Code and data for "Biased correlated random walks and invasive spread: Insights from the alga Codium fragile"

• Main Authors: Karine Gagnon, Stephanie Peacock, Mark Lewis, Yu Jin
• Format: Dataset
• Language: English
• Subjects: envir; geo; Northwest Atlantic
• Online Access: https://doi.org/10.7939/R3JM23W91; https://era.library.ualberta.ca/items/d01dc048-c752-4b66-817d-e442fbd3d6d8

Invasive species are a major threat to global biodiversity. Understanding what influences the spread of invasive species after introduction is key to minimizing impacts on native ecosystems and has been the subject of much applied and theoretical work. Thus far, models of spread have not accounted for autocorrelated movement in the dispersal of invading organisms, even though many invasions are driven by wind or ocean currents that are autocorrelated. We considered the invasion of the alga Codium fragile in the northwest Atlantic. Long-distance dispersal of Codium is via wind-driven movement of fragments that produce oxygen to maintain buoyancy. An individual-based model for the dispersal that included the stochasticity in sunlight was used to generate a distribution of dispersal distances. We then applied an integrodifference model for population spread from which spreading speed (km/yr) was calculated. We found that spreading speed to the south steadily increased with increasing dispersal time, but the spreading speed to the north was greatest for intermediate dispersal times. We hypothesized that this resulted from autocorrelation in wind speed and direction, with southerly winds dominating in the long run. We confirmed this result using a partial differential equation model for dispersal derived from a correlated random walk with bias. This more general approach may have application to other species where dispersal is driven by autocorrelated processes.