Introduction

Translocation of plants and animals can be used to achieve biodiversity conservation goals. When an organism is translocated to a currently unoccupied location within its indigenous range, it is called a reintroduction. Reintroductions are increasingly used to improve the viability of imperiled species (Taylor and others, 2017). As the number of attempted reintroductions grows, there are notable success stories (for example, black-footed ferrets [Mustela nigripes]; Jachowski and others, 2011) and high-profile reintroductions that have failed (for example, the Brown Treecreeper [Climacteris picumnus] in Australia; Bennett and others, 2012). Many studies have attempted to compile data on these efforts and ascertain what separates the successes from the failures to learn from and improve future reintroductions.

Reintroductions could fail if the proposed site is not abiotically and biotically suitable for the target species. The climatic suitability of a site is highly correlated to the probability of success for conservation translocations (Bellis and others, 2020). Unlike translocations outside a species’ historical range, climatic suitability might be less of a concern in the near term for reintroductions to formerly occupied sites. For reptiles, common factors that explain failed reintroductions include (1) poor survival during the inactive season (for example, overwintering mortality in temperate climates; Roe and others, 2010), (2) high dispersal-related mortality after release (Germano and Bishop, 2009), (3) low habitat quality at the reintroduction site (Germano and Bishop, 2009), and (4) decreased survival during the establishment phase compared to wild conspecifics (Bertolero and others, 2018). Feasibility studies that investigate if the ecological community can support the target species are vital to ensure the reintroduction does not waste valuable resources and expose individuals to a high risk of mortality in the new environment.

The San Francisco gartersnake (SFGS, Thamnophis sirtalis tetrataenia) is a subspecies of the common gartersnake listed as endangered under the U.S. Endangered Species Act (ESA; U.S. Fish and Wildlife Service, 1973, as amended) and the California Endangered Species Act (California Code of Regulations, 1971). San Francisco gartersnakes persist in 13 known populations in San Mateo County and northwestern Santa Cruz County, California (U.S. Fish and Wildlife Service, 2020). Many of these populations are geographically isolated, have low abundance and genetic diversity, and likely have limited or no genetic or demographic connectivity to other SFGS populations (Wood and others, 2020). The recovery plan for SFGS (U.S. Fish and Wildlife Service, 1985) calls for conservation actions to protect six extant populations known at the time and create four new populations. The recovery plan also calls for maintaining a minimum of 200 adult SFGS in a 1:1 sex ratio at each of 10 populations for at least 15 consecutive years for the species to be considered recovered (U.S. Fish and Wildlife Service, 1985). Most extant SFGS populations for which abundance has been estimated fall below this minimum threshold of 200 adults and have low genetic diversity (Wood and others, 2020). Given the fragmentation of habitat and existence of dispersal barriers (for example, roads, urban areas) within the historical range of SFGS, it is unlikely that snakes will be able to recolonize unoccupied but suitable habitat without human intervention. Considering the threats facing SFGS, reintroducing this subspecies to unoccupied suitable habitat within its historical range could be necessary to ensure its viability in the 21st century.

Despite the urbanization and development of the San Francisco Bay area since the 19th Century, large swaths of open space are protected in rural parts of San Mateo County, including state parks and several preserves managed by the Midpeninsula Regional Open Space District (MROSD), land trusts, and other nongovernmental organizations. San Francisco gartersnake populations exist in at least two preserves managed by MROSD: (1) Cloverdale Ranch Open Space Preserve (CR) near the coast in southwestern San Mateo County and (2) Russian Ridge Open Space Preserve (RR) further inland in the Santa Cruz Mountains (Halstead and others, 2011; Kim and others, 2018; Wood and others, 2020). It is possible that suitable habitat exists within other open space preserves managed by MROSD that is not occupied by SFGS at the time of this publication. The need to create additional populations of SFGS to increase the viability of this subspecies is made clear in the species’ recovery plan (U.S. Fish and Wildlife Service, 1985) and more recent species status assessment (U.S. Fish and Wildlife Service, 2020). One promising site for the reintroduction of SFGS to establish a new population is the La Honda Creek Open Space Preserve (LHC) owned by MROSD and in rural San Mateo County near the communities of La Honda and Woodside (fig. 1). The LHC is bordered to the south and east by State Highway 84 and to the north by Skyline Blvd (State Highway 35). The size of LHC, the wetland and upland habitats present, the ecological communities contained within, and the proximity to existing SFGS populations all combine to make this preserve a promising site for the reintroduction of SFGS.

We compiled data on the herpetofaunal community and habitat within LHC to evaluate if this preserve is abiotically and biotically suitable for the establishment of an SFGS population by comparing LHC to the CR and RR preserves that support extant SFGS populations. We then used a demographic model for SFGS to simulate the viability of reintroduced populations of SFGS at LHC under different scenarios to determine if introducing all snakes at one pond or dividing individuals among multiple ponds is more likely to result in a viable population. Finally, we discuss the data needed to monitor the status of reintroduced SFGS populations at LHC, evaluate if the reintroduction is successful, and take alternative actions if necessary.

Study Area

La Honda Creek Open Space Preserve is a 2,515 hectare (ha; 6,215 acre) preserve in San Mateo County, California, that encompasses historical ranches, farms, and timberlands (fig. 2). Elevations at LHC range from approximately 58 to 713 meters (m; 190 to 2,340 feet), with generally higher elevations in the northern half of the preserve and lower elevations in the southern half. La Honda Creek Open Space Preserve is managed to provide habitat for wildlife along with compatible ecological grazing and agricultural uses, hiking, equestrian, and bicycling trails for public recreation. Two Conservation Management Units also exist within LHC that are closed to the public. Most uplands at LHC are coniferous or evergreen hardwood forests with coast redwoods (Sequoia sempervirens) and Douglas fir (Pseudotsuga menziesii), followed by open grasslands and shrub-dominated scrub. Many wetlands are present in LHC, including more than 25 ponds and three main creeks with associated tributaries. Substantial parts of LHC were historically ranch land, and grazing is used as a management tool on four pastures within the preserve. Grazing is used to reduce fuel loads to mitigate fire risk and to manage grassland vegetation (Midpeninsula Regional Open Space District, 2012). One of the objectives of the LHC’s Master Plan is to “Protect and enhance populations of listed wildlife species” (Midpeninsula Regional Open Space District, 2012). One such listed species is the California red-legged frog (CRLF, Rana draytonii), which is listed as threatened under the U.S. Endangered Species Act (U.S. Fish and Wildlife Service, 1996) and breeds in several ponds at LHC (see “Prey Community at La Honda Creek Open Space Preserve”).

