial budget for maize + legume intercrops and maize
alone cropping systems at Domboshava, Zimbabwe.

Fitted Regression Equations for Added maize value (Zim S)
and added cost of fertilizer as a function of fertilizer N applied
(x) in the subsequent year (maize + legume - maize).
Sensitivity analysis on net benefits of a maize-cowpea

and maize-sunnhemp intercrops as influenced by cowpea

or sunnhemp seed cost changes.

Price and labor data used for cost-benefit analysis.

ix

PAGE

55

81

82

83

84

85

86

87

88

89

9O

94

FIGURE

1.1

1.2

1.3

1.4

1.5

1.6

2.1

2.2

2.3

2.4

2.5

LIST OF FIGURES

TITLE

Effect of relay-intercropping legumes and fertilizer N
on maize grain yield in the interseeding year (1996).

Effect of relay-intercropping legumes and fertilizer N
on maize grain yield in the interseeding year (1997).

Effect of relay-intercropping legumes and fertilizer N

on maize grain N content in the interseeding year (1996).

Effect of relay-intercropping legumes and fertilizer N

on maize grain N content in the interseeding year (1997).

Subsequent maize grain yield (1997) response to
fertilizer N and previous year’s cropping system.

Effect of relay-intercropped legumes on subsequent
maize grain N uptake.

Rainfall data for Chiduku, Chinyika and Domboshava
for the l996/97 cropping season.

Subsequent maize grain yield response to groundnut
and fertilizer N applied at Domboshava.

Second maize crop after groundnut response to
fertilizer N applied at Domboshava.

Subsequent maize grain yield response to groundnut
and fertilizer N applied at Chinyika.

Maize grain N content response to groundnut
and fertilizer N applied at Domboshava.

PAGE

25

26

27

28

29

3O

49

56

57

58

59

26

2.7

2.8

3.1

Maize grain N content response to groundnut
and fertilizer N applied at Chinyika.

Maize total above-ground biomass response to
groundnut and fertilizer N applied at Domboshava.

Maize total above-ground biomass response to
groundnut and fertilizer N applied at Chinyika.

Effect of rate of N application on added maize value (ZimS)

and added fertilizer cost (Zim$) at Domboshava following
maize-legume intercrops.

xi

PAGE

60

61

62

91

CHAPTER ONE

RELAY-INTERCROPPING OF ANNUAL LEGUMES INTO A MAIZE
(Zea mays L.) SYSTEM IN ZIMBABWE

ABSTRACT

Declining maize (Zea mays L.) yields in the smallholder cropping systems of
northern Zimbabwe present the need to develop a more sustainable production system.
Increased maize production will continue to emphasize the use of inorganic fertilizers.
However, rising real prices of inorganic fertilizers is driving smallholder maize
production towards lower levels of inorganic fertilizer inputs. Meanwhile, legume
intercrops are a source of plant N that can be produced within local environments and
offer a practical complement to inorganic fertilizers.

This article evaluates the impact of relay-intercropping a food legume - cowpea
(Vigna unguiculata), and a tropical forage legume - sunnhemp (Crotolaria juncea) into
smallholder maize in Zimbabwe. Field studies were conducted on a loamy sand
(Typic Kandiustalf) soil at a mid altitude site for two years. The objectives of the
study were to quantify (i) biomass and N yield of legumes intercropped into maize, (ii)
the impact of the legumes on companion maize grain yield and N uptake, and (iii) the
response of the subsequent maize crop to relay-intercropped legumes.

Herbage biomass as dry matter yield ranged from 0.6 to 4.6 Mg ha'1 for

cowpea and 0.9 to 2.9 Mg ha’1 for sunnhemp, over the two years. Highest above-
ground N yield for cowpea was 154 kg N ha"1 compared to 82 kg N ha‘1 for
sunnhemp. Companion maize grain yields were not reduced when legumes were relay-
intercropped into maize fertilized at zero and 60 kg N ha'1 in each of the two years.
However, maize yields were reduced 20 to 34 % when maize-legume intercrops were
fertilized at 120 kg N ha". In the subsequent year, maize grain yields were increased
by 20% following maize-legume intercrops when no fertilizer N was applied,
compared to maize following maize. From fertilizer replacement value calculations,
legumes reduced fertilizer needs of a subsequent maize crop by up to 36 kg N ha".
This study suggests that annual herbaceous legumes relay-intercropped into
moderately fertilized (>0-60 kg N ha") maize can maintain yields of the companion
maize crop while enhancing yields in a subsequent maize crop. The study did not

assess long-term effects of the increased crop diversity and cover crops.

INTRODUCTION

Declining maize (Zea mays L.) yields in the smallholder cropping systems of
Zimbabwe present the need to develop a more sustainable production system. To
continue to increase maize production will require emphasis on the use of inorganic
fertilizers (W addington and Heisey, 1997). However, legume cover crops are a source
of plant nutrients that can be produced within local environments and they offer a
practical complement to inorganic fertilizers.

In Zimbabwe, green manures were heavily researched from the 1920’s to
1940’s (Metelerkamp, 1988), and large-scale commercial farms (the second sector in a
dichotomous agriculture) used green manures widely. Although there was no deliberate
effort (with documented evidence) to promote the use of green manures by the
smallholder sector, there were informal reports that some smallholders did use some
green manures such as sunnhemp (Crotolaria juncea L.) to maintain soil fertility
(Hikwa et al., 1997). This practice continued until real prices of inorganic fertilizers
fell in the 1950’s and green manures became uneconomic (Tattersfield, 1982).
However, rising real prices of inorganic fertilizers in the recent past and concerns
about the sustainability of current smallholder cropping systems have once again

attracted interest in green manures (Hikwa and Mukurumbira, 1995).

Although green manures have once again become popular among agricultural
scientists, the grong of legume sole-crop green manures in fallows have been
rejected by smallholders because of labor and land constraints. At current fertilizer
costs, most smallholders will grow maize with very little or no fertilizer. There
remains potential to integrate legume cover crops in the existing cropping systems as
intercrops. If legumes are intercropped in a timely manner, competition with the maize
crop for resources (light, water, nutrients) can be minimized while legume herbage N
can be accumulated. This technology is unlikely to directly benefit the companion
crop, but has potential to increase the yields of a subsequent maize crop (J eranyama,
1995)

If food legumes are not to be a net drain on N from the system, they must fix
at least as much N as is removed from the field in grain or other produce when the
legume is harvested (Giller et al., 1994). When an abundant supply of mineral N is
available in the soil profile, legumes preferentially utilize soil N at the expense of N2-
fixation (Allos and Bartholomew, 1956). In such cases, the legume removes more
nitrogen from the soil than it fixes from the atmosphere, which results in a net loss of
available nitrogen for the companion crop. However, as soils in the smallholder farms
have low inherent fertility (Grant, 1970; Mashiringwani, 1983), percent N from N2-
fixation tends to be high and legumes often contribute N to the system in excess of
their own requirements.

When food legumes are used to supply biological N to soil, those with a low

N harvest index (N in harvested grain per unit total above-ground N) will be most

valuable as they are associated with less N removal from the field in harvestable grain.
Values of the N harvest index vary from 90% in soybean to only 25—40% in some
genotypes of cowpea, groundnuts (Arachis hwogea), and pigeon pea (Cajanus cajan)
(Giller et al., 1994). Unlike food crops, forage legumes are usually not intercropped
with cereals in the region (Okigbo and Greenland, 1976), probably because forages do
not contribute directly to the food security of farmers (Kumwenda et al. 1996). There
is often a direct conflict between the need to assure immediate food supply and the
need to assure future food supply by building up soil fertility over a long period.
Kumwenda et al. (1996) noted that farmers discount the value of a benefit that will
only be achieved several years from when investments were made. The most suitable
legumes from a soil fertility perspective are often the most difficult to‘adopt and
usually offer no value for human consumption.

Smallholder farmers rarely plant crops solely for use by livestock. Producers
cannot afford to take risks in subsistence agriculture, so priority of resources, such as
labor, is given to staple food crops. Accordingly, and ironically, legume forages
currently have little place in crop-based systems even though these systems potentially
offer the best opportunities for legume introduction (Thomas and Sumberg,

1995)

The percent of nitrogen from Nz-fixation in intercropped legumes is generally
higher than that of legume monocrops in a given environment as the supply of soil N
available to the legume is reduced by competition for N from the cereal crop

(Rerkasem and Rerkasem, 1988). However, total yield of fixed N is often reduced as

the grain legume occupies less land area and is subject to competition for resources,
particularly for light, from the taller cereal crops (Nambiar et al., 1983).

Two common methods of assessing the N contribution by legume to a cropping
system are total N content in legume herbage biomass and fertilizer replacement value
(FRV) (Hesterman, 1988). The method based on total N legume biomass assumes that
all legume N produced is mineralized and is available to the subsequent cereal crop. In
fact some studies suggest that only 10-30% of the N incorporated in the legume
material is absorbed by the following crop (Ladd et al., 1983; Harris and Hesterman,
1987), with the excess accounted for in soil organic matter, in the inorganic soil N
pool and by losses due to denitrification and leaching.

Fertilizer replacement value (FRV) is defined as the quantity of N fertilizer
required to produce a yield in a crop that does not follow a legume that is identical to
that produced by incorporation of a legume (Hesterman, 1988). Reported FRV’S for
maize-cowpea or maize-black gram (Vigna mango) intercrops in the subsequent year
ranged from 31 to 54 kg N ha‘1 (Nair et al., 1979; Singh, 1983).

There remains controversy as to whether cereals benefit directly from N2 fixed
by intercropped legumes, or whether the enhanced N uptake sometimes observed in
intercrops is simply due to a ’sparing’ of soil N by the legume for use by the cereal
(Agboola and Fayemi, 1972; Remison, 1978; Pandey and Pendleton, 1986).

Legume cover crops are included in cropping systems because they reduce soil
erosion (Giller and Cadisch, 1995), suppress weeds (Exner and Cruse, 1993) and fix

biological N (Giller et al., 1994). However, legume cover crops can also deplete soil

moisture necessary for grain production in semi-arid areas (Baduruddin and Meyer,
1989) and compete for light and nutrient with the main crop (Ofori and Stern, 1987).
There is therefore a need to develop cover crop strategies that comply successfully
with the overarching necessity of water conservation in dryland cropping systems.
Moisture conservation is however, not the major thrust of this study.

Annual herbaceous legumes may provide opportunities for cover crops that
provide biological N and at the same time generate fewer water deficit problems than
longer-lived legumes such as the woody perennials currently being promoted in
agroforestry. However, herbage biomass N from herbaceous plants may be insufficient
to overcome soil N deficiencies in smallholder farms. The integration of small amounts
of inorganic N fertilizer with the organic materials (legume cover crop) offers a
strategy to meet the N needs of smallholder farms (Jama et al., 1997; Waddington and
Heisey, 1977).

Objectives. The objectives of this study were to (i) quantify legume herbage
biomass and N accumulation of a food and a forage legume relay-intercropped into
maize, (ii) evaluate impact of relay-intercropped legumes on the companion maize
crop, and (iii) evaluate response of a subsequent maize crop to relay-intercropped

legumes and compare this with the response to fertilizer N.

MATERIALS AND METHODS

This study was conducted in 1996 and 1997 at Domboshava (elevation approx.
1500 masl), 31 km north east of Harare. The area has Typic Kandiustalf which is
generally infertile with coarse-grained loamy sand soil. Trophic soil properties were
characterized at the beginning of the study (Table 1.1).

Maize (Zea mays L. 3-way cross hybrid R215) was hand-planted at two seeds
per station on tractor disc-plowed land. Designated plots were received 22 kg N ha",
28 kg P ha’1 and 19.5 kg K ha‘l applied as Compound D (8 14 7) before planting.
Additional fertilizer of 0, 60 and 120 kg N ha‘1 as NH,NO3, was applied in designated
plots as a dollop next to each station about 48 days after planting (V6 growth stage of
maize) when the soil moisture level in the ground was at field capacity. Maize plant
spacing was 0.9 m between rows and 0.5 m within row giving a plant population
density of 44, 444 plants per hectare. Each plot was 10 m x 6.4 m.

Weeds were removed by hand as necessary. Two legumes, cowpea (Vigna
unguiculala L.) - a food legume, and sunnhemp (Crotolaria juncea L.) - a forage
legume, were relay-intercropped into maize (two legume rows between adjacent maize
rows) in each of two years. Legumes were planted as intercrops at about 28 days after
maize planting (V4 stage) in pre-assigned plots. Legumes were seeded at in-row

spacings of 10 cm, achieving a plant population of 111, 000 plants per hectare.

The legume seeds were not inoculated with rhizobia before planting, which
corresponds to local farmer practice. Also, an unfertilized plot with maize alone was
included in each replication as control.

Above-ground herbage biomass of legumes was sampled at 45, 60 and 75 days
after planting (DAP) the legumes using hand secateurs from a 0.09 m2 quadrant.
Weeds were hand separated from legumes and legume herbage was dried at 60 C
temperature for at least three days for dry matter yield.

Photosynthetically active radiation (PAR) was measured in 1997 at 50 and 75
DAP (legume) using a Licor sensor meter. Readings were taken just above the maize
canopy, just above understory legume and on the ground under the legume.

Maize grain was harvested from a 1.8 m x 4 In section of the center two rows
in each plot so that the area harvested was 7.2 m2. Grain yields were adjusted to 125 g
kg‘1 moisture content. Maize stover was harvested from a single center-most row in a
0.9 m x 4 In section and yields were expressed as dry matter per hectare.

In the subsequent year, maize following maize-legume intercrops was
established on the same plots. This crop was fertilized with two split applications of 0,
46, 92 and 138 kg N ha", initially as Compound D at planting and NH4NO3 as side

dress at about V6—V8 maize growth stage.

