x nu .» v r- ‘..'J :_—-;.-..._$ 333333334153 3363?? ,3"'3 '3‘33‘33 ‘33.3 3:3 «3 . ‘31, “1’- .- .‘Jod . v ... .zl ....» > —‘4_( ‘ . r .-..._.-- - 1.6;.” 133“; 3 35:, ‘3; I u #1,“. k3" 't: ; ‘33 33333; i3? {m 331' 333' "3333.133 . 33333333323 .3'} 3' {33“ 31313- :31: 33%|”; . {3333133 3 r til 333%! .n '3. " 3‘ ,, ‘ 33.332133 .1 £33“ 5.3;: Hi " P.” 317333 p L3 |W 3323 3,; ”333333‘1333‘ ‘33 .‘bn, ‘t.‘;33_';,;'; 33: 3.35:3 a"- - , a» ' w cut w»..4 ......g ' _ . $2. 335 33533 33(3U333 3 3 3;3 9'33 ' o , 3:33;; 33 33333313333 "3‘3 33.3333 “In ‘3 33:13:33! ‘:3 ““33 33333. 3; 3v 3“ 3 "3 3333331 ‘3: . 1 333‘”; 333‘- 33013.3333 rim”, *1 3 3333 32.3”33; :33: :1, 33333333 5 $ ‘ L— w;3": v 3.31 v fin??? 3.3513 3;"; mi? {I r. {33453.3 333.33} J 333 '3" 3313;313' 3'33"; ”3:333:33: ' :‘i: 33333111; ”333""; 3'33 3n . NHL 3%“: I ”.4 3- 3.1;“ 3339 . 231-3; L3 3’ ‘3‘ . ' .11 3 3 33. "‘13.”. 23$ 3 1 333333:333;; ;3 3 -: ., 9:; ‘~ 3 "13L _ h3g3“, :- ‘ 3E3"; 3 31333? 5:3“ ‘ Q; \‘ -. n - w “‘ ‘ u 3'.— 4...»: _ .. I 3 3 a... «I ...;‘._I I" '1 ~ «g... _ u ...... m 33:9 3 1'3“!“ 3 1 33? V 3'. N" 3:"_.j,.' ' I""~:.. y 3 3 >333 , g3; ' l r 3339'“- W"‘ 3:13;} . . ‘ 3 3 , 3153:. ”in; 333 . 335433.13 b; ' 3 n . -L. ...,.~ - v. npgyw ‘4 .- ,. m" < 0-1 '5‘" , 3 ‘ “'32:; -111- A. --‘ o ' r 3533;553:338: g IN’A. . ‘ Irtcats lilllllllll lllllllIllllllllllllllllll 31293 01716 3456 This is to certify that the dissertation entitled PRACTICE OF SELECTED AGROFORESTRY TECHNOLOGIES: FARMER PERCEPTIONS OF INFLUENTIAL FACTORS presented by Charles MacPhery Masangano has been accepted towards fulfillment of the requirements for Ph.D. degreein Agric. & Ext. Educ. Major professor Dr. F. R. Whims September 30, 1997 Date MS U is an Affirmative Action/Equal Opportunity Institution 0- 12771 LIBRARY Michigan State University PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. MTE DUE IMTE DUE MTE DUE 1M mm“ PRACTICE OF SELECTED AGROFORESTRY TECHNOLOGIES: FARMER PERCEPTIONS OF INFLUENTIAL FACTORS By Charles MacPhery Masangano A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Agricultural and Extension Education 1997 ABSTRACT PRACTICE OF SELECTED AGROFORESTRY TECHNOLOGIES: FARMER PERCEPTIONS OF INFLUENTIAL FACTORS by Charles MacPhery Masangano This study was designed to determine the factors that influenced the farmer in the decision to practice selected agroforestry technologies in Malawi. The influence of the Malawi Agroforestry Extension Project, food productivity, control of soil erosion, fuelwood productivity and profitability were investigated. The influence of demographic variables which included: gender; age; income; education; landholding size and land and tree tenure system was also investigated. The study was conducted in Njolomole Extension Planning Area of Ntcheu RDP in Malawi. A sample of 449 respondents was drawn from the study population of 9236 farm households. Data collection was by interviews which were conducted by five male and four female interviewers. A total of 399 questionnaires were used for data analysis. Exposure to the Malawi Agroforestry Extension Project and control of soil erosion significantly influenced the practice of contour hedgerows. Food productivity significantly influenced the practices of planting and caring for trees in garden boundaries and planting and caring for trees in croplands. Profitability significantly influenced the practice of planting and caring for trees at homesteads. Gender was significantly related to the practices of planting and caring for trees in garden boundaries and planting and caring for trees in croplands. Male farmers practiced planting and caring for trees in garden boundaries and croplands. Farmers’ income was significantly related to the practices of woodlots and planting and caring for trees in garden boundaries. Land tenure system significantly influenced the practice of contour hedgerows. Copyright by CHARLES M. MASANGANO 1997 Dedication For the encouragement and support they gave me as a young child and throughout my school years, this work is dedicated to my father, MacPhery Masangano, mother, Ivy Masangano and brother, Synod Mpaya. Without their words of encouragement and physical support I would not have gone to this level. To my wife, Jane and children, MacPhery, Joshua, Caleb and James, I say thank you for your patience and endurance during my studies here at Michigan State University. iv Acknowledgments I would like to extend my gratitude and thanks to my advisor, Dr. Frederick R Whims, for his guidance, encouragement and support. He was not only a major professor to me, he was a personal fi'iend. I would also like to thank the other members of the committee Drs. F. Bobbitt, M. A. Gold and S. J Levine for their advice during the preparation of this document. Specifically I would like to thank Dr. Bobbitt for his advice on the statistical analyses, Dr. Gold for his general advice emanating from his deep knowledge in agroforestty and Dr Levine for his constructive critics which were very essential for the quality of this document. I would also like to thank Dr. Luke Reese for allowing me to use and helping me to Operate the various equipment which was necessary for preparation and presentation of this document. The ABE. secretaries Diana Davis, Gloria Bateman and Mary Pierce who always allowed me to use the various equipment and sometimes provided advice on formatting of this document. TABLE OF CONTENTS Page LIST OF TABLES .......................................................................................................... vi LIST OF FIGURES ....................................................................................................... xv ACRONYMS ................................................................................................................ xvi CHAPTER I INTRODUCTION ......................................................................................................... 1 Background .......................................................................................................... 1 Introduction to Malawi ............................................................................. 2 Problem statement ................................................................................................ 5 Objective of the study ........................................................................................... 5 Specific research hypotheses ................................................................................. 5 Conceptual model ................................................................................................. 7 Significance of the study ....................................................................................... 9 Definition of terms ................................................................................................ 9 Adoption ................................................................................................... 9 Agroforestry ............................................................................................. 9 Alley cropping ......................................................................................... 10 Contour hedgerows ................................................................................. 10 Faidherbt'a albida ................................................................................... 10 Fodder banks .......................................................................................... 11 Living fences ........................................................................................... ll Profitability ............................................................................................. 1 1 Project .................................................................................................... 12 Smallholder farmers ................................................................................ 12 Limitations ......................................................................................................... 12 II REVIEW OF RELATED LITERATURE AND RESEARCH .................................... 14 Introduction ....................................................................................................... 14 Food production in Malawi ................................................................................. l4 Advantages of agroforestry technologies ............................................................ 17 Provision of organic matter and nutrients ................................................ 18 Nitrogen fixation ......................................................................... 18 Effect on agronomic crop yields .............................................................. 19 Effect on soil erosion .............................................................................. 19 Effect on fuelwood availability ................................................................ 20 Other benefits ......................................................................................... 21 Disadvantages of agroforestry technologies ........................................................ 22 vi Competition ............................................................................................ 22 Time taken to see returns ........................................................................ 23 Complexity ............................................................................................. 24 Labor demanding .................................................................................... 24 Agroforestry technologies recommended in Malawi ............................................ 25 Cultivation under F. albida ..................................................................... 25 Contour hedgerows and/or grass strips .................................................... 26 Fodder banks .......................................................................................... 27 Alley cropping ......................................................................................... 27 Homestead, boundary and woodlot planting ............................................ 28 Living fences ........................................................................................... 28 Problems of low farmer practice of agroforestry technologies ............................. 29 Profitability ................................................ , ............................................ 40 Food productivity .................................................................................... 33 Soil erosion ............................................................................................. 34 F uelwood production .............................................................................. 35 Difl’usion process .................................................................................... 36 Agricultural extension in Malawi ............................................................. 36 The Malawi Agroforestry Extension Project ............................................ 37 Demographic factors ............................................................................... 40 Gender .................................................................................................... 40 Age ......................................................................................................... 43 Education ................................................................................................ 43 Income .................................................................................................... 44 Landholding size ..................................................................................... 44 Land and tree tenure system .................................................................... 45 Summary of chapter ............................................................................................ 47 III DESIGN AND METHODOLOGY ........................................................................... 48 Introduction ....................................................................................................... 48 Design ................................................................................................................ 48 Population .......................................................................................................... 49 Sampling procedures .......................................................................................... 50 Instrumentation and data collection ..................................................................... 50 Model specification and data analysis .................................................................. 51 Analytical framework .............................................................................. 52 Model specification. ................................................................................ 53 Data analysis ........................................................................................... 55 IV ANALYSIS AND RESULTS ................................................................................... 57 Introduction ....................................................................................................... 57 Demographic information about the sample ............................................ 59 vii A distribution of respondents practicing selected agroforestry technologies ................................................................................................................... 60 A comparison of respondents who were and those who were not exposed to the Malawi Agroforestry Extension Project regarding their practice of selected agroforestry technologies ....................................................................................................... 63 A comparison of male and female respondents practicing selected agroforestry technologies ...................................................................... 68 Respondent perceptions of the effect of selected agroforestry technologies on food crop yields ..................................................... 72 Respondent perceptions regarding the effect of selected agroforestry technologies on soil erosion ............................................... 82 Respondent perceptions regarding the efl‘ect of selected agroforestry technologies on fuelwood availability .......................................................................................................... 92 Respondent perceptions regarding the efi‘ect of selected agroforestry technologies on profitability ............................................. 103 Hypotheses testing using logistic regression models .......................................... 113 Logistic regression model for the practice of contour hedgerows ........................................................................................................ 115 Logistic regression model for the practice of woodlots .......................................................................................................... 1 l7 Logistic regression model for the practice of planting and caring for trees in garden boundaries ............................................ 118 Logistic regression model for the practice of planting and caring for trees in croplands .......................................................... 120 Logistic model for the practice of planting and caring for trees at homesteads .......................................................................... 122 viii V DISCUSSION AND RECOMMENDATIONS ......................................................... 124 Introduction ...................................................................................................... 124 Review of problem and purpose of the study ..................................................... 124 Review of research methods ............................................................................. 125 Discussion of results and findings ..................................................................... 126 Hypothesis 1: Exposure to the MAFE project ................................................... 126 Hypothesis 2: Food productivity ....................................................................... 127 Hypothesis 3: Soil erosion ................................................................................ 128 Hypothesis 4: Fuelwood production .................................................................. 128 Hypothesis 5: Profitability ................................................................................. 128 Hypothesis 6: Demographic factors ................................................................... 129 1. Gender .............................................................................................. 129 2. Age ................................................................................................... 130 3. Income .............................................................................................. 13 1 4. Land tenure system ........................................................................... 131 5. Landholding size ............................................................................... 133 6. Tree tenure system ............................................................................ 133 General recommendations ................................................................................. 133 Recommendations for further research .............................................................. 137 Personal observations and reflections ................................................................ 138 BIBLIOGRAPHY ........................................................................................................ 140 APPENDICES A Respondent exposure to information and demonstrations ............................................................................................. 149 B. Additional details about logistic regression models ....................................... 153 C. Interview questionnaire ................................................................................ 159 ix LIST OF TABLES TABLE 2.1 Estimated time when food stores were exhausted in the 1987/88 season in Salima ADD ......................................................... 2.2 Yields of maize from closely spaced hedgerows of Leucaena varieties and the control at Nkhande ........................................................ 4.1 A distribution of respondents practicing and not practicing and not practicing selected agroforestry technologies in Njolomole EPA of Ntcheu RDP ............................................................... 4.2 Estimated plot sizes where selected agroforestry technologies were practiced in Njolomole EPA of Ntcheu RDP ................................... 4.3 A comparison of respondents exposed to the Malawi Agroforestry Extension Project and those not Regarding the their practice of selected agroforestry designed for soil improvement in Njolomole EPA of Ntcheu RDP ..................................... 4.4 A comparison of respondents exposed to the Malawi Agroforestry Extension Project and those not regarding their practice of contour hedgerows in Njolomole EPA of Ntcheu RDP ...................................................................................... 4.5 A comparison of respondents exposed to the Malawi Agroforestry Extension Project and those not regarding their practice of selected agroforestry technology designed for the production of fruit, firelwood and other products in Njolomole EPA of Ntcheu RDP .............................................................. 4.6 A comparison of respondents exposed to the Malawi Agroforestry Extension Project and those not regarding their practice of selected agroforestry technology designed for the production of livestock feed and protection in Njolomole EPA of Ntcheu RDP ....................................................................................... 4.7 A comparison between male and female respondents regarding their practice of selected agroforestry technologies designed for soil improvement in Njolomole EPA othcheu RDP ............................... Page ................. 17 ................. 19 ................. 61 ................. 62 ................. 6S ................. 66 ................. 67 ................. 68 ................. 69 4.8 A comparison between male and female respondents regarding their practice of contour hedgerows in Njolomole EPA of Ntcheu RDP ............................................................................................................. 70 4.9 A comparison between male and female respondents regarding their practice of the selected agroforestry technologies designed for the production of fi'uit, fuelwood and other products in Njolomole EPA of Ntcheu RDP ........................................................................... 71 4.10 A comparison between male and female respondents regarding their practice of the selected agroforestry technologies designed for the production of livestock feed and protection in Njolomole EPA of Ntcheu RDP ............................................................................. 72 4.11 A distribution of respondent perceptions regarding the effect of selected agroforestry technologies designed for soil improvement on food crop yields in Njolomole EPA of Ntcheu RDP ........................................................................................................... 73 4.12 A comparison of respondent perceptions between respondents who were and those who were not exposed to the Malawi Agroforestry Extension Project, regarding the effect of selected agroforestry technologies designed for soil improvement on food crop yields in Njolomole EPA of Ntcheu RDP ........................................... 74 4.13 A comparison of male and female respondent Perceptions regarding the efi‘ect of selected agroforestry technologies designed for soil improvement on food crop yields in Njolomole EPA of Ntcheu RDP ............................................................................. 75 4.14 A distribution of respondent perceptions regarding the effect of contour hedgerows on food crop yields in Njolomole EPA othcheu RDP ...................................................................................................... 76 4.15 A comparison of respondent perceptions between respondents who were and those who were not exposed to the Malawi Agroforestry Extension Project, regarding the effect of contour hedgerows on food crop yields in Njolomole EPA of Ntcheu RDP ......................... 76 4.16 A comparison of male and female respondent Perceptions regarding the efl‘ect of contour hedgerows on food crop yields in Njolomole EPA of Ntcheu RDP ................................................................ 77 xi 4.17 A distribution of respondent perceptions regarding the effect of selected agroforestry technologies designed for the production of fruit, fuelwood and other products on food crop yields in Njolomole EPA of Ntcheu RDP ........................................................ 78 4.18 A comparison of respondent perceptions between respondents who were and those who were not exposed to the Malawi Agroforestry Extension Project, regarding the effect of selected agroforestry technologies designed for the production of fruit, fuelwood and other products on food crop yields in Njolomole EPA of Ntcheu RDP ............................................................................. 79 4.19 A comparison of male and female respondent perceptions regarding, the effect of selected agroforestry technologies designed for the production of fruit, fuelwood and other products on food crop yields in Njolomole EPA of Ntcheu RDP ............................ 80 4.20 A distribution of respondent perceptions regarding the effect of selected agroforestry technologies designed for the production of livestock feed and protection on food crop yields in Njolomole EPA of Ntcheu RDP ............................................................................................. 81 4.21 A comparison of respondent perceptions between respondents who were and those who were not exposed to the Malawi Agroforestry Extension Project, regarding the effect of selected agroforestry technologies designed for the production of Livestock feed and protection on food crop yields in Njolomole EPA of Ntcheu RDP .............................................................................................. 81 4.22 A comparison of male and female respondent Perceptions regarding the effect of selected agroforestry technologies designed for the production of livestock feed and protection on food crop yields in Njolomole EPA of Ntcheu RDP ................................................................ 82 4.23 A distribution of respondent perceptions regarding the effect of selected agroforestry technologies designed for soil improvement on the control of soil erosion in Njolomole EPA of Ntcheu RDP ...................................................................................................... 83 4.24 A comparison of respondent perceptions between respondents who were and those who were not exposed to the Malawi Agroforestry Extension Project, regarding the effect of selected agroforestry technologies designed for soil improvement in reducing soil erosion in Njolomole EPA of Ntcheu RDP ......................................... 84 xii 4.25 A comparison of male and female respondent Perceptions regarding the effect of selected agroforestry technologies designed for soil improvement in reducing soil erosion in Njolomole EPA of Ntcheu RDP ............................................................................. 85 4.26 A distribution of respondent perceptions regarding the efi‘ect of contour hedgerows in the control of soil erosion in Njolomole EPA of Ntcheu RDP ......................................................................... 86 4.27 A comparison of respondent perceptions between respondents who were and those who were not exposed to the Malawi Agroforestry Extension Project, regarding the effect of contour hedgerows in reducing soil erosion in Njolomole EPA of Ntcheu RDP .............................................................................. 86 4.28 A comparison of male and female respondent Perceptions regarding the efl‘ect of contour hedgerows in reducing soil erosion in Njolomole EPA of Ntcheu RDP ............................................................ 87 4.29 A distribution of respondent perceptions regarding the effect of selected agroforestry technologies designed for the production of fruit, fuelwood and other products in the control of soil erosion in Njolomole EPA of Ntcheu RDP ................................. 88 4.30 A comparison of respondent perceptions between respondents who were and those who were not exposed to the Malawi Agroforestry Extension Project, regarding the efl‘ect of selected agroforestry technologies designed for the production of fruit, fuelwood and other products in reducing soil erosion in Njolomole EPA of Ntcheu RDP .............................................................................. 89 4.31 A comparison of male and female respondent Perceptions regarding, the effect of selected agroforesz technologies designed for the production of fruit, fuelwood and other products in reducing soil erosion in Njolomole EPA of Ntcheu RDP ........................................................................................................... 90 4.32 A distribution of respondent perceptions regarding the efl‘ect of selected agroforestry technologies designed for the production of livestock feed and protection in the control of soil erosion in Njolomole EPA of Ntcheu RDP ................................................... 91 xiii 4.33 A comparison of respondent perceptions between respondents who were and those who were not exposed to the Malawi Agroforestry Extension Project, regarding the effect of selected agroforestry technologies designed for the production of livestock feed and protection in reducing soil erosion in Njolomole EPA of Ntcheu RDP ....................................................... 91 4.34 A comparison of male and female respondent Perceptions regarding the effect of selected agroforestry technologies designed for the production of livestock feed and protection in reducing soil erosion in Njolomole EPA of Ntcheu RDP ..................................... 92 4.35 A distribution of respondent perceptions regarding the efl‘ect of selected agroforestry technologies designed for soil improvement on firelwood availability in Njolomole EPA of Ntcheu RDP ...................................................................................................... 94 4.36 A comparison of respondent perceptions between respondents who were and those who were not exposed to the Malawi Agroforestry Extension Project, regarding the efl‘ect of selected agroforestry technologies designed for soil improvement on fuelwood availability in Njolomole EPA of Ntcheu RDP .................................... 95 4.37 A comparison of male and female respondent perceptions regarding the effect of selected agroforestry technologies designed for soil improvement on tirelwood availability in Njolomole EPA of Ntcheu RDP ......................................................................... 96 4.38 A distribution of respondent perceptions regarding the effect of contour hedgerows on fuelwood availability in Njolomole EPA of Ntcheu RDP ...................................................................................................... 97 4.39 A comparison of respondent perceptions between respondents who were and those who were not exposed to the Malawi Agroforestry Extension Project, regarding the effect of contour hedgerows on fitelwood availability in Njolomole EPA of Ntcheu RDP ........................................................................................................... 97 4.40 A comparison of male and female respondent Perceptions regarding the effect of selected agroforestry technology designed for the control of soil erosion on fuelwood availability in Njolomole EPA of Ntcheu RDP ........................................................ 98 xiv 4.41 A distribution of respondent perceptions regarding the effect of selected agroforestry technologies designed for the production of fi'uit, fuelwood and other products on fuelwood availability in Njolomole EPA of Ntcheu RDP ........................................................ 99 4.42 A comparison of respondent perceptions between respondents who were and those who were not exposed to the Malawi Agroforestry Extension Project, regarding the effect of selected agroforestry technologies designed for the production of firm, fuelwood and other products on fuelwood availability in Njolomole EPA of Ntcheu RDP ........................................................................... 100 4.43 A comparison of male and female respondent perceptions regarding the effect of selected agroforestry technologies designed for the production of fruit, fuelwood and other products on fuelwood availability in Njolomole EPA of Ntcheu RDP .................................. 101 4.44 A distribution of respondent perceptions regarding the efl'ect of selected agroforestry technologies designed for the production of livestock feed and protection on fuelwood availability in Njolomole EPA of Ntcheu RDP ...................................................... 102 4.45 A comparison of respondent perceptions between respondents who were and those who were not exposed to the Malawi Agroforestry Extension Project, regarding the effect of selected agroforestry technologies designed for the production of Livestock feed and protection on fuelwood availability in Njolomole EPA of Ntcheu RDP ...................................................... 102 4.46 A comparison of male and female respondent Perceptions regarding the effect of selected agroforestry technologies designed for the production of livestock feed and protection on fuelwood availability in Njolomole EPA of Ntcheu RDP .................................. 103 4.47 A distribution of respondent perceptions regarding the effect of selected agroforestry technologies designed for soil improvement on profitability in Njolomole EPA of Ntcheu RDP ......................................................................................................... 104 4.48 A comparison of respondent perceptions between respondents who were and those who were not exposed to the Malawi Agroforestry Extension Project, regarding the effect of selected agroforestry technologies designed for soil improvement on profitability in Njolomole EPA of Ntcheu RDP ............................................................................................ 105 4.49 A comparison of male and female respondent perceptions regarding the effect of selected agroforestry technologies designed for soil improvement on profitability in Njolomole EPA of Ntcheu RDP .................................................................................................... 106 4.50 A distribution of respondent perceptions regarding the efi‘ect of contour hedgerows on profitability in Njolomole EPA of Ntcheu RDP .......................................................................................................... 107 4.51 A comparison of respondent perceptions between respondents who were and those who were not exposed to the Malawi Agroforestry Extension Project effect of contour hedgerows on profitability in Njolomole EPA of Ntcheu RDP ................................................ 107 4.52 A comparison of male and female respondent Perceptions, regarding the effect of contour hedgerows on profitability in Njolomole EPA of Ntcheu RDP ..................................................... 108 4.53 A distribution of respondent perceptions regarding the efl'ect of selected agroforestry technologies designed for the production of fruit, firelwood and other products on profitability in Njolomole EPA of Ntcheu RDP ..................................................... 109 4.54 A comparison of respondent perceptions between respondents who were and those who were not exposed to the Malawi Agroforestry Extension Project, regarding the efi‘ect of selected agroforestry technologies designed for the production of fruit, fuelwood and other products in Njolomole EPA othcheu RDP ................................................... 110 4.55 A comparison of male and female respondent perceptions regarding the effect of selected agroforestry technologies designed for the production of fruit, firelwood and other products on profitability in Njolomole EPA of Ntcheu RDP .................................. 111 xvi 4.56 A distribution of respondent perceptions regarding the efi‘ect of selected agroforestry technologies designed for the production of livestock feed and protection on profitability in Njolomole EPA of Ntcheu RDP ..................................................... 