Developing a multiple spatial scale model to predict the distribution of Oklahoma's freshwater mussel assemblages with an emphasis on the small rivers of Southeastern Oklahoma.

The Cooperator Science Series was initiated in 2013. Its purpose is to facilitate the archiving and retrieval of research project reports resulting primarily from investigations supported by the U.S. Fish and Wildlife Service (FWS), particularly the Wildlife and Sport Fish Restoration Program. The online format was selected to provide immediate access to science reports for FWS, state and tribal management agencies, the conservation community, and the public at large. Freshwater mussels represent an imperiled taxa worldwide. The National Strategy for the Conservation of Native Freshwater Mussels (1998) identified ten concerns related to the conservation of freshwater mussels including: 1) increasing knowledge of habitat suitability, 2) identifying specific mussel relocation and introduction sites, and 3) evaluating distributions and population dynamics of species. The objectives of this project were 1) determine the distribution and presence of mussel beds in the Muddy and Clear Boggy rivers, and 2) determine environmental factors at multiple spatial scales related to the distribution of a subset of individual mussel species found in the Muddy and Clear Boggy rivers. Because the Muddy and Clear Boggy rivers are deep and turbid lotic systems, the first step was to assess a method (sidescan sonar) to assist in identifying the location of mussel beds in hazardous (e.g., deep, turbid, and with a large influx of woody debris) portions of the rivers. The validation confirmed that ~60% of the sites had mussel beds and ~80% had some mussels or shells present. Water depth was significantly related to our ability to predict mussel-bed locations: predictive ability was greatest at depths of 1–2 m, but decreased in water > 2-m deep suggesting use of a tow fish would improve the likelihood of detection. We determined that sidescan sonar was an effective tool for preliminary assessments of mussel presence during times when they are located at or above the substrate surface and in relatively fine substrates excluding fine silt. Sidescan sonar data were then combined with traditional survey data to build models predicting mussel-bed locations, species presence, and species densities. Generally, we found our predictions of mussel-bed locations and species presence to be inadequate, probably due to the dispersed nature of beds in both rivers. The only exception was the model predicting Wabash pigtoe Fusconaia flava presence (pseudo R2 = 0.36). Densities of mussel species with few host fishes showed significant relationships with increasing densities of their respective host fishes (bleufer Potamilus purpuratus, R2 = 0.41; fragile papershell Leptodea fragilis R2 = 0.46). Models predicting freshwater mussel densities performed much better than presence models with drainage area, width:depth ratios, and % of shale geology selected most often as influential variables. Models predicting Wabash pigtoe densities suggested, unlike most species, this species was quite tolerant of modified land use as densities were exceptionally high in areas with relatively high percent agriculture and pasture land use (66%-73%). Finally, we assessed the movements of different mussel species to provide some insight into how some of these species are able to make movements in response to some forms of environmental perturbation. Several mussel species (bleufer, Wabash pigtoe, threeridge- Amblema plicata, fragile papershell, and yellow sandshell- Lampsilis teres) were tagged in summer and autumn 2012. External tagging via PIT tags occurred at four sites and was initiated during the baseflow period (August-September 2012). Movement did not differ by species but did with time, with individuals moving the most during the early portion of the reproductive period (March, ~ 4 meters). Only fragile papershell showed a positive relationship between movement and discharge (Adj R2=0.27, P=0.06) but this relationship was only found at one site (CB5). This relationship is likely the result of low sample size and the significant influence of one data point suggesting it should be interpreted with caution. The results of our study provide information on what environmental factors are most likely to influence species densities, which can guide conservation initiatives. This research can help managers decide what areas or species may be most suitable for reintroductions and where and what improvements can be made to the landscape to benefit mussel conservation. Other keywords: modelling; Muddy River; Clear Boggy River