Animal-mediated pollination is a critical ecosystem service performed by the subset of food webs distinguished by the exchange of trophic resources for reproductive services. Despite the great importance of pollination to the generation of vast amounts of biodiversity and agricultural productivity, the interdependence between this mutualistic subset and its larger context of trophic interactions is poorly understood. To address this, we explored how the dynamics of plant-pollinator relationships interact with the trophic relationships of non-mutualistic species (including herbivores, omnivores, and carnivores) to determine overall system stability and species’ persistence. We did this by developing a mechanistic network model that integrates trophic and reproductive behaviors based on the nonlinear dynamics of consumer-resource interactions. Our model distinguishes the biomass of floral rewards and vegetation and mathematically formalizes how plant productivity and animal foraging fidelity, visitation frequency, and consumption rates affect the dynamics of the whole system and its component species. Our simulations varied the structure of plant-pollinator networks mimicking pollination systems with and without human disturbance and explored the implications of various ecological hypotheses about whether benefits of reproductive services accrue to plants in a binary, linear, saturating, or unimodal fashion.
While empirically realistic pollination networks sometimes destabilized food webs due to ‘paradox of enrichment’ dynamics, the inclusion of plant-pollinator interactions more frequently increased the stability of food webs. Saturating and unimodal accrual of benefits from pollination led to differing effects on community biomass distribution but the two types of accrual had indistinguishable effects on stability. However, saturating and unimodal functional responses stabilized food web dynamics more than binary and linear responses by preventing the amplification of population oscillations (paradox of enrichment) from occurring. These results represent an important step towards understanding the ecological mechanisms that stabilize food webs and predicting the dynamics of ecological networks that include plant-pollinator interactions.