Status of the Northern Population of the Copper-bellied Watersnake, Nerodia erythrogaster neglecta, The

Northeastern Naturalist, 2009 by Attum, Omar, Lee, Yu Man, Kingsbury, Bruce A

Abstract

The Nerodia erythrogaster neglecta (Copper-bellied Watersnake) population north of 40 degrees north latitude is categorized as threatened under the US Endangered Species Act. Previous efforts to use mark-recapture to estimate population size have been unsuccessful because there were so few initial captures. We therefore evaluated the status of the largest presumed subpopulation of this population by using strip transects to provide a conservative (maximum) estimate of density and population size of adults. Our results indicate that the largest presumed adult Copper-bellied Watersnake metapopulation north of 40 degrees north latitude is extremely small, with a total size of 94 � 22 adults, and is characterized by a population density of 1.17 � 0.27 adult snakes/ha (95 % CI = 0.73-1.87 snake/ha). This Copper-bellied Watersnake population appears to be in danger of extirpation within the foreseeable future. Recovery will require protecting existing wetland complexes and restoring wetlands and the surrounding upland habitats.

Introduction

Nerodia erythrogaster neglecta Conant (Copper-bellied Watersnake) is a subspecies of Nerodia erythrogaster (Forster in Bossu) (Plain-bellied Watersnake) (Conant 1 949). The Copper-bellied Watersnake has been in decline across its range for some time (USFWS 1997), and populations now persist only in isolated pockets. Populations in Indiana, Michigan, and Ohio north of 40 degrees latitude are listed as threatened under the US Endangered Species Act, and these occur mostly in a small area that borders northwestern Ohio and adjacent southern Michigan. To the best of our knowledge, there are no published size estimates for these metapopulations. Efforts to estimate population size using mark-recapture have failed because of a paucity of captured snakes (BA. Kingsbury, unpubl. data). We therefore used strip transects to attempt to provide a conservative estimate of the density and population size of the largest Copper-bellied Watersnake (referred to only as watersnake later herein) metapopulation, which is believed to contain most of the individuals of this watersnake north of 40 degrees north latitude.

Materials and Methods

This study was conducted in northwestern Ohio and adjacent southern Michigan on public and private lands that are believed to contain the largest metapopulation of watersnakes (B.A. Kingsbury, unpubl. data). The study area contains a variety of wetland sizes and types, including both ephemeral wetlands and larger, permanent wetlands that had often been artificially created and have water levels controlled for game fishes and birds. These wetlands are fed by spring, river, lake, and local runoff. The St. Joseph River flows through the study site. The upland landscape consists of a matrix of forested, shrub-scrub, old-field, agricultural, and residential habitats and roads.

We sampled 34 wetlands, which ranged in size from 0.02-5.26 ha and totaled 34.99 ha surface area. We treated each wetland shoreline as a transect by walking along the shoreline edge or wading through water when necessary for the entire length of each wetland (shoreline length range = 57-2357 m; total shoreline length surveyed = 20,366 m). The transects were 2 m wide, with 0.5 m occurring on the shore and 1.5 m in the water. Each wetland was surveyed four times between 26 April-13 June 2006, when snakes were most likely to be observed because of little vegetation growth and conspicuous basking. We only observed adults during the surveys and therefore our estimates do not include neonates or juveniles.

We assumed that animals on the transect were always detected because the observer was presumed to eventually flush snakes near the shoreline while walking the transect. For one wetland, we only surveyed 83 m of 373 m of shoreline because of dense vegetation and deeper water.

Wetland size and shoreline length were obtained from past studies (Attum et al. 2008; Roe et al. 2003, 2004). The surveyed wetlands included all the areas known to have the highest watersnake densities, based on previous surveys and telemetry studies (Attum et al. 2008; Roe et al. 2003, 2004). There were an additional 29 ha of low-use or marginal wetland habitat that we did not survey due to time-limit and additional man-power constraints (Attum et al. 2008; Roe et al. 2003, 2004)

Several localities, of an unknown total number of wetland ha, where the watersnake is known to have occurred north of 40 degrees latitude were not surveyed as these localities are mostly represented by historic records, with no recent observations (B.A. Kingsbury, unpubl. data). We believe including these areas in our estimates would have increased population size by only a few individuals and would have decreased density estimates.

We estimated population size and density from the strip-transect data using DISTANCE 5.0 software (Olson et al. 2005, Thomas et al. 2006). In order to use DISTANCE 5.0 to estimate density with strip-transect data, we set the number of intervals to 1 and then set the interval to half the total width (1 m) of the 2-m wide strip transect. We pooled the data for our analysis, but used the multiplier function in DISTANCE to divide the estimates by four to account for multiple surveys.