Decline and extirpation of an endangered Panamanian stream frog population (Craugastor punctariolus) due to an outbreak of chytridiomycosis

Abstract

We conducted a mark-recapture study of three subpopulations of Craugastor punctariolus at a mid-elevation site in central Panama between 1999 and 2005. The study spans a period over which the pathogenic fungus Batrachochytrium dendrobatidis (Bd) was absent from the site, invaded the site and decimated all amphibian populations, and now persists. We quantified natural demographic parameters prior to and during an event of mass mortality due to chytridiomycosis caused by Bd. Prior to the event of mass mortality, all three subpopulations of C. punctariolus were large (19–68 animals/200 m), showed a stable age-size distribution, and had high survival. All age-size classes of this species co-occurred on boulder clumps along streams, and adults showed high site fidelity and were likely territorial. Following detection of Bd at this site in late September 2004, four dead C. punctariolus were found infected with Bd and all three subpopulations completely disappeared from this site within 2 months. The association of all age-size classes with microhabitats appropriate for survival and growth of Bd likely contributed to the rapid and severe degree of decline of this species at this site. These data provide insight into the patterns and mechanisms of decline within a species due to Bd.

Introduction

Approximately 43% of amphibian species are declining (Stuart et al., 2004) and several species have gone extinct (Wilson and McCranie, 2003, La Marca et al., 2005, Lips et al., 2006). The Neotropics are home to more than half of the world’s 6000 described amphibian species (Frost et al., 2006), with high species endemism at middle and upper elevations (Duellman, 1999). It is these middle and upper elevation amphibian communities that have declined most severely (e.g. Lips et al., 2004, Stuart et al., 2004). Declines in upland areas of the Neotropics and Australia have been especially severe and geographically and taxonomically broad, with approximately half an amphibian assemblage being extirpated within a year, and an 80% reduction in the abundance of the remaining species (Pounds et al., 1997, Williams and Hero, 1998, Lips et al., 2003, Stuart et al., 2004, Bustamante et al., 2005, La Marca et al., 2005).

Two Neotropical clades that exemplify the montane, riparian amphibian decline scenario are the genus Atelopus (La Marca et al., 2005) and the Craugastor rugulosus species group (Campbell and Savage, 2000, McCranie and Wilson, 2002, Lips, 1999, Lips et al., 2004, Lips et al., 2006). Atelopus declines have been well documented, in part because they are colorful, diurnal frogs that are easy to observe (Lötters, 1996). In a comprehensive review of 113 species of Atelopus, La Marca et al. (2005) found that all species that occurred above 1000 m had declined and 75% of those species had completely disappeared. Less well known has been the decline and disappearance of the C. rugulosus species group, a clade of drab, nocturnal stream frogs found predominantly in upland areas from Mexico to Panama. Twenty-six (79%) of the 33 species in this clade are critically endangered, endangered, or vulnerable (IUCN et al., 2008), and the best available data suggest that this entire group has experienced severe population declines throughout Central America (i.e. Campbell, 1998, McCranie and Wilson, 2002, Savage, 2002, Lips et al., 2003, Lips et al., 2004). Only two populations are currently known to occur at historic abundances, Craugastor ranoides in the low elevation dry forest habitat on the Santa Elena Peninsula of Costa Rica (Puschendorf et al., 2005), and low to mid-elevation populations of C. punctariolus between El Valle de Anton and the Panama Canal watershed (R. Ibáñez, pers. comm.).

In many cases, amphibian die-offs and declines have been attributed to chytridiomycosis caused by the fungus Batrachochytrium dendrobatidis (Bd hereafter). These declines are non-random with respect to timing (Laurance et al., 1996, Lips et al., 2006, Kriger and Hero, 2006), physical habitat (Ron, 2005), and ecology of affected species (Williams and Hero, 1998, Lips et al., 2003, Murray and Hose, 2005), with species most closely associated with riparian habitats in cool moist forests at greatest risk. Not all populations respond to the same degree or at the same rate following the introduction of Bd to a population (e.g. Brem, 2006, Lips et al., 2006, Rowley and Alford, 2007b). Differences in species responses within a community have been attributed to differences in susceptibility (Daszak et al., 2000), micro-environmental conditions (e.g. Brem, 2006), and population connectivity (e. g. Robertson et al., 2008).

In order to understand the mechanisms involved in population fluctuations, it is necessary to know the abundance of each sex and age class and the rates of change of demographic parameters. Long-term data sets (e.g. Alford and Richards, 1999, Houlahan et al., 2001) based on mark-recapture techniques (Pollock et al., 1990, Alford and Richards, 1999, Schmidt, 2003, Muths et al., 2003, Grafe et al., 2004) can provide estimates of population abundance, survival, dispersion, sex ratio, and age structure necessary to detect population trends. A long-term data set of multiple demographic units (e.g. male, female, and juvenile subunits) can indicate whether groups respond uniformly to environmental perturbations (Stearns, 1976, Crouse et al., 1987, Collins and Storfer, 2003, Grafe et al., 2004). For example, in Costa Rica and Panama many amphibian populations have declined to less than 20% of typical abundances and have shown no subsequent recovery in the past 10 to 15 years (e.g. Lips et al., 2003, Lips et al., 2006). This pattern suggests that many adults are affected, that no reproduction occurs, or that all stages are affected rapidly and simultaneously. Capture–recapture studies have been conducted on various amphibians (e.g. Richter and Seigel, 2002, Muths et al., 2003, Schmidt, 2003, Scherer et al., 2005), although rarely for tropical anurans (but see Funk et al., 2003, Grafe et al., 2004).

Here we describe the natural population dynamics and habitat use of three subpopulations of Craugastor punctariolus over a 6 year period and document changes in demographic parameters following the arrival of Bd at the site in 2004. We estimated the amount of time from the first detectable decline in survival to the time when the population completely disappeared from our study area. We also include data related to the basic natural history of this poorly known endangered species because it has declined precipitously throughout its range and is likely to go extinct in the near future.