THE IMPACT OF CLIMATE VARIABILITY AND CHANGE ON SWEET POTATO PRODUCTION IN EAST AFRICA By Saul Daniel Ddumba A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Geography Doc tor of Philosophy 2018 ABSTRACT THE IMPACT OF CLIMATE CHANGE AND VARIABILITY ON SWEET POTATO PRODUCTION IN EAST AFRICA By Saul Daniel Ddumba Over decades, food production in East Africa has been affected by a changing climate, limited use of fert ilizers and pest control, inadequate food storage facilities and complex marketing channels that together have led to malnutrition, hunger , and poverty. The six most important food crops feeding the region include cassava, maize, plantains, sweet potatoes , potatoes and paddy rice. Of all these crops, relatively little is known about how climate influences sweet potato growth, development, and yield. Deterministic simulation models for sweet potatoes exist but are relatively young or still in development. Re levant data for climate impact assessments are scarce: detailed agronomic data for sweet potato cultivars grown in East Africa are limited; representative high - quality climate data for the region are scarce , and soils data is only available at coarse spati al resolution. The major objective of this research was therefore to assess the impact of climate variability and change on sweet potato production in East Africa. This was achieved by: (i) developing a modeling framework for use in a deterministic sweet p otato crop model, SPOTCOMS, for East Africa; (ii) analyzing trends of historical climate and sweet potato root yields for the period 1980 - 2009; (iii) developing local climate change scenarios for East Africa for the current time slice 2010s, near future 20 30s, mid - future 2050s and distant future 2050s using two representative concentration pathways 4.5 and 8.5 for four Global Climate Models, CSIRO, MIROC5, MRICGC3 - M and NorESM - 1; and (iv) estimating the impact of projected future climate change on sweet po tato production using SPOTCOMS model. Crop coefficients where determined from field trials for four sweet potato cultivars, NASPOT 1, NASPOT 10 0, NASPOT 11 and SPK004. Results from the calibration and evaluatio n of SPOTCOMS model gave an index of agreemen t (IA) of 0.94 and 0.7, a modeling efficiency of 0.9 and 0.31, and a mean bias error of 1.16 t/ha and 0.5 t/ha respectively. Trend analysis indicated that East Africa had warmed on average by 1.5 0 C, the rainfall for the February - June season had declined b y more than 60 mm while rainfall for the August December season had increased for most parts of East Africa by more than 50 mm over the past 30 years. The results of future climate projections from Global Climate Models showed mixed results for precipita tion and more distinct results for temperatures. Temperatures in the region were projected to rise by 0.8 0 C, 1.2 0 C and more than 3 0 C in the 2030s , 2050s, 2070s respectively and precipitation is projected to consist of more increases in the short rainfall i ntensity than the long rains for all the three future time slices. The s ensitivity analysis showed that SPOTCOMS was sensitive to increase in precipitation and temperature for all the four sweet potato cultivars, NASPOT 1, NASPOT 10 0, NASPOT 11 and SPK004 . The projected increase in sweet potato yield in the region coincides with areas that will experience increases in precipitation and temperature. Models with the larger radiative forcing of RCP8.5 showed an overall higher increase in precipitation, temper atures and therefore higher increases in sweet potato yield. All the four cultivars ( NASPOT 1, NASPOT 10 0, SPK004 and NASPOT 11) showed similar spatial distribution of yields but SPK004 had lower yields for both historical and projected future periods. Re sults from this study are useful to all stakeholders interested in sweet potat o production in East Africa and the rest of the tropics. iv ACKNOWLEDGMENTS Coming to Michigan State University to pursue a Ph .D. was one of the most rewarding experiences for me and my family. My journey began wit h the generous support I received from the people of USA through the Fulbright Science and Technology Program. And, during the course of my program, I have received support from the Borlaug LEAP, the Global Center for Fo od Systems Innovation, MSU Agricultural Extension and other financial sources at MSU. I will forever be grateful to the people of USA . But like any good journey, there are always people who make it worthwhile. On top of the list is my advisor, Prof. Jeffre y A. Andresen. I am very grateful to Jeff for being a good mentor to me both in my academic career and together with his family, they have supported me greatly. I also greatly appreciate my guidance committee ; Prof. Julie Winkler, Prof. Sieglinde Snapp, Pr of. Nathan Moore and Prof. Jennifer Olson. Thank you so much for the support you have given me dur ing my Ph .D. journey. I am equally grateful to the faculty of t he Department of Geography who ha ve contributed to my program in many different ways. I also wish to extend my appreciation to Dr. Gopal Alagarswamy and the CLIP Lab for ment oring me in crop modeling and for supporting me in many other ways during the entire period of my Ph .D. program at MSU. I also thank Dr. S. Mithra, Dr. Robert Mwanga, Dr. Gore ttie Sengendo, Dr. Robert o Quiroz, Mr. Isaac Mpembe and Makerere University for supporting me in many different ways during my Ph .D. program . Finally, I would like to thank my dear wife, Rebecca Nassimbwa Ddumba and our three handsome boys Jotham Ddumba , Evan Ddumba and Andrew Ddumba for supporting me and for keeping me motivated . And to my dad, Livingstone Sengooba , my mum Annette Ndagire , and my siblings thank you. I dedicate this work to God, for blessing me with life. v TABLE OF CONTENTS LIST OF TABLES ................................ ................................ ................................ ................... vii LIST OF FIGURES ................................ ................................ ................................ .................... x CHAPTER 1 ................................ ................................ ................................ ............................... 1 AGROCLIMATOLOGICAL ASPECTS OF SWEET POTATO PRODUCTION IN EAST AFRICA ................................ ................................ ................................ ................................ .. 1 1.1 Introduction ................................ ................................ ................................ .................... 1 1.2 Climate change and variability, and physical characteristics of East Africa ..................... 4 1.3 Regional agriculture ................................ ................................ ................................ ....... 6 1.4 Climate requiremen ts for sweet potatoes ................................ ................................ ........ 9 1.5 Characteristics and adaptive potential of sweet potato cultivars in Uganda ................... 11 1.5.1 Introduction ................................ ................................ ................................ ........... 11 1.5.2 Sweet potato farming systems in Uganda ................................ ............................... 12 1.5.3 Sweet potato varieties in Uganda ................................ ................................ ........... 14 1.5.3.1 Maturity ................................ ................................ ................................ .......... 18 1.5.3.2 Yield ................................ ................................ ................................ ............... 18 1.5.3.3 Taste ................................ ................................ ................................ ................ 20 1.5.3.4 Abundance ................................ ................................ ................................ ...... 20 1.5.3.5 Resistance to pests and diseases ................................ ................................ ....... 21 1.5.4 Farmer support and dissemination of plant ing materials ................................ ......... 23 1.5.5 Constraints to sweet potato production in Uganda ................................ .................. 25 1.5.6 Participatory and adaptive research on sweet po tato production ............................. 25 1.5.7 Conclusion ................................ ................................ ................................ ............. 28 1.6 Research gaps and limitations ................................ ................................ ....................... 29 1.7 Research questions and objectives ................................ ................................ ................ 30 CHAPTER 2 ................................ ................................ ................................ ............................. 31 APPLICATION OF A PROCESS - BASED MODEL FOR SWEET POTATO GROWTH DEVELOPMENT IN EAST AFRICA ................................ ................................ ................... 31 2.1 Introduction ................................ ................................ ................................ .................. 31 2.2 Materials and methods ................................ ................................ ................................ .. 33 2.2.1 Treatments and experimental design ................................ ................................ ...... 33 2.2.2 Selection of cultivars ................................ ................................ .............................. 35 2.2.3 Plant management ................................ ................................ ................................ .. 36 2.2.4 Plant trait destructive monitoring ................................ ................................ ........... 36 2.2.5 Soils ................................ ................................ ................................ ...................... 36 2.2.6 Secondary agronomic and climate data ................................ ................................ .. 41 2.2.7 Trend analysis for the collected field data ................................ .............................. 43 2.2.8 Calibration and validation of SPOTCOMS crop - model ................................ .......... 43 2.2.8.1 The SPOTCOMS model ................................ ................................ .................. 43 2.2.8.2 Determination of cultivar parameters for SPOTCOMS ................................ .... 48 2.2.8.3 Calibration and testing of SPOTCOMS ................................ ........................... 49 vi 2.2.8.4 Sensitivity analysis of cultivar coefficients in SPOTCOMS ............................. 50 2.2.8.5 Evaluation of model performance ................................ ................................ .... 51 2.3 Results ................................ ................................ ................................ ......................... 53 2.3.1 Variation of Sweet potato data fro m the field ................................ ......................... 53 2.3.2 Cultivar coefficients ................................ ................................ ............................... 60 2.3.3 Sensitivity of cultivar coefficients from field data ................................ .................. 63 2.3.4 Calibration of the model ................................ ................................ ........................ 66 2.4 Discussion ................................ ................................ ................................ .................... 71 2.5. Conclusion ................................ ................................ ................................ .................. 73 CHAPTER 3. ................................ ................................ ................................ ............................ 77 THE IMPACT OF CLIMATE CHANGE AND VARIABILITY ON SWEET POTATO PRODUCTION IN EAST AFRICA ................................ ................................ ....................... 77 3.1 Introduction ................................ ................................ ................................ .................. 77 3.2 Methodology ................................ ................................ ................................ ................ 79 3.2.1 Description of the study area ................................ ................................ .................. 80 3.2.2 Data sources ................................ ................................ ................................ .......... 81 3.2.2.1 Historical data ................................ ................................ ................................ . 81 3.2.2.2 Selection of projected future climate scenario and preparation of projected future climate data ................................ ................................ ................................ ....... 88 3.2.2.3 Wet/dry GCM models and hot/cool GCM models ................................ ............ 92 3.2. 3 Trend analysis ................................ ................................ ................................ ........ 95 3.2.4 Sensitivity analysis ................................ ................................ ................................ 96 3.2.5 Climate change impact assessment ................................ ................................ ......... 97 3.3 Results ................................ ................................ ................................ ......................... 99 3.3.1 Trend analysis ................................ ................................ ................................ ........ 99 3.3.2 Sensitivity analysis ................................ ................................ .............................. 108 3.3.2 Projected future climate and sweet potato production model results ..................... 111 3.3.2.1 Future climate projections for East Africa ................................ ...................... 111 3.3.2.3 Future crop yields ................................ ................................ .......................... 124 3.3.2.3.1 Historical sweet potato yield ................................ ................................ ... 124 3.3.2.3.2 Febru ary - June (FMAMJ) yield projections ................................ .............. 128 3.3.2.3.3 August - December (ASOND) yield projections ................................ ........ 133 3.4. Discussion ................................ ................................ ................................ ................. 138 3.5 Conclusion ................................ ................................ ................................ ................. 142 CHAPTER 4. ................................ ................................ ................................ .......................... 144 SUMMARY AND CONCLUSION ................................ ................................ ................. 144 APPENDICES ................................ ................................ ................................ ........................ 150 APPENDIX A: Sensitivity analysis ................................ ................................ ..................... 151 APPENDIX B: Mean changes i n sweet potato yield ................................ ............................ 163 APPENDIX C: Projected changes (%) for annual rainfall ................................ ................... 169 APPENDIX D : Projected mean temperature ................................ ................................ ........ 183 BIBLIOGRAPHY ................................ ................................ ................................ ................... 195 vii LIST OF TABLES Table 1. 1 Crop rotations of sweet potatoes in Uganda. Source: Basha asha et al., 1995. ............ 13 Table 1. 2 Non Orange varieties. ................................ ................................ ............................. 15 Table 1. 3 Orange - fleshed varieties ................................ ................................ ........................... 17 Table 1. 4 Characteristics of selected non - orange fleshed varieties ................................ ............ 19 Table 1. 5 Characteristics of selected orange - fleshed varieties (abundance not known) .............. 22 Table 1. 6 Organizations helping farmers in sweet potato production ................................ ........ 24 Table 2. 1 Properties of sw eet potato cultivars used in this study ................................ ............... 35 Table 2. 2 Soil characteristics for Pit One ................................ ................................ .................. 39 Table 2. 3 Soil characteristics for Pi t Two ................................ ................................ ................. 39 Table 2. 4 Soil nutrient analysis results ................................ ................................ ...................... 40 Table 2. 5 Description of experimental data sites used in the determinatio n of cultivar coefficients and validation of the crop model ................................ ................................ ............ 42 Table 2. 6 Soil characteristics used in SPOTCOMS ................................ ................................ ... 47 Table 2. 7 Su mmary of cultivar parameters determined from field experiments ......................... 61 Table 2. 8 Sensitivity analysis of cultivar coefficients ................................ ............................... 64 Ta ble 2. 9 Calibration (under irrigation) and evaluation ................................ ............................. 67 Table 2. 10 Correlation coefficients between simulated root yield and observed root yield at selected locations. ................................ ................................ ................................ ..................... 67 Table 3. 1 Basic Descriptions for the locations in Uganda 85 Table 3. 2 Basic Descriptions for the lo cations in Kenya ................................ ........................... 86 Table 3. 3 Basic Descriptions for the locations: Tanzania ................................ .......................... 87 Table 3. 4 Description of global circulation models used ................................ ........................... 89 viii - 2009 by site and season. ................................ ................................ ................................ ................................ .... 106 Table 3. 6 Summary data for Sensitivity analysis Dagoretti Corner (DC), Kenya ..................... 151 Table 3. 7 Summary data for Sensitivity analysis Mbeya (MB), Tanzania ............................... 155 Table 3. 8 Summary data for sensitivity analysis Namulonge (NAM), Uganda ........................ 159 Table 3. 9 Ensemble mean changes in sweet potato root yield (t/ha) under future cl imate scenarios for 4 GCMs; CSIRO, MIROC5, MRI - CGCM3, NorESM1 - M for August - December (ASOND) season for NASPOT 1: Base yield, 2030 - rcp 4.5, 2050 - rcp4.5, 2070 - rcp4.5, 2030 - rcp8.5, 2050 - rcp8.5, 2070 - rcp8.5 ................................ ................................ ................................ ................. 163 Table 3. 10 Ensemble mean changes in sweet potato root yield (t/ha) under future climate scenarios for 4 GCMs; CSIRO, MIROC5, MRI - CGCM3, NorESM1 - M for August - December (ASOND) season for NASPOT 11 ................................ ................................ ................................ ............ 164 Table 3. 11 Ensemble mean changes in sweet potato root yield (t/ha) under future climate scenarios for 4 GCMs; CSIRO, MIROC5, MRI - CGCM3, NorESM1 - M for August - December (ASOND) season for NASPOT 10 ................................ ................................ ................................ ............ 166 Table 3. 12 Ensemble mean changes in sweet potato root yield under future climate scenarios for 4 GCMs; CSIRO, MIROC5, MRI - CGCM3, NorESM1 - M for August - December (ASOND) season for SPK004 ................................ ................................ ................................ ............................. 167 Table 3. 13 Changes in annual rainfall for 2030s ................................ ................................ ..... 169 Table 3. 14 Changes in annual rainfall (mm) for 2050s ................................ ............................ 170 Table 3. 15 Changes in annual rainfall (mm) for 2070s ................................ ............................ 171 Table 3. 16 Changes in the February - June (FMAMJ) rainfall for 2030s ................................ ... 173 Table 3. 17 Changes in the February - June (FMAMJ) rainfall for 2050s ................................ ... 175 Table 3. 18 Changes in the February - June (FMAMJ) rainfall for 2070s ................................ ... 176 Table 3. 19 Changes in the August - December (ASOND) rainfall for 2030s ............................. 178 Table 3. 20 Changes in the August - December (ASOND) rainfall for 2050s ............................. 179 Table 3. 21 Changes in the August - December (ASOND) rainfall for 2070s ............................. 181 Table 3. 22 Annual mean temperatures for 2030s for the 4 GC Ms under two representative concentration pathways, RCP4.5 and RCP 8.5 ................................ ................................ ......... 183 ix Table 3. 23 Annual mean temperatures for 2050s for the 4 GCMs under two representative concentration pathways, RCP4. 5 and RCP 8.5 ................................ ................................ ......... 184 Table 3. 24 Annual mean temperatures for 2070s for the 4 GCMs under two representative concentration pathways, RCP4.5 and RCP 8.5 ................................ ................................ ......... 186 Table 3. 25 Mean temperatures for February - June in 2030s for the 4 GCMs under two representative concentration pathways, RCP4.5 and RCP 8.5 ................................ .................. 187 Table 3. 26 Mean temp eratures for February - June in 2050s for the 4 GCMs under two representative concentration pathways, RCP4.5 and RCP 8.5 ................................ .................. 189 Table 3. 27 Mean temperatures for February - June in 2070s for the 4 G CMs under two representative concentration pathways, RCP4.5 and RCP 8.5 ................................ .................. 191 Table 3. 28 Mean temperatures for August - December in 2030s for the 4 GCMs under two representative concentration pa thways, RCP4.5 and RCP 8.5 ................................ .................. 192 Table 3. 29 Mean temperatures for August - December in 2050s for the 4 GCMs under two representative concentration pathways, RCP4.5 and RCP 8.5 ................................ .................. 193 Table 3. 30 Mean temperatures for August - December in 2070s for the 4 GCMs under two representative concentration pathways, RCP4.5 and RCP 8.5 ................................ .................. 194 x LIST OF FIGURES Figure 1. 1 Factor s affecting crop production and associated consequences in East Africa ........... 3 Figure 1. 2 Map of East Africa ................................ ................................ ................................ .... 5 Figure 1. 3 Top 8 crop s in East Africa ................................ ................................ ......................... 7 Figure 1. 4 Average annual production for the top eight staple crops in East Africa, Source: FAO, 2013 ................................ ................................ ................................ ................................ ............ 8 Figure 2. 1 Design of the experiment. (a) The layout of field plots by cultivar type, V1, V2, V3, V4 for the four replications; REP 1, REP 2, REP 3 AND REP 4. V1, V2, V3, V4 represent the sweet potato cultivars NASPOT 1, NASPOT 10 O, NASPOT 11 and SPK 004. (b) The layout of vines for a single cultivar plot. The crosses represent plants in a column and colors are used to only emphasize that 4 plants in the same column are sampled every sampling date 34 Figure 2. 2 Vine length:V1 = NASPOT 1, V2 = NASPOT 10 0, V3 = SPK004 (Ejumula), V4 = NASPOT 11 ................................ ................................ ................................ .............................. 55 Figure 2. 3 Number of branches:.V1 = NASPOT 1, V2 = NASPOT 10 0, V3 = SPK004 (Ejumula), V4 = NASPOT 11 ................................ ................................ ................................ ... 56 Figure 2. 4 Number of leaves and leaf arear: V1 = NASPOT 1, V2 = NASPOT 10 0, V3 = SPK004 (Ejumula), V4 = NASPOT 11 ................................ ................................ ...................... 57 V1 = NASPOT 1, V2 = NASPOT 10 0, V3 = SPK004 (Ejumula), V4 = NASPOT 11 ............... 58 Figure 2. 6 Flesh weights, and correlation coefficients for storage roots. V1 = NASPOT 1, V2 = NASPOT 10 0, V3 = SPK004 (Ejumula), V4 = NASPOT 11 ................................ .................... 59 Figure 2. 7 Dry wieghts and correlation coefficients for storage roots. V1 = NASPOT 1, V2 = NASPOT 10 0, V3 = SPK004 (Ejumula), V4 = NASPOT 11 ................................ .................... 60 Figure 2. 8 Range of cultivar parameters in 2012 and 20 13 season. The orange dot represents the mean of simulated root yield ................................ ................................ ................................ ..... 62 Figure 2. 9 Sensitivity analysis of cultivar coefficients ................................ .............................. 66 Figure 2. 10 Graphs of simulated against observed data ................................ ............................. 69 Figure 2. 11 Regression plot for model results using historical sweet potato yields for the period 2004 - 2009 ................................ ................................ ................................ ................................ . 70 xi Figure 3. 1 Flow chart describing major project objectives, tasks, processes, and input data types 8 0 Figure 3. 2 Study area. (a) Locations of study sites, (b) soil map of Africa (source: FAO, 2012) 84 Figure 3. 3 Spatial distribution of stochastically generated curr ent mean annual rainfall over East Africa ................................ ................................ ................................ ................................ ........ 93 Figure 3. 4 Spatial distribution of stochastically generated current mean annual temperature over East Africa ................................ ................................ ................................ ................................ 94 Figure 3. 5 Trend in average rainfall over the February to June (FMAMJ) season for selected locations over the period 1980 - 2009 across East Africa ................................ ......................... 101 Figure 3. 6 Trend in average rainfall over the August to December (ASOND) season for selected locations over the period 1980 - 2009 across East Africa. Th e * represent a significant trend at 0.05 level ................................ ................................ ................................ ................................ ........ 102 Figure 3. 7 Trend in average temperature over the February to June (FMAMJ) season for selected locations over the period 1980 - 2009 across East Africa. *, **. And *** represent a significant trend at 0.05, 0.01, 0.001 levels respectively. ................................ ................................ .......... 103 Figure 3. 8 Trend in average temperature over the August to December (ASOND) season for selected locations over the period 1980 - 2009 across East Africa. *, **. And *** represent a significant trend at 0.05, 0.01, 0.001 levels respec tively. ................................ ......................... 104 Figure 3. 9 Climate sensitivity of sweet potato crop model (SPOTCOMS) for the August to December season for 3 locations; Dagoretti Corner in Kenya, Mbeya in Tanzania and Namulonge in Uganda for two sweet potato cultivars NASPOT 1 and SPK 004. The simulated sweet potato yield was as a result of rainfall and temperatures which had been either increased or decreased by a given proportional change in rainfall and a changed amount in bot h minimum and maximum temperatures. ................................ ................................ ................................ ........................... 109 Figure 3. 10 Climate sensitivity of sweet potato crop model (SPOTCOMS) for the February to June (FMAMJ) season for 3 locations; Dagoretti Corner in Keny a, Mbeya in Tanzania and Namulonge in Uganda for two sweet potato cultivars NASPOT 1 and SPK 004. The simulated sweet potato yield was as a result of rainfall and temperatures which had been either increased or decreased by a given proportional change in r ainfall and a changed amount in both minimum and maximum temperatures. ................................ ................................ ................................ .......... 110 Figure 3. 11 Historical mean annual temperature and relative changes of mean annual temperatures for the 2030s for RC P4.5 for four GCMs ................................ ................................ ................. 112 Figure 3. 12 Historical mean annual temperature and relative changes of mean annual temperatures for the 2030s for RCP8.5 for four GCMs ................................ ................................ ................. 113 xii Figure 3. 13 Historical mean annual temperature and relative changes of mean annual temperatures for the 2070s for RCP4.5 for four GCMs ................................ ................................ ................. 114 Figure 3. 14 Hist orical mean annual temperature and relative changes of mean annual temperatures for the 2070s for RCP8.5 for four GCMs ................................ ................................ ................. 115 Figure 3. 15 Historical mean annual rainfall and relative changes of mean annual rainfall for the 2030s for RCP4.5 for four GCMs ................................ ................................ ............................ 116 Figure 3. 16 Historical mean annual rainfall and relative changes of mean annual rainfall for the 2030s for RCP8.5 for four GCMs ................................ ................................ ............................ 117 Figure 3. 17 Historical mean annual rainfall and relative changes of mean annual rainfall for the 2070s for RCP4.5 for four GCMs ................................ ................................ ............................ 118 Figure 3. 18 Historical mean annual rainfall and relative changes of mean annual rainfall for the 2070s for RCP8.5 for four GCMs ................................ ................................ ............................ 119 Figure 3. 19 Historical mean seasonal temperature and relative changes of mean temperatures for August to December for the 2030s for RCP4.5 for four GCMs . ................................ ............... 120 Figure 3. 20 Historical mean seasonal temperature and relative changes of mean tem peratures for August to December for the 2030s for RCP8.5 for four GCMs ................................ ................ 121 Figure 3. 21 Historical mean seasonal temperature and relative changes of mean temperatures for August to December f or the 2070s for RCP4.5 for four GCMs ................................ ................ 122 Figure 3. 22 Historical mean seasonal temperature and relative changes of mean temperatures for August to December for the 2070s for RCP8.5 for four G CMs ................................ ................ 123 Figure 3. 23 Distribution of average sweet potato yield for the August to December season for the period 1980 - 2009 ................................ ................................ ................................ .................. 126 Figure 3. 24 Distribution of average sweet potato yield for the February to June season for the period 1980 - 2009 ................................ ................................ ................................ .................. 127 Figure 3. 25 Historical average yield and relative changes in yield of SPK 004 sweet potato cultivar for the February to June season in the 2030s, 2050s and 2070s for RCP 4.5 for CSIRO - Mk3.6 ................................ ................................ ................................ ................................ ..... 129 Figure 3. 26 Historical average yield and relative changes in yield of SPK 004 sweet potato cultivar for the February to June season in the 2030s, 2050s and 2070s for RCP 8.5 for CSIRO - Mk3.6 ................................ ................................ ................................ ................................ ..... 130 xiii Figure 3. 27 Historical average yield and relative changes in y ield of NASPOT 1 sweet potato cultivar for the February to June season in the 2030s, 2050s and 2070s for RCP 4.5 for CSIRO - Mk3.6 ................................ ................................ ................................ ................................ ..... 131 Figure 3. 28 Figure 6 Historical average yield and relative changes in yield of NASPOT 1 sweet potato cultivar for the February to June season in the 2030s, 2050s and 2070s for RCP 8.5 for CSIRO - Mk3.6 ................................ ................................ ................................ ......................... 132 Figure 3. 29 Historical average yield and rel ative changes in yield of SPK 004 sweet potato cultivar for the August - December season in the 2030s, 2050s and 2070s for RCP 4.5 for CSIRO - Mk3.6 ................................ ................................ ................................ ................................ ..... 134 Figure 3. 30 Historical average yield an d relative changes in yield of SPK 004 sweet potato cultivar for the August - December season in the 2030s, 2050s and 2070s for RCP 8.5 for CSIRO - Mk3.6 ................................ ................................ ................................ ................................ ............... 135 Figure 3. 31 Historical average yiel d and relative changes in yield of NASPOT 1 sweet potato cultivar for the August - December season in the 2030s, 2050s and 2070s under RCP 4.5 for CSIRO - Mk3.6 ................................ ................................ ................................ ......................... 136 Figure 3. 32 Historical aver age yield and relative changes in yield of NASPOT 1 sweet potato cultivar for the August - December season in the 2030s, 2050s and 2070s for RCP 8.5 for CSIRO - Mk3.6 ................................ ................................ ................................ ................................ ..... 137 1 CHA PTER 1 AGROCLIM ATOLOGICAL ASPECTS O F SWEET POTATO PRODUCTION IN EAST AFRICA 1.1 Introduction Agricultural productivity is strongly dependent on climate variability and change. A recent report by the Intergovernmental Panel on Climate Change (IPCC) concluded that glob al mean temperatures will increase between 1.1 0 C and 4.8 0 C from present values by the end of the 21 st century (IPCC, 2013) . Rainfall may become more variable and erratic with a possible increase in the number and severity of extreme events, especially in tropical areas. The warming is largely associated w ith increases in atmospheric greenhouse gas concentrations, largely carbon dioxide (CO 2 ), and, methane (CH 4 ) and nitrous oxide (N 2 O). Global atmospheric concentrations of carbon dioxide, methane and nitrous oxide have increased markedly since 1750 as a res ult of human activities and now far exceed pre - industrial values. CO 2 increased from a pre - industrial value of 280 ppm to 379 ppm in 200 5 (IPCC, 2007b) . There is a high spatial heterogeneity across different regions of the world in terms of projected climate trends and resulting impacts. For example, (IPCC, 2013) noted that the tropical Indian O cean is likely to feature a zonal pattern with reduced warming and decreased rainfall in regions east of the ocean and enhanced warming and increased rainfal l in regions west of the ocean including East Africa. The Indian Ocean dipole of the interannual variability is very likely to remain active, leading to climate extremes in East Africa, Indonesia and Australia (IPCC, 2013) . 2 The El Niño - Southern Oscillation (ENSO) is very likely to remain a dominant mode of interannual variability in the future and the regional rainfall variability it induces may increase (IPCC, 2013) . The level of confidence in ENSO projection is very low in East Africa (IPCC, 2013) , however, average temperatures are projected to increase by more than 1 0 C and with more erratic and variable rainfall possible by 2025 (FEWSNET, 2010, 2012; Hepworth & Goulden, 2008) . Changing climate is likely to lead to unprecedented impacts on agriculture. For example, crop productivity in lo wer latitudes, especially the seasonally dry and tropical regions, is projected to decrease for even small local temperature increases of 1 - 2°C, which would increase the risk of hunger (IPCC, 2007a) . The frequ ency of droughts and floods is projected to increase, which would also lead to negative effects on crop production. Relatively more adverse impacts are projected in the lesser developed countries because they tend to be located in already warm tropical are as, rely on climate - sensitive sectors like agriculture, have relatively low incomes, and weak adaptive capacity (Heltberg, Siegel, & Jorgensen, 2009; Mendelsohn, Dinar, & Williams, 2006) . Africa has been identified as a region of high crop production risk and projected yield losses due to climate change (Parry, Rosenzweig, & Livermore, 2005; Rosenzweig & Parry, 1994; Schlenker & Lobell, 2010) . Crop yields for some countries could decrease by up to 50% by 2020 as a result of declining available agricultural land and shorter growing season len gth (IPCC, 2007a) . Over decades, food production in East Africa has been affected by a changing climate, limited use of fertilizers and pest control, inadequate food storage facilities and complex marketing c hannels that together have led to chronic malnutrition, hunger , and poverty (Figure 1 .1 ). We are also uncertain about how future changes in climate will affect crop production. In order to address these challenges, there is a need to invest in research, to encourage the practice of conservation 3 agriculture, and to use climate - smart technologies like growing high yielding, drought tolerant, and nutritious crops. East Africa is struggling with a food crisis for various reasons. More than 70% of crops g rown in East Africa are rain - fed with little or no application of fertilizers. Supplemental irrigation and application of fertilizers would increase crop yields but the majority of farmers are smallholders who cannot afford the cost of the irrigation infra structure and farm inputs. Farmers rely on the rainy seasons to grow crops in order to feed their families and earn income by selling surplus crops to markets in urban centers. Figure 1. 1 Factors affecting crop production and associated consequences in East Africa Source: Generated by this research 4 The proposed research will assess the impact of climate change and variabilit y on sweet potatoes in East Africa. The climate in this region is highly variable characterized by a seasonal movement of the Intertropical Convergence Zone (ITCZ) (Og allo, 1989) , proximity to the Indian Ocean (Goddard & Graham, 1999) , and variable topography and presence of Lake Victoria (Anyah, Semazzi, & Xie, 2006) . The climate , therefore, supports grow ing of a wide range of food and cash crops in the region. 1.2 Climate change and variability, and physical characteristics of East Africa The topography of East Africa varies from 0 m on the coast of the Indian Ocean to 5, 890 m at the highest peak of M t. Ki l imanjaro . The regional climate is controlled by the presence of the Intertropical Convergence Zone; Indian Ocean; Variable topography (Ogallo, 1989; Goddard and Graham, 1999; Anyah et al., 2006). The region receives a bimodal annual rainfall ranging between 500 mm to over 2,500 mm (FEWSNet, 2010, 2012) and mean annual temperatures range between 8.1 0 C (at high elevations) to 32 0 C (FEWSNet, 2010, 2012). Figure 1. 2 below shows the major physical features and the countries that constitute East Africa. 5 Figure 1. 2 Map of East Africa The length of the long rains season, known to originally fall in the months March to May, has decreased while that of the short rains, originally known to be in the months of September to Decembe r, has increased over the past 30 years (FEWSNet, 2010; 2012). East Africa has experienced increased in the incidence of droughts and floods (Uganda Meteorological Department, 2010) and temperatures have increased by 0.8 0 C over the past 30 years (FEWSNet, 2010, 2012). As a result of these reported changes and variability in climate, crop failures have been reported in the region ( e.g. , Jassogne et al., 2013) which implies that agricultural production systems in East Africa are very sensitiv e to climate. Gl obal warming arising from an increase in greenhouse gases (GHGs) is likely to impact the regional climate of East Africa (IPCC, 2013). Temperatures are projected to increase by an average of 1.5 0 C from their current numbers, and rainfall will become more e rratic and variable in the next 20 years (IPCC, 2013). The long rains growing season will shorten in most areas while 6 short rains will lengthen (IPCC 2007; McSweeney et al., 2008; Hepworth and Goulden, 2008; FEWSNET, 2010, 2012; Cook and V icky 2013). Incre ases in GHGs will directly impact crop species especially C 3 crops through CO 2 enrichment (Bhattacharya et al 1985; 1990; Siqinbatu et al., 2013). Droughts, floods, shortening of growing seasons, increasing temperatures will lead to high crop production ri sks and large yield losses in East Africa (Rosenzweig and Parry, 1994; Parry et al., 2005; Thornton et al., 2009; Schlenker and Lobell, 2010; Moore et al., 2012; Arndt et al., 2012). 