Depression of metabolic rate has been recorded for virtually all major animal phyla in response to environmental stress. The extent of depression is usually measured as the ratio of the depressed metabolic rate to the normal resting metabolic rate. Metabolic rate is sometimes only depressed to approx. 80% of the resting value (i.e. a depression of approx. 20% of resting); it is more commonly 5–40% of resting (i.e. a depression of approx. 60–95% of resting); extreme depression is to 1% or less of resting, or even to an unmeasurably low metabolic rate (i.e. a depression of approx. 99–100% of resting). We have examined the resting and depressed metabolic rate of animals as a function of their body mass, corrected to a common temperature. This allometric approach allows ready comparison of the absolute level of both resting and depressed metabolic rate for various animals, and suggests three general patterns of metabolic depression.

Firstly, metabolic depression to approx. 0.05–0.4 of rest is a common and remarkably consistent pattern for various non-cryptobiotic animals (e.g. molluscs, earthworms, crustaceans, fishes, amphibians, reptiles). This extent of metabolic depression is typical for dormant animals with ‘intrinsic’ depression, i.e. reduction of metabolic rate in anticipation of adverse environmental conditions but without substantial changes to their ionic or osmotic status, or state of body water. Some of these types of animal are able to survive anoxia for limited periods, and their anaerobic metabolic depression is also to approx. 0.05–0.4 of resting. Metabolic depression to much less than 0.2 of resting is apparent for some ‘resting’, ‘over-wintering’ or diapaused eggs of these animals, but this can be due to early developmental arrest so that the egg has a low ‘metabolic mass’ of developed tissue (compared to the overall mass of the egg) with no metabolic depression, rather than having metabolic depression of the entire cell mass. A profound decrease in metabolic rate occurs in hibernating (or aestivating) mammals and birds during torpor, e.g. to less than 0.01 of pre-torpor metabolic rate, but there is often no intrinsic metabolic depression in addition to that reduction in metabolic rate due to readjustment of thermoregulatory control and a decrease in body temperature with a concommitant Q10 effect. There may be a modest intrinsic metabolic depression for some species in shallow torpor (to approx. 0.86) and a more substantial metabolic depression for deep torpor (approx. 0.6), but any energy saving accruing from this intrinsic depression is small compared to the substantial savings accrued from the readjustment of thermoregulation and the Q10 effect.

Secondly, a more extreme pattern of metabolic depression (to <0.05 of rest) is evident for cryptobiotic animals. For these animals there is a profound change in their internal environment – for anoxybiotic animals there is an absence of oxygen and for osmobiotic, anhydrobiotic or cryobiotic animals there is an alteration of the ionic/osmotic balance or state of body water. Some normally aerobic animals can tolerate anoxia for considerable periods, and their duration of tolerance is inversely related to their magnitude of metabolic depression; anaerobic metabolic rate can be less than 0.005 of resting. The metabolic rate of anhydrobiotic animals is often so low as to be unmeasurable, if not zero. Thus, anhydrobiosis is the ultimate strategy for eggs or other stages of the life cycle to survive extended periods of environmental stress.

Thirdly, a pattern of absence of metabolism when normally hydrated (as opposed to anhydrobiotic or cryobiotic) is apparently unique to diapaused eggs of the brine-shrimp (Artemia spp., an anostracan crustacean) during anoxia. The apparent complete metabolic depression of anoxic yet hydrated cysts (and extreme metabolic depression of normoxic, hypoxic, or osmobiotic, yet hydrated cysts), is an obvious exception to the above patterns.

In searching for biochemical mechanisms for metabolic depression, it is clear that there are five general characteristics at the molecular level of cells which have a depressed metabolism; a decrease in pH, the presence of latent mRNA, a change in protein phosphorylation state, the maintenance of one particular energy-utilizing process (ion pumping), and the down-regulation of another (protein synthesis). Oxygen sensing is now the focus of intense investigation and obviously plays an important role in many aspects of cell biology. Recent studies show that oxygen sensing is involved in metabolic depression and research is now being directed towards characterising the proteins and mechanisms that comprise this response. As more data accumulate, oxygen sensing as a mechanism will probably become the sixth general characteristic of depressed cells.

The majority of studies on these general characteristics of metabolically depressed cells come from members of the most common group of animals that depress metabolism, those non-cryptobiotic animals that remain hydrated and depress to 0.05–0.4 of rest. These biochemical investigations are becoming more molecular and sophisticated, and directed towards defined processes, but as yet no complete mechanism has been delineated. The consistency of the molecular data within this group of animals suggests similar metabolic strategies and mechanisms associated with metabolic depression.

The biochemical ‘adaptations’ of anhydrobiotic organisms would seem to be related more to surviving the dramatic reduction in cell water content and its physico-chemical state, than to molecular mechanisms for lowering metabolic rate. Metabolic depression would seem to be an almost inevitable consequence of their altered hydration state.

