The importance of current velocity, scale, and patchiness for aquatic invertebrate movement and colonization

Abstract

The field of landscape ecology focuses on spatial patterns of heterogeneity and their effects on ecological processes. Another main focus is on scale, and how the interplay between patterns and processes changes with scale. Landscape ecology has traditionally been concerned with terrestrial heterogeneity, particularly how the arrangement of vegetation affects ecological processes. This dissertation, however, focuses on streambed landscapes, which, though largely unexplored, represent an excellent opportunity to study landscape ecology concepts such as heterogeneity, scale hierarchy and the effects of landscape structure at various scales. Chapter 1 uses percolation theory as a framework to examine how the proportion of useable habitat in a streambed landscape interacts with current velocity to affect the movement patterns of two benthic grazers. The physical structure and arrangement of patches in a landscape can influence animal movement. Most movement studies have occurred in terrestrial landscapes, though aquatic landscapes are equally heterogeneous and animals living there encounter patches differing in movement resistance. Furthermore, the pervasive and variable effects of unidirectional flowing water are unique in streambed landscapes, and the constraints imposed on animal movement by this factor are poorly understood. Using percolation theory as a framework, this study asks how the proportion of useable habitat in a streambed landscape interacts with current velocity to affect movement patterns of two benthic grazers that differ in body size and habitat preferences. Ten experimental streambed landscapes of varying proportions of two algal types were constructed: tall, structured filamentous stands engineered by the chironomid larva Pagastia sp., and low, smooth diatom turfs. Movement paths of larval glossosomatid caddisflies (Agapetus boulderensis) and pulmonate snails (Physa sp.) were recorded on these plots at two current velocities, and analyzed their movement patterns using four metrics: net displacement, movement rate, mean vector length, and upstream orientation. Both landscape structure and current velocity affected animal movement. Agapetus net displacement, movement rate, and upstream orientation were significantly related to the proportion habitat, current velocity, and their interaction; mean vector length was affected by proportion habitat and its interaction with current velocity. Specifically, Agapetus traveled farther as smooth habitat increased, but did so only in slow flow conditions. Swifter flows caused slower Agapetus movement using more upstream-oriented paths, yet only in completely smooth landscapes. Agapetus movement paths showed increasing straightness with proportion habitat at both flow levels. Physa movement rate and mean vector length were affected by proportion habitat, current velocity, and their interaction; net displacement was affected by proportion habitat; and upstream orientation was affected by proportion habitat and current velocity. Increasing proportions of smooth habitat allowed Physa to travel farther using more upstream-oriented paths. Faster flows caused Physa to move faster, a pattern demonstrated only in the smoothest landscape. Physa's mean vector length indicated a straighter path at the slow current velocity than at the faster velocity, but again only in the smoothest landscape. Landscape ecology has mainly remained above the water's surface, and this study is a first step towards understanding the interactions between habitat heterogeneity, animal movement, and fluid media. Chapter 2 broadens the extent of Chapter 1, employing invertebrate community measures to focus on three major spatial scales of the study stream. Hierarchical ANOVA and multiple regression with response variables species richness, species diversity, invertebrate abundance, and faunal ash free dry mass (AFDM) were used. These analyses helped determine how the aquatic invertebrate community responds to multiple scales of patchiness and habitat structure, as well as the spatial scales at which landscape heterogeneity is most important to the distribution of these animals. Experiments were performed over approximately one month, with invertebrate measurements taken from observations spaced three times over the course of the experiment, as well as from a final destructive sampling of the study stones. In general, results from the multiple regression indicated that percentage and AFDM of algae were the best predictors of species richness and diversity, invertebrate abundance, and faunal AFDM. Specifically for the multiple regression by time, percent algae was significant at all three time intervals for species richness, species diversity and for time steps 2 and 3 for invertebrate abundance. Other variables in the regression, such as mean depth, were significant for species richness and diversity at time step 1, and also for invertebrate abundance at time step 2. Mean current velocity was significant for species diversity at time step 1, and habitat (riffle vs. run) was significant for invertebrate abundance at time step 3. Results from the final multiple regression based on the destructive sampling indicated similar results: AFDM of algae was a significant predictor of species richness and species diversity, as were habitat and mean current velocity. Invertebrate abundance was significantly affected by mean depth and habitat, while none of the variables measured was a significant predictor of AFDM fauna. Results from the hierarchical ANOVA generally show that of the variables measured, those at the surface scale were the best predictors of species richness, species diversity, and invertebrate abundance. Specifically, in the analysis by time, the finest (surface) scale was significant for species richness, diversity, and invertebrate abundance at all time steps. The intermediate (stone) scale was significant at time step 2 for species richness but not significant for species diversity or invertebrate abundance. The broadest (channel unit) scale was only significant for invertebrate abundance at time step 3. In the final analysis of destructive sampling, the surface scale was significant for species richness, diversity, faunal AFDM and invertebrate abundance. The stone scale was only significant for species richness and diversity, while the channel unit scale only significant for species diversity. This study integrates fine-scale stream ecology with the landscape ecology principle of scale-dependence to indicate the scales most important to benthic invertebrate distribution. While the distribution of invertebrates on the streambed can vary significantly between spatial scales, this study indicates that the fine scale explains the most variation in the response variables.