Modeling the Habitat Range of Phototrophs in Yellowstone National Park: Toward the Development of a Comprehensive Fitness Landscape

The extent to which geochemical variation shapes the distribution of phototrophic metabolisms was modeled based on 439 observations in geothermal springs in Yellowstone National Park (YNP), Wyoming. Generalized additive models (GAMs) were developed to predict the distribution of phototrophic metabolism as a function of spring temperature, pH, and total sulfide. GAMs comprised of temperature explained 38.8% of the variation in the distribution of phototrophic metabolism, whereas GAMs comprised of sulfide and pH explained 19.6 and 11.2% of the variation, respectively. These results suggest that of the measured variables, temperature is the primary constraint on the distribution of phototrophs in YNP. GAMs comprised of multiple variables explained a larger percentage of the variation in the distribution of phototrophic metabolism, indicating additive interactions among variables. A GAM that combined temperature and sulfide explained the greatest variation in the dataset (53.4%) while minimizing the introduction of degrees of freedom. In an effort to verify the extent to which phototroph distribution reflects constraints on activity, we examined the influence of sulfide and temperature on dissolved inorganic carbon (DIC) uptake rates under both light and dark conditions. Light-driven DIC uptake decreased systematically with increasing concentrations of sulfide in acidic, algal-dominated systems, but was unaffected in alkaline, cyanobacterial-dominated systems. In both alkaline and acidic systems, light-driven DIC uptake was suppressed in cultures incubated at temperatures 10°C greater than their in situ temperature. Collectively, these quantitative results indicate that apart from light availability, the habitat range of phototrophs in YNP springs is defined largely by constraints imposed firstly by temperature and secondly by sulfide on the activity of these populations that inhabit the edges of the habitat range. These findings are consistent with the predictions from GAMs and provide a quantitative framework from which to translate distributional patterns into fitness landscapes for use in interpreting the environmental constraints that have shaped the evolution of this process through Earth history.