VULNERABILITY AND IMPACTS OF CLIMATE CHANGE ON FOOD CROPS IN RAISED ATOLL COMMUNITIES: A CASE STUDY OF BELLONA COMMUNITY IN .

by

Joseph Maeke

A thesis submitted in fulfillment of the requirements for the Degree of Master of Science in Climate Change.

Copyright © 2013 by Joseph Maeke

Pacific Centre for Environment and Sustainable Development (PACE-SD) Faculty of Science, Technology and Environment The University of the South Pacific

July, 2013

DECLARATION OF ORIGINALITY

Statement by Author

I, Joseph Maeke hereby declare that this thesis is the account of my own work and that, to the best of my knowledge; it contains no material previously published, or submitted for the award on any other degree or diploma at any tertiary institution, except where due acknowledgement or reference is made in the text.

Signature: Date: 19th July 2013

Name: Joseph Maeke

Student ID No: S01004381

Statement by Supervisor

This research in this thesis is performed under my supervision and to my knowledge is the sole work of Mr. Joseph Maeke.

Signature: Date: 19th July 2013

Name: Prof. Elisabeth Holland

Designation: Principal Supervisor

DEDICATION

Dedicated to my best friend and wife, Samantha Annonna Maeke for the endless support from the initial stage of this thesis until its completion. Thank you ACKNOWLEDGEMENT

This thesis would have not been possible without the scholarship and financial support of PACE-SD (Pacific Centre for Environment and Sustainable Development) through the AusAID Future Climate Leaders Project (FCLP) of which I am grateful. I am indebted to my initial principal supervisors Dr. Morgan Wairiu and Dr. Dan Orcherton who directed and advised me during the initial stage of the thesis until their departure from PACE-SD. I would also like to thank Prof. Elisabeth Holland who is willing take on the principal supervisor role and supported me in the completion and submission of this thesis. My sincere gratitude goes to my co-supervisor Mr. Viliamu Iese and advisor Dr. Upendra Singh who coached and trained me on DSSAT crop modelling simulation and analysis. For facilitating my DSSAT training abroad, administrative and logistical matters, I would like to thank Mr. Sumeet Naidu the coordinator of the FLCP for his support.

I would also like to acknowledge the support rendered by various agencies in such as the Geology Division, Ministry of Mines for using their GeoChem lab, Statistic office-Ministry of Finance, Solomon Islands Visitors Beaureau, Solomon Island Meteorological Services and Climate Change Division-Ministry of Environment, Climate Change, Disaster Management and Meteorology.

To Mr. Gareth Quity, thank you for your assistance and support in instructing me on DSSAT initially when I had no idea on what crop modelling is all about. My outmost gratitude to the people and community of Bellona who participated in the household survey interviews, focus group discussion and for allowing their garden plots to be observed. In particular, I would like to thank Mr. Wilson Tongabaea for his rich knowledge on Bellonese farming systems and my tour guides and interpreters: Baiabe Tuitupu, Eric Mamu, Ezekiel Tuhenua and Michael Saosogo. I would also like to thank the Renbel Provincial Secretary Mr. Adrian Tuhanuku for giving me permission on behalf of the Renbel Province to conduct this research in Bellona. I am also indebted to my committed research assistant, Mr Jeamond Gumizama for his invaluable assistance in getting the most out of my household survey and focus group interviews as well as field data collection.

i

ABSTRACT

Climate change is a serious threat for Solomon Island communities’ food security. Therefore it is very important to understand the current vulnerability of Solomon Island communities’ food security to adverse impacts of climate change in order to develop important adaptation strategies to improve food security and reduce the impacts. This study examines the impacts of climate change and extreme events on food crop production in Bellonese communities in the Solomon Islands. Household interviews and surveys were conducted for 25% of households in four Bellona wards (West and East Ghonghau, Sa’aiho and Matangi) to collect and document information on food security systems and livelihood. The crop management information from the farmers were also used as inputs to calibrate and run the DSSAT (Decision Support System for Agrotechnology transfer) version 4.5 crop models to simulate the impacts of climate change (current, future) on taro (Colocasia esculenta), cassava (Manihot esculenta) and corn (Zea mays) growth and yield in three wards (West and East Ghonghau, Sa’aiho). Crop production impacts already observed in the four Bellonese wards included wilting, decline in crop yield, tuber/corm size, survival rate of young seedlings, crop growth, early or delay in maturity, abnormality of fruits/tubers, increase in pest and diseases, rotting, loss of some crop varieties, and loss of tuber flavor. Crop model projections for the three wards based on climate, atmospheric CO2, and soil type indicated that for Sa’aiho; C. esculenta yields are projected to decline by 12%, M. esculenta yields are projected to increase by 15.6%, and Z. mays yields are projected to decline by 13.2% for 2090. In both West and East Ghonghau, C. esculenta yields are projected to increase by 4.6% and 6.7% respectively, M. esculenta yields are projected to increase by 18.5%, and 24.1% respectively, and Z. mays are projected to decline by 8.3% for 2090. Based on crop model simulations of currently grown cultivars, C. esculenta cultivation will be restricted to the two fertile soil types, Kenge ungi and Kenge toaha found in West and East Ghonghau respectively. Z. mays yields are projected to decline. M. esculenta yields are projected to be the most resilient of the three crops, with sustained yields for all sites in all climate projections. The DSSAT crop model simulations also agreed with farmers reported impacts of El Niño induced of 1997. To improve food security in a changing climate in the studied sites, it is recommended for farmers to continue to implement their traditional

ii sustainable cultivation and adaptation strategies which include; shifting cultivation, bush fallow practice, crop rotation, intercropping, diversification of crops, shade maintenance, mulching, planting fast yielding/resilient crops, adjust planting dates, increase number of plots, change planting sites, increase mound size, and planting distribution according to soil fertility and crop type. Furthermore, according to the crop models simulations farmers yield in Sa’aiho could be improved if they plant other C. esculenta cultivars; Bun long, Lehua, and Tausala-Samoa, with projected increased yields by 4.7-6.5 fold (3027-4209kg/ha) for 2090 and Z. mays cultivars; GL 482, PIO 3457 orig., WASH/GRAIN-1with projected increased yields by 3.9-4.4 fold (1040-1173kg/ha) for 2090. For West Ghonghau farmers yield could improve if they plant C. esculenta cultivars; Bun long, Lehua, and Tausala-Samoa, with projected increased yields by 2.3-3 fold (3574- 4972kg/ha) for 2090. Use of Z. mays cultivars, GL 482, PIO 3457 orig., WASH/GRAIN-1 with projected increased yields by 3.6-4.4 fold (962-1173kg/ha) for 2090. For East Ghonghau, farmers yield could improve if they plant C. esculenta cultivars, Bun long, Lehua, and Tausala-Samoa, with projected increased yields by 2-2.5 fold (3666-4681kg/ha) for 2090. Use of Z. mays cultivars, GL 482, PIO 3457 orig., WASH/GRAIN-1 with projected increased yields by 3.7-4.4 fold (965-1173kg/ha) for 2090. Overall, the study demonstrates the usefulness of crop modeling tools to assess the impacts of climate change on food crops in Bellona and to make actionable recommendations to increase food security and community resilience under a changing climate.

iii

TABLE OF CONTENTS

 Acknowledgement………...………………………………………………….. i Abstract……..………………………………………………………………… ii List of tables……...…………………………………………………………… x List of figures…………………………………………………………………. xii Chapter One: Introduction………………………………………………….. 1 1.1 Background…………………………………………………………... 1 1.2 Research problem context…………………………………………..... 3 1.3 Objectives of the research……………………………..…………….. 5 1.4 Research questions…………………………………………………... 6 1.5 Methodological approach……………………………………………. 7 1.6 Structure of thesis…………………………………………………… 8 Chapter Two: Literature review…………………………………………….. 9 2.1 Climate change at the global and national (Solomon Islands) level…. 9 2.1.1 Global perspective……………………………………………. 9 2.1.2 Solomon Islands perspective…………………………………. 10 2.2 Theoretical proponents and supporting knowledge of climate change and its impacts on agriculture and food crops……………… 15 2.3 Current knowledge on the impact of climate change on food crops. 16

2.3.1 Impact of increasing carbon dioxide concentration (CO2)……. 17 2.3.2 Impact of increasing temperature on crops…………………… 20 2.3.3 Impact of interactions between increasing temperature and carbon dioxide………………………………………………... 23 2.3.4 Impact of rainfall variation……………………………………. 24 2.4 Identified impacts of extreme events on food crops at both global and national level…………………………………………………….. 25 2.4.1 Definition of extreme event…………………………………… 25 2.4.2 Observations and evidence of extreme events globally………. 26 2.4.3 Observations, evidence and impacts of extreme events nationally for Solomon Islands………………………………. 27 2.5 Projection of future climate change and extreme events…………….. 31 2.5.1 Global climate change projections……………………………. 31

iv

2.5.2 Global extreme events projections……………………………. 32 2.5.3 Solomon Islands climate change projections…………………. 35 2.5.4 Solomon Islands extreme event projections…………………... 36 2.6 Use of crop models to simulate impacts of climate change on crops………………………………………………………………….. 38 2.6.1 Rationale for using DSSAT crop models…………………….. 38 2.6.2 History of the birth and use of DSSAT……………………….. 39 2.6.3 Description of DSSAT………………………………………... 40 2.6.4 Applications of DSSAT………………………………………. 43 2.6.5 DSSAT use and application globally…………………………. 43 2.6.6 DSSAT use and application in Pacific islands………………... 44 2.6.7 DSSAT limitation and uncertainties…………………………... 45 2.6.8 Other crop models…………………………………………….. 46 2.7 Previous studies conducted in Bellona, Solomon Islands…………… 50 Chapter Three: Methods and tools………………………………………….. 53 3.1 Introduction………………………………………………………….. 53 3.2 Background information on study site………………………………. 53 3.2.1 Geographic location…………………………………………... 53 3.2.2 Social, cultural, political structure and demography…………. 54 3.2.3 Geology……………………………………………………….. 56 3.2.4 Climate………………………………………………………... 59 3.2.5 Land use………………………………………………………. 60 3.2.6 Cultivated food crops and cropping systems…………………. 61 3.3 Methods and tools used for collecting and analyzing data and information on Bellonese perspective on climate change and extreme events……………………………………………………….. 63 3.3.1 Household survey……………………………………………... 63 3.3.2 Focus group interview………………………………………… 66 3.3.3 Data analysis of household survey and focus group interview.. 68 3.4 Methods and tools used in DSSAT data collection, input, analysis and simulations………………………………………………………. 68 3.4.1 Climate data collection………………………………………... 68 3.4.2 ENSO (El Niño) year’s weather data………………………...... 69

v

3.4.3 Soil data……………………………………………………….. 69 3.4.4 Crop management data……………………………………….. 71 3.4.5 Input of climate data into Weatherman tool………………….. 72 3.4.6 Input soil data into Soil Build file (SBuild)…………………... 73 3.4.7 Input crop management data into XBuild tool……………….. 73 3.4.8 Calibration of genetic co-efficients of cultivars of taro, cassava and corn……………………………………………… 75 3.4.9 Local varieties of taro, cassava and corn used in DSSAT simulation…………………………………………………….. 78 3.4.10 Simulation and evaluation of impacts of future climate change……………………………………………………….. 79 3.4.10.1 Summary of PCCSP projection used………………… 79 3.4.10.2 Simulations and treatments…………………………... 80 Chapter Four: Vulnerability of households, food crops and cropping system to climate change and extreme events…………… 82 4.1 Introduction…………………………………………………………... 82 4.2 Household characteristics……………………………………………. 83 4.2.1 Housing structure, available service and assets owned………. 85 4.2.2 Sources of income……………………………………………. 90 4.3 Food crops and cropping system…………………………………….. 93 Chapter Five: Bellonese perception on the impacts of climate change on food crops and cropping system…………………………… 100 5.1 Introduction…………………………………………………………... 100 5.2 Past and current change in climate and correlating impacts………… 101 5.3 Past and current change in extreme events (cyclones and ) and correlating impacts……………………………………………… 111 5.3.1 Observation on the trend of cyclones and droughts………….. 111 5.3.2 Impact of cyclones on crops………………………………….. 113 5.3.3 Impact of droughts on crops………………………………….. 117 Chapter Six: Bellonese adaptation and coping strategies to climate change and extreme events…………………………………………… 120 6.1 Introduction………………………………………………………….. 120 6.2 Adaptation strategies………………………………………………… 120

vi

6.2.1 The use of shifting cultivation and fallow practice…………… 120 6.2.2 Crop rotation, intercropping and diversification of crops grown………………………………………………………… 122 6.2.3 Maintaining shade trees after clearing, burning and mulching……………………………………………………… 123 6.2.4 Planting fast yielding and resilient crops……………………... 124 6.2.5 Adjust planting dates…………………………………………. 125 6.2.6 Increase number of plots, change planting site and increase mound size…………………………………………………….. 126 6.2.7. Planting distribution and method…………………………….. 127 6.3 Coping strategies…………………………………………………….. 127 6.3.1 Pruning of crops……………………………………………… 127 6.3.2 Using common table salt as a pest deterrent………………….. 128 6.3.3 Gathering and consumption of wild crops and harvest conservation…………………………………………………... 128 6.3.4 Change in eating habits……………………………………….. 129 6.3.5 Barter system…………………………………………………. 129 6.3.6 Relief from external sources …………………………………. 130 6.3.7 Migration, remittance, selling of handicrafts and family support………………………………………………... 130 6.3.8 Traditional warning signs to extreme events…………………. 131 Chapter Seven: Simulating and evaluating future impacts of climate change on taro, cassava and corn……………………………… 132 7.1 Introduction…………………………………………………………... 132 7.2 Simulation results for taro-Tango Sua (Colocasia esculenta) yield…. 133 7.2.1 Potential production and attainable yields for ambient (2012).. 133 7.2.2 Ambient (2012) attainable versus projected attainable for temperature, rainfall and carbon dioxide…………………….. 138 7.2.2.1 Yield projection under temperature and rainfall simulations…………………………………………… 138

7.2.2.2 Yield projection under temperature, rainfall and CO2 simulations…………………………………………… 140 7.3 Simulation results for cassava –Lioka B1 (Manihot esculenta) yield… 142

vii

7.3.1 Ambient (2012) potential production versus ambient (2012) attainable yield for cassava…………………………………... 142 7.3.2 Ambient (2012) attainable versus projected attainable yields for temperature, rainfall and carbon dioxide………………… 145 7.4 Simulation results for corn (Koni)-Zea mays yield………………….. 148 7.4.1 Ambient (2012) potential versus ambient (2012) attainable yield for Koni……………………………………………….... 148 7.4.2 Ambient (2012) attainable versus projected attainable yields for Koni under temperature, rainfall and carbon dioxide…….. 150 7.5 Overview of the projected yield in Tango Sua, Lioka B1 and Koni..... 151 7.6 Simulation of yields of taro (Tango Sua), cassava (Lioka B1) and corn (Koni) during El Niño years…………………………………… 154 7.7 Simulating ENSO impacts (El Niño1997 drought) versus Bellonese perspective on taro-Tango Sua………………………………………. 158 7.8 Errors and uncertainties associated with simulations……...... 166 Chapter Eight: Recommended adaptation measures...... 168 8.1 Use of DSSAT to improve crop production………………………….. 168 8.1.1 Selection of best cultivars to improve yield in 2030, 2055 and 2090………………………………………………………….. 168 8.1.2 Change of cultivars to improve yield during ambient and ENSO (El Niño) condition of taro and corn…………………. 173 8.2 Maximise use of mulch on Malanga soil to improve soil fertility…… 178 8.3 Use of legume plants as cover crops to improve soil fertility and crop production………………………………………………………. 178 8.4 Use of sea grass and seaweed to improve soil fertility and control Pest/Disease………………………………………………………….. 178 8.5 Cultivate corn on Malanga soil and taro on Kenge toaha/ungi……… 179 8.6 Use of cultural and biological control to manage taro beetle………... 180 8.7 Maximise cultivation of cassava since it is more resilient to extreme events………………………………………………………………... 180 8.8 Increase awareness, education and improve decision making of farmers………………………………………………………………. 181

viii

8.9 Some current Bellonese adaptation practices or measures that are encouraged and should be maintained………………………………. 181 8.10 Weather station in Bellona………………………………………….. 182 Chapter Nine: Conclusions, limitations and recommendations...... 183 9.1 Conclusion…………………………………………………………… 183 9.2 Limitation of study…………………………………………………... 188 9.3 Recommendations for further studies………………………………... 188 Bibliography...... 189 Appendix A: Descriptions of Bellonese indigenous soils…………………... 200 Appendix B: Common crops cultivated in Bellona………………………... 201 Appendix C: Household survey questionnaire…………………………….. 202 Appendix D: Focus group & key informant interview guide……………… 210 Appendix E: Data sheet for collecting crop management data…………… 217 Appendix F: Definition of genetic coefficients……………………………... 218 Appendix G: List of Climate models used in PCCSP projection…………. 219

ix

LIST OF TABLES

Table 1: List of major cyclones affecting Solomon Islands…………………… 28 Table 2: Summary of minimum data set for DSSAT model application……… 42 Table 3: Crops that are normally intercropped together in Bellona…………… 63 Table 4: Sampling size of household survey………………………………….. 64 Table 5: Sections of the questionnaire and the type of data and information collected………………………………………………… 65 Table 6: Questions used in the focus group interview guide………………….. 67 Table 7: Sources of soil information and data………………………………… 70 Table 8: Summary of crop management data…………………………………. 74 Table 9: Calibrated genetic coefficients of each crop……………………….. 77 Table 10: Summary of projection extracted from PCCSP report……………... 80 Table 11: Household characteristics of 59 households surveyed in Matangi, East Ghonghau, West Ghonghau and Sa’aiho……………………… 83 Table 12: Housing type, available amenities and asset ownership……………. 86 Table 13: 45 year averages of number of garden plots per households in major wards on Bellona island……………………………………... 97 Table 14: 45 year averages for cropping fallow in Bellona island……………. 97 Table 15: Household observations of changes in temperature and rainfall…… 101 Table 16: Impact of temperature and rainfall change on crop aspects………… 102 Table 17: Bellonese planting calendar………………………………………… 125 Table 18: Simulation results for Tango Sua yields for ambient and projected changes in temperature, rainfall and carbon dioxide……………….. 134 Table 19: Physical and chemical characteristics of Kenge toaha (KT), Kenge ungi (KU) and Malanga (MA)………………………………. 134 Table 20: Simulation results for Lioka-B1 yields (dry weight) for ambient

(2012) and projected changes in temperature, rainfall, and CO2….... 143 Table 21: Simulation results for Koni yields (dry weight) for ambient

(2012) and projected changes in temperature, rainfall, and CO2…… 148 Table 22: Growth stages and stress of Tango Sua under ambient and 1997 drought in West Ghonghau…………………………………………. 160 Table 23: Growth stages and stress of Tango Sua under ambient and 1997

x

drought in East Ghonghau…………………………………………… 160 Table 24: Growth stages and Stress of Tango Sua under ambient and 1997 drought in Sa’aiho………………………………………………….. 161 Table 25: Maturity days, weight per corm and survival density of Tango Sua in three sites………………………………………………………… 165

xi

LIST OF FIGURES

Figure 1: Methodological approach used…………………………………….... 7 Figure 2: Observed global mean temperature…………………………………. 9 Figure 3: Annual mean temperature for Honiara 1950-2009………………….. 12 Figure 4: Temperature trend for Auki station…………………………………. 12 Figure 5: Temperature trend for Henderson station…………………………… 13 Figure 6: Annual rainfall trend for Honiara…………………………………… 14 Figure 7: Rainfall trend for 7 meteorological stations………………………… 14 Figure 8: TOPEX/Poseidon satellite altimetry data indicating average rate of global sea level rise (1993-2010)………………………………… 29 Figure 9: Various versions of DSSAT cropping system model……………….. 40 Figure 10: Diagrammatic representation of DSSAT components and modular structure…………………………………………………………… 41 Figure 11: Geographic location of Bellona raised atoll……………………….. 54 Figure 12: Population map by wards in Bellona……………………………… 56 Figure 13: Landform and phosphate deposit on Bellona……………………… 58 Figure 14: Cross section of Bellona landform and soil types…………………. 59 Figure 15: Land use change on Bellona between 1966 and 2006…………….. 60 Figure 16: Intercropping of taro and banana in West Ghonghau……………… 63 Figure 17: GPS location of households that were surveyed in Bellona……….. 64 Figure 18: Formatted climate raw data in excel spreadsheet………………….. 73 Figure 19: Past and present household sources of income…………………….. 90 Figure 20: Traditional woven baskets in Matangi…………………………….. 91 Figure 21: Informal labour activity in West Ghonghau……………………….. 93 Figure 22: Percentage of food crops cultivated amongst the 59 households within the 4 wards of Bellona……………………………. ……….. 94 Figure 23: Mean percentage of food crops cultivated amongst the 59 households in Bellona……………………………………………… 95 Figure 24: Responses of the five focus groups expressed as percentage on the impacts of increase temperature trend on food crops……… 103 Figure 25: Responses of the five focus groups expressed as percentage on the impacts of increase rainfall trend on food crops…………… 106

xii

Figure 26: Tunnels created by Bongu…………………………………………. 108 Figure 27: Large cavities caused by Bongu…………………………………… 108 Figure 28: Wilting of young taro crop………………………………………… 108 Figure 29: Completely eaten root and corm by Bongu……………………….. 108 Figure 30: Timeline of cyclones and droughts occurrence over past 30 years... 112 Figure 31: Percentage of respondents of the five groups to the impacts of cyclones on their food crops………………………………………. 113 Figure 32: Supporting stakes used for pana and yam in East Ghonghau……… 114 Figure 33: Lengalenga dish……………………………………………………. 116 Figure 34: Pota preparation…………………………………………………… 116 Figure 35: Percentage of respondents of the five groups to the impacts of droughts on their food crops……………………………………. 118 Figure 36: Percentage of households engaging in fallow, crop rotation and intercropping practice……………………………………………… 121 Figure 37: Sequence of crop rotation in Bellona……………………………… 122 Figure 38: Taro planted under trees with rich mulch of dead leaves and twigs.. 124 Figure 39: Newly planted Trimanisi sweet potatoes in East Ghonghau………. 125 Figure 40: An Australian White Ibis pictured in Tangakitonga, West Ghonghau………………………………………………………….. 131 Figure 41: LAI for Tango Sua and nitrogen leaching at West Ghonghau, East Ghonghau and Sa’aiho…………………………………………….. 136 Figure 42: Leaf area index for Tango Sua and total soil nitrogen at West Ghonghau, East Ghonghau and Sa’aiho…………………….. 136 Figure 43: Leaf area index and yields for Tango Sua at West Ghonghau, East Ghonghau and Sa’aiho…………………………………………….. 137 Figure 44: Projected and ambient (2012) yields of Tango Sua in West Ghonghau………………………………………………………….. 138 Figure 45: Projected and ambient (2012) yields of Tango Sua in East Ghonghau………………………………………………………….. 139 Figure 46: Projected and ambient (2012) yields of Tango Sua in Sa’aiho…….. 139 Figure 47: Potential and attainable yields of Lioka B1 in West/East Ghonghau and Sa’aiho…………………………………………….. 144

xiii

Figure 48: LAI of Lioka B1 and nitrogen leaching in West/East Ghonghau and Sa’aiho………………………………………………………… 144 Figure 49: Ambient (2012) and projected attainable yields for Lioka B1 in East Ghonghau…………………………………………………….. 146 Figure 50: Ambient (2012) and projected attainable yields for Lioka B1 in West Ghonghau……………………………………………………. 146 Figure 51: Ambient (2012) and projected attainable yields for Lioka B1 in Sa’aiho……………………………………………………………... 146 Figure 52: Excess water stress and LAI of Koni in West/East Ghonghau and Sa’aiho…………………………………………………………….. 149 Figure 53: Total soil water and nitrogen uptake in East/West Ghonghau and Sa’aiho………………………………………………………… 150 Figure 54: Projected trend of taro (Tango Sua) yield in three wards in Bellona…………………………………………………………….. 153 Figure 55: Projected trend of cassava (Lioka B1) yield in three wards in Bellona…………………………………………………………….. 153 Figure 56: Projected trend of cassava (Lioka B1) yield in three wards in Bellona…………………………………………………………….. 153 Figure 57: Simulated yields of Tango Sua for the past 11 El Niño Years……… 154 Figure 58: Simulated yields of Lioka B1 for the past 11 El Niño Years………. 155 Figure 59: Simulated yields of Lioka B1 for the past 11 El Niño Years………. 156 Figure 60: Simulated yields of Koni for the past 11 El Niño Years…………… 157 Figure 61: Simulated nitrogen uptake and water stress during the 1992 El Niño…………………………………………………………….. 157 Figure 62: Three main taro development stages………………………………. 159 Figure 63: Yield of Tango Sua under drought and ambient condition in West Ghonghau………………………………………………………….. 162 Figure 64: Yield of Tango Sua under drought and ambient condition in East Ghonghau………………………………………………………….. 162 Figure 65: Yield of Tango Sua under drought and ambient condition in Sa’aiho…………………………………………………………….. 163 Figure 66: LAI for Tango Sua under drought and ambient condition in West Ghonghau………………………………………………………….. 164

xiv

Figure 67: LAI for Tango Sua under drought and ambient condition in East Ghonghau………………………………………………………….. 164 Figure 68: LAI for Tango Sua under drought and ambient condition in Sa’aiho…………………………………………………………….. 164 Figure 69: Simulation yields at ambient (2012) of best 3 recommended taro cultivars……………………………………………………………. 169 Figure 70: Simulation yields for 2030 of best 3 recommended taro cultivars… 169 Figure 71: Simulation yields for 2055 of best 3 recommended taro cultivars… 170 Figure 72: Simulation yields for 2090 of best 3 recommended taro cultivars… 170 Figure 73: Simulation yields at ambient (2012) of best 3 recommended corn cultivars……………………………………………………………. 171 Figure 74: Simulation yields for 2030 of best 3 corn cultivars………………... 172 Figure 75: Simulation yields for 2055 of best 3 corn cultivars………………… 172 Figure 76: Simulation yields for 2090 of best 3 corn cultivars………………... 173 Figure 77: Recommended 3 best taro cultivars versus Tango Sua simulated under El Niño Years in West Ghonghau…………………………… 174 Figure 78: Recommended 3 best taro cultivars versus Tango Sua simulated under El Niño Years in East Ghonghau……………………………. 174 Figure 79: Recommended 3 best taro cultivars versus Tango Sua simulated under El Niño Years in Sa’aiho……………………………………. 174 Figure 80: Yield of Koni in normal planting date (June 15th) and adjusted planting date (May 15th) in West Ghonghau……………… 175 Figure 81: Yield of Koni in normal planting date (June 15th) and adjusted planting date (May 15th) in East Ghonghau……………… 175 Figure 82: Yield of Koni in normal planting date (June 15th) and adjusted planting date (May 15th) in Sa’aiho………………………. 176 Figure 83: Recommended 3 best corn cultivars versus Koni simulated under El Niño Years in West Ghonghau………………………………….. 177 Figure 84: Recommended 3 best corn cultivars versus Koni simulated under El Niño years in East Ghonghau…………………………………… 177 Figure 85: Recommended 3 best corn cultivars versus Koni simulated under El Niño years in Sa’aiho…………………………………………… 177

xv

Photo Source Disclaimer

All photographs were produced by the author except where due acknowledgement or reference is made.

xvi

CHAPTER ONE: INTRODUCTION

1.1 Background

Climate has a fundamental influence on almost all forms of life, including food crops.

Food crops depend on adequate climate variables such as rainfall, temperature, humidity and other environmental (soil, water availability and topography) and external factors (invasive species, pest and disease) for growth, development and production (Wairiu et al., 2012). Therefore, any changes or shift in the climate variables will affect the overall growth, development and the production of the food crops. It is projected that by 2100, temperature will rise by about 2-6oC, while rainfall is predicted to increase in the tropics and higher latitudes, but declines in the already dry semi-arid to arid mid-latitudes as well as in the interior of large continents

(Turral et al., 2011; IPCC, 2007d). Thus, the evidence of climate change observed is already without doubt and the future trends in Carbon dioxide (CO2), temperature including sea level rise conforms to the upper limit of the models stipulated in the

IPCC AR4. According to Turral et al., (2011) with temperature rises, the rate of photosynthesis also increases but declines after it reaches a maximum, whilst respiration rate continues to rise to the point that a plant dies. This implies that should a temperature exceed an optimum, the plant or crop productivity will decline. Increase in temperature will also affect the growing season of crops, for instance, in the northern temperature regions, such growing season will be longer while elsewhere it will be reduced (Turral et al., 2011). Water yield and productivity will also decline due to increased rate of evapotranspiration as a result of increasing temperature. Changes in rainfall will also affect water availability for plant growth and production. For

1 instance reduction in rainfall predicted for already dry regions of the mid latitudes and continental interior (e.g. Africa) will heavily impacted on their crop yield and production (Turral et al., 2011).

Though the effects of climate change will be felt almost everywhere on the planet,

Mimura et al., (2007, p. 689) clearly stated with high confidence that the “Small islands, whether located in the tropics or higher latitudes, have characteristics which make them especially vulnerable to the effects of climate change, sea-level rise, and extreme events”. These characteristics include their small area size, remoteness, isolation, location, exposure to extreme events or natural hazards, high or increasing population, low adaptive capacity, and poor infrastructure and governance (Barnett and Adger, 2003). Mimura et al., (2007) further substantiated with high confidence that climate change will adversely affect the subsistence and commercial agriculture on small islands. They stated that coastal agriculture is very likely to be harmfully impacted by sea level rise causing inundation, intrusion and soil salinization whilst extreme events such as flooding and drought will negatively affect inland agriculture.

However, as a consequence of climate change, the highest reductions in agriculture potential will be felt by the small Pacific Island countries amongst other developing countries (ADB, 2009). Nevertheless, it is the atoll islands and their inhabitants that are of great risk as climate change, and extreme events are potentially undermining their national sovereignty (Barnett and Adger, 2003). Livelihood and food security in many of the Pacific island countries is largely dependent on their natural resources, and any climate extreme or sudden onset of any climate related anomalies directly affects their livelihoods. Low-lying atoll countries such as , Kiribati and

Marshall islands are already experiencing the impacts of climate change and extreme

2 events, particularly on their food crops and food security, water as well as coastal fisheries resources. In Marshall Islands during the El Niño season of 1997–1998, there was a significant reduction in most crop yields (FAO, 2008).

1.2 Research problem context

Solomon Islands, though made up of higher volcanic islands, also share common problems with atoll countries because it has both atoll and low lying islands.

Food security is under threat by climate change and for Solomon Islands it is the most urgent and pressing issue. The Solomon Islands Government in 2008 and the Solomon

Islands National Adaptation Program of Action (SINAPA) highlighted climate change as a priority issue that required urgent action (Talo, 2008). Many communities on low-lying atolls are now experiencing climate change impacts through sea level rise, salt water intrusion and extreme events (cyclones, droughts etc.) which affected the main staple food crops. One of such atoll communities is Bellona, located on the southern side of Solomon Islands. Although this atoll is classified as “raised atoll”,

Bellona community have already experienced climate risks such as low rainfall and frequent cyclones which affected their food security and livelihoods (PACC, 2006;

Rasmussen et al., 2009). Drought events normally occur even with short periods of no rain; nonetheless, the high exposure to cyclones is a key climatic hazard to the people’s livelihoods (Reenberg et al., 2008).

The Bellona community lacks adequate infrastructure, making service delivery by both the national and provincial government difficult. Moreover, they face food security uncertainties in the future since they’ve already experienced a series of

3 extreme events related to climate change. One of the main concerns for the communities of Bellona is how climate change is affecting food security. The atoll is frequently experiencing widespread damages to its socio-economic infrastructure and food crops caused by tropical cyclones as it is situated in the path of cyclone. For example, in the past five decades, four major cyclones have directly impacted the atoll, severely causing extensive damages to crops, infrastructure (water tanks) and rural livelihoods (Reenberg et al., 2008). Food production problems and security issues have also increased as a result of cyclonic events. In many cases, periods of drought normally follow after cyclones (Reenberg et al., 2008).

Food productions on the atolls are affected by extreme events such as cyclones, droughts, as well as changes in rainfall patterns and temperatures. These parameters should be evaluated in terms of food crop production, which may highly exacerbate the negative impacts on crop growth development and productivity. To predict such impact of future climate change on crop growth, crop simulation models such as

Decision Support System for Agro-technology Transfer (DSSAT) can be used. This tool uses data from soil, crops, management inputs, atmosphere and weather under cropping system and simulates growth, development and yield over time (Jones et al.,

2003).

For this proposed study, two main methodological approaches were used. The first approach follows the theoretical and practical postulates of Rasmussen et al., (2009) and Reenberg et al., (2008) which involved a survey of production systems, household questionnaires and key informant and group interviews on the atoll community of

Bellona. Such approach will obtain perceptions of the local population in terms of

4 impacts of climate change and extreme events on food security. This study builds on the findings of Rasmussen et al., (2009) and Reenberg et al., (2008) conducted on the same atoll community about 6 years (2006-2007) ago and indicates the current information on vulnerability and impacts on food security systems as well show changes or trends that have occurred. The second approach focused on the use of

DSSAT crop model version 4.5 which incorporate climate data (daily maximum/minimum temperature, rainfall, and sunshine hours), soil chemical and physical profile information as well as crop management as input (Jones et al., 2003).

Simulations were generated for the impact of future change of temperature, rainfall and carbon dioxide on the taro (Colocasia esculenta), cassava (Manihot esculenta) and corn (Zea mays).

1.3 Objectives of the research

The three objectives of the study are as follows:-

(i) To understand and identify vulnerability, impacts and adaptations to climate

change and extreme events on food crops and cropping systems on the atoll

community of Bellona (Solomon Islands),

(ii) To simulate and evaluate the impacts of future change of climate:-rainfall,

temperature, carbon dioxide on taro (Colocasia esculenta), cassava (Manihot

esculenta) and corn (Zea mays) yield by using DSSAT crop simulation model,

(iii) To identify and recommend relevant adaptation strategies, options or

measures to improve food crop production and management.

5

1.4 Research questions

The subsequent research questions directed the research process.

(i) How vulnerable are food crops, cropping system/management and human and

social structures to climate change and extreme events?

(ii) What are past and current impacts of climate change and extreme events on

food crops and cropping system in Bellona?

(iii) In what ways have the Bellonese community/farmers responded to the

impacts of climate change and extreme events on food crops including

cropping management system in terms of both adaptation and coping

strategies?

(iv) What are the expected impacts of future climate change on (taro, cassava and

corn) crop yields?

(v) What crop management options will be used to essentially maximize yield,

crop production and management during stressful events of future climate

scenarios and ENSO (El Niño) events?

6

1.5 Methodological approach

Bellonese perspective on Climate Modelling Impact of climate change on food

A Data collection B

Household Focus group Crop Soil dataClimate DataClimate data Survey interviews management -physical/-30yrs rainfall(30 yrs) -25% sampling -different age data chemical-30yrs temperature-Rainfall, Temp groups/gender, -planting date properties-30 yrs SunshineSunshine hours hours -spacing etc

DSSAT Crop modelling Data analysis

Environment Output Sensitivity modification tool Analysis tool -PCCSP future climate scenario Vulnerability Impact Adaptation Optimize crop management and PlantGro & Summary maximize yield Past & current vulnerabilities, impacts and adaptation measures Predicts Future Impact ild

Using the results from Bellonese perspectives (A) together with results from DSSAT crop

modelling (B) , identification and recommendation of relevant (integrated) climate change related,

adaptation strategies, options or measures to improve food crop production and management were

made.

Figure 1: Methodological approach used

7

1.6 Structure of thesis

There are a total of nine chapters in this thesis.

Chapter one: Provides an introductory to the research work. It outlines the background information, context of the research problem, objectives and the research questions that guided the research process.

Chapter two: Reviews the relevant literature on climate change and extreme events related to agriculture, food crops and cropping systems and provides a brief overview of climate change on the global and national scale. This chapter also describes the use of DSSAT cropping system tool and other relevant crop models.

Chapter three: Outlines the study area as well as the research methods, tools and techniques used to collect and analyse the data and information.

Chapter four: Provides the vulnerability of households, food crops and cropping system to climate change and extreme events.

Chapter five: Presents the Bellonese perception on the impact of climate change and extreme events (cyclones and droughts) on food crops and cropping system.

Chapter six: Outlines an overview of the Bellonese adaptation and coping strategies to climate change and extreme events

Chapter seven: Presents the results of the model simulations of the impacts of El

Niño and future climate change on taro (C. esculenta), cassava (M. esculenta) and corn (Z. mays).

Chapter eight: Provides recommended adaptation measures that may be used by the

Bellonese communities to improve crop management and production.

Chapter nine: Concludes and summarise the main findings of the research study, limitations and recommendations for further studies.

8

CHAPTER TWO: LITERATURE REVIEW

2.1 Climate change at the global and national (Solomon Islands) level

2.1.1 Global perspective

2.1.1.1 Temperature change

Evidence by means of direct observation in global average air and ocean temperatures has clearly indicated that the climate system is warming (IPCC, 2007c). Such warming has been attributed to the observation of increasing global sea level rise and also widespread melting of snow and ice (IPCC, 2007c). Over the past 100 years (1906 to

2005) the global mean surface temperature (estimating linear trends), has increased by about 0.4 ± 0.18°C (Figure 2).

Figure 2: Observed global mean temperature (Source: IPCC, 2007a)

9

Nevertheless, the rate at which warming has been for the last 50years is nearly twice that over the 100 years period. That is over the last 100 years, the rate per decade is about 0.07°C ± 0.02°C per decade compared to 0.13°C ± 0.03°C per decade for the last

50 years. Accordingly, IPCC (2007a) reported that since 1850 when instrumental records started, the years 1995-2006 were classified as the 12 warmest years.

2.1.1.2 Precipitation and rainfall change/variation

IPCC (2007a) categorizes precipitation as increasing precipitation in high latitudes while decreasing in most subtropical regions. They further substantiated that from observations precipitation have distinctively increased in eastern parts of South and

North America, Northern Europe and Northern and Central Asia. The opposite were observed in parts of Southern Asia, Southern Africa, the Mediterranean and Sahel.

Looking at this observation and trend, it is obvious that precipitation including rainfall is not the same everywhere, it is highly variable and many other factors may influence it. For instance, topography, geographical location and even wind convergence zones like the ITCZ or SPCZ. IPCC (2007a) corroborated that precipitation is both spatially and temporally exceedingly variable.

2.1.2 Solomon Islands perspective

2.1.2.1 General climate of Solomon Islands

Solomon Islands is seasonal and has two distinct tropical climates, marked by a wet season that falls between November and April and a dry season which is normally from May to October (PCCSP, 2011b). These two seasons are mainly influenced by

10 the trade winds which are dominant and prevail within the Solomon Islands.

In terms of temperature, it is normally constant all over the year with little seasonal variation with however significant variation only occurs in July to August when cooler air prevails from the south (PCCSP, 2011b). Though the air temperatures indicated little seasonal variation they are correlated to the seas surface temperatures.

There are three main climate systems that affect rainfall distribution in Solomon

Islands namely, the West Pacific Monsoon, the South Pacific Convergence Zone and the Inter-tropical Convergence Zone. The eastern Solomon Islands experiences constant rainfall throughout the year whilst in the west, rainfall normally occurs from

November to April (PCCSP, 2011b). However, with the influence ENSO, there are significant variations of rainfall each year.

