From Hope to Alert: Demography of a Remnant Population of the Critically Endangered Atelopus Varius from Costa Rica

Home , Atelopus cruciger, Atelopus laetissimus, Atelopus spumarius, Atelopus varius

Studies on Neotropical Fauna and Environment

ISSN: 0165-0521 (Print) 1744-5140 (Online) Journal homepage: http://www.tandfonline.com/loi/nnfe20

From hope to alert: demography of a remnant population of the Critically Endangered Atelopus varius from Costa Rica

José F. González-Maya, Diego A. Gómez-Hoyos, Ivan Cruz-Lizano & Jan Schipper

To cite this article: José F. González-Maya, Diego A. Gómez-Hoyos, Ivan Cruz-Lizano & Jan Schipper (2018): From hope to alert: demography of a remnant population of the Critically Endangered Atelopus varius from Costa Rica, Studies on Neotropical Fauna and Environment, DOI: 10.1080/01650521.2018.1460931 To link to this article: https://doi.org/10.1080/01650521.2018.1460931

Published online: 24 Apr 2018.

Submit your article to this journal

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=nnfe20 STUDIES ON NEOTROPICAL FAUNA AND ENVIRONMENT, 2018 https://doi.org/10.1080/01650521.2018.1460931

From hope to alert: demography of a remnant population of the Critically Endangered Atelopus varius from Costa Rica José F. González-Maya a,b, Diego A. Gómez-Hoyos a, Ivan Cruz-Lizanoa and Jan Schipper a,c aProyecto de Conservación de Aguas y Tierras, ProCAT Colombia/Internacional-Fundación Sierra to Sea Costa Rica, Puntarenas, Costa Rica; bInstituto de Ecología, Universidad Nacional Autónoma de México, Ciudad Universitaria, México DF, México; cArizona Center for Nature Conservation–Phoenix Zoo, Phoenix, AZ, USA

ABSTRACT ARTICLE HISTORY Harlequin frogs have suffered severe declines across the Neotropics. We present a population Received 16 May 2017 assessment for a recently discovered population of Atelopus varius from Costa Rica. Using mark- Accepted 29 March 2018 recapture methods from September 2011 to February 2013, we estimated survival and recruit- KEYWORDS ments parameters using Cormack-Jolly-Seber models. We obtained 222 captures and estimated Amphibian decline; low recruitment rates and high seniority. Given estimates of population growth rates close to capture-recapture; chytrid; zero, the observed population seems to be stable during the study. However, contrary to recruitment; seniority; trout expectations for seasonally reproducing species like A. varius, we did not find an increase in recruitment rates between dry and rainy season. We provide details on ongoing threats for the population, as well as propose conservation actions to mitigate these threats.

RESUMEN Las Ranas Arlequín han sufrido declives severos en el Neotrópico. Presentamos una evaluación poblacional para una población recientemente descubierta de Atelopus varius de Costa Rica. Usando métodos de marca-recaptura desde septiembre de 2011 hasta febrero de 2013, estimamos parámetros de supervivencia y reclutamiento usando modelos Cormack-Jolly-Seber. Obtuvimos 222 capturas y estimamos bajas tasas de reclutamiento y alta ‘antigüedad’ (seniority). Dadas que las tasas de crecimiento poblacional se acercaron a cero, la población observada parece estar estable durante el estudio. Sin embargo, contrario a lo esperado para especies con reproducción estacional como A. varius, no encontramos un incremento en las tasas de reclutamiento entre las estaciones lluviosa y seca. Proveemos detalles de amenazas actuales para la población, así como propuestas de acciones de conservación para mitigar estas amenazas.

