Advances in Integrated Soil Fertility Management in sub-Saharan Africa:Challenges and Opportunities

Advances in Integrated Soil FertilityManagement in sub-Saharan Africa:Challenges and Opportunities

Edited by

Andre BationoTropical Soil Biology and Fertility (TSBF) Institute of the International Centre forTropical Agriculture (TSBF-CIAT), Nairobi, Kenya

Boaz WaswaTropical Soil Biology and Fertility (TSBF) Institute of the International Centre forTropical Agriculture (TSBF-CIAT), Nairobi, Kenya

Job KiharaTropical Soil Biology and Fertility (TSBF) Institute of the International Centre forTropical Agriculture (TSBF-CIAT), Nairobi, Kenya

Joseph KimetuTropical Soil Biology and Fertility (TSBF) Institute of the International Centre forTropical Agriculture (TSBF-CIAT), Nairobi, Kenya

A C.I.P. Catalogue record for this book is available from the Library of Congress.

ISBN-13 978-1-4020-5759-5 (HB)ISBN-13 978-1-4020-5760-1 (e-book)

Published by Springer,P.O. Box 17, 3300 AA Dordrecht, The Netherlands.

www.springer.com

Printed on acid-free paper

Top left- Response to P fertilizer by soybeansTop Middle- Millet cowpea rotation system: a promising technology in the drylands

Top Right- Integrating high value crops such as tomatoes in research portfolio of AfNetBottom left- Response of maize to fertilizer in farmer managed trials

Bottom Middle- Participatory research: A tool widely used by AfNet ScientistsBottom Right- Tapping the potential of the banana-based cropping systems

All Rights Reserved© 2007 Springer

No part of this work may be reproduced, stored in a retrieval system, or transmittedin any form or by any means, electronic, mechanical, photocopying, microfilming, recording

or otherwise, without written permission from the Publisher, with the exceptionof any material supplied specifically for the purpose of being entered

and executed on a computer system, for exclusive use by the purchaser of the work.

Contents

Preface xiii

Part I: Setting the scene/Introductory papers

1. A critical analysis of challenges and opportunities for soil fertility restoration in Sudano-SahelianWest Africa 1Schlecht E., Buerkert A., Tielkes E. and Bationo A.

2. Appropriate technologies to replenish soil fertility in southern Africa 29Mafongoya P.L., Bationo A., Kihara J. and Waswa B.S.

3. Available technologies to replenish soil fertility in Eastern Africa 45Okalebo J.R., Othieno C.O., Woomer P.L., Karanja N.K., Semoka J.R.M., Bekunda M.A., MugendiD.N., Muasya R.M., Bationo A. and Mukhwaua E.J.

4. Integrated Agricultural Research for Development: contributing to the Comprehensive AfricaAgricultural Development Programme (IAR4D in CAADP) 63Ralph von Kaufmann

5. From Thousands to Millions: Accelerating Agricultural Intensification and Economic Growth inSub-Saharan Africa 75Maatman A., Wopereis M.C.S., Debrah K.S. and Groot J.J.R.

6. Soil science, population growth and food production: some historical developments 85Alfred E. Hartemink

7. Development of an arable layer: A key concept for better management of infertile tropical savannasoils 99Amézquita E., Rao I.M., Hoyos P., Molina D., Chavez L.F. and Bernal J.H.

8. Food Security in Africa: The Challenges of Researchers in the 21st Century 105James B. Kung’u

9. Background, Current status and the African Context of the International Nitrogen Initiative 115Mateete Bekunda, James Galloway, Keith Syers and Mary Scholes

Part II: Integrated Nutrient Management (INM)

10. Within-Farm soil fertility gradients affect response of maize to fertiliser application in westernKenya 121Vanlauwe B., Tittonell, P. and Mukalama, J.

11. Characterisation of soil degradation under intensive rice production in Office du Niger zone of Mali 133Minamba Bagayoko, Mamadou K. N’Diaye, Mohamed Dicko and Bréhima Tangara

12. Soil fertility issues in the Blue Nile Valley, Ethiopia 139Per Ivar Våje

vi Contents

13. Overcoming phosphorus deficiency in soils of Eastern Africa: recent advances and challenges 149Generose Nziguheba

14. Relative contribution of trees and crops to soil carbon content in a parkland system in BurkinaFaso using variations in natural 13C abundance 161Bayala J., Balesdent J., Marol C., Zapata F., Teklehaimanot Z. and Ouedraogo S.J.

15. Mineral fertilizers, organic amendments and crop rotation managements for soil fertilitymaintenance in the Guinean zone of Burkina Faso (West Africa) 171B.V. Bado, A. Bationo, Lompo, F, M.P. Cescas and M.P Sedogo

16. Effect of planting technique and amendment type on pearl millet yield, nutrient uptake, andwater use on degraded land in Niger 179Fatondji D., C. Martius, C.L. Bielders, P.L.G. Vlek , A. Bationo and B. Gerard

17. Short-term effects of cover crops on stem borers and maize yield in the humid forest ofsouthern Cameroon 195Chabi-Olaye A., Nolte C., Schulthess F. and Borgemeister C.

18. Improving cereal productivity and farmers’ income using a strategic application of fertilizers inWest Africa 201Ramadjita Tabo, Andre Bationo, Bruno Gerard, Jupiter Ndjeunga, Daniel Marchal, BassirouAmadou, Mallam Garba Annou, Diakala Sogodogo, Jean-Baptiste Sibiry Taonda, OusmaneHassane, Maimouna K. Diallo and Saidou Koala

19. Advances in improving Agricultural Profitability and Overcoming Land Degradation in Savannaand Hillside Agroecosystems of Tropical America 209M. Ayarza, E. Barrios, I. Rao, E. Amézquita and M. Rondón

20. Integrating legumes to improve N cycling on smallholder farms in sub-humid Zimbabwe:Resource quality, biophysical and environmental limitations 231Regis Chikowo, Paul Mapfumo, Peter A Leffelaar and Keu E Giller

21. Meat and bone meal as nitrogen and phosphorus fertilizer to cereals and rye grass 245Alhaji S. Jeng, Trond Knapp Haraldsen, Arne Grønlund and Per Anker Pedersen

22. Screening Legume Green Manure for Climatic Adaptability and Farmer Acceptance in theSemi-Arid Agro-ecological Zone of Uganda 255J.B. Tumuhairwe, M.C. Rwakaikara-Silver, S. Muwanga and S. Natigo

23. Nutrient flows in smallholder production systems in the humid forest zone ofsouthern Cameroon 261J. Kanmegne, Smaling E.M.A., Brussaard L., Gansop-Kouomegne A. and A. Boukong

24. Management of improved fallows for soil fertility enhancement in the western highlandsof Cameroon 277Meppe F., Bilong P. and Nwaga D.

25. Integrated Soil Fertility Management: Use of NUTMON to Quantify Nutrient Flows in FarmingSystems in Central Kenya 283Kathuku A.N., S.K. Kimani, J.R. Okalebo, C.O. Othieno and B. Vanlauwe

26. Optimizing Soil Fertility Gradients in the Enset (Ensete ventricosum) Systems of the EthiopianHighlands: Trade-offs and Local Innovations 289Tilahun Amede and Endale Taboge

27. Consequences of Field Management and Soil Erosion on the Sustainability of Large ScaleCoffee Farming in Kiambu 299Okoth P.F., Ng’ang’a J.K. and Kimani P.K.

Contents vii

28. The Use of Erosion Proxies for the Spatial Assessment of Erosion in a Watershed and Modellingthe Erosion Risk in a GIS 311Okoth P.F., Oketch P.A. and Kimani P.K.

29. Bean Improvement for Low Soil Fertility Adaptation in Eastern and Central Africa 325Lunze L., Kimani P.M., Ngatoluwa R., Rabary B., Rachier G.O., Ugen M.M.,Vicky Ruganza andAwad elkarim E.E.

30. Combining Tithonia diversifolia and minjingu phosphate rock for improvement of P availabilityand maize grain yields on a chromic acrisol in Morogoro, Tanzania 333S.T. Ikerra, E. Semu and J.P. Mrema

31. Improving food production using ‘best bet’ soil fertility technologies in the Central highlandsof Kenya 345Daniel Mugendi, Monicah Mucheru-Muna, Jayne Mugwe, James Kung’u andAndre Bationo

32. Effects of organic and mineral sources of nutrients on maize yields in three districts ofcentral Kenya 353Kimani S.K., Esilaba A.O., Odera M.M., Kimenye L., Vanlauwe B. and Bationo A.

33. Effect of Combining Organic Leafy Biomass and Inorganic fertilizer on Tomato Yields andnematodes control in Arenosols in Kinshasa Area 359Mafuka M. M., Nsombo M., Nkasa C. Ibwenzi K. and Taba K.

34. Nutrient Dynamics On Smallholder Farms In Teghane, Northern Highlands Of Ethiopia 365Assefa Abegaz, Herman van Keulen, Mitiku Haile and Simon J. Oosting

35. Nitrogen-15 Recovery in Cropped Soil Cores Fertilized With Potassium Nitrate and CloverResidues 379Anne W. Muriuki, Larry D. King and Richard J. Volk

36. Evaluation of nitrogen fixation using 15N dilution methods and economy of a maize-tepary beanintercrop farming system in semi-arid SE-Kenya 389Chris A. Shisanya and Nkanata M. Gitonga

37. Evaluation of cowpea genotypes for variations in their contribution of N and P to subsequentmaize crop in three agro-ecological zones of West Africa 401R.C. Abaidoo, J.A. Okogun, G.O. Kolawole, J. Diels, P. Randall and N. Sanginga

38. Variability of cowpea breeding lines to low phosphorus tolerance and response to externalapplication of Phosphorus 413A.K. Saidou R.C. Abaidoo, B.B. Singh, E.N.O. Iwuafor and N. Sanginga

39. The potential benefits of Azolla, Velvet bean (Mucuna pruriens var. utilis) and N fertilizers in riceproduction under contrasting systems in eastern Uganda 423Crammer K. Kaizzi, Henry Ssali, Angella Nansamba and Paul, L.G. Vlek

40. Tied-ridging and integrated nutrient management options for sustainable crop production insemi-arid eastern Kenya 435Miriti J.M., Esilaba A.O., Bationo A., Cheruiyot H., Kihumba J. and Thuranira E.G.

41. Economic evaluation of local inputs in Meru South District, Kenya 443Monicah Mucheru-Muna, Daniel Mugendi, Jayne Mugwe and James Kung’u

42. Effect of Rock Phosphate, Lime and Green Manure on Growth and Yield of Maize in a NonProductive Niche of a Rhodic Ferralsol in Farmer’s Fields 449Tabu I.M., Bationo A., Obura R.K. and Khaemba J.M.

viii Contents

43. Changes in Soil Organic Matter as Influenced by Organic Residue Management Regimes inSelected Experiments in Kenya 457Waswa B.S., Mugendi D.N., Vanlauwe B. and Kung’u J.

44. Effects of manure application on crop yield and soil chemical properties in a long-term field trialin semi-arid Kenya 471Kihanda F.M., Warren G.P. and Micheni A.N.

