Unmarked individuals in mark-recapture studies : comparisons of marked and unmarked southern elephant seals at Marion Island

Abstract

The presence of unmarked individuals is common in mark–recapture study populations; however, their origin and significance in terms of population dynamics remain poorly understood. At Marion Island, southern Indian Ocean, where virtually all southern elephant seal Mirounga leonina pups born annually (1983–2008) were marked in a long-term mark–resight study, large numbers of unmarked seals occur. Unmarked seals originate either from marker (tag) loss or from immigration. We aimed to identify patterns in the occurrence of marked and unmarked individuals that will allude to the possible origin and significance of the untagged component of the population, predicting that tag loss will add untagged seals to mainly adult age categories whereas migrating untagged individuals will be mostly juveniles.We fitted a generalized linear model using the factors month, year and age-class to explain the relative abundance of untagged seals (tag ratio) from 1997 to 2009. Site usage of untagged seals relative to tagged seals was assessed using a binomial test. Untagged seals, predominantly juveniles, were present in the highest proportions relative to tagged seals during the winter haulout (tagged seals/total seals less than 0.3) and the lowest proportion (approximately 0.5) during the female breeding haulout, increasing in relative abundance from 1997 to 2009. Untagged seals were distributed evenly across suitable haulout sites while tagged seals displayed high local site fidelity and occurred in greater numbers at or near large breeding beaches. Untagged seals are considered to be mostly migrant seals that disperse from other islands within the southern Indian Ocean and haul out at Marion Island during non-breeding haulouts in particular. Some of these seals immigrate to the breeding population, which can be a key component of the local population dynamics.We emphasize the need for mark–recapture studies to evaluate the role of the unmarked component of a population, thereby inducing a more confident estimation of demographic parameters from the marked sample.