112 4.57 A comparison of respondent perceptions between respondents who were and those who were not exposed to the Malawi Agroforestry Extension Project, regarding the efi‘ect of selected agroforestry technologies designed for the production of livestock feed and protection on profitability in Njolomole EPA of Ntcheu RDP ............................................... 112 4.58 A comparison of male and female respondent Perceptions regarding the efi‘ect of selected agroforestry technologies designed for the production of livestock feed and protection on profitability in Njolomole EPA of Ntcheu RDP ................................................ 113 4.59 Maximum likelihood estimates for the practice of contour hedgerows in Njolomole EPA of Ntcheu RDP ...................................................... 117 4.60 Maximum likelihood estimates for the practice of woodlots in Njolomole EPA of Ntcheu RDP ....................................................................... 1 18 4.61 Maximum likelihood estimates for the practice of planting and caring for trees in garden boundaries in Njolomole EPA of Ntcheu RDP ............................................................................................ 120 4.62 Maximum likelihood estimates for the practice of planting and caring for trees in croplands in Njolomole EPA of Ntcheu RDP .......................................................................................................... 121 4.63 Maximum likelihood estimates for the practice of planting and caring for trees at homesteads in Njolomole EPA of Ntcheu RDP .......................................................................................................... 122 Al Number of respondents who received information and demonstrations of selected agroforestry technologies in Njolomole EPA of Ntcheu RDP ............................................................................ 168 xvii A2 A comparison of respondent perceptions between respondents who were and those who were not exposed to the Malawi Agroforestry Extension Project regarding their exposure to extension information designed to promote the practice of selected agroforestry technologies ....................................................................................... 150 A3 A comparison of respondent perceptions between respondents who were and those who were not exposed to the Malawi Agroforestry Extension Project on whether they had seen demonstrations mounted by extension staff on the selected agroforestry technologies ........................................ 151 A4 A comparison of respondent perceptions of whether they were influenced by information and demonstrations provided by extension to practice the selected agroforestry technologies between respondents who were and those who were not exposed to the Malawi Agroforestry Extension Project ................................................................................................................... 152 xviii LIST OF FIGURES 1. Conceptual model with five independent variables ........................................................ 7 2. Demographic factors influencing practice of agroforestry technologies .......... - .................................................................................... 8 xix TGB ACRONYMS Agricultural Development Division Alley cropping Contour hedgerows Extension Planning Area Faicflrerbia albida Fodder banks Improved fallows Living fences The Malawi Agroforestry Extension Project Relay cropping Rural Development Project Planting and caring for trees in croplands Planting and caring for trees in garden boundaries Planting and caring for trees at homesteads Woodlots XX Chapter I INTRODUCTION Background Hunger, environmental degradation and high population growth rate are growing problems the world is facing today. Experts predict an additional four billion people on earth with 95 percent of them living in developing countries by the year 2025. Food demand will triple with the increased population (Anonymous, 1995). An obvious question is “where will this increased food be produced?” Land for agricultural production is very limited and poverty so rampant. Increasing numbers of small producers are failing to produce enough for their basic needs. Already 1.3 billion people, most of whom being women and children are living in poverty. The green revolution of the 1960s and 19703 saved millions fi'om starvation. The efi'orts of donor organizations like USAID, World Bank, Ford and Rockfeller Foundations played a key role in establishing international research organizations and training scientists from developing nations. These efforts helped to strengthen national policies as well as research and extension organizations resulting in adoption of technologies by small farmers. Population pressure has however caught up with the gains made in the green revolution. Low agricultural production combined with rampant environmental degradation raise Malthusian concerns of wide spread poverty and hunger (Anonymous, 1995). Introduction to Malawi Malawi is a small narrow country occupying the southern part of the East Afiican Rift Valley, within 9 to 17 degrees south latitude and 33 to 36 degrees east longitude. It is divided into the Northern, Central and Southern Regions, with 25 administrative districts. The country covers a total area of 11.9 million hectares with 9.4 million hectares of land and the rest under. lakes. The total available agricultural land is 5.3 million hectares (Bunderson et a1, 1993; Malawi Government, 1987). The per capita GNP in 1992 was USD 200.00 with a national debt service ratio of 23 percent. Forty eight percent of the people do not have access to safe drinking water. Just like other developing countries Malawi is faced with major challenges of how to increase smallholder agriculture production and maintain food self sufiiciency at household level in the face of increasing population and limited land resources. Malawi has an average population density of 89 people per square kilometer, with a population growth rate of 3 .7 percent. Eighty eight percent of her people live in nrral areas and are basically engaged in agriculture. This rapidly increasing population is putting a lot of pressure on the limited and un—expanding land resources. Landholding sizes continue to decrease as farmers sub- divide their land in order to share it among their children. Estimated land holdings are very small, averaging 0.5 hectares in the Southern Region, 0.7 hectares in the Central Region and 1.1 hectares in the Northern Region (Bunderson et a1, 1993). Other estimates of landholding size in the country are even more conservative than these (Jones et al, 1993). Farmers practice continuous cropping with maize, the main staple as the dominant crop. Crop rotation is almost nonexistent. Low incomes and high fertilizer prices on the 3 other hand result in low usage of inorganic fertilizers. Only 13 percent of those cultivating one hectare or less use inorganic fertilizers (Malawi Government, 1987). Declining soil fertility is a major land degradation problem as evidenced by the declining crop yields (Banda et a1, 1994; Smale, 1991). As observed elsewhere (Guggenheim and Spears, 1991), one strategy farmers use in order to deal with the problem of land pressure is expanding their farm lands by opening marginal areas. Most of these marginal areas were originally natural forests with steep slopes. Soils on these steep slopes are usually very shallow and susceptible to erosion (Park, 1992). Cultivation of these marginal lands is therefore causing major problems of soil erosion and depletion of natural forests. Once the tree cover has been opened for cultivation, the land quickly gets degraded and its productivity reduced. Bands et a1 (1994) conducting a study on crop productivity and soil erosion on a 44 percent slope, reported maize yields as low as 300 ngha. They also observed that soil loss due to erosion was as high as 80 tons per hectare per year. At such high rates of soil loss, the top soil gets washed away within a few years of opening the land. The total national forest cover, estimated to be 38 percent of the total land in 1987, is being depleted at the estimated rate of 3.5 percent per year in Malawi. Land clearing for agricultural production is the major cause of deforestation (Dewees, 1995). Agricultural land is also estimated to be expanding at the rate of 3.5 percent per year. Other causes of deforestation include: firelwood consumption for domestic energy needs; tobacco curing and other industrial uses like brick making and sale of fuelwood and charcoal to urban centers. 4 To maintain food self sufficiency, while protecting the precious natural resources of land and forests, Malawi needs immediate intervention of existing farming practices by the majority of its smallholder farmers. Technologies which are capable of maintaining and or improving soil fertility at low cost and reducing the rate of environmental degradation while increasing food production are required. Agroforestry when used as a complementary approach to conventional technologies has potential to increase food production while improving soil fertility and reducing environmental degradation. Research in Malawi has generated a number of agroforestry technologies promising to provide solutions to the problems of low food production and soil and environmental degradation. Farmer practice of these technologies has however been very low. Efl‘orts to increase the number of farmers practicing these technologies led to the introduction of a pilot project called the Malawi Agroforestry Extension (MAFE) Project in 1992. This was a five year project, implemented in partnership between Malawi Government and Washington State University using firnds provided by the USAID. It was piloted in five of the 175 extension planning areas (EPA) of the country. One of the EPAs where the project was piloted was Njolomole EPA of Ntcheu Rural Development Project (RDP). Despite all these efforts by the Malawi Government and all its partners, low farmer adoption of agroforestry technologies is still a major problem and studies to identify reasons for this situation have not been conducted. This study was therefore aimed at identifying some of the farmer perceptions of the factors influencing adoption of the technologies. 5 Problem statement The main problem in this study was the lack of firmer practice of selected agroforestry technologies in the face of low crop yields, high rates of soil and land degradation and depletion of natural forests. The study was therefore an attempt to investigate the reasons for the low firmer practice. Objective of the study The purpose of this study was to identify farmer perceptions of the factors influencing farmer decision to practice agroforestry technologies in Njolomole EPA of Ntcheu RDP in Malawi. Two specific objectives of the study were: 1. To investigate the influence of five selected factors: firmer exposure to the MAFE project; food productivity; soil erosion control; fuelwood productivity and profitability on firmers’ decision to practice agroforestry technologies in Njolomole EPA of Ntcheu RDP. 2. To investigate the influence of six demographic factors: gender", age; education; income; landholding size and land and tree tenure system on farmer decision to practice agroforestry technologies in Njolomole EPA of Ntcheu RDP. Comparisons were made between firmers who were exposed to the MAFE project and those who were not. Similar comparisons were made between male and female farmers. Specific research hypotheses The specific research hypotheses were: 1. Farmer exposure to the MAFE project influenced their decisions to practice selected agroforestry technologies in Njolomole EPA of Ntcheu RDP. 6 2. Farmer perceptions regarding the efi‘ect of selected agroforestry technologies on food crop yields influenced their decisions to practice those agroforestry technologies. 3. Farmer perceptions regarding the effect of selected agroforestry technologies on the control of soil erosion influenced their decisions to practice those agroforestry technologies. 4. Farmer perceptions regarding the effect of selected agroforestry technologies on the availability of fuelwood influenced their decisions to practice those agroforestry technologies. 5. Farmer perceptions of the effect of selected agroforestry technologies on profitability influenced their decisions to practice those technologies. 6. The following demographic factors: gender; age; highest level of formal schooling attained; the amount of income farmer received in 1996; landholding size; land and tree tenure system; influenced their decision to practice the selected agroforestry technologies. As the conceptual model in figures 1 shows, independent variables were: the effect of Malawi Agroforestry Extension Project; food productivity; soil erosion; fuelwood production and profitability. Demographic variables investigated are shown in Figure 2 and they included: gender; age; education; landholding size; income and land and tree tenure system. F armers’ decision to practice the specific agroforestry technology was the dependent variable. There were eleven agroforestry technologies studied. CONCEPTUAL MODEL PROFITABILITY FOOD . . PRODUCTIVITY ADOPTION i SOIL FUELWOOD AVAILABILITY EROSION PROJECT Figure 1: Conceptual model with adoption as the dependent variable. l GENDER I LANDHOLDING SIZE AGE ADOPTION INCOME l EDUCATION LAND AND TREE TENURE SYSTEM (LAND ANDTREE OWNERSHIP) Frgure2: Demographic factors influencing adoption of agroforestry technologies 9 Significance of the study The study represents the first major efi‘ort to identify the factors that influence firmer decisions to practice selected agroforestry technologies in Malawi. Malawi has made a lot of efi‘orts to come up with agroforestry technologies that seem to have promising impact on the productivity of smallholder farmers while also preserving the precious natural resources Of the country. The benefits of these efi‘orts can never be felt unless the technologies are practiced by firmers. The findings Of this study provide an understanding of the factors influencing the decision to practice the technologies and are instrumental in designing a strategy for increasing the rate of adoption of the technologies. The findings and recommendations made in this study will be made available to both the Ministry of Agriculture and Livestock Development and the Ministry Of Forestry and Natural Resources for them to be able to redesign their research and extension efforts for increasing the adoption of the selected agroforestry technologies. Definition of terms Adoption: The word adoption in this study was defined in terms of firmer perceptions of practicing the selected technologies. A firmer was considered an adopter of a particular agroforestry technology if he/she said that he/she practices the technology. Agroforesty: The International Center for Agroforestry (ICRAF) has defined the term agroforestry as “a collective name for land-use systems and practices where woody perennials are deliberately integrated with agricultural crops or animals on the same land management unit either in spacial mixture or in temporal sequence” (ICRAF annual report, 10 1992). The integration must result in significant ecological and economic interactions between the woody and non woody components. Alley cropping: This is an agroforestry system in which agronomic crops are grown in alleys formed by hedgerows of woody shrubs or trees. The hedgerows are usually cut back at planting and kept pruned during cropping season to reduce shedding and competition with the crops (Kang et al, 1981). The pnrnings are normally applied to the agronomic crop as green manure to provide the crop with nutrients afier decomposition. Contour hedgerows: This refers to the practice whereby hedgerows of woody shrubs or trees are grown on contour. Agricultural crops are normally grown between the hedgerows. The main purpose for contour hedgerows is soil and water conservation. The hedgerows serve as nrnofi’ checks and therefore reduce soil erosion (ICRAF annual report, 1992). The hedgerows can also be used in the same manner as in alley cropping, where the agronomic crop benefits from leaf application and/or nitrogen fixation in the case of leguminous and other nitrogen fixing woody shrubs or trees. These hedgerows were most often reinforced by strips of grass planted just above the hedgerow. When this was done, they were referred to as contour bufi‘er strips. In some cases only the grass strips were used without the hedgerows and this was referred to as contour vegetation strips (Bunderson et al, 1995). Faidlrerbr’a albida: This is a widely distributed tree in croplands of the semi-arid zone of Western Afiica, in the unimodal upland plateau of Southern Afiica and in Eastern Africa (ICRAF annual report, 1992). Increased crop yields have been Observed for some crops when grown under this tree. The positive effect of F. albida on crop production is 11 attributed to improved soil fertility, improved soil physical conditions brought about by better water retention, microbial populations and micro-environmental conditions produced by the trees (ICRAF annual report, 1992). The tree is unique in that it sheds ofi‘ its leaves in the rainy season and hence competes little for light and water with the crops growing beneath it. Fodder banks: This term refers to an agroforestry technology where by protein rich woody shrubs and trees like Leucaena Ieucocephala and Gliricidia sepirnn are planted to provide abundant high quality feed to livestock. Living fences: Living fences are normally established on the boundaries of homesteads, crop fields or as enclosures for livestock dwelling places. They are most often established in order to maintain privacy or to keep out domestic and wild animals. “Their main function is to eliminate the need to construct and replace dead fences every year, or the cost of purchasing and maintaining wire fences” (Bunderson et al, 1995). Ifmultipurpose trees are used, living fences can provide fi'uits, fodder, manure, wood for fuel, poles and handles for firm tools. Profitability: The term profits was defined in financial terms as the surplus income a firmer realized afier all production costs had been accounted for. Accordingly the term profitability was defined in relative terms as the degree to which one technology gave more financial profits than another technology. This definition was limited to financial benefits. All other benefits which were not financially quantifiable at firm level like those relating to environmental protection were not included in this definition. Secondly, profitability in this study was measured in terms of firmer perceptions only. 1 2 Project: The term project referred to the Malawi Agroforestry Extension (MAFE) project. Smallholder farmers: Agricultural producers in Malawi are generally broken down into two groups of small holders and estate producers (Malawi Government, 1987). Smallholder firmers generally refer to “all the households associated with customary land tenure system” (Mkandawire, 1992), and “are involved in firming primarily for subsistence needs” (Masangano, 1989). Malawi Government segments these smallholder farmers into three groups of those with less than 0.7 hectares, those with between 0.7 and 1.5 hectares and those with more than 1.5 hectares of land. Limitations The data collection was done by interviews using a team of five male and four female interviewers. Limitations associated with this type of instrument such as interviewer bias, response bias were acknowledged as weaknesses of this study. The selected agroforestry technologies considered in this study included, systematic interplanting with F. albida, alley cropping, relay cropping, improved fillows, contour hedgerows, woodlots, planting and caring for trees in garden boundaries, planting and caring for trees in craplands, planting and caring for trees at homesteads, fodder banks and living fences. The choice of these technologies were made because the technologies were currently recommended by Malawi Government to farmers in the study area. Findings of this study are therefore only generalizable to these technologies only. Although this study compared the perceptions of farmers who were exposed and those who were not exposed to the MAFE project, caution must be exercised when 13 interpreting the results. The MAFE project was piloted in the same EPA fiom which firmers who were not exposed to the MAFE project were sampled. There was no barrier to control those farmers who were not exposed to the project fiom observing what was happening to those firmers who were exposed to the MAFE project. Distance could have been one barrier imposed on the farmers not exposed to the project, however the ecological conditions of the EPA, notably the degree of soil erosion problems, rendered it more suitable for studying both categories of farmers. Chapter 2 REVIEW OF RELATED LITERATURE AND RESEARCH Introduction Malawi like most developing countries is facing a critical problem of how to increase smallholder agricultural production and maintain food self-sufiiciency at household level with the high rate of population growth and limited land resources. Malawi has a total population approaching twelve million (Malawi Government, 1993) with one of the world’s highest population growth rates of 3 .7 percent (Bunderson et al, 1993) The country’s fertility rate is estimated at 7.6 (Malawi Government, 1993) with a life expectancy of 44 years and infint mortality rate of 149 per 1000 (Bean Cowpea Collaborative Research Support Program Report, 1996). Food production in Malawi Malawi’s economy is basically agricultural based involving 88 percent of the labor force and producing 40 to 50 percent of the GNP as well as contributing 85 to 90 percent of the foreign exchange earnings (Malawi Government, 1987). Malawi has had a record of food self-suficiency until the early 19803. This situation has however, been changing lately. The country is now failing to keep pace the rising food demands due to limited land resources and low production levels. The current food position of Malawi is poor as demonstrated by the high incidence of child malnutrition among the 55 percent of rural households who have land holdings of less than one hectare (Kydd, undated; Sahn and Anrlpragasam, 1991; Quinn et al, 1988). Landholding sizes are getting smaller and more fi'agmented while demand for food 14 15 continues to rise. By 1987, thirty five percent of the small holders had 0.7 hectares of land or less, and forty percent had between 0.7 and 1.5 hectares of land (Malawi Government, 1987). A more recent estimate by Jones et a1 (1993) indicate an average landholding size of only 0.48 hectares. These land holdings are too small to support an average household of five. Consequently farmers do not practice crop rotations on their firms. Continuous cropping with food crops is common practice. Maize being the main staple tends to dominate the cropping pattern. Intercropping of maize with other food crops like beans, groundnuts, cassava and others is a common phenomenon. This is leading to declining soil fertility and low productivity. The situation is worsened by the low usage of inorganic fertilizers by most smallholder firmers. Kydd (undated) reported the average inorganic fertilizer application rate by smallholder farmers as 11 kg/ha and that this was largely skewed towards the large landholding farmers. A large majority of the smallholder firmers can not afford fertilizers due to very high prices. The cost of inorganic fertilizer in Malawi is much higher than other countries in the region. Lele (1988) observed that the 1987 fertilizer maize price ratio in Malawi was three times that of Kenya, making the use of inorganic fertilizer in Malawi largely unattractive for most smallholder farmers. Low currency rates, lack of subsidies and high transportation costs are largely to blame for the high cost of inorganic fertilizers. Low production levels are a common feature among smallholder firmers, especially those with less than one hectare of land (Lele, 1988; Mkandawire, 1988). Smale (1991) working with farmers in Malawi reported maize yields as low as 0.8 metric tones per hectare fiom unimproved maize varieties that were not fertilized, increasing to 1 6 1.4 tons with fertilization. Hybrid maize with fertilizer on the other hand gave yields of 2.5 tons. Despite the large relative increases in yields resulting fi'om fertilizers and improved varieties, only twenty four percent of the maize area was grown to improved varieties. Ninety six percent of the improved maize growers fertilized their crop as compared to only fifiy seven percent of the unimproved maize growers. High cost of seeds for improved varieties as well as their associated inputs like fertilizers, were the major constraints to firmer’s ability to adopt improved agricultural technologies. Carr (1988) stated that; “past initiatives in a number of agricultural programs have sought to overcome the constraint of land through the intensification of maize production with a consequent release of resources for other cash crops. Intensification has been fostered through the expanded use of fertilizers and improved maize seed, supported by a growing credit and extension service. These efforts have resulted in the use of fertilizers by more than 25% of the firming population with about 16% receiving credit. Improved maize seed has not proved widely popular to date and its use has stagnated at less than 10% of the maize area. The programs have assisted farmers with above average land resources, but have so far had little impact on the majority of the resource poor households which face the most serious problem of filling soil fertility and seasonal undemutrition.” Conventional technologies available in the country are very expensive and obviously not suitable for the smallholder firmers. The productivity of these farmers is therefore getting poorer. Previous studies have demonstrated that a large majority of these farmers are not able to feed themselves fiom one end of the cropping season to the other. In a study conducted in Salima Agricultural Development Division, Mkandawire (1988) showed that as high as eighty six percent ofthe firm families run out offood in the months just before the next harvest, see Table 2.1 below. 17 Table 2.1: Months in Which Family Food Stores were Exhausted in the 1987/88 Season in Salima ADD Month Cumulative number of families Percent of total who had run out of food May 87 21 5 June 87 27 7 July 87 41 11 August 87 67 17 September 87 9O 23 October 87 147 38 November 87 164 43 December 87 216 56 January 88 233 61 February 88 3 12 - 81 March 88 330 86 Source: Mkandawire, 1988 Note: Crop growing season in Malawi starts from November to April. These farmers need technologies which can increase their food production and reduce environmental degradation at low cost. Agroforestry technologies have been heralded as capable of helping to increase crop yields and maintain or even improve the environment at low cost. Advantages of agroforestry technologies. Agroforestry technologies have several characteristics which can help mitigate some of the problems discussed above. The following are some of the commonly documented advantages of agroforestry technologies. 1 8 Provision of nutrients and organic matter Some agroforestry practices supply essential organic matter (Beer, 1988; Shanna et al, 1994) and other nutrients such as magnesium, potash, calcium and sulfur which can become limiting under continuous cultivation without inputs of these elements. The supply of plant nutrients is achieved through the process of nutrient recycling from deep to upper soil layers. The woody tree species, especially those that are deep rooted absorb nutrients from deep soil layers, where the rooting system of the agronomic crops can not reach. These nutrients are translocated to above ground parts including leaves. Litter fill and decomposition completes the process of nutrient recycling fiom deep to upper soil layers while also adding organic matter (MacDicken and Vergara, 1990). Application of pnrnings gives the same effect as leaf fall. Kang et al (1984) observed that “six years of alley cropping Leucaena with maize and cowpeas on low fertility entisol had given very encouraging results. Periodic addition of Leucaena pnrnings helped to maintain high levels of soil nutrients and organic matter”. These authors observed that plots which were receiving prunings contained twice the amount of soil organic matter as compared to the plots where prunings were removed. Similar observations have been documented by other authors (Beer, 1988; Campbell et al, 1994; Shanna et al, 1994). Nitrogen fixation Some tree species especially those with nitrogen fixing characteristics seem to provide nitrogen to the agronomic crops. When grown under suitable conditions, some nitrogen fixing woody species like Leucaena Ieucocephala and Gliricr'dr'a septum have been observed to provide as much as 110 kg N/ha in one year (Kang et al, 1986). 1 9 Effect on agronomic crop yields Some agoforestry technologies seem to have a positive influence on yields of the agonomic crops. Studies show that yields of agronomic crops tend to be higher when grown under F. albeda (Randwanski and “fickens, 1967; Saka et al, 1994). Banda et al (1994) observed increases in yield of maize gown in strips between hedgerows of Leucaena motes. As Table 2.2 shows, maize gown between hedgerows ofLeucaena species gave yields of 1369kg per hectare and higher over a number of years while the highest yield obtained from a maize monocrop was only 815 kg per hectare. Table 2.2: Yields of Maize from Closely Spaced Hedgerows ofLeucaena Varieties and the Control at Nkhande in Malawi. Treatment Maize yield KG/ha 1986 1987 1989 1990 1991 Peru 2479 3818 4106 2143 2295 HawaianGiant 2033 1530 3652 1571 1369 Cunningham 1639 1936 2136 1558 1832 Control 815 455 318 309 152 Source:Bandaetal, 1994 Effect on soil erosion Some agoforestry technologies such as contour strips with woody hedgerows sigrificantly reduce soil erosion and runoff on steep slopes. Banda et al (1994) observed that cultivation of steep land (slope of 44 percent) resulted in erosion of 266 tons of soil per hectare in six years. However, hedgerows ofLeucaena Ieucocephala reduced the 20 amount of soil loss to 15 tons and in some cases less 10 tons per hectare on the same slope in the same period of time. Similar observations have been reported elsewhere (Lal, 1989; Dharrnasena, 1994). Reduction of the rainfall erosivity by the multi storied canopy of the agoforestry systems (MacDicken and Vergara, 1990) and reduction of runoff velocity by stems and surface roots are the main factors that contribute to the reduction of soil erosion. Effect on fuelwood availability The woody biomass obtained fi'om agoforestry trees can be a major source of firele for both domestic household needs as well as a source of cash. Bunderson et al (1991) reported woody biomass yields from Leucaena Ieucocephala as high as 2.5 tones per hectare in one year. Such levels of woody biomass can supply a substantial amount of firelwood. Fuelwood is a major source of energy in Malawi, providing 90 percent of the firel requirements (Malawi Government, 1987). Fuelwood consumption has been estimated at the rate of 1.1 cubic meters (0.66 tons) per capita per year in the rural areas. Almost all the preparation of meals which require cooking or roasting is done using firewood. Alternative sources of energy are either too expensive or not available to the majority of the people. Electricity for example, is only available in the urban areas and even there, most of the people can not afford it. The cost of petroleum products like kerosene is beyond the reach of the low income population. Firewood is considered the cheaper alternative (Dewees, 1995). The task of food preparation is mostly done by women and as a result they are the 2 1 ones who normally go out to fetch firewood. With the high rate of deforestation (3 .5 percent per annum) the forest area is diminishing and the distances that women travel to fetch firewood are becoming longer. This has direct impact on their labor demands as is evidenced by seasonal differences in their collection and usage of fuelwood behavior. Coote et al (1993) observed a remarkable reduction in the amount of firewood that women collected in the rainy season when they are busy with field activities as compare to the dry season. These authors observed that women collected and used 6 to 9 head loads offirewood per month in the rainy season as compared to 10 to 15 head loads in the dry season. This has important implications on the number and types of meals prepared. Women cook less often and family members eat fewer warm meals when firewood is scarce (Agarwal, 1986; Molnar and Screiber, 1989;). Women switch to foodstufl‘s and ingredients which require less cooking. Fuelwood-dependent food processing and preservation activities like smoking meats and fish or parboiling cereals and pulses decrease. Consequently the problem of fuelwood shortage may have serious impacts on the nutritional status of people. Shortage of fuelwood also tends to afl‘ect soil fertility as women start using dung and other firm residues for cooking instead of applying them in the field as manure. Other benefits ' Other benefits include provision of forest products like timber and poles for both domestic purposes and cash. Usage of fiuit trees in agoforestry provides the advantage of provision of finite. (Bunderson et al, 1994). Similarly usage of certain tree species, especially the leguminous ones like Leucaena Ieucocephala and Gliricidia septum provide 2 2 abundant high quality feed to livestock (ICRAF annual report, 1992; Bunderson et al,, 1994). Overall, aggegate production fiom agoforestry systems has been geater than systems of monoculture with forest or agonomic crops alone (Harwood, 1979; Wilson and Kang, 1981). MacDicken and Vergara (1990) discussed perceived economic and social benefits of agoforestry technologies as: 1. Crop diversity and reduced risk; diversification of the production base reduces risks of complete crop failure as well as economic impacts of price fluctuations of any single crop. 2. Increased income opportunities resulting fiom the intensification of agoforestry practices. Distribution of labor demand as well as income also tends to be more extended than the seasonal monoculture systems. 3. Increased variety of products may also have potential for improved human nutrition. Disadvantages of agroforestry technologies. This section provides a brief discussion of a number of disadvantages of agoforestry technologies: Competition Although not well understood the major weakness of agoforestry technologies cited in the literature is competitive interaction between the woody species and the agonomic crops. The woody species and the agonomic crops interact for light, water and nutrients. This interaction can be complementary or competitive. Some evidence shows that the competitive efi‘ect tends to have more impact on agonomic crop yields in some agoforestry technologies. Alley cropping is the most criticized in this respect. 