Cogent Food & Agriculture (Jan 2019)
Modelling the impact of agrometeorological variables on regional tea yield variability in South Indian tea-growing regions: 1981-2015
Abstract
As tea (Camellia sinensis L.) yield strongly determined by local environmental conditions, thus assessing the potential impact of the seasonal and inter-annual climate variability on regional crop yield has become crucial. The present study assessed the region-level tea yield variability at different temporal scales utilising observed climate data for the period 1981–2015, to understand how the climate variability influences tea yields across the South Indian Tea Growing Regions (SITR)? Using statistical models, step-wise multiple regression (SMLR), seasonal autoregressive integrated moving average (SARIMAX), artificial neural network (ANN) and vector autoregressive model (VAR), the relations between meteorological factors and crop yield variability was measured. The higher explaining ability of ANN and VAR models over SMLR and SARIMAX shows that the multivariate time series models are better suited for capturing the nonlinear short-term fluctuations and long-term variations. The analysis showed considerable spatial variation in the relative contributions of different climate factors to the variance of historical tea yield from 3 to 95%. Climate variability explained ~84.8% of the annual tea yield variability of 1.9 t ha−1 y−1, over 106.85 thousand ha translates into an annual variation of ~0.02 million ton in tea production over the study area. Among the climatic factors, temperature variability identified to be the most serious factor determining the tea yield uncertainty than rainfall variability in South India (SI). Hence, the study recommends the policymakers to develop imperative regional specific adaptation strategies and effective management practices (for temperature related issues) to reduce the negative impact of climate change on crop yields.
Keywords