2.1.2.2 Temperature change

Using the period from 1950-2009, there is evidence of a warming annual and seasonal trend in the mean air temperature at Honiara, the capital of Solomon Islands as in Figure 3 (PCCSP, 2011b). Talo (2008) also highlighted that there is an upward warming trend for Solomon Islands. This is indicated by the warming trend of annual mean temperatures of two meteorological stations (Auki and Henderson as in Figures

4 and 5) which is stipulated to be consistent with the warming trend that is also experienced in Pacific region.

11

Figure 3: Annual mean temperature for Honiara: 1950-2009 (Source: PCCSP, 2011b)

Figure 4: Temperature trend for Auki station (Source: Talo, 2008)

12

Figure 5: Temperature trend for Henderson station (Source: Talo, 2008)

2.1.2.3 Rainfall change

Rainfall trend or change according to the recent analysis by PCCSP (2011b) for

Honiara station indicated a non-statistically significant trend (Figure 6) for the period between 1950 and 2009. However, according to Talo (2008) despite the fact that there is upward warming trend of temperature, rainfall records show the opposite, a downward trend (Figure 7).

13

Figure 6: Annual rainfall trend for Honiara (Source: PCCSP 2011b)

Figure 7: Rainfall trend for 7 meteorological stations (Source: Talo, 2008)

14

2.2 Theoretical proponents and supporting knowledge of climate change and its impacts on agriculture and food crops

There is an increasing body of scientific evidence that gives a collective picture of a warming world and that climate change is unequivocal and will have significant implications for agricultural systems, food crops and food security (IPCC, 2007d;

Turral et al., 2011). The IPCC AR4 clearly stated with very high confidence that the small islands are highly vulnerable to the adverse effects of climate change (Mimura et al., 2007). This is because small islands have characteristics such as limited size, remoteness, isolation, exposure to extreme events or natural hazards, high or increasing population, low adaptive capacity, and poor infrastructure and governance

(Barnett and Adger, 2003; IPCC, 2007d; Mimura et al., 2007). In terms of subsistence and commercial agriculture on small islands, Mimura et al., (2007) further substantiated that it is very likely that climate change will adversely affect them.

Coastal agriculture will be affected by increase in sea level, coastal flooding, salt water intrusion, salinization of soil and reduction in water supply whilst inland agricultural systems will be adversely affected by the effects of drought and flooding.

Despite the fact that there is limited historical data available for the Pacific islands, authors agree that there is a general warming trend and drier conditions are becoming more frequent within the last hundred years due to global warming (Wairiu et al.,

2012; FAO, 2008). There have been an increase from 0.8 to 1.0oC of annual and seasonal ocean surface and island air temperature since 1910 over the most part of the

South Pacific (Folland et al., 2003). Additionally, the South Pacific in comparison to previous past records during the twentieth century had experienced warmer and drier climate (Hay et al., 2003). Due to the observed rise in El Niño events since 1970, the 15

Central Equatorial Pacific is experiencing an increase of about 30% in rainfall and

0.4oC in (FAO, 2008). In terms of future climate projections, by 2100, temperature will increase between 1-3.1oC whilst sea level will increase to a range of about 0.18m-0.58m in 2090-2099 relative to 1980-1999 (IPCC, 2007d). The frequency of tropical cyclones with intense wind speeds and rainfall is also projected to increase (IPCC, 2007d).

However, according to Dixon (2009) climate change may be beneficial in the short term as is evident by a study conducted in Argentina (Magrin et al., 2005) which indicated that increased precipitation and minimum temperatures with decline in maximum temperature have benefitted the yields of wheat, maize, sunflower and soybean. Henceforth, with climate change versus agriculture, food crops and food security, there will be countries or regions that will be adversely affected whilst others will be benefitting. Typically, small islands countries will be the most affected whilst those countries located on the mid to higher latitudes who will be benefitting from climate change (FAO, 2008). Thus, though climate change is a global problem, there are both losers and winners, with losers being the small islands as well as other developing countries like Africa (Dixon, 2009).

2.3 Current knowledge on the impacts of climate change on food crops

There were several avenues by which climate change and climate variability may have impact on food crops or agriculture, specifically affecting crop growth, development and yield production (Easterling et al., 2007). Those are, increasing carbon dioxide concentration, increased temperatures, changes in precipitation patterns and frequency

16 of extreme events (Turral et al., 2011). Numerous studies have also substantiated these knowledge and findings and are discussed below.

2.3.1 Impacts of increasing carbon dioxide concentration (CO2)

2.3.1.1 Trend of CO2 concentration

IPCC (2007a) confirmed that atmospheric CO2 has significantly increased due to the consequences of human activities since the birth of the industrial revolution in 1750.

Human activities such as those of fossil fuel use and change in land use have been primarily attributed to such an increase (IPCC, 2007a). It has been found that as of

2005, the global atmospheric CO2 has reached 379ppm. This is an increase of about

99ppm since pre-industrial era which only has about 280ppm (IPCC, 2007a). Recently, the concentration of CO2 that has been recorded at Mauna Loa Observatory in Hawaii in April 2013 has reached about 398.40ppm (Tans and Keeling, 2013). Nonetheless, the question is what importance does this increasing CO2 concentration has in terms of food crops growth, development and production?

2.3.1.2 Role and importance of CO2 on crops

CO2 plays a significant function in photosynthesis; a process by which plants or crops use to produce their food from whereby CO2 is one of the key ingredients for this process. Generally, an increase of CO2 will also increase crop yield as a result of increase photosynthesis (Allen and Prasad, 2004). Thus, the amount of CO2 that is freely available for crops in the atmosphere may determine the growth, development and production of the crops among other factors. 17

It was agreed and corroborated by several studies that increase in CO2 will have positive effects on crop growth and yield however, such will rely on the type of species, management factors (fertilizer, water applications), and photosynthetic pathways

(Allen and Prasad, 2004; Ainsworth and Long, 2005; Norby et al., 2003; Kimball et al.,

2002; Jablonski et al., 2002)

2.3.1.3 Impact of increasing carbon dioxide

Increasing CO2 will enhance the photosynthesis process which in turn will improve the yields of crops (Allen and Prasad, 2004). Nevertheless, such benefits of increased CO2 on crop yield will be offset when temperatures reached a maximum (Turral et al.,

2011). This implies that should a temperature exceed an optimum, the plant or crop productivity will decline. The optimum refers to the best or most favourable temperature at which a plant or crop photosynthesizes at its best which in turn resulted in increased growth, yield or productivity. The optimum or optimal temperature is not the same for all plants or crops. For instance, for Cassava (Manihot esculenta) the optimum temperature for optimum growth and production is 25-30oC while for Maize it is between 28-37.5 oC (El-Sharkawy, 2004; Crafts-Brandner and Michael E.

Salvucci, 2002).

On the contrast, it must be noted that different plants have different photosynthetic pathways and their responses in terms of growth and yield to increased CO2 will be different. For instance, C3 crops which includes crops like grain cereals (rice and wheat), grain legumes (beans and peanuts), and tuber crops (cassava and potato) will respond distinctively to increasing CO2 concentrations (Allen and Prasad, 2004; Dixon,

18

2009). The rationale behind this is because C3 crops have an enzyme known as

Rubisco (ribulose 1,5-bisphosphate carboxylase/oxygenase) that may easily attach to

CO2 or O2 which means that an increase in CO2 will allow the enzyme to out compete dissolved O2 for binding sites therefore resulting in an increase rate of photosynthesis

(Allen and Prasad, 2004). On the contrast, C4 crops have the enzyme phosphoenolpyruvate carboxylase (PEPcase) which moderates CO2 concentrations.

Such differential responses between C3 and C4 crops are illustrated by experimental research results that indicated the yields of C3 crops increased by about 10-25% while only 0-10% for C4 crops at 550ppm atmospheric CO2 concentrations (Easterling et al.,

2007). According to Dixon (2009) other plants that have Crassulacean acid metabolisms (CAM) pathway such as those of the members of the cactus family

(Cactaceae) including crop like pineapple, might respond positively to elevated CO2 concentrations.

The impacts of increased CO2 on crops quantified in experimental conditions or models however may not accurately reflect the response of crops on real farm conditions. This is mainly due to numerous limiting factors such as soil type, water and air quality, drying effect due to increased evapotranspiration and wind, competition for resources (nutrients, space, and water), pest, diseases and weeds that are uncertain

(Tubiello et al., 2007; Tubiello and Ewert, 2002; Karnosky, 2003; Ainsworth and

Long, 2005; Gifford, 2004). This is further substantiated by Lobell and Field (2008) stating that there is considerable uncertainty in models in evaluating the impacts of increased CO2 concentrations on crop yields. They further concluded that the underlying principal rationale for such uncertainty is due to the limited experimental data available for responses of crops cultivated under actual field settings with regards

19 to increasing temperature. Thus, it is clear that the gap in literature in terms of the impact of increased CO2 concentrations on crops is that there are limited studies that included the limiting factors such as soil type, water and air quality, competition for resources, pest, diseases and weeds. Furthermore, there are also limited studies being conducted on crops response to increasing CO2 concentrations on the normal field conditions. According to Tubiello et al., (2007), there are more studies on the main cereal grains and in temperate regions. There is however, a limited number of a study or data available on other food crops especially on the developing countries in sub-tropical regions, and even more so in the Pacific Island countries.

Overall, though increased CO2 will boost photosynthesis in turn enhance crop yields as reported by controlled experiments, it is obvious that other factors have to be considered before acknowledging the beneficial aspect of elevated CO2.

Consequently, given the positive correlation between CO2 and temperature, doubling of the CO2 concentrations would mean that temperatures would also increase thereby offsetting the likely benefits (Turral et al., 2011). The benefits of elevated CO2 on food crops will also be negated by water and thermal stress linked to climate change and variability coupled with other changes like loss of soil fertility, weeds, invasive species, and erosion (Wairiu et al., 2012; Lal, 2004)

2.3.2 Impact of increasing temperature on crops

2.3.2.1 Role and importance of temperature on crops

Temperature is one of the crucial climatic variables that are important in influencing quality and quantity of crops as well as the site to which it can be cultivated.

20

According to Turral et al., (2011) with increasing temperatures, the photosynthesis rate also increases however it declines after the temperature reaches a maximum whilst respiration rate continues to rise to the point that a crop dies. In other words, should a temperature exceed an optimum, the crop growth and productivity will decline. Thus, any shift in temperature would have profound impacts on crop production (Masters et al., 2010). With certain temperature range, for instance between 10°C and 30°C for wheat, the crop development normally increases in a linear relationship however at high temperatures, overall productivity will decline as this may result in heat stress and high sterility (Masters et al., 2010).

2.3.2.2 Impact of increasing temperature

Increasing temperatures also influences the crops to require more water as a result of increased evaporation from both crops and the soil that in turn cause additional stress to the crops. Shifting in temperature also could lead to high pest reproductive rates and increasing frequency of new diseases and invasive species (Padgham, 2009).

The impact of increasing temperatures on crops will not be the same on different regions. For instance, in temperate regions (Europe) such shift in temperature would result in adequate cropping areas migrating pole ward. This pole ward shift will be beneficial for crops that have tropical origins for instance rice which are only cultivated in temperate regions during warm seasons (Masters et al., 2010).

Furthermore, the growing period for determinate crops like cereals will be reduced while increases for indeterminate crops like root crops.

21

Increased temperature is anticipated to influence positive impact on northern Europe whereby there will be increasing areas for cropping and higher crop production with introduction of new varieties. On the other hand, such shift in temperature may offer a higher risk nutrient leaching and turnover of organic matter as well as possibility of pest and disease outbreak (Masters et al., 2010). Areas of southern Europe will experience increase water shortage, lower of crop yields and diminution of adequate land for cultivating crops. Nevertheless, increase in temperature will also affect the growing season of crops, for instance, in the northern temperature regions where agricultural potential is bounded by cold temperature stress, such growing season will be longer while elsewhere it will be reduced (Turral et al., 2011; Masters et al., 2010).

The impact of increasing temperature in the tropical region is different from its impact in temperate regions. According to Masters et al., (2010), it is clear that with any temperature increase crop yield decreases, especially for crops that are originated within the tropics. This is because such tropical crops like maize and rice are already within their higher range for favourable growth (Masters et al., 2010). Other tropical areas where coffee is cultivated will experience low quality and less production as increasing temperature will lead to reduction of high quality areas. Regions that have broad variation in day-time temperatures for instance semi-arid regions will have the utmost effect even with little shift in the mean annual temperature (Masters et al.,

2010).

The IPCC (2007c) further categorizes and summarize the impact of increasing temperature on crops according to latitudinal regions. They concluded that with a local shift of local mean temperature of up to 1-3°C , mid-high latitude areas will experience

22 that crop productivity will increase to some extent. On the contrast, crop productivity will decline even with a slight local temperature increase of 1-2°C on lower latitudes, specifically on the dry and tropical regions (IPCC, 2007c). Such impact on lower latitudes will further increase the threat of food insecurity and hunger.

In the Pacific region, despite the limited availability of historical data or published literature to quantify the effects of increasing temperature over time on food crops or agriculture, there have observed impacts on agricultural crops associated with temperature increase (Wairiu et al., 2012). For instance, there was significant reduction in the tuber formation of sweet potatoes at temperatures above 34oC in

Papua New Guinea (PNG). Taro production in Solomon Islands has been reported to decline as a result of increasing temperatures (Talo, 2008). Increase in temperature does not only affect the growth and yield of crops but also influence the spread of diseases. In PNG, increase in temperature was observed to increase coffee rust infestation which has serious effect on the coffee production. Taro yields were also reduced at higher altitudes as a result of increase incidence of taro leaf blight disease

(Wairiu et al., 2012). According to FAO (2008), increase in temperature has increase pest incidence affecting yam in .

2.3.3 Impact of interactions between increasing temperature and carbon dioxide

Increasing temperature does not only directly affect the crops physiology, biology, ecology and distribution, it also influence the CO2 effects on crops. Several studies since the knowledge have been put forward by IPCC TAR have confirm this interaction between temperature and CO2 (Easterling et al., 2007). For example, it was

23 found out that increase temperature throughout the flower stage of crops might reduce the effects of CO2 in terms of reduction in grain quantity, quality and size (Thomas et al., 2003; Baker, 2004; Caldwell et al., 2005). Other indirect impact of high temperatures on CO2 effects is on increasing the demand for water by crops. A study conducted on rain-fed wheat in China illustrated that further water was required to offset the adverse impact with increasing temperature over 1.5°C at 450 ppm CO2

(Xiao et al., 2005). Though increased CO2 concentrations might be beneficial for C3 crops as compared to C4 crops as have been initially described in section 2.1 above, an increase in temperature may cause an opposite effect however there remains uncertainty with the net effects (Easterling et al., 2007).

2.3.4 Impact of rainfall variation

2.3.4.1 Importance of rainfall on crops

Rainfall is fundamental to crop growth and production. Agricultural sectors highly depend on rainfall for its water needs and irrigation demand. Globally about 100% of pasture land and over 80% of total agricultural land is dependent on rain fall

(Easterling et al., 2007). Shifts in annual precipitation normally modified ecosystem functions, especially in marginal lands or areas that have been degraded, eroded or in the process of desertification.

2.3.4.2 Impact of rainfall on crops

Decreasing rainfall compounded with increasing temperature and evaporation adversely affects water availability for crops in turn increasing the demand for

24 irrigation (Easterling et al., 2007). On the other hand, even increasing precipitation also has negative impacts. For instance, in the USA increase of extreme precipitation by 2030 would lead to a loss of US$3 billion per year as a result of high soil moisture

(Rosenzweig et al., 2002). Other studies confirmed that increase rainfall resulting in flood and soil erosion leads to high crop losses (Monirul and Mirza, 2002; Nearing et al., 2004).

The extent to which the increasing rainfall will impact rain-fed cropping systems will be different for different climatic regions. For instance, crop productivity for rain-fed farming in high latitudes such as North America and Europe will increase while those of mid-low latitudes will be affected (Turral et al., 2011). Projection for rain-fed crop yields in Africa by 2020 could be reduced by 50%.

In the Pacific region, agricultural production is mostly rain fed and thus the effect of rainfall trend and distribution is significant. However, according to Wairiu et al., (2012) the amount of rainfall requirements and tolerance is not the same for each type of crop.

They highlighted that for optimal agriculture production an annual rainfall of

1800-2500mm is essential. For example in southern mountainous region of PNG where rainfall is excessive at more than 4000mm annually, there have been noticeable lower yields of sweet potatoes (Allen and Bourke, 2009). The variability in rainfall and change in rainfall pattern in the Pacific region has also affect the normal planting date or time, harvest periods, yield storage, and reduce overall yield (Wairiu et al.,

2012). In Vanuatu, farmers highlighted that in the past 3-4 years due to increase rainfall, some of their crops produce flowers earlier than normal while early fruiting is observed in other crops (FAO, 2008). However, according to Wairiu et al., (2012), within the Pacific region, flooding and waterlogging accounts for significant impact

25 on both agricultural crops and infrastructure. For instance, in Fiji the cost associated with flooding on sugar cane production was estimated to be around US$13.4 million.

2.4 Identified impacts of extreme events on food crops at both global and national level 2.4.1 Definition of extreme event

Extreme event is an event that is rare at a particular place and time which may include floods, droughts, or heavy rainfall over a season (IPCC, 2007e; IPCC, 2007f).

However, IPCC (2012, p. 5) defined extreme event (weather or climate) as, “the occurrence of a value of a weather or climate variable above (or below) a threshold value near the upper (or lower) ends of the range of observed values of the variable”

Initially, the IPCC (2001) have stated that there will be increase frequency and intensity of extreme events and more widespread during the 21st century. This has also been supported by the IPCC (2007). Further evidence indicated that the frequency of heat waves were observed to increase in North America and Europe, while increase in extreme precipitation was observed in the United States (Allison et al., 2009). More recently there is confidence on the evidence that was obtained since

1950 that extreme events have changed (IPCC, 2012). However due to limited data available for all the different regions of the world, extreme events are referred to as rare and difficult to assess their frequency and intensity.

2.4.2 Observations and evidence of extreme events globally

There is observational evidence since 1950 that supports the change in extreme

26 events (IPCC, 2012). They have found out that globally there is a general increase in the number of warm days and nights. Such have been observed in North America,

Europe and Australia. There is also an observed increase in the number and length of warm spells and heat waves since mid-twentieth century. Though there are regional and sub regional variations in terms of heavy precipitation trends, IPCC (2012) conclude that statistically the number of heavy precipitations has increased in many regions. In terms of tropical cyclones, the movement of cyclone tracks have been observed to shift pole ward in both the Northern and Southern hemispheres.

Observations have also indicated that the intensity and length of droughts varies among the different regions. For instance, in Southern Europe and West Africa, the drought events are longer and intense as oppose to regions of Central North America and North West Australia where droughts events were shorter, less frequent and less intense. Due to limited evidence and observation in flood events, IPCC (2012) could not substantiate the change in frequency or intensity of extreme flood events. On the contrast, they stipulated that there is evidence and observation that indicated an increase in the extreme coastal high water in the late twentieth century as a have been related to the increase in mean sea level.

Observed impacts and evidence of extreme events as summarize by IPCC AR4, includes reduction in the length of growing season with harmful effects on crops during warmer and drier conditions (IPCC, 2007c).

2.4.3 Observations, evidence and impacts of extreme events nationally for Solomon Islands 2.4.3.1 Cyclones

Solomon Islands is highly vulnerable to extreme events and experiences severe

27 tropical cyclones and long dry spells associated with ENSO (Talo, 2008). In

Solomon Islands, occurrence falls between November and April. An average of about 10 cyclones per decade (1 per year) and are more frequent with El

Niño years with about 13 per decade (PCCSP, 2011b). Cyclones affect both commercial and subsistence crops in Solomon Islands. As an example, rice and palm oil production have been critically affected by in 1986 in Solomon

Islands, leaving the country heavily dependent on imported of rice since then (Talo,

2008). The total cost of the damages on both agricultural crops and infrastructure was around $US100 million (SIG, 2012). Apart from the immediate impacts of cyclone damage, there is also increase outbreak of pest and disease subsequent to physical damage that affects newly grown crops. In 2002, (category 5) severely devastated two most remote islands in Temotu province leaving behind tons of debris and severe damages to agriculture productivity, food and water supply (SIG,

2012; Talo, 2008). Cyclone Beni in 2006 also caused severe damages to agriculture and food garden incurring a cost of about $US0.5 million (Table 1).

Table 1: List of major cyclones affecting Solomon Islands (Source: SIG, 2012)

28

2.4.3.2 Extreme sea levels

According to satellite altimetry data (Figure 8), Solomon Islands is experiencing much higher sea level rise at a rate of about 8-10mm per year (Hoegh-Guldberg and

Bruno, 2010). Evidence of such increase in sea level has been mostly experienced by low lying outer islands of Ontong Java and Reef islands in north and eastern

Solomons (SIG, 2012). However, according to PCCSP (2012b), ENSO has significant influence on the sea levels in Solomon Islands. They mentioned that during La Niña sea levels are normally higher by about 0.1m and lower by about the same height during El Niño.

Solomon Islands

Figure 8: TOPEX/Poseidon satellite altimetry data indicating average rate of global sea level rise (1993-2010). Insert: Solomon Islands is located in the extreme sea level rise zone (Source: Hoegh-Guldberg and Bruno, 2010)

A king tide in 2008 caused extensive damage to coastal infrastructure, water sources and food gardens in northern Choiseul, Ontong Java and other parts of Solomon 29

Islands (SIG, 2012). stands, taro plots, household and cultural properties of coastal communities were lost due to this extraordinary tide that came in the form of high swells. Coastal erosion and loss of coastal land areas have been experienced and currently these areas are exposed and vulnerable to the effects of the king tides (SIG,

2012).

2.4.3.3 Droughts

Droughts that have affected many parts of Solomon Islands are closely linked with El

Niño (PCCSP, 2011b). Since most food crops system in Solomon Islands are rain-fed, prolong dry conditions have adverse effects on food crop growth, production supply.

For instance a severe drought that affected the country during the 1997/98 El Niño episode resulted in many food gardens severely affected leading to food shortages; which prompted distribution of supplies by the government and donors (SIG, 2012).

The severe drought in 2004 caused food and water shortages in Temotu province located on the eastern Solomons.

2.4.3.4 Extreme rainfall

SIG (2012) reported that extreme, intense and prolonged rainfall was a major problem in the Solomon Islands, affecting different sectors and ecosystems. Massive flooding in 2008 caused by extreme rainfall resulted in severe flooding in , Central

Islands, Western, and . In 2010, a first ever highest daily rainfall of

251.8mm recorded since its recording started, cost nine lives on Guadalcanal and massive damage to agriculture (food crops, garden areas), infrastructure and health

30 amounted to millions of dollars (SIG, 2012). Root crop such as potato is highly vulnerable to extreme rainfall. It was reported that the yield of sweet potato which is currently the main staple food in most rural areas have been declining which is believed to be associated with extreme rainfall (SIG, 2012). Vegetative growth was observed to be higher while tuber growth declines. Such effects were noticeable on windward areas of larger islands.

2.5 Projection of future climate change and extreme events

2.5.1 Global climate change projections

2.5.1.1 Projected temperature

In terms of projected temperatures, the IPCC (2007a) used 6 emission scenarios to outline what the warming will be like in the twenty first century based on the

1980-1999 temperature recordings. The 6 scenarios are based on different levels of emissions predicted to be occurring in the future based on socio economic, population, technology advance, and behavioral parameters of the countries in the globe. For the lowest emission scenario known as B1, the best likely estimate is 1.8oC which lies between the projected ranges of 1.1-2.9oC. The intermediate scenario A1B has a best estimate of 2.6oC which ranges between 1.7-4.4oC. On the contrast, the highest emission scenario A1F1 has likely ranges of 2.4-6.4oC with a best estimate of 4oC.

Thus, globally by 2100, the temperature is projected to be in temperature ranges of

1.1oC to 6.4oC or in the best estimate of 0.6oC to 4.0oC.

31

2.5.1.2 Projected rainfall

With improved abilities to make more precise global projections since the IPCC TAR a better understanding of precipitation patterns have also improved. IPCC (2007c) stated that there is “very likely” to be increasing precipitation in high latitude regions whilst in most subtropical land regions there is “likely” to be a decline. More so, there will be an increase in precipitation in higher latitudes and decrease in lower latitudes therefore average river runoff will be higher in high latitude and lower in mid-lower latitude regions (Bates et al., 2008).

2.5.2 Global extreme events projections

IPCC (2012) highlighted that by the end of the twentieth century; there will be substantial warming in temperature extremes. They concluded with certainty that the frequency and magnitude of temperature extremes will increase as well as the length, frequency and intensity of warm spells or heat waves.

In terms of precipitation, many areas of the globe would experience heavy precipitation to increase in frequency by twentieth century. Such regions include those of the tropical regions and high latitudes. Globally, cyclone frequency will either be the same or decrease, however the intensity of its wind speed is projected to increase but not on all ocean basins (IPCC, 2012).

On the other hand, drought events will increase in intensity in the twentieth century in some areas in some seasons as a result of decline in precipitation and increase in

32 evapotranspiration (IPCC, 2012).These areas are Southern and Central Europe,

Central America, Central North America, Mexico, North East Brazil, Southern

Africa and the Mediterranean region. The areas affected by drought events are projected to likely increase (IPCC, 2007c). Expected impact of drought will be on land degradation meaning there will be reduction of crop yields, physical damage to crops resulting in crop failure and the risk of wild fire.

Though there are projections of changes in temperature and precipitation, changes in flood events for the future is uncertain due to limited evidence and due to the fact that regional changes are complex (IPCC, 2012).

Extreme coastal high water will increase in the future as a result of increase in mean sea level rise. Thus, coastal areas those currently experiencing effects of inundation, coastal erosion and salinization are likely to continue experience those effects in the future.

Projections by IPCC (2012) for large scale natural climate variability such as ENSO

(El Niño Southern Oscillation) remains uncertain as there is low confidence in projection modelling and the frequency of such is inconsistent.

Generally, with such observed evidence and projections of extreme events by IPCC

(2012) the next question is what implications such changes would have on food crops.

According to IPCC (2007c) local crop productivity mainly in subsistence areas at low latitudes is projected to adversely be affected by the increase frequency of droughts and floods. They further projected that with changes on more frequent extreme hot days as well as nights, yields of crops will change in both colder and

33 warmer environments. For colder environments, the yield of crops will increase while in warmer environments the yields will decline. With the decline in yield, there is also the likelihood of the outbreaks of insects which may further aggravate the reduction in yields. Further decline in yields as a result of heat stress and fire hazard is also projected for warmer environments as projected increase in the frequency of warm spells and heat waves is very likely.

Besides a projected increase in warm spells, the frequency of heavy precipitation will increase over most land areas which in turn will induce damage to crops (IPCC,

2007c). Such damage will be likely caused by increase in soil erosion and flooding events which may cause physical uprooting of crops. Soil may also be unsuitable for crops as there may be water logging on flat areas which may cause root rot or the inability to cultivate the land.

In terms of tropical cyclone, the intensity is likely to increase and will be expected to cause physical damage to crops by wind throw and uprooting of crops and tree crops

(IPCC, 2007c).

The incidence of extreme sea level is projected to likely increase; however this excludes tsunamis, and will have impact on food crops due to salinization of cultivated land and ground water (IPCC, 2007c; Easterling et al., 2007)

Overall, Easterling et al., (2007) summarizes that in addition to the impacts of climate change the projected changes in both the frequency and severity of extreme events will have significant effects on food crops and food security. Increase frequency of extreme

34 events like heat stress, droughts and flooding will reduce crop yields beyond those impacts as a result of shifts in the mean variables alone. Increase in extreme events will damage crops at developmental stages like during flowering and also affecting the timing of field applications reducing the farm inputs efficiency (Antle et al., 2004;

Porter and Semenov, 2005). More so, increase in extreme events may also influence the outbreaks of pest and diseases of plants (Alig et al., 2004; Gan, 2004).

2.5.3 Solomon Islands climate change projections

The recent projections conducted by the PCCSP (2012a) for Solomon Islands have used about 18 out of 24 climate models derived from CMIP3 data base using 3 emission scenarios:-

 B1-low emission

 AIB-medium emission

 A2-high emission

These projections were made for three 20 year timeframe, 2030, 2055 and 2090 in relation to 1990. However, the projections were not specifically focus on a certain island, province, or town; rather it provides a representation of the overall geographic region of Solomon Islands (PCCSP, 2011a; PCCSP, 2011b).

2.5.3.1 Projected temperature

It is projected with “high confidence” that both air and sea surface temperature will definitely increase all through the 21st century. The rationale for this is because all the models used agree with such increase and that such an increase is constant with the

35 increasing GHG concentrations (PCCSP, 2011b). The rate of increase as simulated by the models projected that by 2030 both annual and seasonal temperatures would rise by about 1oC. However by 2090 using the A2 high emission scenario it is projected that a temperature rise of 2.5oC is expected.

2.5.3.2 Projected rainfall

Both the wet and dry season including annual average rainfall over the 21st century is projected to increase with “high confidence’. This is substantiated by the fact the all the models used agree with this increase by 2090 and that the increase in rainfall is projected for the equatorial pacific (IPCC, 2007b; PCCSP, 2011b). Nevertheless, by

2030 rainfall change is projected to be minimal however it is by 2090 that such change would be significant under the high A2- high emission scenario.

Overall, it is clear that there is discrepancy between the projected increase in rainfall with the decreasing trend observed in most of the Solomon Islands Meteorological stations as reported by Talo (2008). Such inconsistency may be attributed by the two factors, first the influence of local parameters such as topography which is not covered by the models, and second the projections does not represent a specific location within

Solomon Islands but an average of whole region (PCCSP, 2011b).

2.5.4 Solomon Islands extreme event projections

2.5.4.1 Extreme temperature and rainfall

There is very high confidence amongst the models that there will be an increase in the frequency and intensity of hot days over the course of 21st century (PCCSP, 2011b). It

36 is expected that by 2055 there will be an increase of about 1°C (B1 low scenario) and by 2090 an increase of about 2.5oC (A2 high scenario). Similarly, over the course of

21st century, the intensity and frequency of extreme rainfall is also projected with high confidence to increase (PCCSP, 2011b). The models simulated that by 2055 under B1 low scenario, an increase of about 15mm is expected and by 2090, an increase of about

30mm under A12 high scenario.

2.5.4.2 Drought

Drought events on the other hand are projected with medium confidence by the majority of models to decrease over the course of 21st century (PCCSP, 2011b). The models projected that mild drought will happen about 7-8 times in every 20 years in

2030 and decline thereafter by 2090. Moderate and severe drought events are projected to stay the same at about once or twice in every 20 years but with low confidence.

2.5.4.3 Cyclones

The number of cyclones in the South West Pacific thus covering Solomon Islands is projected to decrease with medium confidence over the course 21st century (PCCSP,

2011b). This reduction is due to the fact that most of the assessment techniques used indicated a decrease in number of cyclones in the South West Pacific and many studies substantiated a reduction in the number of cyclones globally (PCCSP, 2011b; Knutson et al., 2010)

37

2.5.4.4 Extreme sea levels

Extreme sea levels are expected to increase due to the influence of interannual variability like ENSO, with the influence of winds, waves and weather phenomena

(PCCSP, 2011b). Such sea level extremes will also be likely to increase in frequency since mean sea level is projected with very high confidence to increase over the course of 21st century.

2.6 Use of crop model to simulate impacts of climate change on crops

2.6.1 Rationale for using DSSAT crop models

There are several principal reasons by which the development and use of crop models like DSSAT (Decision Support System for Agrotechnology Transfer) have initiated.

According to Jones et al., (2003) four reasons have brought about the development of

DSSAT namely:-

 Increasing requirement for agricultural decision making as a result of

increasing demands for agricultural products compounded with additional

pressures on natural resources, land and water.

 Limitation in traditional agronomic research methods to satisfy the increasing

demand for agricultural decision making

 Traditional agronomic experiments are site and time-restricted, costly and

takes a lot of time. The results they produce are only site and season specific.

 The need to have relevant and accessible research findings that can be used

effectively.

38

However, the overall motive behind developing DSSAT was the need to make adequate decisions on transferring production technology from one place to others by integrating knowledge about soil, climate, crops and management (Jones et al., 2003).

They further stated that DSSAT provides a system approach by which research is undertaken that integrates understanding into models that allows prediction of the system behavior for any given conditions.

2.6.2 History of the birth and use of DSSAT

In the 1980s, there were already other crop simulation models that were developed before DSSAT. Each of these crop models were developed for only specific crops, for instance, CERES models for maize and wheat, SOYGRO model for soybean and

PNUTGRO model for peanut (Jones et al., 2003). These crop models were however mainly used in laboratories. Since these crop models were using different file and data structures as well as different modes of operation it was difficult to run cropping system analysis (Jones et al., 2003). Generally, the limitations with these early crop models are difficulties associated with compatibility, maintenance and addition of new crops. Thus, the need to develop a compatible models emerged which initially resulted in the design and development of DSSAT. The first release of DSSAT was in 1989 with its version 2.1 with crops like maize, wheat, soybean and peanut. Additional crops such as potato, rice, dry beans, sun flower and sugar cane were also included in later DSSAT version development (Hoogenboom et al., 1994; Hoogenboom et al.,

1999; Jones et al., 1998). The second release was in 1994 version 3.0 (Tsuji et al., 1994) and third in 1998 version 3.5 (Hoogenboom et al., 1999). The fourth release was the version 4 in 2004 and version 4.02 in 2006. Currently, at the time of writing this paper,

39

DSSAT crop model is at version 4.5 released on October 2010 compatible with a total of about 28 crops (Hoogenboom et al., 2010b) and the latest version 4.6 is proposed to be officially released by 2013.

V4.6 (2013)-28 crops

V4.5 (2010)-28 crops DSSAT versions V4.02 (2006)

V4 (2004)- >18 crops

V3.5 (1998)-16 crops

V3.0 (1994)

V2.1 (1986)

Figure 9: Various versions of DSSAT cropping system model

2.6.3 Description of DSSAT

DSSAT cropping system model (CSM) incorporates soil, crop management, and weather data and simulates growth and development of crop over time (Hoogenboom et al., 2010a). The current version 4.5 has the ability to simulate such parameters on 28 different crops under cereals (6), legumes (7), root crops (4), oil crops (1), vegetables

(5), fiber (1), forages (2), sugar/energy (1), and fruit crops (1). The model also simulates other factors such as carbon and nitrogen process, soil water and management practices (Hoogenboom et al., 2010a). There are three main components in DSSAT (Figure 10), namely the (1) main driver program that controls the sequence of simulations, (2) land unit module that manages simulations affecting land and (3) primary modules. The primary modules mainly functions with the weather, soil,

40 soil-plant-atmosphere, plant growth interface and management components. These components describe and simulate the changes over time in soil and plants that occur as a result of interaction between weather and management inputs on a single land unit

(Hoogenboom et al., 2010a; Jones et al., 2003). There are 6 main operational steps in the main program (Figure 10) namely; run initialization, season initialization, rate calculations, integration, daily output, and summary output.

Figure 10: Diagrammatic representation of DSSAT components and modular structure (Source: Hoogenboom et al, 2010a)

Overall, DSSAT requires a set of data known as minimum data set (MDS) as input for it to operate. However, there are different minimum data sets (MDS) required for various model applications. Table 2 below summarizes the various levels of MDS and its various model applications.

41

Table 2: Summary of the minimum data set for DSSAT model application Level 1 2 3 Crop model Run model for Testing and Development of operation particular experiment evaluation of model for application model Crop Appraisal Analysis Knowledgment production & e.g. determine e.g. determine why enhancement modelling whether the yield was the yield may not e.g. provide data to expected been as expected enhance understanding on level 2 Minimum Data Environment Use Use level 2 data. set Daily Environmental, Research report & minimum/maximum soil & crop data for publications. temperature, level 1 Detailed experiments precipitation & Crop measurement including response to sunshine hours. Yield and yield temperature, water Soil components like and Nitrogen. Soil surface & profile biomass, seed size Conduct specific information & number. experiments to Crop data Phenology-date of address knowledge Type, cultivar, flowering, gaps. planting date, maturity and row/planting space, harvest. irrigation, fertilizer, Growth analysis tillage & chemical Leaf, stem, seed at application. regular intervals. Soil measurements Soil moisture at different depths. Nitrogen, carbon and phosphorus.

For this study level 1 MDS was used including some MDS of level 2 such as yield, maturity date and soil moisture at different depths.

42

2.6.4 Applications of DSSAT

DSSAT plays an important role in assisting decision makers in analyzing complex decisions by reducing both the time and resources needed (Tsuji, 1998). The crop model has been used by researchers globally for wide range of purposes and applications (Jones et al., 2003). Various published studies that have used DSSAT crops models were comprehensively documented by Jones et al., (2003). These included the use of DSSAT on Crop Management, Fertilizer management, Irrigation

Management, Precision Management, Climate Change, Climate variability, Food security, Pest Management, Yield forecasting, sustainability, Tillage management,

Variety evaluation, Genomics, environmental pollution and education

2.6.5 DSSAT use and application globally

Lets us look at the use of DSSAT on Climate change in different regions. In Asia

(North West India) the responses in yield of both rice and wheat to climate change was simulated using DSSAT (Lal et al., 1998). They found out that though the yields of both crops increased under elevated CO2 levels, such beneficial effect was offset by the increase of temperature of 2-3oC. In another study in central India, using the

CROPGRO model in DSSAT, they examined the growth and yield of soybean crop in terms of the impact of thermal and moisture stresses (Lal et al., 1999). Their results indicated that soybean is sensitive to higher heat units, its yield respond substantially to rainfall variation and its growth and development severely affected by prolonged dry spells. The study also investigated the impact of future climate change (doubling of CO2) on soybean and their result was similar to results found for rice and wheat.

43

That is though the yields improve, an increase in temperature cancels out the benefits.

In Europe, they use DSSAT to evaluate the impact of projected climate change in

Bulgaria on grain yield of maize and winter (Alexandrov and Hoogenboom, 2000).

They found out that with current levels of CO2 the model projected a decrease in yields as a result of higher temperature and low precipitation causing shorter growing season.

In North America, they investigated the effects of mean temperature increases of

1-4oC on wheat yields (Rosenzweig and Tubiello, 1996). They found out that yield changes are negative under temperature change and current levels of CO2 while they range from positive to negative under elevated CO2 and temperature change. Another study in Mexico, Central America evaluated the potential impact of climate change on rain fed maize using DSSAT-CERES-maize (Conde et al., 1997). They use the crop model to estimate the yields of maize crops at 7 sites in Mexico both under baseline and climate change scenarios. In Argentina, South America they examine the impact of climate change on the production (growth and development) of wheat, maize and soybean (Magrin et al., 1997). Their results showed that there is variation in yield responses of the three crops. For instance, a general rise in yield of soybean and decrease in maize is expected to occur in relation to different sites. In terms of wheat, the yield would increase in southern and western regions whilst decline towards the north.

