Aim This study aimed to evaluate the probability of suitable habitats in the USA for two adventive orchid bee species (Eulaema polychroma (Mocsáry) and Euglossa viridissima Friese), one of which has become established in southern Florida despite the absence of its associated orchid hosts.
Location North and Central America, northern South America and the Caribbean.
Methods Using positive occurrence data within the native range of both orchid bee species, Maxent species distribution modelling was employed to evaluate the probability of suitable habitats in the USA. The power of predictability for the model was tested using partitions of the data.
Results Our results show the absence of suitable habitat in southern Arizona for E. polychroma to maintain populations there, as well as establishing the northernmost limit for the species at around 29°N in north-western Mexico. Suitable habitat was found for E. viridissima in various locations throughout southern Florida. This species is predicted to spread to occupy roughly the southern half of the Florida Peninsula.
Main conclusions The findings indicate that species distribution modelling is useful for evaluating records of species occurrence outside of their native range. Our results indicate that the isolated record of a male of E. polychroma from southern Arizona should not be considered representative of an established population in the absence of further males and females from the same region. Conversely, E. viridissima has successfully become established in south-eastern Florida after a seemingly accidental introduction first noticed in the summer of 2003. We discuss the naturalization of E. viridissima in Florida, the probability of suitable habitat across the Caribbean (where orchid bees are otherwise natively absent today) and the absence of perfume orchids (Orchidaceae). Lastly, we discuss the implications of these results for understanding the biology and biogeography of Euglossini.