Plant and Chemical Analysis. Total N in maize grain, stover and legume herbage was

 

determined by a modified micro-Kjeldahl method. Dry plant materials were ground in

a Wiley mill to pass through a 2-mm screen. Plant samples of 0.1 g were digested in 4

ml of 18 M HZSO4 with 1.5 g K2S04 and 0.075 g Se catalyst. Following digestion total
NH,+ was determined by spectrophotometry. Total N yield was calculated as the

product of dry matter yield and nitrogen concentration.

Statistical Analysis.

In the relay-intercropping year, the experiment was planted as a randomized
complete block design (RCBD) with treatments replicated three times. Analysis of

variance (ANOVA) was used to analyze treatment differences for grain yield, total-

"”1

above—ground biomass, grain N content and total N uptake of maize. When N applied

K’s—r..—
:4

was significant, response was further partitioned into linear and quadratic trends from
single degree of freedom comparisons and regression equations determined in Proc
Reg of SAS (SAS, 1997).

Legume herbage biomass was analyzed as a repeated measures experiment with
a first order auto-regression correlation type [AR( 1)] over sampling periods in Proc
Mixed of SAS (SAS, 1997). Due to a significant (PS 0.05) three way interaction of
legume x N applied x sampling period, a reduced model of herbage biomass and N
yield was used within a sampling period. In the reduced model, herbage biomass and
N yield was analyzed as an RCBD, with treatments replicated three times.

In the subsequent year, experiment was a RCBD with a split-plot arrangement,
replicated three times. The first year cropping system [maize + lst-yr. N or maize-
legume + 1st-yr. N] were whole plots and 2nd-yr. N rates were subplots. Analysis of

variance using Proc GLM (SAS, 1997) was used to identify treatment effects.

10

Response to N fertilizer rate in the 2nd-yr. were determined by evaluating linear and
quadratic trends from single degree of freedom comparisons. Whenever trends were
significant, regression equations were calculated to determine fertilizer replacement

values (FRV) of relay-intercropped legumes in the preceding year.

11

RESULTS AND DISCUSSION

Legume herbage biomass. A significant legume x nitrogen interaction was
observed in both years for herbage biomass and N yield. Hence legumes herbage
biomass (as dry matter, DM) and N yields are presented separately for each N rate
(Tables 1.2 and 1.3). Herbage biomass ranged from 0.6 to 4.6 Mg ha"1 for cowpea
and from 0.9 to 2.9 Mg ha‘1 for sunnhemp. These herbage biomass yields are similar
to those recorded with pigeon pea (Cajanus cajan L.) when grown alone (2.07-3.21
Mg ha‘1 , Kwatpata 1984) and when intercropped with maize in Malawi (3.0 Mg ha"
Sakala, 1994).

Response of legumes to N fertilizer were determined by evaluating linear and
quadratic trends from a single degree of freedom comparisons. In 1996, only
sunnhemp was linearly related to N applied. Sunnhemp herbage biomass yield was
positively correlated (r = 0.55; PS 0.1) with N rate. However, cowpea and N applied
were not significantly correlated (P_<_ 0.1), but herbage biomass was lowest at the
maximum N applied (Table 1.2). Differences in response by the two species are
mainly due to different grth habits. Cowpea will get more shade from fast covering
high N rate maize, but sunnhemp grows up (erect) into the maize canopy and
intercepts more light. In 1997, both cowpea and sunnhemp linearly responded to N

applied and were negatively correlated with N rate, with r = -0.55 and -0.41

12

respectively, at P_<_ 0.1. Herbage biomass tended to decline with increased N rate
(Table 1.3).

The negative correlation of herbage biomass to applied N could be explained
by the direct effects of shading by associated maize on dry matter production by the
legume. Because dry matter production in crops depends on the efficiency of
interception of photosynthetically active radiation (PAR) (Biscoe and Gallagher, 1977;
Monteith, 1977), shading of the legume understory resulted in low herbage biomass. In
this study, PAR recorded in 1997 shortly before tasselling were 87, 78 and 61 percent
for understory legume in maize-legume intercrops fertilized with 0, 60 and 120 kg N
ha‘1 , respectively (data not shown). The PAR recorded indicated a fair amount of

shading in the understory legume at 120 kg N ha".

L_egume N yield. Legume N yields ranged from 15 to 154 ng ha‘l for cowpea and 23
to 82 kg N ha'1 for sunnhemp (Tables 1.2 and 1.3). Giller and Wilson (1991) have
shown that tropical legumes grown in pure cultures can often accumulate 100-200 kg
N ha‘1 in 100-150 days. The range of N yields reported in our study are somewhat
lower, partly because the legumes were intercropped and were allowed to grow for a
maximum of only 75 days. Legume N yields in intercrops are usually lower than those
from sole legumes because intercrops occupy less land area and are subject to
competition for resources, particularly for light, from the taller cereal crop ( Nambiar

et a1. , 1983).

In 1997, N yields of cowpea and sunnhemp were negatively correlated to N

13

rate (r = -0.56 and -0.52 ; PS 0.1, respectively). Giller and Cadisch (1995) reported a
decrease in N2 fixation by legumes due to the ’problem’ of excessive plant-available
N. Our results suggest that N yields are reduced in the presence of an increased
inorganic N pool and due to the shade effects by maize fertilized at high N rates (<60
kg N ha“. Furthermore, Eaglesham et a1. (1983) concluded from pot studies that
fertilizer N applications in excess of 25 kg N ha‘1 would be likely to inhibit N fixation

of cowpea under field conditions.

Relay-intercropped maize yields. Due to a significant (PS 0.05) cropping system x

nitrogen rate interaction, data are presented as response of maize to cropping system
and fertilizer N. Because legume x nitrogen interactions were not significant, and
main effects of cowpea and sunnhemp were not significantly (PS 0.05) different, data
were averaged across the two legumes. Least square equations for maize grain yield,
grain N content and total above-ground biomass in response to applied nitrogen in
maize-legume intercrop systems and sole maize were calculated (Table 1.4).
Relay-intercropping cowpea and sunnhemp into maize fertilized with zero or 60
kg N ha", was not associated with a significant (PS 0.05) grain yield reduction (Figs
1 .l and 1.2). However, relay-intercropping legumes into maize fertilized with 120 kg
N ha" resulted in significant grain yield reductions of 18% in 1996 (Fig 1.1) and 32%
in 1997 (Fig 1.2), respectively compared to unfertilized sole maize.
Results for zero and 60 kg N ha’1 which are representative of the range of N

rates that smallholders in Zimbabwe often use suggests that legumes could be relay-

14

intercropped into maize without decreasing maize grain yields. In fact, yields were
slightly improved at these N rates, however, not significantly so. Also, our results
agree with those of Haizel (1974) working with maize-cowpea, and Andrews (1972)
and Rees (1986) with sorghum-cowpea intercrop systems, in which no maize yield
suppressions nor increases were observed.

Maize grain yield reductions at 120 kg N ha'1 likely resulted from competition
for resources such as moisture, light and nitrogen with the legume. The maize grain
yield reduction of 18-32% when maize was relay-intercropped with legumes in our
study corresponds to those by other researchers reporting declines in unfertilized maize
yields when intercropped with cowpea of 31% (Haizel, 1974), 33% (W anki et a1.
1982) and 18% (Ofori and Stern, 1986). However, unfertilized cereal yields have been
increased in some studies by 11% (Agboola and Fayemi, 1971) and 45% (Remison,
1978) in maize-cowpea intercrop systems in West Africa.

Competition between species in intercropped systems for growth-limiting
factors is regulated by basic morpho-physiological differences and agronomic factors
such as the proportion of crops in the mixture, fertilizer applications and relative time
of planting (Harper, 1961; Trenbath, 1976).

Maize grain N uptake. Maize grain N uptake was Similar or slightly greater
with the intercrop than with the control (sole maize) when no N fertilizer was applied
(Figs 1.3 and 1.4). At 60 kg N ha", intercropped maize was associated with a
significantly higher grain N uptake in both years. At 120 kg N ha", grain N uptake in

the intercrop system was reduced by 20% in 1996 and 34% in 1997 (Figs 1.3 and 1.4).

15

intercropped into maize without decreasing maize grain yields. In fact, yields were
slightly improved at these N rates, however, not significantly so. Also, our results
agree with those of Haizel (1974) working with maize-cowpea, and Andrews (1972)
and Rees (1986) with sorghum-cowpea intercrop systems, in which no maize yield
suppressions nor increases were observed.

Maize grain yield reductions at 120 kg N ha'1 likely resulted from competition
for resources such as moisture, light and nitrogen with the legume. The maize grain
yield reduction of 18-32% when maize was relay-intercropped with legumes in our
study corresponds to those by other researchers reporting declines in unfertilized maize
yields when intercropped with cowpea of 31% (Haizel, 1974), 33% (Wanki et a1.

1 982) and 18% (Ofori and Stern, 1986). However, unfertilized cereal yields have been
increased in some studies by 11% (Agboola and Fayemi, 1971) and 45% (Remison,
1 978) in maize-cowpea intercrop systems in West Africa.

Competition between species in intercropped systems for growth-limiting
factors is regulated by basic morpho-physiological differences and agronomic factors
such as the proportion of crops in the mixture, fertilizer applications and relative time
of planting (Harper, 1961; Trenbath, 1976).

Maize grain N uptake. Maize grain N uptake was similar or slightly greater
with the intercrop than with the control (sole maize) when no N fertilizer was applied
(Figs 1.3 and 1.4). At 60 kg N ha“, intercropped maize was associated with a
significantly higher grain N uptake in both years. At 120 kg N ha", grain N uptake in

the intercrop system was reduced by 20% in 1996 and 34% in 1997 (Figs 1.3 and 1.4).

15

In cereal-legume intercropping, the legume component is capable of fixing
atmospheric N2 under favorable conditions and this is thought to reduce competition
for N with the cereal (Trenbath, 1976). In the absence of an effective Nz-fixing
system, both the cereal and intercropped legume compete for available soil N (Ofori et
al. 1987). Another theory states that in the presence of adequate to excessive soil N,
legumes switch off N2 fixation and utilize the readily available soil N. This theory
seems to adequately explain the grain N uptake pattern in our study. Our results
suggest that legumes were actively fixing N2 when plots were fertilized with zero or
60 kg N ha", but Nz-fixation was reduced by the higher N rate. Decreased N2 fixation

resulted in competition for soil N, hence reducing maize grain N uptake.

Subsgruent Maize Response to Legume and Fertilizer Nitrogen

Maize Yields: No Nitrogen Applied. Maize following maize that had been relay-
intercropped with a legume produced 20% higher grain yields than maize following
maize with no legumes (continuous maize) (Fig 1.5). There was no significant legume
x nitrogen fertilizer interaction on subsequent maize grain yield and total above-ground
biomass. Therefore data presented are means of two legumes (cowpea and sunnhemp).
Most studies of maize-cowpea intercropping conducted in Zimbabwe have not assessed
effect in a subsequent year, in spite of reporting either an intercrop advantage (e. g.
Mariga, 1990) or disadvantage (eg. Shumba et al., 1990; Natarajan and Shumba,
1990) in the intercropping season. This study is an attempt to provide an assessment

on effects of intercropped legumes in the subsequent season.

16

Maize improvements of 33% following pigeon pea in Malawi (Kwatpata,
l 984) and 21% following sunnhemp in Tanzania (Temu, 1982) have been reported.
Agboola (1980) working with one season fallows of pigeon pea, mucuna (Calopogium
mucunoides L.) and cowpea in a sub humid province of Nigeria observed increases of
subsequent maize crop yields of 10-30%. Nair et a1. (1979) found a maize grain yield
increase of 34% in the year following a maize-cowpea intercrop.

Grain N content in the subsequent year was significantly affected by the
legume type. Maize following maize intercropped with cowpea was associated with a
l 6% higher grain N content, while sunnhemp reduced maize grain N uptake by 50%
when no N fertilizer was applied (Fig 1.6). Increased maize grain N uptake associated
with the previous cowpea crop seems to be a response to an enlarged soil N pool. A
decline of about 50% with sunnhemp may be due to a net soil N immobilization.
However, we cannot make formal conclusion, as soil N pools were not measured in

this study.

Maize response to Fertilizer and Legu_me Nitrogen. There was a positive response of
maize grain yield (GR), grain N content (GRN) and total above-ground N uptake (TN)
to fertilizer N and legume from the previous year. Fitted regression equations were
calculated for maize grain yield, grain N content and total N uptake as a function of
fertilizer N applied (Table 1.5).

Maize grain yields were greater at all fertilizer N levels following a maize-

legume intercrop compared to continuous maize, but not significantly so at the highest

17

N level. Maximum yield benefits of the maize-legume intercrop in the subsequent year
appeared to be realized when maize was fertilized with 46-92 kg N ha'1 , while at
higher N levels diminishing returns were apparent (Fig 1.5). Lack of maize response
to higher N rate in 1996/97 was due to incessant and excessive rainfall received
especially in the months of January and February (rainfall data is presented in Fig 2.1).

Grain N content of maize following the maize-cowpea intercrop was always
higher than that of the continuous maize, but not always at a statistically significant
level (PS 0.05). However, with maize following the maize-sunnhemp intercrop, grain
N content was always lower than that of continuous maize (Fig 1.6). A possible
explanation of this response is (i) immobilization of sunnhemp, absorbing inorganic
sources of N or (ii) lack of synchrony between legume N release and maize grain N
uptake. However, if the latter occurred, then we would expect grain N uptake of the
maize following maize-sunnhemp intercrop to be similar to that of continuous maize,
but this was not the case in our study.
Fertilizer Replacement Values. For the cropping systems under study to be acceptable
to both the farmers and researchers, there must be a convincing yield or N uptake
improvement in the subsequent year, with no yield reduction in the intercropping
season.

One way to assess improvements in the subsequent year is to evaluate fertilizer
replacement values (FRV). For an FRV to be valid, yield of a subsequent crop
following the legume should be Significantly higher than that of the non-legume

control. Based on this criterion, FRV could only be calculated based on grain yield,

18

grain N content and total N uptake of maize.

