Mulches in smallholder maize systems in the Limpopo ... · MULCHES IN SMALLHOLDER MAIZE SYSTEMS IN THE LIMPOPO PROVINCE OF SOUTH AFRICA: UNTANGLING THE EFFECTS OF N THROUGH EXPERIMENTATION

MULCHES IN SMALLHOLDER MAIZE SYSTEMS IN THE

LIMPOPO PROVINCE OF SOUTH AFRICA: UNTANGLING THE

EFFECTS OF N THROUGH EXPERIMENTATION AND

SIMULATION

By

Seshuhla Rebinah SASA

Thesis submitted in fulfilment of the requirement for the degree of

Master of Agricultural Science

School of Agriculture, food and wine

Faculty of sciences

University of Adelaide, Australia

July 2009

ii

Table of contentsTable of contents……………………………………………………………………….…. ii

List of tables……………………………………………………………………………..... iv

List of figures………………………………………………………………………….…....v

Declaration………………………………………………………………………………..vii

Acknowledgements……………………………………………………………………….viii

Abstract………………………………………………………………………………….…ix

Chapter 1 .......................................................................................................................... 1General Introduction........................................................................................................ 1CHAPTER 2..................................................................................................................... 4Literature review .............................................................................................................. 4

2.1 Suitability of legumes for enhancing subsequent crop production ............................4

2.2 Criteria for selecting legume crops .............................................................................4

2.3 Biological Nitrogen fixation in legumes ......................................................................5

2.4 Legume biomass accumulation ...................................................................................7

2.5 Factors affecting legume growth and nitrogen fixation .............................................8

2.6 Effects of mulch on growth parameters of subsequent crops. .................................10

2.7 Decomposition of mulches and green manure..........................................................12

2.8 Factors affecting decomposition of mulches .............................................................12

2.9 Effects of legumes on subsequent crops....................................................................15

2.10 Effects of mineral nitrogen fertiliser on crops..........................................................16

2.11 Simulation modelling for agricultural research .......................................................17

2.12 Conclusion .................................................................................................................18

CHAPTER 3................................................................................................................... 21Cereal and legume management by subsistence farmers in Limpopo province of South Africa: Socio-economic and farming details. ................................................................. 21

3.1 Introduction...............................................................................................................21

3.2 Materials and methods..............................................................................................22

3.2 Results and discussion...............................................................................................24

3.3 Conclusion .................................................................................................................44

CHAPTER 4................................................................................................................... 46The effects of fertiliser, legumes and grass mulches applied to a maize crop in Limpopo province .......................................................................................................................... 46

4.1 Introduction...............................................................................................................46

4.2 Material and methods ...............................................................................................47

4.3 Results .......................................................................................................................52

4.4 Discussion ..................................................................................................................65

4.5 Conclusion .................................................................................................................70

iii

CHAPTER 5................................................................................................................... 73Using closed pot incubations to investigate the N and C mineralization in crop residues of varying quality............................................................................................................ 73

5.1 Introduction...............................................................................................................73

5.2 Materials and methods..............................................................................................74

5.3 Results .......................................................................................................................78

5.4 Discussion ..................................................................................................................92

5.5 Conclusions..............................................................................................................101

Chapter 6 ...................................................................................................................... 103General discussions ...................................................................................................... 103

Literature cited…………………………………………………………………………..107

Appendices……………………………………………………………………………….117

iv

List of tablesTable 2.1. The amount of nitrogen kg ha-1 fixed by different legumes crops ...……………6

Table 2.2. The amount of biomass (t ha-1) produced by different legume crops...…….…...7

Table 2.3. C:N ratio in different crops……………………………………………….........13

Table 2.4. Lignin percentages in different crops…………………………………….…….14

Table 3.1. Soil chemical analysis for the soil profiles………………………………..…....23

Table 3.2. Soil particle analysis……………………………………………………………24

Table 3.3. Smallholder farmers’ ages and their level of education (%)…...…………..…..26

Table 3.4. Types of cropping methods applied by farmers………………………………..35

Table 3.5. Timelines for maize cropping…………………………………………..……...35

Table 3.6. Ploughing equipment used and number of ploughing…………………….…...36

Table 3.7. The % of farmers using types of manure, source and time application ……….37

Table 3.8. The % of farmers knowing about the potential benefits of N fixation bylegumes, application of skill and their response to N fixation information

…………………….…………………………………………………………..…...41

Table 3.9. The farmers source of information and farmers group membership ………..…44

Table 4.1. Treatments designed and implemented at GaKgoroshi and Gabaza…….…..…48

Table 4.2. Rainfall (mm) during 2007-2008 and long term average (LTA)……..………..52

Table 4.3. Soil chemical analysis for the soil profiles………………………………..…...52

Table 4.4. Soil particle size analysis……………………………………………..………..53

Table 4.5. The frequency of water and N deficient factor >0.5 during flowering (FS)

to end of grainfill (SE), and soil water evaporation above average

(137 mm) ………………………………………………………………..………...65

Table 5.1. Properties of soils used in the incubation………………………………………75

Table 5.2. Quantities of residue C, N and C:N ratio incorporated into soils……………...79

Table 5.3. The efficiency of residue C utilisation by microbes following the

amendment of 4 residues and 2 soil types ………….………………………….....83

Table 5.4. Ammonium and nitrate concentrations in Tarlee and Waikerie soils after

the incorporation of canola, wheat, pea and mucuna……………………………..85

Table 5.5. The efficiency of residue C utilisation for Waikerie soil using 3 types of

residue and 2 methods of residue application…………………...………………..89

Table 5.6. Ammonium and nitrate-N (mg N kg-1) for the incorporation and mulch of

the different plant materials in Waikerie soil…………...…………………………91

v

List of figuresFigure 3.1: Rainfall map of the Limpopo province of South Africa………………………23

Figure 3. 2. Family size grouped in the number of people per household……………...…27

Figure 3.3. Income sources per household and the % of farmers receiving themes………28

Figure 3.4. Income constraints face by smallholder farmers……………………………...29

Figure 3.5. Types of livestock owned by farmers…………………………………………30

Figure 3.6. % land allocation to maize as compared to legumes and other crops…………32

Figure 3.7. Major constraints faced by farmers……………………………………………34

Figure 3.8. Number of weeding applied by smallholder farmers………………………….38

Figure 3.9. Types of residue management………………………………………………...39

Figure 3.10. Problems associated with legume derived N by smallholder farmers……….42

Figure 3.11. Reasons for non-membership by smallholder farmers………………………44

Figure 4.1. Maize plant height as influenced by different soil fertility management

Practices…………………………………………………………………………...54

Figure 4.2. Maize dry-matter as influenced by different soil fertility management

practices……………………………………………………………………………55

Figure 4.3. The relationship between plant height and drymatter as influenced by

different soil fertility management………………………………………………...56

Figure 4.4. Comparison between the observed and simulated maize biomass during

the 2007-2008 growing season…………………………………………………..57

Figure 4.5. Cumulative distribution functions for maize dry matter with different

mulch and N fertiliser treatments during long term period

(1970-2008) …………………………………………………………………...…..58

Figure 4.6. Cumulative distribution function for maize grain yield (1971-2008)…………59

Figure 4.7. Average soil water deficit factor for maize growth during the 2007- 2008

growing season…………………………………………………………………….60

Figure 4.8. Nitrogen deficit factor on maize grain yield during the 2007-2008

growing season…………………………………………………………………….61

Figure 4.9. Correlation between long-term in-crop rainfall and N stress during

flowering to start of grainfill (A) and from start to end of grainfill (B)…………...63

Figure 4.10. Correlation between long term growing season rainfall and soil water

stress from flowering to start of grainfill…………………………………………..64

Figure 5.1. Cumulative C mineralisation for Tarlee and Waikerie soil for 98 days period

amended with wheat , canola , mucuna and pea or a no-residue control……...….80

vi

Figure 5.2. Microbial biomass C for Tarlee and Waikerie soils with and without the

application of canola, wheat, pea and mucuna during the 98 day incubation

period……………………………………………………………………………...82

Figure 5.3. The percentage C for Tarlee and Waikerie residue treatments at the end of

the incubation……………………………………………………………………...84

Figure 5.4. Cumulative C mineralisation in incorporated and mulched wheat, mucuna

and pea in Waikerie soil for 119 days……………………………………………..87

Figure 5.5. Microbial biomass C for Waikerie soil with and without the application of

wheat, pea and mucuna during the 119 day incubation period……………………88

Figure 5.6. The percentage C remaining for incorporated and mulched residues in

Waikerie soil at the end of incubation……………………………………………..90

vii

Declaration

NAME: Seshuhla Rebinah SASA PROGRAM: Master of Agricultural Science

This work contains no material which has been accepted for the award of any other degree

or diploma in any university or other tertiary institution and, to the best of my knowledge

and believe, contains no material previously published or written by another person, except

where due reference has been made in the text.

I give consent to this copy of my thesis, when deposited in the University Library, being

made available for loan and photocopying, subject to the provisions of the Copyright Act

1968.

I also give permission for the digital version of my thesis to be made available on the web,

via the University’s digital research repository, the Library catalogue, the Australian

Digital Theses program (ADTP) and also through web search engines, unless permission

has been granted by the University to restrict access for a period of time.

SIGNATURE Date

Regular Thesis

viii

AcknowledgementsI would like to acknowledge the invaluable support given by my supervisors Prof. Gurjeet

Gill and Dr. Anthony Whitbread. I thank the Australian Center for International

Agricultural Research (ACIAR) for funding my studies, the University of Adelaide for

giving me the opportunity to study at the institution and the CSIRO for the skills I achieved

and the support the staff members have provided, emotionally, spiritually and physically. I

appreciate the support given by Dr. John Hargreaves on APSIM training and the laboratory

assistance offered by Bill Darvoren. I acknowledge the support of my ex- supervisor Dr.

Bill Bellotti during his time in Adelaide University.

I would like to thank the smallholder farmers in Gabaza and GaKgoroshi for offering me

land to conduct my field experiment. The biggest appreciation is given to my employer,

The Department of Agriculture, Limpopo province, South Africa for allowing me this

opportunity to study abroad. I aknowledge the support I got from my colleague, Mr J.J.

Mkhari for collecting the remaining data on my behalf, Prof. J.J.O. Odhiambo from the

University of Venda for arranging theplanting of mucuna to be used in my field experiment

and Sankie Lephale for assisting in data collection. The success of this study is dedicated

to my dearest friends and colleagues (Dobane Sebola, Veronica Matheba and Meta

Matsebe) who remained the pillar of my support where I couldn’t reach. I thank my

mother, Aletta Sasa has for always been courageous to me.

ix

Abstract

In Limpopo Province of South Africa, poor soil fertility and low crop yields are serious

problems facing resource poor smallholder farmers. A survey of over 60 farmers in 2

villages (Gabaza and GaKgoroshi) found that most of the smallholder farmers were women

(68%), elderly (50% above 68 years of age) and had not attended school or only attended

up to the primary level (80%). Very few farmers kept livestock (usually in small numbers)

and most grew cereal and legume crops (on 1ha of land) for home consumption and

livestock feed, with legumes being planted on 13% of the land. The study showed that 80%

of farmers were not fully aware of the benefits of legumes in fixing nitrogen (N) and

improving yield.

A field study at the survey village of Gabaza found that the application of fertiliser N and

grass mulch combination and fertiliser N plus guarbean mulch significantly increased plant

height and maize shoot growth at 4 and 8 weeks after planting. However, when grass

mulch was without N fertiliser, there was no increase in maize growth relative to the

control (0N).

A farming systems simulation model (Agricultural Production Systems sIMulator -

APSIM) was used to simulate this field study as well as over the long-term (1971 to 2008).

Simulation analysis showed poor average maize yield (<3000 kg ha-1) with the application

of grass residues even when used with 30 kg N fertiliser. However, the application of

guarbean residues as mulch with or without N fertiliser and as green manure increased

maize yields to >4000 kg ha-1. Simulation showed that the grass mulch with or without the

addition of N fertiliser reduced water stress and soil water evaporation but increased N

stress during the reproductive phase of the crop in most seasons. When guarbean mulch

was used as green manure by itself, or mulch plus N fertiliser, N stress was reduced but

water stress and soil water evaporation were increased which could have been due to faster

decomposition of legume mulch as compared to grass mulch. Addition of N fertiliser

reduced N stress to maize but increased water stress and soil water evaporation similar to

the guarbean mulch because of high soil evaporation.

APSIM analysis clearly showed the importance of N x soil water interactions in

determining maize growth and yield at Gabaza. Therefore, two studies were undertaken in

the laboratory in Australia to determine the dynamics of carbon (C) and N where residues

of different qualities [canola (C:N 43), wheat (26), pea (9) and mucuna (14)] were applied

x

to clay loam (Tarlee) or sandy (Waikerie) soils. In experiment 1, where residues were

incorporated into the two soils, the cumulative CO2-C evolution for the wheat and canola

treatments at the end of the incubation period were fairly similar but significantly higher

than for pea, mucuna and the control. In general, the application of residues increased

microbial biomass C more than the control, with highest increases up to 1.48 and 1.56 mg

C g-1 soil for canola and wheat in Tarlee soil, respectively and 0.82 mg C g-1 soil for pea in

Waikerie soil. Even though the Tarlee soil showed greater C release than Waikerie soil, the

C turnover from the residues between the 2 soils was not significantly different except for

pea residues. Canola and wheat residues were found to immobilise N whereas N content

increased in both soils with the application of legumes (pea and mucuna).

In experiment 2, mucuna, pea and wheat residues were either incorporated or applied as

surface mulches on Waikerie soil. Initially the CO2-C release was higher for incorporated

than mulched residues and CO2-C released was higher for pea residues. However, at the

end of the incubation more CO2-C was released with the application of wheat residue

indicating differences between residue types in the pattern of soil respiration. Microbial

biomass C was higher for incorporated than mulched residue treatments; pea residue

showed the highest biomass C for incorporated (0.78 mg C g-1 soil) whereas mucuna had

the highest microbial biomass (0.11 mg C g-1 soil) treatments. The method of residue

application resulted in a significant difference in C turnover between residues, with pea

residue showing significant increase in C utilisation than mucuna and wheat. The pea

residues, which had the lowest C:N, increased soil mineral N more than other treatments in

both incorporated and mulched treatments. Lower mineralisation of N observed in residues

of high C:N ratio compared to the control could be due to immobilisation of N. Therefore,

understanding the nutrient dynamics of different crop residues could play an important role

in the management of residues in different soil types. Based on these results it can be

concluded that legume residues have the potential to improve soil fertility and crop yields

in dryland farmers’ fields in Limpopo. Extension programs aimed at increasing farmers’

knowledge of the benefits of N fixation by legumes may increase their adoption and

thereby improve soil fertility and maize yield.

1

Chapter 1

General Introduction

Limpopo province is one of South Africa’s nine provinces and it covers an area of about

12.46 million (m) hectares (ha) which accounts for 10.2% of South Africa’s total land area

of 122.3 m ha (de Villiers et al., 2007; Lehohla, 2004). The province has a total estimated

population of 5,273,642 which constitutes 11.8% of the country’s total population of

44,819,778 (Lehohla, 2004). Limpopo is a rural province relying on agriculture, mining

and tourism for economic growth. Agriculture in Limpopo province is divided into two

distinct systems: commercial and subsistence agriculture, which occupy 14.7% and 14% of

the province’s land, respectively (Department of environmental affairs and tourism, 2008).

The two systems of agriculture produce similar crops and livestock; however, they differ in

the scale of operation and method of production.

Farmers in Limpopo province grow a variety of crops like cereals and cash crops in order

to meet the demand of the growing population. According to Statistics South Africa

(2001), the population of Limpopo increased from 4,929,368 in 1996 to 5,273,642 in 2001.

Commercial farmers practise large scale farming using the most advanced production

technology. Large scale farming systems range in area from 600 to 2000 ha according to

Lehohla (2004). These commercial farmers operate large farms, which are well organized

and situated on prime land whereas subsistence agriculture is practised by smallholder

farmers in rural areas on land ranging from half (0.5) to two (2) ha, which are rain fed. The

discussion here will be focussed on the performance of subsistence agriculture.

Farming under the smallholder system is characterised by a low level of production

technology and small sized farm holdings with production primarily for subsistence, with

little marketable surplus. For example, Whitbread and Ayisi (2004) mentioned yields of

<500kg ha-1 in maize, which is a staple food in Limpopo province. The smallholders are

faced with the problems of poor soil fertility and variable rainfall. The poor yields have

raised interest in research on improving yields (Misiko, 2007; Murh et al. 2002; Tittonell et

al. 2007). However, the crops chosen for these studies are not necessarily those which are

most commonly grown by smallholder farmers in Limpopo.

The crops grown by smallholder farmers are commonly cereals, cash crops and legumes,

which they normally intercrop but with cereal crops occupying a larger area of the field

2

than other crops. According to Food and Agriculture Organisation (2004) (cited in Peoples

et al., 1995), global allocation of arable land greatly favoured cereals (48%) with legumes

only occupying 11% land. This is similar to the situation in Limpopo province where

maize is grown with legumes (maize occupying larger portion of the land). Research has

focussed on cereal crops of barley (Hordeum vulgare), wheat (Triticum aestivum) and oats

(Avena sativa), and legumes such as peas (Pisum sativum), soybean (Glycine max), velvet

bean (Mucuna pruriens), jack bean (Canavalia ensiformis), pigeon pea (Cajanus cajan),

chickpea (Cicer arietinum), lablab (Lablab purpureus) and crotalaria (Crotolaria juncea)

(Carsky et al., 2001; Saxena, 1986; Wortmann and McIntyre, 2000) smallholder farmers

grow maize (Zea mays), millet (Pennisetum glaucum), sorghum (Sorghum bicolor),

bambara groundnut (Vigna subterranea), cowpea (Vigna unguiculata), peanuts (Arachis

hypogaea), sugar bean (Phaseolus limensis), pumpkin (Cucurbita pepo) and watermelon

(Citrullus lanatus). These crops are grown for human consumption and the stover is left in

the field for livestock feed during the winter season.

Rainfall in Africa is highly variable in amount and distribution from region to region as

well as from year to year. For example, Carsky et al. (2001) reported very different total

annual rainfall in Kaduna and Bauchi, two sites in northern Nigeria. Rainfall in Limpopo

ranges from 300 to 750mm per annum. Whitbread and Ayisi (2004) indicated variations in

rainfall in three locations of Limpopo province during the 1998 to 2002 growing seasons.

Given this scenario, the integration of legumes into cropping systems has the potential of

improving water use efficiency of the subsequent crop during low rainfall seasons through

nitogen (N) supply (Armstrong et al., 1997). In many studies cereal crop yield increased in

plots that were previously planted to legumes than non-legume crops (Fofana et al., 2004;

Mapfumo et al., 2005; Mpangane et al., 2004; Schultz, 1995). In addition to the variability

in rainfall, the quality of the soil also impacts upon smallholder agriculture in Limpopo.

The soils in Limpopo province are poor and highly degradable, as de Villiers et al. (2007)

pointed out, over 30% of soils in Limpopo are sandy in texture (less than 10% clay) and

almost 60% of the soils have low organic matter content and low levels of N (de Villiers et

al., 2007). The benefits of inorganic fertilisers, farmyard manure and inclusion of legumes

in the cropping systems to improve soil fertility have been supported by studies conducted

by Giller (2001); Mapfumo et al. (2005) in Zimbabwe as well as Koenig and Cochran

(1994) in the United States of America. The improvement of soil fertility in Limpopo

province was observed by Mpangane et al. (2004) when legumes were included in the

3

maize cropping systems. The extent to which crop residues influence plant growth is

determined by the amount of biomass, decomposition and nitrogen mineralisation rates,

and the timing of N release. Therefore, there is a need to understand the N and carbon (C)

dynamics of crop residues under different methods of residue incorporation.

The studies conducted so far have not compared relative effectiveness of legume derived

and mineral nitrogen in increasing maize yield in Limpopo which has a highly variable

climate. The literature review will cover research identifying the suitability of legumes for

enhancing subsequent crop production, effects of mulch on growth parameters of

subsequent crops, decomposition of mulches and factors affecting decomposition, and

effects of legume residue on subsequent crop growth. The review will assist in selecting

legumes better suited for efficiency and practicality in providing N to maize crops in a

variable climate.

4

CHAPTER 2

Literature review

2.1 Suitability of legumes for enhancing subsequent crop production

Legumes are crops that fix atmospheric nitrogen, through symbiotic relationship with

rhizobia in the soil, into forms that plants can absorb. They show different growth

characteristics in the field which are associated with their importance and intended uses in

the farming systems. Legumes can be used to improve soil fertility through nitrogen

fixation, soil cover and weed control (Koenig and Cochran, 1994; Misiko, 2007; Murh et

al. 2002; Ramakrishna et al. 2006). However, in Limpopo, legumes are normally grown for

personal food consumption by smallholders and there is little awareness of their

importance for enhancing subsequent crop productivity. Therefore, there is a need to

determine the criteria for selecting legumes for supplying nitrogen to dry land maize crops

in Limpopo province.

2.2 Criteria for selecting legume crops

Legumes planted in different regions differ with climate and soil types; however, their

intended use by farmers also contributes to their selection in farming systems. Legumes are

important in providing good quality protein, providing nutritious fodder for livestock and

improving soil fertility through nitrogen fixation (Saxena, 1986; Zaroug, 1986). Despite

the importance of food legumes for soil fertility, smallholder farmers in Limpopo only

grow food legumes for home consumption and livestock fodder. Fodder legumes are in fact

more efficient soil fertilisers than food legumes (Murh et al., 2002; Peoples et al., 1995;

Prasad and Power, 1997; Saxena, 1986). Although fodder legumes could be introduced to

farmers in Limpopo, it might be difficult to persuade them to adopt these legumes as they

normally prefer to grow food legumes.

Legumes (both food and fodder) improve soil fertility when incorporated as mulch, green

manure and cover crops. These benefits were demonstrated by Giller (2001) and Koenig

and Cochran (1994) in the USA, where soil fertility was improved by the introduction of

legumes. However, smallholder farmers in Limpopo do not use legumes for green manure

and cover crops but grow food legumes such as bambara groundnut (Vigna subterranean),

cowpea (Vigna unguiculata), peanuts (Arachis hypogaea) and sugar bean (Phaseolus

5

limensis), which are consumed as either green leaf vegetables or harvested as grain.

Smallholder farmers in Limpopo province rely on legumes in the natural vegetation for

livestock feed.

2.3 Biological Nitrogen fixation in legumes

Nitrogen (N) is required by plants and animals for growth and survival. It is found in

abundance in the atmosphere, occupying about 80% of the atmosphere in a gaseous form

of N2, which is not readily available for use by plants or animals. The dynamics of nitrogen

fixation are described by Giller (2001), Sarrantonio (1991) and Prasad and Power (1997).

Legumes fix nitrogen from the atmosphere (N2) through their roots and provide it to

subsequent crops when their residues are mineralised by microbes (Ayisi and Mpangane,

2004; Koenig and Cochran, 1994; Prasad and Power, 1997). The residues are particularly

useful in the form of organic manures, due to their high nitrogen content which is more

likely to become readily available for uptake by other plants than nitrogen in many other

crop residues (Armstrong et al. 1999). However, the amount of nitrogen fixed by legumes

can vary considerably (Carsky and Ndikawa, 2009; Saxena, 1986). Therefore, selection of

legumes for improving soil fertility for a particular area requires careful background

research.

Although legumes fix atmospheric nitrogen that is required by subsequent crops, the

amount of nitrogen provided by legumes in Limpopo province is limited as the above

ground material and seeds are consumed by either people or livestock.

For Limpopo Therefore, it would be important to select legumes that are capable of fixing

more nitrogen under variable climatic conditions. As a general rule, forage legumes fix

higher amounts of nitrogen than grain legumes (Table 2.1).

6

Table 2.1. The amount of nitrogen (kg ha-1) fixed by different legumes crops.

Forage Climatic

zone

Legume crop kg N ha1

year1

Author

Temperate Clovers 23-620 Prasad and Power (1997)

Lucerne/ alfalfa 164-386 Prasad and Power (1997); Peoples et al.

(1995)

Tropical Stylosanthes

guianensis

30-196 Prasad and Power (1997); Murh et al.

(2002); Peoples et al. (1995)

Tick clover 700 Prasad and Power (1997)

Grain

legumes

Temperate Vetch and Tick

beans

57-190 Prasad and Power (1997)

Peas 46- 244 Prasad and Power (1997)

Lupins 128- 288 Prasad and Power (1997); Peoples et al.

(1995)

Tropical Lentil 35-107 Prasad and Power (1997); Peoples et al.

(1995); Saxena (1986)

Pigeonpea 41-235 Prasad and Power (1997); Peoples et al.

(1995)

Cowpea 73-354 Prasad and Power (1997)

Soybean 17-450 Prasad and Power (1997); Peoples et al.

(1995)

Cluster beans 37-196 Prasad and Power (1997)

Groundnut 33-206 Prasad and Power (1997); Peoples et al.

(1995)

Chickpea 41-270 Prasad and Power (1997); Saxena,

1986); Peoples et al. (1995)

Mungbean 224 Prasad and Power (1997)

Faba bean 53- 330 Peoples et al. (1995)

Common bean 0-125 Peoples et al. (1995)

Green gram 9-112 (Peoples et al. 1995)

Black gram 21- 140 (Peoples et al. 1995)

7

2.4 Legume biomass accumulation

Legume biomass is important in improving soil organic matter. Different legume crops

tend to produce different amounts of biomass. Legume biomass accumulation is a function

of several factors such as rainfall, temperature and soil. For example, Murh et al. (2002)

and Odhiambo (2004) showed the differences in the amount of biomass produced by

different legume species. In addition to the differences in legume varieties, climatic

conditions also affect biomass production in legumes. Such variation in growth was

demonstrated by Caamal-Maldonado et al. (2001); Carsky and Ndikawa (2009); Wortmann

and McIntyre (2000) as well as Odhiambo (2004) under bimodal and unimodal rainfall,

respectively (Table 2.2). In the Table below, Stylosanthes guianensis shows more biomass

accumulation than other legumes in bimodal rainfall areas

Table 2.2. The amount of biomass (t ha-1) produced by different legume crops(Information was sourced from works of Murh et al. (2002) and Wortmann and McIntyre (2000)

In addition, the type of cropping system used also affects biomass production. According

to Barthes et al. (2004), more biomass was produced in intercrops than sole and fertilised

Legume Growing season rainfall Soil pH

Biomass t ha-1

N fixed kg ha-1

Author

Centrosema macrocarpum

Bi-modal during April to m i d-August and from August to October

5.1 6.6 70 M u r h e t a l . (2002)

Stylosanthes guianensis

Bi-modal during April to m i d-August and from August to October

5.1 11 116 M u r h e t a l . (2002)

Mucuna pruriens

Bi-modal- March to June and August to December

4.9 6.3 155-230

Wortmann and McIntyre, (2000)

Canavalia

ensiformis

Bimodal- March to June a n d A u g u s t t o December

4.9 9.8 133 Wortmann and McIntyre, (2000)

Crotolaria Bimodal- March to June a n d A u g u s t t o December


Glycine max Bimodal- March to June a n d A u g u s t t o December


Lablab

purpureus

Bimodal- March to June a n d A u g u s t t o December

4.9 4.7 83-140 Wortmann and McIntyre, (2000)

8

crops. Although the data demonstrate the amount of biomass that legumes can provide, the

results refer to crops grown in bimodal rainfall areas. Smallholder farmers in Limpopo

grow bambara groundnut, cowpea, peanuts and sugar beans only in summer. Sometimes

the crops fail because of the variable climatic conditions. Therefore, legumes that produce

more biomass under variable climatic conditions need to be selected and introduced in

smallholder cropping systems.

2.5 Factors affecting legume growth and nitrogen fixation

Legume growth is affected by a variety of factors (e.g. temperature, soil pH and moisture),

which in turn affects nitrogen fixation. Environmental and soil factors that enhance legume

growth, also affect nitrogen fixation.

2.5.1 Temperature

Legume crops require optimum temperature, between 15 and 25°C, for optimum growth

and nitrogen fixation as high temperatures between 30 and 40°C inhibit root growth and

ultimately reduce nitrogen fixation (Sarrantonio, 1991). Low temperatures delay root hair

infection and reduce nodulation. Survival of bacteria in soils at high temperature appears to

be improved by the presence of clay particles and soil organic matter. High temperatures

can inhibit nodulation and if nodulation occurs, can inhibit the activity of N2 fixation on

legumes. Cool temperatures lead to delayed development of plants, including the delay in

the formation of nodules and so decreased rates of N2 fixation.

2.5.2 Soil N

Nitrogen fixation in legumes depends on environmental factors such as soil moisture and

temperature. High soil nitrate supply leads to legumes deriving N from the soil rather than

biological nitrogen fixation. Thonnissen et al. (2000) observed low NO3 in plots planted

with legumes and this is explained by the effectiveness of legumes to assimilate NO3

derived from soil organic matter mineralisation. However, Tian et al. (2000) reported

accumulation of N by legume crops planted in soils lacking nitrogen in south-western

Nigeria. Generally, legumes are able to fix more nitrogen in poor soils than in fertile soils.