45. Nutrient Recovery from Solid Waste and Linkage to Urban and Peri-Urban Agriculture inNairobi, Kenya 487Mary Njenga, Stephen Kimani, Dannie Romney and Nancy Karanja,

46. Optimising crop productivity in legume-cereal rotations through nitrogen and phosphorusmanagement in western Kenya 493Kihara J., Kimetu J.M., Vanlauwe B., Bationo A., Waswa B. and Mukalama J.

47. Evaluation of the Potential of Using Nitrogen Fixing Legumes in Smallholder Farms ofMeru South District, Kenya 503Jayne Mugwe, Daniel Mugendi, David Odee and John Otieno

48. Improved cassava varieties increase the risk of soil nutrient mining: an ex-ante analysis forwestern Kenya and Uganda 511Fermont A.M., Obiero H.M., van Asten P.J.A., Baguma Y. and Okwuosa E.

49. Partial balance of nitrogen in a maize cropping system in humic nitisol of Central Kenya 521J.M. Kimetu, D.N. Mugendi, A. Bationo, C.A. Palm, P.K. Mutuo, J. Kihara, S. Nandwa andK. Giller

50. Integrated Soil Fertility Management Technologies: A Counteract to Existing Milestone inObtaining Achievable Economical Crop Yields in Cultivated Lands of Poor SmallholderFarmers in Malawi 531Moses Wazingwa Munthali

51. Fertilizer nitrogen recovery as affected by soil organic matter status in two sites in Kenya 537Karunditu M.W., Mugendi D.N., Kung’u J. and Vanlauwe B.

52. Performance evaluation of various agroforestry species used in short duration improved fallowsto enhance soil fertility and sorghum yields in Mali 547Bocary Kaya, Amadou Niang, Ramadjita Tabo and André Bationo

53. Nutrient balances for different farm types in Southern Mali 557S. Kanté, E.M.A. Smaling and H. van Keulen

54. Increasing the Productivity and Sustainability of Millet Based Cropping Systems in the SahelianZones of West Africa 567Traoré Samba, Birama S. Coulibaly, Amadou Koné, Minamba Bagayoko and Zoumana Kouyaté

55. Comparative short-term effects of different quality organic resources on maize productivity undertwo different environments in Zimbabwe 575Florence Mtambanengwe, Paul Mapfumo and Bernard Vanlauwe

56. Improving soil fertility through the use of organic and inorganic plant nutrient and crop rotationin Niger 589Adamou A., Bationo A., Tabo R. and Koala S.

57. Biological system for improving the availability of Tilemsi phosphate rock for wheat(Triticum aestivum L) cultivated in Mali 599A.H. Babana and H. Antoun

Contents ix

58. Managing manure heaps with agro-organic wastes and cover to reduce nitrogen losses duringstorage on smallholder farms 611Gichangi E.M., Karanja N.K. and Wood C.W.

59. Soil characteristics and the performance of sorghum (Sorghum bicolor (L) Moench) on tin minespoils of the Jos Plateau, Nigeria 619A.A. Yusuf, A.F. Eben-Johnson, V.O. Chude and I.Y. Amapu

60. The development of a prototype land information system for the northern Guinea savanna ofNigeria as a basis for agro-technology transfer 629Tabi F.O., Ogunkunle A.O. and Diels J.

61. What role can planted fallows play in the humid and sub-humid zone of West andCentral Africa? 647Stefau Hauser, Christian Nolte and Robert J. Carsky

62. Balanced Nutrient Management System Technologies In The Northern Guinea Savanna OfNigeria: Validation And Perspective 669Kolawole G.O., Diels J., Manyong V.M., Ugbabe O., Wallays K., Dercon G., Iwuafor E.N.O.,Falaki A.M., Merckx R., Deckers J., Tollens E., Vanlauwe B. and Sanginga N.

63. Soil Erosion and Soil Inorganic-N Depletion as Influenced by Live hedges in Arable Steep-landsof the Central Highlands of Kenya 679James K. Mutegi, Daniel N. Mugendi, Louis V. Verchot and James B. Kungu

64. Relationships between rhizobial diversity and host legume nodulation and nitrogen fixation intropical ecosystems 691Abdullahi Bala and Ken E. Giller

65. Limestone, Minjingu Phosphate Rock and Green Manure Application on Improvement of AcidSoils in Rwanda 703Nabahungu N.L, Semoka J.M.R and Zaongo C.

66. Evaluating performance and yield stability of some groundnut (Arachis hypogaea L.) varietiesunder irrigation in three agroecological zones of the Senegal River Valley 713Madiama CISSE and Souleymane DIALLO

67. Assessment of the contribution of tied ridges and farmyard manure application to sorghumproduction in semi-arid areas of Tanzania 723Kabanza A.K. and Rwehumbiza F.B.R.

68. Evaluation of Gliricidia sepium, Casuarina junghuhniana and Faidherbia albida tree species forimprovement of crop production and fuelwood supply in Muheza districts, Tanzania 731Joel L. Meliyo, Marandu A.E.T. and Munuo E.

69. Mineral N distribution in the soil profile of a maize field amended with cattle manure and mineralN under humid sub-tropical conditions 737Nyamangara J.

70. Intensity cultivation induced effects on soil organic carbon dynamic in the western cotton area ofBurkina Faso 749Badiori Ouattara, Korodjouma Ouattara, Georges Serpantié, Abdoulaye Mando, Michel P.Sédogo and André Bationo

71. Assessment of improved soil fertility and water harvesting technologies through community basedon-farm trials in the ASALs of Kenya 759Gichangi E.M., Njiru E.N., Itabari J.K., Wambua J.M., Maina J.N. and Karuku A.

x Contents

72. Profitability of agro-forestry based soil fertility management technologies: the case of small holderfood production in Western Kenya 767J.M. Maithya, L.N. Kimenye, F.I. Mugivane and J.J. Ramisch

73. Integrated natural resources management a strategy for food security and poverty alleviation inKwalei village, Lushoto district, Tanzania 781Joel L. Meliyo, K.F.G. Masuki and J.G. Mowo

74. Effects of total inorganic nitrogen and phosphorus availability on Maize Yields in the First postTephrosia vogelii fallow 787Mkangwa C.Z., Maliondo S.M.S. and Semoka J.M.R.

75. Environmental hazards in African agriculture: factors influencing application of agrochemicals inNakuru district, Kenya 795Lagat J.K., Wangia S.M., Njehia B.K. and Ithinji G.K.

76. Assessment of indigenous soil and water conservation technology for smallholder farms insemi-arid areas in Africa and close spaced trash lines effect on erosion and crop yield 805Isaiah I.C. Wakindiki, B.O. Mochoge and Meni Ben-Hur

Part III: Belowground biodiversity

77. Prosopis africana (Guill., Perrot et Rich.) Taub and Entada africana (Guill. et Perrot.) leaf litterdecomposition and impact of biomass transfer on millet (Pennisetum glaucum (L.) R. Br.)growth and development on station in Niger 815Larwanou Mahamane, Harouna Niandou Abdel-Aziz, Abasse Tougiani and Niang Amadou

78. Soil microbial biomass carbon and nitrogen as influenced by organic and inorganic inputs atKabete, Kenya 827Baaru M.W., Mugendi D.N., Bationo A., Verchot L. and Waceke W.

79. Evaluating effect of mixtures of organic resources on nutrient release patterns and uptake by maize 833F.O. Ayuke, N.K. Karanja and S.W. Bunyasi

80. Mycorrhizal associations as indicators of forest quality after land use practices 845Onguene N.A.

81. Biomass production, N and P uptake of Mucuna after Bradyrhizobia and Arbuscular mycorrhizalfungi inoculation, and P-application on acid soil of Southern Cameroon 855Jemo M., Nolte C. and Nwaga D.

82. Evaluating the effect of Bacillus and Rhizobium. bi-inoculant on nodulation and nematode controlin Phaseolus vulgaris L. 865Karanja N.K., Mutua G.K. and Kimenju J.W.

Part IV: Participatory Approaches and Scaling up/out

83. Integrated Soil Fertility Management Technologies: review for scaling up 873Michael Misiko and Joshua Ramisch

84. Costs and Returns of Soil Fertility Management Options in Western Kenya 881Paul L. Woomer

85. Modeling farmers’ decisions on integrated soil nutrient management in sub-Saharan Africa:A multinomial Logit analysis in Cameroon 891Guy Blaise Nkamleu

Contents xi

86. Opportunities for and constraints to adoption of improved fallows: ICRAF’s experience in thehumid tropics of Cameroon 905Degrande A., Asaah E., Tchoundjeu Z., Kanmegne J., Duguma B. and Franzel S.

87. The Effect of Socio-Economic Factors on a Farmer’s Decision to Adopt Farm Soil ConservationMeasures. An Application of Multivariate Logistic Analysis in Butere/Mumias District, Kenya 915Victoria E. Anjichi and L.W. Mauyo

88. Farmer’s perception of planted calliandra tree fallows for shortening fallow cycles in southernCameroon 921Christian Nolte, Bernard Ondo Zo’o and Jean Paul Dondjang

89. Policies, Institutions and Market Development to Accelerate Technological Change in theSemiarid Zones of Sub-Saharan Africa 933Mark Winslow, Barry Shapiro and John Sanders

90. Factors Influencing Choice and Adoption of Integrated Soil Fertility Management Technologies inCentral Kenya Highlands 941Mureithi B.M., Kimani S.K., Odera M.M., Mwangi E.M. and Gachanja E.

91. Social capital and adoption of soil fertility management technologies in Tororo district, Uganda 947Lule Ali, N. M. Mangheni, P. C. Sanginga, R.J. Delve, F. Mastiko and R. Miiro

92. Adoption of Leguminous Trees/Shrubs, Compost and Farmyard Manure (FYM) As Alternatives toImproving Soil Fertility in Trans Nzoia District-Kenya 955Nekesa. A.O., Okalebo, J.R. and Kimetto, S.

93. Participatory Diagnosis in the Eastern Drylands of Kenya: Are Farmers aware of their SoilFertility Status? 961Kimiti J.M., Esilaba A.O., Vanlauwe B. and Bationo A.

94. On-Farm Evaluation and Scaling-up of Soil Fertility Management Technologies inWestern Kenya 969Odendo, M., Ojiem, J., Bationo, A. and Mudeheri, M.

95. The Resources-to-Consumption System: A Framework for Linking Soil Fertility ManagementInnovations to Market Opportunities 979Sanginga Pascal C.; Susan Kaaria, Robert Muzira, Robert Delve, Bernard Vanlauwe, Jonas andNteranya Sanginga

96. Scaling Up Options on Integrated Soil Fertility management in Western Kenya: The Case ofCOSOFAP: Challenges and Opportunities 993Qureish Noordin, John Mukalama, Daniel Rotich, Electine Wabwile and John Lynam

97. Socio-Economics of Soil Conservation in Kericho District, Kenya 1001Kipsat M.J.

98. Market Integration and Conduct Analysis: An Application to Cattle Markets in Uasin GishuDistrict, Kenya 1013Serem A.K, Maritim H.K, Kipsat M., Kisinyo P. and Nyangweso P.

99. Factors determining integrated soil fertility management in central Kenya highlands: ParticipatoryLearning and Action (PLAR) model analysis 1019Odera M.M., Kimani S.K., Esilaba A.O., Kaiyare J.M., Mwangi E. and Gachanja E.

