Winterkill in Lotic Systems May Be an Important Driver of Amphibian Population Declines
Studies of frogs overwintering in ice-covered ponds and lakes have documented large winterkills, suggesting that winter can be a time of great stress and may be a serious threat to rare and endangered frog species that overwinter in these systems. Lotic systems are assumed to be better oxygenated and therefore less subject to overwinter mortality than lentic systems, but few studies of the winter ecology of ranids in stream ecosystems exist. We investigated the habitat use and survival of Columbia Spotted Frogs (Rana luteiventris) overwintering in a stream system in the Toiyabe Mountains of Nevada. We radio-tracked 13 Spotted Frogs from October 2000 to March 2001 to locate hibernacula and document winter habitat use. During the winter, frogs moved up to 191 m under the ice, often upstream against the current. We marked 1,763 frogs from 2000 to 2001 and found evidence of winterkill (population decreases of 66% to 86.5%) at four out of five sites while the fifth site showed no evidence of winterkill. This corresponded to an actual observation of winterkill of 88% of the frogs at one of the sites, making this one of a few studies directly tying observed wintertime mortality to mark–recapture estimates of decreases in ranid populations and the first to do so for amphibians overwintering in a lotic system. Amphibian winter ecology in lotic systems, the severity and frequency of winterkill events in these systems, and their effect on population size and structure is in need of further study over a wide range of species and should be considered in conservation plans for ranids in lotic systems.
Location of the study sites in the Toiyabe Mountains of Nevada, USA. The black dots in Warner's Reach, Warner's Seep, Jamie's Reach, and Pasture C are all sites where Rana luteiventris were caught and then released with implanted radio transmitters. Mark–recapture work was conducted at Jamie's Reach, Pastures B, C, D, Ledbetter Meadows, Cow Canyon Creek (marked by an “X”), and Farrington Ranch (not indicated).
Frog movement during the 2000–2001 winter season at Warner's Reach (A), Jamie's Reach (B), and Warner's Seep (C). Movements were measured weekly, though frogs did not necessarily move every week. Note that the scales differ in the three panels. The stream flows from south to north (bottom to top). In (C), the dashed line indicates the boundary between wet and dry land.
November–March movements of individual frogs in Warner's Seep (A), Warner's Reach (B), Pasture C (C), and Jamie's Reach (D). Positive bars indicate the distance moved upstream (south) since the last date the frog was located. Negative bars indicate the distance moved downstream (north) since the last date the frog was located. White bars indicate movement in or through open water. Black bars indicate movement in or through ice-covered water.
Population estimates from surveys of Jamie's Reach (A), Pastures B, C, and D combined (B), Ledbetter Meadows (C), Farrington Ranch (D), and Cow Canyon Creek (E). Since Pastures B, C, and D neighbor each other and are therefore not independent, we felt it would be best to present the population estimates of these three pastures as a single unit. Standard error bars and mean population estimates calculated using program JOLLY (see Pollack, 1990). Where standard error bars are not given, the program was unable to compute the standard error. The dates of each sampling period are given. White bars are results of surveys done in 2000. Shaded bars are results of surveys done in 2001.
Temperatures at locations of frogs with transmitters at Warner's Reach (A), Warner's Seep (B), Pasture C (C), and Jamie's Reach (D). Each symbol represents an average of measurements taken by four (Warner's Reach), three (Warner's Seep and Jamie's Reach), or two (Pasture C) frogs during the course of locating the frogs via telemetry (between 0930 h and 1530 h). Measurements were taken of the air temperature at 1 m and 1 cm above the water near telemetered frogs, and also 1 cm above the surface of the substrate. October–May monthly air temperature averages (E). The top pair of lines represent the monthly maximum temperatures, the middle lines the monthly average temperatures, and the bottom pair represent the minimum temperature for the month. Dashed lines indicate the average for each month over the years 1994–2006. The solid lines represent these months for 2000–2001 winter. Measurements were taken at the Washington, NV Weather Station, 47 km southwest from the study site. October–May total liquid precipitation (F). The dashed line indicates the average for each month over the years 1994–2006. The solid line represents these months for 2000–2001 winter. Measurements were taken at the Washington, NV Weather Station, 47 km southwest from the study site.
Contributor Notes
Associate Editor: M. J. Lannoo.