Range‐wide persistence of the endangered arroyo toad (Anaxyrus californicus) for 20+ years following a prolonged drought

Abstract Prolonged drought due to climate change has negatively impacted amphibians in southern California, U.S.A. Due to the severity and length of the current drought, agencies and researchers had growing concern for the persistence of the arroyo toad (Anaxyrus californicus), an endangered endemic amphibian in this region. Range‐wide surveys for this species had not been conducted for at least 20 years. In 2017–2020, we conducted collaborative surveys for arroyo toads at historical locations. We surveyed 88 of the 115 total sites having historical records and confirmed that the arroyo toad is currently extant in at least 61 of 88 sites and 20 of 25 historically occupied watersheds. We did not detect toads at almost a third of the surveyed sites but did detect toads at 18 of 19 specific sites delineated in the 1999 Recovery Plan to meet one of four downlisting criteria. Arroyo toads are estimated to live 7–8 years, making populations susceptible to prolonged drought. Drought is estimated to increase in frequency and duration with climate change. Mitigation strategies for drought impacts, invasive aquatic species, altered flow regimes, and other anthropogenic effects could be the most beneficial strategies for toad conservation and may also provide simultaneous benefits to several other native species that share the same habitat.


This species was listed under the Endangered Species Act in 1994
after reported declines from approximately 75% of formerly occupied habitat across its range in California (Jennings & Hayes, 1994;Sweet, 1992;USFWS, 1994). Its status was retained upon reevaluation in 2014 (USFWS, 2015). Anthropogenic threats identified by various authors include off-highway vehicle (OHV) use, dam/hydrological operations, disease, and invasive species (Ervin et al., 2006;Funk et al., 2014;Madden-Smith et al., 2003;Miller et al., 2012;Ramirez, 2003;Robeson, 2015;Sweet, 1992). In the United States, the arroyo toad historically occupied 25 watersheds along mostly coastal and a few desert drainages from Monterey County to San Diego County (Ervin et al., 2013;USFWS, 2015). The 1999 USFWS Recovery Plan for the arroyo toad lists 20 (but actually 19 due to a misidentification; see Ervin et al., 2013) populations at specific locations that must be self-sustaining for a downlisting consideration (USFWS, 1999, pp. 75-76). According to the Recovery Plan, selfsustaining populations are defined as "having successful recruitment equal to 20% or more of the average number of breeding individuals in seven of ten years of average to above-average rainfall amounts with normal rainfall patterns" (USFWS, 1999, p. 76).
Given this prolonged period of drought, there has been growing concern that the number of consecutive years of drought may have surpassed the lifespan of the species , and recruitment may have been severely diminished due to lack of surface water (especially in ephemeral watersheds), resulting in possible population declines and local extirpations. Additionally, evidence of direct mortality of toads due to drought was reported during a telemetry study that included observations of desiccated toads found under the sand in which they had burrowed (Gallegos, 2011(Gallegos, -2013(Gallegos, , 2016. These concerns prompted collaborative, range-wide surveys for the arroyo toad in 2017 that continued with several additional surveys through 2020.
We investigated population status by surveying known historical arroyo toad locations within the United States and compared the locations where toads were detected/not detected to locations where they were extant in 1999 (the time the recovery plan was written; USFWS, 1999) and in 2014 (the time of the last reevaluation of their status; USFWS, 2014USFWS, , 2015. To cover the extent of the historical locations within the United States, we formed a collaboration of researchers to comprehensively survey as many historical sites as possible from Monterey County to San Diego County from 2017 to 2020 and combined our detection/non-detection findings.

