Effects of experimentally simulated pollinator decline on recruitment in two European herbs

Abstract

Summary Pollinator decline causes great concern because reduced pollinator availability may negatively affect plants' seed production and in turn influence seedling recruitment and densities of plant populations. We used a novel experimental approach that simulates pollinator decline to assess the effects of a 4‐year reduction in pollinator availability on reproduction and recruitment in two common European herbs, Centaurea jacea and Leucanthemum vulgare. The experiment strongly reduced visitation rates (by 59%) and seed set (by 23%) in Leucanthemum, but this did not cause reduced seedling recruitment, suggesting that other factors, such as water availability or the existence of a seed bank, are important for recruitment in this population of Leucanthemum. The experiment also significantly reduced visitation rates to Centaurea (by 33%), but did not affect seed set or seedling recruitment. In this species, however, seed set increased with visitation rates and local recruitment increased with seed production, indicating that a consistent pollinator decline may negatively affect Centaurea population densities. Synthesis. Our experiment reduced pollinator availability without artificially affecting quality components of the pollination process. Therefore, this approach may represent a realistic simulation of a pollinator decline. To our knowledge, this is the first study that uses an experimental reduction in pollinator visitation to assess its effects on population recruitment in plants. This study shows that a large pollinator decline may affect population densities, but only in populations where seedling recruitment is limited by seed production. Our results also highlight the importance of going beyond visitation rates and seed production to study the impact of pollinator decline on plant populations.