Extended Data Table 3 Effects of plant competitors for pollinators at multiple spatial scales on pollen limitation
From: Competition for pollinators destabilizes plant coexistence
- Two-sided analysis of variance (ANOVA) tests with background competitor species as a fixed effect (with d.f. = 5) show that pollen limitation (quantified by the log-ratio of per germinant seed production under ambient pollination over per germinant seed production with pollen supplementation) differed significantly as a function of the identity of the competitor species within the same plot as the focal individuals for all species except for self-compatible B. arvensis. This “within-plot” effect indicates that plant neighbors within a plot affected the pollen limitation of focal plants. However, plants in neighboring plots may also attract pollinators away from focal individuals, effectively competing for pollinators at larger spatial scales. Moreover, the background species in the array of neighbouring plots differed between individual plots due to the random assignment of competitor identity to plot. Multiple linear regression analyses with the number of surrounding plots with each background species as predictor variables (two-sided; each with 1 d.f.), however, showed that pollen limitation was not significantly affected by the identity of the competitors for pollinators at any spatial scale in our study larger than the focal plot (i.e., within an n x n plot grid centered on the focal plot, where n = 3, 5, 7). In addition, two-sided ANOVA tests show that pollen limitation was not significantly affected by whether focal plants were in plots at the edge or interior of our plot array (a fixed effect with d.f. = 1), suggesting that pollinators were not, for example, favoring plants in plots nearest to the surrounding matrix. Together, these analyses suggest that while pollinators must be making foraging decisions at larger spatial scales than those of our plots, pollinator decisions about which plants to visit within a plot most strongly determined plant performance.
