Despite the predictions of the modeling approaches described in the recent reports of IPCC ((crop simulation models, agro-ecological zone (AEZ) and the model of the Ricardian approach)), which say that Canada, a temperate region, will probably play a more important role in feeding the world if the A2 scenario of the Special Report on Emissions Scenarios (SRES) comes to be true (Cline, 2007; IPCC, 2007), Canadian agriculture is facing many problems recently arising from climate change and variability. Hence, to cope with climate change and variability, it is not only mitigation that is important but also adaptation. And when it comes to adaptation, it is the climate variability that interests us rather than the increase in global average temperatures. The main characteristics of the vulnerability and adaptation to climate change are those related to climate variability and extremes, and not just change in average conditions (Chiotti et Johnston 1995; Means et al., 1997; Smit et al., 1997; Smithers et Smit, 1997; Karl et Knight, 1998; Berz, 1999; Hulme et al., 1999; Mendelsohn et al., 1999; Wandel et Smit, 2000; IPCC, 2001; Smit et Pilifosova, 2007). Research during the 1990s has emphasized the need to recognize the variability (or heterogeneity) of inherent spatial conditions (agro-climatic, soil resources, cultural values, …) in which agriculture developed, and therefore the importance of validating the indicators of adaptation and analyzing them in more detail to take into account the regional differentiation of agro-climatic conditions in relation to vulnerability and adaptive capacity (Bryant et al., 2007). For example, drought and excess rainfall were the most common impacts of climatic conditions identified by a sample of farmers in southern Ontario, representing 80% of responses (Smit et al., 1996). In addition, still talking about Canada, it is generally recognized that climate change has the potential to have the greatest impact on the Prairies and in central British Columbia, which is reflected in the hydrographs of streams in snowmelt in response to recent climate variability, and which may affect the timing of water availability (Leith and Whitfield, 1998; Whitfield, 2001). In addition, adaptation studies go beyond crop yields modeling to integrate adaptation which implies in particular that farmers can use some adaptation practices best suited to different climate scenarios (Bryant et al., 2000).
Adaptation refers to the responses of individuals, groups and governments to climatic stimuli or the effect of reducing vulnerability or susceptibility to negative impacts or potential damage associated with climate change (Carter et al., 1994; Watson et al., 1996; Pielke, 1998; Tol et al., 1998; UNEP, 1998; Wheaton and MacIver, 1999; Smit et al., 2000; UKCIP, 2003; Pilifosova and Smit, 2007). In addition, it is oriented to take advantage of opportunities associated with climate change (at least in some regions) (Carter et al., 1994, Watson et al., 1996; Pielke, 1998; Tol et al., 1998, UNEP, 1998 , Wheaton and MacIver, 1999; Smit et al., 2000; Pilifosova and Smit, 2007). For example, in Canada, most adaptation options are changes in agricultural practices and current public policy decision-making processes concerning a series of changing climatic conditions (including climate variability and extremes) and non-climate conditions (political, economic and social) (Smit and Skinner, 2002). Regarding climate change, adaptation is important from two perspectives – one is related to the assessment of impacts and vulnerabilities, the other is concerned with the development and evaluation of response options (Frankhausser 1996; Yohe et al., 1996; Tol et al., 1998; UNEP, 1998; Smit et al., 1999; Pittock and Jones, 2000; Pilifosova and Smit, 2007).
An article written by Barry Smit, Robert Blain and Philip Keddie in 1997 represents an example of farmers’ adaptation to climatic variability through the use of corn hybrid selection in Southern Ontario. This example of adaptation, crop development, comes under the different types of technological developments (Smit and Skinner, 2002). Crop development means the development of new crop varieties, including hybrids, to increase the tolerance and suitability of plants to temperature, moisture and other relevant climatic conditions (Smit and Skinner, 2002). In fact, hybrid varieties are developed by combining genetically different parents in order to enhance such attributes of disease and mould resistance, stalk strength, maturity time, and yield (Aldrich et al., 1975; Tollenaar et al., 1994). Corn hybrid varieties are available for a wide range of climatic conditions, including accumulated, measured as Corn Heat Units (CHU) (Smit et al., 1997).
The article of Smit el al. (1997) takes two sample counties in Southern Ontario, Lambton County and Wellington County, to allow a comparison of responses to climatic variability between farmers from different agricultural systems, specifically to show on what basis farmers choose the hybrid varieties. To do so, climate data were obtained for three weather stations in each of the two study counties to map the variations in CHUS for the different regions of Ontario for the period 1973-1993. The CHU map indicates the heat, relative to corn development needs, accumulated at a given location in an average year. As a result, farmers are advised to plant hybrid varieties that match the average CHUS at their location. It is important to note here that yield and maturity are very important in corn production because of the spatial variations in growing season length, and considerable resources have been devoted to hybrid development of these traits (Joseph and Keddie, 1985); hence, the importance of labeling and classifying corn hybrids according to their CHU designation (Brown and Bootsma, 1994). For each location, hybrids were classified into one of five categories according to their CHU rating relative to the recommended (i.e., average) CHU at that location. While input requirements do not vary significantly among corn hybrids, there is a correspondence between maturity length (heat requirements) and yield (Daynard 1994). Hybrids with lower heat requirements (earlier maturing or shorter-season varieties) generally have lower yields. Hybrids developed for higher levels of accumulated heat (later-maturing or longer-season varieties) invariably have higher yields, so long as they reach maturity.
