Cornell University Ithaca, New York, United States
Abstract: Plant-plant facilitation is an important driver of local scale interactions, and work is beginning to test if such positive interactions can predict larger scale community patterns. Pollinator facilitation, the increase in pollinator visits due to the presence of attractive neighboring plants, is prevalent in nature and can substantially influence plant fitness, but the extent to which it can predict species co-occurrence remains an open question. Here we employed two Bidens (Asteraceae) species with dramatically divergent floral displays, B. cernua and B. frondosa, to test if facilitation occurs and is concordant with co-occurrence between the congeners. Whereas B. cernua has a showy, sunflower-like inflorescence (radiate), B. frondosa has entirely lost the petal-like ray florets (discoid). In growth chamber experiments we found that despite extensive selfing, both Bidens species yielded increased numbers of viable seeds when supplemented with hand pollination, indicating that animal pollination could confer fitness benefits. Through field observations of 27 natural communities, common gardens, and transect sampling, we tested the hypotheses that 1) showy Bidens facilitates pollination of its unshowy congener, 2) increased pollination leads to increased fitness, and 3) this positive interaction can explain congener co-occurrence at the landscape-level. In natural populations, co-occurrence with showy B. cernua increased visitation to unshowy B. frondosa flowers by nearly 60% and increased lifetime reproductive success by more than 96%. Pollinator facilitation was again found in common gardens, albeit to a lower magnitude (25%). In common gardens co-occurrence did not affect B. frondosa’s reproductive success. Across 147 sites in New York State, we found that radiate and discoid Bidens spp. strongly aggregate despite the overall pattern for Bidens spp. to be spatially dis-aggregated. Our results demonstrate that a single facilitative interaction, mediated by divergence in floral display, can explain co-occurrence among congeners across the landscape. By explicitly testing an indirect facilitative mechanism and using it to predict species co-occurrence, our study underscores the importance of positive interactions, even among close relatives, across scales in ecology.