Methods

Simulating Reintroductions

We simulated reintroducing SFGS to LHC using a demographic population model developed and published by USGS biologists (Rose and others, 2023). Survival, growth, and fecundity of SFGS are size-dependent (Rose and others, 2022; Schoenig and others, 2023a). San Francisco gartersnakes grow rapidly when they are young, with snakes’ growth rate declining as individuals reach greater snout-vent lengths (SVL). Survival is highest for snakes of intermediate size (300–550 millimeters [mm] SVL) and decreases for larger adult females (550–700 mm SVL), with the survival rates of the largest (greater than 700 mm SVL) and smallest (less than 300 mm SVL) snakes highly uncertain based on available data (Rose and others, 2022). We incorporate that uncertainty in the model for our population simulations. The probability of reproduction and fecundity for female SFGS increase with size, with longer snakes more likely to be gravid and to give birth to larger litters of neonate snakes (Schoenig and others, 2023a). Details on the size-based demographic model for SFGS are available in Rose and others (2023), and a diagram representing the model is shown in figure 3. For consistency in our simulations, we assumed a neonate survival rate (from birth to 1 year old) of 0.30, which resulted in a slowly growing population in a stochastic environment in previous simulations for SFGS (Rose and others, 2023). We acknowledge that neonate survival could be higher or lower than this value in the wild, but fixing neonate survival to one value is sufficient for comparing relative outcomes among different reintroduction scenarios.

Based on the results of Rose and others (2023), we simulated reintroduced populations of SFGS in which head-started, 1-year-old juvenile snakes are released to the wild. In head-starting, gravid adult female snakes are temporarily removed from donor populations and held in captivity until they give birth. The adult female snakes are then returned to the site from which they were collected. The neonate snakes are raised in captivity before being released to the wild at an age and size expected to increase their chances of survival (Nguyen and others, 2023). Head-starting provides reptiles the opportunity for increased growth and survival compared to neonates born in the wild (Tuberville and others, 2019). Previous population simulations indicated that releasing head-started juvenile snakes resulted in the greatest viability of reintroduced and source populations (Rose and others, 2023). Our population model and simulations only track female SFGS, but we assume that an equal number of female and male juvenile SFGS are released on average.

For the purposes of simulation, we considered each wetland to represent a potential site for a subpopulation of SFGS. We simulated six scenarios for reintroduction, with the number of distinct subpopulations at which snakes were released varying from one to six. We did not model connectivity and dispersal of snakes among subpopulations. Instead, we modeled each subpopulation in isolation. Treating each subpopulation as distinct likely produces conservative estimates of quasi-extinction risk because dispersal of snakes from a nearby subpopulation could “rescue” a declining subpopulation from extinction (Brown and Kodric-Brown, 1977). The total number of head-started female juvenile SFGS released each year was held constant at 36, to limit the number of adults required to produce juveniles each year to 10 adult females. This number of adult females required to produce enough neonates for rearing and release, assuming a 1:1 sex ratio among neonates, was estimated in an earlier study on SFGS reintroduction strategies and is intended to minimize effects on SFGS donor populations (Rose and others, 2023). The number of individuals was divided evenly among subpopulations. For example, if 36 total juvenile SFGS were released in each scenario, this would be 36 SFGS at 1 subpopulation, 18 SFGS per subpopulation for 2 subpopulations, or 12 SFGS per subpopulation for 3 subpopulations (table 1). For the scenario with 5 subpopulations, 4 subpopulations had 7 SFGS released each year and the 5th subpopulation had 8 SFGS released each year, a total of 36 SFGS released annually. The scenarios were run under four durations of reintroduction efforts. An equal number of SFGS were reintroduced annually for 5, 10, 15, or 20 years. The size of juvenile SFGS released was drawn from a normal distribution based on the mean (338.9 mm) and standard deviation (46.4 mm) of expected SVL for 1-year-old juvenile SFGS from a von Bertalanffy growth function fitted to capture-recapture data (Rose and others, 2023). It is possible that head-started SFGS reared in captivity could reach larger sizes at 1 year old, but without data on growth in captivity, we made the conservative assumption that captive-reared snakes would be a similar size to those in wild populations.

Table 1.    

Results from reintroduction scenarios for San Francisco gartersnakes (Thamnophis sirtalis tetrataenia) at La Honda Creek Open Space Preserve.

[Subpopn, subpopulation; Mean p_qext, mean probability of quasi-extinction during the 30-year simulation for subpopulations in that scenario; Mean N₃₀, mean abundance of SFGS after 30 years for all simulations; Mean t_qext, mean time to quasi-extinction for simulations that go quasi-extinct in that scenario]

Table 1.    Results from reintroduction scenarios for San Francisco gartersnakes (Thamnophis sirtalis tetrataenia) at La Honda Creek Open Space Preserve.

