Skip to main content
NCBI home page
Integrative and Comparative Biology logoLink to Integrative and Comparative Biology
. 2016 Apr 21;56(1):62–72. doi: 10.1093/icb/icw009

Adaptation to Low Temperature Exposure Increases Metabolic Rates Independently of Growth Rates

Caroline M Williams *,1, Andre Szejner-Sigal †, Theodore J Morgan ‡, Arthur S Edison §, David B Allison , Daniel A Hahn †
PMCID: PMC4930064  PMID: 27103615

Abstract

Metabolic cold adaptation is a pattern where ectotherms from cold, high-latitude, or -altitude habitats have higher metabolic rates than ectotherms from warmer habitats. When found, metabolic cold adaptation is often attributed to countergradient selection, wherein short, cool growing seasons select for a compensatory increase in growth rates and development times of ectotherms. Yet, ectotherms in high-latitude and -altitude environments face many challenges in addition to thermal and time constraints on lifecycles. In addition to short, cool growing seasons, high-latitude and - altitude environments are characterized by regular exposure to extreme low temperatures, which cause ectotherms to enter a transient state of immobility termed chill coma. The ability to resume activity quickly after chill coma increases with latitude and altitude in patterns consistent with local adaptation to cold conditions. We show that artificial selection for fast and slow chill coma recovery among lines of the fly Drosophila melanogaster also affects rates of respiratory metabolism. Cold-hardy fly lines, with fast recovery from chill coma, had higher respiratory metabolic rates than control lines, with cold-susceptible slow-recovering lines having the lowest metabolic rates. Fast chill coma recovery was also associated with higher respiratory metabolism in a set of lines derived from a natural population. Although their metabolic rates were higher than control lines, fast-recovering cold-hardy lines did not have faster growth rates or development times than control lines. This suggests that raised metabolic rates in high-latitude and -altitude species may be driven by adaptation to extreme low temperatures, illustrating the importance of moving “Beyond the Mean”.

Introduction

Metabolic cold adaptation describes a macrophysiological pattern whereby respiratory estimates of metabolic rates of ectotherms from cold environments, typically from high latitudes or altitudes, tend to be elevated relative to those from warm environments (Addo-Bediako et al. 2002). Metabolic cold adaptation has been an important and hotly debated topic in ecological and evolutionary physiology for nearly a century (Fox 1936; Bullock 1955; Clarke 1993). Although there are numerous examples of clinal variation in metabolic rate both within and across species that are consistent with metabolic cold adaptation, particularly in terrestrial arthropods and fish (Wohlschlag 1960; Block and Young 1978; Chappell 1983; Torres and Somero 1988;Ayres and Scriber 1994; Addo-Bediako et al. 2002; Terblanche et al. 2009; White et al. 2012), other studies find no association between habitat temperature and metabolic rates (Clarke 1993; Clarke and Johnston 1999; Steffensen 2002; Lardies et al. 2004). Our purpose in this manuscript is not to assess whether metabolic cold adaptation is a general macrophysiological rule. Although it is not always found, the pattern is general enough that it merits consideration. Rather, our goal is to discuss some of the ultimate, selective mechanisms that may drive metabolic cold adaptation in the context of habitats with high thermal variation, such as those at high latitudes and altitudes (Sunday et al. 2011).

When a pattern consistent with metabolic cold adaptation is found, it is most often ascribed to another macrophysiological rule, countergradient variation (Conover et al. 2009). Countergradient variation describes a pattern of local adaptation in organismal life histories wherein geographic variation in genotypes counteracts environmental influences, reducing phenotypic variation along an environmental gradient (Levins 1968; Conover and Schultz 1995). In the context of metabolic cold adaptation, this implies that body size and generation time are preserved along latitudinal or altitudinal thermal gradients, through elevation of rates of growth and development to counter the slowing effects of cool temperatures and short seasons. Countergradient variation is well-enough supported by both intraspecific and interspecific comparisons in terrestrial arthropods, fish, and amphibians to be considered a general macrophysiological rule (Schultz et al. 1996; Gotthard et al. 2000; Laugen et al. 2003; Gaston et al. 2009).

