Selection strategies for genotypes superior in their e[CO2] response cannot use traditional field methods because there are no natural [CO2] gradients available that could be exploited (Ainsworth et al., 2008a). Empirical selection under an increasing [CO2] apparently has not been successful in developing cultivars with optimum response to e[CO2]. Studies on wheat and oat cultivars demonstrated that some early 20th century cultivars can capitalise on rising [CO2] better than some modern, late 20th century cultivars (Manderscheid and Weigel, 1997; Ziska and Blumenthal, 2007; Ziska et al., 2004). This may be surprising, given that [CO2] was continually increasing in parallel to the ongoing cultivar selections, yet traits selected by breeders during the 20th century are not the same traits that are associated with superior
response to rising [CO2]. For example, selection for yield stability may have decreased flexible tillering response beneficial under e[CO2] (Ziska et al., 2004).