Highest FRV’S were calculated based on grain N content and the least on grain
yield (data not shown). Because the lowest FRV of 18 kg N ha’1 was obtained with
grain yield, this suggests modest residual N benefits derived from the cropping system.
Singh ( 1983) estimated N benefits to subsequent cereal crops after cereal-legume
intercrops. He obtained N fertilizer equivalents of 3 kg ha'1 with soybean,

3 1 kg ha‘1 with greengram, 46 kg ha’1 with grain cowpea and with groundnut, and

54 kg ha’1 with forage cowpea.

19

Table 1.1. Trophic soil properties at the beginning of study at Domboshava.

 

Soil texture Loamy sand
pH (CaClz) 4.5

C (%) 0.46
Mineralizable N (ppm) 24.13
P20, (ug g") 9.10
CECT (me %) 1.92
TEB1 (me %) 5.17

 

I Cation exchange capacity
1 Total exchangeable bases

20

Table 1.2. Effect of fertilizer N on legume herbage biomass (DM) and N yield at 75

 

 

 

DAP in 1996.
N applied Cowpea Sunnhemp
(kg ha“)
DM N yield DM N yield
(Mg ha") (kg ha") (Mg ha") (kg ha“)
0 1.05 27.21 0.89 22.71
60 1.93 50.42 2.34 57.02
120 0.60 15.28 2.92 76.66
CV (%) 32 24 40 22
LSD (0.05) 0.84 22.45 1.21 24.58

 

21

Table 1.3. Effect of fertilizer N on legume herbage biomass (DM) and N yield at 75

 

 

 

DAP in 1997.

N applied Cowpea Sunnhemp

(kg ha“)
DM N yield DM N yield
(Mg ha") (kg ha") (Mg ha") (kg ha")

0 4.57 154.32 2.86 82.07

60 2.72 104.46 1.80 l 51.01

120 2.01 73.28 1.32 44.24

CV (%) 28 25 32 19

LSD (0.05) 1.68 31.54 1.36 24.34

 

22

Table 1.4. Fitted Regression Equations for Maize grain yield (GR), grain N content
(GRN) and total above-ground biomass (TDM) in Maize-Legume Intercrop
Systems as a function of N fertilizer applied (x).

 

 

Cropping system EquationT R2 Sign
Intercropping year : 1996
Maize alone GR = 1.36 + 0.01(x) 0.75 0.003
GRN = 15.88 + 0.13(x) 0.73 0.003
TDM = 3.05 + 0.019(X) 0.61 0.01

Maize + legume GR = 1.45 + 0.025(x) - 0.0002(x2) 0.78 0.01
GRN = 16.72 + 0.29(x) - 0.002(x2) 0.79 0.01
TDM=3.54 + 0.04(x) - 0.0003(x2) 0.68 0.03

Intercropping year: 1997

Maize alone GR = 2.12 + 0.04(x) 0.63 0.01
GRN = 21.2 + 0.5(x) 0.62 0.01
TDM = 7.75 + 0.04(x) 0.44 0.05

Maize + legume GR = 1.92 + 0.103(x) - 0.0007(x2) 0.74 0.0001
GRN = 17.38 + 0.98(x) - 0.006(x2) 0.73 0.0001
TDM=7.54 + 0.116(x) - 0.0008(x2) 0.42 0.017

 

TLinear equations were based on 7 degrees of freedom, while quadratic equations
were based on 6 degrees of freedom.

23

iii—‘93

Table 1.5 : Regression equations for maize grain yield (GR), grain N content (GRN),
total above-ground biomass (TDM) and total N in above-ground biomass (TN)
as a function of N fertilizer applied (x) in the subsequent year
(maize + legume-maize).

 

 

Cropping system EquationT R2 Sign

Maize - Maize GR = 1.22 + 0.045(x) - 0.002(x2) 0.63 0.01
GRN =15.41 + O.334(x)-O.0013(x2) 0.71 1e-04
TN =37.45 + 0.24(x) 0.58 0.001

Maize + cowpea - Maize GRN = 22.78 + 0.153(x) 0.51 0.001
TN = 41.28 + 0.228(x) 0.48 0.01

Maize + sunnhemp-Maize GRN = 5.99 + 0.48(x) - 0.0022(x2) 0.75 0.002
TN = 32.11 + 0.519(x) - 0.0024(x2) 0.78 0.001
Maize + legumes: - Maize GR = 1.94 + 0.039(x) - 0.0002(x2) 0.7 le-04

 

TLinear equations were based on 10 degrees of freedom, while quadratic equations
were based on 9 degrees of freedom.

1Refers to combined effect of cowpea and sunnhemp and equation is based on 21

degrees of freedom.

24

1

F.-.-—.._
i , .
.1

 

 

9: 1a.._-‘|C.Dlt.ni
s

 

 

O Maize alone
A Maize + legume
"7" O
on
E
1:
21‘.»
>4
a 0
Lu.
an
is 1 -
c6
2
0 T I I 1 T
0 30 60 90 120

N applied (kg ha")

Figure 1.1 Effect of relay-intercropped legumes and fertilizer N on maize grain yield
in the interseeding year (1996).

25

 

O Maize alone
7 7 A Maize + legume

Maize grain yield (Mg ha")

 

ya-.. 3.:
s ' ' ‘
L

 

 

l I T 1 I 1

0 30 60 90 120

N applied (kg ha'l)

Figure 1.2 Effect of relay-intercropped legumes and fertilizer N on maize grain yield
in the interseeding year (1997).

26

Maize grain N content (kg ha!)

 

34

32‘

O Maize alone
A Maize+legume

304
28‘
26-
244
224
204
18-

16~

 

 

 

 

14 I I I I r
0 30 6O 90 120

N applied (kg ha'l)

Figure 1.3 Effect of relay-intercropped legumes and fertilizer N on maize
grain N content in the interseeding year (1996).

27

I l r‘_m--Gfi‘afi1

ATE— wxv Eat—Cu 2 52:1 0.2va

 

 

 

90

O Maize alone
A Maize+legume

80 n

70 -

40

l

30 a

Maize grain N content (kg ha'l)
‘6

l

 

20

 

. .- -7-11

 

 

10 T l 1 l I
0 30 60 90 120

N applied (kg ha'l)

Figure 1.4 Maize grain N content response to relay-intercropped legumes
and fertilizer N in the interseeding year (1997).

28

Maize grain yield (Mg ha!)

 

 

 

 

 

O maize - maize
A maize + legume - maize
4 _i
3 _.
2 ..
o
1 i
0 I I T I I
o 30 60 90 120
N applied (kg ha“)

Figure 1.5 Subsequent maize grain yield (1997) response to fertilizer N

and previous year's cropping system.

29

150

I” R nud‘saj.w
. I

6O

 

40“

30‘

20‘

Grain N Uptake (kg ha'l)

 

 

O Maize —maize
A Maize + cowpea - maize
I Maize + sunnhemp - maize

 

 

Figure 1.6. Effect of relay intercropped legumes on subsequent maize grain N uptake

I I

50 100

N applied (kg ha")

30

150

CONCLUSIONS

Herbage biomass and N yield of intercropped legumes were influenced by N
application. A high fertilizer N rate (120 kg ha") was associated with decline in
legume herbage biomass and subsequent N yields compared to 60 kg N ha'1 .

Companion maize grain yields were not reduced when legumes were relay-
intercropped into maize fertilized with zero and or 60 kg N ha". However, at 120 kg
N ha", yields were reduced by 18-32%. Also, grain N uptake was not reduced at zero
and 60 kg N ha‘1 fertilizer levels in the intercropping system, but were reduced by 20-
34% at 120 kg N ha“.

In the subsequent year, maize grain yields were increased by 20% following
legume intercrops compared to continuous maize. Maize grain N uptake following the
maize-cowpea intercrop was increased by 16% while grain N uptake following the
maize-sunnhemp intercrop was reduced by as much as 50%. Overall, legumes reduced
fertilizer needs of subsequent maize crop by up to 36 kg N ha“.

This research suggests that annual herbaceous legumes have a unique niche in
smallholder farms in Zimbabwe. If they are relay-intercropped into moderately
fertilized (>0-60 kg N ha") maize, the yields of companion maize can be maintained,

at the same time that subsequent maize crop yields are enhanced.

31

REFERENCES

Agboola, AA, and AA. Fayemi. 1971. Preliminary trials on the intercropping of
maize with different tropical legumes in Western Nigeria J. Agric. Sci.77:
219-225.

Agboola, AA, and AA. Fayemi. 1972. Fixation and excretion of nitrogen by tropical
legumes. Agron J. 64: 409-412.

Agboola, AA. 1980. Effect of different cropping systems on crop yield and soil
fertility in the semi-humid tropics. In Organic recycling in Africa (Rome: FAO)
p. 87-105.

Andrews, DJ. 1972. Intercropping with sorghum in Nigeria. Exp. Agric. 8: 139-150.

Allos, HF. and W.V. Bartholomew. 1956. Relacement of symbiotic fixation by
available nitrogen. Soil Science 87, 61-66.

Baduruddin, M., and D.W. Meyer. 1989. Water use by legumes and its effect on soil
water stress. Crop Sci. 29: 1212-1216.

Biscoe, P.V., and J.N. Gallagher. 1977. In J.J. Landsberg and CV. Cutting (eds).

Environmental Eflects of Crop Physiology. pp. 7 5-100. Academic Press, New
York.

Eaglesham, A.RJ., S.Hassouna, and R. Seegers. 1983. Fertilizer-N effects on N2
fixation by cowpea and soybean. Agron. J. 75: 61-66.

Exner, D.N, and RM. Cruse. 1993. Interseeded forage legume potential as winter
ground cover, nitrogen sourse, and competitor. J. Prod. Agric. 6: 226-231.

Giller, KB and K]. Wilson. 1991. Nitrogen Fixation in Tropical Cropping Systems.
CAB International, Wallingford.

Giller, K.E, J .F. McDonagh and G. Cadisch. 1994. Can Biological Nitrogen Fixation
Sustain Agriculture in the Tropics? In J .K. Syers and BL. Rimmer (ed.) Soil
Science and Sustainable Land Management in the Tropics. Dep. of Agricultural
and Environmental Sci., University of Newcastle upon Tyne, UK.

32

I' _ Ll' “r“.- _
0 I
I

Giller, KB. and G. Cadisch. 1995. Future benefits from biological nitrogen fixation:
An ecological approach to agriculture. Plant and Soil 174: 225-277.

Grant, PM. 1970. Restoration of productivity of depleted sands. Rhod. Agric. J. 67:
131-137.

Haizel, K.A. 1974. Maize-cowpea intercropping study in Kumasi. Ghana J. Agric. Sci.
7: 169-178.

Harper, IL. 1961. Approaches to the study of plant competition.Symp. Soc. Exp].
Biol. 15: 1-39.

Hesterman, 0B. 1988. Exploiting forage legume for nitrogen contribution in cropping
systems. p. 155-166. In W.L. Hargrove (ed). Cropping strategies for sefficent
use of water and nitrogen. ASA-CSA-SSSA Spec. Publ. 51, ASA-CSA-SSSA.
Madison, WI.

Harris, GB, and 0B. Hesterman. 1987. Recovery of nitrogen-15 labelled alfalfa
residue by a subsequent corn crop. 1n: J.F. Power (ed.) The role of legumes in
conservation tillage systems. Soil Cons. Soc. Am. Ankey, IA. pp 58-59.

Hikwa, D. and L. Mukurumbira. 1995. Highlights of previous, current, and proposed
soil fertility research by Department of Research and Specialist Services
(DR&SS) in Zimbabwe. In SR. Waddington (ed.), Report on the First Meeting
of the Network Working Group, Soil Fertility Rsearch Network for Maize-
Based Farming Systems in Selected Countries of Southern Africa. Lilongwe,
Malawi, and Harare, Zimbabwe. The Rockefeller Foundation Southern Africa
Agricultural Sciences Program and CIMMYT. Pp. 44-50.

Hikwa, D., M. Murata, F. Tagwira, C. Chiduza, H. Murwira, L. Muza, and S.
Waddington. 1997. Performance of green manure legumes on exhausted soils in
northern Zimbabwe: A soil fertility network trial. In: SR. Waddington et al.
(eds), Proceedings of the Soil F ert Net Results and Planning workshop. Africa
University, Mutare, Zimbabwe.

Jama, B., RA. Swinkels, and R.L. Buresh. 1997. Agronomic and Economic Evaluation
of Organic and Inorganic sources of Phosphorus in Western Kenya. Agron. J.
89:597-604.

Jeranyama, P. 1995. Medic planting date effect on dry matter and nitrogen

accumulation when clear-seeded or intercropped with corn. MS thesis, Dept. of
Crop and Soil Science, Michigan State University.

33

n‘L:.:

I'IT'.
If

Kwatpata, MB. 1984. Shifting cultivation problems and solutions in Malawi. In: The
future of Shifting cultivation in Africa and the task of Universities (Rome:
FAO) pp. 77-85.

Kumwenda, J.D.T., SR. Waddington, S.S. Snapp, R.B. Jones, and M.J. Blackie. 1996.
Soil Fertility Management Research for the Maize Cropping Systems of
Smallholders in Southern Africa: A Review. NRG Paper 96-02. Mexico, D.F.:
CIMMYT.

Ladd J.N, J.M. Oades, RB. Jackson and J.H.A. Butler. 1983. Utilization by wheat
crops of nitrogen from legume residues decomposing in soils in the field. Soil
Biol Biochem 13: 251-238.

Mariga, IX. 1990. Effect of cowpea planting date and density on performance of a
maize-cowpea intercrop. Zimbabwe J. Agric. Res. 28: 125-131.

Mashiringwani, NA. 1983. The present nutrient status of the soil in communal areas
of Zimbabwe. Zimb. Agric. J. 80: 73-75.