Both within and between species, fast growth rates are often correlated with higher rates of respiratory metabolism, presumably due to greater rates of both anabolic and catabolic intermediary metabolism needed to support fast growth (Arendt 1997; Metcalfe and Monaghan 2001; Stoks et al. 2006; Glazier 2015). Because of this correlation between metabolic rates and growth rates, latitudinal/altitudinal local adaptation studies that show a pattern consistent with metabolic cold adaptation tend to attribute that pattern to countergradient selection on growth rates and development time (Ayres and Scriber 1994; Conover and Schultz 1995; Addo-Bediako et al. 2002; White et al. 2012). However, local adaptation to high latitude and altitude sites with high seasonal and thermal variation could entail selection on the ability to be hardy to many stresses beyond time limitation for growth and reproduction. Organisms must frequently balance multiple abiotic and biotic challenges (Sokolova and Pörtner 2007; Todgham and Stillman 2013); thus, we expect that local adaptation at high-latitudes and -altitudes will be the product of multifarious selection acting to mitigate both short, cool growing seasons and stressful thermal extremes; necessitating that our conceptual framework for metabolic cold adaptation “moves beyond the mean” to consider the ecological, evolutionary, and mechanistic consequences of thermal variability.

When challenged by temperatures below their critical thermal minimum for movement, many ectotherms lose neuromuscular coordination and enter into a state termed chill coma (David et al. 2003; MacMillan and Sinclair 2011). If the duration of low temperature is sufficiently brief (usually less than 12 h), organisms will typically recover fully coordinated movement shortly after returning to warmer temperatures and ultimately continue critical processes like growth and reproduction (MacMillan and Sinclair 2011). The time taken to recover neuromuscular coordination and resume activity after rewarming is termed the chill coma recovery time. Among ectotherms, chill coma and recovery from chill coma have been best studied in terrestrial insects. Chill coma recovery time is ecologically relevant because the time spent recovering from chill coma represents lost opportunities for foraging, dispersal, mating, and reproduction; and increased vulnerability to predation, parasitism, and environmental stress (due to loss of ability to behaviorally avoid inhospitable conditions) (David et al. 2003; MacMillan and Sinclair 2011). Reinforcing this view, chill coma recovery time shows clear patterns of local adaptation in many terrestrial arthropods wherein populations or species from high-latitudes and -altitudes have faster chill coma recovery time than those from more thermally stable environments (David et al. 2003; Sinclair et al. 2012). The rationale for this pattern is that insects living in high-latitude and -altitude sites experience greater thermal variability and may frequently enter into chill coma when temperatures fall below critical limits for activity either overnight or as cold fronts sweep in, and then recover from chill coma as temperatures warm again during daytime or as cold fronts dissipate.

A series of quantitative trait locus, experimental evolution, and genome-wide association studies have shown that chill coma recovery is a strongly heritable quantitative trait underlain by many genomic regions, each with a relatively small effect (Morgan and MacKay 2006; Norry et al. 2007; MacKay et al. 2012; Williams et al. 2014). From a physiological perspective, insects lose ionic and osmotic homeostasis during chill coma and whole-organism recovery of coordinated movement is correlated with the restoration of ionic and osmotic gradients (MacMillan and Sinclair 2011; MacMillan et al. 2012; Andersen et al. 2013; Findsen et al. 2013; MacMillan et al. 2015). Reestablishing ionic and osmotic gradients, and eventually restoration of coordinated movement, is energetically intensive requiring both the availability of ATP and the ability of downstream intermediary metabolic processes to use that ATP to recover homeostasis. Several studies have shown that cold exposure disrupts intermediary metabolism (Lalouette et al. 2007; Michaud and Denlinger 2007; Koštál et al. 2011) and that low-temperature acclimation can reduce the magnitude of cold-induced perturbations to intermediary metabolism (Overgaard et al. 2007; Colinet et al. 2012). We have also recently shown that cold adaptation via artificial selection for fast chill coma recovery time in the fly Drosophila melanogaster increased the robustness of biochemical networks of intermediary metabolites, permitting enhanced maintenance of metabolic processes during cold exposure and recovery (Williams et al. 2014). Given that cold disrupts intermediary metabolism, and that plastic and evolutionary responses to cold enhance the ability to maintain metabolic processes in the cold, could selection for cold-hardiness in highly thermally variable habitats drive the evolution of respiratory metabolic rate in a pattern that is consistent with metabolic cold adaptation, independent of selection on growth rates and development time?

Here, we tested whether fast recovery from chill coma was associated with higher respiratory metabolic rates across two complementary sets of lines of the fly D. melanogaster, both derived from a mid-latitude site in Raleigh, NC, USA, that experiences substantial seasonal thermal variation. The first set was a series of isogenic lines that represent naturally segregating variation in chill coma recovery time (D. melanogaster Genetic Reference Panel; MacKay et al. 2012). The second was a set of replicated experimental evolution lines artificially selected for fast or slow chill coma recovery time, as well as unselected control lines. We tested whether chill-coma recovery time was associated with a shift in the relationship between temperature and metabolic rate by testing lines across a series of temperatures from very low (0°C) to relatively high (25°C). Across both sets of lines, fast chill coma recovery was associated with higher respiratory metabolism at temperatures above 15°C. Using the artificial evolution lines, we further tested whether lines selected for fast chill coma recovery time that had higher respiratory metabolic rates, also had higher growth rates and shorter development times. Although lines selected for fast chill coma recovery had higher respiratory metabolism than either control or slow-recovering lines, cold-hardy fast-recovering lines did not have higher growth rates or development times than control lines. Because chill coma recovery time is a cold-hardiness trait expected to be a target of selection in habitats with high thermal variation, we propose that selection on cold-hardiness could drive patterns of respiratory metabolism that are consistent with metabolic cold adaptation, independent of countergradient variation in life history traits.