9

2.5.3 Soil pH

Most leguminous crops require neutral to slightly acidic soil pH for growth (Giller and

Wilson, 1991) although nodulation may decrease at more acidic soil pH. Soil acidity

affects survival and growth of rhizobia and nitrogen fixation by the symbiotic association

with the legume. At very low soil pH, calcium content decreases and aluminium

concentration increases in the soil, inhibiting root growth and ultimately reducing

nodulation (Giller and Wilson, 1991). This was exemplified in studies conducted by

Becker and Johnson (1998) where the amount of nitrogen fixed by different legumes

increased when the soil pH was between 4.8 and 6.2 and reduced drastically when the soil

pH reached 3.9.

2.5.4 Moisture

Legumes are intolerant to shortage and excess of water. In dry soil, abnormal root hairs

occur resulting in restricted infection and poor nitrogen fixation, and when the soil wets,

new root hairs develop and nitrogen fixation increases. In studies conducted by Jensen

(1987), increments in the amount of nitrogen fixed were observed when rain was received

and decreased in the absence of rainfall. Legumes grow in a variety of climatic conditions

and soil types. The variation in climatic conditions leads to variation in legume

performance. For example, chickpea and lentil showed lower shoot nitrogen content in

studies conducted by Saxena (1986) during seasons of snowfall than in seasons without

snow. In Limpopo province, rainfall could be an important factor in nitrogen fixation.

The number of rhizobia in soil declines drastically as soil dries. Grain legumes with deep

rooting systems, such as cowpea, are grown in climates with limited rainfall as they can

withstand periods of drought, as long as roots manage to penetrate sufficiently deep into

the soil before the drought begins (Giller, 2001). Survival of rhizobia during long periods

of flooding is also of particular importance in cropping systems.

2.5.5 Soil rhizobia

Legume crops build a symbiotic relationship with bacteria. Most legumes have a specific

rhizobia strain that maximises N2 fixation. The rhizobia population in the soil play a role in

N2 fixation as different rhizobium bacteria infect specific legumes for nodulation. Studies

conducted by Tian et al. (1992), where soybean was inoculated with different strains of

10

rhizobial bacteria, resulted in best performance in seed inoculated with a local strain rather

than in new strain. Thus appropriate rhizobia strains are important for inoculation of

legume seeds. Therefore, to ensure that an effective rhizobia strain is present when

planting a legumes species, the seed grown by smallholder farmers need to be inoculated.

2.6 Effects of mulch on growth parameters of subsequent

crops.

Mulching affects subsequent crop performance by manipulating soil conditions including

moisture, temperature and weed control.

2.6.1 Soil moisture

Mulch reduces soil water evaporation and increases soil moisture. The reduction of soil

water evaporation in mulched soil is supported by studies conducted by Tian et al. (1993)

in field plots in Nigeria where soil moisture in mulched plots was higher than in un-

mulched plots. The same scenario was observed by Lal (1978); Maurya and Lal (1981) in

Nigeria and Ramakrishna et al. (2006) in Vietnam, who reported increased soil moisture in

the straw mulched than un-mulched plots. The farming systems utilised by smallholder

farmers in Limpopo do not incorporate mulching, the soil is left bare for crop residues to

be consumed by livestock. Therefore it is important to make farmers aware of the

importance of mulch in conserving soil water in the cropping system.

2.6.2 Soil temperature

Mulch in the cropping systems affects soil temperature in different ways. During hot

weather mulch will reduce soil temperature and increase the temperature during cold

weather. For example, Lal (1978); Maurya and Lal (1981) and Tian et al. (1993) observed

lower soil temperatures in mulched plots than in un-mulched. However, higher soil

temperatures were reported in mulched plots in studies conducted by Ramakrishna et al.

(2006) ranging from 37.7˚C to 25˚C in straw mulched plots and from 34˚C to 21˚C in un-

mulched treatments during autumn and spring cropping seasons. Generally, presence of

mulch tends to reduce soil maximum temperature but slightly raise the minimum

temperature. The difference in soil temperature is caused by the type of mulching material

11

that is used. In many studies, non legume mulch was found to affect soil temperature more

than legume mulch (Tian et al. 1993).

2.6.3 Weed control

Mulch can play an important role in the control of weeds in cropping systems. For

example, less weed infestations were reported in wheat straw mulched plots than in un-

mulched plots in Vietnam by Ramakrishna et al. (2006). The benefits of suppressing weeds

by mulch were reported by Caamal-Maldonado et al. (2001) using jackbean and velvet

bean mulch. As the amount of straw mulch was applied uniformly, the effectiveness of

different amounts of straw mulch was not measured. Smallholder farmers in Limpopo

control weeds by hand hoeing, which is time consuming and labour intensive. The

introduction of mulch to control weeds could benefit these farmers. However it is

important to investigate the amount of mulch required to control weeds.

2.6.4 Soil fertility

Mulching increases soil organic matter which provides nutrients to the soil when

decomposing. For example, Costa et al. (1990) reported increased soil inorganic nitrogen

in mulched pots compared to un-mulched pots when conducting experiments in a

greenhouse in Brazil. In the study conducted by Wortmann and McIntyre (2000), increases

in soil nitrates were observed when canavalia, mucuna, cowpea and crotalaria were grown

in field experiments. Soil nitrogen increase was also reported by Sainju and Singh (2001)

in legume incorporation in maize cropping systems. The studies conducted by Tian et al.

(1993) in Nigeria showed higher soil N in plots which had legume mulch applied than in

plots using cereal mulch or the control plots. The ability of mulch to improve soil fertility

through the addition of organic matter depends mostly on its decomposability.

2.6.5 Crop yield increase

Under water-limited and warm environments, application of mulch usually increases crop

yield due to reduced temperature and increased soil moisture retention (Lal, 1978; Maurya

and Lal, 1981). The increase in yields is supported by studies conducted by Ramakrishna et

al. (2006) where a higher groundnut yield was observed in mulched than in un-mulched

plots. In the same study, polyethylene mulch increased groundnut yield by 94% over the

12

un-mulched plots, 46.8% more than in chemical mulch and 25% more than in straw mulch

due to reduced moisture evaporation and weed control.

In addition to reduced moisture mulch improve crop yield through nutrient supply. Tian et

al. (1993) reported higher maize yield in legume mulched plots than in the control plots in

Nigeria. This scenario was also observed by Hauser and Nolte (2002) in Cameroon and by

Tian et al. (2000) in Nigeria where maize yield increased following legume fallows in

comparison to the control plots. This increase in crop yield following the application of

legume mulch could be due to N supply from the mulch (Armstrong et al., 1997;

Mpangane et al., 2004; Thonnissen et al., 2000).

2.7 Decomposition of mulches and green manure

Decomposition is the breaking down of organic matter. The release of N from

decomposing organic material, or mineralisation of N, results from the activity of micro-

organisms in breaking down the material. Legume crop material containing a small

proportion of N relative to the dry weight (C:N ratio) has a limited amount of N available

for growth of the micro-organisms and any mineralised N will be utilised immediately by

the micro-organisms. Decomposition is dependent on the enzymatic cleavage of chemical

bonds within the plant material as soluble low molecular weight substances such as

glucose or amino acids are rapidly attacked by micro-organisms (Giller, 2001). The

breakdown of more complex substances takes longer because insoluble polymeric

materials tend to be cleaved primarily by slow-growing micro-organisms (Giller, 2001).

The release of N into the soil (net mineralisation) for use by plants is thus a balance of the

process of mineralisation and immobilisation.

The breaking down of organic matter depends on several factors. Some of the factors that

affect decomposition of mulch are C:N ratio, temperature, time of the year (season) and

method of application (incorporated in the soil or left on the soil surface).

2.8 Factors affecting decomposition of mulches

2.8.1 C:N ratio

The C:N ratio is the amount of carbon (in grams) in relation to the amount of nitrogen (in

grams) in the plant dry matter. Crops differ in quality in terms of C:N ratio, leaf structure,

13

secondary metabolites, polyphenols and tannins. Plant residues with high C:N ratio

decompose slowly. In one study, Giller (2001) noted that crops with a C:N ratio >20:1

have initial net immobilisation of N, whereas residues with a smaller C:N ratio decompose

rapidly with net mineralisation of N occurring right from the beginning. Legume residues

have low C:N ratio and tend to decompose more rapidly than non-legume crops. This was

exemplified by Grunwald and van Bruggen (2000) who found that vetch (Vicia dasycarpa)

and oats (Avena sativa) had a C:N ratio of 13.3 and 33.6, respectively, and according to

(Giller, 2001), legumes commonly have a C:N ratio of less than 20:1. Therefore, the C:N

ratio provides an indication of how rapidly a plant material is likely to be decomposed. In

addition to the C:N ratio, the decomposition of crop dry matter during the growing season

depends upon temperature and rainfall.

The differences in C:N ratios between legumes and non-legumes were explored in studies

conducted by Grunwald and van Bruggen (2000); Koenig and Cochran, (1994) and Tian et

al. (1992) as shown in Table 3 below where barley and oats show high C:N ratios as

compared to legume crops.

Table 2.3. Average C:N ratio in different crops.

Crop C:N ratio Author

Alfalfa 19 (Koenig and Cochran, 1994))

Faba bean 18 (Koenig and Cochran, 1994)

Barley 39 (Koenig and Cochran, 1994))

Rape 7 (Koenig and Cochran, 1994)

Vetch 13 (Grunwald and van Bruggen, 2000)

Oats 33 (Grunwald and van Bruggen, 2000)

Mucuna pruriens 7 (Tian et al. 1992)

Centrosema pubescens 8 (Tian et al. 1992)

The C:N ratio is determined by the lignin content in crops; if lignin content is high, the

C:N ratio increases resulting in poor decomposition and low nutrient release. Therefore,

crops with low C:N ratio such as legumes are important in providing N to cereal crops.

14

2.8.2 Lignin content

Both the physical structure and chemical composition of plant residues determines whether

or not the plant residue is resistant to decomposition. The dynamics of green plant material

decomposition is explained by Giller (2001). Green legumes materials contain little lignin,

which is laid down in plants as a structural component in secondary thickening of cell

walls. Thus green manure decomposes more rapidly than grain legume stover or woody

tissues. Decomposition of shoots in forage legumes or prunings of legume trees is rapid as

40% or more of the N in legume shoot material can be released in less than two weeks after

addition to the soil (Grunwald and van Bruggen, 2000). As leaves age, the N content

decreases and lignin content increases, so that older tissues decompose more slowly

(Giller, 2001). Older plant residues tend to be physically harder and therefore less readily

attacked by the soil fauna, which play an important role in decomposition or breaking up

the residues into smaller fragments with a greater surface area for microbial attack.

The age of crops and the different parts in the crop affect the amount of lignin content in

the crop. The differences in lignin content of different crops were demonstrated by Giller

(2001); Koenig and Cochran (1994) and Tian et al. (1992). In the table below, Mucuna

pruriens shows high lignin content whereas barley has the lowest lignin content.

Table 2.4. Lignin percentages in different crops

Crop Lignin % Author

Alfalfa 11.1 Koenig and Cochran (1994)

Faba bean 8.1 Koenig and Cochran (1994)

Barley 7. 8 Koenig and Cochran (1994)

Rape 5.6 Koenig and Cochran (1994)

Mucuna 16.8 Tian et al. (1992)

Centrosema pubescens 10.1 Tian et al. (1992)

2.8.3 Climate

Temperature and rainfall (moisture) determine the amount of dry matter produced which in

turn will be decomposed. Decomposition is more rapid under warm temperatures than cold

temperatures. This is indicated in Koenig and Cochran (1994) studies where less dry matter

was lost in alfalfa, faba bean, barley and rape through decomposition when the temperature

was low (below 0˚C) and with zero precipitation; yet the loss increased with increasing

15

temperature and rainfall. Therefore, temperature and rainfall are important when

considering the rate of dry matter decomposition in the field.

2.9 Effects of legumes on subsequent crops

The benefits of legumes to subsequent crops depend on the integration of several factors.

The benefits of legume green manure to subsequent crops depend on soil fertility, as

according to Giller (2001), the benefits of green manure are likely to be less on fertile soils.

The reduced benefits of green manure on fertile soil is supported by findings of Grunwald

and van Bruggen (2000) where decomposition of organic matter was higher in the

conventional system than in the organic system.

2.9.1 Subsequent crop yield

Legumes increase grain yield and dry matter of cereal crops if included in the cropping

system such as in intercrops or in rotation. Increases in maize yield were reported by

Maluleke et al. (2004) when maize was intercropped in relay with lablab. However, maize

grain yield showed a general decrease in other studies (Caamal-Maldonado et al., 2001;

Maluleke et al., (2004); Papastylianou, (1986); Houndekon et al., 1998) when maize was

intercropped simultaneously with lablab, jack bean and velvet bean.

Legumes improve dry matter accumulation in cereal crops. Higher dry matter levels were

demonstrated in studies conducted by Ayisi and Mpangane (2004) and Carsky et al.

(2001), and when maize was planted simultaneously with cowpea and lablab than when

planted following mucuna, crotolaria and native fallow. However, maize yield increased

when mucuna was planted as green manure rather than when removed for hay in studies

conducted by Jiri et al. (2004). The amount of leaves that fall during the growth of long

duration legumes has an impact on growth of cereal crops. According to Giller (2001), the

residual effects of harvested legumes come from leaves lost during the growth of legumes

or from decomposition of roots and legumes.

Therefore it is argued that including legumes in cereal cropping systems has both positive

and negative impacts on cereal grain yield and dry matter accumulation. However, the type

of legumes used also contributes to the cereal performance.

16

2.9.2 Weed control by legumes

Legumes control weeds in smallholders’ fields by means of suppression. The success of

controlling weeds was demonstrated by Becker and Johnson (1998) and Caamal-

Maldonado et al. (2001) where mucuna, jumbibean and wild-tamarind reduced the density

of spear grass (weed) and suppressed natural weed growth. Despite the reduction of weeds

by legumes, yield of the intended crop is also reduced when grown as an intercrop. The

smothering of young maize grain yield was reported in studies conducted by Houndekon,

et al. (1998). Introduction of legumes by small holder farmers in Limpopo could reduce the

need for hand-hoeing for weed control.

2.9.3 Soil cover

Legumes are utilised in the cropping systems as food, fodder, cover, green manure, mulch

or pasture. Legumes provide soil cover (Houndekon et al., 1998; Giller, 2001) as they

rapidly cover the field producing a significant amount of aerial canopy and adding organic

matter to the soil. A legume such as mucuna showed more rapid growth and produced

greater cover for the soil than lablab and cowpea in studies conducted in the Limpopo

province by Odhiambo (2004)). However, legume crops have to compete for space with

staple foods such as maize which tends to restrict the farm area under legumes. There is a

need to investigate methods that could be used to increase area under legumes and to

demonstrate benefits of leaving some legume biomass in the field to improve soil fertility.

2.10 Effects of mineral nitrogen fertiliser on crops

Increases in the amount of fertiliser applied to crops often increases crop yields. This was

observed in studies conducted by Tian et al. (1992) in Vietnam where shoot dry matter,

shoot N and the amount of N fixed and grain yield increased with the increase in fertiliser

N. The same scenario was also reported by Hauser and Nolte (2002) in Cameroon where

grain yield of maize was higher with the increase in inorganic fertiliser than when maize

was not fertilised. Whitbread and Ayisi (2004) reported increased maize yield with the

application of N fertiliser in Limpopo. Recently, simulation modelling tools have been

developed to use in decision making for improving soil fertility and crop productivity

(Godwin et al., 1984; Hayman et al., 2008; Keating et al., 2003; McCown et al., 1996;

Nelson et al., 1998; Probert et al., 1998b; Probert et al. 1998a; Whitbread and Glem,

2004).

17

2.11 Simulation modelling for agricultural research

Smallholder farmers depend on rainfall for crop productivity; however, rainfall is highly

variable and unequally distributed in regions such that by the time farmers place the seed

in the soil it would be too late in the season and crops may not reach maturity and could

suffer moisture stress due to limited rainfall at the time they mature. Studies have been

conducted to determine management practices for improved crop productivity; however,

the response to such practices are area and time specific i.e. practices applied in one area

during a particular season do not yield the same results in another area in the same season

or in the same area in another season (Whitbread and Ayisi, 2004). To address these

constraints, APSIM (Agricultural Production System sIMulator) model has been employed

to simulate crop performance under different management practices (Andren et al., 1992;

Godwin et al., 1984). APSIM is a model designed to address long term resource

management issues. It resulted from a need for tools that provide accurate information on

crop production in relation to climate, genotype, soil and management. APSIM is a

modelling framework that is used in the simulation of discrete management units within

production systems. APSIM was developed to simulate biophysical processes in farming

systems, particularly as they relate to the economic and ecological outcomes of

management practices in the face of climate risk.

In studies conducted in Australia by Whitbread and Glem (2004) to investigate the

potential production of grain sorghum across a range of seasons using APSIM, the biomass

growth and grain production was simulated with high degree of precision. However, when

Whitbread et al. (2004) employed APSIM in Zimbabwe precise simulation was observed

for maize biomass but not for grain as maize cobs were stolen while still green and this

scenario was not accounted for by the model. In addition, this model was able to determine

possible management practices for alleviating poverty by farmers as scenarios for poor and

rich farmers were precisely simulated. Another successful study conducted using APSIM

was done by Whitbread and Ayisi (2004) in Limpopo province where APSIM simulated

maize biomass growth and grain production with high degree of precision.

Simulation modelling as a research tool has gained increased support in recent years.

Some of the advantages of simulation modelling include:

Ability to place component research (in this case decomposition of legume residues

and supply of N to maize) in the wider farming systems context.

18

Ability to represent climate variability and analyse the likely impact of variable

climate on experimental results.

Linking of science from different domains (crop physiology, soil science, climate

science).

Provides a methodology for taking experimental results and extrapolating these

results to other locations (soil types), other years (longer historical record), and a

wider range of management treatments (eg. different rates and timing of N fertiliser

applications, different legume residue application methods, etc.).

Creates the possibility of incorporating new research results (eg. this research) into

the model science and in this way makes research available to a wider audience of

scientists.

Communication of research results to a diverse range of audiences, for example,

what are the management implications of the current project for small holder

farmers?

2.12 Conclusion

The reduction of crop yields in smallholder farmers’ fields has been the trigger for research

studies with a focus on improving soil fertility. Studies conducted thus far have

incorporated legumes in the farming systems based on intended use by researchers (e.g.

mulch); however, smallholder farmers use legumes for home consumption and livestock

feed which would reduce the benefits in terms of soil fertility from the legume. The effects

of legumes in improving soil fertility and subsequent crops yields tend to differ according

to differences in legume qualities for effective decomposition and nutrient release. As the

effects of legumes on crop yield and soil fertility differ from year to year and from region

to region due to climate, it becomes difficult for research to specify legume practices for

the particular area and year. Researchers are faced with an ongoing challenge to develop

mechanisms for better incorporation of legumes into smallholder farming systems.

In addition to incorporating legumes in the cropping systems, research has developed

simulation modelling tools to predict results of practices in the field and highlights possible

decisions to be taken in management. The success of these tools requires relevant long-

term data to validate the models. Therefore, research is urgently needed to identify

legumes that quickly provide more biomass and fix significant amounts of nitrogen in poor

19

soils of Limpopo. Development of such a technology has the potential to increase food

security and financial well being of smallholder farmers in this region.

21

CHAPTER 3

Cereal and legume management by subsistence farmers in Limpopo province of South Africa: Socio-economic

and farming details.

3.1 Introduction

The major challenge facing the smallholder farming communities is a low crop yield which

is influenced by several factors (Ndove et al., 2004). Farming under the smallholder system

is characterised by low level of production technology and small size of farm holdings of

approximately 1.5 hectare per farmer, with production primarily for subsistence and little

marketable surplus (Chigariro, 2004; Department of Agriculture Limpopo, 2007; Kashem

and Jones, 1988). Smallholders commonly intercrop maize (occupying a larger portion of

land) with other crops, including legumes for food security; however, food self-sufficiency

is generally not reached (Snapp et al., 2002). The integration of legumes in cereal cropping

systems has shown to improve yields (Maasdorp et al., 2004; Maluleke et al., 2004; Ndove

et al., 2004). Such integrations involve legume intercrops, rotations, fallows and green

manures (Misiko et al., 2007; Snapp et al., 2002).

The integration of legumes in cereal cropping systems addresses common problems

encountered by smallholders such as weed and pest infestation, moisture loss, lack of

labour and nutrient loss. The use of mulch to reduce the loss of soil moisture in rain-fed

areas has proved to be successful; however, this is possible in bimodal rainfall patterns

because in this type of rainfall smallholders are able to collect plant residues during the

long rainfall seasons for use during the short rainfall seasons and the other way round

(Misiko et al., 2007; Ramakrishna et al., 2006; Tian et al., 1993).

Low crop yields in the smallholder farming systems led to focus of research on mineral

fertilizers specifically for N supply as many soils are N limited; however, increments of N

fertilizers resulted in reduced crop yields (Fofana et al., 2004). In addition to reduced crop

yields, fertilisers were expensive and unaffordable for farmers (Kashem and Jones, 1988;

Snapp et al., 2002). Such results led to focus on using organic inputs for N supply (Fofana

et al., 2004). The use of organic materials such as green manure, crop residues, compost

and animal manure have been shown to improve soil fertility and crop yield (Fofana et al.,

2004; Maluleke et al., 2004). The success of applying these materials is influenced by their

22

potential use as food, fodder or cash crops (Chigariro, 2004; Misiko et al., 2007; Ndove et

al., 2004). Several studies were conducted to create awareness of the use of organic

material technologies through on-farm trials and experimentation; however, adoption of

technologies was not adopted by all farmers (Fischler and Wortmann, 1999; Fofana et al.,

2004; Misiko et al., 2007; Ndove et al., 2004).

The farmers who were reluctant to adopt the technologies mentioned major constraints

related to lack of access to supplies at times of need such as money, market and

mechanical inputs (Fischler and Wortmann, 1999). A similar trend was observed in

Bangladesh by Kashem and Jones (1988) in the adoption of hybrid seed. In studies

conducted by Snapp et al. (2002) in Malawi farmers who did not adopt the technology

opted to fallow their fields mentioning several reasons including lack of seed, labour and

fertilisers. However, farmers who adopted technology initiated experimentations which

involved their own traditional practices (Fischler and Wortmann, 1999). Although the use

of organic materials have shown positive benefits through experimentation, adoption of the

practices by farmers is still low (Maasdorp et al., 2004; Sofranko et al., 1988) .

Investigations into farmers family backgrounds and living conditions can enable

researchers to assist the smallholders in planning farming activities. Farmers participation

in on-farm trials increases their understanding and interests in skills application of soil

fertility improvement practices and crop yield increasing technologies. Even though

several agricultural technologies were introduced, adoption by smallholders in Limpopo is

still limited.

The objective of the current study was to determine socio-economic and farming

characteristics of farmers in Limpopo province.

3.2 Materials and methods

3.2.1 Site description

a Location and biophysical environment

The study was conducted in two villages GaKgoroshi/Sechaba and Gabaza of Limpopo

province, South Africa. GaKgoroshi (-23.721120°, 29.135534°) is situated 50km west of

Polokwane city and falls under Aganang municipality of the Capricorn district. Gabaza (-

23

23.991964°, 30.334951°) is situated 140km south east of Polokwane city and 40km from

Tzaneen town, and falls under Greater Tzaneen municipality of the Mopani district. The

annual summer rainfall received in the two villages ranged between 400 to 600 and 600 to

800mm in GaKgoroshi and Gabaza, respectively (Figure 3.1).

.GabazaGaKgoroshi.

Figure 3.1: Rainfall map of the Limpopo province of South Africa

The soils in the two villages are characterised by low N and P, with acidic pH less than 5.5

(Table 3.1).

Table 3.1. Soil chemical analysis for the soil profiles

Depth pH N P K Ca Mg Na Cl Zn

(cm) Mg kg-1

Gabaza 0-15 5.4 4.3 1.7 47.0 674.3 303.7 7.0 4.7 1.0

15-30 5.3 1.7 1.3 30.0 715.0 290.0 6.0 3.0 0.9

GaKgoroshi 0-15 5.3 3.3 12.7 98.7 267.7 78.3 2.0 1.0 0.8

15-30 5.2 3.3 11.0 96.0 336.0 89.3 1.67 1.0 0.6

24

Table 3.2. Soil particle analysis

Depth Sand Silt Clay

(cm) %

Gabaza 0-15 63 11 26

15.30 63 10 27

30-60 59 12 30

GaKgoroshi 0-15 87 3 10

15.30 85 4 11

30-60 78 6 16

b Sampling method

Sixty farmers from the two villages (thirty farmers per village) were interviewed using

open-ended and close-ended questionnaires (Appendix 1). The interviews were conducted

during April/May 2008. The farmers were invited for a meeting in their respective meeting

areas by the local extension officer and they were interviewed individually. The invitation

was specifically directed to farmers growing field crops in their fields. However, some

farmers who grew crops in their backyards also came for the interviews but their data was

not included in the analysis. The purpose of the meeting was to gather information

regarding their socio-economic and farming details and this was explained to them before

the start of interviews.

3.2 Results and discussion

3.2.1 Socio- economic details

Among socio-economic issues, information was collected on the proportion of gender in

farming communities, age and educational qualification, length of farming experience,

family size, income sources and constraints associated with raising household income.

a Proportion of gender in farming communities

The subsistence farming community of the Limpopo province consists of more women

than men. This study shows that 38.3% of the sixty farmers interviewed were males

whereas 61.7 % were females. Similar results of more female than male farmers were

observed by Ndove et al. (2004) in Dan village which is about five kilometres from

Gabaza. The information on higher number of females than men participating in farming

25

activities was also supported by the report from the Department of Agriculture Limpopo

(2007) which mentioned that women constituted 80% of the smallholder farmers in the

province. The reason for greater female participation was related to the males working in

the cities, looking after livestock and females being widows or unmarried.

b Age and educational qualification

Nearly 50% of the farmers in the two villages were aged above sixty-eight (68) years and

did not attend school or attended up to primary level. Table 3.3 shows that there were no

farmers below thirty-eight (38) years of age in the two villages. In GaKgoroshi, 50% of the

farmers were above 68 years and 6.7% were in the age group range between 38 to 47

years, and in Gabaza, 46.7% of the farmers were aged above 68 years and 13.3% were

between 38 to 47 years of age. This study shows that about 80% of the farmers were

elderly (above 50 years of age). Ndove et al. (2004) reported 63 years as the average age of

respondents with 45% of respondents in the 61-70 age range in Dan. However, in studies

conducted by Pandey and Van Minh (1998) in northern Vietnam only 12% of the farmers

were in the elderly category. The same scenario was observed in the study conducted by

Kashem and Jones (1988) in Bangladesh where 60% of the farmers were under 40 years. In

addition to old age, low education level is also a characteristic of the smallholder farmers

in Limpopo.

Smallholder farmers were mostly illiterate or had attended school up to primary level;

however, there were some differences between the 2 villages. Table 3.3 shows that in

Gakgoroshi 13.3% of farmers had no schooling whereas in Gabaza 50% did not go to

school. In GaKgoroshi 67% have attended up to primary level and 20% went until

secondary whereas in Gabaza 43.6% attended up to primary level with 6.7% up to

secondary. The low level of education of smallholder farmers was also reported by Ndove

(2004) in which 64% of the farmers who were interviewed did not attend school at all,

14% studied up to primary level and 22% attended adult literacy school. The low literacy

situation of farmers was also reported in Bangladesh by Kashem and Jones (1988) where

two-thirds of farmers were reported as not having received any formal education. The old

age and low level of education of the smallholder farmers may reduce the rate of adoption

of agricultural technology.

26

Table 3.3. Smallholder farmers’ ages and their level of education (%)

Range of farmers ages in years Level of education

38-47 48-57 58-67 Above 68 No School Primary Secondary

GaKgoroshi 6.7 13.3 30.0 50.0 13.3 67.7 20.0

Gabaza 13.3 10.0 30.0 46.7 50.0 43.3 6.7

c Family size

Family size in this study comprises individuals living in one household. Farmers in

Limpopo have different family sizes, ranging from less than four to more than twelve

family members. The households have typical family structures consisting of two to three

generations. In the two generations households, members consisted of parents and children

whereas in three generations households, members include grandparents. The adults

comprised people aged 18years and older, and people below 18 years were regarded as

children. During the interviews it was realized that in most households, older children have

already left because they had their own families. Figure 3.2 shows that about 56.7% of

farmers in Gabaza had 4 to 6 family members whereas in GaKgoroshi 43.3% had 7 to 9

members with some farmers (13.3%) still having 10 to 12 members. In both the two

villages, a small number of farmers (6.7%) had less than four members. In studies

conducted in Zimbabwe, Maasdorp et al. (2004) mentioned average household sizes of 5-6

members. Pandey and Van Minh (1998) reported an average household size of 7.2

members in Cao Bang province of northern Vietnam. Kashem and Jones (1988) reported

family sizes of ten members in Bangladesh. Large family sizes provide an indication for

more labour per household; however, in these two villages, large families comprised

mostly of children who spent most of their times in schools, resulting in limited availability

of labour.

