棉 花 学 报 Cotton Science 2008，20(2)：116-122
Study on the Key Meteorologic Factors-Based Model for Main Characters of Cotton Fiber Quality Abstract：Cotton fiber length, strength and micronaire are the main quality characters determined by cultivar characteristics, environment and cultivation, and could be used to making sure cotton fiber quality. The objective of this research was to determine the effects of meteorology on fiber properties of cotton and found key meteorologic factors-based model of fiber length, strength and micronaire. Experiments were random block design with three replications with different cultivars and conducted in different places. Data was collected from 1994 — 1995 in Shanxi and Shandong province which use three cultivars ， and from 2001 — 2002 in 37 places including cotton variety test region of the Yangtze river, Shanxi and Shandong province which involves 20 cultivars. All experiments are the same in treatments so that experimental errors of soil and environment can be reduced. Based on known research achievement in relationship of cotton fiber quality and cultivars and meteorologic factors, by analyzing experimental data above, the key meteorology factors which significantly affected fiber length, strength, micronaire was determined. The results showed that fiber length was significantly affected by daily average minimum temperature during bolling stage, and so as fiber strength was daily average maximum temperature and relative humidity, and fiber micronaire was daily night temperature and daily average precipitation, respectively. Then, key meteorologic factors-based models of fiber length, strength and micronaire were set up, respectively. These non-liner models calculate both quality of cultivar variables (fiber length, strength, micronaire) and environmental variables (temperature, solar radiation, relative humidity). The RMSE (Root Mean Square Error) of key meteorologic factors-based models of fiber length, strength and micronaire were 0.985 mm, 1.003 cN·tex-1 and 0.233, respectively. Meanwhile, the 1：1 figs of simulated and observed fit well. All results showed that the models could make good prediction. Furthermore, compared to existing models, these models was proved the best in not only emphasizing conclusive effects of cultivar characters but also emboding influence of meteorologic factors on the main cotton fiber quality characters. Finally, the conceptual limits of the models are discussed.
|