Determination of best fit model for the distribution and crop loss associated with Bacterial wilt of tomatoes

Abstract

Tomato is the second leading crop in Kenya in terms of production and value after potato. It is widely used as vegetable across the world. However, tomato varieties are attacked by bacterial wilt which devastates farmers. The bacterium is soil borne and persists in contaminated soils. It can be vector transmitted and has wide host’s range of over 50 plant species making it difficult to control. Bacterial wilt attack on tomato farm results to losses of more than 90%. Its manifestation changes with varying conditions and farm management practices. Bacterial wilt infects roots and stems of many plants that are considered alternative hosts. This enables it to continue spreading across the tomato growing regions. The study used logistic model to determine the distribution, of bacterial wilt. The study used secondary data obtained from the plant clinics on reported cases of bacterial wilt. In this study, disease incidents and distribution were inferred from cases presented by farmers to plant clinics. The response variables used were presence of bacterial wilt and estimated crop loss in the farm. The explanatory variables used in the model were weather data (minimum daily temperature, maximum daily temperature, minimum relative humidity, maximum relative humidity and precipitation), development stage of tomato and agro ecological zones (AEZ). Data from the year 2012 to 2013 obtained from Plantwise Online Management System (POMS) on tomato crop was used. R-Statistical Programming Package was used for the logistic analysis. The results showed that relative humidity and minimum temperature significantly influenced bacterial wilt incidence and distribution as inferred from cases presented by farmer to plant clinics. In the analysis, agro-ecological zones, LH2, LM3, LM4, LM7, UM2 and UM4 significantly influence tomato losses in the farm due to attack by bacterial wilt. In the AEZ, the coefficient estimates are positive showing an increase in the bacterial wilt incidents. The disease incidents as presented cases brought to clinics by farmers varied in each county. Kirinyaga showed the highest incidents of the disease followed by Nakuru and Embu at 20.6%, 20.2% and 19.1% respectively. Bacterial wilt was found to be present in all the counties irrespective of difference in AEZ.