27

Family size

0.0

20.0

40.0

60.0

80.0

100.0

<4people 4-6people 7-9people 10-12people

Num

ber o

f far

mer

s (%

)

GakgoroshiGabaza

Figure 3.2. Family size grouped in the number of people per household.

d Income sources for household

Household income sources in this study include wages (non- agricultural income),

remittances, income in kind, pension (government grants for elderly people- R1,200),

children’s grants (from government- R200), money from crops and livestock sales, and

unemployment incident fund (for those who worked for institutions providing it). This

study shows that farmers received income from different sources (Figure 3.3). Most of the

household incomes in these two villages come from old age pension, remittances as well as

children grants (Figure 3.3). In GaKgoroshi it was 73.3, 66.7 and 50% whereas in Gabaza

90, 43.3 and 50% came from the 3 sources, respectively. Ndove et al. (2004) also

mentioned government social grants as the only income source of smallholder farmers’

income. In studies conducted by Misiko et al. (2007) in Kenya, 10% of farmers received

off farm income in the form of pension, salary, remittance and business income, whereas

90% received it as wage labour, few children being herds boys, maids, and manual labour

in urban centres. This is different from the studies conducted by Pandey and Van Minh

(1998) in Vietnam where most income was from labour, selling animals and forest

products.

There were some farmers receiving income from wages (3.3%) and in kind (6.7%) in

Gabaza (Figure 3.3). In GaKgoroshi, 10% of farmers received income from unemployment

28

incident fund (UIF). On average, only 20% of household income was derived from

agricultural production (both crops and livestock). In the two villages, about 60% of the

households had family members who were working and earning salaries (remittances).

Ndove et al. (2004) mentioned that produce from farmers’ fields in Dan was used for home

consumption and very little was traded. In studies conducted by Misiko et al. (2007), 55%

of farmers received on-farm income from cash crops and limited milk production and 35%

relied on selling napier grass, stover and firewood. Pandey and Van Minh (1998) reported

that 33% of the income of Vietnamese farmers was sourced from selling livestock and only

5% of income was from salary. The ability of farmers to earn on-farm income is dependent

on local rainfall. The two villages in Limpopo province received a variable unimodal

annual rainfall during summer months ranging between 400 and 800mm resulting in one

planting season. This was different from the study conducted by Misiko et al. (2007) in

Kenya where the annual rainfall was bimodal ranging between 1500 and 2000mm and

providing for two planting seasons.

Family income sources

0.0

20.0

40.0

60.0

80.0

100.0

Wag

es

Remitta

nces

Incom

e in k

ind

Pensio

ns

Child g

rants

Crop sa

les

Lives

tock

UIF

% o

f far

mer

s

GakgoroshiGabaza

Figure 3.3. Income sources per household and the number of farmers receiving the

incomes.

e Main constraints associated with increasing household income

Farmers in GaKgoroshi and Gabaza are faced with several socio-economic and bio-

physical constraints associated with increasing household income. The main socio-

economic constraints identified by farmers as associated with lack of increasing income

29

included unemployment, low level of education and old age (Figure 3.4). The other main

constraint identified by farmers was drought which is a bio-physical constraint. Constraints

related to agricultural inputs (e.g. fertiliser, weeds, tractor) were not identified by most of

the farmers. In Gabaza, there were few cases showing the inclusion of orphans and

sicknesses as constraints (3.3%) whereas there were cases of water-logging (3.3%), lack of

fertiliser application (3.3%) and unavailability of tractor at planting (3.3%) in GaKgoroshi.

In contrast, Misiko et al. (2007) mentioned malaria, labour constraints, seasonal hunger

and urgent annual demands to pay school fees as major constraints in increasing income in

Kenya.

Income constraints

0

20

40

60

80

100

Old ag

e

Low le

vel e

d.

Unemplo

ymen

t

Drough

t

Waterlo

gging

Orphan

s

Overpo

pulat

ion

Mixed c

roppin

g

Wee

ds

No fert

. App

l.

No trac

tor

Sickne

ssNon

e

% o

f fam

ers

GaKgoroshi Gabaza

Figure 3.4. Income constraints face by smallholder farmers

3.2.2 Farming enterprise

Information was collected on type of livestock owned, size of arable land owned, crops

planted, land size allocated to maize as compared to legume crops, maize yield in good and

poor rainfall seasons and major production constraints faced by farmers in the fields.

a Livestock

Most smallholder farmers owned livestock that included cattle, goats, sheep and poultry

which were kept mainly for subsistence purposes to produce meat, milk and eggs. Figure

3.5 below shows that poultry is owned more than cattle, goat and sheep by 80 and 70% of

30

smallholder families in GaKgoroshi and Gabaza, respectively; however, the poultry kept

was mostly in small numbers less than ten, similar to other livestock (data not shown).

Cattle and goats were owned by 46.7 and 30% of farmers in GaKgoroshi, and 26.7 and

10% farmers in Gabaza, respectively, also in small numbers (<10) (Figure 3.5). The same

findings were mentioned by Ndove et al., (2004) where individual ownership of livestock

ranged between 1 and 100 head. The farmers who were interviewed in Gabaza mentioned

that they owned no sheep. The farmers who were interviewed in GaKgoroshi did not own

any pig because it was realized during the interviews that they were mostly members of the

Zion Christian Church as they put the identification emblem on their clothes. This church

proscribes its members from the consumption of alcoholic beverages, smoking and eating

pork Byrnes (1996). In the two villages, there were still families that did not own any

livestock, 6.7% in GaKgoroshi and 20% in Gabaza.

0

20

40

60

80

100

Cattle Goat Sheep Pig Poultry None

% o

f far

mer

s ow

ning

live

stoc

k GakgoroshiGabaza

Figure 3.5. Types of livestock owned by farmers.

b Size of arable land owned

The sizes of arable land that was owned by smallholder farmers differed, ranging from 1 to

6 ha with 1.6 as the mean for land holding. Farmers in GaKgoroshi and Gabaza who

owned 1ha of land which was less than the land holding mean, were more than those who

owned 2 ha and more. 43.3 and 66.7% of farmers in GaKgoroshi and Gabaza had 1ha of

land, respectively. In Gabaza, there were those farmers who owned 5 to 6ha of land

(3.3%), whereas in GaKgoroshi there were no farmers who owned 5ha and more. In

studies conducted by Ndove et al. (2004) in Dan village, individual allocations were

between 0.5 and 1ha which were obtained through permission to occupy rights. The size of

31

farms in the two villages was found to be similar to subsistence farm sizes in Malawi

where farm sizes ranged from 1 to 2ha (Snapp et al., 2002). In studies conducted by

Pandey and Van Minh (1998) in Vietnam farm sizes were less than 1ha which were

received from government policy to enhance food security.

c Types of crops planted

Smallholder farmers planted a variety of crops in their fields during the summer season

starting from November and December in Gabaza and GaKgoroshi, respectively. The first

crops were harvested in March and April in Gabaza and GaKgoroshi, respectively. Most of

the smallholder farmers in these 2 villages grew maize (Zea mays) which is the staple food,

whereas cowpea (Vigna unguiculata), common bean (Phaseolus vulgaris) and bambara

(Vigna subterranean) were grown on <10% of the land (Figure 3.6). In GaKgoroshi 43.3%

of farmers planted common bean (Phaseolus vulgaris) whereas in Gabaza farmers did not

grow the bean. Groundnut (Arachis hypogaea) was grown by farmers mostly in Gabaza

(96.7%) than GaKgoroshi (6.7%).

The same crops were also grown in Swaziland other than mungbean, sorghum, sweet

potatoes, cotton and garden vegetables (Sithole and Apedaile, 1987). These legumes were

also found to be commonly grown by farmers in Kenya (Misiko et al., 2007). This is

different from Malawian farmers who grew maize as sole without intercropping (Snapp et

al. 2002). Legume crops grown by farmers in Malawi include pigeonpea (Cajanus cajan),

soybean (Vigna unguiculata), mucuna (Mucuna pruriens) and Tephrosia vogelii (Snapp et

al. 2002). Other legumes found to be grown by Kenyan farmers were peas (Pisum

sativum), soyabean (Glycines max), jack bean (Canavalia ensiformis) and lablab (Lablab

purpureus) (Misiko et al. 2007). The inclusion of legume crops in the fields rested on their

multiple potential uses by farmers.

Maize is commonly harvested and consumed at immature stage as green maize or later

when mature, was milled and cooked into porridge in Limpopo. Cowpea was consumed as

leaves and seeds whereas bambara, groundnut and common bean are consumed as seeds

only. These vegetables are eaten together with maize meal or alone for protein. The

remaining residues of these legumes and maize were normally left on the ground after

harvest. The types of legume crops grown by subsistence farmers differ in terms of area

and uses. Legume crops grown by farmers in Malawi had other uses in addition to home

32

consumption such as sale for cash, fuel wood, soil improvement, weed suppression and

fish killing according to Snapp et al. (2002). Understanding of the different uses of

legumes in smallholder farms by researchers increase the successful adoption of legumes

practices by smallholder farmers.

d Allocation of land to maize compared to legumes crops

Smallholder farmers planted maize on bigger areas as compared to legumes and other

crops. Figure 3.6 shows that the area of land allocated to maize was more than the land

allocated to legumes and other crops. Other crops include watermelon and pumpkin. In

GaKgoroshi and Gabaza, 89.1 and 85.4% of land was allocated to maize, 13 and 12% to

legumes, and 9.2 and 2.4% to other crops, respectively. The same land allocation to crops

was observed by Snapp et al. (2002) in Malawi where maize occupied 50 to 70% of the

cropped area. Land occupied by corn and upland rice in Vietnam was over 90% with

legumes occupying only 0.42% (Pandey and Van Minh, 1998).

0.0

20.0

40.0

60.0

80.0

100.0

Maize Legume crops Other crops

% o

f lan

d al

loca

ted

to c

rops

GaKgoroshiGabaza

Figure 3.6. The % of land allocation to maize, legumes and other crops

e Maize yield in good and poor rainfall seasons

Maize yield differs in the two villages. The difference is observed both during high and

low rainfall years. In periods of high rainfall, smallholder farmers receive more yield than

in period of low rainfall. Maize yield during years of high rainfall ranges between 0.5 and

33

1 ton per ha (t ha-1); however, there are farmers who receive 3 and more t ha-1 of maize. In

years of low rainfall, smallholders in the two villages received less than 1t ha-1; however,

some farmers receive 2 to 2.5t ha-1.

Maize yield in the two villages is generally low resulting in food insecurity and no

marketing opportunities. In studies conducted in Swaziland by Sithole and Apedaile

(1987) 9.8 bags (80 kg bag-1) of maize yield on average were sold per family per year.

Misiko et al. (2007) reported that 90% of farmers in Kenya experienced food insecurity as

farm harvests lasted for almost three to four months. Most farmers in these two villages

produced less than 1.1t ha-1 of maize which is less than what is needed to reach the food

self-sufficiency level of 200kg of cereal grain per capita as determined by Food and

agriculture organisation (2004), as most of the households constituted 5 to 8 members on

average. This was supported by Snapp et al. (2002) who mentioned that self-sufficiency in

cereal grain production requires an average harvest of 200kg per adult.

f Major constraints associated with increasing maize yields by farmers

Farmers were asked to mention the constraints they were facing in relation to increasing

the maize yields in their fields. Smallholder farmers mentioned several challenges

associated with increasing maize yield in their fields. Fig. 3.7 shows termites being the

most serious constraint in these two villages as mentioned by 83.3 and 96.7% of farmers in

GaKgoroshi and Gabaza, respectively. In GaKgoroshi, farmers also face challenges of

drought and wild birds, whereas in Gabaza, farmers face challenges of weeds (especially,

Striga hermonthica)) and drought. Farmers in Gabaza mentioned the unavailability of

tractors during the planting season as a major constraint. This was supported by Ndove et

al. 2004) where unavailability of tractor was regarded as a major constraint resulting in

late planting by farmers. This was different from the constraints faced by farmers in

Malawi where they mentioned livestock damage to crops instead of wild animals (Snapp et

al., 2002). In Zambia, another different scenario was observed by Francis and Rawlins-

Branan (1987) where weeds and pests were considered minor problems whereas financial,

labour and input availability were considered major constraints. In studies conducted by

Adesina et al. (2000) farmers noted lack of information on improved crop management as

a constraint.

34

0

20

40

60

80

100

Wild bi

rds

Wild

anim

als

Termite

s

Wee

ds

Drough

t

Swamp

Rootw

orm

Poor s

oilPes

ts

Plantin

g meth

od% o

f far

mer

s fa

cing

con

stra

ints

GaKgoroshiGabaza

Figure 3.7. Major constraints faced by farmers

The problem of pests was mentioned by some farmers, 10% in GaKgoroshi and 43.3% in

Gabaza as a major constraint; however, pests were considered a major problem by farmers

in Zimbabwe (Snapp et al., 2002). The farmers in the two villages did not mention lack of

fertiliser as a constraint despite the poor maize yields from their soils. Lack of fertiliser as

a constraint in increasing crop yield was observed by Langley et al. (1988).

3.2.3 Crop management and timelines

a Cropping method

Smallholders have been farming in their fields for long periods using the traditional

planting method of broadcasting seed and intercropping under rain-fed conditions. The

different methods of cropping are applied differently by smallholders. The types of

cropping methods are random pure stand, row pure stand, mixed random intercrop and row

intercrop. In this study, random pure stand meant when one type of seed was broadcasted

in the field by hand during planting, row pure stand was when one type of seed was placed

in rows during planting, normally through the use of a planter, mixed random intercrop

was when different crop seeds were mixed together and broadcast by hand during planting,

and row intercrop was when seeds of different crops were planted in different rows using a

special planter. In table 3.4 most farmers in GaKgoroshi applied mixed random intercrop

method of planting whereas in Gabaza most farmers applied random pure stand and mixed

35

random intercrop. The same cropping method was mentioned by Ndove et al. (2004) in

Dan where farmers applied seeds by broadcast.

Farmers in GaKgoroshi mixed maize seeds with all other seeds whereas in Gabaza, farmers

did not mix bambara and groundnuts with maize. In Gabaza, twenty-six of the farmers

applied the random pure stand cropping type when planting bambara and groundnuts;

twenty-eight of the farmers intercropped maize with other crops. Only one farmer in

GaKgoroshi planted maize using random pure stand.

Table 3.4. Types of cropping methods applied by farmers

RPS RoPS MRI RoI

GaKgoroshi 3.3 10.0 76.7 16.7

Gabaza 86.7 10.0 93.3 0.0

RPS = Random Pure Stand; RoPS = Row Pure Stand; MRI = Mixed Random; Intercrop;

RoI=Row intercrop

b Cropping timelines

The cropping systems in the two villages are dependent on the timing and amount of

rainfall. Table 3.5 shows the different cropping activities performed by farmers in the two

villages. The cropping activities are represented by alphabets in the table where, P, W, H,

F, G and Po stand for planting, weeding, harvesting, fallowing, grazing and ploughing,

respectively. Smallholder farmers in GaKgoroshi have a longer fallowing period starting

from July and ending in November, than farmers in Gabaza which starts in July and end in

October. The fallowing period in the two villages was different from the fallowing period

in studies conducted by Misiko et al. (2007) in Kenya. The fallow period is influenced by

the availability of rainfall as in most studies where bimodal rainfall is received, farmers

practice crop rotation and utilise the two growing seasons (Misiko et al., 2007).

Table 3.5. Timelines for maize cropping

Months February toApril

May to July

August to October

November to January

GaKgoroshi P + W + H1 H2 + F F + G F + Po + P

Gabaza P + H1 H2 + F F + G + Po P +W

P= planting; W= weeding; H1= harvesting green crops; H2= harvesting mature crops; F=

fallowing; G= grazing; Po= ploughing; + sign = and

36

Soil tillage. Soil preparation is performed at different times. Some farmers till their soil

once at planting whereas others till the soil twice, during ploughing and at planting. The

soil is tilled by either tractor or animal traction during October to December, depending on

the earliest rainfall. Table 3.6 shows that in GaKgoroshi, 90% of the farmers use tractor to

plough their fields, 6.7% use animal traction and 3.3% combines tractor with animal

traction, whereas in Gabaza, all 100% of the farmers use tractor and none use animal

traction. Ndove et al. (2004) mentioned only the use of tractors during planting in Dan. In

the study conducted by Sithole and Apedaile (1987) in Swaziland, farmers used tractors

and animal traction. This was different from the field ploughing in Malawi where farmers

ploughed their fields by hand using hand hoes (Snapp et al. 2002). Land preparation or

tillage using hand hoes among Vietnam farmers was a common practice due to small farm

sizes of less than 1ha (Pandey and Van Minh, 1998).

The number of times that the fields were ploughed also differed. In the two villages, the

number of farmers who ploughed their fields once is more than those who plough their

fields twice. In GaKgoroshi 66.7% of farmers plough the fields once and 33.3% of them

plough it twice and in Gabaza 63.3% plough the fields twice and 36.7% plough them

twice.

Table 3.6. Ploughing equipment used and number of ploughing

GaKgoroshi Gabaza

Tractor 90.0 100

Animal traction 6.7 0

Tractor and animal traction 13.3 0

Ploughing once 66.7 63.3

Ploughing twice 33.3 36.7

Among the farmers who plough their fields twice, some apply manure and others do not

apply anything during ploughing. In GaKgoroshi 40% of the farmers apply manure

whereas farmers in Gabaza do not apply manure. Some of the farmers apply manure

during the second ploughing. However, some of the farmers apply inorganic fertilisers and

manure during planting.

Manure was applied to the fields by 56.7% of farmers in GaKgoroshi and 23.3% in

Gabaza. In GaKgoroshi 70.6% of them and 57.1% in Gabaza applied cattle manure which

37

was the most commonly used manure (Table 3.7). In studies conducted by Sithole and

Apedaile (1987) in Swaziland, 1.7% of farmers did not fertilise their fields. In these two

Limpopo villages, 43% of the farmers did not apply any manure in Gakgoroshi whereas

77% of the farmers did not add manure in Gabaza. The table further shows that the manure

applied in GaKgoroshi was mostly from farmers’ own livestock (88.2%) whereas in

Gabaza some of the manure was accessed through purchase (57.1%). Few farmers from the

two villages accessed manure from neighbours, 11.8 and 14.3% in GaKgoroshi and

Gabaza, respectively.

Table 3.7. The % of farmers using types of manure, source and time application GaKgoroshi Gabaza

Type of manure Cattle 70.6 57.1

Chicken 11.8 42.9

Cattle/chicken 11.8 0.0

Cattle/goat 5.9 0.0

Source Own animal 88.2 28.6

Neighbour 11.8 14.3

Buy 0 57.1

Time of manure application At planting 5.9 0.0

After harvest 58.8 85.7

At ploughing 35.3 14.3

Manure in these 2 villages is applied at different times to the fields, at planting, after

harvest or during ploughing. Table 3.7 shows that 58.8% of farmers in Gakgoroshi and

85.7% in Gabaza apply manure after harvest. The amount of manure applied by

smallholder is generally low, the reason being that they own small number of livestock

which makes it difficult for them to accumulate enough manure. In addition, famers do not

have enough income to buy manure, resulting in infrequent application in their fields,

mostly after 3 to 5 years. The amounts of manure applied by most farmers in GaKgoroshi

and Gabaza are 2t ha-1 and 5t ha-1, respectively. Ndove et al. (2004) indicated that farmers

in Dan applied manure at a rate between 0.5 and 1. 5t ha-1. In studies conducted by Misiko

et al. (2007) in Kenya, only 10% of smallholder farmers had enough and could sell farm

yard manure and 85% had little or nothing at all.

38

Planting. The planting of crops is done during November to early January, with

GaKgoroshi normally starting to plant their crops later than Gabaza because of low and

late rainfall. Farmers in the two villages also applied synthetic fertilisers to their crops

during planting. Around 50% of the farmers in GaKgoroshi and Gabaza apply fertiliser to

their crops. The practice of not applying fertiliser to crops by most of farmers was also

mentioned by Ndove et al. (2004). In studies conducted by Sithole and Apedaile (1987),

80% of farmers applied chemical fertilisers to their crops. About 62.5% and 45.5% of the

farmers, who applied fertiliser to their crops in GaKgoroshi and Gabaza, knew the name of

the fertiliser they applied, respectively, which is associated with farmers’ low level of

education.

Weed control. Weed control is the commonly achieved by farmers using hand hoes after

crops have fully emerged during December and January depending on the time of planting.

The number of times the weeds are removed from the fields differs with households due to

labour availability and ability to hire weeding labour. Figure 3.8 shows that all farmers in

GaKgoroshi weed their fields only once whereas in Gabaza there are farmers who weed

their fields twice (56.7%) and to some even three times (3.3%). The reason for one

weeding activity in GaKgoroshi is because of the poor soil characteristics preventing both

crops and weeds to survive as compared to Gabaza where the soil is fertile and weeds

identified as the second major problem following termites (Fig. 3.9). In addition to fertile

soil in Gabaza, high rainfall also would cause weeds to emerge later in the season.

0

20

40

60

80

100

1 2 3Number of weeding

% o

f far

mer

s in

wee

ding GaKgoroshi

Gabaza

Figure 3.8. Number of weeding applied by smallholder farmers

39

Harvesting. Harvesting (H1) starts in February by cutting consumable green leaves of

cowpea, watermelon and pumpkin, immature pods of bambara, groundnuts and cowpeas

and also immature pumpkin fruits. The first harvest ceases with the reduction in

palatability of green crops. During the second harvest, mature crops are collected from the

field, processed for storage and consumption. This harvest ends during June. The

remaining crop residues are left in the fields and will be consumed by livestock when time

is announced officially by the tribal authority.

Residue management. The farmers were provided with a list of residue management

practices to identify the ones they apply. The types of residues management practices listed

were leave on the field and graze in situ (L/G), remove to feed livestock (R), burn (B), sell

(S), and leave and plough under (L/P). Figure 3.9 shows that most of the farmers in the two

villages, 100% in GaKgoroshi and 93.3% in Gabaza leave/graze in situ. There are no

farmers who remove crop residues to feed livestock, burn or sell. In Gabaza 6.7% of

farmers practiced plough under whereas in GaKgoroshi there are no farmers who ploughed

under residues. This is different from Malawi (Snapp et al., 2002) where 40-70% of

farmers incorporated maize residues and the remaining 40-60% of farmers burnt the

residues. The practice of incorporating and burning in Malawi was also performed in

legume residues as 50-70% of farmers incorporated groundnut residues and 30-50% burnt

the residues (Snapp et al. 2002). Subsistence farmers still require information and training

in crop residue management.

0

20

40

60

80

100

L/G R B S L/P

Types of residue management

% o

f far

mer

s

GaKgoroshiGabaza

40

Figure 3.9. The % of farmers and types of residue management. L/G, R, B, S, and LP

represent leave on the field and graze in situ, remove to feed livestock, burn, sell, and leave

and plough under, respectively.

3.2.4 Potential benefits of nitrogen fixation by legumes and skill application

When explaining the purpose of the meeting before the start of the interviews, the potential

benefits of legumes and their ability to fix atmospheric N were also explained to farmers

and discussions were held for further understanding of the potential benefits of legumes by

farmers. Realising that farmers leave legume residues in the fields after harvest for

livestock (Figure 3.9), they were then asked questions about their knowledge in terms of

whether they have ever heard or never heard about the application of legume residues for

soil fertility and the purpose of planting legumes in their fields. Since the potential benefits

and the process of nitrogen fixation by legumes was explained to farmers, they were

requested to say their interest in the application of this skill. This study looked into the

farmers who have ever heard or never heard about nitrogen fixation, those who applied or

never applied the skill on nitrogen fixation and those who were interested in applying the

skill in future.

Farmers who had heard about nitrogen fixation and improved soil fertility through legumes

were asked if they have ever applied or never applied the skill. Table 3.8 shows that most

farmers (80-87%) had not heard about nitrogen fixation. Amongst the farmers who had

heard about nitrogen fixation, only a few of them (3.3-6.7%) applied the skill but most

were now interested in this topic. Ndove et al. (2004) reported the courage and interest in

participating in skill application among farmers through trial demonstrations and

experimentation. This was motivated by the involvement of extension personnel in the

project. According to Adesina et al. (2000) 59% of farmers in Cameroon who had heard

about the skill of alley farming never applied it but 41% of the farmers adopted and applied

the skill.

The lack of knowledge in the potential benefits on nitrogen fixation by legumes and poor

application of skills by farmers raise concerns about the role and knowledge of extension

personnel in skills and technology transfer as farmers indicated that they contact the

extension office for agricultural advice. In studies conducted by Ndove et al. (2004) skills

41

and technology transfer to farmers by local extension officers and researchers was easy and

beneficial to farmers working in groups. Misiko et al. (2007) observed successful skill

applications on legume integration in farmers fields through field demonstrations,

organised farmers groups, farmer field schools groups and other interested farmers.

Table 3.8. The % of farmers knowing about the potential benefits of N fixation bylegumes, application of skill and their response to N fixation information

Responses GaKgoroshi Gabaza

Heard 13.3 20.0

Never heard 87.7 80.0

Applied 3.3 6.7

Never applied 96.7 93.3

Interested 100 90.0

Not interested 0.0 10.0

3.2.5 Problems associated with legume derived N

Smallholder farmers are faced with several problems associated with legume derived

nitrogen. The farmers were given a list of problems associated with legumes as source of

nitrogen and they were requested to identify the most important problems. The farmers in

the two villages (100%) were faced with challenges of labour and 93.3 and 100% in

GaKgoroshi and Gabaza identified competing demands for legume crops as feed for

livestock also as a major problem (Figure 3.10). The challenge of extra labour

unavailability was supported by Ndove et al. (2004) in Dan in which farmers were mostly

elderly. Farmers in Vietnam did not identify unavailability of labour as a problem because

they use exchange labour practice for land preparation, sowing and harvesting where more

than forty farmers are involved in one piece of land before working on the next field

(Pandey and Van Minh, 1998). The problems of competing demands were also identified

by farmers in Malawi where the introduction of pigeonpea was mostly threatened by the

common practice of open grazing of livestock after maize harvest (Snapp et al., 2002).

About 30% of farmers in Gabaza were still challenged by lack of land to plant maize

whereas in GaKgoroshi farmers did not regard land as a challenge. The problem of land

shortage in Gabaza was similar to that mentioned by Snapp et al. (2002) in Malawi where

farmers are less willing to give up part of their maize land for legume production. Local

seed was found to be popularly used by farmers in Kenya because of lack of funds to

42

purchase certified seeds (Misiko et al., 2007). The situation was similar to the scenario

observed by Adesina et al., (2000) in Cameroon where few farmers (6%) mentioned lack

of seed as a constraint. This was different from studies conducted by Snapp et al. (2002) in

Malawi, where lack of access to seeds was frequently noted by farmers. This was similar

with farmers in Malawi (Snapp et al., 2002) and Zambia (Francis and Rawlins-Branan,

1987) who frequently noted lack of cash to buy seed.

There were no farmers in the two villages who were concerned about the adaptation

because they used their own traditional seeds. The same scenario was found by Ndove et

al. (2004) who reported that farmers, local extension personnel and the research team

observed an increase in the incidence of aphid infestation after the introduction of exotic

legume seed.

0.0

20.0

40.0

60.0

80.0

100.0

Land

for m

aize

Extra l

abou

r

No ada

pted l

egum

es

Compe

ting d

eman

ds

Lack

of fu

nds

Drough

tFarm

ers'

% fa

cing

pro

blem

s

GaKgoroshiGabaza

Figure 3.10. Problems associated with legume derived N

3.2.6 Farmers general source of information and farmers group development

Farmers received cropping information from different sources. Table 3.9 shows that most

of the smallholder farmers 76.7% in the two villages received information from extension

office and 73.3 and 86.7% in GaKgoroshi and Gabaza, respectively depended on other

farmers. Ndove et al. (2004) reported the promoted consultation among farmers, extension

personnel and community leaders in Dan. Farmers in Cameroon obtained information

43

mostly from non-governmental organizations, extension service and other farmers than

from research services (Adesina et al., 2000). The local farmers sources of information

were different from the Zambian farmers who received information from many sources

including non-governmental organisations such as farmer training centres, community

development and old cooperatives in addition to extension office and other farmers

(Francis and Rawlins-Branan, 1987).

In most cases, farmers who had several sources of information were involved in farmers

groups and organisations. Smallholders in GaKgoroshi are members of the Kgorosecha

farmers organisation whereas farmers in Gabaza are not members of any farmers

organisation. Farmers organisation in this study refers to the different farmers groups

coming together to form one inclusive group. The poor application of skills by farmers

from the two villages is still questionable, despite the fact that they received information

from the extension office.

Table 3.9. The farmers source of information and farmers group membership

Source of information GaKgoroshi Gabaza

Extension office 76.7 76.7

NGO’s 0 3.3

Other farmers 86.7 73.3

Organizational membership

Member 76.7 20.0

Non-member 23.3 80.0

Fig. 3.11 shows that 10% of smallholders in GaKgoroshi who are not members of the

farmers’ groups either lack knowledge about organizational development whereas others

(16.7%) did not consider joining the available organisation. Smallholder farmers in Gabaza

were not members of any farmers organization because 33.3% mentioned that there was no

organisation and that 30% lacked knowledge about farmers’ organisation development.

This was different from the studies conducted by Ndove et al. (2004) where participants

belonged to the Dan farmers association which participated in a multidisciplinary Rural

and Development pilot project, a partnership between the Australian Centre for

International Agricultural Research (ACIAR) and South Africa. This was similar to studies

conducted by Misiko et al. (2007) in Kenya where participating farmers originated from

44

the previously initiated Folk Ecology (FE) program of the Tropical Soil Biology and

Fertility Institute (TSBF) of the International Centre for Tropical Agriculture (CIAT). The

intervention of the outside institutions proved to have a positive impact on farmers’

knowledge and skills application through farmers groups development. Misiko et al.

(2007) reported a positive adoption of the FE program by farmers in Kenya through

establishment of dialogue between the TSBF-CIAT and farmers.

