Survey for Morro Bay Kangaroo Rat: A Rare Mammal of Uncertain Status

The Morro Bay kangaroo rat Dipodomys heermanni morroensis (Heteromyidae) occupies sandy soils bordering the eastern and southern margins of Morro Bay, specifically in and near Los Osos in western San Luis Obispo County, California (Figure 1). This small (adult body mass ∼ 65 g), granivorous, nocturnal, burrowing rodent (Roest 1984) was listed as endangered pursuant to the U.S. Endangered Species Conservation Act in 1970 (USFWS 1970), the California Endangered Species Act (1971) in 1971, and the U.S. Endangered Species Act (ESA 1973, as amended) in 1973. In addition, it was designated as fully protected pursuant to the California Fish and Game Code in 1970 (California Fish and Game Commission 1970). Field research from the late 1950s to mid-1980s (reviewed in Kofron and Villablanca 2016) documented a rapid population decline. Despite many searches, including one from 2008 to 2012 by two of us (Kofron and Villablanca 2016) that covered most of the known geographic range, the Morro Bay kangaroo rat has not been captured or seen in the wild since 1986. The last captive individual died in 1993 (Thompson et al. 1995).

Kofron and Villablanca (2016) reviewed the biology and conservation status of the Morro Bay kangaroo rat, including taxonomy and genetics, soil type affinity and burrow systems, history of decline, suspected primary causes of decline, breeding in the wild and in captivity, habitat restoration, and threats. Based on field surveys, they speculated that if the Morro Bay kangaroo rat still exists, then it is at extremely low density in a few isolated colonies. In other words, if it still exists, it will be difficult to detect. Therefore, Kofron and Villablanca (2016) recommended that searches continue in the public lands where they observed the greatest concentrations of potential signs in 2011, specifically at Pecho South (Morro Dunes Ecological Reserve West, and Montaña de Oro State Park; Figure 1) and Junior High/Santa Ysabel (Morro Bay State Park). Although these areas are only part of the geographic range, they are expansive.

The Morro Bay kangaroo rat is the only kangaroo rat in the vicinity of Morro Bay. Of the nine subspecies of D. heermanni, the Lompoc kangaroo rat D. heermanni arenae is the closest geographically (Kelt 1988), with a separation distance of 19 km (Souza 1958). In addition, of the four subspecies of D. heermanni that are most proximate geographically, the Lompoc kangaroo rat is the closest genetically (Villablanca 2007). Although Benedict et al. (2019) dispute the subspecies taxonomy within D. heermanni, the populations referred to the Lompoc kangaroo rat are still the most proximate geographically and genetically, regardless of their taxonomic status.

Previously, Duggan et al. (2011) used wildlife scent-detection dogs and then trapped at alert sites for ground squirrels, subsequently recommending this two-stage approach for cryptic rodents. Accordingly, Kofron and Villablanca (2016) recommended searching for the Morro Bay kangaroo rat using 1) a wildlife scent-detection dog trained on the Lompoc kangaroo rat, and 2) baited camera traps at sites with potential signs of the Morro Bay kangaroo rat. In this manner, photographic images would provide direct evidence of its presence. Most recently, Thomas et al. (2020) reported that scent dogs and camera traps are highly effective at detecting threatened, cryptic small mammal species existing at low density. Herein, we 1) report on a survey for the federally listed Morro Bay kangaroo rat using a wildlife scent-detection dog and baited camera traps in 2016 and 2017, 2) discuss the implications of the results, and 3) make recommendations for future actions.

We used a wildlife scent-detection dog (male Dutch Shepherd; Figure 2) to search for the Morro Bay kangaroo rat in 2016. In 2014, while the dog was a juvenile, we selected it for training because of its enthusiasm for the outdoors and its gentle demeanor, which are desirable qualities for purposeful labor in the field and for potential encounters with special-status wildlife species. Two of us (LO, AT) are experienced professional handlers of wildlife scent-detection dogs (Boroski and Oliver 2018; Swiecki et al. 2018; Oliver et al. 2020), and we previously trained this dog to search for the San Joaquin kit fox Vulpes macrotis mutica and for carcasses of bats and birds at energy project sites in central California.

