Temporality of species coexistence and bounded population size

Background/Question/Methods

Biodiversity remains a major puzzle to ecologists. While spatial heterogeneity of the environment has been well demonstrated as a critical mechanism to maintain species coexistence, the role played by environmental variability in time is less understood. To fill this gap, many models have been proposed, such as storage effect and nonlinear competitive interactions. A common feature to all, frequency-dependent growth rate is required for species coexistence. Here, I ask if there exists any condition inherent to the structure of ecosystems with temporal variability that can support coexistence in the absence of frequency-dependent growth. A simple theory was first developed, based on invasibility criteria and accounting finite population size limited by resource. It predicts that two competing species can coexist if the arithmetic mean relative fitness is larger than one and the harmonic mean relative fitness is less than one. For empirical test, a microcosm was engineered. Two strains of Escherichia coli, one resistant to chloramphenicol and the other to tetracycline, were competed in a serial transfer of batch cultures whereby the two antibiotics were alternatively present in sub-inhibitory dosage. The competition dynamics, monitored with flow cytometry, provided direct test of the theory.

Results/Conclusions

When transfers (dilution of the growing culture into a fresh medium with the alternative antibiotic) occurred once a fixed population size was reached, the competing strains coexisted stably. In contrast, one competitor drove the other to extinction if transfers occurred at a fixed time interval. Calculations based on the experimental data were consistent with predictions of the theory. Intuitively, the fixed population size serves as a balancing mechanism between the competitors. Consider the tetracycline resistant strain. In a batch where tetracycline is present, it is the superior competitor with the maximum growth rate. However, the faster it grows, the faster it drives the population to reach the limit, and the less time it spends in this favorable environment. In a batch where chloramphenicol is present, the strain now becomes the inferior competitor with retarded growth rate. This slow growth allows the population to stay in the current batch longer and thus experience more propagation before transferring to the next. In conclusion, while current ecological models of temporal coexistence are largely focused on complex interactions between organisms, e.g, frequency-dependent growth, there exist mechanisms ingrained to the abiotic structure of ecosystems that promote diversity. This work identifies bounded population size as such mechanism and demonstrates with experiments.