Seasonal energetics: are insects constrained by energy during dormancy?

Most environments, including tropical and aquatic environments, face seasonal variation in key abiotic and biotic parameters which fundamentally affects all organisms. As such, life has evolved to persist through periods of seasonal stress and to exploit periods of seasonal abundance (McNamara and Houston, 2008; Nelson et al., 2010). In seasonal environments, some parts of the year tend to be severely resource limited, making it difficult to meet life-sustaining energetic demands. One common strategy for surviving the resource-limited periods that exist across major phylogenetic divides is to enter a dormant state (Wilsterman et al., 2021). In this dormant state, energy consumption is minimized through metabolic suppression and survival is primarily driven by consumption of energy stored before entering the resting stage (Guppy and Withers, 1999; Hahn and Denlinger, 2007; Short and Hahn, 2023). In insects, common forms of dormancy are winter hibernation (diapause; see Glossary) and summer hibernation (aestivation; see Glossary) (Denlinger, 2022; Koštál, 2006).

The topic of energetics during dormancy in insects has received considerable attention over the last century, including several important papers in Journal of Experimental Biology (Bale and Hayward, 2010; Denlinger et al., 1984; Schneiderman and Horwitz, 1958; Storey and Storey, 2012). A major reason for this attention is the staggering durations of dormancy that insects can sustain. Although dormancy lasting for a few months is most common, some species remain dormant for years. For instance, the midge Contarinia tritici (Diptera) routinely overwinters for 3 years (Barnes, 1952), the beetle Leptinotarsa decemlineata (Coleoptera) can be dormant for up to 9 years (Tauber and Tauber, 2002), and the moth Prodoxus y-inversus (Lepidoptera), incredibly, can stay dormant for 30 years (Powell, 2001). This wide span of dormancy durations begs the general questions: how energetically constrained are insects during their dormancy?; and what factors ultimately drive energy use during dormancy?

A number of studies suggest that dormancy energetics is key to fitness (reviewed in Sinclair, 2015; Varpe, 2017; Wilsterman et al., 2021). Not only do insects need energy to sustain, repair and protect themselves during the dormant stage but also they need access to energy stores for post-dormancy processes (Hahn and Denlinger, 2007, 2011). These include the completion of development (in the case of egg, larval/nymphal or pupal dormancy; Dhillon and Hasan, 2018), reproductive maturation (in the case of adult dormancy; Bosch et al., 2010), regeneration of tissues (e.g. flight musculature) degraded prior to dormancy (Lebenzon et al., 2022; Stegwee et al., 1963) and dispersal or migration (Milner et al., 1992). Some insect species have the ability to feed at reduced rates during dormancy or at high rates after dormancy is broken (Hahn and Denlinger, 2007). However, many capital breeding insects (i.e. those that primarily use stored energy for reproduction; see Glossary) do not have this capacity and must survive both dormancy and post-dormancy solely on energy stores accumulated pre-dormancy (Varpe, 2017). Intrinsic variation in life-history traits [e.g. dormancy life stage, post-dormancy dispersal, position along the capital to income breeding (see Glossary) axis] will thus influence how energetically constrained a dormant insect is and affect the link between energetics and fitness. However, to our knowledge, no comprehensive phylogenetic or life-history-related analyses have been performed predicting vulnerabilities of certain dormancy strategies to depletion of energy reserves during dormancy.

