SOIL FERTILITY REGENERATION THROUGH IMPROVED FALLOW SYSTEMS IN SOUTHERN MALI By BOCARY KAYA A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2000 To my beloved wife, Arunya Limbanyen, and to our lovely daughter, Aissata L.Y. Kaya, born here in Gainesville during this program. My studies would not have been successfully completed without your inestimable devotion, love, care, and understanding. Youmean everything to me, and I have no words good enough to express my feeling toward you. Please accept this simple expression: "Thank you." ACKNOWLEDGMENTS I would like to express my sincere gratitude to Dr. P.K.R. Nair, chairman of my supervisory committee, for his constant guidance, advice, and his professional and moral support during the course of my entire study program at the University of Florida. I also would like to thank the other members of my supervisory committee -Drs. P.E. Hildebrand, G. Kidder, P.E. Linehan, and L.E. Sollenberger. This research certainly would have been more difficult, if not impossible, were it not for the support and encouragement that Dr. Nair and the other committee members extended on many occasions. I am very grateful to Dr. R.C. Littell, professor of statistics at the University of Florida, for his patience, help, and guidance at the difficult moments of the statistical analysis of the data. I would like to express my sincere gratitude to the US Agency for International Development (USAID) and the Malian national research institute, Institut d'Economie Rurale (lER), for having funded this program through the Programme d'Appui a la Recherche Agronomique (IER-PARA/USAID). I 111 extend thanks to the Government of Mali for granting me study leave to complete this program. My gratitude goes to Drs. M. Doumbia, soil scientist, and A. Gakou, agronomist of the lER, and their colleagues for the local supervision and guidance they providedme at the N'Tarla research station and the Sotuba soils and plant laboratory. Many thanks to the Sikasso farming systems research team (ESPGRN) for logistical support and for sharing data and other information vital for this research. I would also like to thank the International Foundation for Science, Stockholm, the Rockefeller Foundation, New York, and the Center for Latin American Studies at the University of Florida, Gainesville, for the partial funding I received from these organizations for my research in Mali. During my five years at the School of Forest Resources and Conservation of the University of Florida, I have had the chance to interact with many past and present (agro)forestry graduate students-Abuid Mwale, Alyson Degang, Andrea Albertin, Bill Gibbs, Daniel Mugendi, Donald Mee, Eddie Ellis, Frank Merry, Jimmy Knowles, John Bellow, Mike Bannister, Paul Thangata, Robert Miller, and Shinjiro Sato-who have been extremely helpful and understanding. My sincere thanks and appreciation go to each of them. IV During itiy stay in Gainesville, Florida, I was very fortunate to have met Paul Campbell, his wife. Jinny, and their two sons, Philip and Steven; Stanley and Elizabeth Muchmore; Mary Teegardin; and Perry Keidel, his wife, Monica Kypreos, and their daughter, Alexandra. Despite our social and religious differences, all these individuals were generous in sharing their love, friendship, and understanding with me and my family. Finally, I would like to express my sincere gratitude to my friends, Bokary Allaye Kelly and Amadou Haidara, to my family in Mali, and to my wife's family in Bangkok, Thailand, for having demonstrated unconditional love and devotion to me and to my family here in the U.S. during my Ph.D. program at the University of Florida. V TABLE OF CONTENTS Page ACKNOWLEDGMENTS iii LIST OF TABLES ix LIST OF FIGURES xiii ABSTRACT XV CHAPTERS 1 GENERAL INTRODUCTION 1 2 LITERATURE REVIEW 9 Introduction 9 Traditional Tree-based Land-use Systems of Semiarid SSA 10 New Approaches to Tree-based Land-use Systems 27 Directions for the Future of Tree-Based Soil Fertility Improvement in Semiarid SSA 41 Use of Linear Programing (LP) as Decision Making Support 44 Conclusion 50 3 SOIL PROPERTIES AND MAIZE PRODUCTION IN AN ON-FARM TRIAL OF IMPROVED FALLOWS WITH THREE SPECIES AT N'GOUKAN, MALI 52 Introduction 52 Materials and Methods 54 Description of the Study Area 54 Experiment Setup, Management, and Sampling Scheme 55 Results 51 MPT Survival, Growth Rate, and Production of DM .... 