Modeling
the Population Dynamics of a Declining Amphibian Population
Using Mark-Recapture Methods
- Barry Noon, Dept. of Fisheries &
Wildlife Biology,
Colorado State Univ. (USA) under US-India ‘Fulbright Fellowship’ Programme
ABSTRACT
Available evidence suggests that declines in amphibian populations have been occurring for decades. However, concern has grown in recent years because declines appear to be even more prevalent than previously recorded. Reductions in geographic distribution, declines in population size, and extinctions have been documented in species of amphibians from around the world. In Colorado (USA), the boreal toad (Bufo boreas) was once widely distributed and locally abundant in high elevation lakes and wetlands. However, surveys conducted from 1986 to 1992 reported range-wide population declines and extinction of several local populations of this species. During these surveys, boreal toads were observed at only 10 of 59 historic localities in the Front and Park Ranges of Colorado. Based on these reports and others, the boreal toad was listed as an endangered species by the State of Colorado in 1994.
Many causal hypotheses have been put forward to explain amphibian population declines, including the introduction of exotic diseases. Recently, Batrachochytrium dendrobatidis, commonly referred to as chytrid fungus, was identified on live and dead boreal toads collected as far back as 1998. This fungus has been implicated in mass mortality events associated with amphibian population declines and extinctions from around the world. Its role as an agent of disease in amphibians has been confirmed by the fulfillment of Koch’s postulates, a set of criteria used to determine the pathogen responsible for a particular disease.
Our research, conducted in Rocky Mountain National Park (RMNP), Colorado from 1992-2002, evaluated possible causes of decline in the populations of boreal toads. Specifically, we evaluated the relationships between climatic factors and introduced disease (chytrid fungus) and population survival rates using mark-recapture methods. Prior to analysis we developed a candidate set of plausible models relating disease and climate to survival rates in the form of a general linear model. These models were subsequently explored for the ‘best’ explanatory model on the basis of AIC model selection procedures implemented in the computer program MARK. Based on a hierarchical process of model selection our best model found a strong relationship between annual survival rate and prevalence of the chytrid fungus pathogen.
- Dr. Barry Noon