Spatial and Temporal Variability in Growth of Giant Gartersnakes: Plasticity, Precipitation, and Prey
The growth rate of reptiles is plastic and often varies among individuals, populations, and years in response to environmental conditions. For an imperiled species, the growth rate of individual animals is an important component of demographic models, and changes in individual growth rates might precede changes in abundance. We analyzed a long-term dataset on the growth of Giant Gartersnakes (Thamnophis gigas) to characterize spatial and temporal variability and evaluate potential environmental predictors of growth. We collected data on the growth in snout–vent length (SVL) of Giant Gartersnakes over 22 yr (1995–2016) from eight sites distributed throughout the Sacramento Valley of California, USA. The von Bertalanffy growth curves indicated male Giant Gartersnakes grew faster toward shorter, asymptotic SVL than did females. Nearly equal variability in growth was attributable to differences among years and among sites. From 2003–2016 we collected data on precipitation, temperature, and the abundance of fish and anuran prey at each site and used these variables as predictors in growth models of Giant Gartersnakes. Snake growth was positively related to the amount of precipitation that fell during the prior water year and the abundance of anurans at a site. Fish and frog abundance interacted to affect snake growth: at low abundances of one prey type, the other positively affected growth, but the slope of this relationship decreased as alternative prey abundance increased. Our results highlight the plasticity of growth in this threatened snake species, point to potential environmental drivers of growth, and provide valuable data for demographic modeling efforts.Abstract
Location of study sites in the Central Valley of California, USA. Circles represent capture locations of snakes from which we obtained growth-increment data.
Growth curves for an average female (solid line) and male (dashed line) Giant Gartersnake. Thick black lines represent median growth curves, thin black lines represent growth curves based on 95% credible intervals. Grey lines are 10 randomly generated growth trajectories per sex to show individual variation. Dashes along the y-axis represent the size of individuals from which we measured growth data.
Effect of site on the von Bertalanffy growth coefficient, k. Circles represent posterior median estimates of k from each site; error bars show the 95% credible intervals. The horizontal dashed line represents the average estimate of k. Abbreviations for site names: BC = Badger Creek; CD = Colusa Drain; CO = Colusa NWR; GI = Gilsizer Slough; BF1 = Butte Farm 1; BF2 = Butte Farm 2; NAT = Natomas Basin; SA = Sacramento NWR.
Effect of year on the von Bertalanffy growth coefficient, k. Circles represent posterior median estimates of k for each year; error bars show the 95% credible intervals. The horizontal dashed line represents the average estimate of k. No data were available for 2006.
Effect of site on asymptotic length. Circles represent posterior median estimates of asymptotic length from each site; error bars show the 95% credible interval. The horizontal dashed line represents the average estimate of asymptotic length. Abbreviations for site names: BC = Badger Creek; CD = Colusa Drain; CO = Colusa NWR; GI = Gilsizer Slough; BF1 = Butte Farm 1; BF2 = Butte Farm 2; NAT = Natomas Basin; SA = Sacramento NWR.
Interaction between fish and frog abundance and the growth coefficient, k. (a, b, c) Panels plot k vs. standardized frog abundance, for low (−2), average (0), and high (2) levels of standardized fish abundance, respectively. (d, e, f) Panels plot k vs. standardized fish abundance for low (−2), average (0), and high (2) levels of standardized frog abundance, respectively. Solid lines represent posterior median estimates and dashed lines represent 95% credible intervals. Dashes along the x-axis represent standardized fish and frog abundance values from study sites.
Contributor Notes