Farmers choose their hybrid varieties prior to the growing season, presumably knowing the average heat at their location, but faced with the uncertainty inherent in year to-year variations or in growing-season conditions. Farmers in Wellington generally chose lower- CHU-rated varieties than did their counterparts in Lambton, reflecting the shorter average growing season in Wellington. Each year, and given the experience of previous years, farmers have to choose hybrid varieties to plant, not knowing whether the growing season will be long (warm), short (cool), or about average. For example, after the high CHU year of 1991, farmers chose significantly more longer-maturing and potentially higher-yielding, but riskier varieties. On the other hand, after the lower CHU years of 1992 and 1993, farmers’ hybrid selections became markedly more conservative. This tendency for more conservative hybrid choice following the lower CHU years is consistent across locations. Farmers in Wellington County generally chose more risky hybrids, and perhaps for this reason did not become even more risky in their choice up to 1992 to the degree apparent in Lambton. Nonetheless, in both counties, farmers chose shorter season (lower CHU) hybrids after 1992. Furthermore, this trend is apparent regardless of the size of farm or the area farmers planted in corn, and independent of enterprise orientation (Blain et al., 1995).
To conclude, selection of corn hybrid varieties (according to their maturity length or heat requirements) represents a means of coping with, or purposefully adapting to, a variable climate regime. And hybrids mean that they are not necessarily genetically modified organisms (GMOs). In addition, the key climatic condition for corn growth and maturity is the accumulated growing season heat, measured as Corn Heat Units (CHUS). So important is this attribute that hybrids are classified and labeled according to their CHU designation (Brown and Bootsma, 1994). However, mapping CHU variations is not sufficient because it does not provide insights into the relative risk positions of individuals. It does not allow determination of whether the trends reflect large changes in hybrid selection by a few producers or widespread changes among most farmers (Smit et al., 1997). As a result, a relative risk index is needed for each respondent at each year. And to assure the results of the average risk index, open-ended questions are needed. It is worth noting here that a conservative farmer may choose a short-season hybrid that has a greater probability of maturing but has lower yields; another farmer may choose a later-maturing hybrid that has a higher expected yield, but is more risky because it requires a higher level of accumulated heat to reach maturity. Also, it is true that farmers are advised to plant hybrid varieties that match the average CHUS at their location. However, farmers frequently select hybrids above or below the recommended CHU range. We should remember here that decisions in agriculture are influenced by a complex mix of economic, cultural, political, and environmental factors, most of which are variable and beyond the control of individual farmers (Ilbery, 1985; Smit et al., 1996). A farmer’s selection of corn hybrids involves consideration of many of these factors. Moreover, risk management is rarely limited to one action (in this case, choosing a hybrid mix). Other strategies to deal with climate-related risks, such as crop diversification or crop insurance, may help explain some of the apparently risky corn hybrid choices (Smithers and Smit, 1996). Any reduction (or removal) of crop insurance subsidies would mean that risks would be more fully borne by farmers, in which case a more careful consideration by farmers of the likelihood of certain CHU accumulations might be warranted – rather than weighting expectations heavily on the conditions of the previous year. For example, there is a broad adjustment to the prevailing climatic regime, as evidenced by the farmers in Wellington County choosing shorter-season varieties than farmers in Lambton. In addition, the study shows that technological developments are not the panacea for agriculture under climatic variability and change, even in technologically advanced commercial farming systems. Here, one should keep in mind that the excess of technical services, especially the physical capital, can cause damage to agriculture and the physical atmosphere, emitting more greenhouse gases. The second law of thermodynamics states that “all physical processes, natural and technological, proceed in such a way that the availability of the energy involved decreases” (Daly and Townsend, 1992). So, 100% efficiency does not exist. The first and the second laws of thermodynamics make it clear that all the energy used on the face of the Earth, renewable and non-renewable forms of energy, will ultimately be degraded to heat (Daly and Townsend, 1992). There would seem to be opportunities to reduce vulnerabilities to climatic variation not by developing new hybrids for this purpose, but by clarifying the nature of (and probabilities associated with) climatic variability, so that individuals can select hybrid mix strategies consistent with their risk preferences, rather than this seemingly reckless gambling with nature. Given the unpredictability of specific growing season conditions, farmers have little choice but to gamble, yet weighting choices so much on the last throw of the dice seems to be a poorly informed basis for decisions when the probabilities associated with climatic variability are well known.
Furthermore, the study of Smit et al. (1997) is broadly consistent with those from much of the work on human responses to environmental hazards (Kunreuther and Slovic, 1986; Burton et al., 1993; Palm, 1995), which has shown that adaptations are most powerfully influenced by most recent experiences, and that recognition of earlier experiences declines rather quickly with time.