100. Spatial Pricing Efficiency and Regional Market Integration of Cross-Border Bean (PhaseolusVulgaris L.) Marketing in East Africa: The Case of Western Kenya and Eastern Uganda 1027L.W. Mauyo, J.R. Okalebo, R.A. Kirkby, R. Buruchara, M. Ugen, and H.K. Maritim

xii Contents

101. Assessment of farmers’ perceptions of soil quality indicators within smallholder farms in thecentral highlands of Kenya 1035Mairura F.S., Mugendi D.N., Mwanje J.I., Ramisch J.J. and Mbugua P.K.

102. Initiating Rural Farmers to Participatory Research: Case of Soil Fertilization in Bushumba,East of DR. Congo 1047Paulin Njingulula Mumbeya and Ngongo Mulangwa

103. Farmers’ participation in soil fertility management research process: Dilemma in rehabilitatingdegraded hilltops in Kabale, Uganda 1051Muzira R.N., Kabale farmers’ groups, Pali P., Sanginga P. and Delve R.

104. Integrated Soil Fertility Management and Poverty Traps in Western Kenya 1061J.K. Ndufa, G. Cadisch, C. Poulton, Q. Noordin and B. Vanlauwe

105. Moving methodologies to enhance agricultural productivity of rice-based lowland systemsin sub-Saharan Africa 1077M.C.S. Wopereis and T. Defoer

Preface

Lack of access to food and its availability is of central concern in Africa and a fundamental challenge for humanwelfare and economic growth. Low agricultural production results in low incomes, poor nutrition, vulnerabilityto risks and lack of empowerment. The New Partnership for Africa’s Development (NEPAD) targets an averageannual increase of 6% in agricultural productivity to ensure food security and sustained national economies. Landdegradation and soil fertility or nutrient depletion are considered as the major threats to food security and naturalresource conservation in SubSaharan Africa (SSA).

What is needed is to break the cycle between poverty and land degradation in Africa by employing strategies thatempower farmers economically and promote sustainable agricultural intensification using efficient, effective andaffordable agricultural technologies. Such affordable management systems should be accessible to the poor, small-scale producers and the approach should be holistic and dynamic in order to foster both technical and institutionalchange. Farmers’ and local entrepreneurs’ capacities to invest in sustainable land management need to be enhancedby moving farmers in a market-oriented agriculture.

Mineral fertilizers are the most efficient way to reverse soil nutrient depletion (Africa losses $ 4 billion per yeardue to soil nutrient mining), and thus improved livelihoods. Whereas in the developed world, overuse of fertilizer andmanure has damaged the environment, in SSA the low and insufficient use of fertilizer and other organic amendmentshas led to overexploitation of soil nutrient resources (nutrient mining) causing environmental degradation.

The Tropical Soil Biology and Fertility (TSBF) Institute of CIAT is a research programme whose main aim is tocontribute to human welfare and environmental conservation in the tropics by developing adoptable and suitable soilmanagement practices that integrate the biological, chemical and socioeconomic processes that regulate soil fertilityand optimize the use of organic and inorganic resources available to the land users. The African Network for SoilBiology and Fertility (AfNet), being a network of scientists in Africa, is the single most important implementingagency of TSBF in Africa. AfNet’s main goal is to strengthen and sustain stakeholder capacity to generate, share andapply soil fertility management knowledge and skills to contribute to the welfare of farming communities. AfNetthus facilitates and promotes collaboration in research and development among scientist in Africa for the purpose ofdeveloping innovative and practicable resource management practices for sustainable food production in the Africancontinent. The AfNet members share TSBF goals and approaches. TSBF conducts research in tropical countries,but always in collaboration with national scientists.

AfNet conducts its research agenda through the new paradigm of Integrated Soil Fertility Management (ISFM),a holistic approach to soil fertility research that embraces the full range of driving factors and consequences ofsoil degradation – biological, chemical, physical, social, economic, and political. Within the more than 80 sites ofnetwork trials in different agro-ecological zones distributed in the East, South, Central and West Africa regions,increasingly emphasis is placed on both basic and adaptive research. AfNet has increased its efforts in scalingup results of best bet soil fertility management technologies to more farmers and communities employing a widerange of dissemination tools. The network is also very much involved in training, capacity building and informationdissemination activities.

With a total membership of 10 researchers in 1989, AfNet has a total membership of over 350 persons in 2005. Itis an African-wide network with 180 members from East and Central Africa, 80 from Southern Africa and 90 fromWest Africa. AfNet membership is drawn from National Agricultural Research and Extension Services (NARES),international agriculture research centers, universities and individuals from various disciplines mainly soil science,social science, agronomy and technology exchange.AfNet is managed by a scientific committee ofAfrican Scientist.

xiv Preface

In an effort to achieve its objective of collaborative research, AfNet organizes biannual meetings where net-work members are offered the opportunity to participate, share, exchange and publish results emanating from theirresearch activities in soil biology and fertility. In May 2004, AfNet organised a symposium in Yaoundé, Cameroonthat brought together over 150 participants.

The objectives of the symposium were:

1. To review recent research achievements on integrated soil fertility management (ISFM) and ecosystem services2. To develop strategies on scaling-up soil fertility enhancing technologies3. To increase stakeholder awareness of new initiatives in natural resource management including integrated

agricultural research for development (IAR4D)

The symposium was organized under three major themes, namely:

1. State of the art on integrated soil fertility management and ecosystem services2. Opportunities and challenges for scaling up/out integrated soil fertility management innovations3. Increasing stakeholder awareness of new initiatives in natural resource management and developing strategies

for implementation

This book presents papers of the symposium and is divided into four major parts: Part I: Setting the scene: Thepapers under this section give a general overview of soil fertility management issues in Africa. Part II: IntegratedNutrient Management (INM) presents information on the state of land degradation in Africa, INM approachesand principles, recent achievements on INM technologies and management of carbon and nutrient cycles. Part III:Belowground biodiversity presents information on managing soils for enhanced ecosystem services and managingsoil genetic resources for enhanced biodiversity. Part IV: Participatory approaches and scaling up/out, presents infor-mation on success stories on development and adoption of sustainable land management, constraints to adoptionof ISFM technologies, socio-economics and policy, key concepts of farmer participatory approaches and scalingup/out, opportunities and challenges of participatory market analysis for enabling rural innovations, IntegratingISFM in IAR4D to include market access, enabling policies and their interactions and gender analysis.

The book presents experiences learnt by the various researchers in their effort to achieve sustainable food pro-duction and natural resource utilization. It is hoped that the wealth of information the book contains will contributepositively to the design and implementation of future research and development programmes that will lead tosustainable food production and poverty eradication in Africa.

We thank the donor agencies – the Canadian International Development Agency (CIDA), The Rockefeller Foun-dation, The Ford Foundation, The Syngenta Foundation, The Forum for Agricultural Research in Africa (FARA),the International Foundation for Science (IFS) and the Technical Centre for Agricultural and Rural Cooperation(CTA) – for their intellectual, moral and financial support towards this conference. We also thank the ICRAF officeat Yaoundé, Cameroon for helping in the organizing of this symposium.

Joachim VossDirector General of CIAT

A critical analysis of challenges and opportunities for soil fertility restoration in

Sudano-Sahelian West Africa

E. Schlecht1, A. Buerkert2,*, E. Tielkes3 and A. Bationo41Institute for Animal Production in the Tropics and Subtropics, University of Hohenheim, Stuttgart, Germany;2Institute of Crop Science, University of Kassel, Witzenhausen, Germany; 3Centre for Agriculture in theTropics and Subtropics, University of Hohenheim, Stuttgart, Germany; 4Tropical Soil Biology and Fertility(TSBF) Institute of CIAT, c/o ICRAF, Nairobi, Kenya; *Author for correspondence (e-mail: buerkert@uni-kassel.de)

Key words: Fertilizers, Legumes, Modelling, Organic amendments, Soil fertility management, West Africa

Abstract

Since the 1970s, research throughout West Africa showed that low soil organic matter and limited avail-ability of plant nutrients, in particular phosphorus and nitrogen, are major bottlenecks to agriculturalproductivity, which is further hampered by substantial topsoil losses through wind and water erosion. A fewwidely recognized publications pointing to massive nutrient mining of the existing crop–livestock productionsystems triggered numerous studies on a wide array of management strategies and policies suited to improvesoil fertility. Throughout Sudano-Sahelian West Africa, the application of crop residue mulch, animalmanure, rockphosphates and soluble mineral fertilizers have been shown to enhance crop yields, wherebyyield increases varied with the agro-ecological setting and the rates of amendments applied. In more humidareas of Western Africa, the intercropping of cereals with herbaceous or ligneous leguminous species, theinstallation of fodder banks for increased livestock and manure production, and composting of organicmaterial also proved beneficial to crop production. However, there is evidence that the low adoption ofimproved management strategies and the lack of long-term investments in soil fertility can be ascribed to lowproduct prices for agricultural commodities, immediate cash needs, risk aversion and labour shortage ofsmall-scale farmers across the region. The wealth of knowledge gathered during several decades of on-stationand on-farm experimentation calls for an integration of these data into a database to serve as input variablesfor models geared towards ex-ante assessment of the suitability of technologies and policies at the scale offarms, communities and regions. Several modelling approaches exist that can be exploited in this sense. Yet,they have to be improved in their ability to account for agro-ecological and socio-economic differences atvarious geographical scales and for residual effects of management options, thereby allowing scenarioanalysis and guiding further fundamental and participatory research, extension and political counselling.

Soil fertility – the perpetual issue

Owing greatly to the two major Sahelian droughtsin the early 1970s and 1980s, the poor productivity

of agropastoral systems in Sudano-Sahelian WestAfrica (SSWA) has raised worldwide concern andsubsequently stimulated numerous research anddevelopment projects dealing with issues of soil

This article has been previously published in the journal “Nutrient Cycling in Agroecosystems” Volume 76 Issues 2–3.

A. Bationo (eds.), Advances in Integrated Soil Fertility Management in Sub-Saharan Africa: Challenges and Opportunities, 1–28.© 2007 Springer.

fertility, land degradation and desertification(Schlecht and Hiernaux 2004). For the WestAfrican Sahel in particular, Dutch scientists dem-onstrated that above an average annual precipi-tation of 250 mm, low soil organic matter (SOM)and limited availability of plant nutrients, in par-ticular phosphorus (P) and nitrogen (N), are majorbottlenecks to plant production (Penning de Vriesand Djiteye 1982; Breman and De Wit 1983).

Even if one has to be careful to generalize acrossSSWA given the widespread occurrence of ero-sion-related micro-habitats (such as depressionswith clay and nutrient accumulations), there isevidence of a clear eco-regional gradient in thephysico-chemical factors that limit plant growth.While at an average annual rainfall <300 mm andsoil pH values between 5.5–6.5 water availability ismost limiting, from 300 to 600 mm of rainfall andsoil pH from 4.1 to 4.5, low P and N availabilityare primarily hampering biomass production. Fi-nally, at <600 mm annual precipitation and soilpH>4.5, N, sulphur (S) and potassium (K) defi-ciencies prevail. Across the same rainfall gradient,the soils’ texture successively increases in claycontent from below 5% to more than 15% (Bue-rkert et al. 2000).

