Does similarity in floral morphology, volatiles, or color explain floral visitor communities better than phylogenetic relatedness? A test with naïve pollinators at a common garden outside of the native range of Abronia spp

Eric LoPresti, Plant Biology, Oklahoma State University, Stillwater, OK and Marjorie Weber, Michigan State University, East Lansing, MI

Background/Question/Methods
Suites of morphological, color, and volatile chemical traits of flowers play a large role in attraction and repellence of clades of floral visitors; the repeated selection of similar suites of morphological traits by taxonomically-similar pollinators underlies the controversial concept of “pollination syndromes”. A strong way to test the pollination syndrome concept is to examine interactions occurring with floral visitor communities which are naïve, yet taxonomically-similar to those in native ranges. Over the past few years, we have characterized the floral morphology, color, and volatile profiles of nearly all sand verbenas (Abronia spp.: Nyctaginaceae), a small but ecologically- and morphologically-diverse clade of Western US wildflowers. In this experiment, I planted a common garden of sand verbenas in Michigan, outside of their native range, but with similar floral visitor taxa present and conducted floral visitor observations as well as pollination effectiveness assays.
Results/Conclusions
I found that similarity of broad suites of traits did not predict floral visitor communities well, but neither did phylogenetic similarity. Instead, certain traits (floral number per inflorescence, floral height) had much larger effects on floral visitor communities than other traits that we hypothesized would have large effects (volatile abundance or composition). I then qualitatively compared this to floral visitor communities in native ranges, where data was available, and found that these were very similar, suggesting the traits function similarly across contexts. The pollination success data was idiosyncratic and not highly predictable from overall visitor communities or pollination in native ranges. This probably indicates that pollination in the common garden was done by few species, a hypothesis which I could not strongly test. Recent experimental evolution research has found that selection by pollinators often causes rapid changes in single traits. This study complements those findings by looking across the evolutionary history, morphological, chemical, and color diversity in a small clade and finding similar conclusions that few traits drive floral visitor communities, and that these communities are not predictable solely based on evolutionary relatedness.