A study on when to conclude a long term fertilizer trial on coconut yield

Abstract

Yield data obtained for twenty consecutive from a fertilizer trial carried out at the coconut research institute in Sri Lanka on a mature coconut plantation was subjected to an analysis with a view to ascertatining the relative influence of repeated application of fertilizer on coconut yield. The inter-annual correlation coefficients between successive pairs of years showed that the correlations first increased, then reached a plateau, then increased again and finally ended up in an asymptote. From the eighth year onwards the inter-annual correlation between succeeding pair of years remained stati0onary. The average correlation was 0.9351. On the basis of these results, an eight to ten year period is considered sufficient to under stand the full response to fertilizer in an experiment. After this period, it appears no useful purpose would be achieved.