Ecological niche modeling of interactions in a buzz-pollinated invasive weed

Abstract

Plant species that are alien and invasive represent an important threat to biodiversity and ecosystem function. These plants compete with native flora for resources and pollinator services. Although most invasive plant species do not depend on pollinators for reproduction and can reproduce uniparentally, some depend on pollinator-assisted reproduction and therefore rely on recruitment of pollinators in the new habitat. Here, we studied Solanum rostratum , a high-risk invasive weed widely distributed around the world that depends on buzz-pollination. It has a floral morphology that depends on medium and large-sized buzzing bees for pollination. The main aim of this study is to model the distribution of an invasive plant, Solanum rostratum, using ecological niches based on its buzz-pollinator distribution. Thus, we hypothesized that S. rostratum is more likely to establish successfully in geographic regions where the appropriate buzz pollinators are also found. We considered the realized and potential distribution of S. rostratum and its pollinators based on occurrence data and ecological niche models combining both abiotic (climatic conditions) and biotic (potential pollinator distribution) factors using the MaxEnt algorithm. Five climatic variables related to temperature and precipitation were important in determining the realized distribution of S. rostratum, while seasonal temperature was the most important variable in determining pollinator distribution. The distributions of the bumblebees ( Bombus fraternus , B. griseocollis , and B. impatiens ) were closely related to the realized distribution of S. rostratum . However, the potential distribution of S. rostratum is wider than that of the pollinator included here, suggesting that other factors such as modifications in the reproductive system or recruitment of other buzzing and non-buzzing bees might be occurring in some invasive populations. We also found that S. rostratum has the potential to expand to unoccupied regions in South America, South Africa, Eastern Europe, and the Middle East. In conclusion, modeling of niches used by invasive species considering both climatic variables and biological interactions such as pollination would be very useful to study and predict the distribution of invasive plants which depend on pollinators to reproduce and establish in new habitats.