Energy use is influenced not only by intrinsic factors but also by extrinsic factors, such as dormancy microclimate [e.g. in soil, above soil, in caves, in trees, within aggregations (see Glossary)] (Danks, 1987; Szejner-Sigal and Williams, 2022) and macroclimate (e.g. latitude or altitude; Roberts et al., 2021; Williams et al., 2012). For instance, dormancy duration is regulated by exposure to low temperatures in many insects (Hodek and Hodková, 1988; Süess et al., 2022; Trimble et al., 1990). Low temperatures additionally assist in energy conservation through the temperature dependence of metabolic rate. Thus, warmer winters may lead to energetic stress both directly, by increasing metabolic rate, and indirectly, by extending the duration of dormancy stages that require cold before they can be terminated (Bale and Hayward, 2010; Lehmann et al., 2017; Nielsen et al., 2022). However, low temperatures also come with a risk of cold stress, leading to the evolution of robust cold-tolerance mechanisms in insects that undergo dormancy (e.g. cryoprotectants, antifreeze proteins and membrane remodeling; Teets and Denlinger, 2013; Teets et al., 2023). Thus, the optimal dormancy temperature is likely to be a compromise between the competing costs of cold stress (where warmer temperatures are beneficial) and energetic stress (where colder temperatures are beneficial) (Irwin and Lee, 2002, 2003).

Furthermore, the interactions between extrinsic (e.g. abiotic and biotic environmental conditions) and intrinsic factors (e.g. life history) are of major importance. For instance, the variation in anatomical complexity among species that are dormant in different life stages should strongly influence their thermal sensitivity. An insect undergoing adult dormancy is expected to maintain more complex and developed tissues, and thus have high baseline energetic demands. But insects undergoing early pupal dormancy may have less energetically demanding tissues, resulting in lower baseline energetic demands (Boggs, 2009; Sinclair, 2015). Indeed, there is evidence supporting that thermal sensitivity and thermal tolerance are life-stage dependent and that thermal sensitivity increases through ontogeny (Folguera et al., 2010; Freda et al., 2019). In addition, there is evidence that life-stage-specific thermal sensitivities and tolerances are driven by life-stage-specific ecological pressures (Boychuk et al., 2015; Kingsolver and Buckley, 2020; Zhang et al., 2015). Although it is tempting to consider one particular life stage or developmental stage as more optimal for dormancy than another, overall phylogenetic signals (see Glossary) are variable and differences in dormant life stages are common within insect families (Fig. 1) (Denlinger, 2022; Powell, 1987). As such, this supports the hypothesis that the timing and duration of extant strategies reflect a compromise between competing selective pressures shaping dormancy in the context of an insect's life-history evolution.

Overall, research over the last century has operated under the strong assumption that energy stores are depleted during dormancy and that this depletion has serious negative fitness consequences. The fact that different insect species show dormancies that can vary in length between 1 and 30 years, however, adds to the evidence that the extent of energetic stress can vary, or perhaps be modulated. There are indeed several exciting new lines of research in the field of diapause energetics that suggest such flexibility. For instance, dormant insects may sense their energy store size and adjust utilization rate dynamically, as recently reviewed by Short and Hahn (2023). Further, insects may shift oxidation substrates depending on metabolic demands, and even alternate between aerobic and anaerobic metabolism in a regulated manner (Chen et al., 2021). Perhaps even more dramatically, new research shows extensive mitophagy (see Glossary) during dormancy (Lebenzon et al., 2022), indicating that insects may be able to effectively regulate their overall aerobic capacity (and thus energy utilization rate) during dormancy. Thus, there is much to explore about these and other new physiological mechanisms and constraints across broad phylogenetic divides and ecological contrasts in insects.

The mechanisms underlying the observed variation among species in dormancy energetics are, we believe, an important emerging research topic (Denlinger, 2023). Energy use can provide an important tool for understanding what limits success during dormancy, but it is reliant on several key assumptions (Fig. 2): (1) energy is being used during dormancy and that will be reflected in a substantial decrease in energy stores; (2) energy use is a temperature-dependent process, and higher temperatures lead to higher rates of energy use; (3) energy stores during dormancy are related to fitness. There is large variation in how insects use energy during dormancy, challenging our ability to make broad generalizations (Hahn and Denlinger, 2011; Short and Hahn, 2023). The breadth of physiological capabilities and life-history strategies make it important to consider not only where results match our expectations but also, perhaps more importantly, where they do not. Only then can we gain a more complete understanding of the role of energy use in dormancy and make more accurate generalizations about the species or dormancy strategies most vulnerable to ongoing climate warming. In this Commentary article for the centenary of Journal of Experimental Biology, we explore when energy store size and utilization seem to matter, and when they do not, and highlight gaps in our current understanding. We also hope to highlight some promising new directions that the field could take during the coming century.