Release duration	Subpopns	Snakes per subpopn	Mean pqext	Mean N30	Mean tqext
5	1	36	0.253	373	18.8
5	2	18	0.318	324	16.8
5	3	12	0.368	284	15.4
5	4	9	0.406	260	14.4
5	5	7 (8)	0.448	229	13.2
5	6	6	0.474	220	12.6
10	1	36	0.149	443	21.3
10	2	18	0.221	377	20.7
10	3	12	0.246	333	20.1
10	4	9	0.290	304	19.2
10	5	7 (8)	0.329	272	18.8
10	6	6	0.354	262	18.2
15	1	36	0.083	490	25.3
15	2	18	0.138	421	24.2
15	3	12	0.170	382	23.6
15	4	9	0.195	350	22.5
15	5	7 (8)	0.226	316	22.2
15	6	6	0.248	302	22.1
20	1	36	0.057	528	27.3
20	2	18	0.077	464	26.6
20	3	12	0.104	402	25.9
20	4	9	0.125	370	25.7
20	5	7 (8)	0.158	344	25.5
20	6	6	0.164	326	25.1

For each scenario, we ran 1,000 simulations with different parameter values for SFGS growth, survival, and fecundity to reflect parametric uncertainty (in other words, uncertainty about the true value of growth, survival, and reproductive rates) and environmental stochasticity (in other words, random annual variation in survival and growth rates). Demographic stochasticity, or randomness in the birth and death of individuals, is an important factor in the probability of extinction for small populations (Caughley, 1994; Melbourne and Hastings, 2008). To account for demographic stochasticity, we simulated reproduction and survival for each individual SFGS using random draws from Poisson and Bernoulli distributions, respectively. We ran each simulation for 30 years and calculated the number of SFGS surviving in the final year. We defined a quasi-extinction threshold of five individuals. If the population of SFGS was less than or equal to five individuals at any point during the simulation, it was considered to have gone extinct and the simulation stopped. We recorded the year in which the subpopulation went extinct for each simulation and calculated the average time to quasi-extinction for each scenario in which the subpopulation went quasi-extinct (in other words, simulations in which the subpopulation did not go quasi-extinct after 30 years did not factor into the average time to quasi-extinction calculation). Finally, we calculated the quasi-extinction probability using the proportion of the 1,000 simulations in which each subpopulation went extinct. For scenarios with greater than one subpopulation, we calculated the quasi-extinction probability for each individual subpopulation.

Demographic population models are, by design, a simplification of the real dynamics of populations. Predictions from demographic population models can be inaccurate because of several factors that exist in nature and are not captured by the model. Rather than focus on the absolute quasi-extinction probability predicted from any single simulation, we use the results of the population simulations to compare the viability and abundance of populations among different scenarios (McCarthy and others, 2003).

Results

Records of SFGS from 1983 reported by Barry (1994) were within 1.5 km (0.9 mi) of the eastern and western borders of LHC (fig. 2). These nearby occurrences, combined with the existing population of SFGS at nearby RR 3.7 km (2.3 mi) to the southeast, and the number of wetlands present at LHC, indicate it is likely that SFGS historically occurred within the present-day boundary of LHC. If SFGS previously occupied LHC, the cause of the species’ putative extirpation is unknown. One possible explanation for the extirpation of SFGS is that incompatible land uses (for example, intensive grazing, agriculture, or forestry) negatively affected aquatic and terrestrial habitats and prey populations.

San Francisco gartersnake populations occur in cooler, fog-influenced coastal regions, and warmer inland regions of San Mateo County (U.S. Fish and Wildlife Service, 2020). The LHC is likely suitable climatically for SFGS given its proximity to the extant population at RR and the more coastal population at CR. Mean total annual precipitation at LHC from 1991 to 2020 averaged 800 mm, slightly lower than the 924 mm at RR, but higher than the average precipitation of 680 mm at CR (PRISM Climate Group, 2022). The climatic effect of the Pacific Ocean, approximately 7 km (4.4 mi) west of the western edge of LHC, moderates temperatures in this region, limiting extreme low temperatures in the winter and high temperatures in the summer.

Climate change could affect the suitability of LHC for SFGS in the future, and these effects will likely be similar to those at other sites occupied by SFGS. Increased variability in precipitation in the 21st Century could increase the risk of droughts, wildfires, and heavy rainfall events (Diffenbaugh and others, 2015; Swain and others, 2018; Gershunov and others, 2019; Swain, 2021). Long-term droughts could negatively affect amphibian populations (Moss and others, 2021) that are prey for SFGS. The potential effects of wildfires on SFGS are more complex. Although wildfires could cause direct mortality and disturb habitat occupied by SFGS, fires could prevent succession and maintain the open grassland and scrub habitats used by SFGS. A study of an SFGS population after a controlled burn indicated that the survival probability of SFGS in burned areas was likely unchanged after the fire (Halstead and others, 2019). The effects of heavy rainfall and flooding on SFGS populations are unknown. Heavy rainfall can stimulate breeding by amphibians (Jensen and others, 2003), producing abundant prey for semiaquatic snakes in wet years. Overall, the volatile interannual fluctuations in rainfall predicted by climate models represents increased environmental stochasticity, which, in general, is expected to increase extinction probability for populations (Caughley, 1994; Doak and others, 2005). Given the limited geographic range of SFGS, the risks posed by climate change are likely to affect most existing SFGS populations, and suitable but unoccupied habitats, in similar ways.

Overall, the habitat present at LHC is similar to areas occupied by SFGS populations in the Santa Cruz Mountains, with one notable exception regarding terrestrial habitat immediately adjacent to wetlands at LHC (see next paragraph). The LHC has more than 25 wetlands, including natural, improved, and constructed ponds. Some wetlands are permanent and hold water year-round, and some are seasonal, filling with water during winter and spring rain and drying during the summer and autumn. The wetlands at LHC range from 0.01 ha (0.03 acres) to 1.03 ha (2.55 acres) in surface area. The Midpeninsula Regional Open Space District has improved several wetlands in the past 15 years to provide better habitat for native amphibians, and has developed a plan to make wetlands suitable for SFGS should the subspecies be reintroduced (Midpeninsula Regional Open Space District, 2012). The LHC has several wetlands with emergent vegetation that are used for breeding by native amphibians (see “Prey Community at La Honda Creek Open Space Preserve”). These wetlands are adjacent to upland grassland and scrub (fig. 2) that could be used for dispersal or refuge during inactive periods.

