Name: Mariam Zachariah
Name of supervisor: Prof. Arpita Mondal
Department: Civil Engineering
The topic of research: Rainfall deficits and cropping choice drive crop loss in Marathwada,
Description of research work:
Rainfall deficits and cropping choice drive crop loss in Marathwada, India.
Understanding the role of climate variability in driving crop loss is an important research area, particularly for the agrarian economy of India. Among other social, economic and political factors, crop loss also contributes to ensuing crises such as bankruptcy, famine and suicides. The wide variability in agroclimatic features over the country warrant studies that focus on small target areas, for discerning crop sensitivities to the regional climate factors. To this end, we quantified the effects of growing season rainfall and temperature on selected crops grown in Marathwada, in the state of Maharashtra. This region is a microcosm of India’s agrarian culture with the community engaged primarily in agriculture and allied activities. Situated in the rain shadow region of the Western Ghats, Marathwada is also highly prone to droughts, and a resultant hotspot of agrarian crisis.
On observing the acreages of the crops grown in the region during the rain-fed Kharif season, from 1997-2014, we found a gradual shift from food crops such as sorghum and pearl-millet, to cash crops such as sugarcane and soybean (Fig.1a). In order to understand whether this observed shift is reconcilable with the native climate conditions, we disaggregated and compared the sensitivities of these crops to the growing season rainfalls and temperature. In the first part of the study, we used Mixed-Effects Linear Regression (MELR) models that express crop yields as a function of rainfall, temperature and other derived variables such as soil moisture and irrigation, while also accounting for unobserved soil characteristics and improvements in cropping practices via. intercepts in space and time. The sensitivities were then, inferred from the regression coefficients (Fig.1b). In the next part, these findings were verified using a process-based model, namely the Decision Support System for Agrotechnology Transfer (DSSAT v4.7), from customized experiments that allow either rainfall or temperature to evolve as observed, while all other meteorological inputs are kept constant at their respective long-term daily means. The sensitivities, in this case, were inferred from the coefficient of variation in yields from these experiments (Fig.1c). Our results showed that the cash crops that are increasingly cultivated in recent years are water intensive, while the declining food crops are not sensitive to rainfall or temperature.
The daily average temperature distributions for the various crop development stages under future warming scenarios of 1.5 and 2oC above pre-industrial levels, were also found to lie within the optimum temperature ranges for these crops, suggesting that temperatures are expected to have minimal effect on crop yields, even under future climate change. This led us to conclude that rainfall deficits, rather than temperature is likely to be the primary driver of crop loss in the region, and seemingly aggravated by the shift from the native drought-resilient food crops to rainfall-intensive cash crops. In this regard, strategies for promoting and sustaining the cultivation of drought resistant food crops seems to be one of the potential ways forward, for the crisis-ridden Marathwada region. Such measures will be useful for addressing both immediate farmer crisis in the region, and threats to food security in the long term. The simplicity of the methodology proposed in this work allows for similar studies focusing on other regions, and including other relevant drivers of crop loss, depending on the region. This study is published in Environmental Research Letters, and is available for download at https://iopscience.iop.org/article/10.1088/1748-9326/ab93fc.
Fig.1 (a) Area under cultivation of the important crops grown in Marathwada during the Kharif season, from 1997-2014. (b) The regression coefficients for rainfall and temperature of selected crops, from standardized MELR models. (b)The coefficient of variation in yields of these crops due to changes in rainfall and temperature from DSSAT model.