Predicting cotton boll maturation period using degree days and other climatic factors

Degree days are often used for cotton (Gossypium hirsutum L.) growth monitoring and management. The objectives of this research are to determine if 15.5°C is an accurate lower‐threshold temperature to monitor the boll maturation period (BMAP) for cotton in the northern, rainfed region of the U.S. Cotton Belt, to investigate other climatic factors in this cotton region that may improve the accuracy of the current degree day system for cotton, and to evaluate degree day models that include both an upper‐ and lower‐threshold temperature. Cotton was planted at three different timings in 2001 and 2002 to provide different climatic regimes during the BMAP. On 10 typical plants per plot, all first‐position flowers were individually tagged with date of flower opening and were then harvested at full maturity. Daily weather data consisted of maximum, minimum, and average air temperature; maximum and average soil temperature; average soil moisture; maximum and average solar radiation; and maximum and average photosynthetically active radiation. The 17°C degree day model, which used 17°C as the lower threshold, provided the best adjusted r2 (0.2715) of all the single‐variable models; the degree day 15.5°C model had an adjusted r2 of 0.2276. The best model using both upper and lower temperature thresholds was DD3017, using 30 and 17°C as the thresholds, and had an adjusted r2 of 0.2452. Adding average, minimum, and maximum air temperatures to the DD15.5, DD17, and DD3017 models reduced coefficient of variation and mean square error and increased adjusted r2 values.