Summary
Measuring habitat suitability is important in conservation and in wildlife management. Measuring the abundance or presence–absence of a species in various habitats is not sufficient to measure habitat suitability because these metrics can be poor predictors of population success. Therefore, having some measure of population success is essential in assessing habitat suitability, but estimating population success is difficult. Identifying suitable proxies for population success could thus be beneficial. We examined whether faecal corticosterone metabolite (fCM) concentrations could be used as a proxy for habitat suitability in common gartersnakes (Thamnophis sirtalis). We conducted a validation study and confirmed that fCM concentrations indeed reflect circulating corticosterone concentrations. We estimated abundance, reproductive output and growth rate of gartersnakes in field and in forest habitat and we also measured fCM concentrations of gartersnakes from these same habitats. Common gartersnakes were more abundant and had higher reproductive outputs and higher growth rates in field habitat than in forest habitat, but fCM concentrations did not differ between the same two habitats. Our results suggest either that fCM concentrations are not a useful metric of habitat suitability in common gartersnakes or that the difference in suitability between the two habitats was too small to induce changes in fCM concentrations. Incorporating fitness metrics in estimates of habitat suitability is important, but these metrics of fitness have to be sensitive enough to vary between habitats.
Methodology
Abundance estimates
We monitored the habitat use of common gartersnakes between field and forest habitats at Queen's University Biological Station, Ontario, Canada (44°33′N, 76°21′W). We set up five pairs of 2500 m2 plots in adjacent fields and forests and placed eight plywood coverboards (360 cm2) on each plot (Halliday and Blouin-Demers, 2015). We systematically walked back and forth across each plot and checked under all of the coverboards twice per day over three consecutive days every 2 weeks from 5 May to 16 July 2013. We uniquely marked each individual gartersnake using ventral scale branding (Winne et al., 2006). We counted the number of different individual gartersnakes caught in each field and forest grid throughout the study and compared abundance in field vs. (adjacent) forest habitat using a paired t test in R (package, stats; function, t.test; R Core Team, 2014).
Fitness and faecal corticosterone “metabolite estimates
We examined the growth rate, reproductive output and fCM concentrations of 20 female common gartersnakes in enclosures in Pontiac County, Québec, Canada (45°29′N, 75°55′W). We built 2.67 m × 2.67 m × 1.33 m frames out of lumber and created walls for the enclosure by attaching polyethylene vapour barrier to the frames. We buried the bottom 10 cm of the walls of the enclosures in the ground to prevent snakes from escaping. We built six of these enclosures in field habitat and an additional six in forest habitat. Each enclosure contained a 60 cm × 60 cm plywood coverboard for shelter. In late May 2014, we placed 10 females in forest enclosures and the other 10 females in the field enclosures; snakes were housed either in pairs or on their own. We captured all female snakes for this experiment in fields and wetlands near Ottawa, Ontario (n = 3) and close to our enclosures in Québec (n = 17). In the laboratory, before putting females in the enclosures, we gave each female the opportunity to mate with three or four males collected from the same area. We observed every female mate during these encounters. We fed half of the snakes in each habitat two large earthworms per week and we fed the other half of the snakes four large earthworms per week to determine whether fCM concentrations could be an indicator of habitat quality only with certain caloric intake levels. We systematically placed females in each food and habitat treatment combination based on their body size to ensure that each combination had similar distributions of body sizes, because body size may influence growth, reproductive frequency and reproductive output. We measured the body temperature of each snake with an infrared thermometer immediately before feeding it. We measured the mass and snout–vent length (SVL) of each snake once per week and also attempted to collect a faecal sample from each snake once per week. Faecal samples were stored in individual sealed tubes and were kept on ice until they were returned to the laboratory, where they were frozen at −80°C until they were processed. Snakes started giving birth in late August and finished by early September. We removed all snakes from the enclosures after the snakes gave birth and returned all mothers and their offspring to their point of capture.
We extracted fCMs from faecal pellets following the methods of Berkvens et al. (2013). Briefly, we placed thawed and homogenized (using a metal spatula cleaned with 80% methanol) faeces in 2.0 ml tubes with 80% methanol at a ratio of 0.1 g of faeces per 1.0 ml of methanol. We agitated this mixture on a magnetic stir plate at room temperature for 18 h. Following this period, we centrifuged each tube at 210g for 10 min, decanted the liquid (henceforth referred to as the extract) into a new tube and stored the extract at −80°C until we measured fCM content by commercial radioimmunoassay (RIA). The extraction efficiency for this procedure was estimated by splitting a sample in two, spiking one half of the sample with a known amount of corticosterone, measuring corticosterone/fCM content in each half and comparing recovered corticosterone with the amount added, for six independent samples; using this method, extraction efficiency was 50%. This value is relatively low compared with ∼90% extraction efficiency for the faeces of ground squirrels (Spermophilus beldingi; Mateo and Cavigelli, 2005), but on par with the 49% extraction efficiency reported for faeces of African house snakes (Lamprophis fuliginosus; Berkvens et al., 2013). The present study is the first, to our knowledge, in which fCM concentrations were measured in faecal samples of gartersnakes. For this reason, we conducted a validation experiment (see online supplementary material Appendix S1). Briefly, we compared fCM concentrations with plasma corticosterone concentrations in gartersnakes before and after they were fed corticosterone-injected earthworms. We demonstrated that fCM concentrations reflect plasma corticosterone concentrations.