2.6.6 DSSAT use and application in Pacific Island countries

It is unfortunate that currently there were limited or lack of published studies in the

Pacific Island countries that use DSSAT in terms of investigating the impacts of climate change on crops. According to Wairiu et al., (2012) a major gap in Pacific

44

Islands region is the limited understanding and predictions of climate change and its effects on agriculture and food supply for the next couple of decades. They recommended that use of integrated modelling is required to increase and enhance the understanding of the future effects of climate change on agriculture and food productions. Though DSSAT is new and is currently beginning to be used in Pacific

Islands, a recent Master’s thesis by Quity (2012) use DSSAT to simulate the impacts of climate change on taro in higher volcanic island of Isabel in Solomon Islands. The present study is similar to the study conducted by Quity (2012); however it uses additional local varieties of crops like cassava (Lioka B1), corn (Koni) and taro (Tango

Sua). Another difference is that the present study was conducted on raised atoll island and is situated on the path of cyclone. Therefore, it would provide results which future studies would build on it and compare the impacts of climate change on crops on both higher volcanic islands as well as on raised atoll island.

2.6.7 DSSAT limitation and uncertainties

One of the main limitations of DSSAT is that only few crop models are available within the system (Jones et al., 1998). Most available crop models in DSSAT are those typically cultivated in developed countries, which are defined as high valued crops. However, there are limited representation of those crops cultivated in the developing world such as sweet potatoes (Ipomea batatas) and yams (Dioscorea alata/ Dioscorea esculenta) (Singh et al., 1998). Jones et al., (1998) puts forward that uncertainties associated with these cropping system models (CSMs) is due to not responding to or taking into consideration all the management and environmental factors in real farm or field conditions. For instance, CSMs have limited capabilities

45 in predicting the effects of diseases and pests, intercropping, other soil conditions and they may not be very useful in severe environmental stress conditions, which may limit the yield of crops (Jones et al., 2003). Understanding the relationship or interaction between these factors, the environment and management options to be expanded or included in the CSMs capabilities is an ongoing challenge faced by crop modelling scientists (Boote et al., 1996).

Another important uncertainty and limitation associated with DSSAT is that the

CSMs are constrained to homogenous field conditions (Jones et al., 1998). This means that the CSM is like a one point model that assumes and simulates a plant or crop as an individual, thus, it assumes that all plants in the farm are the same. Jones et al., (1998) further states that the CSM assumes that competition is also the same in all plants (above and below ground biomass). However, in a real field or farm condition, where there are some mix crops, the CSM does not fully represent the relationship between the simulated crops and other species.

2.6.8 Other crop models

There are various other crop models that have been developed in the last 20 years which are able to simulate crop growth, yield and water use under both crop management and weather data (Utset, 2009). They are APSIM (Agricultural

Production Systems Simulator), AGROSIM (Agro-ecosystem Simulation)

CROPSYST, DEMETER, HERMES, and STICS (Simulateur Multidisciplinaire

Pour les Cultures Standard).

46

2.6.8.1 APSIM

This crop model was developed by the Agricultural Production Systems Research

Unit in Australia basically to improve farm-based management and decision making under climate change (Keating et al., 2003). The model incorporates a wide variety of crops, pastures and trees including management controls and soil processes. One advantage of APSIM over DSSAT is that it can account for intercropping which means APSIM may simulate the effect of one crop on the other. Several studies have used APSIM in intercropping simulations. For instance APSIM was used to evaluate intercropping of cowpea and maize (Carberry et al., 1996) and yield of canola with competition of weed –Raphanus raphanistrum (Robertson et al., 2002). In the Pacific region, APSIM has already been in used in Fiji to evaluate the impact of climate change on cassava yield (Wairiu et al., 2012).

2.6.8.2 AGROSIM

AGROSIM is a German crop model developed at the Institute of Landscape Systems

Analysis of the Leibniz centre for Agricultural Landscape Research (ZALF) and primarily validated for use in moraine landscapes in East Germany (Kersebaum et al.,

2009). The model simulates crop growth process and can be applied to farm based crop management strategies, impacts of climate change on yields and various implications of different land use. AGROSIM was used for describing soil-crop interaction in winter cereals (wheat, barley and rye) and sugar beet in terms of growth, development, influence of CO2 and nitrogen with water stresses (Mirschel and Wenkel, 2007). Another study by Kersebaum et al., (2009) uses both AGROSIM

47 and HERMES models simulate the effects increasing CO2 on the growth and yields of winter wheat, barley and sugar beets.

2.6.8.3 CROPSYST

This crop model was developed by the Washington State University in the United

States of America and is described as a multi-year, multi-crop, daily step simulation model (Stöckle et al., 2003). It is an effective Environmental and Agricultural tool to simulate the effect of climate, crop management inputs and soils on both the environment and crop productivity. CROPSYST have been used by several studies to simulate yield and biomass production of cereals in response to nitrogen and water treatments at different locations (Stöckle et al., 1997; Pala et al., 1996; Stöckle et al.,

1994).

2.6.8.4 DEMETER

DEMETER is a complex German model which use detailed biochemical equations to simulate and compute effects of elevated CO2 on crop development parameters such as on leaf temperature and stomata (Tubiello and Ewert, 2002). This crop model has been used to simulate water balance, phenological development, interaction of leaf photosynthesis to elevated CO2 as well as yield responses to free air CO2. For example, DEMETER has been used to simulate the effect of increasing CO2 concentration to 550ppm on the yield and growth of wheat in Arizona, USA

(Kartschall et al., 1995). Another similar study uses DEMETER to evaluate the biomass and yield response of wheat under drought, excess water and elevated free

48 to air CO2 (Grossman-Clarke et al., 2001). Their study highlighted that DEMETER has effectively simulated and described both the qualitative and quantitative behavior of the wheat crop.

2.6.8.5 HERMES

HERMES is another German crop model similar to AGROSIM which is one dimensional but has multi-year and multi-crop capabilities (Kersebaum et al., 2009). It is mainly used in simulation of agro-ecosystems specifically for water and nitrogen

(Kersebaum, 1995). This study use HERMES to calculate water and nitrogen uptake from the responses to water and nitrogen availability in the soil profile as well as a from the root distribution.

2.6.8.6 STICS

Developed in France around 1996, STICS uses daily weather data and simulates crop growth, yield, nitrogen balances and soil water (Brisson et al., 2003). Unlike other crops models, STICS is not a fixed model but is more an interactive tool and depend on parameters of existing models. However, it is a very flexible model and because it uses generic factors, STICS adaptability to various crops is one of its main features

(Brisson et al., 2003). STICS have been used in maize crop precision farming

(Bruckler et al., 2000), as well as to improve crop management under water and nitrogen treatments (Justes et al., 2001). It was also used in West Indies to evaluate the effect of tillage and irrigation on banana cultivation (Brisson et al., 1998).

49

2.7 Previous studies conducted in Bellona

2.7.1 Agriculture, subsistence, livelihoods and climate change

A detail survey and study on the subsistence production systems in Bellona was initially conducted by Christensen (1975) from 1965-1967. This comprehensive work investigates and describes the local subsistence basis of shifting cultivation and forms the foundation for preceding studies that are mostly conducted by the

University of Copenhagen, Denmark under the CLIP project. For instance,

Walkenhorst (2005) conducted crop diversity and genetic erosion and compare with the results of Christensen (1975). He found out that the overall crop diversity and intra specific diversity of most crops has decline in the last 40 years as a result of declining survival rate and increasing turnover rate of crop varieties. Another study by Reenberg et al., (2008) highlighted that though subsistence and shifting agriculture since the mid 1960s have still been widespread and forms an important part of the Bellonese livelihoods, the importance of agriculture has decreased. This is because other income generating activities like small business and employment have been increasingly practiced. However, Bellona is remotely located from the rest of the Solomon Islands and is also on the path of cyclone which makes it highly exposed to climate change, climatic extremes and socio-economic stresses

(Rasmussen et al., 2009; Reenberg et al., 2008). Nevertheless, according to

Rasmussen et al., (2009) the strong traditional Polynesian social structure which is based on close cultural bonds, sharing, food distribution, and remittances has helped

Bellonese cope with external stresses. In terms of land use, Birch-Thomsen et al.,

(2010) concludes that the pattern of land use have remained the same however the intensity of land use has increased. Land use has remain the same due to lack of

50 technological or agricultural technique input, while land intensity increase as a result of innovation of crop choice like sweet potatoes. Additionally, another study by

Mertz et al., (2011) indicated that the fallow period has increased as a result of changes in crop choices, redistribution of cultivated areas. They pointed out that soil degradation is evident in the centre of the island due to high intensity of land use.

Apart from terrestrial agriculture, marine resources have also important subsistence, cultural and commercial value (Thaman et al., 2010). The study by Thaman et al.,

(2010) documented the various marine biodiversity and ethno biodiversity of Bellona and the importance of such biodiversity in terms of preservation, conservation, food security amidst globalization. They also highlighted that the rich Bellonese ethno biodiversity is critically threatened and recommend that such should be protected.

2.7.2 Soil studies

Initial soil survey study on Bellona was conducted in the mid 1960s on the description of highly porous soil (Dalsgaard, 1970). This study analysed physical characteristics of 8 soil profiles on the main soils on Bellona. Building on this study,

Breuning-Madsen et al., (2010) worked with farmers in Bellona and developed an indigenous soil classification for the soils in Bellona. They provide the descriptions and definition of various soil types and the classification system used. Our present study uses this indigenous soil classification system because; (1) it is convenient and easy to use in the field when interviewing local farmers and; (2) it has minimum soil data required to run the DSSAT model which is not provided by Dalsgaard (1970).

Another soil study further builds on Breuning-Madsen et al., (2010) and conducted micro morphological, chemical and mineral compositions of the three main soils on

51

Bellona which are normally used for cultivation (Borggaard et al., 2012). Their results indicated that the soils on Bellona lie on coral limestone dolomite and contain different amounts of mineral and element composition. Overall, they concluded that even though the soils are fairly fertile, they have low amounts of Potassium (K) and they recommended further crop experiment to substantiate their results.

52

CHAPTER THREE: METHODS AND TOOLS

3.1 Introduction

This chapter is divided into three sections. Section 1 provides an overview of the study area with background information on the geography, geology, demography, land use, cultivated crops, and socio-economic parameters of the study site. Section 2 outlines the methods and tools used to collect, collate and analyze the data and information of Bellonese perspectives on the vulnerability, impact and adaptation of climate change and extreme event on food crops and cropping system. Section 3 provides the methods and tools used in DSSAT model simulations.

3.2 Background information on study site

3.2.1 Geographic location

Bellona Island is located at the South of Solomon Islands, lying at approximately latitude 11oS 11’ and longitude 159o 15’ East (Christensen, 1975; Hansell and Wall,

1976). The island is situated at about 170km from the country’s capital, Honiara which is located on the island of Guadalcanal (Figure 11). Bellona Island is a small uplifted atoll which is described by some authors as elliptical in shape (Christensen,

1975; Breuning-Madsen et al., 2010; Borggaard et al., 2012) or resembles the shape of a canoe (SIG, 2001).

The total area of the island is about 20km2 with a length of about 10km and width of

2km. A 10km stretch of road runs along the centre of the island connecting Sa’aiho

53 ward in the west and Matangi ward in east. Major villages, hamlets and communities reside along this road. The location of villages along the centre of the island is because the coastal areas are dominated by raised steep cliff rocks of about 50-70m which is unsuitable for human settlement.

Figure 11: Geographic location of Bellona raised atoll (Source: Birch-Thomsen et al., 2010, (SIG, 2003)

3.2.2 Social, cultural, political structure and demography

Historically, the Bellonese people originated from and Futuna believed to be

Polynesian sea-farers travelling from approximately 2500km east of Bellona (SIG, 2001; Birch-Thomsen et al., 2010). According to SIG (2001), it is believed that the Bellonese have a single ancestor known as Kaitu’u; partly god and

54 partly human, who travelled with seven clans in two canoes looking for land. The two canoes are then known as canoe-shaped islands of Bellona and Rennell.

Cultural practices of the Bellonese people have eroded as a result of influence from christianity, interaction and contact with Melanesian culture and foreigners (SIG,

2001). As a result of this, the present day generation has vague remembrance of the way of life of their ancestors prior to the period of Christianity contact (SIG, 2001).

The influence and dominance of Christianity beliefs and principles contributed much to the loss of original cultural practices. Nowadays, only the elders retain some of the cultural practices, for instance the practice of nose pressing and humiliating oneself as demonstrating respect (SIG, 2001). The Bellonese practiced the patrilineal system where males have the dominant role in the Clan, Sub-clan or family. Thus, land ownership is by mainly by way of inheritance thus only the patrilineal kin is entitled to inherit land (Christensen, 1975). In terms of traditional art work, women engaged in mat, basket weaving while men do art work by wood sculpture (SIG, 2001; Baiabe,

2012).

In more recent history, Bellona Island combined with to form Rennell and Bellona Province, the smallest province in the Solomon Islands. Politically there are 6 wards in Rennell and 4 in Bellona. The wards in Bellona are Matangi, East

Ghonghau, West Ghonghau and Sa’aiho. Each ward has a provincial member who represents the people of each ward and brings their concerns or issues to the provincial executive members during their meetings. The province is represented in the provincial level by a premier while at the national parliament by an honorable member which is elected to serve for a 4 year term period (SIG, 2001).

55

The total population of Bellona according to the 2009 census was 1009 (Figure 12).

There were more women compared to men by about 5%. Most of the populace is concentrated in central Bellona specifically in East and West Ghonghau while the lowest populace is on Matangi ward, East Bellona (Figure 12).

Figure 12: Population map by wards in Bellona (Source: (SINSO, 2009)

3.2.3 Geology

According to Christensen (1975) and Breuning-Madsen et al., (2010), Bellona is a former raised atoll that originated as a result of tectonic activities. The geological age

56 of Bellona is younger than the Pleistocene historical period (Crespin 1956a; Crespin,

1965b; White and Warin, 1962).

Christensen (1975) mentioned that such tectonic activity which resulted in the uplift might have occurred in the late Tertiary geological period (65 million to 2.6 million years ago). Currently, the raised corals create an atoll rim on the island with a height of about 55m and the highest of about 70m while the lowest middle part of the depression

(previously the lagoon floor) is about 7m (Christensen, 1975; Breuning-Madsen et al.,

2010). Some historical records indicated that the center of the island was a lake inhabitated by both bird for breeding and fish which originated from the ocean (Wolff,

1969). Thus, subsequent to further uplift activity which causes the lake to drain, the vast amount of bird droppings and decomposition of fish forms layer of guano on the island (Breuning-Madsen et al., 2010). These substances over time through the process of diagenesis changed into various forms of elements like calcium, iron and aluminum phosphates which are now located on central depression, basins and plateaus on the island (White and Warin, 1959; Breuning-Madsen et al., 2010). There is rich deposit of phosphate in between the basins and on the cavities in chimneys with the limestone outcrops as a result of the geological process that had taken place. This gives rise to the two distinctive landforms that can be seen today; the limestone coral that forms a double rim on the island and the phosphate deposits. According to Breuning-Madsen et al., (2010) there are three types of phosphate (Figure 13) namely, (1) the oolitic phosphate located west north west (WNW) as chimneys and in the east south east

(ESE) as blanket deposits, (2) phosphate rock located on the southeast and (3) phosphate clay which is located on the central part depression and forms a thick blanket over the limestone. Hansell & Wall (1976) conducted soil mapping on Bellona and classified the soils according to the USDA soil classification as Rendolls,

57

Tropofolists and Eutrorthox. Rendolls and Tropofolist are located on the rim limestone outcrop whilst Eutrorthox is located on the central depression areas (Figure 14).

Cultivation on Bellona is mainly on the central depression areas on the Eutrorthox soils. A study conducted by Dalsgaard (1970) concludes that the soils on Bellona have high porosity and particle density and an overall low bulk density. The soils on

Bellona were later categorized by Breuning-Madsen (2010) into indigenous classification. This study stated five main soil types (Kenge toaha, Kenge ungi,

Malanga, Hingo hingo and Kenge mea), however only the first four soils are mainly used for cultivation as the fifth only occurs as sub-soil. The descriptions of these indigenous soils are given in appendix A.

Figure 13: Landform and phosphate deposit on Bellona (Source: Breuning-Madsen et al., 2010)

58

Figure 14: Cross section of Bellona landform and soil types (Source: Hansell & Wall, 1976)

3.2.4 Climate

There are no established weather stations established on Bellona Island by the

Solomon Islands Meteorological Services to date. Thus, there were no continuous recording of daily rainfall, temperature, sunshine hours, wind speed, and solar radiation. However, the only available daily rainfall, minimum and maximum temperatures for Bellona was recorded from March 1965 to May 1967 by Christensen

(1975). Drawing from these recorded weather parameters and climate records of nearby islands, it can be concluded that Bellona has a temperature of around 25-30oC, with extreme temperature of about 33oC and yearly precipitation exceeds 2.5m

(Breuning-Madsen et al., 2010). Overall, Bellona experiences a tropical wet equatorial climate with high humidity. According to Reenberg et al., (2008) and

Breuning-Madsen et al., (2010), though there is adequate precipitation on Bellona, it is

59 not sufficient for farming or agriculture due to the limited capacity of the soils to withhold water, the excessive drainage of soils which causes rain water to easily infiltrate and also due to non availability of surface waters (streams, rivers, lakes).

3.2.5 Land use

A study conducted by Birch-Thomsen et al., (2010) classifies land use on Bellona into six categories and compared land use changes over 40 years between 1966 and 2006.

Figure 15: Land use change on Bellona between 1966 and 2006 (Source: Birch-Thomsen et al., 2010)

The study used satellite imagery and remotely sensed data to distinguished land use changes. The six land use categories are: arable land, non-arable land, coconut 60 plantation, village area, traditional cultivation (food crops) and cultivation on former coconut plantation (Figure 15).

Land use change on Bellona observed between 1960 and 2006 has been attributed to both biophysical conditions and socioeconomic parameters such as migration, remittance, employment, reliance on processed food, change or crop choice

(Birch-Thomsen et al., 2010; Reenberg et al., 2008; Walkenhorst, 2005). However, according to Birch-Thomsen et al., (2010) there is no significant change on the pattern and types of land use change since it remained almost the same over the last

40 years period. This might be due the lack of introduction of new production techniques and technologies for agriculture use as well as the continuous practice of shifting cultivation (Birch-Thomsen et al., 2010). Additionally, there is also non-use of fertilizers or Agrotechnology inputs that may boost or intensify land use.

3.2.6 Cultivated food crops and cropping systems

The main food crops cultivated on Bellona are: greater yam (Dioscorea alata), lesser yam or pana (Dioscorea esculenta), taro (Colocasia esculenta), giant taro (Alocasia macrorriza), sweet potatoes (Ipomea batatas), potato yam (Dioscorea bulbifera), pacific yam (Dioscorea nummularia), cassava (Manihot esculenta), banana (Musa spp.), slippery cabbage (Hibiscus manihot), melon (Citrullus lanatus), pumpkin

(Cucurbita pepo) and corn (Zea mays). A list of some of the varieties of these food crops is summarized in appendix B.

61

The study conducted by Mertz et al., (2011) provides insights into the change that occurred on the cultivation quantity and intensity of the food crops in Bellona. This study indicates that there has been substantial decline in the cultivation of both the greater and lesser yams (pana) by around 25% for the last 40 years (1960-2006). The other species of yams were not identified during the assessment period (2006-2007).

The decline in cultivation is also observed in taro while sweet potato has been identified as the dominant crop being cultivated. Other crops that the Bellonese people cultivated more are water melon, cassava, pumpkin and corn. Cultivation of slippery cabbage somehow has neither increased nor decreased.

The cropping system in Bellona is a more fallow based type where there is no fertilizer, chemical or pesticides application. Even animal manure was not used on the island as the practice of burning and fallow periods were employed to enhance soil fertility.

Christensen (1975) mentioned different traditional ways in which the Bellonese people classify or distinguish their gardens based on the time of planting, type of crops used, mono-cropping and intercropping. For instance, Launga’ eha refers to gardens that are planted within the planting season of the local calendar. Yam gardens are referred to as ‘Umanga uhi’ while taro gardens are ‘Umanga tango’ and so forth.

Intercropping of crops is referred to as Tanu puupuu which means mixed planting. In terms of intercropping, the normal practice is to have a main crop planted together with secondary associates. The choice of crops used for intercropping is based on milieu of the main crop and the timeframe required for full development (Christensen,

1975). Table 3 outlines the common forms of inter planting observed in Bellona during the period of the assessment (July 2012) of this study which is similar to what

(Christensen, 1975) documented with the only additional change being those

62 in columns 2, 4 and 5. For example, sweet potato is commonly intercropped with corn and additionally to intercropping of taro and banana vice versa; slippery cabbage is also being intercropped.

Table 3: Crops that are normally intercropped together in Bellona

Main crops Banana Yam Taro Sweet potatoes

Secondary Taro, giant taro Taro, giant taro Banana and Corn associated and slippery and banana slippery crops cabbage cabbage

.

Figure 16: Intercropping of taro and banana in West Ghonghau

3.3 Methods and tools used for collecting and analyzing data and information on Bellonese perspective on climate change and extreme events

3.3.1 Household survey

63

A 25% sampling (Table 4) on the total number of households in the four wards of

Bellona was surveyed using household survey questionnaires. This percentage of sampling is similar to that of the method used by Rasmussen et al., (2009). The households that were surveyed were marked with a Garmin GPS-72H and coded according to the ward name with the household number (Figure 17). For example,

SH01 refers to “S”- name of ward Sa’aiho, “H01”stands for Household number which is 1. The GPS points were then uploaded onto a Google earth map (Google, 2012).

Table 4: Sampling size of household survey. District or No. of Household 25% sampling of Household coding Ward household used Sa’aiho 57 14 SH01-SH14 West Ghonghau 76 19 WH01-WH19 East Ghonghau 70 18 EH01-EH18 Matangi 32 8 MH01-MH07 Total 235 59

Figure 17: GPS location of households that were surveyed in Bellona (Source: Google Earth, 2012)

64

The questionnaire focused on the community livelihoods (land use, economy, mobility etc.). Specific questions were asked in terms of climate change impacts, extreme events that provided important indications of household and community vulnerability.

Table 5 below provides the summary of the sections of the questionnaires and the type of data and information collected. A copy of the household questionnaire is appended in Appendix C.

Table 5: Sections of the Questionnaire and the type of data and information collected

Section Data and Information

1:Household Provided data on number of household heads, gender, age,

Background marital status, origin, educational level, job type per household

Information heads and the time in which each household member had lived in

their village.

2:Socio-economic Provided data on both the present and past (within the last 30 characteristics years) of most important sources of income and expenses per

household.

Provided information whether cash-flow has been increased or

decreased.

3:Household Provided data on the housing living standard (building material living standards type), services available (e.g. water tank, solar, sanitation), and resource base asset/goods owned, and their sources of food supply (e.g.

household gardens, marine resources, processed food).

4:Food crops and Provided data/information on main crops cultivated; the number cropping system of plots per household, soil type, farming inputs used and length

of fallow periods.

65

5:Impacts of Provided information on whether the households witnessed any climate change on significant changes in climate over the past 30 years and the food crops and trend of that change. labour activity Specifically the section offered information whether such

changes in climate (temperature & rainfall) have effect on their

crops with emphasis on crop type and crop aspect. The response

or adaptation measures employed by the household are also

addressed.

Provided information whether the change of climate affect

labour activity.

6:Impacts of Provided information on whether the households witness any

Extreme events on cyclones and drought over the last 10, 20 and 30 years.

Food crops and Specifically the section offered information whether such labour activity cyclones or droughts have effect on their crops with emphasis on

the three most affected crop types. The response or adaptation

measures employed by the household are also addressed.

3.3.2 Focus group interview

A focus group interview was conducted on five groups namely:

1. Elderly people (>50yrs)

2. Men (26-50yrs)

3. Women (26-50yrs)

4. Young people (16-25yrs)

5. Farmers

66

Each group consists of about 6 members and they represent the different wards in

Bellona. The focus group interviews used an in-depth semi-structured interview guide

(appendix D) which obtained detailed information and perceptions on climate change and extreme events impacts on food crops. This method of interview provides a broader view and differences in Bellonese knowledge, observations and perceptions of climate change/extreme events impacts, vulnerability and adaptive capacity. The main questions used in the focus group interview guide are in Table 6.

Table 6: Questions used in the focus group interview guide

Section Questions

1  Name of group type and participants

2  In the last 30 years, what climate change events do they

remember? For young people use last 10-15 years. The use of

timelines proved easier to collect such information.

 For each event what were the main impacts and adaptation

measures undertake?

3  In the last 30 years, what extreme weather events do they

remember? For young people use last 10-15 years. The use of

timelines proved easier to collect such information.

 For each event what were the main impacts and adaptation

measures undertaken (both short & long term)?

4  Have they notice any change in the frequency and intensity

(strength) of cyclones and drought?

 Have these changes led to changes in choice of crops or varieties

cultivated? Give reasons.

67

 What were the main constraints/difficulties in changing cropping

systems?

3.3.3 Data analysis of household survey and focus group interview

After the collection of the household survey questionnaires and focus group data, the data was manually typed into Microsoft excel spreadsheet for analysis. The analysis conducted was on descriptive statistics specifically on frequencies to categorize and identify issues on household parameters like gender, age, educational level, income sources, major expenses, assets owned, sources of food, type of crops and crop management used, and the impact of climate change and extreme events on the four wards in Bellona. Data was represented in summary graphs and tables for the ease of interpretation and explanation.

3.4 Methods and tools used in DSSAT data collection, input, analysis and simulations

3.4.1 Climate data collection

Due to its close proximity to Bellona Island (and the fact there is no other weather station that collects daily weather data on Bellona), climate data was obtained from the

Henderson Weather Station through the Solomon Islands Meteorological Office in

Honiara. The data collected are for daily rainfall, maximum and minimum temperature and sunshine hours from 1982 to 2012. The 30 year period of data was chosen as such period may provide a good representation of the climate. Subsequent to the collection of those data in the form of electronic copy (CD), these were then

68 collated and formatted in an adequate style in Microsoft excel before being exported into DSSAT weatherman file.

3.4.2 ENSO (El Niño) year’s weather data

In terms of El Niño years within the past 30 years, eleven (11) El Niño years were selected from the PCCSP (Pacific Science Climate Change Program) report. The

PCCSP is a component of the International Climate Adaptation Initiative which is a research partnership between Australian government and 14 Pacific island countries to enhance the understanding of climate system and effective adaptation (PCCSP,

2011b). The eleven El Niño years selected from the PCCSP report were 1982, 1987,

1991,1992,1993,1994, 1997, 2002, 2004, 2006 and 2009. Based on these 11 years, the weather data for each year were obtained from the Solomon Islands

Meteorological Office in Honiara.

3.4.3 Soil data

Secondary soil information (chemical, physical and profile description) from previous studies and findings were collected from four main studies that were conducted in

Bellona. Those studies are summarized below in Table 7.

69

Table 7: Sources of soil information and data

Author Year Title Data/Information used

Dalsgaard 1970 Description of some highly porous Soil type and profile

soil profiles on the Island of descriptions.

Bellona, an emerged atoll in the

Solomon Islands

Hansell 1976 Land Resources Study-Vo lu me 8 Geology of Bellona and and Wall Outer Islands soil types including cross

sectional images.

Breuning- 2010 An indigenous soil classification Soil types with physical

Madsen et system for Bellona Island – a raised and chemical data. al., atoll in Solomon Islands Including profile

descriptions.

Borggaard 2012 Composition of characteristic soils Geological information of et al., on the raised atoll Bellona, Bellona.

Solomon Islands

Soil and geological information from these sources were used in providing background information of the study site. The soil data (chemical and physical) and descriptions on the profile layers were initially collated in excel spreadsheet before being exported into the S-build soil database in DSSAT.

Due to the fact that the published soil studies appended in Table 7 did not provide soil moisture content or water retention data, soil sampling and analyses on the main soil types on Bellona was conducted in this research. Sampling was conducted on the three

70 main agricultural soil types (Kenge ungi, Kenge toaha and Malanga) at 4 different depths (0-15cm/ 15-30cm/30-45cm & >40cm). Samples of about 100grams were collected, air-dried and weighed first before oven dried to a constant mass in an oven

(Laboratory Oven-Thermoline Scientific Model: TO-235F) at about 100-110oC. The difference of the mass between “before” and “after” oven drying was calculated as soil moisture content. This data is required as part of the DSSAT minimum data requirements for soil data and information.

3.4.4 Crop management data

This study is not an experimental field research thus crop management data and information were mainly collected from Bellonese farmers and based on their local knowledge. Thus, crop management data was collected using two approaches. First, it was collected during the focus group interviews with the farmers of Bellona.

Second, it was collected during site visit to some of the garden plots owned by the farmers. Data sheets as included in appendix E were used to collect such data and information. Crop management data or parameters collected were:

1. Cultivar name

2. Site location (name)

3. Slope

4. Drainage

5. Soil type

6. Planting date

7. Planting method

8. Row spacing

71

9. Planting density

10. Plot area size

11. Length of maturity

12. Plot area

13. Estimated yield

14. Management input (use of fertilizer, mulch, pesticides, insecticides, shade and manure)

The information obtained from the site visits is not limited to the use of the data sheets however additional information and observations were also recorded in notebooks. Digital photos and GPS locations were also taken as part of the collection of crop management data. These collected data and information were later tabulated in an Excel spread sheet with corresponding photos attached for easier understanding and verification before manually inputting into the Crop Management Tool (XBuild) in DSSAT.

3.4.5 Input of climate data into Weatherman tool

The climate data was formatted in excel spread sheet subsequent to importing in

DSSAT Weatherman database. The raw climate data were formatted as illustrated in

Figure 18 before they were exported into DSSAT using Weatherman tool function. It should be noted that where there is unavailability of data a default value of -99.0 was used.

72

Figure 18: Formatted climate raw data in excel spreadsheet

3.4.6 Input soil data into Soil Build File (SBuild)

The soil data and information collected from both the secondary sources (past soil studies) and soil sampling were initially arranged in an excel spread sheet prior to manually inputting into the Soil Build file (SBuild) in DSSAT. The data and information were arranged first by their name, location, slope, altitude, drainage, and their general descriptions (soil colour and texture). Then, they were tabulated in profiles soil horizons with various depths (cm) and their physical and chemical characteristics provided. Next, these data were manually inputted into SBuild file by creating new soil profiles for each soil and entering the available data and information from the excel spreadsheet.

3.4.7 Input crop management data into XBuild Tool

Similarly, like the soil data, crop management data and information were first arranged and formatted in an excel spread sheet before manually entered in the

XBuild database in DSSAT. Crop management data and information were tabulated 73 according to varieties or cultivar planted and their physical descriptions, soil type used, slope of land, planting dates, planting depths, planting density, row spacing, maturity time, main pest/disease, fertilizer/organic manure used, pesticides used, and expected crop yields. Subsequent to this tabulation, each crop plot visited was each coded with designated experiment number and name. The coding name used for the experiment name is appended below.

PE- Institute name: PACE

BE- Site code: Bellona

Year:-2012:12

Experiment #:01-09

The experimental files were then created by opening the crop management data tool and inserting crop management information manually for each crop type and saving each according to the allocated experiment codes. Table 8 below shows crop management data for each crop, site and treatments used.

Table 8: Summary of crop management data

Crop Location Variety Soil type Planting Planting Treatments type date depth W. Lioka Kenge 1st June 15cm Potential production & Ghonghau B1 ungi 2012 Water and Nitrogen limiting Cassava W. Lioka Kenge 1st June 15cm Climate projection Ghonghau B1 ungi 2012 (Temperature & Rainfall-2030, 2055 & 2090)

st W. Lioka Kenge 1 June 15cm CO2 projection (2030, Ghonghau B1 ungi 2012 2055 & 2090) W. Tango Kenge 15th 20cm Potential production, Ghonghau Sua ungi March 50% light & Water and 2012 Nitrogen limiting 74

Taro W. Tango Kenge 15th 20cm Climate projection Ghonghau Sua ungi March (Temperature & 2012 Rainfall-2030, 2055 & 2090)

th W. Tango Kenge 15 20cm CO2 projection (2030, Ghonghau Sua ungi March 2055 & 2090) 2012 Koni Kenge 15th June 4cm Potential production & E.Ghonghau toaha Water and Nitrogen limiting Koni Kenge 15th June 4cm Climate projection Corn E.Ghonghau toaha (Temperature & Rainfall-2030, 2055 & 2090) E.Ghonghau Koni Kenge 15th June 4cm CO2 projection (2030, toaha 2055 & 2090)

3.4.8 Calibration of genetic co-efficients of cultivars of taro, cassava and corn

Given the fact that the local varieties of taro, cassava and corn in Bellona are not readily available in DSSAT crops database, calibration on the model for existing cultivars for each crop was conducted. The overall reason of calibrating is that the local varieties are not available in DSSAT therefore we calibrate the existing varieties genetic coefficients to create our new local varieties within DSSAT. This will ensure that whatever simulations we perform will closely resembled what we may expect the local variety to perform in the field. However, the genetic coefficients were not calibrated based on experimental field data but on local Bellonese farmers’ knowledge.

The genetic coefficients were calibrated manually using mainly the crop information data on maturity days, corm/tuber/grain initation time and yields to ensure that the

75 model simulations closely mimic the maturity days and yields estimated by the farmers. The DSSAT model for all three crops has never been validated for Bellona.

The summary of the steps involved are appended below:-

1. The Sensitivity Analysis function was used to identify which cultivar within

DSSAT has close resemblance to the local Bellona variety in terms of: -

maturity days, corm/tuber/grain initation time and yield. Note that when

using sensitivity analysis, the data for soil, climate and crop management will

reflect those of Bellona while only the cultivars used will differ. The

sensitivity analysis was conducted for all the available cultivars in DSSAT

database for each of the crops (taro, cassava and corn). The Overview and

Evaluate OUT. Files were used to identify and select which existing cultivars

closely resemble the local Bellona varieties.

2. Subsequent to the selection of the best suited cultivars for each crop types,

genetic coefficient for each crop was calibrated. Before calibration, the

genetic coefficients of the closely resembled cultivars is copied and pasted in

the genetic file of each crop type and a new local variety name is given. Next,

the genetic coefficient numbers or values under coded coefficients (Table 9

and Appendix F) were manipulated by either decreasing or increasing the

values until the factors such as maturity days, corm/tuber/grain initation time

and yield closely resemble those of the field values. The Overview OUT.

Files were used to identify and compare these factors. Note that each time the

genetic coefficient values were manipulated, the model was run. Table 9

below summarizes the various cultivars selected, new local varieties created

and the genetic coefficient variables that were calibrated for each crop types.

76

The definition for each genetic coefficient code is provided in appendix F.

Table 9: Calibrated genetic coefficients of each crop

Crop A. Best suited cultivar Genetic coefficients calibrated type B. Local varieties SRFR LA1S LAWS PHINT LLIFA

Cassava A.Mcol-22 0.3 300 180 22 10

B.Lioka-B1 0.35 350 220 20 20

P4 P5 G2

Taro A.Tausala-Samoa 900 700 1.00

B. Tango Sua 500 480 2.10

P1 P5 G2 G3

Corn A.Sotubaka 300.0 930.0 500.0 6.00

B. Koni 180.0 940.0 320.0 5.00

In terms of cassava, five genetic coefficients of cultivar A.Mcol-22 appended below were calibrated to ensure that the simulated yield and maturity days of local variety

Lioka B1 closely resembled the yield and maturity days data provided by the farmers.

 SRFR-storage root fraction of assimilate used for non-root growth

 LA1S-Area/leaf (cm2) of the first leaves when growing without stress.

 LAWS-Leaf area/weight ratio when crop growing without stress (cm2/g)

 PHINT-Interval between leaf tip appearances for first leaves

 LLIFA -Leaf life from full expansion to start senescence (phyllocrons)

For taro, only three genetic coefficients of Tausala-Samoa appended below were calibrated.

 P4-rapid vegetative growth  P5-cormel and corm growth

77

 G2- growth partitioning coefficient affecting petiole growth

In terms of corn four genetic coefficients of Sotubaka were calibrated. They were;

 P1-Thermal time from seedling emergence to the end of the juvenile phase

(expressed in degree days above a base temperature of 8oC) during which the

plant is not responsive to changes in photoperiod.

 P5-Thermal time from silking to physiological maturity (expressed in degree

days above a base temperature of 8oC).

 G2-Maximum possible number of kernels per plant.

 G3-Kernel filling rate during the linear grain filling stage and under optimum

conditions (mg/day).

3.4.9 Local varieties of taro, cassava and corn used in DSSAT simulation 3.4.9.1 Taro

For taro the local variety Tango Sua was used in the simulations. Tango Sua was chosen because it is one of the most common taros cultivated in Bellona and the crop management information was collected from plot in West Ghonghau. Tango Sua was first calibrated using the genetic coefficient of Tausala-Samoa cultivar which is available in DSSAT (Table 9). Tausala Samoa was chosen because it has close resemblance to Tango Sua when running sensitivity analysis in terms of maturity days, corm initation time and yield under the field conditions of Bellona.

3.4.9.2 Cassava

In terms of cassava (Manihot esculenta) we used the local variety which has unknown 78 name however some Bellonese referred to as Lioka white as it has white tubers. Thus the author took the liberty of naming it Lioka B1 (Bellona 1). The choice of simulating this variety is because it is one of the most common varieties cultivated by the households. Lioka B1 was first calibrated using the genetic coefficient of A.Mcol-22.

3.4.9.3 Corn

It is unfortunate during the field visit that the farmers were not able to provide the name of corn variety that was cultivated in Bellona. The farmers mentioned that only one variety of corn was cultivated during the time of this field visit. Thus, the author took the liberty of using Bellonese term “Koni” to refer to this variety. Koni was first calibrated using the genetic coefficient of Sotubaka cultivar which is available in

DSSAT.

3.4.10 Simulation and evaluation of impacts of future climate change 3.4.10.1 Summary of PCCSP projections used

The PCCSP future climate projections scenarios (Table 10) were used in DSSAT to simulate the likely impacts of climate on the yields of (taro) Tango Sua, (cassava)

Lioka B1 and (corn) Koni in Bellona. The PCCSP projection used for these simulations is based on the CMIP3 global climate output for three 20 year period centered on 2030, 2055 and 2090 (PCCSP, 2011a). There were three projections used by PCCSP namely B1 (low), A1B (medium) and A2 (high) future gas emissions scenarios. However, in this study the A2 (high) scenario was used in an attempt to see what would be the response of the three crops under this worst case scenario.

79

Table 10: Summary of projection extracted from PCCSP report (Source: (PCCSP, 2011a) o Year Temperature ( C) Rainfall (%) Carbon dioxide (ppm)

2030 1 8 450

2055 1.8 13 550

2090 3.3 21 750

3.4.10.2 Simulations and treatments

The projections were included in the “Environment Modification” component in

DSSAT for Tango Sua, Lioka B1 and Koni. Model runs (simulations) were conducted for Tango Sua for four treatments (2, 3, 4, and 5) listed below while for Lioka B1 and

Koni four treatments excluding treatment number 2 (1, 3, 4, and 5) were conducted.

The reason is because Tango Sua is normally cultivated under shade, thus an additional treatment (2) of 50% canopy light was included.

1. “Ambient (2012) Potential production” –with 100% light and where there is no

consideration of water and nitrogen or other environmental parameters.

2. “Ambient (2012) Potential production” with canopy light of 50% was an

additional treatment only conducted for Tango Sua and not the other two crops

since the practice of taro cultivation in Bellona includes preservation of trees

within the plot. Thus, such practice prevents direct 100% sunlight reaching the

crops.

3. “Ambient (2012) Attainable production” where there is consideration of water,

and nitrogen.

4. “Future climate projection” where only temperature and rainfall change for

2030, 2055 and 2090 was used.

80

5. “Future climate and CO2 projection” includes temperature, rainfall and CO2

changes for 2030, 2055 and 2090.

