| Global change is threatening the mutualisms between species, and hence the function of ecosystems. Pollinator declines have raised concerns about persistence of plants dependent on insects for their reproduction, but our knowledge on the role of landscape-mediated pollinator response diversity in ensuring plant reproduction, a key ecosystem function, across space and time is lacking. It has been shown for plant communities that negative co-variation in the abundance of plant species has the potential to stabilize ecosystem functioning in space and time, through response diversity among species and populations. We know little about the importance of such mechanisms for stabilizing pollination services by insects. Mobile organisms respond to both year-to-year variability in climate, but also to the landscape-scale resource dynamics, according to their traits. Pollination systems being rather generalized, we hypothesize that such response diversity has strong buffering potential for the reproductive outcomes of plants.
Here we propose to assess the buffering role of pollinator diversity for wild plant reproduction by quantifying the link between landscape-scale land-use dynamics, pollinator response diversity and stability of pollination of wild plants using observations from five consecutive years in 18 semi-natural grasslands. The grasslands are situated in landscapes differing in the availability and temporal variability of floral resources, in a large part driven by land-use and crop rotation. This has shown to cause variability in local pollinator densities due to dilution/concentration processes. We will collect data to extend an existing three-year dataset on plant-pollinator interactions and seed-set of 14 plant species in permanent plots. We will quantify availability and year-to-year variability of floral resources (incl. mass flowering crops, flowering field edges, trees and hedges) using a combination of spatial data and ground observations. We will scale up plant-pollinator network information by identifying pollen from the collected pollinators using a recently developed AI-based system. Hierarchical multi-species and structural equation models will be used to link landscape dynamics to measures of spatio-temporal variation in pollinator availability, and on to the spatio-temporal variation in focal ecosystem function, to wild plant seed production.
Our results will provide a better understanding of biodiversity-ecosystem functioning relationships across scales, and shed light on the still poorly captured insurance value of biodiversity at landscape scales. Our assessment of the consequences of landscape-scale land-use heterogeneity in space and time will enable us to assess across which landscape-scale land-use constellations we expect pollinator communities to be able to deliver pollination services to wild plants and thus contribute to sustaining biodiversity in habitats of high conservation value.
|