Specifically, we trained our wildlife scent-detection dog to recognize the scent of Lompoc kangaroo rats by using their fecal pellets; we tested the dog, and then conducted maintenance training over 10 mo. We obtained fresh fecal pellets from Lompoc kangaroo rats that were captured in washed and autoclaved Sherman XL live traps near Black Lake (37 km southeast of Morro Bay), San Luis Obispo County. We used a reward-based training method (Smith 2018) that associated alerting (giving a visible response to the handler) to the scent of fecal pellets of the Lompoc kangaroo rat with delivery of a favorite toy to the dog, along with affection from the primary handler/owner (Figure 2). Training comprised discriminating the scent of Lompoc kangaroo rats from the scents of co-occurring small mammal species, including the California pocket mouse Chaetodipus californicus, California mouse Peromyscus californicus, and deer mouse P. maniculatus. We verified the dog's ability to detect and alert to not only the scent of fecal pellets of Lompoc kangaroo rats, but also to the scent of live Lompoc kangaroo rats in the field. We used a line of washed and autoclaved Sherman XL live traps (spaced at 40-m intervals at Black Lake) that contained Lompoc kangaroo rats, other small mammals, or were empty. The handler was blind to which traps contained which treatment. The dog responded correctly at every trap, alerting at every trap that contained a Lompoc kangaroo rat while ignoring every trap that contained some other small mammal or was empty. In final testing, the dog found and alerted to 82% of hidden targets with fecal pellets of the Lompoc kangaroo rat (n = 40), while committing zero false-positive errors. Our training and use of the wildlife scent-detection dog in the field was in accordance with best standards and practices (Boroski and Oliver 2018).

After training and testing were complete, we searched with the wildlife scent-detection dog on 9 to 11 August 2016 for D. heermanni in four historically occupied areas that are currently on public lands (Figure 1): Pecho South, Junior High/Santa Ysabel, Bayview (Morro Dunes Ecological Reserve East) where the last capture of a Morro Bay kangaroo rat occurred in 1986, and in vicinity of the parking lot in north Hazard (Montaña de Oro State Park). Our goal was to search specifically on public lands where the Morro Bay kangaroo rat existed in the 1980s, and where we or James H. Brown (mammalogist, University of New Mexico, Albuquerque, personal communication 2015) saw potential signs of kangaroo rats in 2011 and 2013. Then, between 1 September 2016 and 17 March 2017, we operated baited camera stations at two alert sites and two sites of special interest (where the dog checked several burrows repeatedly but did not alert). Each baited camera station contained two Moultrie I-45 Game Spy digital cameras with infrared flash on u-posts 1 to 2 m distant from and aimed at a vertical feeding tube that dispensed rolled oats via gravity as an animal removed this bait from a pile on the substrate. For each of the four stations, we counted a trap night only if both cameras functioned properly and if some bait remained the next day.

Because of the expansive landscapes at Pecho South and Bayview, we were able to search for Morro Bay kangaroo rats with the wildlife scent-detection dog in only a small sample of the numerous patches of habitat (Figure 2). Therefore, our detection probability was biased downwards, relative to the high detection probability shown in Thomas et al. (2020), who were able to search the entire potentially occupied area. In contrast, our search was complete at Junior High/Santa Ysabel and in vicinity of the parking lot at north Hazard. The dog alerted at two sites in Pecho South (35°18′ 0.93″N, 120°51′ 48.14″W; 35°18′ 14.98″N, 120°52′ 4.83″W) and showed special interest at two sites in vicinity of the parking lot at north Hazard (35°17′ 57.26″N, 120°52′ 19.43″W; 35°17′ 56.73″N, 120°52′ 18.9″W). The four sites were in open, low-growing coastal sage scrub with elements of Morro manzanita Arctostaphylos morroensis in Montaña de Oro State Park, and comprised typical habitat for the Morro Bay kangaroo rat (Stewart and Roest 1960). Our records showed that we previously observed potential signs of Morro Bay kangaroo rats near the two alert sites in 2011, and we received a reliable report of potential signs near the two sites of special interest in 2013 (JH Brown, personal communication 2015). The dog did not alert or show special interest at Junior High/Santa Ysabel or at Bayview.

After eliminating some trap nights because of camera malfunctions or because all bait had been removed, we had collected approximately 4,000 photographic images from eight cameras over 84 cumulative nonconsecutive nights. The images of small mammals documented no Morro Bay kangaroo rats. The small mammals observed at each of the four sites were similar, including California pocket mouse, California mouse, and deer mouse. In sum, our survey for the Morro Bay kangaroo rat produced negative results. However, a degree of uncertainty remains because the scent-detection dog committed no false-positive errors during final testing, and it was excited and persistent in searching at each alert site. The possibilities are the scent dog committed false-positive errors in searching for the Morro Bay kangaroo rat, or alternatively the scent dog detected the Morro Bay kangaroo rat but subsequent camera trapping failed to detect it.