The term “ambient” used in both ambient potential and ambient attainable production simply refers to the current year (2012) when the data for crop management and weather files were collected and simulated. Thus, ambient generally refers to the weather and crop management parameters or conditions in the year 2012.

In terms of analysis of the simulations, three files were used:-

1. Overview Output (Overview. OUT)

2. PlantGro Output (PlantGro. OUT)

3. Evaluate Output (Evaluate. OUT)

These three outputs were used to analyze and determine the impacts of the future climate scenario on Tango Sua, Lioka B1 and Koni growth and yields. The output files were analysed using both the “view” and “plot” function. The view function provides written summary and tabulates results while the plot function offers graphical representation of the simulation analysis in terms of time-series or scatter plots.

However, the outputs of graphics were then exported to Microsoft excel whereby the graphs were easily named and analysed.

Other simulation using ENSO (El Niño) years from the past 30 years daily climate data were conducted to investigate the impacts of ENSO events on Tango Sua, Lioka B1 and Koni. Eleven (11) El Niño years within the past 30 years as per PCCSP (2011b) were selected and simulated against the three crops. Sensitivity Analysis Option of the model was also used to optimize crop management and to select possible cultivars for Bellona Island. This was conducted essentially with the aim of maximizing yield during stressful conditions of future climate scenarios and ENSO events.

81

CHAPTER FOUR: VULNERABILITY OF HOUSEHOLDS, FOOD CROPS AND CROPPING SYSTEM TO CLIMATE CHANGE AND EXTREME EVENTS

4.1 Introduction

The term vulnerability, within a climate change context has been used widely and has different implications. However, for this study, vulnerability is defined as the extent to which a system is not able to cope with and is prone to the negative effects of climate change and extremes (Parry et al., 2007). Three components that define vulnerability are; (1) sensitivity of a system to changes in climate, (2) adaptive capacity of the system and (3) the degree of exposure to climate hazards (Parry et al.,

2007). Sensitivity refers to the extent to which a system is affected by climate change or extremes, whether it is beneficial or unfavorable. The ability of a system to cope with or adjust to the impacts of climate change is the adaptive capacity; while exposure refers to the extent to which a system is exposed to climate change and extremes. In this study, the term “system” refers to both human and social structures as well as food crops and cropping systems. This is because there is an inter-relationship between humans (farmers), social structures and crop production

(Berry et al., 2006). In terms of food crops, the sensitivity components are covered in

Chapter 7 whereby the sensitivity of three crops (Tango Sua, Lioka B1 and Koni) is evaluated using DSSAT crop modelling. This chapter focuses on the adaptive capacity and the degree of exposure of both the households and the food crops/cropping system. It aims to present a description of the households in the four wards in Bellona in terms of their livelihoods, socio-economic activities, types of cultivated crops and crop production systems. These parameters are important indicators of adaptive capacity. 82

4.2 Household characteristics

This section describes the characteristics of households in Bellona in terms of gender,

age, level of education, assets owned, income sources, food sources and the number

of garden plots that provided important indications of household, community,

farmers, food crops and cropping system vulnerability in terms of their adaptive

capacity.

Table 11: Household characteristics of 59 households surveyed in Matangi, East Ghonghau, West Ghonghau and Sa’aiho Wards Total/ Parameters Description Matangi East West Sa’aiho Average Ghonghau Ghonghau Number of 8 18 19 14 59 households Percentage Male 4 7 18 6 59.3% Head of Female 4 11 1 8 40.7% household Gender Household Age >55 yrs 3 6 8 5 22 head Age Mean age 56 50 50 48 51 Highest level Primary 4 10 9 4 45.8% of Education Secondary - 2 6 5 22.1% completed College 1 5 4 3 22.1% (Household Tertiary - - - - - head) None 3 1 - 2 10.2%

Table 11 indicates that more than half of the households are headed by male.

However, female-headed households are not uncommon in Bellona where in this

case represents about 40% of households surveyed. Most of these female-headed

households are caused by husbands who have moved to Honiara (urban area) or are

83 no longer living with their wives. About 37.5% of these female heads are either widows, single mothers or divorced. This study showed that female-headed households are more vulnerable to the impacts of climate change and extreme events because of the following reasons:-

 They lack the support and labour activity from their husbands as this is a

society where men are mostly responsible for initial clearance, preparation

and ploughing of the land.

 Males or husbands are traditionally responsible for financially supporting the

household.

 Most, if not all of the households are subsistence based and are very much

dependent on their gardens for their livelihoods, making them vulnerable to

the impacts of climate change and extreme events.

The average age of all the household heads is 51 years. Even though this is below the retirement age of 55 in Solomon Islands, overall, 37.3% of household heads are 55 years old or older. This indicates that such age of household heads are inactive and may soon become a liability to the family. For example, such age group can cause a decline in crop or gardening productivity as by requiring the more active family member to look after them instead of spending more time gardening. In other words, the age factor of the household heads may influence the household and their crop production, making them vulnerable to the impacts of climate change and extreme events.

About 45.8% of the household heads receive primary education, 22.1% attend secondary education, 22.1% attend college or tertiary education while 10.2% receive

84 no form of formal education. The data indicates that 89.8% of the household heads receive some form of formal education and are literate. The significance of majority of households being literate is that they can easily understand written or spoken information provided to them either by provincial or national agricultural officers and NGOs working on crop production. Literacy will also enhance their understanding of how to implement or make informed decisions or ways to improve their crop productivity in terms of their exposure to climate change impacts and extreme events. Studies also put forward that the literacy factor can reduce food crops vulnerability as any ideas, technologies or information communicated via written materials from relevant agencies to the farmers such as factsheets, pamphlets or brochures can be easily understood and adhered to (Mutsvangwa, 2010).

4.2.1 Housing structure, available service and asset owned

About 93.1% of households in the wards in Bellona have permanent housing structure (Table 12). Permanent houses are those with manufactured materials, for instance corrugated roofing iron, timber or concrete walls and floors, as well as steel or concrete posts. Semi-permanent and traditional housing structures made from either a mixture of both manufactured and local materials (semi-permanent) or only local materials (traditional) account for only about 6.8% of the total households surveyed.

85

Table 12: Housing type, available amenities and asset ownership Wards Total/ Parameters Description Matangi East West Sa’aiho Average Ghonghau Ghonghau Permanent 7 18 17 13 93.1% Housing Semi-permanent 0 0 1 1 3.4% Type Traditional 1 0 1 0 3.4% Tank 8 16 16 14 91.5% Well 0 0 0 0 0 Available Generator 1 5 4 1 18.7% service/ame Solar power 8 11 4 13 61% nity Pit toilet 7 17 18 13 93.2% Flush toilet 1 4 6 2 22.% Bicycle 7 9 17 12 76.3% Fiber canoe 0 0 0 5 8.5% Wooden canoe 0 0 0 6 10.2% Assets Outboard motor 0 0 0 4 6.8% owned Radio 2 9 14 11 61% Kerosene Stove 3 6 8 5 37.3% Tractor 0 0 0 0 0

The type of housing structure is important because it signifies the level of vulnerability the households have towards climate change and extreme events. This is because the stronger the housing structure is, the less vulnerable and exposure it has from climatic events. From these results, it is evident that the housing structures in Bellona are less vulnerable to some extent. However this is not to say that because most percentage of houses has permanent structures, they are safe and can withstand extreme events like cyclones. The strength and duration of such climatic extremes, the standard of housing structures, the location and orientation of houses, are some of the crucial factors to consider. For instance, some of permanent housing structures in

Bellona are built on stilts. These may become vulnerable when exposed to strong

86 winds and cyclones given that the island is located on the active path of cyclones.

The importance of looking at housing structures in terms of vulnerability to food crops and cropping systems is that:-

1. Having a permanent structured house/building will ensure that the farmers

can concentrate on farming practices rather than rebuilding every time when

there extreme events occur. Thus, more time can be allocated to labour input

into crop production rather than on rebuilding housing structures.

2. Good and adequate building structures are also important for storage of crops

harvest and planting materials (e.g. seeds or tubers) not only from climatic

related events but also from pest like rodents, beetles or insects and diseases.

The study found that reducing vulnerability is augmented by adequate housing structure and provision of basic amenities. Table 12 revealed that more than 90% of households have access to rain water tanks. This is because there is no surface water like streams or rivers on the island thus households depended on their water tanks for domestic and household necessities. Only about 18.7% of the households have access to generators while 61% own solar power for providing electricity for lighting, functions and entertainment. This percentage difference may be attributed to the high cost of fuel and the fact that solar power is more economical. In terms of sanitation,

93.2% of the households have access to pit toilets whilst only 22% have flush toilet.

The difference in this two toilet types can be explained by the fact that water availability is a significant problem thus people resort to pit toilets which do not require water.

87

The significance of looking at the available amenities in terms of vulnerability to food crops and cropping systems is that:-

1. The availability of adequate amenities will influence the crop management

inputs into which in turn have direct impacts on crop production and quality.

For instance, about 91% of households have tanks for water collection and

storage because there are no surface waters or water supply available. This

lack of access to abundant water supply means that the household only have

enough water for their domestic uses (washing, cooking and drinking) and

limited water for crop irrigation. Thus, it can be argued that because of the

geological makeup of Bellona Island (limestone), the vulnerability of the food

crops and cropping systems is high. More so, given that most soils on Bellona

have less water retention capacity, water stress is a major concern especially

when there is limited availability of water supply for irrigation. Thus, this

resulted in the lack of irrigation in crop production practice in Bellona and

that their farming is rain-fed.

2. Access to adequate sanitation amenities such toilet and clean water is also

important in terms of labour input into crop production. Health related

problems that arise from sanitation problem can result in lower labour input

or lack of labour activity for a certain period of time. The early stages of

crops are essentially important for their growth and production. Thus, if a

farmer is sick and leaves his young crops for a week or more without

weeding, crops are put at risk from competition with weeds and attacks by

pest. Only 22% of the households had access to adequate sanitation;

suggesting the vulnerability of food crops has instructed labor input and

management on the farm.

88

The households on Bellona own basic assets like bicycle, canoe, outboard motor, and radio. Such assets are important indicators of the welfare status and they are useful because they can be used as source of income by lending them to others to use at a charge, or they may improve accessibility to information and transportation

(Mutsvangwa, 2010). Bicycles are a very significant asset for people who live on

Bellona. This is because the island has a stretch of road from one end of the island to another with no public transport services. The results show that about 76.3% of the households have bicycles with a minimal number of bicycles per household of 1 and maximum of 4. While bicycles are needed for daily transportation for the people, they are very important for transportation of farmers to their plots, planting materials and crop harvest.

Use of bicycles for transport, enhances farm labor input by reducing time taken if the former were to choose to walk to the garden plots. During our field work we noted that most farmers used their bicycles to transport their planting material and harvest to and from their plots.

Only about 18.7% of the households have access to canoes. Of these, 10.2% are wooden canoes whilst only 8.5% are fiberglass canoes. To power fiberglass canoes, only 6.8% have outboard motor engines. The uses of canoes or boats are important in terms of fishing as well as for transportation to the nearby island of Rennell for holiday, education, health, family visits and transportation of goods. Radio is an important asset that serves as mode for information access. There are radio programs offered by Ministry of Agriculture via the Solomon Islands Broadcasting Corporation

(SIBC) radio station. An example of this is called “Farmer’s corner” which provides awareness information in agricultural issues, new variety of crops, pest or diseases

89 etc. Daily weather reports and updates including weather forecasts and cyclone warnings are offered by the Solomon Islands Meteorological Services via the SIBC radio. The results show that only about 61% of the households have access to radio.

This means that about 39% of households in Bellona do not have access to radio meaning that their vulnerability and exposure to climatic related events like cyclones or new pest/disease breakout is fairly significant.

4.2.2 Sources of income

The sources of past and current income include copra, informal or casual labour, remittance, pension, salary (formal employment), market, fishing, farming, carving and weaving (Figure 19).

Household sources of income Copra Casual Labour Remittance S Pension o u Salary r Small Business c Market Past e Fishing Present s Farming

Carving Weaving 0 5 10 15 20 25 30 35 Percentage % Figure 19: Past and present household sources of income

Apart from salary income, most households earn their money from selling handicrafts like basket, mats and wooden carvings (Figure 20).

90

Figure 20: Traditional woven baskets in Matangi about SBD$150-250 each

It is important to note that income from farming or agriculture accounts for the least income sources (1.7%). This is because most food crops are mainly for subsistence and only a small proportion can be sold as raw or cooked during the market days which are not regular.

Casual or informal labour is also a form of income generating source. Examples of this include providing assistance in gardening, carrying goods/materials from airport/ship drop areas, cleaning/brushing compounds, and other small projects that occur in ad hoc bases at a certain charge. An example of a project during the time of this field work is informal labour participation (Figure 21) in manual excavation for a

Mobile Telecommunications Tower project contracted by Solomon Telekom Limited.

Fishing is another significant source of income which accounts for about 8.4%.

Communities in the western ward of Bellona, Sa’aiho are highly engaged in fishing activities, compared to the rest of the wards on Bellona. This is because it is situated in an area which is conveniently accessible to coast (low limestone cliffs). 91

Furthermore, since it is on leeward side of the island, there are minimal strong winds and waves.

Some households (8.4%) are engaged in small retail outlets. These retailers sell basic food items like rice, noodles, canned food, batteries, salt, biscuits, cigarettes, beetle nuts, and matches amongst other basic items. However, due to limited transportation accessibility, unscheduled shipping services and high cost to the main urban centres in Honiara, operating business activities on Bellona proved difficult. From the interviews, shipping services to the island solely depends on charters of business people or projects thus only one ship travels to Bellona every 3-6 months. While domestic air transport is available, it only allows a maximum of about 16kg of baggage and any excess are either extremely expensive or not permitted. These factors of limited, irregular and costly transportation by both air and sea are very significant in terms of the vulnerability of the Bellonese communities’ livelihoods to climate change and extreme events. The remoteness of Bellona Island and minimal economic activities thereby are parameters that make the Bellonese communities highly vulnerable to climate change and extreme events.

92

Figure 21: Informal labour activity in West Ghonghau: Manual excavation for the foundation of Solomon Telekom mobile telecommunications network tower.

4.3 Food crops and cropping system

There are various food crops cultivated in Bellona. The main root crops are taro, yam, pana (lesser yam), cassava, sweet potatoes and these accounts for most percentage of the cultivated crops (Figure 22). However, majority of the households (Figure 23) cultivate sweet potatoes (78.8%). This is similar to the findings of a study carried out on Bellona which stated that sweet potatoes are becoming the most cultivated crop in

Bellona (Reenberg et al., 2008). The reasons why the Bellonese cultivates more of sweet potatoes according to the households are:-

 sweet potatoes have short maturity timeframe (3 months)

 less labour input is required for land preparation and management

 sweet potatoes are less sensitive to soil types and can grow on most soils in

Bellona and on used land areas

 sweet potatoes can be cultivated in front or back yard of houses thus easily

accessible during extreme climatic events like cyclone 93

 sweet potatoes can be cultivated continuously straight after harvesting and

require less fallow periods

Percentage of crops cultivate among the four wards in Bellona (59 households surveyed) 100 90 P Matangi 80 e 70 r East Ghonghau e 60 n 50 West Ghonghau t 40 a 30 Sa'aiho g 20 e 10

0

Crop types

Figure 22: Percentage of Food crops cultivated amongst the 59 households within the 4 wards of Bellona

Other crops cultivated include banana, slippery/chinese cabbage, pumpkin, pawpaw, corn, water melon and bean. Banana is the second most cultivated crop which accounts for about 71.12%, yet in most cases it is intercropped with other crops like taro and slippery cabbage. Vegetables and fruits like chinese cabbage, bean, pawpaw, water melon and pumpkin are least cultivated because they are not the main staple food crops and some are only planted during certain seasons.

94

Mean percentage of crops cultivated in Bellona (59 households surveyed) Pawpaw, Water melon, 10.7 Pumpkin, Banana 4.9 Corn, 12.2 Bean, 3.1 Sweet Potatoes 7.6 Chinese Taro Cabbage, 2.7 Pana Yam Banana, 71.1 Slippery Cassava cabbage, 31.9 Slippery cabbage Cassava, 16.4 Chinese Cabbage Yam, Sweet Potatoes, Pumpkin 8.4 78.8 Pana, 30.8 Pawpaw Corn Water melon Taro, 55.2 Bean

Figure 23: Mean percentage of food crops cultivated amongst the 59 households in Bellona

The significance of looking at the percentage of types of food crops cultivated and the rationale behind them is that it provides the extent to which these crops are vulnerable to climate change and extreme events. For instance, the main staples on

Bellona are sweet potatoes, taro, yam, pana and cassava. In the past 50 years, yam and pana were the main staple planted because they give good yields in both quantity and quality (Christensen, 1975). On the contrary, current results indicate that yam and pana are only cultivated by 8.4% and 30.8% of the households respectively whilst sweet potatoes account for around 78.8% of the households. According to the interviews conducted, the reason for is the increased cultivation of sweet potatoes is because the farmers noticed a decline in the yield and the quality of the yam and pana tubers. Respondents commented that a changing climate and unpredictable weather conditions affected the growth and yield of yams and pana; therefore they shifted towards cultivating a more resilient crop (sweet potatoes) which provides 95 high and fast yields. However, this is not the case with cassava which is only cultivated by 16.5% of households. Cassava is reported to be a more resilient and less vulnerable crop to climate extremes. Yet, the low percentage of cultivation is because the Bellonese only cultivates cassava in small quantities for puddings and for the times when their harvest of other crops like taro is not adequate. Cassavas are often cultivated as boundary crops in some garden plots. Thus, it can be argued that while cassava is less vulnerable to climate change and extreme events, its adaptive capacity has not been exploited significantly by the Bellonese.

The average number of plots or garden field per household is recorded and compared with the records of 1962-7 (Christensen, 1975) and 2006-7 (Mertz et al., 2011) as depicted in Table 13 below. Table 13 reveals that all the wards experienced a decline in the trend of number of plots per household. Two possible explanations for this decline are; (1) farmers rely more on food imports from Honiara that supplement their diet thus cultivate their crops in fewer locations or (2) due to changes in access to land

(Birch-Thomsen et al., 2010; Mertz et al., 2011). The significance of this finding is that it indicates the exposure of the land/soil to various changes in climate and extreme events where in the long run results in declined soil fertility in some parts of the island.

A study revealed that that the centre of the island of Bellona is more fertile and has more inhabitants compared to the outer rim of island (Mertz et al., 2011). The high fertility in the centre is believed to be because in the past vast amount of bird droppings and decomposition of fish forms layer of guano on the centre of the island

(Breuning-Madsen et al., 2010). Therefore, in the low fertile areas, the farmers use less plots because even though much is invested into labour productivity, crop productivity declines. The results for 2006 compared to our results 2012 clearly showed that

96

Ghonghau ward which is in the centre of the island has more plots compared to the other wards which may signify the land exposure and fertility issue.

Table 13: 45 year averages of number of garden plots per households in major Wards on Bellona Island Ward 1962-1966 2006-2007 2012

Matangi 7.8 4.3 2.8

Ghonghau (East & West) 6.2 4.6 3.1

Sa’aiho 6.3 4.6 2.1

Total 6.8 4.5 2.7

The cropping system of using fallow periods to enable the land to replenish itself is commonly practiced among the households in Bellona. Fallow periods reported by the households range from 2 to 10 years depending on the type of food crops previously planted or which shall be planted. Table 14, outlines the average fallow period from 1962-6 records (Christensen, 1975), 2006-7 records (Mertz et al., 2011) and our results in 2012. It is obvious that the fallow periods increased in 2006-7 compared to 1962 however declined in 2012. The 2012 results closely resembled those of the 1962 records for taro, pana and yam however declined for sweet potatoes.

Table 14: 45 year averages for cropping fallow in Bellona Island Crop Type Average fallow Average fallow Average fallow 1962-6 2006-7 2012 Sweet potatoes 4.2 7.8 1.7 Taro 5.5 16.6 5 Pana (lesser yam) 5.5 13.8 5.5 Yam (greater yam) 5.5 9.8 6

97

The decline in the fallow periods may be caused by increased pressure placed on the land availability by increased land intensification and extended cropping periods which consequently affect soil fertility and crop production (Bourke et al., 2006;

Jansen et al., 2006). However, a study highlights that such decline in fallow periods are influence by the redistribution of land use intensity which is varied in Bellona

(Mertz et al., 2011). Furthermore, they asserts that fallows periods increase in areas around the rim of the island where soils are less fertile however in the centre of the island where most of the population live and the soils are more fertile, fallow periods are shorter.

In terms of vulnerability, the study by Mertz et al., (2011) showed that soils on Bellona

Island have been impacted by short follow periods; possibly implying that adaptive capacity of the soil to replenish itself, has been limited mainly due to conversion of remnant natural forest areas; small-scale brushing and burning and clearing of fallow vegetation. This could also be a major reason why the cultivation of sweet potatoes having becoming the dominant crop in Bellona. Moreover, the impacts of climate change and extreme events which are evident in the exposure of soil to climatic hazards like drought or extreme rainfall have the potential to further exacerbate soil fertility and crop production. The cropping system on Bellona as have been reported in past and present have generally remained the same (Christensen, 1975;

Birch-Thomsen et al., 2010; Mertz et al., 2011). The farming practice is generally brushing, burning and fallow vegetation where there are no inputs of fertilizer, animal manure, pesticides, insecticides and irrigation. All of the households interviewed including the focus groups substantiated that their farming practices do not include any chemicals, manure or irrigation input. This is another indication of vulnerability in

98 terms of the cropping systems, crops and soil increase exposure and decrease in adaptive capacity to climate extreme and hazards. Since no technological, chemical or biological inputs are used in the Bellonese cropping systems, it indicates that the adaptive capacity to cope with soil fertility issues or management of pest or disease is limited.

99

CHAPTER FIVE: BELLONESE PERCEPTION ON THE IMPACT OF

CLIMATE CHANGE ON FOOD CROPS AND CROPPING SYSTEM

5.1 Introduction

This chapter describes the Bellonese perspective in terms of the impact of climate change and extreme events on food crops and cropping system. The IPCC defines impact as the effects that climate change has on both the natural and human systems

(IPCC, 2007d).The word impact will be described within the context of climate change and extreme events related to food crops, cropping systems, and how this affects livelihoods of the Bellonese people. It is important to include livelihoods because in Bellona, there is direct relationship between the livelihoods and food crops. Both the direct impacts and indirect impacts of climate change will be presented and discussed. Examples of direct impacts include impacts on crops caused by changes in temperature and rainfall (wilting or rotting) while indirect impacts include changes in pest, disease, weeds or decline labour input (Turral et al., 2011).

The information provided is the result of the 59 households’ survey and 5 focus group (elderly, men, women, young people and farmers) interviews. Section 5.2 provides the Bellonese perception and observation of the past and current occurrence of climate change and its impacts on food crops. Section 5.3 outlines their perception in terms of extreme events mainly cyclones and drought and their impacts on food crops.

100

5.2 Past and current change in climate and correlating impacts

The households were asked if they had noticed or experienced any long term change in temperature and rainfall over the last 30 years and its impacts on the food crops.

The results in Table 15 indicate that almost all of the households confirmed that there is actual change observed in the temperature (96.6%) and rainfall (98.3%) over the past 30 years. The other 3.4% and 1.7% of the households respectively either stated that temperature and rainfall remained the same or that they were unsure.

Table 15: Household observations of changes in temperature and rainfall

Parameter Change in Temperature Change in Rainfall

Percentage of household 96.6% 98.3% saying “Yes” Percentage of household 3.4% 1.7% saying “No” or “Unsure”

For changes in temperature and rainfall, there are some impacts that are the same for both parameters. Both changes in temperature and rainfall cause reduction in crop growth, yield and tuber production, increase in pest and disease infestation as well as tuber deformation. Changes in temperature are reported by the respondents to have caused the wilting of crops, reduction in tuber size and deformation in fruits of most crops. The changes in rainfall on the contrary induce rotting in tubers, corms and roots, physical damage to newly planted crops, loss of taste or flavor in tubers and increase in vegetative growth but reduction in tuber production.

101

Table 16: Impact of temperature and rainfall change on crop aspects Impact on Crop aspect by Impact on Crop aspect by Rainfall Temperature Change change Decline in growth Decline in yield Decline in yield Rotting of roots/tubers Wilting of crops Decline in growth Reduction in tuber size Physical damage to newly planted crops Abnormality in tuber shape/deformation Increase in pest & disease infestation Increase in pest & disease infestation Tuber deformation Deformation of fruits Loss of taste or flavor in tubers Increase in vegetative growth but decline in tuber production.

On a more detailed approach, the 5 focus groups (men, youth, women, farmers and elderly) were asked about the trend of temperature and rainfall witnessed over the past 30 years and their impacts on food crops. All of the 5 focus groups clearly indicated that they have observed increase trends in temperature and rainfall over the last 30 years. Note that the youths were asked only for the last 10-20 years due to their age limit.

The trend of temperature change observed increased over the past 30 years, have various impacts on food crops. Figure 24 outlines the various impacts as expressed in overall percentage the focus groups. The percentage is calculated as the number of focus group responded “yes” to a particular impact divided by the total number of focus group multiplied by 100. For example if all of the focus group reported wilting as an impact of increasing temperature, it would be calculated as (5/5 x 100) which is

100%. The impacts include:-

 Wilting of crops

 Decline in yield

102

 Decline in yield quality

 Decline in tuber size

 Decline in survival rate

 Decline in crop growth

 Loss or change in taste or flavor

 Early maturity

 Abnormality or deformation of fruits/tubers

 Increase in pest and diseases.

Impact of increase in temperature on food crops

Wilting

Decline in yield 20 40 100 Decline in yield quality

60 Decline in tuber size

Decline in survival rate 40 100 Decline in growth

40 Loss or change of taste

Early maturity 40 60 40 Abnormality in fruits and tubers Increase in Pest/Disease

Figure 24: Responses of the five focus groups expressed as percentage on the impact of increase temperature trend on food crops in Bellona

All of groups (100%) identified wilting and declined in yield as the most significant impact on the crops by the temperature increase. Sixty percent (60%) claimed decline in yield quality and early maturity as adverse impacts whilst 40% mentioned decline in tuber size, survival rate, growth, loss or change in taste (flavor), and abnormality

103 in fruits shape. The wilting of crops refers to the loss of the firmness of the non-woody parts of the crops like leaves or young shoots due to loss of water in crop cells. This is proposed to be the impact of increase in temperature where there is increase in water loss from both soil and plant from increased evaporation and transpiration. The results indicate that all the groups observed wilting as the major impact which clearly substantiates significant impacts of climate change in crop production. According to the observations of the Bellonese communities, the increase in heat stress which results in wilting of crops negatively affected the survival rate of newly planted crops. They stated that the impacts of the increased temperature include scorched young crops, leaves of the crops turning from green to brown resulting in stunted growth as well as crops that die as a result of permanent wilting.

All the groups also argued that the decline in yield of crops correlates with the increase in temperature. This is because reductions in crop yields observed are observed to be related to the impacts like wilting, declined in survival rate, decline in tuber size, early maturity and decline in crop growth and increase in pest/diseases.

For example, wilting and the decline in survival rate of newly planted crops reduced yield because crop density or population decreased resulted in low production. Early maturity, decline in crop growth and increase in pest and disease also reduced yield because less number of tubers or fruits are produced, whilst the decline in tuber size affects the mass of the yield. The Bellonese respondents observed that with early maturity, the size of the tubers of root crops and fruits were smaller with some deformity. For instance, the elderly and women’s group mentioned that the size of yams 40-50 years ago were so large that they had to stand upright while peeling or grating them. They stated that while those yams were about 2-4 feet in length,

104 present size of yams rarely reach a foot and a half. In terms of the morphology and shape of the yams, irregularities in the shape were observed where shapes of yams today are irregular with many crevices while those in the past reflect more of a whole cylindrical shape.

Similar to temperature, rainfall trend was reported by all the focus groups to have increased over the past 30 years, having serious impacts on the Bellonese food crops.

The impacts reported by the focus groups include:-

 Decline in yield

 Shifting of planting date

 Rotting

 Loss of crop varieties

 Increase in Pest and Disease

 Shortage of planting materials

 Impact on flowering and fruiting of crops

 Delay in maturity

 Loss of taste or flavor

 Decline in growth

 Decline in survival rate

 Increase in weeds

Figure 25 outlines the various impacts reported by the focus groups expressed as percentage. For instance, all of the five focus groups (5/5 x 100 =100%) state a decline in yield, rotting, increase in pest and disease, delay in maturity and loss of taste (flavor) as the most severe impacts of increased rainfall. Most of the groups

(80%) identify the impacts on flowering/fruiting and decline in survival rate as the

105 second most severe impact on the crops. Meanwhile, 60% claim the loss of some crop varieties and 40% identify decline in growth and increase in weeds as some of the adverse impacts of increased rainfall.

Impact of increase in rainfall on food crops

Decline in yield

Rotting 40 100 80

40 100 Loss of crop varieties

100 Increase in Pest/Disease 60

Shortage of planting material 100 100 Effect on flowering and 80 100 fruiting Delay in maturity

Figure 25: Responses of the five focus groups expressed as percentage on the impact of increase rainfall trend on food crops in Bellona

Decline in yield is (by far) the most significant impact because it not only affects the production of the crops but the livelihoods of the people who depend on it. Rotting was reported to be severe because of excess water retained in soil over long periods caused by heavy rainfall. The Bellonese people reported that rotting affects the roots of the crops like slippery cabbage, pumpkin, water melons, and banana. It also affects the tubers of sweet potatoes, yam and pana with corms of taro. Some of the members of the groups also mention that the stem of the bananas are also subject to rotting during prolonged rainy conditions. It is unclear if the retention of water on the stem provides favourable conditions for diseases or pest to fungus to cause the rotting process. For instance a study conducted in Sri Lanka identified that a fungus

Marasmiellus sp that causes stem rot in bananas is commonly growing and spreads

106 during rainy season (Thiruchchelvan et al., 2012). The study further claims that this fungus affects bananas of all ages where young bananas experience wilting of leaves and death while in middle to mature bananas experience hindered fruiting or reduced size of the bunches. All the groups interviewed highlight that the increase in rainfall further exacerbate the population and infestation of pests. The most common pests identified are Bongu (Taro beetle, Papuana spp) that feeds on almost all root crops and the Takuku slug (Leidyula floridana) which feeds mainly on plant tissues-leaves of most vegetables and some root crops. Respondents also asserted that the increase in rainfall is somewhat linked to the increase in the population of these pests as well as their infestation. Increase in rainfall with excess water in the soil has been reported as providing an adequate condition for these pests to attack the plants. This is because of three reasons;

1. The water causes the soil to be softer and more loose around the tubers or corm

providing the adequate accessibility for a pest like Bongu to easily attack the

tubers or corm.

2. Excess water induces rooting of tubers or corms allowing easy targets for the

pests.

3. Excess water also enhances the rooting of not only tubers, corms but other dead

vegetation like tree trunks, twigs leaves which are favourable habitats and

microhabitats for these pests thus will allow more reproduction and

distribution.

During the field visit (July 2012), the impacts of Bongu were clearly evident on

taro corms and young taro crops (Figure 26). There were tunnels which form large

cavities identified in the taro corm as a result of the burrowing activities of Bongu

107

(Figure 27). Some of the young taro crops even resulted in wilting and death as the

damage caused was very severe (Figure 28 and Figure 29).

Figure 26: Tunnels created by Bongu Figure 27: Large cavities caused by Bongu

Figure 28: Wilting of young taro crop Figure 29: Completely eaten root and corm by Bongu

Delay in maturity and loss of taste (flavor) in food crops was also reported by all of the focus groups. They stated that the delay in maturity is a consequence of increased rainfall which damages the flowers of some vegetables (pumpkin and water melons) and fruit trees (pawpaw, bread fruit, and coconut). The increase in rainfall affects the pollination process thus resulting in delayed of expected time of harvest and reduced the resulting yield because only few flowers have survived and pollinated. Root crops

108 too have been reported to have delayed maturity as excess water in the soil reduces the soil fertility, slowing the growth of the crops to reach maturity. However, sweet potatoes are reported to have increase vegetative growth but less in tuber formation and production as a result of increase or changes in rainfall. The Bellonese reported that increased rainfall results in increased vines growth but reduced tubers so the young shoots are cut off to be consumed as cabbage. Results from interviews indicated that sunlight induces tumor formation and production. Increased rainfall often implies the growth of fewer tubers. Root crops such as sweet potatoes, taro, yam and pana were reported to be “less tasty” or lost their flavor, as a consequence of receiving too much rain. All the groups reported that that root crops planted on soils that retain more water lose their taste (flavor) compared to the ones planted on the more limestone, loosely packed soils. The amount of mineral or elements content in a soil play an important role as some minerals or elements like potassium, phosphorus and magnesium may determine to some extent the taste or flavor of tubers (Flis et al.,

2012). This study highlights that the there is also a significant relationship between the location of the potatoes grown and their flavor or taste (Flis et al., 2012). Therefore, this present study believes that the increase in rainfall may affect taste or flavor in root crops through the impacts of leaching where increase in water infiltration or percolation can influence the washing away of nutrients, minerals or elements that could have the potential to influence the taste or flavor of the tubers. The physical impact of increase rainfall as indicated by the Bellonese respondents reduces the survival rate of the newly planted and young crops. This reduction is believed to have been caused by the increased surface runoff and overland flow which washes away young and newly planted crops. The impact is significant on crops like water melon

(Citrullus lanatus), pumpkin (Cucurbita pepo), sweet potatoes (Ipomea batatas), and

109 slippery cabbage (Hibiscus manihot / Abelmoschus manihot). Two third of the groups

(60%) reported that some varieties of crops were lost as a result of increased or change in rainfall. The varieties of crops reported include Sinamu (Dioscorea nummularia) an old type of yam, Boiato (Dioscorea pentaphylla) also an ancient yam and Abubu

(Dioscorea bulbifera) a bulbil-bearing yam with thorn-less vines. A report on these three yams during the period of 1962-1966, states that while all of the yams are be old and ancient, only Sinamu is common whereas Boiato and Abubu are rare (Christensen,

1975). A further study carried out on crop diversity and genetic erosion states that among the 31 farms investigated, Sinamu is found in 21 farms, Abubu in 17 farms while Boiato was only found in 1 farm (Walkenhorst, 2005). This same author suggests that in comparison to Christensen’s (1975) study in 1960s, there was a decline in varieties of Abubu from 10 to 3 and Boiato from 5 to 1. His study concluded that pests, introduction of new crops, change of farming practice and climate change, are the reasons for the loss of crop varieties. In our study, 60% of the focus group stated that change or increase in rainfall influences the loss of the yam varieties; this though could not be substantiated with confidence. This is because other issues like change of farm management practice, lifestyle or choice of food could be responsible for the loss of the yam varieties (Walkenhorst, 2005). In fact, a detailed study on yam diversity and climate change would be useful in determining with confidence whether or not climate change is responsible for the loss of the yam varieties. About 40% of the focus groups identified increase in weeds as a consequence of change and increase in rainfall. This is because the prolonged days of rainfall prevented the farmers from attending to their gardens thus resulting in reduced labour input. For example, increased rainfall and prolonged days with rain prevents farmers from burning their gardens for new cultivation thus further increases the need for more labour input as the

110 weeds and shrubs recolonize the once cleared land. Some stated that increase rainfall encourages the growth of weeds thus increasing their competition against crops for space and nutrients.

5.3 Past and current change in extreme events (cyclones and droughts) and correlating impacts 5.3.1 Observation on the trend of cyclones and droughts

The focus groups were asked to identify the cyclones and drought events experienced in the past 30 years and to establish a timeline of the occurrence based on their observation and the impacts that these extreme events on their food crops. They were also asked if they had noticed any changes on the frequency and intensity of these two extreme events. Figure 30 below is a summary timeline drawn from the respondents of the five groups on the occurrence and strength of cyclones and droughts. It has been observed by the focus groups that in the past years prior to 2001, major cyclones normally occur within a span or interval of 7 years. For instance in

Figure 30 there is a 7 years interval between Cyclones Keri, Namu and Nina. The trend of 7 year interval was reported as allowing the Bellonese to predict when the next cyclone will happen and therefore prepare ahead before it strikes. However, after cyclone Nina, the interval of time span of occurrence of cyclones not only became inconsistent but declined. Thus, generally, it is reported that occurrence of cyclones have been observed to decline however the strength has been noticed as variable. According to all focus groups, several small storms or cyclones happened between the major cyclones outlined in Figure 30 however their impacts were not significant. For example, the men and elderly group reported a form of strong wind referred to as Hungi kenge which occurred around the 1990s. Hungi kenge was said

111 to be very destruction to crops and has the potential to dig up the soil because of its extreme strength.

Weak M Moderateoderate StrongStrong Severe

1979 CycloneCyclone KeriKeri Drought-1979Drought-1979

Y 19861986 CycloneCyclone Namu E Drought-1986Drought-19866 CycloneCyclone Nina A 19931993 R Drought-1997/1998Drought-1997/19988 20012001 CycloneCyclone AbigailAbigail

2010 CycloneCyclone Ului CycloneCyclone Droughtt

Figure 30: Timeline of cyclones and drought occurrence over past 30 years

Focus groups reported that droughts normally occurred after cyclones, while some within the groups mentioned that it occurs both prior to and after cyclones. The latter

(droughts occurring both prior and after cyclone) is similar to what was recorded by a survey carried out by Reenberg et al., (2008). All the focus group mentioned that they noticed change in frequency and strength of drought. However, in terms of frequency, they reported a decline in the long term drought (months) however an increase in the short term (weeks) drought. In terms of drought strength ( the impact of heat stress associated with drought on food crops as observed by the Bellonese), they mentioned that though long term droughts were declining in strength, the short term droughts are increasing in strength. Short-term droughts were reported to adversely affect crops, mostly because of extreme heat stress on plants.

112

5.3.2 Impacts of cyclones on crops

The five groups identified the following as impacts of cyclone on their crops:-

1. Uprooting of crops 2. Rotting 3. Physical damage on crops by fallen trees and debris 4. Shortage of planting material 5. Washing away of young and newly planted crops 6. Death of tree crops 7. Increase in pest and diseases 8. Impact on copra industry 9. Increase in heat stress after cyclone causing wilting of crops 10. Decrease in fruit size

The percentage of the respondents of the five groups on the list of impacts is represented in Figure 31. All of the groups (100%) identified uprooting of crops, rotting, physical damage by fallen trees or debris, washing away of young crops and increase of pest and disease as significant impacts. Uprooting of crops is said to be caused by the extreme strength of the wind. Not only were the crops uprooted but also the supporting stakes for crops like pana and yam (Figure 32).

Percentage of respondents by focus groups on the impacts of cyclones on food crops Uprooting of crops

Rotting 40 20 100 80 Damage by fallen trees

100 Shortage of planting material 100 Washing away of young crops

100 Death of tree crops 80 80 Increase in pest/disease 100 Effect on copra industry

Figure 31: Percentage of respondents of the five groups to the impacts of cyclones on their food crops 113

Figure 32: Supporting stakes used for pana and yam in East Ghonghau

This has significant impact on the two root crops as they heavily rely on the stakes for support; thus such uprooting adversely affects the yield. The impacts of winds associated with cyclones physically damaged crops, by causing nearby trees to fall, crushing understory crops of vegetation. The intense amount of rainfall associated with cyclones are said to induce rooting of tubers and corms of root crops as well as physical washing away of newly planted or young crops like sweet potatoes, cassava, water melon, slippery cabbage and corn. Excess water in soil has been reported to be the catalyst for rotting of tubers; providing best condition for pest like Bongu that infect the crops like taro.

