Sensitivity of rainfed cotton to the changes in climatic factors and CO 2 concentrations

The continuing rise in atmospheric carbon dioxide concentration is one of the main causes for global warming and a resultant change in the climate. Global average air temperature is expected to be 1.8-4.0 °C above the current level in the 21 st century that might impact agriculture dramatically in terms of larger variability in food, feed and fiber production (Aggarwal, 2008). Temperature is the main factor affecting plant growth and development and higher air temperatures certain level significantly affect crop productivity (Sacks and Kucharik, the climate model prediction indicates frequent flood and drought-like in the future. High-temperature compounding with the changes in rainfall patterns will adversely affect crop and

The continuing rise in atmospheric carbon dioxide concentration is one of the main causes for global warming and a resultant change in the climate. Global average air temperature is expected to be 1.8-4.0 °C above the current level in the 21 st century that might impact agriculture dramatically in terms of larger variability in food, feed and fiber production (Aggarwal, 2008). Temperature is the main factor affecting plant growth and development and higher air temperatures above a certain level significantly affect crop productivity (Sacks and Kucharik, 2011). Most of the climate model prediction indicates frequent flood and drought-like situations in the future. High-temperature compounding with the changes in rainfall patterns will adversely affect crop production. It is an important aspect to be understood that the relationship between rainfall and crop productivity as the rainfall directly influencing the growth and productivity of crops. Crops grown under rainfed condition are already finding difficult to cope with temperature and rainfall changes induced by the climate change. Cotton is an agro-industrial crop grown in developing and developed countries including China, United States, India, Pakistan and Brazil. It contributes more than fifty percent of all fibers used for clothing and household furnishings. Changes in climate is likely to impact cotton growth and productivity due to increasing CO 2 concentration, temperature and changes in rainfall behaviours (Bange, 2007). Effect of climate change on productivity of crops has been largly studied by employing crop models coupling with climate change scenarios (Tao et al., 2008). For adapting the crop systems in view of climate change, it is necessary to know the effect of climate change on crop production. Quantifying the effect of various climatic factors on cotton paves the way for developing suitable adaptation strategies and thereby minimize the deleterious effect of varying climatic factors. The present investigation aimed to assess the response of cotton to the changes in temperature, rainfall and CO 2 concentrations over Tamil Nadu.

CROPGRO-Cotton model
The CROPGRO-Cotton model of DSSAT (V. 4.7), designed by International Benchmark Sites Network for Agrotechnology Transfer (IBSNAT) was used for assessing the sensitivity of rainfed cotton to the changes in climatic factors and CO 2 concentrations.

Estimation of genetic coefficients in CROPGRO-Cotton (DSSAT) model
In the CROPGRO-Cotton model, parameterization was done for cotton cultivar SVPR4 using the weather data, soil information, crop management details and biometric observations pertain to the experimental field located at the Agricultural College and Research Institute, Madurai. Data obtained from the experiments conducted in two years with two dates of sowing (21 st September and 6 th October) at two nitrogen levels (N80 and N60) were used in the crop simulation model for deriving the genetic co-efficient of cotton cultivar SVPR4.

Genetic coefficients
In the crop simulation model, traits of the plants and physiological features are included in the form of genotype coefficientsto to describe the physiological processes such as plant growth and development, photosynthesis, yield and yield attributes for individual crop varieties. The genetic coefficients in the crop simulation models that are linked with the development stages and yield of the crop were derived by iteration, through manipulating the related coefficients to simulate the numbers of days for the occurrence of different phenological events and yield closely matching to the   Time required for cultivar to reach final pod load under optimal (PODUR) -9.0; The maximum ratio of (seed/ (seed+shell)) at maturity (THRSH) -70.0; Fraction protein in seeds (g(protein)/g(seed)) SDPRO-0.153; Fraction oil in seeds (g(oil)/g(seed)) SDLIP-0.120

Evaluation statistics of CROPGRO-Cotton model
The reability of CROPGRO-Cotton model simulations was tested using the different statistical indexes such as i. root mean square error (RMSE), ii. coefficient of determination (r 2 ), iii. index of agreement (d).