Introduction (Carvajalino-Fernández et al. 2008; Flechas et al. 2012; Rueda-Solano 2012; González-Maya et al. Frogs of the Atelopus genus, known as harlequin frogs, 2013a; Gómez-Hoyos et al. 2014, 2017). These remnant are distributed from Costa Rica down to Bolivia and the populations offer a unique opportunity for the study Guianas in the south and east of their range, respectively and development of conservation actions, providing (Lötters 1996, 2007; La Marca et al. 2005). Except for hope for other Atelopus species (González-Maya et al. some populations in Colombia, the Amazon and the 2013a; Perez et al. 2014; Gómez-Hoyos et al. 2017). Guianas, harlequin frogs have suffered from rapid and Ofthefourharlequinfrogspeciesrecordedin severe population declines across their entire distribu- Costa Rica, only Atelopus varius has been rediscov- tion, leading to concerns regarding the potential extinc- ered from two localities during 2008 and 2015 tion of the entire genus (Lötters 2007). According to the (González-Maya et al. 2013a; Barrio-Amorós & IUCN Red List of Threatened Species (IUCN 2012), 74 Abarca 2016). For these surviving populations, man- of the 93 species are categorized as Critically agement and conservation actions seems warranted, Endangered, from which a large proportion is consid- since several factors considered to pose potential ered potentially extinct (Lötters 2007). threats for amphibians – such as pathogens, habitat Despite these dark prospects, some species that were degradation and invasive species – are operating in thought to be extinct, or were classified as data defi- these areas (JFG-M & DAG-H, pers. obs.). cient, have been ‘rediscovered’ and studied Furthermore, post population decline surveys are

