High resolution distribution modelling of a threatened short-range endemic plant informed by edaphic factors.

Short-range endemic plants often have edaphic specialisations that, with
their restricted distributions, expose them to increased risk of
anthropogenic extinction. Here, we present a modelling approach to
understand habitat suitability for Ricinocarpos brevis R.J.F.Hend.
& Mollemans (Euphorbiaceae), a threatened shrub confined to three
isolated populations in the semi-arid south-west of Western Australia. The
model is a maximum entropy species distribution projection constructed on
the basis of physical soil characteristics and geomorphology data at
approximately 25 m2 (1 arc-second) resolution. The model predicts the
species to occur on shallow, low bulk density soils that are located high
in the landscape. The model shows high affinity (72.1% average likelihood
of occurrence) for the known populations of R. brevis, as well as
identifying likely locations that are not currently known to support the
species. There was a strong relationship between the likelihood of R.
brevis occurrence and soil moisture content that the model estimated at a
depth of 20 cm. We advocate that our approach should be standardised using
publicly available data to generate testable hypotheses for the
distribution and conservation management of short--range endemic plant
species for all of continental Australia.