Host plant distributions and climate interact to affect the predicted geographic distribution of a Neotropical termite

Species distribution models (SDMs) have been widely used in the scientific literature. The majority of SDMs use climate data or other abiotic variables to forecast the potential distribution of a species in geographic space. Biotic interactions can affect the predicted spatial distribution of a species in many ways across multiple spatial scales, and incorporating these predictors in an SDM is a current topic in the scientific literature. Constrictotermes cyphergaster is a widely distributed termite in the Neotropics. This termite species nests in plants and more frequently nests in some arboreal species. Thus, this species is an excellent model to evaluate the influence of biotic interactions in SDMs. We evaluate the influences of climate and the geographic distribution of host plants on the potential distribution of C.cyphergaster. Three correlative models (MaxEnt) were built to predict the geographic distribution of the termite: (1) climate data, (2) biotic data (i.e., the geographic distribution of host plants), and (3) climate and biotic data. The models that were generated indicate that the potential geographic distribution of C.cyphergaster is concentrated in the Cerrado and Caatinga regions. In addition, path analysis and multiple regression revealed the importance of the direct effects of biological interactions in the geographic distribution of the termite, while climate affected the distribution of the termite mainly through indirect effects by influencing the geographic distributions of host plants. The current study endorses the importance of including biological interactions in SDMs. We recommend using biotic predictors in SDM studies of insect species, mainly because insects have important environmental services and biotic interaction data can improve the macroecological studies of this group. Resumo Os modelos de distribuicAo de especies (SDMs) tem sido amplamente utilizados na literatura cientifica. A maioria dos SDMs usa dados climaticos ou outras variaveis abioticas para predizer a distribuicAo potencial de uma especie no espaco geografico. Interacoes bioticas podem afetar a distribuicAo espacial predita de uma especie de varias maneiras em multiplas escalas espaciais, incorporar esses preditores em um SDM e um topico atual. Constrictotermes cyphergaster e uma especie de cupim com ampla distribuicAo no Neotropico. Esta especie nidifica em especies de plantas, principalmente de porte arboreo. Assim, esta especie e um excelente modelo para avaliar a influencia das interacoes bioticas em SDMs. Avaliamos a influencia do clima e da distribuicAo geografica das plantas hospedeiras sobre a distribuicAo potencial de C.cyphergaster. Tres modelos correlativos (MaxEnt) foram construidos para predizer a distribuicAo geografica do cupim: (1) dados climaticos, (2) dados bioticos (ou seja, a distribuicAo geografica das plantas hospedeiras) e (3) dados climaticos e bioticos. Os modelos que foram gerados indicam que a distribuicAo geografica potencial de C.cyphergaster esta concentrada no Cerrado e na Caatinga. Alem disso, a analise de caminho e a regressAo multipla revelaram a importancia dos efeitos diretos das interacoes biologicas na distribuicAo geografica dos cupins, enquanto o clima afetou a distribuicAo do cupim principalmente atraves de efeitos indiretos influenciando a distribuicAo geografica das plantas hospedeiras. O presente estudo confirma a importancia de incluir interacoes bioticas em SDMs. Recomendamos o uso de preditores bioticos em estudos SDM de especies de insetos, principalmente porque os insetos prestam servicos ambientais importantes e os dados de interacAo biotica podem melhorar os estudos macroecologicos deste grupo.