2 3 A review conducted by Sanchez (1995) showed that out of eight alley cropping studies conducted in widely differing ecological zones of the world, only three had a positive interaction (net complementarity) when yields of the agonomic crops were the only economic benefit considered. The other five studies showed a negative efi‘ect on crop yields. In One of the studies, the negative efi’ect was as much as -58 percent. In other words, there was 58 percent crop yield reduction in alley cropping as compared to the sole crop. The competitive effect seems to be more serious where soil moisture seems to be most limiting. It should however be pointed that, the equation used for calculating these interaction effects only considered yields of agonomic crops as the only economic benefit of the alley cropping. Other benefits like production of fuelwood, poles, timber and fruits were not considered. Nevertheless, the data provide very strong evidence about the negative effects of alley cropping. Time taken to see returns Agoforestry technologies have the disadvantage of not providing or only providing limited early returns. Farmers are usually required to wait for a minimum of three years before they start reaping the benefits of agoforestry (Carter, 1996; Sanchez, 1995). This disadvantage is worsened by the fict that agoforestry technologies tend be associated with the risk of reduction of food crop yields as discussed above. Most of the farmers are operating at a bare survival level, what they produce is usually just enough for consumption fiom one harvest to the next. They are not able to produce surpluses for use in later years. They therefore can not afford the risk of having a reduction in their food 2 4 production in any particular year while waiting for benefits which are supposedly to come much later. Complexity Another problem associated with agoforestry technologies is that of complexity. Rogers (1983) defined the concept of complexity as the degee to which an innovation is perceived as dificult to understand and use. Sanchez (1995) described complexity as one of the four major issues associated with agoforestry. Agoforestry is a multi-disciplinary area combining several disciplines including forestry and agriculture. The interactions which occur between the woody species and the agonomic crops are complex and dificult to understand. Farmers engaging in agoforestry need to have a deep understanding of these interactions and their consequences to farm productivity. All these fictors make practicing of agoforestry technologies complex. This complexity may be influencing farmer adoption of agoforestry technologies. Innovations which are more complicated tend to be adopted more slowly than those which are more readily understood by most members of the social system (Rogers, 1983). Labor demanding Some agoforestry tasks tend to demand labor during periods when the labor is also needed in other production activities. A good example of such tasks is pruning. The woody species or trees in alley cropping or contour hedgerows need fi'equent pruning in order to reduce competition for light in the crop gowing season (Bunderson et al, 1991; Kang and Wilson, 1981; Wendt et al, 1993). This pruning can sometimes be as fi'equent 25 as every three weeks. The times these prunings are supposed to be done is also the time when agronomic crops are in the field, demanding labor for other tasks like weeding. Farmers may therefore perceive shortage of labor at such times as a constraint to their adoption of recommended agoforestry technologies. Agroforestry technologies recommended in Malawi Research in Malawi has identified and recommended a number of agoforestry technologies to smallholder farmers. These technologies are discussed in this section. Cultivation under I". albida. Cultivation of food crops under F. albida trees is one of the agoforestry technologies recommended in Malawi. F. albida is a large leguminous tree which is commonly found in most parts of Malawi. It is a unique tree in that it sheds 03‘ its leaves in the main gowing season and this reduces competition with the food crop for light and water. The shedding of the leaves also enriches the top soil with nutrients and organic matter as well as influencing the micro-environment in terms of improved rainfill infiltration (Saka et al, 1994). Observations made in Malawi in 1993, showed that up to 125 ng/ha was mineralized in alluvial soils under F. albida over a period of four months (ICRAF, 1993). Other studies have shown that yields of food crops like maize, sorghum and pearl millet can increase by 50 to 250 percent when gown under F albida (Randwanski and Wrckens, 1967; Saka et al, 1994). F. albida is also a good source of firelwood. 2 6 Contour hedgerows and/or grass strips Contour hedgerows and/or contour gass strips is another agoforestry technology recommended in Malawi. This technology is designed to control soil erosion. Forty five percent of the total land in Malawi is on slopes of at least 12 percent (Shaxon et al, 1977). Soil erosion is a major problem on these sloping lands. Erosive rains, poor crop management and inadequate crop cover all result in increased soil erosion. On the. other hand, the high population gowth rate that Malawi has been experiencing has increased the problem of land pressure. Landholding sizes for most of the smallholder farmers are too small to support their family food requirements. As a result of this, most of the farmers expand their farming areas by opening land which is on steep slopes originally deemed unsuitable for cultivation. Some firmers cultivate land as step as 45 percent slope. Problems of soil erosion and consequently, land degadation are so serious in these types of land that if left unchecked, a major crisis will occur. Adoption of conventional soil conservation measures has on the other hand been very poor because they are expensive, labor intensive, and do not show short term benefits (ICRAF annual report, 1992). Studies in Malawi as well as in other countries have shown that contour hedgerow intercropping is very efi‘ective in controlling erosion (Banda et al, 1994; Bunderson et al, 1994) and at a much lower cost (ICRAF annual report, 1992). The contour hedgerows check runofi‘ coming fiom above. When farmers use fi'uit trees on the contour ridges, they also benefit fiom the fiuit while usage of leguminous woody species provide added advantages of: a) Provision of biologically fixed nitrogen to the companion crop. 2 7 b) Provision of gen manure or mulch to the companion crops, hence recycling plant nutrients from deeper soil layers. c) Provision of fivorable conditions for soil macro- and micro-organisms. d) Provision of woody biomass which can be used for fuelwood, poles and timber. e) Provision of prunings for browse. Fodder banks Fodder banks of protein-rich woody legumes provide a lot of high quality feed to livestock as well as a ready supply of firelwood and even building materials (Bunderson et al, 1994). Studies conducted in Kenya showed that the economic advantages of using Leucaenaleafaslivestockfeed fordairyanimalswasthreeandahalftimesbetterthan using it as mulching for fertilizer for food crops (ICRAF annual report, 1992). Alley cropping The major advantages associated with alley cropping include: 1. Provision of biologically fixed nitrogen to the companion crop especially when leguminous woody species have been used (Kang et al, 1986). 2. Provision of geen manure or mulch to the companion crop (Beer, 1988; Kang et al, 1986; Sharma et al, 1994). 3. Recycling of nutrients fi'om deeper soil layers where the roots of agonomic crops can not reach. Agonomic crops benefit from these nutrients through litter fill from the woody species (Beer, 1988; Kang et al, 1986; Sharma et al, 1994). 4. Provision of fuelwood and other materials like poles, staking sticks and browse. Bunderson et al ( 1991) reported woody biomass yields fi'om Leucaena leucocephala fiom 2 8 an alley cropping trial as high as 2.5 tones per hectare. Alley cropping has however been criticized as a technology which can sometimes be more harmfirl than beneficial to firmers. The criticism stems fi'om the issue of competition between the woody species and the agonomic crops. Competition for light, water and nutrients has most often been observed to result in reduction of yields of the agonomic crops. Homestead, boundary and woodlot planting. Fruit and other multipurpose tree(MPT) species may be planted along firm boundaries, roads and homesteads. Trees may also be planted in small woodlots or orchards, depending on the availability of land. These trees can be important sources of food, fuelwood, poles, timber and fodder for both domestic use as well as sale for cash. Some of the tree species may even have important medicinal or pesticide value (Bunderson et al 1995). Living fences Living fences are planted mainly to eliminate the need to constnrct and replace dead fences every year or the cost of purchasing and maintaining wire fences. They are planted for a number of reasons including; keeping out domestic and wild animals, enclosing domestic animals and demarcating farm or garden boundaries and homesteads. Depending on species, living fences do also provide other uses like; fodder, geen manure, fi'uits, fuelwood as well as privacy (Bunderson et al, 1995). 29 Problems of low farmer practice of agroforestry technologies in Malawi Despite all the efforts made in Malawi and the advantages that the agoforestry technologies have, their adoption has remained very poor. Furthermore studies to identify the reasons for this poor response have not been done. Adoption of a particular innovation can be influenced by a number of factors including the specific characteristics of the innovation, the diffusion process and the characteristics of the social system. This section will briefly discuss each of these factors, as they relate to the adoption of selected agoforestry technologies in Malawi. Rogers (1983) discussed five innovation specific characteristics including; relative advantage, compatibility, complexity, triability and observability. These characteristics have profound influence on the clients decision to adopt the particular technology (Ryan and Gross, 1943). Rogers (1983) defined the concept of relative advantage as the degee to which an innovation is perceived as better than the idea it supersedes. The degee of relative advantage may be measured in economic terms like increased profitability and productivity. It can also be measured using other fictors like socio-prestige, convenience and satisfaction. Perceived relative advantage tends to weigh more than the physical advantage in the decision to adopt the innovation. This study investigated whether firmers perceived agoforestry technologies as having some relative advantage over their current practices. Specifically the study investigated whether firmers thought that agoforestry technologies were better than their previous or current practices in terms of: profitability; crop productivity; production of fuelwood and control of soil erosion. 3 0 Profitability One reason for low adoption suggested in the literature is that farmers do not perceive agoforestry technologies as more profitable than the monocultures of agonomic crops. Several authors have suggested that farmer adoption of new and improved technologies depends on their perception of the profitability of the technologies (Atta- Krah and Francis, 1986; Monu and Omole, 1991; Moris, 1991; Osuji 1991; Wendt et al, 1994). Roling (1993) working with Kenyan farmers, observed that small firmers can innovate incredibly fist if they perceive the progarn or technology to be profitable. When the farmers were provided with the right type of seeds packaged in appropriate bags, fertilizers being widely available and prices of their produce were attractive, adopted the technology very fast. This suggests that scientists need to evaluate their technologies more carefully before recommending them to firmers. The few economic studies on agoforestry technologies have generated mixed results. Akyeampong and I-Iitimana (1996) conducted an economic appraisal of alley cropping with Leucaena diversrfolia in Bunmdi. They observed that alley cropping on maize gave negative net present values. Even when they conducted a sensitivity analysis with the price of maize increased by 50 percent, the net present values were still negative. In another study, where bananas, beans were intercropped with nine different species of trees, some treatments showed positive net present values while others showed negative net present values (Akyeampong et al, 1995). A cost benefit analysis of agoforestry in Indonesia showed that maize gown under agoforestry systems was not economically beneficial to firmers (Silver, 1991). With this 31 kind of mixed results, two questions of interest would include; what are the perceptions of firmers relating to the profitability of agoforestry technologies? Secondly, how do the perceptions of farmers relating to the profitability of agoforestry technologies influence their decision to adopt those agoforestry technologies? An important variable to be considered when one talks about profitability is the availability of good markets for the products fi'om the technologies (Sanchez, 1995). Successful marketing of the agoforestry products is a necessity if farmers are to benefit fiom the technologies. The problem is that most of the non timber products of the agoforestry technologies do not have well developed markets. As a result, most of the financial appraisals tend to show that agoforestry technologies are not profitable (Peters et al, 1989). However, evidence fi'om other literature show that marketing of non timber products can make agoforestry more profitable (Clay, 1992; Hosier, 1989; Guggenheim and Spears, 1991; Mercer and Soussan, 1992; Peters et al, 1989). In Malawi, the market for some of the agoforestry products has been so much controlled to the extent that farmers may not perceive the profitability of the technologies. Dewees (1995) discussing issues of forestry policy in Malawi indicated that there exists a large market for firewood and charcoal in the main urban centers. Urbanization rate in Malawi is estimated at 2,000 people per week. One problem of this high rate of urbanization concerns energy supply. Firewood and charcoal are the major sources of energy in the urban sector. This is due to the unavailability and high costs of other sources of energy. The urban population depends heavily on the market for the supply of firewood and charcoal. The retail value of 32 the urban annual consumption of firewood and charcoal in 1993 was estimated at MK370.5l million and MK16.4 million respectively (Ng’ong’ola, 1992). Although the rural areas may not seem to have a market for these products, the urban market is likely to spread deeper into those areas of the country as supply becomes scarce in the areas close to the urban centers. However, the problem seems to do with the wood energy policy of the country. The government has been encouraging farmers to plant trees to meet the woodfuel market demand. Tree planting has been subsidized and in some cases lucrative bonuses have been paid to farmers in order to make tree planting financially attractive compared to crops. Secondly the government has committed itself to undertake monitoring and control of woodfirel trafiicking where it is being done on a commercial scale (Malawi Government, 1987). As a result of this, woodfirel markets have been controlled through confiscation on road blocks and on routes to main urban centers (Dewees, 1995). The third aspect of the policy is that while woodfuel prices are not controlled, prices for woodfuels fi'om government plantations are at a level below prices for woodfuels fi'om customary and private land (Malawi Government, 1987). This is done in order to discourage farmers fi'om selling fuelwood from natural forests. The policy is however affecting prices for fuelwood from planted forests making tree planting for firelwood production unprofitable. These policies present two major contradictions: a) Government is encouraging investment in tree planting while keeping prices from its ’ MK370.5 million was equivalent to US $97.5 million while MK16.4 million was equivalent to US 84.3 million at the exchange rate of 1993. 3 3 plantations low. There is no way farmers can expect to get higher prices for their woodfuel products. Government plantations account for 89,300 hectares while farmers’ plantations account for less than 10,000 hectares. The supply from government plantations completely suppresses the prices paid to farmers. This situation could be a major constraint to implementing efi‘ective tree gowing programs especially if the major objective is to produce fuelwood (Mercer and Soussan, 1992). b) The other contradiction lies in the fict that government controls the movement of fuelwood products to the market while on the other hand encouraging farmers to invest in tree planting. Although the main objective is to protect indigenous natural forests, firmers may be interpreting this as a government way of discouraging any markets for fuelwood and therefore not be willing to invest in agoforestry technologies. Food productivity With the problems of small landholdings, most of the smallholder farmers engage in firming for subsistence. Their main concern is to produce enough food to feed their families from one harvest to the next. Sometimes these firmers fiil to achieve this objective and are forced to engage in other activities like doing piece work in order to generate the food they need to cover up for the deficit (Mkandawire, 1988; Masangano, 1989). For these farmers, a beneficial technology would have to be one which helps them to get enough food for the whole year. In some studies, agoforestry technologies have been reported to maintain and or increase crop yields of their companion crops by maintaining or even improving soil fertility and soil micro-environment (Banda et al, 1994; Saks et al, 1994). Such 34 observations are however confounded by the fact that other study findings tend not to support these findings (Gosh et al, 1989). One problem with agroforestry is that the woody species use up part of the land which could be used for the food crop. If yields of such food crops do not increase sufficiently to compensate for the yields lost from the area planted to the trees, then the firmer risks a reduction in his/her total food production. Secondly is the problem of competition between the agonomic and tree crops for gowth fictors like light, water and nutrients. As discussed in the section on disadvantages of agroforestry technologies above, this competition has sometimes resulted in reduction of food crop yields. The risk of reduction of food production may be too costly for the firmer to accept and the technology is therefore, unlikely to be accepted. Soil erosion The problem of land pressure is forcing most smallholder firmers to open marginal land for cultivation. Such land is most often on very steep slope with very easily erodible soils. As a result of this, soil erosion is a major problem in Malawi. A large amount of soil is lost every year with the fertile top soil lost within the first few years of cultivation (Banda et al, 1994). This problem is causing serious land degadation and continuous decline in crop yields. It is however not known whether farmers perceive soil erosion as a major problem on their farms. Ifthey do, the next question is whether they perceive the selected agoforestry technologies as a solution to this problem. 35 Fuelwood production Shortage of firelwood is becoming a crisis in many developing countries. Fuelwood is the main source of energy in the developing world. F AO estimates show that more than seventy five percent of the energy consumption in most developing countries is from woodfirel (Agarwal, 1986). However, with the increasing problem of deforestation, fuelwood is becoming more scarce. PeOple are becoming more desperate for energy in most developing countries. Journeys to gather fuelwood which used to take an hour or less in the past, now take as much as a firll day (Eckholrn, 1975). The consequences of these fuelwood shortages are many. Molnar and Screiber (1989) listed some of them as follows; 1. Women who are the main firelwood collectors spend more time collecting firelwood and hence spend their time less productively. 2. Children spend more time helping with fuelwood collection and less time in school. 3. Women cook less often and family members eat less meals. 4. Women use dung and crop residues for cooking and therefore having less farm yard manure to apply in their fields. 5. Fuelwood-dependent food processing and preservation decrease. 6. Wood based income generating activities decrease or become less profitable. Agoforestry technologies can help mitigate the problem of fuelwood shortage. It is however not known whether farmers think that the recommended agoforestry technologies can help them reduce their firelwood shortage problems. 3 6 Diffusion process The difl‘usion process can briefly be defined as the process by which an innovation is communicated through certain channels over time among members of a social system (Rogers, 1983). As the definition implies, the process takes time to occur. Most often, focused efi‘orts have to be made to speed up the process. National agricultural extension progarns are examples of such efforts. Several approaches have been used in the agicultural extension progams. Pickering ( 1989) described five extension approaches used in Afiica. These include what he called the commodity-based approach, community development-cum-extension approach, innovation centered approach, goup based approach and farmer focussed approach. The firmer focussed approach also been referred to as the training and visit system (Benor and Baxter, 1984; Benoer et al, 1984). A sixth extension approach was described by Collinson (1981) as the firming systems research and extension approach. Agricultural extension in Malawi Agicultural extension in Malawi is implemented by the Ministry of Agiculture and Livestock Development through eight Agicultural Development Division (ADDs). These ADDs are sub-divided into 30 Rural Development Projects (RDPs) which are sinner divided into 173 Extension Planning Areas (EPAs). The extension approach followed is called the block extension method which is basically a modified training and visit system. The approach requires that an extension worker divide the working area, which is normally called a section, into eight sub-sections. These sub-sections are called blocks. The extension worker is to work with all the 3 7 farmers in each block as a goup and is encouraged to visit each block at least once every two weeks. Each block should to have a demonstration garden where firmers are provided with agicultural advice in a practical way. The main delivery method for the agicultural advice is through demonstrations, meetings, short courses, lectures, radio and printed materials like posters and newsletters for the few who can read them. The main advantages of the block extension approach approach include: a wider cross-section of the farmers are contacted simultaneously; sharing of ideas is enhanced through farmer interaction; and, extension workers are more easily supervised. Although, the block extension approach has been effective for some endeavors, it has not been effective in promoting firmer practice of selected agoforestry technologies. It was therefore decided to introduce a new extension approach for the promotion of firmer practice of selected agoforestry technologies. The approach was introduced as a pilot project called the Malawi Agoforestry Extension Project. The Malawi Agroforestry Extension Project In an effort to increase the adoption of recommended agoforestry technologies, the Malawi Government with cooperation from Washington State University and financing fi'om USAID introduced a pilot agoforestry extension project called the Malawi Agoforestry Extension Project (MAFE). It was a five year project extending from August, 1992 to September 1996. The project was irnplimented by several organisations including Ministry of Agriculture and Livestock Development, Ministry of Forestry and Natural Resources, Washington State University and the USAID. It was piloted in parts of certain EPAs of five of the twenty five districts of the country. The districts involved 38 were Mzimba in the Northern Region, Dowa and Ntcheu in the Central Region and Mangochi and Chikwawa in the Southern Region. The goal and objectives of the project were; Project goal: Enhance the economic well-being of smallholder farmers by improving food production eficiency and sustainability with reduced degadation of natural resources. Project objectives: 1. Develop and refine an agoforestry delivery system for implementation nationwide. 2. Adapt agoforestry technologies to farmer problems in order to facilitate adoption, impact and long term viability. 3. Train staff and farmers on agoforestry technologies and practices. 4. Develop and strengthen linkages between appropriate governmental and non- governmental organization. The extension approach used in the project was more of the firming systems research and extension approach as described by Collinson (1989). In this approach, technologies developed by research are tried and managed under farmer condition. Researchers, extension workers as well as farmers work together when trying the technologies at the firm. The main objective being to test and adapt the technologies to farm situations The basic concept of the project was to empower and motivate firmers to help themselves in a manner that was ecologically sound and economically attractive (Bunderson et al, 1992). This in essence meant that the firmers were the center of project activities. In other words, while the technologies may have been already available for 39 practice, the project emphasized on the process rather than the product. Farmer involvement or participation in the planning, decision making and implementation of the project was central. Their needs, objectives and priorities were emphasized. Their indigenous and traditional knowledge was also recogrized. Farmers were considered as partners in the implementation process, rather than imposing the knowledge gained fi'om research. Where necessary the technologies as designed by research were adapted to suit farmers’ needs, objectives, priorities and constrains. There are several advantages to this approach including: 1. When firmers are involved in the planning and decision making process, they are likely to be deeply committed to the project. Knowles (1980) said that “human beings tend to be committed to a decision to the extent that they have participated in making it. In other words firmers in the project were more likely to remain committed to practicing the technologies to the extent that they were involved in planning and deciding how to implement them. Involving the farmers in the planning and decision making process would have made them feel empowered that they were determining their own destiny and that feeling is very rewarding in itself (Adams, 1975; Dewey, 1966; Freire, 1970). 2. Iffirmers were involved in identifying their own needs, then the technologies they decided to practice would have been those that directly addressed those needs. The technologies and the leaming experiences they gained would be more appropriate to them (Dewey, 1966). Knowles (1980) indicated that people tend to be more deeply motivated to learn those things they see the need to learn. 3. Emphasis on indigenous and traditional knowledge would have made possible 40 adaptation of the technologies to local needs and conditions (Brokensha et al, 1980). Also adults have a backgound which is rich with experience. They define themselves in terms of the experience which they have accumulated. When they find themselves in situations which their experience is not being used, or its worth is minimized, they feel rejected as persons (Knowles, 1980). By tapping on the farmer’s knowledge and experience, they would have felt more accepted in the project and therefore be more willing to practice the technologies. Demographic factors Evidence fi'om the literature shows that some demogaphic factors do influence firmer adoption of recommended agicultural technologies. Such demographic fictors include: gender; age; level of education; level of income; landholding size and land tenure system (land ownership). This part of the literature review will isolate some of the current knowledge about the relationship of these demogaphic factors and adoption of recommended technologies. Gender Gender difi‘erences in developing countries have been observed to have fundamental influences on agoforestry technologies and other tree planting progams. Some of the difi'erences between men and women relate to issues of responsibilities for food and crop production, household food preparation, child care and land and tree tenure issues. Men and women have three distinct domains of responsibility for food and cash cropping in most developing country situations. The level of differentiation may vary from 4 1 society to society, but generally women tend be more concerned and responsible for household food provision and other household chores like child care (Chavangi et al, 1988). On the other hand, men tend to be more interested in cash generation (Molnar and Screiber, 1989). With the kind of responsibilities that women have to fulfill, their interests and knowledge about trees is very different from that of men. Women tend to be more interested in trees products like fruit, herbal medicines and firewood, while men tend to be interested in planting trees for cash generating products like poles and timber (Agarwal, 1986; Chavangi et al, 1988; Rocheleau, 1992; Scherr, 1994). Tree production progams whose primary focus is cash generation have tended not to be supported by women. Similarly tree production programs whose primary focus is production of food and firelwood have tended not to be supported by men. Indigenous knowledge on trees and tree products between men and women also differs a lot (den Biggelaar, 1994a; Rocheleau, 1991; Rocheleau, 1992). The difi'erences are mostly due to the same difi‘erences in responsibilities. These gender differences have sigrificant influences on the successful implementation of agoforestry technologies. Women are the main source of labor for crucial activities like watering tree nurseries. However, these women have been observed to be unwilling to provide such labor where the main purpose for such labor was cash generation (Rocheleau, 1991). The other major difi'erence between men and women relates to land and tree tenure. In some societies in developing countries, land tends to be under the control of men (Chavangi et al, 1988; Footmann and Nabane, 1992; Jacobson, 1992) while in other 42 societies, land is supposed to be controlled by women (Rocheleau, 1987; Rocheleau 1988). Tree rights also tend to vary from society to society. In some cases the person who has rights to the land, also has exclusive rights to trees gowing on it. In other cases, a person may have exclusive rights to the land but having no rights or only part of the rights to the trees gowing on his land. These difi’erences have fundamental influences on the adoption of agoforestry technologies. Fortrnann and Nabane (1992) observed that women in Mhondoro District in Zimbabwe were not willing to plant trees because they did not have rights of tenure to the land. In the event of divorce, a woman lost all the rights of usage of trees which she planted on her husband’s land even if she was living in the same locality. On the other hand, men of the Luhya tribe in Kenya did not allow their women to plant trees on their land for fear of losing their rights to the land (Chavangi et al, 1988). Malawi has two distinct customary systems of land tenure. Most societies of the Southern and Central Regions practice matrilineal system of maniage, where the husband joins the wife. By definition, women control the land in these societies. Land is transferred fiom the mother to the daughters and not to the sons. People of the Northern Region practice patrilineal system of marriage, where the wife joins the husband. By definition, land is'controlled by men in this system. Land is transferred from the fither to the sons and not to daughters. These differences may have sigrificant influences on the adoption of recommended agoforestry technologies. 