Consistent shifts in pollinator-relevant floral coloration along Rocky Mountain elevation gradients

• Published in: The Journal of ecology 2018-09, Vol.106 (5), p.1910-1924
• Main Authors: Gray, Michelle, Stansberry, Marcus J., Lynn, Joshua S., Williams, Charles F., White, Thomas E., Whitney, Kenneth D.
• Format: Article
• Language: English
• Subjects: bee vision; Brightness; Color; Coloration; Colour; Communities; ecological filtering; Elevation; Flowers; fly vision; Geography; Gradients; Interactions; Landscape; Landscape ecology; Low temperature; Mountains; natural selection; Plant reproduction; plant–pollinator interactions; Pollination; Pollinators; Reflectance; reflectance spectra; Saturation; Saturation (color); signalling; Species; Spectrophotometers; ultraviolet; Ultraviolet radiation; Visual flight; Visual perception; Wavelengths
• ISSN: 0022-0477; 1365-2745
• DOI: 10.1111/1365-2745.12948

1. Floral colour is a fundamental signal that shapes plant-pollinator interactions. Despite theoretical reasons why floral colours might shift in representation along biotic and abiotic gradients, few studies have examined community-level shifts in colour, and even fewer significant patterns have been detected. 2. We examined floral colour on six replicated transects spanning 1,300 m in the Rocky Mountains of Colorado, USA. Along these transects, there is a hypothesized shift from bee-dominated to fly-dominated pollination with increasing elevation. The reflectance of flowers of 110 forb and shrub species was measured using a spectrophotometer, and was used to estimate three components of colour (hue, saturation and brightness) in relevant pollinator visual spaces. Percent cover data were collected from 67 sites and used to obtain community-weighted mean (cwm) estimates of floral colour. 3. We found strong patterns of elevational change in floral colour. Reflectancecwm of shorter wavelengths (UVB through human blue, 300-500 nm) generally decreased linearly with elevation, while reflectancecwm of longer wavelengths (human green through red, 500-700 nm) showed hump-shaped patterns with highest reflectance at intermediate elevations. With respect to pollinators, saturationcwm increased significantly with elevation in both bee and fly visual spaces, while brightness contrastcwm showed a hump-shaped pattern in bee space and a decline with elevation in fly visual space. For hue, cover of species perceived as bee-blue declined with elevation, while cover of bee-UV-green species showed a hump-shaped pattern. In comparison, we detected no elevational shifts in floral hues as perceived by flies. 4. Synthesis. Hue patterns are consistent with the hypothesis that bee pollinators have shaped the geography of floral colour. The roles of fly pollinators and of abiotic drivers are more difficult to infer, although the drop in floral brightness at high elevations is consistent with predictions that low temperatures and more intense ultraviolet radiation should favour increased pigment concentrations there. Our results indicate that floral colour can be dynamic yet predictable across the landscape, a pattern that provides opportunities to tease apart the ecological and evolutionary drivers of this important plant trait.