Published online 24 Apr 2018 2 J. F. GONZÁLEZ-MAYA ET AL. critical for amphibian conservation in general (Perez With the aim of avoiding transmission of Bd between et al. 2014), because they help to identify threat localities, we used exclusive wear and boots for our factors driving population declines, and help refine study area. the management actions required. Here we estimated population parameters for an A. varius population in Data analysis Talamanca, Costa Rica and identified the main threats for population persistence. Based on our We used the Cormack-Jolly-Seber (CJS) model based results we specify urgent research needs and propose on live animal recaptures in an open population management measures for A. varius. (Lebreton et al. 1992) in order to estimate apparent survival. Apparent survival is the probability that the animal remains alive and available for recapture and, Materials and methods technically, not the survival probability of marked ani- Study area mals in the population (White & Burnham 1999). To confirm that the CJS model adequately fitted the col- Our study site is located in the Las Tablas Protected lected data, we used bootstrap goodness-of-fit test Zone (LTPZ), Puntarenas Costa Rica. LTPZ is located (GOF). Analyses were conducted in program MARK in Southeastern Costa Rica, specifically in the 8.0 (White & Burnham 1999). Models were constructed Talamanca mountain range, one of the regions with based on apparent survival (Phi) and recapture rates higher levels of endemism of the country (González- (p) with constant (.), temporal (t: each monthly capture Maya et al. 2013b). LTPZ spans over 19,000 ha, and it session) and seasonal variation (season: dry or rainy is part of La Amistad Biosphere Reserve (González- month during the surveys) of such parameters, and for Maya & Mata-Lorenzen 2008). Our study site is located two age groups (immature individuals, including juve- in the Cotón river, at 1300 m asl, with a mean annual niles and sub-adults: ˂ 27 mm and mature individuals: precipitation of 3500 mm and mean annual tempera- ˃ 27.1 mm). We used two seasons (dry: December to ture of 19°C, corresponding to very humid premontane March; rainy: April to November) as a temporal varia- tropical forest (González-Maya & Mata-Lorenzen tion within models, because Costa Rica has a marked 2008). In this site, the landscape is dominated by bi-seasonal climatic pattern and A. varius is a season- mature and secondary forest, with some scarce patches ally reproductive species. A multi-state model was not of crops and pastures. considered because it assumes that all individuals make the transitions at the same time, which is difficult to Survey test for small sample sizes such as in our study. Seniority (Gamma) was estimated through Pradel Our surveys were conducted during the day on a ‘seniority only’ model. Since this model computes 3 km transect along both sides of the Cotón river. encounter histories identical to the CJS model, we use The transect was surveyed monthly between the ĉ achieved by the GOF test of the CJS model. September 2011 to February 2013, by a herpetologist Seniority probability (SP) is defined as the probability and a trained field assistant. The surveyors walked of an individual to be alive and in the population at along the river margin covering at least 2 m from the times i and i – 1 (Pradel 1996). Thereby, SP is useful river bank, searching post-metamorph individuals as for identifying the proportion of the population that well as eggs and tadpoles. Detected post-metamorph was previously in the population before and during individuals were captured, measured (i.e. weight and reproductive season. In order to estimate the para- snout-to-vent length) and their dorsal and ventral meters involved in population growth, we estimated surfaces were photographed. Samples from the skin recruitment and survival, and their interaction, as a of each individual were obtained using cotton swabs baseline characteristic of the population (Nichols for the study of Batrachochytrium dendrobatidis (Bd) et al. 2000). Also, the estimation of these parameters presence (Kriger et al. 2006), that were stored and is useful to plan conservation actions for threatened refrigerated dry on sterile vials for posterior testing. species (Schmidt et al. 