Our survey on public lands known to previously contain Morro Bay kangaroo rats, and that had potential signs in 2011 and 2013, produced no evidence (photographic images) of this taxon, despite using a trained wildlife scent-detection dog and baited camera traps at alert sites. Unfortunately our data combined with all other existing data do not allow us to conclude whether the Morro Bay kangaroo rat is extinct or extant. Essentially, the international standard has not been met to make a definitive determination of extinction. That is, a species should be considered extinct only when there is no reasonable doubt that the last individual has died (International Union for the Conservation of Nature 2012). We acknowledge that because of resource limitation we surveyed only a small sample of the numerous patches of habitat in the expansive landscape.

In addition, we acknowledge the nonsymmetrical inferential logic of our survey, whereby detection of the target species documents its presence, but the lack of detection does not necessarily document its absence. Altman and Bland (1995) criticized studies that purport to prove that something is absent because some studies, like ours for the Morro Bay kangaroo rat, may not have been sufficiently large. We wish our own survey efforts could have been increased by a factor of 10 because substantially more habitat exists on public and private lands that we did not search or did not search sufficiently. Therefore, even when combining our data with all other existing data, it is impossible to exclude the possibility that the Morro Bay kangaroo rat still persists at low density in one or more isolated colonies. This is because of the patchy nature of the remaining habitat in the expansive landscape, the secretive lifestyle of this rodent, a lack of access to private lands, and the possibility it is persisting in an area we did not search.

Because the Morro Bay kangaroo rat had not been captured for 30 y despite many efforts, we previously reported it as possibly extinct (Kofron and Villablanca 2016). However, we caution that more than one-third of mammal species previously considered extinct or possibly extinct have been rediscovered (Fisher and Blomberg 2011), which means that having considered these species as extinct or possibly extinct was incorrect. Uncertainty about whether a rare species is extinct or extant is common because rare species are often difficult to detect (Lee et al. 2017), and it is almost impossible to determine with any degree of certainty whether or not a species is extinct (Roberts 2006). However, the rediscovery of species considered to be extinct is also relatively common (Akcakaya et al. 2017), and each year two or three missing species are found again (Lee et al. 2017), such as the San Quintin kangaroo rat D. gravipes in Mexico (Tremor et al. 2019) and the silver-backed chevrotain Tragulus versicolor in Vietnam (Nguyen et al. 2019). Nonetheless, small missing mammals like the Morro Bay kangaroo rat are disadvantaged because fewer resources (money, time, effort) are expended seeking small missing mammals than larger missing mammals (Fisher 2011; Lee et al. 2017). Importantly, Lee et al. (2017) found the probability of rediscovering a missing mammal most strongly correlates with small body size, like the Morro Bay kangaroo rat. Although small body size decreases the extinction rate, it also decreases detectability and increases the length of time to rediscovery if still extant.

The “known geographic range” (Figure 1) of the Morro Bay kangaroo rat is attributed to Stewart and Roest (1960), who described the geographic range as coinciding with sandy loam soils. We and others have relied on their determination, although the soil description has been problematic. Stewart and Roest (1960) included “the low hills just east of Los Osos Creek” (Extension in Figure 1) as part of the geographic range, which we now know is the only part with sandy loam soil. Most of the currently recognized geographic range is Baywood fine sand (Figure 1; Kofron and Villablanca 2016). The error and confusion might seem semantic. However, because the Morro Bay sand spit, a peripheral area, is comprised mostly of Baywood fine sand (Figure 1; Kofron and Villablanca 2016), in reality it may be habitat and part of the geographic range that is not currently recognized. In the original taxonomic description of the Morro Bay kangaroo rat, Grinnell (1922) stated the following regarding its habitat: “sandy ground in the immediate vicinity of Morro Bay” with altitudinal range from “sea level up to 250 feet” (76 m). Including the sand spit as part of the geographic range would conform to Grinnell's description. In addition, while camera trapping in the foredunes of the Callender Dunes (38 km south of Morro Bay, San Luis Obispo County), we recently obtained photographic images of Lompoc kangaroo rats at the shoreline. The predominant foredunes plant there is European searocket Cakile maritima, which is likewise predominant on the sand spit along with coastal scrub. The available information leads to the following interpretation: the sand spit is part of the geographic range, it has suitable habitat, and it has never been searched for Morro Bay kangaroo rats (Grinnell 1922; Stewart and Roest 1960; Kofron and Villablanca 2016).

