Turtle communities in the Upper Mississippi River System

The fisheries component of the Long-Term Resource Monitoring Program (LTRMP), which utilizes many sorts of collecting methods, began sampling in 1989. Beginning in 1992, fisheries component specialists at each of six field stations began to systematically record data on turtles collected as a by-product of fisheries monitoring. Herein we summarize data collected on turtle communities in the UMRS from 1992 to 1995. Our analysis is important because it is the first based on long-term collections that encompass a wide geographic area made in any large river. Overall, we caught 4,414 individual turtles of 10 species from all reaches and in all years. Species of chelonians collected included the common map turtle (Graptemys geographica), the false map turtle (G. pseudogeographica), the Ouachita map turtle (G. ouachitensis), the painted turtle (Chrysemys picta), the red-eared slider (Trachemys scripta elegans), the river cooter (Pseudemys concinna), the common snapping turtle (Chelydra serpentina), the smooth softshell (Apalone mutica), the spiny softshell (A. spinifera), and the stinkpot (Sternotherus odoratus). The number of species collected tended to increase from north to south. We found the southern reaches to contain more species, but with reduced diversity indices due to decreased evenness in the sample and with increased density as measured by catch per unit (CPU) effort. Apparently, increased density accrued from increased number of individuals from one or two dominant species present rather than a general increase in all species present. Species diversity as measured by number of species collected did not vary with habitat type. We sampled tributaries, backwaters, impoundments, channel borders, side channels, and tailwater zones. Habitat was an important determinant in the sorts and proportions of turtle species collected. Backwaters and impoundments were dominated by C. picta and T. s. elegans. In contrast, species of Graptemys dominated main channel border and side channel borders. Tailwater habitats differed from other habitats in species composition due to the importance of A. spinifera. Species composition within particular habitats was strongly affected by sampling location within the system. Among all the gear types used by fisheries components at the LTRMP field stations, fyke nets were by far the most effective at catching turtles. This gear type accounted for 65% of all turtles collected. Temporal variation accounted for a significant amount of variation in carapace length for some species. Turtles collected earlier in the year (i.e., June, July, or August) tended to be larger than those collected later in the year (i.e., September or October). Overall our models on the influence of variables on turtle size accounted for 28 to 48% of the variance excepting the A. spinifera model, which accounted for 82% of the variance in carapace length. Sexual size dimorphism was found for many of the species collected. In such cases males were smaller than females. Dimorphism was most pronounced in the two softshell species and least pronounced for the Stinkpot and common snapping turtle. Navigation dams such as those found in the UMRS create impoundments that have different ecological features than the more natural habitat types such as backwaters, channel borders, etc. We found effects that may be attributable to impoundment. We found no effect on the number of species in impoundments as all major habitats for all reaches combined contained nine species. However, species diversity indices were greater for backwater habitats than impoundments. The difference seems to be due to greater unevenness in impoundments where one or two species dominate collections. Despite this, the number of individuals collected in impoundments was fewer than those collected in backwater habitats. This suggests that protection of backwater habitats from alteration and maintaining their riverine nature is important in maintaining species diversity and density among aquatic turtles in the UMRS.