2005; Muths et al. 2011). The Data on location (i.e. geographic coordinates), time Pradel survival and recruitment (PSR) model was used and substrate were recorded for each individual, and to estimate recruitment. Recruitment accounts for the a unique ID was assigned to each capture. Given that number of individuals added to a breeding population each individual of A. varius exhibits a unique spot- (Muths et al. 2011). In most bufonid species, recruit- ting pattern, capture histories were built based on ment is not equal to reproduction given that indivi- dorsal/ventral pictures of each captured individual. duals spend multiple years as juveniles or immatures STUDIES ON NEOTROPICAL FAUNA AND ENVIRONMENT 3 before joining the breeding population (Schmidt et al. usually above 40 mm, our sample include at least 22 2005; Muths et al. 2011); according to our field obser- mature females. We did not detect any A. varius eggs vations, in a seasonal breeding species such as A. var- or tadpoles during our surveys, or individuals under ius, new individuals entering the breeding population 18 mm SVL (Figure 1). should come from cohorts near to complete two years According to bootstrap GOF the CJS model ade- of development. Survival and recruitment parameters quately fits the data (ĉ = ~ 1). Due to uncertainty in were used to calculate Atelopus varius population selecting the best-fitting models, the model averaging growth (Lambda), or population size change rate, method was applied to the four top-ranked models which has values >1 indicating growth, and values <1 (Table 1); although model 5 was also competent indicating decline (Muths et al. 2011). (delta AICc = 1.912), the estimate was not reliable. The best-fitting models were selected based on the Considering model averaging, estimated apparent sur- Akaike information criteria with a correction for small vival was (estimate ± standard error [95% confidence sample sizes (AICc) and derived measures (Burnham & interval]) 0.87 ± 0.09 [059–0.97] in the rainy season Anderson 1998; Cooch & White 2015). When we and 0.78 ± 0.16 [0.37–0.95] in the dry season. The found uncertainty about the best fitting model, we capture probability varied over time from used a model-averaging method with the top-ranked 0.011 ± 0.01 [0.002–0.069] to 0.1 ± 0.098 [0.013– models (delta AICc ˂ 2), including only those models 0.48]. Regarding the estimation of SP, the two best- with reliable estimations. All statistical results are pre- fitting models explaining the data both included a sented with the respective standard error (±SE) and capture probability depending on season, and gamma 95% confidence intervals (LCI–UCI). depending on season or constant, respectively (Table 1). Although the selection of models including Threat assessment both a seasonal SP as well as a constant SP, and there- fore the difference in SP between seasons, was not Swab samples were analyzed to confirm Bd presence in substantially important, lower parameters were esti- the population. Of 222 samples, we selected 15 random mated from model averaging for the dry season sub-samples and these were tested for Bd using stan- (0.5 ± 0.17 [0.20–0.79]) compared to the rainy season dardized methods (Hyatt et al. 2007). We extracted (0.88 ± 0.07 [0.66–0.97]). Capture probability was dif- DNA from the swabs with PrepMan Ultra (Applied ferent between seasons; however, estimations were Biosystems, Carlsbad, CA, USA), and analyzed the imprecise (rainy: 0.01 ± 0.007 [0.005–0.04], dry: samples using the standard real-time quantitative poly- 0.095 ± 0.05 [0.03–0.26]). Meanwhile, the best-fitting merase chain reaction assay. We followed the standar- model including the estimation of recruitment dized procedures in Boyle et al. (2004) with the following exception: the nucleic acids were extracted μ using 50 l PrepMan, a negative control (H2O) and a positive control swab dipped in a broth of Bd culture from Costa Rican strain JGA01. The presence of other potential threats such as habitat perturbation and invasive species (e.g. rainbow trout) were registered (though not quantified) during surveys in order to provide a first insight on possible influence of such pressures, since these threats have been reported as potential drivers of decline in other populations (Pounds et al. 2010).

