Project Description
After 2 decades of relative stability, the population of northern fur seals in the Pribilof Islands has declined at an annual rate of 6% since 2000. Estimates of the size and trend of the population currently are based on two indices: annual counts of breeding bulls and biennial surveys of marked pups on St. George and St. Paul Islands. There is no long term marking program in place to collect demographic life histories of a large number of animals with which to understand possible causal mechanisms for the current population decline. NMML is beginning a comprehensive long-term tagging program to allow the estimation of survival and reproductive rates, with methodological innovations to address the issues of tag loss and low re-sighting rates. This will entail development of new statistical methods to estimate tag loss and incorporate it into survival estimates, and full statistical power analyses to determine the tagging and resighting effort necessary to obtain useful demographic estimates. A pilot tagging project on adult females was initiated in fall 2007 (FY08) to evaluate flipper tags, with assessment of the tags and innovative new resighting methods planned for 2008 and 2009.
Issues & Justification
By the 1970s the population of northern fur seals (Callorhinus ursinus) on the Pribilof Islands had declined to one-third of their historic numbers. They were declared depleted under the Marine Mammal Protection Act in 1988. After nearly two decades of relative stability in major rookeries, censuses of pups and breeding bulls indicated a new decline at a 6% annual rate beginning in 2000. The reason for the current decline is unknown, and possible demographic mechanisms cannot be assessed without detailed life-history information that is typically based on longitudinal studies of tagged animals.
Pup production estimates are generated by shearing hair from the heads of a sample of pups and later counting the proportion of marked pups, using mark-recapture analysis to estimate the pup population. The marking of pups by shearing is a time-intensive method and the mark soon disappears as the hair grows back. Permanent marks, such as brands, have potential harmful effects on the thermoregulatory function of the pelage of fur seals, and tags applied to the flippers have been difficult to re-sight and are prone to loss at rates that invalidate survival estimates. A workshop held in September 2005 (report available) at the National Marine Mammal Laboratory ( NMML) provided information from experts in the fields of both pinniped and terrestrial mark-recapture science to evaluate the current analytical methods of estimating population numbers and vital rates based on the re-sights of marked animals, and technological improvements in marking methods.
This tagging program is designed to address previously intractable problems with marking and re-sighting northern fur seals. This will involve the development of new statistical methods to estimate rates of tag loss and incorporate these into estimates of survival. The feasibility of producing useful demographic estimates within budgetary constraints will also be determined with statistical power analysis in light of new statistical methods. We must also evaluate the efficacy of currently available tags in terms of both their retention and their readability under possible re-sighting/recovery programs. e intend to revisit this method and discuss other methods of marking pups, including permanent or near-permanent marks, that would allow longitudinal sampling throughout the life time of the fur seal. This multi-pronged approach has enormous potential for improving demographic studies of fur seals, and pinnipeds generally, providing life history data such as survival, reproductive success, age of first reproduction, and fidelity etc.
Goals
- Development of new statistical methods for estimating and compensating for tag loss in survival estimation
- Statistical assessment of sample size requirements for useful demographic precision
- Pilot assessment of two types of flipper tags
- Development of long-term study plans to estimate age-specific survival and reproduction of northern fur seals
Methods
Analytical
Demographic models have been constructed to determine the possible range of changes that must occur in survival and reproductive rates to account for the observed rate of population decline in comparison to rates that were estimated in the 1950s for a stable population. Preliminary statistical power analyses have been conducted that indicate sample sizes for both pups and adults are obtainable under scenarios not involving tag loss. The tag loss problem is being studied along two lines: a new analytical method that allows the abandonment of the untenable assumption that loss of each tag is independent of the other, and a Bayesian framework for mark-recapture estimation of survival that allows the incorporation of tag loss, including prior information from independent studies. The possibility of incorporating prior information about tag loss into survival estimates is important because it allows independent studies of tag loss to be seamlessly added to our study design as new technical methods for permanent marking of fur seals become available. We are anticipating the ability to mark a sample of fur seals with double flipper tags and some other near-permanent, independent device, such as implanted Radio Frequency ID (RFID) or VHF radio implant, to estimate the proportion of seals that lose both flipper tags. This is the current weakness in virtually all demographic studies based on double- tagging. Such permanent marks are not generally feasible as the sole basis for demographic estimates because the application, and sometimes the recovery of those marks in sufficiently large samples are prohibitively expensive.
Pilot Field Study
Fur seals will be marked with two types of numbered cattle ear tags: colored plastic or engraved metal. In fall of 2007 and 2008, adult females will be marked with both types of tags and the relative retention rates of the two tags will be compared in the summer of 2008 and 2009. Field methods for the resighting of these tags will be assessed, testing high resolution digital cameras and image-stabilizing optics, as well as rookery locations that provide good resighting visibility and adequate sample sizes.