Russell T. Nagata

Heat units to predict tomato harvest in the southeast USA

Abstract Planting and first harvest dates of tomato ( Lycopersicon esculentum Mill.) from 2 seasons in 3 years at eight locations in Georgia, North Carolina and South Carolina formed 38 environments which were used to determine the most reliable method to predict fast harvest date of tomato based on daily maximum and minimum air temperature. Eleven methods of calculating heat units were chosen for comparison based on their performance as described in the literature. The most reliable method was defined as the one with the smallest coefficient of variation (CV). CVs were calculated for each method over both seasons and locations, for each season over all locations, each location over all seasons, and for each season at each location. All heat unit summation methods had smaller coefficients of variation (CV) than the standard method of counting days from planting to first harvest. Heat unit summation methods improved harvest date prediction accuracy compared with the counting day method for tomatoes in the South Atlantic Coast (SAC) region. Prediction using location/season specific models were less variable than the models over all seasons and locations. Incorporating daylength improved model prediction accuracy when applied over all locations and seasons, all locations by season, and all seasons by location. Based on the results of this study, the heat unit summation technique recommended for this region (where the location and season specific models are not available) is the reduced ceiling method multiplied by daylength.

Heat units, solar radiation and daylength as pepper harvest predictors

Abstract Daily maximum and minimum air temperature, total solar radiation and daylength data from seven locations during three seasons of 3 years were used to compare 52 heat unit accumulation models with counting days as a harvest prediction method for pepper. The best model was defined as the one with the least variation, i.e. smallest coefficient of variation (CV). CVs were calculated for each method over all seasons and locations, for each method over all locations for each season, and for each method in each season at each location. In all cases heat unit accumulation methods were better than counting days. The location and season specific model was the most accurate, but the analysis over all seasons and locations did result in smaller CVs than counting days, so improved harvest prediction can be achieved by using regionally developed models.