Effective conservation and management of sensitive species requires knowledge of their ecology. For example, understanding habitat requirements can identify environmental constraints that govern persistence of a species, where on the landscape a species will occur, and habitat-related causes of declines. Understanding demography and vital rates can identify reasonable expectations for abundances, survival, and potential for population growth, and identify sensitive life stages. In addition, quantitative evaluations of management, conservation, and recovery actions can provide valuable insights into the success or failures of these programs and how to improve their effectiveness. In reality, these types of information typically are not available. Although assumptions about the ecology of a species may exist, quantitative data commonly is lacking, and management, conservation, and recovery actions are rarely evaluated for their success. Thus, management decisions are often based on incomplete information.

Amphibian populations have undergone severe declines and extinctions worldwide, including in relatively pristine areas. Yet, basic ecological data are often lacking for this taxon, which has been poorly studied relative to other vertebrates like birds and mammals. This dissertation addresses information gaps in the ecology of two federally listed anurans in the Sierra Nevada, CA, the threatened Yosemite toad (Anaxyrus canorus) and the endangered Sierra Nevada yellow-legged frog (SNYLF, Rana sierrae), which have both declined in distribution and abundance. These species occur in patchily distributed, high elevation, montane, aquatic systems. The wet meadows, lakes, ponds, and streams available to these species vary greatly in size, hydrology, topography, and isolation. Sierra Nevada weather is highly variable and unpredictable, spanning gradients from wet to drought and warm to cold years. The two species have different life history strategies that allow them to persist in these uncertain environments. In Chapters 1 and 2, I addressed habitat selection, first investigating Yosemite toad habitat selection in wet meadows, where the species most commonly breeds (Chapter 1), and then investigating habitat selection of SNYLF in streams (Chapter 2), where the species has not been well-studied. In Chapter 3, I investigated the demography of SNYLF in streams. In addition to quantifying the ecology of SNYLF in streams, in Chapters 2 and 3 I investigated the success of a recovery action—population augmentation—by comparing habitat selection and demography of captive-reared frogs and wild frogs.

The Yosemite toad inhabits variable and unpredictable environments with short growing seasons. To investigate Yosemite toad breeding habitat selection in these environments, in Chapter 1 I quantified habitat characteristics in wet meadows at multiple scales including egg mass deposition microhabitats, tadpole microhabitats, breeding areas within wet meadows, and wet meadows. Using logistic regression, I quantified the relationship between breeding occupancy and habitat variables measuring surface water characteristics, water temperature, cover, habitat complexity, physical environment, and climate. I found that habitat selection by the Yosemite toad reflects choices that promote fast development which may be an adaptation to its short growing season and uncertain environment. Yosemite toads bred in very shallow, warm water with low gradient shorelines, and habitat characteristics associated with water and temperature were important at all scales. Overall, Yosemite toads selected for very small differences in these habitat components. For example, water depths where they laid eggs were consistently very shallow, averaging 0.04–0.05 m. Probability of occupancy increased with water temperature at all scales. Breeding areas within meadows ranged from potholes to flooded vegetation with standing or flowing water. Breeding meadows tended to have flatter slopes, west-facing aspects, and complex habitats in the form of springs, inlets, outlets, and breedinghabitat type richness. The quantitative data describing important habitat components summarized in the chapter can be used to design appropriate management and conservation for the Yosemite toad across its range.

In Chapter 2, I addressed two information needs for SNYLF that inhabit streams in its northern range, habitat selection and the effectiveness of a recovery action. Because available habitats in streams can change across space and time, I investigated SNYLF habitat requirements across spatial and temporal scales. I quantified habitat use across diverse streams (third-order perennial streams to first-order intermittent headwater creeks), habitat types within streams (pool, riffle, cascade), and seasonal flows (high water at snowmelt to reduced flows late in the summer). Available habitat changed with flows. Perennial streams retained a variety of habitats across flows whereas intermittent streams dried to just a few pools. These results suggest that SNYLF is an aquatic generalist in streams, able to meet its ecological requirements in a variety of habitats, with some site-specific constraints affecting habitat selection. Frogs used all seasonal habitat types but selected deeper pools most often relative to available habitats in intermittent streams and selected riffles or showed no preference in perennial streams. Frogs avoided fast deep-water microhabitats, but otherwise, preferred flowing water or deeper water without flow. I also compared habitat selection between wild and captive-reared frogs that were released as a population augmentation in one intermittent stream. Captive-reared frogs selected similar habitats as wild frogs, suggesting captivity does not alter habitat selection behaviors post-release. This also suggests that reintroduction designs for this species can be based on wild frog preferences. Results in this chapter improve our understanding of how SNYLF habitat use varies among different types of streams and differs seasonally in response to changing habitat conditions, which in turn can inform designs of effective management, conservation, and recovery programs, including reintroductions of frogs in streams.

