Assessment of the effects of high or low temperature stress on insects is generally undertaken in one of two ways. In the first approach, groups of individuals are exposed to a given temperature for a fixed period, following which recovery (% survival in the group) is assessed at a more 'normal' temperature after several hours or days (Krebs and Loeschcke 1995a). These are often referred to as mortality assays. Alternatively, individual organisms are exposed to temperatures changing at a given rate and the temperature at which they are knocked down or show spasms is recorded (Roberts et al. 1991; Klok and Chown 1997; Gibert and Huey 2001). The latter method is often modified such that a constant temperature is maintained and the time the insects take to be knocked down is recorded (Hoffmann et al. 1997), or time to recovery after a set exposure is examined (David et al. 2003). In addition, in a few studies, a combination of temperature change and exposure time has been used (Worland et al. 1992). Lutterschmidt and Hutchison (1997) provide a critique of the major methods for the assessment of high temperature limits and refer to the fixed temperature assays involving survival assessments as 'static methods', and the changing temperature assays as 'dynamic methods'. Although Lutterschmidt and Hutchison (1997) come out in favour of the dynamic, critical thermal maximum (CTmax) method as a standard measure of thermotolerance, they argue that both methods are useful for assessing tolerance, though they caution that these two methods might be measuring different things.
This idea has been particularly well explored in a variety of Drosophila species (Hoffmann et al. 2003b). In lines of D. melanogaster selected for knockdown resistance to heat in a 'knockdown tube' (Huey et al. 1992), knockdown resistance improves substantially compared with control lines. However, there is no effect of selection on knockdown of individuals assessed singly in smaller vials, or on survival in mortality assays (Hoffmann et al. 1997). Similarly, selection with and without hardening affects hardening in the 'knockdown tube', but not in the smaller vials. These results have been verified with isofemale lines (lines developed from a single mated female—see Hoffmann and Parsons 1988) and hybrid lines of D. serrata and D. birchii (Berrigan and Hoffmann 1998). Thus, at the intraspecific level there appear to be no genetic correlations between knockdown and mortality measures of heat resistance. That is, mortality, recovery, and knockdown measures of thermotolerance may be associated with different mechanisms and genes (Gilchrist et al. 1997; Hoffmann et al. 1997). In support of this idea, McColl et al. (1996) have shown that selection for increased resistance to knockdown with hardening results in allelic changes in the stress genes hsr-omega and hsp68 in D. melanogaster. In turn, the absence of correlated changes in other stress resistance traits (high temperature mortality—usually associated with variation in hsp70 (Feder 1999), desiccation, cold) when knockdown times (with and without hardening) are altered, suggests that these genes are not associated with general stress resistance. These findings support the idea that different mechanisms of thermal tolerance are associated with different genes (Sorensen and Loeschcke 2001).
In a recent comparison of six Drosophila species, Berrigan (2000) found significant correlations between thermotolerance assessed using knockdown in the 'knockdown tube', knockdown in small vials, and mortality assays. The presence of correlations in these measures between species, but not within species, suggests that the interspecific correlations are a consequence of correlated selection regimes, and that differences in thermotolerance might be adaptive responses to the environment. A similar, strong correlation between cold stupor (knockdown temperature) and LLT has been found among eight other Drosophila species (Hori and Kimura 1998), providing additional support for this idea.
These findings raise the issue of which measure should be used for assessing thermotolerance. To date, most studies have been based on mortality assays (Berrigan and Hoffmann 1998). However, Hoffmann et al. (1997) and Sorensen and Loeschcke (2001) have argued that knockdown assays may be more relevant to investigations of the ecology and evolution of organisms because stress levels in these assays are less severe than in those involving mortality. They are, therefore, more likely to be relevant to the survival and fertility of insects in the field. While this is probably true for highly mobile stages (e.g. winged adults, or fast-moving hemipteran nymphs), more sedentary stages (larvae, pupae), or those restricted to a given habitat (galling insects, inhabitants of necrotic fruit) might not have the luxury of a behavioural response to extreme temperatures (Huey 1991). For example, larvae of D. melanogaster may be routinely exposed to potentially lethal temperatures for substantial periods (Feder 1996; Feder et al. 1997a; Roberts and Feder 1999) with little prospect of escape (Fig. 5.4). In consequence, mortality assays may be more relevant to these stages than to adult flies, which may never experience such high temperatures (Feder et al. 2000b). Arguably, there may also be situations where mobile stages cannot respond behaviourally to extreme temperatures. Thus, characterization of the full response of an insect to potentially lethal temperatures could potentially require assessment of several measures of tolerance. Some measures will indicate the thermal limits to activity, while others will provide an estimate of the time for which a given stressful temperature (including subzero temperatures) might be endured. While one of these measures might be more convenient for laboratory purposes than the others (Gibert and Huey 2001), it should be recognized that these measures represent assessments of different aspects of an organism's physiology (Feder 1996; Sorensen and Loeschcke 2001), the ecological relevance of which may be difficult to ascertain without suitable information on the natural history of the species involved (Sorensen et al. 2001). A particular problem in this regard, especially in the context of low temperature tolerance, is the exposure time used in any given assessment.
Was this article helpful?