Methods

Fly stocks

We used two complementary genetic resources, both originating from a natural population in Raleigh, NC, USA. The first was the D. melanogaster Genetic Reference Panel (MacKay et al. 2012); a fully genotyped panel of inbred lines representing genetic variation segregating within the population at the time of founding the lines. We chose six lines from either tail of the distribution of chill coma recovery times, representing naturally occurring combinations of alleles associated with extreme cold-hardiness and -susceptibility. The second genetic resource was a set of replicate experimental evolution lines derived from the same base population, and selected in the laboratory for fast versus slow recovery from chill coma (Williams et al. 2014). The selection process generated substantial and genetically fixed differences in chill coma recovery times, with hardy lines recovering after an average of 6.1 and 5.8 min (replicate lines 1 and 2), compared with 12.4 and 23.7 min for susceptible flies (Williams et al. 2014). All flies were reared on standard cornmeal–agar–molasses medium under controlled density as previously described (Williams et al. 2014).

Respirometry

For respirometry experiments, >1-day-old female flies were sorted into groups of 20 under light CO2 anesthesia and then left to recover for 2–4 days. All experiments were performed on 5- to 8-day-old mated females, reared at 25°C. We used a Sable Systems International (SSI) respirometry system (Las Vegas, NV, USA) with an Oxzilla II oxygen (O2) analyzer (SSI) and a LiCor 7000 carbon dioxide (CO2) analyzer (Lincoln, NE, USA). We collected data using a UI2 interface (SSI) at a frequency of 1 Hz. Incurrent air was scrubbed of water vapor and CO2 using a drierite–ascarite–drierite column, excurrent gas was scrubbed of water vapor before entering the CO2 analyzer using a magnesium perchlorate column, and CO2 and water vapor were scrubbed before the O2 analyzer using an ascarite–magnesium perchlorate column. We corrected data to a reading taken through an empty baseline chamber at the beginning and end of each recording to correct for instrument drift.

Thermal performance curves for metabolic rate

We used stop-flow respirometry to measure VCO2 and VO2 as estimates of metabolic rate in groups of 10 female flies at 0°C, 15°C, 20°C, and 25°C, with each group of flies measured only once at one temperature (Supplementary Material). Respiratory exchange ratios (RERs) were calculated as the ratio of VCO2:VO2. Differences in activity levels among lines could confound metabolic rate estimates, so we estimated minimum costs of transport for each line in separate experiments (Supplementary Material), and subtracted these costs from metabolic rate estimates to ensure that activity was not driving the patterns we saw. We did not measure activity at 0°C because all flies are in chill coma (inactive) within a few minutes of exposure to 0°C. After each measurement, groups of flies were frozen and weighed to 0.01 µg using a microbalance.

All statistical analyses were performed in R 3.2.2 (R Core Team 2015). Preliminary data exploration was performed as recommended by Zuur et al. (2010). VO2 was log10-transformed to improve normality. We fit general linear mixed models describing VO2 as a function of temperature, natural logarithm of temperature, and temperature as a second or third order polynomial, and compared the fit using AIC (where ΔAIC > 2 is justification for preferring a more complex model). After the most parsimonious form of temperature was determined, we added random effects of syringe, run, and replicate population nested within cold hardiness, plus nested permutations of these random effects. We ascertained which combination of random effects was most parsimonious using AIC as above. We then fit additional fixed effects of cold hardiness and the interaction between selection and temperature (assessing whether the thermal sensitivity of metabolic rate differed among lines), with mass as a covariate. We simplified the model (including pooling factor levels) using AIC as above to determine the minimal adequate model (Crawley 2007).