Metelerkamp, HRR 1988. Review of crop research relevant to semiarid areas of
Zimbabwe. In Cropping in the Semiarid Araes of Zimbabwe. Proceedings of a
workshop. Harare, Zimbabwe: Agritex, Department of Research and Specialist
Services (DR&SS), and Gemeinschaft fur Technische Zusarnmenarbeit (GTZ).
Pp. 190-315.

Monteith, IL. 1977. Climate and the efficency of crop production in Britain. Philos.
Trans. R Soc. London Ser. B. 281: 277-294.

Nair, K.P.P., U.K. Patel, RP. Singh, and MK. Kaushik. 1979. Evaluation of legume
intercropping in conservation of fertilizer nitrogen in maize culture. J. Agric.
Sci. 93: 189-194.

Nambiar, P.T.C., MR. Rao, M.S. Reddy, C.N. Floyd, P.J. Dart, and R.W. Wiley.
1983. Effect of intercropping on nodulation and Nz-fixation by groundnut
(Arachis hypogea).Expl. Agric. 19: 77-86.

Natarajan, M., and EM. Shumba. 1990. Intercropping research in Zimbabwe: Current
status and outlook for the future. In: SR. Waddington, A.F.E. Palmer, and GT.
Edje (eds) Research Methods for Cereal/Legume Intercropping. Mexico, D.F.:
CIMMYT pp. 190-193.

Ofori, F. and W.R Stern. 1987. Cereal-legume intercropping systems. Advances in
Agronomy, vol 41: 41-90.

34

-11...." III

Okibgo, B.N. and DJ. Greenland. 1976. Intercropping in tropical Africa. p. 63-101. In
RI. Papendick, P.A. Sanchez, and GB. Triplett (ed.) Multiple cropping. ASA
‘Spec. Publ. 27, Madison, WI.

Pandey, RK., and J.W. Pendleton. 1986. Soybeans as green manure in a maize
intercropping system. Exp. Agric. 18: 125-138.

Rees, DJ. 1986. Crop growth, development and yield in intercropping sorghum with
cowpea in semi-arid conditions in Botswana. H. Exp. Agric. 22: 169-177.

Remison, S. U. 1978. Neighbour effects between maize and cowpea at various levels
ofN and P. Exp. Agric. 14: 205-212.

Rerkasem, K. and B. Rerkasem. 1988. Yields and nitrogen nutrition of intercropped
maize and ricebean (Vigna umbellata (Thumb) Ohwi and Ohashi). Plant and
Soil 108: 151-162.

Sakala, W.D.M. 1994. Crop management interventions in traditional maize pigeon pea
intercropping systems in Malawi. MSc thesis. Lilongwe, Malawi: Bunda
College of Agriculture, University of Malawi.

SAS Institute, Inc 1997. SAS/STAT Users Guide, Release 6.12 Ed. Cary NC.

Singh, SP. 1983. Summer legume intercrop effects on yield and nitrogen economy of
wheat in the succeeding season. J. Agric. Sci. 101: 401-405.

Shumba E.M., H.H. Dhliwayo, and 0.2. Mukoko. 1990. The potential of maize-
cowpea intercropping in low rainfall areas of Zimbabwe. Zimbabwe Journal of
Agricultural Research 28: 33-38.

Tattersfield, JR. 1982. The role of research in increasing food crop potential in
Zimbabwe. Zim Sci. News vol. 16 no. 1 : 6-11.

Thomas, D., and J .E. Sumberg. 1995. A review of the evaluation and use of tropical
forage legumes in sub-Saharan Africa. Agriculture, Ecosystems and
Environment 54: 151-163.

Trenbath, BR 1976. In Mutliple cropping (RI Papendick, P.A. Sanchez, and GB.
Triplett. eds. ) pp. 129-169. Spec. Publ. ASA. Madison, WI.

Temu, A.E.M. 1982. Rotational green manuring with Crotolaria in the Southern
highlands of Tanzania. Southern Highlands Maize Improvement Program,
Progress Report, Tanzania. Mimeo.

35

TI‘T‘

Waddington, SR, and P.W. Heisey. 1997. Meeting the nitrogen requirements of maize
grown by resource-poor farmers in southern Africa by integrating varieties,
fertilizer use, crop management and policies. In G.O. Edmeads, M. Banziger,
HR. Mickelson, and CB. Pena-Valdivia (eds). Developing drought-and-low N-
tolerant maize. Proceedings of a symposium, March 25-29, CIMMYT, El
Batan, Mexico, D.F.: CIMMYT.

36

 

CHAPTER TWO

MAIZE (Zea mays L.) YIELD AND NITROGEN UPTAKE FOLLOWING
GROUNDNUT (Arachis hypogaea L.) ON SMALLHOLDER FARMS IN
NORTHERN ZINIBABWE

ABSTRACT

?.-_"4“
1

The more systematic rotation of maize (Zea mays L.) with groundnut (Arachis
hwogaea L.) has been proposed as a way of maintaining soil fertility and preventing
maize productivity declines associated with smallholder cropping systems of
Zimbabwe. The objectives of this study were (i) to evaluate the impact of a groundnut
crop on subsequent maize yield and N uptake, (ii) to evaluate the response to fertilizer-
N for maize in rotation with groundnut compared to continuous maize and (iii) to
determine the fertilizer replacement value of groundnut. Treatments included
continuous maize and a maize-maize-groundnut rotation, receiving variable fertilizer-N
rates in a split-plot arrangement. Maize yields were improved by 0.1-2.2 Mg ha'1
following groundnut compared to continuous maize. The low groundnut yield on farm
were associated with little yield improvements for a subsequent maize crop. In a
second maize crop after groundnut on station, yields were increased by 0.8 Mg ha'1
and grain N uptake was increased by 53% relative to continuous maize. Fertilizer

needs were reduced by 72 kg N ha’1 when maize followed groundnut.

37

 

INTRODUCTION

Declining soil fertility in the smallholder farms of Zimbabwe is partly a result
of continuous maize (Zea mays) production and partly due to inadequate nutrient

inputs and management, exacerbated by unreliable rainfall distribution and marginal

”‘1

economics. Traditionally, African agricultural systems restored soil fertility lost during

cropping by extended fallows with natural vegetation (Araki, 1993; Blackie and Jones,

“1..-
1‘.

1993). Increasing population pressure on limited agricultural land has rendered
fallowing a non-viable option, while continuous maize has become a common cropping
system in Zimbabwe (Kumwenda et al., 1996). Continuous cropping of maize at
reasonable levels of productivity (>1 Mg ha'1 grain yields) can likely not be sustained
without substantial additions of nutrients (Grant, 1970; MacColl, 1989). One
alternative to reduce over-dependence on chemical fertilizers is to grow maize in
rotation with a legume such as groundnut (Arachis hypogaea).

Although maize remains the major crop and a staple food, groundnuts are
important in the diets of smallholder farmers. Groundnuts are the second most
important crop in sub-humid smallholder farmers in Zimbabwe and are widely
recognized for their nutritive value, particularly for young children. The kernels are
eaten raw, boiled or roasted, and made into confectionery and snack foods. They are

also used in soups or made into sauces for meat, rice or maize meals. The vegetative

38

residues from groundnut are excellent forage. However, despite the benefits of
groundnut as a crop, its production in the smallholder farms has declined (Shumba,
1983; Dendere, 1987).

Reasons for groundnut production declines includes a government policy of
over-promoting maize for a long time by making maize pricing and marketing more
favorable and the lower priority placed on groundnut by male farmers because of the
high labor demand in processing (and as such the crop is referred to as a "female crop"
(Shumba, 1983)). Farmers also have taste preferences and perceptions of profitability

that seem to favor maize.

‘1nuwu.M‘ ”1‘11

Legume-cereal rotations in the region have long been recognized for restoring
soil fertility and increasing crop productivity (MacColl, 1989; Mukurumbira, 1985).
Rotations shift biological balance in the soil, reducing build up of specific pests and
diseases and sustains productivity of the cropping system (Kumwenda et al., 1996). In
a long-term maize-legume rotation trial in South Africa, unfertilized maize grain yields
were improved by 2 Mg ha'1 in rotation with field pea (Pisum arvense Poir) compared
to unfertilized continuous maize (Nel et al., 1996).

Legume residues are particularly useful as organic green manures due to their
high N content and because this N is more readily available for plant uptake than N
from non-legumes (Giller et al., 1994). The capabilities of grain legumes to contribute
to soil fertility differ among species. The greater the efficiency of the legume in
translocating N to the harvested components (such as grain) the smaller the

contribution of N to the soil profile. Giller et al. (1994) showed that groundnut which

39

contains large amounts of N in its residues at harvest and prior leaf-fall can supply
more N for subsequent crops than other grain legumes such as soybean. McDonagh et
al. (1993) found that groundnut residues containing 100-130 kg N ha“, when
incorporated into the soil reduced fertilizer needs of a subsequent maize crop by 60 kg
N ha’1 .

For grain legumes to play an important role in the maintenance of soil fertility,
they must leave behind substantially more N from Nz-fixation than the amount of soil

N that was removed in the harvested portion of the crop. Clearly the two purposes

.1

served by the legume crop, one to provide forage and grain yield and the other to

r.-

leave residual N are somewhat contradictory. The role of grain legumes in contributing
N to cropping systems is bound to be compromised by the breeding priority of
optimizing the efficiency of conversion of N into the grain removed (Hanzell and
Vallis, 1977). A smaller grain yield of legume may be the price to pay if the amount
of legume N restituted to the soil profile is to be increased.

In the current study, the effect of including a groundnut crop in rotation with
maize was evaluated. Specific objectives were: (i) to evaluate the impact of a
groundnut crop on subsequent maize yield and N uptake, and response to fertilizer-N
of maize in rotation with groundnut compared to continuous maize and, (ii) to

determine the fertilizer replacement value of groundnut.

40

MATERIALS AND METHODS

Since the 1992/93 cropping season, the CIMMYT Maize Program in Harare has
conducted a set of long-term trials on crop productivity and soil fertility trends in
maize-groundnut systems under current smallholder management (W addington et al.,
1996). Two distinct agro-ecological zones, natural regions (NR) II1 and III (Vincent
and Thomas, 1961), were chosen for this study. These typical represent sub-humid
maize cropping areas of Zimbabwe.

This experiment was established at Chinyika and Domboshava (both NR II) and
Chiduku (NR III) smallholder areas. Soils in these locations are predominantly sandy
soils derived from granite and classified as Typic Kandiustalf. The sites have been
cropped for various length of time, ranging from 14 years at Chinyika to over 60 years
in Chiduku. Trophic soil properties were characterized at the beginning of the study in
1994 at each site (Table 2.1).

The experiment was arranged in a randomized complete block design replicated
three times at each site. The treatments (and cropping systems) in this study were ;

1. Continuous maize - Fertilizer level of 275 kg ha’1 Compound D (8-14-7) as basal
and a side dress of 70 kg N ha" as NH4NO3 so that total N applied was 92 kg ha‘1

2. Continuous maize - no fertilizer applied

 

1A broad classification of Natural regions is based on rainfall: NR I 900-1200mm p.a
NR 11 750-900mm p.a; NR HI 650-750mm p.a; NR IV 450-650mm pa and
NR V < 650mm p.a.

41

4.11
tht

sub-1

Rani
Ions
P0PU
C0n1
pIaCl

DCXI

appr

apar
seed

Spar

thet
11(Ere

midd

3. Maize-Maize-Groundnut rotation - fertilizer applied to maize only as in treatment 1

4. Maize-Maize-Groundnut rotation - no fertilizer applied.
In year four (maize after groundnut), cropping systems 1 and 3 were further Split into
sub-plots and subjected to fertilizer applications of 0, 46, 92 or 138 kg N ha".

Maize (Zea mays L. var. ’R215’) was hand-planted at two seeds per planting
station on plots that had been moldboard plowed. Each sub-plot was 5 x 5.4 m with 6
rows of maize, planted at 0.9 m between rows and 0.5 m within rows, giving a plant
population density of 44, 440 plants per hectare. Initial fertilizer was applied as
Compound D at about one week after plant emergence (corresponding to farmer
practice). Additional fertilizer was applied in the form of NH,,NO3 as a surface dollop
next to each plant station at V6 maize growth stage when the soil moisture level
approximated field capacity.

Groundnut (Arachis hypogaea var. ’Spanish’) were planted in rows of 0.45 m
apart and 0.25 In between planting stations in a row. Groundnut were planted at two
seeds per station, giving approximately 160 000 plants per hectare. Because the
Spanish cultivars exhibit promiscuous nodulation in the soils under study and because
farmers do not inoculate, seeds were not inoculated with rhizobia before planting.

Maize grain and groundnut kernels were harvested from a 2.7 x 3 m section of
the three center rows of each plot, so that the area harvested was 8.1 m2. Grain yields
were adjusted to 125 and 100 g kg‘1 moisture content for maize and groundnut,
respectively. Total above-ground biomass of maize was measured from two adjacent

middle rows from an area of 2.7 m2.

42

 

 

Kjelda
Wiley
K350,
spectrc

nitrcge

respect
fourth

system.
compai
rifSpons
linear g
trends i

replaCe.

Plant Chemical Analysis.

Total N in maize grain and stover was determined by a modified micro-
Kjeldahl method. Dry plant materials were ground to pass through a 2-mm screen in a
Wiley mill. Plant samples of 0.1 g were digested in 4 ml of 18 M H2SO4 with 1.5 g
KZSO4 and 0.075 g Se catalyst. Following digestion, total NH,+ was determined by
spectrophotometry. Total N yield was calculated as the product of dry matter yield and

nitrogen concentration.

Statistical Analysis.

ANOVA in Proc GLM (SAS, 1997) was used to analyze treatment effects with
respect to maize grain yield, grain N content and total above-ground biomass in the
fourth year. Experiment was a RCBD with treatments arranged in a split plot, cropping
systems were main plots and fertilizer-N rates subplots. Cropping systems were also
compared using orthogonal contrasts (=1 df comparisons). Subsequent maize
responses to fertilizer-N rates within a cropping system were determined by evaluating
linear and quadratic trends from single degree of freedom comparisons. Whenever
trends were significant, regression equations were calculated to determine fertilizer

replacement values (F RV) of groundnut in rotation with maize.