1.3 Regional agriculture East Africa grows a number of crops including ones used mainly for income generation such as coffee, cotton, tea, sugarcane and various food crops. The six most important food crops feeding the region include cassava, maize, plantains, sweet potatoes , potatoes and paddy rice (FAO, 2013). Figure 1. 3 s hows that crop production for all the six major food crops has been increasing over the past 3 decades with cassava and maize yields being always above other food crops in the region. 7 Figure 1. 3 Top 8 crops in East Africa On a country level, sweet potatoes are among the top four food crops grown in each of the three East African countries used in this research as shown in Figure 1. 4. In Uganda, the top four most important food crops from highest to low importance are plant ains, cassava, sweet potatoes and maize; in Kenya , they are maize, potatoes, sweet potatoes , and cassava; and in Tanzania, they are cassava, maize, sweet potatoes and paddy rice. Out of the three countries, Uganda has the highest annual sweet potato produc tion of more than 2,000 tonnes (Figure 4a) while Kenya has the least production of about 500 tonnes (Figure 4b). In terms of annual trends, Tanzania showed the highest rise in production from the year 2000 to 2010 with the present production levels even re aching and surpassing those of Uganda (Figure 4d). 0 5 10 15 20 25 30 35 40 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 Production (Megatons) Year Cassava Maize Plantains Sweet potatoes Potatoes Rice, paddy 8 (a) (b) (c) (d) Figure 1. 4 Average annual production for the top eight staple crops in East Africa, Source: FAO, 2013 Besides being one of the four most important s tapes in the region sweet potatoes plays several significant roles in the agricultural systems in East Africa. It is a crop grown by resource - poor farmers, it is very nutritious, richer in vitamin A than cassava, maize and plantain (Mwanga et al., 200 3) , it can sustain families up to six months through piecemeal harvesting, it can be used 0 2000000 4000000 6000000 8000000 10000000 Av. annual production (tonnes) Food crops Uganda 1991-2000 2001-2010 0 2000000 4000000 6000000 8000000 10000000 Av. annual production (tonnes) Food crops Kenya 1991-2000 2001-2010 0 2000000 4000000 6000000 8000000 10000000 Av. annual production (tonnes) Food crops Tanzania 1991-2000 2001-2010 0 2000000 4000000 6000000 8000000 10000000 1980 1984 1988 1992 1996 2000 2004 2008 2012 Production (tonnes) Year Sweet potato production in East African countries Uganda Tanzania Kenya 9 as both food and animal feed (Bashaasha, Mwanga, p'Obwoya, & Ewell, 1995) , it requires relatively little soil nutrients (Slathers et al., 2005) and it is a drought and heat tolerant crop (Gomes & Carr, 2003) . 1. 4 Climate requirements for sweet potatoes (Togari, 1950) and (Villordon, LaBonte, Solis, & Firon, 2012) emphasized that the early - season (first 20 days) growing environment has a direct and significant impact on storage root initiation and thus final yield. Temperature stress is one of the most crucial limitations on crop growth and development (Wrigley, 1994) and causes irreversible damages to the plant processes and thus affecting final yield. Temperature and CO 2 levels can significantly affect sweet potato growth (Loretan et al., 1994) . Higher storage root weights and numb ers are favored by growing sweet potatoes in varied temperatures and thermoperiods such as growth in a sand culture at 29 0 C (16 - h light period) and 20 0 C (dark) compared to those grown at a constant temperature of 29°C (same photoperiod, (Kim, 1961; Loretan et al., 1994) . In a study by (Spence & Humphries, 1972) , working with rooted single leaves of sweet potatoes , obtained optimum storage root formation and development was obtained when t he soil temperature was 25 0 C, whereas soil temperatures of 15 and 350C were inhibitory to storage root formation. Growth was reduced 50% with a temperature reduction of from 21.1 to 15.6°C (Jarret & Gawel, 1991) . Reduction of photoperi od from 16 to 4 h produced smaller, slightly chlorotic, but otherwise normal plants (Jarret & Gawel, 1991) . (Gajanayake, Reddy, Shankle, Arancibia, & Villordon, 2014) is the first study that quantified the functional relationships of root initiation, growth, development, and biomass partitioning of sweet potato in response to a wide range of temperatures, where the crop is grown 10 today and expected to grow in future under projected changes in climate. They found that maximum total root numbers produced increased with temperature from 20/12 to 30/22°C, but differed only between the two extreme temperature treatments (20/12 0 C and 40/32°C). When the temperature increased from 17.4 to 36.4°C, time to 50% total root initiation was reduced by 4.5 days. Having a well - developed root system during the early growth stages of the crop was extremely crucial for both development of the shoot system as well as for the process of storage root development of sweet potato . But this is study looked at th e first 59 days after planting. Elevated CO 2 concentrations were also reported to increase sweet potato storage root yields possibly as a result of a shift in the distribution of photosynthate from the leaves to the storage roots (J. R. Allen, Bhattacharya, Lu, Pace, & Rogers, 1985; Bhattacharya, Biswas, Battacharya, Sionit, & Strain, 1985) . The cessation of root elongation, primary and secondary cambia growth, an increase of radial growth or bulking by an increased rate of cell division, cell proliferation, and cell expansion is associated with the deposition of starch (Ravi, Nascar, Makeshkumar, & Binoy Babu, 2009) . Among environmental factors, high air temperature causes reduction in sto rage root formation and growth through the changes in phytohormones synthesis and activation (Ravi & P., 1999) and dry matter partitioning and bulking of storage roots in sweet potato and tubers in potato (Solanum tuborosum L.) (Van Dam, Kooman, & C., 1996) . Greater yields depend on an early development of source (leaf area) for optimum light interception and sink (both initiation and enlargement of storage roots) in sweet potato . Apart from the sink organs, enhancement of source organ functions and capacity are crucial to increas ing sweet potato yield. To achieve high dry matter production through photosy nthesis process, it is important to develop an optimum leaf area. Also, the photosynthesis in sweet potato is sensitive to 11 elevated temperature (>35°C). According to Indira and Kaberathumma (1988), three important physiological events in the growth phase o f sweet potato are responsible for final crop productivity, namely storage root initiation, storage root development, and storage root maturity. Both air and soil temperature regulate the competition between shoot and storage root growth in sweet potato (Gajanayake et al., 2014; Ravi et al., 2009) . 1.5 Characteristics and adaptive potential of sweet potato cultivars in Uganda 1. 5 .1 Introduction Due to the high adaptive potential of sweet potatoes, Africa is among the top producers with the highest pr oduction coming from Uganda in East Africa (Bashaasha et al., 1995) . Even with high production and high nutritious values of sweet potatoes in Uganda, malnutrition problems still do exist in the country. For example, 2 0% of children and 19% of women have vitamin A deficiencies (Opinion Research Corporation Macro International Inc (ORC Macro), 2006) . Vitamin A deficiency (VAD) has negative effects such as increased exposure of children to common illness, stunted growth, development, vision and reduce d immune systems (Tumwegamire et al., 2007) and VAD claims between 15,000 to 60,000 lives annually (Ruel, 2001) . These statistics paint a picture that something is probably wrong. One possibility could be that the crops produced and consumed, espe cially sweet potato varieties may not be rich enough in vitamins, or that vitamin - rich crops are produced and consumed during a particular season leaving families to live without any source of vitamins for the rest of the time. There is , therefore, a need for all stakeholders to devise means to address this problem, one way of which is to have a collection of the crop characteristics and then identify ways to help families have a quality diet across the year. 12 The aim of this paper is to integrate and ident ify the most productive sweet potato cultivars from existing literature using characteristics including maturity period, yield, root quality, resistance to pests and diseases, root flesh color, and abundance. Having a collection of information about the sw eet potato varieties that have been researched on is valuable to researchers interested in crop breeding, modeling and pathology studies. The paper begins by providing an overview of the sweet potato farming systems in Uganda and a detailed analysis of the sweet potato cultivars based on published work and the organizations that help in the dissemination of planting materials and any other relevant information to the farmers. The paper presents the major constraints experienced by farmers and finally, highl ights the role of participatory and adaptive research in developing suitable crop varieties and general sweet potato production systems in Africa. 1. 5 .2 Sweet potato farming systems in Uganda Farming systems of sweet potato in Uganda vary from mono - crop ping to mixed cropping systems. Sweet potato is either mono - cropped or inter or relay cropped with other crops such as legumes e.g., beans, or cereals like maize, millet , and sorghum (Ewell and Mutuura 1994: Bashaasha et al 1995). This enables households e specially in rural areas to have a constant food supply and various food options during the year. Sweet potato is normally grown in crop rotations which are decided upon differently by farmers in the region in Uganda and more importantly, according to the resources and priorities of a given household (Bashaasha et al., 1995). The crop rotations of sweet potato potatoes for the different ecological z ones in Uganda are provided in T able 1 .1 . It is mainly grown by women in small householder farming systems fo r food and income generation (Bashaasa et al., 1995; (R. W. Gibson, Mwanga, Namanda, Jeremiah, & Barker, 200 9) . 13 Table 1. 1 Crop rotations of sw eet potatoes in Uganda. Source: Bashaasha et al., 1995. Region Crop rotation High latitude zones (e.g., Kabale) fallow > sorghum > sweet potato > beans/maize fallow > peas > sweet p otato > sorghum > irish potato (e.g., Mbale) cotton > millet > sweet potato /cassava Pastoral dry to semi - arid rangeland zone fallow > maize/beans > cassava/millet > sweet potato > beans/maize or millet > cotton > cassava > fallow Northern and eastern short - grassland zone fallow > cotton > millet > sesame > cassava/fallow or fallow > sesame > cassava > sweet potato > maize/fallow Southern and western tall - grassland zone Fallow > sweet potato > maize/bean >millet > cassava > sweet potato or Fallow > maize/beans > sweet potato > cassava > millet The main source of planting materials for growing sweet potatoes by households is by use of domestic wastewater to grow vines and planting sweet potato roots in a nursery, watering them for4 - 6 and then transplanted to actual plots. Sweet potato is planted on ridges, mounds of about 0.5m high and occupying an area of about 1 square meters (Gibson et al., 2008) or on flat ground. It is planted following the bimodal part of the rainfall between March - M ay and October December (Ewell and Mutuura 1994; Bashaasha et al 1995). Planting materials are mainly vines and sometimes sizeable roots accumulated by households using irrigation and domestic wastewater (Gibson et al 2009). In drier areas, planting mate rials are conserved by planting vines in areas with available water such as wetlands, areas near water holes, and sometimes by physical watering in the backyard (Gibson et al., 2009). Various sweet potato varieties with varying maturation times are grown on the same plot during the same season and sometimes at different times in order to allow for a long time of harvest. Most subsistence households practice in - ground storage and piecemeal harvest whereby only enough roots are harvested (Ebregt, Struik, Odongo, & Abidin, 2007; Hall, Bocket, & Nahdy, 1998; Smit, 1997) . This form of staggering planting and planting of different varieties enable the 14 households to go through dry periods by having mature roots stored in the ground for up to six months and piecemeal harvesting. However, the sweet potato production still undergoes a number of challenges and constraints that have led to low yields in Uganda. 1. 5 .3 Sweet potato varieties in Uganda The choice and preference of sweet potato varieties by farmers differ from individual to individual. Farmers use characteristics such as yield, time to reach maturity, root color, root size, root shape, root quality, sweetness, pest and disease resistance, and marketability (Basha asha et al. 1995). There has been a considerable amount of research on the characteristics of sweet potato varieties grown in Uganda. Some studies provide detailed information about the characteristics of varieties while others provide much less informatio n. The varieties are either local landraces existing within communities or officially released landraces got from other communities and distributed across the country, and others are bred and tested at the National Crops Resources Research Institute (NaCRR I) in Namulonge and later released to communities. In this paper, we review ed the literature and collect characteristics on all th e recorded varieties and we made comparisons of the varieties. The synthesize d information from this review i s useful for rese archers and other stakeholders interested in understanding sweet potato product ion in Uganda. This work focused on varieties that had been published in articles but was noted that there we re many more varieties existing in Uganda whose characteristics have not yet been documented. The varieties we re presented under two categories; non - orange fleshed sweet potato varieties (Table 1 .2 ) and orange - fleshed sweet potato (OFSP) varieties (Table 1.3 ). Of the non - OFSP, the released varieties include: Tanzania, Bw anjule, New Kawogo, Tororo 3, Wagabolige, and Sowola (Mwanga et al., 2001); NASPOT 1, NASPOT 2, NASPOT 3, NASPOT 4, and 15 NASPOT 6 (Mwanga et al., 2003); NASPOT 11 (Gibson et al. 2011); and, Vita A and Kabode (Namanda et al., 2011). It is not very clear from the literature whether the last two varieties are non - OFSP or OFSP. The released OFSP include NASPOT 5 (Mwanga et al., 2003); Ejumula and Kakamega or SPK 004 (Mwanga et al., 2005); and NASPOT 7, NASPOT 8, NASPOT NASPOT 9). Table 1. 2 Non Orange varieties. Cultivar Root Flesh Color References Dimbuka Cream Mwanga et al. 2005; Yanggen and Nagujja 2006; Mwanga et al 2007a; Gibson et al. 2011; Tumwegamire et al. 2011b Dimbuka - Bukulula Cream Mwanga et al. 2009 Kyebandula Cream Bashaasha et al. 1995; Gibson et al. 1997; Hall et al. 1998; Mwanga et al. 2003a; Tumwegamire et al. 2011a Old Kawogo Cream Bashaasha et al. 1995; Gibson et al. 1997; Mwanga et al. 2003a; Magabi Bashaasha et al. 1 995; Sukali Bashaasha et al. 1995; Bitambi Cream Bashaasha et al. 1995; Gibson et al. 1997; Mwanga et al. 2003a; Tumwegamire et al. 20011a; Tumwegamire et al. 2011b Tanzania - RLr Light Orange/ Pale Yellow Bashaasha et al. 1995; Gibson et al., 1997; Mwanga et al. 2001; Mwanga et al. 2003a: Mwanga et al. 2005; Yanggen and Nagujja 2006; Gibson et al. 2009; Mwanga et al. 2009; Tumwegamire et al. 2011a; Tumwegamire et al. 2011b Mukazi Cream Hall et al. 1998; Tumwegamire et al. 2011a; Tumwegamire et al. 2 011b Araka Cream Mwanga et al. 2005; Namanda et al. 2011; Tumwegamire et al. 2011a Bwanjule - RLr White Mwanga et al. 2001; Mwanga et al. 2003a; Mwanga et al. 2009; Tumwegamire et al. 2011a 16 Cultivar Root Flesh Color References N ew Kawogo - RLr White/Cream Gibson et al. 1997; Mwanga et al. 2001; Mwanga et al. 2003a; Mwanga et al. 2005; Yanggen and Nagujja 2006; Gibson et al. 2009; Mwanga et al. 2009; Gibson et al. 2011; Tumwegamire et al. 2011a; Tumwegamire et al. 2011b Tororo - 3 - RL r Cream Mwanga et al. 2001; Mwanga et al. 2003a; Tumwegamire et al. 2011a; Tumwegamire et al. 2011b Wagabolige - RLr White/Cream Mwanga et al. 2001; Mwanga et al. 2003a; Tumwegamire et al. 2011a Sowola - RV Cream Mwanga et al. 2001; Mwanga et al. 2003a; Mwa nga et al 2005; Mwanga et al. 2007a; Mwanga et al. 2009; Tumwegamire et al. 2011b Bunduguza Light Yellow Mwanga et al 2005; Yanggen and Nagujja 2006; Mwanga et al. 2009; Tumwegamire et al. 2011a; Tumwegamire et al. 2011b NASPOT 1 - RV Pale Yellow Mwanga et al. 2003a; Gibson et al. 2008; Mwanga et al. 2009; Gibson et al. 2011 NASPOT 2 - RV Cream Mwanga et al. 2003a; Gibson et al. 2008; Mwanga et al. 2009 NASPOT 3 - RV Cream Mwanga et al. 2003a; Gibson et al. 2008; Mwanga et al. 2009 NASPOT 4 - RV Pale Yel low Mwanga et al. 2003a; Gibson et al. 2008; Mwanga et al. 2009 NASPOT 6 - RV White Mwanga et al. 2003a; NASPOT 11 - RV Gibson et al., 2011 Vita A - RV Namanda et al. 2011 Kabode - RV Namanda et al. 2011 Araka - lr White Mwanga et al 2005; Mwang a et al. 2009; Namananda et al. 2011; Tumwegamire et al. 2011b Osukut - lr Namanda et al. 2011 Silk Rwanubende Hall et al. 1998 Yosefu Hall et al. 1998 Muguma Hall et al. 1998 Kahungezi Hall et al. 1998 Kalebe Cream Hall et al. 1998; Y anggen and Nagujja 2006; Tumwegamire et al. 2011a; Tumwegamire et al. 2011b 17 Table Cultivar Root Flesh Color References Osapat Yellow Mwanga et al. 2005; Yanggen and Nagujja 2006; Tumwegamire et al. 2011a; Tumwegamire et al. 2011b; Yada et al. 2011 Kassim Mwanga et al. 2005 Kampala Mwanga et al. 2005; Gibson et al. 2008; Tumwegamire et al. 2011b Kenya Mwanga et al. 2005; Yanggen and Nagujja 2006; Koromojo Mwanga et al. 2005 Otada Mwanga et al. 2005; Tumwegamire et al. 20 11a; Tumwegamire et al. 2011b Liralira Mwanga et al. 2005 Table 1. 3 Orange - fleshed varieties Cultivar Root Flesh Color References NASPOT 5 - RV Orange Mwanga et al. 2003a; Mwanga et al. 2005; Mwanga et al. 2009 NASPOT 7 - RV Intermediate Orange Mwanga et al. 2009 NASPOT 8 - RV Pale Orange Mwanga et al. 2009 NASPOT 9 O - RV Intermediate Orange Mwanga et al. 2009 NASPOT 10 O - RV Dark Orange Mwanga et al. 2009 Ejumula - RLr Orange Mwanga et al. 2005; Yanggen and Nagujja 200 6; Mwanga et al. 2007a; Mwanga et al. 2007b; Namanda et al. 2011; Tumwegamire et al. 2011a; Tumwegamire et al. 2011b Kakamega (SPK 004) Orange Mwanga et al 2005; Yanggen and Nagujja 2006; Mwanga et al. 2007a; Mwanga et al. 2007b; Namanda et al. 2011; Tumw egamire et al. 2011a Abuket 1 Orange Gichuki et al. 2005; Tumwegamire et al. 2011a; Tumwegamire et al. 2011b Abuket 2 Light Orange Gichuki et al. 2005; Tumwegamire et al. 2011a; Tumwegamire et al. 2011b Kala Orange Mwanga et al 2005; Yanggen and Nagujja 2006; Mwanga et al. 2007a; Mwanga et al. 2009; Tumwegamire et al. 2011a; Yada et al. 2011 Edule Orange Yanggen and Nagujja 2006; Gweri Orange Gichuki et al. 2005 18 1.5 .3.1 Maturity An analysis of information in Table 3 from the literature shows that So wola takes 100 - 120 days to reach maturity while others like Tanzania, Dimbuka - Bukulula, Bwanjule, Dimbuka, Tororo - 3, Wagabolige, NASPOT 1, NASPOT 2 and NASPOT 6 take between 120 - 130 days to mature. Mukazi and Kyebandula were also ranked among the early mat uring varieties in a study carried out by Hall et al . ( 1998) but no specific maturity date was provided. For the OFSP in Table 1. 4, on the other hand, NASPOT 10 O and NASPOT 7 take 110 days and 115 days respectively to reach maturity while NASPOT 8, Ejumul a, Kakamega, NASPOT 5 and NASPOT 9 take between 120 - 150 days. 1. 5 .3.2 Yield In terms of yield, non - OFSP varieties in Table 3 ga ve the highest yield with an average yield of 29t/ha for NASPOT 1, Sowola (25.6 t/ha), Dimbuka - Bukulula and NASPOT 3 (25t/ha), NASPOT 6, Wagabolige and Araka (23.9). A most recently released variety NASPOT 11 was also reported to have yielded as much as NASPOT 1(Gibson et al., 2011). The rest of the non - OFSP varieties had below 21t/ha. The OFSP varieties with the highest yield w er e NASPOT 5 with 23t/ha and NASPOT 7 with 20.4t/ha. The root dry matter content for most sweet potato varieties is over 30% but Mukazi has the highest value of 35.8%, followed by NASPOT 3 with 35%, Sowola, and NASPOT 6 with 34% (Table 3). For the OFSP va rieties, the highest root dry matter content is given by Kakamega with 33.3%. This is a very important ratio for breeders to understand the proportion of biomass partitioning for a sweet potato plant. A high ratio shows that a good proportion of nutrients are used by the plant in developing roots. 19 Table 1. 4 Characteristics of selected non - orange fleshed varieties Cultivar Maturity Av. Yield (kg/ha) Root Quality Abundan ce Resistance to pests & diseases Root dry matter content (%) Taste sweetness - Carotene (µg/100g) Weevil s SPVD Alternaria Stem Blight Mukazi Early 8.1 35.8 not recorded High Sowola 100 - 120 25.6 34 moderate HS MR Tanzania 120 22.9 32 moderate High S MR MR Dimbuka - Bukulula 128 25 32.4 moderate 13.3 - 24.1 S S MR Kyebandula Early high high not recorded High Bwanjule 120 - 150 21.4 30 moderate MR R Dimbuka 120 - 150 19.7 31.5 moderate 24 - 32 S S MR Tororo - 3 120 - 130 18 31 moderate MR MR Wagabolige 120 - 150 24.1 33 moderate R NASPOT 1 120 - 150 29 33 sweet High S MR NASPOT 2 120 - 150 21 29 sweet S R NASPOT 6 120 - 150 24 34 moderate MR MR New Kawogo 130 - 150 23.3 33 moderate High MR HR HS NASPOT 3 130 - 150 25 35 sweet MR R NASPOT 4 130 - 150 21 33 sweet MR R Araka 23.9 29 R NASPOT 11 ~28 S MR Bunduguza ~23 31.5 High HS High Susceptibility, S Susceptibility, MR Moderately Resistant, R Resistant, HR High Resistance 20 1. 5 .3.3 Taste Taste is a very important fa ctor used by consumers of sweet potatoes in Uganda. A variety with a sweet taste is more preferred to one with less sweet taste. For the non - OFSP, the released varieties NASPOT 1, NASPOT 2, NASPOT 3 and NASPOT 4 have sweet tastes and therefore may be consu med more than the local landraces (see Table 1.4 ). NASPOT 5 in Table 4 has a sweeter taste than the rest of the OFSP varieties. 1. 5 .3.4 Abundance Abundance is also an important factor used by farmers to determine the varieties they plan on planting in a given season. The more abundant varieties vary for different regions in the country. However, the most common varieties across the country include Mukazi, NASPOT 1, Kyebandula, New Kawogo, Bitambi , Tanzania and Bunduguza ( Bashaasha et al 1995; Hall et al., 1998; Mwanga et al. 2001; Mwanga et al 2003a; Bua et al., 2005). The most common local varieties per region from highest to low abundance include: Central Dimbuka, Kalebe , New Kawogo, Silk, Munyera , Kyebandula; Eastern Bunduguz a, Araka, Kigayire, Sil k, Tanzania; Northern Liralira, Koromojo, Nyakenya, Ombivu, Nyaromayo; Western Kyebandula, Kahogo, Mugumira, Kyokyokyemba (Bua et al 2005). Tanzania is a regional variety as it is widely grown in East African countries and locally known by different na mes. For example, it is called SPN/O in Tanzania, Enaironi in Kenya, Chingovwa in Zambia and Kenya in Malawi (Mwanga et al 2001). 21 1.5 .3.5 Resistance to pests and diseases The literature considers mainly three categories; weevils, sweet potato virus d isease (SPVD) and A lternaria stem blight (see Tables 1. 3 and 1. 4). Bwanjule, Tororo - 3, NASPOT 6, New Kawogo, NASPOT 3 and NASPOT 4 and NASPOT 5 are moderately resistant to sweet potato weevils. Most orange - fleshed such as sweet potato (OFSP) varieties, and Sowola, Tanzania, Tororo - 3, NASPOT 1 and NASPOT 6 are moderately resistant to SPVD. New Kawogo was reported to have a high resistance and Bwanjule, Wagabolige and NASPOT 2 are also resistant to SPVD. All OFSP varieties, some non - OFSP varieties such as Tan zania, Dimbuka - Bukulula , and Dimbuka are moderately resistant to a lternaria stem blight. Therefore, successful production of a variety in an area should put into consideration the nature and type of pests and diseases present in the area. Based on all the defined characteristics in the preceding sections, four cultivars were selected for the later part of this study (Table 1.5). That is, for crop - model development and for assessment of the impact of climate change on sweet potato production in East Africa. 22 Table 1. 5 Characteristics of selected orange - fleshed varieties (abundance not known) Cultivar Maturity (days) Av. Yield (t/ha) Root Quality Resistance to pests & diseases Dry matter content (%) Taste sweetness - Carotene (µg/100g) Weevils SPVD Alternaria Stem Blight NASPOT 10 O 110 16 Moderate 185.6 - 342.8 S MR MR NASPOT 7 115 20.4 Moderate 44.3 - 192.7 S MR MR NASPOT 8 120 17.8 32 Moderate 85.6 - 219.3 S MR MR Ejumula 120 - 150 18.8 30.1 Moderate 206.3 S S MR Kakamega 120 - 150 14.9 33.3 Moderate 376 - 3760 S MR MR NASPOT 5 120 - 150 23 30 Sweet MR MR NASPOT 9 125 16.5 30.1 Moderate 206.3 - 460.3 S MR MR HS High Susceptibility, S Susceptibility, MR Moderately Resistant, R Resistant, HR High Resist ance 23 1. 5 .4 Farmer support and dissemination of planting materials The importance of this section is to highlight the organizations playing a significant role in the distribution of planting materials for the different varieties and providing general suppo rt to sweet potato farmers in Uganda. The organization is distributed across the country, see Table 1.6 . The organizations are government, not for profit, or community - based organizations (CBO). The organizations help in the multiplication and disseminati on of new and improved sweet potato varieties, promotion of orange - fleshed sweet potato to avert vitamin A deficiency. They also help in sensitization of farmers and all stakeholders about the importance of growing and consuming orange fleshed varieties th at are rich in vitamin A. Examples of such institutions include the ministry of Health in collaboration with Volunteer Efforts for Development Concerns (VEDCO) and the National Agricultural Research Organization (NARO) successfully conducted sensitization s in Luwero district in central Uganda in 2001 (VEDCO, 2001) and Save the Child (NGO), Makerere Universit y, Department of Agricultural Extension, and World Vision sensitized stakeholders (Mwanga, Stevenson, & Yencho, 2005) . 24 Table 1. 6 Organizations helping farmers in sweet potato production Region Name of organization Type of Organization (NGO/CBO) Reference Central Bu ganda Cultural and Development Foundation (BUCADEF) CBO Yanggen and Nagujja 2006; Gibson et al. 2008; Gibson et al. 2009 Central Volunteer Effort for Development Concerns (VEDCO) CBO VEDCO 2001 Central Tusitukire wamu Kabulanaka Farmers' Association (TUK AFA) Gibson et al. 2011 Central Rakai District Farmer's Association (RADFA) CBO Yanggen and Nagujja 2006 Central Community Enterprise Development Organization (CEDO) Yanggen and Nagujja 2006 Central Adventist Development and Relief Agency (ADRA) NGO Yanggen and Nagujja 2006 Central Concerned Women (COWO) NGO Yanggen and Nagujja 2006 Central Masaka District Development Organization (MADDO) NGO Yanggen and Nagujja 2006 Save the Child NGO Mwanga et al. 2005 Country - wide National Agricultural Resea rch Organization NGO Gibson et al. 2008; Gibson et al. 2009 East Soroti Sweet potato Producers Association (SOSPPA) CBO Gibson et al. 2009 East Soroti Cathoric Diocese Development Organization (SOCADIDO) NGO SOCADIDO 2001 Western Hoima District Farmers Association ( HODIFA ) CBO Yanggen and Nagujja 2006 Western Sub - county Offices Government Yanggen and Nagujja 2006 Western District Agricultural Office Government Yanggen and Nagujja 2006 Western Africare NGO Yanggen and Nagujja 2006 Northern James Arwat a Foundation (JAF) NGO Yanggen and Nagujja 2006 Country - wide Makerere University Government Mwanga et al. 2005 Country - wide World Vision NGO Mwanga et al. 2005 (Mwanga et al., 2005) 25 1.5 . 5 Constraints to sweet potato production in Uganda Sweet potato production is affected by a number of constraints that compr omise its potential in being very productive. The average yield of sweet potato (4t/ha) is much lower than the potential yield of 25t/ha. Various constraints responsible for the low production include ; pest and diseases, drought, vine scarcity, lack of ca pital, high labor requirements, poor yields, low prices, animal destroy the crop , lack of land, poor markets, and crop rotting (Yanggen & Nagujja, 2006) . The major disease constraints affecting s weet potatoes include the sweet potato virus disease (SPVD), weevil, and physiological lack of vigor (R. W. Gibson et al., 2009) . Major sweet potato pests include sweet potato weevil ( Cylas punticolis , C.brunneus, and C. formicarius ), butterflies, mole rats, other rodents and wild beasts (Bashaasha et al., 1995; Ewell & Mutuura, 1994; R. W. Gibson et al. , 2009) . Intercropping sweet potato with corn, soybean and corn + soybean reduces the damage of sweet potato weevil (Yaku, Hill, & Chiasson, 1992) while pesticides and fertilizer application, and rouging of the diseased plants (Bashaasha et al., 1995; R. W. Gibson et al., 2009) are the main ways to regulate virus diseases of the crop. Like in most developing countries, fertilizer app lication in sweet potato farming systems in Uganda is very low leading to poor yields (Anon, 1993) . However, even in instances of low farm inputs, some sweet potato varieties give a higher yield than others. 1.5 . 6 Participatory and adaptive research on sweet potato production Participatory breeding and adaptive research ha ve unique contributions to make to the constraints faced by sweet pot ato smallholder farmers in Uganda and African countries in general. Farmers, community - based and non - profit organizations, and agricultural extension have been 26 instrumental in enhancing production and research in sweet potato agricultural systems in Uganda . Some literature has highlighted the involvement of farmers in participatory plant breeding and participatory variety selection (Richard. W. Gibson, Byamukama, Mpembe, Kayongo, & Mwanga, 2008; Richard. W. Gibson, M pembe, & Mwanga, 2011) with positive results in terms of the final product from the research and also from the farmers experience and fulfillment in contributing towards the whole research process. There is a wide body of literature on participatory rese arch approach but only a few selected articles are used in this section to briefly orient the reader. Then a few success stories for sweet potatoes and other crops are presented to highlight the importance of this approach in enhancing agricultural product ion. In an agricultural context, participatory research can be functional oriented and empowering involving farmers (Okali et al., 1994) and other stakeholders such as extension officers, non - governmental organizations , and scientists. Participatory rese arch improves crops and genetic diversity, the efficiency of the research services in identifying adaptive technologies, and empowers rural communities to influence the agendas of and to benefit from the knowledge in formal research ( Sutherland 1998; Morri s and Bellon, 2004; Humphries et al., 2005 ). In designing a model of participatory research for bean improvement, Bulter et al (1994) observed that f needs on their f arms, and whether they will use the new variety or not. Therefore, in order to achieve the most of out of participatory research, the research should be collaborative (Sperling et al., 1993; Bentley, 1994; Sutherland 1998; Witcombe et al., 2005a), contract ual, consultative and collegiate (Sutherland, 1998), decentralized (Ashby & Sperling, 1994; Berg, 1997; Morris & Bellon, 2004) , and should lead to a specific local adaptation and intra - varietal diversity (Berg, 1997) . 27 Participatory plant breeding (PPB) and participatory varietal selection (PVS) has been widely used on various crops in Africa. Scientists and farmers used PVS to identify preferred sweet potato varieties e.g., (Richard. W. Gibson et al., 2008; Kapinga et al., 1998) , and PPB to lead to the official release of a new sweet potato variety called NASPOT 11 (Richard. W. Gibson et al., 2011) . More details on PPB and PVS can be found in (Sperling L, Ashby, Smith, Weltzien, & McGuire, 2001; Sperling L, M. E. Loevinsohn, & Ntabomvuras, 1993; Witcombe, Gyawa li, Sunwar, Sthapit, & Joshi, 2005) . PPB was also successfully used in bush beans in Rwanda (Sperling & Scheidegger, 1996) and in Hondur a s (Humphries, O. Gallardo, J. Jimene z, & F. Sierra, 2005) , cassava production in Tanzania (De Waal, F. R. Chinjinga, L. Johansson, F. F. Kanju, & Nathaniels., 1997) and in Ghana to develop superior cassava cultivars (Manu - Aduening et al., 2006) . On - farm participatory research has also been used in improving s oil organic matter and soil nutrient management in southern Africa (Kanyama - Phiri, Snapp, Kamanga, & Wellard, 2000; Kerr, Snapp, Chiwa, Shumba, & Msachi, 2007; Snapp, Mafongoya, & Waddington, 1998; Snapp, Rohrbach, Simtowe, & Freeman, 2002) . In most of these examples, the involvement of farmers in research helps scientists to consider other factors such as taste, color, and farmer preferences that would not have otherwise been considered in formal plant breeding. Therefore, participatory research should be encouraged and used more often to help in advancing adapting technologies in smallholder farmers. This research approach will play a significant role in helping communities to adapt to the impacts of climate chan ge on agriculture , especially for the most vulnerable communities. However, participatory research has some challenges such as high costs, some attributes such as taste and color may be hard to measure, and sometimes there is a need for additional training by scientists such as learning of farmers languages (Morris & Bellon, 2004) . Some of these challenges can be addressed by working with scientists 28 within the area who may know the local language and caref ully planning the research process before actual implementation. 1.5 .7 Conclusion There is still a need to conduct more research on sweet potatoes in Uganda and in East Africa in general. This will not only help in reducing malnutrition problems experien ced in the countries, but it will also prepare the country against the impacts of the projected changes in climate which are likely to affect many crops. From the characteristics of varieties discussed above, it is important to note that no single characte ristic can be used to determine the best variety to select although such knowledge is useful in determining the potential and impact one variety might have over the other. For example, some people may prefer varieties with a sweet taste while others may be interested in early maturing varieties. Under such scenarios, the use of participatory research approaches is important in coming up with the best option. Equally important, is the involvement of governmental and community - based organizations, and researc h institutions that help in supporting the farmers. All the success stories of released varieties in Uganda have been attained as a result of a joint effort from all organizations. Finally, as more growth characteristics become available for different var ieties of sweet potato , there is need to carry out sweet potato modeling to assess the impact of changes in environmental conditions especially climate and management practices on sweet potato production. The information to be generated from modeling work will be highly valuable especially to the sweet potato breeders who normally take over 10 years of crop breeding before a new variety that can perform well in the changing environmental conditions is released. 29 1.6 Research gaps and limitations This study made an attempt to address the following research gaps. Agronomic data for sweet potatoes for the whole growing process from planting to harvesting was scarce. Therefore, experiments were set up in two growing seasons for four identified cultivars, two or ange - fleshed (SPK 004 and NASPOT 10) and two non - orange fleshed cultivars ( NASPOT 1 and NASPOT 11) in Uganda and the existing historical sweet potato data was supplemented with one collected from field plots. Relevant data for climate impact assessments ar e scarce: detailed agronomic data for sweet potato cultivars grown in East Africa are limited; representative high - quality climate data for the region are scarce , and soils data is only available at coarse spatial resolution. suitable climate data source w as identified to cater for the poor climate data characterized by missing records and soils data were also collected from both field plots and from existing historical soils databases. A number of gridded satellite climate data products were compared with the observed climate records which could be found in the three East African countries, Uganda, Kenya , and Tanzania. Deterministic simulation models for sweet potatoes exist but are relatively young or still in development. A process - based crop model for sw eet potatoes was non - existent and yet it was the suitable research tool required for this study. This study , therefore , modified, calibrated and used an existing crop model, with the consultation of the original model developer from India. Relatively litt le is known about how climate influences sweet potato growth, development, and yield. 30 1. 