Most of the focus groups (80%) highlight a shortage of planting materials, death of tree crops and impact on their copra industry as significant. The shortage of planting material is mainly applies to root crops like pana and yam. This is because excess water influences rotting of these two crops, inducing shortage of planting material for the next planting season as tubers are the planting materials. Death of tree crops like 114 pawpaw and coconut has been tied to the impact of the strong winds that normally tear off the heads of these tree crops or uproot them. Cyclones are reported to be one of the main causes of crippling the copra industry on Bellona. It was reported that cyclone Namu and Nina in 1993 extensively destroyed most of the coconut plantations on Bellona. According to one of the members of the men’s group, Mr.

Eric Mamu, speaking in Pidgin English; I quote;

“After cyclone Nina if you stand lo western side lo Bellona ba u save lukim go nomoa the houses lo West Ghonghau lo centre lo island ya. Kolasp every tall coconut ota destroyed by Nina”(Mamu, 2012).

Translating this, Mr. Eric Mamu suggests here that cyclone Nina destroyed most of the taller such that it is possible to see houses in central Bellona even from the western end of the island. Due to extensive damage caused by cyclone Nina, copra farmers were discouraged to continue with the copra industry. While the decline in copra production interest is also linked to the inconsistency in shipping services and the correlating high costs like the cost for freighting, cyclone is claimed to be the dominant cause for crippling the copra industry. For example, copra production ceased in the mid 1980’s when cyclone Namu (1986) occurred in Bellona which devastated most of the coconut palms (Reenberg et al., 2008). Some of the key informants mentioned that coconut palms are responsible for much of the damages on both their crops and houses. They argued that if coconut palms were not introduced as a cash crop, much of the damages on Bellona would have been avoided or minimal.

Coconut palms were reported to be easily brought down by strong winds of cyclones and pose a high risk to human lives, infrastructure and crops. There are still coconut

115 plantations observed at Bellona at the time of the field visit but are only use for food, building materials and handicrafts.

Only two focus groups (40%) mentioned the impact of heat stress on crops. This is because most shade trees or vegetation were destroyed by cyclones. This impact was reported by the men and farmers’ groups who reported that after cyclone, crops tend to wilt because of the impact of heat stress caused by direct sunlight on the crops.

They added that when natural shade trees and canopies which filter direct radiation of sun have been lost, their crops were exposed to heat stress. Only the women’s group identified the reduction of fruit size as an impact of cyclones. Examples of the fruits of tree crops like coconut, bread fruit and pawpaw were mentioned. From the women’s observation, after cyclones these crops bear fruits that are smaller than usual and some have abnormalities like deformations and crevices. The group also mentioned that coconut palms produced less number nuts which are mainly is used for preparing Bellonese dishes (Figures 33 and 34) like Tokonaki or Songo

( puddings) , Lengalenga ( a mixture of taro corms and leaves mixed with coconut cream), Pota ( taro leaves and coconut cream) and for drinking.

Figure 33: Lengalenga dish Figure 34: Pota preparation

116

5.3.3 Impact of droughts on crops

The five Bellonese focus groups were asked to identify the impacts of droughts on their food crops. Their responses clearly indicated 8 types of impacts which are:-

1. Wilting of crops,

2. Death of crops,

3. Reduction in size of fruits and tubers,

4. Abnormality in crop growth,

5. Early maturity,

6. Decline in yield,

7. Deformation of tubers, and;

8. Delay in maturity of bananas.

The respondents from the five focus groups were then summarized in percentage form which is represented in Figure 35. All of the groups (100%) acknowledged wilting, death of crops, early maturity, and reduction in size of fruits and tubers as well as decline in yield as significant impacts of drought on crops. Wilting as a result periods of no rain negatively affects most of the crops including vegetables like pumpkin (Cucurbita pepo), water melon (Citrullus lanatus), slippery cabbage

(Hibiscus manihot / Abelmoschus manihot) and other root crops. The fact that the

Bellonese cropping system is dependent on rainfall for irrigation and that they do not have surface water implies that manual irrigation during drought events is not possible. This however causes the extent of wilting by heat stress to be very severe, as a result, death of crops especially the young or newly planted is common especially when additional input of water to sustain adequate crop growth during drought events is insufficient.

117

Percentage of respondents by groups on the impacts of drought on food crops

Wilting

200 Death of crops 40 100 Reduction in sizes of fruits & tubers 100 Abnormality in crop growth

Early maturity 100 Decline in yield

Deformation of tubers 100 Delay in maturity of bananas 100 80

Figure 35: Percentage of respondents of the five groups to the impacts of droughts on their food crops

Early maturity in root crops (tubers) was observed by all the five focus groups, yet they mentioned some crops were shown to produce smaller than normal tuber at maturity.

This has led to a belief within the Bellonese communities that drought causes decline in the yields of their food crops; both in sizes and the number of tubers or fruits. An example of this is one farmer’s account quoted below:-

“Lo time wea hem no rain fo long time ya osem drought ya, me only save harvestem nomoa onefala or sometimes threefala small fruit kumara lo one fala hill. Fruit blo kumara ya smol osem nma fist blo me ya. But lo time we hem no drought ya me save tekem abaot 7 or 10 fala fruit osem lo one hill”(Kiwa, 2012)

Translating this, Kiwa (2012) is implying that during extreme drought, he can only harvest 1-3 small sized (smaller than his fist size) tubers per mound of sweet potatoes.

118

However, during ambient conditions, he can harvest about 7 to 10 tubers per mound.

The size of banana fruits and bunch was also reported to reduce in size by some of the participants.

Abnormality in crop growth and physical structure is observed by most of the groups

(80%) where they mentioned that some of their crops indicate stunted growth and some loss their rigidity. A common example given by the participants is that of sweet potatoes where during droughts, the vines of the sweet potatoes grew to about less than 1 foot or just cover the area within the mound circumference then stop growing.

Some of groups (40%) mentioned that droughts also affect the shape of morphological characteristic of their root crops. This impact is dominant in yams as have been described previously in section 5.2 under the impact of increased temperature. Only the elderly people’s group identified drought as having adverse impact on bananas by delaying their maturity. That is, depending on the varieties of bananas, normally, it takes about 9 to 12 months for maturity. However with the drought impact, it is said that it takes more than the normal harvesting timeframe for their bananas to reach maturity.

119

CHAPTER SIX: BELLONESE ADAPTATION AND COPING STRATEGIES TO CLIMATE CHANGE AND EXTREME EVENTS

6.1 Introduction

This chapter outlines the responses of the surveyed households of Bellonese communities including farmers’ perspectives in terms of adaptation and coping strategies to impacts of climate change and extreme events on their food crops and cropping systems. Within the context of the study, adaptation strategies are described as being the long-term or planned measures taken by the farmers or Bellonese communities, to adapt to the impacts of climate change and extreme events (IPCC, 2007d; Senbeta, 2009). Coping strategies are described as short term responses taken after the impacts of climate change and extreme events have occurred (Mutsvangwa, 2010). However, coping strategies may develop into adaptation strategies or sometimes it can be difficult to distinguish between the two strategies because they may overlap (Berkes and Jolly, 2001; Senbeta, 2009). For instance, mulching can be an adaptation strategy implemented before the impact of drought or it may also be a coping strategy implement during the drought event. Traditional early warning signs are important adaptation strategies that the Bellonese use to predict the onset of extreme events such as droughts and cyclones. Nevertheless, this tradition is slowly lost over time and only a few elderly people still have this traditional knowledge.

This chapter will identify the adaptation strategies and describe their significance in the first subsection. The second subsection will focus on identifying and evaluating the coping strategies and finally elaborate on the traditional early warning signs.

6.2 Adaptation Strategies

6.2.1 The use of shifting cultivation and fallow practice

The history of shifting cultivation in the Pacific islands started about 4000 years ago (Kirch, 1996). Bellona has a long history of the practice of shifting cultivation according to several studies in the past (Christensen, 1975; Reenberg et al., 2008; Breuning-Madsen et al., 2010) and more recent (Mertz et al., 2011). Responses from 120 both the household survey and focus groups interviews reveal that shifting cultivation and the use of bush fallowing and burning is still widely practiced in Bellona. About 95% of the households reported that they use fallow periods ranging from 2 to 10 years (Figure 36). The shifting cultivation and fallow periods have been identified as adaptation strategies because the process of shifting from one area to the next allows the soil in used areas to replenish and uphold its fertility (Christensen, 1975).This is because during fallow period wild vegetation extracts nutrients or ions from the soil and concentrates them in their tissues. Thus, since the practice of Bellonese cropping system is fallow burning, nutrients and ions from the wild plants are being circulated back to the soil for the new crops to use via burning and rotting.

The use of fallow period periods and shifting cultivation as an adaptation measure is seen as a highly significant strategy by all the households and focus groups none of whom reported the use of fertilizers or any other form of manures as crop management input.

Percentage of households enaging in fallow, crop rotation and intercropping practice

74.6 95 Fallow Crop rotation Intercropping

100

Figure 36: Percentage of households engaging in fallow, crop rotation and intercropping practice

In other words, since this form of strategy (fallow and shifting cultivation) naturally allows the soil to regain fertility, it indeed is a more environmental friendly approach. Using agro chemicals such as fertilizers have the potential to improve crop productivity, but also contribute to greenhouse gas emissions that influence climate change. Shifting cultivation and using fallow periods also enhances crop productivity. 121

6.2.2 Crop rotation, intercropping and diversification of crops grown

Crop rotation is the practice of cultivating different sequence of crops on the same plot of land. It is a beneficial approach to balance, manage and improve the fertility of the soil, improve soil structure, avoid excessive depletion of soil nutrients, as well as control weeds, pest and diseases (Jackson et al., 2011; Berhanu and Gerald, 1998). All of the households interviewed in Bellona reported having practiced some form of crop rotation. A common practice among the Bellonese in terms of crop rotation is that they cultivate sweet potatoes, corn and melon subsequent to other major root crops like yam or pana. This is because of the resilient attributes of sweet potatoes to nutrient demand and other reasons as provided in Chapter 4 section 4.2.3. Figure 37A is a sequence of crop rotation reported by the Bellonese respondents during our study which is similar to the study by Walkenhorst (2005) depicted in Figure 37B. The study by Walkenhorst (2005) supported our findings that sweet potatoes were commonly used in crop rotations normally subsequent to other main crops such as yam or pana.

Taro, banana or yam, pana Banana or Yam, Pana

4-6 months fallow Three months fallow

Melon Melon

Corn and Sweet potatoes Sweet potatoes

Fallow-1-6 years Fallow

A B

Figure 37: Sequence of crop rotation in Bellona (A) in our study in 2012 and (B) by Walkenhorst (2005)

About 74.6% of the households practiced intercropping. Intercropping is a crop management practice where more than one crop type is planted on a given area. In Bellona, it is a common practice to plant a main crop with secondary associates. The choice of crops used for intercropping in Bellona is based on milieu of the main crop and the timeframe required for full development (Reenberg et al., 2008; Christensen,

122

1975). Table 4 in the Chapter 3, section 3.2.6 outlines the common forms of intercropping observed in Bellona during the period of our assessment (July 2012) which closely resembles that observed in the 1960s (Christensen, 1975).

Intercropping as an adaptation strategy is a very significant approach as it maximizes and utilizes different soils, providing additional yield without additional work since different crops may create local environment that other crops may exploit better (Reenberg et al., 2008; Christensen, 1975). Thus, preservation of crop rotation and intercropping in Bellona as an adaptation strategy is crucial to adapt to the changes in climate and severity of the extreme events.

6.2.3 Maintaining shade trees after clearing, burning and mulching

In Bellonese cropping system, the extent to which clearing and burning is undertaken to prepare an area for a garden plot is dependent on the crops that will be planted. For example, for yam, pana and taro gardens, some trees will be retained for the purpose of providing shade and also as support stakes for yam and pana. Figure 38 below illustrates a taro farm in West Ghonghau planted under Hau trees (Hibiscus tiliaceus) which are not cleared or removed during bonga (brushing or clearing) and baakani (burning) of the garden area. The Hau trees are left on the garden as they are also considered important by the Bellonese for soil fertility recovery. The dead leaves, twigs and ash remains from the bonga and baakani are left as mulch on the soil. These are maintained to prevent the soil from over drying and ensuring that soil moisture is adequate. Mulch has been reported to not only protecting the soil and crops from heat stress or extreme temperatures but also aids in adding nutrients to the soil.

123

Figure 38: Taro planted under Hau trees (Hibiscus tiliaceus L) with rich mulch of dead leaves and twigs maintaining the soil moisture and preventing the impact of extreme temperatures

6.2.4 Planting fast yielding and resilient crops

The cultivation of early or fast yield crops has become increasingly common in Bellona. Such crops include sweet potatoes (Ipomea batatas), water melon (Citrullus lanatus), Chinese cabbage (Brassica sp.) and corn (Zea mays). Cassava (Manihot esculenta) is also cultivated but not on a large scale. Varieties of sweet potatoes (Ipomea batatas) cultivated in Bellona have increased from 6 varieties reported in 1960s to 12 varieties in recent years (Christensen, 1975; Walkenhorst, 2005). The most common variety being planted at present is the variety Trimanisi (Figure 39) which means three months referring to the time it takes from planting to harvest. Results indicated that about 80% of the households cultivated sweet potatoes (Chapter 4, section 4.2.3). Thus, it can be said that cultivating a fast or early maturity crop like sweet potatoes is an important strategy to adapt to the impacts of climate change and extreme events because it will take less time to harvest. Not only is the time span short but sweet potatoes and other crops like water melon, corn or cassava require less labour and shorter fallow periods input compared to yam or pana. Cassava is considered a climate resilient crop in Bellonese communities, even though it is less common. 124

Figure 39: Newly planted Trimanisi sweet potatoes in East Ghonghau

6.2.5 Adjust planting dates

Traditionally, Bellonese people have a planting calendar for their crops. A detail elaboration of this calendar is provided by (Christensen, 1975). Thus, a comparison is made on the planting dates as collected from farmers in the present study (2012) to what has been recorded in the 1960s.

Table 17: Bellonese planting calendar

Crop Christensen 1960s 2012 Planting Harvesting Planting month Harvesting month Pana November April September- February (lesser yam) Yam August May July April (greater yam) Taro Any month-no - Any month-no - season season Banana Any month-no - Any month-no - season season

Only pana (lesser yam) and yam (greater yam) are seasonal crops in Table 17. This implies that these two crops require certain time of the year to be planted in relation to the correlating favourable weather condition. There is a slight shifting of planting

125 date for both pana and yam currently practiced as compared to the 1960s calendar (Table 17). According to the responses of the key informant interviews in focus groups, pana is currently planted in September. This is two months earlier than what had been practiced in the 1960s. The planting date of yam has shifted a month back compared to that of about 50 years ago. However, the planting dates given for 2012 are not the same for all the respondents. For instance, some mentioned that pana is being planted in January and harvested in June. Nevertheless, the essence of shifting or adjusting these planting dates is, according to the farmers, the result of two factors. The two factors are (1) the changing and unpredictable weather pattern that they observed and (2) the sensitivity of yam and pana to soil fertility and extreme climatic events. Consequently, planting dates must be adjusted such that the crop yield will be sufficient and support their consumption requirements.

6.2.6 Increase number of plots, change planting site and increase mound size

About 29% of the households indicate using the strategy of increasing the number of their plots or planting sites to ensure that when extreme events occur, ample supply of food supply is available That is, while some areas of their plots may be more exposed or susceptible to wind damage or increased surface runoff, they can depend on the other less vulnerable plots. However, our results indicated in Table 13 Chapter 4 in section 4.3 that there is a decline in the number of plots per households compared to studies in 1960s and 2011. This might be the reason that only an insignificant amount of 29% mentioned increasing plots as an adaptation strategy.

In relation to this, 36% of the households state that instead of continuous planting on vulnerable areas, they shift their garden plots to less vulnerable and more fertile areas. This is why most plots are located on the centre of the island in Ghonghau ward where it is believed to be more fertile and less susceptible to both climatic and non climatic factors such as pest and diseases. There is rather an insignificant number of households (9%) that consider increasing the mound size in sweet potatoes yams and pana, will promote higher yields. They asserted that the bigger size mounds, the greater the chances of protecting the tuber and roots from desiccation during drought

126 and the less vulnerable the crops are to increased or heavy rainfall. Furthermore, bigger size mounds are claimed to provide ample space where more tubers can be accommodated and may provide more nutrients compared to smaller size mounds.

6.2.7. Planting Distribution and method

Some of the key informants in the focus group interviews indicate that they also cultivate their crops in a pattern such that crops are planted in line with the soil fertility and depth. This is similar to studies carried out on Bellona in the past (Christensen, 1975; Reenberg et al., 2008). This form of planting pattern is referred to as dividing (tohitohi) the garden into sections (potu) according to soil fertility. This is considered as a significant adaptation strategy as it regulates the planting of crops according to their fertility requirements. For instance, it allows sensitive crops like yam or pana to be cultivated in highly fertile areas on the plot while establishing more resilient crops like sweet potatoes on less fertile areas. One of the farmers mentioned that the planting depth of crops is vital in terms of yield production and its exposure to climate extremes. He stated that the planting depth of the taro varieties depends on the bearing size of the corms. For example, the varieties Tango Sua and Kamaamangu which have bigger corms have a planting depth of about 30cm, whilst for Tango Ngeka which bears smaller corms the planting depth is about 15cm. He further claimed that if the bigger taro varieties (Tango Sua and Kamaamangu) are not planted according to the required depth of 30cm, they are easily damaged by strong winds and their corms will be exposed to heat stress and pests.

6.3 Coping strategies

6.3.1 Pruning of crops

All the focus groups mention that the pruning of sweet potatoe vines is normally done when massive growth of vines during prolong rainy seasons is observed. That is, sweet potatoes during prolonged periods of rainfall experience limited tuber production but increased growth of vines. The Bellonese believes that pruning the vines initiates tuber initiation and high yield production. Despite the pruning off of

127 excess vines, tuber production normally is low during prolonged rainy season; thus, the pruned vine cuttings are used as vegetables where they are cooked as cabbage. In a way, they cope by maximizing the vegetative part of the sweet potatoes when tuber production is limited. Another coping strategy practiced on Bellona is that during cyclones or extreme storms, when bananas trees are blown down to the ground, their stems are being cut to initiate regrowth.

6.3.2 Using common table salt as a pest deterrent

Though the Bellonese do not use pesticides, insecticides or chemicals for pest or disease control, two groups (farmers and elderly people) mention the use of salt to help control or prevent the infestation of a pest insect known locally as Akaua; an insect that destroys taro leaves and stalks as well as bananas. However, no study has been conducted to analyze the effectiveness of this strategy.

6.3.3 Gathering and consumption of wild crops and harvest conservation

During extreme events such as cyclones or drought when most cultivated crops are severely affected many Bellonese people collect and gather wild crops as substitute for consumption. One such important crop is the wild yam commonly known as “Black battery” which is claimed to have high calories and can sustain a person for longer periods of time. Christensen, 1975 found that there are other wild fruits that are collected for consumption including Morinda-Indian mulberry tree (nguna) and Spondias fruits-Tahitian apple (bii). Most of these edible fruit trees are not cultivated but were preserved or not removed when the vegetation is cleared. Nguna and Bii are also considered by Bellonese as important plants for the recovery of soil fertility in fallow gardens.

Another important coping strategy practiced can be referred to as “conservation or sustainable harvesting”. This has been reported during our study as well also in a study carried out about 50 years ago (Christensen, 1975). The Bellonese normally leaves small tubers of yam unharvested to be utilized later during periods of short food supply or when cyclones or drought destroy their food crops. Edible fruit trees

128 such as Isi (Polynesian chestnut), Banga (Cut nut), Ghaapoli (Ficus spp.), and bii (Tahitian apple) are often preserved and not cleared during garden forest clearance. These fruit trees provide for the Bellonese during times of food supply shortage and disaster. Others in the focus group interviews state that they also rely on their old garden plots which have been left to fallow as some of root crops developed new tubers.

6.3.4 Change in eating habits

Our study indicate that about 83% of the households rank processed foods as first and second most important source of food. These processed foods include imported rice, Chinese noodles, and canned tuna. Although our study did not focus on the exact details on the types of food consumed on a daily basis, it is obvious that the eating style of the Bellonese has change compared to the last 50 years. This is supported by a study carried out on Bellonese communities where the survey revealed that about 90% of households eat rice as the main staple (Reenberg et al., 2008).The shift in habits from the traditional or agricultural crops to more on processed food may not be an advantage to preserving traditional crops and farming techniques, yet is also a coping strategy during periods of extreme events like drought or cyclones. The key informants in focus groups state that food supplied (relief) after an extreme event (cyclone) plays a great role in influencing the Bellonese to change their eating style. This is to say that this changing eating style is seen a coping strategy where during extreme events, the Bellonese impose less pressure on their food crops supply. However, the changing preference of eating to processed food may influence the change of attitude towards the traditional Bellonese farming practice or preservation of the traditional crops.

6.3.5 Barter system

According to the Bellonese men and elderly people’s focus group sessions, traditional barter systems are an important climate change coping strategy. They claim that only a few people actually practice this today; mostly focusing on garden fruit crops and fisheries.

129

6.3.6 Relief from external sources

Relief in terms of food supply, tools, building materials, tanks, seeds or planting materials of fast yield crops (sweet potatoes, water melons, cabbages) are usually provided by the National or Provincial Government, NGOs (Red Cross, World Vision), National Disaster office and other Foreign Aid donors. Such reliefs are significant coping strategies in that they ensure that the Bellonese have enough food supply and resources to last them until they are able to commence new cultivation and recover from the extreme event. However, as have been reported by the focus groups, it is mostly during cyclones that such reliefs are received. They claim that during droughts, there is limited external relief. They mentioned that the responsive attitude of external relief towards cyclones outweighed their attitude towards drought on Bellona.

6.3.7 Migration, remittance, selling of handicrafts and family support

Bellonese people have a strong cultural, social and communal relationship which is a very crucial coping strategy for such a remote island community that is highly exposed to extreme events and climate change. Migrations to the main urban centre, Honiara, and other centres for employment on both a short and long term basis is not uncommon. On a short term basis, the Bellonese travel to Honiara to sell their handicrafts like wooden carvings and woven baskets then they send money back to their families. The Bellonese wooden carvings and woven baskets are some of the main attractions for tourism in Honiara and the Bellonese obtain good value for these local products. Some of the Bellonese who do not travel to Honiara send their products to their relatives who will sell them on their behalf and send back money, food or other necessities. Others migrate on a long term for employment where they are able to send money and food supplies back to their families in Bellona. Such a cultural and communal identity is a very significant coping strategy for the Bellonese people who are living in remote location, exposed to climate extremes and have limited access to better infrastructural services and economic activities.

130

6.3.8 Traditional warning signs to extreme events

During an interview, a Bellonese elder Tongabaea (2012) suggested that traditional early warning signs of extreme events such as cyclones or drought are significant adaptation measures because they allow the Bellonese to prepare for when cyclones or droughts do occur. In terms of cyclones early warning signs include:-

1. Pana vines growing rapidly and reaching the end of the support stake and then falling to the ground. This serves as a signal to farmers that a big storm or cyclone is soon to happen. 2. Fish not responding to baits and catching fish proved very difficult because it seems as though the fish are very sensitive and alert to their surrounding, ocean current or waves. This also signifies to the fishermen that a big storm or cyclone is approaching. 3. Fruit or nut trees such as Cut nut (Barrigtonia spp.) and Isi (Polynesian chest nut) bearing extremely large nuts and bunches compared to normal condition indicates that a storm or cyclone will soon strike (Tongabaea, 2012).

For droughts, only one traditional early warning sign was mentioned by Tongabaea (2012). That is, the Ibis (Australian Ibis- Threskiornis molucca) pictured in Figure 40 that normally feeds on the ground will fly abnormally high as compared to the normal flight height. This is an indication that a prolonged period of dry weather or no rainfall is approaching.

Figure 40: An Australian White Ibis pictured in Tangakitonga, West Ghonghau 131

CHAPTER SEVEN: SIMULATING AND EVALUATING FUTURE IMPACTS OF CLIMATE CHANGE ON TARO, CASSAVA AND CORN

7.1 Introduction

Simulations in DSSAT were conducted for the three crops; Colocasia esculenta (Tango Sua), Manihot esculenta (Lioka B1) and Zea mays (Koni).The simulations were performed using SUBSTOR-Aroid model for Tango Sua, CROPSIM-cassava model for Lioka B1 and CERES-Maize for Koni on the impacts of current or ambient (2012) climate and future change in climate in terms of temperature, rainfall and carbon dioxide on the yields of taro, cassava and corn. The future change of the climate variables are in accordance with the PCCSP climate change projections for 2030, 2055 and 2090 which has been described in Chapter 3 section 3.4.7.6. Simulations were performed for each of the three crops using four treatments (1, 3, 4 5) for Lioka B1 and Koni and four treatments (2, 3, 4 5) for Tango Sua. Details of the treatments are described in previous Chapter 3 section 3.4.7.6.2.

List of treatments 1. Ambient potential production (100% light) 2. Ambient potential production (50%) light (only for Tango Sua) 3. Ambient attainable (Nitrogen and Water turn on) 4. Attainable for projected temperature, rainfall (2030/2055/2090)

5. Attainable for projected temperature, rainfall and CO2 (2030/2055/2090)

Simulations were performed for the three crops for three sites namely East Ghonghau, West Ghonghau and Sa’aiho based on the same weather and crop management data but with different soil type data. The three sites were selected because they represent the three main soil types (Kenge ungi, Kenge toaha and Malanga) that are commonly used for cultivation. This is mainly to see how each of the three crops responds to the impact future change in climate based on three different soils.

This chapter presents and discusses the results of these simulations on each of the crops. Further, simulations were also conducted on the impacts of climate variability

132 related to ENSO (El Niño) for the past 30 years on the yield of the three crops. Eleven (11) El Niño years within the past 30 years as per PCCSP report were selected and simulated against the three crops. In addition, a brief analysis comparing the Bellonese perspectives and model simulations specifically on taro (Tango Sua) in terms of the most severe drought in 1997 is also made at the end of the chapter. The reason of choosing Tango Sua is that it is one of the most important food crops and staple in Bellona and the Crop model for taro (SUBSTOR-Aroid) is available in DSSAT.

Other Crop models for important traditional food crops like lesser yam or pana (Dioscorea esculenta) and greater yam (Dioscorea alata) are not available in DSSAT therefore were not used in the simulations. Cassava or Lioka B1 (Manihot esculenta) is only consumed as an option and is quite resilient to climatic stress or extreme event while corn or Koni (Zea mays) is not the main staple and is cultivated by few households only, thus the analysis was restricted to Tango Sua.

7.2 Simulation results for taro-Tango Sua (Colocasia esculenta) yield

7.2.1 Potential production and Attainable yields for Ambient (2012)

Table 18 appended below provides a summary of the simulation results for all the four treatments amongst the three sites, East Ghonghau, West Ghonghau and Sa’aiho. The potential yield (50% sunlight) referring to the yield the farmer expects to harvest if factors such as water or nitrogen stress do not affect the taro, is the same (2677 kg/ha) for all three sites. This is because the model does not take into account the limiting impacts of nitrogen and water which are important components for growth and corm production of taro. The 50% sunlight was used because in Bellonese cropping system, taro is cultivated in plots which have trees preserved for shading purpose. This potential yield clearly indicates that if the condition is perfect (where conditions are rarely so), where no other variables affect the growth and development of Tango Sua, the attainable yield would have increased by about 54.2, 79.3 and 263.7% for East Ghonghau, West Ghonghau and Sa’aiho respectively. However, this is not the case in a cropping system where there are no fertilizer or irrigation inputs

133 by farmers and where rainfall and/or heat stress plays a major role in the growth, development and production of crops.

Table 18: Simulation results for Tango Sua yields for ambient and projected changes in temperature, rainfall and carbon dioxide

Field Estimate of attainable Harvest (Dry 1,489.40 weight-kg/ha) by Farmers Site E.Ghonghau W.Ghonghau Sa’aiho Soil type Kenge toaha Kenge ungi Malanga Potential (50% light) 2677 2677 2677 Ambient Attainable ( Nitrogen and (2012) 1736 1493 736 water limiting ) Projection 2030 Attainable 1405 1298 565 under 1142 1074 471 Temperature 2055 Attainable & Rainfall 2090 Attainable 735 728 333 Projection 2030 Attainable 1634 1451 644 under 1745 1514 659 Temperature, 2055 Attainable Rainfall & CO2 2090 Attainable 1852 1561 648

The Bellonese cropping system is rain-fed and there are no additional inputs of technological or chemical parameters to improve yield production. Therefore, the margin between the potential and attainable yield is large and significant. The close analysis of the attainable yields of all three sites reveals that the attainable yield for Tango Sua in Sa’aiho as the lowest of all. Tango Sua only produces a dry yield of about 736 kg/ha in Sa’aiho while for East and West Ghonghau it is 1736kg/ha and 1493kg/ha respectively. The possible reason for this is due to the difference in soil fertility and soil structure in these three sites. To identify the possible reason of soil fertility, the physical and chemical characteristic of these three soils were analysed.

Table 19: Physical and Chemical characteristics of Kenge toaha (KT), Kenge ungi (KU) and Malanga (MA) (Source: Breuning-Madsen et al., 2010) Soils Clay Silt Sand pH Org C Total P Ca Mg K + + + (%) (%) (%) (H2O) (%) N (%) (mg/kg) (cmol (cmol (cmol kg-1) kg-1 kg-1 KT 47 45 8 7.1 4.9 0.37 57.6 31.1 11.1 0.5 KU 40 44 16 6.9 5.8 0.31 78.9 27.4 7.2 0.3 MA 12 21 67 6.9 1.6 0.20 64.1 21.3 4.2 0.2

134

The East and West Ghonghau site have the soil types Kenge toaha (KT) and Kenge ungi (KU) (soft top soil clay loam) which are more fertile and have a higher capacity for soil water compared to Malanga (MA) (Oolitic coarse sand with silty clay) which is located in Sa’aiho as depicted in Table 19 (Breuning-Madsen et al., 2010; Borggaard et al., 2012).Table 18 shows that when compared to MA, KT and KU have higher percentages of essentials elements of Carbon, Nitrogen, Potassium, Calcium and Magnesium, which are essentials for growth, development and production of corms and leaves. This result is in favor with the claims of the Bellonese farmers that KU and KT are the best and fertile soils for taro because of its soil structure being soft and normally produce soft corms and nutritious taro leaves. This is similar to the earlier findings of a study by Breuning-Madsen et al., (2010).

The model simulations further indicate the extent of nitrogen leaching at the three sites which affect the leaf area index (LAI) of Tango Sua (Figure 41). LAI measurement is an important indicator of taro because it shows the area and size of leaves which have important function on the interception of solar radiation and production of biomass (de Oliveira Bernardes et al., 2011). However, the size of the leaf or LAI depends on the availability of nutrients in the soil (Taiz and Zeiger, 2008).Therefore, we look at the nitrogen leaching in the three sites with three different soils and the LAI of Tango Sua. This also indicated the influence of LAI on the level of yield or biomass production (Figure 43). It is apparent from Figure 41 that LAI for Tango Sua at East Ghonghau is the highest on days (137-140) with 0.943 and declines thereafter to 0.637 on maturity day (188). The lowest LAI is found in Sa’aiho with the highest peak of 0.527 on days (148-150) with a decline to (0.381) on maturity day (188). LAI for West Ghonghau is 0.863, and is slightly lower than that of East Ghonghau. The difference in LAI between the three sites as an influence from nitrogen leaching caused by high rainfall is presented in Figure 41. Sa’aiho has the highest leaching that almost reaches 0.8 kg/ha. The nitrogen leaching is observed to correspond with the rainfall trend.

135

Leaf area index vs nitrogen leaching for Tango Sua at W.Ghonghau, E.Ghonghau and Sa'aiho

90 1

80 0.8 70

60 0.6

50 0.4 40

30 0.2 Precipitation (mm/day) (mm/day) Precipitation Leaf area index (LAI) area index (LAI) Leaf 20 0 10

0 -0.2 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 PRECIP mm/d N Leached kg/ha(SoilNi)-West Ghonghau N Leached kg/ha(SoilNi)-Sa'aiho N Leached kg/ha(SoilNi)-East Ghonghau LAI-West Ghonghau LAI-Sa'aiho LAI-East Ghonghau

Figure 41: LAI for Tango Sua and nitrogen leaching at West Ghonghau, East Ghonghau and Sa’aiho

Leaf area index for Tango Sua vs total soil nitrogen in West/East Ghonghau and Sa'aiho

90 1 80 0.9 70 0.8 0.7 60 0.6 50 0.5 40 0.4 30 0.3

Precipitation (mm/d) (mm/d) Precipitation 20 0.2 10 0.1 0 0 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 PRECIP mm/d Tot soil N kg/h-West Ghonghau Tot soil N kg/h-Sa'aiho Tot soil N kg/h-West Ghonghau LAI-West Ghonghau LAI-Sa'aiho LAI-East Ghonghau

Figure 42: Leaf area index for Tango Sua and total soil nitrogen at West Ghonghau, East Ghonghau and Sa’aiho

Even though the three sites received almost the same daily rainfall, leaching for Sa’aiho is found to be higher. This is because Sa’aiho has Malanga soil which has low

136 nitrogen (Table 19) compared to Kenge toaha and Kenge ungi. The fact that Malanga contains oolitic coarse sand (67% sand) makes it highly porous and more vulnerable to leaching during high or prolonged rainfall. In figure 30, the leaching of nitrogen in Sa’aiho has a direct relationship with the increased trend of rainfall. For instance, between days 15-20, 20-30, and 120-130, an increase in rainfall also saw an increase in leaching which affects the LAI of Tango Sua cultivated on Sa’aiho. The result of increased leaching of nitrogen has caused the total nitrogen in the three soils to decline (Figure 42). The trend of LAI in the three sites also reflects the yields of Tango Sua because the higher the LAI, the more interception of solar radiation and high biomass and yield production (Manner and Taylor, 2011) (Figure 43).

Yield vs LAI of Tango Sua at W.Ghonghau, East Ghonghau and Sa'aiho 2000 1 1800 0.8 1600 1400 0.6 1200 1000 0.4 800 0.2 600

400 Indeax (LAI)area Leaf 0 200 Corm yield dry yield dry Corm (kg/ha)weight 0 -0.2 0 50 100 150 200 Days after Planting Tuber kg dm/ha-West Ghonghau Tuber kg dm/ha-Sa'aiho Tuber kg dm/ha-East Ghonghau LAI-West Ghongha LAI-Sa'aiho LAI-East Ghonghau

Figure 43: Leaf area index and yields for Tango Sua at West Ghonghau, East Ghonghau and Sa’aiho

East and West Ghonghau, which are located towards the centre of the island, have fertile soils that are less susceptible to leaching, producing higher LAI and yields compared to Sa’aiho. Overall, it is evident that the model simulations on the attainable yields conform to the farmers’ estimation of yield production per site. More so, the rationale behind the difference between the yield quantity production as a function of soil fertility and structure is similar to the farmers’ perception, past soil study (Breuning-Madsen et al., 2010) and model simulation. However, the model provides a more detail picture of when and to what extent the impact of variables such as rainfall

137 and leaching has on the LAI and yield of Tango Sua.

7.2.2 Ambient (2012) attainable versus projected attainable for temperature, rainfall and carbon dioxide 7.2.2.1 Yield projection under temperature and Rainfall Simulations

The changes in temperature and rainfall used for the three future year’s projection were as follows: 2030: increase of 1oC and 8% in rainfall 2055: increase of 1.8oC and 13% in rainfall 2090: increase of 3.3oC and 21% in rainfall

The rainfall and temperature projection simulation results of Tango Sua for the three sites (Figures 44, 45 and 46) indicate a decline in yield trend compared to the ambient (2012) attainable yield.

Projected yield of Tango Sua vs ambient in West Ghonghau 1800 1600 1400 1200 1000 800 600 400 200 Corm yield dr weight (kg/ha) dr weight yield Corm 0 0 20 40 60 80 100 120 140 160 180 200 Days after planting

Ambient (2012) 2030-Temp & Rain 2055-Temp & Rain 2090-Temp & Rain 2030- CO2 2055-CO2 2090-CO2

Figure 44: Projected and ambient (2012) yields of Tango Sua in West Ghonghau

138

Projected yield of Tango Sua vs ambient in East Ghonghau 2000 1800 1600 1400 1200 1000 800 600 400 200 0 (kg/ha) dry weight yield Corm 0 20 40 60 80 100 120 140 160 180 200 Days after planting

Ambient (2012) 2030-Temp & Rain 2055-Temp & Rain 2090-Temp and Rain 2030-CO2 2055-CO2 2090-CO2

Figure 45: Projected and ambient (2012) yields of Tango Sua in East Ghonghau

800 Projected yield vs ambient for Tango Sua in Sa'aiho 700 600 500 400 300 200 100 (kg/ha) dry weight yield Corm 0 0 20 40 60 80 100 120 140 160 180 200 Days after Planting

Ambient-2012 2030-Temp & Rain 2055-Temp & Rain 2090-Temp & Rain 2030-CO2 2055-CO2 2090-CO2

Figure 46: Projected and ambient (2012) yields of Tango Sua in Sa’aiho

The trend of decline in yield continues with the increasing temperature and rainfall over the three future year’s period (2030, 2055 and 2090). For example, in Figure 44, the yield for Tango Sua cultivated in West Ghonghau is simulated to decline by 13.1%, 28.1% and 51.2% for 2030, 2055 and 2090 respectively. This means that with

139 an increase of 1oC in temperature and 8% in rainfall by 2030, there will be a corresponding reduction of about 13.1% or 195 kg/ha of Tango Sua yield in West Ghonghau. By 2090, an increase of about 3.3oC in temperature and 21% in rainfall will result in a substantial decline of about half (765 kg/ha) of what is currently harvested in 2012 (1493 kg/ha).

In East Ghonghau, the decline in yield is slightly higher than West Ghonghau. Figure 45 indicates that a projected decline of 19.1%, 34.2% and 57.7% is expected by 2030, 2055 and 2090 in comparison to ambient (2012) yields of each site. For Sa’aiho, a projected decline of 23.2%, 36% and 54.8% from the ambient (2012) yield was projected for 2030, 2055 and 2090.

Overall, the results indicate that Tango Sua cultivated on all three sites is projected to have a declining yield ranging between 13.1-23.2% by 2030, 28.1-36% by 2055 and 51.2-57.7% by 2090. However, the most vulnerable site of the three is Sa’aiho (Figure 46) which has the highest percentage of yield reduction due to its limited soil fertility and loosely packed soil structure that is susceptible to leaching as explained in previous section 7.2.1. The decline in taro yield under projected changes in rainfall and temperature for 2030, 2055 and 2090 is also in line with decline in yield of taro varieties such as Tausala-Samoa and Lehua simulated for high volcanic island (Isabel) in Solomon Islands (Quity, 2012).

7.2.2.2 Yield projection under temperature, rainfall and CO2 simulations

The CO2 concentrations used for the three future years were 450ppm for 2030,

550ppm for 2055 and 750ppm for 2090. However, when CO2 is included with the temperature and rainfall projections, the trend of yield production indicate a different path compared to simulations under temperature and rainfall only. The trend in yield is similar for West and East Ghonghau. That is, it drops slightly by about 2.8-5.9% in 2030 and slightly increases by 0.5-1.4% in 2055 and 4.6-6.7% 2090. On the contrary, the yield for Sa’aiho declines by 12.5%, 10.5% and 12% respectively in 2030, 2055 and 2090.