Energy stores and their utilization rates are considered key dormancy adaptations in many species across the tree of life, and have been thoroughly reviewed in insects by Hahn and Denlinger (2007, 2011), Sinclair (2015) and Short and Hahn (2023). Here, we present a few cases from the field and the laboratory that clearly support the assumptions (Fig. 2) regarding energy use during dormancy.

Insects often accumulate energy stores prior to the onset of the adverse season in which they are dormant, in order to survive long periods without feeding. These energy stores cover metabolic and maintenance costs and are expected to result in a decline of energy stores throughout diapause (Fig. 3). For instance, in the beetle Aulacophora nigripennis, lipid stores decrease 75% by the end of winter (Watanabe and Tanaka, 1998). However, the rate of energy use is not always linear through dormancy, suggesting that different stages of diapause have different energetic demands. For example, the fly Sarcophaga crassipalpis shows a sharp drop in lipid energy stores during early diapause, followed by no change in lipid stores during the last portion of diapause, suggesting a switch to reliance on non-lipid reserves (Adedokun and Denlinger, 1985).

In addition to maintaining baseline metabolic costs, there are also energetic costs associated with tissue damage, repair and stress resistance (Hahn and Denlinger, 2011). For example, the cricket Gryllus pennsylvanicus shows a marked increase in metabolic rate shortly after an acute cold stress, linking energy use to re-establishing homeostasis and membrane potentials, and to cold tolerance more generally (MacMillan et al., 2012). Štětina et al. (2018) also found evidence of energetic costs associated with recovery from multiple types of cold stress in the fly Chymomyza costata. Many insects also produce cryoprotectants, small molecular weight compounds such as polyols, to reduce cold stress during winter (Teets and Denlinger, 2013; Teets et al., 2023). These cryoprotectants can be metabolically derived from carbohydrate stores (i.e. glycogen), which can also fuel glycolysis and other processes to maintain baseline metabolic demands (Sinclair, 2015), providing a dynamic energy pool that can be allocated to maintenance and/or stress tolerance. How flexibly insects can switch between energy pools or use cryoprotectants as metabolic fuel when other energy pools are low remains understudied. More generally, in what sort of environment (predictable/stochastic or variable/stable) would it be favorable to evolve such flexibility?

Accounting for thermal effects on metabolic rates and energy use has advanced our understanding of overwintering energetics (reviewed in Sinclair, 2015). The mechanisms driving metabolic rate-temperature relationships follow an exponential function that is described by the metabolic level (i.e. the intercept of log-transformed metabolic rate–temperature function, or metabolic intensity) and thermal sensitivity (i.e. slope or rate of increase) (Fig. 2). Dormancy is often accompanied by either lower intercepts – which reflects metabolic suppression – and/or changes in thermal sensitivity, which determine how much energy usage rates change with temperature (Colinet et al., 2015; Irwin and Lee, 2003; Kovac et al., 2022). Thermal sensitivity of metabolism is reduced during dormancy in the fly Rhagoletis pomonella (Toxopeus et al., 2021), but it is still found to have a Q₁₀ coefficient (see Glossary) of between 2 and 3, meaning that metabolic rate doubles to triples for every 10°C increase in environmental temperature. Temperature–metabolic rate relationships are easily incorporated into additive models (see Glossary) to provide estimates of total energy use over extended periods of time (as seen in Fig. 3), and can yield accurate models of energy use through winter (Roberts et al., 2021). However, the extent to which metabolic rate–temperature relationships are malleable in seasonal environments has only recently started to become clear (Williams et al., 2015, 2012). How thermal sensitivity and/or suppression changes as a consequence of thermal acclimation, stress exposure, diapause stage or life history remains largely unknown. Further, how do metabolic rate–temperature relationships compare between diapause and aestivation?

