Quantifying the Expansion of an Invasive Plant Species, Dog-strangling Vine (Vincetoxicum rossicum), in Environmental and Geographic Space Over the Past 130 Years

Abstract

Invasive plant species are an increasing global threat to native biodiversity. Effective management depends on accurate predictions of their spread. However, modelling the geographic distribution of invasive species, particularly with methods like correlative species distribution models (SDMs), is challenging. SDMs operate under the assumption that species are in equilibrium with their environment (i.e., they occur in all suitable environments); this assumption is more likely to be violated for a species that is still in the process of colonizing suitable environments. SDMs also assume that environmental constraints are the most important factors determining a species' distribution. However, these assumptions are not commonly assessed, and when violated can have consequences for model reliability. I investigated SDM performance and equilibrium in the invasive Vincetoxicum rossicum vine in northeastern North America. Vincetoxicum rossicum has a long, detailed history of occurrence records in its invaded range, which enabled me to observe trends in equilibrium and model performance over a relatively long time scale. I tested the hypotheses that: 1) invasive species approach equilibrium in environmental and geographic space over time; 2) SDM performance will increase as V. rossicum approaches environmental equilibrium; and 3) range expansion in the early stages of an invasion is primarily a function of dispersal rather than environmental constraints, while the reverse is true in later stages. I found that V. rossicum has reached equilibrium in environmental space, but is still expanding its geographic range. SDM performance was poor in the first 30 years following introduction, but then improved as V. rossicum approached environmental equilibrium. SDMs were outperformed by spatial dispersal models in the earliest time period, however, the reverse was true for all subsequent time periods. Overall, these results suggest that V. rossicum’s distribution is becoming more stable and more predictable over time and that models built using the most recent data for this species, will be the most transferable across time and space. Additionally, my findings highlight the need for researchers modelling invasive species’ distributions to consider the inherent assumptions, biases, and unique features related to SDMs and SDMs of invasive species.