Maximum Entropy Modelling of Parthenium Hysterophorus L. Under Current and Projected Climate Change Within the Maasai Mara Ecosystem

Abstract

Land managers are increasingly using Species Distribution Models (SDMs) in modeling potential geographical distribution of Invasive Alien Plant Species (IAPS). These models show current distribution and also areas prone to invasion by the species if the conditions affecting their survival are the same as those in their native areas. Therefore, species niche requirements are the basis of development of these models. Among the numerous SDMs that are in use, Maxent approach was applied in this study because it operates on presence-only data. Consequently; it is simple, fast, economical and convenient to construct meaningful species distribution models from the data that included coordinates of the IAPS. Further, GIS and Remote Sensed data that was required was easily available for extracting a set of environmental predictions that affect the fundamental and residual niches of the IAPS. These environmental predictions include precipitation, elevation and temperature for the area of study. Socio-economic factors such as wildlife population and livestock density were obtained from KWS. This study addressed one of the major IAPS that are thought to pose the highest risks to the Mara ecosystem; Parthenium hysterophorus L. For this study, a stratified random sampling method was used. Study area was stratified on the basis of the road network that is the main agent of IPS dispersal. Plots of 10 by 10m were set on either side of the road at every 1km travelled. GPS coordinates of the sampled Parthenium hysterophorus L., for use in the Maxent modelling were obtained by using an application developed by the Regional Centre for Mapping of Natural Resources for Development (RCMRD) called ‘Invasive Species Mapper’ available from google play store. A combination of ArcMap 10.5 and the Maxent modelling software were used to manipulate the data collected into the final product; a prediction map showing the potential geographical locations of the spread of the invasive species under study. Habitat suitability maps for the present conditions and for future were created using present environmental and socio-economic factors that impact on the growth and distribution of the IPS and extrapolated climatic scenarios of RCPs 2.6 and 8.5 for the year 2050 respectively. Results showed range expansion in the current and the projected scenarios posing imminent threat to native vegetation. Models offer a proactive management strategy for the management of IAPS. Key words: Species Distribution Models, Maxent, Niche, IAPS, GIS, Remote sensed data, ArcMap 10.5 software, targeted sampling, prediction map, RCPs.