Application of Potential Biological Removal methods to seabird populations

The Potential Biological Removal (PBR) approach was developed in response to the United States Marine Mammal Protection Act, to identify populations experiencing human-caused mortality at levels that could result in population depletion. The PBR is calculated from the maximum population growth rate (r_max) and a lower estimate of the population size (N_min), as PBR = ½ r_maxN_minf , where f (typically between 0.1 and 0.5) is a “recovery factor” that may be set lower to allow a population to recover faster, or to provide additional protection to the population. If the human-caused mortalities are less than the PBR, then a depleted population will be able to recover so that, given sufficient time, it has a 95% probability of being over half the carrying capacity.

When assessing the potential impact of human-caused mortalities on seabird populations, the PBR has been used as a guide to the productivity of the seabird populations. Applying the PBR to seabirds is difficult as neither the maximum growth rate nor the total population size can be directly measured. Instead, approximations must be used that allow estimation of these parameters from readily available data.

In this report, we used simulations of seabird demography to assess the accuracy of these approximations. This approach involved three main steps. First, we simulated the population dynamics for 12 types of seabirds, representing a range of species breeding in New Zealand. For each species type, we estimated the maximum human-caused mortality rate that the populations could incur, while still being able to recover to above half the carrying capacity, with 95% probability, in the presence of both environmental and demographic stochasticity. Second, we generated a PBR estimate using an approximate maximum growth rate and population size. The PBR estimate included a parameter ρ, calibrated so that the base PBR (PBR_b; evaluated with f = 1 and with the total population, N, rather than the conservative estimate, N_min) had only a 5%-probability of exceeding the maximum human-caused mortality. Finally, we explored the effect of errors or bias in the demographic parameters used for the calculation of the PBR, to provide guidance in setting the value of the recovery factor, f.

The analysis showed that the approximate base PBR derived from demographic parameter estimates tended to overestimate the maximum human-caused mortality. Inclusion of a calibration factor, ρ, was required to adjust the PBR approximations to meet the management criterion; ρ varied between 0.17 and 0.61, depending on the species types. In general, the calibration factor was smaller for species with slower growth rates, such as albatrosses, and higher for species with higher growth rates, such as shags and penguins. Previous estimates of the PBR for seabird populations that did not include this calibration factor are likely to have overestimated the human-caused mortalities that the populations could incur.

The choice of f values will depend on what errors in the underlying parameters are considered plausible, and on requirements for the recovery time of depleted populations. In this report, some exploration of the consequences of incorrect estimates of the parameters is given, but an explicit recommendation for the choice of f values is not made.

With the inclusion of the additional calibration factor, ρ, the method for calculating the PBR described here provides a simple way for determining whether fishing-related mortalities are sufficiently low that seabird populations are able to recover to and/or remain at above half the carrying capacity in the long term.