Metabolic rate of individual flies during cold exposure and recovery

We used open-flow respirometry to continuously measure VCO2 as an estimate of metabolic rates in individual D. melanogaster during cold exposure and recovery (Supplementary Fig. S1). It is not possible to measure VO2 in individual flies due to the relatively low sensitivity of oxygen compared with carbon dioxide analyzers. From these data, we estimated metabolic rates at the following time points: (1) Before cold exposure (25°C); (2) during cold exposure (0°C); (3) rewarming, the time taken for the temperature to completely equilibrate to 25°C (∼20 min); (4) early recovery, the first 1.5 h after temperature had equilibrated; and (5) late recovery, the final 1.5 h of 4 h recovery. These time points are indicated on Supplementary Fig. S1 (1–5 along top axis). For each time period, metabolic rates for each individual were averaged and analyzed using mixed general linear models as described above (Thermal performance curves for metabolic rate), except that size (thorax length; Supplementary Material) was measured using a dissecting microscope with eye piece reticule and used as a covariate instead of mass. To characterize variability in metabolic rates over time, we calculated rolling standard deviations on 1500 s windows for each individual (zoo package; Zeileis and Grothendieck 2005). The size of the window is shown in the black box in Supplementary Fig. S1. These rolling standard deviations were averaged over the early and late recovery time points as previously described, and analyzed using the same linear model framework.

Growth rates and development time

Eggs were collected from population cages of each line using grape juice agar plates with a thin coat of liquid yeast. Plates were left for 2 h in the cages and then immediately submerged in 70% ethanol to stop egg development and stored in a 4°C chamber. Groups of 100 eggs were dried and weighed (Cahn C-35 microbalance [±1 μg], Orion Research, Inc., Boston, MA, USA) to obtain the average egg dry mass for each line, which was used as a starting mass to calculate growth rates. Groups of 30 eggs from each line of flies (collected from the growth medium) were each placed into five replicate vials for each temperature treatment, and were checked for pupation or adult emergence at 5 pm each day. Date of pupation or adult emergence was recorded and used to calculate pupal and adult development time. Adult flies were collected the day they emerged and frozen at −20°C. After all flies had emerged, they were dried individually at 60°C for 48 h and a random subset of the population of six females per replicate was weighed. The use of the random subset was required due to the large population size of the experiment. Growth rates were obtained for each line by calculating the difference between adult and egg dry mass divided by the total development time. Growth rate data were analyzed using general linear mixed models as described above (Thermal performance curves for metabolic rate), with starting mass of eggs as a covariate.

Results

Thermal performance curves for metabolic rate

Cold-hardy and cold-susceptible flies did not differ in mass when reared at 25°C (normal rearing conditions for all flies used in respirometry experiments) (P > 0.1, Fig. 1A,B). In flies from all lines, VO2 increased with temperature (F2,45 = 371.7, P < 0.0001), with the relationship best described by a second degree polynomial (Fig. 1C,D). In the experimental evolution lines, VO2 was more thermally sensitive in cold-hardy flies, with higher VO2 at warm temperatures but lower VO2 at 0°C in cold-hardy compared with -susceptible flies, and control flies were intermediate (hardiness × temperature: F4,98 = 10.6, P < 0.0001, Fig. 1C). At 0°C, metabolic rates were lowest in cold-hardy flies and highest in -susceptible, with controls intermediate (Fig. 1E). The mass-scaling of metabolic rate varied as a function of temperature: respiration rates increased with mass at warm temperatures, but showed progressively shallower scaling exponents as temperature decreased, being completely mass-independent at 0°C (mass × temperature: F2,98 = 21.4, P < 0.0001; Supplementary Fig. S2A). We replicated these patterns using lines of flies originating from the DGRP: hardy flies had higher thermal sensitivity of VO2, with higher VO2 than susceptible lines at warm temperatures, but both hardy and susceptible lines had similar VO2 in the cold (hardiness × temperature: F2,136 = 3.0, P = 0.05, Fig. 1D,F). Mass scaling exponents for metabolic rate again decreased as a function of temperature, parallel to the pattern seen in the experimental evolution lines (mass × temperature: F2,136 = 5.5, P = 0.005; Supplementary Fig. S2B). RERs were extremely variable at 0°C, and ranged from 0.3 to 1.3 (Supplementary Fig. S3A). Across the warmer temperatures (15–25°C), RERs decreased with increasing temperature (Experimental Evolution lines: F1,19 = 7.9, P = 0.01, Supplementary Fig. S3B; DGRP: F1,35 = 12.1, P = 0.001, Supplementary Fig. S3C), with values centered around 1 at 25°C but increasing to a mean of 1.1 by 15°C. For the Experimental Evolution lines, hardy flies had lower RERs than control and susceptible (which were pooled in the best model; F1,82 = 8.9, P = 0.004), while by contrast in the DGRP hardy flies had higher RER than susceptible flies (F1,108 = 6.9, P = 0.001).

Fig. 1.