Another issue of major concern throughout theSahelian and Sudanian zones1 were the topsoil lossand redistribution processes caused by wind andwater erosion and their effects on the soils’ nutri-ent status and productivity. Both topics receivednew momentum in the early 1990s with a series ofwidely discussed nutrient budget assessmentspointing to massive nutrient mining of the existingcrop–livestock production systems (Stoorvogeland Smaling 1990; Van der Pol 1992; Krogh 1997).The controversial discussion about the validity ofthese budgets at different scales triggered a newseries of studies on nutrient and natural resourcemanagement strategies and policies which aimed atimproving soil fertility and consequently crop andlivestock performance in SSWA (Bationo et al.1998). Erosion control by mulch application orridging, a better integration of livestock and cropproduction and targeted use of organic amend-ments were all seen as prerequisites to maintain

soil fertility (McIntire et al. 1992; Powell andUnger 1998). Despite controversy about the nec-essary technical and political approaches, therewas widespread consensus that any increase in theregion’s agricultural production level would re-quire substantial inputs of mineral fertilizers (VanKeulen and Breman 1990; Buerkert and Hiernaux1998; Breman et al. 2001).

Concurrently with these more fundamentalstudies, a considerable number of technologieswere generated to improve the productivity ofAfrican soils, but farmers have not or onlypartly implemented many of these (Bationo et al.1998; Haigis 2004). Increasingly, social scientists,economists and agronomists started to acknowl-edge the efficiency of farmers’ low external inputstrategies to maintain the productivity of selectedparts of their fields (Brouwer et al. 1993; Lamersand Feil 1995; Lamers et al. 1995; Sterkand Haigis 1998), and provided contrasting evi-dence to previous predictions of imminentdoomsday-scenarios (Scoones and Toulmin 1998;Mazzucato and Niemeijer 2001; De Ridder et al.2004).

The existence of such gaps between scientificfindings and farmers’ reality, however, onlypartly explains the low adoption rates observedfor recommendations to enhance the soil pro-ductivity of farmers’ fields. Aiming at conclu-sions about how to better match our presentknowledge with farmers’ short- and long-termneeds of increased productivity, the achieve-ments and shortcomings of relevant studies onsoil fertility maintenance in the crop–livestocksystems of SSWA were critically reviewed.Hereby, the analysis concentrated on the com-parability of results of similar studies acrosslocations and climatic gradients, and the quan-tification of possible residual effects of technol-ogies on soil fertility and productivity. Particularemphasis was put on the understanding of thephysical and social buffering capacity of the landuse systems under study. It was also evaluatedwhether the studies accounted for the require-ment and availability of amendments, labourand capital needed for specific technologies atthe farmers’ level. In addition, the usefulness ofexisting bio-economic and land use models forex-ante analysis of the biophysical and socio-economic benefits of technologies from the plotto the regional scale was examined.

1In the present context, the term Sahelian zone is used synon-

ymously for the semi-arid zone (average rainfall 250–

600 mm a�1) and the term Sudanian zone for the sub-humid

zone (600–900 mm a�1).

2 A critical analysis for soil fertility restoration

Restoring soil fertility: the multitude of technical

solutions

It is a commonly accepted paradigm that above anaverage annual rainfall of 250 mm, the low soilproductivity in rainfed agriculture of SSWA is dueto a combination of unfavourable factors. Thesecomprise a very poor chemical and physical statusof the highly weathered and predominantly acidsandy soils (low P and N reserves and plantavailability, low SOM, low pH, low water holdingcapacity, rapid crust formation), erratic rainfall,the effects of abrasive storms and the occurrence ofinsect pests and low prices for the produced com-modities. The latter hamper subsistence-orientedfarmers’ investments in measures to enhance soilfertility, thereby providing a widespread impres-sion of ‘soil mining’ (Van Keulen and Breman1990; Breman et al. 2001).

Technical approaches to overcome low soilproductivity include (i) mulching with crop resi-dues, (ii) hand-spreading of household wastes,animal manure and compost, (iii) corralling oflivestock on fields, (iv) intercropping or rotation ofcereals with legumes and (v) application of rock-phosphates and soluble N and P fertilizers. Here-by, the proportion of yield increase varies with theagro-ecological setting (soil types and rainfall)and the rates and frequencies of applying theseamendments.

In traditional land use systems, declining SOMand nutrient status of continuously cropped landwas regularly restored by observing several yearsof fallow. Fallowing improves the soil nutrientstatus through the capture of mineral-rich soilparticles (in particular of K and Ca) eroded bywind near the fallow or in distant source areas(Herrmann 1996; Sterk et al. 1996; Rajot 2001;Wezel and Haigis 2002). Mineralisation of fallowshoot and root biomass increases SOM, and fallowlegumes may help to raise soil N by symbioticN2-fixation. Despite these well-known effects offallowing on the soil fertility status, rapid popu-lation growth leading to shortage of suitablecropland prevents the continued widespread use ofthis practice throughout the region (Graef 1999;Wezel and Haigis 2002; Schlecht and Buerkert2004). Another technology with little practicalrelevance for large parts of the drier areas ofSahelian West Africa is that of planted wind-breaks. Although the latter are often assumed to

have positive effects on plant establishment, SOMand nutrient availability, experimental data on thecompetition for water and nutrients and theireffects on biomass production are scarce (Leihneret al. 1993; Brenner et al. 1995; Smith et al. 1997;Michels et al. 1998). Even though successful inreducing erosion (Michels et al. 1998), the instal-lation and effective management of such wind-breaks requires high capital and labour inputs thatcan only be provided by external subsidies such asdevelopment projects or government programs(Lamers et al. 1996). In addition, a more wide-spread use of windbreaks can create land tenureproblems (Neef 1999). Breman and Kessler (1997)have summarized that windbreaks seem helpful toimprove crop establishment on sandy soils (par-ticularly in the southern Sahelian zone) at specificlocations with a shallow ground water table, wherecompetition for moisture with annual crops isavoided. In the Sudanian zone, however, maxi-mum benefits for crop production are obtainedwhere the canopy cover is 15–20%, trees arehomogeneously distributed and have a high ratioof trunk height to crown diameter (achieved bypruning). The impact of agroforestry practicesseems particularly important on slopes and in theupper part of watersheds given their positive ef-fects on soil stabilisation, and at desert margins forthe protection of productive land from wind ero-sion (Breman and Kessler 1997).

Soil tillage has also been advocated to fostercrop growth by enhancing soil porosity and aer-ation, infiltration rate, water holding capacity andnutrient mineralisation from organic material.With yield increases of 22–103%, these benefitswere particularly evident in the Sudanian zone(Nicou and Charreau 1985). In the Sahelian zone,Klaij and Hoogmoed (1989) observed improvedestablishment, seedling survival and early growthof pearl millet (Pennisetum glaucum L. R. Br.)due to reduced wind erosion after ridging, andSubbarao et al. (2000) reported an 11-year aver-age grain yield increase of 83% with ridging forthe same crop. Across zones, benefits of soil till-age certainly depend on the soil type, but thelabour investment involved may only be profit-able with higher-valued crops such as cotton(Gossypium hirsutum L.) or paddy rice (Oryzassp.; Jansen 1993). At the farm level an importantconstraint to tillage on Sahelian soils is the in-creased draught power needed for ploughing the

3 E. Schlecht

bone-dry soils at the onset of the rainy season, atime when draught animals are often ill-fed andtherefore weak. Due to their locally limited rele-vance in the West African context, the afore-mentioned three technologies will not beconsidered further in this analysis.

Effects of organic amendments on soil quality andcrop yields

The clay fraction of most Sahelian and Sudaniansoils is characterized by low sorption mineralsdominated by kaolinite and therefore the SOMconcentration largely determines the total cation-exchange capacity, pH and aluminium toxicity ofthese soils (Bationo and Buerkert 2001). However,given the synchrony between high soil tempera-tures and rainfall as well as termite activitythroughout the Sudano-Sudanian zone, native oradded SOM is rapidly mineralised under landcultivation, even in the drier areas. On most soilsmedium- and long-term SOM therefore remainslow, even if elevated amounts of plant biomass arerecycled during fallow periods or with mulchingand manuring (Feller and Milleville 1977; DeRidder and Van Keulen 1990; Renard et al. 1993;Bacye et al. 1998).

Buerkert et al. (2000, 2002) have summarizedhow short-term effects of mulch application on soilphysical and soil chemical properties and finally oncrop growth depend in magnitude on the specificsoil type (such as its clay content) and the climaticconditions (intensity and length of the rainy sea-son). Mulching effects certainly also depend on thekind of organic material applied (branches ofwoody species, crop residues such as stover,household waste, compost and hand-spread man-ure) even if to our knowledge comparative data onthe effects of different mulch materials have notbeen published.

Mulching of crop residues and other plant material

Mulching, that is the superficial application ofplant material, is most commonly done with res-idues of cereal crops, such as millet, sorghum ormaize stover. For the Sahelian zone reportedphysical effects of annual application rates of2000 kg ha�1 (typically used in researchers’ exper-iments but rarely available on farmers’ fields)

comprise the following: reduced wind and watererosion effects on seedling emergence and earlygrowth (Michels et al. 1995a, b), breakage ofsurface crusts and a decrease in soil surface pen-etration resistance (Hoogmoed and Stroosnijder1984; Buerkert and Stern 1995) by a stimulationof termite activity, higher root length density(Hafner et al. 1993), increased formation of stablesoil aggregates enhancing soil porosity and waterinfiltration, higher water holding capacity and adecreased maximum soil surface temperature(Buerkert and Lamers 1999; Figure 1). Throughcarbon addition and/or a decelerated decomposi-tion of native SOM, mulching can also lead torelative increases in SOM levels and an increase inpH compared to unmulched control plots(Bationo and Buerkert 2001). On sandy soils,mulching has been reported to cause significantchanges in soil chemistry such as increases ofcation exchange capacity (CEC), P availabilitythrough decomplexation of P from Al and Fechelates through ligand exchange mediated byorganic acids (Kretzschmar et al. 1991). Also im-portant are pH and K increases through residuemineralisation or capturing of Harmattan trans-ported dust (Herrmann et al. 1993; Rebafka et al.1994; Buerkert and Lamers 1999; Sinaj et al. 2001).An experiment conducted by Buerkert and Lamers(1999) allowed to differentiate between the physicaland chemical components of mulch-induced effectson millet growth. Mulched application of milletstover at 2000 kg ha�1 decreased eolian losses oftopsoil to a similar extend as an equivalent soilcover of 10% with inert polyethylene tubes. Inthree subsequent years, total dry matter (TDM)increase of millet above the bare control was 35,108 and 283% when stover was applied but only 6,44 and 13% with plastic mulch. These differencesreflected the plant-nutritional components of resi-due mulch effects. Yield increases due to mulchingobtained in other studies throughout the region aresummarised in Table 1.

Wezel and Boecker (1999) measured that a singlemulching with twigs of Guiera senegalensisJ. F. Gmel. at 1000 and 2000 kg ha�1 can lead toaverage millet grain yield increases of 68 and 83%over 2 years. Given the widespread growth ofGuiera on farmers’ fields, these authors concludedthat regular applications of 1000 kg ha�1 of itsbranches at least on low-productivity parts of fieldscould substantially increase millet production.