140

For West and East Ghonghau, although the increase in temperature and rainfall should have reduced the yield of Tango Sua, the projected increase in CO2 concentration over the two future year periods (2055 and 2090) caused a slight increase of about 9-21kg/ha and 68-120kg/ha. However, in Sa’aiho a loss in yield of 92kg, 77kg and 118Kg is projected in 2030, 2055 and 2090. Some possible reasons as to why there is a projected slight decrease in Tango Sua yield in 2030 and slight increase in yield for 2055 and 2090 in West and East Ghonghau are as follows. The o CO2 concentration used for 2030 is 450 ppm under an increase of 1 C in temperature and 18% increase in rainfall. This concentration of 450 ppm might not be sufficient to offset the negative impacts of the temperature increase of about 1oC on the increase in growth and in production of corms.

However, the CO2 concentration of 550ppm and 750ppm despite an increase of 1.8 oC and 3.3oC indicated beneficial carbon dioxide fertilization for Tango Sua with a production ranging from 21-118kg/ha. Increase in CO2 has benefits to crops known as carbon dioxide fertilization or carbon dioxide fertilization effect (CFE) by enhancing photosysnthesis and reducing water loss per unit of leaf area (Lobell and Field, 2008) . That is, carbon dioxide fertlization increases growth and boosts productivity in crops. Taro is a C3 plant which means it has the capacity to efficiently use elevated CO2 levels to maximise its growth and production.

Though there are limited literature on the responses of taro (Colocasia esculenta) on elevated CO2 levels which we may substantiate our results, there are studies that do support that C3 plants experience increased yields under increased CO2 levels. For example, at 550ppm CO2 concentrations, there was an increase of about 10-20% for C3 crops (e.g. wheat and rice) under unstressed conditions (Ainsworth et al., 2004; Gifford, 2004; Long et al., 2004). Using other crop modelling system such as AFRC-Wheat, APSIM, CERES, CROPGRO, CropSyst, LINTULC and SIRIUS.

Tubiello et al., (2007) state that plant growth and yield response to 550ppm CO2 indicate increased yield of about 5-20%. However, the impacts of elevated CO2 on photosynthesis and yield may be negated by the combined impacts of water and nutrient stress. This is evident in the results among the three sites whereby there was an increase of yields by 4.6-6.7% for 2090 in West and East Ghonghau which have

141 more fertile soils (Kenge ungi and toaha) while a decline by 12% in Sa’aiho which have less fertile soil (Malanga). In terms of the impact of water stress, given that the simulations use a projection of rainfall increase by 8%, 13% and 21% for 2030, 2055 and 2090; the impact of water stress is not significant among the three sites.

Though the results of the study indicate an increase of about 0.5-1.4% and 4.6-6.7% under 550ppm and 750ppm in comparison to the ambient (2012) yields in West and East Ghonghau, the simulations does not take into account other parameters that could affect taro growth. Various studies mention that a weakness of yield projections such as the one used by this study is the fact that it does not consider other parameters that may affect the growth and yield of crops like pest, disease, weeds, or soil fertility (Tubiello and Ewert, 2002; Karnosky, 2003; Gifford, 2004; Ainsworth and Long, 2005; Ziska and George, 2004).

Further results (Table 18) indicated that both soil type and fertility play an important role in the future yields of taro under different climate scenarios. For instance, West and East Ghonghau sites which have more fertile soils (KT and KU) tend to respond positively to CFE at 550ppm and 750ppm of CO2. Tango Sua cultivated on Sa’aiho that has less fertile soil (MA) responds negatively to all three levels of CO2 concentration. Farmers’ interviews revealed that KU and KT soils are more fertile and suitable for almost all crops however KT is more preferred for taro cultivation. On the other hand, they claim that Malanga soil is less fertile, and commented that it is only used for crops like sweet potatoes, corn and water melons. This is a serious problem for farmers who have only Malanga soil or limited plots on KU or KT soils for cultivating taro.

7.3 Simulation results for cassava –Lioka B1 (Manihot esculenta) yield 7.3.1 Ambient (2012) potential production versus ambient (2012) attainable yield for cassava

The potential yield (dry weight) simulated by the model is 6617kg/ha for Lioka B1 grown in all three sites (Table 20). The potential yield is the same for all three sites because;

142

 the variety of cassava-Lioka B1 is the same for all three sites,  the weather (rainfall, temperature and solar radiation) is the same for all sites and,  the water and nitrogen were turned off, thus the soil characteristic of each site is not being considered in this potential production simulation.

Table 20: Simulation results for Lioka-B1 yields (dry weight) for ambient (2012) and projected changes in temperature, rainfall, carbon dioxide

Field Estimate of attainable Harvest (Dry 2802.47 weight) Site E.Ghonghau W.Ghonghau Sa’aiho Soil type Kenge toaha Kenge ungi Malanga Potential (100% light) 6617 6617 6617 Ambient Attainable ( Nitrogen and 3494 2805 2577 (2012) water limiting ) Projection 2030 Attainable 3599 2988 2720 under 3565 3019 2815 Temperature 2055 Attainable & Rainfall 2090 Attainable 3442 2905 2765 Projection 2030 Attainable 3761 3042 2730 under 4046 3260 2872 Temperature, 2055 Attainable Rainfall & CO2 2090 Attainable 4337 3324 2979

However, under attainable simulation, differences are observed in the yield of Lioka B1 in all three sites. This is because the limiting factors of nitrogen and water are turned on meaning that the nitrogen and water content of the three different soils were taken into account by the model simulation (Figure 47). Graph results (Figure 47) indicate that there is a slight difference in attainable yields of Lioka B1 on the three sites as compared to the potential yield. Lioka B1 performed better in East Ghonghau with a yield of 3494kg/ha however declines slightly by about 19.7% in West Ghonghau and 26.2% in Sa’aiho.

143

7000 Potential vs attainable yields for Lioka B1 6000 5000 4000 3000 2000 1000 Tuber dry weight (kg/ha) 0 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 Days after Planting

Potential Harvest wt-East Ghonghau Harvest wt-West Ghonghau Harvest wt-Sa'aiho

Figure 47: Potential and attainable yields of Lioka B1 in West/East Ghonghau and Sa’aiho

70 LAI of Lioka B1 vs nitrogen leaching and precipitation 3

60 2.5

50 2

40 1.5

30 1 LAI (m2/m2) (m2/m2) LAI

Precipitation (mm/d) (mm/d) Precipitation 20 0.5

10 0

0 -0.5 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150

PRECIP mm/d (Weather) Run 1 N Leached kg/ha(SoilNi) Run 1 N Leached kg/ha(SoilNi) Run 2 N Leached kg/ha(SoilNi) Run 3 LAI (PlantGro) Run 1 LAI (PlantGro) Run 2 LAI (PlantGro) Run 3

Figure 48: LAI of Lioka B1 and nitrogen leaching in West/East Ghonghau and Sa’aiho

The possible reasons for the decline in West Ghonghau and Sa’aiho are soil type, structure and fertility. The site in Sa’aiho has Malanga soil (oolitic coarse sand) which is susceptible to leaching and is less fertile compared to Kenge toaha soil found in East Ghonghau site. Nitrogen leaching in the three sites with Sa’aiho having the highest leaching levels that peak with the high rainfall on days 43, 46, 52, 67 and 129 (Figure 48). This depicts an observed trend of significant leaching of Nitrogen in Sa’iaho of about 28.3kg/ha compared to only 9.8kg/ha and 7.4kg/ha in East and West

144

Ghonghau could be why both the LAI and yield of Lioka B1 is smaller in Sa’aiho. For example, with the increased trend of both rainfall (53.8mm) and nitrogen leaching (25.2kg/ha) on day 46, LAI of Lioka B1 in Sa’aiho drops from its maximum of 1.9 on day 45 and then gradually falls till the end of maturity day (155). While, LAI of both West and East Ghonghau increases to their maximum of 2.1 and 2.6 on day 93 before gradually declining.

A declining trend of yields observed here is directly proportional to the decline in soil fertility and structure. Malanga soil in Sa’aiho is found to be more vulnerable to the impacts of weather parameter like rainfall.

7.3.2 Ambient (2012) attainable versus projected attainable yields for temperature, rainfall and carbon dioxide

The changes in temperature, rainfall and CO2 concentrations used for the three future year’s projection were as follows: o 2030: increase of 1 C and 8% in rainfall with 450ppm CO2 o 2055: increase of 1.8 C and 13% in rainfall 550ppm CO2 o 2090: increase of 3.3 C and 21% in rainfall 750ppm CO2

In Figures 49, 50 and 51 the projected yields of Lioka B1 in the three sites are plotted against the ambient yield of 2012. The graphs indicate a common trend of increased Lioka B1 yields under both: (1) temperature and rainfall projections only and (2) temperature, rainfall and CO2 for 2030, 2055 and 2090. However, the magnitude of increase in yield is different for all three sites. In East Ghonghau (Figure 49), under temperature and rainfall projections, the highest yield peaks at 2030 at an increase of about 3% from ambient yields in 2012. While for West Ghonghau and Sa’aiho the yield peaks in 2055 at an increase of about 7.6 and 9.2% respectively. The results indicate that Lioka B1 cultivated in all three sites perform better under increased temperature and rainfall. However, West Ghonghau and Sa’aiho show their maximum yields at an increase of 1.8 oC in temperature and 13% in rainfall by 2050. On the other hand, Lioka B1 in East Ghonghau performs at its maximum at an increase of 1 oC and 8% in temperature and rainfall by 2030.

145

5000 Ambient and projected yields of Lioka B1 in East Ghonghau 4500 Ambient (2012) 4000 Temp & Rain -2030 3500 Temp & Rain-2055 3000 2500 Temp & Rain-2090 2000 CO2-2030 1500 CO2-2055 1000 Yield dry Yield dry weight (kg/ha) 500 CO2-2090 0 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 Days after planting

Figure 49: Ambient (2012) and projected attainable yields for Lioka B1 in East Ghonghau

3500 Ambient and projected yields of Lioka B1 in West Ghonghau Ambient (2012) 3000 Temp & Rain -2030 2500 Temp & Rain-2055 2000 Temp & Rain-2090 1500 CO2-2030 1000 CO2-2055

500 CO2-2090 Yield dry Yield dry weight (kg/ha) 0 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170

Days after planting

Figure 50: Ambient (2012) and projected attainable yields for Lioka B1 in West Ghonghau

3500 Ambient and Projected yields of Lioka B1 in Sa'aiho

3000 Ambient (2012) 2500 Temp & Rain-2030

2000 Temp & Rain-2055 Temp & Rain 2090 1500 CO2-2030 1000 CO2-2055 Yield dry Yield dry weight (kg/ha) 500 CO2-2090

0 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 Days after planting

Figure 51: Ambient (2012) and projected attainable yields for Lioka B1 in Sa’aiho 146

In terms of simulating the projected CO2 concentration with the temperature and rainfall, all three sites display similar trends of yield increase in Lioka B1. Their yields increase with increase in CO2 concentration. All of the sites have their highest peak in yield of Lioka B1 by 2090 under 750ppm concentration of CO2, and an increase of 3.3 oC and 21% in temperature and rainfall. However, the highest yield gainer site in comparison to the ambient (2012) yield is East Ghonghau with an increase of about 24.1% while Sa’aiho has the lowest increase of about 15.6%. From these results, it is projected that Lioka B1 will produce high yields compared to what is harvested currently (2012) even if the projected changes in temperature, rainfall and CO2 occur by the end of this century.

As previously mentioned in the taro (Tango Sua) results, a possible reason for this increase in yield is the C3 photosynthetic pathway. Cassava is also a C3 plant therefore its C3 photosynthetic pathway tends to maximize the elevated CO2 concentration for both its growth and tuber production. However, in terms of CO2 projection, the results show that Lioka B1 performs better in East and West Ghonghau than in Sa’aiho. For example, by 2090, Lioka B1 in East Ghonghau will produce around 1358kg/ha yield (dry weight) more than what is projected for Sa’aiho.

This indicated that though CO2 fertilization plays an important role in yield production, nutrients in terms of soil fertility also plays an important role in a changing pattern of weather or climate in the future. Sa’aiho has the least fertile soil (Malanga) with lower nutrients among the three sites and as a result it has lower increase in comparison to ambient (2012) yield of 15.6% compared to 18.5% and 24.1% of West and East Ghonghau. In terms of the impact of water stress, given that the simulations use a projection of rainfall increase by 8%, 13% and 21% for 2030, 2055 and 2090; the impact of water stress is not significant among the three sites.

Overall, the results indicate a trend of increasing yield of Lioka B1 in all three sites o with high temperature (3 C), rainfall (21%) and 750ppm CO2.This is similar to the findings of Witter (1995) which highlighted that cassava plant increase by about

150% of dry matter with high temperature and high CO2 at 750ppm.

147

7.4 Simulation results for corn (Koni)-Zea mays yield

7.4.1 Ambient (2012) potential versus ambient (2012) attainable yield for Koni

Table 21 below outlines the summary results for Koni simulation. Under potential simulation, the potential yield for corn is 304 kg/ha for all three sites (Table 21). However when water and nitrogen stress were simulated, the attainable yield was 288kg/ha for East and West Ghonghau and remains at 304kg/ha for Sa’aiho. The reason for the decline in both East and West Ghonghau is because of excess water stress. Figure 52 shows that the two sites experience higher excessive water stress ranging from 0.73-0.76 associated with higher rainfall (118mm-224mm) on days 37 and 52.

Table 21: Simulation results for Koni yields (dry weight) for ambient (2012) and projected changes in temperature, rainfall, carbon dioxide

Field Estimate of attainable Harvest (Dry 280 weight) Site E.Ghonghau W.Ghonghau Sa’aiho Soil type Kenge toaha Kenge ungi Malanga Potential (100% light) 304 304 304 Ambient Attainable ( Nitrogen and 288 288 304 water limiting ) Projection 2030 Attainable 288 288 288 under 280 280 280 Temperature 2055 Attainable & Rainfall 2090 Attainable 264 264 264 Projection 2030 Attainable 288 288 288 under 280 280 280 Temperature, 2055 Attainable Rainfall & CO2 2090 Attainable 264 264 264

Note that the minimum stress value is 0 and maximum is 1. Therefore, the fact that East and West Ghonghau have stress values of around 0.7 significantly indicates that there is excess water in the soils which may affect nutrient uptake. The total water in soils of East and West Ghonghau as depicted in Figure 42 shows they have about 343-396mm of water which is 35% more than that in Sa’aiho. The soil type (KT/KU) of East and West Ghonghau having higher percentage of clay (40-47%) compared to Malanga soil (12% clay) in Sa’aiho influences the excess water stress because of its

148 high water retention capacity.

Excess water stress and LAI of Koni 300 0.8

0.7 250 0.6 200 0.5

150 0.4

0.3 100

Precipitation (mm/d) (mm/d) Precipitation 0.2 50

0.1 Max=1) stress Water (Min=0,

0 0 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 Days after planting

PRECIP mm/d LAI-East Ghonghau Excess H20 stress-East Ghonghau LAI-West Ghonghau Excess H20 stress-West Ghonghau LAI-Sa'aiho

Figure 52: Excess water stress and LAI of Koni in West/East Ghonghau and Sa’aiho

More so, the excess amount of water in soils of East and West Ghonghau when compared to Sa’aiho negatively affects the nitrogen uptake (Figure 53) resulting in the lower yield and LAI (Figure 52). The biomass produced by Koni in West and East Ghonghau only accounts for 970-984kg/ha which is about 20% less than biomass production in Sa’aiho (1166kg/ha). The total Nitrogen uptake during the growing season is about 15.5kg/ha for both sites compared to 18.5kg/ha for Sa’aiho. This is about 20% more Nitrogen uptake by Koni in Sa’aiho which directly correlates to 20% production increase in biomass. This means that Koni cultivated on oolitic coarse sandy soil (Malanga) is more resilient to extreme rainfall compared to clay loam soil (Kenge toaha/ungi) on Bellona Island. This is because of the low water retention capacity of Malanga and its excessive drainage. However, it must be also noted that low water retention of Malanga may also increased the nitrogen leaching which may in turn also reduce yield.

149

450 Total soil water and nitrogen uptake by Koni 20 400 18 350 16 14 300 12 250 10 200 8 150 6 Nitrogen uptake (kg/ha) (kg/ha) uptakeNitrogen 100 Total soil water (mm) (mm) water soil Total 4 50 2 0 0 0 10 20 30 40 50 60 70 80 90 100 Days after Planting PRECIP mm/d Total soil water-East Ghonghau Total soil water-West Ghonghau Total soil water-Sa'aiho N uptake kg/ha-East Ghonghau N uptake kg/ha-West Ghonghau N uptake kg/ha-Sa'aiho

Figure 53: Total soil water and nitrogen uptake in East/West Ghonghau and Sa’aiho

7.4.2 Ambient (2012) attainable versus projected attainable yields for Koni under temperature, rainfall and carbon dioxide

The changes in temperature, rainfall and CO2 concentrations used for the three future year’s projection were as follows: o 2030: increase of 1 C and 8% in rainfall with 450ppm CO2 o 2055: increase of 1.8 C and 13% in rainfall 550ppm CO2 o 2090: increase of 3.3 C and 21% in rainfall 750ppm CO2

Unlike taro-Tango Sua and cassava- Lioka B1, the results of corn (Koni) indicate a different trend in its yield for future climate scenarios. Under projected changes in temperature and rainfall, the yield of Koni remains the same even for 1oC increase in 2030 however declines gradually by 2.8% in 2055 and 8.3% by 2090 in East and West Ghonghau sites in comparison to the ambient (2012) yields. More so, in Sa’aiho Koni yields declines by 7.9% and 13.2% by 2030 and 2090 respectively. This implies that the corn yield in Bellona decreases with an increase in temperature of beyond 1.8oC to 3.3oC.

150

A recent study by Tao and Zang (2011) on both irrigated and rain fed maize in China show a similar result where under increasing temperature, the yields of maize decreased by 1% to about 33.7% in different sites in China at 1oC, 2oC and 3oC.

Under CO2 considerations, they found that that decrease values ranged from 0.7% to

-25.9%. With CO2 projections, our results indicate that yields of Koni remained the same as under the temperature and rainfall projections. From this result, it is observed that Koni does not respond positively to elevated levels of CO2 like Tango Sua and Lioka B1. This could be the fact that corn is a C4 plant which has a mechanism located in its leaves that even at the ambient CO2 concentration, the mechanism induces CO2 saturation of photosynthesis (Allen and Prasad, 2004).

Therefore, in terms of corn or other C4 plants, RuBisCO-Ribulose-1,5-bisphosphate carboxylase-oxygenase is restricted to bundle sheath cells which have CO2 concentration of about 3-6 times the atmospheric concentration of CO2 (von Caemmerer and Furbank, 2003). With this high concentration, even with future elevated CO2 levels of 450ppm, 550ppm and 750ppm, there will be no uptake of it because the higher concentration will saturate RuBisCO (Tao and Zhang, 2011). The result (Table 21) clearly indicated that Koni respond only to temperature change or is sensitive to temperature and not on elevated CO2. This is why yields of Koni in all three sites declined when temperature is increased by 1, 1.8 and 3oC; however remain constant when elevated CO2 concentrations of 450, 550 and 550ppm were simulated.

Various authors argue that at 550ppm, C4 crops will experience yield increase of only between 0-10% unlike C3 crops which find future elevated CO2 levels beneficial (Easterling et al., 2007; Gifford, 2004; Ziska, 2003; Ainsworth et al., 2004; Long et al., 2004). However, although lacking in photosynthesis advantages, the beneficial effect of elevated CO2 levels on corn is on water use efficiency (Witter, 1995). Overall, compared to C3 crops, future increased levels of CO2 will see less yields in C4 crops like corn.

7.5 Overview of the projected yield in Tango Sua, Lioka B1 and Koni

Overall, Figures 54, 55and 56 summarize the trend of projected yield of all three crops for 2030, 2055 and 2090 under projected increase in temperature, rainfall and 151

CO2 in relation to the ambient year of 2012. Tango Sua yields will slightly decline by 2.8% and 5.9% by 2030, and increase by 1.4% and 0.5% by 2055 and 4.6% and 6.7% by 2090 for West and East Ghonghau. On the contrast, in Sa’aiho, Tango Sua yield will decline by 12.5% in 2030, 10.5% in 2050 and 12% by 2090.

Meanwhile, Lioka B1 yields will increase with increase in temperature, rainfall and

CO2 on all three sites. For West and East Ghonghau Lioka B1 will increase by 8.4% and 7.6% in 2030, 16.2% and 15.8% in 2055 and 18.5% and 24.1% by 2090. In Sa’aiho, the yields would decline by 5.9%, 11.4% and 15.6% by 2030, 2055 and 2090, respectively.

On the contrast, there will be a decline in Koni yields in all three sites. Both West and East Ghonghau have been projected to have similar decline in yield of about 2.8% and 8.3% by 2055 and 2090. The model does not simulate any changes in yield by 2030 in both these sites. In Sa’aiho, there is a projected decline of 5.3%, 7.9% and 13.2% by 2030, 2055 and 2090.

In general, these results signify that taro (Tango Sua) can still be cultivated on Bellona by 2090 however; it may only be restricted to two soil types (Kenge ungi and Kenge toaha) which are more fertile and less susceptible to leaching and nutrient loss. Nevertheless, it should be noted that our projection is based on attainable yield which does not consider the impacts of pest or disease which are indirect impacts of climate change and extreme events. Cassava (Lioka B1) is projected to be the most resilient of the three crops and may yield higher root production with the projected change in temperature, rainfall and CO2. Cassava is more resilient to pest attack, drought and may be cultivated on almost any type of soil even in areas where irrigation is not possible like Bellona and where long dry season is common (Witter, 1995). Corn (Koni) is the most vulnerable of the three crops and may not be performing well with the future climate and CO2 scenarios. This is because with its C4 photosynthetic pathway, the effect of elevated CO2 concentrations will be on water use efficiency and not on increase photosynthesis like C3 crops (Witter, 1995).

152

2000 Projected trend of Tango Sua yield in Bellona 1800 1600 1400 1200 1000 800 600 400

Yield dry Yield dry weight (kg/ha) 200 0 2012 2030 2055 2090 Year East Ghonghau West Ghonghau Sa'aiho Linear (East Ghonghau) Linear (West Ghonghau) Linear (Sa'aiho)

Figure 54: Projected trend of taro (Tango Sua) yield in three wards in Bellona

5000 Projected trend of Lioka B1 yield in Bellona 4000

3000

2000

1000 Yield dry Yield dry weight (kg/ha) 0 2012 2030 2055 2090 Year East Ghonghau West Ghongau Sa'aiho Linear (East Ghonghau) Linear (West Ghongau) Linear (Sa'aiho) Figure 55: Projected trend of cassava (Lioka B1) yield in three wards in Bellona

Projected trend of Koni yield in Bellona

310 300 290 280 270 260 250

Yield dry Yield dry weight (kg/ha) 240 2012 2030 2055 2090 Year

East Ghonghau West Ghongau Sa'aiho Linear (East Ghonghau) Linear (West Ghongau) Linear (Sa'aiho)

Figure 56: Projected trend of cassava (Lioka B1) yield in three wards in Bellona 153

7.6 Simulation of yields of taro (Tango Sua), cassava (Lioka B1) and corn (Koni) during El Niño years

Results from the simulation of past 11 years of El Niño are presented in Figures 57-59 for Tango Sua, Lioka B1 and Koni on all three sites. The model was run with same input values of three crops and soil data for the three sites but with different daily rainfall, temperature for the El Niño years ranging from 1982-2009. The 11 years are 1982, 1987, 1991,1992,1993,1994, 1997, 2002, 2004, 2006 and 2009. The simulation was conducted for the yields of the three crops in three different sites.

For Tango Sua, there is a linear trend of decline in the yield as the simulation moves from 1982 to 2009 (Figure 46). However, the linear trend of decline in yield is not significant because as indicated by the low R2 values for all three sites which is around 0.1. This is because of the variation in yield of taro in relation to the strength of the El Niño at the particular years. For example all the three sites experience a drop in yield in 1982, 1991, 1994, 1997 and 2009 as the influence of moderate to strong El Niño years which causes water stress and in turn affect the yields of Tango Sua. However, for the three sites, Sa’aiho is the most vulnerable which indicated by a higher loss of yield in the most severe El Niño.

2500 Trend of Tango Sua yields in ENSO years

2000

y = -10.919x + 23454 1500 R² = 0.1139

y = -13.368x + 28109 1000 R² = 0.1145

Yield dry Yield dry weight (kg/ha) 500 y = -14.027x + 28578 R² = 0.1042 0 1980 1985 1990 1995 2000 2005 2010 2015

West Ghonghau Sa'aiho East Ghonghau Linear (West Ghonghau) Linear (Sa'aiho) Linear (East Ghonghau)

Figure 57: Simulated yields of Tango Sua for the past 11 El Niño years Key: El Niño Strong… Moderate… 154

For instance, during the 1997 El Niño which is one of the most severe El Niño affecting the Solomon Islands, Tango Sua in Sa’aiho lost about 62% in yield compared to 51% in West Ghonghau and 26% in East Ghonghau in comparison to the Ambient yield (2012). This is mainly due to the limited water holding capacity of Malanga soil and its low fertility which caused severe water stress during drought in El Niño years.

For Lioka B1, there is a positive linear trend indicating an increase in yield as simulation progress from 1982 to 2009 (Figure 58). This might be because cassava (Lioka B1) is more tolerant to drought compared to taro (Tango Sua). However, during the El Niño in 1992, there was a significant yield decline in West and East Ghonghau in comparison to Sa’aiho. There was a significant decline of about 73.6% and 78.5% in yield in both West and East Ghonghau compared to about 24% decline in Sa’aiho from the ambient (2012) yields. The reason for this significant decline in these two sites is mainly due to their soil structure and texture (Kenge ungi and toaha) being more clay. This means that during the 1993 drought, dryness of clay particles make it more compact thus causing extractable water in East and West Ghonghau to be lower than that of Sa’aiho (Figure 59).

4500 Trend of Lioka B1 in El Nino years y = 51.962x - 101030 4000 R² = 0.1475 3500

3000 y = 38.036x - 73707 R² = 0.1548 2500

y = 22.527x - 42844 2000 R² = 0.3354 1500 1000 Yield dry dry Yield weight (kg/ha) 500 0 1980 1985 1990 1995 2000 2005 2010 2015

East Ghonghau West Ghonghau Sa'aiho

Linear (East Ghonghau) Linear (West Ghonghau) Linear (Sa'aiho)

Figure 58: Simulated yields of Lioka B1 for the past 11 El Niño years

155

250 1992 El Nino vs extractable water, cumulative runoff and LAI of Lioka B1 2.5

200 2

150 1.5

100 1 LAI (mm/mm)LAI Precipitation(mm/d) Precipitation(mm/d) 50 0.5

0 0 0 20 40 60 80 100 120 140 160 180 Days after planting Extr water mm-West Ghonghau Extr water mm-Sa'aiho Extr water mm-East Ghonghau N uptake kg/ha-West Ghonghau N uptake kg/ha-Sa'aiho N uptake kg/ha-East Ghonghau Cum Runoff mm-West Ghonghau Cum Runoff mm-Sa'aiho Cum Runoff mm-East Ghonghau PRECIP mm/d x 2 LAI-West Ghonghau LAI-Sa'aiho LAI-East Ghonghau

Figure 59: Simulated yields of Lioka B1 for the past 11 El Niño years

More so, during periods of rain during the growth season as on days 49, 119, 152 since the soil surface of Kenge ungi and toaha are already dry and compact, cumulative surface runoff is higher compared to Malanga soil (oolitic coarse sand) in Sa’aiho (Figure 59). This increased water run offs on East and West Ghonghau mean limited water infiltrating into the soil which Lioka B1could have use for growth. The water stress in these two sites influences the limited uptake of nitrogen as depicted in Figure 59 which in turn results in low LAIs and crop yield during the 1992 drought. While for Lioka B1 in Sa’aiho, less cumulative runoff and more extractable water mean higher nitrogen uptake, LAI and yield during the 1992 drought. Thus, even though Tango Sua might be vulnerable during drought in Sa’aiho (Malanga soil), Lioka B1 performed better in Sa’aiho during drought event despite having limited soil fertility.

For Koni, there is a positive trend of yield among the ENSO years. However, this trend is not significant because R2 ranged from 0.15-0.2. However, the yield and germination of Koni is affected by moderate -strong El Niño in 1987, 1992 and 1997 (Figure 60). Similarly to Lioka B1, Koni performed much better in Sa’aiho (Malanga soil) compared to East and West Ghonghau (Kenge toaha/ungi soils). For instance in

156 the 1992, Koni yield in West Ghonghau was about 14kg/ha dry weight of grain, however, in 1997 severe drought, there was no germination of Koni which resulted in crop failure in both in East and West Ghonghau.

350 Trend of Koni in El Nino years y = 2.3429x - 4382.9 R² = 0.2115 300

250 y = 5.3694x - 10500 R² = 0.1551 200 y = 4.4297x - 8601.9 R² = 0.1562 150

100

50 Yield dry Yield dry grain weight (kg/ha) 0 1980 1985 1990 1995 2000 2005 2010 2015

West Ghonghau Sa'aiho East Ghonghau Linear (West Ghonghau) Linear (Sa'aiho) Linear (East Ghonghau)

Figure 60: Simulated yields of Koni for the past 11 El Niño years Key: X ….Crop failure because of lack of germination within 15 days of sowing

25 1.2 Water stress and Nitrogen uptake in 1992 Drought for Koni in three sites 1 20

0.8 15 0.6 10 0.4

5 Nitrogen uptake (kg/ha) uptake Nitrogen 0.2 Water stress (Min=0, Max=1) stress (Min=0, Water 0 0 0 10 20 30 40 50 60 70 80 90 100 N uptake kg/ha-West Ghonghau N uptake kg/ha-Sa'aiho N uptake kg/ha-East Ghonghau H20 stress-West Ghonghau H20 stress-Sa'aiho H20 stress East Ghonghau

Figure 61: Simulated nitrogen uptake and water stress during the 1992 El Niño

The significant loss in yield of Koni in 1992 drought in West Ghonghau by about 95% as compared to ambient yield (2012) is due to high water stress experienced in West Ghonghau. Figure 61, indicates that of the three sites, West Ghonghau and East Ghonghau experience water stress at about 10 days after planting, which is critical for Koni growth. However, the stress continues throughout the growing season for

157

West Ghonghau compared to the other two sites. In East Ghonghau, water stress declined after day 40 while Sa’aiho has no water stress.

Our result also indicates that with the water stress, Nitrogen uptake is also affected. For example in Figure 61, West Ghonghau having water stress all throughout the growing season of Koni also has the lowest Nitrogen uptake ranging from 0.02-0.26 kg/ha. Studies confirm that drought or water stress in soil causes a lower availability of nutrient at the surface of roots because there is a decline in mineralization of bound nutrients and transport of nutrients by diffusion or mass flow in soil (Buljovcic and Engels, 2001; Walworth, 1992; Seiffert et al., 1995). Buljovcic and Engels (2001) mention that there is significant decline of Nitrogen uptake in corn with a decrease by about 5% of soil water content for 3 days. In the present study given that water stress is significant throughout the growing Koni season in West Ghonghau, almost for 90days, the diminished Nitrogen uptake (in turn) causes the low yield of 14kg/ha simulated in 1992 El Niño. On contrary, Sa’aiho having no water stress has the highest yield of about 304kg/ha with a maximum Nitrogen uptake of 19.9kg/ha.

However, the severe 1997 drought as simulated by the model indicated crop failure in both West and East Ghonghau as a result of lack of germination within 15 days of sowing. This lack of germination during this 1997 drought is due severe water stress. Water stress has been found to reduce the percentage and rate of germination and growth of seedlings (Khayatnezhad et al., 2010; Achakzai, 2009). This is because, seeds need water which is essential to initiate germination and lack of water or rainfall during drought or stress condition may delay the completion of germination thus resulted in crop failure (Tesche, 1975; Achakzai, 2009).

7.7 Simulating El Niño (El Niño 1997 drought) versus Bellonese perspective on impacts on taro-Tango Sua

According to the model simulations (Table 22-24) below, there are seven main development or growth stages of taro, namely, root formation, emergence of first leaf, establishment, rapid vegetation growth, end corm, maturity and harvest. However, from the Bellonese perspective, they consider only three main stages (establishment, vegetative and maturity) of taro development which are similar to the unpublished

158 findings of Quity (2012) on Isabel Island. These stages of growth and development as perceived by farmers are observed by the physical form of the taro (Tango Sua) growth. Such development stages are clearly illustrated by Singh et al., (1992) in Figure 62.

Figure 62: Three main taro development stages (Singh et al., 1992)

The results of the simulation (Table 22-24) below indicated that for all three sites, under ambient and drought, the root formation commences 1 day after planting while emergence of 1st leaf occurs 2 days after planting. The establishment, rapid vegetation and end of corm occurs 69, 129 and 182 days after planting under ambient condition. In drought conditions, the establishment and rapid vegetation stage are the same as ambient however the end of corm takes additional 2 days making it 184 days for all three sites. Maturity and harvesting is simulated to be on day188 after planting for all three sites under ambient however differs under drought conditions. For instance, in both East and West Ghonghau (Tables 23 and 24) Tango Sua reached maturity earlier on 187 days after planting while in Sa’aiho it took 189 days.

According to interviews with farmers, the emergence of first leaves normally occurs about 3-5 days while the whole establishment stage takes about 1-1.5 months after planting. It takes 2-3 months for the vegetative stage when the taro will have massive physical growth and producing numerous leaves. The final stage of maturity occurs 5-6 months after planting depicted by shrinking stem and only few leaves left on the crop. There are almost similar parameters of taro (Tango Sua) development between 159

Bellonese perspectives and model simulations, however, the model provides us with more detailed information as well as other factors that may influence the growth and development of taro such as Leaf area index (LAI), water and nitrogen stress.

Table 22: Growth stages and Stress of Tango Sua under ambient and 1997 drought in West Ghonghau Growth stage Crop age LAI *Water stress *Nitrogen stress Ambient 1997 Ambient 1997 Ambient 1997 Ambient 1997 Planting 0 0 0 0 0 0 0 0 Root form 1 1 0 0 0 0 0 0 1st leaf 2 2 0 0 0 0 0 0 Establishment 69 69 0.32 0.42 0 0.06 0.03 0.07 Rapid 129 129 0.83 0.23 0 0.46 0.19 0.65 vegetation End of corm 182 184 0.61 0 0 0.37 0 0.58 Maturity 188 187 0.58 0 0 0 0 0.33 Harvest 188 187 0.58 0 0 0 0 0 *NB: Stress 0= Min, 1=Max

Table 23: Growth stages and Stress of Tango Sua under ambient and 1997 drought in East Ghonghau

Growth stage Crop age LAI *Water stress *Nitrogen stress Ambient 1997 Ambient 1997 Ambient 1997 Ambient 1997 Planting 0 0 0 0 0 0 0 0 Root form 1 1 0 0 0 0 0 0 1st leaf 2 2 0 0 0 0 0 0 Establishment 69 69 0.34 0.46 0 0.06 0 0 Rapid 129 129 0.92 0.23 0 0.55 0.15 0.54 vegetation End of corm 182 184 0.67 0 0 0.36 0 0.49 Maturity 188 187 0.64 0 0 0 0 0.33 Harvest 188 187 0.64 0 0 0 0 0

160

Table 24: Growth stages and Stress of Tango Sua under ambient and 1997 drought in Sa’aiho Growth stage Crop age LAI *Water stress *Nitrogen stress Ambient 1997 Ambient 1997 Ambient 1997 Ambient 1997 Planting 0 0 0 0 0 0 0 Root form 1 1 0 0 0 0 0 1st leaf 2 2 0 0 0 0 0 Establishment 69 69 0.17 0.13 0 0.03 0.29 0.45 Rapid 129 129 0.47 0.18 0 0.13 0.26 0.72 vegetation End of corm 182 184 0.40 0.03 0 0.26 0.05 0.36 Maturity 188 189 0.38 0.02 0 0 0 0 Harvest 188 189 0.38 0.02 0 0 0 0

In terms of the impact of drought on Tango Sua, we compare the results of the simulation for the drought year 1997 which was one of the worst in Bellona to ambient (2012) in conjunction with the Bellonese perspective. The simulated results represented in Figures 63-65 show that during the drought event in 1997, Tango Sua only yielded about 441kg/ha, 477kg/ha and 184 kg/ha in West/East Ghonghau and Sa’aiho respectively. While in 2012 it has about 1493kg/ha, 1734kg/ha and 726 kg/ha in those three sites. This signifies a substantial decline in yield of about 70.5%, 72.5% and 74.7% respectively for the three sites during the 1997 drought event. Though the farmers did not provide a numerical figure on the yield of Tango Sua they harvested during the 1997 drought, they substantiated that they experienced a serious decline in the production of taro. The difficulty in obtaining a numerical figure on taro yield for Bellona for 1997 is that their harvest is not measured and they conduct their harvest not in a onetime situation but continuously. Thus, there is no written record of taro yield.

However, it is evident that taro production is negatively affected by drought and the use of crop modelling further provides us with the details that the farmers fail to provide. This indicated the significance of integrating the local perceptions of the Bellonese farmers and crop modelling for better understanding on what parameters actually affect their crop production and be better equipped to find ways to manage or adapt to such impacts. A typical example is that in Figure 63 where in West Ghonghau, the model showed that in 1997, rainfall was severely reduced on days 50 to 155 after planting causing water stress which is depicted in Table 22 during the 161 establishment (0.06), rapid vegetation growth (0.46), and end of corm stages (0.37). These stages are crucial for both the growth and production of taro corms. However due to the drought event of about 3.5 months, lower physical growth and corm production was observed in Tango Sua. Water stress caused by drought reduces the yield because it impairs the growth of organs and their final size (Blum, 1996). However, looking at ambient (2012) Table 22 clearly indicates that there was no water stress (0) for all the main development stages. Thus, this is the reason why the margin between attainable yields in 1997 is around 70% lower than the attainable yields in 2012.