7 Research questions and objectives The major objective of the research was to assess the impact of climate variability and change on sweet potato production in Eas t Africa. The study addressed the following research questions and their corresponding objectives. Question 1. What are major climatic constraints, currently and in the recent past, to sweet potato production in East Africa? Objective 1: Develop a modelin g framework for use in a deterministic sweet potato crop model, SPOTCOMS, for East Africa . Objective 2: Develop a historical climate and soils database for East Af rica for the period 1980 2009 . Objective 3: With the SPOTCOMS crop model, identify the maj or climatic constraints for sweet potato production in East Africa . Question 2. How might sweet potatoes be impacted by projected future changes in climate in East Africa? Objective 4: Develop local climate change scenarios for East Africa for near - term fu ture (2041 - 2070) and distant future (2071 - 2100) periods Objective 5: Estimate the impact of projected future climate change on sweet potato production in the region Question 3. Which areas are (historically and in the future) most suitable for sweet potat o production in East Africa? Objective 6: Identify areas most suitable for sweet potato production in East Africa using historical and future climate data 31 CHAPTER 2 APPLICATION OF A PRO CESS - BASED MODEL FO R SWEET POTATO GROWTH DEVELOPMENT IN EAST AFRICA 2.1 Introduction As one of the ten most important staples that the feed the world (FAO, 2014) , sweet potato has not yet made significant progress in crop modeling research compared to maize, rice, wheat, and potatoes. Unlike most staples, however, sweet potato , especially the orange - fleshed cultivars, are very rich in vitamin A, and some have high amounts of calcium, iron , and zinc (Tumwegamire et al. 2011), stores well in soil as a famine reserve crop, is drought tolerant (Gomes & Carr, 2003) , grows well in low - nutrient soils and can sustains families up to 6 months on a piece - meal harvesting. I n areas with declining land availability, sweet potato is a valuable crop due to its relatively high production per unit area, multipurpose functions as both food and animal feed, and low input requirements (Bashaasha et al., 1995; Bovell - Benjamin, 2007; Diop, 1998; Jiang, Jianjun, & Wang, 2004) . S weet potato is a lso valuable crop in areas with declining land availability due to its relatively high production per unit area and low input requirements (Bashaasha et al., 1995; Bovell - Benjamin, 2007; Diop, 1998; Jiang et al., 2004) . East Africa was selected for this study owing to the high level of importance that sweet potato has in that region. Almost every commu nity grows sweet potato for either home consumption only or for both home consumption and income generation. The household communities grow ing who grow sweet potatoes are l argely the rural poor and therefore any 32 intervention of improving production whether coming from government or research would be of great value to the people. There is limited application of fertilizer to the largely rain - fed sweet potato growing in East Africa. Usually , with limitations in land size, sweet potatoes are normally intercrop ped with beans and sometimes maize, but for better yields, agricultural systems in the crop are mono - cropped tend to do much better. The nutrition value and high production from sweet potato would help in i mproving the health and ensure food security amon g the communities. One approach to perform ing such evaluations is to employ crop simulation models which provide researchers with an advantage of a more controlled assessment of weather and climate, soils, and/or other crop variables than field experime nts. Existing sweet potato models include Sweet potato COMputer Simulation (SPOTCOMS, (Mithra & Somasundaram, 2008) a process - based model, CLICROP (Arndt, Farmer, Strzepek, & Thurlow, 2012) an empirical/statistical regression - based model, and an International model for policy analysis of agricultural commodities and trade (IMPACT) (Rosegrant et al., 2008) an economic - based model. Process - based models, however, have strengths and some limitations. For example, SPOTCOMs is based on the detailed representation of sweet potato growth and d evelopment processes and can be used for impact assessments at any location with some level of calibration, but many of the detailed input data and other information required by the model may be difficult to obtain. On the other hand, empirical models typi cally use climate - yield relationships and have an advantage over process - based simulation models as they may capture the effects of cultural and economic limiting factors on crop yields. The major drawback of statistical models is that they cannot be used in assessments under conditions that may lie outside of the empirical range of conditions the model was developed with. 33 The main objective of the present study was to develop a sweet potato modeling framework for Uganda using a process - based SPOTCOMS mod el. The main objective was achieved by determining the sweet potato cultivar parameters followed by a model calibration and validation process. The project was aimed at investigating the response and productivity of sweet potatoes under different environme ntal conditions. There is, therefore, a need to evaluate sweet potato 2.2 Materials and methods 2.2.1 Treatments and experimental design A split - plot f ield design was used with four replications (Ekanayake, 1989). Genotypes were assigned randomly to main plots. Individual genotype plots were 14 m long and 2.7 m wide; rows were 1 m apart with 0.3 m within row planting distance. There were 14 rows in each plot. Non - rooted sweet potato apical stem cuttings of approximately 30 cm length were planted on ridges which were 1.0 m apart and at a plant - to - plant spacing of 30 cm. The cuttings were planted on 29 August 2012 and on June 15, 2013 , for the first and sec ond growing seasons respectively, 2013 for the first season and November 14, 2013 , for the first and second growing seasons. Figure 2.1 show s the design/layout o f the plots and blocks. 34 Figure 2. 1 Design of the experiment. (a) The l ayout of field plots by cultivar type, V1, V2, V3, V4 for the four replications; REP 1, REP 2, REP 3 AND REP 4. V1, V2, V3, V4 represent the sweet pot ato cultivars NASPOT 1, NASPOT 10 O, NASPOT 11 and SPK 004. (b) The l ayout of vines for a single cultivar plot. The crosses represent plants in a column and colors are used to only emphasize that 4 plants in the same column are sampled every sampling date. REP 1 REP 4 B V 4 V 2 V 3 V 1 V 2 V 3 V 4 V 1 REP 2 REP 3 A V 1 V 3 V 4 V 5 V 3 V 1 V 2 V 4 1 2 3 4 5 6 7 8 (a) (b) WAY 35 2.2.2 Selection of cultivars Orange - fleshed cultivars, NASPOT 10 O (Mwanga et al. 2009) and SPK 004 (Kakamega) (Mwanga et al., 2007) and non - orange fleshed cultivars, NASPOT 1( Gibson et al., 2008; Gibson et al., 2011) and NASPOT 11( Gibson et al., 2011 ) were planted in two seasons from August 29, 2012 to January 28, 2013 and June 15, 2013 to November 14, 2013. The four cultivars were selected for this study because of their popularity and competitive traits as shown in Table 2.1. For example , the two n on - orange cultivars, NASPOT 1 AND NASPOT 11 are high yielding varieties, NASPOT - Carotene. Table 2. 1 Properties of sweet potato cultivars used in this study Cultivar trait Cultivars NASPOT 1 1,2,3,4 NASPOT 10 3 NASPOT 11 4 SPK OO4 5, 6, 7, 8, 9, 10 Root flesh color Pale yellow Dark orange Cream Orange Maturity (days) 120 - 150 110 120 - 150 120 - 150 Average root yield (kg/ha) 29 16 28 14.9 - Car otene (µg/100g) 185.6 - 342.8 376 - 760 Taste Sweet Moderately sweet sweet Moderately sweet Resistance to Weevils S S S S Resistance to SPVD MR MR MR MR Resistance to Alternaria stem blight MR MR 1 Mwanga et al., 2003a; 2 Gibson et al., 2008; 3 Mwanga et al., 2009; 4 Gibson et al., 2011; 5 Mwanga et al., 2005; 6 Yanggen and Nagujja, 2006; 7 Mwanga et al., 2007a; 8 Mwanga et al., 2007b; 9 Namanda et al., 2011; 10 Tumwegamire et al., 2011a 36 2.2.3 Plant management A nursery bed for multiplying the sweet potato vin es was set up earlier for the four cultivars from which vine cuttings were collected and used in the experimental plots. The four sweet potato cultivars were grown in field trials during the period 2011 - 2013 Namulonge, in Uganda. The field trials were moni tored regularly and moderate irrigation was applied in case two weeks passed without the field went without receiving rainfall. Diammonium phosphate (DAP) was broadcasted on the mounds at earlier stages of planting sweet potato in order to boost the nitrog en and phosphorous levels in the soils. This was done in order to allow the sweet potatoes to grow under water - stress - free conditions. 2.2.4 P lant trait destructive monitoring Alternate rows leaving border rows were sampled for non - destructive measurem ents for the three replicates . The fourth replicate was left undisturbed up to the end of the full maturity period in order to compare the root yields with the data which was being collected during destructive sampling. The phonological data that was coll ected from the field during the growing season s included the following sweet potato attributes: vine length, number of leaves, leaf area, number and size of roots, fresh and dry weights of vine, leaves , and roots, canopy cover. The data were measured every 15 days , according to Kooman et al., (1996), as a part of the monitoring of the growth and development of the sweet potato crop. 2.2.5 Soils The soils used in to run the model for Namulonge in 2012 and 2013 were analyzed before the experiments were set up to identify the soil classification, bulk density, PH and the nitrogen (N), 37 phosphorous (P) and potassium (K). The actual concentrations of soils used in earlier experiments of 2004 to 2009 were not recorded and therefore an assumption has been taken to use the soil concentrations of 2012 as representative amounts. The soil classifications for the three locations were determined using the Harmonized World Soils Database (HWSD) version 1.2 (FAO/IIASA/ISRIC/ISSCAS/JRC, 2012) . The HWSD is a 30 arc - second raster database with over 16000 different soil mapping units that combines existing regional and national updates of soil information worldwide (SOTER, ESD, Soil Map of China, WISE) with the information contained within the 1: 5,000,000 scale FAO - UNESCO Soil Map of the World. Soil analysis tests were carried out on August 21, 2012 , before preparing the land for the experiment. A 40 m 2 plot of land was sampled and was divided into two small plots of about 20 m 2 . Two pits of about 2 m 2 wide and 1.5 m deep were dug in the middle of each plot. A number of soil properties including color, structure, consistenc y , porosity, depth, texture among others were described as shown in Tables 2.1 and 2.2. In addition, four soil samples (two from each plot) were sampled with the help of a pos t hole Auger. A transect was demarcated diagonally across the big plot and two spots were sampled within each small plot, these were mixed and quarter sampled to get a representative sample at both top (0 - 30 cm) and sub (30 - 60 cm) depths. Generally, the entire plot was characterized by blackjack and Conyza floribunda as the dominant vegetation. The plot had been under fallow for about two years and the land was gently sloping towards the valley. The new formation and inclusions included evidence of minera lization. Table 2.3 shows the soil characteristics of the first pit and the second pit. The soil bulk density was 1.36 g cm - 3 and the volumetric soil water content, at field capacity and wilting point, were 0.35 and 0.17 m 3 m - 3 respectively. The pH of the soils w as ranging from 3 to 5.5 which indicates acidic soils which are the preferred soils for sweet potato growth (Stoddard et al. 2013). The soil textural 38 class also generally ranged from sand to loamy sand, the most suitable texture for sweet potato gro wth. The concentrations of nitrogen, phosphorous and potassium found in the soils were 79.13 kg/ha, 58.36 kg/ha, 90.19 kg/ha respectively. When compared to the required NPK concentration for optimum sweet potato growth which is 84.06 kg/ha, 224.17kg/ha, a nd 336.25 kg/ha, the soil was more deficient in phosphorous which are a n essential element in the development of root biomass (Stoddard et al., 2013). Diammonium phosphate (DAP) fertilizer was, therefore, applied at a rate of 50 kg/ha, six weeks after plan ting the trial using soil nutrient recommendations for sweet potatoes (Stoddard, 2013). We could not apply the fertilizer at planting, which is the most suitable time because soil analysis results were not available at an earlier time. However, from the av ailable literature , we feel that the fertilizer was still effective at six weeks after planting. A description of the soil variables is very important in running SPOTCOMS crop model, especially the field capacity (FC), permanent wilting point (PWP) and alb edo, as they are linked to the amount of water available for the crop and the amount solar radiation used in photosynthesis (Gijsman, Jagtap, & J.W., 2002) . The number of nutrients in form of nitrogen (N), phosphorous (P) and potassium (K) present in the soil also directly influence the amount of crop yield got by the crop at the end of the growing season. Nitrogen aids in the development of aerial parts of sweet potato and an excess of it leads to increase in the number and size of leaves and rapid stem growth (Ustimenko and Bakumovsky, 1982). Potassium , on the other hand, increases the rate of photosynthesis and therefore aff ects the quantity of tuber yield (Biswas and Mukherjee, 1994). Ref: Ustimenko, C.G.V. and Bakumovsky, 1982. Plants growing in the tropics and subtropics. Mir publishers. 39 Table 2. 2 Soil characteristics for Pit One Property Hor izon A Horizon B Horizon C Depth 0 - 20cm 20 - 50cm 50 and above Boundary sharpness Clear Diffuse Diffuse Moisture Very moist Moist Moist Color Reddish black Dusky red Dark reddish brown Texture Loam Clay loam Sandy clay loam Structure: (a) Strength (b) Shape Structureless Weakly developed Weakly developed Granular Crumb Consistency Loose Firm Friable Porosity Fine porous Fine porous Fine porous Fauna None None None Drainage Perfect Perfect Imperfect Compactness /Cementation Loose Firm Friable Root distribution (a) Size (b) Quantity (c) Shape (d) Nature (e) Health (f) Age Small Small Small Frequent Few Few Free growing Free growing Free growing Fibrous Fibrous Fibrous Alive Alive Alive Young Young Young Table 2. 3 Soil characteristics for Pit Two Property Horizon A Horizon B Depth 0 - 40cm 40 and above Boundary sharpness Diffuse Diffuse Moisture Moist Moist Color Dusky read Dark reddish brown Texture Loam Clay loam Structure: (a) Strength (b) Shape Weakly developed Weak Granular Crumb Consistency Friable Extremely firm Porosity Fine Fine Compactness /Cementati on Loose Very compact Fauna None None Drainage Perfect Imperfect Root distribution (a) Size (b) Quantity (c) Shape (d) Nature (e) Health (f) Age Small Sm all Frequent Few Free growing Free growing Fibrous Fibrous Alive Alive Young Young 40 Table 2. 4 Soil nutrient analysis results Details pH O.M N Av.P K Ca Mg Na Textural %ages Textural class %ages mg/kg C.moles/k g Sand Clay Silt Next to pit 1, Top (0 - 30cm) 5.1 2.1 0.1 0.8 0.7 5.9 2.1 0.11 56 27 17 Sandy clay loam Next to pit 1, Sub (30 - 60cm) 4.9 0.8 0.0 0.4 0.7 4.7 1.4 0.1 38 51 11 Clay loam Next to pit 2, Top (0 - 30cm) 5.0 2.1 0.1 1.3 0.6 6.9 2.7 0.1 56 26 18 Sandy loam Next to pit 2, Sub (30 - 60cm) 5.0 1.5 0.1 0.3 0.7 5.6 1.9 0.2 39 48 13 Sandy clay Critical levels 5.5 3.0 0.2 15.0 0.2 4.0 0.5 <1.0 41 2.2.6 Secondary agronomic and climate data The study also used secondary data collected from past fi eld trials from five locations, Namulonge, Masaka, Serere, Soroti and Kabale all located in Uganda, and later used in crop model evaluation. The locations were strategically located across Uganda in regions with varying climates (Table 2.1) and they were t he ones with historical agronomic data which also could limit the choice of site location . Out of the five locations, Kabale was the coolest and was also located at the highest elevation of 2,000 m. Serere was the hottest while the wettest and driest site s were Namulonge and Masaka respectively. The locations had slightly different soil textures at the top - soil and subsoil layers, varying bulk densities, and varying elevation. Historical data for climate and sweet potato root yield was collected from all t he sites. 42 Table 2. 5 Description of experimental data sites used in the determination of cultivar coefficients and validation of the crop model Site name Lat Lon Elevation (m) Annual rainfall (mm) Topsoil (0 - 30 cm) Textu re classification Bulk Density (kg/dm3) PH Soil Base Saturation (%) Fraction Silt Clay Sand Kabale - 1.25 29.98 2,000 1,018 Clay (light) 1.34 5.3 59 16 47 37 Namulonge 0.53 32.62 1,160 1,242 Clay loam 1.36 5.5 65 24 31 45 Masaka - 0.30 31.67 1,3 10 1,200 Clay loam 1.36 5.5 65 24 31 45 Serere 1.49 33.46 1,085 1,250 Clay(light) 1.42 4.9 40 16 40 44 Soroti 1.72 33.62 1,100 1,365 Clay(light) 1.42 4.9 40 16 40 44 Site name Lat Lon Average min. temp. ( 0 C) Average max. temp. ( 0 C) Subsoil (30 - 100 cm ) Texture Classification Bulk Density PH Soil Base Saturation Fraction Silt Clay Sand Kabale - 1.25 29.98 10 23 Clay (light) 1.34 5.3 44 14 54 32 Namulonge 0.53 32.62 16 28 Clay 1.43 5.4 69 21 45 34 Masaka - 0.30 31.67 17 28 Clay 1.43 5.4 69 21 45 34 Serere 1.49 33.46 18 31 Clay (light) 1.42 5.1 49 15 46 39 Soroti 1.72 33.62 13 28 Clay (light) 1.42 5.1 49 15 46 39 43 2.2.7 Trend analysis for the collected field data Trend analysis was performed between seasons and among cultivars on t he eight field variables including vine length, number of leaves, leaf area, number of branches, length of branches, and number of storage roots, f r esh weight and dry weight of storage roots. Correlation analysis was performed between seasons and among the four cultivars using the formula: (1 ) Where r is the correlation coefficient, x is the index time series during the season and y is the same variable in the second growing season for a similar cultivar or for any of the other three cultivars. 2.2. 8 C alibration and validation of SPOTCOMS crop - model 2.2.8 .1 The SPOTCOMS model The present research used a process - based sweet potato model, SPOTCOMS (Sweet POTato COMputer Simulation) developed by (Mithra & Somasundaram, 2008) . The model simulates phenological development in relation to photothermal time, net assimilation, resource allocation to different plant organs - below and above ground, transpiration, and soil water dyna mics on a daily time step. The model simulates crop phenology as a function of growing degree days and divides sweet potato growth into three phases. That is, the first phase from planting to tuber initiation, middle phase from tuber initiation to the begi nning of tuber bulking and the final phase from the beginning of tuber bulking to harvest (Mithra & Somasundaram, 2008) . The parameters that drive the model were determined using eight equations, equation 1 to equatio n 8 shown below. Equation 1 defines the growth process sweet potatoes, equation 2 defines the growth stages of sweet potatoes, equation 3 describes the development of vines, equation 4 describes the 44 development of roots (also defined as tubers in the equat ion), equation 5 defines branching of the crop, equation 6 specifies the number of leaves on a sweet potato plant and, equations 7 and 8 define the leaf area of sweet potatoes. (2 ) Where: GDD is the growing degree days accumulated on the i th day after planting (DAP), i = 28 days under t ropical conditions TMEAN i is the mean temperature on i th DAP ( 3 ) Where phs2gdd The difference between 4 weeks and 7 weeks after planting (4 ) Where VL i is vine length on the i th DAP and GDD i is GDD on i th DAP (5 ) Where nTBR i is the number of tubers on i th DAP ( 6 ) Where BR i is the number of branches on i th DAP and LF i is the number of leaves on branches on i th DAP 45 (7 ) (8 ) Where ALA i is the average leaf area on i th DAP for a cultivar for the whole growing season (9 ) A few modifications were performed on the model in order to make it more robust and applicable to various parts of the world with limited climate data. The major modifications in the model included the removal of daily maximum and minimum humidity in the weather file and the new variable of daily insolation incident on a horizontal surface (in MJ/m^2/day) was included. The soil file required by the model requires the nitrogen (N), phosphorous (P) and potassium (K) concentrations found in the soil before planting and the amount added during the growing season as fertilizers. Four soils classifications were added to the existing four soil classificatio ns. The added soil classes were loamy sandy, loam, sandy clay, and sandy clay loam while the original classes included sandy, sandy loam, clay loam, and clay. The soils used in SPOTCOMS are based on three variables namely field capacity (FC), permanent wil ting point (PWP) and albedo. FC, the volumetric soil water content at drained upper limit in a soil layer (cm 3 [water]/cm 3 [soil]) and PWP, volumetric soil water content in a soil layer at lower limit (cm 3 [water]/cm 3 [soil]), were determined in accordance to (Saxton, Rawls, Romber er, & Papendick, 1986) and (Gijsman et al., 2002) . Finally, the crop evaporation (ET c ) was modified by using the crop coefficients (K c ) reported at the initial, middle and end of the growing season in (R. G. Allen, Pereira, Raes, & Smith, 1998) . The details of data used to run the model are discussed below. 46 The modifications made included, change of type of solar radiation data from sunshine hours to watts per square meter, the output of more variables such as potential evapotranspiration (PET) , evapotranspiration of the crop (Et c ), and soil moisture or root available water (RAWtr). Also, in the second phase of sweet potato growth, the model assumes that a sweet potato plant has 1 storage root, that is, when the parameter, tgrate equals one, one storage root is produced and the second phase begins at the end of 2 nd week and ends at the end of 7th week. For phsgdd , the minimum temperature (Tmin) should be taken as 23 0 C, maximum temperature (Tmax) as 32 0 C and mean daily temperature (Tmean) as 27.5 0 C for all locations in the tropics. Two parameters were added; R2R the row to row spacing (100 cm) and P2P the plant to plant spacing (30 cm). Table xxx shows the soil descriptions that the model currently has. The initial conditions for SPOTCOMS mode l were: Optimum Temperature for Sweet potato=25.0 0 C Base Temperature for Sweet potato=8.14 0 C Maximum Temperature for Sweet potato=38.0 0 C Leaf duration=47 days Extinction coefficient=0.8 PLMX optimum =45.0; PLMX optimum is the maximum photosynthetic efficie ncy at optimum temperature (Kg/ha/hr) Maintenance coefficient at 20 0 C =0.01 Initial level of water available in the soil=5.0 mm Root depth: 1.25 m 47 Table 2. 6 Soil characteristics used in SPOTCOMS Soil type Field capacity Perman ent wilting point Albedo Sandy 0.120 0.045 0.370 Sandy loam 0.230 0.110 0.250 Clay loam 0.335 0.205 0.215 Clay 0.360 0.220 0.140 Loamy sandy 0.161 0.059 0.300 Loamy 0.267 0.083 0.230 Sandy clay 0.333 0.195 0.255 Sandy clay loam 0.204 0.066 0.292 The model runs using a set of data inputs including weather data (minimum and maximum temperature, precipitation and solar irradiance), soil properties, and agronomic data. The weather data used consisted of daily rainfall for the period 2004 to 2013 repor ted at the weather stations of Namulonge and Soroti. Due to the lack of a data at Serere, and since the distance between Serere and Soroti is only 25km, the rainfall data from Soroti was used as a representative precipitation for Serere. Due to the absenc e of data at weather stations in the study sites, daily minimum temperatures, maximum temperatures, and solar radiation w ere provided by the National Aeronautics and Space Administration - Prediction Of Worldwide Energy Resource (NASA - POWER, 2014) for the period 2004 - 2013. The selection of this dataset was also justified by previous work which showed that solar radiation data from this source were often much better than station data (J. W. White, Hoogenboom, Wilkens, Stackhouse Jr, & Hoel, 2011) . SPOTCOMS estimates the effect of potassium stress on tuber yield using Mitscherlich s equation ( Biswas and Mukherjee, 1994): (10 ) where: i TBR K =Potassium stress on tuber production, 48 C K = Constant, K=Quantity of K applied (Kg/ha). equation (Biswas and Mukherjee, 1994): (11 ) where: i TWT N =Potassium stress on tuber production, C N = Constant, N=Quantity of N applied (Kg/ha). One major limitation of SPOTCOMS model was its lack of sensitivity and failure to terminate when the temperature exceeds 38 0 C, the maximum temperature for sweet potato growth. In case a location had temperatures greater then 38 0 C, the warmest locations had to be removed. Luckily enough, for this study, no location had tempera tures exceeding 38 0 C. 2.2. 8 .2 Determination of cultivar parameters for SPOTCOMS In order to determine the cultivar parameters required in running the model, sweet potato trials were set up at Namulonge Crops Resources Research Institute (NaCRRI) in Ugand a using a complete randomized block design for two growing seasons, August 2012 - February 2013 and July 2013 - December 2013. The second growing season trial was exactly the same design as the first trial consisting of four sweet potato cultivars, NASPOT 1, NASPT 10 O, NASPOT 11 and Kakamega (SPK004). For each experimental trial, we set up four replications for each cultivar, 49 three of which were used for taking measurements of four selected sweet potato plants using destructive sampling during the growing season and the fourth was left undisturbed until the final harvesting at maturity. In both experiment trials for the two seasons, irrigation, a herbicide , and Diammonium phosphate fertilizer were applied to ensure that the crop grow s without in non - limitin g conditions. The plant attributes measured and recorded in the trials included the length of stems, number of roots, number of leaves, leaf area, number of branches, the wet and dry weight of roots, the wet and dry weight of stems and wet and dry weight o f leaves. The data used to evaluate the performance of the model was taken from root yield data that was reported from field trials conducted at Namulonge, Serere and Soroti for the seasons shown in Table 1 for the period 2004 to 2009 under rainfed conditi ons (Mwanga et al., 2007) (Mwanga et al., 2010) . Calibrating the model under non - limiting conditions is a recommended procedure which enables the model to accurately simulate crop yield under rain - fed conditions (Ruiz - Noguera, Boote, & Sau, 2001) . 2.2.8 .3 Calibration and testing of SPOTCOMS Calibration and testing of SPOTCOMS w ere performed using data sets from the two experiments of 2012 and 2013 growing seasons which had observations of the phenology of swe et potatoes. The datasets from previous experiments were conducted under rainfed conditions and recorded root yield, biomass and vine yield at harvest and therefore were not suitable candidates for calibration. In the determination of suitable crop paramet ers to use in the model, the average of plant attributes w as determined from four combinations of replications and w as then used to determine crop parameters using equations 1, 2, 3, ..., 8. The combinations only considered three replications, the fourth r eplication was to be used for model testing and verification only and did not undergo destructive sampling during the growing season. The four combinations included: replication1 and 2 (Rep 1&2); replications 1 and 3 (Rep 1&3); replications 2 and 3 (Rep 2& 3); 50 and replications 1, 2 and 3 (Rep1,2&3). The computed sets of parameters were then used to run the model separately and root yields of simulated results for the four cultivars were evaluated using a descriptive statistical assessment between the observ ed root yields recorded from the fourth replication in the two seasons for the four sweet potato cultivars. The purpose of this assessment was to select a combination of parameters which provided the best fit of simulated root yield to the observed values. The combination with the best fit was used to run rest of model simulations used to evaluate the performance of the model using rainfed root yield data. Moreover, since all the four combinations of crop parameters were actually determined from the field d ata and therefore considered legitimate values for crop parameters, they were all ranked and the maximum and minimum values formed the range for specific sweet potato parameters. This information was considered useful for future scientists who would be int erested in understanding or investigating the limits to which a specific parameter can be changed. 2.2. 8 .4 Sensitivity analysis of cultivar coefficients in SPOTCOMS The purpose of the performing sensitivity analysis on cultivar coefficients was to genera te output variability associated with the variability of input, and also to assign the simulated output variability to the model coefficients that affect it most (Pathak, Fraisse, Jones, Messina, & Hoogenboom, 2 007; Ruget, Brisson, Delécolle, & Faivre, 2002) . Local sensitivity was used to provide a normalized measure in the comparing all model coefficients derived in section 2.3.4.1a. Local sensitivity was determined for model responses using the base and the +/ - 5% changes in the base value. The relative change in output and the change in parameter was used to calculate the sensitivity indices. All the eight coefficients determined were individually used to determine the sensitivity indices using the equation pro posed by (Pathak et al., 2007) below. 5 1 (12 ) Where i represent individual coefficients: phsgdd , phs2gdd, Vlen, tgrate , br gap , lfactor , lafactor , or larea . Y is simulated storage root yield using the initial set of d etermined coefficients and Y i is the i ) while keeping all other model parameters at their base values. 2.2.8 .5 Eval uation of model performance The purpose of model ev aluation in this section was to assess the performance of the model for sweet potatoes grown under rainfed conditions. This was an important step because sweet potatoes in Uganda is largely grown under rainfed conditions (Ddumba, Andresen, & Snapp, 2014) . A description of the locations used in the present study was presented in Table 1. In both model testing and evalu ation, seven descriptive statistical parameters were used. The coefficient of determination (R 2 ) is the second order of the Pearson correlation coefficient which explains the extent of agreement between the simulated and observed values (an R 2 equal to 1sh ows a very strong agreement and 0, a very week agreement). The slope of regression, a , was used to describe the relative systematic error in the simulated yields (an a equal to 1 is an optimal value). The present study hypothesized that there should be an agreement between model - simulated root yield and the observed root yield for a specific cultivar shown by a slope of regression a greater than zero and approaches 1 for optimal model fit to the observed values. The mean bias error (MBE) which is an indicat or of the average systematic error as described in (Davies & McKay, 1989) was determined. The mean absolute bias error (MABE) defined as the average of absolute differences between simulated and observed values (ranges from 1 to infinity) an d is used to calculate the 52 average magnitude of simulated errors, irrespective of their direction was determined according to (Shaeffer, 1980) . The root mean square error (RMSE) that describes the average absolute deviation between the simulated and modeled values and t he index of agreement (IA) (Willmott & Wicks, 1980) were determined. The IA is a standardized measure of the degree of model simulation error and proportionality between predictions and observations with a range of 0 to 1, where an IA closer to 1 indicates higher simulation agreement and an IA equal to 0 indicates no agreement at all. IA is more consistent than the linear correlation coefficient, but sensitive to extreme values, due to the squared differences. Finall y, the modeling efficiency (ME) according to (Nash & Sutcliffe, 1970) was determined. ME is a normalized measure of the relative magnitude of the data variance compared with the residual variance (noise). An ME equal to 1 is an optimal value or perfect fit, an ME equal to 0 means simulated values are as accurate as the mean of the observed data, and negative values mean that simulated values are worse than the mean of observed data. The corresponding equations of the describe d statistics are shown in equations 10 to 15. (1 3 ) Where y is the simulated yield, a the slope of the regression line, x the observed root yield, b the intercept and e the error of simulation, Ho: a< 0 , the null hypothesis our research hypothesis, if a is significant, then there is an agreement between simulated root yield and observed root yield (14 ) 53 (15 ) (16 ) (17 ) (18 ) Where Oi and Si are observed and simulated values respectively, n is the number of samples and is the mean of the observed values. Data analyses were performed using SYSTAT software ( Systat Software Inc, 2007) . 2.3 Results 2.3.1 Variation of Sweet potato data from the field Analysis trends of plant attributes such as length of vines and number of storage roots for the entire growing season w ere consistent with the growth of plants, as we expect ed the plant to increase in biomass with an increase in days after planting (Figure s 2.2 to Figure 2.7 ). SPK004 had the longest vine length followed by NASPOT 1, while NASPOT 10 and NASPOT 11 leng ths were similar (Figure 2.2 ). The vine length between the two seasons was very similar although 54 season 1 length was a slightly longer. The correlation coefficients indicated that the vine lengths between cultivars and seasons were similar. The number of leaves for different cultivars across the two g rowing seasons varied similarly over the growing season (Figure 2.4 ). However, SPK004, which had the longest vine length, has a little bigger number of leaves. There were high correlations of vines between cultivars acro ss different seasons (Figure 2.4 ). L eaf area was one of the major variables that distinguished between cultivars. Each cultivar had a different size with SPK004 having the smallest size of leaves and NASPOT 10 0 had the lar gest size of leaves . This pattern is consistent between the two growi ng seasons. The graphs for number of branches showed varying numbers of branches ove r the growing season (Figure 2.3 ). However, since during the sweet potato crop was just in the early stages of growth, the number of branches towards harvest should be th e ones that are representative of the actual number of branches. Therefore, NASPOT 10 0 was the one cultivar that showed a consistent number of branches of 6 in both seasons. The other 3 cultivars showed an average of 4 branches in the first growing season and 6 branches in the second season. At harvest , the number of storage roots recorded in season one w as 3,4,5,6 for NASPOT 1, SPK004, NASPOT 11, and NASPOT 10 0 respectively while the number recorded in season 2 was 3 for NASPOT 11 and 4 for NASPOT 10 0, SPK004 and NASPOT 1 (Figure 2.5). The cultivars showed a high relationship between each other with the magnitude of 0.8. The flesh and dry weigh t s of cultivars in seasons at harvest were double those of the second season (Figure 2.6 and Figure 2.7 ). The cu ltivars in the first season had 3 kg/plant, 4kg/plant and 6 kg/plant for NASPOT 1, SPK004 and the two cultivars ( NASPOT 11 and NASPOT 10 0) respectively. The dry weights 55 were almost similar for like cultivars I the two growing seasons . The correlations were high for among cultivars and b etween seasons . (a) (b) Correlati ons Season 2 V1 V2 V3 V4 Season 1 V1 0.97 0.94 0.92 0.96 V2 0.97 0.96 0.96 0.98 V3 0.96 0.92 0.96 0.95 V4 0.98 0.93 0.94 0.96 (c) Figure 2. 2 Vi ne length:V1 = NASPOT 1, V2 = NASPOT 10 0, V3 = SPK004 (Ejumula), V4 = NASPOT 11 0 100 200 300 400 16 34 49 65 82 97 113 131 152 Length (cm) Days after planting Vine length for season 1 V1 V2 V3 V4 0 100 200 300 400 16 34 49 65 82 97 113 131 152 Length (cm) Days after planting Vine legth for season 2 56 (a) (b) Correlations Season 2 Branches V1 V2 V3 V4 Season 1 V1 0.59 0.29 0.54 0.66 V2 0.64 0.40 0.67 0.74 V3 0.60 0.35 0.55 0.66 V4 0.48 0.14 0.47 0.43 (c) Fi gure 2. 3 N umber of branches:.V1 = NASPOT 1, V2 = NASPOT 10 0, V3 = SPK004 (Ejumula), V4 = NASPOT 11 0.0 5.0 10.0 15.0 16 34 49 65 82 97 113 131 152 Number of branches Days after planting Number of branches for season 1 V1 V2 V3 V4 0.0 5.0 10.0 15.0 16 34 49 65 82 97 113 131 152 Number of branches Days after planting Number of branches for season 2 57 (a) (b) Correlations Season 2 for leaves V1 V2 V3 V4 Season 1 V1 0.91 0.80 0.89 0.88 V2 0.88 0.79 0.81 0.83 V3 0.88 0.77 0.85 0.84 V4 0.86 0.81 0.82 0.83 (c) Figure 2. 4 Number of leaves and leaf arear: V1 = NASPOT 1, V2 = NASPOT 10 0, V3 = SPK004 (Ejumula), V4 = NASPOT 11 0 200 400 600 800 16 34 49 65 82 97 113 131 152 Numver of leaves Days after planting Number of leaves for seaon 1 V1 V2 V3 V4 0 200 400 600 800 16 34 49 65 82 97 113 131 152 Number of leaves Days after planting Number of leaves for season 2 58 (a) (b) Correlations Season 2 V1 V2 V3 V4 Season 1 V1 0.86 0.70 0.83 0.92 V2 0.91 0.82 0.92 0.99 V3 0.93 0.85 0.93 0.91 V4 0.69 0.56 0.70 0.83 (c) Figure 2. 5 V1 = NASPOT 1, V2 = NASPOT 10 0, V3 = SPK004 (Ejumula), V4 = NASPOT 11 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 16 34 49 65 82 97 113 131 152 Number of tubers Days after planting Number of storage roots for seaon 1 V1 V2 V3 V4 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 16 34 49 65 82 97 113 131 152 Number of tubers Days after planting Number of storage roots for season 2 59 (a) (b) Correlations Flesh weight Season 2 V1 V2 V3 V4 Season 1 V1 0.95 0.90 0.96 0.96 V2 0.96 0.94 0.95 0.98 V3 0.98 0.94 0.98 0.99 V4 0.94 0.88 0.92 0.97 (c) Figure 2. 6 Flesh weights, and correlation coefficients for storage roots. V1 = NASPOT 1, V2 = NASPOT 10 0, V3 = SPK004 (Ejumula), V4 = NASPOT 11 0.0 2.0 4.0 6.0 8.0 16 34 49 65 82 97 113 131 152 weight(kg/plant) Days after planting Flesh weight for season 1 V1 V2 V3 V4 0.00 2.00 4.00 6.00 8.00 16 34 49 65 82 97 113 131 152 Weight (kg/plant) Days after planting Flesh weight for season 2 60 (a) (b) Correlations Season 2 Dry weight V1 V2 V3 V4 Season 1 V1 0.94 0.83 0.95 0.95 V2 0.97 0.91 0.94 0.98 V3 0.98 0.92 1.00 0.98 V4 0.96 0.86 0.91 0.97 (f) Figure 2. 7 Dry wieghts and correlation coefficients for storage roots. V1 = NASPOT 1, V2 = NASPOT 10 0, V3 = SPK004 (Ejumula), V4 = NASPOT 11 2.3 .2 Cultivar coefficients Mean cultivar coefficients were computed for replication are shown in Table 25. For each cultivar, the cultivar coefficients were determined from three replications and the fourth replication was used as a control replication. Us ing the different combinations of one pair of replications, a number of cultivar coefficients were determined from which a mean coefficient and the maximum and minimum values were also determined. On plotting the range of the different coefficients, 0.0 0.1 0.2 0.3 0.4 16 34 49 65 82 97 113 131 152 Weight (kg/plant) Days after planting Dry weight for season 1 V1 V2 V3 V4 0.0 0.1 0.2 0.3 0.4 16 34 49 65 82 97 113 131 152 Weight (kg/plant) Days after planting Dry weight for Season 2 61 Figure 2.8 shows that the coefficient which is a function of vine length ( vlen ) had the biggest range for all cultivars while lafactor and laarea , both of which are function of leaf area had the largest variation between different cultivars. Table 2. 7 Summary of cultivar parameters determined from field experiments Summary of cultivar coefficients vlen tgrate br_gap lfactor lafactor larea NASPOT 1 (V1) Mean 0.12484 0.01094 0.00226 0.15377 100.3 46.4 Minimum 0.09138 0.00736 0.00150 0.14440 94.4 44.0 Maximum 0.16501 0.01473 0.00265 0.16976 109.4 49.9 NASPOT 11 (V2) Mean 0.09776 0.01237 0.00246 0.09453 104.4 51.0 Minimum 0.06388 0.00920 0.00225 0.00955 97.8 48.6 Maximum 0.12911 0.01595 0.00280 0.11750 113.4 54.6 SPK 004 (Kaka mega) (V3) Mean 0.12784 0.00972 0.00245 0.21925 54.8 24.2 Minimum 0.10241 0.00614 0.00180 0.20702 45.9 20.5 Maximum 0.16848 0.01289 0.00303 0.22553 64.8 27.8 NASPOT 10 (V4) Mean 0.09115 0.00696 0.00253 0.12020 75.3 35.2 Minimum 0.06553 0.00614 0.00182 0.11135 70.7 33.3 Maximum 0.13615 0.00767 0.00353 0.14837 80.6 37.3 For all cultivars, phsgdd = 543.2, phs2gdd = 407.4, P2P = 0.3 m, R2R = 1 m 62 (a) (b) (c) (d) (e) (f) Figure 2. 8 Range of cultivar par ameters in 2012 and 2013 season. The orange dot represents the mean of simulated root yield 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 NASPOT 1 (V1) NASPOT 11 (V2) SPK 004 (Kakamega) (V3) NASPOT 10 (V4) vlen Cultivars vlen by cultivar 0 0.002 0.004 0.006 0.008 0.01 0.012 0.014 0.016 0.018 NASPOT 1 (V1) NASPOT 11 (V2) SPK 004 (Kakamega) (V3) NASPOT 10 (V4) tgrate Cultivars tgrate by cultivar 0 0.0005 0.001 0.0015 0.002 0.0025 0.003 0.0035 0.004 NASPOT 1 (V1) NASPOT 11 (V2) SPK 004 (Kakamega) (V3) NASPOT 10 (V4) br_gap Cultivars br_gap by cultivar 0 0.05 0.1 0.15 0.2 0.25 NASPOT 1 (V1) NASPOT 11 (V2) SPK 004 (Kakamega) (V3) NASPOT 10 (V4) lfactor Cultivar lfactor iby cultivar 0 20 40 60 80 100 120 NASPOT 1 (V1) NASPOT 11 (V2) SPK 004 (Kakamega) (V3) NASPOT 10 (V4) lafactor Cultivar lafactor by cultivar 0 10 20 30 40 50 60 NASPOT 1 (V1) NASPOT 11 (V2) SPK 004 (Kakamega) (V3) NASPOT 10 (V4) larea Cultivar larea by factor 63 2.3.3 Sensitivity of cultivar coefficients from field data With reference to the average sensitivity function, results of the sensitivity analysis of cultivars indicated that the coefficients which are a function of growing degree days, phsgdd , phs2gdd were the most sensitive for all the four cultivars (Table 2.8). On considering other coefficients that are cultivar specific, the most sensitive cultivar coefficie nts were tgrate for both NASPOT 11 and NASPOT 1, vlen and br_gap for SPK004, and vlen and lafactor for NASPOT 10 0. Figure 2.9 shows the sensitivity of the cultivars which was determined by taking the average of all individual cultivars. The graph emphasizes the earlier observed trends of phs2gdd being the most sensitivity with a negative effect while larea was the most sensitive when considering other coefficients with the exception of those which are a function of growing degree days. 64 Table 2. 8 Sensitivity analysis of cultivar coefficients Coefficie - nts Optimum coefficients Rep 2&3_sn1 5% increase of the parameter Simulated yield sn1 increase 5% decrease of the parameter Simulated yield sn1 decrease NASPOT 1 (V1) phsgdd 543.2 570.36 39. 6 - 2.35 516.04 41.6 1.44 - 0.45 phs2gdd 407.4 427.77 39.4 - 2.43 387.03 42. 2 1.19 - 0.62 vlen 0.1514 0.15897 39. 4 - 2.43 0.14383 39.6 2.3 2 - 0.06 tgrate 0.00736 0.007728 39.7 - 2.28 0.006992 39.1 2.55 0.13 br_gap 0.0015 0.001575 39.5 - 2.38 0.001425 39.4 2.43 0.02 lfactor 0.144397 0.151617 39. 5 - 2.40 0.137178 39. 5 2.40 0.00 lafactor 100.23 105.2415 39.4 - 2.42 95.2185 39. 6 2.36 - 0.03 larea 46.68 49.014 39. 5 - 2.41 44.346 39. 6 2.36 - 0.02 Simulated yield 44. 9 NASPOT 11 (V2) Coefficie nt Optimum coefficients Rep 2&3_sn1 5% increase of parameter Simulated yield sn1 5% decrease of parameter Simulated yield sn1 decrease phsgdd 543.2 570.36 43.7 - 2.08 516.04 45. 5 1.37 - 0.36 phs2gdd 407.4 427.77 43.8 - 2.03 387.03 47. 2 0.66 - 0.68 vlen 0.12911 0.13556 44.0 - 1.94 0.12265 44.1 1.91 - 0.01 tgrate 0.01595 0.0167 4 44.2 - 1.87 0.015152 43. 9 2.01 0.07 br_gap 0.0027 0.002835 44. 1 - 1.92 0.002565 44. 1 1.92 0.00 lfactor 0.11750 0.12337 44.0 - 1.94 0.11162 44.1 1.91 - 0.02 lafactor 101.84 106.932 44. 1 - 1.94 96.748 44.0 1.96 0.01 larea 49.4 51.87 44 .0 - 1.96 46.93 44. 1 1. 93 - 0.02 Simulated yield 48. 