4 3 Age There is mixed evidence about the relationship between age and firmers’ willingress to innovate. Some studies show that there is a sigrificant negative relationship between farmer’s age and their willingress to adopt new innovations (Akinola, 1987; Akinola and Young, 1985; Polson and Spencer, 1991). This means that the younger the firmers, the higher the rate of adoption of new innovation. Bongunjoko (1991) on the hand, observed a positive relationship between firmers’ age and the adoption rate of new technologies. This particular observation meant that the older the firmer, the higher the adoption rate. Other studies found no relationship between the firmers’ age and adoption rate (Onyenwaku and Mbuba, 1991; Opare, 1977). Education Wharton (1963) discussed the importance of education on farmer’s ability to adopt new technologies. He argued that a literate person is much easier to train as compared to an illiterate person. Education, apart from helping farmers to acquire simple skills like writing, reading and simple calculations which facilitate communication, it increases firmer’s inquisitiveness which results in self discovery of knowledge concerning farming. It widens the scope of decision making, stimulates motivation and induces fi'ustration which usually leads to heightened personal and political activity with some important political consequences. Education enables the farmer to engage in the general process of improved rationality or thinking through problems and not merely accepting them as unchangeable givens. With education, firmers are able to question their value systems and cultural 4 4 weaknesses. Educated people have a good attitude to change and are most often willing to seek and try new information. Noor (1980) indicated that if education is relevant to the needs of the people, it leads to more production. Despite these assertions, studies which have tried to investigate relationships between education and firmer adoption have produced mixed results. A positive relationship between firmer’s level of education and rate adoption of innovations was observed in some cases (Akinola, 1987; Atala, 1984; Marsh and Coleman, 1955; Napier et al, 1984; Osunji, 1981). In other cases there was no relationship between firmers’ level of education and rate of adoption of innovations (Ashby, 1982; Bongunjoko, 1981; Fett, 1971; Onyenwaku and Mbuba, 1991). Income Some innovations require capital in order to be implemented and are therefore not suitable for resource poor farmers. A good example is that of improved maize varieties. For a farmer to gow an improved maize variety, he/she needs to have sufiicient amount of capital for purchasing seed and other inputs like fertilizers. Most of the resource poor firmers would not have the capital required for the implementation of such technologies (Carr, 1988; Mkandawire, 1988; hfltandawire and Chipande, 1988). Landholding size Shortage of land is a major constraint under which smallholder farmers in Malawi operate. Average landholding sizes are very small (Bunderson et al, 1993; Jones et al, 1993; Malawi Government, 1987) so that it is very difficult for firmers to produce enough food to feed their families (Masangano, 1989; Mkandawire, 1988). Shortage of land has sometimes been a hindrance to farmers’ ability to adopt new technologies (Carr, 1988; 45 Mkandawire and Chipande, 1988). Studies in various places have shown a positive relationship between landholding size and willingness to adopt new innovations (Akinola, 1987; Onyenwaku and Mbuba, 1991; Opare, 1977; Polson and Spencer, 1991). Mkandawire (1988) observed that firmers with very small landholding sizes filled to utilize credit packages2 provided by the Ministry of Agiculture. The packages were too large for their landholding sizes. Adoption of agoforestry technologies could be influenced by the problem of small landholding sizes in Malawi. One way that adoption would be influenced by this problem would be the mere fict that the woody species use up some of the land. This implies a reduction of the land area left for crop production and consequently a reduction in crop production. Land and tree tenure system There are three land tenure systems in Malawi including: customary land, leasehold land and fi'eehold land. The customary land is regulated by customary law. Essentially land belongs to the community under the custodianship of the local chiefs who allocate it to their subjects. Once allocated, the occupants have rights to utilize the piece of land indefinitely and also have rights of gratuitous transfer and inheritance. (Mkandawire, 1983) provided a good description of customary land in Malawi. Land under this category is regarded as not having a market value, except the value fi'om its capacity to produce a crop. Otherwise the occupants can not sell the land nor dispose it ofi‘ to an 2 A credit package consists of a collection of inputs like fertilizers, seeds and chemicals recommended for a particular crop on a specific hectarage. Such a package is issued to farmers on loan. 4 6 outsider (non member of the community). By custom the occupants do not own the land which they cultivate, but have rights of control over it. Its possession transcends an individual’s lifetime as it is held to belong to the living, the dead and the unborn. Occupants deemed to be behaving in a manner contrary to the community’s norms can have their land forfeited by the village headrrran. Leasehold land is held on lease for 21 years, with a number of specified standard covenants relating to such matters as agricultural use, soil conservation and tree planting. The lease is reviewed every seven years (Malawi Government, 1987). Most of the leasehold land belongs to large estate owners, most of which being foreigrers who obtained it in the pre—colonial and colonial periods. Very little land has been converted to freehold system in the post-colonial era. One district, Lilongwe, has a special extended firnily fi'eehold system. This was introduced in 1972 on a pilot basis and only had limited success (Mkandawire, 1983). The bulk of the land (59,281 out of the 71,475 square kilometers) is under customary land tenure. This is the land on which smallholder firmers occupy and cultivate. Perceived weaknesses of this land tenure system include a lack of security of tenure and lack of concern among the farmers on long term matters such as deforestation and soil erosion (Malawi Government, 198 7). One implication of this would therefore be that since farmers don’t perceive secure rights of land tenure, they may not be interested to invest in long term technologies like those of agoforestry. Associated with the land tenure issue is tree tenure system. Being that the customary land belongs to the whole community, each member of the community has 4 7 equal access to any natural trees gowing on it. This is especially manifested in virgin land where members of the community just go in and out to cut trees or harvest any of the tree products like fruits and medicinal plants. Only in those few cases where the village headman regulates the usage of the natural forests is the situation difi‘erent. In most cases, such regulations do not exist. This may have posed problem to the adoption of agoforestry technologies. Farmers may have thought that they may not have security to the trees which they plant and therefore not be willing to adopt the agoforestry technologies. Summary of chapter In summary, this chapter has provided a review of the literature on: food situation in Malawi; advantages and disadvantages of agoforestry technologies; agoforestry technologies recommended and problems of low firmer practice of selected agoforestry technologies in Malawi. The country is facing problems of shortage of land for cultivation, land degadation and low food productivity. One consequence of these problems is fiilure to meet the national food requirements. To address these problems, Malawi is advocating the practice of agoforestry technologies. However, firmer practice of those agoforestry technologies is very low and fictors which could be likely reasons for this situation have been discussed. Chapter III DESIGN AND METHODOLOGY Introduction The methodology of this study was desigred to obtain farmer perceptions of fictors influencing firmer decision to practice selected agoforestry technologies. Data was collected fiom two goups of farmers including firmers who were exposed to the Malawi Agoforestry Extension (MAFE) and fiom farmers who were not exposed to the MAFE project in Njolomole EPA othcheu RDP. Perceptions of the two goups of farmers regarding the influence of the selected fictors on firmer decision to practice the selected agoforestry technologies were compared. Male and female farmer perceptions were also compared. Interviews were the primary data collection technique. This chapter provides a detailed description of the methodology of the study. It is divided into the following sections: desigr; population; sampling procedures; instrumentation and data collection methods; model specification and data analysis. Design The dependent variable in the study was farmer practice of selected agoforestry technologies with farmer exposure to the MAFE project, firmer perceptions regarding the efi‘ect of selected agoforestry technologies on food crop yields, farmer perceptions regarding the effect of selected agoforestry technologies on the control of soil erosion, farmer perception regarding the effect of selected agoforestry technologies on availability of fuelwood and firmer perceptions regarding the effect of the selected agoforestry technologies on profitability as the independent variables. The influence of some 48 4 9 demogaphic variables on farmer decision to practice the selected agoforestry technologies were also investigated. The demogaphic variables included were, gender, firmer’s age, highest level of formal schooling attained, total income received in the previous year (1996), landholding size, land and tree tenure system. Population Malawi is divided into eight agicultural development divisions (ADDs). These divisions are firrther divided into 30 rural development projects (RDPs) which are further divided into 173 extension planning areas (EPAs). This study was conducted in Njolomole EPA othcheu RDP in Lilongwe ADD. This EPA was chosen for several reasons. It is one of the EPAs where the MAFE project was piloted. It is also an EPA which lies on the Kirk Range where most of the land is on steep slopes and erosion hazard is major problem. ’It is also one of the highly populated EPAs such that most of the firmers are cultivating land which is not suitable for agiculture. The need for agoforestry technologies is very high in this EPA It was therefore chosen as an ideal place for this study. The EPA has 9,236 farm households with eight village extension workers, one soil conservation assistant, one women’s progams assistant and two supervisors. The crops which were grown in the EPA included maize, beans, goundnuts, soya beans, potatoes and cassava (Project crop estimate report, 1997). The EPA experiences high rainfall with annual precipitation ranging fi'om 1000 mm to 1800 mm. Seasonal food shortages is a major problem in the EPA. Thirty percent of the farm households had run out of their food stocks by January 1997. 50 The population of the study included all 9,236 firm households. This population was divided into two goups including 64 farm households who were exposed to the Malawi Agoforestry Extension Project and 9,172 households who were not exposed to the MAFE project. Sampling procedures The village extension workers with the help of the enumerators used in this study compiled a list of all the firm households in the EPA This efi’ort produced a list of 7,748 firm households. The samples used in this study were therefore drawn fi'om this list of 7,748 firm households. The first sample to be drawn was respondents not exposed to the MAFE project. A total of 385 firm households were sampled using random numbers. This sample size was determined using tables provided by the Educational and Psychological Measurement (1970). The other sample constituted all the 64 firm households who were exposed to the MAFE project. These two samples were split into halves using systematic sampling procedure. One male member of each household was interviewed fi'om half of the samples and one female member of each household was interviewed fiom the other half of the samples. Instrumentation and data collection methods An interview questionnaire was used to measure respondent’s perceptions the influence of the selected agoforestry technologies on food crop yields, the control of soil erosion, fuelwood availability and exposure to the MAFE project (see appendix F). The instrument was pilot tested in Bembeke EPA of Dedza Hills RDP. This EPA has similar geogaphical and climatical features to Njolomole EPA The purpose of pilot testing the 51 instrument was to test the suitability and reliability of the instrument. A test-retest reliability r of 0.79 was obtained. The instrument was also reviewed by one agoforestry researcher and one extension specialist and with minor changes made they both agreed that it measured what was intended to be measured. Data collection was by interviews. A team of five male and four female interviewers was recruited by the researcher in February 1997. These interviewers were trained by the researcher with assistance of two extension stafl‘ in the first week of March 1997. The main objectives of the training were to acquaint the interviewers with the objectives of the questionnaire, the selected agoforestry technologies and interview techniques. Each item on the questionnaire was thoroughly discussed and mock interviews were conducted. Issues of general ethics including confidentiality of the information, obtaining respondent’s consent and building rapport were also discussed. Actual interviews were conducted from the second week of March to the second week of April 1997. The male interviewers interviewed male respondents while the female interviewers interviewed the female respondents. Interviewers had to make prior arrangements with the respondents making sure that the interviews were conducted at convenient times for the respondents. Respondents were interviewed in the absence of members of the family of the opposite sex to avoid responses which were influenced by their presence. This was done where necessary by politely asking such other members of the fimily to leave as the interviews were being conducted. A total of 402 out of the targeted 449 respondents participated in the interviews. This consisted of 64 respondents (100%) exposed and 338 respondents (87.8%) not 52 exposed to the MAFE project. Three questionnaires were rejected because they were not properly completed. Reasons for non-respondents were refirsals and untraceable cases. Out of the 399 respondents whose responses were finally used in the analysis, 203 were male and 196 were female. Model specification and data analysis a) Analytical framework The choice of an appropriate model for this study was mainly dictated by the fict that firmer practice of a particular agoforestry technology was a binary dependant variable. It only took two forms of either practicing or not practicing. The appropriate models used in such cases are multiple discriminant analysis or linear probability models (Hair et al, 1995). These models specify a linear relationship between the probability of a particular event and the various predictors. The independent variables took two different forms of measurement. Landholding size, education and age were measured as continuous variables, while total income received in 1996, respondent exposure to the MAFE project, respondent perceptions regarding the influence of selected agoforestry technologies on food crop yields, control of soil erosion, fuelwood availability, respondent perceptions of soil erosion problems and shortage of firelwood were all measured either as dichotomous or categorical. The assumptions which the discriminant analysis model requires are essentially violated by this kind of combination of independent variables, (Hair et al, 1995; Nonrsis, 1994). Most specifically the requirement of multivariate normality can not be logically assumed with this combination of variables. Linear probability models were therefore preferred for the 5 3 study. Logistic regression model was therefore chosen for the study. Conceptually, the following was the behavioral model used to examine factors influencing firmer decision to practice of a particular agoforestry technology: Yl =1/1-e‘ (1) where Yl is probability of practice of agoforestry technology, 2 is the linear combination F B0+B1X1+B2X1+ ....... B,X, where B is the likelihood estimate (coemcient) and e is base of the natural logarithm, approximately 2.718. The probability of not practicing the particular agoforestry technology was estimated by; Yo =1-YI b) Model specification The main objective for the statistical analyses was to test the following hypotheses; 1. Respondents’ exposure to the MAFE project did not influence their decision to practice selected agoforestry technologies. 2. Respondents’ perceptions regarding theeffect of selected agoforestry technologies on food crop yields did not influence their decision to practice those agroforestry technologies. 3. Respondents’ perceptions regarding the efi‘ect of selected agoforestry technologies on the control of soil erosion did not influence their decision to practice those agoforestry technologies. 4. Respondents’ perceptions regarding the effect of selected agoforestry technologies on 5 4 their firelwood availability did not influence their decision to practice those agoforestry technologies. 5. Respondents’ perceptions regarding the effect of selected agoforestry technologies on profitability did not influence their decision to practice those agoforestry technologies. 6. The following demogaphic factors: gender, respondent’s age, respondent’s highest level of formal schooling attained, the amount of income the respondent received in 1996, landholding size, land tenure system and tree tenure system did not influence to their decision to practice selected agoforestry technologies. The model was specified as follows; Y, = e‘/1+e‘ where Y,=practice of a specific agoforestry technology. where z=B,,-1-B,X,+B,X,-l-B,X,+B,X,-l-B,X,+B,,X,,+B,X.,+B,X,+B,,X,,+B,,,)(,,,+B,,X,,+B,,X,2 +B,,X,,+B,,X,,+B,,X,,+B“X“, X,=exposure to the MAFE project, X,=respondent perception of the influence of the practice of the selected agoforestry technology on food crop yields, X,=respondent perceptions as to whether they experienced soil erosion problems on their firm X,whether respondents practiced the selected agoforestry technology in order to control soil erosion, X,=respondent perceptions as to whether the practice of the selected agoforestry reduced 5 5 soil erosion, X,=respondent perceptions as to whether they experienced problems of shortage of fuelwood, X,=whether respondents practiced the selected agoforestry technology in order to increase fuelwood, X,=respondent perceptions as to whether the practice of the selected agoforestry technology increased fuelwood, X,=respondent perceptions as to whether the practice of the selected agoforestry technology was more or less profitable than their previous practices, X,,=gender X,,=respondent’s highest level of formal education attained, X,,=respondent’s age, X,,=total income received in 1996 X,,=landholding size, X, ,=land tenure system and X,6=tree tenure system. c) Data analysis The data were analyzed using Statistical Package for the Social Sciences, advanced statistics version 6.1. The forward likelihood-ratio method was used in order to identify and remove the variables which were not important in the prediction of farmer practice of the technology. “The forward likelihood-ratio method involves estimating the model with each variable eliminated in turn and looking at the change in the log likelihood when each 56 variable is deleted. The likelihood-ratio test for the null hypothesis that the coefi'rcients of the terms removed are 0 is obtained by dividing the likelihood of the reduced model by the likelihood of the full model” (Norusis, 1994). The analysis for each practice was run several times, each time excluding all those variables which were removed by the preceding analysis. This was done in order to come up with better fitting and stronger predictive models. The alpha level was set at p=.05. Respondent’s highest level of formal schooling was constantly discarded in the five models due to problems of missing data. Chapter IV ANALYSIS AND RESULTS Introduction The main purpose of this study was to identify the factors influencing farmer’s decision to practice selected agoforestry technologies in Njolomole EPA othcheu RDP. Two specific objectives of the study were: 1. To investigate the influence of five selected factors: firmer exposure to the MAFE project; food productivity; soil erosion control; fuelwood productivity; and, profitability, on firmers’ decisions to practice selected agoforestry technologies in Njolomole EPA of Ntcheu RDP. 2. To investigate the influence of six demogaphic fictors: gender, age; education; income; landholding size and land and tree tenure system, on firmer decision to practice agoforestry technologies in Njolomole EPA of Ntcheu RDP. These objectives were firrther broken down to a set of six hypotheses. This chapter presents the data and results of the study starting with demogaphic information about the sample, descriptive statistics and finally the chapter discusses predictive models using logistic regessions. The chapter is divided into the following sections; 1. Introduction 2. Demogaphic information about the sample 3. Distribution of respondents practicing selected agoforestry technologies. 4. A comparison of respondents who were and those who were not exposed to the MAFE 57 5 8 Project regarding their practice of selected agroforestry technologies. 5. A comparison of male and female respondents practicing selected agoforestry technologies. 6. Respondent perceptions regarding the effect of selected agoforestry technologies on food crop yields. 7. Respondent perceptions regarding the effect of selected agroforestry technologies on the control of soil erosion. 8. Respondent perceptions regarding the efl‘ect of selected agoforestry technologies on fuelwood availability. 9. Respondent perceptions regarding the effect of selected agoforestry technologies on profitability. 10. Hypotheses testing using logistic regession models. Comparisons of respondent perceptions of the efi‘ect of selected agoforestry technologies on food crop yields, soil erosion, firelwood availability and profitability were made between respondents who were and those who were not exposed to the MAFE project. Similar comparisons were made between male and female respondents. The agoforestry technologies were gouped according to the basic purposes for which they were desigred and implemented. The practices of systematic interplanting with F. albida, alley cropping, relay cropping and improved fillows were gouped together as technologies designed for soil improvement. The practice of contour hedgerows was discussed separately as a technology desigred for the control of soil erosion. The practices of woodlots, planting and caring for trees in garden boundaries, 59 planting and caring for trees in croplands and planting and caring for trees at homesteads were gouped together as technologies designed for the production of fiuit, firele and other wood related products. Although fodder banks and living fences were technologies designed for two different purposes, they were gouped together for convenience. Fodder banks were basically designed for production of livestock feed while living fences were desigred for the provision of protection and privacy. Most of these agoforestry technologies fulfill more than one purpose and could have fitted in several of these goups. However they were restricted to these four goups mostly for purposes of convenience in reporting the results. Demographic information about the sample. A total of 402 out of the targeted 449 respondents participated in the study. This consisted of 64 respondents who were exposed and 338 respondents who were not exposed to the MAFE project. Three questionnaires were rejected because they were not properly completed. Reasons for non-respondents included, 20 refirsals and 27 untraceable cases. The age of the respondents ranged fiom 18 to 94 years with a mean age of 43.6 years. The number of years that the respondents spent in formal schooling ranged fiom zero to 14 with a mean of 3.97 years. They had very small land holdings with an average landholding size of 1.22 hectares ranging from a minimum of 0.1 to a maximum of 8 hectares. Based on recall information, seventy percent of the respondents received an 60 income which was less than 2,000.00 Malawi Kwacha3 fiom various sources in 1996. A DISTRIBUTION OF RESPONDENTS PRACTICIN G SELECTED AGROFORESTRY TECHNOLOGIES. There were relatively higher proportions of respondents who practiced technologies desigred for the control of soil erosion and the production of fruit, fuelwood and other products in the EPA (Table 4.1). Contour hedgerows were practiced by 39.6% of the respondents, woodlots by 41.6% planting and caring for trees in croplands by 49.9% and planting and caring for trees at homesteads by 56.9% of the respondents. The rest of the technologies were practiced by less than forty percent of the respondents. Respondents cited lack of exposure to information and demonstrations by extension staff as the major reasons for not practicing the selected agoforestry technologies. Appendix A Table A1 provides a distribution of respondents who were provided with information and or saw demonstrations on the selected agoforestry technologies as mounted by the extension staff. Thirty nine percent of the respondents cited lack of input as the reason for not practicing the agoforestry technologies. The most commonly cited input in short supply was F. albida seedlings, which resulted directly into a low proportion of respondents who practiced systematic interplanting with F. albida Other fictors cited less frequently as reasons for fiilure to practice selected agoforestry technologies were shortage of land, livestock and termite damage and some farmers believed that trees gew too slowly for their needs. 3 One United States Dollar was equivalent to 15.29 Malawi Kwacha at the time of the field data collection. (Published exchange rates by the Reserve Bank of Malawi, Nation Newspaper, April 23, 1997). 61 Table 4.1: A Distribution of Respondents Practicing Selected Agoforestry Technologies in Njolomole EPA othcheu RDP. Agoforestry Number Percent Number not Percent not technology practicing practicing practicing practicing FA 15.5 337 84.5 AC 35 8.8 364 91.2 RC‘ 52 13.0 347 87.0 IF’ 88 22.1 3 1 1 77.9 CH 158 39.6 241 60.4 WL 166 41.6 233 54.8 TGB 1 l 1 27.8 288 72.2 TC 199 49.9 200 50.1 TH 227 56.9 172 43.1 FB 30 7.5 369 92.5 LP 79 19.8 320 80.2 Note:FAisF.Albr'da,ACisalleycropping,RCisrelaycropping,IFisimprovedfallows,CHiscontourhodgerows,WLis woodlots,TGBisplantingandcaringfortreesingardenboundaries,TCisplantingandcaringfortrsesincroplands,mis phnfingmdunngfmueesathaneueads,FBisfodderbanksmdLFisfivingfencu. ‘ Most respondents did not seem to understand this technology and efforts to explain it were limited to verbal description due to lack of posters and pictures. This problem seems to have resulted in over-reporting of the number of respondents practicing relay cropping in this study. Some agoforestry experts in Malawi think that the proportion of farmers practicing this technology is much less than reported here (Itirnu 1997 in press). The other technologies did not suffer this problem since each interviewer had a copy of the field manual for the agoforestry practices in Malawi which has colorful pictures of the technologies. 5 Some respondents practiced woodlots on pieces of land they were not currently using for crop production as a way of preserving it until the time they need it for cultivation. Most of those respondents considered such a practice both as woodlots and improved fallows. This may have resulted in repeat reporting in this study and hence over-reporting the number of improved fallows in the EPA The literature shows that there were very few respondents practicing improved fallows in the EPA (Bunderson et al, undated). In order to provide a general picture of the extent to which the technologies were practiced, the respondents were asked to indicate the plot sizes on which they practiced the agroforestry technologies. On the other hand, five percent of respondents had their plots paced by the interviewers in order to come up with interviewer estimates of plot sizes. The interviewer estimates were compared with the respondents’ estimates and these two estimates had an error margin of fifteen percent. Generally the respondents tended to under estimate their plot sizes. Table 4.2 below provides information about respondents estimates of plot sizes on which the selected agoforestry technologies were practiced. Table 4.2: Estimated Plot Sizes Where the Selected Agoforestry Technologies were 62 Practiced in Njolomole EPA othcheu RDP. Agoforestry Minimum Maximum Mean plot size technology (hectares) (hectares) (hectares) FA 0.02 1.60 0.56 AC 0.01 1.40 0.66 RC 0.08 1.60 0.48 IF 0.04 6.00 0.63 CH 0.08 3.00 0.73 WL 0.01 2.40 0.39 TGB 0.04 4.00 0.60 TC 0.04 2.40 0.74 TH 0.04 1.60 0.21 FB 0.08 3.00 0.77 LF 0.10 1.60 0.21 Note: FAisF. albida. AC isalleycropping,RCisrelaycroppingandIFisimpmved fallows,Chiscontourhedgerows,WL iswoodlots,TGBistreesingardenboundaries,TCistreesincroplands,THistreesathomesteads,FBisfodderbanksand LF is living fences. 6 3 The technologies were generally practiced in very small plots with mean plot sizes of less than one hectare. The smallest plot sizes being for the practices of planting and caring for trees at homesteads and living fences. These two technologies were mostly practiced around firmers dwelling compounds. Such compounds usually occupy small pieces of land. A COMPARISON OF RESPONDENTS WHO WERE AND THOSE WHO WERE NOT EXPOSED TO THE MAFE PROJECT REGARDING THEIR PRACTICE OF SELECTED AGROFORESTRY TECHNOLOGIES. This section presents findings on the influence of exposure to the MAFE project on respondent’s decisions to practice the selected agoforestry technologies. Comparisons are made between respondents who were exposed to the MAFE project and those who were not regarding their practice of the selected agoforestry technologies. The discussion is divided into four sections according to the purposes for which the technologies were designed. There were generally higher percentages of respondents exposed to the MAFE project who perceived that they were exposed to information (Appendix A Table A2) and seen demonstrations provided by extension stafi‘ (Appendix A Table A3) than those of respondents not exposed to the project. 6 4 a) Comparison of farmers exposed to the Malawi Agroforestry Extension Project and those not regarding their practice of selected agroforestry technologies designed for soil improvement. There were significantly higher percenteges of respondents who were exposed to MAFE project and practiced the selected agoforestry technologies designed for soil improvement as compared to the percentages of respondents who were not exposed to MAFE project (p=<.05), see Table 4.3 below. Systematic interplanting with F. albida was practiced by 60.9% of the respondents exposed to the MAFE project as compared to 6.9% of the respondents not exposed. Alley cropping was practiced by 29.7% of those exposed as compared to 4.8% of those not exposed. Relay cropping was practiced by 32.9% of those exposed as compared to 8.7% of those not exposed and improved fillows were practiced by 32.8% of those exposed as compared to 20.0% of those not exposed. An evaluation study conducted in all five areas where the MAFE project was piloted showed that systematic interplanting with F. albida was practiced by fifty eight percent, alley cropping was practiced by seventy two percent and improved fallows was practiced by nine percent of the respondents exposed to the MAFE project (Bunderson et al, undated). 65 Table 4.3: A Comparison of Farmers Exposed to the Malawi Agoforestry Extension Project and those not Regarding their Practice of Selected Agoforestry Technologies Desigred for Soil Improvement in Njolomole EPA of Ntcheu RDP. Agoforestry Exposed to MAFE Not exposed to MAFE project technology-soil project (F64) (n=335) unprovement number number not number number not practicin practicing practicing practicing FA" 39 ' 25 23 312 (60.9) (39.1) (6.9) (93.1) AC” 19 45 16 3 19 (29.7) (70.3) (4.8) (95.6) RC" 23 41 29 306 (32.9) (64.1) (8.7) (91.3) IF ' 21 43 67 268 (32.8) (67.2) (20.0) (80.0) df-l for all technologies. ” significant at p-.01, ‘ significant at p-.05 Note: Nrnnbers in the brackets are percentages. F. A. is Faidherbr‘a albida, AC is alley cropping, RC is relay cropping and [F is improved fallows. b) Comparison of respondents exposed and those not exposed to the Malawi Agroforestry Extension Project regarding their practice of contour hedgerows. The percentage of respondents exposed to the MAFE project practicing contour hedgerows was sigrificantly higher than that of respondents who were not exposed to the MAFE project (p=<.01) Table 4.4. The technology was practiced by 85.9% of the respondents exposed to MAFE project as compared to only 30.7% of those not exposed. The evaluation study conducted in all the five areas where the MAFE project was piloted revealed that fifty six percent of the farmers exposed to the project practiced contour hedgerows (Bunderson et al, undated). The percentage of respondents exposed to the MAFE project practicing contour hedgerows was much higher in Njolomole EPA as 6 6 compared to the other areas where the MAFE project was piloted. Conversely the percentage of respondents exposed to the MAFE project and practicing alley cropping in Njolomole EPA was much lower than that of respondents from other piloted areas. This can be explained by the fact that being in the Kirk Range, farmers were more likely to practice contour hedgerows for the sake of controlling soil erosion rather than alley cropping which is more suitable in flat lands. Table 4.4: A Comparison of Respondents Exposed and those not Exposed to the Malawi Agoforestry Extension Project Regarding their Practice of Contour Hedgerows in Njolomole EPA of Ntcheu RDP. Agoforestry Exposed (n=64) Not exposed (n=33 5) technology-soil erosion control number number not number number not practicing practicing practicing gracticing . C u 55 9 103 232 (85.9) (14.1) (30.7) (69.3) df-l ” significant at p-.Ol, ‘ sigrificant at p-.05 Note: Numbers in brackets are percentages. CH is contour hedgerows c) Comparison of respondents exposed and those not exposed to the Malawi Agroforestry Extension Project regarding their practice of selected agroforestry technologies designed for the production of fruit, fuelwood and other products. The percentages of respondents exposed to the MAFE project who practiced woodlots, trees in garden boundaries and trees in croplands were sigrificantly higher than the percentages of respondents not exposed to the MAFE project (p=<.05) see Table 4.5. Trees at homesteads was the only agoforestry technology which was widely practiced by both the respondents exposed to the MAFE project (64.1%) and those who were not exposed to the project (55.5%). 67 Table 4.5: A Comparison of Respondents Exposed and those not Exposed to the Malawi Agoforestry Extension Project Regarding their Practice of Selected Agoforestry Technologies Desigred for the Production of Fruit, Fuelwood and other Products in Njolomole EPA othcheu RDP. Agoforestry Exposed (n=64) Not exposed (n=335) technology-fink and other products number number not number number not practicinL practicing practicgg practicing WL” 39 25 127 ' 208 (60.9) (39.1) (37.9) (62.1) TGB' 25 39 86 249 (39. 1) (60.9) (25.7) (74.3) TC” 57 7 142 193 (89.1) (10.9) (42.4) (57.6) TH 41 23 186 149 (64.1) (35.9) (55.5) (44.5) df-l foralltechnologies. ” significantatp-Dl,‘ significantatp-.05. NotezNumbersinbracketsarepercentages. WLiswoodlots,TGBisplantingu1dcuingforueesingudenboundaries,TC isplantingandcaringfortreesincroplandsandTHisplantingandcaringfortreesathornWads. d) Comparison of respondents exposed and those not exposed to the Malawi Agroforestry Extension Project regarding their practice of selected agroforestry technologies designed for production of livestock feed and protection. The percentages of the respondents exposed to the MAFE project who were practicing both fodder banks and living fences were sigrificantly higher than the percentages of respondents not exposed to the MAFE project (p=<.01) see Table 4.6. 