A method to predict weekly strawberry fruit yields from extended season production systems
In major strawberry (Fragaria x ananassa) production areas, fruit are harvested continuously for 4 to 6 mo. During the season, weekly yields vary. To improve weekly forecasts, a yield prediction equation was developed for ‘Strawberry Festival’ using input variables derived from flower counts and temperature data over two seasons in Florida. Weekly yields are dependent on the number, size, and quality of ripe fruit. Variation in the mean fruit number per plant for a 1-wk interval was explained by the equation 1.502 x adjFC + 0.375 x adjFC x (TFDP – TFDP-Harvest week) with good accuracy (rcent2=0.93), where adjFC = the mean number of flowers per plant counted during a prior 1-wk interval determined from temperature data and (TFDP – TFDP-Harvest week) is a measure of the temperature trend during the harvest week. The size of fruit and percentage of fruit culled due to poor quality could be predicted from flower count trends, but correlations between actual and predicted values were low (radj2=0.48 for fruit weight and radj2=0.18 for fruit culled due to size and shape). No input variables were identified to predict disease or damage. The final yield equation, mean yield (g plant–1) = 19.624 x adjFC + 5.343 x adjFC x (TFDP – TFDP-Harvest week), included only the input variables used to predict fruit number. The coefficient of determination (rcent2) for a regression of actual yields on yields estimated using this equation was 0.89. Based on historic temperatures, an expected value for (TFDP – TFDP-Harvest week) and flower counts from an expected flowering interval were also determined for 1-wk harvests. The coefficient of determination was 0.84 for yield estimates based on these values.