The poor membership by farmers of farmers organisations and groups raises concerns

about effectiveness of the extension personnel in farmers organisational development in

the two villages. The interest of smallholder farmers for participating in agricultural

activities is influenced by observable practical benefits. Ndove et al. (2004) observed

consistent increase in the use of inorganic fertilisers by farmers who had never used

fertilisers and the adoption of row planting over the traditional method of broadcasting

seeds through farmer participatory program during 1999 to 2001 study.

Reasons for non-membership

0.0

20.0

40.0

60.0

80.0

100.0

No organiza

tion

Lack

of kn

owledge

In pro

cess

No enc

ourag

ent

Working

prev

iously Sick

Did no

t con

sider

% o

f far

mer

s

GaKgoroshiGabaza

Figure 3.11. Reasons for non-membership by smallholder farmers

3.3 Conclusion

Understanding the socio-economic status and farming details of smallholder farmers

systems in Limpopo province can help in determining the types of technology suitable for

45

their farming systems. Considering their old age, low level of education, large family size,

unemployment and small number of livestock including situations where no livestock is

kept, simple technology which involve locally available and accessible resources will help

in addressing issues related to low crop yields, lack of labour and funds to purchase inputs.

Conducting experiments that demonstrate the potential use of legumes in nitrogen fixation

and N release for soil fertility and crop yield improvement can change farmers traditional

way of growing legumes on a small portion of land relative to land grown to maize crop.

Additionally, farmers will harvest legume crop residues and litter to use as manure rather

than leaving them in the fields to be grazed by livestock. Working with groups of farmers

can assist in increased technology transfer and sharing of knowledge among farmers rather

than working with individual farmers. The smallholder farming systems require the

involvement of youth, more training in cropping systems and nitrogen fixation, and

encouragement in the formation of organizational development. The farmer’s time

allocated to their field as compared to their off-farm activities could also be considered.

46

CHAPTER 4

The effects of fertiliser, legumes and grass mulches applied to a maize crop in Limpopo province

4.1 Introduction

The smallholder farmers of Limpopo Province growing crops under rainfed conditions face

challenges of low crop yields mainly due to poor soil fertility, frequent droughts and

prolonged periods within the growing season of low rainfall. The results from chapter 3

show that intercropping maize and legumes or growing them as sole crops with legumes

grown on very small piece of land compared to maize, is a common practice by

smallholder farmers. The residues of these crops are normally left in the fields for grazing

by livestock, which could suggest value they place on livestock.

Many soil fertility and crop yields improvement technologies have been tested in the

region including fertiliser application, farmyard manure and the inclusion of nitrogen

fixing crops (Giller, 2001; Maluleke et al., 2004; Mpangane et al., 2004; Ndove et al.,

2004). Many of these technologies are unaffordable to resource poor smallholder farmers,

particularly the use of inorganic fertiliser. Organic fertilisers such as farmyard manure are

usually limited in supply and applied in small quantities at irregular times. The inclusion of

legumes in the cropping systems, either in rotation with cereal crops such as maize,

intercropped or applied as a green manure or mulch have been shown to improve crop

yields where fertility is limiting (Thonnissen et al., 2000; Ayisi and Mpangane, 2004; Jiri

et al., 2004)

In addition to the soil fertility improvement technologies, the simulation models are used to

determine possible crop yields and changes in soil fertility under variable climatic

conditions and soil types (Hayman et al., 2008; McCown et al., 1996; Nelson et al., 1998).

Model parameters derived from locally measured data have shown to provide accurate

simulations of results (Probert et al., 1998b; Shamudzarira and Robertson, 2002;

Whitbread and Clem, 2004). Accurate adjustments of parameters increased models’ ability

in determining production results (Keating et al., 2003; Whitbread and Ayisi, 2004; Wolf

et al., 1989); however, the comprehensive data sets required for model parameterisation are

rarely available for developing countries. The hypothesis of this study was that legume

mulch and N fertiliser increase crop yields more than non legume mulch and; therefore,

47

proper parameterisation of APSIM (Agricultural Production Systems sIMulation- APSIM)

model shows the best possible crop management options under different soil types and

variable climatic conditions.

The aim of this study was to compare maize performance when legume mulch, grass

mulch and N fertilizer are applied in locations of variable climate, in Limpopo province

using experimentation and simulation model (Agricultural Production Systems sIMulation-

APSIM).

4.2 Material and methods

4.2.1 Site description

On-farm field experiments were conducted in the villages of GaKgoroshi (-23.721120°,

29.135534°) and Gabaza (-23.991964°, 30.334951°) in the Limpopo province of South

Africa. The field experiment at GaKgoroshi was destroyed twice by livestock and could

not recover due to low rainfall and is not reported on further in this chapter. During the wet

season of 2007/08, an experiment was established where mulch, fertiliser N or a

combination of mulch and fertiliser N was applied to plots prior to sowing a maize crop.

The experiment was designed as a randomized complete block initially with nine

treatments and three replicates. The treatments consisted of two levels of nitrogen (0 and

30 kg N ha¹־) applied as Limestone Ammonium Nitrate (28% N) and three types of mulch

residues applied at 10t ha¹־ before planting [(thatch grass, Hyparrhenia hirta (0.5%N);

mucuna, Mucuna pruriens (3.3%N) and guar bean, Cyamopsis tetragonoloba (2.3%N)].

There were also treatments where N fertiliser and crop residues were combined and where

residues remained either as surface mulch or incorporated.

There were 4 fertiliser treatments with no mulch applied: Control (0N), 30 kg N ha¹־

(30N), 60 kg N ha¹־ (60N), 90 kg N ha¹־ (90N). Fertiliser application was split with half

the total amount to be applied at sowing and 6 weeks after planting for the 30N and 60N

treatments. The 90N treatment received 30 kg N ha-1 at sowing and 6 weeks after planting

with no further application possible due to drought in GaKgoroshi. In Gabaza the 30N

received 15 kg N ha-1 at sowing while 60N and 90N received 30 kg N ha-1, with no further

fertiliser application due to lateness in the season

48

The surface mulch treatments consisted of pre-sowing application of 10 t ha-1 of grass,

mucuna or guarbean mulch with no additional N fertiliser (grass_0, mucuna_0 or

guarbean_0) and guarbean mulch plus 30 kg N ha¹־ (grass_30, guarbean_30). The green

manure incorporated treatments consisted of guar bean incorporated into the soil with

handhoes prior to sowing with no additional N fertiliser (mucuna_Inc or guarbean_Inc).

Table 4.1. Treatments designed and implemented at GaKgoroshi and Gabaza

Treatments GaKgoroshi Gabaza

As designed Implemented Implemented

0N 0N 0N

30N 30N 15N

60N 60N 30N

90N 60N -

Grass_0 Grass_0 Grass_0

Grass_30 Grass_30 Grass_30

Guarbean_0 Guarbean_0 Guarbean_0

Guarbean_30 Guarbean_30 Guarbean_30

Guarbean_Inc Guarbean_Inc Guarbean_Inc

A basal phosphate (P) fertiliser treatment was applied as single super-phosphate at 30 kg P

ha¹־ to all plots during planting. N fertiliser was placed under the seed at sowing. Maize

seeds (ZM423) were sown at 60 000 plants ha-1 on the 29 January 2008. The experiment

was replanted on the 28 February 2008 because the plants from the first sowing did not

emerge due to drought. The crops were weeded twice, at twelve (12) days after planting

and at 4 weeks after planting.

4.2.2 Soil sampling and analysis

Prior to applying the treatments, soil sampling occurred at each of the sites using a hand

auger to extract soil from the depth increments (0-15cm, 15-30cm, 30-60cm and 60-90 cm

depth). A soil sample was taken from each replication and bulked into a single composite

sample for each depth. Samples were air-dried and analysed for a range of physical and

chemical properties (Table 5). The soil was analysed for pH (1:5 water), N (NH 4+ + NO3

-)

using 1:5 Ext. 0.1N K2SO4, P using 1:75 Ext. Bray 2; Cl using 1:2 Ext. 0.1KNO3+

; Ca,

Mg, K and Na using 1:10 Ext Ammonium Acetate- 1N, pH7; Zn with 1:4 Ext.-0.1N HCl;

49

Organic carbon using Walkley-Black; Exchangeable acidity and Aluminium with 1:10

Ext.- 1N KCl and soil texture using hydrometer. Bulk density was measured at the same

depths as the soil samples by driving rings (2.5 cm radius and 5cm length) into the soil and

removing intact cores. These were then dried at 105oC for 48 hours, cooled and then

weighed.

4.2.3 Crop measurements

Plant height was determined by measuring the extended leaf at four (4) and eight (8) weeks

from three plants/plot in each treatment selected at random. Dry matter was collected at 8

and 12 WAP by randomly sampling 3 plants at two locations within each plot. The plants

were cut at soil surface, oven-dried at 55oC and weighed.

At 8 weeks, composite leaf samples (the youngest fully expanded flag leaf of six maize

plants per treatment) were collected from 0N, 30N, guarbean_0, guarbean_30,

guarbean_Inc, grass_0 and grass_30. These samples were air dried and analysed to check

the effects of the different treatments on N content.

The crops did not reach maturity because of drought and also the late planting did not

allow enough time for the crop to reach maturity before the winter.

4.2.4 Simulation analysis

Using the Agricultural Production Systems sIMulator (APSIM) model (Keating et al.,

2003; McCown et al., 1996), maize growth was simulated for Gabaza using long term

(1970-2008) weather records. APSIM is a modelling framework and it is described as the

tool for exploring management strategies that can improve the economics of agricultural

production systems and the consequences for the soil resource and the environment.

a. Simulation of maize performance

Using the dry matter data from the field experiment at Gabaza in 2007/08 as a validation

dataset, the APSIM model was parameterised to simulate the growth of maize until 12

weeks after planting. A long term simulation (1970-2008) was then run to determine the

effects of the crop residues that were used in the experiment on maize yield, soil water and

mineral N. Weather data (rainfall, max and min temperature and radiation) was from long

50

term measurements obtained from the Bureau of Meterology sites at Thabina, Letaba and

Mopani. Rainfall data collected by the local extension officer during the 2007/08 growing

season was also used.

b. Validation of simulation for Gabaza 2007/08

In order to represent the effect of the grass weeds that were growing on the field site prior

to the crop being sown, this simulation was initialised on the 01 October 2007 with a weed

crop (late cultivar, summer grass) planted at sowing density of 7 plants m-2 in 20mm depth

and 25cm row spacing. The soil was tilled on the 16 January 2008 with soil organic matter

module used and disc as tillage type. On the 29 January 2008 the weeds were killed by disc

tillage and residues incorporated into the soil. Mulch was applied as surface organic matter

at 10 000 kg ha¹־ either as grass with C:N ratio of 88 or mucuna with C:N ratio of 19.3

applied on a fixed date (29 January 2008). Fertiliser used in the simulation was NH4NO3

which represented limestone ammonium nitrate (28%N) applied at rates of 0, 15 and 30 kg

N ha¹־. The maize crop (SC501 variety which mostly closely represents the ZM423 variety

used in the field experiment) was sown on the 28 February 2008 at 6 plants/m2 in 0.9 m

rows. Soil mineral N was initialised to 21.5 kg N ha-1, the value measured in the field

experiment. Soil water and surface organic matter were not reset on sowing with the values

determined by the model.

c. Long term simulation for Gabaza (1970-2008)

The soil was tilled using the variable rule tillage on an event with soil organic matter as

module and disc as tillage type. Sowing was done using a variable rule with sowing

triggered when 30 mm rain fell over three days when there was at least 30 mm plant

available water in the soil during 01 November to 15 January sowing window. The maize

crop (SC501 variety which mostly closely represents the ZM423 variety used in the field

experiment was sown at 6 plants/m2 in 0.9 m rows and harvested at maturity. Mulch was

applied as surface organic matter at 10 000 kg ha-1 either as grass with C:N ratio of 88 or

mucuna with C:N ratio of 19.3 at sowing. The fertiliser used was also NH4NO3 with 28%N

applied at sowing. Soil mineral N was initialised to 21.5 kg N ha-1, the value measured in

the field experiment. Soil water was not reset and the model determined its own soil water

settings following the long term weather records. Soil organic matter and N were reset to

the initialisation settings on a fixed date (01 October each year).

51

d. Soil characterisationAPSIM requires a fully characterised soil profile. The soil water for this simulation was

characterised by determining the plant available water capacity (PAWC), bulk density

(BD), drained upper limit (DUL) and crop lower limit (CLL) of the soil. The plant

available water capacity is defined as the difference between the upper water storage limit

of the soil and the lower extraction limit of a crop over the depth of rooting by Ratcliff et

al. (1983) cited by Dalgliesh and Cawthray (1988). For this simulation, PAWC was

determined using the formula described by Dalgliesh and Cawthray (1988). The PAWC for

the rooting which was assumed to be 90cm deep was 66 mm.

The DUL is defined by Ratcliff et al. (1983) as the highest field measured water content of

a soil after it had been thoroughly wetted and allowed to drain until drainage became

practically negligible. This was determined by multiplying the gravimetric water % (for

each depth interval) by the bulk density (g/cc). The CLL being defined as the lowest field-

measured water content of a soil after plants had stopped extracting water and were or near

premature death or become dormant as a result of water stress (Godwin et al., 1984;

Ratcliff et al., 1983) was determined by multiplying the gravimetric water % (when the

crop is mature or stressed) by the bulk density (g/cc) according to the formula used by

Dalgliesh and Cawthray (1988).

The gravimetric water content (%) is defined (Gardner, 1985) as the mass of water (g)

relative to the mass (g) of oven dry (at 105oC) soil. This was calculated using the formula:

[(wet weight (wt) of sample- dry wt of sample)/dry wt of sample- container tare)] x 100

following the method described by Dalgliesh and Cawthray (1988). The bulk density (g/cc)

is defined as the ratio of the mass of dry solids to the bulk volume of the soil (Blake and

Hartge, 1986). This was measured by dividing dry soil wt (g) by total volume of soil (cc)

according to Dalgliesh and Cawthray (1988). The BD was determined using the core

method which was described by Blake and Hartge (1986) and calculated according to

Dalgliesh and Cawthray (1988).

52

Table 4.2. Rainfall (mm) during 2007-2008 and long term average (LTA)

Jul Au

g

Sep Oct Nov Dec Jan Feb Mar Apr Ma

y

Jun

2007/0

8

39.

6

1.2 60.

1

7.5 150.

5

30.0 20.5 31.0 67.1 19.

4

12.0 n.d

.

LTA 8.2 6.2 20.

1

47.

7

93.2 121.

3

137.

3

122.

9

109.

7

39.

1

14.4 6.5

n.d. = not determined

4.2.5 Statistical analysis

The data for plant height and drymatter were analysed using the analysis of variance in

JMP (JMP, 2005; SAS Institude Inc, 2005). The treatment means in all the analyses were

compared by Tukey-Kramer HSD. Treatment means were declared significant at P = 0.05

using the Tukey-Kramer HSD (honestly significant difference) (JMP, 2005).

4.3 Results

4.3.1 Soil chemical composition

The results from the soil chemical composition analyses are presented in Table 4.2. The

soil is a sandy clay loam with a uniform texture at least to 60 cm (Table 4.3). While it

contains a reasonable concentration of soil organic C content in the topsoil and adequate

cations, plant available P is very low (Table 4.2).

Table 4.3. Soil chemical analysis for the soil profiles

Depth pH C Mineral

N

Pa Kb Cab Mgb Nab Clc Znc S-

SO4

cm % Mg kg-1

0-15 5.4 1.2 4.3 1.3 47.0 674.3 303.7 7.0 4.7 1.0 7.0

15-30 5.3 1.1 1.7 1.3 30.0 715.0 290.0 6.0 3.0 0.9 8.0a 1:7.5 extractant Bray 2b 1:10 extractant ammonium acetate 1N pH7c 1:2 extractant 0.1N KNO3d 1:4 extractant 0.1N HCl

53

Table 4.4. Soil particle size analysis

Depth Sand Silt Clay

cm %

0-15 63 11 25

15-30 63 10 26

30-60 59 11 30

4.3.2 Maize plant height

Generally, maize plant height responded to 30kgNha-1 fertiliser applications. At 4 weeks

after planting (WAP), plant height was highest for the 30N, guarbean_30 and grass_30 but

with no significant differences observed among these treatments (Figure 4.1). Significantly

lower plant heights were observed where no fertiliser had been applied with the exception

of the 15N treatment which was not significantly different from the 30N or the

guarbean_Inc treatments. The 0N, grass_0 and guarbean_0 resulted in lowest heights. The

height between these treatments was not significantly different (Figure 4.1).

At 8 WAP, the effects of the different treatments on plant height was the same as at 4

WAP with 30N, guarbean_30 and grass_30 showing high and significantly different plant

height than other treatments (Figure 4.1). The plant height in the 0N, guarbean_0 and

grass_0 treatments was low.

54

0

20

40

60

80

100

120

4wks 8wksTime after planting

Pla

nt h

eigh

t (cm

)

0N15N30NGuarbean_0Guarbean_30Guarbean_IncGrass_0Grass_30

0

20

40

60

80

100

120

4wks 8wksTime after planting

Plan

t hei

ght (

cm) 0N

15N30NGuarbean_0Guarbean_30Guarbean_IncGrass_0Grass_30

d d d

bc

ab a a

c

d

cdd

bcab

ab ab

c

Figure 4.1. Maize plant height as influenced by different soil fertility managementpractices. Vertical error bars represent standard deviation. Treatment meansfollowed by the same alphabet were not significantly different at P = 0.05.

4.3.3 Maize dry matter

Maize dry matter was affected by the different treatments similar to the plant height. At 8

WAP the dry matter was significantly higher in the 15N, 30N, guarbean_30 and grass 30

treatments (Figure 2). Low dry matter weights were measured in the 0N, guarbean_0 and

grass_0 treatments (Figure 4.2).

At 12 WAP maize dry matter was highest in the 30N, guarbean_30 and grass_30

treatments showing similar effects as at 8 WAP. The 15N treatment, guarbean_Inc and

guarbean_0 showed improved dry matter at this stage. Maize growth in the 0N and grass_0

treatments still showed low performance.

55

c

0

1000

2000

3000

4000

5000

8 wks 12 wks

0N15N30NGuarbean_0Guarbean_30Guarbean_IncGrass_0Grass_30

Mai

ze d

rym

atte

r(kg

ha-

1 )

bc

a a

bc

aab

bc

a

c

aa

ab

a

a

bc

a

Figure 4.2. Maize dry-matter as influenced by different soil fertility managementpractices. Vertical error bars represent standard deviation. Means followed by the same letter are not significantly different at P = 0.05

4.3.4 Relationship between maize plant height and drymatter

There is a positive relationship between plant height and dry matter (Figure 4.3) indicating

that the differences observed in plant height and drymatter were caused by the differences

in treatments. The plant height and drymatter were affected similarly in the various

treatments. The lowest plant height and dry matter were observed in the 0N, grass_0 and

guarbean_0 while the highest plant height and drymatter were achieved with the

application of grass_30 followed by guarbean_30, then by guarbean_Inc and 30N (Figure

4.3). Thus, the same treatment that affects plant height will affect drymatter the same way.

56

y = 0.029x + 38.204R2 = 0.92

0

10

20

30

40

50

60

70

80

90

100

0 500 1000 1500 2000

Plant drymatter (kg ha-1)

Plan

t hei

ght (

cm)

Figure 4.3. The relationship between plant height and drymatter as influenced by different soil fertility management

4.3.5 N % in maize leaves

The N % in maize ranged from 1.35 to 1.90. The percentage N in maize leaves was 1.55,

1.73, 1.70, 1.35, 1.90, 1.90 and 1.65 in the 0N, 30N, grass_0, grass_30, guarbean_0,

guarbean_30 and guarbean_Inc respectively.

4.3.6 Maize yield

Maize crops did not reach maturity because of drought and also due to the on-set of cold

temperatures prior to maturity. The simulation also showed that maturity was not reached

by the 31 May 2008 and therefore no simulated grain yields are reported for the 2007/08

growing season.

4.3.7 Growing season simulated maize crop response to fertiliser, grass and guarbean mulch

In general, the simulation shows better prediction of drymatter at 12 WAP (Figure 4.5).

The simulation over-predicted drymatter for the control, grass_0 and guarbean_0 at 8weeks

(Figure 4.5). The better predictions appeared with the grass_30 and guarbean_30

treatments. The same trend occurred at 12weeks for the control and grass_0N treatments

57

with the predicted drymatter being more than the observed (Figure 4.5). The application of

30N, grass_30, guarbean_0 and guarbean_Inc showed the observed to be in agreement

with the predicted at 12weeks (Figure 4.5). With the application of guarbean_30 the

drymatter was under-predicted (Figure 4.5).

12 WAP

0500

10001500200025003000350040004500

Control

15N

30N

Grass_

0

Grass_

30

Guarbea

n_0

Guarbea

n_30

Guarbea

n_Inc

8 WAP

0500

10001500200025003000350040004500 Observed

Simulated

Mai

ze d

rym

atte

r (kg

ha-

1 )

Figure 4.4. Comparison between the observed and simulated maize biomass duringthe 2007-2008 growing season.

58

4.3.8 Simulated long term maize production and its response to mulch type and fertiliser

a. Maize dry matter

The simulation of maize dry matter showed lower dry matter production under grass and

grass_30 applications than all treatments in most seasons (Figure 4.6). Maize in these two

treatments produced less (<4000 kg ha-1) dry matter during the poorest seasons

(approximately 20% of seasons) (Figure 4.6). The maximum yield in these two treatments

reached 6424 and 7326 kg ha-1, respectively, showing that in all the seasons, maize

produced less dry matter (<8000) in the two treatments, which is less or similar to the

control (Figure 4.6). The grass mulch showed to reduce maize dry matter even when

combined with fertiliser (Figure 4.6).

With the application of guarbean with or without fertiliser, or when incorporated into the

soil resulted in maize producing more (>8000 kg ha-1) dry matter in 26% of the seasons

(Figure 4.6). The 30N treatment resulted in maize producing dry matter >8000 kg ha-1 in

16% of the seasons. The highest dry matter yield (10146 kg ha-1) was produced with the

application of guarbean_30 treatment (Figure 4.6).

0.0

0.2

0.4

0.6

0.8

1.0

0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000 11000

Maize drymatter (kg ha-1)

Prob

abili

ty o

f exc

eeda

nce

0N30NGuarbean_0Guarbean_IncGuarbean_30Grass_0Grass_30

Figure 4. 5. Cumulative distribution functions for maize dry matter with differentmulch and N fertiliser treatments during long term period (1970-2008)

59

b. Maize grain yield

Similar to the dry matter, the predicted grain yield in grass_0 and grass_30 was the lowest

of all treatments (Figure 4.7). The grass mulch reduced grain yield even with the addition

of fertiliser (Figure 4.7). The application of guarbean and N fertiliser increased grain yield

in some of the seasons (Figure 4.7). The control, 30N, guarbean_30, guarbean_Inc and

grass_0 treatments show poor maize grain yields (<1000 kg ha-1) in 39% of the seasons,

and 3 crop failures, except 2 for grass_0 (Figure 4.7). When grass_30 and guarbean_0 were

applied, maize yield were poor (<1000 kg ha-1) in 29 and 24% of the seasons, respectively,

with 2 crop failures (Figure 4.7). The highest yields with the application of grass_0 and

grass_30N were 2415 and 2976 kg ha-1 in the best seasons, respectively; which is nearly

similar to the control treatment (2929) (Figure 4.7). With the application of guarbean_30,

guarbean_Inc, guarbean_0 and 30N, maize yield reached 4398, 4341, 4448 and 3846 kg

ha-1, respectively, in the best seasons (Figure 4.7).

0.0

0.2

0.4

0.6

0.8

1.0

0 1000 2000 3000 4000 5000

Maize grain yield (kg ha-1)

Prob

albi

lity

of e

xcee

danc

e

0N30NGuarbean_0Guarbean_IncGuarbean_30Grass_0Grass_30

Figure 4.6. Cumulative distribution function for maize grain yield (1971-2008)

60

4.3.9 Untangling the effects of N and water using modelled N and water stress.

In calculating actual plant growth rate, APSIM uses 2 factors, soil water stress and soil N

stress, to modify potential growth of the crop as determined by radiation. In APSIM, soil

water and N dynamics will be influenced by residue quality or fertiliser application and

using these stress factors, the main drivers can be determined.

a. Soil water deficit factor

The soil water stress is determined by the soil water deficit factor (swdef). The swdef of 1

indicates complete stress and zero (0) is no stress. There are differences in terms of soil

water stress in different treatments.

In general, more water stress appears during the start to end of grain fill than during

flowering to start of grain fill in other treatments except for grass_0 where stress was

experienced in the 2 stages (Figure 4.8). The grass mulch treatments reduced water stress

in the 2 stages except for the grass_30 where water stress is medium water stress occurred

during start to end of grain fill (Figure 4.8). Low water stress in the grass_0 treatment was

almost certainly because growth was comparably poor (Figure 4.2). More water stress was

experienced in the 0N (0.55), 30N (0.62), guarbean_0 (0.54), guarbean_30 (0.62) and

guarbean_Inc (0.62) treatments (Figure 4.8).

0

0.2

0.4

0.6

0.8

1

0N 30N

Grass_

0

Grass_

30

Guarbea

n_0

Guarbea

n_Inc

Guarbea

n_30

Flowering-Grainstart Grainstart-end

Soil

wat

er d

efic

it fa

ctor

Figure 4.7. Average soil water deficit factor for maize growth during the 2007- 2008growing season

61

b. Soil N factor

Similar to the soil water stress, the N stress is affected differently by the application of the

various treatments. In general, the simulation predicted more nitrogen stress (Figure 4.9) in

treatments that showed less or no water stress (Figure 4.8), with more stress appearing at

flowering to start of grainfill.

The grass_0 and grass_30 showed more N stress than other treatments during flowering to

start of grain fill (0.66 and 0.64) and from start of grain fill to end of grain fill (0.64 and

0.60), respectively (Figure 4.9). There is no N stress observed with the application of 30N,

guarbean_0, guarbean_30 and guarbean_Inc at the two stages (Figure 4.9).

Figure 4.8. Nitrogen deficit factor on maize grain yield during the 2007-2008 growing season

d. Long term effects of growing season rainfall on N and water stress

The N stress following the application of grass mulches was not affected by the amount of

rainfall at the two stages (Figure 4.10A and B). The grass mulch treatments showed high N

stress in virtually every season within the observed rainfall range. The 30N and 0N showed

a positive correlation with rainfall (Figure 4.10) indicating N stress in wet seasons. With

the application of guarbean mulch, the N stress showed a weak correlation with rainfall

(Figure 4.10). The application of guarbean_0 showed reduction in N stress during low

rainfall period (<600mm) and increased N stress during higher rainfall period (Figure 4.10)

0

0.2

0.4

0.6

0.8

1

0N 30N

Grass_

0

Grass_

30

Guarbea

n_0

Guarb

ean_In

c

Guarbea

n_30

Flowering-Grainstart Grainstart-end

N d

efic

it fa

ctor

62

when N fixed would be inadequate to meet plant demand. The 30N, guarbean_Inc and

guarbean_30 reduced N stress during seasons that received less rainfall (<400mm) and

increased stress when rainfall increased above 400mm (Figure 4.10). The effect of rainfall

on N stress at start to end of grainfill was less than at flowering; however, the pattern of

stress in relation to rainfall was similar in all the treatments.

With water stress, the reduction was more with the application of grass mulches; however

some stresses were observed at low rainfall seasons, <200mm and 300mm for grass_0 and

grass_30, respectively (Figure 4.11). The guarbean mulches and 30N increased water stress

more than grass mulches during low rainfall seasons; however the stress showed to

decrease with the increase in rainfall (Figure 4.11). The same effects of rainfall on water

stress were observed during the start to end of grainfill.