43

RESULTS AND DISCUSSION

Groundnut yield_s_. Groundnut kernel yields (harvestable portion) for the three
locations ranged from 0.16 to 0.34 Mg ha‘1 and the vegetative biomass (haulrns)
ranged from 0.65 to 1.6 Mg ha" (Table 2.2). Because kernels are harvested and
removed from the field, net plowdown N is therefore primarily based on the haulms
and fallen leaves. One of the signs of kernel physiological maturity is the browning
and senescence of leaves. Plowdown N values tend to underestimate actual
contribution of groundnut to soil N because only the green or intact vegetative material
is measured. In this study, plowdown N was between 16 and 40 kg N ha’1 (Table 2.2).
Our values of plowdown N compare well to those calculated from Suwanarit et a1.
(1986) and Dakora et al. (1987), of 42 and 38 kg N ha", respectively.

Maize yields following groundnut (No Fertilizer). Because of significant (PS
0.05) treatment x location interactions, data are presented as treatment effects within
location (Table 2.3). Unfertilized maize grain yields following groundnut were
improved by 2.2, 0.3 and less than 0.1 Mg ha'1 at Domboshava, Chinyika and Chiduku
respectively (Table 2.3). However, yield increase at Chiduku of 8% was not
significantly (PS 0.05) better than unfertilized continuous maize. Although there were
positive yield increases in maize grain yield following groundnut at Chinyika and

Chiduku, these yields remained low (less than 1 Mg ha") and were only 16% of those

44

obtained at the more fertile Domboshava site.

In Zimbabwe, Mukurumbira (1985) evaluated maize grain yields following
several food legumes which included groundnut and fallow. Yield of maize was
greater following groundnut (6.2 Mg ha") than unplanted fallow (4.3 Mg ha“) or
maize (3.9 Mg ha"). Maize grain yield after groundnut represented a 60%
improvement over continuous maize. Our results at Chinyika and Domboshava
corroborate those of Mukurumbira (1985).

In a second maize crop following groundnut, grain yield was increased by 0.8
Mg ha", representing a 69% improvement over continuous maize when no fertilizer
was applied (Table 2.4). Grain N uptake and total above-ground biomass were each
improved by 53% over continuous maize (Table 2.4). Although maize grain yields
were higher in the second crop following groundnut relative to continuous maize, these
yields were considerably smaller than yields of the first maize crop following
groundnut. Yields at the highest fertilizer level of 138 kg ha’1 were about 1 Mg ha’1
lower than the unfertilized first maize crop after groundnut (data not shown).

In an attempt to explain treatment x location interaction, grain yield differences
between maize following groundnut and continuous maize were correlated to soil pH,
soil phosphorus, mineralizable N, cation exchange capacity, total exchangeable bases,
groundnut haulm biomass and plowdown N, combined January and February rainfall,
and total rainfall as suggested by Matlon (1984) and accomplished by Shumba et al.
(1992) in Zimbabwe. Multiple regression was not used in this analysis. Combined

January and February rainfall was negatively correlated (r= 08 PS 0.05) to yield

45

 

 

differe
percen
paraml
includ
than 5
that cl
promc
corrot
that ti
and 11
240 n
more
Iainfa
distril

Zimb

a pos
Regn
conte
appli

IESpQ

Nat]

differences. This parameter (combined January and February rainfall) explained 56
percent of the variability in yield differences (r2=0.56 pg 0.05) and the other
parameters were not significantly correlated to grain yield differences. Effects of
including groundnut in rotation with maize were offset at locations that received more
than 530 mm in the months of January and February. Incessant and excessive rains
that characterized the 1996/97 season throughout January and the first half of February
promoted widespread waterlogging that affected growth of maize. Our results
corroborate findings of Shumba et al. (1992) who concluded in a tillage study in NRIIl
that tillage x location interaction was mostly explained by January rainfall (r7=0.76)
and that tine tillage increased maize grain yields at locations that received less than
240 mm of rainfall in January and depressed yield at those locations that received
more than 240 mm. However, in our study it was the combined January and February
rainfall that explained most of the variability, further suggesting that rainfall
distribution plays a key role on the effectiveness of smallholder technologies in
Zimbabwe. Rainfall data for the study areas is presented in Figure 2.1.

Maize yields following ggoundnut (Variable N applied). There was not always
a positive maize response to groundnut in rotation when fertilizer-N was applied.
Regression equations for each cropping system for maize grain yield (GR), grain N
content (GRN) and total above-ground biomass (TDM) as a function of fertilizer N
applied were obtained at Domboshava and Chinyika (Table 2.5). At Domboshava

response of a second maize crop following groundnut was also evaluated.

1Natural region 11 70-900mm rainfall per annum.

46

Maize grain yields following both groundnut and continuous maize linearly
responded to increasing N rates at Domboshava (Figure 2.2). At all N rates, yields
were significantly (P_<_ 0.05) higher following groundnut than following maize.

At Chinyika, continuous maize showed a quadratic response while maize
following groundnut linearly responded to variable N rates. At N 2 30 kg N ha",
continuous maize was always associated with a yield higher than that of maize
following a groundnut crop (Figure 2.4).

Maize grain N content for the two cropping systems and at both locations
linearly responded to increasing N rates (Figures 2.5 and 2.6). At Domboshava, grain
N content following groundnut was increased by 41% at each N rate relative to
continuous maize. However, the reverse was true for Chinyika at which grain N
content for continuous maize increased more than grain N content for maize following
groundnut as N rates increased (Figure 2.6). Results for Chinyika Show that the
common assumption that a groundnut crop improves N availability and enhances yield
in a subsequent year may not always be correct, on smallholder fields where soil
fertility is low.

Total above-ground biomass of maize following groundnut at Domboshava was
2 Mg ha'1 higher than following continuous maize when no fertilizer was applied
(Figure 2.7). Differences in total above-ground biomass for the two cropping systems
narrowed and became non significant as fertilizer N rate increased (Figure 2.7). At
Chinyika, total above-ground biomass for continuous maize showed a quadratic

response to increasing N rates while that following groundnut was linear (Figure 2.8).

47

respOI
Conu

NHOv

NHov
find:
increi
ruue
Mghe
eflbd

Aida

left 11
Ifgur

soYbe

response to increasing N rates while that following groundnut was linear (Figure 2.8).
Continuous maize produced significantly (PS 0.05) higher biomass yields than maize
following groundnut at all N rates except at 138 kg N ha".

Fertilizer Replacement Value. Fertilizer replacement values of first year maize

following groundnut was 72 kg N ha'1 based on grain yield and 18 kg N ha‘1 based on
total above-ground biomass (Table 2.6). We note that groundnut residual effects
increased yields of a second maize crop following groundnut. Fertilizer replacement
values for this second maize crop ranged from 31 to 49 kg N ha". In each year, the
highest FRV’S were based on grain yield. MacColl (1989) also presented a beneficial
effect on the second and even a third maize crop following groundnut in central
Malawi.

Fertilizer replacement values obtained in our study are similar to those reported
by other researchers for groundnut using maize as a test crop. In northern Ghana,
groundnut and cowpea (Vigna unguiculata) had FRV’s of 60 kg N ha’1 (Dakora et al.,
1 987). Jones (1974) evaluated residual effects of groundnut on a subsequent maize
crop in savanna areas of Nigeria and obtained an equivalent of 43-73 kg N ha’1 when
no fertilizer nitrogen was applied to maize. Maize yields in short rotations with
legumes at Bunda, Malawi were found to be better after legumes with poor grain yield
but vigorous vegetative grth such as lablab (Lablab purpureus) and these generally
left more residual N than groundnut or soybean (Glycine max) (MacColl, 1989).

Legume equivalent values were 52, 26 and 0-14 kg N ha‘1 for lablab, groundnut and

soybean, respectively in that study.

48

4O

5
3

G
3

5 0 5
2 2 1
AEEV 3052.6!

 

0
41

5

Fig. 2.1.12

$635011.

 

400

 

350 -
300 ,
250 .

200 -

Rainfall (mm)

150 -

100 4

50-

 

 

Nov

 

 

 

 

Dec Jan Feb March April

 

 

I Chiduku
. Domboshava
g Chinyika

 

 

 

Fig. 2.1. Rainfall data for Chiduku, Chinyika and Domboshava for the 1996/97

8688011.

49

 

Table 2.1. Trophic soil properties at Chinyika and Chiduku in Zimbabwe at the
beginning of the study

 

 

Characteristic Chinyika Chiduku
Soil texture loamy sand sandy clay loam
pH (CaClz) 4.3 4.7

C (%) 0.52 1.35
Mineralizable N (ppm) 18.63 41.63
P20s (ug g“) 9.76 8.20

K (pg g") 3.00 8.25
CEC" (me °/.) 3.01 8.25
TEBt (me %) 6.38 8.48

 

T Cation exchange capacity
3 Total exchangeable bases

50

Kern:
Haulr

P10WI

Table 2.2. Groundnut kernel, haulm and plowdown N yield at three locations in

 

 

 

 

Zimbabwe.
Parameter Chinyika Chiduku Domboshava
kg ha'1
Kernel l 6 l l 93 3 3 7
Haulms 650 980 1600
Plowdown N 16 22 40

 

51

Table 2.3. Effect of cropping system on unfertilized maize grain yield at three
locations in northern Zimbabwe.

 

 

 

 

Cropping systemT Chinyika Chiduku Domboshava
Mg ha'1

MZ-MZ-Mz-m 0.39 0.71 2.46
Mz-Mz-Gn-m 0.68 0.77 4.61

CV (%) 26 4 20
M
Mz-MZ-Mz-m VS * NS *
Mz-Mz-Gn-_M_z
Percent increase1 74 8 87

 

TUnderlined letter indicates year for yield. Mz=maize, Gn=groundnut
* Significance at P: 0.05. NS=non significant at PS 0.05.
‘ Refers to increase due to rotation over continuous maize.

52

Table 2.4. Effect of cropping system on unfertilized second maize crop following
groundnut at Domboshava, Zimbabwe.

 

 

Cropping systemT Grain Yield Grain N Content Total biomasst
(Mg ha") (kg ha“) (Mg ha")

Mz-Mz-Mz-Mz-m 1.20 13.16 3.19
Mz-Mz-Gn-Mz-Mz_ 2.03 20. 12 4.89

cv (%) 18 24 16
Contrast
Mz-Mz-Mz-M; VS * * *
Mz-Mz-Gn-M;
Percent increase1 69 53 53

 

1‘Underlined letter indicates year for yield. Mz=maize, Gn=groundnut
* Significance at PS 0.05. NS=non significant at PS 0.05.

3 Total above-ground biomass of maize

' Refers to increase due to rotation over continuous maize

53

Table 2.5. Fitted Regression Equations for maize grain yield (GR), grain N content
(GRN) and total above-ground biomass (TDM) in two maize cropping systems
as a function of N fertilizer applied (x).

 

 

Cropping SystemT Regression Equation’ R2 Signif
Domboshava
Mz-Mz-Mz-_M_; GR=2.46 + 0.03(x) 0.59 0.05
TDM=7.65 +0.03(x) 0.95 0.03
Mz-Mz-Gngm GR=4.61 + 0.04(x) 0.98 0.01
TDM=8.2 + 0.03(x) 0.96 0.02
Mz----MzMzMzM_z_ GR=1.33 + 0.01(x) 0.75 0.05
GRN=13.96 + 0.l8(x) 0.72 0.01
TDM=3.93 + 0.03(x) 0.65 0.02
Mz-Mz-Gn-Mz-m GR=1.92 + 0.01(x) 0.84 0.001
GRN=19.68 + 0.18(x) 0.83 0.002
TDM=5.75 + 0.02(x) 0.48 0.05
Chinyika
Mz---MzMz_M_z_ GR=0.25 + 0.02(x)-0.00007(x2) 0.96 0.001
GRN=3.8 + 0.15(x) 0.89 0.01
TDM=2.74 + 0.069(x)-0.0004(x2) 0.82 0.01
Mz-Mz-Gn-Mz GR=0.34 + 0.009(x) 0.88 0.001
GRN=2.87 +0.l2(x) 0.87 0.001
TDM=2.76 + 0.027(x) 0.79 0.003

 

T Underlined letter represent the crop and year for reported yields. Mz=maize,
=groundnut.
1 Linear equations are based on 6 degrees of freedom, while quadratic equations are
based on 5 degrees of freedom

54

Table 2.6. Fertilizer replacement values of a maize and groundnut rotation cropping
system at Domboshava.

 

 

Parameter Mz-Mz-Gn-_M_;T Mz-Mz-Gn-Mz-Mz
-------- kg N ha"-----------

Grain yield 72 49

Grain N content - 31

Total above-ground 18 32

biomass

 

1 Underlined letter shows the test crop from which data is derived. Mz=maize,
Gn=groundnut

55

Maize grain yield (Mg ha'l)

10

 

 
     

O Mz—Mz-Mz-M;
A Mz-Mz-Gn-M;

 

 

 

1 J 1 HI I

0 30 60 90 120

N applied (kg ha“)

Figure 2.2 Subsequent maize grain yield response to N and rotation
with groundnut at Domboshava

56

Maize grain yield (Mg ha“)

 

 

O Mz-Mz-Mz-Mz-M;
A Mz—Mz-Gn-Mz-Ml

    

 

 

—1
—1

I J I

0 30 60 90 120 150

N applied (kg ha'l)

Figure 2.3 Maize grain yield response to N and rotation with
groundnut at Domboshava

57

Maize grain yield (Mg ha")

 

1.0

O Mz-Mz-Mz—Mz
A Mz-Mz-Gn-M;

p
VI
1

 

 
   

 

 

0.0 I I l I I
0 3O 60 90 120

N applied (kg ha'l)

Figure 2.4 Maize grain yield response to N and rotation with
groundnut at Chinyika

58

150

Grain N content (kg ha’l)

45

40

35

30

25

20

15

10

 

—l

o Mz-Mz-Mz-M;
A Mz—Mz-Gn-M;

—i

d

 

 

 

—-4
—1

0 30 60 90 120

N applied (kg ha'l)

Figure 2.5 Maize grain N content response to N and rotation
with groundnut at Domboshava.