Two key requirements for SFGS habitat beyond wetlands are (1) the presence of mammal burrows and other subterranean refuges and (2) emergent vegetation and shrubs along the edge of wetlands. Mammal burrows and subterranean refuges provide stable temperatures and moisture for snakes to overwinter during colder months, aestivate during summer heat, and avoid predators. The overwintering habitat of SFGS is not well-studied, but it is thought that snakes move into upland habitat during the autumn and take refuge underground in mammal burrows during late autumn and winter before returning to foraging habitat in the spring (U.S. Fish and Wildlife Service, 2020). The proximity of shrub-covered hillsides adjacent to wetlands could be important for allowing SFGS to migrate from foraging grounds used in the spring and summer to refuges used during the fall and winter months and for promoting movement between wetlands. On a shorter time scale, snakes need to retreat from foraging habitats to gestate their young, digest their prey, and shed their skin. During these activities, snakes are vulnerable to predators and take cover underground or under thick vegetation, emerging to bask near these refuges. One outstanding question regarding refuges for SFGS at LHC is whether sufficient mammal burrows and other underground structures are present near wetlands. The historical agricultural land use in LHC has resulted in decreased abundance of small mammal burrows in some areas (J. Andersen, Midpeninsula Regional Open Space District, written commun., 2023). A field visit by a USGS biologist (E. Schoenig) in November 2023 confirmed that small mammal burrows are present in the uplands near wetlands used by amphibians for breeding at LHC. The fact that many other snake species occur at LHC, including other gartersnakes (see below), indicates that sufficient burrows and other upland refuges are present at LHC for SFGS.

The second key requirement for SFGS terrestrial habitat is cover immediately adjacent to wetlands provided by emergent wetland vegetation (for example, cattails [Typha spp.], tules [Scheonoplectus acutus]), riparian trees (for example, willows [Salix spp.]), and scrub (for example, coyote bush [Baccharis pilularis], blackberry [Rubus spp.]) further upland. Several ponds at LHC are adjacent to coyote bush scrub that could provide cover and refuge for SFGS during their inactive periods, and willow and tule stands grow along some pond edges. Field surveys of wetlands at LHC by USGS in November 2023 indicated that at many water bodies, there was little vegetative cover near the water’s edge. Grazing by cattle in and around wetlands results in only short herbaceous cover that is avoided by SFGS, and direct grazing on emergent vegetation minimizes ideal foraging area. Both types of grazing were evident at some LHC wetlands along with heavy traffic by cattle that can collapse burrows and cause compacted ground in the immediate wetland vicinity. As a result, some aquatic features at LHC are surrounded by a ring of bare ground that creates a deterrence and potential risk for snakes moving between their foraging habitat and upland refugia. The habitat surrounding wetlands could be made more suitable for SFGS if the access of cattle to most parts of the water body were limited by exclosures that buffered the water’s edge and extended to the adjacent upland areas to provide protected connectivity between foraging and upland habitats. The value of buffers that limit land use affecting surrounding wetlands has been demonstrated for many amphibians and semiaquatic reptiles (Burke and Gibbons, 1995; Semlitsch and Bodie, 2003).

The community of snakes that occurs at LHC also is relevant for evaluating the suitability of the preserve for SFGS. All the snake species known to occur at CR, except for SFGS, have been present at LHC, along with one additional species (forest sharp-tailed snake, Contia longicauda) that also is present at RR (Amanda Sparkman, Westmont College, written commun., 2023). These include racer (Coluber constrictor), gophersnake (Pituophis catenifer), ring-necked snake (Diadophis punctatus), rubber boa (Charina bottae), and northern Pacific rattlesnake (Crotalus oreganus oreganus), and two other gartersnakes: Santa Cruz gartersnake (T. atratus atratus) and coast gartersnake (T. elegans terrestris). Given that SFGS co-occur with these other snake species in many of the remaining SFGS populations, it is likely that the SFGS could coexist with the snake community at LHC.

Prey Community at La Honda Creek Open Space Preserve

The MROSD has done extensive surveys for CRLF breeding at several wetlands within LHC since 2008, with the most complete data available from 2016 to 2023. Not all wetlands are monitored, and the counts that follow likely represent a subset of the total number of wetlands used for breeding by CRLF at LHC. From 2016 to 2023, the number of ponds at LHC with evidence of CRLF breeding ranged from a minimum of 5 wetlands in 2022 to a maximum of 11 wetlands in 2019. In total, CRLF breeding has been documented in 13 ponds on at least one occasion from 2016 to 2023 (fig. 4). Ponds Driscoll Ranch (DR)03, DR04, DR05, DR06, DR07, and DR08 had the most consistent evidence of CRLF breeding from 2016 to 2023 based on egg mass counts (table 2), larval counts (table 3), and counts of recently emerged metamorphs (table 4). Five of these six ponds (all but DR03) are within an approximately 1-km radius of one another in the center of LHC, to the east and west of Harrington Creek (fig. 4). The intervening habitat between these five central ponds is largely grassland (herbaceous cover) and scrub, with forest surrounding Harrington Creek and its tributaries. In contrast, pond DR03 is near the eastern edge of LHC, approximately 300 m west of State Highway 84. In addition, CRLF breeding is presumed to occur in DR19, DR20, and DR26 each year, but these wetlands are not regularly surveyed (table 2; J. Chung, Midpeninsula Regional Open Space District, written commun., 2023).

Table 2.    

Summary of California red-legged frog egg mass counts at La Honda Creek Open Space Preserve from 2008 to 2023.

[Unpublished data (not publicly available at the time of publication) were used with permission from the Midpeninsula Regional Open Space District on November 25, 2024. Abbreviations: DR, Driscoll Ranch; NS, site was not sampled for egg masses that year; —, no data; LHC, La Honda Creek Open Space Preserve]

Table 2.    Summary of California red-legged frog egg mass counts at La Honda Creek Open Space Preserve from 2008 to 2023.