8 65 Table 2.8 ) SPK 004 (Kakamega) Coefficie nt Optimum coefficients Rep 2&3_sn1 5% increase of parameter Simulated yield sn1 5% decrease of parameter Simulated yield sn1 decrease phsgdd 543.2 570.36 35.5 - 0.56 516.04 35.9 0.37 - 0.10 phs2gdd 407.4 427.77 37.6 0.58 387.03 38.5 - 1.07 - 0.25 vlen 0.14035 0.14737 33.9 - 1.45 0.13333 38.4 - 1.01 - 1.23 tgrate 0.00614 0.006 447 36.6 0.01 0.005833 36.5 0.02 0.02 br_gap 0.0018 0.00189 34.9 - 0.93 0.00171 35.5 0.60 - 0.17 lfactor 0.225527 0.23680 35.2 - 0.74 0.214251 35.4 0.63 - 0.05 lafactor 46.17 48.4785 37.4 0.45 43.8615 35.7 0.49 0.47 larea 20.97 22.0185 36.4 - 0.07 19.9215 3 4.1 1.34 0.63 Simulated yield 36.6 NASPOT 10 Optimum coefficients Rep 2&3_sn1 5% increase of parameter Simulated yield sn1 5% decrease of parameter Simulated yield sn1 decrease phsgdd 543.2 570.36 40.7 - 1.02 516.04 43.2 - 0.16 - 0.59 phs2gdd 407.4 427.77 40.6 - 1.05 387.03 43.4 - 0.24 - 0.65 vlen 0.13615 0.142958 40.7 - 1.01 0.12934 42.9 0.00 - 0.51 tgrate 0.00644 0.0 06762 41.2 - 0.77 0.006118 40.6 1.04 0.13 br_gap 0.0019 0.001995 40.8 - 0.94 0.001805 40.7 1.00 0.03 lfactor 0.111346 0.11691 40.6 - 1.04 0.10578 40.5 1.09 0.02 lafactor 72.53 76.1565 40.7 - 1.00 68.9035 42.8 0.02 - 0.49 larea 34.06 35.763 42.6 - 0.10 32.357 41.0 0.86 0.38 Simulated yield 42.9 66 Figure 2. 9 Sensitivity analysis of cultivar coefficients 2.3.4 Calibration of the model The experiment results presented here were for only 7 out 8 records corresponding to the fou r cultivars that were grown in two seasons. One record was removed for the second season for NASPOT 11 because it had an unrealistic value which was extremely large and therefore could have been erroneously measured. The results from the seven data points o f the field trial data are shown in Table 2.7. -1.5 -1 -0.5 0 0.5 1 Parameter NASPOT 10 SPK004 NASPOT 11 NASPOT 1 67 Table 2. 9 Calibration (under irrigation) and evaluation Sweet potato coefficients n MBE (t/ha) MABE (t/ha) R2 RMSE IA ME a p - value Experimental results from irrigated field tri als 7 1.16 2.58 0.4 18 3.18 0.940 0.901 0.894 0.001 Number of branches 7 0.4 0.9 0.52 1.2 0.74 0.13 0.757 0.029 Number of leaves 7 202.6 223.5 0.21 263.8 0.37 - 8.87 0.901 0.384 Number of storage roots 7 0.3 1.5 0.28 2.1 0.33 - 0.09 0.331 0.398 Rain - fed historical root yield 32 0.5 8.1 0.2444 9.6 0.70 0.31 0.3097 0.041 Table 2. 10 Correlation coefficients between simulated root yield and observed root yield at selected locations. Location Correlation Coefficient Namulonge 0.6 8 Serere 0.55 Soroti 0.44 Masaka 0.67 Kab a ale 0.48 In Figure 2.10 , we compared the model simulated variables with data collected from actual field plots in order to examine the performance of the model. The model performed well in simulating the roo t yield of NASPOT 1, NASPOT 10 O and Kakamega - SPK004 since these values are very close to the 1:1 line but the model overestimates the root yield for NASPOT 11 (Figure 2.10 a). The model is able to reconstruct the branching for the four cultivars pretty w ell although it slightly overestimates the number of branching for NASPOT 1 and NASPOT 11 (Figure 2.10 b). Please note the values shown in these graphs are averages and therefore some of them are fractions. 68 The biggest weakness of the model was in reconstr ucting the number of leaves for the four cultivars. Figure 2.8c shows that the model underestimated the number of leaves for all the cultivars. This , therefore, means that our model parameter equations for leaves need further agronomic experiments in order to make the necessary modifications in the model. However, this drawback is not of a big concern for sweet potato modeling in East Africa because in this region mostly root tubers are consumed by humans and leaves ar e fed to animals. In Figure 2.10 c, it can be observed that the model performs well in simulating the number of root tubers for three cultivars; NASPOT 1, NASPOT 10 O and Kakamega - SPK004. The model , however, simulates a slight underestimate of the number of root tubers for NASPOT 11. Results from the model evaluation using 32 historical rain - fed sweet potato field data points are summarized in Table 2.7. The regression between simulated and observed historical yield data showed a significant relationship at the 95% level of significance. The m ean bias error (MBE) was 0.5t/ha, the mean absolute bias error (MABE) was found to be 8.1 t/ha, the index of agreement was 0.7 and the modeling e fficiency was 31%. In Figure 2.11 a and b, the observed sweet potato points were plotted on a scatter plot in or der to identify any relationships that could exist. The data points were organized by loca tion and by cultivar (Figure 2.11 a, b) and another scatter plot did not represent any specific cultivar or location (Figure 2.11 c). The figures show that data was ge nerally distributed across the 1:1 line for all the locations except for Soroti. The cultivar - type scatter plot showed an even distribution of crop yields on the 1:1 line. 69 (a) (b) (c) (d) Figure 2. 10 Graphs of simulated against observed data y = 0.5799x + 293.22 R² = 0.205 y = x 0 100 200 300 400 500 600 700 800 900 0 250 500 750 Simulated number of leaves Observed number of leaves Number of leaves y = 0.7503x + 1.4202 R² = 0.5167 y = x 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 Simulated number of branches Observed number of branches Number of branches y = 0.7885x + 1.1434 R² = 0.278 y = x 0 1 2 3 4 5 6 0 1 2 3 4 5 6 Number of simulated storage roots Number of observed storage roots Number of storage roots y = 0.6423x + 10.992 R² = 0.418 y = x 0 10 20 30 40 50 60 70 0 10 20 30 40 50 60 70 Simulated weight of storage roots Observed weight of storage roots Flesh weight of storage roots 70 (a) (c) (b) Figure 2. 11 Regression plot for model results using historical sweet potato yields for the period 2004 - 2009 y = x 0 4 8 12 16 20 24 28 32 36 40 44 48 0 4 8 12 16 20 24 28 32 36 40 44 48 Simulated root yield (t/ha) Observed root yield (t/ha) Organized according to location of experiement Namulonge Serere Soroti Kabale Masaka y = 0.3097x + 18.974 R² = 0.2444 y = x 0 4 8 12 16 20 24 28 32 36 40 44 48 0 4 8 12 16 20 24 28 32 36 40 44 48 Simulated sweetpotato root yield (t/ha) Observed sweetpotato root yield (t/ha) Model evaluation with all points 0 4 8 12 16 20 24 28 32 36 40 44 48 0 4 8 12 16 20 24 28 32 36 40 44 48 Simulated root yield (t/ha) Observed sweetpotato root yield (t/ha) Model performance with observed root yield organized according to cultivar type NASPOT 1 NASPOT 10 0 NASPOT 11 SPK004 (Kakamega) 1:1 line 71 2.4 Discussion The differences were reported for different measured variables from the field trials were largely be tween the two seasons and to a lesser extent among cultivars within the same season. One major factor that changed between the two seasons was the amount of precipitation received. In the second season, received less precipitation than the second season. A nd even if irrigation was available in both seasons, there could have been a possibility that the timing of irrigation in the less wet season could not be applied at that exact time when water was most needed for. It was interesting to observe the differen ces in the physical attributes of the four cultivars. For example, the leaf area and vine length were the two most distinct features between cultivars, especially with SPK004 having very long stems and very little leaves. The cultivar coefficients determi ned in this study were quite distinct among the four cultivars especially the ones that were unique to a particular cultivar namely vlen , tgrate , br_gap, lfactor , lafactor , and larea . A comparison of these coefficients with those obtained by (Somasundaram & Mithra, 2008) tgrate vlen with the Indian cultivars, Sree Arum, Sree Bhadra, and Sree Rethna. This implies that cultivars from diff erent locations had similarities because they we re all sweet potatoes although there could be one or more features that ma d e a particular cultivar able to grow in a particular environment. Also, the range of values of coefficients determined reported in t his present study will be useful in future studies especially when researchers will be interested in establishing the maximum values to which the coefficients can be extended. This kind of information can be useful in the investigation of a drought toleran t crop or any other crop characteristics of that may be of interest. Results from model calibration and evaluation are very promising. SPOTCOMS did very well in simulating sweet potato root yields. The model evaluation results were also within a decent 72 ra nge compared to other similar studies using crop models. However, on comparing the model results with other variables such as the number of leaves, the number of branches and number of storage roots, SPOTCOMS showed the largest weakness in simulating the n umber of leaves (Figure 2.8). Also, when we looked at the internal processes of the model such as the way the model handles the actual evapotranspiration and the readily available water, we noticed that the model was not responding as it would be expected. The root mean square error (RMSE) of 3.18 t/ha for storage root yield simulated that was achieved in this study falls within the range of RMSE reported in previous studies which reported 2.88 3.42 t/ha in SPOTCOMS {Mithra, 2008 #23} and 1.14 4.17 t /ha in MADHURAM {Somasundaram, 2008 #21} in India. For the rest of the other crop variables, the modified SPOTCOMS in this study either fell below or above the range of values reported in the two previous studies. For example, this study reported an RMSE o f 1.2 branches for the number of branches, 263.8 leaves for number of leaves, and 9.6 root tubers for the number of root tubers while previous studies by Somasundaram (2008) and Mithra and Somasundaram (2008) reported ranges of 3.53 6.91 branches, 5.51 15.93 leaves and 0.97 1.67 tubers. It was noted that whereas the model has some level of accuracy especially in simulating root yield, SPOTCOMS still needed to be improved in order to capture the growth process of sweet potatoes in East Africa. A comp arison with other crop models for potatoes and yam which are sister crops to sweet potatoes indicated varying ranges or RMSE for the root tuber yield but still falling within the RMSE values obtained in this study. For studies involving potatoes, Lenz - Wied emann et al., (2009) reported an RMSE of 1.6t/ha using DANUBIA model, Angulo et al., (2013) reported a range of 1.13 2.65 t/ha when working with LINTULS - FAST model, while other studies reported 6.74t/ha, 8.7t/ha, 0.74 1.48 t/ha and 1.08 1.19 t/ha fo r the models REGCROP (Gobin, 2010), 73 Potato Calculator (Jamiesen et al., 2009), SOLANUM (Condori et al., 2010), and LINTUL - NPOTATO (van Delden et al., 2001 ) respectively. In two studies involving the use of crop models for yam , the RMSE reported for CROPSYS TVB - Yam (Marcos et al., 2011) and EPIC - Yam (Srivastava and Gaiser, 2010) were 0.5 t/ha and 8.78 25 t/ha. All these values as reported from previous studies were not so unique compared to our value of 3.18 t/ha. Another major limitation of SPOTCOMs was t hat the model continued to run normally even at high temperatures exceeding the maximum temperature of 38 0 C above which sweet potato growth is expected to be inhibited. For many crops, increases in maximum temperatures severely lead to a reduction in yield and fail ure of reproductive processes (Thorntorn et al., 2014). For instance, yield reduction of 1.7% w as reported when each degree day was spent above the maximum temperature under drought conditions in maize (Lobell et al. 2011). In another study involv ing rice production, rice yields reduced by 90% when the temperature was increased to 32 0 C during the night compared with 27 0 C (Mohammed & Tarpley, 2009). At the time of the study, the solution to increased temperatures beyond threshold values was to caref ully assess the growing season temperatures at the location and ensure that locations which temperatures above that threshold value were not used for running the model. 2.5. Conclusion Overall, our sweet potato model, SPOTCOMS, performed well in recons tructing the growth of the sweet potatoes cultivars. Second, the East African region now has the first calibrated sweet potato process - based model, having all the required parameters and coefficients for four sweet potato cultivars, which can be used in a ny form of impact assessment studies. Third, our experimental trials involved correction of various types of sweet potato growth data that is readily 74 available for future reference and studies by any interested scientist. There is currently no comprehensiv e dataset like the one we collected from our first season and from the soon to be completed second season experiment. One of the major achievements of this study is the determination of the sweet potato crop coefficients required for running SPOTCOMS mod el. The field experiments conducted at Namulonge across the two seasons in 2012 and 2013 provided the required dataset on the growth of sweet potato that made it possible to determine the coefficients. Whereas this was done at only one location under two s easons, it is recommended that follow - up studies be conducted across the whole East Africa region for many more sweet potato cultivars including the four cultivars used in this study. One major advantage or value that this study has attempted to achieve wa s the use of four representative high yielding cultivars, two of which were non - orange cultivars ( NASPOT 1 and NASPOT 11) and the other two were orange cultivars ( NASPOT 10 0 and SPK004 - Kakamega). The other contribution of this study was the modification of the previous SPOTCOMS model to be able to input weather data of any size for any number of years and some other minor additions of variable outputs such as the potential evapotranspiration (ET), actual evapotranspiration (Etc) and the root available wat er ( rwtr ). These modifications imply that SPOTCOMS can now be run for a single site for multiple seasons. The model was tested across Uganda in over four locations with varying climate and altitude. It should be noted that sweet potato cultivar coefficien ts were determined using two seasons at a single location in Uganda following minimum crop model r equirements as suggested by (Boote et al 2009 ). This study is the first process - based study on sweet potatoes on the African continent and provides the founda tion upon which subsequent studies can refer in order to continue with sweet potato modeling in the region. The sensitivity and range of the crop cultivar coefficients 75 w ere determined in this study and therefore provides a good basis for similar modeling w ork in other regions with varying climates. The cultivar coefficient sensitivity analysis is also useful for studies which may be focusing on identifying a suitable cultivar for a given question of interest. For example, in (Pathak et al., 2007) , a similar sensitivity analysis was performed on cotton crop in order to determine an ideal cultivar that would give high yields under a highly variable climate. The performance of SPOTCOMS under the model evaluation results also showed that the model has a high potential in simulating sweet potato production in the region. This shows a lot of promise in the appli cation of the model in answering various questions such as those associated with the effect of temperature, rainfall , and soils on the growth of sweet potatoes . Moreover, the model can now be used for investigating the impact of climate change on sweet pot ato production in the East African region, as will be demonstrated in the later chapter. The results from model evaluation also demonstrated that the model would be a good tool in studies involving various sweet potato cultivars as was shown on the four c ultivars used in this study. Like most models, SPOTCOMS has some limitations that will require to be addressed in future studies. For example, SPOTCOMS needs to be set such that it has a threshold beyond which if the temperature is exceeded, sweet potato growth would be inhibited. This has not yet been set in the model and therefore, the researcher is mandated to manually remove or not to consider locations with high temperatures exceeding 38 0 C, the maximum temperature for sweet potato growth . Second, alth ough the model is sensitiv e to both temperatures and soil moisture, SPOTCOMS did not appear to give a corresponding sensitivity on the actual evapotranspiration. In other words, in a case where the ETc would be elevated, the model did not show a correspond ing variation in the root available water. This is one major area that requires revisiting in the model. Third, SPOTCOMS does not yet consider CO2 intake which is known to equally affect sweet 76 potato growth just like temperature and soil moisture. Fourth, the model does not account for the effects of weeds, pest , and diseases and therefore, the model normally tends to overestimate yield because of this weakness . Finally, the model currently uses basic soil routines and does not account for the variation of soil nutrients in the various soil profiles as has been significantly developed in other crop models such as the DSSAT crop models (J. W. Jones et al., 2003) . This too will have to be worked on in the future. One major shortcoming for SPOTCOMS, which is not uncommon in an other process - based crop model , is that the model does not consider the effect of pests and diseases. This , therefore, means that the assumption is made that pest and disease management was carefully implemented i n the fields, although this is not normally the case in reality. 77 CHAPTER 3. THE IMPACT OF CLIMAT E CHANGE AND VARIABI LITY ON SWEET POTATO PRODUCTION IN EAST A FRICA 3.1 Introduction Climate change is a critical global environmental challenge affecting various ecosystem services (Bage, 2007) leading to extreme weathe r events such as droughts, floods, erratic and unreliable rainfall. Rural agriculture - based livelihood systems that are already vulnerable to climate variability and change face immediate risk of increased crop failure, new patterns of pests and diseases, reduction in water and pasture availability, lack of appropriate seeds and planting material, and loss of livestock. This has led to a reduction in agricultural yields and worsening food insecurity (Parry et al., 2005; IPCC, 2007). These challenges were, in part, the motivation behind selecting East Africa for this study and the need to calibrate a sweet potato model for the region which would consequently be used to quantify the impact of climate change on sweet potato production and for other application in future research. Like most regions of the world, East Africa faces unprecedented challenges due to climate 0 C in the next 20 years to 4.30C by 2080 (Hepworth and Goulde n, 2008) leading to changes in the ecosystem functioning. This will , in turn, lead to changes in the distribution of agro - ecological zones and soil moisture, and shortening of the growing seasons (Hulme, 1996). C3 plants such as roots and tubers 78 will be th e most preferred to C4 plants such as cereals whose yields especially for maize will be greatly reduced. Luckily enough, roots crops such as sweet potatoes, are grown in most parts of East Africa and it is an important staple in most countries of this re gion. Sweet potato is among the four most important staple crops (FAO, 2012) in East Africa. Sweet po t ato is suited in ensuring food security and fighting poverty because of its efficient production of calories per unit land, even under low rainfall, poor soil fertility conditions and projected shortened growing seasons where other crops may fail. Furthermore, because of their high carbohydrate content, they have the potential to be transformed from purely subsistence food crops to industrial and commercial crops as has been achieved in Brazil and Thailand. Potential effects of climate change on root crops production are difficult to assess not only because of the uncertainties of the magnitude of the changes in climatic variable but also due to uncertaint ies on how the crops respond to weather and climate, soil, management and other related factors. Sweet potato is known to be drought resistant and able to do well under marginal conditions. Unfortunately, there is little inf ormation on how well sweet pota to perform s , and for support farming communities under situations of extreme weather events, such as drought and shortened growing seasons. This study was aimed at understanding the nature of climate change and variability in East Africa and how it impac ts sweet potato production. The specific objectives of this study were threefold : To analyze the trends in the historical climate and sweet potato production across the East African region ; To perform a sensitivity analysis on temperature and water require ments of sweet potatoes , and To assess the impact of climate change and variability on sweet potato production across East Africa . 79 3.2 Methodology The steps followed to conduct research under Chap ter three are shown in the flow cha r t in Figure 3.1. First, a n extensive literature review on Global Circulation Models (GCMs) from the fifth phase of the Coupled Model Intercomparison Project (CMIP5) (Taylor, Stouffer, & Meehl, 2009) was conducted in order to identify models which were performing better in reconstructing Afri pathways (RCP 4.5 and RCP 8.5) . Climate data from the GCMs w ere downscaled by use of a weather generator, MarkSim and daily climate records for projected future cli mate and current climate where organized. This climate data and other data sets were then used as input datasets in a sweet potato model, SPOTCOMs to generate a simulation of sweet potato data which was then used to assess the impact of climate change and variability on sweet potato production on historical, current and projected future timescales . The detail s of the analysis methods were discussed in the subsequent sections. 80 Figure 3. 1 Flowchart describing major project obj ectives, tasks, processes, and input data types 3.2.1 Description of the study area The topography of East Africa varies from 0 m on the coast of the Indian Ocean to 5, 890 m at the highest peak of Mt. Ki l imanjaro . The regional climate is controlled by t he presence of the Intertropical Convergence Zone; Indian Ocean; Variable topography (Ogallo, 1989; Goddard and Graham, 1999; Anyah et al., 2006). The E ast African climate is also controlled by local features R eview of literature on performance of CMIP5 GCMs in Africa Selected 4 representative GCMS and 2 RCPs No. of GCM RCP CSIRO - Mk3.6 RCP 4.5 & RCP 8.5 MIROC5 MRI - CGCM3 NorESM1 - M Inputs: Latitude/Longitude & elevation for locations Monthly rainfall Daily average temperature Temperature phase angle Generation of projected future climate scenarios by stochastic downscaling of GCM data for 38 locations in Africa using MarkSim Historical daily weather data input (1980 2009) Detailed soil data input Detailed crop cultivar data input Crop simulation with projected future climate and assessing impact of climate change on crop production Assessing the impact of historical observed climate variability on crop production using SPOTCOMS Assessing crop - y ield sensitivity to changes in temperature and precipitation in SPOTCOMS Crop management data input SPOTCOMS 81 such as Lake Victoria, other smaller lakes and topographic features including large mountains like Mt. Ki l imanjaro in Tanzania and Mt. Rwenzori in Uganda. The region receives a bimodal annual rainfall ranging between 500 mm to over 2,500 mm (FEWSNet, 2010, 2012) and mean annual temperatures range betwe en 8.10C (at high elevations) to 320C (FEWSNet, 2010, 2012). Due to the favorable climate and good soils, agriculture is a major economic activity in the region and it sustains the majority of the population in the region. 3.2.2 Data sources 3.2.2.1 Histo rical data The study used observed daily precipitation data for 13 sites in Uganda, 12 sites in Kenya and monthly precipitation data for 12 sites from Tanzania, for 1980 - 2009 as shown in Figure 3.2 . These sites were strategically located in their respect ive countries and provide a nationwide representation of the different climatological zones. In order to have data for other locations of the study region, gridded precipitation datasets from sources that use non - tradition methods including remote sensing were compared with observed data from sites shown in Figure 2 in order to determine suitability to include in the study. The targeted gridded datasets evaluated included; CHIRPS dataset (Funk et al., 2013) with a 0.0250 and 0.050 resolution, NASA - POWER (NASA, 2013) dataset with a 10 resolution and AgMIP Coordinated Climate - Crop Modeling Pro ject (C3MP) data (Ruane et al., 2015). The NASA - POWER and the C3MP gridded datasets had daily precipitation, daily maximum and minimum temperature and solar radiation while the CHIRPS dataset only had daily rainfall. From the analysis, t he C3MP daily data set was selected because it better captures rainfall distribution and actual sequence of extreme events than other data products (Ruane et al., 2015) . 82 The AgMIP Coordinated Climate - Crop Modeling Project (C3MP) data is a historical (1980 - 2010) climate ser ies from a bias - shifted version of the NASA Modern Era Retrospective - analysis for Research and Applications (MERRA; (Rienecker et al., 2011) ) dataset. These s - ERRA data (Ruane, Goldberg, & Chryssanthacopoulos, 2015) are based on the MERRA and MERRA - Land (R eichle et al., 2011) outputs and are shifted to eliminate apparent monthly biases in comparison to an ensemble of gridded observational data from weather stations and satellites. These s - MERRA climate series also incorporate the NASA - GEWEX Solar Radiation Budget daily radiation data (Jeffrey W. White , Hoogenboom, Stackhouse Jr, & Hoell, 2008; Y. Zhang, Rossow, & Stackhouse, 2007) . The soils used in this study were provided by the harmonized world soils database (HWSD). The HWSD is a 30 arc - second raster database with over 16000 different soil mappin g units that combines existing regional and national updates of soil information worldwide (SOTER, ESD, Soil Map of China, WISE) with the information contained within the 1:5 000 000 scale FAO - UNESCO Soil Map of the World (FAO, 2012). The resulting raster database consists of 21600 rows and 43200 columns, which are linked to harmonized soil property data. The use of a standardized structure allows for the linkage of the attribute data with the raster map to display or query the composition in terms of soil units and the characterization of selected soil parameters (organic Carbon, pH, water storage capacity, soil depth, cation exchange capacity of the soil and the clay fraction, total exchangeable nutrients, lime and gypsum contents, sodium exchange percenta ge, salinity, textural class and granulometry). R eliability of the information contained in the database is variable: the parts of the database that still make use of the Soil Map of the World such as North America, Australia, West Africa and South Asia ar e considered less reliable, while most of the areas covered by SOTER databases are considered to have the highest reliability 83 (Central and Southern Africa, Latin America and the Caribbean, Central and Eastern Europe). Tables 3.1, 3.2, and 3.3 , in the appen dix, show that basic information about the study sites including the average rainfall, average temperature, and the soil characteristics. 84 (a) (b) Figure 3. 2 Study area. (a) Locations of study sites , (b) soil map of Africa (source: FAO, 2012) 85 Table 3. 1 Basic Descriptions for the locations in Uganda No . Name of location Average r ainfall (mm) Average Tmin ( O C) Average Tmax ( O C) Average Tmean ( O C) Soil texture Annual (mm ) FMAMJ (mm) ASON D (mm) Annual Tmin ( O C) Tmin FMAMJ ( O C) Tmin ASOND ( O C) Av. Annual Tmax ( O C) Av. Seasonal Tmax: FMAMJ ( O C) Av. Seasonal Tmax: ASOND ( O C) Av. Annual Tmean ( O C) Av. Seasonal Tmean: FMAMJ ( O C) Av. Seasonal Tmean: ASOND ( O C) 1 Gulu 1379 5 61 650 18.0 18.5 17.7 30.9 31.5 30.5 24.5 25.0 24.1 Clay (light) 2 Jinja 1370 645 58 3 16.3 16.8 16.0 27.6 27.7 27.5 21.9 22.2 21.7 Clay loam 3 Kasese 968 3 90 52 1 19.0 19.1 19.0 29.7 30.0 29.5 24.4 24.5 24.3 Loam 4 Kabale 1232 51 6 61 8 12.0 12.3 11.9 24.4 24.3 24.5 18.2 18.3 18.2 Clay (light) 5 Kitgum 1117 47 1 48 3 18.4 19.2 17.8 32.5 33.1 32.0 25.4 26.1 24.9 Sandy clay 6 Mbarara 990 376 549 14.7 14.9 14.7 27.8 27.8 27.8 21.2 21.4 21.2 Sandy clay 7 Masindi 1214 478 61 5 18.7 19.0 18.6 29.6 30.0 29.0 24.2 24.5 23.8 Clay (light) 8 Serere 129 4 59 7 573 19.3 19.9 19.0 30.5 30.9 30.1 24.9 25.4 24.6 Clay (light) 9 Soroti 12 50 565 54 9 19.9 20.4 19.6 30.7 31.0 30.4 25.3 25.7 25.0 Clay (light) 10 Tororo 161 6 768 68 2 16.8 17.3 16.5 28.9 29.1 28.8 22.9 23.2 22.6 S andy loam 11 Masaka 1214 59 2 527 16.4 16.6 16.5 27.0 27.1 26.9 21.7 21.8 21.7 Clay loam 12 Entebbe 135 4 670 56 3 18.4 18.8 18.2 27.1 27.3 27.0 22.8 23.1 22.6 Clay (light) 13 Arua 128 5 455 66 1 18.6 19.1 18.2 30.2 30.8 29.5 24.4 24.9 23.9 Sand 14 Namulo n ge 125 9 53 9 60 1 17.2 17.6 16.8 28.5 28.6 28.4 22.8 23.1 22.6 Clay loam 86 Table 3. 2 Basic Descriptions for the locations in Kenya No. Name of location Average rainfall (mm) Average Tmin ( O C) Average Tmax ( O C) Average Tmean ( O C) Soil texture Annual (mm) FMAMJ (mm) ASOND (mm) Annual Tmin ( O C) Tmin FMAMJ ( O C) Tmin ASOND ( O C) Av. Annual Tmax ( O C) Av. Seasonal Tmax: FMAMJ ( O C) Av. Seasonal Tmax: ASOND ( O C) Av. Annual Tmean ( O C) Av. Seasonal Tmean: FMAMJ ( O C) Av. Seasonal Tmea n: ASOND ( O C) 1 Dagoretti Corner 892 496 311 14.0 14.6 13.9 26.9 27.2 27.0 20.5 20.9 20.5 Clay (heavy) 2 Eldoret 1091 475 419 11.5 11.8 11.2 25.3 26.1 24.7 18.4 18.9 18.0 Clay (heavy) 3 Garissa 347 134 187 23.0 23.7 22.6 34.9 35.7 34.4 28.9 29.7 28.5 C lay (heavy) 4 Kisumu 1494 724 588 17.8 18.3 17.5 29.4 29.6 29.4 23.6 23.9 23.5 Clay (light) 5 Lamu 872 579 188 25.3 25.9 25.0 30.2 30.8 29.8 27.8 28.3 27.4 Clay loam 6 Lodwar 200 106 68 23.7 24.0 23.7 35.9 36.4 35.7 29.8 30.2 29.7 Sandy loam 7 Makindu 637 222 369 18.3 18.9 18.1 30.2 30.8 30.0 24.2 24.9 24.0 Sandy clay loam 8 Mandera 257 139 109 23.7 24.4 23.2 35.8 36.6 35.2 29.7 30.5 29.2 Sand 9 Marsabit 614 280 285 18.5 19.1 18.3 30.3 30.9 30.0 24.4 25.0 24.1 Clay (light) 10 Narok 751 412 228 10.6 11.4 10.1 24.8 25.0 24.9 17.7 18.2 17.5 Silt loam 11 Voi 620 237 332 20.9 21.5 20.6 31.5 32.1 31.3 26.2 26.8 25.9 Sandy clay loam 12 Wajir 312 164 128 23.0 23.7 22.5 34.9 35.7 34.2 28.9 29.7 28.4 Sand 87 Table 3. 3 Basic Desc riptions for the locations: Tanzania N o. Name of location s Average rainfall (mm) Average Tmin ( O C) Average Tmax ( O C) Average Tmean ( O C) Soil texture Annual (mm) FMAMJ (mm) ASOND (mm) Annual Tmin ( O C) Tmin - FMAMJ ( O C) Tmin - ASOND ( O C) Av. Annual Tma x ( O C) Av. Seasonal Tmax: FMAMJ ( O C) Av. Seasonal Tmax: ASOND ( O C) Av. Annual Tmean ( O C) Av. Seasonal Tmean: FMAMJ ( O C) Av. Seasonal Tmean: ASOND ( O C) 1 Arusha 1058 683 27 1 7.7 8.3 7.5 20.3 20.1 20.9 14.0 14.2 14.2 Loam 2 Bukoba 2082 1088 77 7 18.3 18.5 18.2 26.2 26.3 26.1 22.2 22.4 22.2 Sandy clay 3 Dar es Salaam 1112 63 6 385 21.7 22.2 21.1 31.4 30.9 32.0 26.5 26.6 26.6 Loamy sand 4 Dodoma 60 3 290 17 3 17.0 17.4 17.0 29.8 29.3 30.8 23.4 23.3 23.9 Loam 5 Kigoma 1049 471 42 6 19.5 19.6 19.8 28.4 28.1 28.9 24.0 23.9 24.4 Loam 6 Mbeya 1062 519 29 6 14.0 14.4 14.1 25.9 25.1 27.4 20.0 19.7 20.7 Clay (light) 7 Mtwara 104 9 60 3 22 6 21.0 21.1 21.0 31.2 30.3 32.7 26.1 25.7 26.9 Sandy clay loam 8 Musoma 99 2 515 36 6 16.6 16.7 16.7 29.2 28.8 29.8 22.9 22.7 23.2 Sand y loam 9 Mwanza 1011 44 4 440 18.8 18.8 19.2 27.7 27.7 28.0 23.3 23.2 23.6 Sandy clay loam 10 Same 55 4 319 175 18.5 19.0 18.2 30.5 30.5 30.8 24.5 24.8 24.5 Clay (heavy) 11 Songea 108 8 57 1 242 16.8 17.1 16.8 27.7 26.8 29.3 22.3 22.0 23.1 Sandy clay loam 12 Tabora 97 6 455 338 17.5 17.2 18.3 30.1 29.2 31.4 23.8 23.2 24.8 Sandy clay loam 88 3.2.2.2 Selection of projected f uture climate scenario and preparation of projected future climate data Future climate conditions for 38 locations in East Africa were us ed from the fifth phase of the Coupled Model Intercomparison Project (CMIP5) (Taylor, Stouffer, & Meehl, 2009) for the early century period 20 16 - 2045 , mid - century period 20 36 - 2065 and late century period 20 56 - 2085 which this research termed as the 2 030s, 2050s, and 2070 s for early century period, mid - century period , and late century period respectively . The baseline period which is also the current climate was taken as 1980 - 2009. Future climate data was selected for two representative concentration pathways (RCP 4.5 and 8.5) from 4 GCM models which were assessed under the Coordinated Regional Climate Downscaling Experiment (CORDEX) for the African region (Giorgi, Jones, & Asrar, 2009; C. Jones, 2013) . The model selection for this study was based on the model evaluation by Giorgi et al (2009), Jones (2013) and (Vincent O. Otieno & Richard O. Anyah, 2013; Vincent O. Otieno & R. O. Anyah, 2013) . T able 3.1 describes the four models therefore selected for this study . These were: 1. CSIRO - Mk3.6 (Rotstayn et al., 2009) with a horizontal resolution of 1.875 x 1.875 which uses a Rotstayn convective scheme (Rotstayn, 1998; Rotstayn et al., 2012) ; 2. MIROC5 (Watanabe et al., 2011) having a horizontal resolution of 1.4 x 1.4 and uses an Arakawa and Shubert convective scheme (Arakawa & Shubert, 1974) ; 3. MRI - CGCM3 (Yukimoto S et al, 2006) with a horizontal resolution of 1.125 x 1.12148 using a prognostic Arakawa - Shubert convective scheme (Pan & Randall, 1998) ; and, 4. NorESM1 - M (Bentsen et al., 2012; Seland, Iversen, Kirkevag, & Storelvmo, 2008) with a horizontal resolution of 2.5 x 1.895 and uses a Zhang and McFarlane con vective scheme (G. J. Zhang & McFarlane, 1995) . 89 Table 3. 4 Description of global circulation models used Name of model Horizontal r esolu tion Model expanded name Model group (or center) Reference CSIRO - Mk3.6 1.875 x 1.875 Commonwealth Scientific and Industrial Research Organization Mark, version 3.6.0 Commonwealth Scientific and Industrial Research Organization (CSIRO)/Queensland Climate C hange Centre of Excellence (QCCCE) Rotstayn et al., 2009 MIROC5 1.4 x 1.4 Model for Interdisciplinary Research on Climate, version 5 Japan Agency for Marine - Earth Science and Technology (JAMSTEC) Watanabe et al., 2011 MRI - CGCM3 1.125 x 1.12148 Meteorolog ical Research Institute Coupled Atmosphere Ocean General Circulation Model, version 3 Meteorological Research Institute (MRI) Yukimoto S et al, 2006 NorESM1 - M 2.5 x 1.895 Norwegian Earth System Model, version 1 (mid resolution) Norwegian Climate Centre (N CC) Bentsen et al., 2012; Seland, Iversen, Kirkevag, & Storelvmo, 2008 Biases do exist in all the convective schemes employed in GCMS and these e rrors of limit their utility for climate prediction and projection. In a study on tropical climates based on an intermodel empirical orthogonal function (EOF) analysis of tropical Pacific precipitation, Li and Xie (2014) found out that the excessive equatorial Pacific cold tongue and double intertropic al convergence zone (ITCZ) stood out as the most prominent er rors of the current generation of CGCMs. And that t he equatorial Pacific cold tongue bias was associated with deficient precipitation and surface easterly wind biases in the western half of the ba sin in CGCMs, but the errors we re absent in atmosphere - only models , indicating that the errors aro se from the interaction with the ocean via Bjerknes feedback. And f or the double ITCZ problem, excessive precipitation 90 south of the equator correlated well with excessive downward solar radiation in the Southern Hemisp here (SH) midlatitudes, an error traced back to atmospheric model simulations of cloud during austral spring and summer. Site - specific future temperature and rainfall data were stochastically downscaled for the four GCMs (CSIRO - Mk3.6, MIROC5, MRI - CGCM3, N orESM1 - M) for representative con centration pathways 4.5 and 8.5 emission scenarios using MarkSim (Peter G. Jones & Thornton, 2013) . MarkSim is a spatially explicit da ily weather generator that uses a third - order Markov chain process to generate daily rainfall, radiation, and temperature (P.G. Jones & Thornton, 2000) . It requires geographical coordinate and altitude to downscale and generate s da ily future data of a given site (Peter G. Jones & Thornton, 2013) . S tochastic weather generation has an advantage of mapping large - scale deterministic predictors fo r precipitation at small scales (Maraun et al. 2010; Chiew et al. 2010 ) to produce realizations of the expected small - scale rainfall field. Stochastic rainfall downscaling ( Ferraris et al. 2003 ) aims at generating synthetic spatiotemporal precipitation fie lds whose statistical properties are consistent with the small - scale statistics of observed precipitation, based only on knowledge of the large - scale precipitation field. Stochastic downscaling also has the potential for estimating uncertainties in rainfal l scenarios, by generating large ensembles of synthetic small - scale precipitation fields that can be compared with measured data (Brussolo et al. 2008). The major disadvantage is that stochastic downscaling is not a substitute for physically based models, because it relies on statistical models and algorithms to generate climate does not use physical process. In the weather generator, monthly climate anomalies (absolute changes) for monthly rainfall, mean daily maximum temperature and mean daily minimum tem perature w as calculated 91 for each time slice relative to the baseline climatology (1961 1990). The point of origin was designated 1975, being the midpoint of the 30 - year climate normal (Jones & Thornton, 2013). In this study, future temperature and rainfal l changes were downloaded from the website, ( http://gismap.ciat.cgiar.org/MarkSimGCM/ ) for 4 time slots 2010 - 2020 (2010s) as the control period, the early century period 2030 - 2040 (2030s), mid - centur y period 2050 - 2060 (2050s) and late century period 2070 - 2080 (2070s) which are centered around 2015, 2035, 2055 and 2075 respectively. We used the WorldClim dataset (P eter G. Jones & Thornton, 2013) as the base period and 2015 (2010 - 2020) as the control period. Differences between averages in temperature and percentage change in precipitation were used to describe future climatic changes in relation to the control peri od. This study , therefore, used three different groups of climate data. The historical observed data for the period 1980 2009, the stochastically generated current climate data for 30 years, and stochastically projected future climate data. The stochasti cally projected future climate data for the two representative concentration pathways, RCP4.5 and RCP8.5 was for the early century period 2016 - 2045, mid - century period 2036 - 2065 and late century period 2056 - 2085 which this research referred to as the 2 030s, 2050s, and 2070s for early century period, mid - century period, and late century period respectively. The historical observed data was used to examine climate trends for the historical period; the stochastically generated current climate was compare d with historical observed climate to test on the similarities between these two datasets especially on capturing seasonal variations , and the stochastical generated future climate data was used to generate the rela t ive changes in climate. Then all these d atasets were run in SPOTCOMS and a climate change impact assessment on sweet potato production was performed. 92 3.2.2.3 Wet or dry GCM models and hot or cool GCM models The spatial distribution of precipitation between the historical long - term mean World Clim dataset and the current period of 1980 - 2009 for the four GCMS showed small contrast. All the GCMs showed a consistent pattern in the distribution of precipitation with the Eastern parts of East Africa encompassing most areas in Kenya and high - elevat ion areas of Arusha through central Tanzania showing the least amount of annual precipitat ion below 500 mm (Figure 3.3 ). The GCMs showed slight variation and almost hard to make a distinction between RCP4.5 and RCP8.5. The temperatures from current control period of 2015 are also qui te similar in the spatial distribut ion and variation (Figure 3.4 ) . For all the models, the Eastern strip of East Africa and northern parts are the warmest with an average temperature in the range of 28 - 30 0 C. Generally, all GCMs showed similarly high temperatures . 93 CSIRO - Mk3.6 MIROC5 MRI - CGCM3 NorESM1 - M Figure 3. 3 Spatial distribution of stochastically generated current mean annual rainfall over East Africa for the period1980 - 2009 Rainfall (mm) Rainfall (mm) Rainfall (mm) Rainfall (mm) 94 CSIRO - Mk3.6 MIROC5 MRI - CGCM3 NorESM1 - M Figure 3. 