Transition to density dependence in a reintroduced ecosystem engineer

Abstract

When does a reintroduced population of animals become self-regulating? Quantifying this is critical in determining when interventions can be tapered off, or when they may need to be reinstated. We tracked the growth trajectory of a reintroduced population to establish whether it was irruptive and/or had transitioned to self-regulation. In 2012, we reintroduced 32 eastern bettongs (Bettongia gaimardi), a potoroid marsupial from Tasmania, Australia, to a 485 ha exotic predator-proof fenced reserve in the Australian Capital Territory. We established a 92 cage trap monitoring network to track population growth between the Austral Autumn 2014 and Summer 2018. We used capture-recapture models to track changes in the population through time, and modelled ‘bettong weight’, ‘pouch occupancy’ and ‘age of pouch young’ with population variation, to establish potential associations with changes in population size. The estimated population grew from 32 individuals in 2012 to 100 in 2014, 192 in Autumn 2016, and then declined to 151 in Summer 2018. Estimated survival of adults was high—above 92% between most sessions. Adult female weights ranged between 0.485 and 2.428 kg, and adult males between 0.470 and 2.775 kg. Our study showed density dependence was achieved over the 6 year period. Low adult mortality, and variable pouch occupancy related to female weight, suggested that food availability had influenced lactation in females, with flow-on impacts on juvenile survival. Long-term, broad-scale population dynamics were probably driven by a mix of direct (e.g. disease, harvesting for other reintroductions), and indirect (i.e. climate dependent availability of nutritious food) influences on population size.

Appendix

See Figs. 8, 9, 10 and Tables 1, 2, 3 and 4.

Fig. 8

Histograms overlaid with density plots of the distribution of the estimated age of pouch young in each primary survey. Histogram bins are 10 days long. Percentage figures (rounded to 2 significant figures) above the bars represent the percentage of the population of pouch young falling within that age group

Fig. 9

Histograms overlaid with density plots of the distribution of the estimated age of pouch young in each primary survey. Histogram bins are 55 days long. Percentage figures (rounded to 2 significant figures) above the bars represent the percentage of the population of pouch young falling within that age group

Fig. 10

Bettong weight against time with Tukey honesty test of significance. We applied a Bonferroni correction due to multiple comparisons in the same model (p = 0.05/10 sessions; i.e. p = 0.005). Letters above the upper confidence limits denote significant differences between the estimates in each session

Table 1 Bettong trapping sessions and trapping events in each primary trapping session

Table 2 Pradel robust design capture-recapture model selection table using all data combined, regardless of sex

Table 3 Pradel robust design capture-recapture model selection table using group attributes function to assign two groups to the data (female and male)

Table 4 Survival and recruitment estimates using the Pradel robust design model