63

Figure 4.9. Correlation between long-term in-crop rainfall and N stress duringflowering to start of grainfill (A) and from start to end of grainfill (B).

y = 0.0005x - 0.1202R2 = 0.6116

y = 0.0001x - 0.0101R2 = 0.0692

0

0.2

0.4

0.6

0.8

1

0 200 400 600 800 1000 1200

y = 5E-05x + 0.5743R2 = 0.0092

y = 6E-05x + 0.0567R2 = 0.0082

0

0.2

0.4

0.6

0.8

1

y = 1E-04x + 0.5802R2 = 0.0806

y = 6E-05x + 0.0594R2 = 0.0059

0

0.2

0.4

0.6

0.8

1 Grass_30 Guarbean_30Linear (Grass_30) Linear (Guarbean_30)

y = 0.0004x + 0.1929R2 = 0.2661

y = 1E-05x + 0.6575R2 = 0.0017

y = 0.0001x - 0.0124R2 = 0.0775

0

0.2

0.4

0.6

0.8

1 0N Grass_0Guarbean_0 Linear (0N)Linear (Grass_0) Linear (Guarbean_0)

y = 0.0004x - 0.0937R2 = 0.5232

y = 0.0001x + 0.0039R2 = 0.0481

0

0.2

0.4

0.6

0.8

1

0 200 400 600 800 1000 1200

30NGuarbean_Inc

Linear (30N)Linear (Guarbean_Inc)

y = -2E-05x + 0.6169R2 = 0.0025

y = 0.0005x + 0.0798R2 = 0.3699

y = 0.0001x - 0.0039R2 = 0.0402

0

0.2

0.4

0.6

0.8

1

Rainfall (mm)

N d

efic

it fa

ctor

BA

64

Figure 4.10. Correlation between long term growing season rainfall and soil water stress from flowering to start of grainfill (A) and from start to end of grainfill (B)

e. The long-term frequency of N and water stress at flowering to start of grain fill and at start to end of grain fill, and in-crop soil water evaporation

The frequency of water and stress was calculated from the total number of observations (n

= 38). The frequency showed less cases of water stress factor above 0.5 for the grass

mulches than the guarbean mulches, 30N and the control in both stages which could be due

to poor crop growth and water use (Table 4.4). For the N stress, the grass mulch showed

more cases of N stress factor above 0.5 even with the addition of fertiliser, whereas the

guarbean mulch showed least cases of N stress above 0.5 (Table 4.4). The frequency of soil

y = -0.001x + 0.8242R2 = 0.5264

y = -0.001x + 0.8133R2 = 0.536

0 200 400 600 800 1000 1200

30NGuarbean_IncLinear (Guarbean_Inc)Linear (30N)

y = -0.0004x + 0.3014R2 = 0.2756

y = -0.0009x + 0.7541R2 = 0.454

0 200 400 600 800 1000 1200

Grass_30Guarbean_30Linear (Grass_30)Linear (Guarbean_30)

y = -0.001x + 0.7203R2 = 0.4955

y = -0.0002x + 0.118R2 = 0.102 y = -0.001x + 0.8108

R2 = 0.5442

0 200 400 600 800 1000 1200

0NGrass_0Guarbean_0Linear (0N)Linear (Grass_0)Linear (Guarbean_0)

y = -0.0004x + 0.3385R2 = 0.1534 y = -0.0008x + 0.871

R2 = 0.3534

0

0.2

0.4

0.6

0.8

1

0 200 400 600 800 1000 1200

y = -0.001x + 0.815R2 = 0.3858

y = -0.0002x + 0.176R2 = 0.0421

y = -0.0008x + 0.8824R2 = 0.3352

0

0.2

0.4

0.6

0.8

1

0 200 400 600 800 1000 1200

y = -0.001x + 0.939R2 = 0.481

y = -0.0009x + 0.9478R2 = 0.4696

0

0.2

0.4

0.6

0.8

1

0 200 400 600 800 1000 1200

A B

Growing season rainfall

Soil

wat

er d

efic

it fa

ctor

65

water evaporation was determined as the amount of water loss above average (137 mm)

during the in-crop period as affected by the different treatments. The frequency of soil

water evaporation above average was not detected for grass mulch treatments whereas

more water evaporated from the control and the 30N treatments than the guarbean

treatments (Table 4.4).

Table 4.5. The frequency of water and N deficient factor above 0.5 during floweringto start of grain fill (FS) and from start to end of grainfill (SE), and soil waterevaporation above average (137 mm)

Water stress N stress Soil water evaporation

FS SE FS SE In-cropTreatments Frequency > 0.5 Frequency > 137

mm0N 16 10 7 8 33

30N 18 13 0 1 33

Guarbean_0 19 14 1 1 27

Guarbean_Inc 20 13 1 1 29

Guarbean_30 21 13 3 1 28

Grass_0 3 1 37 35 0

Grass_30 6 4 36 34 0

4.4 Discussion

4.4.1 Maize drymatter and height

Maize drymatter tends to follow the plant height pattern, being high in treatments where

plant height is greater. The increased drymatter and plant height with the application of

grass_30 could be attributed to the less water stress induced by the mulch and reduced N

stress caused by the uptake of readily available N from the added N fertiliser by the crops.

Azooz et al. (1995) demonstrated that mulch reduces soil water evaporation and helps to

retain moisture for the crops. In addition to the beneficial effects of grass mulch, the 30 kg

N fertiliser added to this mulch could have contributed to the direct uptake of the nitrogen

by plants for growth (Green and Blackmer, 1995). The same scenario was observed by

Ramakrishna et al. (2006) where groundnut biomass was significantly different in rice

straw mulch than in un-mulched plots.

66

With the application of the grass mulch with no added fertiliser, the reduced plant height

and drymatter is likely to be due to N stress induced by the absence of fertiliser N as well

as the potential for N tie-up by the addition of this high C:N ratio mulch. In the study

conducted by Ncube et al. (2009) the simulation predicted more N stress with the

incorporation of sorghum mulch. The unavailability of plant available soil N in this mulch

treatment is therefore due to N immobilisation during decomposition.

The increased drymatter and plant height following the application of guarbean mulch

could be due to reduced N stress observed with the application of this mulch. Ncube et al.

(2009) observed less N stress in legume incorporated plots as compared to sorghum plots.

This was confirmed in studies conducted by Recous et al. (1995) and Tian et al. (1993),

where maize dry matter was the highest in plots mulched with legume than maize stover

and rice straw because of the high N content in legume mulch. The significant difference

observed in guarbean_Inc plots as compared to guarbean_0, particularly at 8 WAP, is

likely to be due to greater N mineralisation of incorporated residues than when placed on

the soil surface. Mafongoya and Nair (1997) reported similar effects of incorporated mulch

as compared to surface placed legume mulch.

The non significant differences in maize drymatter between the control, grass_0 and

guarbean_0 could be due to low N content of the soil (Table 5.1), immobilised nitrogen by

microbes during decomposition of grass mulch (Green and Blackmer, 1995) making N to

be unavailable to crops, and N stress induced by the slow decomposition of guarbean

mulch during early stages of crop development (Ncube et al., 2009).

The higher performance of maize in grass-30 over guarbean mulches on maize dry matter

could be attributed to the high C:N ratio in grass mulch than in legumes (Hadas et al.,

2004; Leblanc et al., 2006; Mubarak et al., 2002) which decomposed slowly resulting in

more retained soil moisture according to Tian et al. (1993). In addition to the C:N ratio in

this treatment, the application of N fertiliser would have overcome the N deficiency. Maize

dry matter yield in the 15N and 30N fertiliser plots was significantly higher than the

control due to increased N supply from the fertiliser. This is in agreement with the results

obtained by Chikowo et al. (2004) where maize receiving N fertiliser produced more dry

matter than maize in the control plot.

67

4.4.2 Simulated and observed maize dry matter during 2007/08 growing season

Although the experiment described in this chapter provides a limited dataset to test the

performance of APSIM, work done in neighbouring villages by Whitbread and Ayisi

(2004) and in the region reported by Shamudzarira and Robertson (2002) and Ncube et al.

(2009) provide evidence that APSIM is a reliable simulation tool. At 8 WAP, the over-

prediction of drymatter for the control and grass_0 in this experiment could be attributed to

the poor simulation by the model of the N stress induced by the application of the grass

mulch and increased soil water evaporation from the control treatment (Table 4.4). This

was in contrast with the results obtained by Ncube et al. (2009) where the simulation

predicted a large reduction of sorghum drymatter with the incorporation of sorghum

residues. The over-prediction with the application of 30N, guarbean_0 and guarbean_Inc

could be due to underestimation of water stress experienced from these treatments in the

simulation. The better prediction for the grass_30 could be attributable to the model

considering the reduced water stress (Figure 4.8) and N stress (Figure 4.9) induced by the

application of the 30 kg N fertiliser to the grass_30 treatment. With the guarbean_30

treatment the better prediction could be due to the reduced N stress (Figure 4.9) and

increased N supply from the guarbean mulch and 30 kg N fertiliser. The over-prediction of

drymatter for the 0N treatment could be due to the model simulating lower than actual N

and water stress for this treatment. This was similar to the results obtained by Ncube et al.

(2009) in which total biomass of sorghum was over-predicted for the no residues treatment.

At 12 WAP, the better prediction of drymatter in the guarbean_0, guarbean_Inc and 30N

could be associated with the ability of the model to simulate the adequate N supply from

these treatments. Perhaps this result indicates that N dynamics in the system requires some

attention. With the application of guarbean_30 the under-prediction of drymatter could be

attributed to the model simulating N depletion during the 8 WAP when more rain

(67.1mm) was received which could have led to increased mineralisation of N from this

treatment and increased water stress due to increased crop growth at a later stage. The

model could be also simulating slow release of N from guarbean mulch. The best

prediction with the grass_30 indicates that the model was able to eliminate the water stress

and considered the increased N supply from this treatment. Some lack of agreement

between the predicted and the observed yield has also been reported in previous studies

conducted by Probert et al. (1995). The accurate prediction of results requires accurate

parameterisation of N and water following the application of mulches.

68

4.4.3 Long term simulated maize grain drymatter and yield, soil water deficit factor, and N deficit factor

Maize drymatter and grain yield

The maize drymatter and grain yield as simulated over the long term period were the

lowest with the application of grass mulch even when fertiliser was added to the mulch, but

with the application of guarbean mulches, both drymatter and maize yield were increased.

The low drymatter and poor yield with the application of grass mulches could be

attributable to the high N stress following the application of the grass mulch (Figure 4.9

and 4.10). The poor crop performance could not be attributed to the water stress as the

model showed that soil water evaporation did not exceed the average amount of 137mm

with the application of grass mulches (Table 4.4). These results were consistent with the

work done by Probert et al. (1995) and Ncube et al. (2009) in which crops experienced

severe N stress following the incorporation of sorghum residues.

With the application of guarbean mulches, increased drymatter could be due to less N

stress observed with the application of this mulch. However, crop growth in this treatment

would have been reduced greater soil water evaporation due to faster decomposition of the

mulch and water stress during grain fill caused by greater water use by the larger canopy

(Table 4.4). This was similar to the results obtained by Ncube et al. (2009) who reported

reduced N stress with the application of legumes.

The control treatment showed lower drymatter and yield than the 30N treatment which

could be attributed to increased N stress (Figure 4.9). The same scenario was observed by

Whitbread and Ayisi (2004) where the simulation predicted poor maize grain yields for

treatments with no N inputs than for 30kgN treatments. Thus, the application of grass

normally reduces plant growth due to N immobilisation whereas N fertiliser and guarbean

mulch increase plant growth due to N mineralisation.

4.4.4 Growing season rainfall and N deficit factor

The N stress following the application of grass mulch, with and without N fertiliser was

not affected by the amount of rainfall received during the season (Figure 4.10). There is no

correlation between the N stress and rainfall which means that crops grown under grass

mulch will be deprived of N even during high rainfall seasons. This was supported by the

69

in-crop growing season soil water evaporation which shows no cases of moisture

evaporation above average (Table 4.4). The same scenario was observed by Probert et al.

(1995) where stubble retention deprived crop of N. This was supported in studies

conducted by Ncube et al. (2009). There is high frequency of seasons (36 and 35 on

average out of 38 seasons for the grass_0 and grass_30, respectively) in which N stress

experienced was >0.5 in grass mulch (Table 4.4). The low maize height and poor drymatter

accumulation at the two stages clearly indicates presence of N stress with the application of

grass mulch.

When guarbean mulch was used, the weak positive correlation between N stress and

rainfall could suggest that the mulch decomposed slowly resulting in slow release of N

during the low rainfall seasons. The N stress following legume mulch tends to decrease

with the decrease in rainfall as mineralised N is unlikely to be leached out of the root zone.

As the rainfall received over the period of simulation was low (<600mm) in most seasons

(Figure 4.11), the crops could have utilised the N from the legume mulch resulting in

improved plant height and drymatter. The N stress frequency shows that the N stress was

mostly eliminated with the application of guarbean as the N stress >0.5 was experienced

only in 1 out of 38 seasons. The high frequency of seasons during which in-crop soil water

evaporation was above the average amount (137 mm) indicates that crops would suffer

water stress rather than N stress when legume mulches are utilised in this environment

(Table 4.4).

The positive correlation between rainfall and N deficit factor in the control and 30N

treatments indicates that increased rainfall will increase N stress (Figure 4.11) most likely

due to leaching of N. Table 4.4 shows that there are few cases where N stress was above

0.5, suggesting that crops are more likely to suffer water stress than N stress when fertiliser

N is used. This is supported by the greater frequency of seasons during which in-crop soil

water evaporation was higher than average amount (137 mm). This was consistent with the

results obtained by Whitbread and Ayisi (2004) in which N supply showed to be more

limiting than water in high rainfall seasons. Similar results were reported by Shamudzarira

and Robertson (2002).

70

4.4.5 Growing season rainfall and soil water deficit factor

The weak negative correlation between water stress and rainfall with the application of

grass mulch suggests that grass mulch reduced water stress for crops even during low

rainfall seasons, thus the crops are more likely to experience more N stress than water

stress. This trend is indicated by the lower frequency (2 and 5 on average out of 38 seasons

for grass_0 and grass_30, respectively) of water stress >0.5 (Table 4.4). These results are

consistent with the findings by Probert et al. (1995) where stubble retention improved

water conservation even in low rainfall seasons.

The 0N, 30N and guarbean treatments showed strong positive linear relationship between

water stress and the amount of rainfall. This relationship indicates that during low rainfall

seasons crop growth is strongly influenced by water stress whereas during high rainfall

seasons crops suffer more from water stress than N stress. This is supported by the high

frequency of seasons in which soil water evaporation exceeded the average amount of

137mm (Table 4.4).

4.5 Conclusion

The use of grass and legume mulch had large effects on maize growth. The application of

grass mulch without the addition of N fertiliser produced the lowest total shoot DM and

maize grain yield. Addition of N fertiliser to grass mulch provided some increase in grain

yield but the yields tended to be lower than the legume mulch treatments. These

differences between the grass mulch + 30N treatments were particularly striking under

high yielding seasons which had higher rainfall. Persistent N stress experienced with the

application of grass mulch appears to have a large impact on yield especially under yield

potential situations. The addition of N fertiliser to grass mulch caused some increase in

maize growth and grain yield but the performance was much lower than that when legume

mulches were used which prevented N stress factor from reducing growth. Some

divergence between observed and simulated maize DM particularly where grass mulch was

used indicates need for fine-tuning N dynamics in the model. However, the observed shoot

DM was much lower than the predicted which further highlights the importance of N stress

in Limpopo.

According to the simulation, soil water evaporation was reduced by grass mulch, probably

due to slower breakdown of the grass mulch due to its high C:N ratio. The legume mulch

71

had little effect on soil water evaporation, probably due to rapid breakdown of leguminous

material in this environment and benefits from it are largely associated with N supply. The

results from this experiment show that the use of legumes can overcome the N supply

deficiency in smallholder farmers cropping systems. Based on the experimental results, it

is concluded that the APSIM module could be helpful in predicting maize growth and yield

u n d e r d i f f e r e n t m a n a g e m e n t r e g i m e s i n L i m p o p o .

73

CHAPTER 5

Using closed pot incubations to investigate the N and C mineralization in crop residues of varying quality

5.1 Introduction

Most of the soils in smallholder farmer fields in Limpopo are infertile and crop yields tend

to be low. The use of organic inputs and mineral fertilisers has been shown to improve

crop yields (Jiri et al., 2004; Rivero et al., 2004). The application of compost and use of

legumes as manure and cover crops or as intercrops has also been shown to increase yields

(Armstrong et al., 1999; Chikowo et al., 2004; Whitbread and Ayisi, 2004). Crop yields

tend to be variable which is related to the variability in rainfall.

The results in chapter 4 show that the type, timing and method of application of mulch

affect the crop performance. The addition of fertiliser N improved maize performance

compared to when it was not applied which is an indication of the low fertility status of the

soil. When grass mulch was applied by itself, there was no improvement in crop DM as

compared to the control (0N). However, addition of N rich guarbean mulch gave a

significant improvement in maize DM relative to the control. Addition of mulch along with

30 kg N ha-1 increased maize DM by 24% (guarbean) to 36% (grass) as compared to the

treatment supplied with the same amount of N but no mulch. Even though these treatment

differences were statistically non-significant due to large background variation, it provides

some evidence for the importance of N x soil water interactions in affecting maize growth

in this environment. This could mean that when N fertiliser was added to the soil, it

overcame the immobilisation effect normally caused by high C:N ratio residues such as

grass and the yields improved due to the combined effect of reduced soil evaporation and

increased N availability. When N fertiliser was not added to grass mulch the maize crop

performance did not improve as compared to 0N (control); perhaps the benefits of reduced

soil evaporation from the grass mulch were offset by the N tie up because of high C:N ratio

grass residue. Previous studies have shown reduced crop growth from the application of

high C:N ratio residues due to reduced N availability (Ambus and Jensen, 1997; Aulakh et

al., 1991).

74

There was a large disparity in maize DM between the observed and predicted values for

the two different types of mulch. In the absence of N fertiliser, simulated maize DM under

grass mulch was two-fold greater than the observed crop growth (chapter 4). Where

guarbean mulch was used under 0N, differences between observed and simulated were

much smaller but the latter had higher maize DM. When N fertiliser was applied (30 kg ha-

1) there was a close agreement between the observed and simulated DM for grass mulch

but for guarbean mulch, the observed DM was considerably greater than the simulated

DM. These results clearly indicate that there is a knowledge gap in our understanding of N

and water dynamics in the soil. Therefore, two studies were undertaken under controlled

environment conditions to determine the influence of soil type (sandy or clay soil) and

method of residue application on the decomposition dynamics of different crop residues.

5.2 Materials and methods

Two separate incubation experiments were conducted at the CSIRO laboratories, Urrbrae,

Adelaide under controlled environmental conditions. A range of plant materials obtained

from common field crops or pastures were tested in a closed incubation system. In

experiment 1, the incubations tested the decomposition of 4 residues incorporated into 2

soil types. In experiment 2, the decomposition was tested on 3 types of residues with 2

methods of application on 1 soil type. The control treatments that contained no residues

were included in both experiments.

Soils. The 2 soils used in these experiments were Tarlee and Waikerie, collected from the

surface layers (0-10cm) of a red brown earth soil at Tarlee in the highly productive mid

north of South Australia and from a calcaresol at Waikerie in the low rainfall Murray

Mallee of South Australia, respectively. Based on particle size (Table 5.1), the surface

texture of the Tarlee soil is classified as a clay according to Hazelton and Murphy (2007),

had high field capacity, near neutral pH (7.13), contained more clay , organic C and N

content (Table 5.1). The Waikerie soil is classified as light loamy sand (Hazelton and

Murphy, 2007) and had an alkaline pH (7.9), contained more sand particles, less OC and

very low N content (Table 5.1).

75

Table 5.1. Properties of soils used in the incubation

Bulk

density

Field

capacity

Clay Silt Sand Organic

C

N

Soil type pH

(CaCl2)

g/cc mm/mm %

Tarlee 7.13 1.34 0.25 43.0 22.0 35.0 2.3 0.19

Waikerie 7.91 1.53 0.8 7.0 2.0 91.0 0.6 0.02

5.2.1 Experiment 1

The experiment was conducted in Urrbrae, CSIRO laboratory during July to October 2008

over a period of 14 weeks. The two soils, Tarlee and Waikerie were used in this

experiment. The soils were air-dried and sieved to pass through 2mm mesh. Samples of

500g of soils were weighed into polyethylene bags, wetted to 50% water holding capacity

and pre-incubated at 25oC for 2 days before amendment with residues. Pre-incubation is

considered necessary because sudden changes in environmental conditions such as

rewetting of soil and favourable temperature that enhances the occurrence of N and CO2

flushes are normally neutralised. Crop stubbles were canola (Brassica rapa), wheat

(Triticum aestivum), pea (Pisum sativum) and mucuna (Mucuna pruriens). They were

chopped to 4-5cm lengths, weighed to 3.8g/500g of soil which is equivalent to 10t ha-1 and

mixed into the soil Control treatments consisting of pure soils without plant material were

also included (TC and WC for Tarlee and Waikerie control treatments, respectively). Four

replicates of each treatment were set up. The C and N contents of plant residues are given

in Table 5.2.

5.2.2 Experiment 2

This incubation was conducted during November to March 2009. The soil used was

Waikerie, it was air-dried and sieved to pass through <2mm mesh. Samples of 500g of

soils were weighed into polyethylene bags, wetted to 50% water holding capacity and pre-

incubated at 25oC for 2 days at 50% field capacity before amendment with residues. Plant

residues were wheat (Triticum aestivum), pea (Pisum sativum) and mucuna (Mucuna

pruriens). They were chopped to 4-5cm lengths and weighed to 3.8g 500g-1 of soil which

is equivalent to 10t ha-1. Some of the plant residues treatments were applied as mulch and

76

others were incorporated into the soil. In this experiment the plant residue treatments

received extra 2ml H2O g-1 of residue during incubation. Control treatment consisting of

pure soil without plant residues was also included. Four replicates of each treatment were

set up. C and N contents of plant residues are given in table 5.2.

a. Incubation

The amount of water needed to bring the soil to 75% field capacity was added to the soils.

The plant residues were thoroughly mixed with the soils and transferred to the glass jars.

The base of each glass jar was tapped gently to allow the contents to settle. Each glass jar

had a vial containing NaOH for capturing CO2 and these were changed on a regular basis

(details in the next section). The jars were sealed and incubated at 25oC in the dark. For

the first experiment the incubation was done for 14 weeks and 17 weeks for the second

experiment. The reason for the differences in time was to allow the microbial activity to

take place upto a stage where no further C was mineralised. At 1, 2, 4, 8 weeks and on

completion of the incubation, soil samples were taken without disturbing the remaining

soils in the glass jars to measure soil mineral N, microbial N and microbial respiration.

b. Carbon analysis

A vial (No. 6) containing 25ml of 0.4 M NaOH solution was put in each glass jar to

capture carbon dioxide (CO2) emitted from soil respiration. Two glass jars without soil

containing only vials with NaOH were included as blanks for each sampling time. The

glass jars were incubated in a controlled incubation room at a constant temperature of 25oC

in the dark. For the 1st experiment, samples were collected on day 1, 2, 4, 7 for the first

week and on weekly basis thereafter for 14 weeks. In the 2nd experiment, samples were

collected on day 1, 2, 3, 4, 5, 6, 7, 8 9, 10, 12, 14, 16, 18, 21, 23, 25, 28, 30 and on weekly

basis thereafter until the mineralisation settled at week 17. The samples were precipitated

with BaCl2 and then titrated with hydrochloric acid (HCl) and the readings were used to

calculate the amount of cumulative C released as CO2. The cumulative C released as CO2

was calculated as :

Cumulative C loss = n∑RiTii=0

77

Where n is the number of incubation days, Ri is the mean respiration rate (g C day-1 g-1

soil) between two sampling dates, Ti is the days between two successive respiration

measurements (Liu et al., 2009).

c. Microbial C

Microbial C analysis followed the chloroform fumigation extraction method. For

fumigation, 11g and 12g of soil from each treatment for Waikerie and Tarlee soil was

weighed into the small glass jars during sampling, respectively. The jars (without lids)

containing soil samples were then placed in the desiccator containing chloroform and an

evacuation process was performed. The desiccator was placed in a 25ºC dark room for 7

days. After 7days the desiccator was removed from the dark room and evacuation was

performed again. The samples were extracted with 30ml of 0.5 M K2SO4, filtered through

No. 42 Whatman filter paper. The extracts were frozen pending analysis. The un-fumigated

samples were treated similar to the fumigated but the difference was that the procedure did

not use chloroform. For the 1st experiment, samples were collected at 2, 4 and 8 weeks and

for the 2nd experiment the samples were collected at 1, 2, 4, 8, 14 and 17 weeks . Microbial

biomass C was then determined following ninhydrin-reactive method by Joergensen and

Brookes (1990) and Sparling et al. (1993). Microbial biomass C was calculated as MB-C

(µg C/g dry soil) = flush of fumigation X kEC.

d. C turnover

Carbon turnover described as the efficiency of substrate utilisation for microbial growth

was expressed as the ratio of additional CO2-C plus microbial biomass C to residue C

(Chotte et al., 1998). This is calculated on a 0 to 1 scale where 0 is poor efficiency and 1 is

high efficiency (Chotte et al., 1998). The starting and ending soil C are the amount of C in

the 2 soils at the start and end of incubation after the addition of residues. The gain/loss is

the amount of C gained or lost in the controls and in soils amended with residues at the end

of incubation. The amount of C in residues is the C at the start of incubation. The

cumulative CO2-C is the amount of C released as CO2 from microbial activities at the end

of incubation. Residual CO2-C is calculated as the amount of C released as CO2 from

treatments amended with residues minus the amount in the control treatments, whereas

residual microbial biomass C is calculated as the amount of microbial biomass C from

residues minus that in the control. The C turnover is the total amount of C from residues

78

that is utilised efficiently through microbial activity [(additional CO2-C + additional MB-

C)/residue C].

e. Mineral nitrogen analysis

Soil samples of 20g were collected from each treatment and visible residue particles were

removed. Samples were collected at weeks 1, 2, 4, 8 and 14 for the 1st experiment whereas

for the 2nd experiment the last samples were collected at 17 weeks. The samples were

extracted with 60ml of potassium chloride (KCl) solution by shaking for 1 hour and

filtered through the No. 42 Whatman filter paper. The N was determined by segmented

flow colorimetry extraction following the 2M KCl extraction. Nitrate was dialysed,

reduced to nitrite by Cd reduction and the resultant nitrite reacted with N-1-

napthylethylenediamine dihydrochloride (NEDD) with sulphanilamide and NH4+ was

separated from interferences by gas diffusion and determined after reaction with sodium

salicylate and dichloro-isocyanurate (DCIC) (Rayment and Higgingson, 1992). The

extracts were analysed for NH+4 and NO-

3 using thermal conductivity detection by

Matejovic (1997).

f. Statistical analysis

Analyses of variance (ANOVA) were used to evaluate the statistical significance of the

treatment effects using JMP program. For CO2 , analysis of the cumulative CO2 at the

completion of the incubation was analysed at the 14 weeks for experiment 1 using two-way

ANOVA and at 17 weeks for experiment 2 using one-way ANOVA. The microbial C and

mineral N were analysed for each sampling date using two-way analysis of variance for

experiment 1 and one way analysis of variance for experiment 2. The treatment means in

all the analyses were compared by Tukey-Kramer HSD. Treatment means were declared

significant at P = 0.05 using the Turkey-Kramer HSD (honestly significant difference)

(JMP, 2005; SAS Institude Inc, 2005).

5.3 Results

The quantities of C, N and C:N ratio for the residues used in the two experiments as

presented in Table 5.2 show that canola had the highest C:N ratio followed by wheat,

mucuna then pea. Not surprisingly, the legumes pea and mucuna had higher %N than

wheat and canola (Table 5.2).

79

Table 5.2. Quantities of residue carbon, nitrogen and C:N ratio incorporated into soilsResidue %C %N C:N ratio

Canola 43 0.1 43.1

Mucuna 41 2.9 13.9

Pea 40 4.4 9.1

Wheat 43 1.6 26.0

5.3.1 Experiment 1

a. Carbon mineralisation

Over the entire period of the incubation, the C mineralisation was significantly higher in

the Tarlee soil than Waikerie soil with and without the amendment of plant residues

(Figure 5.4). The cumulative CO2-C measured at the end of the incubation (98 days) was

0.29 and 1.52mg CO2-C g-1 for Waikerie control and Tarlee control treatments,

respectively (Figure 5.1). During the 7 days of incubation, cumulative CO2-C release was

rapid in all treatments where residue was applied with the highest and significant release

occurring for both soils from the application of pea (Figure 5.1). The application of residue

showed significant differences in CO2-C release between the 2 soils throughout the

incubation (Figure 5.1).

At 21 days of incubation the CO2-C was significantly different in all other treatments and

soil types except for the application of mucuna and wheat in Tarlee soil (Figure 5.4). At 28

days pea and wheat showed no significant differences in CO2-C for the 2 soils (Figure 5.4).

At the end of incubation there was no significant difference in CO2-C with the application

of wheat, canola and pea to Tarlee soil (Figure 5.4). The CO2-C was increased significantly

in wheat and canola than pea for Waikerie soil (Figure 5.4). The rate of CO2-C release in

both soils decreased with time during incubation with mucuna releasing less CO2-C than

other residues (Figure 5.1).

80

Tarlee soil

0

1

2

3

4

0 10 20 30 40 50 60 70 80 90 100

ControlCanolaWheatPeaMucuna

Waikerie soil

0

1

2

3

4

0 10 20 30 40 50 60 70 80 90 100Incubation period (days)

Cum

mul

ativ

eC

min

eral

isat

ion

(mg

CO

2-C

g-1

soil)

I

I

Figure 5.1. Cumulative C mineralisation for Tarlee and Waikerie soils amended withwheat, canola, pea and mucuna. Vertical bars represent LSD at P = 0.05

While the absolute amount of CO2-C release from all treatments in the Tarlee soil were

greater than in the Waikerie soil, subtracting the CO2-C release of the control from the

residue treatments in each soil type showed that the amount of CO2-C released from the

residues was greater for Waikerie than Tarlee soil; however, there were no significant

differences (Table 5.3). At the end of incubation the CO2-C for wheat was high and

significantly different from pea and mucuna residues (Table 5.3). There were no significant

differences following the application of residues to the 2 soils.

81

b. Microbial biomass C

As with the CO2-C release, the microbial biomass C for Tarlee control was significantly

higher that of the Waikerie control treatment (Figure 5.2). For the Tarlee soil, microbial

biomass in all treatments increased rapidly during the 28 days of incubation after which it

declined whereas with the Waikerie soil, the biomass C peaked at 14 days and declined

thereafter (Figure 5.2). The differences in microbial biomass C for the Tarlee soil were not

significant following the application of residues (Figure 5.2). For the Waikerie soil, the pea

residue increased microbial biomass C significantly than all other treatments throughout

the 98 days of incubation (Figure 5.2). There were no significant differences in microbial

biomass C between canola, wheat and mucuna for individual soils; however, the

application of these residues to Tarlee soil showed significant increase in microbial

biomass C than when applied to Waikerie soil (Figure 5.2). The absolute value of the

microbial biomass C was significantly high for Tarlee soil treatments than Waikerie soil

treatments. When the microbial biomass C of the control soils was subtracted from each

soil type, the application of pea and canola resulted in significant differences between the

soils whereas wheat and mucuna showed no significant differences in the 2 soils types

(Table 5.3).