| Study area
Our study area included all the watersheds in the United States within the known range of the arroyo toad that were delineated in the 1999 Recovery Plan (USFWS, 1999). A few locations have been updated and revised from the original Recovery Plan to account for corrections made after publication (Ervin et al., 2013;USFWS, 2014). The revised total includes 25 watersheds spanning Monterey to San Diego counties. Multiple sites within these watersheds were surveyed to determine presence ( Figure 2

| ME THODS
We collaborated as 37 partners from 19 various state and government agencies, consulting groups, universities, and independent researchers, to survey for arroyo toads at historical locations. We compiled range-wide comprehensive and current data for the detection/non-detection of the species at as many sites as possible that were listed in the 1999 Recovery Plan and other literature or databases (mostly the USFWS 2014 Species Report). Most of our surveys were conducted during 2017, but several sites initially skipped for logistical reasons were surveyed 2018-2020.
Incidentally, the winter of 2018-2019 had more rainfall compared to the surrounding years; therefore, toads were expected to be more easily detectable during that year. Location descriptions and Global Positioning System (GPS) coordinates from the Recovery Plan, grey literature, and from biologists who had been to the sites were used to determine the precise multiple locations to survey within the 25 watersheds. Because arroyo toads have been documented as having an average dispersal distance of ~3 km (USFWS, 1999), long swaths of habitat were surveyed within each documented location to account for movement of arroyo toads up and down waterways even if some of the habitat was marginal. A location within a watershed was regarded as a "site" if it had historical records of arroyo toads described for that specific location within the watershed. We also considered the ~3 km average dispersal distance (USFWS, 1999) and any geographic barriers (i.e., mountains, urban development) to establish which locations we could regard as being a single "site" versus more than one "site" within a watershed. Our team of collaborators developed a spreadsheet of arroyo toad sites to survey, categorized by watershed and recovery unit. The spreadsheet included the following: (1) a list of all known arroyo toad historical sites based on literature (mostly from the 1999 Recovery Plan and USFWS 2014 Species Report), and (2) fields for participants to provide date surveyed, specific location, and age class observed. Collaborators throughout southern California conducted surveys at as many locations as possible, mostly according to their proximity to nearby sites. Participants conducted daytime and/or night surveys during the breeding and active season of the toad (generally April-July depending on elevation, latitude, and local climate). Surveyors walked the creeks at historical locations surveying visually and dip-netting for tadpoles during the day. There was no minimum or maximum number of linear meters walked; the presence of suitable (and even marginal) arroyo toad habitat dictated the length of creek surveyed.
If no toads or larvae were detected during the day, most surveys were continued at nighttime along the same length of creek and in the same manner by looking but also listening for calling adults.
One survey (day or day/night) per site was made, although occasionally different participants happened to overlap the same site.
Data from all participants were compiled and number of locations where toads were detected/not detected were compared to number of locations where toads were recorded as extant in the 1999 Recovery Plan. Years that toads were last documented from all sites were also compiled (Appendix).
We also examined weather data, reports, unpublished data, and gray literature from past surveys conducted by U.S. Geological Survey (USGS), U.S. Forest Service (USFS), U.S. Fish and Wildlife F I G U R E 2 Of the 115 known sites in the United States, 88 were surveyed as part of this effort and arroyo toads were detected at 61 (see also Appendix) Service (USFWS), and other partners to assess whether any anomalous events may have affected population presence or detectability (i.e., major local weather events or anthropogenic changes to the habitat) at any sites. To gain perspective on climate change in the region, we compiled literature and online climate data for California's South Coast Drainage. Our precipitation and temperature profiles were produced from data on the NOAA website by selecting the "divisional" tab at the top, then "time series," then parameters from the drop-down menus for temperature or precipitation, annual average, bounding years , "state" (California) and "division" (south coast drainage). Precipitation measurements were converted to millimeters and temperature was converted to degrees Celsius. These data were graphed in Microsoft Excel™ to show the difference in temperature and precipitation from the mean over time.

| RE SULTS
Of the more than 70 individuals asked to participate in surveys, we received data and input from 37 researchers and citizen scientists. Our review of published and gray literature, and unpublished data, did not uncover any localized novel impacts that might suggest causes for a population crash or extirpation (besides the known drought). Potential threats such as OHV use, hiking, camping, bathing, trash, and exotic species were recorded at nearly all sites and all years surveyed. Also, several sites with known toad populations had been closed to public use for a prolonged period to protect the species from direct anthropogenic impacts (USFS, personal communication). We did not quantify prevalence of disturbances or collect data on disturbances over time.

