As the geographical centers of riverscapes, rivers support fish populations at local and basin-wide scales. However, refinements to fish sampling protocols and theories underpinning basin-wide community ecology have been slowed by the inherent complexity and immensity of rivers. I conducted three extensive studies in non-wadeable tributaries of the lower Missouri and middle Mississippi rivers (Missouri, USA) to illuminate processes structuring riverine fish communities. In Chapter 2, I tested the efficacy of a six-gear fish community sampling protocol by repeatedly sampling nine sites in spring, 07, and fall (N = 36 surveys). I identified an efficient four-gear sub protocol that consistently detected 90% of observed species richness at sites and only required 52% of initial survey effort. In comparison, an electrofishing-only protocol detected lower percentages of fish richness, varied seasonally between 07 and fall, and was nearly twice as variable. In Chapters 3 and 4, I contrasted fish communities inhabiting the Grand (10 sites, prairie region) and Meramec (12 sites, Ozark region) river systems. Chapter 3 examined tributary use by large-river specialist fishes (LRS), a guild of fishes that likely disperse into tributaries from the Missouri (Grand R.) and Mississippi (Meramec R.) rivers. I tested whether mean annual discharge consistently structured richness of LRS fishes at sites ([alpha] richness) within tributaries, and if habitat and downriver connectivity to the Missouri and Mississippi rivers explained additional variation in LRS [alpha] richness. Although species-discharge relationships were positive, discharge effect sizes varied between rivers revealing discharge did not consistently structure LRS [alpha] richness. After accounting for river-specific effects of discharge, downriver connectivity explained residual variation in LRS [alpha] richness, indicating dispersal into tributaries likely structured LRS [alpha] richness. Consequently, LRS [alpha] richness solely estimated from discharge might be underestimated in connected network branches and overestimated in isolated mainstem reaches. Chapter 4 expanded the research scope beyond LRS species to test whether regional connectivity (distance to dispersal source) or site-level habitat diversity (multivariate dispersion of nine habitat variables) explained α richness of three stream size guilds at sites: LRS species, headwater species likely sourced from [less than or equal to]3rd Strahler order streams, and core riverine fishes. In both river systems, downriver connectivity (distance upriver from mouth of mainstem river) and habitat diversity positively related to LRS- (R2 = 0.44 in Grand R, 0.91 in Meramec R.) and core-species (R2 = 0.37 in Grand R., 0.57 in Meramec R.) richness, respectively. Headwaters within 25 km of sites positively related to headwater species richness in the Grand River system (R2 = 0.85), but not in the Meramec River system where headwater richness was better explained by an inverse relationship with discharge (R2 = 0.32). Increasing LRS richness (13â€"17 spp.) downriver caused sites supporting the highest total species richness ([greater than or equal to]75th percentiles) to skew towards lower-midcourse reaches, rather than in midcourses where core-species richness and habitat diversity peaked. Because riverine fish richness manifests from regional dispersal and local habitat diversity, conserving areas of high richness will likely require management actions aimed at local and regional scales. Altogether these studies collecting 146 species and 145,147 individuals revealed non-wadeable tributaries are key riverscape elements that provide diverse riverine habitats and corridors for members of multiple regional fish species pools.
How many fish? Comparison of two underwater visual sampling methods for monitoring fish communities