4 A critical analysis for soil fertility restoration

Figure 1. Effects of different amounts of crop residue mulch on millet root length density (a; Hafner 1993; horizontal bars show one

LSD0.05), soil temperature and soil resistance (b and c; Buerkert and Lamers, 1999; vertical bars show standard errors of the mean) on

acid sandy soils of south-western Niger.

Table 1. Effects of different application frequencies and amounts of cereal crop residues on cereal grain yields in different agro-

ecological zones of Sudano-Sahelian West Africa as reported by different authors.

Amendment Amount (kg ha�1) Grain yield increase

above control

Application

frequency

Zone, soil type References

Millet crop residues 2000 Millet: 20–>100% Annually Semi-arid

(Burkina Faso),

Arenosol

Bationo et al.

(1992, 1993, 1995)

Millet crop residues 2000

(control = 500)

Millet: 1st year 6%,

2nd year 20%,

3rd year 32–40%

Annually Semi-arid (Niger),

Arenosol

Buerkert et al.

(2000, 2002)

Millet crop residues 4000 Millet: 250% Annually Semi-arid (Niger),

Arenosol

Bationo and

Mokwunye (1991)

Millet crop residues

+ fertilizer

4000 + 13 kg P ha�1

broadcast

Millet: 790% Annually Semi-arid (Niger),

Arenosol

Bationo and

Mokwunye (1991)

Millet crop residues 2000 Millet: 1st year 20%,

2nd year 108%,

3rd year 283%

Annually Semi-arid (Niger),

Arenosol

Buerkert and

Lamers (1999)

Millet crop residues All CR from

last harvest

Millet: 40–238% Annually Semi-arid (Niger),

Arenosol

Yamoah et al. (2002)

Millet crop residues

+ fertilizer

All CR from

last harvest

+ 13 kg P ha�1

+ 30 kg N ha�1

broadcast

Millet: 90–>400% Annually Semi-arid (Niger),

Arenosol

Yamoah et al. (2002)

Maize straw Not clarified Soybean: 21% Annually Semi-arid

(Senegal),

Arenosol

Dommergues and

Ganry (1986)

Maize: 44%

5 E. Schlecht

Several other authors have reported mulchingwith various grasses, leaves and small branchescollected from fallow vegetation, bushes and treesto be a common practise of farmers in differentregions of SSWA (Lamers and Feil 1995; Slinger-land and Masdewel 1996; Hien et al. 1998).However, it appears as if the reasons why thesepractises so far have attracted little researchattention are the heterogeneity and the scarcity ofthe available mulch material as well as the largeamount of labour needed for its collection andtransportation (Slingerland and Masdewel 1996).

Manuring, corralling and compost application

Due to limited access to mineral fertilizers, manurefrom livestock is an important source of SOM andnutrients for crop production throughout SSWA,even if its availability in sufficient quantity may bea major constraint at many places (Powell et al.1998). In the West African context, manure avail-ability, quality, storage and application, therefore,has received much scientific attention (Williams etal. 1995; Fernandez-Rivera et al. 1995; Harris 2002;Schlecht et al. 2004). In the predominantly exten-sive West African production systems, livestockgraze the vegetation of natural pastures and fallowsyear-round, whereas cereal stover and weeds onfields are grazed only after grain harvest until thestart of the new cropping season. Although varyingwith the grazing regime, close to 50% of the daily(24 h) amount of faeces and urine are excretedduring the night rest period and in the earlymorning when animals get up (Schlecht et al. 1995,1998). If the night resting place is at a fixed loca-tion, the dry manure and bedding material, feedleftovers and other organic compound waste or ashcan be easily collected and transported to the field.This so-called farmyard manure varies greatly inquantity and quality, depending, among others, onthe organic material included besides the faeces (DeJager et al. 1998; Harris 1998; Hoffmann et al.2001; Harris 2002). Early work by Bationo andMokwunye (1991) reported four-fold increases inmillet grain yields after the application of 20 t ha�1

of dry farmyard manure. A particular way of tar-geted application of farmyard manure is the zaıtechnique, which consists in digging small plantingholes and filling them with manure before the cropis sown into the hole. The advantages of this

technique consist in the simultaneous increase ofmanure derived SOM, mineralised nutrients andcollected rainfall in the immediate rooting envi-ronment of crops (Fatondji 2002).

However, if no bedding material is used at theanimals’ resting place, the nutrients from urine arepercolating into the soil and cannot be recovered(Powell et al. 1998). To concentrate manure on afield scheduled for cultivation, livestock can also becorralled or tethered there overnight. In this way,no labour is required for manure handling, andurine is applied to the soil along with the faeces,supplying supplementary N and instantaneouslyincreasing the pH of the wetted topsoil by up to4 units, whereby sheep urine appeared to be moreeffective than urine from cattle (Brouwer andPowell 1998; Ikpe et al. 1999). Measurements ofdiurnal and seasonal variations in urine pH ofcattle, sheep and goats grazing Sahelian pasturerevealed that at least in the evening, the urine ofsheep was mostly lower in pH than the urine ofcattle and goats (Table 2). Although monthly dif-ferences in species-specific urine pH were at maxi-mum 0.47 units, the variation between minimumand maximum diurnal values determined within aspecies reached up to 3.55 pH units. However, thereported increase in soil pH due to urine applica-tion is not primarily linked to urine pH as such butto the release of hydroxide anions during ureahydrolysis (Powell et al. 1998), which implies thatchanges in soil pH mainly depend on the ureacontent of the urine. Although N-containingby-products of the metabolism of proteins andnucleic acids are also excreted in urine, ureaaccounts for approximately two-thirds of urine N(Bristow et al. 1992). The N excretion via urinethereby depends on the dietary supply of solubleand non-soluble proteins and intake of salt andwater. Per kilogram live weight, daily urine-Nexcretion in cattle, goats and sheep fed with dry andgreen Sahelian roughages was determined at 50–420, 84–460 and 72–600 mg, respectively (Schlecht1995; Schlecht et al. 1998), from which the effi-ciency of urine to increase soil pH can be deduced.

For faeces, reported nutrient concentrationsper kilogram of dry matter vary between 9 and25 g N, 0.09 and 2.7 g P, and 2.5 and 37 g K(Williams et al. 1995). Outside of corrals, anaverage yearly rate of return to village cropland,pastures and fallows through faeces of 0.09–0.23 kg P ha�1 was estimated from grazing and

6 A critical analysis for soil fertility restoration

excretion behaviour of livestock (Schlecht et al.2004). Daily excretions of faecal N were in therange of 63–230 mg kg�1 live weight in cattle,160–210 mg kg�1 in sheep and 140–170 mg kg�1

in goats (Schlecht 1995; Schlecht et al. 1998),whereby seasonal conditions, the ingested diet(green or dry grass) and the presence or absenceof secondary plant compounds (especially tan-nins) have an equally important impact on vari-ations in faecal N-concentrations than thelivestock species (Powell et al. 1994, 1999; Somdaand Powell 1998; Schlecht et al. 1998).

The effects of different manuring and corrallingtreatments on crop yields in the region vary widely(Table 3). As for crop residue mulching, underly-ing causes for the positive effects of manureapplication are increased soil porosity and aggre-gate stability, increased water infiltration and wa-ter holding capacity, decreased eolian soil losses,increased SOM, pH, CEC and nutrient availability(Powell et al. 1996, 1998, 1999).

Similar to other organic amendments, theapplication of compost has been reported toincrease soil pH, CEC and P and N availability onpoorly buffered soils of SSWA. In the sub-humidzone of Burkina Faso, surface application ofcompost at 5 t ha�1 led to grain and strawyield increases of Sorghum (Sorghum bicolorL. Moench) of 46–69% and 16–20%, respectively,above the unamended control in the year ofapplication (Ouedraogo et al. 2001). Similarly,Dommergues and Ganry (1986) reported yieldincreases of 13 and 54% above the control insoybean (Glycine max L. Merr.) and maize (Zeamays L.) grain, respectively, when applying1.5–2 t ha�1 of compost. While being very effec-tive in increasing yields, compost preparationrequires the availability of substantial amounts ofunused organic material, labour inputs for theestablishment of a suitable pit, and regularwatering and turning. Therefore this technology islikely to be of major practical importance onlywhere farmers are better-off or receive externalsupport, where livestock are managed in zero-grazing systems and all manure is deposited at thehomestead, and in peri-urban or wetter areas withhigher biomass availability.

The dependency of manuring and compostingtechniques on the availability of significantamounts of biomass (as feed or compostablematter) is also reflected by the fact that theseT

able

2.Diurnalandseasonalvariationsin

thepH

ofurinevoided

bycattle,sheepandgoats

grazingSahelianpasture

(Schlechtunpublished

data).

Perioda

January

(mid-dry

season)

March

(late

dry

season)

June

(earlyrainyseason)

August

(mid-rainyseason)

October

(post-harvestseason)

Decem

ber

(earlydry

season)

Cattle

Goats

Sheep

Cattle

Goats

Sheep

Cattle

Goats

Sheep

Cattle

Goats

Sheep

Cattle

Goats

Sheep

Cattle

Goats

Sheep

Morning

Meanb

A8.22

8.32

8.18b

8.27a

8.25a

A8.42ab

A8.31a

8.37a

A8.36

A8.14b

A8.12

A8.13b

A8.22

8.32

8.18b

8.20a

A8.36ab

A8.35b

SD

0.14

0.39

0.15

0.07

0.10

0.16

0.12

0.15

0.11

0.12

0.26

0.18

0.14

0.39

0.15

0.14

0.12

0.06

N12

411

810

11

97

14

12

12

12

12

411

910

6

Noon

Mean

B7.89b

8.01b

7.35b

8.29a

8.13

A8.23a

B8.22

8.35aa

B8.17ab

B7.60b

B7.10b

B7.89b

8.01b

7.35b

8.10

B8.21

8.23a

SD

0.18

0.06

1.26

0.13

0.14

0.04

0.03

0.17

0.14

0.62

0.66

0.18

0.06

1.26

0.09

0.14

0.08

N8

22

611

27

517

10

88

22

76

7

Afternoon

Mean

B7.92ba

8.02ba

7.34bb

8.23aa

8.15aa

B7.57bb

8.26aa

8.25aa

C8.07ab

B7.61c

B7.92ba

B7.35bb

B7.92ba

8.02ba

7.34bb

8.15a

B8.21aa

B8.09ab

SD

0.13

0.18

0.76

0.12

0.16

0.62

0.04

0.15

0.14

0.29

0.23

0.85

0.13

0.18

0.76

0.13

0.13

0.19

N15

19

18

17

18

13

13

15

16

14

14

19

15

19

18

15

15

12

aPeriods:morning6:31–11:00h;noon11:01–14:30h;afternoon14:31–19:00h.

bPresentationofstatisticalresults:A,Bindicate

differencesin

pH

betweenperiodsoftheday,within

speciesandmonths;a,bindicate

differencesin

pH

betweenmonths,within

speciesandperiodsoftheday;a,bindicate

differencesin

pH

betweenspecies,within

monthsandperiodsoftheday.

Level

ofsignificance:P£0.05.

7 E. Schlecht

Table

3.Effects

ofdifferentapplicationfrequencies

andamounts

offarm

yard

manure,singlesuperphosphate

(SSP)andcompostoncerealgrain

yieldsin

differentagro-ecological

zones

ofSudano-SahelianWestAfricaasreported

bydifferentauthors.