160 El Nino (1997) vs ambient (2012) of Tango Sua yields in West Ghonghau 1600 140 1493 1400 120 1200 100 1000 80 800 60 600

441 Yield (kg/ha)

Rainfall (mm/d) (mm/d) Rainfall 40 400 20 200 0 0 0 20 40 60 80 100 120 140 160 180 200 Days after Planting Rainfall (mm/d) 2012 Ambient Rainfall (mm/d) 1997 Drought Tuber dm kg/ha Ambient (2012) Tuber dm kg/ha Drought (1997)

Figure 63: Yield of Tango Sua under drought and ambient condition in West Ghonghau

160 El Nino (1997) vs Ambient (2012) of Tango Sua yields in East Ghonghau 2000 140 1800 1734 1600 120 1400 100 1200 80 1000 60 800 600 Yield (kg/ha) 40 477 400

Precipitation (mm/d) (mm/d) Precipitation 20 200 0 0 0 20 40 60 80 100 120 140 160 180 200 Days after planting PRECIP mm/d-2012 PRECIP mm/d-1997 Tuber kg dm/ha-Ambient (2012) Tuber kg dm/ha -Drought (1997) Figure 64: Yield of Tango Sua under drought and ambient condition in East Ghonghau 162

160 El Nino (1997) vs Ambient (2012) Tango Sua yields in Sa'aiho 800 700 140 726 120 600 500 100 400 80 300

60 184 Yield (kg/ha) 200 Precipitation (mm/d) (mm/d) Precipitation 40 100 20 0 0 -100 0 20 40 60 80 100 120 140 160 180 200 Days after planting

PRECIP mm/d-Ambeint (2012) PRECIP mm/d-Drought (1997) Tuber kg dm/ha-Ambient (2012) Tuber kg dm/ha-Drought (1997)

Figure 65: Yield of Tango Sua under drought and ambient condition in Sa’aiho

It terms of growth, we look at the LAI of Tango Sua. Since LAI is a measure of the leaf area it provides estimates as to (1) photosynthetic and transpiration surface production and (2) crop growth. Figures 66-68 outlines that under drought event Tango Sua indicated an early peak (0.42) and (0.46) on day 69 during establishment stage and decline to (0.23) on day 129 during rapid vegetation stage for West and East Ghonghau. While for Sa’aiho, LAI peaks at Rapid vegetation stage (0.18) and declines thereafter to 0.02 at maturity. This indicated a low LAI at all three sites during drought compared to ambient condition where LAI peaks at day 129 with 0.83, 0.92 and 0.47 during the rapid vegetation growth and declines to 0.58, 0.64 and 0.38 at maturity and harvest stages. Overall, the LAI indicated that Tango Sua cultivated during drought has less leaf area which in turn signifies that the photosynthetic productivity is limited as well as crop growth.

163

1 LAI for ambient and drought (1997) of Tango Sua in West 0.9 Ghonghau 0.8 0.7 0.6 0.5 0.4 0.3 LAI (mm2/mm2) 0.2 0.1 0 0 20 40 60 80 100 120 140 160 180 200 Days after planting Ambient(2012) Drought (1997)

Figure 66: LAI for Tango Sua under drought and ambient condition in West Ghonghau

1 LAI for ambient (2012) and drought (1997) of Tango Sua in East 0.9 Ghonghau 0.8 0.7 0.6 0.5 0.4 0.3

LAI (mm2/mm2)LAI 0.2 0.1 0 0 20 40 60 80 100 120 140 160 180 200 Days after planting Ambient (2012) Drought (1997)

Figure 67: LAI for Tango Sua under drought and ambient condition in East Ghonghau

0.6 LAI for ambient (2012) and drought (1997) of Tango Sua in Sa'aiho 0.5 0.4 0.3 0.2

LAI (mm2/mm2)LAI 0.1 0 0 20 40 60 80 100 120 140 160 180 200 Days after planting

Ambient (2012) Drought (1997)

Figure 68: LAI for Tango Sua under drought and ambient condition in Sa’aiho

164

Nevertheless, the size of LAI is not the same for all varieties of taro and may be influence by planting density and water management (Manner and Taylor, 2011). An adequate LAI of 3 is noted to ensure sufficient photosynthesis (Manner and Taylor, 2011), however a study by Pardales and Belmonte, 1984 reported LAI of up to 4.7. The Low optimum LAI of Tango Sua (0.83, 0.92 and 0.47) under ambient condition for all three sites compared to the optimum LAI of 3 is because of the higher planting space between crops of around (110cm) and a planting density of about 0.8 plant/m2. Thus since crop yield productivity is positively correlates with LAI (Manner and Taylor, 2011) and that the increase in taro population density increases LAI (Pardales and Belmonte, 1984), the Bellonese farmers should consider increasing planting density by reducing planting space.

The reduction in LAI and crop growth as a consequence of drought is in line with the Bellonese perspective on drought impact. The Bellonese farmers mentioned that some of the impacts of drought on their crops including taro are; wilting, stunted or slow growth, decline in corm production, and early maturity with small size corms. LAI is also important for the Bellonese farming system and subsistence because they consume the leaves together with corms in their traditional dish known as Lengalenga. Wilting, death of crops, early maturity and small size of corms as reported by Bellonese during the drought event is closely represented by the model. For example, Table 25 showed that the maturity days, weight of corm and number of taro crop per meter square of area is less than that of ambient year as simulated by the model.

Table 25: Maturity days, weight per corm and survival density of Tango Sua in three sites

Site & Maturity day Unit weight at No. at maturity/ m2 Event maturity (g/unit) Site WG EG SA WG EG SA WG EG SA Ambient 188 188 188 170 200 74 5 5 6 Drought 187 187 189 55 60 16 3 3 5

The weight of a corm of Tango Sua under drought condition is 55g, 60g and 16g which is a reduction of about (67.6%, 70% and 78.4%) from what is harvested in an ambient condition. This clearly corroborates what the farmers reported; regarding the

165 size of corm harvested during drought event is much smaller compared to normal conditions. Additionally, the model also substantiates the wilting and death of taro as observed by farmers because the number of taro at maturity has reduced from 5 in ambient condition to 3 in both West and East Ghonghau in 1997 drought event. This signifies a loss of about 40% of taro crops due to drought event which in turn reduces the yield.

7.8 Errors and uncertainties associated with simulation

Due to the fact that the present study did not conduct field experiments to collect detailed crop management data, no statistics or errors are available within the simulations. The study relied on farmers’ knowledge to run the model and simulate the impacts. Therefore, there are likely sources of errors and uncertainties associated with this approach.

In terms of taro simulations, the solar radiation was manually adjusted within DSSAT to mimic the partially shaded taro cropping system used in Bellona. The local farmers normally cultivate taro under Hau trees (Hibiscus tiliaceus), which are not cleared or removed during bonga (brushing or clearing) and baakani (burning) of the garden area. This adjusted value for solar radiation is not the actual measured value on site. It was estimated and is subject to error and uncertainty. There is also uncertainty or error associated with the taro model in general as reported by Singh et al., (1998) which may affect our results. This is due to the following reasons:-  It has not been well validated for a while  Only few experiments were conducted for aroids or taro  Limited availability of vital accurate data for understanding crop development and growth in the literature, as taro is normally cultivated in developing countries and is referred to as low value crop.

The traditional cultivation of corn (Zea mays) on Bellona is normally by intercropping with sweet potatoes (Ipomea batatas), however intercropping is not provided for within DSSAT (Jones et al., 2003). Therefore, our simulated yield results in terms of corn may not be accurate. A future study should consider using

166

APSIM in terms of intercropping because it has components that incorporate intercropping (Carberry et al., 1996).

Overall, the calibration process used in the present study uses manually adjusted values (genetic coefficients) based on Bellonese farmers’ knowledge and not on controlled experimental field data. The genetic coefficients were calibrated manually using crop information data on maturity days, corm/tuber/grain initation time and yields to ensure that the model simulations closely mimic the maturity days and yields estimated by the farmers. The DSSAT model for all three crops has never been validated for Bellona. Such calibration process of using non-experimental data may produce simulation results which have some degrees of error. However, this is not statistically represented because we did not conduct a controlled experiment.

Nevertheless, these uncertainties do not indicate that this study has failed. This is because generally, DSSAT CSMs are deterministic in nature and use partial differential equations (Hoogenboom, 2012). According to Hoogenboom (2012), the overall uncertainty in DSSAT is controlled by supplying the CSMs with different series of long term weather data. Thus, in the present study, a 30 year (1982-2012) weather data was used to ensure uncertainty is minimized.

167

CHAPTER EIGHT: RECOMMENDED ADAPTATION MEASURES

Using the results from Bellonese perspectives in Chapters 4-6 together with results from DSSAT crop modelling (Chapter 7), various adaptations actions which may improve food crop management and production for the Bellonese communities were identified. These also include some good adaptation measures done by Bellonese farmers currently which should be encouraged and maintained. Secondary sources were also used to integrate useful approaches that may be beneficial for Bellonese cropping system.

8.1 Use of DSSAT to improve crop production 8.1.1 Selection of best cultivars to improve yield in 2030, 2055 and 2090

Using the sensitivity analysis of DSSAT model, simulation was conducted on all the varieties of the taro and corn in the model against ambient (2012) and future years of 2030, 2055, 2090 years to identify the best three varieties of each crop. Cassava was not included because the results in (Chapter 7) indicate that Lioka B1 yields will not be negatively affected by projected changes in climate and CO2 and because it is a more resilient crop.

According to the DSSAT sensitivity analysis simulation results in taro that performed well in ambient (2012) and future years (2030, 2055 2090) for all three sites were; (1) Bun long (2) Lehua (3) Tausala-Samoa.

Bun long cultivar produced higher yields in both West and East Ghonghau, while Lehua is more favourable to Sa’aiho under ambient and future simulations (Figures 69-72). It is projected that Bun long will yield an increase of about 4972kg/ha and 4681 kg/ha in comparison to the yields of Tango Sua by 2090 for West and East Ghonghau. Lehua is expected to yield an increase of 4209kg/ha by 2090 for Sa’aiho (Figure 72). Though Bun long and Lehua cultivars are more beneficial in all three sites in terms of higher production, they take longer (236 to 241days) to reach maturity under ambient (2012). That is, it takes these two cultivars about 48-53 days more than the local variety Tango Sua which takes 188 days to reach maturity. Thus, this means

168

that Tausala-Samoa though has lower yields than Bun long and Lehua may be the best cultivar for Bellona since it takes 224 days to mature which is closer to Tango Sua. However, there is possibility that all three cultivars can be cultivated together to maximize both crop production and early maturity.

8000 West Ghongau Ambient 2012 East Ghonghau Ambient 2012 A 8000 B 6000 6000

4000 4000

2000 2000

0 0 0 50 100 150 200 250 300 0 50 100 150 200 250 300 Days after planting Days after planting Tuber kg dm/ha-Lehua Tuber kg dm/ha-Lehua Tuber kg dm/ha -Bun long Tuber kg dm/ha-Bun long Tuber kg dm/ha-Tausala-Samoa Tuber kg dm/ha-Tausala-Samoa Tango Sua Tango Sua Sa'aiho Ambient 2012 6000 C Figure 69: Simulation yields at ambient (2012) of best three recommended taro 4000 cultivars for: 2000 A- West Ghonghau 0 0 50 100 150 200 250 30 B-East Ghonghau Days after planting C-Sa’aiho Tuber kg dm/ha-Lehua Tuber kg dm/ha-Bun long Tuber kg dm/ha-Tausala-Samoa Tango Sua

West Ghonghau 2030 East Ghonghau 2030 8000 A 8000 B 6000 6000

4000 4000 2000 2000 0 0 0 50 100 150 200 250 0 50 100 150 200 250 Days after planting Days after planting Tuber kg dm/ha-Lehua Tuber kg dm/ha -Lehua Tuber kg dm/ha-Bun long Tuber kg dm/ha-Bun long Tuber kg dm/ha-Tausala-Samoa Tuber kg dm/ha-Tausala-Samoa Tango Sua (2030) Tango Sua (2030)

Sa'aiho 2030 6000 C Figure 70: Simulation yields for 2030 of best three recommended taro cultivars for: 4000

A- West Ghonghau 2000 B-East Ghonghau 0 C-Sa’aiho 0 50 100 150 200 250 Days after planting Tuber kg dm/ha-Lehua Tuber kg dm/ha-Bun long Tuber kg dm/ha-Tausala-Samoa Tango Sua (2030)

169

West Ghonghau 2055 East Ghonghau 2055 8000 A 8000 B 6000 6000

4000 4000

2000 2000

0 0 0 50 100 150 200 250 0 50 100 150 200 250 Days after planting Days after planting Tuber kg dm/ha-Lehua Tuber kg dm/ha-Lehua Tuber kg dm/ha-Bun long Tuber kg dm/ha-Bun long Tuber kg dm/ha-Tausala-Samoa Tuber kg dm/ha-Tausala-Samoa Tango Sua (2055) Tango Sua (2055)

Sa'aiho 2055 6000 C Figure 71: Simulation yields for 2055 of best recommended three taro cultivars for: 4000 2000 A- West Ghonghau B-East Ghonghau 0 0 50 100 150 200 250 C-Sa’aiho Days after planting Tuber kg dm/ha-Lehua Tuber kg dm/ha-Bun long Tuber kg dm/ha-Tausala-Samoa Tango Sua (2055)

West Ghonghau 2090 East Ghonghau 2090 8000 A 8000 B 6000 6000 4000 4000 2000 2000

0 0 0 50 100 150 200 250 0 50 100 150 200 250 Days after planting Days after planting Tuber kg dm/ha-Lehua Tuber kg dm/ha-Lehua Tuber kg dm/ha-Bun long Tuber kg dm/ha-Bun long Tuber kg dm/ha-Tausala-Samoa Tuber kg dm/ha-Tausala-Samoa Tango Sua (2090) Tango Sua (2090)

Sa'aiho 2090 Figure 72: Simulation yields for 2090 of best 6000 C 5000 three recommended taro cultivars for: 4000 3000 A- West Ghonghau 2000 1000 B-East Ghonghau 0 0 50 100 150 200 250 C-Sa’aiho Days after planting

Tuber kg dm/ha-Lehua Tuber kg dm/ha-Bun long Tuber kg dm/ha-Tausala-Samoa Tango Sua (2090)

170

In terms of Corn, according to DSSAT simulations results (Figures 73-76), the best three cultivars ranked accordingly that may perform well in ambient (2012), and future years of 2030, 2055 and 2090 for the three sites are:

1. GL 482 2. PIO 3457 orig 3. WASH/GRAIN-1

In ambient (2012) condition, WASH/GRAIN-1 performed better in both West and East Ghonghau by an increase of 1182-1206kg/ha (Figure 73 A & B) while PIO 3475-orig performed better in Sa’aiho by 1222kg/ha in comparison to Koni (Figure 73 C). In

terms of future changes in temperature, rainfall and CO2, GL 482 displays more

resilience to the changes in climate and CO2 as it yields 1173kg/ha more than the local variety Koni by 2090 in all three sites.

2000 A West Ghonghau Ambient 2012 2000 East Ghonghau Ambient 2012 1470 B 1494 1500 1500 1117 1117 1000 1000 500 500 0 0 0 20 40 60 80 100 0 20 40 60 80 100 Days after planting Days after planting Grain wt kg/ha-WASH/GRAIN-1 Grain wt kg/ha-WASH/GRAIN-1 Grain wt kg/ha-PIO 3475-org Grain wt kg/ha-PIO 3475-orig Grain wt kg/ha-GL 482 Grain wt kg/ha-GL 482 Koni Koni

2000 Sa'aiho Ambient 2012 Figure 73: Simulation yields at ambient C 1526 1500 (2012) of best three recommended corn 1277 cultivars for: 1000 500 A- West Ghonghau 0 B-East Ghonghau 0 20 40 60 80 100 C-Sa’aiho Days after planting Grain wt kg/ha-WASH/GRAIN-1 Grain wt kg/ha-PIO 3475-orig Grain wt kg/ha-GL 482 Koni

171

2000 2000 A West Ghonghau 2030 1437 B East Ghonghau 2030 1437 1500 1500 1277 1391 1000 1000

500 500

0 0 0 20 40 60 80 100 0 20 40 60 80 100 Days after planting Days after planting

Grain wt kg/ha-WASH/GRAIN-1 Grain wt kg/ha-WASH/GRAIN-1 Grain wt kg/ha-PIO 3475-orig Grain wt kg/ha-PIO 3475-orig Grain wt kg/ha-GL 482 Grain wt kg/ha-GL 482 Koni Koni Figure74: Simulation yields for 2030 of best 2000 C Sa'aiho 2030 1455 three corn cultivars for: 1500 1397 1000 A- West Ghonghau 500 B-East Ghonghau

0 C-Sa’aiho 0 20 40 60 80 100 Days after planting Grain wt kg/ha-WASH/GRAIN-1 Grain wt kg/ha-PIO 3475-orig Grain wt kg/ha-GL 482 Koni

2000 West Ghonghau 2055 2000 East Ghonghau 2055 A 1517 B 1517 1500 1500

1000 1000 1392 1392 500 500 0 0 0 20 40 60 80 100 0 20 40 60 80 100 Days after planting Days after planting Grain wt kg/ha-WASH/GRAIN-1 Grain wt kg/ha-WASH/GRAIN-1 Grain wt kg/ha-PIO 3475-orig Grain wt kg/ha-PIO 3475-orig Grain wt kg/ha-GL 482 Grain wt kg/ha-GL 482 Koni Koni

2000 Sa'aiho 2055 Figure 75: Simulation yields for 2055 of best C 1517 1500 three corn cultivars for:

1000 1392 A- West Ghonghau

500 B-East Ghonghau 0 0 20 40 60 80 100 C-Sa’aiho

Days after planting Grain wt kg/ha-WASH/GRAIN-1 Grain wt kg/ha-PIO 3475-orig Grain wt kg/ha-GL 482 Koni

172

2000 A West Ghonghau 2090 2000 B East Ghonghau 2090 1437 1437 1500 1500 1226 1000 1000 1229

500 500

0 0 0 20 40 60 80 100 0 20 40 60 80 100 Days after planting Days after planting Grain wt kg/ha-WASH/GRAIN-1 Grain wt kg/ha-WASH/GRAIN-1 Grain wt kg/ha-PIO 3475-orig Grain wt kg/ha-PIO 3475-orig Grain wt kg/ha-GL 482 Grain wt kg/ha-GL 482 Koni Koni 2000 Figure 76: Simulation yields for 2090 of best C Sa'aiho 2090 1437 1500 three corn cultivars for:

1000 1304 A- West Ghonghau 500 B-East Ghonghau 0 C-Sa’aiho 0 20 40 60 80 100 Days after planting

Grain wt kg/ha-WASH/GRAIN-1 Grain wt kg/ha-PIO 3475-orig Grain wt kg/ha-GL 482 Koni

8.1.2 Change of cultivars to improve yield during ambient and ENSO (El Niño) condition of taro and corn

The three best recommended cultivars Lehua, Bun long and Tausala-Samoa have been simulated under the Past 11 El Niño years as compared to the local variety Tango Sua for all three sites (Figures 77-79). Overall, Bun long cultivar yielded higher production of corm in West and East Ghonghau and Lehua in Sa’iaho. However in terms of the severe drought of 1997, Lehua performed much better than Bun long and Tausala-Samoa. For instance in West Ghonghau, Lehua produced around 3563kg/ha while Tausala-Samoa 2427 kg/ha and Tango Sua only produced 854kg/ha. This indicates that amongst the three varieties, Lehua is more resilient to drought events and water stress conditions.

173

Best three cultivars vs Tango Sua for West Ghonghau 8000

7000 Lehua 6000 Bun long 5000 Tausala-Samoa Tango Sua 4000 3000 2000 1000 Yield dry Yield dry weight (kg/ha) 0 1980 1985 1990 1995 2000 2005 2010 2015

Figure 77: Recommended 3 best taro cultivars versus Tango Sua simulated under El Niño Years in West Ghonghau

9000 Best three cultivars vs Tango Sua for East Ghonghau 8000

7000 Lehua 6000 Bun long

5000 Tausala-Samoa Tango Sua 4000

3000

2000

1000 Yield dry Yield dry weight (kg/ha)

0 1980 1985 1990 1995 2000 2005 2010 2015

Figure 78: Recommended 3 best taro cultivars versus Tango Sua simulated under El Niño Years in East Ghonghau

7000 Best three cultivars vs Tango Sua for Sa'iaho 6000

5000 Lehua Bun long 4000 Tausala-Samoa 3000 Tango Sua

2000

1000 Yield dry Yield dry weight (kg/ha)

0 1980 1985 1990 1995 2000 2005 2010 2015

Figure 79: Recommended 3 best taro cultivars versus Tango Sua simulated under El Niño Years in Sa’aiho

174

In terms of Corn, the first option taken was to adjust the planting date because in Chapter 7 section 7.6 (Figure 60) there is a significant decline of Koni in West Ghonghau during the 1992 El Niño and also crop failure due to lack of germination in the severe drought in 1997 in both West and East Ghonghau. Thus, adjusting the planting date 1 month back from June 15th to May 15th was conducted and simulations indicated a better performance of Koni in all three sites (Figures 80-82).

350 Yield of Koni under adjusted and normal planting date in West Ghonghau 300

250

200

150 Adjusted planting date 100 (May 15th) Normal planting date 50 (June 15th)

0 1980 1985 1990 1995 2000 2005 2010 2015

Figure 80: Yield of Koni in normal planting date (June 15th) and adjusted planting date (May 15th) in West Ghonghau

350 Yield of Koni under adjusted and normal planting date in East Ghonghau 300

250

200 Adjusted planting date 150 (May 15th) Normal planting date 100 (June 15th)

50

0 1980 1985 1990 1995 2000 2005 2010 2015

Figure 81: Yield of Koni in normal planting date (June 15th) and adjusted planting date (May 15th) in East Ghonghau

175

350 Yield of Koni under adjusted and normal planting date in Sa'aiho

300

250

200 Adjusted planting date 150 (May 15th) Normal planting date 100 (June 15th)

50

0 1980 1985 1990 1995 2000 2005 2010 2015

Figure 82: Yield of Koni in normal planting date (June 15th) and adjusted planting date (May 15th) in Sa’aiho

The Koni seeds were germinated in West and East Ghonghau in 1997 under the adjusted planting date as opposed to the June 15th planting date. Thus, the most suitable planting date during the dry season (May to October) each year for Koni would be in May 15th as is simulated by the model for all three sites. The three recommended or best corn cultivars (WASH/GRAIN-1, PIO 3475 orig and GL 482) were also simulated under past El Niño years in comparison to the local variety Koni.

The results (Figures 83-85) indicate that all three recommended cultivars yielded higher yields than Koni. It is identified that WASH/GRAIN-1 performed much better in all three sites under drought conditions compared to PIO 3475 orig and GL 482. For instance, in West Ghonghau (Figure 83), during the severe 1997 drought, WASH/GRAIN produced about 1053 kg/ha grain dry weight compared to 853kg/ha and 559 kg/ha of PIO 3475 orig and GL 482 respectively. Thus, in ranking the three cultivars in accordance to drought tolerance, WASH/GRAIN-1 is the best and highly recommended followed by PIO 3475 orig and GL 482 to be cultivated in Bellona. However, in terms of future climate change, our previous simulations indicate that GL

482 is well suited for changes in elevated temperature, rainfall and CO2.

176

1800 Best three corn cultivars vs Koni under El Niño years in West Ghonghau

1600

1400

1200 WASH/GRAIN-1 1000 PIO 3475 orig

800 GL 482 600 Koni (2x)

400 Yield dry Yield dry weight (kg/ha) 200

0 1980 1985 1990 1995 2000 2005 2010 2015 Figure 83: Recommended 3 best corn cultivars versus Koni simulated under El Niño years in West Ghonghau

1800 Best three corn cultiuvars vs Koni under El Niño years in East Ghonghau 1600 1400 WASH/GRAIN-1 1200 PIO 3475 orig 1000 GL 482 800 Koni (2x) 600 400 200 0 1980 1985 1990 1995 2000 2005 2010 2015

Figure 84: Recommended 3 best corn cultivars versus Koni simulated under El Niño years in East Ghonghau

1800 Best three corn cultivars vs Koni under El Niño years in Sa'aiho

1600

1400 WASH/GRAIN-1 1200 PIO 3475 1000 GL 482 800 Koni (2x) 600

400

200

0 1980 1985 1990 1995 2000 2005 2010 2015

Figure 85: Recommended 3 best corn cultivars versus Koni simulated under El Niño years in Sa’aiho

177

8.2 Maximise use of mulch on Malanga Soil to improve soil fertility

Model simulations identified that during high rainfall, Malanga soil is highly vulnerable to nutrient leaching especially nitrogen. This has caused decline in crop production especially of taro (Tango Sua) in Sa’aiho which is cultivated on the Oolitic coarse sandy loam soil. Thus, it is highly recommended that since the use of fertilizers may not be an environmentally sound option especially in a limestone island like Bellona which has porous soils, increased mulching should be maximised. This is not only useful in supplying nutrients to crops but also acts as a natural barrier against extreme rainfall to fall directly and infiltrate the soil. Mulching with the use of dead leaves, coconut husks, and coconut fronds is recommended to be used as it provides ground cover, conserves soil moisture, increase soil fertility and control weeds (Kumar and Lal, 2012; Ranasinghe et al., 2003)

8.3 Use of legume plants as cover crops to improve soil fertility and crop production

The use of legume cover crops such as creeping peanuts (Arachis pintoi), pigeon peas (Cajanus cajan), cow peas (Vigna sinensis), mungbean (Vigna radiata) centro (Centrosema pubescens) or siratro (Macroptilium atropurpureum) were also recommended to be used to reduce the impacts of heavy rainfall on bare soil. Such legume plants not only provide ground cover but can also increase nitrogen input, use for intercropping, reduce weeds, minimize water loss from evaporation and improve soil fertility (Mousavi and Eskandari, 2011). In terms of taro (Colocasia esculenta) and cassava (Manihot esculenta) the use cow peas (Vigna sinensis) or mungbean (Vigna radiata) is recommended while peanuts (Arachis pintoi) for corn (Zea mays).

8.4 Use of sea grass and seaweed to improve soil fertility and control pest/disease

Using natural or organic means by way of increasing soil fertility and control pest is the best recommended adaptation measure for Bellonese farmers. There are marine plants like sea grass and seaweed which are available in Bellona that could be utilized 178 as mulch, compost or fertilizers to improve soil fertility and control pests. It has been reported by Thaman et al., (2012) that seaweeds such as Caulerpa racemosa, Amphiroa spp, Dictyota spp, Sargassum spp., Turbinaria spp., and sea grasses like Cymodocea serrulata are available in Bellona. Sea grass and seaweed has been used in other parts of world like in Africa (de la Torre-Castro and Ronnback, 2004), Argentina (Eryas et al., 2008) by local coastal farmers and has showed to improve growth and productivity. These marine resources are also found to enhance root growth, germination, resistance to pest, and improve tolerance of crops to stresses (Eryas et al., 2008; Blunden, 1991; Abetz, 1980). In Fiji, coastal farmers in Dakuni, Beqa Island used sea grasses that are collected on the beach as wracks and these have proven beneficial by increasing crop productivity in tomatoes, yagona, water melons and taro. During a field visit by the author with the PACE-SD food security team on the 19th-20th March on Dakuni village, the beneficial impacts of sea grass manure on tomatoes, yagona and water melons was observed. The farmers in Dakuni claimed that the sea grass manure also lengthens the fruiting period of tomatoes and provide resistance from pest and diseases.

Thus, such use of sea grass and seaweed as manure, compost or fertilizer is an essential adaptation strategy that can be used in Bellona since such resource is not only environmentally sound but is cost-free and is available. The use of sea grass or seaweed could also be useful in enhancing crops like taro from the infestation of taro beetle which is very severe on Bellona Island.

8.5 Cultivate corn on Malanga soil and taro on Kenge toaha/ungi

The results in model simulations show that taro (Tango Sua) produced more yields in Kenge toaha and Kenge ungi soils while corn (Koni) is more suitable to be cultivated in Malanga soil under both projected climate change and El Niño years. Thus, it is recommended that Bellonese should reserve their fertile soils (Kenge ungi/toaha) for taro whilst using Malanga soils for corn and cassava.

179

8.6 Use of cultural and biological control to manage taro beetle (Bonggu)

Though chemicals such as imidacloprid and bifenthrin are reported to be effective in controlling the taro beetle and increase taro corm production by about 95% (Lal, 2008) in trial sites in Fiji and Papua New Guinea, it is unclear if such insecticides are environmentally safe for a small atoll Island such as Bellona. Thus, cultural and biological control is recommended to be more applicable for controlling the taro beetle in Bellona.

Cultural control methods that are recommended are the used of wood ash, mulching, intercropping, adjusting planting time, intercropping, trap cropping, use of tolerant varieties (Aloalii et al., 1993), crop rotation, slash and burn vegetation, use repellant plants, ensuring weed free plots, and establishing new plots further away from old plots (Lal, 2008). However, according to Lal (2008), in terms of tolerant or resistant taro varieties, no adequate cultivar is identified.

Biological control by the use of fungus Metarhizium anisopliae on taro planting holes has increased marketable corm yields by about 30% on trial sites Fiji and Papua New Guinea (Lal, 2008). This fungus is reported however to have slow effect on killing Taro beetles; which could be an useful measure to control the extent of effect of taro beetle in Bellona.

8.7 Maximise cultivation of cassava since it is more resilient to extreme events

Our simulations have indicated that cassava has the potential to increase its attainable yields even with the projected change in Climate and CO2. It is more resilient to pest attack, drought and performed well on less fertile soil like Malanga. Thus, cultivation of cassava does not require any inputs like manure or fertilizer. Therefore, increasing the production of cassava can be a feasible adaptation strategy to support other food crops like taro, pana and yam during extreme events. More so though cassava may be harvested within three months, it can be left in ground and harvested continually for up to 18-20 months. Thus, could be use as a food reserve for drought or cyclone events.

180

8.8 Increase awareness, education and improve decision making of farmers

Enhancing Bellonese farmers’ knowledge on farming techniques and integrating with modern knowledge is very crucial in ensuring adequate crop management and production under the changing climate. It is important that the awareness of the scientific evidence of changing climate which may affect the normal planting calendar be transferred to the local farmers so that actions such as (1) shifting of planting dates, (2) exploring tolerant crop varieties and (3) improving crop management inputs such as mulching, composting, shading are implemented. With the uncertainty of future climate change projections; one possible solution is preparing “for the unknown” based farmer knowledge and an integration the scientific understanding of climate change (Tall, 2013).

Though, radio has been used as a medium for providing awareness to farmers both on crop management information, forecasting and early warning system, the use of SMS or text messages could also be a way forward in disseminating such information and alerts. The use of SMS has been found to be effective in improving decision making and crop management in Africa and South Asia (Tall, 2013) and such lessons could be used by relevant agricultural and meteorological agencies in Solomon Islands to disseminate information to farmers on remote islands like Bellona. The benefits of using such telecommunication technology from Africa and South Asia is that it equips farmers with climate information, forecast and early warnings which guided them in making adequate decision under uncertainty and also strengthens their preparedness (Tall, 2013). The use of crop models like DSSAT is also a useful tool that can be used to assist farmers in making adequate decision in terms of improving crop management and production.

8.9 Some current Bellonese adaptation practices or measures that are encouraged and should be maintained

There were some adaptation practices that the Bellonese farmers use which is beneficial to their crop management and production. These practices appended below were recommended and encouraged to be maintained by the Bellonese farmers:-

181

 shifting cultivation  bush fallow practice  crop rotation  intercropping  diversification of crops  shade maintenance  mulching  planting fast yielding/resilient crops  adjust planting dates  increase number of plots  change planting sites  increase mound size  planting distribution according to soil fertility and crop type

8.10 Weather station in Bellona

Having a weather station established on Bellona is crucial for generating reliable weather and climate data which will improve decision making on crop production and management. It is important that such station be established to allow farmers and researchers to obtain relevant data in which important coping and adaptation measures may be developed from understanding the climate trend.

182

CHAPTER NINE: CONCLUSIONS, LIMITATIONS AND RECOMMENDATION

9.1 Conclusion

Climate change affects crops, food security and agriculture on small island developing states like the Solomon Islands. As evidenced in the development of this thesis, this has become an urgent livelihood priority. The extent by which the magnitude of the impacts of climate change including climate variability and extreme events has on remote, isolated atoll island communities is by far the worst. However, there are limited studies apart from vulnerability assessments that evaluate and integrate the local farmers’ perspective and using crop model to simulate the likely future impacts of climate change on their food crops. This study therefore contributes to the understanding of past, current and future impacts and vulnerabilities of climate change on food crops in the raised atoll island community of Bellona.

This final chapter provides the conclusion of the study by addressing each of the three main objectives of this thesis listed below. At the end of the chapter, the limitation of study is highlighted and also several recommendations for future potential studies that may be undertaken.

Objective one: To understand and identify vulnerability, impacts and adaptations to climate change and extreme events on food crops and cropping systems on the Atoll community of Bellona.

 Vulnerability

In terms of vulnerability, the study looks at household characteristics, livelihoods and their food crops. It is identified that more than 90% of the households have permanent structures currently however it is likely that the standard of some these permanent structures may not be adequate for withstanding extreme events like cyclones. Lack of surface water is a major issue in Bellona and thus households entirely depend on rainwater for all their domestic needs making them and their crop system highly

183 vulnerable. Since their crops depend on rainfall for irrigation, their cropping system is extremely vulnerable to drought events. Limited access to efficient sea transport, unscheduled shipping services, minimal economic activities and the remoteness of Bellona Island are some of the significant vulnerability issues faced by the Bellonese.

The results indicate that there is shift in terms of cultivated crops in Bellona as traditional crops like yam and pana which were dominant in the past 50 years have declined in cultivation and productivity. This is because these crops are highly vulnerable to the changing climate and unpredictable weather conditions which affect the growth and yield. The number of garden plots per household currently has declined by average of 39.7% compared to the past 50 years. A possible reason for this declined trend is because of changes in access to land and reliance on imported processed food from Honiara.

Fallow periods have also declined compared to past 50 years for crops like taro and sweet potatoes due to increased pressure on land availability via increased land intensification and extended cropping periods. Land intensification is more pronounced in the centre of the island where soil fertility is high thus making this area highly susceptible to the impacts of climate change and variability. The cropping system in Bellona lacks management inputs like manure, fertilizer, pesticides, and insecticides, and has more or less remained the same from past 50 years. This indicates the vulnerability associated with cropping system, crops and soil to increased exposure and decreased adaptive capacity to climate extremes in terms of soil fertility, crop production, and pest and disease management.

 Impacts

o Temperature and Rainfall The majority of the households (96.6%) and (98.3%) experienced changes in the temperature and rainfall over the past 30 years. In terms of temperature increase, the Bellonese observed wilting, decline in crop yield, quality, tuber/corm size, survival rate of young seedlings, crop growth, early maturity, abnormality of fruits/tubers, and increase in pest and diseases. While increase in rainfall was reported to decline yield,

184 crop growth, survival rate of young seedlings, rotting, loss of some crop varieties, increase in pest/diseases, shifting of planting date, shortage of planting materials, delay in maturity, loss of taste, and increase in weeds. o Extreme events Cyclones The impacts of cyclones on the food crops include, uprooting, rotting, physical damage by fallen trees and debris, shortage of planting materials, washing away of newly planted crops, death of tree crops, increase in pest/diseases, and decrease in fruit size. Drought Droughts were observed by the Bellonese to have caused wilting, death of crops, reduction in fruit/tuber size, abnormality in crop growth, early maturity, decline in yield, and deformation of tubers.

 Adaptation

Bellonese people in the study area have employed several traditional adaptations and coping strategies to enhance their cropping systems and crop production. Adaptation strategies include; shifting cultivation, bush fallow practice, crop rotation, intercropping, diversification of crops, shade maintenance, mulching, planting fast yielding/resilient crops, adjust planting dates, increase number of plots, change planting sites, increase mound size, and planting distribution according to soil fertility and crop type. Coping strategies include; pruning of crops, use of table salt for pest control, use of wild crops as alternative food source, harvest conservation, use of tradition warning signs, change of eating style, barter system, relief from external sources and remittance.

Objective two: To simulate and evaluate the impacts of future change of climate:-rainfall, temperature, carbon dioxide on Taro (Colocasia esculenta), Cassava (Manihot esculenta) and Corn (Zea mays) yield by using DSSAT crop simulation model.

The model simulations indicate that for taro-Tango Sua, the yields will increase for

185 both West and East Ghonghau by 4.6% and 6.7% however decline for Sa’aiho by 12% by 2090. In terms of cassava- Lioka B1, there will be an increase in yields with increase in temperature, rainfall and CO2 in West/East Ghonghau and Sa’aiho by 18.5%, 24.1% and 15.6% by 2090. On contrary, for corn-Koni, the model projected a decline by about 8.3% for both West and East Ghonghau and 13.2% for Sa’aiho by 2090.

These results suggest that taro (Tango Sua) may still be cultivated on Bellona by 2090 however; it may be restricted to only two soil types (Kenge ungi and Kenge toaha) which are more fertile and less susceptible to leaching and nutrient loss. Cassava (Lioka B1) is projected to be the most resilient of the three crops and may yield higher tuber production under projected increase in temperature, rainfall and

CO2. This is because it is more resilient to pest, drought and may be grown on almost any type of soil even in places where irrigation is not possible like Bellona (Witter, 1995). It was concluded that the most vulnerable crop is corn (Koni); which may also be susceptible to yield reductions with the projected future climate and CO2 scenarios. This is because of its C4 photosynthetic pathway that enhances water use efficiency rather than photosynthesis under elevated CO2 concentrations (Witter, 1995).

Objective three: To identify and recommend relevant adaptation strategies, options or measures to improve food crop production and management.

 Recommend best cultivars for drought events and projected climate change

Using DSSAT three best cultivars that were recommended for taro and corn under both El Niño related drought event and future climate change are: Taro:- 1. Bun long 2. Lehua 3. Tausala-Samoa Corn:- 1. GL 482 2. PIO 3457 orig

186

3. WASH/GRAIN-1

The three cultivars of taro have been identified to increase yield ranging from 3574- 4972kg/ha for West Ghonghau, 3666-4681kg/ha for East Ghonghau and 3027-4209kg/ha for Sa’aiho in comparison to the local variety Tango Sua by 2090. In terms of corn, the three selected cultivars increase yields ranging from 962-1173kg/ha for West Ghonghau, 965-1173kg/ha for East Ghonghau and 1040-1173kg/ha for Sa’aiho by 2090.

To improve soil fertility and crop production, it is recommended that:-  The use of mulch should be maximised mainly on Malanga soil  Intercropping with legume plants such as peanuts (Arachis pintoi), cow peas (Vigna sinensis), and mungbean (Vigna radiata).  The use of sea grass and seaweed to improve soil fertility and control pest/disease

It is further recommended that corn should be cultivated on Malanga soil while taro on Kenge ungi and toaha. To control pest and diseases, it is recommended that cultural and biological methods be used in Bellona.

Given the cassava has been observed and projected to be more resilient to changes in climate, and impacts of extreme events, it is recommended that cultivation of this tolerant crop should be maximize.

It is also recommended that the use of mobile SMS and DSSAT; which are important ways to disseminate information, forecasts, alerts and improve decision-making for Bellonese farmers. Along with that, improved awareness campaigns, education and strategically placed early warning systems would also help farmers understand the importance of both traditional knowledge of weather and climate and how that can be blended successfully with scientific knowledge

Finally, it is recommended that having an established weather station on Bellona is crucial for generating reliable weather and climate data to improve decision making on

187 crop production and management.

9.2 Limitation of Study

(i) Rainfall, temperature and sunshine hours’ data from the closest weather station to the study site was used due to the lack of weather station on Bellona Island. (ii) Minimum data requirements for DSSAT crop model in terms of crop management information were obtained from farmer’s response and not actually measured. (iii) A comprehensive model calibration and validation for taro, cassava and corn will require detail experiment data which are unavailable for Bellona. (iv) DSSAT does not take into account intercropping, so our results in terms of corn may not be accurate as corn is normally intercropped with sweet potatoes. (v) Only one crop model software (DSSAT) was used for evaluation. (vi) The model simulation only use 1 PCCSP emission scenario A2 (high) (vii) Our simulation focus on attainable yields therefore does not take in account other parameters that may affect the growth and yield of crops like pest, disease and weeds.

9.3 Recommendation for further studies

The following are recommendations for further future studies: (i) To conduct diversified field experiments in Bellona to further validate the model to obtain a more detail quantitative crop management data. (ii) To create sweet potatoes model as it is becoming an important crop in Bellona and Pacific Islands due to its early maturity and resilience. (iii) To use other crop modelling software such as APSIM, CropSyst, AGROSIM and HERMES to compare how crops respond to different models and identify which one is more efficient and reliable.