Pond	2008	2009	2010	2011	2012	2013	2014	2015	2016	2017	2018	2019	2020	2021	2022	2023
DR02	NS	NS	NS	NS	NS	NS	NS	0	NS	NS	NS	NS	NS	NS	NS	NS
DR03	NS	NS	NS	NS	NS	NS	NS	0	0	3	0	0	6	2	0	1
DR04	NS	NS	NS	NS	NS	NS	NS	0	2	1	0	0	1	2	0	0
DR05	NS	NS	NS	NS	NS	NS	NS	0	5	2	3	2	10	0	3	26
DR06	2	0	0	1	0	2	1	0	6	0	0	6	0	0	0	1
DR07	1	7	0	0	0	1	0	0	2	10	11	8	2	2	0	1
DR08	3	0	0	0	0	1	0	0	3	6	5	2	9	0	6	3
DR09	15	11	10	0	0	0	0	0	0	0	0	0	0	0	0	0
DR10	6	13	7	9	0	2	0	1	3	1	0	0	0	0	1	1
DR11	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS
DR12	0	0	3	0	0	0	1	0	0	0	0	0	0	NS	NS	0
DR13	1	1	0	0	0	0	0	0	0	NS	0	0	0	0	0	0
DR14	1	0	0	0	0	2	1	0	1	1	1	1	1	0	0	2
DR15	1	3	0	3	2	0	0	0	0	0	8	0	4	1	0	5
DR16	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	0	1	0	3	0
DR17	0	0	0	0	0	0	0	0	NS	NS	—	0	0	0	0	0
DR18	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	1	NS	NS
DR19	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	0	NS
DR20	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	0	NS
DR21	NS	NS	NS	NS	NS	NS	NS	0	NS	NS	NS	NS	NS	NS	NS	NS
DR22	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS
DR26	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	4	NS	NS
LHC23	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS
LHC24	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS
LHC25	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS
Slide	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	—	3	1	*4	0	3

^*

California red-legged frog egg masses were removed from that water body and relocated to another.

Table 3.    

Summary of California red-legged frog larvae counts at La Honda Creek Open Space Preserve from 2008 to 2023.

[Unpublished data (not publicly available at the time of publication) were used with permission from the Midpeninsula Regional Open Space District on November 25, 2024. Abbreviations: DR, Driscoll Ranch; NS, site was not sampled for larvae that year; Y, site had larvae but no count is available; LHC, La Honda Creek Open Space Preserve]

Table 3.    Summary of California red-legged frog larvae counts at La Honda Creek Open Space Preserve from 2008 to 2023.

Pond	2008	2009	2010	2011	2012	2013	2014	2015	2016	2017	2018	2019	2020	2021	2022	2023
DR02	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS
DR03	NS	NS	NS	NS	NS	NS	NS	NS	0	3	3	9	3	8	0	NS
DR04	NS	NS	NS	NS	NS	NS	NS	0	0	4	0	0	0	1	0	NS
DR05	NS	NS	NS	NS	NS	NS	NS	0	0	1	3	9	0	5	0	NS
DR06	Y	0	0	0	0	0	1	0	0	1	0	2	0	0	0	NS
DR07	Y	3	2	0	1	0	1	2	1	7	0	0	0	0	0	NS
DR08	0	0	1	0	1	0	0	0	1	2	0	0	0	0	0	NS
DR09	Y	1	4	0	0	0	0	0	0	0	0	0	0	0	0	NS
DR10	Y	2	1	0	2	1	0	0	1	2	0	10	0	0	0	NS
DR11	NS	NS	NS	NS	NS	NS	NS	NS	0	0	NS	NS	NS	NS	NS	NS
DR12	Y	1	14	2	1	2	0	0	1	1	0	0	0	NS	NS	NS
DR13	0	0	0	1	0	0	0	0	0	1	0	0	0	NS	0	NS
DR14	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	NS
DR15	0	0	1	1	1	0	0	0	0	1	0	6	0	0	0	NS
DR16	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	0	0	NS	0	NS
DR17	0	2	1	1	1	0	0	NS	NS	NS	0	0	0	0	0	NS
DR18	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS
DR19	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	0	0	NS
DR20	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	0	NS
DR21	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS
DR22	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS
DR26	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS
LHC23	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS
LHC24	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS
LHC25	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS
Slide	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	0	0	0	0	0	NS

Table 4.    

Summary of California red-legged frog metamorph counts at La Honda Creek Open Space Preserve from 2008 to 2023.

[Unpublished data (not publicly available at the time of publication) were used with permission from the Midpeninsula Regional Open Space District on November 25, 2024. Abbreviations: DR, Driscoll Ranch; NS, site was not sampled for metamorphs that year; Y, site had California red-legged frog metamorphs but no count is available; —, no data; LHC, La Honda Creek Open Space Preserve]

Table 4.    Summary of California red-legged frog metamorph counts at La Honda Creek Open Space Preserve from 2008 to 2023.

Pond	2008	2009	2010	2011	2012	2013	2014	2015	2016	2017	2018	2019	2020	2021	2022	2023
DR02	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS
DR03	NS	NS	NS	NS	NS	NS	NS	NS	33	7	25	198	30	0	11	4
DR04	NS	NS	NS	NS	NS	NS	NS	NS	5	2	2	2	0	0	0	0
DR05	NS	NS	NS	NS	NS	NS	NS	NS	17	11	9	10	19	15	30	5
DR06	0	0	0	3	0	0	NS	NS	11	0	0	0	0	0	0	NS
DR07	0	0	0	0	2	1	2	NS	0	21	1	0	0	0	0	0
DR08	0	0	0	0	0	0	0	NS	0	1	0	0	0	1	0	0
DR09	Y	0	0	4	0	0	0	NS	3	0	0	0	0	0	0	NS
DR10	Y	1	0	8	0	0	0	NS	0	3	0	15	0	0	0	0
DR11	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS
DR12	0	0	0	0	0	0	NS	NS	0	0	0	0	0	NS	NS	NS
DR13	0	0	0	0	0	0	NS	NS	0	0	0	0	0	—	0	NS
DR14	0	1	11	0	0	0	0	NS	0	5	0	0	0	—	0	2
DR15	0	0	0	1	0	0	NS	NS	1	0	0	0	0	—	0	NS
DR16	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	6	0	0	0	NS
DR17	0	0	0	0	0	0	NS	NS	NS	NS	3	0	0	0	0	0
DR18	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	—	NS	NS
DR19	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	0	0	NS
DR20	NS	NS	NS	NS	NS	NS	NS	NS	NS	23	NS	NS	NS	—	0	NS
DR21	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	—	NS	NS
DR22	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	—	NS	NS
DR26	NS	NS	NS	NS	NS	NS	NS	NS	NS	5	NS	NS	NS	—	NS	NS
LHC23	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	—	NS	NS
LHC24	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	—	NS	NS
LHC25	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	—	NS	NS
Slide	NS	NS	NS	NS	NS	NS	NS	NS	NS	NS	3	0	0	4	0	NS

Although Sierran chorus frogs and Pacific newts were not the focus of amphibian breeding surveys at LHC, incidental observations of egg masses, larvae, and metamorphs of each were recorded during CRLF surveys. In addition, many observations of Sierran chorus frogs have been recorded from the southern half of LHC in the GBIF database (fig. 5; Global Biodiversity Information Facility, 2023b, c) and Pacific newts have been observed at wetlands within LHC during surveys by MROSD staff and throughout the Santa Cruz Mountains surrounding LHC based on GBIF data (fig. 5; Global Biodiversity Information Facility, 2023d).