Results We detected 222 individuals of A. varius, which included 103 immatures (23 juveniles and 80 subadults), as well as 113 male and female adults (six undetermined; Figure 1). Uncertainty on the identifica- Figure 1. Distribution of age proportions of Atelopus varius individuals from Las Tablas Protected Zone, Costa Rica. (a) tion between males and sub-adult females did not allow Metamorphs: 7 mm; (b) juveniles: ˂21.5 mm; (c) subadults: discrimination by sex categories. Since it is expected 21.5–27 mm; and (d) adults: 27–43 mm (adult females: that mature females present snout-to-vent length (SVL) 33–43 mm). 4 J. F. GONZÁLEZ-MAYA ET AL.

Table 1. Top-ranked models for Atelopus varius apparent survival, recruitment and seniority probability from Las Tablas Protected Zone, Costa Rica. Model averaging was performed to models with delta AICc ˂ 2. Model AICc Delta AICc AICc weights Model likelihood k Deviance Apparent survival Phi(.)p(season) 178.470 0.000 0.234 1.000 3 84.275 Phi(.) p(t) 179.535 1.065 0.137 0.587 16 56.652 Phi(season)p(t) 179.828 1.358 0.118 0.507 17 54.578 Phi(season)p(season) 179.961 1.491 0.111 0.474 4 83.688 Phi(age)p(season) 180.382 1.912 0.089 0.384 4 84.109 Phi(.)p(.) 181.660 3.190 0.047 0.203 2 89.524 Phi(age)p(t) 181.867 3.397 0.043 0.183 17 56.617 Phi(.)p(season*age) 182.206 3.736 0.036 0.154 5 83.834 Phi(.) p(g) 183.093 4.623 0.023 0.099 3 88.898 Pradel seniority Gamma(season)p(season) 179.635 0.000 0.522 1 4 −935.597 Gamma(.)p(season) 180.215 0.579 0.390 0.748 3 −932.943 Gamma(.)p(.) 184.400 4.765 0.048 0.092 2 −926.702 Gamma(season)p(.) 186.346 6.711 0.018 0.035 3 −926.811 Gamma(.)p(t) 186.815 7.180 0.014 0.028 16 −954.898 Gamma(season)p(t) 188.292 8.657 0.007 0.013 17 −955.770 Gamma(t)p(season) 198.058 18.422 0.000 0.0001 17 −946.004 Gamma(t)p(.) 200.449 20.814 0.000 0.000 16 −941.264 Gamma(t)p(t) 203.244 23.609 0.000 0.000 27 −965.637 Pradel recruitment Phi(.)p(t)f(.) 1209.240 0.000 0.502 1 18 62.807 Phi(.) p(t)f(season) 1211.265 2.025 0.183 0.363 19 62.437 Phi(season)p(t)f(.) 1211.442 2.201 0.167 0.333 19 62.614 Phi(.)p(t)f(t) 1211.689 2.449 0.148 0.294 22 55.532 Phi(.)p(season)f(t) 1228.038 18.797 0.000 0.000 12 95.491 Phi(t)p(t)f(t) 1230.875 21.635 0.000 0.000 33 45.828 Phi(.)p(.)f(t) 1235.956 26.716 0.000 0.000 12 103.409 Phi(season)p(.)f(t) 1238.051 28.811 0.000 0.000 13 103.246 Phi(t) p(.)f(t) 1238.551 29.311 0.000 0.000 26 72.267