59

 

150

Grain N content (kg ha'l)

25

 

 

 
    

 

20 q o Mz-Mz-Mz-m
A Mz—Mz-Gn-M;
15 a
10 1
5 q
0 1 I I l
0 30 60 90
N applied (kg ha")

Figure 2.6 Maize grain N content response to N and rotation
with groundnut at Chinyika.

60

120

 

150

 

9 4 O Mz—Mz-Mz-M;
A Mz-Mz-Gn-M_z ‘

Toal above-ground biomass (Mg ha!)
05

 

 

 

 

5 a

4 a

3 - O

2 1 I I I 1
0 30 60 90 120

N applied (kg ha")

Figure 2.7 Maize total above-ground biomass response to N and rotation
with groundnut at Domboshava.

61

150

 

 
 
 
   

A

7: O Mz-Mz-Mz-M;

£1 6 _ A Mz-Mz-Gn-M;

00

E

g C

C)

E» 5“

.D

e

G

8

in 4 i

O

>

o

.0

£6

"a 3 7

8

Fe

2 l l l I I

 

 

 

0 30 60 90 120 150

N applied (kg ha")

Figure 2.8 Maize total above-ground biomass response to N and rotation
with groundnut at Chinyika

62

CONCLUSIONS

This research showed that maize yields were improved in rotation with
groundnut compared to continuous maize in Zimbabwe. Unfertilized maize grain yields
were improved by 0.1-2.2 Mg ha‘1 following groundnut compared to continuous maize.
For a second maize crop following groundnut, legume residual benefits were of 0.8
Mg ha“. However, improvements over continuous maize were not observed at two of
the three sites when maize was fertilized at the recommended N level due to incessant
and excessive rains received at those sites.

The groundnut crop reduced fertilizer needs of a subsequent maize crop by up
to 72 kg N ha" and by 49 kg N ha“ in a second maize crop at one site. Benefits of
including groundnut were less pronounced at the other two sites.

Our results suggest that benefits of including groundnut in rotation with maize
are sensitive to rainfall events in the smallholder farms of Zimbabwe. Maximum

benefits were obtained at sites with reasonably well distributed rainfall (slightly less
than 530 mm rainfall in the months of January and February, a period that coincide

With anthesis). However, at sites with poorly distributed rainfall benefits were modest

or non-existent.

63

REFERENCES

Araki, S. 1993. Effect on soil organic matter and soil fertility of the chitemene slash-
and-bum practice used in northern Zambia. In K. Mulongoy and R Merclor
(eds) Proceedings: Soil Organic Matter Dynamics and Sustainability of
Tropical Agriculture. Chiscester, U.K.: Wiley-Sayce. pp. 367-375.

Blackie, M.J., and RB. Jones. 1993. Agronomy and increased maize productivity in
southern Africa. Biological Agriculture and Horticulture 9: 147-160.

Dakora F.D., RA. Aboyinga, Y. Mahama, and J. Apaseku. 1987. Assessment of N2-
fixation in groundnut (Arachis hypogaea L.) and cowpea (Vigna unguiculata L.

Walp) and their relative contribution to a succeeding maize crop in Northern
Ghana. NflRCEN Journal 3: 389-399.

Dendere, S. 1987. Contraints to groundnut production and research priorities for
communal areas in Zimbabwe. In: Proceedings of the Second Regional
Groundnut Workshop for Southern Africa. ICRISAT, Patancheru, A.P., India,
pp. 125-129.

Giller, K.E., J.F. McDonagh, and G. Cadisch. 1994. Can biological nitrogen sustain
agriculture in the tropics? In J .K. Syers and D.L. Rimmer (eds) Soil Science
and Sustainable Land Management in the Tropics. Wallingford, U.K.: CAB
International. pp. 713-191.

Grant, PM. 1970. Restoration of productivity of depleted sands. Rhodesia Agricultural
Journal 67: 131-137.

Henzell, BF, and I. Vallis. 1979. Transfer of nitrogen between legumes and other
crops. In A. Ayanaba and P.J. Dart (eds) Biological Nitrogen Fixation in
Farming Systems of the Tropics. Wiley, Brisbane. pp. 73-88.

Jones, M.J. 1974. Effects of previous crop on yield and nitrogen response of maize at
Samaru, Nigeria. Experimental Agriculture 10: 278-279.

Kumwenda, J.D.T., SR Waddington, S.S. Snapp, RB. Jones, and M.J. Blackie. 1996.
Soil Fertility Management Research for Maize Cropping Systems of
Smallholders in Southern Afiica: A review. NRG Paper 96-02. Mexico, D.F.:
CIMMYT.

64

MacColl, D. 1989. Studies on maize (Zea mays L.) at Bunda, Malawi. H. Yield in
short rotation with legumes. Experimental Agriculture 25: 367-374.

Matlon, P.J. 1984. Technology evaluation. In P.J. Matlon et al. (eds) Coming Full
Cycle: Farmers’ Participation in the Development of Technology: 95-118.
Ottawa: IDRC.

McDonagh, J.F., B. Toomsan, V. Limpinuntana, and KB. Giller. 1993. Estimates of
the residual nitrogen benefits of groundnut to maize in Northeast Thailand.
Plant and Soil 154: 267-277.

Mukurumbira, L.M. 1985. Effects of rate of fertilizer nitrogen and previous grain
legume crop on maize yields. Zimbabwe Agricultural Journal 82: 177-179.

Nel, P.C, RO. Barnard, RE. Steynberg, J.M. de Beer, and HT. Groenveld. 1996.
Trends in maize grain yields in a long-term fertilizer trial. Field Crops Research
47: 53-64.

SAS Institute, Inc 1997. SAS/STAT Users Guide, Release 6.12 Ed. Cary NC.

Shumba, EM, SR. Waddington and M. Rukuni. 1992. Use of tine-tillage, with
atrazine weed control, to permit earlier planting of maize by smallholder
farmers in Zimbabwe. Experimental Agriculture 28: 443-452.

Shumba, EM. 1993. Factors contributing to a decline in groundnut production in the
Mangwende-Murehwa District, and the need for a technical research input.
Zimbabwe Agricultural Journal 80: 251-254.

Swanarit A, C. Suwannarat, and S. Chotechanungmanirat. 1986. Quantities of fixed N
and effects of grain legumes on following maize, and N and P status of soil as
indicated by isotopes. Plant and Soil 93: 249-258.

Vincent, V. and RG. Thomas. 1961. An agricultural survey of Southern Rhodesia.
Part 1. Agroecological Survey. Harare, pp. 41-102.

Waddington, S.R, J. Karigwindi, and J. Chifamba. 1996. CIMMYT Maize Soil
fertility and Agronomy Research in southern Africa. Annual Research Report:
CIMMYT-Zimbabwe.

65

CHAPTER THREE

ECONOMIC ANALYSIS OF SMALLHOLDER MAIZE CROPPING SYSTEMS
IN NORTHERN ZIIVIBABWE

ABSTRACT

Continuous cropping of maize, typical of the smallholder farms of Zimbabwe,
at reasonable levels of productivity can not be sustained without substantial additions
of nutrients. Thus agronomists have proposed maize-groundnut rotation and maize-
legume intercrops as alternatives to improve profit margins and at the same time

reduce over-dependence on chemical fertilizers.

This assessment used marginal benefit-cost analysis, a type of partial budgeting
analysis, to identify optimum maize cropping systems from a set of alternatives. Field
experimental data from a Gn-Il/Iz-Mz vs Mz-Mz-Mz trial and a Sale Maize vs Maize-
Iegume intercrop trial were used in this analysis. Sensitivity analysis was also used to

assess the stability of results with changing seed prices of sunnhemp and cowpea, and

yield increase in a groundnut crop.

Continuous maize at 92 kg N ha'1 optimized net benefits over total costs when
compared to maize in rotation with groundnut at variable fertilizer N rates For the
maize-groundnut rotation to equal net benefits of continuous maize, current groundnut
crop yield needed to be more than doubled, assuming labor costs remained constant.

The results suggest that maize-groundnut rotation is less profitable than continuous

66

maize, especially when the maize crop was grown with fertilizer.

The maize-cowpea intercrop at 60 kg N ha’1 was associated with highest net
benefits, compared to maize-sunnhemp intercrop and sole maize at variable N rates.
Both maize-cowpea and maize-sunnhemp intercrops paid back legume seed costs at the
current price in the second season, when a sole maize crop following the intercrops
was grown without fertilizer. Sensitivity analysis revealed that a 50% increase in
either cowpea or sunnhemp seed price would greatly reduce net benefits of the maize-
legume intercrops compared to the control (continuous maize without fertilizer).
Results showed that when moderately fertilized (60 kg N ha") the maize-cowpea

intercrop was more profitable than sole maize or the maize-sunnhemp intercrop.

67

INTRODUCTION

A primary goal of agricultural production, like other business enterprises, is to
provide producers an acceptable level of retum to their capital, labor and management
inputs. Bernsten (1980) identified five factors that affect the profitability of new
technology and thereby influence its acceptability to smallholder farmers. To be
appropriate, new technology must be compatible with (i) the production environment,
(ii) cultural values, (iii) farmers’ goals and, (v) existing institutions.

The "induced innovation" model, as proposed by Hayami and Ruttan in 1977
(Stevens and Jabara, 1988), significantly advances our economic understanding of how
agricultural development is achieved. This theory identifies four key elements
(resource endowment, cultural endowment, technology and institutions) which
determine a country’s rate of agricultural development As these four are interactive,
changes in the levels of one will induce changes in the level of another element. For
example, improvements in marketing institutions will increase the profitability of grain
production by reducing input costs and increasing farm gate prices. As a result farmers
will adopt improved varieties which are now more profitable.

Enterprise and partial budgets have been used to select optimum production
system from a set of alternatives (Hesterman et al., 1986; Shumba et al., 1992).

Marginal benefit-cost analysis, a type of partial budgeting analysis, can be used to

68

assess the profitability of alternative cropping patterns. By estimating the ratio between
the rates of increase of benefit to cost, this technique estimates the rate of return to
capital invested in new technologies (cost of the inputs or test factors which are
variable).

Traditionally, benefit-cost analysis assumes that all farmers have equal access to
resources (land, labor and capital) and similar access to credit and market for inputs
and produce. Based on these assumptions, typically a technology that gives the highest
net benefit is recommended to all farmers (Bernsten, 1980). While the concept of
recommendation domain is widely used to refer to farmers facing similar
circumstances, it is seldomly developed to fully reflect the way in which these
differences influence the profitability of technology among farmers facing different
circumstances (i.e., production environments, access to resources and institutional
support).

There are doubts about the economics of farmers investing more in groundnut
production. Work by Shumba, Bernsten and Waddington (1990) showed that for the
Mangwende communal area of Zimbabwe it was less profitable to invest in the most
promising inputs and practices to raise groundnut productivity (improved seed, early
planting and weeding, NPK fertilizer + gypsum) than to invest in maize production.

Sensitivity analysis is a tool which can be used in benefit-cost analysis to
accommodate farmer-to-farmer differences in resource endowment, levels of
management skills and institutional access that affect profitability.

The objectives of this analysis were to (i) assess the profitability of Gn-A/Iz-A/Iz

69

versus Mz-Il/Iz-Il/Iz cropping systems at variable N fertilizer rates, and (ii) compare the
marginal benefits of maize-legume intercrops and a subsequent sole maize crop at
variable fertilizer N rates. In addition, sensitivity analysis is used to assess the impact
of varying assumption regarding the price of legume seed on the profitability of the

alternative cropping patterns.

70

ECONOMIC ANALYSIS AND ASSUMPTIONS

In Zimbabwe, smallholder maize farmers typically grow continuous maize at
zero or low levels of N. Field experimental data from two trials, namely Gn-A/Iz-A/lz vs
A/[z-Il/Iz-Il/Iz at two fertilizer N levels (0 vs 92 kg N ha") and maize-legume
intercropping at variable fertilizer N levels, were used in a marginal benefit cost
analysis. Costs and benefits of each treatment were compared using partial budgets,
which included only the costs and benefits that varied from the control, following
methods given in CIMMYT (1988). Discounting procedures for calculating net
benefits were adapted from Gittenger (1984). The bank deposit interest rate (in 1996
and 1997) of 23%1 (W addington et al., 1997) was used as a proxy for the opportunity
cost of capital.

Yearn = 1/(1 + interest rate)”1 [1]

The prices of fertilizer and seed (1996) were obtained from local suppliers.
Labor was valued at the wage rate paid to hired farm laborers in the area, which was
determined through a survey in Chinyika and Mangwende Communal Areas to reflect

opportunity cost to family labor. Smallholder farmers have the opportunity to be

 

1Discount rates of as high as 60% have been reported elsewhere in subsistence
agriculture (Bernsten 1998 pers. com). The current bank rediscount rate is 37%
(Reserve Bank of Zimbabwe, August 1998).

71

involved in off-farm activities and to name a few; woodcrafi, basket weaving, gold
panning and quasi-urban employment. All these represent an opportunity cost to on
farm labor. Transport cost were calculated, based on charges from the place of
purchase to the site where the trials were conducted. Costs and labor data are
presented with the budgets.

This analysis in which groundnut forms part of a rotation is when land is
relatively short, labor is plentiful and farmers are at least partly subsistence orientated.
Under favorable management and when groundnut residues are incorporated on sandy
soils versus being fed to livestock, groundnut in rotation can double the following
maize grain yield, particularly when that maize is grown with little or no N fertilizer
(Mukurumbira, 1985; McDonagh et al., 1993).