Fig. 1

Open in a new tab

Size and metabolic rates of female Drosophila melanogaster from Experimental Evolution lines (A, C, E) or the Drosophila melanogaster Genetic Reference Panel (DGRP; B, D, F). For all panels, N = 5–7/line/temperature, with six (DGRP) or two (Experimental Evolution) replicate lines of flies for each level of cold hardiness. (A, B) Boxplots showing mass of lines that vary in cold hardiness. (C, D) Oxygen consumption of cold-hardy (black triangles), control (dark gray circles), or cold-susceptible (light gray squares) female flies. Dotted lines indicate remaining oxygen consumption when the estimated cost of walking was removed. Values are mean ± SEM. (E, F) Magnification of metabolic rates at 0 °C. Significant effects are from general linear models described in results, main effects are not given where there is a significant higher-order interaction term.

To rule out differential activity as a potential driver of the observed differences in metabolic rate, we estimated the distance walked by each line of flies during a typical respirometry recording, and used these data to calculate the average metabolic costs of transport (Supplementary Material). In the experimental evolution lines, hardy flies had similar costs of transport to control flies, while susceptible flies had reduced costs of transport (mostly due to reduced activity levels, Supplementary Fig. S4). In the DGRP, hardy and susceptible flies had similar costs of transport at 20–25°C, but lower costs of transport at 15°C. Thus, in neither set of lines was increased activity of hardy lines driving the increased metabolic rates of cold hardy flies. To confirm this, we subtracted the estimated costs of transport from measured metabolic rates to estimate metabolic costs that were independent of activity (dotted lines in Fig. 1C,D), and reanalyzed these activity-corrected data. Analysis of activity-corrected VO2, or raw or activity-corrected VCO2, gave identical conclusions for all analyses, confirming that the pattern of higher metabolic rates in cold hardy flies was not driven by higher activity levels in these lines (Supplementary Fig. S4).

Metabolic rate during cold exposure and recovery

Metabolic rates of all flies decreased abruptly during cold exposure, then gradually recovered back toward pre-cold rates, but in most cases did not fully recover to before cold levels (e.g. Supplementary Fig. S1). Hardy flies had a distinct trajectory of metabolic rates during cold exposure and recovery compared with susceptible flies: we replicated our finding of higher metabolic rates before cold, dropping to lower rates during cold, and recovering to higher levels than did susceptible flies during recovery (cold hardiness × time: F8,101 = 2.3, P = 0.033; Fig. 2A). Standardizing metabolic rates of each individual to the average “Before cold” metabolic rate for that individual confirmed that hardy flies dropped their metabolic rates to a greater degree during cold exposure, and additionally revealed that although absolute metabolic rates were higher in hardy than susceptible flies during recovery, the metabolic rates of hardy flies remained suppressed relative to their “before cold” rates during recovery, while susceptible flies returned to close to their before cold rates by the late recovery phase (cold hardiness × time: F6,80 = 3.0, P = 0.013; Fig. 2B). Many of the flies showed a distinctly cyclic pattern of respiration that started during either the early or late recovery phases (Supplementary Fig. S5), and the onset of this pattern was independent of the onset of activity (e.g. see Supplementary Fig. S1). There was no significant difference between hardy and susceptible flies in the variance of VCO2 during the early or late recovery period (F2,40 = 3.3, P = 0.095), although it is possible that we lacked power to detect an effect were it there.

Fig. 2.

Fig. 2

Open in a new tab

CO2 production (as a proxy for metabolic rate) of cold-hardy (white), control (light gray), or cold-susceptible (dark gray) or individual female D. melanogaster during cold exposure and recovery. (A) CO2 production rates and (B) change in CO2 production rates compared with the same individual before cold exposure. All time points are relative to a 3-h cold exposure at 0 °C; Early = first 2 h of recovery period, Late = subsequent 2 h of recovery period (see Supplementary Fig. SE for time periods). N = 8 for each box (two replicate selection lines pooled for display purposes, but accounted for in models, see text). Significant effects are from general linear models described in results, main effects are not given where there is a significant higher-order interaction term.

Growth rates and development times

Growth rates increased with increasing temperature, and were similar among lines at the normal laboratory rearing temperature (25°C; Fig. 3). On either side of that temperature, chill-susceptible flies had lower growth rates than hardy and control flies, which were statistically indistinguishable (rearing temperature × cold hardiness [hardy and control pooled]: F3,678 = 8.6, P < 0.0001, Fig. 3). The slower growth rates of susceptible flies resulted from longer development times, while development time did not differ between hardy and control lines (Supplementary Fig. S6).

Fig. 3.