68 Table 4.6: A Comparison of Respondents Exposed and those not Exposed to the Malawi Agoforestry Extension Project Regarding their Practice of Selected Agoforestry Technologies Designed for Production of Livestock Feed and Protection in Njolomole EPA othcheu RDP. Agoforestry Exposed (n=64) Not exposed (n=33 5) technology- livestock feed 39d number number not number number not protection practicing practicing practicing practicing FB" 28 36 2 333 (43.8) (56.6) (0.6) (99.4) LF'”I 30 34 49 286 (46.9) (53.1) (14.6) (85.4) ” n'gnificant at p-.01, ' sigrificant at p-.05. NotezNumbersinbracketsarepercentages. FBisfodderbanksandLFislivingfences. A COMPARISON OF MALE AND FEMALE RESPONDENTS PRACTICIN G SELECTED AGROFORESTRY TECHNOLOGIES. This section presents findings on the influence of gender on the practice of selected agoforestry technologies in Njolomole EPA Comparisons are made between male and female respondents regarding their practice of selected agoforestry technologies. Discussions are divided into the four goups according to the purposes for which the agoforestry technologies were designed. a) Comparison of male and female respondents regarding their practice of selected agroforestry technologies designed for soil improvement. The percentage of male respondents practicing relay cropping and improved fillows were sigrificantly higher than those of female respondents practicing these two technologies (p=<.01) see Table 4.7. No differences were observed between male and 6 9 female respondents regarding their practice of systematic interplanting with E albida and alley cropping. Table 4.7: A Comparison of Male and Female Respondents Regarding their Practice of Selected Agoforestry Technologies Designed for Soil Improvement in Njolomole EPA of Ntchenr RDP. Agoforestry Male (n=203) Female (n=196) technology- soil improvement number number not number number not practicinL practicing practicing practicing . FA 30 173 32 121 (14.8) (85.2) (16.3) (83.7L AC 23 180 12 184 (11.3) (88.7) (6.1) (93.9) RC“ 36 167 16 180 (17.7) (82.3) (8.2) (91.8) IF" " 57 146 3 1 165 (28.1) (71.9) (15.8) (84.2) df-l forall technologies. ” significant at p-.01, ‘ significant at p-.05. Note: Numbersinbracketsareperccntages. FA isFar‘rflnerbr‘a albida,AC isalleyaoppingRC isrelaycroppingandIFis improved fellows. b) Comparison of male and female respondents regarding their practice of contour hedgerows. The percentage of male respondents practicing contour hedgerows was significantly higher than that of female respondents (p=<.05) see Table 4.8. 70 Table 4.8: A Comparison of Male and Female Respondents Regarding their Practice of Contour Hedgerows in Njolomole EPA of Ntcheu RDP. Agoforestry Male (n=203) Female (n=196) technology-soil erosion control number number not number number not practicing practicifl practicing practicing CH‘I 92 1 11 66 130 (45.3) (54.7) (33.7) (66.3) df-l, ” significant at p-.01, ‘ significant at p‘.05. Note: Numbers in brackets are percentages. CH is contour hedgerows. c) Comparison of Male and Female Farmers Practicing Selected Agroforestry Technologies Designed for the Production of Fruit, Fuelwood and other Products. The percentages of male respondents practicing agoforestry technologies designed for the production of fruit, fuelwood and other products were significantly higher than those of female respondents (p=<.01) Table 4.9. Woodlots were practiced by 57.1% of the men and 25.5% of the women. Planting and caring for trees in garden boundaries were practiced by 33.5% of the male and 21.9% of the female respondents. Planting and caring for trees in croplands were practiced by 65.0% of the male and 34.2% of the female respondents while planting and caring for trees at homesteads were practiced by 68.5% of the male and 44.9% percent of the female respondents. 71 Table 4.9: A Comparison of Male and Female Farmers Practicing Selected Agoforestry Technologies Designed for the Production of Fnrit, Fuelwood and other Products in Njolomole EPA of Ntcheu RDP. Agoforestry Male (n=203) Female (n=196) technology-fiuit & other products number number not number number not practicing practicing practicing practicing WL" 1 16 87 50 146 (57. 1) (42.9) (25.3 (74.5) TGB ‘”" 68 135 43 153 (33.5) (66.5) (21.9) (78.1) TC“ 132 71 67 80 (65.0) (35.0) (34.2) (65.8) TH" 139 64 88 108 (68.5) (31.5) (44.9) (55.1) df-l forallteehnologies. “ significantatp-Dl, ‘ signifrcantatp-.05. Note:Numbersinbracketsarepercentages. WLiswoodlots,TGBisplantingandcaringforh'eesingardenboundaries,TC isplantingandcaringfortreesincroplandsandTHisplantingandcaringfortreesathomemads. d). Comparison of male and female respondents practicing selected agroforestry technologies designed for the production of livestock feed and protection. The percentage of male respondents practicing living fences was significantly higher than that of female respondents in the EPA (p=<.01), see Table 4.10. There was no significant difference between the percentages of male and female respondents practicing fodder banks. 72 Table 4.10: A Comparison of Male and Female Respondents Practicing Selected Agoforestry Technologies Designed for the Production of Livestock Feed and Protection in Njolomole EPA othcheu RDP. Agoforestry Male (n=203) Female (n=196) technology- livestock feed & number number not number number not protection practicing practicing practicing practicing FB 20 183 10 186 (9.9) (90.1) (5. D 93.9) LI"I * 57 146 3 l 165 (28.1) (71.9) (15.8) (84.2) df-l forall technologies. ” significantat p-.01, ‘ significantat p-.05. NotezNumbersinbracketsarepercentages.FBisfodderbanksandLFislivingfenccs. RESPONDENT PERCEPTIONS REGARDING THE EFFECT OF SELECTED AGROFORESTRY TECHNOLOGIES ON FOOD CROP YIELDS This section discusses respondents’ perceptions regarding the efi‘ect of selected agoforestry technologies on yields of their food crops. Distributions of respondent perceptions regarding the effect of the selected agoforestry technologies on food crop yields are discussed. Comparisons are made between those respondents exposed to the MAFE project and those not exposed. Similarly, comparisons were made between male and female respondents. There was a high percentage of respondents who were either not sure or declined to respond to items on the effect of selected agoforestry technologies on food crop yields. Two possible reasons for this were that either the respondents were not currently practicing the technologies or they had not practiced the technologies long enough to make informed judgements on them. The discussions are divided into four 7 3 subsections according to the purposes for which the agoforestry technologies were designed. a) Distribution of respondent perceptions regarding the effect of selected agroforestry technologies designed for soil improvement on food crop yields Table 4.11 below provides a distribution of respondents’ perceptions of the efi‘ect of selected agoforestry technologies designed for soil improvement on food crop yields. There were generally higher percentages of respondents who perceived the practices of selected agoforestry technologies designed for soil improvement as having resulted in increased food crop yield as compared to the respondents who perceived the selected agoforestry technologies as having resulted in reduction of their food crop yields. Table 4.11: A Distribution of Respondent Perceptions Regarding the Effect of Selected Agoforestry Technologies Designed for Soil Improvement on Food Crop Yields in Njolomole EPA of Ntcheu RDP. Agoforestry Increased yield Decreased yield Not sure/no response technology- soil improvement FA 42 24 333 (10.5) (6.0) (83.5) AC 24 1 374 (6.0) (0.3) (93.7) RC 37 8 354 (9.3) (2.0) (88.7) IF 71 4 324 (17.8) (1.0) (81.2) Note: Numbers in brackets are percentages. FA is Far'dherbr'a albr'da, AC is alley cropping, RC is relay cropping and IF is improved fallows. 7 4 Table 4.12 shows the percentages of respondents exposed to the MAFE project who perceived the practices of systematic interplanting with F. aIbr'da and relay cropping as increasing their food crop yields were significantly higher than those of respondents who were not exposed to the MAFE project (p=<.01). Table 4.12: A Comparison of Respondent Perceptions between Respondents who were and those who were not Exposed to the Malawi Agoforestry Extension Project Regarding the Efi’ect of Selected Agoforestry Technologies Designed for Soil Improvement on Food Crop Yields in Njolomole EPA of Ntcheu RDP. Agoforestry Exposed (n=64) Not exposed (n=335) technology- §°ll more less NR more yield less NR ““9“?“th yield yield yield FA" 27 3 34 15 21 299 (42.2) (4.7) (4.5) (6.3) AC 18 0 46 6 1 328 (28.1) (1.8) (0.3) RC" 19 0 45 18 8 309 (29.7) (5.4) (2.4) IF 15 0 49 56 4 275 (23.4) (16.7) (1.2) df-l for all the technologies. ” significant at p-.Ol, ‘ significant at p-.05. Note: Numbers in brackets are percentages. NR is non respondents, FA is Faidherbr‘a albida. AC is alley cropping, RC is relay cropping and IF is improved fallows. Table 4.13 shows that the percentage of male respondents who perceived the practice of systematic interplanting with F. albida as decreasing their food crop yields was significantly lower than that of female respondents who perceived the practice of systematic interplanting with F. albida as decreasing their food crop yields (p=<.01). 75 Table 4.13: A Comparison of Male and Female Respondent Perceptions regarding the Effect of Selected Agoforestry Technologies Designed for Soil Improvement on Food Crop Yields in Njolomole EPA othcheu RDP. Agoforestry Male (203) Female (196) technology- ?0“ more yield less NR more yield less NR improvement yield field FA" 27 2 174 15 22 159 (13.3) (1.9 (7.7L (11.2) AC 19 1 183 5 0.00 191 (94) L5) (2.6) RC 27 6 170 10 2 184 (13.3) £10) (5.1) (1.0) IF 49 3 151 22 l 173 (24.1) (1.5) (11.2) (0.5) df=1foralltechnologiea ” agnificantatp-Dl, ‘ significantatp-.OS Note: Numbers in brackets are percentages. NRis non respondents, FA isFar'rflrarbr‘a albida. AC is alley cropping,RC is relaycroppingandIFisirnprovedfallows , b) Distribution of respondent perceptions regarding the effect of contour hedgerows on food crop yields. Table 4.14 shows, the percentage of respondents who perceived the practice of contour hedgerows as helping to increase food crop yields was lnigher (27.3%) than the percentage of respondents who perceived the practice as causing a reduction in food crop yields (2.5%). 76 Table 4.14: A Distribution of Respondent Perceptions Regarding the Efi‘ect of Contour Hedgerows on Food Crop Yields in Njolomole EPA othcheu RDP. Agoforestry Increased yield Decreased yield Not sure/no technology-soil response erosion control CH 109 10 280 (27.3) (2.5) (70.2) Note: Numbers in brackets are percentages. CH is contour hedgerows There was no significant difi‘erence in the percentages of respondents who perceived the practice of contour hedgerows as helping to increase food crop yields between those respondents exposed and those not exposed to the MAFE project (see Table 4.15). Table 4.15: A Comparison of Respondent Perceptions between Respondents who were and those who were not Exposed to the Malawi Agoforestry Extension Project, regarding the Effect of Contour Hedgerows on Food Crop Yields in Njolomole EPA of Ntcheu RDP. Agoforestry Exposed (n=64) Not exposed (n=335) technology-soil . erosion control more yield less yield NR more yield less NR yield CH 39 3 22 70 7 258 (60.9) (4.7) (20.9) (2. 1) df-l, ” significant at p-.01, ‘ significant at p-.05 Note: Numbers in brackets are percentages. NR is non respondents CH represents contour hedgerows. 77 Similarly, there was no significant difference in the percentage of male and female respondents who perceived the practice of contour hedgerows as helping to increase food crop yields (see Table 4.16). Table 4.16: A Comparison of Male and Female Respondent Perceptions regarding the Efl‘ect of Contour Hedgerows on Food Crop Yields in Njolomole EPA of Ntcheu RDP. Agmfmm’ Male (n=203) Female (n=196) technology- zitegosron more yield less yield NR more yield less yield NR CH 76 5 122 33 5 153 (37.4) (2.5) (16.8) (2.6) df-l, ” significant at p-.01, ‘ significant at p-.05 NotezNumbersinbracketsarepereennges.NRisnonrespmdentsandCHiseontourhedgemws. c) Distribution of respondent perceptions regarding the effect of selected agroforestry technologies designed for the production of fruit, fuelwood and other products on food crop yields. The percentage of respondents who perceived the practice of agroforestry technologies designed for the production of fruit, fuelwood and other products as helping to increase food crop yields were higher than those of respondents who perceived the technologies as reducing food crop yields (Table 4.17). The percentage of respondents who perceived the practice of woodlots as helping to increase food crop yields was 29.3% while that of those who perceived it as reducing food crop yields was 4.5 percent. The percentage of respondents who perceived the practice of planting and caring for trees in garden boundaries as helping to increase food crop yields was 18.8% while that of 78 respondents who perceived the practice as reducing food crop yields was 4.5 percent. For planting and caring for trees in croplands 24.3% perceived it as helping to increase food crop yields as compared to 11.3% who perceived it as reducing food crop yields. Similarly, 42.6% of the respondents perceived the practice of planting and caring for trees at homesteads as helping to increase food crop yields as compared to 2.8% who perceived the practice as reducing food crop yields. Table 4.17: A Distribution of Respondent Perceptions Regarding the Efl‘ect of Selected Agroforestry Technologies Designed for the Production of Fruit, Fuelwood and Other Products on Food Crop Yields in Njolomole EPA of Ntcheu RDP. Agroforestry Increased yields Decreased yield Not sure/no response technology- fi'uit, firelwood & other products WL 117 18 264 (29.3) (4.5) (66.2) TGB 75 18 306 (18.8) (4.5) (76.7) TC 97 45 257 (24.3) (11.3) (64.4) TH 170 l l 218 (42.6) (2.8) (54.6) NotezNumbersinbracltetsarepereentages. WLiswoodlots,TGBisplsntingmdcsringforfieesingardenboundaries,TC isplantingandcaringt'ortreesincroplandsandmisplsntingandcaringfortrwsathomeswads. The percentage of respondents exposed to the MAFE project who perceived the practices of woodlots, planting and caring for trees in garden boundaries, planting and caring for trees in croplands as helping to increase food crop yields were significantly higher than the percentages of respondents who were not exposed to the MAFE project (p=<.05) see 7 9 Table 4.18. While there were generally high percentages of respondents who perceived the practice of trees at homesteads as helping to increase food crop yields, there was no significant difference in the percentages between respondents who were and those who were not exposed to the MAFE project. Table 4.18: A Comparison of Respondent Perceptions between Respondents who were and those who were not Exposed to the Malawi Agroforestry Extension Project, regarding the Effect of Selected Agroforestry Technologies Designed for Production of Fruit, Fuelwood and other Products on Food Crop Yields in Njolomole EPA of Ntcheu RDP. Agroforestry Exposed (n=64) Not (n=335) technology- fruit & 0th“ more less yields NR more less NR products yields yields yields WL“ 30 O 34 87 18 230 (46.9L (26.0) (5.4) TGB“ l9 0 45 56 18 261 (29.7) (16.7) Q4) TC“ 38 4 22 59 41 235 (59.4) (6.3) (17.6) (12.2) TH 33 0 3 l 13 7 1 1 187 (51.6) (40.9) (3.3) df-l for all technologies. ” significant at p-.01, ‘ significant at p-.05 Note: Numbers in brackets are percentages. NR is non respondents, WL is woodlots, TGB is trees on garden boundaries, TC istrcesincroplandssndmistrceathomeswads. There were no differences in the percentages of male and female respondents who perceived the practices of woodlots, planting and caring for trees in garden boundaries, planting and caring for trees in croplands and plating and caring for trees at homesteads as helping to increase food crop yields (Table 4: 19). 80 Table 4.19: A Comparison of Male and Female Respondent Perceptions regarding the Efl‘ect of Selected Agroforestry Technologies Designed for the Production of Fruit, Fuelwood and other Products on Food Crop Yields in Njolomole EPA of Ntcheu RDP. Agroforestry Male (n=203) Female (n=196) technology-hit, fuelwood & more yield less NR more yields less NR Other products yields yields WL 84 13 106 33 5 158 (41.4) (6.4) (16.8) (2.6) TGB 44 3 l 144 3 l 3 162 (21.7) (15.3) (15.8) (1.5) TC 73 37 93 24 8 164 (36.0) (18.2) (12.2) (4.1) TH 119 7 77 51 4 141 (58.6) (3.4) (26.0) (2.0) df-l foralltechnologies. ” significantatp-.01, ‘ signifiesntatp-.05 NotezNumbersinbrackebampacenugu.NRisnmrespmdenu,WLiswoodlot,TGBisucesmgardenboundaries,TC istr'cesincroplandsandTHistreesathomesteads. (1) Distribution of Respondent Perceptions Regarding the effect of Selected Agroforestry Technologies Designed for the Production of Livestock Feed and Protection on Food Crop Yields. As Table 4.20 shows, there was a higher percentage of respondents who perceived the practice of living fences (15.8%) as helping to increase food crop yields as compared to the percentage of those who perceived the practice as resulting in reduction of food crop yields (1.0). 81 Table 4.20: A Distribution of Respondent Perceptions Regarding the Effect of Selected Agroforestry Technologies Designed for the Production of Livestock Feed and Protection on Food Crop Yields in Njolomole EPA of Ntcheu RDP. Agroforestry increased yield Decreased yield Not sure/no technology- livestock response feed & protection FB 24 0 375 (6.0) (94.0) LP 63 4 332 (15.8) (1.0) (83.2) Note:Numbersinbracketsarepercentages.FBisfodderbanksandLFislivingfences. The percentage of respondents exposed to the MAFE project who perceived the practice of living fences as helping to increase food crop yield was significantly higher (43.8%) than the percentage of respondents not exposed to the MAFE project (10.4%) see Table 4.21. Table 4.21: A Comparison of Respondent Perceptions between Respondents who were and those who were not Exposed to the Malawi Agroforestry Extension Project, regarding the Effect of Selected Agroforestry Technologies Designed for Production of Livestock Feed and Protection on Food Crop Yields in Njolomole EPA of Ntcheu RDP. Agroforestry Exposed (n=64) Not exposed (n=335) technology- “VCStOCk f0.“ more less NR more less NR and protection yields yields yields yields FB 22 0 42 2 0 333 (34.4) (0.6) LP 28 0 36 35 4 296 (43.8) (10.4) (1.2) df-l for all technologiu. " significant at p=.01, ‘ significant at p-.05. Note: Numbers in brackets are percentages. NR is non respondents, FB is fodder banks and LF is living fences. 8 2 There were no differences in the proportions of male and female respondents who perceived the practices of fodder banks and living fences as helping to increase food crop yields in Njolomole EPA (see Table 4.22). Table 4.22: A Comparison of Male and Female Respondent Perceptions regarding the Efl'ect of Selected Agroforestry Technologies Designed for the Production of Livestock Feed and Protection on Food Crop Yields in Njolomole EPA of Ntcheu RDP. Agroforestry male (n=203) female (n=196) technology- livestock feed more less NR more less NR & protectlon yields yields yields yields FB 20 0.00 183 4 0.00 192 (9.9) (2.0) LP 54 4 145 9 0.00 187 (26.6) (2.0) (4.6) df-l for all technologies. ” significant at p=.01, ‘ significant at p-.05. Note: Numbersin bracketsarepercentages.NRisnonrespondents,FB represents fodder banksandLFrepresents living fences. RESPONDENT PERCEPTIONS REGARDING THE EFFECT OF SELECTED AGROFORESTRY TECHNOLOGIES ON SOIL EROSION. This section of the chapter discusses the perceptions of respondents regarding the effect of selected agroforestry technologies on the control of soil erosion. Distributions of respondents’ perceptions regarding the effect of the selected agroforestry technologies on the control of soil erosion are discussed. Comparisons of respondent perceptions are made between respondents exposed to the MAFE project and those not exposed as well as those of male and female respondents. There were high proportions of respondents who were either not sure or declined to respond to items on the influence of selected agroforestry technologies in reducing soil 8 3 erosion. Two reasons for this were either the respondents were not currently practicing the technologies or they had not practiced them long enough to be able to make informed judgements on them. The discussions are divided into four sections according to the purposes for which the technologies were designed. a) Distribution of respondent perceptions regarding the effect of selected agroforestry technologies designed for soil improvement in the control of soil erosion. The percentage of respondents who perceived the practice of relay cropping as having controlled soil erosion was higher (11.0%) than that of respondents who perceived the practice of this technology as not having reduced soil erosion (2.3%) see Table 4.23. Table 4.23: A Distribution of Respondent Perceptions Regarding the Efi'ect of Selected Agroforestry Technologies Designed for Soil Improvement in the Control of Soil Erosion in Njolomole EPA of Ntcheu RDP. Agroforestry Controlled Did not control Not sure/no technology- soil erosion erosion response improvement FA 39 25 335 (9.8) (6.3) (84.0) AC 20 8 371 (5.0) (2.0) (93.0) RC 44 9 346 (11.0) (2.3) (86.7) IF 28 11 360 (7.0) (2.8) (90.2) Note: Numbers in brackets are percentages. FA is Faidherbl'a albl'da, AC is alley cropping, RC is relay cropping and IF is improved fallows. 8 4 The percentages of respondents exposed to the MAFE project who perceived the practices of alley cropping and improved fallows as helping to reduce soil erosion were significantly higher than the percentages of respondents not exposed to the MAFE project (p=<.01) see Table 4.24. Table 4.24: A Comparison of Respondent Perceptions between Respondents who were and those who were not Exposed to the Malawi Agroforestry Extension Project, regarding the Effect of Selected Agroforestry Technologies Designed for Soil Improvement in Reducing Soil Erosion in Njolomole EPA of Ntcheu RDP. Agroforestry Exposed (n=64) Not exposed (n=335) technology-soil improvement reduced did not NR reduced did not NR erosion reduce erosion reduce erosion erosion FA 23 10 3 1 16 15 304 (35.9) (15.6) (4.8) (4.5) AC" 17 1 46 3 7 325 (26.6) (1 .6) (0.9) (2.1) RC 20 1 43 24 8 303 (31.3) (1.6) (7.2) (2.4) IF" 16 1 47 12 10 3 13 (25.0) (1.6) (3.6) (3.0) df-l for all technologies. ” significant at p=.01, ‘ significant at p=.05. Note: Numbers in brackets are percentages. NR is non respondents, FA is Fairflrerbr‘a. albida, AC is alley cropping, RC is relay cropping and IF is improved fallows. There were no significant differences between male and female respondents who perceived the practices of systematic interplanting with F. albida, alley cropping, relay cropping and improved fallows as helping to reduce soil erosion (Table 4.25). 85 Table 4.25: A Comparison of Male and Female Respondent Perceptions regarding the Efi‘ect of Selected Agroforestry Technologies Designed for Soil Improvement in Reducing Soil Erosion in Njolomole EPA of Ntcheu RDP. Agroforestry Male (n=203) Female (n=196) technology-soil . . improvement reduced dld not NR reduced dld not NR erosion reduce erosion reduce erosion erosion FA 19 11 173 20 14 162 (9.4) (5.4) (10.2) Q1) AC 15 4 184 5 4 187 (7.4) (2.9 (2.6) Q0) RC 31 6 166 13 3 180 (13.3) (3.0) (6.6) Q5) IF 22 6 175 6 5 185 (10.8) (3.0) (3.1) (2.6) ” significant at p-.01, ' significant at p-.05. Note: Numbers in brackets are perwntages. NR is non respondents, FA is Fal‘dherbl'a albl'da. AC is alley cropping, RC is relay cropping and IF is improved fallows. b) Distribution of respondent perceptions regarding the effect of contour hedgerows in the control of soil erosion. The percentage of respondents who perceived the practice of contour hedgerows as having helped to reduce soil erosion was higher (30.3%) than that of respondents who perceived it as not having reduced soil erosion (14.0%) see Table 4.26. 86 Table 4.26: A Distribution of Respondent Perceptions Regarding the Effect of Contour Hedgerows in the Control of Soil Erosion in Njolomole EPA of Ntcheu RDP. Agroforestry Controlled Did not control Not sure/no technology-soil erosion erosion response erosion control CH 121 56 222 (30.3) (14.0) (55.6) Note: Numbers in brackets are percentages. CH is contour hedgerows The percentage of respondents exposed to the MAFE project who perceived the practice of contour hedgerows as having controlled soil erosion was significantly higher than that of respondents who were not exposed to the MAFE project (p=<.01) see Table 4.27. Table 4.27: A Comparison of Respondents Perceptions between Respondents who were and those who were not Exposed to the Malawi Agroforestry Extension Project, regarding the Effect of Contour Hedgerows in Reducing Soil Erosion in Njolomole EPA of Ntcheu RDP. Agroforestry Exposed (n=64) Not exposed (n=335) technology- soil erosion control reduced did not NR reduced did not NR erosion reduce erosion reduce erosion erosion CH" 51 3 10 70 53 212 (79.7) (4.7) (20.9) (15.8) df=1, ” significant at p=.01, ‘ significant at p-.05 Note: Numbers in brackets are Percentages. NR is non respondents, CH is contour hedgerows. 87 There was no significant difference in the percentages of male and female respondents who perceived the practice of contour hedgerows as controlling soil erosion (Table 4.28). Table 4.28: A Comparison of Male and Female Respondent Perceptions regarding the Efl'ect of Contour Hedgerows in Reducing Soil Erosion in Njolomole EPA of Ntcheu RDP. Agroforestry Male (n=203) Female (n=196) technology-soil erosion control reduced did not NR reduced did not NR erosion reduce erosion reduce erosion erosion CH 69 31 103 52 25 l 19 (34.0) (15.3) (26.5) (12.8) df-l, ” significant at p-.01, ‘ significant at p-.05. NotezNumbersinbrackenarepereenhges.NRisnonrespmdaltsuldCHiscmmhedgerm. c) Distribution of respondent perceptions regarding the effect of selected agroforestry technologies designed for the production of fruit, fuelwood and other products in the control of soil erosion. The percentages of respondents who perceived the practice of selected agroforestry technologies designed for the production of fiuit, filelwood and other products as helping to control soil erosion were higher than the percentages of respondents who perceived the practices as not having helped to control soil erosion: woodlots 17.0% compared to 4.3%; trees in garden boundaries 13.0% compared to 6.0%; trees in croplands 18.8% compared to 8.5%; and, trees at homesteads 19.0% compared to 5.5%, see Table 4.29. Table 4.29: A Distribution of Respondent Perceptions Regarding the Efl‘ect of Selected Agroforestry Technologies Designed for the production of Fruit, Fuelwood and Other 88 Products in the Control of Soil Erosion in Njolomole EPA of Ntcheu RDP. Agroforestry Controlled Did not control Not sure/no technology- fi'uit, erosion erosion response filelwood & other products WL 68 17 3 14 (17.0) (4.3) (78.7) TGB 52 24 323 (13.0) (6.0) (81.0) TC 75 34 290 (18.8) (8.5) (72.7) TH 76 22 301 (19.0) (5.5) (75.4) NotezNumbersinbracketsarepelwntages. WLiswoodlots,TGBisplantingandcaringforueesingardenboundaries,TC isplantingandcaringfortreesineroplandsandmisplantingandcaringfortreesathomeswads. The percentages of respondents exposed to the MAFE project who perceived the practices of woodlots, planting and caring for trees in garden boundaries, planting and caring for trees in croplands and planting and caring for trees at homesteads as having controlled soil erosion were significantly higher than those of respondents not exposed to the MAFE project (p=<.01) see Table 4.30. 89 Table 4.30: A Comparison of Respondent Perceptions between Respondent who were and those who were not Exposed to the Malawi Agroforestry Extension Project, Regarding the Efi‘ect of Selected Agroforestry Technologies Designed for the Production of Fruit, Fuelwood and other Products in Reducing Soil Erosion in Njolomole EPA of Ntcheu RDP. Agroforestry Exposed (n=64) Not exposed (n=335) technology- fmit & other reduced did not NR reduced did not NR products erosion reduce erosion reduce erosion erosion WL" 28 1 35 40 16 279 (43 .8) (1 Q (I 1.9) (4.8) TGB" 21 1 42 31 23 281 (32.8) (1.6) (9.3) (6.9) TC" 40 4 20 35 30 270 (62.5) (6.3) (10.4) (9.0) TH“ 32 3 29 44 19 272 (50.0) (4.7) (13.1) (5.7) df'l foralltechnologies. ” significantatp-.01, ’ sigrrificantatp-.05. Note:Numbersinbracketsarepercentages.NRisnonrespondents,WLiswoodlots,TGBistreesongardenboundaries, TCistrwsincroplandsand'IHistreesathomesbads. There were no significant difl‘erences in the percentages of male and female respondents who perceived the practices of woodlots, planting and caring for trees in garden boundaries, planting and caring for trees in croplands and planting and caring for trees at homesteads as having helped to control soil erosion ( Table 4.31). 90 Table 4.31: A Comparison of Male and Female Respondent Perceptions regarding the Efl'ect of Selected Agroforestry Technologies Designed for the Production of F ruit, Fuelwood and other products in Reducing Erosion in Njolomole EPA of Ntcheu RDP. Agroforestry Male (n=203) Female (n=196) technology- fi'ult & other reduced did not NR reduced did not NR products erosion reduce erosion reduce erosion erosion WL 55 12 136 13 5 178 (27.1) (5.23 (6.6) (2.6) TGB 38 l 1 154 14 13 169 (18.7) (5.4) (7.1) (6.6) TC 53 29 121 22 5 169 (263 (14.3) (11.2) (2.5) TH 61 13 129 15 9 172 (30.0) (6.4) (7.7) (4.6) df-l foralltechnologies. “ significantatp-.01, ‘ significantatp-.05 NotezNumbersinhackenanpucenhges.NRisnonrespmdentsWLiswwdlots,TGBisueesongudalboundaries,TC istreesincroplandsand‘l'Histeesathomesbad. (1) Distribution of respondent perceptions regarding the effect of selected agroforestry technologies designed for the production of livestock feed and protection in the control of soil erosion. The percentage of respondents who perceived the practice of living fences as having controlled soil erosion was higher (14.3%) than that of respondents who perceived this practice as not having helped to control soil erosion (3.8%) see Table 4.32. 91 Table 4.32: A Distribution of Respondent Perceptions Regarding the Efi‘ect of Selected Agroforestry Technologies Designed for the production of Livestock Feed and Protection in the Control of Soil Erosion in Njolomole EPA of Ntcheu RDP. Agroforestry Controlled Did not control Not sure/no technology- livestock erosion erosion response feed & protection FB 24 6 369 (6.0) (1.5) (92.5) LF ' S7 15 327 (14.3) (3.8) (82.0) Note:Numbersinbracketsarepercentages.FBisfodderbanksandLFislivingfences. The percentages of respondents exposed to the MAFE project who perceived the practices of fodder banks and living fences as having helped to control soil erosion were higher than those of respondents not exposed to the project (p=<.01) see Table 4.33. Table 4.33 A Comparison of Farmer Perceptions between Farmers who were and those who were not Exposed to the Malawi Agroforestry Extension Project, Regarding the Effect of Selected Agroforestry Technologies Designed for the Production of Livestock Feed and Protection in Reducing Soil Erosion in Njolomole EPA of Ntcheu RDP. Agroforestry Exposed (n=64) Not exposed (n=33 5) technology- hvostook food reduced did not NR reduced did not NR & protection erosion reduce erosion reduce erosion erosion FB" 23 0 41 1 6 328 95.9) (0.3) (1.8) LP” 29 I 34 28 14 293 (45.3) (1.6) (8.4) (4.2) ” significant at p=.01, ‘ significant at p=.05. Note: Numbers in brackets are percentages. NR is non respondents, F3 is fodder banks and LP is living fences. 92 There were no significant differences between the percentages of male and female respondents who perceived the practices of fodder banks and living fences as having helped to control soil erosion (Table 4.34). Table 4.34: A Comparison of Male and Female Respondent Perceptions Regarding the Efi‘ect of Selected Agroforestry Technologies Designed for the Production of Livestock Feed and Protection in Reducing Soil Erosion in Njolomole EPA of Ntcheu RDP. Agroforestry Male (n=203) Female (n=196) technology- livestock food reduced did not NR reduced did not NR & protection erosion reduce erosion reduce erosion erosion FB 18 3 182 6 3 187 (8.9) (1.5) (3.1) (1.5) LF 49 1 1 143 8 4 184 (24.1) (5.4) (4.1) (2.0) df-l forall technologies. ” significantat p-.01, ‘ significant at p-=.05. NotezNumbersinbracketsarepercentages.NRisnonrespondents,FBisfodderbanksandLFislivingfences. RESPONDENT PERCEPTIONS REGARDING THE EFFECT OF SELECTED AGROFORESTRY TECHNOLOGIES ON FUELWOOD AVAILABILITY. This section of the chapter discusses respondent’s perceptions regarding the efi‘ect of selected agroforestry technologies on the availability of firelwood in Njolomole EPA. Distributions of respondent perceptions are discussed, followed by comparisons between perceptions of respondents who were and those who were not exposed to the MAFE project. Similar comparisons are also made between the perceptions of male and female respondents. 93 There were very high proportions of respondents who were either not sure or declined to respond to items on the effect of selected agroforestry technologies on fuelwood availability. Two reasons for this were either the respondents were not currently practicing the technologies or they had not practiced them long enough to be able to make informed judgements on them. The discussions are divided into four sections according to the purposes for which the technologies were designed. a) Distribution of respondent perceptions regarding the effect of selected agroforestry technologies designed for soil improvement on fuelwood availability. The percentages of respondents who perceived the practices of systematic interplanting with F. albida, alley cropping, relay cropping and improved fallows as having helped to increase filelwood availability were not difi‘erent fi'om the percentages of those who perceived the practices as not having helped to increase fuelwood availability (Table 4.35). 94 Table 4.35: A Distribution of Respondent Perceptions Regarding the Efi'ect of Selected Agroforestry Technologies Designed for Soil Improvement on Fuelwood Availability in Njolomole EPA of Ntcheu RDP. Agroforestry Increased Did not increase Not sure/no technology- soil fuelwood fuelwood response improvement FA 33 48 318 (8.3) (12.0) (79.7) AC 17 27 355 (4.3) (6.8) (89.0) RC 39 29 33 l (9.8) (7.3) (83.0) IF 39 58 302 (9.8) (14.5) (75.7) Note: Numbers in brackets are percentages. FA isFeidIurbia albida, AC is alleyclopping, RC is relay cropping andIFis mm flITuli)l‘ewpercentages of respondents exposed to the MAFE project who perceived the practices of selected agroforestry technologies designed for soil improvement as having helped to increase fuelwood availability were significantly higher than those of respondents who were not exposed to the MAFE project (p=<.05) see Table 4.36. 95 Table 4.36: A Comparison of Respondent Perceptions between Respondents who were and those who were not Exposed to the Malawi Agroforestry Extension Project, Regarding the Efl‘ect of Selected Agroforestry Technologies Designed for Soil Improvement on Fuelwood Availability in Njolomole EPA of Ntcheu RDP. Agroforestry Exposed (n=64) Not exposed (n=335) technology- §°fl more fuel- less fuel- NR more fuel- less NR unprovement wood wood wood tirel- ‘ wood FA“ 18 14 32 1 5 34 286 (28.1) (21.9) (4.5) (16.1) AC” 13 4 47 4 23 308 (20.3) (6.3) (1.2) (6.9) RC" 20 1 43 19 28 288 (31.3) (1.6) (5.7) (8.4) IF“I l7 1 46 22 57 256 (26.6) (1.6) (6.6) (17.0) df-l for all technologies. “ significant at p-.01, ‘ significant at p-.05. Note: Numbers in brackets are percentages. NR is non respondents, FA isFar'dherbr‘a albida, AC isalley cropping,RC is relaycroppingandIFisimproved fallowa. The percentages of female respondents who perceived the practices of selected agroforestry technologies designed for soil improvement as not having increased fuelwood were significantly higher than those of male respondents (p=<.05) Table 4.37. 