Brian J. Halstead

Sympatric Masticophis flagellum and Coluber constrictor Select Vertebrate Prey at Different Levels of Taxonomy

Abstract Masticophis flagellum (Coachwhip) and Coluber constrictor (Eastern Racer) are widespread North American snakes with similar foraging modes and habits. Little is known about the selection of prey by either species, and despite their apparently similar foraging habits, comparative studies of the foraging ecology of sympatric M. flagellum and C. constrictor are lacking. We examined the foraging ecology and prey selection of these actively foraging snakes in xeric, open-canopied Florida scrub habitat by defining prey availability separately for each snake to elucidate mechanisms underlying geographic, temporal, and interspecific variation in predator diets. Nineteen percent of M. flagellum and 28% of C. constrictor contained stomach contents, and most snakes contained only one prey item. Mean relative prey mass for both species was less than 10%. Larger C. constrictor consumed larger prey than small individuals, but this relationship disappeared when prey size was scaled to snake size. Masticophis flagellum was selective at the prey category level, and positively selected lizards and mammals; however, within these categories it consumed prey species in proportion to their availability. In contrast, C. constrictor preyed upon prey categories opportunistically, but was selective with regard to species. Specifically, C. constrictor positively selected Hyla femoralis (Pine Woods Treefrog) and negatively selected Bufo quercicus (Oak Toad), B. terrestris (Southern Toad), and Gastrophryne carolinensis (Eastern Narrowmouth Toad). Thus, despite their similar foraging habits, M. flagellum and C. constrictor select different prey and are selective of prey at different levels of taxonomy.

Habitat Suitability and Conservation of the Giant Gartersnake (Thamnophis gigas) in the Sacramento Valley of California

Abstract Resource managers often have little information regarding the habitat requirements and distribution of rare species. Factor analysis-based habitat suitability models describe the ecological niche of a species and identify locations where these conditions occur on the landscape using existing occurrence data. We used factor analyses to assess the suitability of habitats for Thamnophis gigas (Giant Gartersnake), a rare, threatened species endemic to the Central Valley of California, USA, and to map the locations of habitat suitable for T. gigas in the Sacramento Valley. Factor analyses indicated that the niche of T. gigas is composed of sites near rice agriculture with low stream densities. Sites with high canal densities and near wetlands also appeared suitable, but results for these variables were sensitive to potential sampling bias. In the Sacramento Valley, suitable habitats occur primarily in the central portion of the valley floor. Based upon the results of the factor analyses, recovery planning for T. gigas will require an on-the-ground assessment of the current distribution and abundance of T. gigas, maintaining the few remaining natural wetlands and the practice of rice agriculture in the Sacramento Valley, and studying the effects of agricultural practices and land use changes on populations of T. gigas.

Masticophis Flagellum Selects Florida Scrub Habitat at Multiple Spatial Scales

Abstract The use of space by individual animals strongly influences the spatial extent, abundance, and growth rates of their populations. We analyzed the spatial ecology and habitat selection of Masticophis flagellum (the coachwhip) at three different scales to determine which habitats are most important to this species. Home ranges and mean daily displacements of M. flagellum in Florida were large compared to individuals in other populations of this species. Home ranges contained a greater proportion of Florida scrub habitat than did the study site as a whole, and individuals selected Florida scrub habitat within their home ranges. For both selection of the home range within the study site and selection of habitats within the home range, mesic cutthroat and hydric swamp habitats were avoided. Standardized selection ratios of Florida scrub patches were positively correlated with lizard abundance. Several non-mutually exclusive mechanisms, including foraging success (prey abundance, prey vulnerability, and foraging efficiency), abundance of refugia, and thermoregulatory opportunity may underlie the selection of Florida scrub by M. flagellum. Historic rarity and anthropogenic loss and fragmentation of Florida scrub habitat, coupled with the long-distance movements, large home ranges, and selection of Florida scrub by M. flagellum, indicate that large contiguous tracts of land containing Florida scrub will be essential for the persistence of M. flagellum in central Florida.

Abundance and Sexual Size Dimorphism of the Giant Gartersnake (Thamnophis gigas) in the Sacramento Valley of California

Abstract The Giant Gartersnake (Thamnophis gigas) is restricted to wetlands of the Central Valley of California. Because of wetland loss in this region, the Giant Gartersnake is both federally and state listed as threatened. We conducted mark–recapture studies of four populations of the Giant Gartersnake in the Sacramento Valley (northern Central Valley), California, to obtain baseline data on abundance and density to assist in recovery planning for this species. We sampled habitats that ranged from natural, unmanaged marsh to constructed managed marshes and habitats associated with rice agriculture. Giant Gartersnake density in a natural wetland (1.90 individuals/ha) was an order of magnitude greater than in a managed wetland subject to active season drying (0.17 individuals/ha). Sex ratios at all sites were not different from 1 ∶ 1, and females were longer and heavier than males. Females had greater body condition than males, and individuals at the least disturbed sites had significantly greater body condition than individuals at the managed wetland. The few remaining natural wetlands in the Central Valley are important, productive habitat for the Giant Gartersnake, and should be conserved and protected. Wetlands constructed and restored for the Giant Gartersnake should be modeled after the permanent, shallow wetlands representative of historic Giant Gartersnake habitat.