In Chapter 3, I conducted a three-year capture–mark–recapture study to investigate the demography of SNYLF in the same streams. The objectives were to quantify abundances and vital rates in diverse streams, evaluate a population augmentation recovery action in one of the streams, and use the demographic data to inform reintroduction designs. Population abundances were low in three streams supporting the need for recovery actions such as population augmentation. The two largest populations were in intermittent streams, possibly because fish are less common where water is not flowing year-round. Estimates of apparent survival rates of wild adults were relatively high (0.55 –0.90) whereas estimates of annual recruitment were relatively low (0.02–0.26). The oldest frog found in the streams was at least 13 years old whereas few tadpoles or subadults were seen in any of the streams. The high survival rates leading to long-lived adults may facilitate persistence of these stream populations, whereas low recruitment may limit population growth and recovery. The population augmentation was relatively successful with at least 52% of captive-reared frogs surviving their first summer after release and at least 36% surviving their first winter. These results offer promise for the use of reintroductions to augment depleted populations. The demography of wild frogs helps identify considerations for reintroduction designs. For example, the apparent low survival of wild eggs, tadpoles, and subadults, suggests releasing adults, as was done in this study, rather than younger life stages, may be more successful. The results also suggest that further research on factors affecting recruitment and survival of younger life stages is needed.

The Yosemite toad and SNYLF are examples of at-risk amphibians that require basic ecological knowledge to support effective management, conservation, and recovery. To date, these decisions mostly have been based on descriptive information and general field observations. This dissertation addresses key aspects of the ecology of these species as well as presents quantitative data to guide future actions.

Translocation, although often criticized, is gaining popularity as a conservation management tool for imperiled species. Many species today face habitat loss, leaving populations isolated and with few options for recolonizing remaining habitat. Giant gartersnakes (Thamnophis gigas) are a semi-aquatic snake endemic to the Central Valley of California and are currently facing this very predicament. Giant gartersnakes historically inhabited wetlands and sloughs, but after losing >90% of their original wetland habitat and being nearly extirpated from the San Joaquin Valley, remaining populations are scattered among remnant marshes and rely heavily on rice agriculture as surrogate habitat. To recover the species, wildlife managers have prescribed the reintroduction of snakes into the San Joaquin Valley and, given the level of habitat loss throughout the region, translocation will likely be required to accomplish these goals. Translocation has never been attempted in giant gartersnakes, so the focus of this dissertation was to examine the survival, movement, and habitat selection of giant gartersnakes from one wetland donor site and one rice agriculture site to a third restored wetland site all within the Natomas Basin in Sacramento, CA, USA. Survival is often lower in translocated animals and because there have been no previous giant gartersnake translocation studies, our goal in the first chapter was to examine the survival of translocated wild-caught adult snakes to the survival of snakes in each of the two donor sites. We also monitored the survival of captive-reared juvenile snakes over a 2-yr rearing period in captivity and in the first four months following release. We found low survival in translocated adult snakes, consistent with several studies of other snake species; just 8% of translocated adult giant gartersnakes survived >801 days (95% CI = 1–64%), compared with 39% of resident snakes at the donor sites surviving >1,154 days (95% CI = 23–68%). This equated to annualized survival rates of translocated adults (0.32, 95% CI = 0.12–0.82) that were roughly half that of resident snakes ( 0.74, 95% CI = 0.63–0.89). Juvenile survival was more promising, with 76% of juvenile snakes surviving captivity and, once released, 60% surviving (95% CI = 38–94%) for the 4-month monitoring period before winter. We found no effect of average distance moved or surface activity on snake survival. Increased movement after translocation is often linked to survival, which prompted us to examine the movement and home ranges of snakes in the second chapter. In chapter two, we compared 100% and 95% MCP and 95% a-LoCoh home range estimates and examined movement between groups using the metrics of total distance traversed, start-to-end distance of movement paths, sinuosity of movement paths, and three measures of net-displacement. We found that results varied depending on the specific metric, but overall, translocated snakes had intermediate values for home range size, movement distances, and sinuosity compared to the two resident groups. Although translocation affected movement and space use relative to resident snakes, the metrics we used remained similar in individuals before and after translocation, and the landscape type of the study site (rice agriculture versus wetland) appeared to have the greatest effect on home range size and movement. We also did not see any signs of translocated snakes attempting to return to their origin sites, which has been a problem in translocation projects in some other species of snakes. Another important consideration in translocation projects is how well the animals acclimate to the release site, with animals that can acclimate more quickly usually being more successful. We examined this question in part in chapter two by looking at snake movement and built on this further in chapter three by quantifying habitat and vegetation selection of snakes from the donor sites and translocated snakes in the recipient site during the active (April–October) and overwintering (November–March) seasons. Across all three sites, we found that snakes avoided exposed habitat types and selected for habitats that provided cover during the active season. The preferred cover type and strength of selection differed among sites, especially between the wetland donor site and rice canal donor site. During the over-wintering season, habitat selection values were more neutral (neither selected nor avoided) which was expected because snakes are largely brumating underground. Habitat selection of translocated snakes in the recipient site shifted to selecting more typical wetland vegetation, such as tules, especially considering most translocated snakes originated from the canal donor site. The strength of selection still differed between snakes at the wetland donor site and recipient site, however, which may be an effect of natal preferences from the snake’s origin site, existing differences between the two wetland sites, effects of translocation itself, or a combination of factors. Snakes in the recipient site appeared to use rock and rip/rap more heavily compared to the other two sites, especially over winter. This could suggest that man-made refugia benefited translocated snakes in the recipient site, which was restored more recently than the wetland donor site and may therefore not have had as many naturally occurring burrows as the more mature donor wetland. Overall, despite lower survival, snake movement and habitat selection, which are often linked to survival, appeared to show that giant gartersnakes acclimated to the recipient site and did not attempt to emigrate from the recipient site after translocation. Future research building on our findings that improve survival of translocated adult snakes or additional efforts using captive-reared juveniles will help to ensure that larger-scale reintroductions into the San Joaquin Valley and other areas of the giant gartersnake’s range are successful.