Input-response analysis in arecanut - a semiparametric regression approach.

Abstract

The effect of various input factors on the yield of arecanut has been studied based on the field data collected from fanners' gardens in Kasaragod district. Semi parametric regression technique is used to explain input-response relationship. In this approach, both qualitative (discrete) and quantitative (continuous) input variables can be included. The quantitative variables such as fertilizer application (N, P and K), Farm Yard Manure (FYM), Green leaf (GL) application and density (number of palms/ha) and the qualitative variables such as variety (SK local, Mohitnagar, Mangala), type of irrigation (hose, sprinkler) and cropping models(mono, arecanut+banana, arecanut-cocca, arecanut+banana+pepper) were considered as input variables. The average yield (kg/ha)was taken as the response variable. The effect of qualitative variables on yield and the optimum input levels of the quantitative variables were obtained. The estimated optimum levels of N, P and K per palm in the study area are 1OOg, 5Og and 120g, respectively, which are very close to the recommended doses. The yield differences in arecanut under monocrop and mixed cropping systems were not significant and therefore, it is better to maintain arecanut gardens' under mixed cropping system for better return. The yield variations among three varieties were found to be not significant once the effects of other variables are eliminated. No significant difference between the two methods of irrigation was found, though sprinkler gave comparatively more response than hose irrigation.