Study Species and Field Sampling
The four species of Clarkia in this study are sympatric in the Kern River Canyon in Kern County, California. To determine if pollinator foraging behaviors are reflected in the contents of their pollen balls, we collected bees visiting Clarkia from May-June of 2014. We sampled bees in 14 Clarkia communities throughout the four species’ range of sympatry (Figure 1). Clarkia communities varied in Clarkia species richness, and contained either one, two, or four species of Clarkia (Table 1). In each community, we placed four, 20m transects through patches of Clarkia . We sampled all transects in all communities between 15 May and 15 June. Each community was sampled twice on different days: once in the morning (between 8AM and 12PM) and once in the afternoon (between 1PM and 3:30PM). Sampling entailed walking along each transect for 20 minutes and catching bees using a sweep net when they landed on Clarkia. Bees were sacrificed using ammonium carbonate, and we noted the location, date, and Clarkia species bees were visiting when we caught them. We stored, pinned, and identified bee samples to species (or in the absence of species-level resolution, to genus) using Michener et al. (1994). We then scraped the pollen contents off of all collected bees and stored each pollen ball in 90% ethanol in centrifuge tubes at -20°C.
At the end of every netting period, we surveyed Clarkia floral abundance to estimate the relative abundance of each species. To do so, we placed ½m2 quadrats every four meters on either side of transects and counted all open flowers inside the quadrats. Relative abundance of each species was calculated as the proportion of the number of flowers that were open, divided by the total number of flowers we counted at the survey time.
In the summers of 2015 and 2016, we also collected pollen from each species of Clarkia for use in testing our methodological design. To do so, we collected mature anthers from all four species ofClarkia in various communities throughout their range of overlap. We removed pollen from the anthers and stored them in the same manner that we stored pollen ball samples.