Non-native American bullfrogs are not yet known to have established a breeding population at LHC. A small number of bullfrogs have been detected during amphibian surveys completed from 2008 to 2023 and removed to limit their spread into the preserve. American bullfrogs have been captured at pond DR16 and likely dispersed to this pond from the main stem of San Gregorio Creek (J. Chung, Midpeninsula Regional Open Space District, written commun., 2023). There are records of American bullfrogs from the landscape surrounding LHC, with the closest known records being a cluster of American bullfrogs recorded from 2020 to 2023 in a wetland less than 1 km east of the eastern boundary of LHC (Global Biodiversity Information Facility, 2023a; fig. 5). This cluster of American bullfrog occurrences is separated from LHC by State Highway 84, but nevertheless is a potential source for this non-native species to disperse into LHC. Based on GBIF data, another American bullfrog population seems to exist in the Windy Hill Open Space Preserve, approximately 3 km east of the northeastern boundary of LHC (fig. 5).

Population Simulations for San Francisco Gartersnake Populations at La Honda Creek Open Space Preserve

The highest viability for reintroduced SFGS subpopulations at LHC occurred when the duration of reintroduction was greater than or equal to 15 years, and all juvenile snakes were released in a single subpopulation. Dividing the propagule of SFGS juveniles into more subpopulations led to a higher quasi-extinction probability for each subpopulation (fig. 6). When all juvenile SFGS were released as part of a single subpopulation, mean abundance of SFGS per subpopulation at the end of the 30-year simulation was higher than if the same total number of SFGS juveniles were divided up and released into more than one subpopulation (fig. 7). Quasi-extinction probability was much higher when annual reintroductions of SFGS juveniles only took place for 5 or 10 years. A 20-year reintroduction resulted in lower quasi-extinction probabilities and higher abundance of SFGS after 30 years, but the difference between a 15-year or 20-year long reintroduction was smaller, indicating diminishing returns with increasing duration after a certain length of time.

Discussion

This report has focused on the suitability of LHC for reintroducing SFGS and the best approach to translocating captive-reared juvenile snakes to their new home. Following up reintroduction efforts by monitoring the reintroduced population is essential to evaluate the success or failure of the endeavor (Ewen and Armstrong, 2007). As an active management intervention, reintroductions are ideally suited to an adaptive management approach in which data are continually analyzed and decisions reconsidered in the light of new evidence (Runge, 2013). Strategies for collecting data from a reintroduced SFGS population can follow the template used at existing populations. Trapping methods for SFGS are well-established, with drift fences and funnel traps proving effective at capturing juvenile and adult snakes (Halstead and others, 2011; Reeder and others, 2015; Kim and others, 2018). The selection of state variables to measure is an important first step in monitoring a reintroduction program (Nichols and Armstrong, 2012). The abundance of SFGS, and the survival, growth, and fecundity of individual snakes are important metrics for determining population viability (Wood and others, 2020; Rose and others, 2022, 2023). Capture-mark-recapture (CMR) methods, in which individual animals are marked such that they can be identified if recaptured, are preferred to using count data from unmarked animals for monitoring reintroductions (Nichols and Armstrong, 2012).

Capture-mark-recapture methods have been used to estimate the abundance of SFGS (Kim and others, 2018), and CMR data can be used to estimate survival and growth rates of marked SFGS if studies continue for 3 or more years (Rose and others, 2022). Based on prior CMR studies, drift fences with funnel traps could be deployed around the focal wetland(s) at which SFGS are released to capture SFGS moving to and from their foraging habitat. Trapping during the spring (April and May) for approximately 8 weeks has provided adequate numbers of recaptures of marked SFGS to estimate abundance using closed CMR models (Wood and others, 2020). Sampling during the spring also enables researchers to determine the reproductive status and fecundity of female SFGS by using a portable ultrasound to scan for the presence and number of developing embryos (Rose and others, 2023; Schoenig and others, 2023a). Capture-mark-recapture trapping would enable monitoring of the sex ratio of adult snakes, which is a key population metric listed in the species status assessment for SFGS (U.S. Fish and Wildlife Service, 2020).

Estimating growth and survival of SFGS requires sampling several consecutive years, with the precision of these vital rates generally increasing with the length of the study. Although 3 years of CMR data is sufficient to obtain an estimate of the annual survival rate, variation in growth and survival rates among years (Rose and others, 2022) demonstrates that longer periods are necessary to capture fluctuations in response to changing environmental conditions. Another metric that can be used to evaluate individual health is body condition, or the mass of animals relative to their length (Peig and Green, 2009). Lower body condition in translocated animals compared to those in established populations can be an indicator of release costs or poor habitat quality at the recipient site (Platenberg and Griffiths, 1999). Because snake body condition can also decline because of environmental fluctuations such as drought (Rose and Todd, 2017), simultaneously monitoring SFGS at extant populations at CR and RR would be useful to understand if changes in body condition at LHC are unique to the reintroduced population or representative of broader regional trends. Although collecting data from multiple populations of SFGS requires substantial investment of time and money, the information gained about vital rates from reintroduced populations can be valuable enough to outweigh the costs (Runge, 2013).