Output was valued at the purchase price offered by the Grain Marketing Board
(GMB) for grade A white maize and shelled groundnut in 1996 and 1997. Yields were
adjusted downwards by 10 percent to account for a higher management level in
researcher-managed trials, as compared with fields managed by farmers and to correct
for possible yield overestimation in small experimental plots (CIMMYT, 1988). Maize
stover and groundnut haulms were assumed to have no value, although in some areas
these byproducts are fed to livestock. Net benefits/discounted net benefits where
calculated over cash costs and over all costs. Net benefits based on cash costs assumes
that there is no opportunity cost to labor and this may not be a true reflection of
farmer circumstances. The justification for using net benefits over costs include the

high unemployment rate mostly in the urban area or more than 50% of labor force and

72

the relatively few off-farm activities available to smallholder farmers.

In the case of maize-legume intercrops, only the seed cost of either cowpea or
sunnhemp was included in the analysis as this was the only variable seed costs. For the
three cropping patterns evaluated (maize-maize, maize+cowpea-maize and maize+
sunnhemp-maize), the response of maize to varying levels of fertilizer N (0, 46. 92
and 138 kg N ha") was fitted to an exponential curve. Exponential models for maize
response to N were used to derive the curves for added maize value. The added value
at a particular fertilizer level within a cropping system was calculated as its field
benefit less that of continuous maize when no N fertilizer was applied.

ElYl=a+BCXP(-5N)+e, [2]
where Y= added value (ZimS) for treatment, N = N rate applied (kg ha"), e = error.

Marginal benefit cost analysis was used to estimate the marginal rate of return
to alternative treatments (i.e., the difference in net benefits between this treatment and
any other, divided by the difference in costs between those treatnrents). Dominance
analysis (CIMMYT, 1988) was used to identify treatments having similar net benefits
but higher costs than other treatments. Sensitivity analysis was conducted to assess the
stability of results with changing seed prices of sunnhemp and cowpea, and yield

increase in a groundnut crop.

73

ECONOMIC ASSESSMENT

Continuous maize versus a maize-groundnut rotation

Domboshava. In Domboshava, continuous maize was grown for two years at
two levels of N (0 and 92 N ha") and in the third year at variable N rates (0, 46, 92
and 138 kg N ha"). In the maize-groundnut rotation, groundnut was planted in the first
year, followed by maize in the second year at two N levels (0 and 92 N ha") and
maize in the third year at variable N rates (0, 46, 92 and 138 kg N ha").

Year One: (Ln-11424142 vs [lg-11424142

For the first year of continuous maize (1994/95 season), maize fertilized at 92
kg N ha‘1 generated the highest net benefits (Zim$l 2,109) among the fertilizer
treatments (Table 3.1). The marginal rate of return (MRR) of moving from the
farmer’s check (maize with no fertilizer) to 92 kg N ha'1 was 172%. This suggest that
farmers who invested a dollar on fertilizer at Domboshava would recover the money
their initial cash outlay, plus an additional Zim$ 1.72 for each Zim$ 1.00 invested in
fertilizer. In contrast, farmers who planted groundnut in the same season realized a
loss of Zim$ 1,334 (Table 3.1). Since farmers who planted maize without fertilizer
would have earned Zim$ 522 ha“1 , this represents a total reduction in net benefits of

Zim$ 1,856 ha". Groundnut net benefits were lower than the other options because of

1Zim S = USD 0.18

74

low yields and higher labor requirements for planting, weeding and harvesting. For the
groundnut crop to equal net benefits of farmers check, yield needed to be raised to
0.7 t ha'l (more than double the current level).
Year Two: Gn-A_/I_z-ll/Iz vs [Viz-@4142

For year two (1995/96), partial budgets for both a first maize crop following
groundnut and the second maize crop in the continuous maize pattern are presented in
Table 3.2. In continuous maize, moving from zero fertilizer to the recommended rate
of 92 kg N ha’1 was associated with a marginal rate of return of 134%. When maize
was grown following groundnut, the discounted marginal rate of retum associated with
moving from the farmer’s check (continuous maize with no fertilizer) to maize
following groundnut was 692%. This increase in discounted net benefits for maize
following groundnut was mostly due to higher maize yields realized in this cropping
system due to the residual contribution of nitrogen provided by the preceding
groundnut crop. Although not recovered in the same year, each dollar invested in
growing groundnut in the previous year generated an additional ZimS 6.92 in the
subsequent year. When 92 kg N ha'1 was applied to maize following groundnut,
discounted net benefits increased by Zim$ 2,124, which represents a marginal rate of
return of 173% over maize following maize when no fertilizer was applied.

Year Three: Gn-Il/Iz-llfi vs 1142-1142413

In contrast, in year three (1996/97), fertilizer treatments were not justified for

the third maize crop in continuous maize crop, since the discounted net benefits for all

treatments were dominated by the farmer’s check (Table 3.3). However, moving from

75

the farmer’s check (continuous maize without fertilizer) to maize in rotation with
groundnut without fertilizer in the second maize crop had a discounted marginal rate
of return of 560%. Fertilizer additions on the second maize crop (0 vs 46 kg N ha")

following groundnut gave a discounted marginal rate of return of less than 50%.

Combined discounted marginal benefit cost analysis over three seasons

Reflecting on the results of the three seasons under consideration, it is rather
sobering to note that combined discounted marginal rate of return of moving from
farmer’s practice (continuous maize with no fertilizer) to rotating maize with
groundnut when no fertilizer was applied to either the maize or the groundnut crop
was only 4%, due to the low groundnut yield and high labor costs associated with
groundnut resulting in negative net benefits in the first year (Table 3.4). This MRR is
clearly not large enough to attract farmers to change from their current practice to a
new system that incorporates groundnut in rotation with maize. However, this study
showed that applying fertilizer to continuous maize at the recommended level of 92 kg
N ha'1 generated combined discounted marginal rate of return of 104%, which is
sufficiently high to encourage farmers to adopt this recommendation.

Year One: @4142 vs Il_/Iz_-A42

Chinyika. In Chinyika, trials conducted in 1995/96 and 1996/97 compared two
cropping patterns: a groundnut-maize rotation (Gn-Mz) and continuous maize (Mz-
Mz). In the first year (1995/96), the groundnut crop generated a negative net benefit of

Zim$ 1,917 (Table 3.5). This loss represented a decline of Zim$ 3,212 ha“1 in potential

76

net benefits, compared to maize without fertilizer. Highest net benefits were obtained
with continuous maize at 92 kg N ha‘1 (Zim$ 3,587) compared to groundnut and the
farmer’s control (continuous maize without fertilizer). The marginal rate of return of
moving from farmers’ control to 92 kg N ha" was 305%.

Year Two: (in-1141; vs Adz-1L4;

In the second year (1996/97), highest discounted net benefits were obtained
with continuous maize at 92 kg N ha'1 compared to maize following groundnut at
variable N levels (0, 46, 92 and 138 kg N ha") and the control (Table 3.6). For maize
following groundnut, highest net benefits were obtained when no fertilizer was applied
(Table 3.6). At all fertilizer levels in maize following groundnut, discounted net
benefits were dominated by the zero fertilizer treatment (Table 3 .6). However, when
no fertilizer was applied to either the second continuous maize crop or maize following
groundnut, the later generated a higher discounted net benefits (Table 3.6). The
marginal rate of retum for this change was 689%. The increase in discounted net
benefits for maize following groundnut was due to a yield increase of 0.2 t ha”1 over

continuous maize.

Combined marginal benefit cost analysis over two seasons

Summarizing these results over two seasons showed that continuous maize at
92 kg N ha’1 produced the highest benefits (Table 3.7). Combined discounted
marginal rate of return in continuous maize when moving from zero to 92 kg N ha‘1

was 133%. In contrast, over two seasons, the groundnut-maize rotation generated

77

negative discounted net benefits at Chinyika. These negative net returns were due to
low groundnut yield, little yield improvement in maize following groundnut, and the
high labor costs associated with growing groundnut crop. Clearly, the groundnut-maize
rotation is far less profitable than continuous maize, especially when the maize crop is
grown with fertilizer.
Costing of Labor

The profitability of the groundnut-maize rotation in the economic assessment
was highly inflenced by the high labor requirement for producing groundnut. When
labor was assigned a zero monetary value (i.e. Net Benefits over cash costs only) made
the rotation remotely profitable. Use of on going casual labor rate almost always
overestimated the monetary value of labor used to produce groundnut because almost
all the labor is normally supplied by female members of the household.

Maize-Legume Intercropping

In Domboshava, a maize-legume intercropping trial was conducted with the
following cropping patterns; sole maize, maize-cowpea intercrop and maize-sunnhemp
intercrop at three levels of fertilizer N (0, 60 and 120 kg N ha") in the first year. In
the second year, sole maize was grown following the different cropping patterns (in
the previous year) at variable N rates (0, 46, 92 and 138 kg N ha“).

In the first year, among the cropping patterns, highest net benefits (ZimS 2 166
per hectare) were obtained in the maize-cowpea intercrop at 60 kg N ha‘1 (Table 3.8).
For the maize monoculture, the marginal rate of retum for moving from zero to 60 kg

N ha‘1 was 12 percent, which is not sufficient to induce farmers to change their

78

cropping systems.

In contrast, the marginal rate of return of moving from the control (maize alone
without fertilizer) to maize-cowpea intercrop at 60 kg N ha’1 was 103%, suggesting
that for every dollar invested in cowpea and on fertilizer was recovered plus an
additional Zim$ 1.03. Given that a marginal rate of retum of between 50 to 100 % is
usually regarded as acceptable, the rate of return for this treatment is quite attractive.
Maize-sunnhemp intercrop at all the fertilizer levels considered, generated net benefits
that were less than the control (hence were dominated) (Table 3.8). The reduced net

benefits in maize-sunnhemp intercrop were due to the high cost of sunnhemp seed.

Second year following maize-legume intercrops

In the second year, maize was grown alone following the different cropping
patterns from the previous year (season). Fitted exponential regression equations were
calculated for added maize value and a linear equation for added cost for fertilizer
(Table 3.9).

In the first year (Table 3.8), when no fertilizer was applied the maize-legume
intercrops generated lower net benefits (maize-cowpea; Zim$ -254, maize-sunnhemp;
ZimS -409) compared to maize alone. Thus for cropping systems including legumes to
be as profitable as maize alone over two years, subsequent maize must generate net
benefits in excess by these values. When no fertilizer was applied, maize following the
maize-cowpea intercrop and maize following the maize-sunnhemp intercrop produced

additional Zirn$ 1,089 and Zim$ 645 per hectare, respectively (Fig 3.1), which

79

indicates that legume seed costs invested in the first year were recovered and a profit
realized.

Net benefits from N application are maximized when the difference between
the added maize value and added cost for fertilizer are greatest. For maize-sunnhemp,
maximum benefits were obtained at 109 kg N ha'1 (Fig. 3.1). Although maximum net
benefits in maize-cowpea were obtained at 138 kg N ha'1 (asymptote was not reached
at the current maximum N rate but at a higher rate than being considered), the
marginal rate of return of moving from 92 kg N ha'1 to 138 kg N ha‘1 was only 15
percent. This MRR indicates that fertilizer rates higher than 92 kg N ha'1 did not
increase corresponding maize benefits to a level that would be unattractive to farmers.

Sensitivity analysis on maize following the maize-legume intercrops showed
that a 50% increase in legume seed cost would greatly reduce net benefits except for
maize-cowpea at 60 kg N ha" (Table 3.10). A 50% increase in legume seed cost
requires the maize crop following the maize-cowpea and the maize-sunnhemp
intercrops to yield added values of Zim$ 147 and Zim$ 705 per hectare, respectively,
to equal the control (continuous maize without fertilizer). In this scenario the maize
crop following maize-cowpea intercrop generated added value in excess of Zim$
1,000, which is 680% higher than returns required to equal the continuous maize when

no fertilizer was applied.

80

Table 3.1 Partial Budget for the first maize crop in a continuous maize cropping
pattern and the groundnut crop in a groundnut-maize rotation in 1994/95 at
Domboshava, Zimbabwe.

 

 

 

Maize Groundnut
0 kg N ha’1 92kg N ha'1

Adjusted Yield (t 11a")T 1.78 3.87 0.30
Gross Field Benefit (ZDS ha") 2 136 4 644 1 500
Seed cost (ZDS ha") 166 166 350
Planting labor (ZDS ha")* 1 224 1 224 2 106
Fertilizer Cost (ZDS ha")1 0 657 0
Harvest cost (ZDS ha") 224 488 378
Total Cost that Vary (ZDS ha") 1 614 2 535 2 834
Net Benefits over cash costs 1 746 3 333 772
(ZDS ha")
Net Benefits over all costs 522 2 109 -1 334
(ZDS ha“)
Marginal rate of return NA 172% D

 

TYield was adjusted downwards by 10 percent. D=Dominated treatment by the control.
tLabor data for include land preparation, planting and weeding. ZDS = Zimbabwe

dollar

zCost include labor for fertilizer application.

NA=Not applicable because it is the baseline (control).

81

Table 3.2 Partial budget for the second crop and in a continuous maize cropping
pattern and maize following the groundnut crop in a groundnut-maize

rotation in 1995/96 at Domboshava in Zimbabwe.

 

 

 

MZ-fl-MZ Gn-m-Mz

N rate (kg ha") 0 92 0 92
Adjusted Yield (t ha“)l 2.20 4.72 4.15 7.35
Gross Field Benefit (ZDS ha") 2 640 5 664 4 980 8 820
Fertilizer Cost (ZDS ha“)* 0 824 0 824
Harvest Cost (ZDS ha“) 277 595 523 926
Total Cost that Vary (ZDS ha") 277 1419 523 1750
Dicounted NB over cash costs 2 146 3 935 4 049 6 418
(ZDS ha")

Discounted NB over all costs 1 921 3 451 3 624 5 748
(ZDS ha")1

Marginal rate of return (MRR) NA 134% 692% 173%§

141'. yeast-s

 

 

TYield was adjusted downwards by 10 percent

tCost includes fertilizer transport and application labor costs
"Calculation based on discount for year 1 after a groundnut crop= ( 1+ interest rate)".
5 MRR is for moving from Gn-Mz_-Mz without fertilizer to 92 kg N ha", otherwise is
from control (Mz-m-Mz without fertilizer) to the treatment.

