Geoscience Reference
In-Depth Information
compared to their current range of inter-annual variability (Beaumont et al.,
2011). Within the next 60 years, Beaumont et al. (2011) project that almost
all of the Global 200 ecoregions (238 regions considered to support excep-
tional biodiversity [Olson and Dinerstein, 2002]) will face climatic conditions
considered extreme compared to the baseline conditions of the 1961-90 period
(based on the SRES A2 scenario, 4°C by 2100). The entire range of 89 ecore-
gions is projected to experience extreme monthly temperatures even with global
warming of less than 2°C.
So what could we see happen between now and the possible arrival of a
Four Degree World in the second half of this century? Impacts are likely to
accrue in two non-mutually exclusive ways. Some change may be incremental.
For example, gradual increases in mean temperatures or rainfall could result
in gradual changes in ecotonal boundaries between ecosystems, such as those
observed between rainforests and savannahs over the past few decades in
northern Australia, thought to be mediated via changes in rainfall and fire (for
instance, Fensham et al., 2003). But other changes could be abrupt and trans-
formative. There is increasing evidence that complex systems such as ecosystems
have critical thresholds - so-called tipping points - when a rapid transition
occurs between one state and another ( Scheffer and Carpenter, 2003; Scheffer
et al., 2009; Laurance et al., 2011), and that these could be reached at or below
2°C global warming (Leadley et al., 2010a).
There is now some theoretical and empirical evidence that as different
systems approach such tipping points, they exhibit some common early warning
signals (Scheffer et al., 2009), including slower recovery from small perturba-
tions. Empirical research on identifying early warning signals in ecosystems
such as freshwater lakes and rangelands is emerging, but little has been done in
Australia. Identification of early warning signals will be an important tool for
predicting critical transitions, but for most terrestrial systems we do not have
sufficient understanding of the underlying physiological and ecological processes
and feedbacks, and in practice such signals may not be detectable against
background noise and uncertainty until it is too late.
A further complication for predicting exactly when and where such critical
thresholds could be reached is that few of the projections of ecosystem change
consider concurrent changes such as the direct effect of rising atmospheric CO
2
which may differentially affect woody and herbaceous species (Warren et al.,
2011). Increasing water use efficiency (WUE) over time by many plant species
may also mean that some negative impacts projected by large-scale models may
be over-estimated (Loehle, 2011). Other indirect impacts of climate change,
such as changes in fire regimes, are likely to be key to transformational change in
many regions. Fire frequency and intensity is driven largely by weather and fuel
loads. Increasing severity of fire weather is virtually certain in many regions, in
turn increasing the probability of ignition (Williams et al., 2009b). Rising atmos-
pheric CO
2
may result in greater fuel loads in regions where water availability
is not limiting (Williams et al.,
2009b). Climate change will also affect the
