• demography;
  • dispersal;
  • metapopulations;
  • removal experiments;
  • rescue effect


  • 1
     An evaluation of the potentially adverse effects of measurement must be made before concluding that one is dealing with a ‘dispersal sink’.
  • 2
     We conducted a spatially and temporally replicated removal experiment on common brushtail possums (Trichosurusvulpecula) in uniformly suitable old-growth eucalypt forest in south-eastern Australia, that was designed to address the question: does immigration ‘rescue’ populations from extinction?
  • 3
     Despite taking precautions to minimize potential harm, analyses indicated some evidence of an adverse effect of handling on the survival of pouch-young and strong evidence of effects on adult survival. In addition, symptoms of stress associated with handling observed at our site, corresponded to symptoms reported in connection with the long-term (15 + years) trapping study on possums conducted by Efford et al. in the Orongorongo Valley (OV) of New Zealand.
  • 4
     Initial projections from a demographic model indicated that the resident population at our site was not replacing itself (births < deaths), suggesting that the site was a dispersal sink. This was inconsistent with the fact that the site was in prime habitat. Moreover, the measured rate of true immigration in response to experimental removals was not sufficient to maintain the population density. Using data from Efford (1998), our model confirmed his suggestion that the OV site also appears to be a dispersal sink for possums, despite being in prime habitat.
  • 5
     When otherwise undiagnosable deaths among adults were assumed to be due to handling and ‘right-censored’ (excluded), the projection was that the resident population at our site was stable (r ≅ 0), and therefore not in need of ‘rescue’ by immigration. Similarly, when survival estimates for the OV site were corrected by the same amounts, the projection was that the population at that site was also stable.
  • 6
     Most vacancies created by our experimental removals were filled by neighbouring residents that expanded their ranges into the removal areas. We suggest that the artificial ‘removal’ of residents as a consequence of deaths due to handling, may often induce an influx of such apparent immigrants, thereby giving the impression that immigration is ‘rescuing’ populations from extinction.