The exchange of material and individuals between neighbouring food webs is ubiquitous and affects ecosystem functioning. Here, we explore animal foraging movement between adjacent, heterogeneous habitats and its effect on a suite of interconnected ecosystem functions. Combining dynamic food-web models with nutrient-recycling models, we study foraging across habitats that differ in fertility and plant diversity. We found that net foraging movement flowed from high to low fertility or high to low diversity and boosted stocks and flows across the whole loop of ecosystem functions including biomass, detritus, and nutrients in the recipient habitat. Contrary to common assumptions, however, the largest flows were often between the highest and intermediate fertility habitats rather than highest and lowest. The effect of consumer influx on ecosystem functions was similar to the effect of increasing fertility. Unlike fertility, however, consumer influx caused a shift towards highly predator-dominated biomass distributions, especially in habitats that were unable to support predators in the absence of consumer foraging. This shift resulted from both direct and indirect effects propagated through the interconnected ecosystem functions. Only by considering both stocks and fluxes across the whole loop of ecosystem functions do we uncover the mechanisms driving our results. In conclusion, the outcome of animal foraging movements will differ from that of dispersal and diffusion. Together we show how considering active types of animal movement and the interconnectedness of ecosystem functions can aid our understanding of the patchy landscapes of the Anthropocene.

This simulation dataset was simulated using a dynamic food-web model combined with a nutrient recycling model in Matlab. The code used to create this data is also provided.

The model simulated food web and nutrient dynamics in two habitats, with foraging movement of some consumer species between habitats. The habitats differed in number of basal species, with potential scenarios including low, medium, or high diversity of habitat 1 (div1 = 1, 2, or 3 respectively) and low or high diversity of habitat 2 (div2 = 1 or 3). The habitats also differed in fertility, with potential scenarios including low, medium or high fertility (productivity) for habitat 1 or 2 (prod1 and prod2 respectively). We simulated scenarios with no, low, or high linkage (consumer movement) between habitats (link = 0, 1, or 2). There were 1000 replicates for each scenario.

The data here are the mean of the final 100 time steps (they were almost always at equilibrium by this point, but this removed the effect of any cycles). We report the coefficient of variation for each value, so any instances where the simulation was not at equilibrium should be easily identified.