Development of a predictive model for soil temperature and its application to species distribution modeling of ant species in South Korea

Abstract

Soil temperature is an important factor for determining ant species inhabitation, but it has rarely been applied in species distribution modeling due to limited data sources. Hence, this study aimed to develop a predictive model for soil temperature and evaluate the potential distribution of a few ant species in South Korea. The monthly maximum and minimum soil temperatures were predicted by linear regression as a function of the monthly maximum and minimum air temperatures. Then, the developed model was used to predict the potential distribution of the tropical fire ant (Solenopsis geminata Fabricius; TFA), red imported fire ant (Solenopsis invicta Buren; RIFA), and yellow crazy ant (Anoplolepis gracilipes Smith; YCA). Three ant species were selected because they have globally exerted negative impacts on the ecology of the invading area. The developed model could predict soil temperatures at 5, 10, 20, and 30 cm deep with an R2 value of 0.95. The climate suitability of TFA, RIFA, and YCA, predicted by soil temperature in the CLIMEX model, was found to be up to 25, 12, and 65 times higher than the climate suitability predicted by air temperature, respectively. This study showed that the potential distribution predicted based on air and soil temperature could be different for a species inhabiting the ground, suggesting that it is necessary to consider not only the outside environment, but also the soil environment, for an accurate prediction.