| DISCUSS ION
We compiled results for 2017-2020 arroyo toad surveys conducted at all historical watersheds and most of the historical sites, hypothesizing that we would document numerous extirpations due to prolonged periods of drought. Given the lifespan of the toad (average 7-8 years; Fisher et al., 2018), the prolonged duration of the drought, and the comparison to extant sites from 1999 (20+ year duration), we expected to find fewer extant populations than we did.
The short duration of time for our surveys (2017-2020) could have also produced an underestimation the number of extant populations.
Toads were not detected at about 31% of the sites surveyed. Our data show that over the past 20+ years, this species has persisted in ~80% of the watersheds and ~69% of the sites surveyed, but has possibly disappeared from ~31% of these locations. Though this may imply arroyo toad persistence at the majority (~69%) of sites, we do not know if arroyo toad populations are stable or self-sustaining at these sites, which is one of the Recovery Plan metrics (see USFWS, 1999, p. 76). Additional surveys are needed to determine if these extant populations are declining or at risk of extirpation. Overall, we consider the 61 extant sites to be a minimum estimate of extant populations because we did not survey 27 of 115 sites and because it is possible to have missed detection at some sites.
Though this species is known to be adapted to the generally hot, dry climate of southern California, increased drought severity and length may eventually surpass the limits of this species' tolerance. Toads are more terrestrial than frogs and are known to have physiological adaptations for water retention, such as storing water in their bladder or metabolically producing water from their diet (Bundy & Tracy, 1977;McClanahan & Baldwin, 1969). Schmajuk and Segura (1982) show that toads in the Bufo boreas group specifically store more water in their bladder when deprived of it, and Jørgensen (1994) reports that the common toad (B. bufo) can retain up to 20% of its mass as water in the bladder when water deprived. Therefore, the xeric-adapted arroyo toad likely uses this strategy to retain water through prolonged drought. Furthermore, arroyo toads may benefit from moderate drought because suitable conditions for breeding and metamorphosis generally occur in the form of slowmoving braided streams when water levels are low. Though without the typical cycle of flooding and scouring events, habitat that is ordinarily lightly or moderately vegetated can fill in with riparian vegetation-including both native species such as mulefat (Baccharis salicifolia), cattail (Typha spp.), willow (Salix spp.), and invasive species such as giant reed (Arundo donax)-which can overtake areas formerly suitable for arroyo toad breeding (Brehme et al., 2006;Griffin & Case, 2001). While arroyo toads have persisted at most sites despite variable precipitation, our inability to detect the species at approximately 27 sites at which they were previously found suggests that the species is likely continuing to decline. In addition, desiccated adults documented by telemetry during drought years (Gallegos, 2011(Gallegos, -2013(Gallegos, , 2016, suggest that estivating toads are not impervious to drought effects on soil moisture. Negative impacts from recreation, non-native species, and altered hydrological regimes were documented at several locations (Ervin et al., 2006;Madden-Smith et al., 2003;Matsuda et al., 2018;Miller et al., 2012) and may exacerbate the environmental challenges being experienced by these toads. More data are needed to quantify threats such as OHVs, habitat conversions, hydrological changes from dams, disease (i.e., chytrid fungus (Batrachochytrium dendrobatidis); Sweet & Sullivan, 2005), and predation, competition, or habitat manipulation from non-native species (Richmond et al., 2021). Anthropogenic threats may also impact other native species associated with arroyo toads; therefore, addressing these threats may be a tractable and effective way to protect a suite of native species. For example, arroyo toads share or have historically shared habitat with several native common or special status species including western toad (Anaxyrus boreas), two-striped garter snake (Thamnophis hammondii), red-sided gartersnake (Thamnophis sirtalis infernalis), Santa Ana sucker (Catostomus santaanae), unarmored threespine stickleback (Gasterosteus aculeatus williamsoni), California red-legged frog (Rana draytonii), western spadefoot (Spea hammondii), and western pond turtle (Actinemys pallida) (Richmond et al., 2013Richmond, Jacobs, et al., 2014;Sweet & Sullivan, 2005). Anthropogenic alteration of habitat for waterplay (e.g., damming to create pools) and releasing unwanted pets (e.g., turtles, aquarium fish) or game fish for fishing often makes areas incompatible for native species and can promote persistence of non-native species (Miller et al., 2012).
