| Abstract | Bangladesh is frequently cited as one of the countries most vulnerable to climate change, despite the country’s insignificant contribution to climate change. Crop production, especially rice, the main food staple, is the most susceptible to climate change and variability. Any changes in climate will, thus, increase uncertainty regarding rice production as climate is major cause of year-to-year variability in rice productivity. This thesis is motivated partly by the susceptibility of rice farming to climate change and partly by the limited studies of Bangladesh on this topic. The overall aim of this thesis is, thus, to analyse the impact of climate change on rice production at three levels (aggregate-national, disaggregated-climate zone and micro-farm level), and to evaluate the adaptation strategies practised by farmers in a severely drought-prone area.
At the aggregate level, this thesis first investigated national data from secondary sources to examine changes in maximum temperature, minimum temperature and rainfall over the past 60 years. Results from a linear trend model reveal that the time trend is statistically significant for all three major climate variables. This implies climate has changed over the whole period. However, the findings from quantile regression indicate that the explanatory power of the time trend is higher in the higher quantiles than the lower quantiles for all three climate variables. This latter method thus offers a more complete picture of the changing climate at different points of time. Given these changes in climate and using production function theory, an evaluation of the impacts of changing climate on the yields for three rice crops in Bangladesh: Aus, Aman and Boro, was made. The findings confirm that the changes to both maximum and minimum temperatures are statistically significant for Aus and Boro rice. However, changes to the average minimum temperature are found to affect Aus rice production adversely and the average maximum temperature is also negatively related to Boro rice yield. On the contrary, the impacts of maximum temperature and rainfall are more pronounced for Aman rice compared to minimum temperature whose effects are adverse. Given these adverse effects of temperature on rice crops, policy makers should design strategies for the development and use of temperature tolerant rice varieties. However, this analysis of national level data is unable to reveal regional level differences in climate and their differential impacts on rice yield which warrants disaggregated level analysis.
Under the theoretical framework of Just-Pope stochastic production function, the objective of the disaggregated level analysis was to assess the effects of climate change on the yield and variability of Aus, Aman and Boro rice using cross-sectional time series (panel) data. The results reveal that maximum temperature is risk increasing for Aus and Aman rice while it is risk decreasing for Boro rice yield. Minimum temperature is risk increasing for Boro rice and risk decreasing for the Aus and Aman varieties. Finally, rainfall is risk increasing for Aman rice whilst risk decreasing for Aus and Boro rice. Moreover, future climate change is expected to increase the variability of rice yield for all three rice crops. Disaggregated level analysis, thus, provided more information than aggregate level analysis. However, the disaggregated level data is unable to show how individual farmers are affected by climate change which necessities a farm level analysis of impact and adaptation.
The farm level analyses employed data from a survey of 550 farm households in a severely drought-prone area of Bangladesh. Descriptive statistics reveal that net revenue and production loss from Aman rice vary between different subsamples of farmers. For example, mean profit was significantly higher for large and medium farmers compared to small and landless farmers while the latter group of farmers faced higher mean production losses. Integrated farms have higher net revenue compared to rice only farms. Moreover, production losses for highly irrigated farms are lower than for less irrigated farms. Further, results from both mean and median regression on the determinants of profit and production loss indicate significant variables that can be targeted to increase profit or decrease production loss. These include age, years of schooling of household head, household yearly total income, household assets, land tenure, access to agricultural extension services, weather information, electricity and subsidy, percentage of land under irrigation, crop selling at local market, and distance to local or nearby urban market. Government policy initiatives should include support for integrated farming, increasing the provision of education, providing regular weather forecasts, giving subsidies to verysmall and landless farmers, distributing government owned fallow lands to small farmers, and adopting water saving irrigation technologies.
Farmers have taken some adaptation strategies to reduce these adverse effects on rice production. The major adaptation strategies include higher levels of irrigation, cultivation of short-duration rice varieties, changing planting dates, agro forestry, use of different crop varieties and cultivation of non rice crops. Estimates from a multinomial logit model specify that age, gender and education level of household head, household annual total income, household assets, farm size, tenure status, farming experience, access to agricultural credit, availability of subsidies, electricity at home, and farmer-to-farmer extension services all affect adaptation choices. Therefore, policy makers should target these determinants to boost farmers’ adaptation and thereby diminish the adverse effects of climate change.
The analytical framework used in this study has produced robust results. It should be replicated in other developing countries experiencing adverse climate change and having similar characteristics to Bangladesh. |
|---|