Experimental pollinator decline affects plant reproduction and is mediated by plant mating system

Abstract

There is growing concern that current pollinator decline will affect the reproduction of plant species, potentially driving a decline in plant population densities. We experimentally tested whether a reduction in flower visitation caused a reduction in fertilization rate in several species, and whether any reduction in fecundity of species depends on their degree of reproductive dependence on pollinators and their attractiveness for pollinators. We recorded visitation rate, fertilization rate, seed weight, flower size and density of nineteen insect-pollinated perennial herbs inside thirty 2 x 2 m dome-shaped cages covered with fishnet (experimental plots) and in thirty control plots in a Norwegian hay meadow. We used a bagging experiment to estimate the ability of the study species to produce seeds in the absence of pollinators. The visitation rate for fifteen of nineteen study species was lower inside cages than outside and only three of the fifteen species showed significantly reduced fertilization rates in the experimental plots. The magnitude of reduction in fertilization rate was positively related to the degree of pollinator dependence, but not to their attractiveness for pollinators or to the reduction in visitation rate. Seed weight was not affected by the experiment. The lack of an overall effect of reduced pollinator visitation on fertilization rate suggests that some species may be robust to a pollinator decline that could increase pollen limitation on plant reproduction. Our results suggest that species with greater pollinator dependence are more vulnerable to pollinator loss.

Experimental simulation of pollinator decline causes community‐wide reductions in seedling diversity and abundance

“…We hypothesized that responses to the experimental treatment should differ among species depending on their dependence of pollinator visitation for seed production, because the strongest effect of our experimental treatment on reproductive success occurred in the species that were most dependent on pollinators for seed production (Lundgren et al. ), and because self‐incompatible species decline more than self‐compatible species with pollinator loss (Biesmeijer et al. ).…”

“…, Lundgren et al. ). However, this will only affect plant population dynamics if recruitment is also limited by seed production (Crawley , Eriksson and Ehrlén , Bond ).…”

“…Based on our results from previous studies with this experimental design, we expected that species with a high degree of pollinator dependence for seed production (Lundgren et al. ), strong seed limitation on population growth (Lundgren et al. ) and low local conspecific density (Lázaro et al.…”

“…Corresponding Editor: R. J. Mitchell. 3 E-mail: rebekka.lundgren@nmbu.no visitation rates and pollination efficiency decline (Ashman et al 2004, Lundgren et al 2013). However, this will only affect plant population dynamics if recruitment is also limited by seed production (Crawley 1990, Eriksson and Ehrlén 1992, Bond 1994.…”

Effects of long‐term variation in pollinator abundance and diversity on reproduction of a generalist plant

“…Several previous studies use pollination networks to simulate effects of pollinator extinctions (Kaiser‐Bunbury, Muff, Memmott, Muller, & Caflisch, ; Memmott et al, ). Some recent removal experiments also evaluate the potential effects of pollinator declines, but on relatively short time‐scales (Lundgren, ; Lundgren, Lázaro, & Totland, ). Winfree et al () simulated effects of honey bee loss on median pollen deposition in farmed watermelon.…”