Bayesian adaptive survey protocols for resource management

ABSTRACT Transparency in resource management decisions requires a proper accounting of uncertainty at multiple stages of the decision-making process. As information becomes available, periodic review and updating of resource management protocols reduces uncertainty and improves management decisions. One of the most basic steps to mitigating anthropogenic effects on populations is determining if a population of a species occurs in an area that will be affected by human activity. Species are rarely detected with certainty, however, and falsely declaring a species absent can cause improper conservation decisions or even extirpation of populations. We propose a method to design survey protocols for imperfectly detected species that accounts for multiple sources of uncertainty in the detection process, is capable of quantitatively incorporating expert opinion into the decision-making process, allows periodic updates to the protocol, and permits resource managers to weigh the severity of consequences if the species is falsely declared absent. We developed our method using the giant gartersnake (Thamnophis gigas), a threatened species precinctive to the Central Valley of California, as a case study. Survey date was negatively related to the probability of detecting the giant gartersnake, and water temperature was positively related to the probability of detecting the giant gartersnake at a sampled location. Reporting sampling effort, timing and duration of surveys, and water temperatures would allow resource managers to evaluate the probability that the giant gartersnake occurs at sampled sites where it is not detected. This information would also allow periodic updates and quantitative evaluation of changes to the giant gartersnake survey protocol. Because it naturally allows multiple sources of information and is predicated upon the idea of updating information, Bayesian analysis is well-suited to solving the problem of developing efficient sampling protocols for species of conservation concern.

Population Size, Survival, Growth, and Movements of Rana sierrae

Abstract: Based on 2431 captures of 757 individual frogs over a 9-yr period, we found that the population of R. sierrae in one meadow–stream complex in Yosemite National Park ranged from an estimated 45 to 115 adult frogs. Rana sierrae at our relatively low elevation site (2200 m) grew at a fast rate (K = 0.73–0.78), had high overwintering survival rates (44.6–95%), lived a long time (up to 16 yr), and tended to be fairly sedentary during the summer (100% minimum convex polygon annual home ranges of 139 m2) but had low year-to-year site fidelity. Even though the amphibian chytrid fungus (Batrachochytrium dendrobatidis, Bd) has been present in the population for at least 13 yr, there was no clear downward trend as might be expected from reports of R. sierrae population declines associated with Bd or from reports of widespread population decline of R. sierrae throughout its range.

Waste Rice Seed in Conventional and Stripper-Head Harvested Fields in California: Implications for Wintering Waterfowl

Abstract Waste rice seed is an important food for wintering waterfowl and current estimates of its availability are needed to determine the carrying capacity of rice fields and guide habitat conservation. We used a line-intercept method to estimate mass-density of rice seed remaining after harvest during 2010 in the Sacramento Valley (SACV) of California and compared results with estimates from previous studies in the SACV and Mississippi Alluvial Valley (MAV). Posterior mean (95% credible interval) estimates of total waste rice seed mass-density for the SACV in 2010 were 388 (336–449) kg/ha in conventionally harvested fields and 245 (198–307) kg/ha in stripper-head harvested fields; the 2010 mass-density is nearly identical to the mid-1980s estimate for conventionally harvested fields but 36% lower than the mid-1990s estimate for stripped fields. About 18% of SACV fields were stripper-head harvested in 2010 vs. 9–15% in the mid-1990s and 0% in the mid-1980s; but due to a 50% increase in planted rice area...

Active Season Microhabitat and Vegetation Selection by Giant Gartersnakes Associated with a Restored Marsh in California

Abstract Studies of habitat selection can reveal important patterns to guide habitat restoration and management for species of conservation concern. Giant gartersnakes Thamnophis gigas are endemic to the Central Valley of California, where >90% of their historical wetland habitat has been converted to agricultural and other uses. Information about the selection of habitats by individual giant gartersnakes would guide habitat restoration by indicating which habitat features and vegetation types are likely to be selected by these rare snakes. We examined activity patterns and selection of microhabitats and vegetation types by adult female giant gartersnakes with radiotelemetry at a site composed of rice agriculture and restored wetlands using a paired case-control study design. Adult female giant gartersnakes were 14.7 (95% credible interval [CRI] = 9.4–23.7) times more likely to be active (foraging, mating, or moving) when located in aquatic habitats than when located in terrestrial habitats. Microhabitats...

Relative value of managed wetlands and tidal marshlands for wintering northern pintails

Abstract Northern pintail Anas acuta (hereafter pintail) populations have declined substantially throughout the western United States since the 1970s, largely as a result of converting wetlands to cropland. Managed wetlands have been developed throughout the San Francisco Bay estuaries to provide wildlife habitat, particularly for waterfowl. Many of these areas were historically tidal baylands, and plans are underway to remove dikes and restore tidal action. The relationship between tidal baylands and waterfowl populations is poorly understood. Our objective was to provide information on selection and avoidance of managed and tidal marshland by pintails. During 1991–1993 and 1998–2000, we radiomarked and relocated 330 female pintails (relocations, n  = 11,574) at Suisun Marsh, California, the largest brackish water estuary within San Francisco Bay, to estimate resource selection functions during the nonbreeding months (winter). Using a distance-based modeling approach, we calculated selection functions fo...

Temporal and maternal effects on reproductive ecology of the giant gartersnake ( Thamnophis gigas )

Abstract We used mixed-effects models to examine relationships of reproductive characteristics of the giant gartersnake (Thamnophis gigas) to improve population modeling and conservation planning for this species. Neonates from larger litters had lower mass, and mass of neonates also was affected by random variation among mothers. Length of mother did not affect relative mass of litters; however, our data suggest that longer mothers expended less reproductive effort per offspring than shorter mothers. We detected random variation in length of neonates among mothers, but these lengths were not related to length of mother or size of litter. Mean size of litter varied among years, but little evidence existed for a relationship between size of litter or mass of litter and length of mother. Sex ratios of neonates did not differ from 1∶1.