4 Spatial distribution of stochastically generated current mean annual temperature over East Africa Temperatur e ( 0 C) Temperatur e ( 0 C) Temperatur e ( 0 C) Temperatur e ( 0 C) 95 3.2.3 Trend analysis Trend magnitudes of historical climate and derived simulated sweet potato yields and their significance were calculated across East Africa using a non - parametric statistic that follows a methodology by (Sen, 1968) was used. This nonparametric method is less sensitivity to outliers and tests for a trend in a time series wit hout specifying whether the trend is linear or nonlinear (Partal & Kahya, 2006; Yenigun, Gumus, & H., 2008) . And it is a good choice for analyzing trends from variables that are not normally d istributed especially precipitation. This study follows the same steps as those used in (Andresen, Alagarswamy, Rotz, Ri tchie, & LeBaron, 2001) . Briefly, the trend magnitude statistic (B) is defined as [1 5 ] Were D ij = ( x j x i )/( j - i ) for all possible pairs ( x i , x j i < j n , and n the number of observations in the series. For the case of this study, n = 30. The nonparametric Mann - statistic (Kendall, 1975) was used to determine the signif icance of the trends. The null hypothesis, H 0 , was that the data in the series of interest were a sample of n independent and identically distributed variables. The alternative hypothesis, H 1 , of the two - sided test was that the distribution of x i and x j ere analyzed with a two - sided test for trend, with H 0 accepted if the standard normal variate of S was less than or equal to the standard normal cumulative distribution function for a given level of significance . The power of this test fo r sample sizes > 10 has been shown to be nearly as great as that of the more traditional t - statistic, which assumes normality (Hirsch, Slack, & Smith, 1982; Kendal l, 1975) . 96 Twelve variables were used in the trend analysis including seasonal precipitation and mean temperature both the long rains season February June (FMAMJ) and the short rains season August December ( ASOND ) and the seasonal yields for the four sweet potato cultivars NASPOT 1 (na1), NASPOT 10 0 (na10), NASPOT 11 (na11) and SPK004 Kakamega (spk). The variables are abbreviated as PPT - FMAMJ, PPT - ASOND, Tmean - FMAMJ, Tmean - ASOND, na1 - ASOND, na1 - ASOND, na10 - ASOND, spk004 - ASOND, na1 - FMAMJ, na10 - FMA MJ, na11 - FMAMJ, spk004 - FMAMJ 3.2.4 Sensitivity analysis Sensitivity analysis of SPOTCOMS model was performed for changes in temperature and precipitation using observed historical climate and soils data . The aim was to analy z e the effect of changes in c limatic variability on sweet potato growth and development, as simulated by SPOTCOMS. The analysis was made for the four sweet potato cultivars, NASPOT 1, NASPOT 10 0, NASPOT 11 and Kakamega (SPK 004) following a methodology used by Katz and Brown (1992), Mearns et al. (1992), Semenov and Porter (1995 . Various research groups (Katz and Brown, 1992; Mearns et al., 1992; Semenov and Porter, 1995) have conducted studies on the effects of climatic change on crop growth and development. Katz and Brown (1992) fou nd that extreme climatological events are relatively more dependent on changes in climatic variability than on changes in mean values, especially for hot spells and droughts. Mearns et al. (1992) and Semenov and Porter (1995) investigated how changes in cl imatic variability could affect wheat production and performed sensitivity analyses using the CERES - Wheat crop simulation model and historical climatic data perturbed to increase the inter - annual variance of the climatic variables. Semenov and Porter (1995 ) 97 A sensitivity analysis was performed to identify the major climatic constraints for sweet potato development and yield. The analysis was carried out for Namulonge in Uganda, Mbeya in Tanzania and Dagoretti Corner in Kenya. These locations were selected as representative sites for climate across East Africa. The choice for selection of these locations was based on the premise that the three locations representative average climatic patterns from their respective countries. The daily rainfall totals of eac h of these locations was changed in the range of - /+ 50% at increments of - /+10% and the daily minimum and maximum temperatures were also changed by - /+ 1 0 C increments in the range - /+ 5 0 C following a methodology by Katz and Brown (1992), Mearns et al. (19 92) and Semenov and Porter (1995) . This , therefore , led to 121 different combinations of model runs for a particular location. For temperatures, the maximum and minimum temperatures were either decreased or increased together. For example, when the maximu m temperature was increase d by 2 0 C, the same change was performed on the minimum temperatures. To perform an evaluation o f the sensitivity analysis, graphs were plotted using Microsoft Excel and surface diagrams were plotted using Sigma Plot. 3.2.5 Cli mate change impact assessment In order to investigate the impact of future climate change, the SPOTCOMS was used to stimulate growth, development, and yield of sweet potato across East Africa. SPOTOMS is a process - based crop model that s t imulates growth, development, and yield of sweet potatoes (Mithra & Somasundaram, 2008) . The SPOTCOMS and the experimentation that was used to calibrate it for the East African region was described in chapter two. For all the location s used in this study, the planting date as defined for model simulation was taken to be the beginning of the growing season which was February 1 st for the February to May season and August 1 st for the 98 August to December growing season. The planting date is very important because when it is changed, the model simulates different results as it will be taking different climate data corresponding to the modified growing season. The model was used to simulate the phenology and yield of sweet potato , in response to climatic factors, namely precipitation, maximum and minimum temperatures and solar radiation. SPOTCOMS employs soil data, crop management data, and daily meteorological data as an input to simulate daily leaf area index (LAI) and vegetation status param eters, biomass production, and final yield. The daily meteorological data include solar radiation, rainfall, and maximum and minimum air temperatures. The major soil data include soil type, initial soil water content, relative root distribution, soil pH, b ulk density, and soil organic matter. The crop management data include variety, planting date, plant density, irrigation, and fertilizer application. Crop genetic coefficients included in the model relates to the photoperiod sensitivity (thermal time), gro wth stages of sweet potatoes, development of vines, describes the development of roots, branching of the crop, the number of leaves on a sweet potato plant and the leaf area of sweet potatoes . An assessment of climate change impact on the yield for four re presentative cultivars of sweet potatoes NASPOT 1, NASPOT 10 0, NASPOT 11, and SPK004 (Kakamega) was performed. The model was calibrated and validated for these cultivars using experimental data collected at Namulonge Crops Resources Research Institute (Na CRRI) in Uganda, details of the whole process were presented in chapter two. In this impact assessment, SPOTCOMS was run using three different types of climate data. One dataset was historical observed data for the period 1980 2009, the second dataset was the stochastically generated current climate data for 30 years, and third, stochastically projected future climate data. The stochastically projected future climate data for the two representative 99 concentration pathways, RCP4.5 and RCP8.5 was for the e arly century period 2016 - 2045, mid - century period 2036 - 2065 and late century period 2056 - 2085 which this research referred to as the 2030s, 2050s, and 2070s for early century period, mid - century period, and late century period respectively. 3.3 Res ults 3.3.1 Trend analysis The results from the seasonal trend analys is of precipitation, Figure 3.5 and Figure 3.6 , indicated that the long - rainfall - season (February - June) intensity decreased northwards with highest increases of more than 3.5 mm/yr in so uthern parts of Tanzania and decreases of more than 3.9 mm/yr in northern parts of Kenya and Uganda. The trail of decreasing precipitation largely falls along the eastern arm of the East African rift valley from northern Kenya into Tanzania. The average te mperatures trend of the same season, on the other hand, increased from the south - west direction towards the north - west direction in East A frica (Figures 3.7 and Figure 3.8) . The magnitudes of this in crease varied from 0 0 C on the Indian Ocean to more than 1 .5 0 C over the 30 year period in most parts of Uganda and north - west parts of Tanzania and Kenya . Sweet potato yields in the February - June season, for the three cultivars, NASPOT 1 , NASPOT 10 O and NASPOT 11 (Table 3.5) increased with magnitudes of 0.6 - 3 t/ ha (or 0.02 - 0.11 t/ha/yr) in most parts of Uganda and Tanzania with the exception of the north - eastern parts of Tanzania. And, the yields decreased in most parts of Kenya especially the areas along the eastern arm of the rift - valley by magnitudes of more than 0.11 t/ha/yr or 3.3 t/ha. The SPK004 sweet potato cultivar showed a sligh tly different trend (Table 3.5). The increase in SPK004 yields largely followed an East to 100 West pattern across East Africa. And the decrease in SPK004 yields was not as large as the one shown by the other cultivars. 101 0 200 400 600 800 1000 1980 1987 1994 2001 2008 Rainfall (mm) Year Garissa 0 200 400 600 800 1000 1980 1987 1994 2001 2008 Rainfall (mm) Year Mandera 0 200 400 600 800 1000 1980 1987 1994 2001 2008 Rainfall (mm) Year Eldoret 0 200 400 600 800 1000 1980 1987 1994 2001 2008 Rainfall (mm) Year Arua 0 200 400 600 800 1000 1980 1987 1994 2001 2008 Rainfall (mm) Year Kigoma 0 200 400 600 800 1000 1980 1987 1994 2001 2008 Rainfall (mm) Year Mbeya 0 200 400 600 800 1000 1980 1987 1994 2001 2008 Rainfall (mm) Year Dodoma 0 200 400 600 800 1000 1980 1987 1994 2001 2008 Rainfall (mm) Year Dar es Salaam 0 200 400 600 800 1000 1980 1987 1994 2001 2008 Rainfall (mm) Year Mtwara 0 200 400 600 800 1000 1980 1987 1994 2001 2008 Rainfall (mm) Year Lodwar Figure 3. 5 Trend in average rainfall over the February to June (FMAMJ) season for selected locations over the period 1980 - 2009 across East Africa 102 Figure 3. 6 Trend in average rainfall over t he August to December (ASOND) season for selected locations over the period 1980 - 2009 across East Africa. The * represent a significant trend at 0.05 level 0 200 400 600 800 1000 1980 1987 1994 2001 2008 Rainfall (mm) Year Mtwara 0 200 400 600 800 1000 1980 1987 1994 2001 2008 Rainfall (mma) Year Dodoma 0 200 400 600 800 1000 1980 1987 1994 2001 2008 Rainfall (mm) Year Eldoret * 0 200 400 600 800 1000 1980 1987 1994 2001 2008 Rainfall (mm) Year Arua * 0 200 400 600 800 1000 1980 1987 1994 2001 2008 Rainfall (mm) Year Kigoma 0 200 400 600 800 1000 1980 1987 1994 2001 2008 Rainfall (mm) Year Mandera 0 200 400 600 800 1000 1980 1987 1994 2001 2008 Rainfall (mm) Year Garissa 0 200 400 600 800 1000 1980 1987 1994 2001 2008 Rainfall (mm) Year Mbeya 0 200 400 600 800 1000 1980 1987 1994 2001 2008 Rainfall (mm) Year Dar es Salaam 0 200 400 600 800 1000 1980 1987 1994 2001 2008 Rainfall (mm) Year Lodwar 103 10.0 15.0 20.0 25.0 30.0 35.0 40.0 1980 1983 1986 1989 1992 1995 1998 2001 2004 2007 Temperature (deg C) Year Mtwara * Tmin: FMAMJ Tmax: FMAMJ Tmean: FMAMJ 10.0 15.0 20.0 25.0 30.0 35.0 40.0 1980 1983 1986 1989 1992 1995 1998 2001 2004 2007 Temperature (mm) Year Mandera * 10.0 15.0 20.0 25.0 30.0 35.0 40.0 1980 1983 1986 1989 1992 1995 1998 2001 2004 2007 Temperature (deg C) Year Arua *** 10.0 15.0 20.0 25.0 30.0 35.0 40.0 1980 1983 1986 1989 1992 1995 1998 2001 2004 2007 Temperature (deg C) Year Kigoma *** 10.0 15.0 20.0 25.0 30.0 35.0 40.0 1980 1983 1986 1989 1992 1995 1998 2001 2004 2007 Temperature (deg C) Year Dodoma Tmin: FMAMJ Tmax: FMAMJ Tmean: FMAMJ 10.0 15.0 20.0 25.0 30.0 35.0 40.0 1980 1983 1986 1989 1992 1995 1998 2001 2004 2007 Temperature (deg C) Year Lodwar *** 10.0 15.0 20.0 25.0 30.0 35.0 40.0 1980 1983 1986 1989 1992 1995 1998 2001 2004 2007 Temperature (mm) Year Garissa 10.0 15.0 20.0 25.0 30.0 35.0 40.0 1980 1983 1986 1989 1992 1995 1998 2001 2004 2007 Temperature (deg C) Year Eldoret ** 10.0 15.0 20.0 25.0 30.0 35.0 40.0 1980 1983 1986 1989 1992 1995 1998 2001 2004 2007 Temperature (deg C) Year Mbeya * Tmin: FMAMJ Tmax: FMAMJ Tmean: FMAMJ Figure 3. 7 Trend in average temperature over the February to June (F MAMJ) season for selected locations over the period 1980 - 2009 across East Africa. *, **. And *** represent a significant trend at 0.05, 0.01, 0.001 levels respectively. 104 10.0 15.0 20.0 25.0 30.0 35.0 40.0 1980 1983 1986 1989 1992 1995 1998 2001 2004 2007 Temperature (deg C) Year Arua 10.0 15.0 20.0 25.0 30.0 35.0 40.0 1980 1983 1986 1989 1992 1995 1998 2001 2004 2007 Temperature deg C) Year Kigoma *** 10.0 20.0 30.0 40.0 1980 1983 1986 1989 1992 1995 1998 2001 2004 2007 Temperature (deg C) Year Mbeya * Tmin: ASOND Tmax: ASOND Tmean: ASOND 10.0 15.0 20.0 25.0 30.0 35.0 40.0 1980 1983 1986 1989 1992 1995 1998 2001 2004 2007 Temperature (mm) Year Mtwara *** Tmin: ASOND Tmax: ASOND Tmean: ASOND 10.0 15.0 20.0 25.0 30.0 35.0 40.0 1980 1983 1986 1989 1992 1995 1998 2001 2004 2007 Temperature (deg C) Year Dodoma Tmin: ASOND Tmax: ASOND Tmean: ASOND 10.0 15.0 20.0 25.0 30.0 35.0 40.0 1980 1983 1986 1989 1992 1995 1998 2001 2004 2007 Temperature (deg C) Year Eldoret *** 10.0 15.0 20.0 25.0 30.0 35.0 40.0 1980 1983 1986 1989 1992 1995 1998 2001 2004 2007 Temperature (mm) Year Garissa 10.0 15.0 20.0 25.0 30.0 35.0 40.0 1980 1983 1986 1989 1992 1995 1998 2001 2004 2007 Temperature (deg C) Year Lodwar 10.0 15.0 20.0 25.0 30.0 35.0 40.0 1980 1983 1986 1989 1992 1995 1998 2001 2004 2007 Temperature (mm) Year Mandera Figure 3. 8 Trend in average temperature over the August to December (ASOND) season for selected locations over the period 1980 - 2009 across East Africa. *, **. And *** represent a significant trend at 0.05, 0.01, 0.001 levels respectively. 105 The results from the seasonal trend analysis of precipi tation, Figure 3.8 , indicated that the short - rainfall - season (August - December) intensity inc reased northwards with highest increases of more than 1mm/yr across most parts of Uganda, Kenya and areas surrounding Lake Victoria. There were decreases in precipitation of more than 2.34 mm/yr (> 60mm) in areas alo ng the Indian Coast (Figure 3.8 ). The av erage temperatures trend of the August - December season showed increases of 1.2 0 C in most parts of East Africa with the exception of a few areas along the Indian Ocean and Eastern Kenya . Sweet potato yield trends in the August - December season for the four c ultivars, NASPOT , NASPOT 10 O (Figure 3.3d) and NASPOT 11 and SPK004 Table 3.5 increased from the south to north with magnitudes of 0.04 t/ha/yr to 0.5 t/ha/yr (1.2 - 15 t/ha). Results from significance tests in the trends of variables indicated that the mean temperature for both seasons was largely significant at 0.001 level of significance (S.L) for most areas across East Africa ( Figure 3.8 ). Precipitation for the August - December season (PPT - ASOND) recorded some significance at locations, Gulu (0.01 S.L) and Eldoret (0.05 S.L). The variables for yields for the four cultivars across both seasons did not show significant trends accept at a few scattered locations. 106 Table 3. 5 for the period 1980 - 2009 by site and season. No Location Country Precipitation (mm/yr) Mean Temperature ( 0 C/yr) Simulated sweet potato yield (kg/ha/yr) Short rainfall season ( ASOND ) Long rainfall season (FMAMJ) FMAMJ ASOND FMAMJ ASOND na1 na1 0 na11 spk004 na1 na10 na11 spk004 1 Arua U G A N D A - 1.067 1.350** 0.037*** 0.043 0.015 ** - 0.039 ** 0.004 ** - 0.053 ** 0.267** 0.275** 0.494** 0.158* * 2 Entebbe - 0.587 0.436 ** 0.045*** 0.038*** 0.258 ** 0.200 ** 0.290 ** 0.172 ** 0.093 ** 0.059 ** - 0.035 ** 0.04 9 ** 3 Gulu - 3.256 5.608* * 0.041** * 0.026* 0.199 ** 0.089* * 0.672* * 0.138** 0.089 ** 0.058 ** - 0.056 ** 0.064 ** 4 Jinja 0.320 1.018 ** 0.052*** 0.050*** 0.360* * 0.360* * 0.298 ** 0.254* * 0.355** 0.037 ** 0.059 ** 0.061 ** 5 Kabale - 0.211 - 1.052 ** 0.041*** 0.040 *** 0.153 ** 0.158 ** 0.133 ** 0.160 ** 0.177* * 0.184* * 0.158* * 0.160 ** 6 Kasese - 0.613 1.083 ** 0.046*** 0.028 0.328 ** 0.111 ** 0.374 ** 0.075 ** 0.138 ** 0.111 ** 0.110 ** 0.063 ** 7 Kitgum - 1.650 2.843 ** 0.032** * 0.018 0.216 ** 0.310 ** 0.253 ** 0.160* * 0.066 ** 0. 030 ** 0.110 ** - 0.020 ** 8 Masaka - 5.117 - 4.186 ** 0.054*** 0.048*** - 0.003 ** 0.010 ** 0.014 ** - 0.040 ** - 0.013 ** - 0.039 ** - 0.250 ** 0.004 ** 9 Masindi - 2.700 1.452 ** 0.056*** 0.044*** 0.247 ** 0.180 ** 0.297 ** 0.085 ** 0.124 ** 0.110 ** 0.111 ** 0.104 ** 10 Mbarar a 1.800 1.282 ** 0.044*** 0.040*** 0.240 ** 0.183 ** 0.267 ** 0.111 ** 0.219* * 0.239* * 0.184* * 0.227* * 11 Namulonge 0.183 3.360 ** 0.047*** 0.040*** 0.356* * 0.390** 0.449* * 0.217 ** 0.270 ** 0.237 ** 0.263 ** 0.245 ** 12 Serere - 1.620 3.382 ** 0.039*** 0.028** 0. 351 ** 0.252 ** 0.369 ** 0.062 ** 0.023 ** - 0.021 ** - 0.076 ** 0.044 ** 13 Soroti - 1.091 2.408 ** 0.046*** 0.052*** 0.177 ** 0.097 ** 0.170 ** 0.126 ** 0.001 ** 0.012 ** - 0.059 ** 0.030 ** 14 Tororo 0.100 5.800 ** 0.042*** 0.043*** 0.351* * 0.188 ** 0.676* * 0.005 ** 0.082 * * 0.127 ** 0.058 ** 0.109 ** 15 Arusha T A N Z A N I A - 3.569 - 0.839 ** 0.030** * 0.052*** 0.088 ** 0.090 ** 0.089 ** 0.105 ** 0.104 ** 0.161 ** - 0.008 ** 0.150* * 16 Bukoba 0.024 - 1.727 ** 0.044*** 0.043*** 0.149 ** 0.210 ** 0.197 ** 0.056 ** 0.015 ** 0.044 ** 0.066 ** 0.0 31 ** 17 Dar es Salaam 0.800 - 6.136 ** - 0.005 *** 0.009 - 0.067 ** - 0.049 ** - 0.170 ** - 0.091 ** 0.093 ** 0.089 ** 0.143 ** 0.002 ** 18 Dodoma - 1.433 1.136 ** 0.009 *** 0.027 0.067 ** 0.063 ** 0.059 ** 0.057 ** - 0.088 ** - 0.047 ** - 0.068 ** 0.076 ** 19 Kigoma 0.063 - 3.086 ** 0.034*** 0.031*** 0.021 ** - 0.014 ** - 0.069 ** - 0.008 ** 0.042 ** 0.044 ** 0.202 ** 0.064 ** 20 Mbeya 2.390 - 0.408 ** 0.027* ** 0.021* - 0.019 ** - 0.030 ** - 0.003 ** - 0.001 ** 0.203 ** 0.026 ** 0.383 ** 0.078 ** 21 Mtwara - 1.586 - 2.662 ** 0.020* ** 0.020*** - 0.067 ** - 0. 043 ** - 0.073 ** - 0.023 ** 0.001 ** - 0.005 ** 0.094 ** - 0.036 ** 22 Musoma - 7.080 3.950 ** 0.029*** 0.019 0.322 ** 0.283 ** 0.266 ** 0.274 ** - 0.340 ** - 0.189 ** - 0.603* * - 0.032 ** 107 Table 3.5 ( Cont d ) No Location Country Precipitation (mm/yr) Mean Temperature ( 0 C/yr) Simulated sweet potato yield (kg/ha/yr) Short rainfall season ( ASOND ) Long rainfall season (FMAMJ) FMAMJ ASOND FMAMJ ASOND na1 na10 na11 spk004 na1 na10 na11 spk004 23 Mwanza T A N Z A N I A - 2.700 3.472 ** 0.024** * 0.018* 0.136 ** 0.169 ** 0.215 * * 0.125 ** - 0.015 ** 0.006 ** - 0.193 ** 0.033 ** 24 Same - 1.537 - 0.506 ** 0.010 *** 0.030*** 0.011 ** 0.009 ** 0.010 ** 0.003 ** - 0.157 ** - 0.127 ** - 0.135 ** - 0.152 ** 25 Songea 2.952 - 0.209 ** 0.021* ** 0.044*** - 0.017 ** - 0.019 ** - 0.013 ** - 0.019 ** 0.050 ** 0.040 ** 0.0 74 ** 0.007 ** 26 Tabora - 0.100 0.208 ** 0.021 *** 0.032*** - 0.022 ** 0.004 ** - 0.022 ** 0.023 ** 0.070 ** 0.099 ** 0.175 ** 0.055 ** 27 Dagoretti Corner K E N Y A - 2.961 - 0.280 ** 0.019 *** 0.040** 0.008 ** 0.007 ** 0.008 ** 0.020 ** - 0.071 ** 0.005 ** - 0.214 ** - 0.023 ** 2 8 Eldoret - 3.831 5.256* * 0.029** * 0.038*** 0.224** 0.230** 0.209** 0.257* * - 0.174 ** - 0.204 ** - 0.156 ** - 0.209 ** 29 Garrisa 0.186 2.170 ** 0.007 *** 0.013 0.087 ** 0.064 ** 0.065 ** 0.071 ** 0.004 ** 0.002 ** 0.002 ** - 0.002 ** 30 Kisumu - 1.218 5.323 ** 0.028*** 0 .029*** 0.287 ** 0.292 ** 0.299 ** 0.185 ** 0.067 ** 0.083 ** - 0.061 ** 0.176 ** 31 Lamu - 0.257 0.245 ** 0.022* ** 0.020** 0.038 ** 0.028 ** 0.030 ** 0.026 ** - 0.040 ** 0.028 ** - 0.057 ** 0.007 ** 32 Lodwar 0.741 1.827 ** 0.033*** 0.010 0.022 ** 0.020 ** 0.021 ** 0.024 ** 0. 054 ** 0.047 ** 0.050 ** 0.059 ** 33 Makindu - 0.290 - 0.567 ** 0.020 *** 0.034*** - 0.006 ** - 0.020 ** 0.018 ** 0.025 ** - 0.091 ** - 0.071 ** - 0.071 ** - 0.067 ** 34 Mandera - 1.300 3.079 ** 0.029* ** 0.008 0.259 ** 0.239 ** 0.206 ** 0.225 ** - 0.175 ** - 0.154 ** - 0.150 ** - 0.125 * * 35 Marsabit - 4.925 0.947 ** 0.030** * 0.040*** 0.024 ** - 0.009 ** 0.001 ** 0.026 ** - 0.073 ** - 0.215 ** - 0.287* * - 0.133 ** 36 Narok - 0.792 1.853 ** 0.027*** 0.038*** 0.046 ** 0.047 ** 0.046 ** 0.038 ** 0.093 ** 0.123 ** 0.083 ** 0.092 ** 37 Voi 0.386 - 1.160 ** 0.006 * ** 0.019* 0.104 ** 0.104 ** 0.207 ** 0.079 ** 0.026 ** - 0.008 ** 0.008 ** 0.056 ** 38 Wajir - 0.031 2.250 ** 0.024* ** 0.030* 0.192 ** 0.168 ** 0.161 ** 0.160 ** 0.017 ** 0.017 ** 0.030 ** 0.032 ** * Significant at the 0.05 level. ** Significant at the 0.01 level. *** Sig nificant at the 0.001 level. 0 C/yr); na1, NASPOT 1; na10, NASPOT 10 0; spk004, SPK004 9Kakamega); na11, NASPOT 11; are yield (t/ha/yr) FMAMJ, February June; ASOND, August - December 108 3.3.2 Sensitivity a nalysis The general observation from results in the sensitivity analysis shows that sweet potato root yields increase with increasing precipitation and inc reasing temperatures (Figures 3.9 and 3.10 ). The results also indicate the dry and cooler places rec ord the lower yields than wet and cooler regions. Likewise, dry and hotter regions produce lower yields than wet and hotter regions. On comparing the sensitivity analysis between seasons, some differences were observed in the magnitude of sweet potato yie ld for Dagoretti Corner, Kenya and in Mbeya, Tanzania while no big difference in variation was shown for cultivars at Namulonge, Uganda. Indeed, yields for the August - December season (Figure 3.9) for Mbeya and Dagoretti are only half those of their corresp onding February - June season (Figure 3.10) in the two countries. A comparison between location sensitivity shows that the seasonal distribution of precipitation and maximum and minimum temperatures largely determines the trends in the variation of the yiel d. For example, Namulonge in Uganda shows a wider variation of yields for a different combination of precipitation and temperatures compared to Dagoretti in Kenya and Mbeya in Tanzania. There were no major differences in variation of sensitivity results am ong different cultivars. In most of the graphs, the yield curves appeared to simply continue rising with increasing precipitation and temperature, although they seemed to begin appearing as though they were reaching the optimum threshold above which sweet potato growth is inhibited. An extensive summary of the sweet potato yields is attached to a table in the appendix. 109 (a) NASPOT 1, Dagoretti Corner (b) SPK 004, Dagoretti Corner (c) NASPOT 1, Mbeya (d) SPK oo4, Mbeya (e) NASPOT 1, Namulonge (f) SPK 004, Nam ulonge Figure 3. 9 Climate sensitivity of sweet potato crop model ( SPOTCOMS ) for the August to December season for 3 locations; Dagoretti Corner in Kenya, Mbeya in Tanzania and Namulonge in Uganda for two sweet potato cultivars NASPOT 1 and SPK 004 . The simulated sweet potato yield was as a result of rainfall and temperatures which had been either increased or decreased by a given proportional change in rainfall and a changed amount in both minimum and maximum temperatures. Sweet potato yield (kg/ha) Sweet potato yield (kg/ha) Sweet potato yield (kg/ha) Sweet pot ato yield (kg/ha) Sweet potato yield (kg/ha) Sweet potato yield (kg/ha) 110 (a) N ASPOT 1, Dagoretti Corner (b) SPK 004, Dagoretti Corner (c) NASPOT 1, Mbeya (d) SPK 004, Mbeya (e) NASPOT 1, Namulonge (f) SPK 004, Namulonge Figure 3. 10 Climate sensitivity of sweet potato crop model (SPOTCOMS) for the February to June (FMAMJ) season for 3 locations; Dagoretti Corner in Kenya, Mbeya in Tanzania and Namulonge in Uganda for two sweet potato cultivars NASPOT 1 and SPK 004. The simulated sweet potato yield was as a result of rainfall and temperatures which had been e ither increased or decreased by a given proportional change in rainfall and a changed amount in both minimum and maximum temperatures. Sweet potato yield (kg/ha) Sweet potato yield (kg/ha ) Sweet potato yield (kg/ha) Sweet potato yield (kg/ha) Sweet potato yield (kg/ha) Sweet potato yield (kg/ha) 111 3.3.2 Projected future climate and sweet potato production model results 3.3.2.1 Future climate projections for East A frica Most parts of Kenya and Tanzania and northern Uganda are projected to receive an increase in precipitation of more than 50 mm while western and central Uganda and southern Tanzania will experience a decrease in precipitati on of more than 70 mm in th e 203 0s (Figures3.15 and 3.16 ) . The drier GCM shows a decrease in precipitation in most parts of Tanzania and Uganda South eastern parts of Kenya and parts of TZ coastline are projected to receive more rainfall of 50mm compared to the 2010s by NorESM - 1. Te mperatures are projected to increase in the range 0.9 - 1.8 0 C in western regions of East Africa enclosing Uganda, most parts of TZ and some par ts of Kenya . In the 2070s, central parts of East Africa are projected to receive more than 90mm of annual precipit ation (Figures 3.15, 3.16, 3.17 and 3.18 ). The southern parts of Tanzania will experience a reduction in annual precipitation of more than 75 mm as shown by both GCMs and their corresponding RCPs. For temperatures in the 207 0s , the RCP4.5 scenario (Figures 3.13 ) show s slight increases in temperature with larger values in western regions of East Africa including Uganda and parts of Tanzania of about 1.3 0 C. Higher temperature changes of 2.3 - 4 0 C are projected for most parts of East Africa by MRICGC3 - M RcP8.5 ( Figure 3.14 ) and south - western parts of Tanzani a in NorESM - 1 RCP8.5 (Figure 3.14 ). 112 Annual average temperature CSIRO - Mk3.6 MIROC5 MRI - CGCM3 NorESM1 - M Figure 3. 11 Historical mean annual temperature and relat ive changes of mean annual temperatures for the 2030s for RCP4.5 for four GCMs Changes in mean temperature ( o C) Mean Temperatur e ( o C) Changes in mean temperature ( o C) Changes in mean temperature ( o C) Changes in mean temperature ( o C) 113 Annual average temperature CSIRO - Mk3.6 MIROC5 MRI - CGCM3 NorESM1 - M Figure 3. 12 Historical mean annual temperature and relative cha nges of mean annual temperatures for the 2030s for RCP8.5 for four GCMs Changes in mean temperature ( o C) Mean Temperatur e ( o C) Changes in mean temperature ( o C) Changes in mean temperature ( o C) Changes in mean temperature ( o C) 114 Annual average temperature CSIRO - Mk3.6 MIROC5 MRI - CGCM3 NorESM - M Figure 3. 13 Historical mean annual temperature and relative changes of mean annual temperatures for the 2070s for RCP4.5 for four GCMs Changes in mean temperature ( o C) Mean Temperatur e ( o C) Changes in mean temperature ( o C) Change s in mean temperature ( o C) Changes in mean temperature ( o C) 115 Annual average temperature CSIRO - Mk3.6 MIROC5 MRI - CGCM3 NorESM1 - M Figure 3. 14 Historical mean annual temperature and relative changes of mean annual temperatures for the 2070s for RCP8.5 for four GCMs Changes in mean temperature ( o C) Mean Temperatur e ( o C) Changes in mean temperature ( o C) Changes in mean temperature ( o C) Changes in mean temperature ( o C) 116 Annual rainfall CSIRO - Mk3.6 MO3045 MRI - CGCM3 NorESM1 - M Figure 3. 15 Historical mean annual rainfall and relative changes of mean annual rainfall for the 2030s for RCP4.5 for four GCMs Change s in rainfall (mm) Change s in rainfall (mm) Change s in r ainfall (mm) Change s in rainfall (mm) Rainfall (mm) 117 Annual rainfall CSIRO - Mk3.6 MO3085 MRI - CGCM3 NorESM1 - M Figure 3. 16 Historical mean annual rainfall and relative changes of mean annual rainfall for the 2030s for RCP8.5 fo r four GCMs Change s in rainfall (mm) Change s in rainfall (mm) Change s in rainfall (mm) Change s in rainfall (mm) Rainfall (mm) 118 Annual rainfall CSIRO - Mk3.6 MO7045 MRI - CGCM3 NorESM1 - M Figure 3. 17 Historical mean annual rainfall and relative changes of mean annual rainfall for the 2070s for RCP4.5 for four GCMs Change s in rainfall (mm) Change s in rainfall (mm) Change s in rainfall (mm) Change s in rainfall (mm) Rainfall (mm) 119 Annual rainfall CSIRO - Mk3.6 MO7085 MRI - CGCM3 NorESM1 - M Figure 3. 18 Historical mean annual rainfall and relative changes of mean annual rainfall for the 2070s for RCP8.5 for four GCMs Change s in rainfall (mm) Change s in rainfall (mm) Change s in rainfall (mm) Change s in rainfall (mm) Rainfall (mm) 120 Historical average temperature CSIRO - Mk3.6 MIROC5 MRI - CGCM3 NorESM1 - M Figure 3. 19 Historical mean seasonal temperature and relative changes of mean temperatures for August to December for the 2030s for RCP4.5 for four GCMs . Changes in mean temperature ( o C) Changes in mean temperature ( o C) Changes in mean temperature ( o C) Changes in mean temperature ( o C) Mean Temperatur e ( o C) 121 Historical avera ge temperature CSIRO - Mk3.6 MIROC5 MRI - CGCM3 NorESM1 - M Figure 3. 20 Historical mean seasonal temperature and relative changes of mean temperatures for August to December for the 2030s for RCP8.5 for four GCMs Changes in mean temperature ( o C) Changes in mean temperature ( o C) Changes in mean temperature ( o C) Changes in mean temperature ( o C) Mean Temperatur e ( o C) 122 Historical average temperature CSIRO - Mk3.6 MIROC5 MRI - CGCM3 NorESM1 - M Figure 3. 21 Historical mean seasonal temperature and relative changes of mean temperatures for August to December for the 2070s for RCP4.5 fo r four GCMs Changes in mean temperature ( o C) Mean Temperatur e ( o C) Changes in mean temperature ( o C) Changes in mean temperature ( o C) Changes in mean temperature ( o C) 123 Historical average temperature CSIRO - Mk3.6 MIROC5 MRI - CGCM3 NorESM1 - M Figure 3. 22 Historical mean seasonal temperature and relative changes of mean temperatures for August to December for the 2070 s for RCP8.5 for four GCMs Changes in mean temperature ( o C) Mean Temperatur e ( o C) Changes in mean temperature ( o C) Changes in mean temperature ( o C) Changes in mean temperature ( o C) 124 3.3.2.3 Future crop yields 3.3.2.3.1 Historical sweet potato yield The magnitudes of highest yield have a range of 83 - 100 t/ha for both seasons in most parts of Uganda and north - western parts of Tanzania for NASPOT 1 (Figur es 3.8 a - e), NASPOT 10 (Figures 3.8 f - j), and NASPOT 11 (Figures3.8 p - t) and 58 - 62 t/ha for SPK004 (Figures 3.8 k - o). The least projected combined crop yields were recorded in the region coinciding with the eastern arm of the rift - valley for all cultivars. Generally, the crop yields increase from the southeastern parts of East Africa to the north - western regions. NASPOT 1 combined root yield for both seasons is projected to increase by more than 4t/ha in western and southern Kenya, for MRICGC3 - M RCP 4.5 (Fi gure 3.9a). Higher NASPOT 1 yields of more than 5 t/ha were projected for most parts of East Africa by MRICGC3 - M RCP 8.5 (Fig 3.9b). NorESM - 1 showed a smaller increase in yield of 1 t/ha for most parts of East Africa but southern and southwestern Tanzania ( Figure 3.9c). There was, however, higher increase in yield for NASPOT 1 in 2035 for NorESM - 1 RCP8.5 than NorESM - 1 RCP 4.5 although it is still less than that recorded by MRICGC3 - M RCP8.5 in 2035. In the 2050s and 2070s, there are isolated regions which will have higher increases in yield of magnitude more than 10t/ha mainly in Kenya and northern Tanzania. This is the case were both GCMs MRICGC3 - M (Figures 3.9 f, j) and NorESM - 1 (Figures 3.9 h, i) are in agreement. The SPK004 combined (both seasons added tog ether) yield projections show that in the 2030s , the root yield will have minor increases in yield of about 1 t/ha with a few spots registering higher yields (Figures 3.10 a - d) for both MRICGC3 - M and NorESM - 1. The yield in the 2050s is projected to increas e with magnitudes of not more than 3 t/ha in parts of central and 125 southern - western Tanzania and in areas around Mt. Ki l imanjaro (Figures 3.10 e - h). In the 2070s , the highest combined yield increases in SPK004 were shown by MRICGC3 - M RCP 4.5 (Figure 3.10 i) . There is an agreement in regions with a projected increase in yield of SPK004 for MRICGC3 - M 4.5 (Figure 3.10 i), MRICGC3 - M RCP 8.5 (Figure 3.10 j) and NorESM1 - M RCP 4.5 (Figure 3.10 k). 126 spk_yield_AD na11_yield_AD na1_yield_AD na10_yield_AD Figure 3. 23 Distribution of average sweet potato yield for the August to December season for the period 1980 - 2009 Sweet potato yield (kg/ha) Sweet potato yield (kg/ha) Sweet potato yield (kg/ha) Sweet potato yield (kg/ha) 127 spk_yield_FJ na11_yield_FJ na1_yield_FJ na10_yield_FJ Figure 3. 24 Distribution of average sweet potato yield for the February to June season for the period 1980 - 2009 Sweet potato yield (kg/ha) Sweet potato yield (kg/ha ) Sweet potato yield (kg/ha) Sweet potato yield (kg/ha) 128 3.3.2.3.2 February - June (FMAMJ) yield projections In the 2030s , most areas in East Africa were projected to have increases of 5 t/ha shown by RCP 8.5 for both models in some areas near Lake Turkana in Kenya (Figures 3.25, 3.26, 3.27 and 3.28). In the 2050s , most areas are projected to have increased NASPOT 1 yields with MRICGC3 - M RCP 8.5 and NorESM1 - M RCP 8.5 showing ev en higher yields. In the 2070s , si milar patterns of changes in yield to those of 2055 were projected to be seen. M ost parts of East Africa were projected to receive an increase in sweet potato produ ction in the range of 1 - 3 t/ha GCM (Figures). There were small projected increases in SPK00 4 yield in the 2050s and 2070s (Figures 3.25, 3.26, 3.27 and 3.28 ). Overall, the SPK004 yield will increase by 1 - 3 t/ha across the East African region. 129 Average yield for 1980 - 2009 2030s 2050s 2070s Figure 3. 25 H istorical average yield and relative changes in yield of SPK 004 sweet potato cultivar for the February to June season in the 2030s, 2050s , and 2070s for RCP 4.5 for CSIRO - Mk3.6 Sweet potato yield (kg/ha) Relative changes in sweet potato yield (kg/ha) Relative changes in sweet potato yield (kg/ha) Relative changes in sweet potato yield (kg/ha) 130 spk_yield_Hist_FJ S203s 2050s 2070s Figure 3. 26 Historical average yield and relative changes in yield of SPK 004 sweet potato cultivar for the February to June season in the 2030s, 2050s , and 2070s for RCP 8.5 for CSIRO - Mk3.6 Sweet potato yield (kg/ha) Relative changes in sweet potato y ield (kg/ha) Relative changes in sweet potato yield (kg/ha) Relative changes in sweet potato yield (kg/ha) 131 na1_yield_Hist_FJ 2030s 2050s 2070s Figure 3. 27 Historical average yield and relative changes in yield of NASPOT 1 sweet potato cultivar for the February to June season in the 2030s, 2050s , and 2070s for RCP 4.5 for CSIRO - Mk3.6 Sweet potato yield (kg/ha) Relative changes in sweet potato yield (kg/ha) Relative changes in sweet potato yield (kg/ha) Relative changes in sweet potato yield (kg/ha) 132 na1_yield_Hist_FJ 2030s 205 0s 2070s Figure 3. 28 Figure 6 Historical average yield and relative changes in yield of NASPOT 1 sweet potato cultivar for the February to June season in the 2030s, 2050s , and 2070s for RCP 8.5 for CSIRO - Mk3.6 Sweet potato yield (kg/ha) Relative changes in sweet potato yield (kg/ha) Relative changes in sweet potato yield (kg/ha) Relative changes in sweet potato yield (kg/ha) 133 3.3.2.3 .3 August - December (ASOND) yield projections The SPK004 root yield for 1980 2009 showed the lowest yield recorded for the short rains of magnitudes below 38 t/ha and largely below 29t/ha for most parts of the historical climate and control period (Figure s 3.29). In the 2030s , the whole of the East African region is projected to record increases in sweet potato of at least 1 t/ha with a few areas near the Kenya Tanzania border recording 2 - 4 t/ha (Figures 3.29 and 3.30). In the 2050s and 2070s, most areas across East Africa will register an increase in SPK004 of at least 1 t/ha. Areas along the Kenya Tanzania border are projected to receive a higher projection of 4t/ha or more (Figures 3.29 and 3.30). The NASPOT 1 and SPK004 root yield in 1980 - 2009 sho wed increasing yields from the south - east to north - west direction with largest increases of yields of 44 - 47 t/ha in some parts of Uganda (Figure 3.29 ). In the 2030s , the NASPOT 1 yields are projected to increase across central parts of East Africa with hig her increases shown by MRICGC3 - M RCP4 .5 and 8.5 (Figures 3.31 and 3.32 ). The NorESM1 - M shows very little increase in yield . In the 2050s , NASPOT 1 yields are projected to increase largely in most parts of Kenya and north - eastern Tanzania for both scenarios of MRICGC3 - M (Figures 3. 31 and 3.32). In the 2070s , the NASPOT 1 yields are projected to increase highest across most of the East African region with the largest increase of more than 10 t/ha root yields in south0 - western Kenya and part of northern Tanzan ia (Figure 3. 32 ). The southern part of Tanzania and northeastern parts of Kenya will experience a reduction in yields to the magnitude of 7 - 15 t/ha. 134 Hist orical average yield for 1980 - 2009 2030s 2050s 2070s Figure 3. 2 9 Historical average yield and relative changes in yield of SPK 004 sw eet potato cultivar for the August - December season in the 2030s, 2050s , and 2070s for RCP 4.5 for CSIRO - Mk3.6 Sweet potato yie ld (kg/ha) Relative changes in sweet potato yield (kg/ha) Relative changes in sweet potato yield (kg/ha) Relative changes in sweet potato yield (kg/ha) 135 Average yield for 1980 - 2009 2030s 2050s 207 0s Figure 3. 30 Historical average yield and relative changes in yield of SPK 004 sweet potato cultivar for the August - December season in the 2030s, 2050s , and 2070s for RCP 8.5 for CSIRO - Mk3.6 Sweet potato yield (kg/ha) Relative changes in sweet potato yield (kg/ha) Rel ative changes in sweet potato yield (kg/ha) Relative changes in sweet potato yield (kg/ha) 136 na1_yield_Hist_AD na1_y ield_D35_CO45ad na1_yield_D55_CO45_ad na1_yield_D75_CO45_ad Figure 3. 31 Historical average yield and relative changes in yield of NASPOT 1 sweet potato cultivar for the August - December season i n the 2030s, 2050s , and 2070s under RCP 4.5 for CSIRO - Mk3.6 Sweet potato yield (kg/ha) Relative changes in sweet potato yield (kg/ha) Relative changes in sweet potato yield (kg/ha) Relative changes in sweet potato yield (kg/ha) 137 Average yield over 1980 - 2009 2030s 2050s 2070s Figure 3. 