As the reintroduction of SFGS proceeds, it would also be valuable to monitor the dispersal of SFGS beyond the initial reintroduction site. The movement of reintroduced snakes between nearby wetlands connected through suitable upland habitat could increase the probability of persistence for the entire reintroduced population (Brown and Kodric-Brown, 1977). Installing drift fences and funnel traps at wetlands near the reintroduction site could aid in capturing individuals dispersing to new, unoccupied habitats and following individuals moving between established subpopulations. Capture probability of dispersing snakes might be low if few individuals emigrate from the reintroduction site and density at new wetlands is low. Another possibility for monitoring dispersal to new habitats is the use of drift fences with camera traps (Hobbs and Brehme, 2017; Amber and others, 2021) or automated passive integrated transponder (PIT) tag readers (Walkup and others, 2023), instead of funnel traps. Unlike funnel traps, camera traps and automated PIT tag readers do not need to be checked daily, and the detection of SFGS by a passive sampling method could be used as a decision point to begin using CMR trapping at a pond. Drift fences with camera traps could also be used to encircle the wetland at which SFGS were released, and the dispersal of individual snakes away from the wetland could be tracked through photographs of snakes moving through openings in the drift fence array. Alternatively, radiotelemetry could be used to track the movements of reintroduced snakes. Radiotelemetry was used to track the movements and document mortality of giant gartersnakes (Thamnophis gigas) translocated to constructed wetlands in the Sacramento Valley of California (Nguyen and others, 2023). The smaller size of SFGS compared to giant gartersnakes could necessitate smaller transmitters with shorter battery life, and limit tracking to the largest adult female snakes only. Further, the complex topography and dense vegetation in parts of LHC could make locating SFGS using radiotelemetry challenging.

The focus of this report has been the demographic viability of reintroduced SFGS populations, but maintaining genetic diversity also is key to establishing a viable, self-sustaining population (Schwartz and others, 2007; Van Rossum and Hardy, 2022). Reintroducing SFGS from CR and RR, and potentially other source populations, to LHC would provide increased genetic diversity compared to either donor population alone, and could help alleviate concerns of inbreeding depression in the reintroduced population. In other taxa, the introduction of individuals from a genetically divergent population of the same species has been shown to increase genetic diversity and adaptive potential (Chan and others, 2019). Genotyping all SFGS that are released at LHC, and following up by genotyping SFGS that are born in the reintroduced population, could enable scientists and managers to track which individuals are contributing to future generations (Van Rossum and Hardy, 2022).

The data collected during the monitoring of SFGS demography, genetic diversity, and status of CRLF populations can all be integrated into an adaptive management approach to the reintroduction of SFGS. Advocates of an active adaptive management approach often argue for an experimental approach to reintroduction, in which different methods are attempted simultaneously, and their success is evaluated with empirical data (McCarthy and others, 2012). Regarding an endangered species, there is a heightened concern for the fate of translocated animals because the effects of removing animals based on the viability of the donor population are paramount. Still, adaptive management can be used to design a monitoring plan and to make decisions in response to data collected from the reintroduced population as the reintroduction proceeds. If data on the vital rates of individual SFGS and abundance of the reintroduced SFGS population indicate the population is declining or failing to grow fast enough to be viable over the next 20 years, it might be necessary to revisit the reintroduction strategy and make changes that will lead to greater population growth. For example, although population simulations indicated that the release of captive-reared juvenile SFGS resulted in the highest viability for the donor and reintroduced populations (Rose and others, 2023), monitoring data could indicate that survival of juvenile SFGS is low at the reintroduction site and, as a result, the population is not growing as expected. If juvenile survival in the wild was lower than expected, then this updated information could be incorporated into the demographic model and stakeholders could evaluate if releasing adult SFGS or mixed-age propagules were preferable alternatives. Monitoring data on the abundance and vital rates of SFGS could also indicate the population is growing faster than expected. In that case, the decision could be made to stop the addition of new snakes at the first reintroduction site and shift reintroductions to a new wetland, if the reintroduced population seems demographically and genetically viable.

Translocating animals to a location that is unoccupied requires the commitment of resources from many stakeholders over several years. Given the proximity of LHC to a known SFGS population approximately 4 km to the east at RR, it is worthwhile to consider if SFGS could naturally colonize LHC without human assistance. The intervening habitat between the eastern boundary of LHC and the western boundary of RR includes multiple wetlands, grassland and scrub that could act as a dispersal corridor. The primary obstacle to natural SFGS recolonization of LHC is barriers to dispersal along the eastern boundary of LHC. State Highway 84 runs north to south along a steep canyon near and along the eastern boundary of LHC. The steep slope and shaded redwood forest habitat in the canyon might present a natural deterrent to dispersal for SFGS, which prefer shrub cover and less densely forested woodlands. State Highway 84 could present a barrier to dispersal of SFGS, either behaviorally if snakes avoid crossing an open road where they are exposed to predators, or demographically if vehicles might kill the occasional SFGS attempting to disperse further west. Also, the town of La Honda east of LHC has a series of smaller roads and suburban development that is inhospitable to SFGS. The additional barrier presented by State Highway 84 makes natural colonization of LHC by SFGS less likely than if LHC and RR were connected by an unfragmented landscape.

The International Union for the Conservation of Nature (IUCN) recommends that any proposal for translocating wildlife should consider the potential benefits and costs (International Union for Conservation of Nature, 2013). A primary concern for reintroducing SFGS to LHC is the potential for negative effects on the donor SFGS populations at CR and RR. In simulated reintroductions, minimizing the number of adult snakes removed from the donor populations by releasing captive-reared juvenile snakes resulted in the lowest effect on the demographic viability of simulated donor populations (Rose and others, 2023). Beyond demographic viability, the genetic effects of translocating snakes from CR and RR to LHC must be considered. The CR and RR populations have low genetic effective population sizes (N_e; Wood and others, 2020), an indicator of reduced genetic diversity and possible risk of inbreeding depression. Therefore, the removal of snakes from CR and RR for translocation to LHC could cause further decreases in genetic diversity in these existing populations. Negative effects on donor SFGS populations at CR and RR could be mitigated by only transferring a fraction of captive-born neonates to LHC and returning the rest to the donor population. Also, in the future, reciprocal translocation of SFGS between populations within the same genetic cluster (for example, CR, RR, and Año Nuevo State Park) or even among the southern and northern genetic clusters could be necessary to increase genetic diversity and stave off negative effects of inbreeding depression (Heber and others, 2013; Forsdick and others, 2017; Grueber and others, 2017).

