Mechanisms affecting species coexistence in fluctuating environments can be classified into two groups: the temporal storage effect and relative nonlinearity of competition. Most empirical studies have focused on identifying these mechanisms in isolation, even though they are often likely to operate in unison. We studied the two mechanisms together using nectar-colonizing yeasts as a model system. In floral nectar, yeasts inhabit a dynamic environment where high variability in carbon and nitrogen concentrations causes them to experience large differences in osmotic pressure and resource availability as they disperse from flower to flower. Given trade-offs in tolerance of osmotic pressure, and saturating responses to nitrogen availability, the two mechanisms are both expected to affect coexistence. To test this hypothesis, we first quantified the growth responses of the five most common species of yeast isolated from a hummingbird-pollinated monkeyflower, Mimulus aurantiacus, to different combinations of sucrose and amino acids in nectar. We then fitted response curves to parameterise and simulate models of resource competition under sucrose and amino acid variability. We used the simulation results to make predictions on competitive outcomes and to quantify the combined influence of the temporal storage effect and nonlinearity on invader growth rate.
We found that the five species (Candida rancensis, Metschnikowia gruesii, M. koreensis, M. reukaufii, and Starmerella bombicola) varied in their responses surfaces (sucrose x amino acids), but also had a strong competitive hierarchy. Two distinct genotypes of the dominant species, M. reukaufii, maintained higher growth than other species across a broad spectrum of conditions. Despite the competitive hierarchy, simulation results indicated that variability in sucrose and amino acids concentrations across flowers would allow several pairs of species to coexist via the combined influence of a temporal storage effect and relative non-linearity. Competition experiments are currently underway to test this prediction. Overall our results illustrate the importance of studying multiple coexistence-affecting mechanisms in concert.