32 Historical average yield and relative changes in yield of NASPOT 1 sweet potato cultivar for the August - December s eason in the 2030s, 2050s , and 2070s for RCP 8.5 for CSIRO - Mk3.6 Sweet potato yield (kg/ha) Relative changes in sweet potato yield (kg/ha) Relative changes in sweet potato yield (kg/ha) Relative changes in sweet potato yield (kg/ha) 138 3.4. Discussion presented in the previous section were compared with past work and highli gh ted th e contribution that the study was set to achieve. The nature and accuracy of the results largely depended on the quality of data used and the precision of both the General Circulation Models and the crop model employed in the study. These challenges were a lso discussed in the preceding paragraph. And finally, another perspective on the robustness of the agricultural production problem was put in a broader context by highlighting the role of socio - economic factors in impacting production, much as this study did not consider such factors. The reduction in precipitation during the long rains season (Feb - June) and its increase during the short rains season (August - December) as reported from the trend analysis of the last 30 years is in agreement with findings by the Famine Early Warning Systems Network (FEWSNet, 2010; 2012) who reported a reduction in the length of the long rains season and increase in length of the short rains season. For temperatures, on the other hand, the present study found a n increase in th e mean temperatures in the range 0 0 C - 1.5 0 C which averagely gets to about 0.7 0 C in both rain seasons. This increase is almost similar to the 0.8 0 C increase that was reported in FEWSNet, 2010:2012) which is a sign of consistenc y of the climate data sources employed in the present study. The projected trends were consistent with the observed trends. The GCMs captured well the seasonal variation in climate. The spatial and temporal variation of seasonal yields trends showed that the increase in yields coincid ed with areas which showed increases in precipitation and temperatures which confirm the importance of availability and requirement of moisture and high temperatures for the optimal growth and productivity of sweet potatoes . However, the timing of these c hanges in 139 precipitation and temperatures i s critical to the yields obtained at end the end of the season as (Loretan et al., 1994; Togari, 1950; Villordon et al., 2012) emphasized that the early - season (first 20 da ys) growing environment directly and significantly impacts on storage root initiation and thus final yield. Whereas increases in temperatures are largely good but there can be some impacts because temperature stress can limit on crop growth and developmen t (Wrigley, 1994) thereby causing irreversible damages to the plant processes and thus affecting final yield. Sensitivity analysis showed that SPOTCOMS was sensitive to increase in precipitation and temperature for all the four sweet potato cultivars, NASPOT 1, NASPOT 10 0, NASPOT 11 and SPK004. SPOTCOMS did not clearly show a threshold upon which sweet potato growth would be inhibited. Most of the graphs for the three locations in Uganda, Kenya , and Tanzania showed an increasing trend of crop yields with increasing precipitation and temperature which will require further refinement of SPOTCOMs to include thresholds. The validity of our analysis was however confirmed wh en we clearly plotted graphs of daily maximum and minimum temperatures for a scenario when the temperatures were increased by 5 0 C for the three locations (Figure 3.15) to have an idea on the highest temperatures reached. The graphs showed that all the thre e locations, the temperatures never exceed 35 0 C. Studies have not yet been conducted to fully examine the effect of temperatures on sweet potato growth but (Spence & Humphries, 1972) , obtained optimum storage root formation and development was obtained when the soil temperature was 25 0 C, whereas soil temperatures of 15 0 C and 35 0 C were inhibitory to storage root formation when th ey were working with rooted single leaves of sweet potatoes . Therefore, at about 35 0 C, sweet potato growth is most likely to still be taking place. F uture climate projections from Global Climate Models showed mixed results for precipitation and more dist inct results for temperatures. For example, a large number of regions in 140 East Africa showed increases in annual precipitation of more than 50mm by MRICGC 3 - M, the wettest GCM and a similar magnitude of the decrease in precipitation by NorESM - 1 which is the driest GCM. These results were projected to continue in 2030s , 2050s , and 2070s. On the other hand, temperatures in the region were projected to rise by 0.8 0 C, 1.2 0 C and more than 3 0 C in 2030s , 2050s, 2070s respective which are in agreement with previous work by (Cook & Vizy, 2013; FEWSNET, 2010, 2012; IPCC, 2013) . The projections further showed more increases in the short rains then the long rains for all the three future time slices, findings which are in agreeme nt with results from previous studies which reported shortening of long rains growing season and lengthening of short rains season (FEWSNET, 2010, 2012; IPCC, 2013) . The projected increase in sweet potato yield in the region coincide d with areas that experience increases in precipitati on and temperature. Models with larger radiative forcing of RCP8.5 showed an overall higher increase in precipitation, temperatures and therefore higher increases in sweet potato yield. All the four cultivars ( NASPOT 1, NASPOT 10 0, SPK004 and NASPOT 11) showed similar spatial distribution of yields but SPK004 had lower yields for both historical and projected future periods. This pattern of magnitudes of yields wa s consistent with the observations made in chapter two where SPK004 was the lower yielding cultivar of the four. Overall, sweet potato root yield increase s of 7 t/ha, 10 t/ha and more than 20 t/ha were projected for 2030s , 2050s , and 2070s. This study was the first of its kind in the region and therefore there were no previous studies to use compare with the results on projected yields presented here. But using the I rish potato modeling work, a study involving two GCMs for rainfed conditions in the United States; the Canadian Centre Climate Model Scenario (CCGS) suggested an increase in potato yield, and the Hadley Centre Model Scenario (HGCS) suggested a d ecrease in yield (Tubiello et al. (2002). 141 Another simulation study with a potato model suggested a reduction of the growing season with climate change (Holden and Brereton, 2006; Stockle et al., 2010). U nderstanding the magnitu de of the impacts of clima te change and variability on sweet potato production is complicated by the interaction of numerous biophysical and socioeconomic factors. This study has attempted to focus on the weather/climate/seasonal factors including temperature, precipitation and s ea sonality , and crop factors, namely crop type and crop yield. There are many other factors which need to be considered in order to address and even understand the entire problem facing crop production and agriculture. Other factors include: extreme weather events; e levation/altitude; crop price; agricultural factors such as farm size, irrigation, agricultural expenditure, labor, herb icide/insecticide and equipment ; household demographic and socioeconomic factors such as household size, composition, food con sumption pattern, income, water and sanitation, aggregate income, expenditure, livelihood pattern, off - farm income, livestock rearing, and maternal education, among o thers); and individual factors such as sex, age, morbidity, and diet. Previous studies ha ve combined climate, crop and economic models to examine the impact of climate change on agricultural production and food security, but results have varied widely due to differences in models, scenarios and input data (Rosenzweig and Parry 1994, Nelson et al 2010). Interdisciplinary studies based on primary data at a household level are urgently required to guide effective adaptation, particularly for rural subsistence farmers. There is need to use data from all these factors in order to develop robust stat istical methods to establish and validate causal links, quantify impacts, and make reliable predictions that can guide evidence - based health interventions in the future. 142 3.5 Conclusion This chapter has presented historical trends of rainfall, temperature and sweet potato yields for the period 1980 - 2009. Climate variability was found to be characterized by a reduction in intensity of the long rains, increase in rainfall intensity of the short rains and increase in temperatures by more than 1 degree in the last three decades. The corresponding sweet potato yields were also reported to increase with increasing precipitation and temperature. The sensitivity analysis showed that the model we have developed for East Africa can be used to test the impact of clim ate and other parameters on sweet potato production. The study has quantified the impact of climate change on sweet potato production for the East African region. As one of the main staples of the region, sweet potat o production is projected to largely in crease in future time frames in the century. This implies that with careful attention given to the good management of farming of the sweet potatoes , the high production would significantly contribute to the projected increase in population. The people of E ast Africa will also have to optimize the use of sweet potato vegetative and root production in order to be able to consume most nutrients from both the vegetative parts and the roots. Just like most studies, the modeling tool employed in this study had some limitations and will need further refinement. For example, there is need to include temperature thresholds within the model, the soil water balance model employed in SPOTCOMs needs to be further developed and the readily available water needs further development. The soil module employed in the model will also need be improved to include the different soil profile layers similar to those used in the Decision Support System for Agronomical Tool (DSSAT) crop models (J. W. Jones et al., 2003) . For optimal use of the results of this study, all ke y stakeholders have a role to play from governments to research institutions, to private sector organizations and the donor community. 143 Some of the major challenges experienced during the execution of the research were poor data quality mainly the lack of g ood observation records for both climate and sweet potatoes , and generally limited studies of the relationship of climate on sweet potatoes . These constraints are some of the major stumbling blocks to further research and therefore, this study recommends m ore future research on generally all root tubers. Future follow - up work that is proposed will include the testing, calibration , and evaluation of SPOTCOMS across several locations across East Africa followed by a further assessment of the impact of enviro nment al factors on sweet potato production. Secondly , this study needs to be extended to other sweet potato cultivars compared to the four used in this study. Also, there is need to use several other gridded climate datasets in running SPOTCOMS in order to be confident of the variation in the results. 144 CHAPTER 4. SUMMARY AND CONCLUSI ON This research assessed the impact of climate variability and change on sweet potato production in East Africa by following four major steps. The first step was to d evelo p a modeling framework for use in a deterministic sweet potato crop model, SPOTCOMS, for East Africa; the second step was to analyz e trends of historical climate and sweet potato root yields for the historical period 1980 - 2009; and the third step was to st udy develop ed stochastically generated current , near future 2030s, mid - future 2050s and distant future 20 7 0s climate data scenarios for East Africa using two representative concentration pathways 4.5 and 8.5 for four Global Climate Models, CSIRO, MIROC5, M RICGC3 - M and NorESM - 1; and finally, the fourth step was to estimat e the impact of projected future climate change on sweet potato production using SPOTCOMS model. Some of the major achievements of this study wa s the determination of the sweet potato crop coefficients required for running SPOTCOMS model. The field experiments conducted at Namulonge across the two seasons in 2012 and 2013 provided the required dataset on the growth of sweet potato that made it possible to determine the coefficients. Whereas this was done at only one location under two seasons, it wa s recommended that follow - up studies be conducted across the wh ole East Africa region for more sweet potato cultivars including the four cultivars used in this study. This study was conducted for four representative high yielding cultivars in East Africa. Two of which were non - orange cultivars ( NASPOT 1 and NASPOT 11) and the other two were orange cultivars ( NASPOT 10 0 and SPK004 - Kakamega). The other contribution of this study 145 was the modificatio n of the previous SPOTCOMS model to be able to input weather data of any size for any number of years and some other minor additions of variable outputs such as the potential evapotranspiration (ET), actual evapotranspiration (Etc) and the root available w ater ( rwtr ). These modifications imply that SPOTCOMS can now be run for a single site for multiple seasons. And finally , this study has provided the first quantification of the impact of climate change on sweet potato production in East Africa upon which f uture studies can build. F uture follow - up work that is proposed will include the testing, calibration , and evaluation of SPOTCOMS across several locations across East Africa followed by a further assessment of the impact of environment al factors on sweet potato production. Secondly , this study needs to be extended to other sweet potato cultivars compared to the four used in this study. Also, there is need to use several other gridded climate datasets in running SPOTCOMS in order to be confident of the vari ation in the results. For optimal use of the results of this study, all key stakeholders have a role to play from governments to research institutions, to private sector organizations and the donor community. Some of the major challenges experienced durin g the execution of the research were poor data quality mainly the lack of good observation records for both climate and sweet potatoes , and generally limited studies of the relationship of climate on sweet potatoes . These constraints are some of the major stumbling blocks to further research and therefore, this study recommends more future research on generally all root tubers. Like most models, SPOTCOMS has some limitations that will require being addressed in future studies. For example, SPOTCOMS needs t o be set such that it has a threshold beyond which if the temperature is exceeded, sweet potato growth would be inhibited. This has not yet been set in the model. Second, although the model is sensitiv e to both temperatures and soil moisture, 146 SPOTCOMS did not appear to give a corresponding sensitivity on the actual evapotranspiration. In other words, in a case where the ETc would be elevated, the model did not show a corresponding variation in the root available water. This is one major area that requires r evisiting in the model. Third, SPOTCOMS does not yet consider CO 2 intake which is known to equally affect sweet potato growth just like temperature and soil moisture. Fourth, the model does not account for the effects of weeds, pest , and diseases and ther efore, the model normally tends to overestimate yield because of this weakness . Finally, the model currently uses basic soil routines and does not account for the variation of soil nutrients in the various soil profiles as has been significantly developed in other crop models such as the DSSAT crop models (J. W. Jones et al., 2003) . This too will have to be worked on in the future. One major shortcoming for SPOTCOMS, which is not uncommon in an other process - based crop model , is that the model does not consider the effect of pests and diseases. This , therefore, means that the assumption is made that pest and disease management was carefully implemented i n the fields, although this is not normally the case in reality. A large proportion of the cropping and rangeland area of sub - Saharan Africa is proj ected to see a decrease in growing season length, and most of Africa in the southern latitudes may see losses of at least 20 percent (Thornton et al., 2011). At the same time, the probability of season failure is projected to increase for all of sub - Sahara n Africa, except for central Africa; in southern Africa, nearly all rain - fed agriculture below latitude 15°S is likely to fail one year out of two (Thornton et al., 2011). In terms of timing of growing season onset, Crespo et al. (2011) demonstrate that it may be possible to adapt to projected climate shifts to at least the 2050s in maize production systems in parts of southern Africa by changing planting dates. 147 The findings of this study provide hope as warmer climates will generally accelerate the growth and development of sweet potato, but overly cool or hot weather will also affect sweet potato productivity. Future studies on sweet potato should look at sweet potato yield quality which is greatly related to climate variability and extreme. This should n ot be an exception as some studies have already been conducted on other crops , for example, Porter & Semenov (2005) showed that protein content of wheat grain responded to changes in the mean and variability of temperature and rainfall. Future research s hould also focus on mixed crop - livestock systems which are prevalent in East Africa. Sweet potato is a crop which fits well in this farming system and therefore understanding how climate change and changing climate variability in the future may affect the relationship between crops and livestock is very important. The synergy between sweet potato which is a drought tolerant crop and livestock may be a good solution towards addressing impacts of projected future of climate change and variability on agricultu re. The effects of future changes in climate variability on pests, weeds , and diseases are not well understood (Gornall et al., 2010). But changes in climate variability and in the frequency of extreme events may have substantial impacts on the prevalence and distribution of pests, weeds, and sweet potato diseases. This is an area which needs further research and development especially given that the current crop - models are not capable of modeling the effect of pests, weeds , and diseases. Food security in t he East African region could be enhanced by increasing farm - based storage facilities; improving the transportation system, especially feeder roads that link food production areas and major markets; providing farmers with early warning systems; extending 148 cr edit to farmers; and the use of supplementary irrigation. These socio - economic factors have been shown to be impacted by extreme climate events in rural communities in Ghana (Codjoe and Owusu, 2011). The o range - fleshed sweet potato which is rich in vitamin A will be a valuable crop in the future. This is because the overall availability of food shows some correlation with climate variability. Lloyd et al. (2011) showed that the impact of climate change and increased climate variability on food production wi ll have a negative impact on the prevalence of undernutrition, increasing severe stunting by 55% in East and southern Africa by the 2050s. Therefore, investing in research in crops such as sweet potato which is likely to do well in a future with projected changing climate is of paramount importance. Since sweet potato is grown mostly by poor smallholder farmers, governments can play a crucial role in smallholder agriculture. Governments in East African countries can invest in activities such as storage, tr ace, processing and retailing; implementing and scaling up options that help producers to be more resilient to climate volatility, such as the use of smallholder crop insurance schemes; and establishing safety net programmes for the most vulnerable househo lds, such as has been implemented successfully in Ethiopia (Lipper, 2011). By so doing, the governments could have helped smallholder farmers to be more resilient to the impacts of climate change on crop production. There is a great need to improve the mon itoring of local conditions, not only to provide data and information for improving our understanding and our models but also to guide effective adaptation and to provide information for yield early warning systems and locally appropriate indices for weath er - based crop and livestock insurance schemes. As pointed out in this study, there a number of satellite and land - based weather data sets but they are not a replacement for land - 149 based weather measurement. Governments need to significant ly increase their in vestments in the area of data monitoring so that the research scientific community can conduct research with more precision. 150 APPENDICES 151 APPENDIX A: Sensitivity analysis Table 3. 6 Summary data for Sensitivity analysis Dagorett i Corner (DC), Kenya Combinati on p (%) p (C) DC - na1 - AD - yield (t/ha ) DC - na10 - AD - yield (t/ha ) DC - na11 - AD - yield (t/ha ) DC - spk - AD - yield (t/ha ) DC - na1 - FJ - yield (t/ha ) DC - na10 - FJ - yield (t/ha ) DC - na11 - FJ - yield (t/ha ) DC - spk - FJ - yield (t/ha ) _P - 50T - 5 - 50 - 5 7.4 7.7 7.2 9.2 9.6 9.9 9.3 11.7 _P - 40T - 5 - 40 - 5 7.5 7.8 7.3 9.3 9.9 10.3 9.6 12.3 _P - 30T - 5 - 30 - 5 8.1 8.6 6.8 10.5 10.6 11.4 10.2 12.0 _P - 20T - 5 - 20 - 5 7.9 8.2 7.7 9.7 11.6 12.5 11.2 15.3 _P - 10T - 5 - 10 - 5 8.1 8.4 7.9 10.1 13.2 14.3 12.7 17.5 P_P+0T - 5 0 - 5 8.4 8.7 8.2 10.5 15.4 16.7 14.7 20.4 _P+10T - 5 10 - 5 8.7 9.1 8.5 10.9 18.1 19.5 17.3 22.8 _P+20T - 5 20 - 5 9.2 9.6 8.9 11.3 20.7 22.1 20.0 24.3 _P+30T - 5 30 - 5 9.7 10.1 9.4 11.9 23.0 24.3 22.5 25.8 _P+40T - 5 40 - 5 10.2 10.7 10 .0 12.4 25.1 26.1 24.9 27.4 _P+50T - 5 50 - 5 10.7 11.2 10.6 13.0 27.1 28.1 27.4 28.9 _P - 50T - 4 - 50 - 4 8.6 9.1 8.4 10.8 10.9 11.3 10.7 13.1 _P - 40T - 4 - 40 - 4 8.8 9.3 8.5 11.0 11.4 11.9 11.2 14.1 _P - 30T - 4 - 30 - 4 9.1 9.5 8.8 11.3 12.5 13.2 12.2 15.6 _P - 20T - 4 - 20 - 4 9.4 9.9 9.1 11.7 14.1 14.8 13.7 17.8 _P - 10T - 4 - 10 - 4 9.8 10.3 9.5 12.2 16.2 17.2 15.8 20.2 P_P+0T - 4 0 - 4 10.3 10.8 10.0 12.7 18.7 19.7 18.2 22.4 _P+10T - 4 10 - 4 10.8 11.4 10.6 13.2 21.2 22.3 20.9 24.7 _P+20T - 4 20 - 4 11.4 11.9 11.2 13.7 24.1 25.1 24.2 26.2 _P+30T - 4 30 - 4 11.9 12.4 11.8 14.2 26.4 27.4 27.0 27.7 _P+40T - 4 40 - 4 12.5 13.0 12.3 14.8 28.8 29.1 29.6 28.9 _P+50T - 4 50 - 4 13.1 13.7 13.0 15.5 30.6 30.9 31.9 29.9 _P - 50T - 3 - 50 - 3 9.9 10.4 9.6 12.0 12.4 12.9 12.2 14.8 _P - 40T - 3 - 40 - 3 10.2 1 0.6 9.8 12.3 13.7 14.2 13.4 16.3 _P - 30T - 3 - 30 - 3 10.5 11.0 10.2 12.7 15.5 16.2 15.5 18.9 _P - 20T - 3 - 20 - 3 11.0 11.5 10.6 13.3 17.8 18.7 17.8 21.1 _P - 10T - 3 - 10 - 3 11.6 12.1 11.2 13.9 20.1 20.9 20.1 23.0 P_P+0T - 3 0 - 3 12.3 12.8 11.9 14.5 22.6 23.5 22.9 25 .1 _P+10T - 3 10 - 3 13.0 13.6 12.8 15.2 25.1 25.9 25.8 26.8 _P+20T - 3 20 - 3 13.8 14.4 13.6 15.9 27.4 28.3 28.5 28.2 _P+30T - 3 30 - 3 14.6 15.2 14.5 16.6 29.7 30.1 31.4 29.3 _P+40T - 3 40 - 3 15.5 16.0 15.4 17.3 31.6 31.8 34.0 30.2 _P+50T - 3 50 - 3 16.4 16.9 16. 4 18.0 33.1 33.0 35.9 31.2 152 Combinati on p (%) p (C) DC - na1 - AD - yield (t/ha ) DC - na10 - AD - yield (t/ha ) DC - na11 - AD - yield (t/ha ) DC - spk - AD - yield (t/ha ) DC - na1 - FJ - yield (t/ha ) DC - na10 - FJ - yield (t/ha ) DC - na11 - FJ - yield (t/ha ) DC - spk - FJ - yield (t/ha ) _P - 50T - 2 - 50 - 2 11.1 11.7 10.8 12.8 14.1 14.7 13.8 16.2 _P - 40T - 2 - 40 - 2 11.4 12.0 11.1 13.2 16.1 16.8 15.7 18.5 _P - 30T - 2 - 30 - 2 11.9 12.5 11.6 13.7 18.4 19.4 18.1 20.8 _P - 20T - 2 - 20 - 2 12.5 13.1 12.2 14.4 20.9 22.0 20.6 22.9 _P - 10T - 2 - 10 - 2 13.3 13.9 13.0 15.3 23.5 24.4 23.4 24.9 P_P+0T - 2 0 - 2 14 .3 14.9 14.0 16.2 26.2 27.0 26.5 26.6 _P+10T - 2 10 - 2 15.3 15.9 15.2 17.0 28.9 29.2 29.6 28.1 _P+20T - 2 20 - 2 16.4 17.0 16.4 17.8 31.2 31.2 32.5 29.2 _P+30T - 2 30 - 2 17.4 17.9 17.5 18.5 33.1 32.8 35.2 30.1 _P+40T - 2 40 - 2 18.2 18.7 18.7 19.2 34.6 34.2 37.6 31.1 _P+50T - 2 50 - 2 19.1 19.5 19.7 19.9 36.2 35.5 39.4 32.2 _P - 50T - 1 - 50 - 1 12.2 12.6 11.9 13.5 15.4 16.2 15.1 17.2 _P - 40T - 1 - 40 - 1 12.5 13.0 12.3 14.0 17.8 18.6 17.4 19.6 _P - 30T - 1 - 30 - 1 13.1 13.6 12.8 14.5 20.6 21.4 20.2 22.0 _P - 20T - 1 - 20 - 1 13.8 1 4.2 13.5 15.3 23.4 23.9 23.2 24.1 _P - 10T - 1 - 10 - 1 14.7 15.2 14.4 16.2 26.0 26.4 26.4 26.2 P_P+0T - 1 0 - 1 15.8 16.3 15.7 17.1 28.4 28.8 29.5 27.6 _P+10T - 1 10 - 1 17.0 17.4 17.0 18.1 30.8 31.0 32.5 28.9 _P+20T - 1 20 - 1 18.1 18.5 18.3 18.9 33.1 32.7 35.5 30. 1 _P+30T - 1 30 - 1 19.2 19.4 19.6 19.7 34.7 34.2 38.1 30.9 _P+40T - 1 40 - 1 20.2 20.4 20.9 20.5 36.3 35.7 40.0 31.7 _P+50T - 1 50 - 1 21.2 21.5 22.1 21.4 37.7 36.7 41.9 32.5 _P - 50T+0 - 50 0 13.0 13.5 12.8 14.1 16.7 17.4 16.3 18.2 _P - 40T+0 - 40 0 13.4 13.9 13.2 14.6 19.4 20.1 19.0 20.9 _P - 30T+0 - 30 0 14.0 14.5 13.7 15.2 22.5 23.3 22.2 23.4 _P - 20T+0 - 20 0 14.8 15.3 14.5 16.0 25.6 26.1 25.7 25.3 _P - 10T+0 - 10 0 15.8 16.4 15.6 17.1 28.1 28.6 29.0 27.2 P_P+0T+0 0 0 17.1 17.6 17.0 18.1 30.4 30.9 32.0 28.7 _P+10T+ 0 10 0 18.4 18.9 18.4 19.0 32.9 32.9 34.9 30.0 _P+20T+0 20 0 19.8 20.1 19.9 19.8 34.6 34.5 37.9 30.9 _P+30T+0 30 0 20.7 21.0 21.4 20.7 36.2 35.8 39.8 31.7 _P+40T+0 40 0 21.7 22.0 22.6 21.6 37.6 36.9 41.6 32.4 _P+50T+0 50 0 22.9 23.1 23.7 22.7 38.7 37.9 43.5 33.3 _P - 50T+1 - 50 1 13.8 14.2 13.5 14.6 17.7 18.3 17.2 18.9 _P - 40T+1 - 40 1 14.2 14.6 14.0 15.1 20.6 21.2 20.1 21.6 153 Combinati on p (%) p (C) DC - na1 - AD - yield (t/ha ) DC - na10 - AD - yield (t/ha ) DC - na11 - AD - yield (t/ha ) DC - spk - AD - yield (t/ha ) DC - na1 - FJ - yield (t/ha ) DC - na10 - FJ - yield (t/ha ) DC - na11 - FJ - yield (t/ha ) DC - spk - FJ - yield (t/ha ) _P - 30T+1 - 30 1 14.9 15.2 14.6 15.8 24.1 24.6 23.6 24.0 _P - 20T+1 - 20 1 15.7 16.1 15.5 16.6 27.0 27.2 27.3 25.9 _P - 10T+1 - 10 1 16.9 17.2 16.6 17.7 29.3 29.5 30.7 27.5 P_P+0T+1 0 1 18.2 18.5 18.1 18.7 31.6 31.6 33.5 28.9 _P+10T+1 10 1 19.6 19.9 19.6 19.6 34.0 33.4 36.3 29.9 _P+20T+1 20 1 20.8 20.9 21.2 20.5 35.5 34.9 39.0 30.8 _P+30T+1 30 1 21.9 21.9 22.6 21.3 37.1 36.1 40.9 31.6 _P+40T+1 40 1 23.0 23.0 23.8 22.3 38.5 37.2 42.9 32.4 _P+50T+1 50 1 24.2 24.3 25.2 23.2 39.6 38.3 44.9 33.0 _P - 50T+2 - 50 2 14.4 14.8 14.1 15.1 18.8 19.2 18.2 19.7 _P - 4 0T+2 - 40 2 14.9 15.3 14.6 15.6 21.9 22.2 21.2 22.3 _P - 30T+2 - 30 2 15.6 16.0 15.3 16.3 25.5 25.5 25.0 24.6 _P - 20T+2 - 20 2 16.5 16.9 16.2 17.2 28.1 28.1 28.7 26.4 _P - 10T+2 - 10 2 17.8 18.0 17.4 18.2 30.6 30.7 32.0 27.9 P_P+0T+2 0 2 19.1 19.4 18.9 19.2 33. 1 32.7 34.8 29.4 _P+10T+2 10 2 20.5 20.6 20.5 20.0 35.1 34.5 37.7 30.5 _P+20T+2 20 2 21.7 21.6 22.0 20.9 36.9 35.9 40.5 31.4 _P+30T+2 30 2 22.8 22.7 23.4 21.8 38.2 37.1 42.4 32.1 _P+40T+2 40 2 24.0 23.8 24.7 22.8 39.5 38.2 44.5 32.7 _P+50T+2 50 2 25.2 25.1 26.2 23.8 40.6 39.0 46.2 33.2 _P - 50T+3 - 50 3 15.0 15.4 14.7 15.6 19.6 19.8 19.0 20.3 _P - 40T+3 - 40 3 15.5 15.9 15.2 16.0 22.7 22.8 22.2 22.8 _P - 30T+3 - 30 3 16.3 16.6 15.9 16.8 26.2 26.0 26.1 25.0 _P - 20T+3 - 20 3 17.3 17.5 16.8 17.7 28.8 28.5 29.8 2 7.0 _P - 10T+3 - 10 3 18.6 18.8 18.1 18.7 31.7 31.2 33.2 28.7 P_P+0T+3 0 3 19.9 20.0 19.6 19.6 34.4 33.6 36.2 30.1 _P+10T+3 10 3 21.3 21.3 21.3 20.4 36.4 35.3 39.4 31.0 _P+20T+3 20 3 22.5 22.3 22.8 21.2 37.9 36.5 41.8 31.8 _P+30T+3 30 3 23.6 23.3 24.2 22 .1 39.2 37.5 43.7 32.4 _P+40T+3 40 3 24.8 24.4 25.5 23.1 40.3 38.5 45.6 32.9 _P+50T+3 50 3 26.1 25.7 27.0 23.9 41.3 39.4 47.0 33.1 _P - 50T+4 - 50 4 15.5 15.9 15.2 16.0 20.5 20.4 19.7 20.9 _P - 40T+4 - 40 4 16.1 16.4 15.7 16.5 23.8 23.5 23.1 23.6 _P - 30T+4 - 30 4 16.9 17.1 16.4 17.2 27.5 27.0 27.2 25.6 _P - 20T+4 - 20 4 17.9 18.1 17.4 18.2 30.4 29.8 30.9 27.7 154 Combinati on p (%) p (C) DC - na1 - AD - yield (t/ha ) DC - na10 - AD - yield (t/ha ) DC - na11 - AD - yield (t/ha ) DC - spk - AD - yield (t/ha ) DC - na1 - FJ - yield (t/ha ) DC - na10 - FJ - yield (t/ha ) DC - na11 - FJ - yield (t/ha ) DC - spk - FJ - yield (t/ha ) _P - 10T+4 - 10 4 19.2 19.4 18.6 19.2 33.3 32.6 34.6 29.3 P_P+0T+4 0 4 20.7 20.7 20.2 20.0 36.0 35.0 37.9 30.4 _P+10T+4 10 4 22.0 21.8 21.8 20.8 38.1 36.7 41.0 31.1 _P+20T+4 20 4 23.1 22.9 23.4 21.7 39.5 37.9 43.4 31.8 _P+30T+4 30 4 24.2 24.0 24.7 22.6 40.8 39.0 45.3 32.3 _P+40T+4 40 4 25.5 25.1 26 .1 23.5 41.9 39.9 47.0 32.6 _P+50T+4 50 4 26.9 26.4 27.6 24.5 42.8 40.7 48.2 32.8 _P - 50T+5 - 50 5 16.0 16.3 15.7 16.4 21.2 21.0 20.3 21.4 _P - 40T+5 - 40 5 16.7 16.8 16.2 16.9 24.8 24.4 23.8 23.9 _P - 30T+5 - 30 5 17.4 17.6 16.9 17.6 28.4 27.7 28.0 26.1 _P - 2 0T+5 - 20 5 18.5 18.5 17.9 18.5 31.5 30.6 31.9 28.1 _P - 10T+5 - 10 5 19.9 19.8 19.2 19.5 34.5 33.3 35.6 29.9 P_P+0T+5 0 5 21.3 21.1 20.7 20.3 37.2 35.5 38.9 31.1 _P+10T+5 10 5 22.5 22.2 22.3 21.1 39.0 37.0 42.2 31.9 _P+20T+5 20 5 23.6 23.3 23.8 22.1 40.3 38.0 44.6 32.6 _P+30T+5 30 5 24.8 24.3 25.1 23.0 41.5 39.1 46.4 33.2 _P+40T+5 40 5 26.1 25.5 26.5 23.9 42.6 40.0 47.9 33.5 _P+50T+5 50 5 27.6 26.8 28.1 24.7 43.6 40.8 49.1 33.6 155 Table 3. 7 Summary data for Sensitivity analy sis Mbeya (MB), Tanzania Combinatio n preci p (%) Tem p (C) MB - na1 - AD - yield (t/ha ) MB - na10 - AD - yield (t/ha) MB - na11 - AD - yield (t/ha) MB - spk - AD - yield (t/ha ) MB - na1 - FJ - yield (t/ha ) MB - na10 - FJ - yield (t/ha) MB - na11 - FJ - yield (t/ha) MB - spk - FJ - yield (t/ha ) _P - 50T - 5 - 50 - 5 7.8 8.2 7.5 9.8 9.6 10.4 7.5 12.7 _P - 40T - 5 - 40 - 5 9.2 9.7 8.8 11.3 11.3 12.2 8.8 15.1 _P - 30T - 5 - 30 - 5 10.7 11.2 10.2 12.6 12.9 13.8 10.2 16.8 _P - 20T - 5 - 20 - 5 11.8 12.3 11.5 13.5 14.3 15.2 11.5 18.0 _P - 10T - 5 - 10 - 5 12.9 13.3 12.5 14.3 15.6 16. 7 12.5 19.1 P_P+0T - 5 0 - 5 13.8 14.2 13.4 14.9 16.7 17.9 13.4 19.9 _P+10T - 5 10 - 5 14.5 14.9 14.1 15.5 17.8 19.1 14.1 20.7 _P+20T - 5 20 - 5 15.2 15.5 14.8 15.9 19.0 20.2 14.8 21.3 _P+30T - 5 30 - 5 15.8 16.1 15.3 16.3 20.1 21.3 15.3 22.0 _P+40T - 5 40 - 5 16.3 16.5 15.8 16.7 21.1 22.1 15.8 22.6 _P+50T - 5 50 - 5 16.8 16.9 16.2 17.0 22.0 22.7 16.2 23.0 _P - 50T - 4 - 50 - 4 8.3 8.7 8.0 10.4 11.1 12.1 8.0 15.0 _P - 40T - 4 - 40 - 4 9.9 10.4 9.5 12.1 13.1 14.4 9.5 17.8 _P - 30T - 4 - 30 - 4 11.5 12.0 11.0 13.4 14.9 16.0 11.0 19.6 _P - 20T - 4 - 20 - 4 12.7 13.2 12.3 14.4 16.3 17.5 12.3 20.9 _P - 10T - 4 - 10 - 4 13.8 14.3 13.5 15.2 17.6 19.0 13.5 21.8 P_P+0T - 4 0 - 4 14.8 15.2 14.4 15.8 18.8 20.2 14.4 22.4 _P+10T - 4 10 - 4 15.6 15.9 15.1 16.3 19.9 21.4 15.1 23.1 _P+20T - 4 20 - 4 16.3 16.5 15.8 1 6.7 21.3 22.6 15.8 23.8 _P+30T - 4 30 - 4 17.0 17.1 16.4 17.1 22.5 23.7 16.4 24.7 _P+40T - 4 40 - 4 17.5 17.5 16.8 17.4 23.6 24.6 16.8 25.4 _P+50T - 4 50 - 4 18.0 18.0 17.2 17.8 24.4 25.3 17.2 25.7 _P - 50T - 3 - 50 - 3 8.9 9.3 8.6 11.1 12.8 14.2 8.6 17.5 _P - 40T - 3 - 40 - 3 10.7 11.2 10.2 12.9 15.2 16.5 10.2 20.3 _P - 30T - 3 - 30 - 3 12.5 12.9 12.0 14.2 17.1 18.5 12.0 22.5 _P - 20T - 3 - 20 - 3 13.8 14.3 13.4 15.2 18.7 20.1 13.4 23.8 _P - 10T - 3 - 10 - 3 15.0 15.4 14.7 16.1 20.1 21.7 14.7 24.6 P_P+0T - 3 0 - 3 16.0 16.3 15.6 16.7 21.2 22.9 15.6 25.3 _P+10T - 3 10 - 3 16.9 17.0 16.5 17.3 22.5 24.0 16.5 26.2 _P+20T - 3 20 - 3 17.6 17.7 17.1 17.7 23.9 25.3 17.1 27.0 _P+30T - 3 30 - 3 18.2 18.2 17.7 18.0 25.4 26.7 17.7 28.1 _P+40T - 3 40 - 3 18.8 18.6 18.2 18.4 26.5 27.8 18.2 28.8 _P+50T - 3 50 - 3 19.3 19.1 18.6 18.6 27.7 28.8 18.6 29.2 _P - 50T - 2 - 50 - 2 9.6 10.1 9.3 11.9 14.8 16.3 9.3 19.9 _P - 40T - 2 - 40 - 2 11.7 12.2 11.2 13.8 17.2 18.7 11.2 22.8 _P - 30T - 2 - 30 - 2 13.6 14.0 13.3 15.2 19.4 20.8 13.3 25.0 156 Combinatio n preci p (%) Tem p (C) MB - na1 - AD - yield (t/ha ) MB - na10 - AD - yield (t/ha) MB - na11 - AD - yield (t/ha) MB - spk - AD - yield (t/ha ) MB - na1 - FJ - yield (t/ha ) MB - na10 - FJ - yield (t/ha) MB - na11 - FJ - yield (t/ha) MB - spk - FJ - yield (t/ha ) _P - 20T - 2 - 20 - 2 15.1 15.5 14.9 16.2 21.2 22 .7 14.9 26.5 _P - 10T - 2 - 10 - 2 16.3 16.6 16.3 17.1 22.7 24.4 16.3 27.8 P_P+0T - 2 0 - 2 17.3 17.6 17.3 17.8 24.0 25.8 17.3 28.7 _P+10T - 2 10 - 2 18.2 18.2 18.1 18.3 25.7 27.4 18.1 30.1 _P+20T - 2 20 - 2 18.9 18.9 18.8 18.5 27.3 29.1 18.8 30.8 _P+30T - 2 30 - 2 19. 6 19.5 19.3 18.8 29.3 31.1 19.3 31.4 _P+40T - 2 40 - 2 20.1 19.8 19.7 19.1 31.0 32.5 19.7 32.1 _P+50T - 2 50 - 2 20.7 20.3 20.1 19.4 32.7 33.9 20.1 32.2 _P - 50T - 1 - 50 - 1 10.5 10.9 10.1 12.6 16.8 18.4 10.1 22.5 _P - 40T - 1 - 40 - 1 12.7 13.2 12.4 14.6 19.3 21.0 12. 4 25.3 _P - 30T - 1 - 30 - 1 14.7 15.1 14.7 16.0 21.6 22.9 14.7 27.5 _P - 20T - 1 - 20 - 1 16.3 16.6 16.4 17.0 23.5 25.0 16.4 29.2 _P - 10T - 1 - 10 - 1 17.6 17.8 17.8 17.9 25.5 27.2 17.8 30.6 P_P+0T - 1 0 - 1 18.7 18.8 18.8 18.5 27.1 29.2 18.8 31.8 _P+10T - 1 10 - 1 19.5 19 .5 19.7 19.0 29.3 31.2 19.7 33.1 _P+20T - 1 20 - 1 20.2 20.1 20.3 19.2 31.4 33.2 20.3 33.5 _P+30T - 1 30 - 1 20.8 20.5 20.9 19.4 33.9 35.4 20.9 33.4 _P+40T - 1 40 - 1 21.3 20.9 21.2 19.7 35.8 36.7 21.2 34.1 _P+50T - 1 50 - 1 21.8 21.3 21.7 20.0 37.9 38.0 21.7 34.1 _P - 50T+0 - 50 0 11.3 11.7 11.1 13.2 18.9 20.6 11.1 24.8 _P - 40T+0 - 40 0 13.7 14.2 13.6 15.3 21.5 23.1 13.6 27.7 _P - 30T+0 - 30 0 15.9 16.2 16.0 16.7 23.7 25.1 16.0 29.9 _P - 20T+0 - 20 0 17.5 17.7 17.8 17.7 25.9 27.6 17.8 31.5 _P - 10T+0 - 10 0 18.8 18.8 19.3 18.4 28.4 30.3 19.3 33.2 P_P+0T+0 0 0 19.8 19.8 20.3 19.1 30.6 33.0 20.3 34.3 _P+10T+0 10 0 20.6 20.5 21.1 19.5 33.3 35.2 21.1 35.3 _P+20T+0 20 0 21.2 21.0 21.7 19.6 35.6 37.0 21.7 35.6 _P+30T+0 30 0 21.8 21.3 22.2 19.8 38.4 39.1 22.2 35.7 _P+40T+0 40 0 22.2 21.7 22.6 20.1 40.1 40.0 22.6 35.8 _P+50T+0 50 0 22.6 22.0 23.0 20.4 41.9 41.1 23.0 35.5 _P - 50T+1 - 50 1 12.6 19.6 20.5 13.9 21.8 33.8 20.5 35.4 _P - 40T+1 - 40 1 14.7 15.0 14.8 15.8 23.5 25.2 14.8 29.9 _P - 30T+1 - 30 1 16.8 17.0 17.4 17.1 26.0 27.5 17.4 31.9 _P - 20T+1 - 20 1 18.4 18.6 19.1 18.2 28.8 30.6 19.1 33.7 _P - 10T+1 - 10 1 19.6 19.6 20.5 18.9 31.8 33.8 20.5 35.4 157 Combination precip (%) Temp (C) MB - na1 - AD - yield (t/ha) MB - na10 - AD - yield (t/ha) MB - na11 - AD - yield (t/ha) MB - spk - AD - yield (t/ha) MB - na1 - FJ - yield (t/ha) MB - na10 - FJ - yield (t/ha) MB - na11 - FJ - yield (t/ha) MB - spk - FJ - yield (t/ha) P_P+0T+1 0 1 20.6 20.5 21.5 19.6 34.7 36.6 21.5 36.3 _P+10T+1 10 1 21.3 21.2 22.3 19.9 37.8 39.2 22.3 36.9 _P+20T+1 20 1 21.8 21.6 22.9 20.0 40.3 40.8 22.9 36.8 _P+30T+1 30 1 22.3 21.9 23.3 20.2 42.4 42.3 23.3 36.2 _P+40T+1 40 1 22.8 22.2 23.7 20.5 44.0 42.8 23.7 36.2 _P+50T+1 50 1 23.2 22.5 2 4.1 20.7 45.5 43.5 24.1 35.7 _P - 50T+2 - 50 2 12.9 13.2 12.9 13.9 22.9 24.7 12.9 27.9 _P - 40T+2 - 40 2 15.5 15.6 15.9 16.3 25.8 27.6 15.9 31.3 _P - 30T+2 - 30 2 17.5 17.5 18.4 17.5 28.6 30.4 18.4 33.2 _P - 20T+2 - 20 2 19.1 19.1 20.1 18.6 32.0 33.8 20.1 35.4 _P - 10T+2 - 10 2 20.1 20.1 21.4 19.3 35.9 37.2 21.4 36.8 P_P+0T+2 0 2 21.1 21.0 22.4 19.9 38.9 40.0 22.4 37.4 _P+10T+2 10 2 21.9 21.7 23.1 20.3 42.2 42.1 23.1 37.2 _P+20T+2 20 2 22.4 22.1 23.7 20.3 44.0 43.1 23.7 36.9 _P+30T+2 30 2 22.9 22.4 24.1 20.5 45.5 43.7 24.1 36.3 _P+40T+2 40 2 23.4 22.8 24.5 20.8 46.1 44.0 24.5 36.3 _P+50T+2 50 2 23.8 23.1 24.9 21.0 47.3 44.4 24.9 35.7 _P - 50T+3 - 50 3 13.4 13.7 13.6 14.3 24.8 26.6 13.6 28.7 _P - 40T+3 - 40 3 16.1 16.1 16.8 16.6 28.1 29.8 16.8 32.5 _P - 30T+3 - 30 3 18 .0 17.9 19.1 17.9 31.7 33.3 19.1 34.2 _P - 20T+3 - 20 3 19.6 19.6 20.8 19.0 35.5 37.0 20.8 36.5 _P - 10T+3 - 10 3 20.7 20.6 22.1 19.6 39.6 40.1 22.1 37.6 P_P+0T+3 0 3 21.7 21.6 23.1 20.2 42.5 42.8 23.1 37.9 _P+10T+3 10 3 22.4 22.3 23.8 20.6 45.0 43.8 23.8 37 .4 _P+20T+3 20 3 23.0 22.6 24.3 20.6 46.3 44.2 24.3 37.1 _P+30T+3 30 3 23.5 22.9 24.8 20.8 47.1 44.4 24.8 36.5 _P+40T+3 40 3 23.9 23.2 25.1 21.1 47.3 44.5 25.1 36.4 _P+50T+3 50 3 24.3 23.5 25.5 21.4 48.3 44.7 25.5 35.7 _P - 50T+4 - 50 4 14.0 14.2 14.3 14 .7 26.7 28.5 14.3 29.3 _P - 40T+4 - 40 4 16.6 16.5 17.5 16.9 30.5 32.1 17.5 33.2 _P - 30T+4 - 30 4 18.4 18.4 19.7 18.1 34.6 35.4 19.7 34.9 _P - 20T+4 - 20 4 20.1 20.1 21.4 19.2 38.6 39.6 21.4 36.9 _P - 10T+4 - 10 4 21.2 21.1 22.7 19.8 42.8 42.1 22.7 37.8 P_P+0T+4 0 4 22.2 22.1 23.7 20.5 45.1 44.3 23.7 38.2 _P+10T+4 10 4 23.0 22.7 24.4 20.9 46.9 44.6 24.4 37.4 _P+20T+4 20 4 23.5 23.0 25.0 20.9 47.7 44.7 25.0 37.1 158 Combinatio n preci p (%) Tem p (C) MB - na1 - AD - yield (t/ha ) MB - na10 - AD - yield (t/ha) MB - na11 - AD - yield (t/ha) MB - spk - AD - yield (t/ha ) MB - na1 - FJ - yield (t/ha ) MB - na10 - FJ - yield (t/ha) MB - na11 - FJ - yield (t/ha) MB - spk - FJ - yield (t/ha ) _P+30T+4 30 4 24.0 23.3 25.5 21.1 48.2 44.8 25.5 36.5 _P+40T+4 40 4 24.3 23.6 25.8 21.4 48.0 44.7 25.8 36.4 _P+50T+4 50 4 24.7 23.9 26.2 21.7 48.9 44.9 26.2 35.7 _P - 50T+5 - 50 5 14.5 14.7 15.0 15.1 28.6 30.1 15.0 29.6 _P - 40T+5 - 40 5 17.0 16.9 18.0 17.2 32.8 34.0 18.0 33.6 _P - 30T+5 - 30 5 18.8 18 .8 20.2 18.4 37.1 37.5 20.2 35.1 _P - 20T+5 - 20 5 20.6 20.6 22.0 19.5 41.3 41.5 22.0 37.0 _P - 10T+5 - 10 5 21.7 21.5 23.3 20.1 45.4 43.7 23.3 37.8 P_P+0T+5 0 5 22.7 22.4 24.4 20.8 47.0 45.3 24.4 38.2 _P+10T+5 10 5 23.4 23.1 25.2 21.2 48.0 44.9 25.2 37.4 _ P+20T+5 20 5 23.9 23.4 25.7 21.2 48.4 44.9 25.7 37.1 _P+30T+5 30 5 24.4 23.7 26.2 21.4 48.4 44.9 26.2 36.5 _P+40T+5 40 5 24.7 24.0 26.5 21.7 48.1 44.7 26.5 36.4 _P+50T+5 50 5 25.1 24.2 26.9 22.0 49.0 44.9 26.9 35.7 159 Table 3. 8 Summary data for sensitivity analysis Namulonge (NAM), Uganda Combinati on preci p (%) Tem p (C) NAM - na1 - AD - yield (t/ha) NAM - na10 - AD - yield (t/ha) NAM - na11 - AD - yield (t/ha) NAM - spk - AD - yield (t/ha) NAM - na1 - FJ - yield (t/ha) NAM - na10 - FJ - yield (t/ha) NAM - na11 - FJ - yield (t/ha) NAM - spk - FJ - yield (t/ha) _P - 50T - 5 - 50 - 5 11.5 12.2 12.0 13.7 11.5 12.7 12.0 13.2 _P - 40T - 5 - 40 - 5 12.5 13.1 12.1 15.4 12.5 13.7 13.1 15.5 _P - 30T - 5 - 30 - 5 14.0 14.8 13.6 16.7 13.2 15.1 14.2 16.5 _P - 20T - 5 - 20 - 5 15.3 16.2 14.9 17.8 14.5 16.2 15.2 17.2 _P - 10T - 5 - 10 - 5 16.6 17.5 16.1 18.8 15.5 17.2 16.1 18.1 P_P+0T - 5 0 - 5 17.7 18.6 17.1 19.7 16.6 18.1 16.9 18.7 _P+10T - 5 10 - 5 18.8 19.6 18.0 20.4 17.4 18.7 17.6 19.3 _P+20T - 5 20 - 5 19.8 20.4 18.9 21.1 18.2 19.3 18.4 19.9 _P+30T - 5 30 - 5 20.6 21.1 19.7 21.7 18.9 19.9 19.0 20.3 _P+40T - 5 40 - 5 21.4 21.5 20.5 22.2 19.6 20.3 19.7 20.7 _P+50T - 5 50 - 5 22.1 22.1 21.1 22.6 20.0 20.8 20.2 21.1 _P - 50T - 4 - 50 - 4 12.1 13.0 11.5 16.1 20.5 13.2 12.5 15.5 _P - 40T - 4 - 40 - 4 14.1 14.9 13.5 17.8 12.7 14.7 13.9 16.8 _P - 30T - 4 - 30 - 4 15.9 16.8 15.4 19.3 14.1 16.2 15.2 17.9 _P - 20T - 4 - 20 - 4 17.4 18.5 16.9 20.4 15.5 17.4 16.3 18.7 _P - 10T - 4 - 10 - 4 19.0 20.1 18.4 21.7 16.6 18.7 17.4 19.7 P_P+0T - 4 0 - 4 20.3 21.4 19.5 22.6 17.9 19.7 18.3 20.5 _P+10T - 4 10 - 4 21.6 22.5 20.6 23.4 19.0 20.4 19.2 21.2 _P+20T - 4 20 - 4 22.8 23.3 21.6 24.0 19.9 21. 1 20.1 21.9 _P+30T - 4 30 - 4 23.8 24.1 22.5 24.6 20.8 21.8 20.8 22.3 _P+40T - 4 40 - 4 24.8 24.6 23.3 25.3 21.5 22.3 21.5 22.7 _P+50T - 4 50 - 4 25.5 25.2 24.0 25.7 22.0 22.7 22.0 23.2 _P - 50T - 3 - 50 - 3 14.0 15.2 13.4 18.9 12.5 14.4 13.5 17.1 _P - 40T - 3 - 40 - 3 16 .6 17.7 15.9 21.0 13.8 16.1 15.0 18.6 _P - 30T - 3 - 30 - 3 18.6 19.8 18.2 22.4 15.3 17.9 16.6 19.9 _P - 20T - 3 - 20 - 3 20.3 21.5 19.9 23.4 17.0 19.2 17.9 20.8 _P - 10T - 3 - 10 - 3 22.2 23.4 21.7 24.8 18.3 20.7 19.2 22.0 P_P+0T - 3 0 - 3 23.7 24.9 23.1 25.6 19.9 22.0 20 .4 23.0 _P+10T - 3 10 - 3 25.2 26.0 24.4 26.3 21.2 22.9 21.5 23.7 _P+20T - 3 20 - 3 26.4 26.9 25.5 26.9 22.6 23.9 22.6 24.6 _P+30T - 3 30 - 3 27.4 27.7 26.5 27.1 23.6 24.6 23.4 25.2 _P+40T - 3 40 - 3 28.8 28.4 27.4 28.1 24.5 25.2 24.2 25.4 _P+50T - 3 50 - 3 28.8 29. 4 28.3 28.9 25.1 26.5 24.9 26.8 _P - 50T - 2 - 50 - 2 16.6 18.0 15.9 22.1 26.2 16.0 14.8 19.4 _P - 40T - 2 - 40 - 2 19.4 20.8 18.7 24.2 15.2 18.0 16.7 21.1 160 Combinati on preci p (%) Tem p (C) NAM - na1 - AD - yield (t/ha) NAM - na10 - AD - yield (t/ha) NAM - na11 - AD - yield (t/ha) NAM - spk - AD - yield (t/ha) NAM - na1 - FJ - yield (t/ha) NAM - na10 - FJ - yield (t/ha) NAM - na11 - FJ - yield (t/ha) NAM - spk - FJ - yield (t/ha) _P - 30T - 2 - 30 - 2 21.6 23.1 21.3 25.8 17.1 20.1 18.6 22.6 _P - 20T - 2 - 20 - 2 23.6 24.9 23.3 26.7 19.1 21.7 20.2 23.6 _P - 10T - 2 - 10 - 2 25.8 27.0 25.4 27.6 20.8 23.5 21.8 24.9 P_P+0T - 2 0 - 2 27.2 28.3 26.9 28.3 22.7 25.3 23.3 26.1 _P+10T - 2 10 - 2 28.8 29.4 28.3 28.9 24.4 26.5 24.9 26.8 _P+20T - 2 20 - 2 30.1 30.6 29.6 29.7 26.2 27.4 26.3 27.7 _P+30T - 2 30 - 2 31.1 31.1 30.7 29.6 27.5 28.3 27.2 28.2 _P+40T - 2 40 - 2 32.9 32.3 31.9 30.7 28.8 28.9 28.2 28.4 _P+50T - 2 50 - 2 33.6 32.7 32.6 30.7 29.4 29.4 28.9 28.6 _P - 50T - 1 - 50 - 1 19.2 20.9 18.6 24.9 16.1 18.5 16.8 22.2 _P - 40T - 1 - 40 - 1 22.4 24.0 21.8 27.4 17.3 20.9 19.2 23.9 _P - 30T - 1 - 30 - 1 25.1 26.6 24.8 29.0 19.7 23.2 21.4 25.8 _P - 20T - 1 - 20 - 1 27.1 28.5 27.1 29.7 22.0 25.1 23.5 27.4 _P - 10T - 1 - 10 - 1 29.3 30.4 29.5 30.3 24.1 27.5 25.6 28.6 P_P+0T - 1 0 - 1 30.9 31.9 31.3 31.4 26.7 29.6 27.6 29.6 _P+10T - 1 10 - 1 32.6 33.1 32.9 31.5 28.9 30.5 29.6 30.2 _P+20T - 1 20 - 1 34.2 34.6 34.6 32.1 30.6 31.7 31.1 30.5 _P+30T - 1 30 - 1 35.4 35.2 35.5 31.7 32.6 32.8 32.4 31.0 _P+40T - 1 40 - 1 37.4 36.4 37.3 32.6 33.8 33.4 33.5 31.4 _P+50T - 1 50 - 1 38.1 36.5 38.2 32.4 34.7 33.9 34.8 30.7 _P - 50T+0 - 50 0 22.6 24.3 21.7 27.9 18.5 21.4 19.6 25.0 _P - 40T+0 - 40 0 25.8 27.5 25.4 30.3 20.2 24.0 22.5 27.3 _P - 30T+0 - 30 0 28.9 30.4 29.0 31.6 22.8 27.0 25.6 29.4 _P - 20T+0 - 20 0 31.0 3 2.5 31.5 32.4 25.8 29.8 28.3 30.9 _P - 10T+0 - 10 0 33.3 34.2 34.1 32.7 28.7 32.4 31.1 32.2 P_P+0T+0 0 0 35.2 36.0 36.1 33.6 31.9 33.9 33.7 32.7 _P+10T+0 10 0 36.7 36.3 37.8 33.6 34.2 34.6 35.9 32.9 _P+20T+0 20 0 38.5 37.8 39.8 34.1 35.7 35.9 37.9 32.7 _ P+30T+0 30 0 39.2 37.7 40.9 33.3 37.4 37.1 39.3 33.1 _P+40T+0 40 0 40.9 39.2 42.9 33.9 39.1 37.8 41.1 32.7 _P+50T+0 50 0 41.8 39.3 43.7 33.6 40.4 37.8 41.8 31.5 _P - 50T+1 - 50 1 26.3 28.0 25.6 30.1 21.0 24.8 22.8 27.8 _P - 40T+1 - 40 1 29.8 31.4 29.5 32.1 2 3.5 27.9 26.7 30.7 161 Combinati on preci p (%) Tem p (C) NAM - na1 - AD - yield (t/ha) NAM - na10 - AD - yield (t/ha) NAM - na11 - AD - yield (t/ha) NAM - spk - AD - yield (t/ha) NAM - na1 - FJ - yield (t/ha) NAM - na10 - FJ - yield (t/ha) NAM - na11 - FJ - yield (t/ha) NAM - spk - FJ - yield (t/ha) _P - 30T+1 - 30 1 33.0 34.4 33.5 33.6 26.9 31.8 30.8 32.9 _P - 20T+1 - 20 1 35.3 36.4 36.8 34.4 31.1 34.5 34.4 33.8 _P - 10T+1 - 10 1 37.2 37.2 39.2 34.7 34.1 37.2 38.1 34.6 P_P+0T+1 0 1 39.0 39.1 41.2 35.2 37.3 37.9 40.6 34.8 _P+10T+1 10 1 40.2 39.2 43.0 35.1 39.0 38.9 42.7 34.3 _P+20T+1 20 1 41.4 40.5 44.8 35.4 40.3 39.9 44.5 33.9 _P+30T+1 30 1 42.0 40.4 45.5 34.2 42.0 40.3 45.9 33.9 _P+40T+1 40 1 44.0 42.0 47.1 34.6 43.5 40.3 47.4 33.2 _P+50T+1 50 1 44.8 42.1 48.0 34.2 44.1 39.6 47.8 31.9 _P - 50T+2 - 50 2 29.8 31.2 29.4 32.0 26.8 28.1 26.7 30.7 _P - 40T+2 - 40 2 33.7 34.4 34.1 34.0 27.1 32.3 32.0 33.4 _P - 30T+2 - 30 2 36.8 37.1 38.1 34.9 31.5 36.2 36.8 35.1 _P - 20T+2 - 20 2 38.6 38.9 41.3 35.8 35.7 39.0 41.2 35.5 _P - 10T+2 - 10 2 40.0 39.8 43.4 35.6 39.5 40.2 44.4 36.0 P_P+0T+2 0 2 41.9 41.8 45. 5 36.2 41.3 40.8 46.8 35.2 _P+10T+2 10 2 42.9 41.7 46.9 35.7 42.6 41.2 48.5 34.6 _P+20T+2 20 2 44.2 43.0 48.7 35.7 43.0 42.2 50.4 34.3 _P+30T+2 30 2 44.6 42.4 49.4 34.4 44.6 41.9 51.3 34.2 _P+40T+2 40 2 46.4 43.9 51.1 34.9 45.6 41.6 52.2 33.5 _P+50T+2 50 2 47.2 43.9 52.3 34.4 45.8 40.4 51.7 32.2 _P - 50T+3 - 50 3 34.1 35.4 55.8 34.4 32.1 40.8 54.3 32.2 _P - 40T+3 - 40 3 36.7 37.1 37.8 35.3 34.0 36.0 37.0 35.8 _P - 30T+3 - 30 3 39.2 39.2 41.8 35.6 35.9 39.8 42.6 36.2 _P - 20T+3 - 20 3 41.3 41.3 44.6 36.3 39.9 4 1.6 46.8 36.4 _P - 10T+3 - 10 3 42.9 42.2 46.7 35.9 42.8 42.2 49.2 36.3 P_P+0T+3 0 3 44.6 43.7 49.2 36.4 43.7 42.3 51.1 35.4 _P+10T+3 10 3 45.4 43.6 50.7 35.9 44.5 42.4 51.9 34.7 _P+20T+3 20 3 46.4 44.1 52.9 35.7 44.4 43.2 53.2 34.4 _P+30T+3 30 3 46.3 43 .1 53.2 34.4 46.2 42.7 53.9 34.5 _P+40T+3 40 3 47.6 44.5 55.1 35.0 46.9 42.1 54.6 33.5 _P+50T+3 50 3 48.1 44.4 55.8 34.4 47.0 40.8 54.3 32.2 _P - 50T+4 - 50 4 36.5 36.9 36.8 33.8 46.0 36.7 36.0 35.1 _P - 40T+4 - 40 4 39.5 39.3 41.2 35.8 36.0 40.2 42.3 36.7 _P - 30T+4 - 30 4 41.5 41.3 44.7 36.0 39.9 42.4 47.1 37.2 _P - 20T+4 - 20 4 43.9 43.5 48.0 36.5 43.0 43.4 50.7 36.8 162 Combinati on preci p (%) Tem p (C) NAM - na1 - AD - yield (t/ha) NAM - na10 - AD - yield (t/ha) NAM - na11 - AD - yield (t/ha) NAM - spk - AD - yield (t/ha) NAM - na1 - FJ - yield (t/ha) NAM - na10 - FJ - yield (t/ha) NAM - na11 - FJ - yield (t/ha) NAM - spk - FJ - yield (t/ha) _P - 10T+4 - 10 4 45.2 43.7 50.2 36.0 45.1 42.9 52.8 36.5 P_P+0T+4 0 4 46.7 44.6 52.8 36.5 45.3 42.8 53.9 35.6 _P+10T+4 10 4 47.0 44.2 54.3 36.0 45.6 43.0 54.3 34.9 _P+20T+4 20 4 47.4 44.5 56.0 35.7 45.4 43.6 55.8 34.6 _P+30T+4 30 4 46.8 43.5 56.0 34.5 46.7 43.2 56.3 34.7 _P+40T+4 40 4 48 .1 44.8 57.5 35.0 47.5 42.4 56.7 33.6 _P+50T+4 50 4 48.7 44.7 58.2 34.5 47.6 40.9 56.0 32.3 _P - 50T+5 - 50 5 38.8 38.8 39.8 34.1 38.4 39.2 41.8 36.0 _P - 40T+5 - 40 5 42.1 41.2 44.1 35.9 40.5 42.3 47.4 37.0 _P - 30T+5 - 30 5 43.9 42.9 47.6 36.2 43.5 43.8 50.8 37.3 _P - 20T+5 - 20 5 46.3 44.7 51.5 36.5 45.3 44.0 53.2 37.0 _P - 10T+5 - 10 5 46.6 44.2 53.7 36.0 46.3 43.3 54.8 36.6 P_P+0T+5 0 5 47.4 45.0 55.8 36.5 45.9 43.0 55.6 35.7 _P+10T+5 10 5 47.6 44.4 56.8 36.0 45.9 43.2 55.8 35.0 _P+20T+5 20 5 47.8 44.7 57.9 35.8 45.8 43.8 57.1 34.7 _P+30T+5 30 5 47.1 43.6 57.2 34.5 47.1 43.4 57.6 34.7 _P+40T+5 40 5 48.4 44.9 58.6 35.0 48.0 42.4 57.7 33.6 _P+50T+5 50 5 48.9 44.9 59.2 34.5 47.9 41.1 56.6 32.3 163 APPENDIX B: Mean changes in sweet potato yield Table 3. 