282 Monitoring Threatened Species and Ecological Communities
were spinifex cover, annual rainfall in the prior year and capture rates of D. blythi.
Total rodent population size was used for D. blythi as rodents are a key prey source
for this species; D. blythi is also a potential predator and competitor for S.
youngsoni. Spinifex is the dominant vegetation at all sites and is used for shelter by
both species of dasyurid (Masters 1993). The 1-year lags for annual rainfall and
rodent prey population size were used to account for the time required for small
mammals to respond, via breeding, to rainfall events (Greenville et al. 2012, 2016).
Further details of the modelling approach and detailed results can be found in
Greenville et al. (2016).
Results
Over the period of sampling, 888 D. blythi and 2491 S. youngsoni were captured.
The mulgara f luctuated from zero to almost 10 captures per 100 trap-nights across
all sites throughout the study, and showed considerable synchrony in the timing of
peaks and troughs in numbers (Fig. 21.1). These f luctuations tended to follow
rainfall, suggesting that rainfall might inf luence the capture rate of this species.
Previous MARSS modelling confirmed that rainfall in the prior year had a positive
effect on captures of the mulgara, as did the cover of spinifex and numbers of
Fig. 21.1. Long-term pattern in capture rates (numbers of captures per 100 trap-nights) of the brush-tailed
mulgara Dasycercus blythi at nine sites in the Simpson Desert, western Queensland. The captures rates
(symbols), predicted population size (line) and 95% credible intervals (shading) are modelled using a
one-state Bayesian multivariate autoregressive state–space model with no covariates; names and symbols
represent the nine different sites (see Fig. 21.2 for site locations). Mulgara drawing courtesy of Alison Foster.