Impact And Adaptation Of Climate Change On Cotton Phenology, Yield And Fiber Quality In Xinjiang

Xinjiang is the largest cotton production area in China, which cotton yield and fiber quality is beneficial to national economy and the people’s livelihood. Numerous studies have shown that climate change has had and continue to has significant impact on crop production. Thus, it’s very important to study on impact and adaptation of climate change on cotton phenology, yield and fiber quality in Xinjiang. Based on the crop model associated with the climate change scenario produced by the UK Hadley Centre regional climate model-PRECIS, the cotton production was simulated in Xinjiang under A2and B2emission scenarios. The impact of climate change on cotton production was analyzed based on the adaptation measures such as adjusting sowing date, breeding the new cotton variety. And provides a basis for the management of cotton production by yield prediction. Finally, a digital platform of cotton production based on CottonXL which was developed according to the principles of FSPM (Functional-Structural Plant Model) was developed by using software engineering method and database technology. The digital platform will be an useful tool for cotton production responses to climate and modern agriculture informatization, intelligentialize, mechanization planting. The main conclusions are as following:(1) If keeping other managements unchanged, under warmer climate, the rate of growth and development of cotton was increased and each phenology (emergence, squaring, flowering and boll opening) became earlier, but the date of stop growing delayed. Under A2and B2emission scenarios, as the sowing date delay each phenology became later. And delay of sowing date significantly increased the rate of vegetative growth, but have no significant effect on reproductive growth.(2) If keeping other managements unchanged, the whole growing period (from sowing to die) of cotton would prolong under warmer climate, which may be helpful to dry matter accumulation.The risk of cultivating cotton would reduce with the decrease of variation coefficient of cotton yield. Under A2and B2emission scenarios, delay of sowing date did not significantly change lint yield and the percentage of pre-frost lint in Shihezi. However, in Shache, delay of sowing date decreased lint yield and the percentage of pre-frost lint. To ensure high yields, varieties with longer growing period could be planted as an adaptation measure to warming climate in Shihezi and varieties with shorter growing period could be planted as an adaptation measure to warming climate in Shache.(3) Sowing date did not significantly change fibre quality in the Yellow and Yangtze River regions. However, in Xinjiang, delay of sowing date decreased fibre quality because low temperatures at the end of growing period could prevent bolls from reaching full maturity. The times of main stem and branch topping determine the final number of fruits on a plant. Early topping practice resulted in all the fruits of high quality. As the topping time was delayed, the fibre quality decreased significantly in Xinjiang region and Yellow River region, while the fibre quality was slightly affected in Yangtze River region. There is no significant effect of fill mulching on fibre quality in three regions.(4) The dynamic prediction model of cotton yield was builded based on COSIM model, which could accurately predict phenology and yield of cotton day by day. For a single site, the deviations of prediction result of emergence, squaring, flowering and boll opening date was less than5days, and the accuracy of yield prediction was above80%. The accuracy of regional yield prediction was above90%.(5) A digital platform of cotton production based on CottonXL which was developed according to the principles of FSPM (Functional-Structural Plant Model) was developed by using software engineering method and database technology. By inputting the parameters of climate and production condition, this digital platform realized the functions of plant structure selection, optimizing manage measure, yield prediction, teaching presentation and so on. The tests showed that the digital platform was user-friendly and stable, which could improve the capacity of the crop model in yield prediction and meteorological disaster assessment.