Energy stores are often positively correlated with fitness for survival, future growth and reproduction. High energy stores at the onset of dormancy result in a high probability of survival through winter (Ellers and Van Alphen, 2002; Rozsypal et al., 2021). However, survival may not be enough to assess the complete role of energy stores in fitness. Post-winter energy stores can directly affect growth and reproductive success in the spring and summer. For instance, following eclosion (see Glossary), the mosquito Culex pipiens allocates a fifth of the lipid stores into egg production (Zhou and Miesfeld, 2009). The link between reproductive fitness and post-winter energy stores is even more evident for overwintering capital breeders, in which the adults emerge with vestigial (non-functional) mouthparts. For example, Irwin and Lee (2003) show that larger females (i.e. with higher presumed energy stores) of the fly Eurosta solidaginis produce more eggs after winter than smaller females. Given the connection between body condition (e.g. body size, mass) and fitness components, a large number of studies often omit measuring any classic fitness traits, and instead use energy stores as the proxy for overwintering success and fitness. This lack of empirical support could limit our ability to detect potential nuances that particular life-history strategies have and defy this key assumption.

Although it is abundantly clear that many insects spend significant energy during their dormancy and that energetics strongly influences their fitness (described above), there are a number of instances where the assumptions outlined in Fig. 2 are apparently not met.

Our first assumption is the most intuitive – that energy use should occur – which we would expect as long as the organism is alive and not undergoing extreme cryptobiosis (e.g. anhydrobiosis; see Glossary). However, there are cases reported where this expectation is not observed for the main energy storage. For instance, lipid stores of diapausing pupae of the butterfly Pieris napi do not decrease during diapause even though they were found to have a respiratory quotient (see Glossary) of 0.7, indicating that lipids are the primary metabolic substrate (Lehmann et al., 2016). Similarly low rates of lipid utilization during winter diapause, but high utilization rates during post-diapause, are found in L. decemlineata (Güney et al., 2021; Lehmann et al., 2020; Piiroinen et al., 2011). As it is unlikely that insects are able to survive without using energy, these contradictory results highlight the two core challenges of studying diapause energy use. First, that energy use is dynamic and it can be difficult to differentiate usage during diapause from usage during post-diapause (i.e. development). This is especially problematic as many studies only quantify starting and ending amounts of energy, without assessing utilization rates during winter. Second, by nature, these studies aim to measure small quantities, in both energy stores and especially in metabolic rates. There is the inherent risk of a high error to signal ratio (Williams et al., 2011), and a risk that individual variation can greatly impact our ability to detect differences between groups. Additionally, diapause energy use has mostly focused on lipid use as a proxy of energy use, overlooking the importance of other substrates as winter fuel (Hahn and Denlinger, 2007, 2011; Short and Hahn, 2023). Not detecting a signal that matches an expectation does not necessarily mean that energy use is not occurring, but rather that the power to detect it is lacking. It is important to consider the sensitivity of the assay not only in the experimental design but also in the interpretation of results.

The relationship between temperature and metabolic rates leads to the expectation that warm temperatures cause higher energetic use, but there are cases where the rate or amount of energy use observed does not match our expectation based on temperature alone. One example is the beetle Cerotoma trifurcata, in which lipids (as a proxy for energy stores) were depleted through dormancy, but depletion occurred at the same rate in a control and an experimentally warmed plot (Berzitis et al., 2017). These results may be largely explained by plasticity in the temperature–metabolic rate relationships of organisms acclimated to different environments. For instance, the skipper Erynnis propertius that spent diapause in variable environments had a reduced thermal sensitivity compared with that of individuals that spent diapause in stable conditions (Williams et al., 2012). A reduced thermal sensitivity in variable or warm environments could potentially reduce the total energetic cost of dormancy, making energetic expenditure between individuals from a stable/cool microclimate and individuals from a variable/warm environment appear similar. There is also conflicting evidence for a change in thermal sensitivity during dormancy, where several species have an increasing thermal sensitivity as winter diapause progresses (Nielsen et al., 2022; Toxopeus et al., 2021), whereas, for other species, metabolic level increases, but thermal sensitivity does not change (Lehmann et al., 2015; Roberts and Williams, 2022). Therefore, temperature must be a critical determinant of winter energy use, but there is still much unexplained variation observed between species and time points in dormancy, and more work is needed to characterize the generalizability of patterns of thermal sensitivity through winter.