NA=Not applicable because it is baseline (control). NB=Net benefits

82

L:

Table 3.3 Partial budget for the third maize crop in a continuous maize cropping
pattern and the second maize crop following groundnut in a groundnut-
maize rotation in 1996/97 at Domboshava, Zimbabwe.

 

 

 

Mz-Mz-Mz_ MZ-Gfl-MZ.

N rate (kg ha") 0 46 92 138 0 46 92 138
AdeSth Yield1 1.46 1.69 2.21 2.55 1.93 2.44 3.30 2.97
(t ha")
Gross Field 1752 2028 2652 3 060 2196 2928 3 960 3 564
Benefit (ZDS ha")
Fertilizer Cost’ 0 466 824 1305 0 466 824 1305
(ZDS ha”)
Harvest Cost 184 213 278 321 231 307 416 374 ‘

(ZDS ha")
Total Cost that 184 679 1 102 1626 23 1 773 1240 1679
Vary (ZDS ha")
Discounnted NB 1 159 1032 1208 1 160 1452 1627 2073 1493
over cash costs
(ZDS ha")
Discounted Net‘ 103 6 892 1024 948 1299 1424 1798 1246
Benefit over all
costs (ZDS ha")
NIRR (%)§ NA D D D 560 23 37 D

 

TYield was adjusted by downwards by 10 percent.
*Cost includes fertilizer transport and application labor costs.
1 Calculation based on discount for year 1 after a groundnut crop= (1+ interest rate)'2.

5 MRR = Marginal rate of return. D = Dominated treatment by control or a lesser level

treatment.

NA=Not applicable because it is baseline (control). NB=Net benefits

83

Table 3.4 Summary’I of net benefits associated with two fertilizer rates over three
seasons (1994-1997) in a continuous maize cropping pattern and maize-
groundnut rotation at Domboshava, Zimbabwe.

 

 

 

 

 

Cropping system Mz-Mz-Mz Gn-MZ-Mz

N rate (kg ha") 0 92 0 92
Year Zim$ ha'l

1994/95 522 2109 -1334
1995/96 1921 3451 3624 5748
1996/97 103 6 1024 1299 1798
Total 3479 6584 3589 6212
MRRt (%) 104 4 D

 

Mz=maize, Gn=groundnut.
TMRR=Marginal rate of return. D = Dominated treatment by control or a lesser level

treatment.

1 Tables 3.1, 3.2 and 3.3.

84

Table 3.5 Partial Budget the first maize crop in a continuous maize cropping pattern
and the groundnut crop in a maize-groundnut rotation in 1995/96 at

Chinyika, Zimbabwe.

 

 

 

Groundnut
0 kg N ha'1 92 kg N ha'1

Adjusted Yield (t ha")T 1.78 3.87 0.3
Gross Field Benefit (ZDS ha“) 3 000 6 296 720
Seed cost (ZDS ha“) 166 166 350
Planting labor (ZDs ha")* 1 224 1 224 2 106
Fertilizer Cost (ZDS ha")’I 0 657 0
Harvest cost (ZDS ha") 224 488 378
Total Cost that Vary (ZDS ha") 1 705 2 709 2 637
NB over cash costs (ZD$ ha“) 2519 1 485 531
NB over all costs (ZDS ha") 522 3 587 -1 917
Marginal rate of return NA 305% D

 

TYleld was adjusted downwards by 10 percent. D=Dominated by control.

*Labor data for include land preparation, planting and weeding. ZDS = Zimbabwe
dollar. NA= Not applicable because it is baseline (control).
*Cost include labor for fertilizer application. NB=Net Benefits.

85

Table 3.6 Partial budget for the second maize crop in a continuous maize cropping
pattern and the maize crop following groundnut in a groundnut-maize

rotation in 1996/97 at Chinyika, Zimbabwe.

 

 

 

Mlzm Gil-Ml.

N rate (kg ha“) 0 46 92 138 o 46 92 138
Adjusted Yield1 0.22 0.95 1.37 1.57 0.37 0.58 0.98 1.44
(t rial)

Gross Field 264 1 140 1644 1879 444 696 l 176 1728
Benefit (ZDS ha")
Fertilizer Cost* 0 466 824 1305 0 466 824 1305
(ZDS ha")
Harvest Cost 28 120 173 198 47 73 124 181
(ZDS ha")

Total Cost that 28 586 997 1503 47 419 948 1486
Vary (ZDS ha")

DNB over cash 215 547 667 467 361 285 286 344
costs (ZD$ ha”)

DNB1 over all 192 450 526 306 323 225 185 197
costs (ZDS ha“)

MRR (%) 46 18 D 6895 D D D

 

TYield was adjusted by downwards by 10 percent. MRR= Marginal rate of return.
tCost includes fertilizer transport and application labor costs. D= Dominated by lower

level treatment and or control.

1Discounted Net Benefits. Calculation based on discount for year 1 after a groundnut
crop= (1+ interest rate)“.
5Based on moving from continuous maize at zero fertilizer.
NA=Not applicable because it is baseline (control).

86

Table 3.7 Summary'I of net benefits associated with two fertilizer rates over two
seasons (1995-1997) in a continuous maize cropping pattern and maize-
groundnut rotation at Chinyika, Zimbabwe.

 

 

 

 

 

Cropping system Mz-Mz Gn-Mz

N rate (kg ha”) 0 92 0 92
Year ZimS ha'l

1995/96 1295 3587 -1917

1 996/ 97 192 526 323 185
Total 1487 4113 -1594 -l732
MRRT (%) 133 D D

 

Mz=maize, Gn=groundnut.

TMarginal rate of return. D = Dominated by control or lower level treatment.

‘Tables 3.5 and 3.6.

87

Table 3.8 Partial budget for maize + legume intercrops and monoculture maize
cropping systems at Domboshava, Zimbabwe.

 

 

 

Cropping system Maize alone Maize + cowpea Maize + sunnhemp

N rate (kg ha") 0 60 120 0 60 120 0 60 120

Adjusted Yield' 1.1 1.74 2.3 1.00 2.72 2.11 1.27 2.13 1.82

(t ha“)

Gross Field Benefits 1320 2088 2760 1200 3264 2532 1524 2556 2184
( 3 ha")

Seed Cost‘ 0 0 0 147 147 147 592 592 592
(3 ha“)

Fertilizer Costs1 0 608 1216 0 608 1216 0 608 1216
(3 ha")

Harvest Cost 139 219 290 126 343 266 160 268 229
(3 ha")

Total Costs that Vary 139 827 1506 273 1098 1629 752 1468 2037
(8 ha")

Net Benefits 1181 1261 1254 927 2166 903 772 1088 147
(5 ha")

MRR (%) NA 12 D D 103 D D D D

-_Vrva‘kl-_‘.n 1- -——.

 

 

TYields were adjusted downwards by 10 percent. MRR=Marginal rate of return.
*Seed cost of cowpea or sunnhemp and associated transport charges to farm-gate
‘Include transport and application labor costs.

D=Dominated by control or lower level treatment.

NA=Not applicable beciase it is baseline (control).

88

P In Us .6

Table 3.9. Fitted Regression Equations for added maize value (ZimS) and added cost .
of fertilizer as a function of N fertilizer applied (x) in the subsequent year
(maize + legume - maize).

 

 

 

Cropping system EquationT Standard Error

0: B 5
Maize - Maize Y=4126-4098 exp[ -0.0l8(x)] 679 717 8 x 10'3
Maize + cowpea - Y=5479-4390exp[-0.0109(x)] 12.11 11.67 6 x 10"
Maize

Maize + sunnhemp - Y=3659-3014exp[-0.0302(x)] 60.32 80.51 2 x 10'3
Maize

E Significance
Added Cost Y = 259 + 12(x) 0.98 P _<_ 10“

 

T Exponential equations were fitted to model E[Y] = a + B exp(-6x).

89

Table 3.10 Sensitivity analysis on net benefits of a maize-cowpea and maize-sunnhemp
intercrops as influenced by cowpea or sunnhemp seed cost changes.

 

 

 

 

 

Cropping Maize + cowpea Maize + sunnhemp

systemT

N rate (kg ha“) 0 60 120 0 60 120
Zim$ ha‘l

Current price 927 2166 903 772 1088 147

50% rise 834 2093 830 476 792 -49

100% rise 780 1872 609 180 496 -445

50% decline 1000 2240 977 1068 1384 443

 

TControl values for continuous maize are Zim$ 1181, 1261 and 1254 for 0, 60 and
120 kg N ha‘1 , respectively.

90

Added Maize Value (Zim$ ha")

5000

 

 

 

 

—o— Maize-maize
—v— Maize+cowpea-maize ' V
4000 _ + Maize+sunnhemp—maize
—<>— Added cost '
3000 d
2000 7 o
0
O
O
1000 - ' o
O
O
O
0 " I 1 I l
0 30 60 90 120 150
N Applied (kg ha")

Figure 3.1. Effect of rate of N application on added maize value and added cost
for fertilizer at Domboshava following maize-legume intercrops.

91

CONCLUSIONS

Economic assessment of continuous maize versus a maize-groundnut rotation
showed that continuous maize at 92 kg N ha'1 optimized net benefits over labor and
input costs. However, should a farmer prefer to utilize a maize-groundnut rotation
because of rotational benefits such as pests and diseases reduction, the analysis
suggests that net benefits would be maximized when no fertilizer was applied to either
maize or groundnut, or when the current groundnut crop yields were doubled,
assuming labor costs remained constant.

The results for groundnut-maize rotation underline the need for research to (1)
increase the yield of groundnut on smallholder farms and (ii) reduce the associated
labor costs in producing groundnut without adding much to cash costs. Groundnut
yield improvements and reduced labor requirements can be reduced by intermediate
technologies for planting, harvesting and processing. Use of intermediate technolgies
will initially require institutional support by both public and private sector.

The analysis on maize-legume intercrop has shown that in the first year the
maize-cowpea intercrop, at 60 kg N ha", generated higher net benefits than sole maize,
even when the economic value of cowpea was not included. In the second year, maize
yields following the maize-legume intercrops were high enough to pay legume seed
outlays in the first year, plus additional cash benefits. Sensitivity analysis revealed that
a 50% increase in cowpea and sunnhemp seed costs would greatly reduce net benefits

for all maize-legume intercrops, except for the maize-cowpea intercrop at 60 kg N ha”.

92

REFERENCES

Bernsten, RH. 1980. Pre-screening alternative component technologies to identify
probable constraints to farmer adoption. A paper presented at the Cropping
Systems Economic Training Program, June 2-4 Bogor, Indonesia.

CIMMYT. 1988. From Agronomic Data to Farmer Recommendation: An Economic
Training Manual, completely revised Edition, Mexico, D.F.

Gittenger, J.P. 1984. Economic Analysis of Agricultural Projects. Second Edition. The
Johns Hopkins University Press. Baltimore, USA, pp. 299-361.

Hesterman, O.B., C.C. Sheaffer, and E1. Fuller. 1986. Economic comparisons of crop
rotation including alfalfa, soybean and corn. Agron. J. 78: 24-28.

McDonagh, J.F., B. Toomsan, V. Limpinuntana and KB. Giller. 1993. Estimates of
the residual nitrogen benefit of groundnut to maize in Northeast Thailand.
Plant and Soil 154:267-277.

Mukurumbira, L.M. 1985. Effects of rate of fertilizer nitrogen and previous grain
legume crop on maize yields. Zimb. Agric. Journal 82:177-179.

Shumba, E.M., RH. Bernsten and SR Waddington. 1990. Maize and groundnut yield
gap analysis for research priority setting in the smallholder sector of Zimbabwe.
Zimb. Journal of Agric. Res 28:105-113.

Shumba, E.M., SR Waddington, and M. Rukuni. 1992. Use of tine-tillage, with
atrazine weed control, to permit earlier planting of maize by smallholder
farmers in Zimbabwe. Experimental Agriculture 28: 251-254.

Stevens, RD, and CL. Jabara. 1988. Agricultural Development Principles. The John
Hopkins University Press. Baltimore and London.

Waddington, S.R., J. Karigwindi, and J. Chifamba. 1997. Productivity and profitability
of maize + groundnut rotations when compared to continuous maize under
smallholder management in Zimbabwe. In Proceedings for the Soil F ert Net
Results and Planning Workshop (in press).

93

Table A31 1. Price and labor data used for the cost-benefit analysis.

 

Data

Value

 

GMB buying price for Grade A white maize
GMB buying price shelled groundnut
Compund D (8-14-7)

Ammonium nitrate fertilizer

Maize seed price @25 kg ha‘1
Groundnut seed @70 kg ha'1
Cowpea seed cost

Sunnhemp seed cost

Local daily casual worker wage rate
Labor for Maize

Land preparation

Planting

Fertilizer (basal + top)

Weeding (2 hand weedings)
Cutting/stocking

Removing eats

Shelling

Labor for Groundnut

Land preparation

Planting

Weeding (2 hand weedings)

Pulling

Plucking

Shelling

Discount rates for 1996 and 1997

Zim$ 1 200 per t.
Zim$ 5 000 per t.
Zim$ 2 260 per t.
Zim$ 2 490 per t.
Zim$ 166 per ha
ZimS 350 per ha
Zim$ 184 per 25 kg
Zim$ 494 per 25 kg
Zim$ 18 per day

4 person-days ha"

9 person-days ha’1

8 person-days ha"

24 person-days ha"

12 person-days ha'1

19 person-days ha”1

7 person-days per t grain

4 person-days ha'1

18 person-days ha'1

57 person-days ha"

18 person-days ha'1

20 person-days ha'1

7 pers-days/ 100 kg grain
23%

 

94

“1111111111111111111111111116
31293017721105