82

Tarlee soil

0.0

0.5

1.0

1.5

2.0

2.5

3.0 ControlCanolaWheatPeaMucuna

Waikerie soil

0.0

0.5

1.0

1.5

2.0

2.5

3.0

0 10 20 30 40 50 60 70 80 90 100Incubation period (days)

Mic

robi

al b

iom

ass

C (m

g C

g-1

soil)

I II

Figure 5.2. Microbial biomass C for Tarlee and Waikerie soils with and without the application of canola, wheat, pea and mucuna during the 98 day incubationperiod. Lsd bar represent the significant differences at P = 0.05

c. Carbon turnover

Table 5.3 present the amount of C that was efficiently utilised from residues applied to the

Tarlee and Waikerie soils. The results in Table 5.3 show that the amount of C lost (1.68

mg) from Tarlee soil was through CO2 evolution. With the Waikerie soil more C was lost

and disappeared as it could not be captured. The efficiency of utilising C from residues

differed with the type of residues applied as determined at the end of incubation (Table

5.3). The efficiency was about 0.68, 0.70, 0.65 and 0.52 for Tarlee soil, and 0.66, 0.72,

0.86 and 0.52 for Waikerie soil following the amendment of canola, wheat, pea and

mucuna, respectively (Table 5.3). The application of pea residue increased C turnover

significantly than canola and mucuna (Table 5.3). There were no significant differences

between wheat and pea. The least turnover was observed with the application of mucuna

83

(Table 5.3). The application of pea resulted in significant differences between Tarlee and

Waikerie soils. With the application of wheat, canola and mucuna to the 2 soils the C

turnover was not significantly different. Thus, the soil type did not have effect on residue

utilisation.

Table 5.3. Carbon turnover following the amendment of 4 residues and 2 soil types at the end of incubation.

T is for Tarlee soil and W for Waikerie soil. Means followed by the same letter are not

significantly different at P = 0.05aEnding soil C - starting soil C

bCumulative CO2 * (6/22)

cCumulative CO2-C in residue treatment - Cumulative CO2-C in the control treatment

dMicrobial biomass C of residue treatment – microbial biomass C of the control

e(Residual microbial biomass C + Residue CO2-C)/residue C

Treatment Starting

Soil C

Ending

Soil C

aC

Gain/

loss

Amount

of C in

residues

bCum.

CO2-

C

cResidual

CO2-C

Microbial

Biomass

C

dResidual

Biomass

C

eC

Turnover

Mg C g-1 soil

T_Soil 22.73 21.05 -1.68 0.00 1.52 0.00 1.21 0.00 0.00

W_Soil 3.34 2.62 -0.72 0.00 0.29 0.00 0.05 0.00 0.00

T_Canola 22.73 22.51 -0.22 3.19 3.37 1.85b 1.53 0.32b 0.68b

T_Wheat 22.73 22.27 -0.46 3.18 3.45 1.93ab 1.48 0.28bc 0.70b

T_Pea 22.73 22.35 -0.38 2.98 3.31 1.79bc 1.34 0.13cd 0.65b

T_Mucuna 22.73 23.18 0.45 3.04 2.98 1.46d 1.33 0.12d 0.52c

W_Canola 3.34 3.49 0.16 3.19 2.30 2.01ab 0.14 0.09d 0.66b

W_Wheat 3.34 3.29 -0.05 3.18 2.44 2.15a 0.18 0.13cd 0.72b

W_Pea 3.34 3.67 0.33 2.98 2.09 1.79bc 0.82 0.77a 0.86a

W_Mucuna 3.34 4.51 1.17 3.04 1.84 1.55cd 0.07 0.02d 0.52c

84

d. Total soil %C

The total %C in the soil measured at the end of the 98 day incubation period ranged from

0.26 to 2.36% with Tarlee soil showing greater %C than Waikerie soil (Figure 5.3). The

application of mulches in both soils show significant increase in %C compared to the

control treatment; however, there are no significant differences between treatments for

Tarlee soil (Figure 5.3). With the Waikerie soil, the application of mucuna increased %C

significantly than canola and wheat (Figure 5.3). The %C in pea, canola and wheat

mulches were not significantly different for Waikerie soil (Figure 5.3).

b a a a

ede de cd c

a

0.00

0.50

1.00

1.50

2.00

2.50

Tarlee soil Waikerie soil

%C

in th

e so

il

ControlCanolaWheatPeaMucuna

baa

aa

ccddedee

Figure 5.3. The percentage C for Tarlee and Waikerie residue treatments at the end of the incubation. Means followed by the same letter are not significantly different at P = 0.05

e. Mineral N

The mineral N in the 2 soils was initially dominated by the NH4+-N form which declined

within 14 days of incubation and remained low until the end of the incubation (Table 5.4).

The maximum values for NH4+-N concentrations for Tarlee and Waikerie soils observed at

7 days ranged between 0.4 and 102.1, and 0.3 and 131.2 mg NH4+-N kg-1 soil, respectively,

with pea mulch showing significantly high NH4+-N concentrations (102.1 mg NH4

+-N kg-1

soil) than other residues in the two soils (Table 5.4). There were no significant differences

in NH4+-N concentrations with the application of wheat, canola and mucuna (Table 5.4).

The significant NH4+-N concentrations occur in pea treatment applied to Waikerie soil

throughout the incubation period (Table 5.4). The concentrations of NH4+-N following the

application of canola, wheat and mucuna mulches in the 2 soils were not significantly

85

different throughout the incubation period except for the application of pea to Waikerie

soil (Table 5.4).

The NO3--N concentration is affected by the application of residues on the 2 soils. During

the first 7 days of incubation the NO3--N concentrations were low in the soils (Table 5.4).

The NO3--N concentrations were significantly different between the control treatments

with Tarlee soil showing higher concentrations than Waikerie soil (Table 5.4). The

application of pea and mucuna mulches increased soil NO3--N significantly more than

wheat and canola during the incubation period (Table 5.4). During the first 7 days of

incubation, the application of mucuna mulch increased NO3--N concentrations significantly

more in Tarlee than Waikerie soil while pea mulch resulted in low NO3--N concentrations

(Table 5.4). There were no significant differences in NO3--N concentrations for pea, canola

and wheat mulches following their incorporation into the two soils (Table 5.4). The

application of wheat and canola reduced the NO3--N significantly than the Tarlee control

treatment throughout the incubation period. Applying the same residues to Waikerie soil

did not show significant differences from the Waikerie control soil.

Table 5.4. Ammonium (NH4+-N) and nitrate (NO3

--N) concentrations in Tarlee (T) and Waikerie (W) soils after the incorporation of canola, wheat, pea and mucuna plant materials during the 98 days incubation period.

Treatments NH4+-N (mg kg-1) NO3

--N (mg kg-1)

7 14 28 56 98 7 14 28 56 98

T_Control 0.4d 0.7b 0.0b 0.4b 0.8b 47.2b 59.0b 86.2c 117.9c 159.5b

T_Canola 4.5cd 0.3b 0.4b 0.2b 0.4b 6.6de 8.1c 17.0d 64.4d 118.7d

T_Wheat 13.5cd 3.9b 0.0b 0.0b 0.5b 4.0de 15.4c 31.6d 71.5d 132.0cd

T_Pea 102.1b 6.8b 0.1b 0.0b 2.3b 11.4d 96.1a 148.7b 241.4a 219.4a

T_Mucuna 3.7cd 0.4b 0.2b 0.0b 0.0b 81.0a 101.4a 131.6b 199.8b 206.5a

W_Control 0.3d 0.3d 0.0b 0.8b 0.0b 9.3de 10.9c 14.0d 22.4e 27.6f

W_Canola 0.7cd 0.3b 0.0b 01.3b 0.0b 1.6e 0.2c 3.1d 6.1e 11.9f

W_Wheat 16.6c 0.1b 0.0b 1.3b 0.0b 3.8de 14.5c 18.3d 41.5de 58.2e

W_Pea 131.2a 52.1a 31.0a 36.2a 35.9a 0.3e 92.9a 220.1a 271.3a 138.9c

W_Mucuna 9.6cd 0.1b 0.2b 0.1b 0.1b 27.9c 55.1b 71.8c 120.1c 118.5d

Means followed by the same letter within columns are not significantly different according to the Tukey-Kramer HSD at (P = 0.05)

86

5.3.2 Experiment 2


As in experiment 1, the cumulative CO2-C released increased significantly with the

application of residues compared to the control treatment (Figure 5.4). The release was

initially more rapid for the residues that were incorporated than surface mulched residues

(Figure 5.4). At 7 days, CO2-C released was 0.78, 1.33 and 1.01 for incorporated residues,

and 0.81, 1.19 and 0.70 mg CO2-C g-1 soil for mulched residues in mucuna, pea and wheat,

respectively (Figure 5.4). The pea residue increased the CO2-C significantly more than

wheat and mucuna (Figure 5.4). The incorporation of pea into the soil increased CO2-C

significantly than when placed on the soil surface (Figure 5.4). There were no significant

differences in CO2-C between the 2 methods following the application of mucuna (Figure

5.4). When wheat was incorporated into the soil there was a significant increase in CO2-C

than when mulched (Figure 5.4).

At 21 and 28 days of incubation cumulative CO2-C release was significantly different

following the application of residues and methods of application (Figure 5.4). The CO2-C

release increased significantly for mulched pea than incorporated pea (Figure 5.4). The

CO2-C release in all residue treatments differed significantly between the 2 methods of

application, with mucuna releasing less CO2-C than pea and wheat (Figure 5.4). This was

also comparable with the control treatment. At the end of incubation the incorporated

wheat increased CO2-C significantly more than incorporated pea, whereas with mulched

wheat CO2-C was not significantly different from mulched pea (Figure 5.4). The CO2-C

release was not significantly different between the methods of wheat and mucuna

application (Figure 5.4). With mulched pea the CO2-C remained significantly higher than

incorporated pea (Figure 5.4). The CO2-C release for mucuna remained the lowest

throughout the incubation irrespective of whether it was incorporated or surface mulched

(Figure 5.4).

When the CO2-C from the control was subtracted from the residue treatments, there were

no significant differences in CO2-C between the 2 methods of application following the

addition of wheat and mucuna (Table 5.5). The mulched pea increased the CO2-C

significantly from the incorporated pea (Table 5.4).

87

Incorporated

0.0

0.5

1.0

1.5

2.0

2.5

3.0 ControlMucunaPeaWheat

Mulched

0.0

0.5

1.0

1.5

2.0

2.5

3.0

0 10 20 30 40 50 60 70 80 90 100 110 120

Incubation period (days)

Cum

ulat

ive

C m

iner

alis

atio

n(m

g C

O 2-C

g-1

soil)

I

I

Figure 5.4. Cumulative C mineralisation in incorporated and mulched wheat, mucuna and pea in Waikerie soil for 119 days. Vertical bars represent LSD at P = 0.05


As in experiment 1, the microbial biomass C in all treatments increased rapidly during the

14 days of incubation and declined thereafter (Figure 5.5). The decline was followed by

increase after 28 days, peaking at 98 days and declining to low levels thereafter (Figure

5.5). Like in the previous experiment, incorporated pea showed a significant increase in

microbial biomass C than mulched pea and all other treatments during the first 14 days of

incubation followed by a decrease at 28 days and increase thereafter (Figure 5.5). At 28

days of incubation the mulched pea showed a significant increase in the microbial biomass

C than incorporated pea (Figure 5.5). There were no significant differences in microbial

biomass C between wheat and mucuna for the 2 methods of application (Figure 5.5). These

treatments were significantly different from the control (Figure 5.5).

88

During 28 days of incubation microbial biomass C differed significantly with the

application of residues following the 2 methods of application (Figure 5.5). With the

incorporated pea, mulched mucuna and mulched wheat the microbial biomass C was not

significantly different from the control treatment (Figure 5.5). From 56 days of incubation

the incorporated pea showed increased microbial biomass C significantly from other

treatments (Figure 5.5). The application of pea and wheat showed significant differences in

microbial biomass C when incorporated than mulched (Figure 5.5). Significant differences

in microbial biomass C between incorporated and mulched mucuna were observed at 98

days of incubation (Figure 5.5). At the end of incubation, the incorporated pea showed

significant differences in microbial biomass C from mulched pea (Figure 5.5).

When the microbial biomass C of the control is subtracted from the residue treatments at

the end of incubation the incorporated pea increased the C significantly more than the

mulched pea (Table 5.5). The application of wheat and mucuna showed no significant

differences even with the 2 methods of application (Table 5.5).

0.00.20.40.60.81.01.21.41.61.8

0 10 20 30 40 50 60 70 80 90 100 110 120

Control IMucunaIPea IWheatMMucuna MPeaMWheat

Time (days)

Mic

robi

al b

iom

ass

C (m

g C

g-1

soil)

II I

I

II

Figure 5.5. Microbial biomass C for Waikerie soil with and without the applicationof wheat, pea and mucuna during the 119 day incubation period. Vertical barsrepresent LSD at P = 0.05

89

c. Carbon turnover

Table 5.5 presents the amount of C that was efficiently utilised from residues applied to

Waikerie soil. Incorporating residues in the soil resulted in C gain in the soil while

applying residues on the surface resulted in C loss (Table 5.5). The table showed that the

control soil without residue application lost C more than when residues were added (Table

5.5). The efficiency of utilising C derived from residues for microbial growth was

significantly different between residue treatments with the 2 methods of application (Table

5.5). The incorporation of pea and wheat into the soil increased the C turnover significantly

than when these residues were placed on the soil surface (Table 5.5). With the application

of mucuna, the C turnover was significantly higher when the residue was surface placed

than incorporated (Table 5.5). Thus the method of application had effects on residual C

utilisation.

Table 5.5. The C turnover for Waikerie soil using 3 types of residue and 2 methodsof residue application. I and M are, respectively incorporated and mulched residues. Means followed by the same letter are not significantly different at P = 0.05

Treatment Starting

Soil C

Ending

Soil C

aC

Gain/

loss

Amount

of C in

residues

bCum-

CO2-

C

cResidual

CO2-C

Microbial

Biomass

C

dResidue

Biomass

C

eC

Turnover

mg C g-1 soil

Soil 3.34 2.63 -0.70 0.00 0.37 0.00 0.07 0.00 0.00

I_mucuna 3.34 3.43 0.10 3.18 2.77 1.71c 0.14 0.07b 0.78b

I_pea 3.34 3.57 0.23 2.98 2.44 2.07b 0.78 0.71a 0.93a

I_wheat 3.34 4.23 0.90 3.04 2.08 2.40a 0.08 0.01b 0.57e

M_mucuna 3.34 2.87 -0.47 3.18 2.67 1.87c 0.07 0.00b 0.63d

M_pea 3.34 3.17 -0.17 2.98 2.73 2.36a 0.05 -0.02b 0.78b

M_wheat 3.34 2.80 -0.54 3.04 2.24 2.30ab 0.11 0.04b 0.73c

aEnding soil C - starting soil C

bCumulative CO2 * (6/22)

cCumulative CO2-C in residue treatment - Cumulative CO2-C in the control treatment

dMicrobial biomass C of residue treatment – microbial biomass C of the control

e(Residue microbial biomass C + Residue CO2-C)/residue C

Note: Microbial Biomass C could be derived from native SOM and addied residues

90

d. Total soil C (%)

The total %C measured in the soil at the end of the experiment ranged between 0.26 and

0.41 %C in the incorporated and between 0.26 and 0.32 in mulched treatments, with more

%C measured in incorporated than mulched treatments (Figure 5.6). When mucuna and

wheat were incorporated into the soil the %C increased significantly more than when

mulched (Figure 5.6). With the pea residue, the %C was not significantly different between

the 2 methods of application (Figure 5.6).

0.000.050.100.150.200.250.300.350.400.450.50

Incorporated Mulched

Method of residue application

%C

in th

e so

il re

sidu

es

ControlMucunaPeaWheat

c

a

b b

cc

bc

c

Figure 5.6. The percentage C remaining for incorporated and mulched residues inWaikerie soil at the end of incubation. Means followed by the same letter are not significantly different at P = 0.05

e. Mineral N

The dynamics of mineral N is similar to the first experiment in which the NO3--N is the

dominant form of mineral N over the majority of the 119 day incubation period except at 7

days where NH4+-N exceeds NO3

--N (Table 5.6). The mineral N (NH4+-N and NO3

--N) of

plant residues was comparable across the 2 application methods. The pea treatment

increased NH4+-N concentration significantly more than mucuna, wheat and the control

treatments (Table 5.6). The NH4+-N concentration in pea was not significantly different

following the 2 methods of application at 7 days of incubation (Table 5.6). The NH4+-N in

incorporated pea remains highly significant for mulched pea from day 56 to the end of the

91

incubation period (Table 5.6). There were no significant differences in NH4+-N between

other treatments throughout the incubation (Table 5.6).

Table 5.6. Ammonium and nitrate-N (mg N kg-1) for the incorporation and mulch ofthe different plant materials in Waikerie soil.

I_ and M_ represent incorporated and mulched mucuna, pea and wheat in Waikerie soil, respectively. Control is the soil without residue applications. Means followed by the same letter within columns are not significantly different according to Tukey test at P = 0.05.

The application of pea shows high NO3--N concentration throughout the 112 day

incubation period (Table 5.6). At 7 days of incubation the NO3--N concentration with pea

and mucuna residues was significantly greater than wheat and the control treatments (Table

5.6). There were no significant differences with regard to the 2 methods of residue

application (Table 5.6). From 14 days until the end of incubation the NO3--N concentration

for pea treatment was significantly different in the 2 methods of application (Table 5.6).

With the application of mucuna and wheat the NO3--N concentration was not significantly

different even with the method of application (Table 5.5). There were no significant

Treatments Incubation time (days)

NH4+-N (mg kg-1)

7 14 28 56 98 119

Control 1.4b 1.3c 0.0b 1.4b 2.7b 0.8bI_mucuna 2.7b 0.8c 0.4b 0.4b 0.0d 0.7b

I_pea 111.0a 34.2b 28.1a 25.7a 28.6a 36.0aI_wheat 1.2b 0.0c 0.0b 0.1b 0.0b 0.1b

M_mucuna 13.4b 1.2c 2.4b 0.6b 0.0b 0.0bM_pea 124.1a 69.5a 29.7a 4.6b 0.5b 2.5b

M_wheat 3.3b 0.4c 0.5b 0.9b 0.0b 0.9b

NO3--N (mg kg-1)

7 14 28 56 98 119

Control 10.4b 11.4d 14.3e 17.2d 22.6d 26.1d

I_mucuna 39.5a 43.8c 81.6cd 102.8c 133.5c 131.9cI_pea 50.5a 238.8a 259.4a 263.9a 263.1a 290.5a

I_wheat 17.0b 16.3d 30.5de 48.7d 63.0d 67.5dM_mucuna 42.6a 63.0c 86.2c 113.9c 133.4c 135.1c

M_pea 51.8a 143.2b 178.1b 186.2b 218.6b 211.8bM_wheat 15.3b 31.2cd 34.0cde 41.1d 52.7d 60.6d

92

differences among the control, incorporated and mulched wheat throughout the incubation

(Table 5.6).

5.4 Discussion

5.4.1 Experiment 1


The C mineralisation in fertile soils is generally higher than in less fertile soil due to the

high organic matter content contained in the former. In this experiment the increased C

mineralisation for the Tarlee than Waikerie soil could be attributed to its high organic C

and higher clay content as compared to the Waikerie soil (Table 5.1). Similar results were

obtained by Parfitt and Salt (2001) and Franzluebbers et al. (1995) who found that clay soil

fractions had high C mineralisation compared to sandy fractions, attributing this to the

availability of decomposable soil organic matter held in clay fractions. Organic matter in

clay fractions has been found to be mostly protected from microbial activity and as such,

soil mixing activities such as tillage cause physical disruption of soil aggregates and results

in the exposure of microsites where organic matter was previously inaccessible to microbes

resulting in increased C mineralisation (Mason, 1976).

The application of mulches to Tarlee and Waikerie soils led to a significant increase in

CO2 evolution (Figure 5.1). This is caused by increased microbial activities during the

utilisation of water soluble C compounds and the breakdown of complex compounds into

simple soluble molecules (Mason, 1976). The C mineralisation was rapid during the first

two weeks of incubation and slowed down as decomposition progressed (Figure 5.1).

During the initial application of residues to soil, soluble low molecular weight substances

such as glucose and amino acids are rapidly attacked by microorganisms, whereas

insoluble polymeric materials tend to be cleaved primarily by slow growing

microorganisms (Giller and Wilson, 1991). The observed pattern of CO2 release was

consistent with findings by Franzluebbers et al. (1995) who found C loss from cowpea

leaves to be rapid within 15 days of incorporation in soils with different levels of microbial

biomass. Mubarak et al. (2002) reported rapid C loss (50%) in maize and groundnut within

2 and 3 weeks of incubation, respectively. The rapid increase in CO2 evolution is attributed

to rapid catabolism of simple soluble C compounds present in residues (Sign, 1995).

93

Greater increase in CO2 release in the two soils following the application of pea residue

than other treatments could be attributed to its lower C:N ratio (Table 5.2) which enhances

microbial activity during decomposition. The CO2 evolution for mucuna treatment was

initially higher than wheat and canola but was lower at the end of incubation (Figure 5.1).

As mucuna and pea had high N content and low C:N ratio (Table 5.2), the two were to be

expected to have similar CO2 release but mucuna released CO2 slower than pea (Figure

5.1). This could be attributed to higher phenolic content of mucuna residues. Ver Elst and

Pieterse (2006) found that CO2 release increased significantly with application of legumes

containing low lignin and polyphenol.

The higher amount of CO2 release measured at the end of incubation with the application

of wheat and canola than pea and mucuna was due to unusually low C:N ratio (26) of

wheat straw (Table 5.2). This straw originated from a glasshouse experiment in which the

wheat was well fertilised and watered. This was similar to the study conducted by Jawson

et al. (1989) where straw with a C:N ratio of 14:1 resulted from high fertilisation during its

growth and being harvested at incomplete maturation.

The CO2 evolution after the application of canola was initially lower than other treatments

in the two soils but was higher than pea and mucuna at the end of incubation (Figure 5.1).

This could be attributed to its high C:N ratio (Table 5.2). When a residue having a high

C:N ratio such as canola straw is added to soil, there is a sudden increase in the evolution

of CO2 due to increased microbial activity, which is accompanied by the depression in soil

nitrates (Prasad and Power, 1997). This was in similar to the findings by Jingguo and

Bakken (1997a) where the application of clover increased C mineralisation significantly

than straw within 3 weeks of incorporation, attributing this to the high C:N ratio (82) of the

straw material. This was supported by Mubarak et al. (2002) who observed rapid C

mineralisation in groundnut (C:N ratio 26.9) than maize (C:N ratio 40.6).


The microbial C in the Tarlee control was higher and significantly different from the

Waikerie control soil (Figure 5.2) which could be attributed to high clay content (43%),

organic C content and %N, and the near neutral pH (7.3) of the Tarlee soil (Table 5.1). The

application of residues increased microbial biomass C significantly during the 98 days of

94

incubation compared to the control treatments. The significant increase in microbial

biomass C with the application of pea residues than canola, wheat, mucuna and the control

in the 2 soils throughout the incubation period could be attributable to the low C:N ratio in

pea (Table 5.2). Leblanc et al. (2006) associated the increased microbial biomass C with

the rapid decay of hemicellulose and cellulose in residues.

There were no significant differences in microbial biomass C between canola, wheat and

mucuna for the individual soils; however, their application resulted in significant

differences in microbial biomass C between the 2 soils. The lack of significant differences

between the 3 treatments could be attributed to the unusually low C:N ratio and high N %

contained in wheat residues (Table 5.2). Although mucuna is a legume and has low C: N

ratio, it tends to decompose slowly which could mean that it has high polyphenol content.

Canola, a high C:N ratio residue, releases C at a moderate rate which suggests that it has

more water soluble C and no polyphenol contents.

With the application of canola and wheat residues (non-legumes) the microbial biomass C

was significantly different from the pea, mucuna and control treatments. This could be due

to high C:N ratio in canola and wheat compared to pea and mucuna. Sign (1995) observed

significant increase in microbial biomass C with the application of straw residues

compared to the control treatment. The 'priming effect' caused by the added crop residue

carbon is recognised however the use of unlabelled crop residues does not allow the

separation of contribution from SOM and crop residues. Chotte et al. (1998) reported, from

experiments using a vertisol similar to that from Tarlee, that the majority of the priming

effect on CO2 had generally disappeared after 3d of incubation. The priming effect on

microbial biomass newly developed from SOM was a small proportion of the total increase

in MB (20-30%) only. The two sources of C for any priming effect in CO2 and MB-C

would originate both from native MB and SOM.

c. C turnover

The efficiency of utilising the residues differed slightly among residues. The differences in

the efficiency of utilisation for growth by microbes may have resulted from the differences

in the quality of residues. When same residues were applied to Tarlee and Waikerie soils,

the differences could be attributed to the soil properties. A soil containing high clay

content such as Tarlee tends to protect organic matter from further degradation. This was

consistent with the results obtained by Gregorich et al. (1991) using clay and sandy soils.

95

The ability to recover C lost from the Tarlee soil as CO2-C shows that the efficiency of C

utilisation was low as compared to Waikerie soil where the amount of C lost was not

recovered as CO2 (Table 5.3) suggesting that the efficiency of C utilisation for the

Waikerie soil was abnormally high and could have resulted from the initial microbial

uptake without metabolism. According to the explanation by Chotte et al. (1998) using

labelled substrates on a Vertisol, some substrate 14C becomes rapidly incorporated into the

cell constituents without net cell growth, accompanied by some decomposition of replaced 12C constituents and decreases in measured biolmass. This could be supported by the rapid

increase in microbial biomass C during the first 14 days of incubation. Similar findings by

van Veen et al. (1985) show that a small fraction of glucose 14C was evolved as CO2

whereas most of the added glucose C disappeared in the soil. This was supported by the

work conducted by Gregorich et al. (1991) and Voroney and Paul (1984) who reported that

> 90% of glucose C that disappeared within 1 day had been transformed into biomass and

other metabolites.

The non significant differences with the application of individual residues to the 2 soils

could mean that less C was utilised from residues applied to Tarlee soil. Chotte et al.

(1998) explained that newly formed biomass 14C did not equilibrate with indigenous

biomass C resulting in the latter being unprotected and converted to CO2- C. This greater

conversion of indigenous biomass C into CO2- C after the addition of residues to Tarlee

soil could be attributed to higher death of biomass 12C other than predation resulting in

biomass 12C conversion to CO212C, whereas biomass 14C was not affected (Chotte et al.,

1998). Additionally, as the jars were kept close with optimum moisture and temperature

maintained, the microbial activity could have been maximised. The same scenario was

observed by van Veen et al. (1985) where continuous moist conditions showed similar

decline in biomass C for clay and sandy soils. Lower C turnover for mucuna could be due

to its greater polyphenol content which could be inhibitory to microbial activity (Muller et

al., 2003; Palm et al., 2001). With the wheat residue resulting in similar turnover to

mucuna and canola could be due to the low C:N ratio of this residue.

d. Total soil C (%)

The % C was determined from the concentration of C from the control and plant residues

remaining in the soil at the end of incubation. The increased soil C in the Tarlee than

Waikerie soil could be due to greater organic C contained in the Tarlee soil. Polglase et al.

96

(2000) indicated that the formation of organo-mineral complexes in clay soil protect C

from microbial oxidation. The significant difference in total soil C between residue

treatments applied to Tarlee and Waikerie soils could be due to the high amount of organic

C contained in the former than latter soil (Table 5.1). In the 2 soils, mucuna had shown to

increase soil C than pea, canola and wheat. The same scenario was observed by Blair et al.

(2005) who reported increased soil C with the application of flemingia than medic and rice

straw. The increased soil C with the application of low C:N ratio could mean that the

residues are resistant to decomposition due to high cellulose, hemi-cellulose and lignin

content (Ver Elst and Pieterse, 2006). The non significant difference in soil C between

wheat, canola and pea when applied to Waikerie soil could be due to lower than usual C:N

ratio in wheat which could mean that wheat decomposed at the same rate as the other

residues.

e. Mineral N

The initially high concentrations of NH4+-N with the application of plant material could be

due to microbial decomposition of nitrogenous organic residues into NH4+-N (Haynes et

al., 1986). Greater increase in NH4+ compared to NO3

--N with the application of pea and

mucuna than wheat and canola could be attributable to the low C:N ratio and high N

content in pea and mucuna (Table 5.2). This was similar to the findings by Mubarak et al.

(2002) and Tian et al. (1992). In the initial stages of decomposition, ammonification

exceeds nitrification and this shows a short term accumulation of NH4+ resulting in the

absence of NH4+ thereafter (Costa et al., 1990). The formation of NH4

+ during the initial

application of crop residues to soil as described by Prasad and Power (1997) occurs from

the activity of bacteria, fungi and actinomycetes by breaking down complex organic

molecules releasing amines and amino acids which are then reacted upon by other

heterotrophs which release N in the inorganic NH4+ form. If NH4

+-N is not utilised, it is

further and rapidly oxidised into NO3--N by nitrifying bacteria and becomes dominant

throughout the remaining part of the incubation period.

