LONG-TERM ANALYSIS AND MODELING OF AGROCLIMATIC EFFECTS ON NATIONAL AVOCADO YIELDS IN ISRAEL

The agroclimatic analysis was based on a functional relationship betwe en significant selected climatic parameters and avocado yields in Isra el over a 27 year period. The analysis consisted of three steps. 1. Mu ltiple regressions between every climatic elements and national yields . 2. Multiple regression between the statistically significant climati c variables only and national yields. 3. The model yield output was te sted by cross-validation on 21 frost free years and 6 years with frost . The statistically significant climatic variables explained 72% of th e interannual variability of yield for the frost free series and 83% f or the series with frost. Calculated avocado yields showed good agreem ent with measured values. The standard error of estimate was 1.1 t ha- 1 for years without frost, and 1.4 t ha-1 for years with frost. The mo del provides quantitative yield responses to microclimatic modificatio n and national yield estimates 3 months before harvesting, providing l ead time for management and commercial decisions in the avocado indust ry.