Summary
From late May to mid-September 1982 we investigated habital selection by black rat snakes (Elaphe o. obsoleta) at the Queen's University Biological Station in eastern Ontario. We implanted radio transmitters in 4 male and 3 female snakes and used their daily positions as habitat sampling points. We also sampled the available habitat using randomly selected points. We located the snakes 472 times (>90% success) which produced 107 habitat sample points. All snake sample points were separated into active or inactive based on the length of time the snake remained in that position (less than or greater than 7 days, respectively). During the bird breeding season black rat snakes showed a preference for field habitat although in both the field and deciduous forest the snake points were significantly clustered along the habitat interface. Following the bird breeding season, field and deciduous forest habitats were used in proportion to their availability and the preference for the ecotone was no longer found in field samples. We found only limited evidence of non-random habitat use within habitats with regard to both plant species composition and vegetation structure. Inactive sites were diverse but all were located on the forest-field interface, had open exposure to direct sun and provided shelter for the snakes. We suggest from these results that ideal habitat for black rat snakes is a small scale mosaic of field and forest and that their disappearance from other parts of their range in Canada may be related to the disappearance of such mosaics due to land clearing for agriculture.
Methodology
We conducted the study at the Queen's University Biological Station in eastern Ontario, near the center of the only Canadian population in which these snakes remain relatively common (Cook, 1977) and representing the northern extreme of the species' overall distribution. The field station property lies on a 7 ha point extending into Lake Opinicon and can be divided into 3 well defined habitats: mixed deciduous for- est, coniferous forest (artificially plant- ed stands of red pine, Pinus resinosa) and early successional old field, some of which is grazed (Fig. 1). Between 15 May and 22 July 1982, we captured 4 male and 3 female black rat snakes in the study area, implanted subcutaneous radio transmitters and re- leased them at their point of capture. The male snakes weighed between 375 g and 650 g, a range which include 53% of all male snakes captured over 3 years (N = 102) and covers the central part of the distribution of weights (P. J. Weatherhead, unpublished data). The female snakes weighed between 410 g and 578 g, a range which includes 39% of all female snakes captured (N = 80) and again covers the central part of the distribution (P. J. Weatherhead, unpublished data). Details of the transmitter design, implantation technique and radio tracking procedure are provided by Weatherhead and Anderka (1984). We attempted to locate each snake every day except during rain. Previous observations indicated that the snakes often remained in the same location for. We attempted to locate each snake every day except during rain. Previous observations indicated that the snakes often remained in the same location for much longer than one day and more frequent sampling would only have in- creased the number of times we found the snakes in the same locations. Each time we located a snake, its position was marked for subsequent habitat characterization. New sample points were marked only when the snake had moved more than 2 m from its last re- corded position. If a snake was in a tree a compass bearing was randomly selected and the sample point positioned 3 m from the tree on that bearing. This was done to keep the base of the tree from being included in the sampling quadrat while allowing us to quantify features of the vegetation proximate to the tree. Fewer than 5% of all snake samples were from such points. If a snake returned to a position that had previously been sampled an additional sample was taken at that point although no point was ever sampled more than twice. We conducted searches for the snakes from the time of implantation through mid-September when the snakes became relatively inactive.