Agriculture in an even more sunburnt country - Four Degrees of Global Warming - page 94

southern Australia (often 25 per cent or higher) and about an even chance for
either a decrease or increase in the northern parts of Australia (see Whetton
et al., 2013, Chapter 2 , this volume; Hennessy et al . , 2007). Projected rainfall
changes in the main farming zones vary substantially between seasons, with
decreases particularly likely in winter and spring and with a low likelihood of
increases in rainfall in these seasons (CSIRO and BoM, 2007). While some of
the rainfall decreases projected in the scenarios do not seem large (e.g. 25 per
cent), in a grain-cropping location like Birchip in northern Victoria, such reduc-
tions would result in more than a doubling of the frequency of 'exceptional'
droughts, a reduction by 90 per cent of the frequency of high rainfall years, and
the rainfall historically experienced in half of all years would be experienced in
only 19 per cent of years.
Wheat crops during the last decade have experienced significant declines
during extended dry periods (ABARES, 2012), reducing the amount available
for export. The prospect of increased frequency and severity of droughts (see
Whetton et al., 2013; Braganza et al., 2013 [ Chapters 2 and 3 , this volume]; and
Hennessy et al . , 2008) is likely to exacerbate this situation. The sensitivity of
future crop yield to rainfall changes was assessed by Howden (2002) and Luo et
al. (2005), who found that in most sites there was high sensitivity to reductions
in mean rainfall; however, yield was less sensitive to increases in mean rainfall.
This is consistent with observations of yield across years and regions, with yields
in high rainfall years demonstrating various impacts from diseases, pests, water-
logging and nutrient leaching (e.g. Stephens and Lyons, 1998).
Rainfall has also historically been associated with determining land-use in
Australia, Goyders Line in South Australia being perhaps the most well-known
example (Nidumolu et al., 2012). Reductions in rainfall and increases in
potential evaporation rates appear likely to reduce the area cropped in Australia,
with the inland margins of the cropping belt moving towards the coast and only
small compensatory increases in cropping on the coastal margin occurring due to
other land use pressures (Nidumolu et al., 2012). Thus reductions in rainfall of
the scale indicated above would not only reduce site-based yield, but they would
also reduce the area cropped.
Reductions in rainfall are also likely to favour light-textured, sandy soils rather
than heavy soils with high clay content (van Ittersum et al., 2003; Ludwig and
Asseng, 2006), as in sandy soils, small rainfall events contribute to soil water in
the plant available range, whereas this moisture maybe unavailable to plants in
heavier clay soils (Lawes et al., 2009). Sea level rises may similarly reduce areas
suitable for cropping - particularly with regard to sugarcane - due to salinity
intrusion and flooding (Murphy and Sorensen, 2001).
The regional and industry impacts of the above changes could be marked.
Gunasekera et al. (2008) estimated that with levels of global warming consid-
erably less than the 1 4°C assumed for this topic (0.8-2.8°C), Australia would
see a decline in wheat production of up to 13 per cent and beef production of
19 per cent by 2050. Indicative decreases in production per unit area of core
agricultural commodities are up to 40 per cent under the more extreme scenarios