WEATHER - BASED FACTORS THAT INFLUENCE POPULATION DISTRIBUTION OF MYOMORPH RODENT PESTS IN CEREAL FARMS OF UNIVERSITY OF ELDORET, KENYA.

Abstract

In Kenya, rodents are pests of concern because they cause damage to cereal crops before and after harvest, posing a threat to both food availability and safety. However, information about the ecology of myomorph rodent pests and the infestation of cereals in Kenya is limited. The objective of the research was to identify myomorph rodent species, assess population abundance and determine how weather -based ecological factors of rainfall, relative humidity and temperature influence rodent populations, species richness and distribution of male and female rodents in University of Eldoret farms. Two habitats (Maize and wheat farms) were selected where a 70mx70m grid crop cutting was done in commercial fields for the study. The grid was subdivided into four quarters each with four Sherman’s live traps and four locally woven live traps resulting in 32 trapping points with trap spacing being 10m apart. Peanut butter and sun-dried fish (Rastrineobola argentea) were used as baits. The rodents captured were identified to species level using keys for classifying rodents. Chi square test, independent sample t-test, Poisson, binary logistic and multinomial regression were used to establish the relationship between nominal response variable given one or more predictors. Autoregressive Integrated Moving Average (ARIMA) [1, 1, 1] [2, 0, 0] Time series model was used. Three myomorph rodent species Mastomys natalensis, Arvicanthis niloticus and Lemniscomys striatus were captured during the study period. A total of 924 myomorph rodents were captured with 50.97%, 19.48% and 29.54% captured in year one, two and three, respectively. M. natalensis represented the highest captures (60.61%), then A. niloticus (38.42%) and L. striatus (0.97%). There was a significant variation in infestation in year one (P = 0.001) and no significant variation in year two (P = 0.499) and three (P = 0.127). Species displayed variation in distribution. There was a significant difference in distribution of gender in second year of study (t = -2.625, P = 0.009) and overall, no significant variation in distribution of gender (t = 0.525, P = 0.600) in the two habitats. ARIMA (1, 1, 1) model predicted a higher abundance of rodents between the months of March and July with decline in November to January of the forecasted year with minimal variation. The findings indicate statistically significant difference in species distribution of rodents in maize and wheat fields (t = 3.523, P = 0.001) and weak positive correlation between rainfall and gender distribution (r = 0.171, P = 0.001), a weak positive correlation between relative humidity and gender (r = 0.198, P =0.001) that was statistically significantly different. Rainfall and relative humidity had an effect on distribution of gender. However, there was a weak positive correlation between temperature and gender distribution in year one and two (r = 0.056, P = 0.225; r =0.093 P = 0.214) and weak negative correlation in year three (r = -0.046, P = 0.449 with no statistically significant effect of temperature on gender distribution (P > 0.001). Pearson correlations between ecological factors of rainfall, relative humidity and temperature showed a linear weak correlation with species distribution (r =-0.001, P=0.986).Rainfall and relative humidity were shown to influence species distribution, but temperature did not have notable influence on species distribution in this study. Rainfall, relative humidity and temperature could predict the changes in population, species and gender. Changes in temperature had inverse relationship with gender and species abundance. In conclusion, there existed variation in infestation of myomorph rodents, the type of species identified did not vary and weather-based ecological factors of rainfall, relative humidity and temperature do not have consistent in influence on myomorph rodent pest population, species and gender disparity in cereal farms in the three cropping years.