The rapid decline of NH4+-N after the first 7 days of incubation could be due to the

microbial preferential use of NH4+-N over NO3

--N when both nutrients are available

(Jawson et al., 1989). This was similar to the results obtained in several studies (Green and

Blackmer, 1995; Jingguo and Bakken, 1997a; Thonnissen et al., 2000) where green manure

application increased soil NH4+-N significantly in the first week after application but

97

declining rapidly within 3 weeks. Prasad and Power (1997) explained that the quantity of

energy (soluble C) needed to incorporate one unit of N into plant protein is greater with

nitrate than with ammonium.

The NH4+ and NO3

- N concentrations between pea and mucuna treatments which had

nearly the same low C:N ratio were significantly different (Table 5.4). This result was

similar to the findings by Fosu et al., (2007) who observed differences in mineral N with

the application of sunhemp, mucuna, devil bean and calopo. The differences in mineral N

between pea and mucuna could be due to factors other than the C:N ratio such as cellulose,

lignin, polyphenols and tannins. Fosu et al. (2007) found that high quality residues

containing high amount of cellulose and lignin such as mucuna and devil bean released N

slowly than those which had less cellulose and lignin. This was supported by Palm and

Sanchez (1991) who observed immobilisation of high amount of N for material containing

high % N and high polyphenolic concentration. According to Palm et al. (2001a) the

organic materials of poor quality tend to release a smaller proportion of their N at a slow

continuous rate without a period of rapid mineralisation at a later stage. The increased

mineralisation of N for pea was similar to the findings by Ver Elst and Pieterse (2006) and

Mendham et al. (2004) who reported increased N mineralisation with the application of

high quality residues containing low hemi-cellulose + cellulose, polyphenol and lignin %.

The application of wheat and canola reduced the NO3--N less than the control treatments in

the 2 soils throughout the incubation period (Table 5.4). This was consistent with the work

done by Jingguo and Bakken (1997a) where the nitrate concentration was constantly low

with the application of straw. The low concentrations of NO3--N in the soil with the

application of wheat and canola occur as a result of microbial biomass incorporating

mineral N from the surrounding soil and litter during decomposition of organic residues

with a wide C:N ratio (Haynes et al., 1986). There were no significant differences between

wheat, canola and the Waikerie control soil.

The application of wheat and canola increased NO3--N in Tarlee than Waikerie soil at the

end of incubation (Table 5.4). This could be attributed to the high clay concentration of the

Tarlee soil which protects microorganisms from breaking down organic matter added to

the soil. This was similar to the study conducted by Christensen and Olesen (1998) in

98

which straw addition increased soil N content more in clay than sandy soil particulates as

more N was immobilised in sandy soil. The low concentrations of NO3--N in the soil with

the application of wheat and canola occur as a result of microbial biomass incorporating

mineral N from the surrounding soil and litter during decomposition of organic residues

with a wide C:N ratio (Haynes et al., 1986).

The NO3--N concentrations were significantly different between the control treatments

with Tarlee soil showing higher concentrations than Waikerie soil (Table 5.4). In soils with

high clay content, positively charged NH4+ ion is usually held by negatively charged soil

colloids or fixed by clay minerals, so any activity that involves the breakdown of colloids

expose NH4+-N to oxidation into NO3

--N by microorganisms . The clay content in soil is

known to protect organic matter from microbial activity and any soil disturbing process

will expose the OM for further decomposition. The high amount of NO3- -N in Tarlee soil

is attributed to the release of N during the decomposition of organic matter that was

contained in this soil and was previously inaccessible by microorganisms.

The application of pea and mucuna mulches increased soil NO3--N significantly over the

concentrations found in the wheat and canola treatment during the incubation. This was

similar to the results obtained by Jingguo and Bakken (1997a) in which clover material

showed greater nitrate accumulation in the soil than straw. This was supported by

Mubarak et al. (2002) in whose study groundnut and maize residues contained 3.6 and 14%

on initial N at the end of the incubation, indicating that more N was lost to the soil. The

rapid increase in NO3--N from pea application is expected from the leaching of high water

soluble N contained in the mulch (Jawson et al., 1989; Jensen, 1994).

Other factors that could contribute to the N mineralisation include the pH of the soil and

concentration in microbial community. Decomposition was found typically to proceed

more readily in neutral than in acid soils as most microbial biomass becomes inactive in

acid soils (Haynes, 1986). The soils used in this experiment were alkaline (pH 7.3 and 7.9)

which increased the microbial process of nitrifying NH4+ to NO3

-. According to Prasad and

Power (1997), NH4+ in soils having pH 6.93 to 7.85 is oxidised to NO2

- which accumulates

for extended periods before being oxidised to NO3-. Bacteria and actinomycetes often

dominate in neutral and alkaline conditions, while fungi are more active under acid

99

conditions (Prasad and Power, 1997). The greater tolerance of mineralisation to low pH

than nitrification is reflected in the findings that ammonium is generally the dominant form

of N in acidic soils while nitrate predominates in non acidic soils (Haynes et al., 1986).

5.4.2 Experiment 2


The increased CO2 release with the incorporated than mulched residues could be attributed

to the closer contact with soil of the former. This was similar to the findings by Aulakh et

al. (1991) who reported increased CO2 in incorporated than surface placed residues. The

same scenario was observed by Thonnissen et al. (2000) in which incorporated green

manure decomposed significantly faster than mulched green manure. At the end of

incubation the amount of C mineralised in mucuna, pea and wheat was 2.08, 2.44 and 2.77

mg C g-1 soil for incorporated treatments, and 2.24, 2.73 and 2.67 mg C g-1 soil for

mulched treatments, respectively. The CO2 increased significantly with the application of

wheat more than with pea and mucuna mulches.

However, when the CO2 from the control treatment was subtracted from that which

evolved from the mulches, there were no significant differences between incorporated and

surface placed residues. The lack of significant difference in CO2 which evolved from the

residues between the methods of application could be attributed to the optimum moisture

and temperature maintained in the humid aerobic conditions of sealed incubation jars. This

was consistent with the results obtained by Aulakh et al. (1991) where C from incorporated

and surface placed residues was very similar at optimum moisture content.


The application of residues affected microbial biomass C during the incubation period (119

days) in general. The significant increase in CO2 evolution for incorporated than mulched

pea could be due to direct contact of residues with the soil which creates easy access by

microorganisms. The lack of significant difference between incorporated and mulched

wheat could be attributed to the optimum moisture and temperature conditions of the

incubation environment. The significant increase in microbial biomass C for pea residues

could be due to its low C:N content which makes it easily utilisable by microbes. The

ability of the wheat mulch to increase microbial biomass C similar to mucuna could be

100

attributed to the high amount of water soluble compounds contained in wheat mulch and

the high polyphenol content of mucuna mulch.

At the end of the incubation period, the application of incorporated pea still increased the

microbial biomass C significantly from all other treatments. There were no significant

differences among the rest of the treatments including the control. The reduction in

microbial biomass C in all treatments after 14 days of incubation is attributed to survival of

bacteria due to energy availability (Jingguo and Bakken, 1997b). In situations where N is

limited such as with the application of wheat straw, the bacterial counts and viability

become less as compared to when N rich plant materials are applied, resulting in low

production of microbial biomass C. Luscombe and Gray (1974) and Jingguo and Bakken

(1997b) indicated that in N limited soils, C becomes abundant and as such microorganisms

produce extracellular enzymes to utilise C polymers (resistant C compounds).

c. C turnover

The efficiency of residue utilisation at the end of the incubation period differed

significantly between the residues applied and between the methods of application-

incorporated or mulched. The high efficient utilisation of C in pea residue than wheat and

mucuna could be due to the high % N and low C:N ratio in this residue (Table 5.2). With

the application of wheat residue, the higher utilisation efficiency than mucuna could be due

to the low C:N ratio in wheat (Table 5.2). As with the previous experiment mucuna

showed less C turnover than pea of which was not expected because the 2 residues had

nearly the same similar N content and C:N ratio. This could be attributed to secondary

metabolites such as high lignin and polyphenol contents other than the C:N ratio. Although

mucuna has a low C:N ratio it is regarded as having low plant residue quality index (Tian

et al., 1995) because of the high polyphenol contents (Palm et al., 2001). Polyphenols

make plant material less palatable to microorganisms (Mason, 1976) thus when residue

containing more polyphenols is applied to the soil, it will be broken down at a slower rate.

When crop residues are incorporated in the soil direct contact is increased than when

placed on the soil surface resulting in high C turnover.

d. Total soil C

As in the 1st experiment, mucuna increased total soil C more than other residues when

incorporated in the Waikerie soil. This could be due to the secondary metabolites contained

101

in mucuna which render it resistant to degradation (Muller et al., 2003). In this experiment,

the method of residue placement showed a significant effect on soil C with greater increase

in the incorporated than mulched residues with the exception of pea. The non significant

difference in soil C between mulched residues could be due to high amounts of

polyphenols in mucuna suggesting that it decomposed at a slower rate (Tian et al. 1995).

With the wheat mulch which is a high C:N ratio residue (Table 5.2), this could mean that

most of the C in the wheat mulch was water soluble. According to Mason (1976) material

containing more polyphenol take longer time to decompose while on the other hand

decomposition occurs more rapidly in material with low C:N ratio.

e. Mineral N

In general and similar to the 1st experiment, the application of crop residues affected the

ammonium (NH4+-N) and nitrate (NO3

--N) concentrations in the soil. In this experiment

the significance of residue application differed with time. The increased significant

difference in NO3--N with the incorporated pea than mulched pea could be due to the direct

contact of residue with the soil and N volatilisation of surface placed pea residue. This was

similar to the results obtained by Aulakh et al. (1991) who reported significant differences

following the application of vetch with low C:N ratio. This was supported by the results

obtained by Costa et al. (1990) in which N increased more in incorporated than surface

placed residues.

The non significant difference in mineral N with the application of wheat between the

incorporated and mulched treatments could be due to the optimum moisture and

temperature conditions of the incubation environment. This was similar to the findings by

Aulakh et al. (1991) who reported immobilisation of N in incorporated and less change in

mineral N with the surface placed wheat, soybean and corn which had wide C:N ratio.

5.5 Conclusions

The results of this experiment show that under favourable conditions of adequate moisture

content and optimum temperature the decomposition of high quality residues is extremely

rapid resulting in the release of high mineral N concentrations. When residues are left at

the soil surface N supply can limit decomposition and incorporation of residues into the

soil greatly increase the decomposition rate. Incorporating low C:N ratio residue or leaving

it on the soil surface can release sufficient mineral N to synchronise N supply with crop

102

demand at the early stages of growth. However, incorporating or placing residue with high

C:N ratio can immobilise mineral N for a long period.

The use of high quality residues that decompose and release N slowly will help to reduce

the mineralisation of N. Good management of residues such as mixing high and low

quality residues or applying high quality residues during plant growth could be an option

for good utilisation of high quality residue derived N. The quality (C:N) of residues

primarily controls decomposition rate; however, the soil type and method of residue

application contribute to the rate of decomposition. In sandy soils the N release happens so

quickly resulting in loss of N which render N unavailable during times of high crop

demand. As the soils in smallholder farmer fields in Limpopo are mostly sandy and

infertile, addition of low C:N ratio residues that will decompose slowly like mucuna, can

help to improve the fertility of the soil.

103

Chapter 6

General discussions

Poor crop yield is a major problem in smallholder farmer fields in Limpopo. The use of

organic material such as compost, farmyard manure and legume residues, and the

application of N and P fertilisers have been shown to improve crop yield in situations

where soil fertility was poor. The main factor governing the fertility of the soil is the

amount of organic matter in the soil which is regarded as the main plant nutrient source.

Many soils which are regarded as poor contain low organic matter and nutrient content.

The organic matter content of the soil depends mostly on the amount of organic residue

input and decomposability of plant residues, which is affected by plant residue quality and

environmental factors such as rainfall and temperature. The issues related to the type of

plant residues and method of application of these residues and the dynamics of N, C and

soil water are addressed in this study.

6.1 The socio-economic and farming details of subsistence

farmers in Limpopo

The subsistence farmers in Limpopo are mostly female, elderly, less educated and

unemployed, usually with large families that need to be supported. Growing cereals and

legumes in intercrops is their main approach for achieving sufficient food supply; however,

the world food self-sufficiency level of 200 kg per adult per year is often not reached

because of low crop yields. The low level of N in the soils and their sandy texture are the

main causes of low crop yields. Integrating legumes into the cropping systems has been

shown to increase soil N and crop yields; however, in Limpopo, smallholder farmers grow

legumes for home consumption and livestock feed on crop residues after harvest which

reduces their benefit for soil fertility. This practice could be related to the lack of farmer

knowledge about the potential benefits of crop residues on soil C and N and subsequent

crop yields; however, farmers tend to put more value on livestock than improved soil

health and fertility. As inorganic fertilisers are beyond the reach of most smallholder

farmers, there is a serious need for an extension program aimed at improving soil health

and fertility through the integration of legumes into these cropping systems. This study has

clearly shown that the farming population of Limpopo is quite old (with 60 as the mean

age) which may impose some constraints to the adoption of new agricultural technologies.

104

This ageing trend of smallholder farmers is not sustainable and policy makers need to

explore how agriculture could be made more attractive to local youth.

Smallholder farmers in Limpopo access most of their information on new agricultural

technology from public sector extension services. Serious lack of knowledge of local

farmers about the benefits of legumes in terms of biological N fixation identified in this

study may suggest poor knowledge of this topic among local extension personnel. It could

be argued that there needs to be a ‘train the trainers’ program on biological N fixation

which would then be expected to lead to greater adoption of legumes in local cropping

systems.

6.2 The effects of fertiliser, legumes and grass mulches

applied to maize

The application of fertiliser and mulches had significant effects on maize growth at

Gabaza. Use of grass mulch only improved maize growth when it was applied in

combination with N fertiliser. As stated earlier, smallholder farmers in region cannot afford

fertilisers; therefore, grass mulch option may not be suitable for them even though it could

have benefits in reducing soil evaporation and suppression of weeds. Guarbean mulch on

the other hand, was found to improve maize growth similar to 15 kg N ha-1 applied as

fertiliser. Therefore, guarbean mulch is an affordable technology for these farmers

provided the mulch can be retained on the fields and not grazed by their own or communal

livestock. Those farmers who can afford to add low levels of synthetic N fertilisers could

achieve additional yield benefits by using N fertiliser in conjunction with guarbean mulch.

Due to various constraints, the field experiment undertaken as part of this project could not

be taken through to harvest (chapter 4). Future research should be aimed at investigating

maize response to different types of legume mulches used on their own or in combination

with N fertiliser. It may also be worthwhile to explore crop responses to grass-legume

mixture mulches as each component has its own strength with legume contributing to N

supply and grass providing greater benefits of reduced soil evaporation.

Simulation of maize shoot growth and grain yield for the field experiment enabled

assessment of model performance in this environment. APSIM provided reasonable

prediction of some of the treatments but was particularly inaccurate in predicting maize

growth in the grass mulch treatment. Predicted maize DM in grass mulch was much higher

105

than the actual which could be related to inaccurate prediction of available soil N. It seems

there is a need for improvement in the estimation of available soil N in the model for these

infertile sandy soils. If this follow-up research activity was undertaken, it could improve

the ability of APSIM to predict maize growth in Limpopo.

Simulating the N and water dynamics for the treatments in this study showed that the grass

mulch increased N stress even when N fertiliser was added to the mulch but water stress

was greatly reduced. With the application of guarbean and N fertiliser, N stress was greatly

reduced but water stress during reproductive development of the crop was increased as

more soil water evaporated. Thus, crops suffered more water stress than N stress.

Therefore, this could mean that no single mulch type was able to fulfil all the growth

requirements of maize crop in full. The results raise the possibility of using grass-legume

mulch combinations to reduce water and N stress but further research would be required to

validate this in the field.

Even though this project has identified benefits of using legume mulch to improve maize

yields in Limpopo, adoption of this practice has social dimensions as well. At present

farmers and other members of the village graze the fields with their animals after crop

harvest. It would be necessary to undertake a careful assessment of community attitudes to

changes in communal access to farmers land to grazing animals.

6.3 N and C mineralization in crop residues of varying quality

When crop residues of varying quality [canola (C:N ratio 43.1), wheat (C:N ratio 26.0),

pea (C:N ratio 9.1) and mucuna (C:N ratio 13.9)] were added to the soil, the N and C

mineralisation differed between the Tarlee (clay) and Waikerie (loamy sand) soils. The

application of pea residues increased C mineralisation rapidly during the first 7 days of

incubation compared to wheat, canola, with mucuna showing slow C mineralisation;

however, at the end of incubation, more C mineralisation occurred in wheat and canola

than pea and mucuna. These results suggest that plant residues of low C:N ratio mineralise

rapidly and disappear in the soil during the early stages of crop growth whereas those of

high C:N ratio mineralise slowly and stay in the soil for a longer period (Muller et al.,

2003). Thus, crop residues with low C:N ratio but mineralise slowly such as mucuna (due

to higher phenolic content) are more suitable for sandy soils of smallholders in Limpopo

(Wortmann and McIntyre, 2000). Mineralisation of N from residues such as mucuna is

106

expected to occur more slowly and less likely to leach out of the root zone on sandy soils.

Such residues are also expected to remain as mulch on the soil surface for longer and

provide benefits in terms of reduced soil evaporation.

Raising awareness of smallholder farmers about the beneficial effects of legumes in

supplying N to the soil and ultimately improving crop growth could change smallholder

farmers approach to residue management. The use of legumes tends to increase soil N

whereas grass mulch can increase C content of the soil and reduce crop water stress during

reproductive development mainly due to reduced soil evaporation. The use of slow

decomposing legumes such as mucuna could provide N in synchrony with crop demand

unlike legumes such as peas that decompose rapidly and release N early in crop’s life when

demand for N is still low. It is generally assumed that symbiosis between legumes grown

in Limpopo and endemic rhizobia is effective in biological N fixation. It would be

worthwhile to undertake some field studies to assess the effectiveness of different strains

of rhizobia on these soils especially if legumes such as mucuna are to be introduced into

local cropping systems.

107

Literature cited

Adesina, A.A., Mbila, D., Nkamleu, G.B. and Endamana, D. (2000) Econometric analysis of the determinants of adoption of alley farming by farmers in the forest zone ofsouthwest Cameroon. Agriculture, ecosystems and environment 80, 255-265.

Ambus, P. and Jensen, E.S. (1997) Nitrogen mineralisation and denitrification as influenced by crop residue particle size. Plant and soil 197, 261-270.

Andren, O., Steen, E. and Rajkai, K. (1992) Modelling the effects of moisture on Barley straw and root decomposition in the field. Soil biology and biochemistry 24, 727-736.

Armstrong, R.D., McCosker, K., Millar, G., Kuskopf, B., Johnson, S., Walsh, K., Probert, M.E. and Standley, J. (1999) Legume and opportunity cropping systems in central Queensland: 2. Effect of legumes on following crops. Australian journal of agricultural research 50, 925-936.

Armstrong, R.D., McCosker, K., Millar, G., Walsh, K., Johnson, S. and Probert, M.E. (1997) Improved nitrogen supply to cereals in central Queensland following short legume leys. Australian journal of experimental agriculture 37, 359-368.

Aulakh, M.S., Doran, J.W., Walters, D.T., Mosier, A.R. and Francis, D.D. (1991) Crop residue type and placement effects on denitrification and mineralisation. Soil science society of America journal 55,

Ayisi K.K. and Mpangane, P.N.Z. (2004) Growth and symbiotic activities of cowpea cultivars in sole and binary cultures with maize. In: Whitbreat, A.M. and Pengelly, B.C., (Eds.) Tropical legumes for sustainable farming systems in Southern Afirca and Australia, 115 edn. pp. 92-98. Canberra: ACIAR

Ayisi, K.K. and Mpangane, P.N.Z. (2004) Growth and symbiotic activities of cowpea cultivars in sole and binary cultures with maize. In: Whitbread, A.M. and Pengelly, B.C., (Eds.) Tropical legumes for sustainable farming systems in Southern Africa and Australia, Canberra: Australian Centre for International Agricultural Research.

Azooz, R.H., Lowery, B. and Daniel, T.C. (1995) Tillage and residue management influence on corn growth. Soil and tillage research 33, 215-227.

Barthes, B., Azontonde, A., Blanchart, E., Girardin, C., Villenave, C., Lesaint, S., Oliver, R. and Feller, C. (2004) Effect of a legume cover crop (Mucuna pruriens) on soil carbon in an ultisol under maize cultivation in Southern Benin. Soil use and management 20, 231-239.

Becker, M. and Johnson, D.E. (1998) Legumes as dryland fallow in upland rice-based systems of west Africa. Biology and fertility of soils 20, 358-367.

Blair, N., Faulkner, R.D., Till, A.R. and Sanchez, P. (2005) Decomposition of 13c and 15N labelled plant residue materials in two different soil types and its impact on soil carbon, nitrogen, aggregate stability, and aggregate formation . Australian journal of soil research 43, 873-886.

108

Blake, G.R. and Hartge, K.H. (1986) Bulk density. In: Smith, K.A.M.C.E., (Ed.) Methods of soil analysis, Part 1. Physical and mineralogical methods 2nd edition, pp. 363-375. Madison,USA: American society of agronomy- soil science society of America

Byrnes, R. M. (2008) South Africa- Zion Christian church: A country study. 96.

Caamal-Maldonado, J.A., Jimenez-Osornio, J.J., Torres-Barragan, A. and Anaya, A.L. (2001) The use of allelopathic legume cover and mulch species for weed control in cropping systems. Agronomy Journal 93, 27-36.

Carsky, J. R. and Ndikawa, R. (2009) Identification of cover crops for the semi-arid savanna zone of west Africa. The International Development research Centre.

Carsky, R.J., Oyewole, B. and Tian, G. (2001) Effect of phosporus application in legume cover crops rotation on subsequent maize in the savanna zone of West Africa. Nutrient cycling in agroecosystems 59, 151-159.

Chigariro, B (2004) The farming systems of the Zimbabwean district of Wedza and the role and adoption of forage legumes in small-scale crop-livestock enterprises. IN Whitbread, A. M., Pengelly, B.C. (Ed.) Tropical legumes for sustainable farming systems in Southern Africa and Australia. Canberra, ACIAR.

Chikowo, R., Mapfumo, P., Nyamugafata, P. and Giller, K.E. (2004) Maize productivity and mineral N dynamics following different soil fertility management practices on a depleted sandy soil in Zimbabwe. Agriculture, ecosystems and environment 102, 119-131.

Chotte, J.L., Ladd, J.N. and Amato, M. (1998) Sites of microbial assimilation, and turnover of soluble and particulate 14C- labelled substrates decomposing in a clay soil. Soil biology and biochemistry 30, 205-218.

Christensen, B.T. and Olesen, J.E. (1998) Nitrogen mineralisation potential of organomineral size seperates from soils with annual straw incorporation. European journal of soil science 49, 25-36.

Costa, F.J.S.A., Bouldin, D.R. and Suhet, A.R. (1990) Evaluation of N recovery from mucuna placed on the surface or incorporated in the Brazilian oxisol. Plant and soil124, 91-96.

Dalgliesh, N.P. and Cawthray, S. (1988) Determining plant available water capacity. In: Dalgliesh, N.P. and Foale, M.A., (Eds.) Soil matters: monitoring soil water and nutrients in dryland farming, Towoomba, Queensland, Australia: APSRU

de Villiers, M. C., Pretorius, D. J., Barnard, R. O., van Zyl, A. J., and le Clus, C.F. (2007) Land degradation assessment in dry land areas: South Africa . LADA. Land degradation assessment in dryland project LADA (FAO)

Department of Agriculture Limpopo. (2007) Background. Polokwane,South Africa: http://www.lda.gov.za/index.php?Entity=Background

Department of environmental affairs and tourism. (2008) State of the environment. Department of environmental affairs and tourism. http://soer.deat.gov.za/themes.aspx?m=507

http://www.lda.gov.za/index.php?Entity=Background

109

Fischler, M. and Wortmann, C.S. (1999) Green manures for maize-bean systems in eastern Uganda: Agronomic performance and farmers' perceptions. Agriforestry systems 47, 123-138.

Fofana, B., Breman, H., Carsky, R.J., Van Reuler, H., Tamelkpo, A.F. and Gnakpenou, K.D. (2004) Using mucuna and P fertiliser to increase maize grain yield and N fertiliser efficiencyin the coastal savanna of Togo. Nutrient cycling in agroecosystems 68, 213-222.

Food and Agriculture Organisation. (2004) Drought impact mitigation and prevention in the Limpopo basin. FAO corporate document repository, Discussion paper.

Fosu, M., Kuhne, R.F. and Vlek, P.L.G. (2007) Mineralization and microbial biomass dynamics during decomposition of four legumeinous residues. Journal of biological sciences 7, 632-637.

Francis, P. and Rawlins-Branan, M.J. (1987) The extension system amd small-scale farmers: Acase studyfrom Northern Zambia. Agricultural administration and extension 26, 185-196.

Franzluebbers, K., Weaver, R.W., Juo, A.S.R. and Franzluebbers, A.J. (1995) Mineralization of carbon and nitrogen from cowpea elaves decomposing in soil with different levels of microbial biomass. Biology and fertility of soils 19, 100-102.

Gardner, E.A. (1985) Soil water. In: Rayment, G.E., (Ed.) Identification of soils and interpretation of soil data, Queensland, Australia: Australian society of soil science incorporated

Giller, K.E. (2001) Nitrogen fixation in tropical cropping systems, United Kingdom, CABI publishing.

Giller, K.E. and Wilson, K.J. (1991) Nitrogen fixation in tropical cropping systems, United Kingdom: C.A.B International.

Godwin, D.C., Jones, C.A., Ritchie, J.T., Vlek, P.L.G. and Youngdahl, L.J. (1984) The water and N components of the CERES models. In: Kumble, V., (Ed.) Proceedings of the international symposium on minimum data sets for agrotechnology transfer, ICRISAT centre: ICRISAT

Green, C.J. and Blackmer, A.M. (1995) Residue decomposition effects on nitrogen availability to corn following corn or soyabean. Soil science society of America journal 59, 1065-1070.

Gregorich, E.G., Voroney, R.P. and Kachanoski, R.G. (1991) Turnover of carbon through the microbial biomass in soils with different textures. Soil biology and biochemistry 8, 799-805.

Grunwald, N.J. and van Bruggen, A.H.C. (2000) Short-tern cover crop decomposition in organic and conventional soils: Characterisation of soil carborn, nitrogen, microbial and plant pathogen dynamics. European journal of plant pathology 106, 37-50.

Hadas, A., Kautsky, L., Goek, M. and Kara, E.E. (2004) Rates of decomposition of plant residues and available nitrogen in soil, related to residue compositon through

110

simulation of carbon and nitrogen turnover. Soil biology and biochemistry 36, 255-266.

Hauser, S. and Nolte, C. (2002) Biomass roduction and N fixation of five mucuna pruriens varieties and their effect on maize yield in the forest zone of cameroon. Journal of plant nutrition and soil science 165, 101-109.

Hayman, P., Whitbread, A. and Gobbet, D. (2008) Practicing agronomy in an uncertain climate- Using similation modelling to study seasonal drought and the impact of ENSO in the Southern Australian grains belt. In: Unkovich, M., (Ed.) Proceedings of the 14th ASA Conference, Adelaide South Australia:

Haynes, R.J., Cameron, K.C., Goh, K.M. and Sherlock, R.R. (1986) Mineral nitrogen in the plant-soil system, London: Academic press.

Hazelton, P. and Murphy.B (2007) Interpreting soil test results: What do all numbers mean?, Victoria, Australia: CSIRO.

Houndekon V, Manyong VM, Gogan, C and Vesteeg MN (1998) Collaboration to increase the use of Mucuna in production systems in Benin, IDRC

Jawson, M.D., Elliott, L.F., Papendick, R.I. and Campell, G.S. (1989) The decomposition of 14C-labeled wheat straw and 15N-labeled microbial material. Soil biology and biochemistry 21, 417-422.

Jensen, E.S. (1987) Seasonal patterns of growth and nitrogen fixation in field grown pea. Plant and soil 101, 29-37.

Jensen, S. (1994) Dynamics of mature pea residue nitrogen turnover in unplanted soil under field conditions. Soil biology and biochemistry 26, 455-464.

Jingguo, W. and Bakken, L.R. (1997a) Competition for nitrogen during decomposition of plant residues in soil: effect of spacial placement of N-rich and N-poor plant residues. Soil biology and biochemistry 29, 153-162.

Jingguo, W. and Bakken, L.R. (1997b) Competition for nitrogen during mineralisation of plant residues in soil: Microbial response to C and N availability. Soil biology and biochemistry 22, 163-170.

Jiri, O., Maasdorp, B.V., Temba, E. and Whitbread, A.M. (2004) The role of dual purpose and forage legumes to improve soil fertility and provide high quality forage: Options in a maize-based system in Zimbabwe. In: Whitbread, A.M. and Pengelly, B.E., (Eds.) Tropical legumes for sustainable farming systems in Southern Africa and Australia, Canberra: ACIAR

JMP (2005) JMP for basic univariate and multivariate statistics: a step by step guide . Lehman, A., O'Rourke, N., Hatcher, L. and Stepanski, E.J., (Eds.) Unite States of America: SAS Institute Inc.