comprised constant survival, constant recruitment, as phenomena (Nichols et al. 2000). This is expected for well as a capture probability varying over time long life expectancy species belonging to the Atelopus (Table 1). Survival estimates for this model genus (Lötters 1996; Muths et al. 2011; Lampo et al. (0.86 ± 0.07 [0.66–0.96]) were similar to the estimates 2012). Atelopus varius is considered a seasonal breed- with the CJS model, and recruitment was estimated as ing species (Lötters 1996; Savage 2002); however, 0.23 ± 0.07 [0.12–0.39]. Based on the estimates of population parameters such as seniority and recruit- recruitment and survival, population size rate of ment were either not or weakly explained by seasonal change was 1.1 (0.78–1.35). reproduction, contrary to expectations, possibly due Tests for chytrid fungus were positive for two of the to the low capture rate. Although the population was 15 individuals tested. Additionally, during our surveys stable during the survey, we argue that it can still be rainbow trout (i.e. invasive species) were detected considered vulnerable to local extinction given the low repeatedly. Furthermore, up to a distance of at least recruitment rates and high seniority, and considering 500 m away from the survey area, the vegetation was as baseline the elevated pre-decline population esti- under clearing maintenance, indicating ongoing habitat mates; the species was considered locally abundant in transformation within the study area. the same area at least 20 years before our surveys (Pounds & Crump 1994; La Marca et al. 2005; González-Maya et al. 2013a). Discussion Our apparent survival estimates were similar to Population parameter estimates showed that the A. those estimated for the species in El Copé and Santa varius population from Las Tablas Protected Zone Marta in Panama prior to the arrival of Bd (McCaffery was stable, with survival contributing more than et al. 2015). They were also similar to the survival recruitment as empirically observed on both rates for parameter estimated for a population of A. spumarius our population; the estimation of these demographic from Ecuador that tested negative for Bd (Tarvin et al. components allows questions to be answered regard- 2014). The similar survival estimates among these ing population growth rates and the process that populations, all under Bd pre-outbreak scenarios, indi- represents the most important determinants for such cate adult individuals are potentially not severely STUDIES ON NEOTROPICAL FAUNA AND ENVIRONMENT 5 affected by the infection, as shown in our population at field surveys we frequently detected the presence of least during our survey. Our capture probabilities were the species both in the main river and isolated ponds similar to the lower limit estimates in these localities and tributaries. We hypothesized that rainbow trout from Panama (McCaffery et al. 2015), but were sub- might cause population decline by predating on eggs, stantially lower than those estimated for A. spurrelli and probably to a minor extent on tadpoles and meta- and A. elegans in Colombia (Gómez-Hoyos et al. morphs. Atelopus varius lay egg strings on streams and 2014, 2017), which likely contributed to the consider- rivers (Lötters 1996) and are probably very susceptible ably large confidence intervals of parameter estimates to predation by trout (La Marca et al. 2005; Martín- in our study. Torrijos et al. 2016) and other aquatic predators. No data are available concerning survivorship of We do acknowledge that our results did not test froglets, tadpoles or eggs of the species in our study causality or positive/negative relationships between or elsewhere (Lötters 1996), and we did not detect any potential threats and population parameters. tadpoles or eggs during our surveys. We found no However, we believe these causes have plausible prob- evidence for seasonal differences in recruitment, and ability to be affecting the survival of the population, only weak influence of season on seniority (Table 1), especially since they were already positively related contrary to what would be expected for species with with Atelopus decline both globally and in Costa Rica marked seasonal reproduction (during the dry season) (Blaustein et al. 1994; Berger et al. 1998; La Marca et al. (Lötters 1996; Savage 2002). However, the absence of 2005). individuals below 18 mm SVL could be explained by In order to avoid the local extinction of this popula- their very low capture probability. tion, we propose three main conservation/management We found that threats to the population are still actions: (1) entrance of other strains of Bd must be operating in the area, as was previously detected by avoided by establishing a biosafety protocol, including Lips et al. (2003). Habitat loss has not ceased since disinfection of potential carriers of zoospores (e.g. tires, the reported disappearance of the population in the equipment, boots, etc.) for accessing the river; (2) area. During the last half of the twentieth century our enclosures around located breeding sites to reduce study area was subject to logging and forest cover predation on eggs or tadpoles; (3) habitat loss and change due to the expansion of coffee production and degradation must be stopped (e.g. maintenance of pastures for cattle. Even though in the first decade of legal protection areas in the river banks) to maintain the twenty-first century logging was stopped in the suitable conditions around the core population. All area, and coffee plantations and cattle raising were proposed measures will need to be assessed through considerably reduced, the area inhabited by A. varius permanent population monitoring in order to test its is still subject to a low degree of cattle grazing. effectiveness on reducing the risk for this population. Chytrid fungus was detected and still potentially Finally, the good and encouraging news of the redis- threats the survival of this population. Although covery of this ‘Lazarus’ population also brought the these results are based on a small sample size, and serious challenges of ensuring its survival in spite of should be regarded only as confirmation of presence the continuing threats. The stable/positive growth rate of the fungus in the population, we consider this of the population estimated in our study is promising, information relevant for our study; not only this is although this result has to be interpreted with great the first record to Bd presence in A. varius popula- caution due to the low capture rates in this study. tion from LTPZ, but also it is likely the most impor- However, these low capture rates are at the same time tant threat based on the severe consequences detected an indication that abundance is considerably lower in populations of other species upon the arrival of Bd than estimated before (Pounds & Crump 1994;La (Ryan et al. 2008;Crawfordetal.2010). Marca et al. 2005), which calls for the incorporation Previous studies have suggested habitat degradation of the suggested measures plus continuing close mon- as one of the main causes of Atelopus decline through- itoring and research of the population in order to out its range (La Marca et al. 2005); for instance, obtain more clues on how to secure these unique survival of A. spumarius decreased after selective tree populations in the long term. harvest within its habitat (Tarvin et al. 2014). Also, rainbow trout has shown constant presence in recent years; the species was introduced to Costa Rica as early Acknowledgments – as 1927 1928, but it was extensively cultured only since The authors are grateful for the permanent support from 1988, mainly on the higher parts of Talamanca and Arizona Center for Nature Conservation/Phoenix Zoo. This other mountain ranges (Vargas 2003). During our research was partially funded by ProCAT Internacional and 6 J. F. GONZÁLEZ-MAYA ET AL.