Amendment

Amount

(kgha�1)

Grain

yield

increase

abovecontrol

Application

frequency

Zone,

soiltype

References

Farm

yard

manure

a5000

Millet:99%

Once

every3years

Sem

i-arid

(BurkinaFaso),Arenosol

BationoandMokwunye(1991)

Farm

yard

manure

a

+SSP

5000+

8.7

broadcast

Millet:202%

Once

every

3years/SSPyearly

Sem

i-arid

(BurkinaFaso),Arenosol

BationoandMokwunye(1991)

Farm

yard

manure

a

+rockphosphate

5000+

39.3

broadcast

Millet:163%

Once

every

3years/R

Pyearly

Sem

i-arid

(BurkinaFaso),Arenosol

BationoandMokwunye(1991)

Farm

yard

manure

a20,000

Millet:302%

Once

every3years

Sem

i-arid

(BurkinaFaso),Arenosol

BationoandMokwunye(1991)

Farm

yard

manure

a

+NPK

fertilizer

5000+

100

broadcast

Millet:150–700kgha�1

Yearly

Sem

i–arid

(BurkinaFaso),Arenosol

Dugue(1998)

Manure

corralled

2000

Millet:1st

year350kgha�1

2ndyear90kgha�1,

3rd

year190kgha�1,

4th

year145kgha�1

Once

every5years

Sem

i-arid(N

iger),

Arenosol

Schlechtet

al.(2004)

Manure

corralled

4000

Millet:1st

year500kgha�1

2ndyear177kgha�1,

3rd

year240kgha�1,

4th

year150kgha�1

Once

every5years

Sem

i-arid(N

iger),

Arenosol

Schlechtet

al.(2004)

Compostb

5000

Sorghum:46%

1-yearexperim

ent

Sem

i-arid

(BurkinaFaso),

FerricLixisol

Ouedraogoet

al.(2001)

Compost

1500–2000

Soybean:12%

Yearly

Sem

i-arid(Senegal),

Arenosol

Dommergues

andGanry

(1986)

Maize:

54%

Sorghum:45%

aManure

contained

0.41%

totalPand1.21%

totalN.

bCompostfrom

selected

household

refuses,anim

almanure,cropresidues

andashes;duringaseconddecomposingstep,grasses

wereadded

toincrease

theavailabilityofcarbonas

anenergysource;

averageC:N

ratioofthecompost

was12ataCorgcontentof160gkg�1andapH

of7.

8 A critical analysis for soil fertility restoration

methods of soil fertility restoration are appliedmore frequently in the humid zones of the WestAfrican Sahel: in a village in Western Nigerreceiving 360 mm a�1 of rain, corralling andfarmyard manure application together were prac-ticed on 25% of farmers’ fields (Schlecht andBuerkert 2004) compared to 34% in a village inBurkina Faso receiving 800 mm a�1 of rain (Mertzand Reenberg 1999). In any case, due to thescarcity or alternative uses of amendments as wellas on-farm labour, such practices often cover onlya few smaller fields of a household or parts of anindividual field (Schlecht and Buerkert 2004), andwill thus not allow to enhance soil productivity ata larger scale.

Association of cereals with legumes

The beneficial effects of legume cultivation for soilfertility are well known and can be ascribed toimproved soil physical properties, N2-fixation,enhanced P-availability through secretion ofenzymes or acids in the legume rhizosphere(Bagayoko et al. 2000b) to which phytosanitarycomponents may add.

For groundnut (Arachis hypogaea L.) and cow-pea (Vigna unguiculata L. Walp.) receivingapproximately 800 mm a�1 of rainfall in theSudanian zone of West Africa, N2-fixation wasestimated at 0.6 and 1.7 g N per plant, respec-tively, equivalent to 1.2 and 1.5% of total N in theharvested biomass (Harris 1998). Intercropping ofcowpea with millet may enhance millet grain yieldsby 30% above the control (Bationo et al. 1995).Rotations of legumes with cereals can lead to evenlarger yield increases of cereals on West Africansoils which are, however, strongly dependent uponchoice of the legume species and site conditions.Multi-year trials showed groundnut-inducedincreases in TDM yield of maize and sorghum by83% at 1300 mm a�1 average rainfall, whereas theuse of cowpea in the drier Sahelian zone did notlead to substantial yield differences in rotationcropping compared to continuous cropping ofcereals (Buerkert et al. 2002). In their evaluation oftraditional cropping systems in Africa, Dakoraand Keya (1997) concluded that legume–cerealrotations were by far more sustainable thanintercropping systems. Without providing detailson growing conditions or soil type they reported

a doubling of maize yields after groundnutor cowpea compared to continuous maize andmaize–legume intercropping. Process-orientedresearch revealed that legume-induced increases incereal yields were due to higher early season Navailability, enhanced infection of cereal rootswith arbuscular mycorrhiza, decreased nematodeinfestation, increased soil pH, improved P avail-ability through changes in soil chemistry andenhanced phosphatase release (Bagayoko et al.2000a–c; Alvey et al. 2001, 2003; Marschner et al.2004). Changes were also observed in the bacterialcommunity structure whose effects, however, arestill poorly understood.

Despite experimental evidence of large positivelegume effects on cereal growth in crop rotationsand the conclusion of Bationo and Ntare (2000)that a legume–millet rotation in combination with30 kg N ha�1 to millet appears to be a viableoption for millet production throughout semi-aridWest Africa, only very few farmers in sub-SaharanWest Africa currently practise legume rotations.The large majority of them grow either pure cerealstands or cereals with loosely interplantedlegumes, the latter being introduced as a relay cropfilling in for missing planting hills (Mertz andReenberg 1999; Schlecht and Buerkert 2004).Some of farmers’ apparent dislike of pure legumestands in rotations might come from the perceivedrisk of insect and disease attack of legumes whengrown on a comparatively large surface whoseprevention would require regular spraying ofinsecticides. Another cause may be the low stor-ability of legume seeds (particularly in cowpea)due to post-harvest attacks of seed borers which,under farmers’ conditions, require a quick sale ofthe product at low post-harvest prices. For sub-sistence farmers cereal-dominated intercroppingwith a sparse legume component may thus be atypical risk-averting strategy. In terms of totalsurface area cropped, Mertz and Reenberg (1999)reported deliberate intercropping of cereals andlegumes on 85% of the fields in a village ofBurkina Faso receiving 800 mm a�1 of rain,whereas sparse relay-type intercropping was ob-served on 78% of the investigated fields in aNigerian village receiving 360 mm a�1 of rain(Schlecht and Buerkert 2004).

Thomas and Sumberg (1995) have revieweddata from sub-Saharan Africa on the fertiliserrequirements, persistence, management, use and

9 E. Schlecht

nutritive value of forage legumes. Besides legumeintercropping and rotation, the cultivation oflegumes in fodder banks or ley systems can beeffective to enhance soil fertility along with live-stock nutrition. By modelling four different graz-ing schemes of a Stylosanthes hamata (L. Taub.)ley in the Sudano-Sahelian transition zone, Bosmaet al. (1999) predicted an annual increase in SOMby 500–1500 kg ha�1 above the values obtainedunder the current crop–livestock husbandry sys-tem, however, the time horizon of the modellingruns was not disclosed. To maintain high SOMlevels, their linear programming approach sug-gested an intensified use of cereal residues as fod-der and embedding material in livestock kraalscombined with P-fertilized ley systems, wherebythe latter would allow maintaining higher animaldensities throughout the dry season. However, asportrayed by Sumberg (2002), intensified foragelegume production has not been proven attractivefor small-scale farmers, despite clear evidence thattheir cultivation is possible above 800 mm a�1 ofrainfall in all major agro-ecological zones ofAfrica.

Agroforestry

Even if the methodology used to measure theiractual N2-fixation is not always clear, sub-Saharantree legumes, comprising also the widely spreadAcacia species, are reported to be an importantcomponent in the sustainability of open parkland,agro-forestry and alley cropping systems. Sum-marizing studies carried out across sub-SaharanAfrica, Dakora and Keya (1997) reported that Naddition to maize from pruning of tree twigs andleaves ranged from 200 kg N ha�1 for Calliandracalothyrsus Meissn on an N-fertilized Alfisolin Nigeria to 643 kg N ha�1 for Leucaenaleucocephala Lam. De Wit on an unknown soiltype in Kenya. Reported estimates of yearly N2-fixation ranged from 36 kg N ha�1 for Acaciaholosericea Cunn. ex G. Don to 581 kg N ha�1 forLeucaena leucocephala, whereby the range re-ported for individual tree species was sometimes aswide as the range for all species analysed in oneregion (Dakora and Keya 1997). The latter highfixation rates would certainly be uncommon forthe drier parts of sub-Saharan West Africa wherelow soil moisture during the recurrent drought

spells has been reported to reduce nodule func-tioning in symbiotic legumes through the collapseof lenticels (Pankhurst and Sprent 1975), de-creased nitrogenase activity, reduced respiratorycapacity of bacteroids and a decline in the leg-haemoglobin content of nodules (Guerin et al.1990). High temperatures, commonly foundthroughout this region, can further hamper N2-fixation (Dakora and Keya 1997). As far asperennial shrubs are concerned, much of the typ-ical open parkland system in the Sahelian zone ofWest Africa is characterised by irregular but rela-tively dense Guiera senegalensis stands. Wezel andBoecker (1998) and Gerard and Buerkert (1999)have well documented the yield enhancing effect ofthese shrubs on millet planted in their immediatesurroundings and shown that this effect is at leastpartly due to enhanced N and P availability fromleaf fall and dust accumulation.

In contrast, Breman and Kessler (1997) con-cluded that the integration of woody plant speciesin West African cropping systems does not neces-sarily provide benefits to farmers, because addedvalues from trees such as improved soil fertilitydue to efficient internal nutrient recycling and N2-fixation are lowest in unfavourable biophysicalenvironments where they are most needed. Oftenthere is a strong competition for light, nutrientsand water between woody plants and crops orpastures, or else intensive tree use and droughtsprevent an optimal canopy cover and effectivecontribution of woody species to maintain or raiseSOM levels.

There has been considerable debate about therole that hedgerow systems involving a perennialshrub or tree and an annual species plantedin-between may play in enhancing crop productionthroughout SSWA. General claims that annualcrop yields could be substantially increased afterthe mulched application of leaves and twigs cutfrom legume hedgerows were soon challenged byresearch results from the Guinean and Sudanianzone. It became evident that woody species in alleycropping systems can indeed contribute toincreased N2-fixation, nutrient recycling, reducedleaching and erosion losses and to a stimulation ofsoil faunal activity (Kang 1997; Tian et al. 2003),but it remains questionable whether they can alsobe a remedy against a long-term decline insoil fertility, thereby sustaining crop productionwithout the use of added mineral fertilizers. The

10 A critical analysis for soil fertility restoration

results of studies on root distribution and rootdensity across soil profiles (Akinnifesi et al. 1999;Lose et al. 2003) pointed to the strong competitionfor water and nutrients between annuals andperennials in many agro-forestry systems on thepredominantly nutrient-poor soils of SSWA.Although alley cropping with low-competitive treespecies such as Cajanus cajan L. Millsp. wasreported to increase maize yields up to 50% onhighly weathered unfertilised Ultisols, greatestmonetary benefits were obtained with the appli-cation of mineral fertilizers in simple cassava-maize intercropping systems where nutrient cycleswere reportedly almost closed (Akonde et al. 1996,1997). Comparable data from the Sahelian zone isscarce but results from the above-mentioned trialson ligneous windbreaks point to even strongerinter-species competition on acid-sandy Arenosolswith large depressions in crop yields near hedge-rows (Leihner et al. 1993; Michels et al. 1998).

