Open Access
Report
2011
Special Committee on Seals
Lonergan, M., B. McConnell, C. Duck, and D. Thompson,
We fitted a Bayesian state-space model of British grey seal population dynamics to two sources of data: (1) regional estimates of pup production from 1984 to 2012, and (2) an independent estimate assumed to be of total population size just before the 2008 breeding season. The model allowed for density dependence in pup survival, using a flexible form for the density dependence function, and assumed no movement of recruiting females between regions. This model is identical to that used to provide last year’s advice, and the same “revised” priors were used, including a prior on adult sex ratio. One small change in data was that the total population size estimate was adjusted to account for the fact that the population model is based only on regularly monitored breeding colonies (approx. 94% of the total population). We used the model to predict past the last data point (2012) to give estimates of population size in 2014. Estimated adult population size in 2014 was 95,200 (95% CI 76,400-127,500). The model assumes constant adult (i.e., aged 6+) female survival. The prior distribution has support in the range (0.8, 1.0) with a prior mean of 0.91 (SD 0.05); the posterior mean is an implausibly high 0.99 (SD 0.01). We investigated the effect of constraining the prior to the range (0.8, 0.97). Posterior mean adult survival with this revised prior was 0.95 (SD 0.03); estimated population size with this revised prior was 105,200 (95% CI 87,000-128,800). Female survival is currently assumed to be the same for all ages. We investigated the possible effect of including survival senescence, and concluded that adding it would make no practical difference to the modelled population dynamics. Sex ratio is an important parameter in the model, scaling estimates of adult female population size from the population dynamics model to total population size. The current prior is highly informative (prior mean on ratio of total population:adult females 1.7 SD 0.02). We investigated the consequences of using a less informative prior suggested in a previous briefing paper (prior mean 1.2 SD 0.63). With this prior (and the revised prior on adult female survival), total population size was estimated to be much lower (88,600 with 95%CI 70,200-111,700), but the ratio of total population:adult females was an implausibly low 1.14 (SD 0.09).
Link To Publication
Some links to publications are behind pay-walls and hence might not be readily accessible to the public