Sensitivity analysis
Sensitivity of cotton for the changes in temperature, rainfall and carbon dioxide was assessed with 19 combinations as detailed below i. 5 levels of CO 2 concentrations (360, 450, 540, 630 and 720 ppm), ii. 6 levels of maximum and minimum temperature changes (-2°C, 0°C, 2°C, 4°C, 6°C and 8°C) iii. 8 levels of change in rainfall quantity (25, 50, 75, 100, 125, 150, 175 and 200 %) The simulations were made using the CROPGRO-Cotton model embedded in DSSAT under the 39 years (1981-2019) baseline period and averaged the yield for each level of temperature, rainfall and CO 2 concentration considered in the study.

Statistics on the prediction efficiency of CROPGRO-Cotton model
The performance statistics of the CROPGRO-Cotton model for SVPR4 showed the r 2 values of 0.80, 0.83 and 0.89 for the days taken to anthesis, days to atttain the physiological maturity and yield respectively signifying a good match between actual and modelsimulated data. RMSE values of the period taken for anthesis, maturity and seed cotton yield are found to be 3 days, 2 days and 169 kg ha -1 indicating a good match between simulated and observed values. The model statistics indicated that the d values are 0.86, 0.88 and 0.93 for days taken to anthesis, days taken to maturity and seed cotton yield respectively which indicate good harmony between observed and model-simulated data ( Table 1). The high r 2 and d values (>0.8) indicated the close agreement between observed and model-simulated yield and RMSE was in the acceptable range indicating the satisfactory performance of the model (Bhuvaneswari et al., 2014;Raja et al., 2018).

Influence of CO 2 enrichment on Cotton
DSSAT simulation indicated that cotton crops showed a positive response for the enriched CO 2 concentrations and the magnitude of increase in cotton productivity was higher with an increase in CO 2 concentration. It was found out that cotton yield increased by 8.7, 15.9, 21.1 and 25.4 percent from the current yield (1495 kg ha -1 ) with the CO 2 concentration of 450, 540, 630, 720 ppm respectively (Figure 1). Elevated atmospheric CO 2 could create favourable conditions for the growth and development of cotton to a certain extent.

Response of cotton to the changes in temperature
The cotton yield increased up to 4 °C increase in temperature and a further rise in temperature reduced the yield of cotton. Cotton yield was increased by 12.3 and 8.1 percent with 2 and 4 °C increase in temperature respectively. In contrast, for 6 °C rise in temperature, the cotton yield declined by 15.5 percent and yield reduction was 34.1 percent with 8 °C increase in temperature ( Figure  2). The results are similar to the findings of Hebbar et al. (2013) who stated that productivity remain to be same or increase in central and southern India in future climatic conditions.

Effect of changes in rainfall on cotton
It could be seen from the DSSAT model simulation that the yield was less when the cotton crop receives 25 percent of normal rainfall and showed a gradual increase in yield up to 150 percent of rainfall (Figure 3). Rainfed cotton crop has responded up to an additional 50 percent increase in rainfall from the current rainfall quantity (100%). The cotton yield increased by 4.9 and 7.2 percent with 25 and 50 percent increase in rainfall respectively. Vol. 23,No. 3 The results of the rainfall effect on cotton are in harmony with the study of Cetin Oner and Basbag Sema. (2010).
Enhanced CO 2 concentrations influenced the cotton yield positively. The cotton crop had a beneficial effect owing to increased temperature up to 4 °C and thereafter yield got declined. Rainfall increases up to 50 percent also enhances cotton productivity. It clearly indicated that cotton could be one of the best alternative crops under changing climatic conditions in the future.

ACKNOWLEDGMENTS
The authors would like to thank MoEF& CC and DST-CCP-SPLICE, Govt of India, New Delhi for the financial support to conduct this study. The authors are also thankful to AgMIP team for rendering the technical input for taking up this study.