| Summary: | International audience The Amsterdam albatross (Diomedea amsterdamensis) is one of the rarest bird species of world avifauna, consisting of a single population in the upland plateau of Amsterdam Island (SE Indian Ocean). All breeding birds of the population are today banded and a monitoring program involving mark-recapture procedures has been carried out continuously over the past 16 years. We present the first estimate of risk of decline for the Amsterdam albatross using a stochastic matrix population model, and evaluate the extent to which the measurement errors in demographic estimates may affect the baseline conservation assessment. We also estimate the potential effect that resumption of long-line fisheries in the vicinity of Amsterdam Island (one the alleged causes for its low numbers in the recent past) may have on the persistence of this population. Our results indicate that, in the absence of any impact of long-line fisheries, the Amsterdam albatross is unlikely to experience a decline larger than 20% of the current population abundance over the next 50 years. Our results point out the difficulty to assess with certainty the extinction risk of small populations despite the availability of long term data on their demography. They suggest that a very cautious approach should be taken for the preservation of small populations of long-lived species that cannot sustain any level of incidental by-catch. Any new long-line fishery resuming in the foraging range of the Amsterdam albatross, but especially close to Amsterdam Island, may rapidly put this species at risk of extinction.
|
|---|