The Mikelberg Family Foundation. Special thanks to A. in the lowland forests of Gorgona Island (Colombia, South Fischer, F. Castañeda, R. González and all personnel at Las America). Ecohealth. 9:298–302. Alturas. Thanks to J. Abarca for supporting Bd analyses. Gómez-Hoyos DA, Bolívar GW, Burbano-Yandi CE, García JL. 2014. Evaluación poblacional y estrategia de monitoreo para Atelopus spurrelli en el Parque Nacional Natural Utría, Colombia. Revista Biodiversidad Neotropical. Disclosure statement 4:104–112. No potential conflict of interest was reported by the authors. Gómez-Hoyos DA, Suárez-Joaqui T, Bolívar W, García JL. 2017. Population assessment strategy for Atelopus elegans (Bufonidae) in the Gorgona National Natural Park, Funding Colombia. North-Western J Zool. 13:154–158. González-Maya JF, Belant JL, Wyatt SA, Schipper J, This work was supported by the Mikelberg Family Cardenal-Porras J, Corrales D, Cruz-Lizano I, et al. Foundation; IUCN Amphibian Specialist Group; Las 2013a. Renewing hope: the rediscovery of Atelopus varius Alturas del Bosque Verde; Phoenix Zoo and Mohammed in Costa Rica. Amphibia-Reptilia. 34:573–578. Bin Zayed Species Conservation Fund . González-Maya JF, Mata-Lorenzen J. 2008. Dung-beetles (Coleoptera: Scarabeidae) from the Zona Protectora Las Tablas, Costa Rica. Checklist. 4:458–463. ORCID González-Maya JF, Schipper J, Hoepker A. 2013b. Talamanca montane forests. In: Howarth R, editor. Biomes and eco- José F. González-Maya http://orcid.org/0000-0002-8942- systems: an encyclopedia. Ipswich (MA): Salem Press. p. 5157 1203–1205. Diego A. Gómez-Hoyos http://orcid.org/0000-0002- Hyatt AD, Boyle DG, Olsen V, Boyle DB, Berger L, Obendorf 2894-0403 D, Dalton A, et al. 2007. Diagnostic assays and sampling Jan Schipper http://orcid.org/0000-0002-8338-7874 protocols for the detection of Batrachochytrium dendrobatidis. Dis Aquat Organ. 73:175–192. IUCN. 2012. IUCN red list of threatened species version References 2012.2. Gland (Switzerland): International Union for Conservation of Nature. Barrio-Amorós CL, Abarca J. 2016. Another surviving popula- Kriger KM, Hines HB, Hyatt AD, Boyle DG, Hero JM. 2006. tion of the Critically Endangered Atelopus varius (Anura: Techniques for detecting chytridiomycosis in wild frogs: Bufonidae) in Costa Rica. Mesoam Herpetol. 3:1–7. comparing histology with real-time Taqman PCR. Dis Berger L, Spearea R, Daszak P, Greene DE, Cunningham AA, Aquat Organ. 71:141–148. Goggin CL, Slocombe R, et al. 1998. Chytridiomycosis La Marca E, Lips KR, Lötters S, Puschendorf R, Ibáñez R, causes amphibian mortality associated with population Rueda-Almonacid JV, Schulte R, et al. 2005. Catastrophic declines in the rain forests of Australia and Central population declines and extinctions in neotropical harle- America. Proc Natl Acad Sci. 95:9031–9036. quin frogs (Bufonidae: Atelopus). Biotropica. 37:190–201. Blaustein AR, Wake DB, Sousa WP. 1994. Amphibian Lampo M, Celsa SJ, Rodríguez-Contreras A, Rojas-Runjaic F, declines: judging stability, persistence, and susceptibility García CZ. 2012. High turnover rates in remnant popula- of populations to local and global extinctions. tions of the harlequin frog Atelopus cruciger (Bufonidae): Conservation Biol. 8:60–71. low risk of extinction? Biotropica. 44:420–426. Boyle DG, Boyle DB, Olsen V, Morgan JA, Hyatt AD. 2004. Lebreton J-D, Burnham KP, Clobert J, Anderson DR. 1992. Rapid quantitative detection of chytridiomycosis Modeling survival and testing biological hypotheses using (Batrachochytrium dendrobatidis) in amphibian samples marked animals: a unified approach with case studies. Ecol using real-time Taqman PCR assay. Dis Aquat Organ. Monogr. 62:67–118. 60:141–148. Lips KR, Green E, Papendick R. 2003.Chytridiomycosisinwild Burnham KP, Anderson DR. 1998. Model selection and frogs from southern Costa Rica. J Herpetol. 37:215–218. multimodel inference: a practical information-theoretic Lötters S. 2007. The fate of the harlequin toads – help approach. New York (NY): Springer-Verlag. through a synchronous multi-disciplinary approach and Carvajalino-Fernández JM, Cuadrado-Peña B, Ramírez- the IUCN ‘Amphibian Conservation Action Plan’? Pinilla MP. 2008. Registros adicionales de Atelopus nahu- Mitteilungen aus dem Museum für Naturkunde in Berlin mae y Atelopus laetissimus para la Sierra Nevada de Santa – Zoologische Reihe. 83:69–73. Marta, Colombia. Actualidades Biológicas. 30:97–103. Lötters S. 1996. The neotropical toad genus Atelopus. In: Cooch E, White GC. 2015. Program Mark: a gentle introduc- Vences M, Glaw F, editors. Checklist - biology - distribution. Ithaca (NY): Colorado State University. tion. Koln (Germany): Verlags GbR. Crawford AJ, Lips KR, Bermingham E. 2010. Epidemic dis- Martín-Torrijos L, Sandoval-Sierra JV, Muñoz J, Diéguez- ease decimates amphibian abundance, species diversity, Uribeondo J, Bosch J, Guayasamin JM. 2016. Rainbow and evolutionary history in the highlands of central trout (Oncorhynchus mykiss) threaten Andean amphi- Panama. Proc Natl Acad Sci. 107:13777–13782. bians. Neotropical Biodivers. 2:26–36. Flechas SV, Sarmiento C, Amezquita A. 2012. Bd on the McCaffery R, Richards-Zawacki CL, Lips KR. 2015.The beach: high prevalence of Batrachochytrium dendrobatidis demography of Atelopus decline: Harlequin frog STUDIES ON NEOTROPICAL FAUNA AND ENVIRONMENT 7