Joergensen, R.G. and Brookes, P.C. (1990) Ninhydrin-reactive nitrogen measurements of microbial biomass in 0.5 M K2SO4 soil extracts. Soil biology and biochemistry 22, 1023-1027.

111

Kashem, M.A. and Jones, G.E. (1988) Small farmers' perceptions of obstacles to improved rice cultivation in Bangladesh. Agricultural administration and extension 29, 293-300.

Keating, B.A., Carberry, P.S., Hammer, G.L., Probert, M.E., Robertson, M.J., Holzworth, D., Huth, N.I., Hargreaves, J.N.G., Meinke, H., Hochman, Z., Maclean, G., Verburg, K., Snow, V., Dimes, J.P., Silburn, M., Wang, E., Brown, S. , Bristow, K.L., Asseng, S., Chapman, S., McCown, R.L., Freebairn, D.M. and Smith, C.J. (2003) An overview of APSIM, a model designed for farming systems simulation. European Journal of Agronomy 18, 267-288.

Koenig, R.T. and Cochran, V.L. (1994) Decompositon and nitrogen mineralisation for legume and non-legume crop residues in a sub-arctic agricultural soil. Biology and fertility of soils 17, 269-275.

Lal, R. (1978) Influence of within- and between-row mulching on soil temperature, soil moisture, root development and yield of maize. Field crop research 1, 127-139.

Langley, J.A., Heady, E.O. and Olson, K.D. (1988) The macro implications of a complete transformationof U.S. agricultural production to organic farming practices. Agriculture, ecosystems and environment 10, 323-333.

Leblanc, H.A., Nygren, P. and McGraw, R. (2006) Green mulch decomposition and nitrogen release from leaves of two Inga spp. in an organic alley-cropping practice in the humid tropics. Soil biology and biochemistry 38, 349-358.

Lehohla, P. (2004) Statistics South Africa, Pretoria, South Africa, http://www.statssa.gov.za/publication/Statistics

Liu, L., King, J.S., Booker, F.L., Giardina, C.P., Allen, H.L. and Hu, S. (2009) Enhanced litter input rather than changes in litter chemistry drive soil carbon and nitrogen cycles under elevated CO2:a microcosm study. Global change biology 15, 441-453.

Luscombe, B.M. and Gray, T.R.G. (1974) Characteristics of Arthrobacter grown in continuous culture. Journal of general microbiology 82, 213-222.

Maasdorp, B. V., Jiri, O., Temba, E. (2004) Contrasting adoption, management, productivity and utilisation of Mucuna in two different smallholder farming systems in Zimbabwe IN Whitbread, A. M., Pengelly, B.C. (Ed.) Tropical legumes for sustainable farming systems in Southern Africa and Australia. Canberra, ACIAR.

Mafongoya, P.L. and Nair, P.K.R. (1997) Multipurpose tree prunings as a source of nitrogen to maize under semiarid conditions in Zimbabwe. Agroforestry systems 35, 31-46.

Maluleke, H.M., Ayisi, K.K. and Whitbread, A.M. (2004) Lablab density and planting date effects on growth and grain yield in maize-labalab intercrops. In: Whitbread, A.M. and Pengelly, B.C., (Eds.) Tropical legumes for sustainable farming systems in Southern Africa and Australia , Canberra: Australian Centre for International Agircultural Research

http://www.statssa.gov.za/publication/Statistics

112

Mapfumo, P., Mtambanengwe, F., Giller, K.E. and Mpepereki, S. (2005) Tapping indigenous herbaceous legumes for soil fertility management by resource-poor farmers in Zimbabwe. Agriculture, ecosystems and environment 109, 221-233.

Mason, C.F. (1976) Decomposition, London: Edward Arnold.

Matejovic, I. (1997) Determination of carborn and nitrogen in samples of various soils by the dry combustion. Communications of soil science and plant analysis 28, 1499-1511.

Maurya, P.R. and Lal, R. (1981) Effects of different mulch materials on soil properties and on the root growth and yield of maize (Zea mays) and cowpea (Vigna unguigulata). Field crops research 4, 33-45.

McCown, R.L., Hammer, G.L., Hargreaves, J.N.G., Holzworth, D.P. and Freebairn, D.M. (1996) APSIM: a Novel software system for model development, model testing and simulation in agricultural systems research. Agricultural systems 50, 255-271.

Mendham, D.S., Kumaraswamy, S., Balasundaran, M. , Sankaran, K., Corbeels, M., Grove, T.S., O'Connell, A.M. and Rance, S.J. (2004) Legume cover cropping effects on early growth and soil nitrogen supply in eucalypt plantations in south-western India. Biology and fertility of soils 39, 375-382.

Misiko, M. (2007) Smallholder amendments to cereal-legume rotations in western Kenya. Soil fertility management and the African smallholder. Ph.D. thesis. The Netherlands .

Misiko, M., Tittonnel, P., Giller, K.E. and Richards, P. (2007) Smallholder amendments to cereal-legume rotations in western Kenya . In: Misiko, M., (Ed.) Fertile ground? Soil fertility management and the African smallholder Ph.D. thesis, pp. 15-34. Wageningen,The Netherlands:

Mpangane, P.N.Z., Ayisi, K.K., Mishiyi, M.G. and Whitbread, A.M. (2004) Grainyield of maize grown in sole and binary cultures with cowpea and lablab in Limpopo province of South Africa. In: Whitbread, A.M. and Pengelly, B.C., (Eds.) Tropical legumes for sustainable farming systems in Southern Africa and Australia, Canberra: Australian Centre for International Agricultural Research.

Mubarak, A.R., Rosenani, A.B., Anuar, A.R. and Zauyah, S. (2002) Decomposition and nutrient release of maize stover and groundnut haul under tropical field condition of Malaysia. Communications in soil science and plant analysis 33, 609-622.

Muller, T., Magid, J., Jensen, L.S. and Nielsen, N.E. (2003) Decomposition of plant residues of different quality in soil-DAISY model calibration and simulation based on experimental data. Ecological modelling 166, 3-18.

Murh, L., Tarawali, S.A., Peters, M. and Schultze-Kraft, R. (2002) Soil mineral dynamics and maize grain yields following Centrosema macropcarpum and Stylosanthes guianensis: effects of different rotations and varying levels of N fertiliser. Field crops research 78, 197-209.

113

Ncube, B., Dimes, J.P., van Wijk, M.T., Twomlow, S.J. and Giller, K.E. (2009) Productivity and residual benefits of grain legumes to sorghum under semi-arid conditions in south-western Zimbabwe: Unravelling the effects of water and nitrogen using a simulation model. Field crops research 110, 173-184.

Ndove, T.S., Whitbread, A.M., Clark, R.A. and Pengelly, B.C. (2004) Identifying the factors that contribute to the successful adoption of improved farming practices in the smallholder sector in Limpopo province, South Africa. In: Whitbread, A.M. and Pengelly, B.C., (Eds.) Tropical legumes for sustainable farming systems in Southern Africa and Australia, Canberra:

Nelson, R.A., Dimes, J.P., Paningbatan, E.P. and Silburn, D.M. (1998) Erosion/Productivity modelling og maize farming in the Phillipines uplands part 1: Parameterising the agricultural production system simulator. Agricultural systems58, 129-146.

Odhiambo, J.J. (2004) Biomass production and nitrogen accumulation of some dual purpose legumes in Limpopo province, South Africa. In: Fischer, T., Turner, N., Angus, J., McIntyre, L., Robertson, M., Borrell, A. and Lloyd, D., (Eds.) Proceedings of the International crop science conference, Brisbane: The Regional Institute Ltd

Palm, C.A., Gachengo, C.N., Delve, R.J., Cadisch, G. and GillerK.E. (2001) Organic inputs for soil fertility management in tropical agroecosystems: application of an organic resource database. Agriculture, ecosystems and environment 27-42.

Palm, C.A., Giller, K.E., Mafongoya, P.L. and Swift, M.J. (2001a) Management of organic matter in the tropics: translating theory into practice: Review. Nutrient cycling in agroecosystems 61, 63-75.

Palm, C.A. and Sanchez, P.A. (1991) Nitrogen release from the leaves of some tropical legumes as affected by their liging and poluphenol contents. Soil biology and biochemistry 23, 83-88.

Pandey, S. and Van Minh, D. (1998) A socio-economic analysis of rice prodcution systems in the uplands of northern Vietnam. Agriculture, ecosystems and environment 70, 249-258.

Papastylianou, I. (1986) The role of legumes in agricultural production in Cyprus. In: Beck, D.P. and Materon, L.A., (Eds.) Nitrogen fixation by legumes in Mediterranean agriculture: Developments in plant and soil sciences, Aleppo,Syria: Martinus Nijhoff publishers

Parfitt, R.L. and Salt, G.J. (2001) Carbon and nitrogen mineralisation in sand, silt, and clay fractions of soils under maize and pasture. Australian journal of soil research 39, 361-371.

Peoples, M.B., Herrodge, D.F. and Ladha, J.K. (1995) Biological nitrogen fixation: An efficient source for nitrogen sustainable agriculture production? Plant and soil 174, 3-28.

114

Polglase, P.J., Paul, K.I., Khanna, P.K., Nyakuengama, G.J., O'Connell, A.M., Grove, T.S. and Bataglia, M. (2000) Change in soil carbon following afforestation or reforestation: Review of experimental evidence and development of a conceptual framework . Report No. 20

Prasad, R. and Power, J.F. (1997) Soil fertility management for sustainable agriculture, USA: Lewis publishers.

Probert, E.M., Carberry, P.S., McCown, R.L. and Turpin, J.E. (1998b) Simulation of legume cereal system using APSIM. Australian journal of agrcultural research 49, 317-327.

Probert, M.E., Dimes, J.P., KeatingB.A., Dalal, R.C. and Strong, W.M. (1998a) APSIM's water and nitrogen modules and simulation of the dynamics of water and nitrogen in fallow systems. Agricultural systems 56, 1-28.

Probert, M.E., Keating, B.A., Thompson, J.P. and Parton, W.J. (1995) Modelling water, nitrogen, and crop yield for a long term fallow management experiment. Australian journal of experimental agriculture 35, 941-950.

Ramakrishna, A., Tam, H.M., Wani, S.P. and Long, T.D. (2006) Effect of mulch on soil temperature, moisture, weed infestation and yield of groundnutin northern Vietnam. Field crops research 95, 115-125.

Ratcliff, L.F., Ritchie, J.T. and Casssel, D.K. (1983) Field-measured limits of soil water availability as related to laboratory-measured properties. Soil science society of America 47, 770-775.

Rayment, G.E. and Higgingson, F.R. (1992) Australian handbook of soil and water chemical methods, edn. Port Melbourne: Inkata press.

Recous, S., Robin, D., Darwis, B. and Mary, B. (1995) Soil inorganic N availability: Effect on maize residue decomposition. Soil biology and biochemistry 27, 1529-1538.

Rivero, C., Chirenje, T., Ma, L.Q. and Martinez, G. (2004) Influence of compost on soil organic matter quality under tropical conditions. Geoderma 124, 355-361.

Sainju, U.M. and Singh, B.P. (2001) Tillage, cover crops, and kill-planting date effects on corn yield and soil nitrogen. Agronomy journal 93, 878

Sarrantonio, M. (1991) Methodologies for screening soil improving legumes, USA: Rodale Institute research centre.

SAS Institude Inc (2005) Statistics and graphics guide Release 6. SAS, (Ed.) 6 edn, United States of America: SAS Institute.

Saxena, M.C. (1986) Food legumes in the Mediterranean type of environment and ICARDA's efforts in improving their productivity. In: Beck, D.P. and Materon, L.A., (Eds.) Nitrogen fixation by legumes in Mediterranean agriculture: Developments in plant and soil sciences, Syria: Martinus Nijhoff.

Schultz, E. (1995) Crop production in a rotation trial at Tarlee, South Australia. Australian journal of experimental agriculture 35, 865-876.

115

Shamudzarira, Z. and Robertson, M.J. (2002) Simulating response of maize to nitrogen fertiliser in semi-arid Zimbabwe. Experimental agriculture 38, 79-96.

Sign, H. (1995) Nitrogen mineralisation, microbial biomass and crop yields as affected by wheat residue placement and fertiliser in a semi-arid tropical soil with minimum tillage. Journal of applied ecology 32, 588-595.

Sithole, V.M. and Apedaile, M.P. (1987) Factors influencing the reduction of maize acreage on Swazi nation land. Agricultural administration and extension 27, 201-213.

Snapp, S.S., Rohrbach, D.D., Simtowe, F. and Freeman, H.A. (2002) Sustainable soil management options for Malawi: Can smallholder farmers grow more legumes? Agriculture, ecosystems and environment 95, 115-125.

Sofranko, A.J., Khan, A. and Morgan, G. (1988) Insights into farmer-extension contact: Evidence from Pakistan. Agricultural administration and extension 30, 293-307.

Sparling, G.P., Gupta, V.V.S.R and Chunya Zhu (1993) Release of ninhydrin-reactive compounds during fumigation of soil to estimate microbial C and N. Soil biology and biochemistry 25, 1803-1805

Statistics South Africa. (2001) Census Reporthttp://www.statssa.gov.za/census01/html/default.asp

Thonnissen, C., Midmore, D.J., Ladha, J.K., Olk, D.C. and Schmidhalter, U. (2000) Legume decomposition and nitrogen release when applied as green manures to tropical vegetable production systems. Agronomy journal 92, 253-260.

Tian G, Brussaard, L. and Kang, B.T. (1995) An index for assessing the quality of plant residues and evaluating their effects on soil and crop in the (sub-) humid tropics. Applied soil ecology 2, 179-187.

Tian, G., Kang, B.T. and Brussaard, L. (1992) Effects of chemical composition on N, Ca, and Mg release during incubation of leaves from selected agroforestry and fallow plant species. Biogeochemistry 16, 103-119.

Tian, G., Kang, B.T. and Brussaard, L. (1993) Mulching effect of plant residues with chemically contrasting compositions on maize growth and nutrients accumulation. Plant and soil 153, 179-187.

Tian, G., Kolawole, G.O., Kang, B.T. and Kirchoff, G. (2000) Nitrogen fertilizer replacement indexes of legume cover crops in the derived savanna of west Africa. Plant and soil 224, 287-296.

Tittonell, P., Zingore, S., van Wijk, M.T., Corbeels, M. and Giller, K.E. (2007) Nutrient use efficiencies and crop responses to N, P and manure applications in Zimbabwean soils: Exploring management strategies across soil fertility gradients. Field crops research 100, 348-368.

van Veen, J.A., Ladd, J.N. and Amato, M. (1985) Turnover of carbon and nitrogen through the microbial biomass in a sandy loam and clay soil incubated with [14C(U)]glucose and [15N](NH4) 2SO4) under different moisture regimes. Soil biology and

http://www.statssa.gov.za/census01/html/default.asp

116

biochemistry 17, 747-756.

Ver Elst, P. and Pieterse, P.J. (2006) Qualitiy as a factor influencing the possible utilisation of eight exotic legume species as mulches. South African journal of plant soil 23, 237-245.

Voroney, R.P. and Paul, E.A. (1984) Determination of kc an kn in situ for calibtration of the chloroform fumigation-incubation method . Soil biology and biochemistry 16, 9-14.

Whitbread, A.M., Braun, A., Alumira, J. and Rusike, J. (2004) Using the agricultural simulation model APSIM with smallholder farmers in Zimbabwe to improve farming practices In: Whitbread, A.M. and Pengelly, B.C. (Eds.) Advances in tropical forage and ley legume technologies for sustainable grazing and cropping systems in southern Africa. ACIAR Proceedings No 115, ACIAR Canberra p171-180.

Whitbread, A.M. and Ayisi, K.K. (2004). The description of biophysical environment of 3 maize producing areas in the Limpopo Province of RSA and the validation of APSIM to simulate maize production. In: Whitbread, A.M. and Pengelly, B.C. (Eds.) Advances in tropical forage and ley legume technologies for sustainable grazing and cropping systems in southern Africa. ACIAR Proceedings No 115, ACIAR Canberra p. 17-26.

Whitbread, A.M. and Clem, B. (2004) Grain sorghum production and soil nitrogen dynamics following ley pastures on a vertosol soil in Queensland, Australia. In: In: Whitbread, A.M. and Pengelly, B.C. (Eds.) Advances in tropical forage and ley legume technologies for sustainable grazing and cropping systems in southern Africa. ACIAR Proceedings No 115, ACIAR Canberra. p115-125.

Wolf, J., De Wit, C.T. and Van Keulen, H. (1989) Modelling long term crop response to fertiliser and soil nitrogen. Plant and soil 120, 11-22.

Wortmann, C.S. and McIntyre, B.D. (2000) Annual soil improving legumes: agronomic effectiveness, nutrient uptake, nitrogen fixation and water use. Field crops research 68, 75-83.

Zaroug, M.G. (1986) Development of a regional program on biological nitrogen fixation to enhance agricultural production in the near East. In: Beck, D.P. and Materon, L.A., (Eds.) Nitrogen fixation by legumes in Meduterranean agriculture, Syria: Martinus Nijhoff

117

Appendices

Appendix 1. Survey questionnaire

(University of Limpopo School of Agricultural and Environmental Sciences)

Agricultural baseline survey for smallholder farmers in GaKgoroshi/GaSechaba

1. Enumerator: ………………… Village Name……………………….

2. Respondent Name:………………………………….

3. Gender of respondent: M……………. F…………….

4. Age of respondent

Below 18 28-37 38-47 48-57 58-67 Above 68

5. Highest educational qualification

None Primary Secondary Tertiary

6. How long have you been farming

Less than 5 years 5-10 years Over 10 years

Socio-economic details

7. Name of household head: ……………………………………………………….

8. Gender of household head: Male………… Female………………..

9. Age of household head in Years…………………………………………………

10. Number of people resident at household:………………………………………..

11. Number of adults (18 – 80) resident in household………………….

12. Is household head always resident at home?……….Yes/No

If No above, what occupies him/her away from home?…………………………

13. What off-farm activities do household members engage in? ……………………..

…………………………………………………………………………………………

118

14. Income sources for household

Source R0.00-

R500.00

R501.00-

R800.00

R801.00 and

more

Wages per month

Remittances per month

Income in kind

Pensions per month

Child grants per month

Income from sale of crops in 2006/07 season

Income form sale of crops in 2005/06 season

Income from sale of livestock and poultry in

2007

Income from sale of livestock and poultry in

2006

Others: specify

15. List 3 main constraints for increasing income

16. Do you grow enough food to meet household needs? Yes/No …………………...

17. If No above, explain why? ………………………………………………………..

Farming details

18. Do you own any livestock? Yes /No ……………………………………………...

19. If Yes above, which classes and how many?

Livestock class Number

a) Cattle ………….

b) Goats ………….

c) Sheep ………….

d) Pigs ………….

e) Poultry ………….

f) Others (specify) ………….

20. How much arable land do you own (ha)? ………………………………………...

119

21. How much arable land do you cultivate (ha)? ……………………………………

22. How many fields/plots do you own? ……………………………………………..

23. What are their sizes (ha)? …………………………………………………………

24. List the crops you grow in order of importance:

Crop Area (ha Output (specify units

2006/07 2005/06 2004/05 2003/4 2002/3

25. Do you ever get any surplus for sale? Yes/No……………………………………..

26. If Yes above, for which crops and how much? ……………………………………

………………………………………………………………………………………….

27. If No, above, do you get enough to meet household needs? …………………….....

…………………………………………………………………………………………..

28. What was the highest/lowest maize yield in the past five years

Maize Area (ha) Output (specify units- kg)

Highest 2006/07 2005/06 2004/05 2003/4 2002/3

Lowest

29. List the major constraints to increase maize yield in this area. Order them according to importance.1)……………………………………………………………………………………

2)……………………………………………………………………………………

3)……………………………………………………………………………………

30. What do you consider to be solutions to the problems listed above?1)……………………………………………………………………………………

2)……………………………………………………………………………………

3)……………………………………………………………………………………

4)……………………………………………………………………………………

120

31. For the crops you grow, list the variety you regularly use and time of planting.Crop Variety Time normally

planted

Time harvested

32. Which new crops would you like to grow and why?Crop Reason/s

33. How much seed do you apply per ha?

25kg 50kg 75kg Other

34. How do you grow your crops? (make a X on the appropriate method)

a) Random purestand _______

b) Row purestand _______

c) Mixed intercrop (random) _______

d) Row intercrop ____

35. If intercrop above, indicate the intercrop/s you use.a)…………………………………………….

b)…………………………………………….

c)…………………………………………….

d)…………………………………………….

36. Do you practise crop rotation? Yes/No.

37. If yes above, state the rotation you use …………………………………………….

…………………………………………………………………………………………..

38. If no above, why not? ………………………………………………………………

39. Do you allow for fallow periods on your farm? Yes…….. No………

121

40. If yes, for what period of time …………………….

3-6 Months 6 months to one year More than one year

41. If no, what is the reason?

Shortage of land Never heard of fallow Other

42. What do you use for ploughing your fields?

43. How many times do you plough the field before plantingZero Once Twice Other

44. Which crops do you plant during ploughing?

a)

b)

c)

d)

45. If no crops, what do you apply during ploughing?

46. Do you apply fertilizers? Yes/No………………………………………………….

47. If Yes above, which fertilizer and if not why not …………………………………

48. Do you have easy access to fertilizers Yes / No……………………………………

49. Fertiliser usage

Crop fertilised Fertilizer used Rate applied (kg ha-1) Why not fertilized

50. Do you have information on rates to apply Yes/No ………………………………..

51. Where do you get the information? ………………………………………………...

………………………………………………………………………………………

52. Do you apply any animal/kraal manure?

Yes (specify which)

No (explain why)

Hands Tractor Animal traction Other

122

53. Where do you get you manure?

Own animal Neighbours Others

54. When do you apply the manure? ..............................................................................

55. How frequently do you apply manure?......................................................................

56. How much manure do you apply?..............................................................................

57. What is the difficulty that you face with regards to manure application?

Unavailability Transportation Bulkiness Labour Others (specify)

58. How many times do you weed your fields?

Number of weeding Method of weeding

59. How do you manage your crop residues?

Leave on the

field

Plough under Removed to feed

livestock

Grazed in situ Burn Sell

60. Do you know about green manuring or mulching?

61. Do you plant any legume crops in your field?

62. If yes, what are they?

a)………………………………

b)………………………………

c)………………………………

63. For what purpose do you plant legumes

Fodder Human consumption Soil fertility Medicinal purpose Other

64. Have you heard about nitrogen fixation by legumes in the soil?

Yes No

65. Did you ever apply the skill of nitrogen fixation by legumes in the soil?

Yes No

Yes No

123

66. Would you like to use legumes for soil fertility in future? (suppose the farmer does not

know the purpose of legumes in the soil, and it is explained to him/her)

Yes No

67. What do you identify as problems with legume derived N for maize nutrition

Area of land required for growing

maize

Extra labour required to grow,

harvest and apply legume residue

for maize

Availability of adapted legume

varieties?

Competing demands for legume

residues (for feeding livestock)

Other

Do you have any access to credit? Yes / No ………………………………………

68. If No why not?

Lack of information High interest rate Collateral requirements Others (specify)

69. What is your general source of information on crop production and soils?

70. Are you a member of any farmers’ organisation?

71. If yes, what is the name of the organisation?

72. What position do you hold in the organisation?

Chairperson Secretary Treasurer Member

73. If not a member,why?………………………………………………………………

Extension staff NGO’s Other farmers Other (specify)

124

Appendix 2. Soil properties and initial values for APSIM simulation- Gabaza)

Layer number 1 2 3 4

Layer depth (mm) 150 150 300 300

Water content at air_dry (mm/mm) 0.04 0.08 0.13 0.13

Ll15(mm/mm) 0.11 0.11 0.15 0.18

Plant available water holding capacity (mm) 16.5 16.5 21 12

Crop lower limit (mm/mm) 0.11 0.11 0.15 0.18

Drained upper limit (mm/mm) 0.22 0.22 0.22 0.22

aSaturated water content (mm/mm) 0.51 0.47 0.40 0.41

bswcon 0.5 0.5 0.5 0.5

Bulk density (g/cm3) 1.10 1.23 1.39 1.38

Organic carbon (%) 1.57 1.45 1.00 1.00

pH 5.39 5.33 5.47 5.76

NH4-N (g/g) 0.1 0.1 0.1 0.1

NO3-N (g/g) 4.330 1.670 1.00 1.00

cFinert 0.5 0.7 0.7 0.9

dFbiom 0.03 0.02 0.02 0.01

125

Appendix 3. Soil properties and initial values for APSIM simulation- GaKgoroshi.

Layer number 1 2 3 4

Layer depth (mm) 150 150 300 300

Water content at air_dry (mm/mm) 0.03 0.04 0.08 0.08

Crop lower limit (mm/mm) 0.16 0.15 0.16 0.16

ll15 (mm/mm) 0.05 0.06 0.08 0.08

Plant available water holding capacity (mm) 19.5 22.5 15.0 15.0

Drained upper limit (mm/mm) 0.32 0.31 0.31 0.31aSaturated water content (mm/mm) 0.46 0.43 0.38 0.37bswcon 0.7 0.7 0.7 0.7

Bulk density (g/cm3) 1.41 1.39 1.36 1.33

Organic carbon (%) 0.46 0.42 0.42 0.42

pH 5.32 5.23 5.55 5.87

NH4-N (g/g) 0.1 0.1 0.1 0.1

NO3-N (g/g) 0.3 0.3 0.3 0.3cFinert 0.5 0.7 0.7 0.9dFbiom 0.03 0.02 0.02 0.01aSaturated water content calculated from total porosity – 0.05. Total Porosity (TP) = 1-

(bulk density/particle size density assumed to be 2.65)

bswcon determines the proportion of water above the DUL that will be drained daily.

cFinert describes the proportion of initial organic carbon assumed to be inert. Assuming

that all organic C measured at depth is essentially inert; this quantity is assumed to remain

the same at all depths.

dFbiom describes the initial biom as a proportion of non-inert C. These values are based

on (Probert et al. 1998)

CLL (crop lower limit) was assumed to be 50% of the DLL (drained upper limit)

DUL (drained upper limit) was calculated according to (Dalgliesh and Cawthray, 1988)

BD (Bulk density) was measured according to (Blake and Hartge, 1986)

126

Appendix 4. Soil chemical characterisation for APSIM

Depth N Cl Ca Mg Na K

cm mg/kg (cmol+/kg)

Gabaza 0-15 7.31 6 1210 268 13 70

15-30 3.27 10 1340 280 25 48

30-60 4.59 4 1250 288 18 30

60-90 4.56 4 n.d 255 18 n.d.

GaKgoroshi 0-15 0.42 n.d n.d. 265 140 67

15-30 0.42 n.d. n.d. 280 100 68

30-60 0.41 n.d. n.d. 300 60 69

60-90 0.40 n.d n.d 320 40 70

n.d. = not determined

127

Appendix 5. Soil chemical analysis for the soil profiles under different treatments for

Gabaza

Treatments Soil Pa Ka Caa Mga Naa pHb

Depth mg/kg

0N 0-15 6.3 42 729 292 8.2 5.9

15-30 2.8 26 918 352 10.6 6.2

30-45 1 24 920 371 10.5 6.2

45-60 0.8 24 943 401 11.5 6.3

30N 0-15 5.8 27 706 250 10.2 5.9

15-30 1.9 27 895 332 11.6 6.1

30-45 0.6 20 901 361 13.4 6.3

45-60 0.2 22 952 404 10.6 6.4

60N 0-15 1.1 28 638 248 8.7 5.9

15-30 0.7 25 787 303 12.1 6.0

30-45 0.6 24 912 367 11 6.2

45-60 0.5 22 963 413 12.9 6.3

90N 0-15 14.5 37 766 267 7.4 5.8

15-30 1.2 27 897 334 11.1 6.0

30-45 1.0 23 930 371 10.4 6.1

45-60 0.2 29 1090 479 12.1 6.3

Guarbean_0 0-15 1.2 57 732 295 12.5 5.9

15-30 0.9 29 767 295 12.1 6.0

30-45 0.6 23 842 345 14 6.2

45-60 0.23 22 859 392 12.7 6.24

Guarbean_Inc 0-15 1.17 30 667 275 8.3 5.93

15-30 0.93 27 897 357 13 6.09

30-45 0.46 25 978 437 12.5 6.27

45-60 0.18 25 1048 455 15.4 6.24

Guarbean_30 0-15 1.1 44 696 272 9.3 5.92

15-30 0.42 27 796 290 11.2 6

30-45 0.36 24 967 365 13 6.1

45-60 0.19 25 980 395 12.2 6.12

Grass_0 0-15 1.17 49 730 288 14.8 5.97

15-30 0.76 29 1227 473 20.1 6.14

30-45 0.42 31 992 381 18.7 6.1

128

45-60 0.25 26 1097 435 20 6.08

Grass_30 0-15 2.7 46 667 249 6.6 5.91

15-30 0.38 34 824 304 12.2 6.07

30-45 0.76 26 1026 390 12.8 6.18

45-60 0.42 21 893 368 10.4 6.29a 1:10 extractant ammonium acetateb water

Documents

Mulches in smallholder maize systems in the Limpopo ... · MULCHES IN SMALLHOLDER MAIZE SYSTEMS IN THE LIMPOPO PROVINCE OF SOUTH AFRICA: UNTANGLING THE EFFECTS OF N THROUGH EXPERIMENTATION