Interactive effects of climate and land use on diversity, trophic interactions, and phenology of pollinator communities

Chapter 1 – General introduction. The insect decline documented in the last decades has raised concerns due to the importance of insects for ecosystem functioning and our economic activities. Bees and other flower-visiting insects in the Hymenoptera, Diptera, Lepidoptera and Coleoptera orders provide a variety of functions—from pollination to pest control—, belong to different trophic levels, and respond to different global change drivers at different spatial scales. Climate and land-use changes are major drivers of pollinator declines. However, evidence of their interactive effects is largely missing despite their assumed importance. The LandKlif project sets a novel multiscale space-for-time approach along independent gradients of climate and land use, ideal for filling this knowledge gap. Fieldwork was conducted on 179 sites within four habitat types (forest, grassland, arable land and settlement) embedded in 60 regions of different land-use intensity (seminatural, agricultural and urban). Within this framework, I studied how the community composition and diversity of seven pollinator taxa change with climate and land use at several spatial scales (Chapter 2); how climate–land-use interactions affect the abundance and richness of taxa of a food chain of cavity-nesting insects, considering resource availability and parasitism rates (Chapter 3); and how climate warming influences the emergence phenology of cavity-nesting insects, bee-antagonist synchrony, and bee weight, disentangling genetic adaptation from phenotypic plasticity (Chapter 4). Chapter 2 – Interactive effects of climate and land use on pollinator diversity differ among taxa and scales. Awareness of the importance of non-bee pollinators has increased recently. However, we are still far from understanding synergistic climate and land use effects among pollinator taxa and scales. Here, I used Malaise traps and DNA metabarcoding to unravel the additive and interactive effects of both drivers on community composition, regional gamma diversity, community dissimilarity (beta diversity) and local alpha diversity of seven pollinator taxa. Climate–land-use interactions shaped pollinator community composition and diversity at large spatial scales, while local diversity was affected by their additive effects. Low-intensity land use increased pollinator diversity, while climate effects depended largely on the taxa considered. However, at the largest interregional scale, higher temperatures homogenized most pollinator communities. This chapter underlines (1) the importance of low-intensity land uses at multiple scales for pollinators, (2) of regional forest cover for buffering climate change effects, (3) the role of flowering plant diversity to increase pollinator alpha and beta diversity, and (4) the risks of climate change for community homogenization independently of land-use factors. Chapter 3 – Temperature and land use interactively shape food resources, parasitism rates, and diversity of cavity-nesting bees, wasps, and their antagonists. Climate–land-use interactions are also expected to affect interconnected trophic levels. In this study, I used trap nests to assess combined effects of land use and temperature on a food chain from flowering plants to bees, hunting wasps, and their antagonists (parasitoids, predators and cleptoparasites), while accounting for resource availability. Parasitism rates and bee abundance and richness were influenced by climate–land-use interactions related to urban areas (bees), forest cover (parasitism rates) and landscape configuration (both). Higher trophic levels were more independent of temperature than lower trophic levels, while land use was a strong predictor across taxa. Cropland had generally worse effects than urban areas, especially for antagonists, which were highly dependent on resource availability. However, bee diversity declined in urban areas with higher temperatures across spatial scales. This chapter highlights the importance of landscape configuration at small spatial scales (<600m), of seminatural habitats across spatial scales, and the need of reducing urban temperatures and preserving forest patches in anthropogenic landscapes. Chapter 4 – Climate warming threatens bee fitness by increasing bee-antagonist synchrony and reducing bee body weight. Climate warming effects on pollinating insects are less clear than land-use impacts, and we lack information on changes in the phenology of bees and their antagonists with climate warming. Hence, I performed a climate chamber experiment to investigate how climate warming affects the emergence phenology of cavity-nesting communities, the synchrony of bees and antagonists, and bee weight at emergence, differentiating genetic and plastic adaptations. My study confirms that bee’s prepupal phase determines phenotypic plasticity to temperature cues and extends the finding to whole cavity-nesting communities. Besides, I found genetic adaptation of insects to their climate of origin, both in terms of emergence timing and weight, and increased bee-antagonist synchrony with warmer climates. Also, the weight of bees overwintering as pupa depended both on temperature and on emergence timing, but females might be more able to regulate their emergence weight than males. In conclusion, climate warming can increase bee-antagonist synchrony, with likely negative effects for bee populations, while simultaneously decreasing their emergence weight. Changes in climate can affect bee populations genetically adapted to certain temperatures if microevolution does not occur fast enough. Chapter 5 – General discussion. This work shows additive and interactive effects of climate and land use at multiple spatial scales for different pollinator taxa and parasitism rates. Forest habitats were consistently positive for insect diversity, despite their low flowering plant richness, while in arable land, both floral and pollinator diversity was poor. Flowering plant diversity increased pollinator diversity at local scales and the dissimilarity of pollinator communities at regional scales. The effects of temperature and precipitation varied among taxa, but most communities appeared more homogeneous under warmer regional climates independent of land-use intensity, indicating that not all climate change effects can be buffered by land-use management. However, within regional scales, large forest areas can buffer climate change effects, while urban areas can intensify them. Parasitism rates also increased with the combination of landscape connectivity and warmer temperatures. While I showed experimentally that climate warming can reduce bee body weight and increase bee-antagonist synchrony, observed parasitism rates did not increase with temperature. This and other unanswered questions need further research. For now, I can recommend reducing greenhouse emissions to avoid pollinator community homogenization and the heat-island effect of cities, preserving large forest areas and improving their management for pollinators, and to keep working on making agricultural landscapes more pollinator friendly. Conservation measures must consider synergistic effects of climate and land use to be effective.