Climate variability environmental stress indication across four rain-fed states in nigeria using multivariate analysis

Abstract

Climate parameters can be used to verify already established model for climate variability or change indication (CCI) of environmental  stress (ESI) to ensure agricultural productivity and sustainability. The study verified ESI from 32-year temperature, relative humidity and  solar radiation data across four (Benue, Edo, Niger and Ondo) States in Nigeria using multiple regression model. The ESI was higher than  30 from February to April 2005 and May 2007 in Edo, April 2006, January 2013 and 2014 in Bida. The reliability statistics had a Cronbach's  value of 0.821, so the data had good internal consistency. The data distributions were highly significant (F = 87.355, p = 0.000) from the  Hotelling's t-squared statistic (t 2 ). There was a very strong correlation (0.814) between April and May at 0.01 levels. The model explained  64.2 % variance in the variables. The Durbin-Watson value < 2 indicated positive autocorrelation. The ANOVA indicated a  general significance (p < 0.05) in the model's fitness. The computed ESI was meritorious (KMO = 0.859) and valuable (χ2 = 1494.061, p <  0.05) for the factor analysis. The Principal Component Analysis showed that the seven-month rainy periods under Component 1 with  higher eigenvalues had been having higher ESI than the dry periods under Component 2. The four study States could be having shortage  rainfall distribution and the farmers could easily be getting tired thereby being less productive from ESI. So, there is need for  he farmers across the four States to craft strategies for proper adaptation to effects of the climate variability and change.