Despite most work being focused on winter diapause, comparisons with the energetics in summer diapause (aestivation) may be needed to better assess the assumption that high temperatures lead to higher metabolic rates. Because of the non-linear nature of metabolic rate–temperature relationships, warm temperatures predominantly drive the disproportionately high energy use in variable environments, potentially making aestivating in warm summers even more challenging than cold winters. High winter temperatures can lead to increased mortality (Mech et al., 2018; Nielsen et al., 2022), but whether elevated metabolic rates or energy store depletion drive mortality remains unknown.

There are also overlapping processes occurring during summer and winter diapause, including suppressed metabolic rate and the buildup of energy stores leading up to dormancy (Saulich and Musolin, 2017). Most evidence points toward strong parallels between winter and summer mechanisms driving metabolic suppression, gene expression, arrested development and stress preparation (Llandres et al., 2015; Saulich and Musolin, 2017; Xiao et al., 2013). Reducing the risk of energy depletion at high temperatures is especially important for summer diapause. This can be accomplished by regulating thermal sensitivity through minimal metabolic control, for example. This strategy results in the metabolic rate appearing to be temperature insensitive because the dormant organism's metabolic rate is suppressed to a minimal, life-sustaining metabolic rate that is independent of temperature (Makarieva et al., 2006). The decoupling of temperature and metabolic rate has been observed at ambient temperatures up to 15°C, which is in the lower range of what aestivating insects might experience (Makarieva et al., 2006). While it is possible that minimal metabolic control plays a role in winter diapause as well, the benefit during aestivation should be greater as a result of higher temperatures overall. Alternatively, there may be some costs to decoupling metabolic rate from temperature; for instance, not being able to take advantage of warm spells which are used to repair tissue damage received during cold spells (Koštál et al., 2007) or replenish intermediary metabolite pools (Chen et al., 2021). The involvement of minimal metabolic control in aestivation provides an interesting avenue for further research, and it can only be assumed that decoupling metabolic rates from temperature may also be among the suite of dormancy adaptations.

The wide range of dormancy strategies seen across the vast array of adverse seasonal conditions suggest that dormancy metabolic rate is likely to be optimized for a given life-history context, rather than maximally suppressed. For instance, diapausing pupae of the fly R. pomenella depress their metabolic rate to 90% compared with equal stage non-diapausing individuals (Toxopeus et al., 2021), whereas adults of the butterfly D. plexippus in reproductive diapause retain the ability to fly, thus only suppressing their metabolic rates by ∼10% compared with reproductively active individuals (Chaplin and Wells, 1982). These differing dormancy strategies have resulted in divergent metabolic tactics in these two species. Therefore, given the gradient of metabolic suppression seen among dormant insects, it is necessary to take life-history context into account to accurately model optimal energetic strategies.

The presence of species capable of varying dormancy duration from single to multiple years leads us to question the extent to which energetic stores determine fitness. Although there are extreme examples lasting more than 20 years, this is rarely documented, and it is much more common for prolonged dormancy to last 2–5 years. A rough calculation of energy use of the beetle Chrysomela aeneicollis, using energetic expenditure functions from Roberts et al. (2021), suggests that its energy stores are sufficient only for a single winter diapause. In a single winter diapause, these beetles would use 68% of their pre-winter energy stores, 247% in a 3 month aestivation and 630% in a 2 year prolonged dormancy (Fig. 3). Obviously, these values are not compatible for survival, but are reflective of the more intense metabolic suppression, decrease in thermal sensitivity and/or increase in pre-dormancy energy stores that organisms in prolonged dormancy must exhibit. There are likely to be costs to intense suppression, such as a lack of investment in cold tolerance, so that prolonged dormancy is only favorable in environments with a reduced risk of mortality caused by extrinsic stressors. To better understand diapause and metabolic regulation, following the Krogh principle (see Glossary), it is important to extend our studies into prolonged dormancy (>1 year in duration) because of species’ extreme ability to prevent energy drain. However, given researcher time and funding constraints, it can be difficult to consider species that require multi-year experiments as meeting the core tenet of the Krogh principle ‘animals that are convenient to study’.

