Ecological Units and Spatial Pattern in River Ecosystems.

Abstract

Rivers are advective, largely unidirectional ecological networks whose spatial patterning reflects both catchment and network structural characteristics. I used a headwaters-to-mouth, longitudinal, high-frequency-spatial sampling design to facilitate analyses of the extent and variability of spatial patterning in Midwestern river systems and to test alternate hypotheses about the underlying causes of biological spatial autocorrelation. These analyses establish the conceptual validity of channel segment based classifications used in management settings, and provide guidance for appropriate survey sampling design and statistical analyses in river systems.

In the first study I tested the theoretical assumptions underlying the mapping and practical application of riverine ecological units (EUs) within a river mainstem. EUs require concordance between fish and invertebrate assemblage composition and between biological assemblages and environmental variables. Along the Lower Muskegon River mainstem, fish/invertebrate concordances and many environment/biology concordances were strong, resulting in distinct, homogeneous biological assemblages that persisted through time.

In the second study I tested the same theoretical assumptions of EUs in a variety of disjunct river tributary systems in Michigan and Ohio. Although fish/invertebrate and environment/biology concordances were very strong in all of the tributaries, downstream tributary channels with substantial stream flow were the only contiguous stream segments with similar environmental and biological character. This suggests a better understanding of spatial pattern and processes in headwater streams is needed to guide effective EU delineation in tributaries.

In the third study I explored a common feature of spatial patterning, positive spatial autocorrelation (SAC). SAC was common in both environmental variables and fish assemblage composition, although the magnitude of SAC varied by measure and spatial extent. Strong environment/biology associations accounted for most or all of the SAC in biological assemblages, offering strong support for niche processes as the origin of biotic SAC in these river systems. Likewise, proximity effects on biological assemblages were largely mediated through similarity in the environment.

My work here suggests that EUs do provide realistic units to map, inventory, and classify river segments for practical management, and provide a way to abstract and communicate the complex ecological processes and patterns that are characteristic of river ecosystems.