96 Table 4.37: A Comparison of Male and Female Respondent Perceptions Regarding the Efi‘ect Selected Agroforestry Technologies Designed for Soil Improvement on Availability of Fuelwood in Njolomole EPA of Ntcheu RDP. Agroforestry Male (n=203) Female (n=196) technology- :70“ more fuel- less NR more filel- less filel- NR improvement wood fuel- wood wood wood FA“ 18 (8.9) 13 172 15 35 146 (6.4) (7.7) (17.9) AC" 14 7 182 3 20 173 Q9) (3.4) (1.5) (10.2) RC" 3 l 5 167 8 24 164 (15.3) (2.5) £11) (12.2) IF " 32 20 155 7 38 155 (15.8) (9.9) (3.6) (19.4) df-l for all technologies. ” significant at p-.01, ‘ significant at p-.05. Note: Numbers in brackets are percentages. NR is non respondents, PA is Far‘dherbr’a albida, AC is alley cropping, RC is relaycroppingandIFisimproved fallowa. b) Distribution of respondent perceptions regarding the effect of contour hedgerows on fuelwood availability. The percentage of respondents who perceived the practice of contour hedgerows as having increased fuelwood was not different fi'om that of respondents who perceived the practice as not having increased fuelwood (Table 4.38). 97 Table 4.38: A Distribution of Respondent Perceptions Regarding the Effect of Contour Hedgerows on Fuelwood Availability in Njolomole EPA of Ntcheu RDP. Agroforestry Increased Did not increase Not sure/no technology- soil fuelwood fuelwood response erosion control CH 43 g 43 3 13 (10.8) (10.3) (78.4) Note: Numbers in brackets are percentages. CH is contour hedgerows. The percentage of respondents exposed to the MAFE project who perceived the practice of contour hedgerows as having increased filelwood was significantly higher than that of respondents not exposed to the MAFE project (p=<.01) see Table 4.39. Table 4.39: A Comparison of Respondent Perceptions between Respondents who were and those not Exposed to the Malawi Agroforestry Extension Project, Regarding the Efl‘ect of Contour Hedgerows on the Availability of Fuelwood in Njolomole EPA of Ntcheu RDP. ' Agroforestry Exposed (n=64) Not exposed (n=335) technology-soil erosion control more fuel- less NR more fuel- less firel- NR wood filel- wood wood wood . CH" 28 10 26 15 57 287 (43.8) (15.6) (4.5) (17.0) df-l, ” significant at p-.01, ' significant at p-.05. Note: Numbers ib brackets are. percentages. NR is non respondents and CH is contour hedgerows. As Table 4.40 shows, the percentage of male respondents who perceived the practice of contour hedgerows as having increased filelwood was higher than that of female respondents (p=<.01). 98 Table 4.40: A Comparison of Male and Female respondent Perceptions Regarding the Efi‘ect of Contour Hedgerows on the Availability of F uelwood in Njolomole EPA of Ntcheu RDP. Agroforestry Male (n=203) Female (n=196) technology-soil ‘ erosion more fuel- less fuel- NR more fuel- less NR wood wood wood fuel- wood CH" 33 17 153 10 26 160 (16.3) (8.4) (5.1) (13.3) df-l, ” significant at p-.01, " significant at p-.05. Note: Numbers in brackets are percentages. NR is non respondents and CH is contour hedgerows. c) Distribution of respondent perceptions regarding the effect of selected agroforestry technologies designed for the production of fruit, fuelwood and other products on fuelwood availability. The percentages of respondents who perceived the practices of woodlots, planting and caring for trees in garden boundaries, planting and caring for trees in croplands and planting and caring for trees at homesteads as having increased fuelwood were not different from those of respondents who perceived these technologies as not having increased filelwood (Table 4.41). Table 4.41: A Distribution of Respondent Perceptions Regarding the Efl'ect of Selected Agroforestry Technologies Designed for the Production of Fruit, F uelwood and Other 99 Products on Fuelwood Availability in Njolomole EPA of Ntcheu RDP. Agroforestry Increased Did not increase Not sure/no technology- fruit, fuelwood fuelwood response filelwood & other products WL 91 73 235 (22.8) (18.3) (58.9) TGB 50 64 285 (12.5) (16.0) (71.4) TC 80 76 243 (20.1) ( 19.0) (60.9) TH 98 102 199 (24.6) (25.6) (49.9) NotezNumbersinbracketsarepercentages. WLiswodlotsJGBisplanfingandcaringforheesingudenbmmdariesflC isplantingandcaringfortreesincroplandsandmisplantingandcaringfortreesathomeswads. The percentages of respondents exposed to the MAFE project who perceived the practice of selected agroforestry technologies designed for the production of fi'uit, fuelwood and other products as having increased fuelwood were significantly higher than those of respondents not exposed to the project (p=<.01) see Table 4.42. 100 Table 4.42: A Comparison of Respondent Perceptions between Respondents who were and those who were not Exposed to the Malawi Agroforestry Extension Project, Regarding the Efl‘ect of Selected Agroforestry Technologies Designed for the Production of Fruit, Fuelwood and other Products on Fuelwood Availability in Njolomole EPA of Ntcheu RDP. Agroforestry Exposed (n=64) Not exposed (n=335) technology- fi'uit, filelwood & other more fuel- less fircl- NR more fuel- less fuel- NR products “’00“ W004 wood wood WL" 32 1 31 59 72 204 (50.0) (1.6) (17.6) (21.5) TGB" 22 2 40 28 62 245 (34.4) (3. 1) (8.4) (18.5) TC" 38 2 24 42 74 219 (59.4) (3. 1) (12.5) (22.1) TH" 35 1 28 63 101 177 (54-7) (1.6) (18.8) (30.1) df '1 forall technologies. “ significantat p-.01, ‘ significantatp-.05. Note:Numbusmbrsckenanpercenhga.NRisnmrespmdenn,WLiswoodlots,TGBistraesingardenboundaries, TCistreesincloplandsand'l'Histrwsathomesteads. The percentages of male respondents who perceived the practices of selected agroforestry technologies designed for the production of mm, filelwood and other products as having increased fuelwood were significantly higher than those of female respondents (p=<.01) see Table 4.43. 101 Table 4.43: A Comparison of Male and Female Respondent Perceptions Regarding the Efi‘ect of Selected Agroforestry Technologies Designed for the Production of Fruit, Fuelwood and Other Products on the Availability of Fuelwood in Njolomole EPA of Ntcheu RDP. Agroforestry Male (n=203) Female (n=196) technology-fiuit, filelwood & more fuel- less filel- NR more filel- less fuel- NR other products wood wood wood wood WL" 73 35 95 18 38 140 (36.0) (17.2) (9.2) (19.4) TGB" 36 21 146 14 43 139 (17.7) (10.3) (7.1) (21.9) TC" 64 47 92 16 29 151 (31.5) (23.2) (8.2) (14.8) TH" 75 53 75 23 49 124 (36.9) (26.1) (11.7) (25.0) df-l foralltechnologies. “ signifrcantatp=.01, ‘ significantatp-.05. NotezNumbersinbracketsarepercentages.NRisnonreapondents,WLiswoodlots,TGBistrcesingardenboundaries,TC istreesincroplandssnd'l'Histrwssthomemads. (1) Distribution of respondent perceptions regarding the effect of selected agroforestry technologies designed for the production of livestock feed and protection on fuelwood availability. The was a smaller percentage of respondents (3 .8%) who perceived the practice of fodder banks as having increased fuelwood as compared to the percentage of those who perceived the practice as not having increased fuelwood (7.3%). On the other hand the percentage of respondents who perceived the practice of living fences as having increased filelwood availability was higher (14.5%) than that of respondents (8.5%) who perceived it as not having increased fuelwood see Table 4.44. 102 Table 4.44: A Distribution of Respondent Perceptions Regarding the Effect of Selected Agroforestry Technologies Designed for the Production of Livestock Feed and Protection on Fuelwood Availability in Njolomole EPA of Ntcheu RDP. Agroforestry Increased Did not increase Not sure/no technology- livestock fuelwood fuelwood response feed and protection FB 15 29 355 (3.8) (7.3) (89.0) LP 58 34 307 (14.5) (8.5) (76.9) Note:Numbersinbracketsarepercentages.FBisfodderbankssndLFislivingfcnces. The percentages of respondents who were exposed to the MAFE project who perceived the practices of fodder banks and living fences as having increased filelwood availability were significantly higher than those of respondents not exposed to the project (p=<.01) see Table 4.45. Table 4.45: A Comparison of Respondent Perceptions between Respondents who were and those who were not Exposed to the Malawi Agroforestry Extension Project Regarding the Effect of Selected Agroforestry Technologies Designed for the Production of Livestock Feed and Protection in Njolomole EPA of Ntcheu RDP. Agroforestry Exposed (n=64) Not Exposed (n=335) technology- “VCStOCk feed more fuel- less fuel- NR more fuel- less fuel- NR & protectlon wood wood wood wood FB“ 15 8 41 0 21 314 (23.4) (12.5) (6.3) LP" 29 l 34 29 33 273 (45.3) (1.6) (8.7) (9.9) (If- 1 forall technologies. ” significant at p-.01, ‘ significantat p-.05. Note: Numbers in brackets are percentages. NR is non respondents, F8 is fodder banks and LF is living fences. 1 O 3 The percentage of male respondents who perceived the practices of fodder banks and living fences as having increased fuelwood were higher than the percentages of female respondents (p=<.01) see Table 4.46. Table 4.46: A Comparison of Male and Female Respondent Perceptions Regarding the Effect of Selected Agroforestry Technologies Designed for the Production of Livestock Feed and Protection on the Availability of F uelwood in Njolomole EPA of Ntcheu RDP. Agroforestry Male (n=203) Female (n=196) technology- “VWOCk feed more filel- less fuel- NR more fuel- less filel- NR & protection wood wood wood wood FB" 13 8 182 2 21 173 (6.4) (3.9) (1.0) (10.7) LF" 50 12 141 8 22 166 (24.6) (5.9) (4. 1) (11.2) df-l forall technologies. ” significant at p-.01, ‘ significantat p-.05. Note:Numbersinbrackenanpemhga.NRismnreepondenu,FBhfoddabanbuldLFisfivingfaw. RESPONDENT PERCEPTIONS REGARDING THE EFFECT OF SELECTED AGROFORESTRY TECHNOLOGIES ON PROFITABILITY. This section discusses respondent perceptions regarding the profitability of selected agroforestry technologies in Njolomole EPA of Ntcheu RDP. Distributions of respondent who perceived the selected agroforestry technologies as more profitable than their previous practices are compared to those of respondents who perceived that the agroforestry technologies were less profitable than their previous practices. Comparisons are made between perceptions of respondents who were and those who were not exposed to the MAFE project. Similar comparisons are made between male and female respondents. 104 A large proportion of the respondents were either not sure or did not respond to the items on profitability. The reasons for these were either the respondents did not practice the technologies or did not practice them long enough to be able to make informed judgements on their profitability. The discussions are divided into four sub- sections according to the purposes for which the technologies were designed. a) Distribution of respondent perceptions regarding the effect of selected agroforestry technologies designed for soil improvement on profitability. The percentage of respondents who perceived the practice of improved fallows as more profitable than their previous practices was higher (16.5%) than the percentage of those who perceived it as less profitable than their previous practices (1.8%) see Table 4.47. Table 4.47: A Distribution of Respondent Perceptions Regarding the Effect of Selected Agroforestry Technologies Designed for Soil Improvement on Profitability in Njolomole EPA of Ntcheu RDP. Agroforestry More profits Less profits Not sure/no technology- response mirovement number % number % number % FA 32 8.0 23 5.8 344 86.2 AC 20 5.0 3 0.8 375 94.2 RC 38 9.5 13 3.3 348 87.2 IF 66 16.5 7 1.8 326 82.2 Note: Numbers in brackets are percentages. FA isFar'dherbr’a albida, AC is alley cropping, RC is relay cropping and IF is improved fallows. l O 5 The percentages of respondents exposed to the MAFE project who perceived the practices of systematic interplanting with E albida, alley cropping, and relay cropping as more profitable than their previous practices were significantly higher than those of respondents who were not exposed to the MAFE project (p=<.01) see Table 4.48. Table 4.48: A Comparison of the Perceptions of Respondents Exposed and those not Exposed to the Malawi Agroforestry Extension Project, Regarding the Effect of Selected Agroforestry Technologies Designed for Soil Improvement on Profitability in Njolomole EPA of Ntcheu RDP. Agroforestry Exposed (n=64) Not exposed (n=335) technology- §°fl more less NR more less NR Improvement profits profits profits profits FA" 24 4 36 8 19 308 (37.5) (6.3) (2.3) (5.7) AC" 15 0 49 5 3 327 (23.4) (1.5) (0.9) RC“ 19 1 44 19 12 304 (29.7) (1.5) (5.7) (3.6) IF l4 1 49 52 6 277 (21.9) (1.6) (15.5) (1.8) df '1 for all technologies. ” significant at p-.01, ‘ significant at p-.05 Note: Numbers in brackets are percentages. NR is non respondents. FA is Fairflrerbr'a albida, AC is alley cropping, RC is relay cropping and IF is improved fallows. The percentages of male respondents who perceived the practice of systematic interplanting with R albida as more profitable than their previous practices was significantly higher than that of female respondents (p=.01) see Table 4.49. Table 4.49: A Comparison of Male and Female Respondent Perceptions Regarding the 106 Efi'ect of Selected Agroforestry Technologies Designed for Soil Improvement on Profitability in Njolomole EPA of Ntcheu RDP. Agroforestry Male (n=203) Female (n=196) technology-Soil improvement more less NR more less NR profit profit profit profit FA" 22 3 178 10 20 166 (10.8) (1.5) (5.1) (10.2) AC 16 3 184 4 0 192 (7.9) (1.5) (2% RC 28 8 167 10 5 181 (13.8) (3.9) (5.1) (26) IF 48 5 150 18 2 176 (23.6) (2.5) (9.2) (1.0) df-l for all technologies. ” significant at p-.01, ‘ significant at p-.05 Note: Numbers in brackets are percentages. NR is non respondents. PA is Farwrsrbr'a albida, AC is alley cropping, RC is relay cropping and IF is improved fellows. b) Distribution of respondent perceptions regarding the effect of contour hedgerows on profitability. The percentages of respondents who perceived the practice of contour hedgerows (27.1%) as more profitable than their previous practices was higher than that of respondents who perceived the practice as less profitable than their previous practices (2.5%) see Table 4.50. 107 Table 4.50: A Distribution of Respondent Perceptions Regarding the Effect of Contour Hedgerows on Profitability in Njolomole EPA of Ntcheu RDP. Agroforestry More Less profitable Not sure/no technology-soil profitable response ° tr 1 “08m“ con 0 number % number % number % CH 108 27.1 10 2.5 281 70.4 Note: Numbers in brackets are percentages. CH is contour hedgerows. The percentage of respondents exposed to the MAFE project who perceived the practice of contour hedgerows as more profitable than their previous practices was not significantly different from that of respondents who were not exposed to the MAFE project (Table 4.51). Table 4.51: A Comparison of Respondent Perceptions between Respondent who were and those who were not Exposed to the Malawi Agroforestry Extension Project, Regarding the Efi‘ect of Contour Hedgerows on Profitability in Njolomole EPA of Ntcheu RDP. Agroforestry Exposed (n=64) Not exposed(n=335) technology-soil erosion control more less NR more less NR profits profits profits profits CH 33 1 30 75 9 25 1 (51.6) (1.6) (22.4) (2.7) df-l, ” significant at p-.01, ’ significant at p-.05. Note: Numbers in brackets are percentages. NR is non respondents, CH is contour hedgerows. Similarly there were no significant difi‘erences between the percentages of male and female respondents who perceived the practice of contour hedgerows as more profitable than their previous practices (Table 4.52). Table 4.52: A Comparison of Male and Female Respondent Perceptions Regarding the 108 Efi‘ect of Contour Hedgerows on Profitability in Njolomole EPA of Ntcheu RDP. Agroforestry Male (n=203) Female (n=196) technology-soil erosion control More less NR more less NR profit profit profit profit CH 76 8 1 19 32 2 162 (37.4) (3.9) (16.3) (1.0) ” significant at p-.01, ‘ significant at p-.05. Note:Numbersinbracketsarepercentages.NRisnonrespondentsandCHiscontourhedgerows. c) Distribution of respondent perceptions regarding the effect of selected agroforestry technologies designed for the production of fruit, fuelwood and other products on profitability. The percentages of respondents who perceived the practice of selected agroforestry technologies designed for the production of mu, filelwood and other products as more profitable than their previous practices were higher than the percentages of respondents who perceived these practices as less profitable than their previous practices: woodlots 27.1% as compared to 6.8%; trees in garden boundaries 17.0% as compared to 4.3%; trees in croplands 24.8% as compared to 10.3%; and, trees at 42.1% as compared to 6.5%, (Table 4.53). 109 Table 4.53: A Distribution of Respondent Perceptions Regarding the Profitability of Selected Agroforestry Technologies Designed for the Production of Fnrit, Fuelwood and Other Products in Njolomole EPA of Ntcheu RDP. Agroforestry More profitable Less profitable Not sure/no response technology- fi'uit, filelwood 8‘ number % number % number % other products WL 108 27.1 27 6.8 264 66.2 TGB 68 17.0 17 4.3 314 78.7 TC 99 24.8 41 10.3 259 64.9 TH 168 42.1 26 6.5 205 51.4 NotetNumbersin bracketsarepercentages. WLiswoodlots,TGBisplantingsndcaringfortreesingardenboundaries,TC isplantingandcaringfortrwsincroplandsandfliisphntingandcsringforueesathorneswads. The percentages of respondents exposed to the MAFE project who perceived the practices of woodlots and planting and caring for trees in croplands as more profitable than their previous practices were significantly higher than those of respondents who were not exposed to the MAFE project (p=<.05) see Table 4.54. There were however no significant differences in the percentages of respondents between those who were and those who were not exposed to the MAFE project who perceived the practices of planting and caring for trees in garden boundaries and planting and caring for trees at homesteads as more profitable than their previous practices. 110 Table 4.54: A Comparison of Respondent Perceptions between Respondents who were and those who were not Exposed to the Malawi Agroforestry Extension Project, Regarding the Effect of Selected Agroforestry Technologies Designed for Production of Fnrit, Fuelwood and other Products on Profitability in Njolomole EPA of Ntcheu RDP. Agroforestry Exposed (n=64) Not exposed (n=335) technology- fruit. fuelwood & more less NR more less NR other products profit profit profit profit WL" 27 1 36 81 26 228 42.2) (1.6) (24.2) (7.8) TGB 19 1 44 49 16 270 (29.7) (1 .6) (14.6) (4.8) TC“ 33 2 29 66 39 230 (51.6) (33 (19.7) (11.6) TH 32 1 3 l 136 25 174 (50.0) (1.6) (40.6) (7.5) df-l for all technologies. ” significant at p-.01, ‘ significant at p-.05. NotezNumbersinbracketsarepercentages.NRisnonrespondents,WLiswoodlots,TOBistreesingardenboundaries,TC istreesincroplandsand'l'I-Iistreesathornesteads. There were also no significant differences between male and female respondents who perceived the practices of woodlots, planting and caring for trees in garden boundaries, planting and caring for trees in croplands and planting and caring for trees at homesteads as more profitable than their previous practices (Table 4.55). Table 4.55: A Comparison of Male and Female Respondent Perceptions Regarding the Effect of Selected Agroforestry Technologies Designed for the Production of Fruit, 111 Fuelwood and other Products on Profitability in Njolomole EPA of Ntcheu RDP. Agroforestry Male (n=203) Female (n=196) technology-hit & other more less NR more less NR products profit profit profit profit WL 81 21 101 27 6 163 (39.9) (10.3) (13.8) 9.9 TGB 41 11 151 27 6 163 (20.2) (5.4) (13.9 (3.1) TC 80 33 90 19 8 169 Q94) (16.3) (9.7) (4. 1) TH 1 19 18 66 49 8 139 (58.6) (8.9) (25.0) (4. 1) df=1 for all technologies. “ significant at p-.01, ’ significant at p-.05. Note:Numbersinbracketsarepercentages.NRisnonrespondents,WLiswoodlots,TGBishwsingudenboundaries,TC istreesincroplandsandmistrwsinathomesteads. (1) Distribution of respondent perceptions regarding the effect of selected agroforestry technologies designed for the production of livestock feed and protection on profitability The percentages of respondents who perceived the practices of fodder banks and living fences as more profitable than their previous practices were higher than those of respondents who perceived these practices as less profitable than their previous practices: fodder banks 4.8% as compared to 0.3%; and, living fences 15.8% as compared to 2.0%, (Table 4.56). 112 Table 4.56: A Distribution of Respondent Perceptions Regarding the Efi‘ect of Selected Agroforestry Technologies Designed for the Production of Livestock Feed and Protection on Profitability in Njolomole EPA of Ntcheu RDP. Agroforestry More profitable Less profitable Not sure/no technology- response 1i st k feed d pig 0:: on an number % number % number % FB 19 4.8 1 0.3 379 95.0 LP 63 15.8 8 2.0 328 82.2 Note:Numbeninbracketssnpemennges.FBisfoddabanbmdLFisfivingfalces. The percentage of respondents exposed to the MAFE project who perceived the practice of fodder banks as more profitable than their previous practices (28.1%) was significantly higher than that of respondents who were not exposed to the MAFE project (Table 4.57). Table 4.57: A Comparison of Respondent Perceptions between Respondents who were and those who were not Exposed to the Malawi Agroforestry Extension Project, Regarding the Effect of Selected Agroforestry Technologies Designed for the Production of Livestock Feed and Protection on Profitability in Njolomole EPA of Ntcheu RDP. Agroforestry Exposed (n=64) Not exposed (n=335) technology- “VCStOCk feed ‘ more less NR more less NR & protection profit profit profit profit FB 18 l 45 l 0 334 (28.1) (1.6) LP 14 l 49 52 6 277 (21.9) (1.6) (15.5) (1.8) df'l for all technologies. " significant at p-.01, ’ significant at p-.05. NotezNumbersinbracketsarepercentages.NRisnonrespondents,FBisfodderbanksandLFislivingfences. 113 The percentages of male and female respondents who perceived the practices of fodder banks and living fences as more profitable than their previous practices were not significantly different (Table 4.58). Table 4.58: A Comparison of Male and Female Respondent Perceptions regarding the Efi‘ect of Selected Agroforestry Technologies Designed for the Production of Livestock Feed and Protection on Profitability in Njolomole EPA of Ntcheu RDP. Agroforestry Male (n=203) Female (196) technology- “VCStOCk feed more less NR more less NR & protection profit profit profit profit FB 16 l 186 3 0 193 (7.9) Q5) (15.8) LP 54 7 142 9 l 186 (26.6) (3.4) (4.6) (0.5) df-l forall technologies. ” significant atp-.01, ‘ significant at p-.05. Note:Numbersinbracketsarepercentages.NRisnonrespondents,FBisfodderbsnksandLFislivingfences. HYPOTHESES TESTING USING LOGISTIC REGRESSION MODELS This section of the chapter discusses the results of logistic regression models. The main objective of these analyses was to test the following sets of null hypotheses: 1. Respondents’ exposure to the MAFE project did not influence their decisions to practice selected agroforestry technologies. 2. Respondents’ perceptions regarding the efl‘ect of selected agroforestry technologies on food crop yields did not influence their decisions to practice those agroforestry technologies. 3. Respondents’ perceptions regarding the effect of selected agroforestry technologies on 114 the control of soil erosion did not influence their decisions to practice those agroforestry technologies. 4. Respondents’ perceptions regarding the efi‘ect of selected agroforestry technologies on fuelwood availability did not influence their decisions to practice those agroforestry technologies. 5. Respondent perceptions regarding the effect of the selected agroforestry technologies on profitability did not influence their decisions to practice those agroforestry technologies. 6. The following demographic factors: gender, respondent’s age; respondent’s highest level of formal schooling attained; the amount of income the respondent received in 1996; landholding size; land tenure system and tree tenure system did not influence their decisions to practice selected agroforestry technologies. Since there were eleven agroforestry technologies selected for the study, there should have been eleven logistic regression analyses conducted, one for the practice of each agroforestry technology. However, due to missing data only five logistic regression analyses were included in this chapter. These models are for the practices of: contour hedgerows; woodlots; planting and caring for trees in garden boundaries; planting and caring for trees in croplands; and, planting and caring for trees at homesteads. 115 Logistic regression model for the practice of contour hedgerows. The variables initially entered in the model were: X,= exposure to the MAFE project, X2= respondent perception of the effect of the practice of contour hedgerows on food crop yields, X,= respondent perceptions as to whether they experienced soil erosion problems on their farm X4= whether respondents practiced contour hedgerows in order to control soil erosion, X,= respondent perceptions as to whether the practice of contour hedgerows reduced soil erosion, X6= respondent perceptions as to whether they experienced problems of shortage of fuelwood, X,= whether respondents practiced contour hedgerows in order to increase fuelwood, X.= respondent perceptions as to whether the practice of contour hedgerows increased fuelwood, X,= respondent perceptions as to whether the practice of contour hedgerows was more or less profitable than their previous practices, Xm= gender Xu= respondent’s highest level of formal education attained, Xn= respondent’s age, X13= total income received in 1996 Xu= landholding size, 1 1 6 X,,= land tenure system and X16= tree tenure system. X3, X4 and X, were all measures of respondent perceptions regarding the effect of the practice of contour hedgerows on soil erosion while X,, X, and X. were all measures of respondent perceptions regarding the effect of the practice of contour hedgerows on firelwood availability. The variable Xl3 had seven categories. It was therefore transformed into six dummy variables before entering it into the model. The variables which remained after the variable selection process were Xl= exposure to the MAFE project, X3= respondent perceptions of soil erosion problems on their farm, X4= whether respondents practiced contour hedgerows in order to control soil erosion, X13= land tenure system (table 4.59). The variable X3= respondent perceptions of soil erosion problems on their farm was not removed during the variable selection process but also did not come up as a significant predictor in the final model. There were 186 cases included and the model correctly classified 79.57% of the cases. The -2 log likelihood improved from 220.42 (model with constant alone) to 176.45 with the four variables (Appendix B). Table 4.59 shows that the respondents who were exposed to the MAFE project, who also felt that they had secure rights of land tenure and perceived that the practice of contour hedgerows controlled soil erosion were more likely to practice contour hedgerows. From Table 4.59, the probability of respondent decision to practice contour hedgerows can be estimated as follows e’/1+e' where z=.4928 + .1317x3 -.7525x, -l.4359x, -.9790x,, 117 Table 4.59: Maximum Likelihood Estimates for the Practice of Contour Hedgerows in Njolomole EPA of Ntcheu RDP in Malawi. Variable Estimate SE. Wald df p level X, -.7525 .3039 6.1302 1 .0133“ X3 .1317 .2321 .3218 1 .5705 X, -1.4359 .6412 5.0146 1 .0251“ X” -.9790 .2326 17.7124 1 .0000" constant .4928 .6761 .5313 1 .4661 ” significant at p-.01, ' significant at p-.05. Logistic regression model for the practice of woodlots. The variables entered for the practice of woodlots were the same as those entered for the practice of contour hedgerows but using data for woodlots. The variables which remained after the variable selection process were X,= respondent’s exposure to MAFE project, X.= respondent’s perception as to whether the practice of woodlots increased filelwood, Xw= gender and X,,= total income received in 1996 (table 4.60). There were 164 cases entered and the model correctly classified 84.76% of the cases (see Appendix B ). The -2 log likelihood improved from 172.62 (model with constant alone) to 111.88 in the final model. As Table 4.60 shows, Xlo= gender and Xu= total income received in 1996 came out as significant predictors of respondent’s decision to practice woodlots. Respondent perceptions as to whether the practice of woodlots increased filelwood and respondent’s exposure to the MAFE project did not meet the removal criteria but were also not significant. Table 4.60 shows that 1 1 8 male respondents who received income which was more than K499.00 in 1996 were more likely to practice woodlots. Table 4.60: Maximum Likelihood Estimates for the Practice of Woodlots in Njolomole EPA of Ntcheu RDP in Malawi. Variable Estimate S.E. Wald df p level X, -.8674 .4816 3.2440 1 .0717 X. .3301 .2652 1.5489 1 .2133 X", -1.2438 .2603 22.8403 1 .0000" X1, .8718 .2714 10.3162 1 .0013“ constant -l.7404 4449 15.3004 1 .0001" ” significant at p-.01, ’ significant at p-.05. From Table 4.60 above, the probability of respondent decision to practice woodlots can be estimated by e‘/1+ez where z=-1.7404 +.3301x, + .8718x,, -.8674x, -1.2438x,, Logistic regression model for the practice of planting and caring for trees in garden boundaries. The variables entered for the practice of planting and caring for trees in garden boundaries were the same as those entered for the practice of contour hedgerows but using data for planting and caring for trees in garden boundaries. The variables which remained after the variable selection process were X,=respondent’s exposure to the MAFE project, X2= respondent perception regarding the efl'ect of the practice of planting and caring for trees in garden boundaries on food crop yields, X3= whether respondent perceived that they experienced soil erosion problems on 1 l 9 their farm, Xm= gender, Xn= respondent’s age, X13= total income received in 1996 and Xu= landholding size. There were 348 cases entered and the model correctly classified 86.21% of the respondents. The -2 log likelihood improved from 409.95 (fi'om model with constant only) to 263.46 in the final model (see Appendix B). Respondent perceptions regarding the efi‘ect of the practice of planting and caring for trees in garden boundaries on food crop yields, total income received in 1996, and gender were the variables which came out as significant predictors of the decision to practice planting and caring for trees in garden boundaries (Table 4.61). Respondent perceptions of soil erosion problems on their farms, respondent exposure to the MAFE project, age, and landholding size did not come up as significant predictors but they remained in the model during the variable selection process. Male respondents who received income in excess of K499.00 in 1996 and perceived the practice of planting and caring for trees in garden boundaries as increasing food crop yields were more likely to practice planting and caring for trees in garden boundaries. From Table 4.61, the probability of farmer decision to practice planting and caring for trees in garden boundaries can be estimated by e'/ 1+e’ where z= .1599 + .0998X3 + .0120X12 + .4163X13‘ + .0765Xl4 - .3008Xl -1.9592X2 - .3382x,,. -.9346X,,* Table 4.61: Maximum Likelihood Estimates for the Practice of Planting and Caring for 120 Trees in Garden Boundaries in Njolomole EPA of Ntcheu RDP in Malawi. Variable Estimate S.E. Wald df p level X1 -.3008 .2218 1.8393 1 .1750 X2 -1.9592 .2247 76.0394 1 .0000" X3 .0998 .1780 .3144 1 .5750 X“, -.3382 .1683 4.0409 1 .0444‘ X12 .0120 .0117 1.0481 1 .3060 X1; .4163 .1910 4.7487 1 .0293" XL." -.9346 .4907 3.6281 1 .0568 X“ .0765 . 1567 .23 82 1 .6255 constant . 1599 .7700 .0431 1 .8355 “ significant at p-.01, ‘ significant at p-.05. 'Respondent received income ofK499.00 orother in 1996. 'Respondent received income ofK1000.00-1,999.00 or other in 1996. Logistic regression model for the practice of planting and caring for trees in croplands. The variables which were entered for the practice of planting and caring for trees in croplands were the same as those entered for the practice of contour hedgerows but using data on planting and caring for trees in croplands. The variables which remained after the variable selection process were X,=respondent’s exposure to MAFE project, X,=respondent’s perception of the influence of the practice of planting and caring for trees in craplands on food crop yields, X,=whether respondent perceived that he/she experienced soil erosion problems on his/her fann, X.= whether respondent perceived the practice of planting and caring for trees in 121 croplands as having helped to increase fuelwood and X,o=gender. The model correctly classified 91.03% of the respondents and the -2 log likelihood improved fiom 140.58 (from the model with constant only) to 94.47 in the final model. The number of cases included was 156 (see Appendix B). As Table 4.62 shows, respondent’s perception regarding the efi‘ect of the practice of planting and caring for trees in croplands on food crop yields and gender were significant. Male respondents who perceived the practice of planting and caring for trees in croplands as increasing food crop yields were more likely to practice planting and caring for trees in croplands. Table 4.62: Maximum Likelihood Estimates for the Practice of Planting and Caring for Trees in croplands. Variable Estimate S.E. Wald df p level X, -. 1805 .4913 .1349 1 .7134 X2 -1.1464 .3829 8.9637 1 .0028" X, .7093 .5798 1.4965 1 .2212 X, .0175 .3464 .0026 1 .9596 X1,) 2.1778 .5430 16.0873 1 .0001" constant -5.8199 1.1246 26.7801 1 0000" “ Ignificant at p-.01, ‘ significant at p-.05 From the Table 4.62 above, respondent’s decision to practice planting and caring for trees in garden boundaries can be estimated by e‘/1+ez where z= -5.8199 + .7093x3 + .0175x, + 2.1778xlo -.1805X, -1.1464x, 1 2 2 Logistic regression for the practice of planting and caring for trees at homesteads. The variables entered for the practice of planting and caring for trees at homesteads were the same as those entered for the practice of contour hedgerows but using data for planting and caring for trees at homesteads. ' There were 348 cases entered in the model and it correctly classified 83.33% of the respondents. The -2 log likelihood improved fi'om 481.28 (fi'om model with constant only) to 301.33 in the final model (see Appendix B). The variables which remained afier the variable selection process were X,= whether respondent perceived the practice of planting and caring for trees at homesteads as more or less profitable than his/her previous practices and X,,= age (Table 4.63). Respondent perception as to whether the practice of planting and caring for trees at homesteads was more or less profitable than previous practices was significant at the .01 level. Respondents who perceived the practice of planting and caring for trees at homesteads as more profitable than their previous practices were more likely to practice the technology. Fruits were the main product which were produced and sold from this technology. Table 4.63: Maximum Likelihood Estimates for the Practice of Planting and Caring for Trees at Homesteads in Njolomole EPA of Ntcheu RDP in Malawi. Variable Estimate S.E. Wald df p level X, 1.9363 .1933 100.3037 1 0000" x,2 -.0134 .0097 1.8987 1 .1682 constant -4.0559 .6543 38.4247 1 0000" ” significant at p-.01, ‘ significant at p-.05. 