As impressive as meeting the energetic requirements of prolonged dormancy is, what is perhaps even more astounding are cases where – following a 2–5 year prolonged dormancy – individuals of the same species have higher fecundity than individuals of the same species that diapause for a single winter (Matsuo, 2006; Wang et al., 2006; Wei et al., 2010). Although on the surface this seems to violate our third assumption, that energy stores are linked to fitness, there is likely to be a straightforward answer to this apparent discrepancy. Increased fecundity is probably a result of larger individuals being more likely to survive a prolonged dormancy, and larger individuals having higher fecundity (Short and Hahn, 2023; Wei et al., 2010). For instance, relationships between dormancy duration and size have been found in the bee Perdita portalis (Danforth, 1999) and suggested in the moth Thaumetopoea pityocampa (Battisti et al., 2000). These findings are the most compelling challenge to the assumption that energy use in winter is linked to fitness and, given the simple explanation, there is only weak evidence against this assumption.

Violations of the proposed assumptions provide an interesting chance to reflect on the nature of diapause energetic research. Some insects are known to survive aestivation during the warm conditions of the summer months, insects in prolonged dormancy can survive for decades without feeding, and there are cases where lipid stores are not depleted during months of overwintering. Given the large body of work that supports the assumptions discussed in this Commentary, the existence of several cases that break them leaves us with a chance to appreciate the diversity of energetic strategies and capabilities of insects. Making broad generalizable rules about energy use that cover all insects seems impractical, but expanding the number of species that are studied may allow researchers to identify key evolutionary patterns or life-history traits that are linked to energetic patterns.

We propose several areas that we believe will further improve our understanding of diapause energetics (Fig. 4). First, it is critical to expand our understanding of the energetic costs of investing in stress tolerance (Rozsypal et al., 2021). By identifying the energy sinks during dormancy, we can better understand how life-history strategies in different environments use energy stores and their effect on fitness. For instance, mounting a sufficient level of cold tolerance is a very fitness-relevant trait for many insects in winter, but what is its energetic cost? Can accumulated cryoprotectants be used as an energy pool after the coldest periods subside? Further, what is the energetic cost of recovery from cold stress? Štětina et al. (2018) used a combination of respirometry and metabolomics that can be a useful integrative approach to assess the energetic costs of recovery from different types of cold stress. Similarly, how do multiple stressors impact energetic investment required for stress tolerance and recovery? For instance, during adverse seasons, dormant insects can face both desiccation and temperature stress. Even though stress responses may act synergistically between stressors (i.e. cross-tolerance), the energetic costs of mounting the stress response remains an open and challenging field (Sinclair et al., 2013).

Second, we need to expand our understanding of how energy pools are moved around and how dynamic their interconversion is. There is extensive work on how energy pools are used for some species, and some evidence of interconversion, but the extent is poorly understood (Colinet et al., 2012; Michaud and Denlinger, 2007; Storey and Storey, 1990). Substrate interconversion rate could be approached using isotope tracing studies through the winter to track the fate of various energy pools, as shown in Zhou and Miesfeld (2009).

Third, many of the studies of dormancy energetics stop at post-dormancy energy quantification and correlate this with fitness. However, the extent to which the positive correlation between fitness and post-dormancy energy stores holds in insects is likely to be variable, and may not be as generalizable as previously thought (Moraiti et al., 2012; Wei et al., 2010; Zhao et al., 2022). While there are few studies that link lower reproductive output with a prolonged dormancy (Moraiti et al., 2023), more studies are needed in order to better understand the cross-seasonal fitness impacts of dormancy duration. Including post-winter performance relevant to a life-history context will improve our understanding of the links between energy stores and fitness.