61 Maize Dry Matter and Grain Yield 62 VI Soil Changes 64 Discussion 68 Conclusion 73 4 SOIL PROPERTIES AND MAIZE (ZEA MAYS) PRODUCTION AS AFFECTED BY THE TRANSFER OF DIFFERENT-QUALITY BIOMASS AT N'TARLA, MALI 75 Introduction 75 Materials and Methods 78 Description of the Study Area 78 Experiment Setup, Management, and Sampling Scheme 79 ResultMsaize Dry Matter and Grain Yield .... 8855 Maize N Uptake .... 87 Rate of Decomposition of Biomass 91 Maize N Uptake Efficiency 91 Soil Aspects 94 Discussion 100 Conclusion 107 5 SOIL ORGANIC MATTER DYNAMICS FOLLOWING ADDITION OF BIOMASS OF DIFFERENT FALLOW-IMPROVEMENT SPECIES 109 Introduction 109 Materials and Methods 113 Results 116 Soil Particulate Organic Matter (POM) . 116 Nitrogen Content of Particulate Organic Matter (POM^) 123 Relationship Between POM Variables and Maize Grain Yield 126 Discussion 134 Conclusion 136 6 MODELING CHANGES OF THE FARMING SYSTEMS WITH THE ADOPTION OF IMPROVED FALLOWS USING LINEAR PROGRAMMING (LP) 138 Introduction 138 Materials and Methods 140 Farming Systems in Southern Mali (Koutiala) 140 The Model 143 Vll Results 147 Validation 147 Scenario 1 148 Scenario 2 154 Scenario 3 154 Discussion 161 Conclusion 165 7 SYNTHESIS AND GENERAL CONCLUSIONS 167 Summary and Conclusion 167 Future Research 172 APPENDIX A. MATRIX OF THE LINEAR PROGRAMING MODEL AT YEAR I (SCENARIO 1) FOR TYPE B HOUSEHOLDS IN THE KOUTIALA REGION, MALI 174 APPENDIX B. MATRIX OF THE LINEAR PROGRAMING MODEL AT YEAR 2 (SCENARIO 1) FOR TYPE B HOUSEHOLDS IN THE KOUTIALA REGION, MALI 178 APPENDIX C. MATRIX OF THE LINEAR PROGRAMING MODEL AT YEAR 3 (SCENARIO 1) FOR TYPE B HOUSEHOLDS IN THE KOUTIALA REGION, MALI 182 APPENDIX D. MATRIX OF THE LINEAR PROGRAMING MODEL AT YEAR 4 (SCENARIO 1) FOR TYPE B HOUSEHOLDS IN THE KOUTIALA REGION, MALI 186 REFERENCES 190 BIOGRAPHICAL SKETCH 208 viii LIST OF TABLES Table Page 3.1 Growth rate and production of biomass of MPTs and the contribution of stylosanthes on improved fallows at N'Goukan in the Koutiala region, Mali 63 3.2 Stover production and grain yield of maize in 1998 at N'Goukan in the Koutiala region, Mali 65 3.3 Maize grain yield from contrasts comparing different species and natural grass fallow in 1998 at N'Goukan in the Koutiala region, Mali ... 65 3.4 Soil fertility parameters at N'Goukan, Mali, in June 1998 before planting 66 3.5 Changes in soil fertility parameters at N'Goukan, Mali: difference between June 1998 (before planting) a1n9d97October 1998 (at harvest) values 67 4.1 Dry matter production and grain yield of maize in 1997 and 1998 at the N'Tarla research station, Mali 86 4.2 N uptake of maize under different treatments at the N'Tarla research station experiment, Mali, 88 4.3 N uptake of maize under different treatments at the N'Tarla research station experiment, Mali, 1998 (residuals) 89 4.4 N uptake of maize under different treatments at the N'Tarla research station experiment, Mali, 1998 (biomass applied) 90 4.5 Soil fertility parameters at N'Tarla, in June 1997 at the beginning of the experiment 95 IX 4.6 Soil fertility parameters at N'Tarla, in October 1998 (residual effects) 98 4.7 Changes in soil chemical properties at the N'Tarla research station, Mali expressed as difference between initial (June 1997) and final (June 1998) stages 99 5.1 Particulate organic matter (POM) and nitrogen content of POM (POM^) in the topsoil (0 to 20 cm) at the N'Tarla research station (Mali) in 1997 and 1998 117 5.2 Average weight and N content of particulate organic matter (POM) in surface soil (0 to 20 cm) at different sampling dates at N'Tarla, Mali 120 5.3 Variation of soil particulate organic matter(POM) and POM nitrogen content as differences between averages at different sampling dates at N'Tarla, Mali 120 5.4 Different size fractions of soil particulate organic matter at the N'Tarla research station (Mali) in 1997 and 1998 for residual(R) and biomass applied (B)treatments 121 5.5 POM size class fractions at different sampling date at N'Tarla, Mali 122 5.6 Nitrogen content of different POM fraction size classes at the N'Tarla research station (Mali) in 1997 and 1998 for residual(R) and biomass applied (B)treatments 127 5.7 N content of POM size class fractions at different sampling date at N'Tarla, Mali 128 5.8 Correlation coefficient (r) between total soil particulate organic matter weight and N content with maize grain yield at N'Tarla, Mali 128 5.9 Correlation coefficient (r) between weight and N content of different soil particulate organic matter fraction size classes with maize grain yield at N'Tarla, Mali 128 X