A temperature index model to estimate long-term freeze-risk of satsuma mandarins grown on the northern coast of the Gulf of Mexico

Estimates of long-term freeze-risk aid decisions regarding crop, cultivar, and rootstock selection, cultural management practices that promote cold hardiness, and methods of freeze protection. Citrus cold hardiness is mostly a function of air temperature, but historical weather records typically contain only daily maximum (T(max)) and minimum (T(min)) air temperatures. A mathematical model was developed that used T(max) and T(min) to estimate air temperature every hour during the diurnal cycle; a cold-hardiness index (CHI500) was calculated by summing the hours less than or equal to 10 degrees C for the 500 h before each day; and the CHI500 was regressed against critical temperatures (T(c)) that cause injury. The CHI500 was calculated from a weather station located within 0.1 km of an experimental grove and in the middle of the satsuma mandarin (Citrus unshiu Marc.) industry in southern Alabama. Calculation of CHI500 was verified by regressing a predicted CHI500 using T(max) and T(min), to a measured CHI500 calculated using air temperatures measured every hour for 4 winter seasons (1999-2003). Predicted CHI500 was linearly related to measured CHI500 (r2 = 0.982). However, the slope was a little low such that trees with a CHI500 = 400, near the maximum cold-hardiness level achieved in this study, had predicted T(c) that was 0.5 degrees C lower than measured T(c). Predicted and measured T(c) were similar for nonhardened trees (CHI500 = 0). The ability of predicted T(c) to estimate freeze injury was determined in 18 winter seasons where freeze injury was recorded. During injurious freeze events, predicted T(c) was higher than T(min) except for a freeze on 8 Mar. 1996. In some freezes where the difference in T(c) and T(min) was <0.5 degrees C there were no visible injury symptoms. Injury by the freeze on 8 Mar. 1996 was due, in part, to abnormally rapid deacclimation because of defoliation by an earlier freeze on 4-6 Feb. the same year. A freeze rating scale was developed that related the difference in T(c) and T(min) to the extent of injury. Severe freezes were characterized by tree death (T(c) - T(min) > 3.0 degrees C), moderate freezes by foliage kill and some stem dieback (1.0 degrees C less than or equal to T(c) - T(min) less than or equal to 3.0 degrees C), and slight freezes by slight to no visible leaf injury (T(c) - T(min) < 1.0 degrees C). The model was applied to T(max) and T(min) recorded daily from 1948 through 2004 to estimate long-term freeze-risk for economically damaging freezes (severe and moderate freeze ratings). Economically damaging freezes occurred 1 out of 4 years in the 56-year study, although 8 of the 14 freeze years occurred in two clusters, the first 5 years in the 1960s and 1980s. Potential modification of freeze-risk using within-tree microsprinkler irrigation and more cold-hardy cultivars was discussed.