Super abundant, non-native honey bees (Apis mellifera) decrease the fitness of native, coastal sage scrub plants in Southern California

Dillon J Travis, Biological Sciences, University of California San Diego, San Diego, CA and Joshua R. Kohn, Division of Biological Sciences, University of California, San Diego, La Jolla, CA

Background/Question/Methods
San Diego County supports over 600 species of native bees, a variety of other pollinating insects, and the highest plant species richness in the United States. Although the region has many pollinators, studies describing pollination networks in Coastal Sage Scrub habitats have shown that non-native, primarily feral, Africanized honey bees (Apis mellifera) account for ~75% of all floral visits to native plants. Honey bees are even more dominant (>90% of floral visits) on the most abundantly flowering plant species. Because honey bees visit more flowers on a plant before moving to the next plant than do native pollinators, we hypothesized that seeds produced from honey bee pollination might suffer more from inbreeding depression than seeds produced following pollination by native insects. To determine the impact of honey bee pollination on the reproductive fitness of native southern California plants, we exposed unvisited flowers of three species (Salvia apiana, S. mellifera, and Phacelia distans) to 6 pollination treatments: open-pollination (control), cross-pollination, self-pollination, single visit honey bee-pollination, single visit native insect-pollination, and an exclusionary treatment. We collected and quantified seeds from each treatment and grew them in a greenhouse to assess fitness by measuring germination rate, height (biomass proxy), survival, and number of flowers produced. We used multiplicative fitness functions to estimate the fitness of seeds from each treatment.
Results/Conclusions
All three species exhibited high inbreeding depression with self-fertilized seeds having less than half the fitness of outcrossed seeds although the fitness reduction due to selfing was greater for the two perennial Salvia species than for the annual (P. distans). For all three species, the fitness of seeds from honey bee and open-pollination was not different than that following self-pollination. For two of the species (S. apiana and P. distans) pollination by native insects produced seeds whose fitness equaled that of the outcross treatment and was significantly higher than seeds produced from honey bee pollination. For the third species (S. mellifera) the fitness of seeds resulting from pollination by native insects could not be assessed, owing to insufficient visitation by non-honey bees. Our results show that honey bee pollination produces seeds of low fitness, likely due to increased deposition of self-pollen. This result may be general, given the documented foraging behavior of honey bees in many systems. Estimates of the importance of a pollinator in any system may be inaccurate if the fitness of seeds resulting from their visitation is not assessed.