Pollinator asynchrony drives the temporal stability of flower visitation rates, but not of plant reproductive success

Abstract

Abstract The stability of ecological communities over time, often quantified using their aggregated abundance or biomass, is thought to be influenced by species richness and the asynchrony of species abundance fluctuations. However, the relationship between community stability and the stability of ecosystem functioning provided by this community has been little explored, especially for functions involving several trophic levels such as plant pollination. Understanding the mechanisms driving community stability and how this should be related to the stability of ecosystem functions is important to predict the resilience of communities in front of the increasing threats to biodiversity. Using a 5‐year dataset across 13 Mediterranean shrublands and 12 plant species, we assessed the effect of pollinator richness and asynchrony as drivers of stability in plant visitation rates and plant reproductive success. Additionally, we analysed whether the effects of species richness and visitation rate on plant reproductive success were consistent over the years. The stability of plant visitation rates is driven by the asynchrony of pollinator species abundances, but the effect of pollinator species richness is only indirect, as richer communities also tend to have more asynchronous populations. In contrast, the stability of plant reproductive success is not related to the stability of pollinator visitation rates. A year‐by‐year analysis reveals that plants in the studied system are unlikely to be limited by pollinator abundance except in specific years, blurring any potential effect of visitation rate stability on reproductive success stability. Synthesis: We highlight the importance of asynchronous fluctuating species that are redundant in their function as an important factor for the stability of ecosystems. This is particularly important for pollinators with diverse responses to environmental fluctuations, as their asynchrony between years ensures stable levels of flower visitation rates. However, we warn that final ecosystem functions such as plant reproductive success, which depends not only on pollinators but on plant competition or resource availability, might be decoupled from first order ecosystem functions (i.e. visitation rates and pollen deposition) when pollination is not a limiting factor.