(iv) To consider using other CO2 emission scenarios like A1B (medium) and B1 (low) to see the difference in terms of yield response by the crops.

188

BIBLIOGRAPHY

Abetz, P. 1980. Seaweed extracts: have they a place in Australian agriculture or horticulture? Journal of the Australian Institute of Agricultural Science, 46, 23-29. Achakzai, A. K. K. 2009. Effect of water stress on imbibition, germination and seedling growth of maize cultivars. Sarhad Journal of Agriculture, 25, 165-172. ADB 2009. Building climate resilience in the agriculture sector in Asia and the Pacific, Philippines, Asian Development Bank. Ainsworth, E. A. & Long, S. P. 2005. What have we learned from 15 years of free-air CO2 enrichment (FACE)? A meta-analysis of the responses of photosynthesis, canopy properties and plant production to rising CO2. New Phytologist, 165, 351-372. Ainsworth, E. A., Rogers, A., Nelson, R. & Long, S. P. 2004. Testing the "source– sink" hypothesis of down-regulation of photosynthesis in elevated CO2 in the field with single gene substitutions in Glycine max. Agricultural and Forest Meteorology, 122, 85-94. Alexandrov, V. A. & Hoogenboom, G. 2000. The impact of climate variability and change on crop yield in Bulgaria. Agricultural and Forest Meteorology, 104, 315-327. Alig, R. J., Adams, D., Joyce, L. & Sohngen, B. 2004. Climate change impacts and adaptation in forestry: responses by trees and market choices. American Agricultural Economics Association, 11 Allen, B. & Bourke, R. M. 2009. People, Land and Environment. In: BOURKE, R. M. & HARWOOD, T. (eds.) Food and Agriculture in Papua New Guinea. ANU E-press. Allen, L. H., Jr & Prasad, P. V. V. 2004. Crop Responses to Elevated Carbon Dioxide. Encyclopedia of Plant and Crop Science, 346-348. Allison, I., Bindoff, L. N., Bindschadler, A. R., Cox, M. P., de Noblet, N., England, H. M., Francis, E. J., Gruber, N., Haywood, M. A., Karoly, J. D., Kaser, G., Le Quéré, C., Lenton, M. T., Mann, E. M., McNeil, I. B., Pitman, J. A., Rahmstorf, S., Rignot, E., Schellnhuber, J. H., Schneider, H. S., Sherwood, C. S., Somerville, J. C. R., Steffen, K., Steig, J. E., Visbeck, M. & Weaver, J. A. 2009. The Copenhagen Diagnosis. Updating the World on the Latest Climate Science. Sydney, Australia: he University of New South Wales Climate Change Research Centre (CCRC). Aloalii, I., Masamdu, R., Theunis, W. & Thistleton, B. 1993. Prospects for Biological Control of Taro Beetles, Papuana spp. (Coleoptera: Scarabaeidae), in the South Pacific: In: Proceedings of the Sustainable Taro Culture for the Pacific Conference, September 24–25 1992. Honolulu, Hawaii: Hawaii Institute of Tropical Agriculture and Human Resources Research Extension, University of Hawaii. Antle, J. M., Capalbo, M. S., Elliott, T. E. & Paustian, H. K. 2004. Adaptation, spatial heterogeneity, and the vulnerability of agricultural systems to climate change and CO2 fertilization: an integrated assessment approach. Climatic Change, 64, 289-315. Baiabe, T. 20 July 2012. RE: Personal Interview. Baker, J. T. 2004. Yield responses of southern U.S. rice cultivars to CO2 and 189

temperature. Agricultural and Forest Meteorology, 122, 129-137. Barnett, J. & Adger, W. N. 2003. Climate dangers and atoll countries. Climatic Change, 61, 321–337. Bates, B., Kundzewicz, Z. W., Wu, S. & Palutikof, J. 2008. Climate Change and Water. Technical Paper of the Intergovernmental Panel on Climate Change. Geneva. Berhanu, G. & Gerald, S. 1998. The Economic Importance Of Crop Rotation Systems: Evidence From The Literature," Staff Papers 11690, Economics.: Michigan State University, Department of Agricultural, Food, and Resource Berkes, F. & Jolly, D. 2001. Adapting to climate change: social-ecological resilience in a Canadian western Arctic community. Conservation Ecology, 5, 18. Berry, P. M., Rounsevell, M. D. A., Harrison, P. A. & Audsley, E. 2006. Assessing the vulnerability of agricultural land use and species to climate change and the role of policy in facilitating adaptation. Environmental Science & Policy, 9, 189-204. Birch-Thomsen, T., Reenberg, A., Mertz, O. & Fog, B. 2010. Continuity and change: Spatiotemporal land use dynamics on Bellona Island, Solomon Islands. Singapore Journal of Tropical Geography, 31, 27-40. Blum, A. 1996. Crop responses to drought and interpretation of adaptation. Plant Growth Regulation, 20, 135-148. Blunden, G. 1991. Agricultural uses of seaweeds and seaweed extracts. In: GUIRY, D. M. & BLUNDEN, G. (eds.) Seaweed Resources in Europe: Uses and Potential. Chichester, UK: John Wiley & Sons Ltd. Boote, K. J., Jones, J. W. & Pickering, N. B. 1996. Potential uses and limitations of crop models. Agronomie Journal, 88, 704-716. Borggaard, O. K., Koch, C. B., Elberling, B., Breuning-Madsen, H. & Stemmerik, L. 2012. Composition of characteristic soils on the raised atoll Bellona, Solomon Islands. Geoderma, 170, 186–194. Bourke, R. M., McGregor, A., Allen, M. G., Evans, B. R., Mullen, B. F., Pollard, A. A., Wairiu, M. & Zotalis, S. 2006. Solomon Islands Smallholder Agriculture Study.Volume 1. Main Findings and Recommendations. Canberra: Australian Government AusAid. Breuning-Madsen, H., Bech-Bruun, T. & Elberling, B. 2010. An indigenous soil classification system for Bellona Island – a raised atoll in Solomon Islands. Singapore Journal of Tropical Geography, 31, 85–99. Brisson, N., Dorel, M. & Ozier-Lafontaine, H. Year. Effects of soil management and water regime on the banana growth between planting and flowering Simulation using the STICS model. In: GALAN, S. V., ed. Proceedings of the International Symposium Banan in Subtropics, Acta Hort, 1998. 229-238. Brisson, N., Gary, C., Justes, E., Roche, R., Mary, B., Ripoche, D., Zimmer, D., Sierra, J., Bertuzzi, P., Burger, P., Bussie`re, F., Cabidoche, M. Y., Cellier, P., Debaeke, P., Gaudille`re, P. J., He´nault, C., Maraux, F., Seguin, B. & Sinoquet, H. 2003. An overview of the crop model STICS. European Journal of Agronomy, 18, 309-332. Bruckler, L., Lafolie, F., Ruy, S., Granier, J. & Beaudequin, D. 2000. Modelling the agricultural and environmental consequences of non-uniform irrigation on a maize crop. Agronomie, 20, 609-624. . Buljovcic, Z. & Engels, C. 2001. Nitrate uptake ability by maize roots during and after drought stress. Plant and Soil, 229, 125-135. 190

Caldwell, C. R., Britz, J. S. & Mirecki, M. R. 2005. Effect of temperature, elevated carbon dioxide, and drought during seed development on the isoflavone content of dwarf soybean [Glycine max (L.) Merrill] grown in controlled environments. Journal of Agricultural and Food Chemistry, 53, 1125-1129. Carberry, P. S., Adiku, S. G. K., McCown, R. L. & Keating, B. A. 1996. Application of the APSIM cropping systems model to intercropping systems. In: ITO, C., JOHANSEN, C., ADU-GYAMFI, K., KATAYAMA, K., KUMAR-RAO, J. V. D. K. & REGO, T. J. (eds.) Dynamics of roots and nitrogen in cropping systems of the semi-arid tropics. Japan Int. Res. Centre Agric. Sci. Christensen, S. 1975. Subsistence on Bellona Island (Mungiki): A study of the Cultural Ecology of a Polynesian Outlier in the British Solomon Islands Protectorate, National Museum of Denmark & Royal Danish Geographical Society. Conde, C., Liverman, D., Flores, M., Ferrer, R., Ara'ujo, R., Betancourt, E., Villarreal, G. & Gay, C. 1997. Vulnerability of rainfed maize crops in Mexico to climate change. Climate Research, 9, 17-23. Crafts-Brandner, J. S. & Michael E. Salvucci, E. M. 2002. Sensitivity of Photosynthesis in a C4 Plant, Maize, to Heat Stress. Plant Physiology, 129, 1773–1780. Crespin , I. 1956a. Report on samples from Rennell and Bellona atolls, Solomon Islands. Record 1956/50. Canberra: Australian Bureau Mineral Resources. Crespin, I. 1965b. Report on rock samples from Rennell and Bellona Island. Record 1956/92. Canberra: Australian Bureau Mineral Resources. Dalsgaard, K. 1970. Description of some highly porous soil profiles on the Island of Bellona, an emerged atoll in the Solomon Islands. Geografisk Tidsskrift, Bind 69. de la Torre-Castro, M. & Ronnback, P. 2004. Links between humans and seagrasses—an example from tropical East Africa. Ocean & Coastal Management 47, 361-387. de Oliveira Bernardes, C., da Silva Martins, C. A., Santos Lopes, F., da Rocha, M. J. R. & Teixeira Xavier, T. M. 2011. Leaf Area, Leaf area Index and Light extinction coefficient for Taro culture. Biosfera. Centro Científico Conhecer - Goiânia. Dixon, G. R. 2009. The Impact of Climate and Global Change on Crop Production. In: LETCHER, T. M. (ed.) Climate Change: Observed Impacts on Planet Earth. Elsevier B.V. Easterling, W. E., Aggarwal, K. P., Batima, K. M. P., Brander, L., Erda, S. M., Howden, A., Kirilenko, J., Morton, J. F., Soussana, J., Schmidhuber & Tubiello, N. F. 2007. Food, fibre and forest products. In: PARRY, L. M., O.F. CANZIANI, F. O., PALUTIKOF, P. J., VAN DER LINDEN, J. P. & HANSON, E. C. (eds.) Climate Change 2007: Impacts, Adaptation and Vulnerability. Contribution of Working Group II to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge. El-Sharkawy, A. M. 2004. Cassava biology and physiology. Plant Molecular Biology, 56, 481–501. Eryas, C. M., Defosse, E. G. & Dellatorre, F. 2008. Seaweed Compost as an Amendment for Horticultural Soils in Patagonia, Argentina. Compost Science & Utilization, 16, 119-124. FAO 2008. Climate Change and Food Security in Pacific Island Countries. Rome, Italy. 191

Flis, B., Zimnoch-Guzowska, E. & Mańkowski, D. 2012. Correlations among Yield, Taste, Tuber Characteristics and Mineral Contents of Potato Cultivars Grown at Different Growing Conditions. Journal of Agricultural Science, 4, 197-207. Folland, C. K., Salinger, M. J., Jiang, N. & Rayner, A. N. 2003. Trends and Variations in South Pacific Islands and Ocean Surface Temperatures. Journal of Climate, 16, 2859- 2875. Gan, J. 2004. Risk and damage of southern pine beetle outbreaks under global climate change. Forest Ecology and Management, 191, 61-71. Gifford, R. M. 2004. The CO2 fertilising effect – does it occur in the real world? New Phytologist, 163, 221-225. Google, E. 2012. Map of Bellona Island: 11o17’37.68’’S, 159o47’33.60’’E, elevation 60ft. Grossman-Clarke, S., Pinter, P. J., Jr., K., T , Kimball, B. A., Hunsaker, D. J., Wall, G. W., Garcia, R. L. & LaMorte, R. L. 2001. Modeling a spring wheat crop canopy under elevated CO2 and drought. New Phytologist, 150, 315-335. Hansell, J. R. F. & Wall, J. R. D. 1976. Land Resource Study- Land Resource Study-Volume 8: Outer Islands. Tolworth Tower. Hay, J. E., Mimura, N., Campbell, J., Fifita, S., Koshy, K., McLean, R. F., Nakalevu, T., Nunn, P. & de Wet, N. 2003. Climate Variability and Change and Sea-level Rise in the Pacific Islands: Region A Resource Book for Policy and Decision Makers, Educators and other Stakeholders, SPREP. Apia, Samoa. Hoegh-Guldberg, O. & Bruno, F. J. 2010. The Impact of Climate Change on the World's Marine Ecosystems. Science, 328, 1523-1528. Hoogenboom, G. 2012. Deterministic or stochastic? [Online]. Available: http://dssat.net/734 [Accessed 14/10/2013]. Hoogenboom, G., Jones, J. W., Wilkens, P. W., Batchelor, W. D., Bowen, W. T., Hunt, L. A., Pickering, N. B., Singh, U., Godwin, D. C., Baer, B., Boote, K. J., Ritchie, J. T. & White, J. W. 1994. Crop models. In: TSUJI.G.Y, UEHARA.G & BALAS.S (eds.) DSSAT Version 3. Honolulu,: University of Hawaii. Hoogenboom, G., Jones, W. J., Porter, H. C., Wilkens, W. P., Boote, J. K., Hunt, A. L. & Tsuji, Y. G. 2010a. Decision Support System for Agrotechnology Transfer Version 4.5. Volume 1: Overview. University of Hawaii, Honolulu, HI. Hoogenboom, G., Jones, W. J., Wilkens, W. P., Porter, H. C., Boote, J. K., Hunt, A. L., Singh, U., Lizaso, L. J., White, W. J., Uryasev, O., Royce, S. F., Ogoshi, R., Gijsman, J. A., Tsuji, Y. G. & Koo, J. 2010b. Decision Support System for Agrotechnology Transfer (DSSAT) Version 4.5 [CD-ROM]. University of Hawaii, Honolulu, Hawaii. Hoogenboom, G., Wilkens, P. W., Thornton, P. K., Jones, J. W., Hunt, L. A. & Imamura, D. T. 1999. Decision support system for agrotechnology transfer v3.5. In: HOOGENBOOM, G., WILKENS, P. W. & TSUJI, G. Y. (eds.) DSSAT version 3. Honolulu: University of Hawaii,. IPCC 2007a. Summary for Policymakers. In: SOLOMON, S., D. QIN, M. MANNING, Z. CHEN, M. MARQUIS, K.B. AVERYT, M.TIGNOR AND H.L. MILLER (EDS.) (ed.) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge, United Kingdom and New York, NY, USA. IPCC 2007b. The Physical Science Basis :Working Group II to the Fourth 192

Assessment Report of the IPCC. Cambridge. IPCC 2007c. Summary for Policymakers. Climate Change 2007: Impacts, Adaptation and Vulnerability.Contribution of Working Group II to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge, UK. IPCC 2007d. Climate Impacts, Adaptation and Vulnerability, Working Group II to the Fourth Assessment Report of the IPCC. Cambridge. IPCC 2007e. Annex 1:Glossary. In: BAEDE, A. P. M. (ed.) Climate Change 2007:The Physical Science Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge. IPCC 2007f. Appendix 1:Glossary. Climate Change 2007: Impacts, Adaptation and Vulnerability.Contribution of Working Group II to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge,UK. IPCC 2012. Managing the Risks of Extreme Events and Disasters to Advance Climate Change Adaptation. A Special Report of Working Groups I and II of the Intergovernmental Panel on Climate Change. In: FIELD, C. B., BARROS, V., STOCKER, F. T., QIN, D., DOKKEN, J. D., EBI, L. K., MASTRANDREA, D. M., MACH, J. K., PLATTNER, K. G., ALLEN, K. S., TIGNOR, M. & MIDGLEY, M. P. (eds.). Cambridge, UK, and New York, NY, USA. Jablonski, L. M., Wang, X. & Curtis, S. P. 2002. Plant reproduction under elevated CO2 conditions: a meta-analysis of reports on 79 crop and wild species. New Phytologist, 156, 9-26. Jackson, G., Smith, C., Tahuniari, S., Wame, L., Tsatsia, H., Varuia, G., Muda, F., Watoto, C., Kabu, R., Nanau, A., Lina, V., Ridau, M., Totu, C., Eddie Bosili, Olga Chapangi, Deolyn Komba & Reid, C. 2011. Crop Protection Curriculum and Teaching Methods for Solomon Islands: Improving Plant Protection in Solomon Islands. The Australian Centre for International Agricultural Research (ACIAR). Jansen, T., Mullen, B. F., Pollard, A. A., Maemouri, R. K., Watoto, C. & Iramu, E. 2006. Solomon Islands smallholder agriculture study.Volume 2: Subsistence production, livestock and social analysis. Canberra: Australian Government AusAid. Jones, J. W., Hoogenboom, G., Porter, C. H., Boote, K. J., Batchelor, W. D., Hunt, L. A., Wilkens, P. W., Singh, U., Gijsman, A. J. & Ritchie, J. T. 2003. The DSSAT cropping system model. European Journal of Agronomy, 18, 235-265. Jones, J. W., Tsuji, G. Y., Hoogenboom, G., Hunt, L. A., Thornton, P. K., Wilkens, P. W., Imamura, D. T., Bowen, W. T. & Singh, U. 1998. Decision support system for agrotechnology transfer; DSSAT v3. In: TSUJI, G. Y., HOOGENBOOM, G. & THORNTON, P. K. (eds.) Understanding Options for Agricultural Production. Dordrecht, the Netherlands: Kluwer Academic Publishers. Justes, E., Dorsainvil, F., Brisson, N., Ripoche, D., Thie´beau, P. & Mary, B. Year. Evaluation and improvement of the soilcrop model STICS for predicting water and nitrogen budgets during fallow period between two main crops or under catch crops. In: Proceedings of the Second International Symposium on Modelling Cropping System, Florence (Italy), July 16-18, 2001. Karnosky, D. F. 2003. Impact of elevated atmospheric CO2 on forest trees and forest 193

ecosystems: knowledge gaps. Environment International, 29, 161-169. Kartschall, T., Grossman, S., Pinter, P. J., Garcia, R. L. J., Kimball, B. A., Wall, G. W., Hunsaker, D. J. & LaMorte, R. L. 1995. A simulation of phenology, growth, carbon dioxide exchange and yields under ambient atmosphere and free-air carbon dioxide enrichment (FACE) Maricopa, Arizona, for wheat. Journal of Biogeography, 22, 611-622. Keating, A. B., Carberry, S. P., Hammer, L. G., Probert, E. M., Robertson, J. M., Holzworth, D., Huth, I. N., Hargreaves, G. N. J., Meinke, H., Hochman, Z., McLean, G., Verburg, K., Snow, V., Dimes, P. J., Silburn, M., Wang, E., Brown, S., Bristow, L. K., Asseng, S., Chapman, S., McCown, L. R., Freebairn, M. D. & Smith, J. C. 2003. An overview of APSIM, a model designed for farming systems simulation. European Journal of Agronomy, 18, 267-288. Kersebaum, K. C. 1995. Application of a simple management model to simulate water and nitrogen dynamics. Ecological Modelling, 81, 145-156. Kersebaum, K. C., Mirschel, W., Wenkel, K. O., Manderscheid, R., Weigel, H. J. & Nendel, C. 2009. Modelling climate change impacts on crop growth and management in Germany. In: UTSET, A. (ed.) Climate Variability, Modelling Tools and Agricultural Decision making. New York: Nova Science Publishers, Inc. Khayatnezhad, M., Gholamin, R., Jamaati-e-Somarin, S. & Zabihi-e-Mahmoodabad, R. 2010. Effects of Peg Stress on Corn Cultivars (Zea mays L.) At Germination Stage. World Applied Sciences Journal, 11, 504-506. Kimball, B. A., Kobayashi, K. & Bindi, M. 2002. Responses of agricultural crops to free-air CO2 enrichment. Advances in Agronomy, 77, 293-368

Kirch, P. V. 1996. Late Holocene human-induced modifications to a central Polynesian island ecosystem Proceedings of the National Academy of Sciences of the United States of America, 93 Kiwa, A. 17 July 2012. RE: Personel Interview. Knutson, T. R., McBride, J. L., Chan, J., Emanuel, K., Holland, G., Landsea, C., Held, I., Kossin, J. P., Srivastava, A. K. & Sugi, M. 2010. Tropical Cyclones and Climate Change. Nature Geoscience, 3, 157-163. Kumar, D. S. & Lal, R. B. 2012. Effect of Mulching On Crop Production under Rainfed Condition: A Review. International Journal of Research in Chemistry and Environment, 2, 8-20. Lal, M. 2004. Climate change and small island developing countries of the south Pacific. Special issue on sustainable development. Journal of Contemporay Fiji, 2, 15 -31. Lal, M., Singh, K. K., Rathore, L. S., Srinivasan, G. & Saseendran, S. A. 1998. Vulnerability of rice and wheat yields in NW India to future changes in climate. Agricultural and Forest Meteorology, 89, 101-114. Lal, M., Singh, K. K., Srinivasan, G., Rathore, L. S., Naidu, D. & Tripathi, C. N. 1999. Growth and yield responses of soybean in Madhya Pradesh, India to climate variability and change. Agricultural and Forest Meteorology, 93, 53-70. Lal, N. S. 2008. Taro Beetle Management in Papua New Guinea and Fiji: Final Project Report. Noumea Cedex, . Lobell, D. B. & Field, C. B. 2008. Estimation of the carbon dioxide (CO2) fertilization effect using growth rate anomalies of CO2 and crop yields since 194

1961. Global Change Biology, 14, 39-45. Long, S. P., Ainsworth, E. A., Rogers, A. & Ort, D. R. 2004. Rising atmospheric carbon dioxide: plants FACE the future. Annual Review of Plant Biology, 55, 591-628. Magrin, G. O., Travasso, M. I., Diaz, R. A. & Rodriguez, R. O. 1997. Vulnerability of the agricultural systems of Argentina to climate change. Climate Research, 9, 31-36. Magrin, G. O., Travasso, M. I. & Rodrı´guez, G. R. 2005. Changes in Climate and Crop Production During the 20th Century in Argentina. Climatic Change, 72, 229-249. Mamu, E. 16 July 2012. RE: Personal Interview. Manner, H. I. & Taylor, M. 2011. Farm and forestry production and marketing profile for taro (Colocasia esculenta). In: ELEVITCH, C. R. (ed.) Specialty crops for Pacific Island agroforestry. Holualoa, Hawaii: Permanent Agriculture Resources (PAR). Masters, G., Baker, P. & Flood, J. 2010. Climate Change and Agricultural Commodities: CABI Working Paper 2. CABI. Mertz, O., Birch-Thomsen, T., Elberling, B., Rothausen, S., Bech-Bruun, T., Reenberg, A., Fog, B., Egsmose, M. R. & Breuning-Madsen, H. 2011. Changes in shifting cultivation systems on small Pacific islands. The Geographical Journal, 1-13. Mimura, N., Nurse, L., McLean, R., Agard, J., Briguglio, L., Lefale, P., Payet, R. & Sem, G. 2007. Small islands. Climate Change 2007: Impacts, Adaptation and Vulnerability. Contribution of Working Group II to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge, UK. Mirschel, W. & Wenkel, K. O. Year. Modelling soil-crop interactions with AGROSIM model family. In: KERSEBAUM, K. C., HECKER, J. M., MIRSCHEL, W. & WEGHENKEL, M., eds. Modelling water and nutrient dynamics in soil-crop systems. Proceedings of a workshop held in Müncheberg, Germany, 14-16 June, 2007. 59-73. Monirul, M. & Mirza, Q. 2002. Global warming and changes in the probability of occurrence of floods in Bangladesh and implications. Global Environmental Change, 12, 127-138. Mousavi, R. S. & Eskandari, H. 2011. A General Overview on Intercropping and Its Advantages in Sustainable Agriculture. Journal of Applied Environmental and Biological Sciences, 1, 482-486. Mutsvangwa, E. P. 2010. Climate Change and Vulnerability to Food Insecurity Among Smallholder Farmers: A Case Study of Gweru and Lupane Districts in Zimbabwe. University of the Free State. Nearing, M. A., Pruski, F. F. & O'Neal, R. M. 2004. Expected climate change impacts on soil erosion rates: a review. Journal of Soil and Water Conservation, 59, 43-50. Norby, R. J., Sholtis, D. J., Gunderson, A. C. & Jawdy, S. S. 2003. Leaf dynamics of a deciduous forest canopy; no response to elevated CO2. Oecologia, 136, 574-584. PACC 2006. Pacific Adaptation to Climate Change: Solomon Islands Report of in-country Consultations. PACC. Padgham, J. 2009. Agricultural Development under a changing climate: opportunities and challenges for adaptation. Joint Departmental Discussion Paper. Washington. 195

Pala, M., Sto¨ ckle, C. O. & Harris, H. C. 1996. Simulation of durum wheat (Triticum durum) growth under differential water and nitrogen regimes in a mediterranean type of environment using CropSyst. Agricultural Systems, 51, 147-163. Pardales, J. R., Jr & Belmonte, D. V., Jr 1984. Cultural Management Studies on Upland Taro: Effects of Population Density and Planting Method on Growth and Yield. Philippine Journal of Crop Science, 9, 29-32. Parry, M. L., Canziani, F. O. & Palutikof, P. J. 2007. Technical Summary. Climate Change 2007: Impacts, Adaptation and Vulnerability. Contribution of Working Group II to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. In: PARRY, M. L., CANZIANI, F. O., PALUTIKOF, P. J., VAN DER LINDEN, J. P. & HANSON, E. C. (eds.). Cambridge, UK: Cambridge University Press. PCCSP 2011a. Climate Change in the Pacific: Scientific Assessments and New Research Volume 1: Regional Overview. Australian Bureau of Meteorology and Commonwealth Scientific and Industrial Research Organisation (CSIRO). PCCSP 2011b. Chapter 13. Solomon Islands. In Climate Change in the Pacific: Scientific Assessments and New Research Volume 2: Country Reports. Australian Bureau of Meteorology and Commonwealth Scientific and Industrial Research Organisation (CSIRO). Porter, J. R. & Semenov, A. M. 2005. Crop responses to climatic variation. Philosophical Transactions of the Royal Society B: Biological Sciences, 360, 2021-2035. Quity, G. 2012. Assessing the Ecological Impacts of Climate Change on Root Crop Production on High Islands: A case study in Santa Isabel, in the Solomon Islands. The University of the South Pacific. Ranasinghe, S. C., Premasiri, N. D. R. & Silva, S. R. L. 2003. Effects of Mulches and Cover Crops on Water Status and Gas Exchange of Coconut (Cocos nucifera L.) palms in Gravelly Soils. Cocos, 15, 1-11. Rasmussen, K., May, W., Birk, T., Mataki, M., Mertz, O. & Yee, D. 2009. Climate change on three Polynesian outliers in the Solomon Islands: Impacts, vulnerability and adaptation. Geografisk Tidsskrift-Danish Journal of Geography, 109, 1-13. Reenberg, A., Birch-Thomsen, T., Mertz, O., Fog, B. & Christiansen, S. 2008. Adaptation of Human Coping Strategies in a Small Island Society in the SW Pacific—50 Years of Change in the Coupled Human–Environment System on Bellona, Solomon Islands. Human Ecology, 36, 807–819. Robertson, M. J., Carberry, P. S., Huth, N. I., Turpin, J. E., Probert, M. E., Poulton, P. L., Bell, M., Wright, G. C., Yeates, S. J. & Brinsmead, R. B. 2002. Simulation of growth and development of diverse legume species in APSIM. Australian Journal of Agricultural Research, 53, 429-446. Rosenzweig, C. & Tubiello, F. N. 1996. Effects of changes in minimum and maximum temperature on wheat yields in the central US: A simulation study. Agricultural and Forest Meteorology, 80, 215-230. Rosenzweig, C., Tubiello, N. F., Goldberg, A. R., Mills, E. & Bloomfield, J. 2002. Increased crop damage in the US from excess precipitation under climate change. Global Environmental Change 12, 197-202. Seiffert, S., Kaselowsky, J., Jungk, A. & Claassen, N. 1995. Observed and calculated potassium uptake by maize as affected by soil water content and bulk density. 196

Agronomy Journal, 87, 1070–1077. Senbeta, F. A. 2009. Climate Change Impact on Livelihood, Vulnerability and Coping Mechanisms: A Case Study of West-Arsi Zone, Ethiopia. Masters, Lund University SIG 2001. Rennell and Bellona Province Development Profile. Honiara, Solomon Islands. SIG. 2003. Topograhical Map of Bellona. Honiara, Solomon Islands: Ministry of Lands. SIG 2012. Solomon Islands Second National Communication to the United Nations Framework Convention on Climate Change (UNFCCC)-Draft. Solomon Islands Government. Singh, U., Matthews, B. R., Griffin, S. T., Ritchie, T. J., Hunt, A. L. & Goenaga, R. 1998. Modeling growth and development of root and tuber crops. In: TSUJI, Y. G., HOOGENBOOM, G. & THORNTON, K. P. (eds.) Understanding Options for Agricultural Production. Netherlands: Kluwer Academic Publishers. Singh, U., Tsuji, G. Y., Goenaga, R. & Prasad, H. K. Year. Modeling Growth and Development of Taro and Tanier. In: SINGH, U., ed. Proceedings of the workshop on Taro and Tanier Modelling, 1992. SINSO. 2009. Population Map of Bellona by Wards. Honiara, Solomon Islands: Solomon Island National Statistic Office. Stöckle, O. C., Cabelguenne, M. & Debaeke, P. 1997. Comparison of CropSyst performance for water management in Southwestern France using submodels of different levels of complexity. European Journal of Agronomy, 7, 89-98. Stöckle, O. C., Donatelli, M. & Nelson, R. 2003. CropSyst, a cropping systems simulation model. European Journal of Agronomy, 18, 289-307. Stöckle, O. C., Martin, S. & Campbell, G. S. 1994. CropSyst, a cropping systems model: Water/nitrogen budgets and crop yield. Agricultural Systems, 46, 335-359. Taiz, L. & Zeiger, E. (eds.) 2008. Fisiologia Vegetal Artmed: Porto Alegre. Tall, A. 2013. Climate Services: Empowering Farmers to Confront Climate Risks- Lessons from Africa and South Asia. Research Program on Climate Change, Agriculture and Food Security, CGIAR. Talo, F. 2008. Solomon Islands National Adaptation Programmes of Action. Honiara: Ministry of Environment,Meteorology, Climate Change and Disaster Management Tans, P. & Keeling, R. 2013. Recent Monthly Average Mauna Loa CO2 [Online]. NOAA/ESRL & Scripps Institution of Oceanography. Available: www.esrl.noaa.gov/gmd/ccgg/trends/ [Accessed 04/06/13]. Tao, F. & Zhang, Z. 2011. Impacts of climate change as a function of global mean temperature: maize productivity and water use in China. Climatic Change, 105, 409-432. Tesche, M. 1975. Germination of conifer seeds under the conditions of simulated drought stress in polyethylene glycol (Lutrol). Biochemie und Physiologie der Pflanzen (BPP), 176, 577-584. Thaman, R., Puia, T., Tongabaea, W., Namona, A. & Fong, T. 2010. Marine biodiversity and ethnobiodiversity of Bellona (Mungiki) Island, Solomon Islands. Singapore Journal of Tropical Geography, 31, 70-84. Thiruchchelvan, N., Thirukkumaran, G. & Mikunthan, G. 2012. In Vitro Biological Control of Marasmiellus sp. The casual of Stem Rot of Banana Grown in 197

Jaffna Peninsula, Sri Lanka. Academic Journal of Plant Sciences, 5, 94-101. Thomas, J. M. G., Boote, J. K., Allen, H. L. J., Gallo-Meagher, M. & Davis, M. J. 2003. Elevated temperature and carbon dioxide effects on soybean seed composition and transcript abundance. Crop Science, 43, 1548-1557. Tongabaea, W. 16 July 2012. RE: Personal Interview. Tsuji, G. Y. 1998. Network management and information dissemination for agrotechnology transfer. In: TSUJI, G. Y., HOOGENBOOM, G. & THORNTON, P. K. (eds.) Understanding Options for Agricultural Production. The Netherlands: Kluwer Academic Publishers, Dordrecht. Tsuji, G. Y., Uehara, G. & Balas, S. (eds.) 1994. Decision Support System for Agrotechnology Transfer (DSSAT) Version 3, Honolulu, Hawaii University of Hawaii. Tubiello, F. N., Amthor, J. S., Boote, K. J., Donatelli, M., Easterling, W., Fischer, G., Gifford, R. M., Howdenh, M., Reilly, J. & Rosenzweig, C. 2007. Crop response to elevated CO2 and world food supply: A comment on Food for Thought. European Journal of Agronomy, 26, 215-223. Tubiello, F. N. & Ewert, F. 2002. Simulating the effects of elevated CO2 on crops: approaches and applications for climate change. European Journal of Agronomy, 18, 57-74. Turral, H., Burke, J. & Faurès, J. 2011. Climate change, water and food security. FAO Water Reports. Rome, Italy. Utset, A. 2009. Opportunities and Challenges of Using Modelling Tools for Agricultural Decision-Making Under Climate Change Conditions. In: UTSET, A. (ed.) Climate Variability, Modelling Tools and Agricultural Decision-Making. New York: Nova Science Publishers, Inc. von Caemmerer, S. & Furbank, R. T. 2003. The C-4 pathway: an efficient CO2 pump. Photosynthesis Research, 77, 191–207. Wairiu, M., Lal, M. & Iese, V. 2012. Climate Change Implications for Crop Production in Pacific Islands Region. Food Production- Approaches, Challenges and Tasks. In-Tech Publishing. Walkenhorst, T. 2005. Crop diversity and genetic erosion on the South Pacific Island of Mungiki (Bellona)/ Solomon Islands. Masters, University of Kassel Walworth, J. L. 1992. Soil drying and rewetting, or freezing and thawing, affects soil solution composition. Soil Science Society of America Journal, 56, 433-437. White, C. W. & Warin, N. O. 1962. Survey of Phosphate Deposits in the South-West Pacific and Australian Waters-Bellona Island. In: GROVER, C. J., THOMPSON, B. R., COLEMAN, J. P., STANTON, L. R. & BELL, D. J. (eds.) The British Solomon Islands Geological Record. London: Despartment of Geological Surveys, Honiara, British Solomon Islands Protectorate. White, W. C. & Warin, O. N. 1959. The phosphate deposits of Bellona Island. Record 1959/47. Canberra: Australian Bureau Mineral Resources. Witter, S. H. 1995. Food, Climate, and Carbondioxide-The Global Environment and World Food Production, Florida, USA, Lewis Publishers. Wolff, T. 1969. The Fauna of Rennell and Bellona, Solomon Islands. Philosophical Transactions of the Royal Society of London, Series B 235. Xiao, G., Liu, W., Xu, Q., Sun, Z. & Wang, J. 2005. Effects of temperature increase and elevated CO2 concentration, with supplemental irrigation, on the yield of rain-fed spring wheat in a semiarid region of China. Agricultural Water Management, 74, 243-255. Ziska, L. H. 2003. Evaluation of yield loss in field-grown sorghum from a c3 and c4 198

weed as a function of increasing atmospheric carbon dioxide. Weed Science, 51, 914-918. Ziska, L. H. & George, K. 2004. Rising carbon dioxide and invasive, noxious plants: potential threats and consequences. World Resource Review, 16, 427-447.

199

APPENDIX A: DESCRIPTIONS OF BELLONESE INDIGENOUS SOILS

The following Indigenous soil classifications and descriptions of the four main soils that are used for food crops cultivation in Bellona were adapted from Breuning-Madsen et al., (2010).

Soil Name Description Kenge toaha Firm soil that has topsoil of dark brown which developed in the phosphate clay deposits. It is clayey, stiff and harder to dig, but well structured and containing less strong pellets. It has a good internal drainage. It is often a deep soil resting on reddish brown subsoil (Kenge mea). Kenge ungi Soft topsoil developed in the phosphate clay deposits, friable with medium clay content, and well structured with many strong pellets of about diameter 1–2 mm. Thus, drainage is excessive. It mainly rests on tanahu or phosphate rock. Malanga Soil mainly described as Oolitic coarse sand. The oolites are rounded, strongly cemented phosphate particles. This soil is generally a black shallow layer not more than 30 cm that rests directly on phosphate rock or coral limestone. Soil which is referred to as any mixed soil. Most commonly being a Hingo hingo mix of Kenge and Malanga soil layer with some oolites. The origin of such soil mixture is weathering of the Malanga where some of the oolites in top layer are reduced in particle size. Another form of Hingo hingo is a mixture of Kenge mea (reddish brown soil) with either Kenge ungi or toaha.

200

APPENDIX B: COMMON CROPS CULTIVATED IN BELLONA

Common crop name Scientific name Varieties (Local names) Uhi Langi, Uhi Sungu, Singasinga,Uhi Kemvel, Greater yam Dioscorea alata Uhi Japana & Uhi Trimanisi Lesser yam or pana Dioscorea esculenta Uhingaba Kumara, Uhingaba Lapani Taro Colocasia esculenta Tango Sua, Tango Neka, Tango Tulagi, Tango Too Potato yam Dioscorea bulbifera Ghope Pacific yam Dioscorea nummularia Uhi , Uhi Vanuatu, Uhi Honiara Sweet potatoes Ipomea batatas Patito Gisi, Patito Trimanisi, Patito Drusilla. Cassava Manihot esculenta Lioka Fiji, plus 2 other types with white and yellow tubers (names unknown) Banana Musa spp Huti Puga, Huti Sungu, Huti Ungi, Huti Mai Moana, Huti Mongi Slippery cabbage Hibiscus manihot/ Unknown but have various Abelmoschus manihot types that have white & red stalks Melon Citrullus lanatus Unknown name Pumpkin Cucurbita pepo Unknown name Corn Zea mays Unknown name

201

APPENDIX C: HOUSEHOLD SURVEY QUESTIONNAIRE

202

203

204

205

206

207

208

209

APPENDIX D: FOCUS GROUP & KEY INFORMANT INTERVIEW GUIDE

210

211

212

213

214

215

216

APPENDIX E: DATA SHEET FOR COLLECTING CROP MANAGEMENT DATA

217

APPENDIX F: DEFINITION OF GENETIC COEFFICIENTS

Appended below are definition of genetic coefficients codes used in calibration of the three crops; Tango Sua, Lioka B1, and Koni. The codes and definition are directly sourced from DSSAT version 4.5 program software (Hoogenboom et al., 2010b).

Coefficients Definition

Cassava

LA1S…………… Area/leaf (cm2) of the first leaves when growing without stress. LAWS………….. Leaf area/weight ratio when crop growing without stress (cm2/g) LLIFA ………….. Leaf life from full expansion to start senescence (phyllocrons) PHINT…………. Interval between leaf tip appearances for first leaves SRFR…………... Storage root fraction of assimilate used for non-root growth

Corn P1………………. Thermal time from seedling emergence to the end of the juvenile phase (expressed in degree days above a base temperature of 8oC) during which the plant is not responsive to changes in photoperiod. P5………………. Thermal time from silking to physiological maturity (expressed in degree days above a base temperature of 8oC). G2……………… Maximum possible number of kernels per plant. G3……………… Kernel filling rate during the linear grain filling stage and under optimum conditions (mg/day).

Taro P4………………. Rapid vegetative growth P5………………. Cormel and corm growth G2……………… A growth partitioning coefficient affecting petiole growth

218

APPENDIX G: LIST OF CLIMATE MODELS USED IN PCCSP PROJECTION

List of Climate models used in PCCSP projection (Source: (PCCSP, 2011a)

219