survival and abundance in central Panama prior to and Pradel R. 1996. Utilization of capture-mark-recapture for the during a disease outbreak. Global Ecol Conservat. study of recruitment and population growth rate. 4:232–242. Biometrics. 52:703–709. Muths E, Scherer RD, Pilliod DS. 2011. Compensatory Rueda-Solano LA. 2012. Atelopus carrikeri Ruthven, 1916 effects of recruitment and survival when amphibian (Anura: Bufonidae) una especie policromática. populations are perturbed by disease. J Appl Ecol. Herpetotropicos. 8:61–66. 48:873–879. Ryan MJ, Lips KR, Eichholz MW. 2008. Decline and extirpa- Nichols JD, Hines JE, Lebreton J-D, Pradel R. 2000. tion of an endangered Panamanian stream frog population Estimation of contributions to population growth: a (Craugastor punctariolus) due to an outbreak of chytridio- reverse-time capture-recapture approach. Ecol. 81:3362– mycosis. Biol Conserv. 141:1636–1647. 3376. Savage JM. 2002. The amphibians and reptiles of costa rica: a Perez R, Richards-Zawacki CL, Krohn AR, Robak M, herpetofauna between two continents, between two seas. Grigfith EJ, Ross H, Gratwicke B, Ibañez R, Voyles J. Chicago (IL): University of Chicago Press. 2014. Field surveys in Western Panama indicate popula- Schmidt BR, Feldmann R, Schaub M. 2005. Demographic tions of Atelopus varius frogs are persisting in regions processes underlying population growth and decline in where Batrachochytrium dendrobatidis is now enzootic. Salamandra salamandra. Conservation Biol. 19:1149–1156. Amphib Reptile Conserv. 8:30–35. Tarvin RD, Peña P, Ron SR. 2014. Changes in population Pounds J, Puschendorf R, Bolaños F, Chaves G, Crump M, size and survival in Atelopus spumarius (Anura: Solís F, Ibáñez R, Savage J, Jaramillo C, Fuenmayor Q, Bufonidae) are not correlated with chytrid prevalence. J Lips K [Internet]. 2010. Atelopus varius. Gland, Herpetol. 48:291–297. Switzerland: The IUCN Red List of Threatened Species; Vargas R. 2003. Evaluación de la reproducción de trucha [cited 2017 May 2]. Available from: http://www.iucnred arcoiris (Onchorhynchus mykiss) producida en Costa list.org/ Rica. I Parte. Agronomía Mesoamericana. 14:123–127. Pounds JA, Crump ML. 1994. Amphibian declines and cli- White GC, Burnham KP. 1999. Program MARK: survival mate disturbance: the case of the golden toad and the estimation from populations of marked animals. Bird harlequin frog. Conservation Biol. 8:72–85. Study. 46:S120–S139.