9 Ensemble mean changes in sweet potato root yield (t/ha) under future climate scenarios for 4 GCMs; CSIRO, MIROC5, MRI - CGCM3, NorESM1 - M for August - Decem ber (ASOND) season for NASPOT 1 : Base yield, 2030 - rcp 4.5, 2050 - rcp4.5, 2070 - rcp 4.5, 2030 - rcp8 . 5, 2050 - r cp8.5, 2070 - rcp8.5 Location Base root yield - na1 - ASON D Change in root yield (mm) na1 - av - ASOND - 2030 - rcp45 na1 - av - ASOND - 2050 - rcp45 na1 - av - ASOND - 2070 - rcp45 na1 - av - ASOND - 2030 - rcp85 na1 - av - ASOND - 2050 - rcp85 na1 - av - ASOND - 2050 - rcp85 Aru a 47.9 - 3.9 - 4.6 - 5.0 - 6.1 - 6.1 - 4.1 Arusha 6.5 14.1 13.5 13.1 14.5 13.8 13.8 Bukoba 37.8 - 2.3 - 1.5 - 0.6 - 0.4 - 1.9 0.0 Dagoretti_Corn er 17.0 3.8 2.7 2.0 3.4 1.6 1.5 Dar es Salaam 29.4 0.7 0.0 - 0.4 0.9 - 1.2 0.4 Dodoma 18.7 1.8 1.5 1.1 1.2 1.2 1.1 Eldo ret 15.7 1.1 3.0 5.1 1.7 3.4 5.0 Entebbe 28.3 3.8 4.4 6.5 4.9 5.6 7.0 Garissa 22.6 4.1 3.3 3.5 3.1 2.8 2.3 Gulu 35.6 6.9 7.4 7.9 7.4 7.0 7.3 Jinja 28.3 7.5 6.9 6.5 8.5 8.3 7.3 Kabale 21.3 - 1.9 - 1.4 - 1.2 - 2.7 - 2.9 - 2.1 Kasese 37.3 - 1.6 - 0.8 - 0.3 - 0.4 - 1.7 0.0 Kigoma 29.6 0.7 1.2 - 0.4 0.5 0.4 - 1.0 Kisumu 30.8 4.7 6.0 6.2 5.7 6.4 6.9 Kitgum 35.9 3.6 4.6 5.2 4.0 4.4 5.5 Lamu 21.0 5.9 6.4 6.3 5.0 4.8 4.8 Lodwar 19.3 0.0 0.9 1.1 3.2 3.4 3.4 Makindu 28.3 - 0.2 - 0.6 - 0.8 - 1.6 - 1.5 - 2.0 Mandera 23.1 1.8 1.3 2.0 1.7 1.5 2.1 Marsabit 24.3 1.6 - 1.3 - 2.9 2.0 - 0.8 - 1.9 Masaka 28.5 1.3 2.2 2.5 2.3 3.0 2.8 Masindi 41.0 - 4.1 - 2.5 - 0.7 - 1.9 - 2.5 - 1.4 Mbarara 33.2 - 1.7 - 3.2 - 3.4 - 1.2 - 3.4 - 2.8 Mbeya 20.6 - 1.2 - 2.0 - 2.1 - 1.4 - 1.3 - 2.0 Mtwara 22.1 3.7 3.1 3.0 2 .2 2.4 2.8 Musoma 24.5 4.0 3.2 3.3 4.6 4.2 3.7 Mwanza 30.9 0.0 1.1 0.7 0.9 0.8 0.4 Namulonge 36.9 - 1.7 - 1.4 - 0.8 - 0.6 - 1.9 - 1.2 Narok 11.4 4.6 3.5 2.7 5.3 4.2 3.1 Same 16.4 4.2 3.9 3.4 3.3 3.0 2.3 164 Location Base root yield - na1 - ASON D Change in root yield (mm) na1 - av - ASOND - 2030 - rcp45 na1 - av - ASOND - 2050 - rcp45 na1 - av - ASOND - 2070 - rcp45 na1 - av - ASOND - 2030 - rcp85 na1 - av - ASOND - 2050 - rcp85 na1 - av - ASOND - 2050 - rcp85 Serere 33.3 1.6 3.3 5.2 1.3 4.1 6.2 Songea 20.4 0.6 1.2 0.7 2.0 2.3 1.1 Soroti 35.1 - 0.1 1.9 4.5 2.8 4.5 5.5 Tabora 25.6 0.9 0.8 0.7 2.6 2.4 1.4 Tororo 41.6 - 6.3 - 4.4 - 2.8 - 5.8 - 3.8 - 2.4 Voi 26.3 0.9 0.9 0.4 - 0.2 0.1 - 0.2 Wajir 23.2 5.2 4.8 4.9 5.4 3.6 4.8 Table 3. 10 Ensemble mean changes in sweet p otato root yield (t/ha) under future climate scenarios for 4 GCMs; CSIRO, MIROC5, MRI - CGCM3, NorESM1 - M for August - December (ASOND) season for NASPOT 11 Location Base root yield - na11 - ASOND Change in root yield (mm) na11 - av - ASOND - 2030 - rcp45 na11 - av - ASOND - 2050 - rcp45 na11 - av - ASOND - 2070 - rcp45 na11 - av - ASOND - 2030 - rcp85 na11 - av - ASOND - 2050 - rcp85 na11 - av - ASOND - 2070 - rcp85 Arua 58.3 - 4.2 - 5.3 - 5.7 - 6.7 - 7.2 - 4.2 Arusha 6.2 14.9 14.0 13.3 15.4 14.3 14.5 Bukoba 42.0 - 1.7 - 0.4 0.8 1.0 - 1.0 1.3 Dagoretti_ Corner 17. 0 4.9 3.4 2.6 4.4 2.3 2.2 Dar es Salaam 31.5 2.5 1.5 1.1 2.3 0.2 1.5 Dodoma 18.7 1.8 1.3 0.8 1.3 1.2 1.0 Eldoret 14.8 1.7 3.5 5.3 1.9 3.6 5.4 Entebbe 29.6 5.5 5.9 8.9 6.7 7.3 9.9 Garissa 21.8 5.4 4.2 4.6 4.2 3.5 3.0 Gulu 44.5 4.8 5.8 6.3 5.9 6.2 6.4 Jinja 28.7 10.0 9.8 9.0 11.9 11.2 10.3 Kabale 20.5 - 0.8 - 0.9 - 0.7 - 2.0 - 2.4 - 1.5 Kasese 37.1 - 0.4 0.8 2.1 2.1 0.1 2.4 Kigoma 30.7 2.2 2.3 0.4 2.2 1.9 - 0.2 Kisumu 31.1 8.0 8.9 9.1 9.2 9.3 10.2 Kitgum 34.8 7.2 9.3 10.5 9.1 10.3 12.5 Lamu 21.0 7.6 9.2 9.3 6.6 7.0 7.2 Lodwar 19.1 - 0.2 0.7 1.0 3.0 3.2 3.2 165 Location Base root yield - na11 - ASON D Change in root yield (mm) na11 - av - ASOND - 2030 - rcp45 na11 - av - ASOND - 2050 - rcp45 na11 - av - ASOND - 2070 - rcp45 na11 - av - ASOND - 2030 - rcp85 na11 - av - ASOND - 2050 - rcp85 na11 - av - ASOND - 2070 - rcp85 Makindu 30.6 0.3 - 0.1 - 0.4 - 1.7 - 1.8 - 2.4 Mandera 22.4 1.8 1.1 2.2 2.1 1.6 2.4 Marsabit 24.7 2.7 - 1.1 - 2.8 3.2 - 0.5 - 1.9 Masaka 29.4 2.1 2.7 3.4 3.7 4.0 4.1 Masindi 36.1 3.7 6.6 9.7 7.0 7.2 8.8 Mbarara 33.3 - 0.5 - 2 .8 - 2.2 0.4 - 2.6 - 2.0 Mbeya 21.4 - 1.0 - 1.8 - 1.9 - 1.3 - 1.3 - 2.2 Mtwara 22.5 4.9 4.3 3.6 2.9 3.2 3.1 Musoma 24.2 5.9 4.9 4.7 7.3 7.0 6.1 Mwanza 34.2 0.0 0.8 0.5 1.5 0.8 0.3 Namulonge 31.0 7.9 7.6 8.5 8.7 6.5 8.0 Narok 10.9 5.3 3.8 2.9 6.3 4.7 3.4 Same 15.9 4.7 4.4 3.9 3.6 3.3 2.7 Serere 32.8 4.5 6.2 9.0 4.6 8.2 12.2 Songea 21.2 1.1 1.6 1.2 2.8 2.6 1.5 Soroti 35.4 2.1 4.9 8.0 5.1 7.6 9.7 Tabora 26.6 2.1 1.6 1.3 4.3 3.3 2.0 Tororo 45.4 - 6.8 - 4.8 - 2.2 - 5.8 - 2.9 - 1.2 Voi 28.7 0.5 - 0.1 - 0.8 - 1.0 - 1.2 - 1.3 Wajir 22.6 7.1 6.3 5.9 6.7 4.9 5.6 166 Table 3. 11 Ensemble mean changes in sweet potato root yield (t/ha) under future climate scenarios for 4 GCMs; CSIRO, MIROC5, MRI - CGCM3, NorESM1 - M for August - December (ASOND) sea son for NASPOT 10 Location Base root yield - na10 - ASON D Change in root yield (mm) na10 - av - ASOND - 2030 - rcp45 na10 - av - ASOND - 2050 - rcp45 na10 - av - ASOND - 2070 - rcp45 na10 - av - ASOND - 2030 - rcp85 na10 - av - ASOND - 2050 - rcp85 na10 - av - ASOND - 2070 - rcp85 Arua 44.5 - 0.5 - 1.2 - 1 .6 - 2.6 - 2.7 - 0.7 Arusha 6.7 13.8 13.3 12.9 14.3 13.6 13.5 Bukoba 38.1 - 2.5 - 1.7 - 0.8 - 0.6 - 2.2 - 0.2 Dagoretti_Corn er 17.5 3.4 2.3 1.6 2.9 1.2 1.1 Dar es Salaam 28.4 1.8 1.0 0.6 1.9 - 0.2 1.4 Dodoma 19.3 1.2 1.0 0.6 0.7 0.7 0.5 Eldoret 17.6 - 0.8 1.1 3 .2 - 0.2 1.5 3.1 Entebbe 28.8 3.3 3.9 6.0 4.4 5.1 6.5 Garissa 21.9 4.8 4.0 4.2 3.8 3.5 3.0 Gulu 40.1 2.4 2.8 3.3 2.8 2.5 2.8 Jinja 28.3 7.5 6.9 6.5 8.5 8.3 7.3 Kabale 22.5 - 3.1 - 2.7 - 2.5 - 3.9 - 4.2 - 3.4 Kasese 28.9 6.8 7.5 8.0 8.0 6.7 8.3 Kigoma 29.1 1.2 1.7 0.1 1.0 0.9 - 0.5 Kisumu 31.6 4.0 5.2 5.4 4.9 5.7 6.2 Kitgum 35.9 3.7 4.6 5.2 4.1 4.5 5.5 Lamu 21.0 5.9 6.4 6.3 5.0 4.9 4.8 Lodwar 19.3 0.0 0.9 1.1 3.2 3.4 3.4 Makindu 27.3 0.8 0.4 0.1 - 0.7 - 0.6 - 1.0 Mandera 22.3 2.6 2.1 2.8 2.5 2.2 2.8 Marsa bit 23.9 2.1 - 0.8 - 2.4 2.4 - 0.3 - 1.4 Masaka 28.9 0.9 1.7 2.0 1.9 2.5 2.3 Masindi 36.2 0.8 2.3 4.1 2.9 2.3 3.4 Mbarara 34.1 - 2.6 - 4.2 - 4.4 - 2.1 - 4.4 - 3.8 Mbeya 20.7 - 1.3 - 2.1 - 2.2 - 1.5 - 1.4 - 2.1 Mtwara 22.0 3.9 3.2 3.2 2.4 2.5 2.9 Musoma 24.6 3.9 3.1 3.1 4.5 4.1 3.5 Mwanza 30.6 0.3 1.5 1.0 1.3 1.1 0.7 Namulonge 31.6 3.6 3.9 4.4 4.7 3.4 4.1 Narok 12.0 4.0 3.0 2.1 4.7 3.6 2.5 Same 16.7 3.8 3.6 3.1 2.9 2.7 2.0 Serere 34.0 0.8 2.6 4.4 0.5 3.4 5.4 Songea 20.3 0.7 1.3 0.8 2.1 2.4 1.2 Soroti 35.0 0.0 2 .0 4.6 2.9 4.6 5.5 167 Location Base root yield - na10 - ASON D Change in root yield (mm) na10 - av - ASOND - 2030 - rcp45 na10 - av - ASOND - 2050 - rcp45 na10 - av - ASOND - 2070 - rcp45 na10 - av - ASOND - 2030 - rcp85 na10 - av - ASOND - 2050 - rcp85 na10 - av - ASOND - 2070 - rcp85 Tabora 25.0 1.5 1.4 1.2 3.2 3.0 2.0 Tororo 41.3 - 6.0 - 4.1 - 2.5 - 5.5 - 3.5 - 2.1 Voi 26.3 0.9 0.9 0.4 - 0.2 0.1 - 0.2 Wajir 22.3 6.1 5.7 5.8 6.4 4.6 5.7 Table 3. 12 Ensemble mean changes in sweet potato root yield under future c limate scenarios for 4 GCMs; CSIRO, MIROC5, MRI - CGCM3, NorESM1 - M for August - December (ASOND) season for SPK004 Location Base root yield - spk - ASON D Change in root yield (t/ha) spk - av - ASOND - 2030 - rcp45 spk - av - ASOND - 2050 - rcp45 spk - av - ASOND - 2070 - rcp45 spk - av - ASOND - 2030 - rcp85 spk - av - ASOND - 2050 - rcp85 spk - av - ASOND - 2070 - rcp85 Arua 36.1 - 2.9 - 3.4 - 3.7 - 4.1 - 4.0 - 2.5 Arusha 7.9 13.4 13.4 13.1 13.7 13.3 13.2 Bukoba 34.8 - 4.4 - 4.9 - 4.0 - 2.7 - 4.5 - 3.1 Dagoretti_Corn er 18.0 2.5 1.8 1.5 2.6 1.3 1.1 Dar es Salaam 25 .6 0.3 - 0.7 - 0.8 0.7 - 1.3 0.0 Dodoma 19.1 1.3 1.4 1.0 1.0 1.2 1.0 Eldoret 20.5 0.8 2.6 3.4 - 4.2 - 2.5 3.5 Entebbe 28.3 0.0 0.7 1.7 1.8 2.7 3.6 Garissa 21.0 2.6 2.4 2.4 2.6 2.4 1.9 Gulu 34.5 - 1.4 - 1.4 - 1.2 - 0.3 - 0.2 - 0.3 Jinja 28.8 2.1 1.8 1.5 4.3 4.1 3.2 Kabale 26.2 - 4.5 - 3.5 - 3.3 - 4.5 - 4.9 - 3.8 Kasese 33.7 - 2.2 - 1.9 - 1.8 0.8 0.1 1.5 Kigoma 27.5 - 0.8 - 0.4 - 1.8 - 0.6 - 0.7 - 1.8 Kisumu 30.3 - 0.4 0.5 0.6 1.6 2.4 2.4 Kitgum 32.7 0.0 - 0.3 - 0.1 1.0 1.4 1.9 Lamu 21.4 2.7 2.1 1.6 2.9 1.9 1.6 Lodwar 19.2 0 .0 0.8 1.0 2.9 2.6 2.8 Makindu 24.6 0.0 - 0.1 - 0.2 - 1.0 - 0.7 - 1.0 Mandera 21.7 0.7 0.7 0.8 1.6 1.5 1.9 Marsabit 22.7 1.0 - 0.5 - 1.8 1.9 - 0.3 - 0.9 Masaka 28.6 - 1.2 - 0.6 - 0.2 0.3 0.7 0.2 168 Location Base root yield - spk - ASON D Change in ro ot yield (t/ha) spk - av - ASOND - 2030 - rcp45 spk - av - ASOND - 2050 - rcp45 spk - av - ASOND - 2070 - rcp45 spk - av - ASOND - 2030 - rcp85 spk - av - ASOND - 2050 - rcp85 spk - av - ASOND - 2070 - rcp85 Masindi 33.6 - 1.9 - 1.5 - 1.1 0.0 - 0.7 - 0.3 Mbarara 32.5 - 2.3 - 2.8 - 3.6 - 1.6 - 3.3 - 3.3 Mbeya 18.8 - 0.1 - 0.6 - 0.7 - 0.4 - 0.3 - 0.8 Mtwara 20.9 2.2 1.8 1.9 1.7 1.6 2.1 Musoma 24.2 1.8 1.1 1.7 2.2 2.0 1.8 Mwanza 28.1 - 1.0 0.3 - 0.3 0.1 0.2 - 0.3 Namulonge 31.0 - 0.5 0.2 0.2 2.1 1.4 1.6 Narok 13.7 3.7 3.4 2.4 4.4 3.7 2.9 Same 16.9 3.3 3.3 2.8 2.7 2. 5 1.8 Serere 31.9 - 0.7 - 0.3 - 0.2 - 1.0 0.7 1.8 Songea 19.3 0.2 1.3 0.6 1.6 2.2 1.3 Soroti 32.4 - 1.8 - 1.0 - 0.1 0.8 1.7 2.0 Tabora 23.2 0.1 0.5 0.3 1.4 1.9 1.3 Tororo 35.5 - 4.6 - 3.5 - 3.5 - 4.8 - 3.7 - 2.7 Voi 24.1 0.5 1.4 1.3 - 0.2 0.6 0.3 Wajir 21.3 2.7 2 .8 3.0 3.9 2.6 4.0 169 APPENDIX C: Projected changes (% ) for annual rainfall Table 3. 13 Changes in annual rainfall for the 2030s Location Base line Ann ual rainf all (mm ) Change in rainfall (mm) 2020 - 2049 - CSIRO _ rcp 45 2020 - 2049 - CSIRO _ rcp85 2020 - 2049 - MIRO C_ rcp45 2020 - 2049 - MIROC _ rcp85 2020 - 2049 - MRI - CGCM 3_ rcp45 2020 - 2049 - MRI - CGC M3_ rcp85 2020 - 2049 - NorES M1 - M_ rcp45 2020 - 2049 - NorESM 1 - M_rcp85 Av. RFan - 2030s Arua 1285 6 7 8 7 3 3 20 11 8 Arusha 1058 3 6 - 2 2 2 2 3 1 2 Bukoba 2082 - 5 - 4 - 2 0 - 4 - 3 4 3 - 1 Dagoretti Corner 892 15 14 9 10 6 10 9 8 10 Dar es Salaam 1112 - 3 0 2 4 - 1 1 - 13 - 8 - 2 Dodoma 603 - 2 2 - 4 2 9 2 5 9 3 Eldoret 1091 - 5 - 2 - 4 - 2 - 9 - 10 - 1 3 - 4 Entebbe 1354 11 12 16 16 8 9 25 22 15 Garissa 348 20 3 5 38 32 81 53 23 23 38 Gulu 1379 5 6 9 9 7 7 24 18 11 Jinja 1370 - 9 - 7 - 2 2 - 12 - 11 0 0 - 5 Kabale 1232 - 27 - 25 - 18 - 14 - 23 - 23 - 16 - 19 - 21 Kasese 968 - 11 - 7 - 7 - 2 - 17 - 13 5 - 4 - 7 Kigoma 1049 - 5 - 6 - 4 0 - 11 - 4 - 1 - 4 - 4 Kisumu 1494 - 12 - 14 - 10 - 7 - 26 - 25 - 11 - 8 - 14 Kitigum 1117 3 8 12 13 6 5 22 20 11 Lamu 872 - 1 5 11 13 5 - 2 10 - 4 5 Lodwar 200 25 15 - 12 18 5 60 43 80 29 Makindu 637 - 3 - 4 12 4 - 1 - 8 - 6 - 13 - 3 Mandera 257 - 5 4 12 18 36 3 18 52 17 Marsabiti 614 47 48 45 44 41 47 45 46 45 Masaka 1214 - 11 - 5 - 2 - 3 - 7 - 7 5 4 - 3 Mansindi 1215 0 3 9 6 - 2 2 25 19 8 Mbarara 990 - 16 - 13 - 9 - 9 - 14 - 13 3 - 5 - 9 Mbeya 1062 - 6 - 12 - 1 - 8 - 14 - 12 0 - 3 - 7 Mtwara 1049 - 7 0 - 3 - 7 9 0 - 14 - 9 - 4 Musoma 992 - 5 - 12 - 1 4 - 11 - 8 2 4 - 3 Mwanza 1011 2 2 3 9 1 - 2 11 10 5 170 Location Base line Ann ual rainf all (mm ) Change in rainfall (mm) 2020 - 2049 - CSIRO _ rcp45 2020 - 2049 - CSIRO _ rcp85 2020 - 2049 - MIROC _ rcp45 2020 - 2049 - MIROC _ rcp85 2020 - 2049 - MRI - CGCM 3_ rcp45 2020 - 2049 - MRI - CGCM 3_ rcp85 2020 - 2049 - NorES M1 - M_ rcp45 2020 - 2049 - NorESM 1 - M_rcp85 Av. RFan - 2030s Namulon ge 1259 - 8 - 9 2 1 - 7 - 6 10 10 - 1 Narok 751 12 14 0 15 10 18 12 12 12 Same 554 23 17 15 24 28 20 13 10 19 Serere 1294 - 1 1 6 5 - 1 - 6 9 11 3 Songea 1088 19 17 14 3 12 15 16 27 15 Soroti 1249 6 8 11 8 0 4 13 13 8 Tabora 976 - 13 - 10 - 5 - 1 - 10 - 2 1 0 - 5 Tororo 1616 - 16 - 11 - 8 - 9 - 21 - 22 - 9 - 6 - 13 Voi 620 - 7 - 4 1 0 6 9 1 - 17 - 2 Wajir 312 19 35 34 16 86 39 20 23 34 Table 3. 14 Changes in annual rainfall (mm) for the 2050s Location Basel ine Ann ual rainf all (mm ) Change in rainfall (mm) 2040 - 2069 - CSIRO _rcp45 2040 - 2069 - CSIRO_ rcp85 2040 - 2069 - MIROC_ rcp45 2040 - 2069 - MIROC_ rcp85 2040 - 2069 - MRI - CGCM3_ rcp45 2040 - 2069 - MRI - CGCM3_ rcp85 2040 - 2069 - NorES M1 - M_rcp 45 2040 - 2069 - NorES M1 - M_rcp 8 5 Av. RFa n - 205 0s Arua 1285 5 5 9 7 5 0 21 8 7 Arusha 1058 4 7 0 3 1 4 4 0 3 Bukoba 2082 - 5 - 5 - 4 - 2 - 3 - 5 - 1 - 1 - 3 Dagoretti Corner 892 17 16 8 12 9 13 7 8 11 Dar es Salaam 1112 - 3 0 0 3 - 1 0 - 9 - 11 - 3 Dodoma 603 6 - 1 - 2 2 7 6 4 16 5 Eldoret 1091 - 5 - 5 - 7 - 3 - 8 - 6 0 3 - 4 Entebbe 1354 8 11 17 15 7 12 23 23 14 Garissa 348 21 36 47 29 79 57 25 26 40 Gulu 1379 2 3 6 5 5 5 19 13 7 171 Location Basel ine Ann ual rainf all (mm ) Change in rainfall (mm) 2040 - 2069 - CSIRO _rcp45 2040 - 2069 - CSIR O_ rcp85 2040 - 2069 - MIROC_ rcp45 2040 - 2069 - MIROC_ rcp85 2040 - 2069 - MRI - CGCM3_ rcp45 2040 - 2069 - MRI - CGCM3_ rcp85 2040 - 2069 - NorES M1 - M_rcp 45 2040 - 2069 - NorES M1 - M_rcp 85 Av. RFa n - 205 0s Jinja 1370 - 9 - 6 - 1 - 2 - 14 - 11 - 2 1 - 6 Kabale 1232 - 27 - 25 - 19 - 17 - 23 - 24 - 15 - 19 - 21 Kasese 968 - 13 - 13 - 5 - 3 - 19 - 19 13 0 - 7 Kigoma 1049 - 10 - 9 - 4 - 3 - 7 - 8 - 1 - 12 - 7 Kisumu 1494 - 12 - 13 - 9 - 8 - 23 - 24 - 9 - 6 - 13 Kitigum 1117 2 6 7 4 1 5 18 11 7 Lamu 872 8 12 23 18 17 6 14 1 12 Lodwar 200 38 24 - 3 32 21 76 59 69 39 Makindu 637 4 5 19 7 4 - 7 - 4 - 4 3 Mandera 257 - 4 8 16 17 41 10 29 56 22 Marsabit i 614 39 40 39 44 41 48 36 33 40 Masaka 1214 - 7 - 5 - 5 - 2 - 9 - 9 7 4 - 3 Mansindi 1215 2 2 10 12 - 1 4 22 17 8 Mbarara 990 - 13 - 13 - 11 - 11 - 13 - 17 2 - 7 - 10 Mbeya 1062 - 4 - 9 3 - 2 - 10 - 4 0 2 - 3 Mtwara 1049 3 2 - 1 - 7 7 7 - 5 - 14 - 1 Musoma 992 - 11 - 6 2 3 - 14 - 7 - 3 5 - 4 Mwanza 1011 - 2 1 7 8 0 - 3 4 12 3 Namulon ge 1259 - 7 - 7 0 0 - 6 - 7 9 7 - 2 Narok 751 7 8 - 3 16 - 1 9 3 2 5 Same 554 21 16 15 24 29 18 11 10 18 Serere 1294 - 2 1 2 5 - 8 - 6 7 12 1 Songea 1088 23 25 24 10 17 26 26 25 22 Soroti 1249 1 8 9 8 - 5 2 13 11 6 Tabora 976 - 4 - 3 - 5 5 - 5 2 9 9 1 Tororo 1616 - 14 - 10 - 7 - 7 - 22 - 22 - 7 - 5 - 12 Voi 620 - 10 - 2 - 3 2 6 5 5 - 1 0 Wajir 312 25 38 31 25 104 44 32 34 42 172 Table 3. 15 Changes in annual rainfall (mm) for the 2070s Locatio n Basel ine Ann ual rainf all (mm ) Change in rainfall (mm) 2060 - 2089 - CSIRO _rcp45 2060 - 2089 - CSIRO_ rcp85 2060 - 2089 - MIROC_ rcp45 2060 - 2089 - MIROC_ rcp85 2060 - 2089 - MRI - CGCM3_ rcp45 2060 - 2089 - MRI - CGCM3_ rcp 85 2060 - 2089 - NorES M1 - M_rcp 45 2060 - 2089 - NorES M1 - M_rcp 85 Av. RF an - 207 0s Arua 1285 5 12 13 13 4 4 24 11 11 Arusha 1058 5 12 9 14 - 4 11 10 7 8 Bukoba 2082 - 5 - 2 - 2 1 - 3 - 1 1 0 - 1 Dagorett i Corner 892 20 23 13 15 7 17 17 19 17 Dar es Salaam 1112 - 3 1 2 4 0 - 6 - 8 - 11 - 3 Dodoma 603 - 1 0 - 3 - 3 2 4 1 12 2 Eldoret 1091 0 2 0 1 - 5 - 5 5 10 1 Entebbe 1354 14 15 21 23 10 15 28 24 19 Garissa 348 15 25 47 44 86 40 32 19 38 Gulu 1379 5 8 9 4 5 8 21 11 9 Jinja 1370 - 5 - 3 - 1 5 - 11 - 13 4 3 - 3 Kabale 1232 - 24 - 20 - 1 5 - 14 - 18 - 21 - 11 - 15 - 17 Kasese 968 - 14 - 10 - 3 1 - 19 - 20 15 4 - 6 Kigoma 1049 - 14 - 12 - 3 - 7 - 15 - 10 - 6 - 17 - 11 Kisumu 1494 - 9 - 7 - 5 - 4 - 20 - 18 - 5 - 2 - 9 Kitigum 1117 6 13 11 9 6 14 24 14 12 Lamu 872 2 6 19 16 12 7 9 - 2 9 Lodwar 200 44 30 9 32 19 75 70 61 42 Makindu 637 - 6 - 4 15 9 3 0 0 - 14 0 Mandera 257 - 7 9 13 20 45 11 32 46 21 Marsabit i 614 36 41 34 35 35 38 36 27 35 Masaka 1214 - 3 - 5 - 2 1 - 9 - 6 10 5 - 1 Mansind i 1215 5 6 14 15 7 8 31 19 13 Mbarara 990 - 12 - 10 - 8 - 6 - 13 - 13 7 - 4 - 7 Mbeya 1062 - 3 - 9 - 1 - 7 - 11 - 9 - 4 - 4 - 6 Mtwara 1049 - 2 1 - 4 - 8 7 5 - 5 - 11 - 2 Musoma 992 - 6 - 6 1 2 - 5 - 8 4 6 - 1 Mwanza 1011 - 2 4 10 6 1 - 1 6 7 4 173 Locatio n Basel ine Ann ual rainf all (mm ) Change in rainfall (mm) 2060 - 2089 - CSIRO _rcp45 2060 - 2089 - CSI RO_ rcp85 2060 - 2089 - MIROC_ rcp45 2060 - 2089 - MIROC_ rcp85 2060 - 2089 - MRI - CGCM3_ rcp45 2060 - 2089 - MRI - CGCM3_ rcp85 2060 - 2089 - NorES M1 - M_rcp 45 2060 - 2089 - NorES M1 - M_rcp 85 Av. RF an - 207 0s Namulon ge 1259 0 - 2 1 6 - 3 - 4 12 13 3 Narok 751 9 10 - 3 14 5 9 3 6 7 Same 554 11 12 6 16 24 13 8 9 12 Serere 1294 3 8 11 12 - 1 2 15 18 9 Songea 1088 23 15 16 8 8 18 17 23 16 Soroti 1249 4 10 14 16 4 7 20 22 12 Tabora 976 - 9 - 12 - 5 - 1 - 7 - 6 2 - 1 - 5 Tororo 1616 - 7 - 6 - 3 - 1 - 20 - 17 - 1 0 - 7 Voi 620 - 12 - 9 5 3 3 3 - 10 - 12 - 4 Wajir 31 2 27 36 24 20 98 38 28 33 38 Table 3. 16 Changes in the February - June (FMAMJ) rainfall for the 2030s Locatio n Basel ine - FMA MJ rainf all (mm) Change in rainfall (mm) 2020 - 2049 - CSIRO_ rcp45 2020 - 2049 - CSIRO_ rcp85 2020 - 2049 - MIROC _ rcp45 2020 - 2049 - MIROC_ rcp85 2020 - 2049 - MRI - CGCM3_ rcp45 2020 - 2049 - MRI - CGCM3_ rcp85 2020 - 2049 - NorES M1 - M_rcp 45 2020 - 2049 - NorES M1 - M_rcp 85 Av. RF fj - 203 0s Arua 455 0 7 - 3 1 4 - 3 18 1 3 Arusha 683 8 9 - 4 5 - 4 0 4 4 3 Bukoba 1088 6 8 9 9 8 6 10 11 8 Dagorett i C orner 496 21 19 11 21 1 14 16 12 14 Dar es Salaam 636 6 8 8 10 9 17 - 13 - 3 5 Dodoma 290 - 2 10 - 9 - 9 10 7 3 - 6 0 Eldoret 475 - 4 1 - 2 - 3 - 10 - 5 - 2 1 - 3 Entebbe 670 24 26 26 27 24 22 30 26 26 Garissa 134 35 64 36 36 100 98 35 35 55 Gulu 561 8 6 1 4 10 1 20 7 7 Jinja 645 - 8 - 7 - 10 - 6 - 4 - 12 - 5 - 12 - 8 174 Locati on Basel ine - FMA MJ rainf all (mm) Change in rainfall (mm) 2020 - 2049 - CSIRO_ rcp45 2020 - 2049 - CSIRO_ rcp85 2020 - 2049 - MIROC_ rcp45 2020 - 2049 - MIROC_ rcp85 2020 - 2049 - MRI - CGCM3_ rcp45 2020 - 2049 - MRI - CGCM3_ rcp85 2020 - 2049 - NorES M1 - M_rcp 45 2020 - 2049 - NorES M1 - M_rcp 85 Av. RF fj - 203 0s Kabale 516 - 17 - 16 - 12 - 7 - 6 - 13 - 24 - 17 - 14 Kasese 390 - 10 - 2 - 4 - 2 - 3 - 8 - 7 - 10 - 6 Kigom a 471 2 1 - 1 9 - 9 10 - 4 - 7 0 Kisum u 724 - 5 - 4 - 1 3 - 13 - 13 - 4 - 1 - 5 Kiti gu m 471 6 9 5 13 12 4 26 10 11 Lamu 579 1 5 12 15 5 1 16 2 7 Lodwa r 106 49 32 - 3 28 27 57 45 35 34 Makin du 222 - 4 - 3 14 8 2 - 1 - 8 - 14 - 1 Mande ra 138 8 20 22 21 47 21 16 42 25 Marsab iti 280 64 71 51 41 64 60 58 46 57 Masak a 592 0 7 8 8 8 5 7 9 7 Mans in di 478 6 12 13 8 17 5 33 10 13 Mbarar a 376 - 9 - 4 - 6 - 1 0 - 5 - 4 - 5 - 4 Mbeya 519 - 9 - 10 - 7 - 14 - 15 - 6 - 10 - 11 - 10 Mtwar a 603 - 11 - 7 - 11 - 6 - 6 - 5 - 23 - 15 - 11 Musom a 515 10 - 3 10 13 0 4 7 10 6 Mwanz a 444 8 12 5 17 10 8 9 5 9 Namul onge 539 - 5 - 5 4 - 1 1 - 6 8 4 0 Narok 412 20 27 1 12 13 28 17 20 17 Same 319 12 4 - 6 - 2 2 - 2 - 1 - 4 1 Serere 597 8 11 6 3 9 0 14 9 7 Songea 570 25 15 13 - 2 6 14 15 24 14 Soroti 565 11 12 10 12 14 6 22 16 13 Tabora 455 - 19 - 13 - 11 - 3 - 21 - 4 - 8 - 13 - 11 Tororo 768 - 10 - 3 - 6 - 7 - 7 - 10 - 2 - 4 - 6 Voi 237 - 5 0 - 12 - 5 - 4 18 - 6 - 10 - 3 Wajir 164 27 59 21 8 103 31 7 14 34 175 Table 3. 17 Changes in the February - June (FMAMJ) rainfall for the 2050s Location Baseli ne - FMA MJ rainfa ll (mm) Change in rainfall (mm ) 2040 - 2069 - CSI RO_ rcp4 5 2040 - 2069 - CSIRO_ rcp85 2040 - 2069 - MIROC_ rcp45 2040 - 2069 - MIROC_ rcp85 2040 - 2069 - MRI - CGCM3_ rcp45 2040 - 2069 - MRI - CGCM3_ rcp85 2040 - 2069 - NorES M1 - M_rcp 45 2040 - 2069 - NorES M1 - M_rcp 85 Av. RF fj - 205 0s Arua 455 - 3 0 - 6 - 5 2 - 12 13 - 2 - 2 Arusha 683 10 13 5 9 2 6 11 7 8 Bukoba 1088 7 5 5 7 11 8 4 9 7 Dagoretti Corner 496 29 32 12 25 16 19 15 19 21 Dar es Salaam 636 8 10 9 15 8 16 - 5 - 2 7 Dodoma 290 20 6 7 3 21 20 5 8 11 Eldoret 475 - 14 - 8 - 17 - 11 - 18 - 10 - 10 - 10 - 12 Entebbe 670 22 23 26 25 2 7 26 27 30 26 Garissa 134 49 82 77 58 113 107 65 61 76 Gulu 561 - 5 - 3 - 8 - 7 0 - 6 15 - 3 - 2 Jinja 645 - 9 - 8 - 3 - 1 - 9 - 11 - 3 - 4 - 6 Kabale 516 - 14 - 11 - 10 - 8 - 9 - 12 - 16 - 14 - 12 Kasese 390 - 14 - 10 - 2 3 - 14 - 14 3 1 - 6 Kigoma 471 7 5 13 14 5 10 5 - 9 6 Kisu mu 724 - 3 - 1 2 2 - 6 - 7 4 3 - 1 Kitigum 471 - 4 - 1 - 10 - 7 - 5 - 3 7 - 8 - 4 Lamu 579 6 8 16 13 10 6 12 3 9 Lodwar 106 56 42 18 40 42 63 61 34 45 Makindu 222 23 22 30 26 18 18 5 19 20 Mandera 138 8 25 30 27 60 27 37 54 33 Marsabiti 280 62 64 52 61 71 76 54 4 7 61 Masaka 592 2 6 4 9 8 0 12 8 6 Mansindi 478 0 0 6 14 3 1 24 9 7 Mbarara 376 - 4 - 2 - 1 1 7 - 3 4 - 1 0 Mbeya 519 0 - 1 4 - 5 - 7 8 - 2 4 0 Mtwara 603 2 1 - 6 - 5 1 9 - 7 - 17 - 3 Musoma 515 3 12 18 14 - 1 10 5 8 9 Mwanza 444 14 18 20 24 17 14 14 17 17 176 Table Locati on Basel ine - FMA MJ rainf all (mm) Change in rainfall (mm) 2040 - 2069 - CSIRO_ rcp45 2040 - 2069 - CSIRO_ rcp85 2040 - 2069 - MIROC_ rcp45 2040 - 2069 - MIROC_ rcp85 2040 - 2069 - MRI - CGCM3_ rcp45 2040 - 2069 - MRI - CGCM3_ rcp85 2040 - 2069 - NorES M1 - M_rcp 45 2040 - 206 9 - NorES M1 - M_rcp 85 Av. RF fj - 205 0s Namul onge 539 - 5 - 9 5 2 5 0 14 5 2 Narok 412 17 19 3 22 - 1 12 6 8 11 Same 319 12 10 1 10 11 3 3 2 6 Serere 597 - 5 - 3 - 7 - 4 - 5 - 11 - 1 - 1 - 5 Songea 570 36 31 32 13 13 35 28 32 28 Soroti 565 1 4 - 2 - 3 - 2 1 7 2 1 Tabora 455 5 5 1 7 - 2 16 7 3 5 Tororo 768 - 13 - 8 - 2 - 4 - 8 - 14 - 6 - 9 - 8 Voi 237 4 17 2 13 15 18 10 21 12 Wajir 164 37 65 35 34 125 41 48 44 54 Table 3. 18 Changes in the February - June (FMAMJ) rainfall for the 2070s Location Basel ine - FMA MJ rainf all (mm) Change in rainfall (mm) 2060 - 2089 - CSIRO_ rcp45 2060 - 2089 - CSIRO_ rcp85 2060 - 2089 - MIROC_ rcp45 2060 - 2089 - MIROC_ rcp85 2060 - 2089 - MRI - CGCM3_ rcp45 2060 - 2089 - MRI - CGCM3_ rcp85 2060 - 2089 - NorES M1 - M_rcp 45 2060 - 2089 - NorES M1 - M_rcp 85 Av. RF fj - 207 0s Arua 455 0 7 - 5 4 - 1 - 9 19 0 2 Arusha 683 14 20 18 23 - 3 19 21 17 16 Bukoba 1088 7 8 8 15 12 14 6 6 10 Dagoretti Corner 496 38 41 20 29 19 32 31 34 30 Dar es Salaam 636 6 12 8 15 11 8 - 1 - 2 7 Dodoma 290 18 11 5 2 10 21 1 5 9 Eldoret 475 - 13 - 7 - 12 - 12 - 22 - 13 - 9 - 8 - 12 Entebbe 670 27 25 30 31 26 25 36 28 28 Garissa 134 42 57 86 72 121 73 72 36 70 Gulu 561 - 4 - 4 - 7 - 9 - 1 - 4 12 - 5 - 3 Jinja 645 - 8 - 5 1 5 - 1 - 14 7 3 - 2 Kabale 516 - 14 - 8 - 2 - 3 - 4 - 11 - 13 - 9 - 8 Kasese 390 - 14 - 14 1 8 - 16 - 19 6 2 - 6 Kigoma 471 - 1 2 18 10 - 3 4 5 - 11 3 Kisumu 724 2 5 10 7 - 3 - 3 6 8 4 177 Location Basel ine - FMA MJ rainf all (mm ) Change in rainfall (mm) 2060 - 2089 - CSIRO _rcp45 2060 - 2089 - CSIRO_ rcp85 2060 - 2089 - MIROC_ rcp45 2060 - 2089 - MIROC_ rcp85 2060 - 2089 - M RI - CGCM3_ rcp45 2060 - 2089 - MRI - CGCM3_ rcp85 2060 - 2089 - NorES M1 - M_rcp 45 2060 - 2089 - NorES M1 - M_rcp 85 Av. RF fj - 207 0s Kitigum 471 1 7 - 3 1 1 4 16 0 4 Lamu 579 - 1 4 9 8 0 7 0 2 4 Lodwar 106 60 45 33 33 32 61 71 35 46 Makindu 222 12 17 33 38 23 22 7 6 20 Mandera 138 3 27 23 28 63 28 30 45 31 Marsabiti 280 58 67 51 54 60 65 58 41 57 Masaka 592 8 9 10 13 8 5 12 6 9 Mansindi 478 4 1 8 19 7 4 28 12 10 Mbarara 376 - 3 1 8 9 2 - 1 3 0 3 Mbeya 519 6 - 1 7 - 1 - 7 8 0 - 1 1 Mtwara 603 - 1 - 2 - 6 - 8 2 10 - 5 - 16 - 3 Musoma 51 5 10 13 18 11 17 10 16 16 14 Mwanza 444 12 24 28 15 24 15 14 11 18 Namulonge 539 2 - 5 7 8 9 1 18 12 7 Narok 412 19 25 3 25 12 18 9 14 16 Same 319 2 11 1 11 4 4 - 3 7 5 Serere 597 - 5 0 - 1 2 2 - 9 4 1 - 1 Songea 570 40 29 34 16 10 33 25 31 27 Soroti 565 3 8 5 0 5 4 10 7 5 Tabora 455 2 1 4 9 - 4 12 6 - 3 3 Tororo 768 - 5 - 4 - 1 0 - 6 - 6 - 1 - 3 - 3 Voi 237 1 2 7 10 18 16 2 15 9 Wajir 164 44 61 30 27 115 32 41 46 49 178 Table 3. 19 Changes in the August - December (ASOND) rainfall for the 2030s Location Basel ine - ASO ND - rainf all (mm ) Change in rainfall (mm) 2020 - 2049 - CSIR O_rcp 45 2020 - 2049 - CSIRO_ rcp85 2020 - 2049 - MIROC_ rcp45 2020 - 2049 - MIROC_ rcp85 2020 - 2049 - MRI - CGCM3_ rcp45 2020 - 2049 - MRI - CGCM3_ rcp85 2020 - 2049 - NorES M1 - M_rcp 45 2020 - 2049 - N orES M1 - M_rcp 85 Av. RF ad - 203 0s Arua 661 12 10 13 14 4 8 25 20 13 Arusha 271 - 3 9 9 5 19 14 12 3 9 Bukoba 777 - 17 - 17 - 16 - 10 - 17 - 11 - 2 - 3 - 12 Dagoretti Corner 311 15 12 7 3 21 12 10 10 11 Dar es Salaam 385 - 21 - 16 - 8 - 3 - 15 - 25 - 13 - 21 - 15 Dodoma 173 - 1 2 26 26 31 1 8 42 17 Eldoret 419 - 4 - 5 - 7 3 - 8 - 16 5 10 - 3 Entebbe 563 - 6 - 6 0 5 - 10 - 6 22 22 3 Garissa 187 15 19 41 34 82 24 23 26 33 Gulu 650 5 7 16 16 5 11 34 30 16 Jinja 583 - 8 - 6 6 15 - 17 - 9 7 20 1 Kabale 618 - 39 - 33 - 28 - 19 - 37 - 33 - 10 - 22 - 28 Kasese 521 - 17 - 14 - 18 - 9 - 31 - 23 7 - 4 - 14 Kigoma 426 - 7 - 8 - 11 - 6 - 10 - 13 1 4 - 6 Kisumu 588 - 17 - 21 - 15 - 8 - 36 - 35 - 13 - 8 - 19 Kitigum 483 2 10 20 16 3 7 22 32 14 Lamu 188 10 15 23 22 39 1 10 - 25 12 Lodwar 68 - 17 - 12 - 19 - 6 - 31 103 64 198 35 Mak indu 369 - 2 - 3 9 2 0 - 14 - 1 - 12 - 3 Mandera 109 - 13 - 6 6 25 34 - 12 29 70 17 Marsabiti 285 27 25 39 45 19 33 37 52 35 Masaka 527 - 21 - 16 - 17 - 14 - 23 - 19 2 4 - 13 Mansindi 615 - 5 - 6 3 4 - 16 - 4 21 26 3 Mbarara 549 - 22 - 20 - 17 - 17 - 27 - 22 5 - 6 - 16 Mbeya 29 6 14 1 17 11 - 4 - 8 27 19 10 Mtwara 226 5 15 15 3 50 5 8 4 13 Musoma 366 - 18 - 18 - 14 4 - 20 - 17 3 7 - 9 Mwanza 440 2 3 3 10 - 5 - 6 20 26 7 Namulong e 601 - 14 - 15 - 2 5 - 14 - 8 14 18 - 2 179 Location Basel ine - ASO ND - rainf all (mm ) Change in ra infall (mm) 2020 - 2049 - CSIR O_rcp 45 2020 - 2049 - CSIRO_ rcp85 2020 - 2049 - MIROC_ rcp45 2020 - 2049 - MIROC_ rcp85 2020 - 2049 - MRI - CGCM3_ rcp45 2020 - 2049 - MRI - CGCM3_ rcp85 2020 - 2049 - NorES M1 - M_rcp 45 2020 - 2049 - NorES M1 - M_rcp 85 Av. RF ad - 203 0s Namulong e 601 - 14 - 15 - 2 5 - 14 - 8 14 18 - 2 Narok 227 9 4 2 28 11 16 20 13 13 Same 175 40 36 50 67 79 59 46 31 51 Serere 573 - 12 - 10 3 7 - 12 - 16 7 15 - 2 Songea 242 43 54 35 33 53 36 42 41 42 Soroti 548 - 2 0 14 7 - 14 0 5 15 3 Tabora 338 3 5 10 18 18 13 20 28 14 Tororo 682 - 24 - 19 - 13 - 7 - 34 - 34 - 13 - 3 - 18 Voi 332 - 8 - 6 12 7 18 3 13 - 17 3 Wajir 128 15 14 52 31 79 61 52 51 44 Table 3. 20 Changes in the August - December (ASOND) rainfall for the 2050s Location Basel ine - ASO ND - rainf all (mm ) Change in rainfal l (mm) 2040 - 2069 - CSIR O_rcp 45 2040 - 2069 - CSIRO_ rcp85 2040 - 2069 - MIROC_ rcp45 2040 - 2069 - MIROC_ rcp85 2040 - 2069 - MRI - CGCM3_ rcp45 2040 - 2069 - MRI - CGCM3_ rcp85 2040 - 2069 - NorES M1 - M_rcp 45 2040 - 2069 - NorES M1 - M_rcp 85 Av. RF ad - 205 0s Arua 661 14 12 15 17 9 8 31 19 16 Aru sha 271 1 0 - 7 - 3 6 7 3 - 7 0 Bukoba 777 - 16 - 13 - 14 - 12 - 19 - 17 - 4 - 8 - 13 Dagoretti Corner 311 8 - 1 8 2 6 12 8 1 5 Dar es Salaam 385 - 23 - 17 - 14 - 13 - 15 - 28 - 17 - 28 - 19 Dodoma 173 - 6 - 6 5 11 17 - 2 - 2 40 7 Eldoret 419 2 - 1 0 5 1 - 5 13 18 4 Entebbe 563 - 11 - 4 4 6 - 16 - 3 21 18 2 Garissa 187 - 1 9 27 11 65 25 7 12 19 Gulu 650 6 8 18 16 10 12 27 28 16 180 Location Basel ine - ASO ND - rainf all (mm ) Change in rainfall (mm) 2040 - 2069 - CSIR O_rcp 45 2040 - 2069 - CSIRO_ rcp85 2040 - 2069 - MIROC_ rcp45 2040 - 2069 - MIROC_ rcp85 2040 - 2069 - MRI - CGCM3_ rcp45 2040 - 2069 - MRI - CGCM3_ rcp85 2040 - 2069 - NorES M1 - M_rcp 45 2040 - 2069 - NorES M1 - M_rcp 85 Av. RF ad - 205 0s Jinja 583 - 5 - 2 - 3 2 - 16 - 10 1 8 - 3 Kabale 618 - 39 - 39 - 33 - 26 - 36 - 36 - 17 - 25 - 31 Kasese 521 - 18 - 18 - 12 - 12 - 2 6 - 26 15 - 2 - 13 Kigoma 426 - 24 - 19 - 22 - 22 - 14 - 22 - 9 - 14 - 18 Kisumu 588 - 16 - 22 - 16 - 9 - 37 - 39 - 15 - 7 - 20 Kitigum 483 8 13 20 14 7 11 29 31 17 Lamu 188 30 36 60 47 71 20 39 - 1 38 Lodwar 68 2 1 - 19 11 - 15 140 88 138 43 Makindu 369 - 7 - 7 9 - 6 - 7 - 23 - 8 - 17 - 8 Mandera 109 - 15 - 5 5 10 26 - 2 29 62 14 Marsabiti 285 17 16 21 24 13 21 22 21 19 Masaka 527 - 16 - 16 - 16 - 12 - 26 - 16 2 6 - 12 Mansindi 615 1 - 1 9 10 - 7 2 23 24 8 Mbarara 549 - 24 - 25 - 28 - 25 - 32 - 30 - 2 - 11 - 22 Mbeya 296 8 - 7 11 12 - 3 - 7 9 7 4 M twara 226 12 17 20 4 32 5 9 - 4 12 Musoma 366 - 23 - 21 - 15 - 1 - 23 - 21 - 6 9 - 12 Mwanza 440 - 13 - 4 - 2 - 2 - 11 - 13 - 1 17 - 4 Namulong e 601 - 9 - 7 - 10 - 1 - 14 - 15 3 11 - 5 Narok 227 - 1 2 - 16 16 4 11 15 4 4 Same 175 40 20 32 53 68 42 30 18 38 Serere 573 - 5 0 8 1 1 - 12 - 9 12 19 3 Songea 242 23 42 27 27 44 33 42 26 33 Soroti 548 - 3 4 14 16 - 11 0 14 18 6 Tabora 338 - 4 1 - 8 12 6 5 13 26 6 Tororo 682 - 16 - 14 - 10 - 4 - 34 - 31 - 4 4 - 14 Voi 332 - 18 - 15 - 8 - 3 3 - 2 10 - 13 - 6 Wajir 128 17 9 28 16 95 61 25 36 36 181 Table 3 . 21 Changes in the August - December (ASOND) rainfall for the 2070s Location Basel ine - ASO ND - rainf all (mm ) Change in rainfall (mm) 2060 - 2089 - CSIR O_rcp 45 2060 - 2089 - CSIRO_ rcp85 2060 - 2089 - MIROC_ rcp45 2060 - 2089 - MIROC_ rcp85 2060 - 2 089 - MRI - CGCM3_ rcp45 2060 - 2089 - MRI - CGCM3_ rcp85 2060 - 2089 - NorES M1 - M_rcp 45 2060 - 2089 - NorES M1 - M_rcp 85 Av. RF ad - 207 0s Arua 661 12 16 23 23 7 11 32 24 18 Arusha 271 - 2 2 - 1 4 1 2 - 5 - 5 - 1 Bukoba 777 - 16 - 12 - 15 - 13 - 18 - 17 - 1 - 4 - 12 Dagoretti Corner 311 5 7 8 1 - 1 5 14 8 6 Dar es Salaam 385 - 22 - 18 - 11 - 12 - 16 - 31 - 17 - 27 - 19 Dodoma 173 - 8 - 2 2 6 6 - 7 - 3 42 4 Eldoret 419 11 9 7 15 7 - 12 23 34 12 Entebbe 563 0 3 10 18 - 8 5 22 22 9 Garissa 187 0 9 21 30 67 22 13 17 22 Gulu 650 12 18 21 17 12 15 32 28 19 Jinja 583 2 2 - 5 6 - 19 - 14 4 5 - 2 Kabale 618 - 35 - 32 - 32 - 24 - 31 - 33 - 13 - 21 - 28 Kasese 521 - 19 - 13 - 11 - 8 - 28 - 27 19 2 - 10 Kigoma 426 - 24 - 20 - 26 - 21 - 18 - 20 - 17 - 20 - 21 Kisumu 588 - 17 - 13 - 16 - 7 - 35 - 31 - 10 - 2 - 16 Kitigum 483 11 22 23 22 14 25 37 32 23 Lamu 188 21 24 63 47 78 30 45 - 7 38 Lodwar 68 10 15 - 15 19 - 5 138 99 115 47 Makindu 369 - 15 - 16 0 - 9 - 12 - 14 - 2 - 24 - 11 Mandera 109 - 12 - 6 9 19 34 - 1 43 51 17 Marsabiti 285 16 16 15 17 11 13 18 15 15 Masaka 527 - 15 - 17 - 17 - 9 - 25 - 15 10 5 - 10 Ma nsindi 615 4 8 17 13 4 5 34 28 14 Mbarara 549 - 25 - 24 - 25 - 23 - 28 - 27 1 - 11 - 20 Mbeya 296 1 - 5 1 - 7 - 5 - 17 1 5 - 3 Mtwara 226 3 14 7 3 29 4 9 3 9 Musoma 366 - 20 - 23 - 20 0 - 26 - 23 - 3 6 - 14 Mwanza 440 - 11 - 6 - 4 5 - 12 - 8 4 13 - 2 Namulong e 601 - 1 - 3 - 8 4 - 13 - 13 5 16 - 2 182 Location Basel ine - ASO ND - rainf all (mm ) Change in rainfall (mm) 2060 - 2089 - CSIR O_rcp 45 2060 - 2089 - CSIRO_ rcp85 2060 - 2089 - MIROC_ rcp45 2060 - 2089 - MIROC_ rcp85 2060 - 2089 - MRI - CGCM3_ rcp45 2060 - 2089 - MRI - CGCM3_ rcp85 2060 - 2089 - N orES M1 - M_rcp 45 2060 - 2089 - NorES M1 - M_rcp 85 Av. RF ad - 207 0s Narok 227 - 2 - 2 - 13 11 - 1 7 12 4 2 Same 175 29 15 18 37 71 29 33 9 30 Serere 573 4 8 19 19 - 11 4 23 34 13 Songea 242 29 30 12 22 32 27 33 34 27 Soroti 548 2 6 21 28 - 3 6 24 40 15 Tabora 338 - 10 - 14 - 12 3 5 - 10 1 11 - 3 Tororo 682 - 10 - 7 1 6 - 30 - 27 0 6 - 8 Voi 332 - 18 - 14 4 0 - 4 - 1 - 11 - 27 - 9 Wajir 128 12 12 20 16 95 56 27 35 34 183 APPENDIX D: Projected mean temperature Table 3. 22 Annual mean temperatures for the 20 30s for the 4 GCMs under two representative concentration pathways, RCP4.5 and RCP 8.5 Location Basel ine annu al Tav (C) Mean temperature change ( 0 C) 2020 - 2049 - CSIRO_r cp45 2020 - 2049 - CSIRO_r cp85 2020 - 2049 - MIROC_ rcp45 2020 - 2049 - MIROC_ rcp85 2020 - 2049 - MRI - CG CM3_ rcp45 2020 - 2049 - MRI - CGCM3_ rcp85 2020 - 2049 - NorES M1 - M_rcp 45 2020 - 2049 - NorES M1 - M_rcp 85 Arua 24.4 - 0.9 - 1.0 - 1.0 - 1.0 - 0.8 - 1.1 - 1.4 - 1.2 Arusha 14.0 5.7 5.7 5.8 5.8 5.6 5.7 5.8 5.8 Bukoba 22.2 - 0.5 - 0.6 - 0.7 - 0.7 - 0.8 - 0.8 - 0.8 - 0.8 Dagoretti Corner 2 0.5 - 2.3 - 2.3 - 2.3 - 2.2 - 2.4 - 2.4 - 2.2 - 2.2 Dar es Salaam 26.5 - 0.1 - 0.1 - 0.1 - 0.1 - 0.2 - 0.2 - 0.1 - 0.1 Dodoma 23.4 - 0.4 - 0.4 - 0.3 - 0.2 - 0.5 - 0.6 - 0.5 - 0.5 Eldoret 18.4 - 1.0 - 1.0 - 1.0 - 1.0 - 0.9 - 0.9 - 1.1 - 1.1 Entebbe 22.8 - 0.6 - 0.7 - 0.8 - 0.8 - 0.7 - 0.7 - 1.1 - 0.9 Garissa 28.9 0.7 0.7 0.6 0.7 0.6 0.6 0.7 0.9 Gulu 24.5 - 0.7 - 0.8 - 0.9 - 0.9 - 0.8 - 0.9 - 1.5 - 1.1 Jinja 21.9 0.8 0.8 0.7 0.7 0.9 0.9 0.4 0.6 Kabale 18.2 - 0.4 - 0.5 - 0.5 - 0.5 - 0.6 - 0.6 - 0.6 - 0.7 Kasese 24.4 - 0.7 - 0.8 - 0.7 - 0.7 - 0.1 - 0.1 - 1.0 - 0.9 Kigoma 24.0 0.5 0.4 0.4 0.4 0.4 0.4 0.3 0.2 Kisumu 23.6 - 0.2 - 0.3 - 0.3 - 0.2 0.4 0.4 - 0.3 - 0.3 Kitigum 25.4 - 0.4 - 0.5 - 0.5 - 0.5 - 0.5 - 0.7 - 1.0 - 0.7 Lamu 27.8 - 0.5 - 0.4 - 0.4 - 0.4 - 0.5 - 0.5 - 0.4 - 0.4 Lodwar 29.8 0.0 - 0.1 0.1 0.1 - 0.2 - 0.3 - 0.2 - 0.1 Maki ndu 24.2 - 1.2 - 1.2 - 1.2 - 1.2 - 1.3 - 1.2 - 1.1 - 1.0 Mandera 29.7 - 0.1 - 0.1 - 0.2 - 0.2 - 0.3 - 0.2 - 0.2 - 0.2 Marsabiti 24.4 - 4.0 - 4.1 - 4.1 - 4.0 - 4.2 - 4.3 - 4.1 - 4.0 Masaka 21.7 0.0 - 0.1 - 0.2 - 0.2 0.0 0.0 - 0.4 - 0.3 Mansindi 24.2 - 0.7 - 0.8 - 0.9 - 0.8 - 0.6 - 0.7 - 1 .5 - 1.1 Mbarara 21.2 - 0.4 - 0.5 - 0.5 - 0.5 - 0.3 - 0.3 - 0.7 - 0.7 Mbeya 20.0 - 1.5 - 1.6 - 1.5 - 1.4 - 1.8 - 1.8 - 1.7 - 1.7 Mtwara 26.1 0.7 0.7 0.8 0.8 0.6 0.6 0.8 0.8 Musoma 22.9 0.8 0.7 0.7 0.7 0.7 0.7 0.6 0.6 Mwanza 23.3 0.0 - 0.1 - 0.1 - 0.1 - 0.2 - 0.2 - 0.3 - 0.2 Namulong e 22.8 - 0.4 - 0.5 - 0.6 - 0.6 - 0.4 - 0.5 - 1.0 - 0.8 Narok 17.7 - 0.4 - 0.4 - 0.3 - 0.3 - 0.5 - 0.6 - 0.4 - 0.3 184 Location Basel ine annu al Tav (C) Mean temperature change ( 0 C) 2020 - 2049 - CSIRO_r cp45 2020 - 2049 - CSIRO_r cp85 2020 - 2049 - MIROC_ rc p45 2020 - 2049 - MIROC_ rcp85 2020 - 2049 - MRI - CGCM 3_rcp45 2020 - 2049 - MRI - CGCM3_r cp85 2020 - 2049 - NorES M1 - M_rcp 45 2020 - 2049 - NorES M1 - M_rcp8 5 Same 24.5 - 1.3 - 1.3 - 1.2 - 1.2 - 1.4 - 1.3 - 1.2 - 1.2 Serere 24.9 - 0.5 - 0.6 - 0.6 - 0.6 - 0.5 - 0.5 - 0.9 - 0.8 Songea 22.3 - 0.8 - 0.9 - 0.8 - 0.6 - 1.0 - 1.0 - 0.9 - 0.9 Soroti 25.3 - 0.6 - 0.7 - 0.8 - 0.8 - 0.7 - 0.7 - 1.1 - 0.9 Tabora 23.8 - 0.1 - 0.1 - 0.1 - 0.1 - 0.2 - 0.3 - 0.3 - 0.2 Tororo 22.9 0.1 0.1 0.1 0.1 0.8 0.8 - 0.1 - 0.1 Voi 26.2 - 0.9 - 0.9 - 0.9 - 0.9 - 1.0 - 0.9 - 0.8 - 0.8 Wajir 28.9 - 0.7 - 0.7 - 0.7 - 0.7 - 0.9 - 0.9 - 0.8 - 0.6 Table 3. 23 Annual mean temperatures for 2050s for the 4 GCMs under two representative concentration pathways, RCP4.5 and RCP 8.5 Location Basel ine annu al Tav ( 0 C) Mean temperature change ( 0 C) 2 040 - 2069 - CSIRO_r cp45 2040 - 2069 - CSIRO_r cp85 2040 - 2069 - MIROC_ rcp45 2040 - 2069 - MIROC_ rcp85 2040 - 2069 - MRI - CGCM3_ rcp45 2040 - 2069 - MRI - CGCM3_ rcp85 2040 - 2069 - NorES M1 - M_rcp 45 2040 - 2069 - NorES M1 - M_rcp 85 Arua 24.4 - 0.8 - 0.9 - 0.9 - 0.9 - 0.7 - 1.0 - 1.3 - 1.1 Arusha 14.0 5 .8 5.8 5.9 5.9 5.7 5.7 5.9 5.9 Bukoba 22.2 - 0.4 - 0.5 - 0.6 - 0.6 - 0.7 - 0.7 - 0.7 - 0.7 Dagoretti Corner 20.5 - 2.3 - 2.2 - 2.2 - 2.2 - 2.3 - 2.4 - 2.1 - 2.1 Dar es Salaam 26.5 - 0.1 - 0.1 0.0 0.0 - 0.1 - 0.1 0.0 0.0 Dodoma 23.4 - 0.3 - 0.3 - 0.2 - 0.1 - 0.4 - 0.5 - 0.4 - 0.4 Eldoret 18.4 - 0.9 - 0.9 - 0.9 - 0.9 - 0.8 - 0.8 - 1.0 - 1.0 Entebbe 22.8 - 0.5 - 0.6 - 0.7 - 0.7 - 0.6 - 0.6 - 1.0 - 0.9 Garissa 28.9 0.7 0.8 0.7 0.7 0.7 0.6 0.8 0.9 Gulu 24.5 - 0.6 - 0.7 - 0.8 - 0.8 - 0.7 - 0.8 - 1.4 - 1.0 Jinja 21.9 0.9 0.9 0.8 0.8 1.0 1.0 0.5 0.6 Kabale 18.2 - 0.3 - 0.4 - 0.4 - 0.4 - 0.5 - 0.5 - 0.5 - 0.6 Kasese 24.4 - 0.6 - 0.7 - 0.6 - 0.6 0.0 0.0 - 0.9 - 0.8 Kigoma 24.0 0.6 0.5 0.5 0.5 0.5 0.5 0.4 0.3 Kisumu 23.6 - 0.2 - 0.2 - 0.2 - 0.1 0.5 0.5 - 0.3 - 0.3 185 Location Basel ine annu al Tav ( 0 C) Mean te mperature change ( 0 C) 2040 - 2069 - CSIR O_rcp 45 2040 - 2069 - CSIRO_r cp85 2040 - 2069 - MIROC_r cp45 2040 - 2069 - MIROC_r cp85 2040 - 2069 - MRI - CGCM3_r cp45 2040 - 2069 - MRI - CGCM3_r cp85 2040 - 2069 - NorES M1 - M_rcp 45 2040 - 2069 - NorES M1 - M_rcp 85 Kitigum 25.4 - 0.3 - 0.4 - 0.5 - 0.4 - 0.4 - 0.6 - 0.9 - 0.6 Lamu 27.8 - 0.4 - 0.4 - 0.4 - 0.4 - 0.5 - 0.5 - 0.3 - 0.3 Lodwar 29.8 0.1 0.0 0.1 0.1 - 0.2 - 0.2 - 0.2 - 0.1 Makindu 24.2 - 1.1 - 1.1 - 1.2 - 1.1 - 1.2 - 1.1 - 1.0 - 1.0 Mandera 29.7 0.0 0.0 - 0.1 - 0.1 - 0.2 - 0.1 - 0.2 - 0.1 Marsabiti 24.4 - 4.0 - 4.0 - 4.0 - 3.9 - 4.2 - 4.2 - 4.1 - 3.9 Masaka 21.7 0.1 0.0 - 0.1 - 0.1 0.1 0.1 - 0.3 - 0.2 Mansindi 24.2 - 0.6 - 0.7 - 0.8 - 0.8 - 0.5 - 0.6 - 1.4 - 1.0 Mbarara 21.2 - 0.3 - 0.4 - 0.4 - 0.4 - 0.3 - 0.2 - 0.6 - 0.6 Mbeya 20.0 - 1.5 - 1.5 - 1.5 - 1.3 - 1.7 - 1.7 - 1.6 - 1.6 Mtwara 26.1 0.7 0.7 0.8 0.8 0.7 0.7 0.8 0.8 Musoma 22.9 0.8 0.8 0.8 0.8 0.8 0.8 0.7 0.7 Mwanza 23.3 0.0 0.0 0.0 0.0 - 0.1 - 0.2 - 0.2 - 0.1 Namulonge 22.8 - 0.3 - 0.4 - 0.5 - 0.5 - 0.4 - 0.4 - 0.9 - 0.7 Narok 17.7 - 0.3 - 0.3 - 0.3 - 0.2 - 0.4 - 0.5 - 0.3 - 0.2 Same 24.5 - 1.2 - 1.2 - 1.2 - 1.2 - 1. 3 - 1.2 - 1.1 - 1.1 Serere 24.9 - 0.4 - 0.5 - 0.5 - 0.5 - 0.4 - 0.4 - 0.9 - 0.7 Songea 22.3 - 0.8 - 0.8 - 0.7 - 0.6 - 0.9 - 1.0 - 0.8 - 0.9 Soroti 25.3 - 0.6 - 0.6 - 0.7 - 0.7 - 0.6 - 0.6 - 1.0 - 0.8 Tabora 23.8 0.0 0.0 0.0 0.0 - 0.1 - 0.2 - 0.2 - 0.1 Tororo 22.9 0.2 0.2 0.2 0.2 0. 9 0.9 0.0 0.0 Voi 26.2 - 0.8 - 0.8 - 0.8 - 0.8 - 0.9 - 0.8 - 0.7 - 0.7 Wajir 28.9 - 0.6 - 0.6 - 0.7 - 0.6 - 0.8 - 0.8 - 0.7 - 0.6 186 Table 3. 24 Annual mean temperatures for 2070s for the 4 GCMs under two representative concentration path ways, RCP4.5 and RCP 8.5 Location Basel ine annu al Tav ( 0 C) Mean temperature change ( 0 C) 2060 - 2089 - CSIRO_r cp45 2060 - 2089 - CSIRO_r cp85 2060 - 2089 - MIROC_ rcp45 2060 - 2089 - MIROC_ rcp85 2060 - 2089 - MRI - CGCM3_ rcp45 2060 - 2089 - MRI - CGCM3_ rcp85 2060 - 2089 - NorES M1 - M_rcp 45 2060 - 2089 - NorES M1 - M_rcp 85 Arua 24.4 - 0.8 - 0.9 - 1.0 - 0.9 - 0.7 - 1.0 - 1.3 - 1.1 Arusha 14.0 5.8 5.8 5.9 5.9 5.7 5.7 5.9 5.9 Bukoba 22.2 - 0.5 - 0.5 - 0.6 - 0.6 - 0.7 - 0.7 - 0.7 - 0.7 Dagoretti Corner 20.5 - 2.3 - 2.2 - 2.2 - 2.2 - 2.3 - 2.4 - 2.2 - 2.1 Dar es Salaam 26 .5 - 0.1 - 0.1 0.0 0.0 - 0.2 - 0.1 0.0 0.0 Dodoma 23.4 - 0.3 - 0.3 - 0.2 - 0.1 - 0.4 - 0.5 - 0.4 - 0.4 Eldoret 18.4 - 0.9 - 1.0 - 0.9 - 0.9 - 0.8 - 0.9 - 1.0 - 1.0 Entebbe 22.8 - 0.6 - 0.6 - 0.7 - 0.7 - 0.6 - 0.6 - 1.0 - 0.9 Garissa 28.9 0.8 0.8 0.6 0.7 0.6 0.7 0.7 0.9 Gulu 24.5 - 0.6 - 0.7 - 0.8 - 0.8 - 0.7 - 0.9 - 1.4 - 1.0 Jinja 21.9 0.9 0.8 0.8 0.8 1.0 1.0 0.5 0.6 Kabale 18.2 - 0.3 - 0.4 - 0.4 - 0.4 - 0.5 - 0.5 - 0.5 - 0.6 Kasese 24.4 - 0.6 - 0.7 - 0.7 - 0.7 0.0 0.0 - 0.9 - 0.8 Kigoma 24.0 0.6 0.5 0.5 0.5 0.5 0.5 0.4 0.3 Kisumu 23.6 - 0.2 - 0.2 - 0.2 - 0.2 0.5 0.5 - 0.3 - 0.3 Kitigum 25.4 - 0.3 - 0.4 - 0.5 - 0.5 - 0.4 - 0.6 - 0.9 - 0.6 Lamu 27.8 - 0.4 - 0.4 - 0.4 - 0.4 - 0.5 - 0.5 - 0.3 - 0.3 Lodwar 29.8 0.0 0.0 0.1 0.1 - 0.2 - 0.2 - 0.2 - 0.1 Makindu 24.2 - 1.1 - 1.1 - 1.2 - 1.1 - 1.2 - 1.1 - 1.0 - 1.0 Mandera 29.7 0.0 0 .0 - 0.1 - 0.1 - 0.2 - 0.2 - 0.2 - 0.1 Marsabiti 24.4 - 4.0 - 4.0 - 4.0 - 3.9 - 4.2 - 4.2 - 4.1 - 3.9 Masaka 21.7 0.1 0.0 - 0.1 - 0.1 0.1 0.1 - 0.3 - 0.2 Mansindi 24.2 - 0.6 - 0.7 - 0.8 - 0.8 - 0.5 - 0.6 - 1.4 - 1.0 Mbarara 21.2 - 0.3 - 0.4 - 0.4 - 0.4 - 0.3 - 0.3 - 0.6 - 0.6 Mbeya 20 .0 - 1.5 - 1.5 - 1.5 - 1.3 - 1.7 - 1.7 - 1.7 - 1.6 Mtwara 26.1 0.7 0.7 0.8 0.8 0.6 0.7 0.8 0.8 Musoma 22.9 0.8 0.8 0.8 0.8 0.8 0.8 0.7 0.7 Mwanza 23.3 0.0 0.0 0.0 0.0 - 0.1 - 0.2 - 0.2 - 0.2 Namulong e 22.8 - 0.3 - 0.4 - 0.5 - 0.5 - 0.4 - 0.4 - 0.9 - 0.7 Narok 17.7 - 0.3 - 0.3 - 0.3 - 0.2 - 0.4 - 0.5 - 0.3 - 0.2 Same 24.5 - 1.2 - 1.2 - 1.2 - 1.2 - 1.3 - 1.2 - 1.1 - 1.1 Serere 24.9 - 0.4 - 0.5 - 0.6 - 0.6 - 0.4 - 0.5 - 0.9 - 0.7 187 Location Basel ine annu al Tav ( 0 C) Mean temperature change ( 0 C) 2060 - 2089 - CSIRO_r cp45 2060 - 2089 - CSIRO_ rcp85 2060 - 2089 - MIRO C_rcp4 5 2060 - 2089 - MIROC_r cp85 2060 - 2089 - MRI - CGCM3_r cp45 2060 - 2089 - MRI - CGCM3_r cp85 2060 - 2089 - NorES M1 - M_rcp 45 2060 - 2089 - NorES M1 - M_rcp8 5 Songea 22.3 - 0.8 - 0.8 - 0.7 - 0.6 - 0.9 - 1.0 - 0.8 - 0.9 Soroti 25.3 - 0.6 - 0.7 - 0.7 - 0.7 - 0.6 - 0.7 - 1.0 - 0.9 Tabora 23.8 0.0 0.0 0.0 0.0 - 0.1 - 0.2 - 0.2 - 0.2 Tororo 22.9 0.2 0.1 0.1 0.1 0.8 0.9 0.0 0.0 Voi 26.2 - 0.8 - 0.8 - 0.8 - 0.8 - 1.0 - 0.8 - 0.7 - 0.7 Wajir 28.9 - 0.6 - 0.6 - 0.7 - 0.6 - 0.8 - 0.8 - 0.7 - 0.6 Table 3. 25 Mean tem peratures for February - June in 2030s for the 4 GCMs under two representative concentration pathways, RCP4.5 and RCP 8.5 Location Basel ine Tav FMA MJ ( 0 C ) Mean temperature change ( 0 C) 2020 - 2049 - CSIRO_r cp45 2020 - 2049 - CSIRO_r cp85 2020 - 2049 - MIROC_ rcp45 2020 - 2049 - MIROC_ rcp85 2020 - 2049 - MRI - CGCM3_ rcp45 2020 - 2049 - MRI - CGCM3_ rcp85 2020 - 2049 - NorES M1 - M_rcp 45 2020 - 2049 - NorES M1 - M_rcp 85 Arua 24.9 - 0.8 - 0.9 - 1.0 - 1.0 - 1.0 - 1.0 - 1.5 - 1.2 Arusha 14.2 5.8 5.8 5.9 5.9 5.7 5.8 5.9 5.9 Bukoba 22.4 - 0.6 - 0.7 - 0.8 - 0.8 - 0.9 - 0.9 - 0.8 - 0.9 Dagoretti Corner 20.9 - 2.3 - 2.2 - 2.2 - 2.2 - 2.3 - 2.4 - 2.2 - 2.2 Dar es Salaam 26.6 0.4 0.4 0.4 0.4 0.3 0.3 0.4 0.3 Dodoma 23.3 - 0.4 - 0.4 - 0.3 - 0.2 - 0.5 - 0.6 - 0.6 - 0.6 Eldoret 18.9 - 1.2 - 1.2 - 1.2 - 1.2 - 1.0 - 1.1 - 1.3 - 1.3 Entebbe 23.1 - 0.7 - 0.8 - 0.9 - 0.8 - 0.9 - 0.9 - 1.2 - 1.0 Garissa 29.7 0.8 0.7 0.7 0.7 0.6 0.4 0.8 0.8 Gulu 25.0 - 0.8 - 0.9 - 1.0 - 1.0 - 1.1 - 1.0 - 1.6 - 1.2 Jinja 22.2 0.8 0.7 0.6 0.7 0.8 0.8 0.2 0.5 Kabale 18.3 - 0.5 - 0.6 - 0.7 - 0.6 - 0.7 - 0.6 - 0.6 - 0.7 Kasese 24.5 - 0.5 - 0.7 - 0.6 - 0.6 - 0.2 0.0 - 0.7 - 0.7 Kigoma 23.9 0.5 0.4 0.3 0.3 0.3 0.3 0.3 0.2 Kisumu 23.9 - 0.5 - 0.5 - 0.5 - 0.5 0.4 0.3 - 0.6 - 0.6 Kitigum 26.1 - 0.4 - 0.5 - 0.6 - 0.6 - 0.6 - 0.7 - 1.1 - 0.7 Lamu 28.3 - 0.6 - 0.6 - 0.7 - 0.6 - 0.7 - 0.7 - 0.6 - 0.7 Lodwar 30.2 0.0 - 0.1 0.0 0.0 - 0.3 - 0.4 - 0.2 0.0 188 Location Baseli ne Tav FMA MJ ( 0 C ) Mean temperature change ( 0 C) 2020 - 2049 - CSIRO_ rcp45 2020 - 2049 - CSIR O_rcp 85 2020 - 2049 - MIROC_r cp45 2020 - 2049 - MIROC_r cp85 2020 - 2049 - MRI - CGCM3_r cp45 2020 - 2049 - MRI - CGCM3_r cp85 2020 - 2049 - No rES M1 - M_rcp 45 2020 - 2049 - NorES M1 - M_rcp 85 Makindu 24.9 - 1.0 - 1.0 - 1.1 - 1.1 - 1.2 - 1.1 - 0.9 - 1.0 Mandera 30.5 - 0.1 - 0.2 - 0.2 - 0.2 - 0.4 - 0.4 - 0.3 - 0.3 Marsabiti 25.0 - 3.8 - 3.9 - 3.9 - 3.8 - 4.1 - 4.2 - 4.0 - 3.9 Masaka 21.8 0.0 - 0.1 - 0.3 - 0.3 - 0.1 - 0.1 - 0.5 - 0.4 Mansindi 24.5 - 0.7 - 0.8 - 0.8 - 0.8 - 0.9 - 0.7 - 1.6 - 1.0 Mbarara 21.4 - 0.3 - 0.4 - 0.5 - 0.4 - 0.3 - 0.2 - 0.6 - 0.5 Mbeya 19.7 - 1.7 - 1.7 - 1.7 - 1.6 - 1.9 - 2.0 - 2.0 - 1.8 Mtwara 25.7 1.4 1.4 1.5 1.5 1.3 1.3 1.4 1.4 Musoma 22.7 0.9 0.9 0.8 0.9 1.0 0.8 0.7 0.7 Mwan za 23.2 0.1 0.1 0.0 0.0 - 0.1 - 0.2 - 0.2 - 0.2 Namulonge 23.1 - 0.6 - 0.7 - 0.6 - 0.5 - 0.7 - 0.5 - 1.0 - 0.8 Narok 18.2 - 0.6 - 0.5 - 0.5 - 0.5 - 0.6 - 0.8 - 0.5 - 0.6 Same 24.8 - 1.1 - 1.1 - 1.1 - 1.1 - 1.4 - 1.2 - 1.1 - 1.1 Serere 25.4 - 0.8 - 1.0 - 1.0 - 1.0 - 1.0 - 0.9 - 1.4 - 1.2 Songea 22.0 - 0.8 - 0.8 - 0.8 - 0.7 - 0.9 - 1.0 - 1.0 - 0.9 Soroti 25.7 - 0.8 - 0.9 - 1.0 - 1.0 - 0.9 - 0.9 - 1.3 - 1.1 Tabora 23.2 - 0.1 - 0.2 - 0.2 - 0.2 - 0.2 - 0.5 - 0.4 - 0.3 Tororo 23.2 0.0 - 0.1 - 0.1 0.0 0.7 0.8 - 0.3 - 0.3 Voi 26.8 - 0.6 - 0.6 - 0.6 - 0.7 - 0.8 - 0.7 - 0.6 - 0. 