fishmax: An R package for Bayesian estimation of maximum body size in animals

Maximum body size is a key trait of an animal species, determining its population dynamics, productivity and interactions with other species and the environment. For fish in particular, the maximum body length or correlates strongly to other life history parameters and is commonly used in ecological models and population assessments.

of a species is typically defined as a single value, representing the single largest observed individual or an average of several of the largest individuals. This approach, however, does not explicitly account for sampling effort, where greater sampling intensity generally leads to a higher probability of finding a larger individual, particularly for indeterminate growers.

Here we present an R package for Bayesian estimation of maximum animal body size and its uncertainty using the largest observed individuals from multiple samples. Although developed for maximum body length, the method could be applied to estimate other extremes (e.g. greatest body weight, tallest height) from other taxa that grow indeterminately.

The package uses three alternative methods—extreme value theory and two versions of exact finite sample approach. The latter two are suitable in cases where the underlying population body‐size distribution can be approximated by a truncated normal distribution.

To demonstrate the application of the method we used a hypothetical fishing competition data, but any other records of maximum observed individuals across multiple comparable samples (scientific surveys, trophy hunting, underwater surveys, historical accounts) are also suitable.