TE is difficult to measure in the field since

TE is difficult to measure in the field, since transpiration measurements necessitate expensive facilities like lysimeters, which is particularly challenging for large crops like banana. Therefore TE is estimated by other proxy variables like carbon isotope discrimination (δ13C), which is an indication of stomatal closure throughout a plant\'s life and is thereby closely related to transpiration and carbon uptake for growth (Cabrera-Bosquet et al., 2007, Farquhar and Richards, 1984, Hubick et al., 1986 and Kondo et al., 2004). A good correlation between TE and δ13C was observed in this study as well (Fig. 4). This is mainly attributed to the pronounced effect of water deficit on both δ13C and TE resulting in a good correlation. However, the HBX 41108 of the method is not high enough for further assessment of the correlation between δ13C and TE during water deficit, where selection takes place (Fig. 5). Consequently, selection for high δ13C plants during water deficit would not necessarily lead to the selection of high TE plants. The 13C levels are only determined by the possibility of CO2 exchange with the atmosphere and thus opening and closing of stomata while TE takes into account the real dry matter accumulation and thereby growth. Another possible explanation for the lack of correlation between TE and δ13C as suggested by Turner et al. (2007b). It can be explained by the difference in dry matter allocation between leaves and corm. Since TE was determined by the dry matter accumulated in the aboveground parts of the plant and δ13C was only measured at leaf level, differences in dry matter allocation could influence the correlation between both variables (Turner et al., 2007b). Furthermore, dry matter allocation is known to be a strategy for coping with drought and can differ among cultivars.