Pollen protein content has been demonstrated to be an essential nutritional component for pollinators, particularly bees, and thus important in mediating plant-pollinator interactions. However, it is unknown how investment in pollen protein impacts male and female reproductive success in plants within diverse co-flowering communities. Among-species variation in resources allocated to pollen nutrition could further be constrained by life-history strategies (e.g., survival-reproduction trade-offs) or evolutionary history.

Here, we surveyed pollen protein content for 29 species within a diverse co-flowering community and evaluated the effect of pollen protein on male and female reproductive success across plant species. We also tested the role of life history (annuals vs perennials) and phylogeny in mediating differences in resource allocation to pollen nutrition. 

We found that pollen protein content influences components of male (bee visitor abundance and pollen dispersal) but not female (conspecific pollen deposition and pollen tube growth) reproductive success, suggesting this trait affects plants via male function with the potential for sexual conflicts driven by differential investment on this trait. We found no phylogenetic signal on pollen protein content. However, pollen protein content was higher in annual compared to perennial species suggesting survival versus reproduction trade-offs also contribute to variation in pollen protein at the community level. 

Our study underscores the importance of understanding the ecological and evolutionary drivers of pollen protein content across plant species. Our results further suggest the existence of sexual conflicts and ecological trade-offs mediated by differential investment in pollen nutritional quality, with important implications for community assembly and the structure of plant-pollinator interactions.

This dataset was used to assess how pollen protein content impacts male and female reproductive success acros co-flowering plant species. This study was conducted at a serpentine seep meta-community in the McLaughlin Natural Reserve (38.8582º N, 122.4093º W) in Northern California, USA. We collected pollen protein content from 29 plant species (10 annual and 19 perennial species), using pollen/anthers of ~10 individual plants per species. For each plant species, we obtained male and female reproductive success components. Male components were 1) bee visitor abundance, 2) pollen dispersal and 3) pollen disperser richness, all obtained through sampling flower-visiting bees (n = 621) in the community between 09h00 and 15h00 using entomological nets over 15 total days between May 9th and June 1st, 2021. We also collected bee pollen loads and quantified the number of pollen grains on the body of the bees and identified them, which were used to calculate mean pollen dispersal and pollen disperser richness for each plant. Female components were 1) mean number of conspecific pollen grains (CP) on stigmas and 2) the proportion of CP that formed pollen tubes. Both female components were obtained from 1279 styles collected from the studied plant species (55.6 ± 37.1 styles per species). Additionally, we measured bee body size to understand the role of pollen protein content in attracting bees that require superior nutritional income, and estimated plant species abundance in the flowering season using 40 plots (1x2-meter each) to understand their influence on the considered male and female components.