Fig. 3

Open in a new tab

Growth rates of cold-hardy (black), -susceptible (light gray), and control (dark gray) female flies, from an experimental evolution experiment. N ≥ 5 replicate vials of 30 eggs/vial/line, resulting in ≥30 surviving adults for each line. Values are mean ± SEM of two replicate experimental evolution lines for each level of cold hardiness (accounted for in models as a random effect). Significant effects are from general linear models described in results, main effects are not given where there is a significant higher-order interaction term.

Discussion

Here, we show that lines of D. melanogaster flies with faster recovery from chill coma, a common metric of cold hardiness, also have higher respiratory metabolism at warm temperatures permissive for activity (15–25°C). We demonstrated this pattern in both a set of experimental evolution lines artificially selected for fast or slow recovery from chill coma (compared with unselected controls), and also a series of lines derived from the D. melanogaster Genetic Reference Panel. Both sets of lines were originally derived from a mid-latitude population that represent naturally segregating genetic variation for many traits, including cold hardiness (MacKay et al. 2012). Chill coma recovery time shows clear patterns of local adaptation across terrestrial arthropod species, wherein populations from high latitude and high altitude sites show faster recovery from chill coma than populations from less cold and thermally variable habitats (David et al. 2003; Sinclair et al. 2012). Therefore, selection on the ability to recover activity after cold knockdown at high latitudes and altitudes may drive higher respiratory metabolism and thus contribute to the pattern of metabolic cold adaptation in terrestrial arthropods.

Interestingly, while cold-hardy, fast-recovering lines had higher respiratory metabolic rates than control and susceptible lines at warm temperatures permissive for fly activity (15–25°C), respiratory metabolism dropped to a greater degree upon cold exposure in cold-hardy compared with control or susceptible lines. Artificial selection for cold-hardiness via fast chill coma recovery thus altered the relationship between temperature and metabolic rate, wherein metabolic rates increased more quickly with temperature in hardy lines than control lines, followed by cold-susceptible lines. Similarly, fast-recovering, cold-hardy lines from the Drosophila Genetic Reference Panel also had a higher thermal sensitivity of respiratory metabolism than cold-susceptible lines. Although most studies of metabolic cold adaptation in insects only assess metabolism at one or two shared temperatures (Addo-Bediako et al. 2002), our observations are consistent with studies that have found intraspecific and interspecific variation in metabolic rate–temperature relationships associated with living in cold habitats (Chappell 1983; Terblanche et al. 2009). Our work differs from other studies on metabolic rate–temperature relationships because we include both metabolic rates at relatively high temperatures where flies can be active (15–25°C) and a stressful low temperature that is below the threshold for activity (0°C). Despite the difficulty of quantifying metabolic rates of small insects at low temperatures, we included this measure at 0°C specifically because we think it critical to understand how selection during the cold exposure may act on metabolic rate.

When considering patterns of respiratory metabolism during adaptation to cold stress, selection could alter metabolism at three different points: before cold exposure so that animals have been selected to be better prepared to resist cold stress, during cold exposure, and during recovery from cold so that animals are better able to tolerate perturbation from cold stress (Williams et al. 2014). By following respiratory metabolism across each of these time points, we showed that cold-hardy lines have greater metabolic plasticity than control or susceptible lines. Specifically, cold-hardy lines have higher respiratory metabolism before the cold but they have lower respiratory metabolism at 0°C than control or susceptible lines, suggesting greater metabolic plasticity in response to cold stress. During recovery from cold stress, the respiratory metabolic rates of cold-hardy lines increased again faster than control or susceptible lines, but even after 4 h of recovery hardy lines did not completely recover to their pre-cold respiratory metabolic rates whereas control and susceptible lines did return to near their pre-stress metabolic rates. This suggests that metabolic suppression was present during recovery from cold in hardy flies. This is in line with previous findings that plastic responses to cold exposure do not necessarily increase metabolic costs in chill-susceptible insects (Basson et al. 2012; but see MacMillan et al. 2012).