Another potential risk from the reintroduction of SFGS to LHC is that adding these predators to wetlands could have negative effects on native CRLF populations. The small founder population of SFGS is unlikely to have dramatic effects on CRLF populations in the first 3 to 5 years following translocation. Over time, if the SFGS population increases in abundance, SFGS could represent one of the main predators on CRLF at LHC, along with resident Santa Cruz gartersnakes. Given the long coevolutionary history of CRLF coexisting with gartersnakes, including SFGS, throughout California, it is unlikely that the presence of SFGS alone would drive a population of CRLF to extinction. San Francisco gartersnakes and CRLF coexist at RR and CR, as well as other extant SFGS populations. San Francisco gartersnakes feed on a variety of amphibians and fish (Larsen, 1994; U.S. Fish and Wildlife Service, 2020), and as generalists are unlikely to deplete CRLF populations. Also, large CRLF likely exceed the gape of SFGS, and the existence of a size-refuge could aid in CRLF persistence with gartersnake predators.

To quantify if the addition of SFGS at LHC has a negative effect on CRLF populations, egg mass counts can be used to track the abundance of this threatened frog species. Collecting a time series of egg mass counts at several wetlands before reintroduction would enable the use of a before-after control-impact (BACI) study design (Clinton and others, 2022). Continuing egg mass counts at the focal wetland where SFGS are reintroduced and at surrounding wetlands where no SFGS are present would be a valuable complement to monitoring SFGS abundance. Monitoring CRLF egg masses at sites where no SFGS are present is essential for determining whether any changes in the CRLF egg mass count at the focal wetland(s) where SFGS are reintroduced are the result of broader trends (for example, in response to interannual changes in precipitation) or occurred in response to predation by SFGS. Surveying for all native and non-native pond-breeding amphibians in the focal wetland(s) and wetlands without SFGS would further allow for quantification of community-level effects of SFGS reintroduction (Clinton and others, 2022). Based on the negative effects of American bullfrogs on CRLF documented elsewhere (Fisher and Shaffer, 1996; Lawler and others, 1999; Anderson and others, 2019), the spread of American bullfrogs into wetlands at LHC would represent a much greater threat to CRLF persistence than the reintroduction of SFGS.

In addition to CRLF, other potential amphibian prey species are widespread at LHC, including Sierran chorus frogs and Pacific newts. Sierran chorus frogs are generalists that breed in a variety of wetlands including seasonal and permanent ponds (Stebbins, 2003). Sierran chorus frogs may make up the bulk of SFGS diet during the late winter and spring (Kim and others, 2021), when adult frogs migrate to water bodies to breed and metamorphs (juvenile frogs that have recently undergone metamorphosis from the larval stage) emerge from the water. Sierran chorus frogs produce large cohorts of larvae and metamorphs that leave the water in large numbers in the late spring and early summer. Because gartersnakes are gape-limited predators (Shine, 1991), small Sierran chorus frogs might be particularly important prey for juvenile SFGS (Larsen, 1994; Kim and others, 2021). The high tolerance that SFGS have for the tetrodotoxin produced by Pacific newts suggests that newts are an important part of the diet for SFGS (Brodie and others, 2002). Pacific newts breed in the winter months in ponds, reservoirs, and stream pools. Pacific newt larvae take several months to develop into adult newts and generally emerge from the water in the late summer or early fall. Water bodies that support breeding by a variety of different amphibian species, which each can have different breeding periods and larval stages of varying duration, likely benefits SFGS by providing multiple sources of prey that are abundant at different times of year. The presence of Sierran chorus frogs and Pacific newts at several wetlands within LHC further bolsters the likelihood that the preserve has suitable prey for reintroduced SFGS.

If the reintroduction of SFGS to LHC is pursued in the future, several outstanding questions remain about the methods used to rear juvenile snakes in captivity and release snakes into the wild. For captive-reared snakes, it is unclear if “enrichment” by recreating more natural environments in captivity provides benefits once snakes are released into the wild (Roe and others, 2015). A potentially important decision is whether to use a “soft release” method, in which snakes are first released into enclosures at the reintroduction site, or a hard release in which snakes are released and free to disperse immediately. Evidence for the benefits of a soft release for translocated snakes are mixed (Sacerdote-Velat and others, 2014; Cornelis and others, 2021), and building enclosures that allow for foraging by semiaquatic gartersnakes would be difficult. Finally, translocating wild animals and rearing animals in captivity raise concerns about the risk of transmitting disease (Sainsbury and Vaughan-Higgins, 2012). Snake Fungal Disease (SFD) caused by the pathogen Ophidiomyces ophiodiicola is an emerging risk to wild snake populations (Lorch and others, 2016). Screening captured and captive-reared SFGS for SFD could be important to prevent spread of the fungal pathogen between populations. These choices and other decisions could benefit from convening a panel of experts in gartersnake ecology, animal translocation, and captive-rearing to develop plans for each step of the process, data to collect to evaluate efficacy of chosen methods, and decision points to reevaluate the chosen methods should initial efforts result in negative outcomes for reintroduced SFGS.

Conclusions

La Honda Creek Open Space Preserve (LHC) is a suitable candidate site for a trial reintroduction of San Francisco gartersnakes (SFGS) to unoccupied habitat within its historical range. The climate, habitat, prey community, and land use at LHC are all similar to nearby open space preserves that support extant populations of SFGS. Based on our analysis, the following conditions could increase the likelihood of successful population establishment: annual releases of SFGS for more than 10 years, initial releases at a single wetland (in other words, founding one subpopulation of SFGS), and annual monitoring of the reintroduced management approach, where decisions can be made quickly to change management actions based on the health, genetic diversity, and vital rates of reintroduced SFGS. Furthermore, excluding cattle from the vicinity of the focal wetland at which snakes are reintroduced would likely improve the viability of the population by providing foraging and refuge habitat. Before reintroduction begins, many decisions will need to be made about the number, source, and specific destination of translocated animals. Bringing together the many stakeholders concerned with wildlife conservation, open space preservation, and maintaining working landscapes could be vital to ensuring that the reintroduction of SFGS to LHC is successful.