Hummingbird-pollinated flowers

This system involves the communities of nectar-inhabiting fungi and bacteria that develop in flowers of a hummingbird-pollinated shrub, Mimulus aurantiacus, at the Jasper Ridge Biological Preserve, located about 5 km from Stanford's main campus. This system is ideally suited for studying community assembly for several reasons. First, floral nectar is initially sterile, and microbial species undergo primary succession as various species colonize the initially sterile habitat. Short microbial generation times mean that populations grow quickly in nectar, enabling well-replicated observations of community assembly over many generations. Second, hummingbirds as the microbes' main dispersal vectors enable detailed quantification of immigration history. We are developing methods to ascertain the timings of potential immigration events by monitoring bird visits to flowers using motion-activated cameras (see this Stanford Report article for more detail). Third, because their life is essentially inseparable from the pollinators in terms of immigration to flowers and nectar utilization in flowers, nectar-inhabiting microbes can affect the function of flowers for plant reproduction, potentially resulting in close links between community structure and function.

We have so far found that: (1) microbial species are distributed non-randomly, indicative of dispersal limitation driven by non-random foraging by hummingbirds; (2) microbial communities vary greatly over time as well, with sometimes almost complete species turnover over a flowering season and from one year to another; (3) priority effects are strong enough to affect beta diversity; (4) the magnitude of priority effects are predictable, with phylogenetically closer relatives affecting one another more strongly; (5) much of this variation can be explained by niche differences among species and environmental variability in temperature; and (6) microbial species differ in their effects on nectar chemistry, foraging by hummingbirds and bees, and seed production. Motivated by these findings, we are building mathematical models to predict the effect of climate change on plant-pollinator-microbe interactions. Further, we have de novo sequenced the genome of the dominant yeast, Metschnikowia reukaufii. This analysis has suggested extensive duplication of nitrogen-scavenging genes in this yeast as a basis of priority effects. Overall, knowledge gained through this project has facilitated conceptual synthesis on the mechanisms and conditions for priority effects.

We are continuing to use this system to study the effect of historical contingency on the feedback between community structure and function. As mentioned above, historically induced differences in community structure can influence the flower's functional role by affecting pollinator visits. In turn, changes in pollinator visits can alter microbial immigration history and thus community structure. But how strong is this feedback? How much does it depend on environmental conditions? At what spatial and temporal scales does the feedback happen? These are some of the questions we will investigate over the next several years.

An additional emphasis of this project is integration of research, education and outreach.