The potential benefits of increased metabolic plasticity during adaptation to cold stress become clear when considering the mechanisms that may underlie chill coma and recovery responses. During chill coma insects accumulate osmotic, ionic, and metabolic imbalances that increase in severity with duration spent in chill coma (MacMillan and Sinclair 2011; Andersen et al. 2013; Findsen et al. 2013; MacMillan et al. 2015). Metabolic, ionic, and osmotic homeostasis must be reestablished during recovery from chill coma, along with repair of other types of cellular damage that may accrue with time at low temperature. Reestablishing homeostasis and repairing cold-induced damage is energetically expensive from both the perspectives of energetic currency to do work and anabolic substrates to effect repairs (MacMillan et al. 2012). Given that respiratory rates reflect rates of intermediary metabolism, in this context higher metabolic rates before and after cold stress may support greater substrate flux through critical pathways of intermediary metabolism at warm temperatures to help individuals better resist cold-induced perturbations to metabolism and to recover metabolic homeostasis and repair damage more quickly after cold perturbation. Reduction of intermediary metabolism, and particularly aerobic catabolism, is a common response to many stressors beyond cold (Storey and Storey 1990; Guppy and Withers 1999). The ability of cold-hardy lines to reduce intermediary metabolism during cold stress to a greater degree than control or susceptible lines may help to reduce the magnitude of metabolic perturbation during cold exposure, thus leaving cold hardy flies better prepared to recover homeostasis. Indeed, allied work on these same experimental evolution lines has found that cold-hardy lines suffer less cold-induced perturbations in their levels of metabolites than susceptible lines (Williams et al. 2014), and that this is associated with higher rates of catabolism and anabolism (C. M. Williams, M. D. McCue, N. Sunny, T. J. Morgan, D. B. Allison, and D. A. Hahn, submitted for publication). Unfortunately, the literature on metabolic cold adaptation in terrestrial arthropods is currently lacking in studies that include metabolic responses at all three times (before, during, and after cold stress), although several studies measure metabolic rates either during or after cold stress (Stevens et al. 2010; Basson et al. 2012). We do not know whether the pattern of increased metabolic plasticity we observed here is a general aspect of metabolic cold adaptation, but we encourage other authors to include time series data of respiratory metabolism across a thermal perturbation when considering metabolic adaptation to cold stress.

RERs were close to 1 at room temperature, suggesting that flies were primarily relying on carbohydrate metabolism (Lighton 2008). RERs increased with decreasing temperature in both sets of lines, which may indicate an increasing predominance of anabolic relative to catabolic processes as temperatures cool (Lighton 2008). At low temperatures, RERs were extremely variable. For these experiments we used stop-flow respirometry, which allows gas to build up and then be injected into the system in a bolus, thus low oxygen concentrations are not a possible explanation for this variability at 0°C. Inspection of VCO2 and VO2 at 0°C shows similar levels of variation, further supporting that oxygen measurements had sufficient resolution. It is thus possible that the highly variable RERs at 0°C reflect true biological differences in the response to low temperatures, but this requires further investigation. Differences in RER between hardy and susceptible flies were inconsistent across the Experimental Evolution lines and the DGRP, so we conclude that RER does not seem to be related to cold hardiness in these lines.

One of the caveats for using experimental evolution approaches in the laboratory to study mechanisms of adaptation in the field is that one may inadvertently artificially select for a trait that they did not intend (Gibbs 1999). Is it possible that we artificially selected for high levels of activity that are driving the observed patterns of respiratory metabolism rather than cold tolerance? We quantified activity, distance walked, and walking speed in flies from each of the experimental evolution lines and the naturally derived lines and used these data to estimate costs of transport (Supplementary Material). These data rule out increased activity levels as a driver of increased metabolic rates in hardy flies. Among our experimental evolution lines, the cold-hardy lines did not differ from the control lines in costs of transport, although the susceptible lines did have reduced costs of transport due to lower activity level (Supplementary Fig. S3). Thus, the higher respiratory metabolic rates we observed in hardy versus control lines at temperatures permissive for movement were not driven by activity. In the DGRP, hardy flies had lower activity at cool temperatures; the opposite to the pattern we saw for metabolic rate. We are confident that our results represent a specific response of increased respiratory metabolism to selection for hardiness to cold stress that is consistent with patterns of metabolic cold adaptation.

Abundant nutrient availability in laboratory selection studies can also ameliorate life history trade-offs that may prevent the evolution of phenotypes in nature—therefore in the wild, hardy flies with increased metabolic demands may experience decreased ability to invest in other fitness components such as reproduction or somatic maintenance (Zera and Harshman 2001). In this sense, laboratory selection experiments indicate what can happen in response to a selective pressure, but not necessarily what will happen in the wild (Gibbs 1999). We did not assay other measures of fitness such as fecundity or viability—it is entirely possible that these fitness measures may be reduced in cold hardy flies, indicating a functional trade-off between cold hardiness and reproduction. Alternately, the cost may come in the form of increased nutrient requirements, requiring longer foraging periods or increased efficiency of digestion, absorption, or catabolism.