7 Wajir 29.7 - 0.7 - 0.8 - 0.7 - 0.7 - 1.0 - 1.0 - 0.8 - 0.7 189 Table 3. 26 Mean temperatures for February - June in 2050s for the 4 GCMs under two representative concentration pathways, RCP4.5 and RCP 8.5 Name of location Basel ine Tav FM A MJ ( 0 C ) Mean temperature change ( 0 C) 2040 - 2069 - CSIRO_r cp45 2040 - 2069 - CSIRO_r cp85 2040 - 2069 - MIROC_ rcp45 2040 - 2069 - MIROC_ rcp85 2040 - 2069 - MRI - CGCM3_ rcp45 2040 - 2069 - MRI - CGCM3_ rcp85 2040 - 2069 - NorES M1 - M_rcp 45 2040 - 2069 - NorES M1 - M_rcp 85 Arua 24.9 - 0.7 - 0.8 - 0 .9 - 0.8 - 0.8 - 0.9 - 1.3 - 1.0 Arusha 14.2 5.9 5.9 5.9 5.9 5.7 5.8 5.9 5.9 Bukoba 22.4 - 0.5 - 0.6 - 0.7 - 0.7 - 0.8 - 0.8 - 0.8 - 0.8 Dagoretti Corner 20.9 - 2.2 - 2.2 - 2.2 - 2.2 - 2.3 - 2.3 - 2.1 - 2.1 Dar es Salaam 26.6 0.4 0.4 0.5 0.4 0.4 0.3 0.4 0.4 Dodoma 23.3 - 0. 4 - 0.3 - 0.2 - 0.1 - 0.4 - 0.5 - 0.5 - 0.5 Eldoret 18.9 - 1.0 - 1.1 - 1.1 - 1.0 - 0.9 - 1.0 - 1.1 - 1.1 Entebbe 23.1 - 0.6 - 0.7 - 0.8 - 0.8 - 0.8 - 0.8 - 1.2 - 1.0 Garissa 29.7 0.8 0.8 0.6 0.7 0.5 0.5 0.7 0.8 Gulu 25.0 - 0.6 - 0.7 - 0.8 - 0.8 - 0.9 - 0.8 - 1.5 - 0.9 Jinja 22.2 0. 8 0.7 0.7 0.7 0.8 0.9 0.3 0.5 Kabale 18.3 - 0.4 - 0.5 - 0.5 - 0.6 - 0.6 - 0.5 - 0.6 - 0.7 Kasese 24.5 - 0.4 - 0.5 - 0.5 - 0.5 - 0.1 0.1 - 0.7 - 0.6 Kigoma 23.9 0.5 0.5 0.4 0.4 0.4 0.4 0.4 0.3 Kisumu 23.9 - 0.4 - 0.5 - 0.5 - 0.4 0.4 0.4 - 0.6 - 0.6 Kitigum 26.1 - 0.2 - 0.3 - 0.4 - 0.4 - 0.5 - 0.5 - 0.9 - 0.5 Lamu 28.3 - 0.5 - 0.5 - 0.5 - 0.5 - 0.6 - 0.6 - 0.5 - 0.6 Lodwar 30.2 0.0 - 0.1 - 0.1 - 0.1 - 0.2 - 0.4 - 0.2 0.0 Makindu 24.9 - 1.0 - 1.0 - 1.1 - 1.1 - 1.2 - 1.1 - 1.0 - 1.0 Mandera 30.5 - 0.1 - 0.2 - 0.2 - 0.2 - 0.4 - 0.4 - 0.3 - 0.3 Marsabiti 25.0 - 3.8 - 3.9 - 3.8 - 3.8 - 4.1 - 4.2 - 3.9 - 3.8 Masaka 21.8 0.0 - 0.1 - 0.2 - 0.2 - 0.1 - 0.1 - 0.4 - 0.3 Mansindi 24.5 - 0.5 - 0.6 - 0.7 - 0.6 - 0.7 - 0.5 - 1.4 - 0.9 Mbarara 21.4 - 0.3 - 0.4 - 0.5 - 0.4 - 0.3 - 0.2 - 0.6 - 0.6 Mbeya 19.7 - 1.6 - 1.6 - 1.6 - 1.5 - 1.8 - 1.9 - 1.9 - 1.7 Mtw ara 25.7 1.4 1.4 1.5 1.5 1.4 1.3 1.4 1.4 Musoma 22.7 0.9 0.9 0.9 0.9 1.0 0.8 0.8 0.7 Mwanza 23.2 0.2 0.1 0.1 0.1 0.0 - 0.1 - 0.1 - 0.1 Namulong e 23.1 - 0.5 - 0.6 - 0.6 - 0.5 - 0.6 - 0.5 - 1.0 - 0.7 Narok 18.2 - 0.5 - 0.5 - 0.5 - 0.4 - 0.6 - 0.7 - 0.4 - 0.5 Same 24.8 - 1. 0 - 1.1 - 1.0 - 1.1 - 1.3 - 1.1 - 1.0 - 1.1 Serere 25.4 - 0.7 - 0.8 - 0.9 - 0.8 - 0.8 - 0.7 - 1.2 - 1.0 190 Name of location Baseline Tav FMAMJ ( 0 C ) Mean temperature change ( 0 C) 2040 - 2069 - CSIR O_rcp 45 2040 - 2069 - CSIRO_r cp85 2040 - 2069 - MIROC_r cp45 2040 - 2 069 - MIROC_r cp85 2040 - 2069 - MRI - CGCM3_r cp45 2040 - 2069 - MRI - CGCM3_r cp85 2040 - 2069 - NorES M1 - M_rcp 45 2040 - 2069 - NorES M1 - M_rcp 85 Songea 22.0 - 0.7 - 0.7 - 0.7 - 0.6 - 0.8 - 1.0 - 0.9 - 0.8 Soroti 25.7 - 0.6 - 0.7 - 0.8 - 0.8 - 0.8 - 0.7 - 1.1 - 0.9 Tabora 23.2 - 0.1 - 0.1 - 0.1 - 0 .1 - 0.2 - 0.4 - 0.4 - 0.3 Tororo 23.2 0.1 0.0 0.0 0.1 0.8 0.9 - 0.1 - 0.1 Voi 26.8 - 0.5 - 0.6 - 0.6 - 0.7 - 0.8 - 0.7 - 0.6 - 0.6 Wajir 29.7 - 0.7 - 0.7 - 0.7 - 0.7 - 1.0 - 1.0 - 0.8 - 0.7 191 Table 3. 27 Mean temperatures for February - June in 2 070s for the 4 GCMs under two representative concentration pathways, RCP4.5 and RCP 8.5 Name of location Basel ine Tav FMA MJ ( 0 C ) Mean temperature change ( 0 C) 2060 - 2089 - CSIRO_r cp45 2060 - 2089 - CSIRO_r cp85 2060 - 2089 - MIROC_ rcp45 2060 - 2089 - MIROC_ rcp85 2060 - 20 89 - MRI - CGCM3_ rcp45 2060 - 2089 - MRI - CGCM3_ rcp85 2060 - 2089 - NorES M1 - M_rcp 45 2060 - 2089 - NorES M1 - M_rcp 85 Arua 24.9 - 0.8 - 0.9 - 0.9 - 0.9 - 0.9 - 1.0 - 1.4 - 1.1 Arusha 14.2 5.9 5.9 6.0 6.0 5.8 5.9 6.0 6.0 Bukoba 22.4 - 0.4 - 0.5 - 0.6 - 0.6 - 0.7 - 0.8 - 0.6 - 0.7 Dagoretti Corner 20.9 - 2.2 - 2.2 - 2.1 - 2.1 - 2.2 - 2.3 - 2.1 - 2.1 Dar es Salaam 26.6 0.5 0.5 0.6 0.5 0.5 0.4 0.5 0.5 Dodoma 23.3 - 0.3 - 0.3 - 0.2 - 0.2 - 0.4 - 0.5 - 0.5 - 0.5 Eldoret 18.9 - 0.9 - 1.0 - 1.0 - 1.0 - 0.8 - 0.9 - 1.0 - 1.0 Entebbe 23.1 - 0.5 - 0.6 - 0.8 - 0.8 - 0.7 - 0.7 - 1.2 - 1.0 Garissa 29.7 0.8 0.8 0.5 0.6 0.5 0.5 0.7 0.8 Gulu 25.0 - 0.7 - 0.8 - 0.9 - 0.9 - 0.9 - 0.9 - 1.6 - 1.1 Jinja 22.2 0.8 0.8 0.7 0.7 0.9 1.0 0.4 0.5 Kabale 18.3 - 0.4 - 0.4 - 0.5 - 0.6 - 0.5 - 0.5 - 0.6 - 0.7 Kasese 24.5 - 0.5 - 0.6 - 0.7 - 0.7 0.0 0.1 - 0.8 - 0.8 Kigoma 23.9 0.6 0.5 0.4 0.5 0.4 0.4 0.4 0.3 Kisumu 23.9 - 0.4 - 0.4 - 0.5 - 0.4 0.4 0.4 - 0.5 - 0.6 Kitigum 26.1 - 0.2 - 0.3 - 0.4 - 0.4 - 0.4 - 0.5 - 0.9 - 0.6 Lamu 28.3 - 0.4 - 0.3 - 0.4 - 0.4 - 0.5 - 0.5 - 0.4 - 0.4 Lodwar 30.2 0.0 - 0.1 0.0 - 0.1 - 0.1 - 0.4 - 0.2 - 0.1 Maki ndu 24.9 - 1.1 - 1.1 - 1.2 - 1.1 - 1.3 - 1.2 - 1.0 - 1.0 Mandera 30.5 - 0.1 - 0.1 - 0.1 - 0.2 - 0.3 - 0.4 - 0.3 - 0.1 Marsabiti 25.0 - 3.7 - 3.7 - 3.7 - 3.7 - 3.9 - 4.1 - 3.8 - 3.7 Masaka 21.8 0.0 - 0.1 - 0.1 - 0.1 - 0.1 - 0.1 - 0.3 - 0.2 Mansindi 24.5 - 0.6 - 0.7 - 0.7 - 0.7 - 0.7 - 0.6 - 1.5 - 1.0 Mbarara 21.4 - 0.3 - 0.4 - 0.5 - 0.6 - 0.3 - 0.2 - 0.7 - 0.7 Mbeya 19.7 - 1.6 - 1.6 - 1.7 - 1.6 - 1.8 - 1.9 - 1.9 - 1.8 Mtwara 25.7 1.5 1.5 1.6 1.6 1.4 1.4 1.5 1.5 Musoma 22.7 1.1 1.0 1.0 1.0 1.1 0.9 0.9 0.8 Mwanza 23.2 0.2 0.2 0.1 0.1 0.0 - 0.1 - 0.1 - 0.1 N amulong e 23.1 - 0.4 - 0.4 - 0.6 - 0.6 - 0.5 - 0.6 - 1.2 - 0.9 Narok 18.2 - 0.4 - 0.4 - 0.3 - 0.3 - 0.4 - 0.6 - 0.4 - 0.4 Same 24.8 - 1.0 - 1.0 - 1.0 - 1.1 - 1.3 - 1.1 - 1.0 - 1.0 Serere 25.4 - 0.6 - 0.7 - 0.8 - 0.8 - 0.7 - 0.7 - 1.1 - 1.0 Songea 22.0 - 0.7 - 0.7 - 0.7 - 0.6 - 0.9 - 0.9 - 0. 9 - 0.8 Soroti 25.7 - 0.6 - 0.7 - 0.8 - 0.8 - 0.7 - 0.7 - 1.0 - 0.9 Tabora 23.2 - 0.1 - 0.1 - 0.2 - 0.1 - 0.3 - 0.4 - 0.4 - 0.3 Tororo 23.2 0.3 0.2 0.2 0.1 0.9 1.1 0.0 0.0 Voi 26.8 - 0.6 - 0.6 - 0.7 - 0.7 - 0.9 - 0.7 - 0.6 - 0.6 Wajir 29.7 - 0.6 - 0.7 - 0.6 - 0.6 - 0.9 - 0.9 - 0.8 - 0.6 192 Table 3. 28 Mean temperatures for August - December in 2030s for the 4 GCMs under two representative concentration pathways, RCP4.5 and RCP 8.5 Name of location Basel ine - Tav - ASO ND ( 0 C ) Mean temperature change ( 0 C) 2020 - 2049 - CSIRO_r cp45 2020 - 2049 - CSIRO_r cp85 2020 - 2049 - MIROC_ rcp45 2020 - 2049 - MIROC_ rcp85 2020 - 2049 - MRI - CGCM3_ rcp45 2020 - 2049 - MRI - CGCM3_ rcp85 2020 - 2049 - NorES M1 - M_rcp 45 2020 - 2049 - NorES M1 - M_rcp 85 Arua 23.9 - 0.8 - 0.9 - 1.0 - 0.9 - 0.8 - 1.1 - 1.3 - 1.1 Arusha 14.2 5.6 5 .7 5.8 5.8 5.6 5.5 5.8 5.9 Bukoba 22.2 - 0.5 - 0.6 - 0.6 - 0.6 - 0.7 - 0.7 - 0.7 - 0.7 Dagoretti Corner 20.5 - 2.4 - 2.4 - 2.3 - 2.2 - 2.5 - 2.6 - 2.3 - 2.2 Dar es Salaam 26.6 - 0.7 - 0.7 - 0.6 - 0.6 - 0.8 - 0.7 - 0.6 - 0.5 Dodoma 23.9 - 0.5 - 0.5 - 0.4 - 0.4 - 0.7 - 0.7 - 0.6 - 0.6 Eldoret 18.0 - 0.9 - 0.8 - 0.8 - 0.8 - 0.8 - 0.9 - 1.0 - 0.9 Entebbe 22.6 - 0.4 - 0.5 - 0.6 - 0.7 - 0.5 - 0.5 - 0.9 - 0.8 Garissa 28.5 0.6 0.7 0.5 0.7 0.6 0.6 0.7 0.9 Gulu 24.1 - 0.8 - 0.9 - 0.9 - 0.9 - 0.8 - 1.1 - 1.4 - 1.2 Jinja 21.7 1.1 1.0 0.8 0.8 1.1 1.0 0.5 0.6 Kabale 18.2 - 0.4 - 0.5 - 0.4 - 0.5 - 0.6 - 0.7 - 0.5 - 0.7 Kasese 24.3 - 0.9 - 1.0 - 0.9 - 0.9 - 0.2 - 0.4 - 1.2 - 1.1 Kigoma 24.4 0.4 0.3 0.4 0.4 0.3 0.3 0.2 0.1 Kisumu 23.5 0.0 0.0 0.0 0.0 0.4 0.4 - 0.2 - 0.1 Kitigum 24.9 - 0.4 - 0.5 - 0.6 - 0.6 - 0.6 - 0.8 - 1.0 - 0.9 Lamu 27.4 - 0.4 - 0.3 - 0.2 - 0.2 - 0.4 - 0.3 - 0.2 - 0.1 Lodwar 29.7 0.1 0.1 0.2 0.2 - 0.1 - 0.2 - 0.2 - 0.4 Makindu 24.0 - 1.5 - 1.4 - 1.4 - 1.4 - 1.4 - 1.5 - 1.3 - 1.2 Mandera 29.2 0.1 0.0 - 0.1 - 0.1 - 0.2 0.0 0.0 0.0 Marsabiti 24.1 - 4.3 - 4.2 - 4.4 - 4.2 - 4.5 - 4.5 - 4.4 - 4.3 Masaka 21.7 0.0 - 0.1 - 0.1 - 0.1 0.1 0.0 - 0.4 - 0.3 Mansindi 23.8 - 0.8 - 0.9 - 0.9 - 0.9 - 0.6 - 0.8 - 1.4 - 1.2 Mbarara 21.2 - 0.6 - 0.7 - 0.6 - 0.6 - 0.6 - 0.6 - 0.8 - 0.9 Mbeya 20.7 - 1.6 - 1.6 - 1.5 - 1.3 - 1.8 - 1.8 - 1.6 - 1.7 Mtwara 26.9 - 0.3 - 0.3 - 0.2 - 0.1 - 0.4 - 0.3 - 0.1 - 0.1 Musoma 23.2 0.6 0.5 0.5 0.5 0.4 0.5 0.4 0.5 Mwanza 23.6 - 0.3 - 0.3 - 0.3 - 0.3 - 0.5 - 0.4 - 0.5 - 0.4 Namulong e 22.6 - 0.1 - 0.3 - 0.5 - 0.5 - 0.2 - 0.5 - 0.9 - 0.8 Narok 17.5 - 0.3 - 0.2 - 0.2 - 0.1 - 0.4 - 0.4 - 0.3 - 0.1 Same 24.5 - 1.6 - 1.5 - 1.5 - 1.5 - 1.6 - 1.6 - 1.4 - 1.3 Serere 24.6 - 0.3 - 0.4 - 0.4 - 0.4 - 0.2 - 0.4 - 0.7 - 0.6 Songea 23.1 - 1.0 - 1.0 - 0.9 - 0.7 - 1.1 - 1.1 - 1.0 - 1.1 Soroti 25.0 - 0.5 - 0.6 - 0.6 - 0.6 - 0.5 - 0.7 - 0.9 - 0.9 Tabora 24.8 - 0.2 - 0.2 - 0.1 - 0.1 - 0.4 - 0.4 - 0.3 - 0.3 Tororo 22.6 0.3 0.2 0.3 0.2 0.7 0.7 0.0 0. 0 Voi 25.9 - 1.2 - 1.2 - 1. - 1.2 - 1.3 - 1.2 - 1.1 - 1.0 193 Table 3. 29 Mean temperatures for August - December in 2050s for the 4 GCMs under two representative concentration pathways, RCP4.5 and RCP 8.5 Name of location Basel ine - Tav - A SO ND ( 0 C ) Mean temperature change ( 0 C) 2040 - 2069 - CSIRO_r cp45 2040 - 2069 - CSIRO_r cp85 2040 - 2069 - MIROC_ rcp45 2040 - 2069 - MIROC_ rcp85 2040 - 2069 - MRI - CGCM3_ rcp45 2040 - 2069 - MRI - CGCM3_ rcp85 2040 - 2069 - NorES M1 - M_rcp 45 2040 - 2069 - NorES M1 - M_rcp 85 Arua 23.9 - 0.8 - 0.9 - 0.9 - 0.9 - 0.7 - 1.0 - 1.2 - 1.1 Arusha 14.2 5.8 5.8 5.9 6.0 5.7 5.7 5.9 6.0 Bukoba 22.2 - 0.4 - 0.5 - 0.5 - 0.5 - 0.6 - 0.6 - 0.6 - 0.6 Dagoretti Corner 20.5 - 2.3 - 2.2 - 2.2 - 2.1 - 2.3 - 2.4 - 2.2 - 2.0 Dar es Salaam 26.6 - 0.6 - 0.6 - 0.5 - 0.5 - 0.7 - 0.6 - 0.5 - 0.4 Dodom a 23.9 - 0.3 - 0.3 - 0.2 - 0.1 - 0.5 - 0.5 - 0.4 - 0.4 Eldoret 18.0 - 0.9 - 0.9 - 0.8 - 0.8 - 0.8 - 0.9 - 1.0 - 0.9 Entebbe 22.6 - 0.4 - 0.5 - 0.5 - 0.5 - 0.4 - 0.4 - 0.8 - 0.7 Garissa 28.5 0.8 0.8 0.7 0.9 0.7 0.8 0.8 1.1 Gulu 24.1 - 0.7 - 0.8 - 0.9 - 0.8 - 0.8 - 1.0 - 1.3 - 1.1 Jin ja 21.7 1.2 1.1 1.0 0.9 1.2 1.1 0.7 0.7 Kabale 18.2 - 0.2 - 0.2 - 0.2 - 0.3 - 0.4 - 0.4 - 0.4 - 0.5 Kasese 24.3 - 0.9 - 0.9 - 0.8 - 0.8 - 0.2 - 0.3 - 1.1 - 1.1 Kigoma 24.4 0.6 0.5 0.6 0.5 0.5 0.5 0.4 0.3 Kisumu 23.5 0.1 0.1 0.2 0.1 0.5 0.5 0.0 0.0 Kitigum 24.9 - 0.4 - 0.6 - 0.6 - 0.6 - 0.6 - 0.8 - 1.0 - 0.8 Lamu 27.4 - 0.4 - 0.3 - 0.2 - 0.2 - 0.4 - 0.3 - 0.2 - 0.2 Lodwar 29.7 0.1 0.1 0.3 0.3 - 0.1 - 0.2 - 0.2 - 0.2 Makindu 24.0 - 1.2 - 1.2 - 1.2 - 1.1 - 1.2 - 1.2 - 1.1 - 1.0 Mandera 29.2 0.2 0.1 0.0 0.1 - 0.1 0.1 0.1 0.1 Marsabiti 24.1 - 4.2 - 4.1 - 4.2 - 4.0 - 4.3 - 4.4 - 4.3 - 4.1 Masaka 21.7 0.1 0.0 0.0 0.0 0.2 0.2 - 0.2 - 0.1 Mansindi 23.8 - 0.7 - 0.8 - 0.9 - 0.8 - 0.6 - 0.7 - 1.4 - 1.1 Mbarara 21.2 - 0.5 - 0.6 - 0.4 - 0.5 - 0.3 - 0.4 - 0.6 - 0.7 Mbeya 20.7 - 1.3 - 1.4 - 1.3 - 1.1 - 1.5 - 1.6 - 1.4 - 1.4 Mtwara 26.9 - 0.1 - 0.1 0.0 0.0 - 0.2 - 0.2 0.0 0.0 Musoma 23.2 0.7 0.6 0.7 0.6 0.6 0.6 0.5 0.6 Mwanza 23.6 - 0.1 - 0.1 - 0.1 - 0.1 - 0.3 - 0.3 - 0.3 - 0.2 Namulong e 22.6 0.0 - 0.2 - 0.3 - 0.4 - 0.1 - 0.3 - 0.7 - 0.6 Narok 17.5 - 0.1 - 0.1 0.0 0.0 - 0.2 - 0.3 - 0.1 0.0 Same 24.5 - 1.4 - 1 .3 - 1.3 - 1.3 - 1.4 - 1.4 - 1.2 - 1.1 Serere 24.6 - 0.3 - 0.4 - 0.4 - 0.4 - 0.3 - 0.4 - 0.8 - 0.6 Songea 23.1 - 0.8 - 0.8 - 0.7 - 0.5 - 0.9 - 0.9 - 0.8 - 0.9 Soroti 25.0 - 0.5 - 0.6 - 0.6 - 0.6 - 0.5 - 0.7 - 1.0 - 0.8 Tabora 24.8 0.0 0.0 0.1 0.1 - 0.2 - 0.2 - 0.1 - 0.1 Tororo 22.6 0. 3 0.2 0.4 0.3 0.8 0.7 0.1 0.1 Voi 25.9 - 1.0 - 1.0 - 1.0 - 1.0 - 1.1 - 1.0 - 0.9 - 0.8 Wajir 28.4 - 0.5 - 0.5 - 0.6 - 0.5 - 0.7 - 0.7 - 0.6 - 0.4 194 Table 3. 30 Mean temperatures for August - December in 2070s for the 4 GCMs under two representati ve concentration pathways, RCP4.5 and RCP 8.5 Name of location Basel ine - Tav - ASO ND ( 0 C) Mean temperature change ( 0 C) 2060 - 2089 - CSIRO_r cp45 2060 - 2089 - CSIRO_r cp85 2060 - 2089 - MIROC_ rcp45 2060 - 2089 - MIROC_ rcp85 2060 - 2089 - MRI - CGCM3_ rcp45 2060 - 2089 - MRI - CGCM3 _ rcp85 2060 - 2089 - NorES M1 - M_rcp 45 2060 - 2089 - NorES M1 - M_rcp 85 Arua 23.9 - 0.7 - 0.8 - 0.8 - 0.7 - 0.6 - 0.9 - 1.1 - 0.9 Arusha 14.2 5.6 5.6 5.8 5.8 5.6 5.5 5.8 5.8 Bukoba 22.2 - 0.6 - 0.7 - 0.6 - 0.7 - 0.8 - 0.8 - 0.7 - 0.7 Dagoretti Corner 20.5 - 2.5 - 2.5 - 2.4 - 2.3 - 2.5 - 2.6 - 2.4 - 2.3 Dar es Salaam 26.6 - 0.7 - 0.7 - 0.6 - 0.6 - 0.8 - 0.7 - 0.6 - 0.6 Dodoma 23.9 - 0.5 - 0.5 - 0.4 - 0.2 - 0.6 - 0.6 - 0.5 - 0.5 Eldoret 18.0 - 0.9 - 0.9 - 0.8 - 0.8 - 0.9 - 0.9 - 1.1 - 1.0 Entebbe 22.6 - 0.5 - 0.6 - 0.7 - 0.7 - 0.6 - 0.6 - 1.0 - 0.9 Garissa 28.5 0.7 0. 8 0.8 0.9 0.8 0.7 0.8 1.0 Gulu 24.1 - 0.7 - 0.8 - 0.8 - 0.8 - 0.8 - 1.0 - 1.3 - 1.0 Jinja 21.7 1.0 0.9 0.9 1.0 1.0 1.0 0.6 0.8 Kabale 18.2 - 0.3 - 0.4 - 0.3 - 0.3 - 0.5 - 0.6 - 0.3 - 0.5 Kasese 24.3 - 0.7 - 0.8 - 0.8 - 0.8 - 0.2 - 0.2 - 1.0 - 1.1 Kigoma 24.4 0.5 0.4 0.5 0.4 0.4 0.4 0.3 0.2 Kisumu 23.5 - 0.1 - 0.1 - 0.1 0.0 0.3 0.3 - 0.3 - 0.2 Kitigum 24.9 - 0.5 - 0.6 - 0.6 - 0.6 - 0.6 - 0.8 - 1.0 - 0.8 Lamu 27.4 - 0.4 - 0.4 - 0.3 - 0.3 - 0.5 - 0.4 - 0.3 - 0.2 Lodwar 29.7 - 0.1 - 0.1 0.1 0.2 - 0.3 - 0.3 - 0.3 - 0.1 Makindu 24.0 - 1.3 - 1.2 - 1.2 - 1.1 - 1.3 - 1.3 - 1.1 - 1.0 Mandera 29.2 0.0 0.0 - 0.1 - 0.1 - 0.2 - 0.1 - 0.1 0.0 Marsabiti 24.1 - 4.2 - 4.2 - 4.3 - 4.1 - 4.4 - 4.4 - 4.3 - 4.2 Masaka 21.7 - 0.1 - 0.2 - 0.1 - 0.2 0.0 0.0 - 0.3 - 0.2 Mansindi 23.8 - 0.7 - 0.8 - 0.7 - 0.7 - 0.6 - 0.7 - 1.2 - 1.0 Mbarara 21.2 - 0.4 - 0.5 - 0.4 - 0.4 - 0.3 - 0.3 - 0.6 - 0.6 Mbeya 20.7 - 1.4 - 1.5 - 1.4 - 1.2 - 1.6 - 1.7 - 1.5 - 1.5 Mtwara 26.9 - 0.2 - 0.2 - 0.1 - 0.1 - 0.3 - 0.3 0.0 0.0 Musoma 23.2 0.6 0.5 0.6 0.6 0.4 0.5 0.5 0.5 Mwanza 23.6 - 0.2 - 0.2 - 0.2 - 0.2 - 0.4 - 0.4 - 0.4 - 0.3 Namulong e 22.6 - 0.2 - 0. 3 - 0.3 - 0.3 - 0.2 - 0.2 - 0.7 - 0.5 Narok 17.5 - 0.3 - 0.3 - 0.2 - 0.1 - 0.3 - 0.5 - 0.3 - 0.2 Same 24.5 - 1.5 - 1.4 - 1.4 - 1.4 - 1.6 - 1.6 - 1.3 - 1.2 Serere 24.6 - 0.3 - 0.4 - 0.4 - 0.4 - 0.4 - 0.5 - 0.8 - 0.7 Songea 23.1 - 0.9 - 1.0 - 0.8 - 0.7 - 1.0 - 1.1 - 0.9 - 1.0 Soroti 25.0 - 0 .5 - 0.6 - 0.6 - 0.6 - 0.7 - 0.7 - 1.0 - 0.9 Tabora 24.8 - 0.1 - 0.1 0.0 0.0 - 0.2 - 0.2 - 0.1 - 0.1 Tororo 22.6 0.2 0.1 0.2 0.1 0.6 0.6 - 0.1 0.0 Voi 25.9 - 1.1 - 1.1 - 1.0 - 1.1 - 1.1 - 1.1 - 1.0 - 0.9 Wajir 28.4 - 0.6 - 0.6 - 0.7 0.5 - 0.8 - 0.8 - 0.7 - 0.5 195 B IBLIOGRAPHY 196 BIBLI OGRAPHY Allen, J. R., Bhattacharya, N. C., Lu, J. Y., Pace, R. D., & Rogers, H. H. (1985). Field studies of sweet potatoes and cowpeas in response to elevated carbon dioxide . Retrieved from Allen, R. G., Pereira, L. S., Raes, D., & Smith, M. (1998). Crop evapotranspiration guidelines for computing crop water requirements. FAO Irrig. and Drain. Paper 56 . Retrieved from http://www.fao.org/docrep/X0490E/X0490E00.htm Andresen, J. A., Alagarswamy, G., Rotz, C. A., Ritchie, J. T., & LeBa ron, A. W. (2001). Weather Impacts on Maize, Soybean, and Alfalfa Production in the Great Lakes Region, 1895 1996. Agron. J., 93 . Anon. (1993). Uganda Agriculture: a Wo r ld Bank Country Study . Retrieved from Washington, USA: Anyah, R. O., Semazzi, F. H. M., & Xie, L. (2006). Simulated physical mechanisms associated with climate variability over Lake Victoria basin in East Africa. Monthly Weather Review, 134 (12), 3588 - 3609. Arakawa, A., & Shubert, W. H. (1974). Interaction of a cumulus cloud ensemble wi th the large - scale environment, Part 1. J Atmos Sci, 31 , 674 701. Arndt, C., Farmer, W., Strzepek, K., & Thurlow, J. (2012). Climate change, agriculture and food security in Tanzania . Retrieved from Ashby, J. A., & Sperling, L. (1994). Institutionalisi ng participatory, client - driven research and technology development in agriculture. ODI Agricultural Research & Extension Network Paper No. 49, 21pp. Bashaasha, B., Mwanga, R. O. M., p'Obwoya, C. O., & Ewell, P. T. (1995). Sweet potato in the farming and food systems of Uganda: A farm surve y report . Retrieved from Bentsen, M., Bethke, I., Debernard, J. B., Iversen, T., Kirkeva°g, A., Seland, Ø., . . . Kristjansson, J. E. (2012). The Norwegian earth system model, NorESM1 - M Part 1: description and basic e valuation. Geosci Model Dev Discuss, 5 , 2843 2931. Berg, T. (1997). Devolution of plant breeding. In L. Sperling & M. Loevinsohn (Eds.), Using Diversity. Enhancing and maintaining genetic resources on - farm. IDRC Canada. Bhattacharya, N. C., Biswas, P. K., Battacharya, S., Sionit, N., & Strain, B. R. (1985). Growth and yield response of sweet potato to atmospheric CO2 enrichment. Crop Science, 25 , 975 - 981. 197 Biswas, T.D. and S.K. Mukherjee, 1994. Soil and fertilizer use. In : Textbook of soil science. Ta ta McGraw Hill Publishing Company, pp: 222 - 285. Bovell - Benjamin, A., C. (2007). Sweet potato: A review of its past, present and future role in human nutrition. Advances in Food and Nutrition Research, 52 . Brussolo, E., J. von Hardenberg, L. Ferraris, N. Rebora, and A. Provenzale, 2008: Verification of quantitative precipitation forecasts via stochastic downscaling. J. Hydrometeor., 9, 1084 1094, doi :10.1175/2008JHM994.1. Chiew, F., D. Kirono, D. Kent, A. Frost, S. Charles, B. Timbal, K. Nguyen, and G. Fu, 2010: Comparison of runoff modeled using rainfall from different downscaling methods for historical and future climates. J. Hydrol., 387, 10 23, doi :10.1016/j.jhydrol.2010.03.025. Codjoe S, Owusu G (2011) Climate change/variability and food systems: evidence from the Afram Plains, Ghana. Regional Environmental Change, 11, 753 765. Cook, K. H., & Vizy, E. K. (2013). Projected Changes in East African Rainy Seasons. Journal of Climate, 26. , 5931 5948. Crespo O, Hachigonta S, Tadross M (2011) Sensitivi ty of southern African maize yields to the definition of sowing dekad in a changing climate. Climatic Change, 106, 267 283. Davies, J. A., & McKay, D. C. (1989). Evaluation of selected models for estimating solar radiation on horizontal surfaces. Solar E nergy, 43 , 153 - 168. Ddumba, S. D., Andresen, J., & Snapp, S. S. (2014). Characteristics and adaptive potential of sweet potato cultivars grown in Uganda. International Journal of Agriculture and Forestry, 4 (2), 135 - 143. De Waal, D., F. R. Chinjinga, L. Johansson, F. F. Kanju, & Nathaniels., N. (1997). Village - based cassava breeding in Tanzania. In Farmers Research in Practice. Van Velhuisen, L., Waters - Bayer, A., Ramirez, R., Johnson, D. A. and Thompson, J. (Eds). London: Intermediate Technology Publica tions. Tools and tillage research (pp. 83 - 88). Diop, A. (1998). Storage and processing of roots and tubers in the tropics. In D. J. B. Calverley (Ed.), Food and Agriculture Organization of the United Nations, Agro - Industries and Post - Havest Management Ser vice (pp. 38 - 50). Rome, Italy: Food and Agriculture Organization. Ebregt, E., Struik, P. C., Odongo, B., & Abidin, P. E. (2007). Piecemeal versus one - time harvesting of sweet potato in north - eastern Uganda with special reference to pest damage. NJAS Wagen ingen Journal of Life Sciences, 55 - 1 . 198 Ewell, P. T., & Mutuura, J. (1994). Sweet potato in the food systems of Eastern and Southern Africa. 405 - 412. FAO. (2014). FAOSTAT Production data http://faostat3.fao.org/faostat - gateway/go/to/download/Q/QC/E accessed on June 26, 2014. FAO/IIASA/ISRIC/ISSCAS/JRC. (2012). Harmonized World Soil Database (version 1.2). FAO, Rome, Italy , and IIASA, Laxenburg, Austria. URL : http://webarchive.iiasa.ac.at/Research/LUC/External - World - soil - database/HTML/index.html . Ferraris, L., S. Gabellani, U. Parodi, N. Rebora, J. von Hardenberg, and A. Pro venzale, 2003a: Revisiting multifractality in rainfall fields. J. Hydrometeor., 4, 544 551, doi:10.1175/1525 - 7541(2003)004,0544 : RMIRF.2.0 .CO;2. Fischer G, Shah M, Tubiello FN, van Velhuizen H (2005) Socio - economic and climate change impacts on agricult ure: An integrated assessment, 1990 - 2080. Philos Trans R Soc Lond B Biol Sci 360(1463):2067 2083. FEWSNET. (2010). A climate trend analysis of Kenya - August 2010: U.S. Geological Survey Fact Sheet 2010 - 3074, 4 p. Available at http://pubs.usgs.gov/fs/2010/3074/pdf/fs2010 - 3074.pdf . FEWSNET. (2012). A climate trend analysis of Uganda. Famine Early Warning Systems NETwork Informing climate adaptation series. Accessed on June 14, 2013. A vailable at http://pubs.usgs.gov/fs/2012/3062/FS2012 - 3062.pdf . Funk, C., Peterson, P., Landsfeld, M., D., P., J., V., J., R., . . . Verdin, A. (2013). A Quasi - global precipitation time ser ies for drought Monitoring, USGS Data Product XXX, Gajanayake, B., Reddy, K. R., Shankle, M. W., Arancibia, R. A., & Villordon, A. O. (2014). Quantifying Storage Root Initiation, Growth, and Developmental Responses of Sweetpotato to Early Season Temperat ure. Agronomy Journal, 106 (5), 1795 - 1804. Grace K, Davenport F, Funk C, Lerner AM (2012) Child malnutrition and climate in Sub - Saharan Africa: An analysis of recent trends in Kenya. Appl Geogr 35(1 2): 405 413. Gibson, R. W., Byamukama, E., Mpembe, I., Kayongo, J., & Mwanga, R. O. M. (2008). Working with farmer groups in Uganda to develop new sweet potato cultivars: decentrali z ation and building on traditional approaches. Euphytica, 159 , 217 - 228. Gibson, R. W., Mpembe, I., & Mwanga, R. O. M. (2011). Be nefits of participatory plant breeding (PPB) as exemplified by the first - ever officially released PPB - bred sweet potato cultivar. Journal of Agricultural Science , 1 - 8. 199 Gibson, R. W., Mwanga, R. O. M., Namanda, S., Jeremiah, S. C., & Barker, I. (2009). Re view of sweet potato seed systems in East and Southern Africa . Retrieved from Lima, Peru: Gijsman, A. J., Jagtap, S. S., & J.W., J. (2002). Wading through a swamp of complete confusion: how to choose a method for estimating soil water retention parameter s for crop models. European Journal of Agronomy, 18 75 - 105. Giorgi, F., Jones, C., & Asrar, G. R. (2009). Addressing climate information needs at the regional level: the CORDEX framework. World Meteorological Organization (WMO) Bulletin, 58 (3), 175. Go ddard, L., & Graham, N. E. (1999). Importance of the Indian Ocean for simulating rainfall anomalies over eastern and southern Africa. Journal of Geophysical Research: Atmospheres (1984 - 2012), 104 (D16), 19099 - 19116. Gomes, F., & Carr, M. K. V. (2003). Eff ects of water availability and vine harvesting frequency on the productivity of sweet potato in southern Mozambique. II. Crop water use. Experimental Agriculture, 39 , 39 - 54. Gornall J, Betts R, Burke E, Clark R, Camp J, Willett K, Wiltshire A (2010) Impl ications of climate change for agricultural productivity in the early twenty - first century. Philosophical Transaction s of the Royal Society B, 365, 2973 2989. Hall, A., Bocket, G., & Nahdy, S. (1998). Sweet potato postharvest systems in Uganda: Strategie s, constraints, and potentials. Social Science Department Wo r king Paper No. 1998 - 7 . Heltberg, R., Siegel, P., Bennett , & Jorgensen, S., Lau (2009). Addressing human vulnerability to climate change: Toward a 'no - regrets' approach. Global Environmental Cha nge, 19 (1), 89 - 99. Hepworth, N., & Goulden, M. (2008). Climate Change in Uganda: Understanding the implications and appraising the response, LTS International, Edinburgh. Hirsch, R. M., Slack, J. R., & Smith, R. A. (1982). Techniques of trend analysis for monthly water data Water Resour. Res, 18 , 107 121. Holden, N., Brereton, A., 2006. Adaptation of water and nitrogen management of spring barley and potato as a response to possible climate change in Ireland. Agric. Water Manage. 82, 297 317. Humph ries, S., O. Gallardo, J. Jimenez, & F. Sierra. (2005). Linking small farmers to the formal research sector: lessons from a participatory bean breeding programme in Honduras. ODI Agricultural Research & Extension Network Paper No 142, 14pp. 200 IPCC. (2007a) . Climate Change 2007: Climate Change Impacts, Adaptation , and Vulnerability. In Working Group II contribution to the Intergovernmental Panel on Climate Change Fourth Assessment Report . Geneva, Switzerland: Intergovernmental Panel on Climate Change. IPCC. (2007b). Climate Change 2007: The physical science basis - summary for policymakers In Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change . Geneva, Switzerland: Intergovernmental Panel on Climate Change. IPCC. (2013). Working Group 1: Contribution to the IPCC fifth assessment report (AR5), Climate Change 2013: The physical science basis . Retrieved from Geneva, Switzerland: Jarret, & Gawel. (1991). Chemical and environmental growth regulation of sweet potato in vitro. Tissue and Organ Culture, 25 , 153 - 159. Jiang, X., Jianjun, H., & Wang, Y. (2004). Sweet potato processing and product research and - Harvest Research Proceedings of an International Workshop held in Chengdu, Sichuan, PR China, November 7 8, 2001, International Potato Center (CIP), Bogor, Indonesia. Jones, C. (2013). An overview of CORDEX o ver the past four years. Available at http://cordex2013.wcrp - climate.org/plenary_A1/PL_A1_04_Jones.pdf accessed on 11/20/2013. Paper presented at the Internati o nal conference on regional climate - CORDEX 2013, Brussels, Belgium. Jones, J. W., Hoogenboom, G., Porter, C. H., Boote, K. J., Batchelor, W. D., Hunt, L. A., . . . Ritchie, J. T. (2003). The DSSAT cropping system model. European Journal of Agronomy , 235 - 265. Jones, P . G., & Thornton, P. K. (2000). MarkSim: software to generate daily weather data for Latin America and Africa Agron. J., 93 , 445 453. Jones, P. G., & Thornton, P. K. (2013). Generating downscaled weather data from a suite of climate models for agricultur al modeling applications. Agricultural Systems, 114 , 1 - 5. Kanyama - Phiri, G., Snapp, S. S., Kamanga, B., & Wellard, K. (2000). Towards integrated soil fertility management in Malawi: incorporating participatory approaches in agricultural research. Managin Kapinga, R., de Steenhuijsen, P. B., Heemskerk, W., Chirimi, B., Mutalemwa, M., Kabissa, J., & Kapingu, P. (1998). Participatory research in sweet potato variety evaluation and selection for the diverse environments of the Lake Z one of Tanzania Paper presented at the 7th Triennial Symposium of the International Society for Tropical Root Crops - Africa Branch. 201 Katz, R.W. and Brown, B.G., 1992. Extreme events in a changing climate: variability is more important than aver ages. Clim. Change, 21: 289 - 302. Kendall, M. G. (1975). Rank correlation methods (Vol. 5th ed. ): Charles Griffe n and Company Ltd. Publ., London, UK. Kerr, R. B., Snapp, S. S., Chiwa, M., Shumba, L., & Msachi, R. (2007). Participatory research on legume diversification with Malawinian smallholder farmers for improved human nutrition and soil fertility. Expl. Agric., 43 , 437 - 453. Kim, Y. C. (1961). Effects of thermoperiodism on tuber formation in Ipomoea batatas under controlled conditions. Plant Physiol ogy, 36 , 680. Li, G., & Xie, S. P. (2014). Tropical biases in CMIP5 multimodel ensemble: The excessive equatorial Pacific cold tongue and double ITCZ problems. Journal of Climate, 27(4), 1765 - 1780. Loretan, P. A., Bonsi, C. K., Mortley, D. G., Wheeler, R. M., Mackowiak, C. L., Hill, W. A., . . . David, P. P. (1994). Effects of several environmental factors on sweet potato growth. Advances in Space Research, 14 (11), 277 - 280. Lloyd SJ, Kovats RS, Chalabi Z (2011) Climate change, crop yields and undernut rition: development of a model to quantify the impacts of climate scenarios on child undernutrition. Environmental Health Perspectives, 119, 1817 1823. Manu - Aduening, J. A., Lamboll, R. I., Ampong, M. G., Lamptey, J. N., Moses, E., Dankyi, A. A., & Gibso n, R. W. (2006). Development of superior cassava cultivars in Ghana by farmers and scientists: the process adopted, outcomes and contributions and chang ing roles of different stakeholders. Euphytica, 150 , 47 - 61. Maraun, D., and Coauthors, 2010: Precipita tion downscaling under climate change: Recent developments to bridge the gap between dynamical models and the end user. Rev. Geophys., 48, RG3003, doi:10.1029/2009RG000314 Mearns, L.O., Schneider, S.H., Thompson, S.L. and McDaniel, L.R., 1990. Analysis o f climate variability in general - circulation models -- comparison with observation and changes in variability in 2 x CO, experiments. J. Geophys. Res., 95: 20469 - 20490. Mendelsohn, R., Dinar, A., & Williams, L. (2006). The distributional impact of climate change on rich and poor countries Environment and Development Economics, 11 , 159 - 178. Mithra, S., V. S., & Somasundaram, K. (2008). A model to simulate sweet potato growth. worl Applied Sciences Journal, 4 (4), 568 - 577. 202 Morris, M. L., & Bellon, M. R. ( 2004). Participatory plant breeding research: opportunities and challenges for the international crop improvement system. Euphytica, 136 , 21 - 35. Mwanga, R. O. M., Kigozi, B., Namakula, J., Mpembe, I., Niringiye, C., Tumwegamire, S., . . . Yencho, G. (201 0). Submission to the variety release committee for release of sweet potato varieties . Retrieved from Mwanga, R. O. M., Odongo, B., Alajo, A., Kigozi, B., Niringiye, C., Kapinga, R., . . . Yencho, G. (2007). Submission to the variety release committee fo r release of sweet potato varieties . Retrieved from Mwanga, R. O. M., Odongo, B., Niringiye, C., Zhang, D., Yencho, G. C., & Kapinga, R. (2003). Orange - fleshed sweet potato breeding activities in Uganda. African Crop Science Conference . Mwanga, R. O. M ., Stevenson, P., & Yencho, G. C. (2005). McKnight Foundation Collaborative Research Program: Development of high yielding multiple resistant sweet potato germplasm . Retrieved from NASA - POWER. (2014). The National Aeronautics and Space Administration - Pr ediction Of Worldwide Energy Resource agronomical data portal. ; url: http://power.larc.nasa.gov/cgi - bin/cgiwrap/solar/agro.cgi?email=agroclim@larc.nasa.g ov NASA. (2013). NASAPOWER dataset. url: http://power.larc.nasa.gov/cgi - bin/cgiwrap/solar/agro.cgi?email=agroclim@larc.nasa.gov . Nash, J. E., & Sutcl iffe, J. V. (1970). River flow forecasting through conceptual models, Part I - A discussion of principles. Journal of Hydrology, 10 , 282 - 290. Nelson G Cet al 2010 Food Security, Farming, and Climate Change to 2050: Scenarios, Results, Policy Options(Was hington, DC: International Food Policy Research Institute) Ogallo, L. J. (1989). The spatial and temporal patterns of the East African seasonal rainfall derived from principal component analysis. International Journal of Climatology, 9 (2), 145 - 167. Opin ion Research Corporation Macro International Inc (ORC Macro). (2006). Uganda Demographic Health Survey 2006. Calverton, MD: ORC Macro. Otieno, V. O., & Anyah, R. O. (2013). CMIP5 simulated climate conditions of the Greater Horn of Africa (GHA). Part 1: c ontemporary climate. Climate Dynamics, 41 (7 - 8), 2081 - 2097. Otieno, V. O., & Anyah, R. O. (2013). CMIP5 simulated climate conditions of the Greater Horn of Africa (GHA). Part II: projected climate. Climate Dynamics, 41 (7 - 8), 2099 - 2113. 203 Pan, D. M., & Ran dall, D. A. (1998). A cumulus paramet e rization with a prognostic closure Q J R Meteor Soc, 124 , 949 981. Parry, M., Rosenzweig, C., & Livermore, M. (2005). Climate change, global food supply and risk of hunger Philosophical Transactions of the Royal Soci ety B - Biological Sciences, 360 (1463), 2125 - 2138. Partal, T., & Kahya, E. (2006). Trend analysis in Turkish precipitation data. Hydrol Process, 20 , 2011 2026 Pathak, T., Fraisse, C., Jones, J., Messina, C., & Hoogenboom, G. (2007). Use of global sensit ivity analysis for CROPGRO cotton model development. Transactions of the ASABE, 50 (6), 2295 - 2302. Ravi, V., Nascar, S. K., Makeshkumar, T., & Binoy Babu, P. K. B. (2009). Molecular physiology of storage root formation and development in sweet potato (Ipo moea batatas (L.) Lam.) J. Root Crops, 35 , 1 27. Ravi, V., & P., I. (1999). Crop physiology of sweet potato Hortic. Rev. (Am. Soc. Hortic. Sci.), 23 , 277 338. Reichle, R. H., Koster, R. D., De Lannoy, G. J. M., Forman, B. A., Liu, Q., Mahanama, S. P. P ., & Toure, A. (2011). Assessment and enhancement of MERRA land surface hydrology estimates. Journal of Climate, 24 (24), 6322 - 6338. Rienecker, M. M., Suarez, M. J., Gelaro, R., Todling, R., Bacmeister, J., Liu, E., . . . Kim, G. - K. (2011). MERRA: NASA's modern - era retrospective analysis for research and applications. Journal of Climate, 24 (14), 3624 - 3648. Rosegrant, M. W., Msangi, S., Ringler, C., Sulser, T. B., Zhu, T., & Cline, S. A. (2008). An i nternational model for policy analysis of agricultural c ommodities and trade (IMPACT): Model description : International Food Policy Research Institute Washington, DC, USA. Rosenzweig, C., & Parry, M. L. (1994). Potential impact of climate change on world food supply. Nature, 367 , 132 - 138. Rotstayn, L. D. (19 98). A physically based scheme for the treatment of stratiform clouds and precipitation in large - scale models. II: comparison of modeled and observed climatological fields Q J R Meteorol Soc, 124 , 389 415. Rotstayn, L. D., Collier, M. A., Dix, M. R., Fen J. (2009). Improved simulation of Australian climate and ENSO - related climate variability in a GCM with an interactive aerosol treatment Int J Climatol, 30 , 1067 1088. 204 Rotstayn, L. D., Jeffrey, S. J ., Collier, M. A., Dravitzki, S. M., Hirst, A. C., Syktus, J. I., & Wong, K. K. (2012). Aerosol - induced changes in summer rainfall and circulation in the Australasian region: a study using single - forcing climate simulations Atmos Chem Phys, 12 , 5107 5188. Ruane, A. C., Goldberg, R., & Chryssanthacopoulos, J. (2015). Climate forcing datasets for agricultural modeling: Merged products for gap - filling and historical climate series estimation. Agricultural and Forest Meteorology, 200 , 233 - 248. Ruel, M. T. ( 2001). Can food - based strategies help reduce vitamin A and iron deficiencies, a review of recent evidence ? International F ood P olicy Research Institute. Washington D.C. Ruget, F., Brisson, N., Delécolle, R., & Faivre, R. (2002). Sensitivity analysis of a crop simulation model, STICS, in order to choose the main parameters to be estimated. Agronom y , 22 (2), 133 - 158. Ruiz - Noguera, B., Boote, K. J., & Sau, F. (2001). Calibration and use of CROPGRO - soybean model for improving soybean management under rainfed conditions. Agricultural Systems, 68 , 151 - 173. Saxton, K. E., Rawls, W. J., Romberger, J. S., & Papendick, R. I. (1986). Estimating generalized soil - water characteristics from texture. Soil Science Society of America Journal, 50 , 1031 - 1036. Schlenker, W., & Lobell, D. B. (2010). Robust negative impacts of climate change on African agriculture. Environmental Research Letters, 5 , 8pp. Seland, O., Iversen, T., Kirkevag, A., & Storelvmo, T. (2008). Aerosol - climate interactions in the CAM - Oslo atmospheric GCM and investigation of associated basic shortcomings Tellus A, 60 (3), 459 491. Semenov, M. A., & Porter, J. R. (1995). Climatic variability and the modeling of crop yields. Agricultural and forest meteorology, 73(3 - 4), 265 - 283. Sen, P. K. (1968). E J. Am. Stat, Assoc., 63 , 1379 - 1389. Shaeffer, D. L. (1980). Model evaluation methodology applicable to environmental assessment models. Ecological Modelling, 8 , 275 - 295. Smit, N., E.J.M. (19 97). The effect of the indigenous cultural practices of in - ground storage and piecemeal harvesting of sweet potato on yield and quality losses caused by sweet potato weevil in Uganda. Agriculture, Ecosystems and Environment, 64 , 191 - 200. 205 Snapp, S. S., Ma fongoya, P. L., & Waddington, S. (1998). Organic matter technologies for integrated nutrient management in smallholder cropping systems of southern Africa Agriculture, Ecosystems and Environment, 71 , 185 - 200. Snapp, S. S., Rohrbach, D. D., Simtowe, R., & Freeman, H. A. (2002). Sustainable soil management options for Malawi: can smallholder farmers grow more legumes? Agriculture, Ecosystems and Environment, 91 , 159 - 174. Somasundaram, K., & Mithra, S. V. S. (2008). Madhuram: A simulation model for sweet p otato growth. World Journal of Agricultural Sciences, 4 (2), 241 - 254. Spence, J. A., & Humphries, E. C. (1972). Effect of moisture supply, root temperature, and growth regulators on photosynthesis of isolated rooted leaves of sweet potato (Ipomoea batatas ). Annals of Botany, 36 (1), 115 - 121. Sperling L, Ashby, J. A., Smith, M. E., Weltzien, E., & McGuire, S. (2001). A framework for analyzing participatory plant breeding approaches and results Euphytica, 122 , 439 - 450. Sperling L, M. E. Loevinsohn, & Ntab breeding: local bean experts and on - station selection in Rwanda. Exp Agric, 29 , 509 519. Sperling, L., & Scheidegger, U. (1996). Participatory selection of beans in Rwanda: results methods and institutional issues. IIED Sustainable Agriculture Gatekeeper Series No. SA 51. London: International Institute for Environment and Development. Systa t Software Inc. (2007). SYSTAT 12 statistics. In: Systat Software, Inc. San Jose, CA. Stockle, C.O., Nelson, R.L., Higgins, S., Brunner, J., Grove, G., Boydston, R., Whiting, M., Kruger, C., 2010. Assessment of climate change impact on Eastern Washington agriculture. Climatic Change 102, 77 102. Taylor, K. E., Stouffer, R. J., & Meehl, G. A. (2009). A summary of the CMIP5 experiment design. WCRP, submitted . Togari, Y. (1950). A study of tuberous root initiation in sweet potato Bull. Natl. Agric. Exp. S tn. Tokyo, 68 , 1 - 96. Tubiello, F.N., Rosenzweig, C., Goldberg, R.A., Jagtap, S., Jones, J.W., 2002. Effects of climate change on US crop production: simulation results using two different GCM scenarios. Part I: Wheat, potato, maize, and citrus. Climate Res. 20, 259 270. Thornton PK, Jones PG, Ericksen PJ, Challinor AJ (2011) Agriculture and food systems in sub - Saharan Africa in a four - plus degree world. Philosophical Transactions of the Royal Society Series A, 369, 117 136 206 Tumwegamire, S., Kapinga, R., Mwanga, R. O. M., Niringiye, C., Lemaga, B., & Nsumba, J. (2007). Acceptability studies of orange - fleshed sweet potato varieties in Uganda . Paper presented at the 13th ISTRC Symposium, Arusha, Tanzania. Van Dam, J., Kooman, P. L., & C., S. P. (1996). Ef fects of temperature and photoperiod on early growth and the final number of tubers in potato (Solanum tuberosum L.) Potato Res, 39 , 51 - 62. VEDCO. (2001). Project report: Promotion of orange flesh sweet potato in Luwero District, 2001. VEDCO (Volunteer E ffort for Development Concerns) . Retrieved from Kampala, Uganda: Villordon, A., LaBonte, D., Solis, J., & Firon, N. (2012). Characterization of lateral root development at the onset of storage root initiation in Beauregard sweet potato adventitious roots HortScience, 47 , 961 968. Watanabe, S., Hajima, T., Sudo, K., Nagashima, T., Takemura, T., Okajima, H., . . . Kawamiya, M. (2011). MIROC - ESM 2010: model description and basic results of CMIP5 - 20c3 m experiments GMD, 4 (4), 845 872. White, J. W., Hooge nboom, G., Stackhouse Jr, P. W., & Hoell, J. M. (2008). Evaluation of NASA satellite - and assimilation model - derived long - term daily temperature data over the continental US. Agricultural and Forest Meteorology, 148 (10), 1574 - 1584. White, J. W., Hoogenboo m, G., Wilkens, P. W., Stackhouse Jr, P. W., & Hoel, J. M. (2011). Evaluation of satellite - based, modeled - derived daily solar radiation data for the Continental United States. Agronomy Journal, 103 (4), 1242 - 1251. Willmott, C. J., & Wicks, D. E. (1980). A n empirical method for the spatial interpolation of monthly precipitation within California. Physical Geography, 1 , 59 - 73. Witcombe, J. R., Gyawali, S., Sunwar, S., Sthapit, B. R., & Joshi, K. D. (2005). Participatory plant breeding is better described a s highly client - orientated plant breeding. II. Optional farmer collaboration in the segregating generations. Exp Agric, 42 , 79 90. Wrigley, C. W., C. Blumenthal, P.W. Gras, and E.W.R. Barlow. (1994). Temperature variation during grain filling and changes in wheat - grain quality Aust. J. Plant Physiol., 21 , 875 888 Yaku, A., Hill, S., B., & Chiasson, H. (1992). Effects of intercropping on a population of sweet potato weevil, Cylas formicarius (F.) (COLEOPTERA: CURCULIONIDAE). Science in New Guinea, 18 (3), 123 - 132. Yanggen, D., & Nagujja, S. (2006). The use of orange - fleshed sweet potato to combat Vitamin A deficiency in Uganda. A study of varietal preferences, extension strategies , and post - harvest utilization. Retrieved from Lima, Peru: 207 Yenigun, K., G umus, V., & H., B. (2008). Trends in streamflow of the Euphrates basin. Turkey. Proc Inst Civil Eng Water Manag, 161 , 189 198. doi:110.1680/wama.2008.1161.1684.1189. Yukimoto S et al. (2006). Present - day climate and climate sensitivity in the Meteorologi cal Research Institute Coupled GCM, Version 2.3 (MRI - CGCM2.3). J Meteor Soc Jpn, 84 , 333 363. Zhang, G. J., & McFarlane, N. A. (1995). The s ensitivity of climate simulations to the parameterization of cumulus convection in the Canadian climate cent er gen eral circulation model Atmos - Ocean, 33 , 407 446. Zhang, Y., Rossow, W. B., & Stackhouse, P. W. J. (2007). Comparison of different global information sources used in surface radiative flux calculation: Radiative properties of the surface. J. Geophys. Res. , 112 .