The roles of spatial scale and landscape change in mediating predator effects on stream fish communities

Abstract

The role of predators in ecosystems has not only intrigued and puzzled ecologists over time, but predators are charismatic icons of conservation whose status indicates threats of global change. Through habitat alteration and fragmentation, climate change, and species introductions, predation pressure has been altered globally through the loss of apex predators, introduction of predators, and changes in predator distributions and abundance. While we know predators can influence ecosystems through top-down processes, managing changes in predation pressure requires quantifying effects of predators at scales relevant to management and conservation. In lotic systems, scales relevant to management often span across drainage basins, so predator effects must be quantified across stream networks. Because lotic communities also respond to landscape change, understanding the role of predators across stream networks requires careful consideration of local and broad scale abiotic factors influencing both predators and prey. I combined simulated, experimental, and observational data to 1) assess sampling strategies to determine effects of landscape change on stream fish communities, 2) measure changes in predator consumption rates across spatial scales and the role of prey behavior in driving scaling relationships, and 3) quantify the relationship between the presence of predators and stream fish community structure while controlling for abiotic variability across stream networks. In chapter 2, I compared how the distribution of sample sites (completely random, highly skewed, or uniform distributions) across landscape gradients influenced variability in measured responses of stream fish community metrics. Strong responses (species richness) to environmental gradients were robust to sample distributions, but large sample size and uniform distributions of samples across gradients were necessary to quantify more complex ecological responses (community composition). In chapter 3, I conducted a mesocosm study to quantify differences in per capita consumption across different arena sizes and measured three aspects of prey behavior hypothesized to be important in driving consumption rates: aggregation, movement, and spatial overlap with predators. Per capita consumption was highest in the largest arena relative to the smallest. I hypothesize the positive relationship between consumption and spatial scale was driven by lower group vigilance because prey aggregated less in large arenas. In chapter 4, I compared fish community structure, including richness and abundance of species, at sites in which a predatory fish, largemouth bass (Micropterus salmoides), were present or absent. I first identified which abiotic factors, including both natural stream attributes and anthropogenic landscape changes, drove the presence of largemouth bass and stream fish community structure. I then controlled for important abiotic factors to determine relationships between largemouth bass and stream fish community structure. Richness was higher than predicted based on abiotic factors at sites where bass were present. Several species associated with small impoundments exhibited significant co-occurrence patterns with largemouth bass, likely driving the heightened richness at sites with bass. Complex ecological phenomena such as community responses to predators are difficult to measures, especially in the context of landscape change. These studies highlight the importance of thoughtful study design, the scale-dependence of biotic interactions, and challenges of quantifying responses to predators at scales relevant to conservation and management.