Fourth, pre-dormancy energy stores can be extremely variable within a species, and they are critical for determining survival. Focusing on how intraspecific variation in pre-dormancy energy stores impacts fitness remains a critical question. This is especially true across life-history contexts such as capital breeding and income breeding, which differ greatly in energy use strategies. Understanding variation in energy stores at the onset of dormancy among individuals within and between populations is of particular interest given the rapid environmental changes due to climate change. In particular, increasing extreme weather patterns may directly impact both energy store accumulation and use prior to, during and after dormancy.

Fifth, expanding the number of species with extensive, readily available data on energy use during dormancy across ecological contrasts will allow us to gain a more complete picture of how insects regulate energy use in dormancy. Increasing our understanding of how thermal sensitivity and metabolic suppression change between insects that are in aestivation/diapause and those that are in dormancy/prolonged dormancy will allow researchers to identify how energetic regulation differs between these processes. Further, a more comprehensive comparative dataset will allow us to better understand the evolution of dormancy strategies and associated energetic regulation.

As Journal of Experimental Biology celebrates its centenary, we are faced with the most rapidly changing environment in our species' history and understanding the biotic responses to these changes should be at the apex of our effort as biologists. Identifying energy pools, energy use rates and the fitness impacts of energy use and retention for a broad range of species will allow us to identify patterns associated with broad life-history characteristics. Once life-history-based patterns are identified, we can begin to build landscape-level energetic models (see Glossary) that will help identify when and where dormancy energy use becomes a critical range-limiting factor, and which groups of diapausing insects will be most vulnerable to changing environments.

Energy use during dormancy is undoubtedly an important fitness-associated trait and, as such, maximizing available energy stores after dormancy is likely to be a key fitness-increasing adaptation, either through intrinsic means (e.g. accumulating larger energy stores, further suppressing metabolism, decreasing thermal sensitivity) or through extrinsic means (e.g. selecting more favorable microhabitats). In this light, ongoing climate change is poised to severely impact energy flux during dormancy and could lead to large fitness consequences in vulnerable species and populations. However, predicting which species are most vulnerable is non-trivial because a large number of species (e.g. insects that undergo aestivation or prolonged dormancy) seemingly do not suffer expected energetic consequences of their energetically unfavorable dormancies (either in very warm environments or for very long time spans). Thus, metabolic rate is likely to be a complex trait optimized to be sufficient within certain dormancy conditions, but it may trade off against other traits important to fitness, such as stress tolerance, growth or reproduction. As suggested by Hahn and Denlinger (2007) in their review, treating dormancy energy loss as a trait to be optimized, rather than just minimized, could be a useful way forward. While this thinking has long been appreciated in the hibernation physiology of mammals, where metabolic suppression shows larger variability, with different levels in different strategies (Humphries et al., 2003; Staples, 2014; Storey and Storey, 1990), this thinking is still underappreciated in the field of insect dormancy energetics (but see Chen et al., 2021, for an important exception).

To summarize, insect dormancy strategies are highly variable and diverse – even among closely related species – indicating high degrees of evolvability in this key life-cycle adaptation. A future task for researchers will be to deconstruct this variability and to try to establish what overarching rules or limitations exist for energetics within the broader framework of dormancy evolution. These are critical challenges in the light of ongoing climate change as, at present, we are not able to satisfactorily predict the energetic consequences of changing conditions across the myriad of dormancy strategies that occur in the class Insecta. These constitute important challenges to experimental biologists and ecophysiologists for the coming century.

We would like to thank Lisa Treidel, Andrea Battisti, Martin Schebeck, Daniel Hahn and two anonymous reviewers for their helpful comments that improved the paper. We would also like to thank Mikaela Fredrikson for beetle art.