Page 134 - Climate Change and Food Systems

climate change and food systems: global assessments and implications for food security and trade
found that, for high-end CC scenarios (higher values for radiative forcings), agroclimatic conditions deteriorate in many zones, in terms of increased drought stress and shortening
of the active growing season. The projections showed a marked need for adaptive measures in most zones, particularly for increasing soil water availability or drought resistance of crops. Rainfed agriculture was found likely to be affected by more climate-related risks and the number of extremely unfavourable years in many climate zones was projected to increase; this would result in higher interannual yield variability, constituting a considerable challenge for adaptation of crops and cropping systems. Recent applications of the agroclimatic index approach at national level have been reported by Hakala et al. (2012), Lalic et al. (2013)
and Rötter et al. (2013a). The latter authors used the approach to identify areas most prone to CC risks (i.e. heat and drought), and subsequently applied crop simulation modelling to evaluate alternative adaptation options
in these areas. Trnka et al. (2014) recently detailed an approach for wheat cultivation in Europe under CC, explicitly considering multiple climate-related stress occurrences.
b. Crop-climate or crop-weather models based on empirical statistical approaches, most
often on multiple regression analysis, have
a long tradition (Nix, 1985). With seed yield
or biomass yield as the dependent variable, models often contain a number of independent variables that represent temperature or rainfall characteristics over a certain time span during the crop growth cycle, or a variety of indices derived from weather data and interpreted
in agricultural terms. Statistical crop models have been applied to detect the influence of climate of the recent past on crop production trends (e.g. Lobell et al., 2011) and they are increasingly applied to predict crop yield responses to CC, although usually restricted to time horizons that do not reach too far into the future (Schlenker and Roberts, 2009;
Lobell and Burke, 2010; Lobell and Gourdji, 2012). An advantage of statistical crop yield models is that they take into account the effect of all kinds of yield-limiting factors (e.g. heat and water stress) and yield-reducing factors (such as weeds, pests and diseases, ozone levels, etc.). However, a disadvantage is that these yield-influencing factors usually cannot be separated from each other and are lumped together, which makes them less suitable
for evaluating alternative adaptation options. For this reason, they are useful primarily
for assessing CC impacts under actual farmers’ conditions that are characterized by suboptimum management.
c. Process-based crop simulation models are currently the most widely used tools for predicting crop productivity under CC, and they are applied from field to global scale (Angulo et al., 2013; Nelson et al., 2013; Rosenzweig et al., 2013; Müller and Robertson, 2014). Their advantage over statistical models is the ability to explicitly take into account interactions of genotype by environment by management (GxExM) and to quantify the relative effects of individual factors on crop development, growth and final yield. A range
of crop simulation models of various levels of complexity exists; however, even the relatively simple, most widely used crop simulation models (sometimes called summary models) suffer from high data demand in terms of calibration and validation, which restricts their meaningful application to a limited number
of crops and regions. Lack of data for model testing also makes it impossible to make use of the results from a large number of output variables; often only a few outputs can be utilized with confidence.
Particular capabilities and limitations of crop simulation models with respect to assessing CC effects on crop yields, including their role in integrated assessment methodology, are briefly presented here. The major shortcomings of the
114