The effect of land use on intra-specific variation in floral traits and subsequently plant-pollinator interactions and pollination

In IntraFlor, we want to (i) further investigate and quantify the effect of grassland management and population genetic diversity on intraspecific variation in these flower traits, and (ii) unravel the consequences of such trait plasticity within and between plant populations/communities for individual pollinator behavior and plant reproductive fitness.

Within three working packages (WP), we will:
(i) quantify correlations between different types of grassland management, local population genetic diversity and intraspecific variation in morphological, physiological, and chemical flower traits of selected plant species across space and time,
(ii) investigate how intraspecific variation in flower traits relates to foraging patterns of individual pollinators, and
(iii) determine how intraspecific trait variation affects differences in pollination success of individual plants.

We will combine morphological and physiological measurements (e.g. display size, flower height, petal color) on-site with laboratory analyses of chemical traits (e.g. flower scent, pollen chemistry) and of the genetic structure and diversity of plant populations. Effects of phenotypic variation in traits on pollinator behavior and pollination success will be determined through field observations and a greenhouse experiment. Insights gained from IntraFlor will greatly enhance our understanding of how the interaction between intraspecific variation in floral traits and land-use affect individual pollinator behavior, subsequent interaction patterns and thus the ecosystem function of pollination.

The goal of this project is to identify the effects of land-use intensity on floral trait variability, and the consequent effects on pollinator foraging behaviour and pollination success. Therefore, the results are divided in two parts:

1. Effects of land-use on floral traits and their variation
Our results show that the flowering community, shaped by land-use intensity, has a significant influence on flower traits and their variability, in addition to land-use components. Interestingly, our two focal species responded differently to inter- and intraspecific competition from the surrounding plant community. Fertilization and grazing intensity correlated directly with flower morphology, which in turn correlated with flower traits (Fig. 1). Ranunculus acris responded strongly to interspecific competition: in areas with high overall flower density, R. acris flowers were larger with lower pollen and nectar sugar content (Fig. 1). In contrast, for Trifolium pratense, an increased relative abundance of T. pratense led to larger flowers with higher pollen and nectar sugar content (Fig. 1), presumably because it is a legume and can fix nitrogen, thus balancing nutrient competition with the surrounding plant community. Our results thus show that different species use different strategies to be successful in areas with high inter- or intraspecific competition.

The variability of flower traits changed significantly with the surrounding plant community, but less so with land-use variables. The variation in nectar sugar content decreased with increasing flower density in R. acris, while it also decreased in plots with a high relative abundance of T. pratense. This again suggests that T. pratense is more honest with potential pollinators than R. acris. We also observed that T. pratense exhibited lower color variability in plots with low relative abundance (Fig. 2).

2. Effects of land-use on pollinator foraging and pollination success

We investigated changes in the foraging behavior of bumblebees, particularly on our focal species Trifolium pratense, as a function of land-use intensity. The pollen composition in the pollen baskets of bumblebees, analyzed using metabarcoding, differed significantly between the various land-use intensity categories (Fig. 3).

Plots with high flower density showed higher bumblebee species richness (Fig. 4a). Although the proportion of T. pratense pollen in bumblebee pollen baskets increased in areas with a high relative abundance of T. pratense (Fig. 4a), this had no significant effect on the number of seeds per inflorescence. However, the number of seeds
increased with the overall flower density (Fig. 4c), suggesting that pollination is facilitated in areas with high flower density.

Overall, our results show that different species exhibit different strategies under varying intensities of land use. The plant community shaped by land use influences flower characteristics and their variability, as well as the pollen collected by bumblebees on the grassland plots, and likely have an indirect effect on seed set.