Pollinator preference has been shown to maintain flower color polymorphism in many species. However, pollinators are not always the sole agent of selection on floral traits. In particular, indirect selection may be important if pigments that color flowers or their biochemical precursors have other functional roles. Flavonoids are a ubiquitous class of plant secondary metabolites that affect pollination, herbivory, and tolerance of UV and heat stress. The relative importance of these plant functions varies among populations. This study explored the ecological importance of traits related to flavonoid biosynthesis in the monocot Tradescantia ohiensis. The following hypotheses were tested: 1) Pollinators prefer pigmented (violet) flowers, and preference is site-dependent. 2) Populations with white flower morphs differ abiotically from populations in which all flowers are pigmented.
Results/Conclusions
To address the first hypothesis, pollinator observations were made in experimental arrays of potted plants at two sites. The pollinator community differed between high and low elevation sites, with native bumble bees (Bombus spp.) at the high elevation replaced by honeybees (Apis spp.) at the low elevation. Solitary bees of the family Halictidae and hoverflies (Syrphidae) occur at both sites, but the latter act primarily as pollen robbers. While bumble bees and halictids show a slight preference for violet flowers, hoverflies visit white flowers more. To address the second hypothesis, twenty-five natural populations were surveyed for floral morph frequencies, soil characteristics, temperature, and canopy cover. Populations with white flowers were more likely to occur at high elevations where ambient temperatures are lower. Pollinator preference does not explain spatial patterns of polymorphism, therefore these results suggest that pigments are costly and confer a benefit beyond pollinator attraction. Further experiments are needed to address how loss of pigment influences traits related to heat tolerance.
