Fig 1.
Modelled landscape with koomal presence.
Pre and post 1980 presences recorded within the South-West Australian Floristic Region (SWAFR).
Table 1.
Contribution of bioclimatic variables to the koomal SDM using all variables and five “minimum set” variables alone.
* denotes variable used in minimum set.
Fig 2.
Comparison between MaxEnt koomal SDMs using full suite of 19 bioclimatic variables (a) and selected suite of 5 most significant variables (b).
Lists of both sets of variables are given in Table 1.
Fig 3.
Histograms and line representing the response of koomal to the proportion of remnant vegetation cover as calculated from the three spatial variables.
R2 shows fit to linear model using raw data.
Table 2.
Summary statistics for the response of koomal to the proportion of remnant vegetation cover as calculated from the three variables given in Fig 3.
Table 3.
Model accuracy indicators with remnant vegetation perspective data sets added.
Fig 4.
MaxEnt PD calcualted with (a) the minimum set of 5 bioclimatic variables alone and (b) with the minimum set and 9 km2 variable model (the baseline model).
Fig 5.
Overlays showing baseline with RCP 4.5 (a) and RCP 8.5 (b) projected distributions overlaid to show areas identified in baseline and projected distributions and secure areas.
Secure areas being those that are recognised as current potential distribution and which are predicted to remain as such.
Fig 6.
a) RCP4.5 and b) RCP 8.5 2070 SDMs with jarrah dieback infested areas removed.
Fig 7.
RCP 8.5 2070 projection of koomal distribution (Fig 6b) with current annual average precipitation (mm) overlaid.