The unique case of profound metabolic depression of hydrated Artemia spp. cysts under a variety of conditions could reflect unique mechanisms at the molecular level. However, the available data are not consistent with this possibility (with the exception of a uniquely large decrease in ATP concentration of depressed, hydrated Artemia spp. cysts) and the question remains: how do cells of anoxic and hydrated Artemia spp. differ from anoxic goldfish or turtle cells, enabling them so much more completely to depress their metabolism?

Within the Aphidoidea, most species of Aphididae, as long as they are in small numbers and not carrying plant viruses, do little perceptible damage to their food plants. In species that cause toxicoses, it is usually assumed that some component of the saliva must be responsible. Paradoxically, however, the salivary enzymes of Aphididae are similar to those that already occur in plants – oxidases and enzymes that depolymerize polysaccharides – and the salivary enzymes are injected in very small amounts relative to their counterparts in the plant. Damage to plants triggers defensive, biochemical responses, and it is suggested that the injected enzymes serve mainly to divert or counter responses at the immediate interface of stylets and plant tissues. The saliva of Aphididae contains non-enzymic, reducing compounds which, in the presence of oxidases, can combine with and inactivate defensive phytochemicals – including those released in response to damage and transported in the phloem sieve tube sap on which Aphididae feed. Salivary and gut oxidases deactivate ingested phytochemicals by oxidative polymerization. Aphididae inject saliva into sieve tubes before sustained ingestion of sap, and this saliva has been presumed to condition the sieve tubes, but in what way remains unclear. It is suggested that there is a dynamic biochemical interaction between aphids and plants; that the interaction is usually well balanced for most of the Aphididae; hence, no outcome is readily observable. Where a significant imbalance occurs, however, either the aphid is unable to feed, i.e. the plant is resistant, and/or the aphid does not effectively counter a hypersensitive response. Not all plant responses are disadvantageous to aphids. Gall-forming Aphidoidea trigger and control abnormal growth in the plant to the insects' advantage, possibly by eliciting vigorous oxidation in selective meristematic tissues, thereby limiting supply of molecular oxygen and inhibiting oxygen-dependent growth-controls. Current problems and possible approaches for further research are reviewed.

The relationships between macro-ecological patterns and physiological investigations in insects, especially those dealing with respiratory metabolism, are assessed in an attempt to encourage the development of the interaction between macroecology and physiological ecology. First, we demonstrate that although physiological ecology has been explicitly concerned with a number of issues relating to species boundaries, many questions remain unanswered. We argue that there are essentially two ways in which the relationship between physiological tolerances and species range boundaries have been investigated. The correlational approach involves physiological inference, physiological prediction, isocline analyses and climatic matching, and has often been criticized for a lack of rigour, while the experimental approach seeks to examine experimentally the relationships between physiological variables and range edges. Second, we use the recent debate on processes underlying latitudinal patterns in body size to caution against the conflation of patterns and processes operating at intraspecific and interspecific levels, the dangers inherent in invoking single explanatory variables, and an undue focus on adaptationist (e.g. optimization) rather than non-adaptationist explanations or some combination of the two. We show that both positive and negative relationships between body size and latitude have been found at the intraspecific level and suggest that interactions between temperature-induced heterochrony, and the relationship between habitat durational stability, growing season length, and generation time can be used to explain these differences. Similar variation in documented patterns is demonstrated at the interspecific level, and the mechanisms usually proffered to explain such clines (especially the starvation/desiccation-resistance hypothesis) are discussed. Interactions between various environmental factors, such as host-plant quality, and their effects on size clines are also discussed. Third, we argue that respiratory metabolism, as a measure of ATP cost, and its spatio-temporal variation are critical to many explanations of macroecological patterns. Adaptive changes in metabolism reputedly involve both depression (stress resistance) and elevation of metabolic rate, although recent studies are increasingly calling these ideas into question. In particular, flow-through respirometry is revolutionizing results by allowing careful separation of resting (or standard) and active metabolic rates. These techniques have rarely been applied to studies of metabolic cold adaptation in insects, one of the most polemical adaptations ascribed to high-latitude and high-altitude species. We conclude by arguing that physiological investigations of species tolerances are important in the context of macroecology, especially species distributional patterns and the possible impact of climate change thereon. However, we caution that relationships between abiotic variables, species tolerances, and distributional ranges may be non-linear and subject to considerable modification by the presence of other species, and that many of the pressing questions posed by macroecology have not been addressed by insect physiologists. Nonetheless, we suggest that because an understanding of the dynamics of species distributions is of considerable importance, especially in the context of current conservation problems, insect physiological ecology has much future scope.