123 From Table 4.63 above, the probability of respondent decision to practice planting and caring for trees at homesteads can be estimated by e‘/1+ez where z= -4.0559 + 1.9363x, - .0134x,2 Chapter V SUMMARY, CONCLUSIONS AND RECOMMENDATIONS Introduction Existing high population density, along with increasing population growth is putting extreme pressure on the limited land for cultivation in Malawi. Consequences of this population grth are the high demands for fuelwood as a major source of energy thus causing serious declines in soil fertility, soil erosion and reduction of natural forest resources. Research, initiated to address these issues, has generated specific agroforestry technologies which are being recommended to farmers. Efforts to improve land resources continue to be hindered by failure of farmers to adopt the proposed technologies. Thus, a situation has been created, generating a need for further investigation of the factors influencing farmer adoption of the technologies. Review of the problem and purpose of the study The main problem of the study was to determine the factors that have influenced farmer attitudes towards adoption of selected agroforestry technologies. The purpose of the study was to generate information to be used for redesigning extension efl‘orts to increase farmer practice of the selected agroforestry technologies. The specific research hypotheses were: 1. Farmer exposure to the MAFE project influenced their decisions to practice selected agroforestry technologies. 2. Farmer perceptions regarding the effect of selected agroforestry technologies on food crop yields influenced their decisions to practice those agroforestry technologies. _ 124 125 3. Farmer perceptions regarding the effect of selected agroforestry technologies on soil erosion control influenced their decisions to practice those agroforestry technologies. 4. Farmer perceptions regarding the effect of selected agroforestry technologies on filelwood availability influenced their decisions to practice those agroforestry technologies. 5. Farmer perceptions regarding the effect of selected agroforestry technologies on profitability influenced their decisions to practice those technologies. 6. The following demographic factors; gender, farmer’s age, farmer’s highest level of formal schooling attained, the amount of income the farmer received in 1996, farmer’s landholding size, land tenure system and tree tenure system influenced farmer decisions to practice selected agroforestry technologies. Review of research methods Data collection was done through interviews by a team of five male and four female interviewers. The males interviewed male respondents while the females interviewed female respondents. A total of 399 out of a targeted 449 questionnaires were used for analysis. Sixty four questionnaires came from respondents who were exposed to the MAFE project and 335 from respondents who were not exposed. In terms of gender, 203 questionnaires were from male respondents while 196 questionnaires were fi'om female respondents. Logistic regression analyses were conducted for the practices of contour hedgerows, woodlots, planting and caring for trees in garden boundaries, planting and caring for trees in croplands and planting and caring for trees at homesteads. Logistic regression analyses for the practices of systematic interplanting with F. albida, alley 126 cropping, relay cropping, improved fallows, fodder banks and living fences were not included because of missing data’. Discussion of results and findings Introduction Four agroforestry technologies were more widely practiced than others: contour hedgerows; woodlots; planting and caring for trees in croplands; and, planting and caring for trees at homesteads. These were practiced by at least forty percent of the respondents. Systematic interplanting with E albida, alley cropping, relay cropping, improved fallows, planting and caring for trees in garden boundaries, fodder banks and living fences were practiced by less than forty percent of the respondents. The single most important reason expressed by farmers for not adopting the practice of selected agroforestry technologies was not having been exposed to their benefits. The following section provides a brief summary of the findings for each hypothesis. Hypothesis 1: Exposure to the MAFE project A comparison of farmers exposed to the MAFE project with farmers who were not exposed was made. Farmers exposed to the MAFE project were more likely to adopt selected agroforestry technologies presented through the project endeavors. Farmer exposure to the MAFE project significantly influenced their decisions to practice contour hedgerows (p=<.05). F arrners who were exposed to the MAFE project practiced contour , 7 A minimum of 100 cases was used as criteria for accepting the logistic regression analyses. Any model with less than 100 cases entered was rejected as having insuficient sample size. 127 hedgerows more often than those who were not exposed to the project. Exposure to other technologies influenced practice but were not statistically significant: Contour hedgerows, exposed/adopted 85.9%- not exposed/adopted 30.7%; Woodlots, exposed/adopted 60.9%- not exposed/adopted 37.9%; Planting and caring for trees in garden boundaries, exposed/adopted 39.1%- not exposed/adopted 25.7%; Planting and caring for trees in croplands, exposed/adopted 89.1% -not exposed/adopted 42.4%. Farmer exposure to the MAFE project also resulted in a positive attitude towards the selected agroforestry technologies. The farmers who were exposed to the project perceived the agroforestry technologies as increasing food crop yields, controlling soil erosion and as being profitable. It therefore seems that providing extension services, like those provided by the MAFE project, to all the farmers in Njolomole EPA would increase the number of farmers practicing the selected agroforestry technologies. Planting and caring for trees at homesteads was on the other hand a technology which enjoyed wide popularity both among farmers who were exposed to the MAFE project (64%) and those who were not exposed (56%). Hypothesis 2: Food productivity Farmers who were influenced by the impact of the selected agroforestry technologies on increased food crop productivity were more likely to practice those agroforestry technologies. Two agroforestry technologies were significant ( p=<.01); planting and caring for trees in garden boundaries and in croplands. 128 There were generally higher proportions of respondents who perceived the practices of selected agroforesz technologies as having increased their food crop productivity. Hypothesis 3: Soil erosion Farmer perception regarding the effect of contour hedgerows on soil erosion control significantly influenced their decision to practice the contour hedgerows (p=<.05). Farmers who perceived that the practiced of contour hedgerows controlled soil erosion tended to practice the technology more often. Although not statistically significant, farmers perceived the practices of woodlots, trees in garden boundaries, trees in croplands and trees at homesteads as controlling soil erosion. Hypothesis 4: Fuelwood production Farmer perceptions of the effect of selected agroforestry technologies on filelwood availability did not significantly influence farmer decisions to practice the agroforestry technologies. Hypothesis 5: Profitability Farmer perceptions regarding the efl'ect of the practice of planting and caring for trees at homesteads significantly influenced their decisions to practice planting and caring for trees at homesteads (p=<.01). Farmers who perceived the practice of planting and caring for trees at homesteads as being more profitable than their previous experiences practiced it while those who perceived it as less profitable than their previous experiences did not. The majority of the farmers perceived the practice of planting and caring for trees at homesteads as more profitable than their previous practices. Fnrits were the main 1 2 9 product which was produced and sold. Farmers make decisions to practice or not practice new technologies by calculating the income efi'ect of those new technologies (Lee, 1983). Agarwal (1983) discussing factors that influence diffusion of innovations argued that the nature of innovation characteristics defines the ease or difiiculty of diffusion. She stated that, the advantages of innovations which provide direct high financial benefits are likely to be perceived more readily than those of innovations whose benefits are indirect and non financial. Perhaps farmer perception of the financial profitability of the practice of planting and caring for trees at homesteads was an explanation for its wide popularity. Profitability was, however, defined only in terms of farmer perceptions. There is need to confirm these perceptions with actual financial data obtained under farmer practice. Profitability did not significantly influence farmer decisions to practice the other technologies. Hypothesis 6: Demographic factors 1. Gender: Gender significantly influenced the practice of planting and caring for trees in garden boundaries (p=<.05) and the practice of planting and caring for trees in croplands (p=.<01). Male farmers were more likely to plant and care for trees in garden boundaries and in croplands. The proportions of male farmers practicing the selected agroforestry technologies were higher than those of female farmers. This gender difi'erence in the practice of selected agroforestry technologies was most notable in the practices of selected agroforestry technologies designed for the production of fruit, fuelwood and other 1 3 0 products. This observation was consistent with farmer perceptions of whether the agroforestry technologies designed for the production of fruit, fuel wood and other products helped to increase filel wood availability. Male farmers perceive the practices of woodlots, planting and caring for trees in garden boundaries, planting and caring for trees in croplands and planting and caring for trees at homesteads as helping to increase fuelwood availability while female farmers did not. One plausible explanation for this difi‘erence in perceptions lies in the reasons for which male and female farmers established the agroforestry technologies. As discussed in chapter 11 studies have shown that the purposes for which women plant and care for trees are difi‘erent fi'om those of men (Chavangi et al, 1988). Women tend to be more concerned with preparing and providing food for their families (Molnar and Screiber, 1989; Chavangi et al, 1988; Agarwal, 1986). Their reason for planting and caring for trees is usually for the production of fuelwood for food preparation. Men on the other hand tend to be more concerned with income generation and construction. The products of choice for men are poles and timber. Fuelwood production is considered as a minor product by the male members of the family. This means that while the practices of woodlots, planting and caring for trees in garden boundaries, planting and caring for trees in croplands and planting and caring for trees at homesteads could have made a significant contribution to the fuelwood availability situation in the EPA, women did not have access to such fuelwood. As a result, women could not perceive such technologies as having helped increase fuelwood availability. Age: The influence of age on farmer decisions to practice the selected agroforestry technologies was not significant. 131 3. Income: There was a significant relationship between the amount of income received in 1996 and farmers’ decisions to practice woodlots (p=<.01) and planting and caring for trees in garden boundaries (p=<.05). Farmers who received more than K499.00 in 1996 were more likely to practice woodlots and trees in garden boundaries than those who received less than that amount. A possible reason for this is the capital which is normally required in order to establish and maintain these technologies. Capital is required for the establishment and maintenance of the technologies for such things as purchase of tree seedlings, fertilizers and chemicals for controlling termites. Capital may also be required to pay for hired labor which is usually needed for field preparation at establishment, weeding and protecting the trees fiom fire. It would therefore be necessary to provide credit facilities for those farmers who do not have suficient capital required for establishment and maintenance of the technologies. It should however be noted that farmer perceptions of the profitability of the practices did not come up as a significant predictor of the practices. That is not to say that the practices were not profitable but that the proportions of farmers who perceived the practices as more profitable than their previous practices were not large enough. It would therefore be prudent to conduct economic studies aimed at investigating the financial profitability of the practices. Such studies would provide useful data for designing appropriate credit schemes designed for promoting the practices. 4. Land tenure system: Land tenure system had a significant influence on farmer decisions to practice contour hedgerows (p=<.01). Farmers who perceived that they had secure rights of ownership to the land that they cultivated practiced contour hedgerows 132 more often when compared to farmers who did not have secure rights of land ownership. Security of land tenure is an important requirement for the practice of contour hedgerows. Farmers who indicated that they did not practice the contour hedgerows due to insecure rights of land tenure were predominantly male and cited the system of Chikamwini' as a reason for not being willing to invest in the agroforestry technology. The practice of contour hedgerows is a land improving technology. Its main benefit is the control of soil erosion which leads to the long term productivity of the land. Farmers who perceive the long term productivity of the land as directly beneficial to them are probably more likely to want to practice the technology. While farmers who do not perceive the long term profitability of the land as directly beneficial to them are probably not likely to want to invest in the practice of contour hedgerows. Farmers therefore, nwd to be assured of the long term ownership of their land if they are to practice planting contour hedgerows. Farmers who do not perceive that the land will belong to them for a long time are probably not likely to invest in contour hedgerows. In the case of Njolomole EPA, the major problem lies with chikamwini system. People of Njolomole EPA practice a matrilineal type of marriage system. At marriage the man is supposed to leave his home and go to settle at his wife’s home. The farm land which he works does not in principle belong to him but to his wife. In the case of divorce or death of his wife, he is supposed to go back to his original home and has no ' Chikamwini refers to the system whereby a man follows his wife at marriage. By tradition the man works on land allocated to his wife by her relations. In the case of divorce or death of the wife, the man is required to go back to his original home and therefore loses the hits of any investment made on the land. 133 firrther access to the land he was working during marriage. This creates a situation whereby the male farmers perceive themselves as not having secure rights of land tenure. They therefore, become less willing to invest in land improving technologies like the practice of contour hedgerows. It should also be noted that the benefit of long term productivity of the land was of more interest to the women than men. As already observed, women tend to be more concerned with production and provision of food for the family while men tend to be interested in generating income. Productivity of the land seems to be more of a concern of women than of men. 5. Landholding size: The influence of landholding size on farmers’ decision to practice the selected agroforestry technologies was not significant. Tree tenure system: The influence of tree tenure system on farmers’ decisions to practice the selected agroforestry technologies was not significant GENERAL RECOMMENDATIONS There is need to increase the number of farmers practicing selected agroforestry technologies in Njolomole EPA of Ntcheu RDP. Most of the farmers in the EPA are not able to produce enough food to last them from one harvest to another. Soil erosion and shortage of fuelwood are very serious problems. As the population is growing, the problems of land pressure, declining soil fertility, soil erosion and reduction of natural forests are also growing. Efforts are required address these problems. Implementation of agroforestry technologies is one way of addressing such problems. 1 3 4 Recommendation 1: It is recommended that extension programs designed to promote the practice of the selected agroforestry technologies be provided to all the farmers in Malawi. Specifically, it is recommended that the Government of Malawi, either though its Ministries of Agriculture and Livestock Development and Forestry and Natural Resources or through non governmental organizations, develop a strategic plan of action for providing information and demonstrations on recommended agroforestry technologies. Recommendation 2: It is further recommended that the Malawi extension program design demonstrations in locations where farmers can make observations and access information. The demonstrations should be designed in such a way that they emphasize the advantages of the selected agroforestry technologies. Farmers who perceive the advantages of the technologies such as increased food production, control of soil erosion and profitability will likely practice the technologies. As observed, however, each technology has its own unique advantages. Those technologies which have the benefits of increasing food crop yields are difi‘erent fiom those which have the benefits of controlling soil erosion and also difi‘erent fi'om those which have the benefits of profitability. The demonstrations must therefore be designed to emphasize specific advantages of each technology. Recommendation 3: It is recommended that a collaborative effort be made between the Ministry of Agriculture and Livestock Development and Forestry and Natural Resources for an extensive establishment of agroforestry tree nurseries. These nurseries should be made in places close to the farmers so that tree seedlings such as 135 those of E albida can be made readily available to the farmers at a reasonable cost. This will help to address the problem of shortage of seedlings. Another approach would be to train farmers in nursery management and then provide them with tree seed. This would help to reduce cost of seedlings. Recommendation 4: Special agroforestry extension programs for women need to be implemented. Creativeprograms like woodlots for women might have significant impact on increasing fuelwood availability. Although, selected agroforestry technologies, especially those designed for the production of mm, filelwood and related products are supposed to be an important source of filelwood, such filelwood was not readily made available to female farmers. As a result, these technologies did not reduce the shortage problem of fuelwood in the households. Women are still required to make their long treks in search of firewood from the declining natural forests or government plantations. This is causing significant negative impact on household food availability and nutrition as well as forest and environmental degradation. Implementation of special extension efforts targeted at women farmers would help to reduce these problems. Recommendation 5: It is recommended that efi’orts be made to develop marketing opportunities for agroforestry products, such as, fi'uits. The practice of planting and caring for trees at homesteads is a technology which was very popular partly due to its profitability. Fruits were the major product sold. It is therefore very important to promote markets for fiuit producing technologies. Recommendation 6: It is recommended that government loans be provided to enable 136 farmers with low income to practice the selected agroforestry technologies. Shortage of income was a major factor influencing farmer adoption of selected agroforestry technologies. Farmers who received more income in 1996 tended to practice the selected agroforestry technologies as compared to those who received less income. Income was required as capital for establishing and maintaining selected agroforestry technologies. Recommendation 7: It is recommended that agroforestry programs be introduced in schools, youth organizations and clubs. The Malawi government needs to introduce agroforestry in the curriculums and syllabi of both primary and secondary schools. Early primary school dropouts tend to become smallholder farmers. Targeting them early in the schools may be a very important strategy for increasing the number of farmers willing to practice agroforestry technologies. Recommendation 8: A form of incentives is recommended to encourage farmers to practice those technologies which may not benefit them directly. The value of this is the long term sustainability of the national resource base. Technologies like contour hedgerows are very important for maintaining the long term productivity of the land. Male farmers in the matrilineal system of marriage need to be encouraged to practice selected agroforestry technologies especially contour hedgerows. Incentives may be the answer. The Ministry of Forestry and Natural Resources already provides some incentives in order to encourage farmers to plant trees. Such incentives should be extended for encouraging practices like contour hedgerows. One type of creative incentive would be food for work programs. As observed 137 in chapter II, farmers were not able to produce enough food for their family needs fiom one season to the next. Instead of government providing free food to such firmers, they could be encouraged to implement these technologies in exchange for the food. RECOMMENDATIONS FOR FURTHER RESEARCH Recommendation 1:. This study basically relied on firmer perceptions of the effect of selected factors influencing farmer practice of selected agroforestry technologies. Further research directed at generating physical and economic evidence substantiating firmer perceptions of those fictors influencing the practice of selected agroforestry technologies needs to be conducted. Physical evidence of such factors as profitability, efi‘ect of the specified technologies on improving food crop yields, increasing fuelwood supplies and benefits of controlling soil erosion should be the focus of such research. Although several economic studies pertaining to selected technologies have been reported in the literature very few of the technologies discussed in this study were considered in those studies. Information is lacking about such technologies as: planting and caring for trees in garden boundaries; planting and caring for trees in croplands; and, planting and caring for trees at homesteads. Recommendation 2: It is recommended that efforts be made to develop improved fi'uit varieties for inclusion in the agroforestry programs. Fruit production was one area which made the practice of planting and caring for trees at homesteads more profitable. Improved varieties which would result in higher yields of good quality and palatable fruit would make the technology much more profitable and attractive to farmers. 1 3 8 Recommendation 3: Studies to determine the feasibility of including fi'uit trees in contour hedgerows need to be conducted. This technology is very important in terms of protecting the soil fi'om erosion and ensuring the long term productivity of the land. However, male firmers tended not to see benefits fiom this technology because of the land tenure problems. Inclusion of fi'uit production would make the technology profitable and this would possibly cause male firmers to recognize its benefits. PERSONAL REFLECTIONS AND OBSERVATIONS This section briefly describes personal reflections and observations made by the researcher. These reflections and observations are based on personal experiences and informal observations. 1. The practice of planting and caring for trees at homesteads was widely practiced in the EPA, yet the level of tree management was below recognized standard in most cases. Farmers could have been making much higher profits fiom this technology if they were better informed about tree management. Husbandry practices like pruning, weeding and manuring could possibly have increased fi'uit yields and improved fi'uit quality which would lead into higher profits. 2. The fact that the Malawi Agroforestry Extension Project was piloted to some farmers in the EPA seems to have created animosity among the farmers who were left out. Most of the farmers who did not respond to the study argued that they saw no benefit in providing information when they did not benefit fi'om the programs. 3. Farmers who were not exposed to the MAFE project seemed to perceive that they were being alienated by the project and therefore refused to participate in the interview. 139 4. The number of technologies investigated in this study were too many. IfI were to replicate this study I would reduce the number of technologies investigated. BIBLIOGRAPHY BIBLIOGRAPHY Adams, F. Unearthing Seeds of F ire: The Idea of Highlander. North Carolina: Winston- Salem, 1975. Agarwal, B. Cold Hearths and Barren Slopes: The Woodfiiel Crisis in the Third World. Maryland USA: Riverdale Company, 1986. Akinola, A. A. An Application of Probit Analysis to the Adoption of Tractor Hiring Service Scheme in Nigeria. Oxford Agrarian Studies, 1987, XVI, 72-82. Akinola, A. A. and Young, T. An Application of the Tobit Model in the Analysis of Agricultural Innovation Adoption Process: Study of the use of Cocoa Spraying Chemicals Among Nigerian Cocoa Farmers. Oxford Agrarian Studies. 1985, XIV, 26-5 1 . Anonymous, Annual Survey of Agriculture: Dedza Hills Results for 1987/88. Anonymous, Global Research on the Environmental and Agricultural Nexus for the let Century. University of Florida: Office of International Studies and Programs, 1995. Ashby, J. A. Technology and Ecology: Implications for Innovation Research in Peasant Agriculture. Rural Sociology, 1982, 42 (2), 234-250. Atala, T. K. The Relationship of Socio-economic Factors in Agricultural Innovation and Utilization of Information Sources in Two Nigerian Villages. The. Nigerian Journal ongricultural Extension, 1984, 2. l-l(). Atta-Krah, A. N. and Francis, P. A. The Role of On-Farm Trials in the Evaluation of Alley Farming: in Alley Farming in the Humid and Sub-humid Tropics. Proceedings of an international workshop held at Ibadan. Nigeria: March 10-14, 1986. Axin, G. H. Issues in Farming Systems Research: a Multi-disciplinary Behavioral Science Perspective. Farming Systems Research Group, paper no. 8 Michigan State University, 1981. Banda, A. Z., Maghembe, J. A.. Ngugi. D. N., and Chome, V. A. Effects of Intercropping Maize and Closely Spaced Leucaena Hedgerows on Soil Conservation and Maize Yield on a Steep Slope at Ntcheu. Malawi. Agroforestry systems, 1994, 27, 17-22. Beer, J. Litter Production and Nutrient Cycling in Coffee (Coffee arabica) or Cacao (Theobroma cacao) Plantations with Shade Trees. Agroforestry systems, 1988, 7, 103-114. 140 141 Benor, D. and Baxetr M. Training and Visit Extension. Washington D. C: World Bank, 1984. Benor, D. Harrison, J. Q. and Baxter, M. Agricultural Extension: The Training and Visit System. Washington D. C: World Bank, 1984. Bogunjoko, I. 0. Sources of Information of Improved Farm Practices: A Study of Farmers in Giwa District of Kaduna State. The Nigerian Journal of Agricultural Extension, 1981, 1 (2), 64-71. Brokensha, D. W., Warren, D. M. and Oswald, W. Indigenous Knowledge Systems and Development. Lanham, MD: University Press of America, 1980. Bunch, R. Encouraging Farmer Experiments: in R. Chambers, A. Pacey, A., and L. A. Thrupp, (Eds) Farmer First: Farmer Innovation and Agricultural Research. Intermediate Technology Publications, 1989. Bunderson, W. T., Itimu, O. A., Saka, A. R., Phombeya, H. S. K., and Mbekeani, Y. Optimum Management Practices for Alley Cropping Maize with Leucaena leucocephala in Malawi. Paper presented at the regional agroforestry symposium. Malawi: Lilongwe, June 16-21 1991. Bunderson, W. T., Phiri, G. S., Nhlane, L. M., and Nanthambwe, S. J. Malawi Agroforestry Extension Project: Project Description and Pre-implementation Plan. Malawi Lilongwe, 1992. Bunderson, W. T., Wellard, K., Phombeya, H. S. K., Kanyama-Phiri, G. Y., and Itimu, O. A. Agroforestry as a Low Input Technology to Improve Food Security: Potentials, Field Constraints, and Policy Approach to Research and Extension in Malawi. Paper presented at the workshop on food, agricultural and nutrition policy research in Malawi: setting the priorities, Malawi: Lilongwe, 1993. Bunderson, W. T., Nanthambwe, S. J., and Nhlane, L. M. A Synopsis of the Malawi Agroforestry Extension Project with Observations on the Adoption of Agroforestry Technologies: paper presented at the senior staff ADDFOOD seminar, Salima, Malawi, 1994. Bunderson, W. T., Bodnar, F., Bromley, W. A., and Nanthambwe, S. J. A Field Manual for Agroforestry Practices in Malawi. Malawi Agroforestry Extension Project, publication No. 6, Lilongwe, Malawi, 1995. Campbell, B. M., Frost, R, King, J. A., Mwanza, M., and Mlenga, L. The Influence of Trees on Soil Fertility on Two Contrasting Semi-arid Soil Types at Matopos, Zimbabwe. Agroforestry systems, 1994, 28, 159-172. 142 Carr, 8. J ., Modification and Extension of the National Rural Development Program. A paper presented at the symposium on agricultural policies for growth and development, Mangochi, Malawi, 1988. Chivangi, N. A., Engelhard, R. J. and Jones, V. Culture as the Basis for Implementation of Self-sustaining Woodfuel Development Programs. In: L. Fortmann and J. W. Bruce (Eds.) Whose Trees? Proprietary Dimensions of Forestry. Boulder and London: Westview Press, 1988, 243-253. Clay, J. Some General Principles and Strategies for Developing Markets in North America and Europe for Non Timber Forest Products. In M. Plotkin and L. Farnalore (Eds.) Sustainable Harvest and Marketing of Rain Forest Products, Washington DC: Island Press, 302-309, 1992. Collinson, M. On Farm Research with Farming Systems Perspective: In N. Roberts (Ed). Agricultural Extension in Africa: A World Bank Symposium. Washington D. C: World Bank, 1989. Cochran, W. G. Sampling Techniques. New York: John Wiley and Sons, 72-86, 1977. Coote, H. C., Luhanga, J. M., and Lowore, J. D. Community Use and Management of Indigenous Forests of Malawi: The Case of Chemba Village Forest Area. Zomba: Unpublished report, Forest Research Institute of Malawi, 1993. den Biggelaar, C. Differentiating the Nature of Agroforestry Systems and Agroforestry Knowledge among Farmers in Southern Rwanda: Participatory and Formal Approaches as Complementary Methods. unpublished, Michigan State University, 1994a. den Biggelaar, C. Philosophical Reflections on the Nature of Endogenous Knowledge in Rwanda. unpublished, Michigan State University, 1994b. Dewees, P. A. Forestry Policy and Woodfuel Markets in Malawi. Natural resources forum, 1995, 19(2), 143-152. Dewey, J. Democracy and Education. New York: The Free Press, 81-99, 1966. Fett, J. H. Education, Literacy, Mass Media Exposure and Farm Practice Adoption in Southern Brazil. Rural Sociology, 1971, 36 (3), 359-365. Fortmann, L. and Nabane, N. The Fruits of their Labors: Gender, Property and Trees in Mhondoro District. University of Zimbabwe: CASS occasional paper series, 1992. Freire, P. Pedagogy of the Oppressed. New York: The Seabury Press, 1970. 143 Gartrell, J. W., Wilkening, E. A. and Presser, H. A. Curvilinear and Linear Models Relating Status and Innovative Behavior: A Reassessment. Rural Sociology, 1973, 38. 391-409. Gosh, S. R, Mohan Kumar, B., Kabeerathumma, S. and Nair, G. M. Productivity, Soil Fertility and Soil Erosion Under Cassava Based Agroforestry Systems. Agroforestry Systems, 1989, 8, 67-82. Gupta, A., Scientists' View of Farmers' Practices in India: Barriers to Effective Interaction. In R. Chambers, A. Pacey, and L. A. Thrupp (Eds.) Farmer First: Farmer Innovation and Agricultural Research. Southampton, U. K: Intermediate Technology Publications, 24-30, 1989. Guggenheim, 8., and Spears, J ., Sociological and Environmental Dimensions of Social Forestry Projects. In M. M. Cemea (Ed.) Putting People First: Sociological Variables in Rural Development. Oxford University Press, Second edition, ch 9, 304-339, 1991. Hair, J. F. Jr. Anderson, R. E., Tatham, R L. and Black, W. C. Multivariate Data Analysis: With Readings (4th edition). New Jersey: Prentice Hall, Englewood Cliffs, 13-14. 1995. Hosier, R. H., The Economics of Smallholder Agroforestry: Two Case Studies. World Development, 1989, 17 (11), 1827-1839. ICRAF Annual Report for 1991. Kenya: Nairobi, 1992. ICRAF Annual Report for 1992. Kenya: Nairobi, 1993. Jacobsen, C. Principles and Methods of Extension Work. Israel : CINADCO, 21-30, 1993. Jones, R. B., Bunderson, W. T., and Itimu, O. A. Alley Cropping: A Partial Solution to Sustainable Crop Production. Paper presented at the first crop science congress for Eastern and Southern Africa. Uganda: Kampala, 14-18 June, 1993. Kang, B. T., Wilson, G. F., and Lawson, T. L. Alley Cropping: A Stable Alternative to Shifting Cultivation. lntemational Institute of Agriculture. Nigeria: Ibadan, 1984. Kang, B. T., Wilson, G. F. and Sipkens, L. Alley Cropping Maize (Zea mays L.) and Leucaena (Leucaena leucocephala Lam.) in Southern Nigeria. Plant and Soil, 1981, 63, 165-179. 