Fig. 1. Representative data plots of temperature (grey, upper lines) and CO₂ production (black, lower lines) from Pringleophaga marioni caterpillars exposed to freezing treatments at (A) –5.8°C (in which all caterpillars survive), (B) –6°C (which results in caterpillars being left moribund and then dead) and (C) –18°C (after which no caterpillars show muscle tone or response to stimuli). Breakpoint (b) and temperature of crystallisation (T_c) are marked. Small cycles in the CO₂ trace shown in B are a result of minor temperature fluctuations at the IRGA. Whenever the range of these fluctuations exceeded the mean, the data were discarded. The break in the middle graph represents a change in data files (we were restricted by the memory that the acquisition software was able to allot to data gathering), and the slight difference in trace afterwards (equivalent to a maximum of 2.1 µW) is a result of re-baselining mid-data-acquisition. The masses of the caterpillars used in the presented data were 0.2342 g (A), 0.1977 g (B) and 0.3826 g (C), accounting for the variation in the scale of the right-hand y-axes.