Findings from a preliminary investigation of the effects of aquatic habitat (water) availability on giant gartersnake (Thamnophis gigas) demography in the Sacramento Valley, California, 2014–17

Abstract

The giant gartersnake ( Thamnophis gigas ) is a semi-aquatic species of snake precinctive to the Central Valley of California. Because the Central Valley has experienced a substantial loss of wetland habitat, giant gartersnake populations are largely found in aquatic habitats associated with rice agriculture. In dry years, less water may be available for rice agriculture, resulting in less aquatic habitat, which could have cascading effects on giant gartersnake populations. We present 2 years of data intended to examine how the demography of giant gartersnakes is affected by the availability of aquatic habitat on the landscape (2016–17), along with 2 years of (sparse) preliminary data (2014–15) collected as part of an earlier radio-telemetry study on giant gartersnake movement behavior. We sampled agricultural canals near rice fields for giant gartersnakes at 8 sites distributed throughout the Sacramento Valley. Five sites were sampled from 2014–17, and 3 sites were sampled from 2015–17. In total, we made 2,995 captures of 1,011 snakes from 2014–17. We used these capture data to fit a multi-site Jolly-Seber model to estimate the abundance of giant gartersnakes as well as the daily and annual probability of capture at each site. We used remotely sensed Landsat data to characterize the extent of flooded rice fields surrounding each site in each year. In addition, we collected 175 females from 2014–17 and delivered them to the Sacramento Zoo for health assessments and reproductive exams. The abundance of giant gartersnakes varied among sites, and abundance estimates were more precise in 2016 and 2017 when sampling effort was greatest. The probability of a giant gartersnake being captured at least once in a year was higher in 2016 and 2017 than 2014 and 2015, and recaptures of snakes marked the previous year were highest in 2016 and 2017 as well. Mean annual apparent survival was estimated to be 0.40 but varied among sites from a low of 0.14 to a high of 0.63. Five sites had diverse size distributions that included abundant sub-adult and large adult female snakes. One site had a truncated size distribution with few large adult female snakes, and 2 sites had mostly large adult-sized snakes and few small individuals. Both the probability a female was gravid and a female’s litter size were positively related to the female’s snout-vent length. Somatic growth rates varied more among years than among sites, and females grew faster (in millimeters per day) than male snakes. The proportion of the landscape around each site under active rice cultivation fluctuated over time (generally between 60–90 percent of the landscape was active rice growing, although this proportion was lower for some sites in some years), and variation in rice growing was asynchronous among sites. This study demonstrates that intensive demographic sampling enables estimation of several key demographic variables at each study site. Continued sampling would allow for investigating potential relationships between the amount of rice growing at a site and demographic parameters such as growth, survival, and reproduction.