144 Kang, B. T., Van der Kruijs, A. C. B. M., and Couper, D. C. Alley Cropping for Food Crop Production in the Humid and Sub-humid Tropics, in Alley Farming in the Humid and Sub-humid Tropics. Proceedings of an intemational workshop. Nigeria: Ibadan, March 10-14, 1986. Knowles, M., The Adult Learner: A Neglected Species. Gulf Publishing Company, Houston, ch 3, 29-49, 1973. Knowles, M. S. The Modern Practice of Adult Education: From Pedagogy to Adragogy. Chicago: Associated Press, Follet Publications, 1980. Krejcie, R. V. and Morgan, D. W. Determining Sample Size for Research Activities. Educational and Psychological Measurement, 1970, 30, 607-610. Kydd, J. Maize Research in Malawi: Lessons from Failure. Department of Agricultural Economics, University of London : Wye College, undated. Lal, R. Agroforestry Systems and Soil Surface Management of a Tropical Alfisol. Agroforestry systems, 1989, 28, 202-212. Lele, U. Structural Adjustment, Agricultural Development and the Poor: Some Lessons from the Malawian Experience. Paper prepared as part of the study "managing agriculture development in Africa", World Bank, 1988. Lindeman, E. C. The Democratic Man: Selected Writings of Eduard C. Lindeman. . Boston: Beacon Press, 1956. MacDicken, K. G. and Vergara, N. T. Introduction to Agroforestry. In K. G. MacDicken and N. T. Vergara (Eds.) Agroforestry: Classification and management, John Wiley and Sons, chl, 1—30, 1990. Malawi Government, Statement of Development Policies 1987-1996. Malawi: Zomba, 1988. Malawi Government Malawi Demographic and Health Survey. National Statistical Office, Malawi: Zomba, 1992. Malawi Government, National Population Policy (final draft). Office of the President and Cabinet: Department of Economic Planning and Development, 1993. Marsh, C. P. and Coleman, A. L. The Relation of Farmer Characteristics to the Adoption of Recommended Farm Practices. Rural Sociology, 1955, 288-296. 145 Mercer, D. E. and Soussan, J. (1992) Fuelwood Problems and Solutions. In N. P. Shanna (Ed.) Managing the world 's forests. Iowa, Dubuque: Kendall! Hunt Publishing Co., ch 8, 177-213. Masangano, C. M. A Descriptive Study of Staff and Farmer Perceptions of the Factors Influencing Smallholder Farmer Participation in Extension Activities of Dedza Hills Rural Development Project in the Kaphuka Area (EPA 5). Ms Thesis, Michigan State University, 1989. Mauriya, D. M. Farmers Participatory On-Farm Research Methodology: A Sustainable Model. in the 12th annual farming systems research symposium report. Association for Farming Systems Research/Extension. Michigan State University, 1992. Mkandawire, R. M. Customary Land, the State And Agrarian Change in Malawi: The Case of the Chewa Peasantry in the Lilongwe Rural Development Project. Journal of Contemporary African Studies, 1983, 3, 109-128. Mkandawire, R. M. Smallholder Survival Strategies, Perceptions, and Responsiveness to Extension and Credit in the Salima Agricultural Development Division: study prepared as part of the EEC funded mid term evaluation of Salima ADD phase IV, 1988. Mkandawire, R. M. The Land Question and Agrarian Change in Malawi. In G. C. Z. Mhone (Ed. )Malawi at the Crossroads: The Post-Colonial Political Economy. Harare, Zimbabwe. SAPES books, 1992. Mkandawire, R. M., and Chipande, G. H. R. Smallholder Agricultural Development in Malawi: The Case of a Targeted Approach: with Special References to Cases from Selected Agricultural Development Divisions. a paper presented at the Malawian symposium on agricultural policies for growth and development Mangochi, Malawi. October, 31-November 4, 1988. Molnar, A. and Schreiber, G. Women and Forestry: Operational Issues. Working paper no. 184, Washington DC: The World Bank, 1989. Monu, E. D., and Omole, M. M. Adoption of Recommended Farm Practices by Nigerian Cocoa Farmers. Nigerian Journal of Agricultural Extension, 1991,1 (2), 54-63. Moris, J. Extension Alternatives in Tropical Africa. Agricultural Development Unit, ODI, occasional paper no.7, 1991. Napier, T. L., Thraen, C. S., Gore, A. and Goe, W. R. Factors Affecting Adoption of Conventional and Conservation Tillage Practices in Ohio. Journal of Soil and Water Conservation, May-June 1984, 203-209. 146 Ndufa, J ., Ohlsson, E. and Shepherd, K. D. Participatory Research Methods for Agroforestry Technology Development. in the 12th annual farming systems research symposium report. Association for Farming Systems Research/Extension, Michigan State University, 1992. N g’ong’ola, D. H. Malawi Urban Household Energy Survey. University of Malawi: Bunda College of Agriculture, Lilongwe, 1992. Norusis, M. J. SPSS Advanced statistics 6.1. Chicago, Illinois: SPSS inc, 1-30, 1994. Noor, A., Education Lending for the Poor: World Bank and the Worlds Poorest. 24- 27, June 1980. Onyenwaku, C. E. and Mbuba, A. C. The Adoption of Seed Yam Minisett Multiplication Technique by Farmers in Anambra State, Nigeria. The Nigerian Journal of Agricultural Extension, 1991, 6 (1&2), 26-33. Opare, K. D. The Role of Agricultural Extension in the Adoption of Innovations by Cocoa Growers in Ghana. Rural Sociology, 1977, 42 (1), 72-82. Osuji, L. 0. Institutional Factors Associated with Adoption of New Farm Technologies Among Farmers in Eastern Nigeria. The Nigerian Journal of Agricultural Extension, 1991, 1 (2), 43-53. Palm, C. A. Contribution of Agroforestry Trees to Nutrient Requirements of Intercropped Plants. Agroforestry Systems, 1995, 30 (1&2), 105-124. Park, C. The Tropical Rainforest: History and Environment. In: Tropical Rainforests. London: Routledge, ch 1, 1-30, 1992. Peters, C. M., Gentry, A. H. and Mendelsonn, R. 0. Valuation of an Amazonian Rainforest. Nature, 1989, 339, 655-656. Pickering, D. Agricultural Extension and its Linkage with Agricultural Research: In N. Roberts (Ed). Agricultural Education in Africa: A World Bank Symposium. Washington, D. C: World Bank, 1989. Quinn, V., Chiligo, M. and Gittinger, J. P. Household Food and Nutritional Security in Malawi. A paper presented at the symposium on agricultural policies for growth and development, Mangochi, Malawi, November 1-4, 1988. 147 Randwanski, S. A. and Wickens, G. E. The Ecology of Acacia albida on Mantle Soils in Zalingeni, Jebel, Marra, Sudan. Journal of Applied Ecology, 1967, 4, 569-579. Rhoades, R. The Role of Farmers in the Creation of Agricultural Technology. In: R. Chambers, A. Pacey and L. A. Thrupp (Eds.) Farmer first: farmer innovation and agricultural research. Intermediate Technology Publications, Southampton, UK., 3-9, 1989. Rocheleau, D. E., Women, Trees and Tenure: Implications for Agroforestry Research and Development. In J. B. Raintree (Ed.) Land Trees and Tenure. ICRAF/Madison, Wisconsin, Land Tenure Center 79-119, 1987. Rocheleau, D. E. Women, Trees and Tenure: Implications for Agroforestry. In L. Fortmann and J. W. Bruce (Eds.) Whose Trees? Proprietary Dimensions of Forestry. Westview Press, Boulder and London, 243-253, 1988. Rocheleau, D. E. Gender, Ecology and the Science of Survival: Stories from Kenya. Agriculture and Human Values, 1991, 8 ( 1&2), 156-165. Rocheleau, D. E. Gender Ecology and Agroforestry: Science and Survival in Kathama. Ecogen case study series, Clark University: Graduate School of Geography, 1992. Rocheleau, D., Wachira, K., Malaret, L. and Wanjohi, B. M. Local Knowledge for Agroforestry and Native Plants: in R. Chambers, A. Pacey, and LA. Thrupp, (Eds.), Farmer First: Farmer Innovation and Agricultural Research. Intermediate Technology Publications, Southampton UK., 14-16, 1989. Rogers, E. M. Diffusion of Innovations (Third edition). New York: The Free Press, ch 1-3, 1-133. 1983. Rogers, E. M. and Shoemaker, F. F. Communication of Innovations: A Cross Cultural Approach (2nd edition). New York: The Free Press, ch 4, 136-171, 1971. Roling, N. The Changing Role of Agricultural Extension. Keynote paper presented at the workshop on agricultural extension in Africa. January 24-28, 1993. Ryan, B. and Gross, N. C. The Diffusion of Hybrid Seed Corn in Two Iowa Communities. Rural Sociology, 1943, 8, 15-24. Sahn, D. E. and Arulpragasam, J. Development Through Dualism: Land Tenure, Policy and Poverty in Malawi. Washington, DC: CFNPP Publications Department, 1991. Saka, A. R., Bunderson, W. T., Itimu, O. A., Phombeya, H. S. K., and Mbekeani, Y. The Effects of Acacia albida on Soils and Maize Grain Yields Under Small Farm Conditions in Malawi. Forest Ecology and Management, 1994, 64, 217-230. 148 Sanchez, P. A. Science in Agroforestry. Agroforestry systems, 1995, 30, 5-55. Silver, H. Benefit-cost Analysis of Agroforestry Practices: Tumpangsari and Inmas Tumpangsari in Cepu Forest District: J ava, Indonesia, PhD dissertation, Michigan State University, East Lansing, Michigan, 1991. Sharma, R., Sharma, E. and Purohit, A. N. Dry Matter Production Nutrient Cycling in Agroforestry Systems of Cardamom Grown Under Alnus and Natural Forest. Agroforestry systems, 1994, 27, 293-306. Shaxson, T. F., Hunter, N. D., Jackson, T. R. and Alder, J. R. A Land husbandry Manual for Malawi. Government Printer, Zomba, Malawi 1977. Smale, M., Chimanga cha Makolo, Hybrid Maize, and Composites: An Analysis of Farmers Adoption of Maize Technology in Malawi: CIMMYT economics working paper 91/04. CIMMYT Mexico, 1989-90. Wendt, J. W., Jones, R. B., and Bunderson, W. J. The Long Term Effects of Leucaena Leaf Application on Soil Chemical Properties in an Alley Cropping System. unpublished, Malawi, 1993. Wharton, C. R. Jr. Education and Agricultural Growth: The Educational Implications of the Requirements for Agricultural Progress. Paper presented at the second annual conference of agricultural economics society of Thailand, 23, 1 February, 1963. Wilson, G. E., and Kang, B. T. Developing Stable and Productive Biological Cropping Systems for the Tropics and Subtropics. in B. Stonehouse (Ed.) Biological husbandry: a scientific approach to organic farming. 193-203, 1981. Zinnah, M. M. The Adoption and Impact of Improved Mangrove Swamp Rice Varieties in West Africa: the Case of Guinea and Sierra Leone. PhD Dissertation, University of Wisconsin, Madison, Wisconsin, 1992. APPENDIX A APPENDIX A Respondents’ exposure to information and demonstrations Table A1: Number of Respondents who Received Information and Seen Demonstrations of Selected Agroforestry Technologies in Njolomole EPA of Ntcheu RDP. (N=399) Agroforestry Received Seen demonstration Not sure/no response technology information number % number % number % PA 143 35.8 108 27.1 148 37.1 AC 103 25.8 93 23.3 203 50.9 RC 79 19.8 73 18.3 247 61.9 105 26.3 73 18.3 221 55.4 CH 143 37.1 128 32.1 128 32.1 WL 140 35.1 125 31.3 134 33.6 TGB 104 26.1 86 21.6 209 52.4 TC 94 23.6 82 20.6 223 55.9 TH 135 33.8 108 27.1 156 39.1 FB 89 22.3 85 21.3 225 56.4 LP 88 22.1 76 19.1 235 58.9 149 150 Table A2: A Comparison of Respondent Perceptions Between Respondents who were and those who were not Exposed to the Malawi Agroforestry Extension Project Regarding their Exposure to Extension Information Designed to Promote the Practice of Selected Agroforestry Technologies. Agroforestry Exposed to MAFE project (n=64) Not exposed to MAFE project (n=335) technol ogy received did not not received did not not surelnon information receive sine/non information receive respondent information respondent information FA 54 4 6 89 172 74 (84.4) (6.3) (9.4) (26.6) (51.3) (22.1) AC 54 5 5 49 196 90 (84.4) (7 .8) (7 .8) (14.4) (58.5) (26.9) RC 43 4 17 36 214 85 (67.2) (6.3) (26.6) (10.8) (63.9) (25.4) IF 41 5 18 64 190 81 (64.1) (7 .8) (28.1) (19.1) (56.7) (24.2) CH 58 4 2 90 182 63 (90.6) (6.3) (3.1) (26.9) (54.3) (18.8) WL 44 14 6 96 180 59 (68.8) (21.9) (9.4) (28.7) (53.7) (17.6) TGB 43 7 24 61 203 71 (67.2) (10.9) (37.5) (18.2) (60.6) (21.2) TC 43 18 3 51 216 68 (67.2) (28.1) (4.7) (15.2) (64.5) (20.3) TH 42 18 4 93 190 94 (65.6) (28. l) (6.8) (27.8) (56.7) (28. 1) FB 57 4 3 32 206 97 (89.1) (6.3) (4.7) (9.6) (61.5) (29.0) LP 42 17 5 46 207 82 (65.6) (26.6) (7 .8) (13.7) (61.8) (24.5) FA is F: albedo. AC is alley cropping, RC is relay cropping, IF is improved fallows, CH is contour hedgerows, WL is woodlots,TGBistreesingardenboundaricsTC istreesincroplands,‘I'I-Iistreesathomesteads,FBisfodderbanksandLF is living fences. 151 TableA3: AComparison ofRespmdentPercepfimsBetweurRespondentswhowa-eandthosewhowuenot ExposedtotheMalsudAgroforesfiyExtensioanjectonWhethertbeyhadSeenDemonmfionsMountedby Ernension Staff on the Selected Agroforestry Technologies- Agroforesu'y Exposed to MAFE project (n=64) Not exposed to MAFE project (n=335) l tecno 08y seen not seen not sure/non seen not seen not surelnon respondent respondent FA 54 9 l 54 206 75 (84.4) (14. 1) (1.6) (16.1) (61.5) (22.4) AC 48 1 1 5 45 204 86 (75.0) (17.2) (7 .8) (13.4) (60.9) (25.7) RC 44 l l 9 29 216 90 (68.8) (17.2) (14.1) (8.7) (64.5) (26.9) IF 43 13 8 30 214 91 (67.2) (20.3) (12.5) (9.0) (63.9) (27.2) CH 54 9 l 74 185 76 (84.4) (14.1) (1.6) (22. 1) (55.2) (22.7) WL 43 18 3 82 194 59 (67.2) (28. l) (4.7) (24.5) (57.9) (17.6) TGB 46 ll 7 40 214 81 (71.9) (17.2) (10.9) (11.9) (63.9) (24.2) TC 44 17 3 38 216 81 (68.8) (26.6) (4.7) (l 1.4) (64.5) (24.2) TH 44 16 4 64 205 66 (68.8) (25.0) (6.3) (19.1) (61.2) (19.743)18 FB 52 10 2 33 210 92 (81.3) (15.6) (3.1) (9.9) (62.7) (27.5) LP 44 17 3 32 215 88 (68.8) (26.6) (4.7) (9.6) (64.2) (26.3) FA isF. Albida, AC is alley cropping, RC is relay cropping,IF is improved fallows, CH is contour hedgerows, WL is wood ltos,TGBistreesingardenboundanesTCisneesmcmplmds,Ttheesuhanamads,FBisfoddabanbmdLFk living fences. 152 Table A4: A Comparison of Respondent Perceptions of Whether they were Influenced by Information and Demonstrations Provided by Extension to Practice the Selected Agroforestry Technologies Between Respondents who were and those who were not Exposed to the Malawi Agroforestry Extension Project Agroforestry Exposed to MAFE project(n=64) Nto exposed to MAFE project (n=335) technol ogy influenced not not influenced not influenced not sire/non influenced mire/non respondent respondent FA 45 13 6 24 204 107 (70.3) (20.3) (9.4) (7.2) (60.9) (31.9) AC 21 25 18 13 198 124 (32.8) (39.1) (28.1) (3.9) (59.1) (37.0) RC 25 22 17 9 219 107 (39. 1) (34.4) (26.6) (2.7) (65.4) (31.9) IF 23 20 21 15 209 11 1 (35.9) (31.3) (32.8) (4.5) (62.4) (33.1) CH 54 8 2 47 187 101 (84.4) (12.5) (3.1) (14.0) (55.8) (30.2) WL 29 25 10 57 193 85 (45.3) (39.1) (15.6) (17.0) (57.6) (25.4) TGB 27 20 17 18 215 102 (42.2) (31.3) (26.6) (5.4) (64.2) (30.5) TC 43 20 1 17 226 92 (67.2) (31.3) (1 .6) (5.1) (67.5) (27.5) TH 28 29 7 48 21 1 76 (43.8) (45.3) (10.9) (14.3) (63.0) (22.7) FB 29 20 15 8 204 123 (45.3) (31 .3) (23.4) (2.4) (60.9) (36.7) LP 22 31 11 7 218 l 10 (34.4) (48.4) (17.2) (2.1) (65.1) (32. 8) FA isF. Albida, AC isalleycropping, RC is relay cropping,IF is improved fallows,CHiscontour hedgerows,WL iswood MTGBistreeaingardenboundarieaTCisneesincmplands,misuwaathomesteads,FBisfodderbankaandLFis livingfences. APPENDIX B ”FINDII B 1. LOGISTIC W381“ IODBL POI CODI'I'UOR ”WHO'S lumber of selected cases: 399 lumber rejected because of missing data: 213 lumber o! cases included in the analysis: 106 Dependent veriable.. PRACTCBl Do you practice contour hedgerows? Initial Log Likelihood Function -2 Log Likelihood 220.42735 veriable(s) Intered on step lumber onlno Did you impliment contour hedgerows for erosion control? IROSIGI2 Do you experience soil erosion problems on your term? LANDQIIU Do you have secure rights or land tenure? PROJICT Do you belong to MAID project? lstimation terminated at iteration number 4 because Log Likelihood decreased by less than .01 percent. -2 Log Likelihood 176.109 Goodness o! tit 196.190 Chi-square dt signiticance Ibdel Chi-Square 44.319 4 .0000 Impmnt 44.319 4 .0000 Classification Table tor PRACTCBl Predicted 1.00 2.00 Percent Correct 1 2 Observed 1.00 1 124 10 92.54% 2.00 2 20 24 46.15% Overall 79.57% ----------------------- Variables in the Iquation -—----------------- variable 8 8.3. Weld d1 Sig R lxp(D) eunnoti) -1.4359 .6412 5.0146 1 .0251 -.1169 .2379 IROBION2tl) .1317 .2321 .3210 1 .5705 .0000 1.1407 LANDTINU(1) -.9790 .2326 17.7124 1 .0000 -.2670 .3757 PROJICT(1) -.7525 .3039 6.1302 1 .0133 -.1369 .4712 Constant .4920 .6761 .5313 1 .4661 153 154 APPINDIX B 2. LOGISTIC RIGRISSION IDDIL I08 WOODLOTB number of selected cases: lumber rejected because of missing data: 399 235 Number of cases included in the analysis: 164 Dependent Variable.. beginning Block number -2 Log Likelihood PRACTWLl Do you practice woodlots? 0. Initial Log Likelihood Punction 172.62300 Variable(s) lntered on Step lumber PROJICT GENDER INCGNll "LIUIL2 1.. Do you belong to HAP! project? Kale or female? less than 8499.00 Was fuel shortage solved by‘uoodlots? lstimation terminated at iteration number 5 because Log Likelihood decreased by less than .01 percent. -2 Log Likelihood 111.076 Goodness of Pit 167.310 Chi-Square df significance Model Chi-Square 60.747 4 .0000 Improvement 60.747 4 .0000 Classification Table for PRACTWL1 IXP(B) .4200 .2003 2.3912 1.3911 Predicted 1.00 2.00 Percent Correct 1 2 Observed 1.00 1 125 3 97.66% 2.00 2 22 14 30.09% Overall 04.76h ----------------------- veriables in the Iquation ------------------ Variable D 8.3. Held df 8ig R PROJICT(1) -.0674 .4016 3.2440 1 .0717 -.0049 OINDIR(1) -1.2430 .2603 22.0403 1 .0000 -.3475 INCOH31(1) .0710 .2714 10.3162 1 .0013 .2195 ILPUIL2(1) .3301 .2652 1.5409 1 .2133 .0000 Constant -1.7404 .4449 15.3004 1 .0001 155 APPENDIX D 3. LOGISTIC RIGRIBSION NDDIL FOR T3338 IN GARDIN BOUNDARIES lumber of selected cases: 399 Number rejected because of missing data: 51 Number of cases included in the analysis: 340 Dependent Variable.. Beginning block number 0. -2 Log Likelihood beginning block number 1. PRACTIIi Do you practice planting and caring of trees in garden boundaries Initial Log Likelihood Function 409.94501 Hethod: Inter veriable(s) lntered on Step lumber 1.. XIYIILDC INCONll INCONI3 AG! GINDIR PROJICT IROBIGNZ LANDI Did the practice of planting and caring for trees in garden boundaries increase or decrease your food crop yields? Income received in 1996 (less than x499.oo or otherwise) “ , “ “ (£2,000.00 to £2,999.00 or otherwise) Male or female Do you belong to IAFI project? Do you experience soil erosion problems on your farm? landholding size in hectares lstimation terminated at iteration number 4 because Log Likelihood decreased by less than .01 percent. -2 Log Likelihood 263.459 Goodness of Fit 335.365 Chi-Square df significance lodel Chi-Square 146.406 0 .0000 Improv-ent 146.406 0 .0000 Classification Table for PRACTxxl Predicted 1.00 2.00 Percent Correct 1 2 Observed 1.00 l 56 40 50.33% 2.00 2 0 244 96.03% Overall 06.21% ----------------------- Variables in the lguation ------—------------- Variable XIYIILD4(1) INCOH31(1) AG! GINDIR(1) PROJECT(1) IROSION2I1) LANDN INCOII3 Constant -1.9592 .4163 .0120 -.3302 -.3000 .0990 -.0765 -.9346 .1599 .2247 .1910 .0117 .1603 .2210 .1700 .1567 .4907 .7700 Wald 76.0392 4.7407 1.0401 4.0409 1.0393 .3144 .2302 3.6201 .0431 df H1‘l‘t‘fi‘b‘b'H'H Dig .0000 .0293 .3060 .0444 .1750 .5750 .6255 .0560 .0355 -.4250 .0019 .0000 - e070‘ .0000 .0000 .0000 -.0630 prtD) .1410 1.5164 1.0121 .7130 .7402 1.1049 .9264 .3927 157 APPINDIX D 4. LOGISTIC RRGRRBSION NDDIL FOR T8228 IN CROPLANDS Number of selected cases: 399 lumber rejected because of missing data: 243 lumber of cases included in the analysis: 156 PRACTTCl Do you practice planting and caring of trees in croplands Dependent variable.. beginning block Number 0. Initial Log Likelihood Function -2 Log Likelihood 140.5751 Variable(s) Intered on Step number 1.. PROJICT Do you belong to NAP! project? TCYIBLD4 did trees in croplands increase yield TCFUIL2 was fuel shortage solved by trees in croplands? IROBIOI2 Do you experience soil erosion problems on your farm? GllDlR Isle or female bstimation terminated at iteration number 6 because Log Likelihood decreased by less than .01 percent. -2 Log Likelihood 94.470 Goodness of Fit 456.303 Chi-square df significance lodel Chi-Square 46.105 5 .0000 Improvement 46.105 5 .0000 Classification Table for PRACTTCl Predicted 1.00 2.00 Percent Correct 1 2 O bserved 1.00 l 124 6 95.30% 2.00 2 0 10 69.23% Overall 91.03% ----------------------- Veriables in the bquation ------—------------------ Variable b 8.8. Held df Big 8 pr(b) PROJICT(1) -.1005 .4913 .1349 1 .7134 .0000 .0349 TCTIILD4(1) -1.1464 .3029 0.9637 1 .0020 -.2226 .3170 TCFUbL2(l) .0175 .3464 .0026 1 .9596 .0000 1.0177 lbOSION2(1) .7093 .5790 1.4965 1 .2212 .0000 2.0326 GINDIR 2.1770 .5430 16.0073 1 .0001 .3166 0.0272 Constant -5.0199 1.1246 26.7001 1 .0000 150 APPENDIX B 5 LOGISTIC NOD81. FOR T3828 AT IONICTIADS number of selected cases: 399 number rejected because of missing data: 51 Number of cases included in the analysis: 340 Dependent variable.. PRACTTSl Do you practice planting and caring for trees in croplands? beginning block number 0. Initial Log Likelihood Function -2 Log Likelihood 401.20030 Variable(s) bntered on Step number 1.. m‘ TIPROF2 was the practice of planting and caring for trees in croplands more or less profitable than your previous practices? lstimation terminated at iteration number 4 because Log Likelihood decreased by less than .01 percent. -2 Log Likelihood 301.320 Goodness of Fit 351.100 Chi-Square df Significance Model Chi-Square 179.953 2 .0000 Improvement 179. 953 2 .0000 Classification Table for PnACTTbl Predicted 1.00 2.00 Percent Correct 1 2 Observed 1.00 1 140 44 76.09% 2.00 2 14 150 91.46% Overall 03.33% --------------- ------- Variables in the lguation ----------------------- Variable b 8.8. weld df Sig R lxp(b) AG! -.0134 .0097 1.0907 1 .1602 .0000 .9067 TEPROF2 1.9363 .1933 100.3037 1 .0000 .4519 6.9327 Con-tent -4.0559 .6543 30.4247 1 .0000 APPENDIX C APPENDIX C RESEARCH QUESTIONNAIRE PRACTICE OF SELECTED AGROFORESTRY TECHNOLOGIES: FARMER PERCEPTIONS OF INFLUENTIAL FACTORS IMPORTANT NOTE TO INTERVIEWER: Read these comments to the respondent and be sure to get his or her consent before starting the interview. We are part of a research team from Michigan State University. We are studying farmer perceptions of selected factors influencing farmer practice of recommended agroforestry technologies in N jolomole EPA of Ntcheu RDP. We would like to ask you some questions concerning the factors you consider important for the practice of selected agroforestry technologies. Your responses will be used in designing appropriate research and extension programs on agroforestry. The interview takes about one hour to complete. Your participation in this interview is voluntary. You are free to not participate in the interview. You are also free to not answer any of the questions you choose. You may also discontinue the interview at anytime. Your answers will remain confidential and will not be disclosed in any way that you can be identified. Do you wish to participate? Yes no (circle the right answer) You indicate your voluntary agreement to answer any of the questions by answering them. 159 160 MALE AND FEMALE FARMER PERCEPTIONS OF SELECTED FACTORS INFLUENCING THE PRACTICE OF RECOMMENDED AGROFORESTRY TECHNOLOGIES IN NJ OLOMOLE EPA OF NTCHEU RDP IN MALAWI. RESEARCH QUESTIONNAIRE Part A: Demographic information Note to interviewer: Read the following to the respondent. “In this part of the interview I am going to ask for some demographic information about you”. 1. Name of farmer 2. Village 3. Male female 4. Marital Status married divorced widowed never married other (specify 5. What is your year of birth? 6. What is the highest level of education you attained? 0 primary school (grade) secondary school (form) teacher training (number of years) technical school (number of years) university (qualification) 7a. What is the size of your farm (total acreage f or all gardens)? 7b. How many gardens do you have? 161 8a. Do you consider yourself as having secure rights of ownership of your farm land? yes . no not sure 8b. Do you consider yourself as having secure rights of ownership of trees growing on your farm land? yes no not sure BC. In the case of divorce, who would have control of ? (a) your farm land yourself your spouse your children other relative (specify relationship) other (specify) ' (b) trees you have planted on your farm land yourself your spouse your children other relatives (specify relationship) other (specify) 10. What was your total family income from all sources in the year 1996. (interviewer, just circle the letter representing the category). A. less than K499.00 B. K500.00-999.00 C. K1,000.00-l,999.00 D. K2,000.00-2,999.00 E. K3,000.00-3,999.00 F. K4,000.00-4,999.00 G. K5,000.00-5,999.00 H. above K6,000.00 162 Part B: Practice Note to interviewer: Read the following to the respondent. “In this part of the interview I am going to ask you questions relating to information about the utilization of selected agroforestry technologies in your farming practices. 11a. Do you practice the following agroforestry technologies? yes no Faidherbia albida trees in cropland contour hedgerows fodder banks alley cropping wood lots trees on garden boundaries trees on homesteads improved fallows relay cropping living fences 163 llb. Please indicate the extent to which you practice the following agroforestry technologies. Faidherbia albida number of trees __ estimated area (acres) trees in cropland number of trees total area (acrea) contour hedgerows number of rows—— estimated length (total for all rows in meters) fodder banks number of rows —-estimated length (total for all rows in meters) alley cropping number of rows__ estimated length (total for all rows in metersL_ wood lots number of wood lots—— area in acres (total for all wood lots)— trees on garden boundaries number of trees ——1ength of boundaries with trees (total)-—- trees on homestead number of trees total area (acres)__ improved fallows number of trees total area (acres) relay cropping number of trees _ total area (acres) living fences number of rows — total length (metres)—— 164 12. Which of the following agroforestry technologies were you practicing prior to 1992? (check) Faidherbia albida trees in croplands contour hedgerows fodder banks alley cropping wood lots trees on garden boundaries trees on homestead improved fallows relay cropping living fences 165 Part c: Project Note to interviewer: Read to the respondent. The Malawi Government set up the Malawi Agroforestry Extension Project in 1992. The main objective of the project is to develop an agroforestry extension system which will to promote the adoption and practice of recommended agroforestry technologies nationwide. This project is piloted in five extension planning areas including Njolomole EPA. The project is therefore implementing extension activities to promote the adoption and practice of recommended agroforestry technologies in this EPA. Some of the activities include farmer u'aining, demonstrations, meetings, distribution of posters, pamphlets and other publications on recommended g, agroforestry technologies. ‘ “In this part of the interview I will ask you questions relating to information about the extension activities of the Malawi Agroforestry Extension Project. 13a. Have you received information provided by the Malawi Agroforestry Extension Project on the following agroforestry technologies? yes no Faidherbia albida trees in croplands contour hedgerows fodder banks alley cropping wood lots trees on garden boundaries trees on homestead improved fallows relay cropping living fences 166 13b. Have you seen demonstrations conducted the Malawi Agroforestry Extension Project on the following agroforestry technologies? yes no Faidherbia albida trees in croplands contour hedgerows fodder banks alley cropping wood lots trees on garden boundaries trees on homesteads improved fallows relay cropping living fences 167 Note to interviewer: Ask question 14 only if the respondent is practicing the particular technology 14. Did the information and demonstrations provided influence your decision to practice the following agroforestry technologies? yes no Faidherbia albida trees in croplands contour hedgerows fodder banks alley cropping wood lots trees on garden boundaries trees at homestead improved fallows relay cropping living fences 168 15. For which of the following agroforestry technologies were you invited to participate in the extension activities provided by the Malawi Agroforestry Extension Project? (check) invited Faidherbia albida trees on croplands trees on croplands contour hedgerows fodder banks alley cropping wood lots trees on garden boundaries trees at homestead improved fallows relay cropping living fences Part D:. Profitability Note to interviewer: Read the following to the respondent. "Ihe term profit in this study is defined in financial terms as the surplus income realized at the farm level after all the production costs have been accounted for. Accordingly the term profitability is defined in relative terms as the degree to which one technology gives more profit than other technologies. The term profitability is therefore restricted to financial terms only. _ Other benefits which are not financially quantifiable at farm level like some of the benefits relating to environmental protection are not included in this definition. In this part of the interview I will ask you questions relating to the profitability of the agroforestry technologies that you practice”. Note to interviewer: Please refer only to those technologies which the respondent has indicated that he /she practices when asking questions 16, 17, 18, 19, 20, 21, 22 and 23 169 16. Are the technologies which you practice more or less financially profitable to you than your previous practices? more less If less profitable, why do you profitable profitable continue to practice them? Faidherbia albida trees in croplands contour hedgerows fodder banks alley cropping wood lots trees on garden boundaries trees on homesteads improved fallows relay cropping living fences 17. What produce! products are you getting from the woody species of the following agroforestry technologies? (check) a) Faidherbia albida fruit ——- firewood —— charcoal staking sticks —— poles —— livestock feed medicines —— pesticides timber others (specify) 170 trees in croplands fruits firewood staking sticks __ poles livestock feed charcoal medicines pesticides—-— timber other (specify) b) Contour hedgerows fruits __ firewood charcoal staking sticks — poles livestock feed medicines —— pesticides timber others (specify) c) fodder banks fruits firewood charcoal stacking sticks poles livestock feed medicines — pesticides timber others (specify) d) alley cropping fruits firewood — charcoal staking sticks — poles _ livestock feed medicines pesticides — timber others (specify) 171 e) wood lots fruits — firewood _— charcoal stacking sticks — poles — livestock feed medicines __ pesticides— timber others (specify) 1) trees on garden boundaries fruit — firewood _ charcoal staking sticks poles —— livestock feed medicines —— pesticides— timber others (specify) g) trees on homesteads fruits firewood charcoal staking sticks poles __ livestock feed medicines __ pesticides __ timber others (specify) improved fallows fruit firewood charcoal staking sticks __ poles livestock feed __ medicines __ pesticides __ timber other (specify) 172 relay cropping fruit firewood charcoal staking sticks poles __ livestock feed medicines pesticides timber other (specify) living fences fruits firewood _ charcoal staking sticks __ poles livestock feed medicines pesticides timber others (specify) 18a. How close is the nearest market where you can sell the produce/products? kilometers fruit firewood charcoal staking sticks livestock feed poles timber medicines pesticides others (specify) 173 18b. Do you sell the produce/products at the identified market? If not why? yes no if no why fruit firewood charcoal staking sticks livestock feed poles timber medicines pesticides others (specify) 18c. For the products which you sell at the identified market, are you satisfied with the prices offered? yes no fruit firewood charcoal staking sticks livestock feed poles timber medicines pesticides others (specify) 174 18d. Do you sell your products through middlemen? yes no fruit firewood charcoal staking sticks livestock feed poles timber medicines pesticides others (specify) 18e. If you sell through middlemen, are you satisfied with the prices you get from them? yes no fruit firewood charcoal stacking sticks livestock feeds poles timber medicines pesticides others (specify) 175 Part E: Food productivity Note to the interviewer: Read to the respondent. “Some agroforestry technologies have been reported to increase crop yields, through various processes like improvement of soil fertility and improvement of soil organic matter content. In fact implementation of agroforestry technologies has sometimes been promoted in order to increase crop yields and crop production. In the following part of the interview, I will ask you questions relating your food productivity”. 19. If you practice the following agroforestry technologies, have these practices helped you to increase or have they decreased your food crop yields? Faidherbia albida trees in croplands contour hedgerows fodder banks alley cropping wood lots trees on garden boundaries trees on homesteads improved fallows relay cropping living fences practice increased yields decreased yields not sure 176 Note to the interviewer: Please refer to only those technologies which resulted in decreased yields. 20. Why do you continue to practice the technologies when you are experiencing decreased yields of your food crops? Faidherbia albida trees in cropland contour hedgerows fodder banks alley cropping wood lots trees on garden boundaries trees on homesteads improved fallows relay cropping living fences 177 Part F: Control of soil erosion Note to interviewer: Read the following to the respondent. “Some agroforestry technologies have been reported to be very effective in controlling soil erosion. In fact implementation of agroforestry technologies has in some cases been promoted in order to control soil erosion. In this section of the interview I will ask you question relating to the control of soil erosron. 21. Do you experience soil erosion problems on your farm? yes I'IO I'IOt sure 22. Which of the following agroforestry technologies did you implement for purposes of controlling soil erosion? (check) Faidherbia albida trees in croplands contour hedgerows fodder banks alley cr0ppin g wood lots trees on garden boundaries trees on homesteads improved fallows relay cropping living fences 178 23. Was soil erosion controlled by the use of the following agroforestry technologies?. yes no not sure Faidherbia albida trees in croplands contour hedgerows fodder banks alley cropping wood lots trees on garden boundaries trees on homesteads improved fallows relay cr0pping living fences Part G Fuelwood Note to the interviewer: Read the following to the interviewer.. In some cases agroforestry technologies have been promoted in order to produce fuelwood (firewood and charcoal). In the following section of the interview, I will ask you questions relating to fuelwood. 24. Do you experience problems of fuelwood shortage? yes I'lO 179 25. Which of the following agroforestry technologies did you implement for purposes of producing fuelwood? Faidherbia albida trees in croplands contour hedgerows fodder banks alley cropping wood lots trees on garden boundaries trees on homesteads improved fallows relay cropping living fences yes no 31“.“ ...—.- 1".- I '12.“ O 1. 180 26. Were your problems of fuelwood shortage solved by the following agroforestry technologies? yes no Faidherbia albida trees in croplands contour hedgerows fodder banks alley cropping wood lots trees on garden boundaries trees on homesteads improved fallows relay cropping living fences 27. If you have other reasons why you practice agroforestry technologies apart from those mentioned above, please indicate them.