Maize and climate change in Sri Lanka: progress, trends, and challenges in simulating impacts

摘要

An increase in atmospheric temperature and carbon dioxide concentration, and altered precipitation under climate change are likely to have impacts on agricultural productivity. There are comparable changes in temperature and precipitation in Sri Lankawith the global trends. This paper presents observed relationships between yield and climate and yield predictions under climate change scenarios in Sri Lanka using maize as an exemplar crop that is classified as a coarse cereal. The study was conductedat the Walker Institute for Climate System Research, University of Reading, UK. The recent changes in climate were determined by analysing long term (from 1990 to 2010) daily air temperature and rainfall from six selected maize growing districts. Yieldresponses were statistically analysed by simple linear regression models, describing the relationship between detrended yield, rainfall, and air temperature. GLAM-the general large-area model for annual crops which is sufficiently process-based was usedto simulate the maize yield for baseline and for 5 Global Climate Models (CCSM4, GFDL-ESM2M, HadGEM2-ES, MIR0C5, MPI-ESM-MR) of RCP8.5 scenario for mid-century. Significant (p<0.10) correlations of detrended yield for seasonal mean temperature (Moneragala, r=-0.46; Badulla, r=0.38) and seasonal total rainfall were observed only for some districts (Anuradhapura, r=0.44 to 0.52; Hambantota, r=0.49; Moneragala, r=0.35 to 0.36). Simulated yield trends correlated well (N-S, 0.49-0.65; RMSE, 144-201 kg ha1) with observed values. The GCMs projected a decrease in yield caused by a shorter crop growing season due to higher temperatures and a precipitation deficit. Overall observed crop-climate correlations were exploited within GLAM for simulating maize yield for future climates and suitable adaptation strategies.