Hot Models - predicting the current and future distributions of two ambush bug species, Phymata americana and Phymata pennsylvanica

Abstract

A fundamental question in ecology centers around species ranges and the limits of a physical niche space imposed by environmental variables. Species distribution models (SDM) provide the ability to identify the effect of environmental variables on species distributions and to forecast future environmentally-driven changes. One such SDM is Maxent, a machine-learning software that uses environmental raster grids to predict the probability of species presence within grids.

 

In the present study, I use Maxent to identify the environmental variables maintaining the distribution of two closely related ambush bug species found in parapatric distribution, Phymata americana and Phymata pennsylvanica. I also created binary suitability maps to predict the current and future ranges of both species given climate change scenarios. The model identified similar variables (temperature and precipitation) that contributed to predicting ambush bug ranges, but these variables have different means and consequently result in distinct ranges. To map future distributions, Representative Concentration Pathways (RCP) projecting various greenhouse gas emissions were used. Distributions narrow in latitude and longitude as RCP scenarios worsen for both P. americana and P. pennsylvanica, but P. pennsylvanica had relatively smaller-scale shifts at a lower RCP.

 

These results provide evidence for specific environmental requirements for P. americana and P. pennsylvanica and highlight the effects of climate change on range shifts. Understanding how these abiotic factors affect ambush bug distributions will be fundamental for future research on their taxonomy and conservation.