Maintaining natural shallow braided aquatic systems with sandy substrates and periodic drying may prevent many invasive predatory species from establishing by eliminating the pooled areas in which they are able to persist (Miller et al., 2012). Unfortunately, shallow braided streams and terraces with sandy substrate are also favored as locations for OHV use, which can be especially damaging to toad populations during the breeding and postbreeding season when eggs, larvae, and metamorphs are reliant on surface water (Ervin et al., 2006;Griffin & Case, 2001). OHV use can cause direct mortality by crushing individuals burrowed under the soil or have indirect effects by habitat modification (e.g., soil compaction), thus reducing or preventing friable sands in which they burrow (Griffin & Case, 2001;Sweet, 1992). research on exotic species interactions, toad movements, habitat analyses, and surveying areas within the potential range of the species (Brehme et al., 2006;Ervin et al., 2006Ervin et al., , 2013Fisher et al., 2018;Gallegos, 2011Gallegos, -2013Gallegos, , 2016Madden-Smith et al., 2003;Matsuda et al., 2018;Miller et al., 2012;Ramirez, 2003). This study contributes to the recovery tasks by providing the most comprehensive and up-to-date information on extant arroyo toad populations throughout their range in the United States. However, this study also had several limitations.
Trying to cover the entire United States range while collecting and reporting data in a consistent manner was challenging due to the engagement of so many participants. A more stringent study design with fewer participants may have increased consistency of data collection methods and allowed for more rigorous analyses on occupancy; however, with fewer participants we may not have been able to survey as many sites. Population trend data and multiple visits over multiple years to all sites could also have improved our ability to accurately determine occupancy over time and help investigate one of the Recovery Plan's metrics (to document self-sustaining populations "…equal to 20% or more of the average number of breeding individuals in seven of ten years…" see USFWS, 1999, p. 76). However, our main objective was to try and detect arroyo toads at as many of the historical sites as possible to provide a comprehensive understanding of which populations were still on the landscape and provide a baseline for future stud- ies. This study provides information on which sites still need to be verified for toad persistence, and it may help identify additional sites within the range of the arroyo toad that could be explored for yet-unknown populations (another metric of the Recovery Plan; USFWS, 1999). By identifying currently occupied sites, the study also could lead to new assessments of management at those sites.
These baseline data documenting the current occupancy status of the species, which had not been explored comprehensively or con- This may also involve conducting repeated surveys during optimal years at the 27 non-detection sites and 27 sites not surveyed.
Our comprehensive surveys confirmed that toads are extant at ~69% of sites; toads were not detected at ~31% of sites. Detection at the majority of sites suggests that arroyo toads may be better evolutionarily suited to the effects of drought cycle changes than previously understood. However, we emphasize that any tolerance to drought is not well-studied. We suggest that minimizing anthropogenic impacts (including introduced aquatic invasive species) to historically and currently occupied sites may be the most effective strategy for arroyo toad conservation; this approach can also have positive implications for native species sharing the same habitat. The results of this study can inform recovery planning for the arroyo toad.

ACK N OWLED G M ENTS
We are grateful to the many people who contributed to the develop-

CO N FLI C T O F I NTE R E S T
None. Data curation (supporting); Investigation (supporting). Jaquelyn J.

AUTH O R CO NTR I B UTI O N S
Hancock: Data curation (supporting); Investigation (supporting).

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available within this article (  1991, 1993-4, 1999, 2000, 2007, 2017 3. Santa Ynez River Basin (Santa Barbara County)