Habitat suitability study for green gram production under present and future climatic scenarios in Kibwezi East Kenya

Abstract

The species distribution model was used to predict the suitability of green gram production under the present, RCP 4.5 and 8.5 climate scenarios. An ensemble of a species distribution model comprising six models was developed. Validation of these models revealed that all models were robust with the best model being random forest (RF) with Area Under the Curve (AUC) = 0.98 and Deviance = 0.29 while the least was the generalized linear model (GLM) with AUC = 0.87 and Deviance = 0.71. The green gram habitat suitability greatly decreased under RCP 8.5 climate scenario prediction whereby about half of the agricultural land in the Kibwezi East Sub County was highly unsuitable for green gram production. The Habitat suitability predictions showed that Thange ward out of the four wards in the location was the most suitable for green gram production. However, as per the predictions its suitability for green gram production may be affected by climate change under all climate scenarios. Results from this study give decision-makers a foundational understanding of the likely effects of climate change in the 2050s compared to the present scenario on habitat suitability for green gram production and a basis for creating strategies and policies to enhance adaptation and create resilience to its effects.