Author : Saludes, Ronald B.

Summary : With the assumption that other external factors are non-limiting, coconut yield variations between and within years can be explained largely by weather. Using multiple linear regression analysis, coconut yield models for the selected Mindanao provinces were developed to forecast semestral production. Annual yield forecast can be obtained by simply adding the forecasts of the two semesters of the same year. These models relate total semestral coconut production to monthly water satisfaction indices of the critical months of individual bunches that matured during a particular semester. On the other hand, water satisfaction indices were derived using water balance analysis. Large deviations between actual and predicted values of coconut production were attributed largely to the cumulative effect of other external factors particularly the change in the number of coconut bearing trees overtime, uncertainty in climate representation of some synoptic stations, and in coconut production data. An attempt was made to identify the most influential month for each quarterly coconut production in Bukidnon. Analysis showed that weather condition around the 3rd month after flower setting could explain variations in quarterly production. Abnormal weather conditions (i.e. high moisture deficit or excess) around that period significantly affected the yield. Copra prices in Mindanao were affected mostly by the world market price of coconut oil and not by domestic coconut production. Otherwise, coconut yield forecast could have been a major input in copra price analysis in Mindanao.