Basher et al. (1998) conclude that alpine systems are among the most vulnerable systems in the region. Despite the fact that they cover only a small area, they are important for many plant and animal species, many of which are listed as threatened. These systems also are under pressure from tourism activity. The Australian Alps are relatively low altitude (maximum about 2,000 m), and much of the Alpine ecosystem area and ski fields are marginal. Most year-to-year variability is related to large fluctuations in precipitation, but interannual temperature variations are small compared to warming anticipated in the 21st century. Studies by Hewitt (1994), Whetton et al. (1996b), and Whetton (1998) all point to a high degree of sensitivity of seasonal snow cover duration and depth. For Australia, Whetton (1998) estimates, for the full range of CSIRO (1996a) scenarios, an 18-66% reduction in the total area of snow cover by 2030 and a 39-96% reduction by 2070. This would seriously affect the range of certain alpine ecosystems and species (Bennett et al., 1991). Decreases in precipitation and increased fire danger also would affect alpine ecosystems adversely.
There seems to be little opportunity for adaptation by alpine ecosystems in Australia, which cannot retreat upward very far because of the limited height of Australian hills and mountains. There are various options for the rapidly expanding mountain-based recreation industry, including increased summer recreation and artificial snowmaking. These adaptations would increase stress on alpine ecosystems and water resources.
The New Zealand Alps are of higher altitude (up to 3,700 m); about 9% of the New Zealand landmass is above the treeline. A large number of species (for example, 25% of vascular plants), which often are highly distinctive, grow there. Despite a 0.5°C rise in New Zealand's mean annual temperatures since the 1860s, there has been no significant rise in the treeline or shrubland expansion (Wardle and Coleman, 1992), and it seems unlikely that there will be any significant threat to alpine ecosystems from warming in the medium term.
The Australian State of the Environment Report (1996) states, "Wetlands
continue to be under threat, and large numbers are already destroyed."
For example, Johnson et al. (1999) estimate wetland loss of about 70% in the
Herbert River catchment of Northern Queensland between 1943 and 1996. Wetland
loss is caused by many processes, including water storage; hydroelectric and
irrigation schemes; dams, weirs, and river management works; desnagging and
channelization; changes to flow, water level, and thermal regimes; removal of
instream cover; increased siltation; toxic pollution and destruction of nursery
and spawning or breeding areas (Jackson, 1997); and use of wetlands for agriculture
(Johnson et al., 1999). Climate change will add to these factors through changes
in inflow and increased water losses.
Specific threats to wetlands from climate change and sea-level rise have been studied as part of a national vulnerability assessment (Waterman, 1996). The best example is provided for Kakadu National Park in northern Australia. There are fears that World Heritage and Ramsar-recognized freshwater wetlands in this park could become saline, given current expectations of sea-level rise and climate change (Bayliss et al., 1997; Eliot et al., 1999). Although this analysis is supported by a large data resource, it is speculative, and efforts to develop more definite monitoring tools are needed. However, it does raise the possibility that many other Australian coastal wetlands could be similarly affected. Some of these wetlands may be unable to migrate upstream because of physical barriers in the landscape.
Many inland wetlands are subject to reduced frequency of filling as a result of water diversion for irrigation, and they also may be seriously affected by reductions in seasonal or annual rainfalls in the catchments as a result of climate change (Hassall and Associates et al., 1998). This may threaten the reproduction of migratory birds (some species of which already are under threat), which rely on wetlands for their breeding cycle (Kingsford and Thomas, 1995; Kingsford and Johnson, 1998; Kingsford et al., 1999). Large decreases in inflow predicted for the Macquarie River and several rivers in northern Victoria by Hassall and Associates et al. (1998) and Schreider et al. (1996, 1997) for scenarios that are consistent with the latest AOGCM simulations would have major impacts on wetland ecosystems.
Wetlands in New Zealand are the most threatened ecosystems; they have declined by 85% since European settlement (Stephenson, 1983). The vast majority have been drained or irretrievably modified by fire, grazing, flood control works, reclamation, or creation of reservoirs. Eutrophification, weed invasion, and pollution have greatly reduced their biodiversity (Taylor and Smith, 1997). More than 50% of the 73 significant wetlands that meet the Ramsar Convention standards for international wetlands are in coastal districts and will be impacted by rising sea levels. Most important wetlands are in highly urbanized or productive landscape settings and therefore have limited options for adaptation to decreased size or increased salinization.
Many Australian river systems, particularly in the southeast and southwest, have been degraded through diversion of water via dams, barrages, channels, and so forth, principally for irrigated agriculture. Many New Zealand rivers have been affected by hydroelectric generation; diversion of water for irrigation; agricultural, manufacturing, and urban pollution; and biotic invasion. Recent research has shown that river ecosystems are particularly sensitive to extremes in flow. Most research has been on the effects of flood flows. Droughts, as opposed to floods, have a slow onset and although recovery from floods by river flora and fauna is relatively rapid, recovery after droughts tends to be slow, may be incomplete, and may lag well behind the breaking of the drought (Lake, 2000). Floods and droughts interact with nutrient supply (Hildrew and Townsend, 1987; Biggs, 1996), so the effects of any possible changes in their frequency and magnitude need to be evaluated within the context of other human activities and climate-induced land-use change.
Current ranges of scenarios tend to suggest reductions in mean flow in many Australian rivers, similar to or greater than those in Schreider et al. (1997) and Hassall and Associates et al. (1998). In particular, any tendency toward more frequent or severe El Niño-like conditions beyond that already contained in the CSIRO (1996a) scenarios would further threaten many riverine and inland wetland systems in Australia and New Zealand. Findings of increased drought frequency and severity in eastern Australia under an NCAR CCMO transient simulation (Kothavala, 1999) and 12-35% reductions in mean flow by 2050 in the Murray-Darling basin by the 2050s in Arnell (1999), using results from the HadCM2 and HadCM3 GCM simulations, are cause for concern. Arnell (1999) found reductions in maximum and minimum flows. Walsh et al. (2000) found less severe increases in drought in Queensland, based on simulations with the CSIRO RCM nested in the CSIRO Mark 2 GCM.
Implications of these findings for riverine ecosystems and estuaries (Vance
et al., 1998; Loneragan and Bunn, 1999) and possible adaptations have yet to
be investigated, although reduced diversions from rivers to increase environmental
flows is one possibility. This could be achieved through increased WUE, imposition
of caps on water diversions, or water pricing and trading, but the latter two
measures are controversial and would have strong implications for rural industry
(e.g., see ABARE, 1999). Increased efficiency in water delivery for irrigation
currently is the favored option for restoring environmental flows in the heavily
depleted Snowy River in southeastern Australia.
Other reports in this collection