Mineral fertilizers

There is a wealth of site-specific information aboutthe short-term yield enhancing effects of P, N andmolybdenum (Mo) fertilizers on cereals andlegumes grown on severely nutrient deficient soilsof SSWA (Pieri 1986; Bationo et al. 1986, 1990,1992, 1993; Friesen 1991; Hafner et al. 1992;Rebafka et al. 1993a, 1994; Buerkert and Hiernaux1998; Buerkert et al. 2000, 2001, 2002; Muehlig-Versen et al. 2003), some of which are summarizedin Table 4. Single superphosphate (SSP) appliedannually at 13 kg ha�1 effectively removed P defi-ciency on most soil types tested throughout SSWA.For groundnut, however, Rebafka et al. (1993b)provided evidence that its S-component can lead toreduced Mo uptake and thus growth reductioncompared to a P application as triplesuperphos-phate (TSP). On S-deficient soils above 1000 mma�1 rainfall, however, at the same rate of P appliedSSPmay bemore effective than TSP (Friesen 1991).

Phosphorus-induced yield increases in cerealshave been shown to substantially increase with Napplication. Despite site-specific variation in therelative importance of N and P, experimental evi-dence from Mali strongly suggests that the relativeimportance of N compared with P increases withrainfall from north to south across the Sudano-Sahelian zone and that at most sites P is more

limiting for crop growth than N (Poulain et al.1974). Most of this fertilizer-response research wasset up to examine the immediate effects of a singleor repeated application of soluble N and P.However, surprisingly little is known about theirresidual effects on plant growth over time, al-though such effects seem crucial to predict adop-tion of P-application technologies by small-scalefarmers. Residual effects of mineral P fertilizerswere studied only occasionally, but Gerard et al.(2001) reported a 14% increase in the dry matteryield of herbaceous fallow vegetation two yearsafter the last addition of SSP to millet plots on aluvic Arenosol in the Sahelian zone.

Rockphosphate (RP) from different regionalsources, RP compacted with soluble fertilizers andpartially acidulated phosphate rock (PAPR) havealso been tested at a number of sites throughoutWest Africa and shown to vary widely in theirefficiency relative to soluble P (Bationo et al. 1990;Bationo and Mokwunye 1991; Buerkert et al.2000, 2002). The variation of these effects likelydepends on the origin of the rock, site-specific soilproperties and rainfall, but these parameters havenot yet been systematically incorporated intodecision-support systems to define recommenda-tion domains for RP.

On the other hand there is evidence that onweakly buffered West African soils the use ofmineral fertilizers may also lead to rapid de-creases in SOM and pH, thereby detrimentallyaffecting crop yields in the long run (Pieri 1986;Bationo and Buerkert 2001). This is in contrast toevidence from chemically more fertile soils in thetemperate zones with their much higher bufferingcapacity. To be sustainable, any long-termapplication of mineral fertilizers to Sudano-Sah-elian soils will therefore need to be combinedwith the application of organic matter to com-pensate for the relatively higher SOM turnoverrates on fertilized plots.

Moreover, harmful effects of fertilizer applica-tion on seedling survival have been reported withrainfall scarcity. For trials in farmers’ fields inBurkina Faso, Dugue (1998) reported that theapplication of mineral fertilizer was not profitablein 15–48% of the cases due to water stress duringthe early stages of millet growth. In on-stationexperiments in Niger, seed-placed P application atlow soil moisture decreased plant survival by20–33% (Muehlig-Versen et al. 2003).

11 E. Schlecht

Table

4.Effects

ofdifferentapplicationfrequencies

andamounts

ofrockphosphate

andNPK

fertilizer

oncerealgrain

andlegumeyieldsin

differentagro-ecologicalzones

of

Sudano-SahelianWestAfricaasreported

bydifferentauthors.

Amendmenta

Amount(kgha�1)

Yield

increase

above

controlb

Application

frequency

Zone,

soiltype

References

Rockphosphate

tocereals

(applied

in1984)

39(broadcast)

Milletgrain

34%

(measuredin

1988)

Notgiven

Sem

i-arid(N

iger),

Arenosol

Bationoand

Mokwunye(1991)

Rockphosphate

tocereals

130(broadcast)

Milletgrain

76%

3–10years

Sem

i-arid(N

iger),

Arenosol

Bationoet

al.(1990)

Rockphosphate

tocereals

130(broadcast)

Milletgrain

14�1

93%

10years

Sem

i-arid(N

iger),

Arenosol

Bationoet

al.(1995)

Rockphosphate

tocereals

andlegumes

130(broadcast)

CerealTDM

38�1

72%

(inyear3)

10years

Sem

i-arid(N

iger),

Arenosol

Buerkertand

Hiernaux(1998)

LegumeTDM

13–128%

Rockphosphate

compacted

withsoluble

fertilizers

130(broadcast)

Cerealgrain

69–108%

Annually

Sem

i-arid(N

iger),

Arenosol

Bationoet

al.(1995)

Rockphosphate

tolegumes

39(broadcast)

TDM

31%

over

4years

3-yearly

8sitesacross

WestAfricad

Buerkertand

Hiernaux(1998)and

Buerkertet

al.(2002)

Rangein

year3:21–98%

Rockphosphate

+NPK

tolegumes

39(broadcast)

+4PasNPK

cplaced

TDM

35%

over

4years;

3-yearly/N

PK

annually

8sitesacross

WestAfricad

Buerkertet

al.(2002)

Rangein

year3:6–171%

Rockphosphate

tocereals

39(broadcast)

TDM

39%

over

4years

3-yearly

8sitesacross

WestAfricad

Buerkertet

al.(2002)

Rangein

year3:0–203%

Rockphosphate

+NPK

tocereals

39(broadcast)

+4PasNPK

cplaced

TDM

51%

over

4years

3-yearly/N

PK

annually

8sitesacross

WestAfricad

Buerkertet

al.(2002)

Rangein

year3:65–259%

Pseed

coating(SSP,AHPetc.)

<0.5

mgPseed

�1

Millet,TDM:30%

Annually

Sem

i-arid(N

iger),

Arenosol

Rebafkaet

al.(1993a)

Milletgrain:45%

SSP

+ureato

cereals

13(P)

+30(N

)(broadcast)

Milletgrain:50–>

300%

Annually

Sem

i-arid(N

iger),

Arenosol

Yamoahet

al.(2002)

SSPto

legumes

13(broadcast)

TDM

18%

over

4years

Annually

8sitesacross

WestAfricad

Buerkertand

Hiernaux(1998)

Rangein

year3:41–127%

SSPto

cereals

13(broadcast)

TDM

24%

over

4years

Annually

8sitesacross

WestAfricad

Buerkertand

Hiernaux(1998)

Rangein

year3:59–211%

SSPto

legume

13(broadcast)

Grain

18%

Annually

8sitesacross

WestAfricad

Buerkertet

al.2002

SSPto

cereal

8.7

(broadcast)

Milletgrain

102%

Annually

Sem

i-arid(N

iger),

Arenosol

Bationoand

Mokwunye(1991)

12 A critical analysis for soil fertility restoration

SSPto

cereal

13(broadcast)

Milletgrain

168%

Annually

Sem

i-arid(N

iger),

Arenosol

Bationoand

Mokwunye(1991)

Milet

straw

120%

Generalresponse

functions

onlargedatasets

Grain

yield

ofcereals

(Y,kgha�1)

withN

(kgha�1)

Mughoghoet

al.(1986)

Y=

2545+

14.62N�0

.07N

2

(R2=

0.30)

Humid

zone

Y=

1332+

32.08N�0

.13N

2

(R2=

0.72)

Sub-humid

zone

Y=

617+

14.34N�1

6N

2

(R2=

0.36)

Sem

i-aridzone

Partiallyacidulated

phosphate

rock

(PAPR

50)

EffectivenessofPAPR

50oncereal

yieldsascomparedto

SSP:

(nocontrolyieldsreported)

Bationoet

al.(1986)

56%

Humid

zone

85%

Sub-humid

zone

76%

Sem

i-aridzone

�109%

BurkinaFaso

�142%

Nigeria

aSSPsingle

super

phosphate,AHPAmmonium

hydrogen

phosphate.

bTDM

=totaldry

matter.

cNPK

15-15-15.

dSites:5Arenosolsin

semi-aridNiger,3Alfisolsin

sub-humid

Togo.

13 E. Schlecht

Technology evaluation, transfer and adoption:

neglected issues and open questions

The multitude of described technologies aiming atincreasing crop yields on Sudano-Sahelian soils isin sharp contrast to their low adoption levels onfarmers’ fields. Supposing that the small-scaleagro-pastoralists predominating in this area mustmake rational decisions in order to survive in aharsh environment, questions must be raisedabout which issues may have been neglected inexperiments and feasibility studies performed sofar. It appears that there are at least the following.

Capital availability and labour costs

A number of authors have pointed to the problemof subsistence farmers not having enough capitalto build up fertility on the geologically very oldWest African soils which are much lower innutrients and SOM than soils in other parts of theworld, except Australia (Breman et al. 2001). Inthe late 1990s this has been used to advocate largepublic investment programs such as a World Bankinitiative to ‘recapitalize’ soil fertility in sub-Sah-aran West Africa, partly based on the use of lo-cally available rockphosphates (Mokwunye 1995;Mokwunye et al. 1996; Scoones and Toulmin1999). Even if investment needs may be high forsome of the technologies advocated, there are twoarguments against the claim that lacking capitalavailability is the major obstacle for innovation atthe farmers’ level. First, needs for external capitalare certainly low for mulching techniques, whereasthey may be medium for the application of farm-yard manure and corralling and high for rock-phosphates and the application of manufacturedfertilizers such as SSP or P-coated seeds. Installa-tion and maintenance of fodder-banks and legu-minous leys also tend to be both cost and labourintensive. If low capital was indeed limiting tech-nology adoption one would therefore expect thatfarmers practise mulching much more intensivelythan they do at present where at the end of the dryseason many still burn large amounts of stover ontheir fields to get rid of it as a nuisance for sub-sequent weeding. Second, across the Sudano-Sah-elian zone even small farmers readily invest incounter-season crops, typically planted in small,clay-rich depressions, in rice fields or in peri-urban

agriculture. It thus appears as if, rather than thecapital status of the farmer, the ratio betweeninvestments and expected returns as well as ‘risk offailure’ estimates govern the decision about whe-ther or not an investment is made in soil amend-ments.