Can the negative results from our survey be combined with all other available data to make a definitive determination that the Morro Bay kangaroo rat is extinct? The answer is no, not yet (Altman and Bland 1995; Fisher 2011; Lee et al. 2017). Large areas of public and private lands were previously occupied by the Morro Bay kangaroo rat (Stewart and Roest 1960), and many patches of habitat remain on public lands at Pecho South (75 ha), Bayview (83 ha) and Hazard (74 ha) in 2020, and these need to be more thoroughly searched. In addition, the four private properties to which Kofron and Villablanca (2016) were denied access still need to be searched: the main Buckskin property (26.3 ha; Figure 1) and the property immediately northwest (16.0 ha), one property (5.7 ha) south of Junior High/Santa Ysabel, and one property (21.9 ha) in Baywood Park. Can a determination of “extinct” ever be made if these private properties are not searched? Also, unlike previous researchers (including Kofron and Villablanca 2016), we now consider the Morro Bay sand spit (total area 4.35 km², foredunes 3.75 km²) as potentially part of the geographic range with suitable habitat that has never been searched. Thus, in consideration of all available information, we cannot conclude whether the Morro Bay kangaroo rat is extinct or extant. However, given the situation that some potential habitat has never been searched, and some previously occupied habitat needs to be searched more thoroughly, we conclude that the Morro Bay kangaroo rat must be considered as possibly extant.

Accordingly, we recommend a continuation of search efforts for the Morro Bay kangaroo rat, in particular at the Morro Bay sand spit with camera traps, again at Pecho South and Bayview while attempting to cover a greater sample area with a wildlife scent-detection dog followed by camera trapping, further south in Hazard, and at the four private properties. If extant, the Morro Bay kangaroo will be challenging and difficult to rediscover because of its likely low density and patchy distribution in the expansive landscape, combined with its small size, nocturnal nature, and secretive lifestyle. Nonetheless, searches for a small missing mammal such as the Morro Bay kangaroo rat can be conducted relatively inexpensively, especially if conducted by local experts (Fisher 2011; pers. obs.).

In closing, we point out that kangaroo rats are recognized as keystone species (e.g., Goldingay et al. 1997; Brock and Kelt 2004; Davidson and Lightfoot 2006), which are species intimately involved with maintaining the structure and composition of their ecosystems and whose influence is disproportionate to their numbers or biomass. Keystone species impact other animals and plants across multiple trophic levels and through time (e.g., Hastings et al. 2007; Valone and Schutzenhofer 2007). As a keystone species, the Morro Bay kangaroo rat may impact other federally listed co-occurring species in the landscape: the endangered Morro shoulderband snail Helminthoglypta walkeriana, the federally endangered Indian Knob mountainbalm Eriodictyon altissimum, and the federally threatened Morro manzanita Arctostaphylos morroensis. In consideration of the potential conservation benefits of rediscovering and recovering the Moro Bay kangaroo rat, we conclude that continuing the search effort has substantial merit.

Please note: The Journal of Fish and Wildlife Management is not responsible for the content or functionality of any supplemental material. Queries should be directed to the corresponding authors for the article.

Reference S1.California Fish and Game Commission. 1970. 1970 Supplement to Fish and Game Code. Sacramento: California Fish and Game Commission.

Found at DOI: https://doi.org/10.3996/JFWM-20-022.S1 (691 KB PDF).

Reference S2. Roest AI. 1984. The Morro Bay kangaroo rat: a summary of current knowledge. Essentials of a poster presented at the 64th Annual Meeting of the American Society of Mammalogists, Humboldt State University, Arcata, California.

Found at DOI: https://doi.org/10.3996/JFWM-20-022.S2 (348 KB PDF).

Reference S3. Souza LJ. 1958. Distribution of the Edna kangaroo rat population. Senior thesis. San Luis Obispo: California Polytechnic State University.

Found at DOI: https://doi.org/10.3996/JFWM-20-022.S3 (23.9 MB PDF).

Reference S4.[USFWS] U.S. Fish and Wildlife Service. 1999. Draft revised recovery plan for the Morro Bay kangaroo rat (Dipodomys heermanni morroensis). Portland, Oregon: U.S. Fish and Wildlife Service.

Found at DOI: https://doi.org/10.3996/JFWM-20-022.S4 (13.69 MB PDF).

Reference S5. Villablanca FX. 2007. Morphological and genetic divergence of Morro Bay kangaroo rats: agreement #P0485102 and #S0685103. Report to California Department of Fish and Game, Sacramento.

Found at DOI: https://doi.org/10.3996/JFWM-20-022.S5 (3.7 MB PDF).

We thank the following persons for their assistance: Vince Cicero of California Department of Parks and Recreation; Robert Stafford of California Department of Fish and Wildlife; Brian Boroski, Daniel Duke and Robyn Powers of H.T. Harvey & Associates; and Catherine Darst, Stephen Henry and Roger Root of U.S. Fish and Wildlife Service. Without the contributions of all these persons, this project could not have been accomplished. Partial funding was provided by the U.S. Fish and Wildlife Service to H.T. Harvey & Associates and to California Polytechnic State University. Finally, we thank the reviewers and editors for their time and efforts to make this a better paper.

Any use of trade, product or firm names in this publication is for descriptive purposes only and does not imply endorsement by the U.S. Government.