The timely availability of labour and real labourcosts seems to be another often overlooked factorgoverning investment in soil fertility measures.Labour needs can be classified as low for corrallingand medium for the practices of mulchingand fertilizer application. They are high for the zaıtechnique, compost making and hand-spreading ofcompost and for the application of householdwaste and farmyard manure. However, the scar-city of reliable labour estimates for these technol-ogies complicates the evaluation of their applicabilityand attractiveness; in fact, their neglect is seen as amajor cause for the non-adoption of technologiesthat seemed to be highly promising from a bio-physical point of view (Bationo et al. 1998). This isillustrated by data of Lamers’ (unpublished) whovaluated net returns to labour of different mulch-ing and fertilizing treatments in four villages ofsouth-western Niger. Although annual crop resi-due mulch application at 2000 kg ha�1 without orin combination with application of P and N fer-tilizers resulted in high millet yield increases at thesame sites (Buerkert et al. 2000), the total netreturns to labour were always highest for theuntreated control and only occasionally positivefor mulching and fertilizer treatments at the sitewith the most favourable pedologic and climaticconditions (Figure 2).

One problem in this context may be that eco-nomic modelling which assesses the attractivenessor profitability of technologies does often not de-tail labour and capital needs for the individualoperations related to a technology. In modelapproaches labour costs of activities are oftenconsidered at an aggregated seasonal or annuallevel (Shepherd and Soule 1998; Barbier andCarpentier 2000). Aggregating labour costs forfield preparation, seeding, manure or fertilizerapplication, weeding and harvest, such as in theexhaustive WOCAT2 knowledge base may disguisetreatment effects on labour requirements duringcritical periods. An exception to this is the model

2WOCAT World Overview of Conservation Approaches and

Technologies; http://www.wocat.org

14 A critical analysis for soil fertility restoration

of Van Duivenbooden and Veeneklaas (1993),which considers labour requirements for individ-ual field operations. Unlike Lamers et al. (1998),most modelling papers also do not disclose theabsolute costs and profits to the reader, whichmakes conclusions beyond the site-specific contextvery difficult. Certainly, labour requirements foroperations such as weeding and harvesting alsodepend on the biomass and grain production ofweeds and crops, respectively. Other operations,such as field preparation, sowing and starter fer-tilization require a relatively fixed amount of la-bour per hectare, but will vary between manualand mechanized methods (Hengsdijk and VanKeulen 2002).

Another major issue of concern in the existingeconomic evaluation of soil fertility technologies inSSWA are the assumed costs of farm labourthemselves. Most agricultural economists base thecalculation of the profitability of soil fertilitytechnologies on labour costs at local labour mar-kets, whereas in SSWA many field owners andfarm managers travel to major costal cities to earna living during prolonged periods of the year. Tounravel farmers’ decision making processes andopportunity costs for labour, it may therefore beinteresting to at least alternatively consider costsof farm labour based on these activities. Anotherproblem is that economic modellers may not beaware of side conditions governing the availabilityand accessibility of inputs needed for a particulartechnology. A good example for this is that theexperimentally most widely tested rate of cropresidue mulch, an annual application of2000 kg ha�1, is unrealistically high for farmers in

the Sahelian zone, given the many alternative usesof this resource (Lamers and Feil 1993), and thelike applies to mulching with unpalatable grasses.However, some of these issues have now been re-cognised (Scoones and Toulmin 1998; Schlechtand Buerkert 2004) and taken into account inmore recent research approaches (Buerkert et al.2001; Muehlig-Versen et al. 2003).

Geographical and temporal scales

In recent years, the large short-distance variationin crop growth typically observed on sandySahelian soils has been an intensively studiedphenomenon. Reflecting both, the effects of landuse history and present farmers’ judicious man-agement practices (‘precision farming’), its conse-quences for farmers’ livelihoods and technologyadoption continues to be debated. At an experi-mental level it has been shown that this micro-variability can interact with soil amendment effectson crops or with leaching losses (Buerkert et al.1995; Buerkert and Stern 1995; Brouwer andPowell 1998; Florax et al. 2002; Gandah et al.2003). Whether farmers explicitly exploit themicro-variability of their fields to minimize theeffects of rainfall-related risks of crop failure orhave just learned to cope with it as a consequenceof scarce internal resources to enhance soil pro-ductivity (manure, crop residues, compost, bran-ches from perennial plants) which are applied onspots of highest returns, has been discussedintensely (Brouwer et al. 1993; Haigis 2000;Mazzucato and Niemeijer 2001; Schlecht and

Figure 2. Total net returns to labour of different millet fertilization strategies at four locations in Niger, 1998 (Lamers et al.

unpublished data). P0/Pl3: application of 0/l3 Kg P ha�1. �CR/+CR: application of 0/2000 Kg ha�1 crop residue mulch.

15 E. Schlecht

Buerkert 2004). Whatever the case, any realisticassessment of economic benefits of alternativetechnologies under on-farm conditions and pre-dictions about farmers’ adoption behaviour willrequire a comprehensive understanding of whereand at what scale farmers would likely apply them,at least initially, within their complex land usesystem.

The interaction between short-distance variabil-ity in treatment responses of crop yields at the fieldscale appears to be mirrored at a larger scale. In amulti-site benchmark experiment, Buerkert et al.(2000) showed that the yield increases in TDMafter 4 years of crop residue mulch application at2000 kg ha�1 varied between 13 and 72% at fourSahelian sites (Arenosols; 500–600 mm a�1, millet)but was only 10% at two Sudanian sites (Alfisols;1100– 1300 mm a�1, maize and sorghum). Sub-sequent multi-site time trend analyses indicatedthat mulch effects on cereal dry matter yielddepended largely on the soil type. Hereby, the basesaturation of the sandy Sahelian soils was a majorparameter for the prediction of mulch effects oncereal yields (Buerkert et al. 2001, 2002). Similarresults were also obtained for RP application wherethe magnitude of effects depended on pH or basesaturation, rainfall and the simultaneous applica-tion of soluble P to the pocket at seeding (Buerkertet al. 2001, 2002).

The experimentally demonstrated spatial vari-ability of treatment effects at different geographi-cal scales is accompanied by a temporalcomponent leading to interactions that need to betaken into account in evaluation schemes tounderstand farmers’ management strategies. Soilorganic matter, for instance, can be separated intoa coarse (>50 lm) and a fine (<50 lm) fraction(Feller and Beare 1997). The decomposition of thecoarse SOM fraction by meso- and micro-faunaactivities determines the amounts of nutrientsrecycled to the soil (Manlay et al. 2004). The fineSOM fraction mineralises slowly, as was illustratedby a double-pool exponential model fitted todecomposing organic matter in Sahelian soils(Somda et al. 1995). Such longer-term decompo-sition processes of fine SOM are presumablyresponsible for residual effects of mulching andmanuring treatments. With manure application topearl millet, beneficial effects on grain and stalkyields were observed up to 4 years after applica-tion of even modest rates of manure (Schlecht

et al. 2004; Figure 3). Moderate residual effects ofcrop residue mulch on millet biomass yields werealso determined for up to four cropping periods ina pot trial (Powell et al. 1999) and in a field study(Figure 4).

At an intermediate time scale, the decomposi-tion process of SOM is also influenced by the

Figure 3. Millet grain yield above control as affected by

manure application at 2–14 t ha�1 in year 1 and residual effects

until year 4 (Schlecht et al. 2004).

Figure 4. Effects of annually applied soil amendments on millet

total dry matter and residual effects of these treatments 2 years

after closure of the experiment on the above-ground dry matter

of the fallow vegetation. Treatments were an absolute control,

annual crop residue mulching at 500 (CR500) and 2000

(CR2000) kg ha�1, and treatment means without P (P0) and

annual application of 13 kg P ha�1 (P13). Modified after

Gerard et al. (2001).

16 A critical analysis for soil fertility restoration

N-concentration of the applied material: at>20 g N kg�1 organic material generally decom-poses faster than at lower N levels (Stevenson andCole 1999). With livestock manure, an observedtemporary immobilization of mineral soil N wasattributed to undigested cell wall components inthe faeces, even at higher N levels in the amend-ment (De Ridder and Van Keulen 1990; Powell etal. 1999). In addition, seasonal differences in thedecomposition and mineralisation of applied or-ganic matter occur, which are due to variations insoil temperature, rainfall and soil moisture (Somdaet al. 1995; Esse et al. 2001; Schlecht and Buerkert2004). Van der Pol (1992) raised the problem thattemporal changes in the nutrient release followingmineralisation of applied organic matter are notsufficiently assessed in most studies. This issue wassubsequently taken up by Smaling et al. (1996),who concluded that the spatio-temporal dynamicsof organic matter decomposition and nutrientstocks at different nutrient levels are still poorlyunderstood. To our knowledge, not much haschanged in this respect since then. Such anunderstanding, however, is needed for a betterappreciation and exploitation of the effects ofSOM mineralization on soil productivity and cropgrowth. There is mounting evidence that Africanfarmers are well aware of the residual effects oforganic soil amendments, and are accounting forthese in their land use and cropping strategies(Evequoz et al. 1998; Powell et al. 1999; Schlechtet al. 2004).

The human component of soil fertility management

Dent et al. (1995) and Bationo et al. (1998)pointed to the fact that technology adoptiondepends to a large extend on socio-economic andsocio-cultural factors, which until recently, havebeen largely ignored in the context of soil fertilityresearch in West Africa. Stoorvogel et al. (1995)viewed individual farmers as the final decisionmakers, whose decisions are conditioned by theirresource endowments, household goals and socio-economic environment. Dent et al. (1995) andMazzucato and Niemeijer (2001) stated that inmany agricultural societies, decisions of a singlemanager are the result of agreements with familymembers, friends, neighbours and larger socialnetworks.

In addition to that, the (individual) perceptionof soil fertility is multi-faceted, dynamic and con-textual: Warren (2002) observed that a house-hold’s resource endowment strongly influenced theextend of erosion on individual fields, with well-endowed farmers paying less attention to soilconservation measures than poorer ones. Similarresults were obtained from bio-economic model-ling of nutrient management by Kenyan farmers(De Jager et al. 1998). In addition to that, differentsocial arrangements, different perceptions of thefuture and changing time horizons diversify theappraisal of a particular resource and its state ofdegradation. Therefore, Warren (2002) argues thatthe evaluation of land degradation (or soil fertility)cannot be limited to assessments of soil parame-ters, nutrient budgets or household economics. Ifthis view is accepted, the evaluation of soil fertilityand identification of measures that might help toimprove it becomes very complicated.

At present, the perception on future develop-ment of soil fertility at the regional level ischanging from the pessimistic views of the 1980sand 1990s (Penning de Vries and Djiteye 1982;Van Keulen and Breman 1990; Van der Pol 1992;Stoorvogel et al. 1993) to more optimistic evalua-tions. From the analysis of the correlation betweenarea productivity and long-term average rainfall,rural population density, animal traction index,fertilizer and manure usage, soil and water con-servation measures and agricultural extensionacross Burkina Faso, it appeared that area pro-ductivity was mainly influenced by rainfall, butwas barely related to rural population density oruse of technology and soil amendments (Mazzu-cato and Niemeijer 2001). This was attributed tothe high specificity of soil fertility managementbased predominantly on internal organic re-sources, labour input and social networks. Thesefindings put into question the belief that the lowuse of external inputs in the Sahel will result inrapid region-wide soil depletion (Mazzucato andNiemeijer 2001). The view of the latter authors ispartly based on the comparison of nutrient con-tents of soil samples taken in the same region at aninterval of 30 years. From these it appeared thatthe N, P and K levels of cultivated surface soilswere higher than those of long-term fallowed ones,leading to the conclusion that farmers had devel-oped strategies to maintain or even raise soil fer-tility over time. In our view, two points were

17 E. Schlecht