Because fast growth rates are correlated with higher rates of respiratory metabolism (Arendt 1997; Metcalfe and Monaghan 2001; Stoks et al. 2006; Glazier 2015), higher rates of respiratory metabolism in terrestrial arthropods from high latitudes and altitudes are often attributed to countergradient selection for fast growth during short, cool growing seasons (Ayres and Scriber 1994; Schultz et al. 1996; Gotthard et al. 2000; Laugen et al. 2003). We tested whether our selection regime for cold hardiness that increased respiratory metabolism also produced correlated changes in growth rates and development time. Neither growth rates nor development time differed between cold-hardy and control lines, although cold-susceptible lines had lower growth rates that were attributable to longer development times at all temperatures other than 25°C (Fig. 3). Countergradient selection on life histories with latitude and altitude is a common enough pattern in ectotherms to be considered a general macroecological rule (Conover et al. 2009; Gaston et al. 2009). We agree that selection for fast growth rates via countergradient selection is a potentially important force that could drive patterns of metabolic cold adaptation in terrestrial arthropods (Addo-Bediako et al. 2002; White and Kearney 2013). However, our work shows that artificial selection for cold-hardiness via fast recovery from chill coma can also drive the evolution of higher rates of respiratory metabolism independent of either growth rates or activity. Because chill coma recovery time shows latitudinal clines consistent with local adaptation within and between Drosophila species and other widely distributed terrestrial arthropods (David et al. 2003; Sinclair et al. 2012), we propose that natural selection on rapid recovery from chill coma recovery, and perhaps other energetically costly aspects of cold hardiness, could also be an important force shaping patterns of metabolic cold adaptation. Acclimatization in response to low mean temperatures at high-latitudes or -altitudes could augment this elevation of metabolic rate (e.g. Berrigan and Partridge 1997; Terblanche et al. 2005).

Conclusions

High-latitude and altitude environments are characterized by shorter, cooler growing seasons and greater thermal variability (Sunday et al. 2011). Understanding patterns of local adaptation to these environments, and the potential for climate change to impact organisms in these environments will necessitate “moving beyond the mean” to consider how selection imposed by thermal variation may affect organisms. Here we show that selection imposed by exposure to and recovery from an extreme low temperature event can shape respiratory metabolism in a manner consistent with patterns of metabolic cold adaptation, without altering either activity or growth rates. This work suggests an alternative selective pressure that could lead to higher respiratory metabolism in high-latitude and high-altitude environments and emphasizes the importance of considering the multifarious nature of selection that can be experienced by organisms across the entire lifecycle.

Supplementary data

Supplementary data available at ICB online.

Supplementary Data
supp_56_1_62__index.html (1.3KB, html)

Acknowledgement

Laura Castellanos and Jennifer Kight helped with fly rearing and respirometry.

Funding

This work was supported by the National Science Foundation [IOS-1051890 to D.A.H., A.S.E., and D.B.A., IOS-1051770 to T.J.M.]; the National Institutes of Health [P30DK056336 to D.B.A.]; and the Florida Agricultural Experiment Station [to D.A.H.].

References

  1. Addo-Bediako A, Chown SL, Gaston KJ. 2002. Metabolic cold adaptation in insects: A large-scale perspective. Funct Ecol 16:332–8. [Google Scholar]
  2. Andersen JL, Findsen A, Overgaard J. 2013. Feeding impairs chill coma recovery in the migratory locust (Locusta migratoria). J Insect Physiol 59:1041–8. [DOI] [PubMed] [Google Scholar]
  3. Arendt JD. 1997. Adaptive intrinsic growth rates: An integration across taxa. Quart Rev Biol 72:149–77. [Google Scholar]
  4. Ayres MP, Scriber JM. 1994. Local adaptation to regional climates in Papilio canadensis (Lepidoptera, Papilionidae). Ecol Monogr 64:465–82. [Google Scholar]
  5. Basson CH, Nyamukondiwa C, Terblanche JS. 2012. Fitness costs of rapid cold hardening in Ceratitis capitata. Evolution 66:296–304. [DOI] [PubMed] [Google Scholar]
  6. Berrigan D, Partridge L. 1997. Influence of temperature and activity on the metabolic rate of adult Drosophila melanogaster. Comp Biochem Physiol A Mol Integr Physiol 118:1301–7. [DOI] [PubMed] [Google Scholar]
  7. Block W, Young SR. 1978. Metabolic adaptations of Antarctic terrestrial micro-arthropods. Comp Biochem Physiol A Physiol 61:363–8. [Google Scholar]
  8. Bullock TH. 1955. Compensation for temperature in the metabolism and activity of poikilotherms. Biol Rev 30:311–42. [Google Scholar]
  9. Chappell MA. 1983. Metabolism and thermoregulation in desert and montane grasshoppers. Oecologia 56:126–31. [DOI] [PubMed] [Google Scholar]
  10. Clarke A. 1993. Seasonal acclimatization and latitudinal compensation in metabolism: Do they exist? Funct Ecol 7:139–49. [Google Scholar]
  11. Clarke A, Johnston NM. 1999. Scaling of metabolic rate with body mass and temperature in teleost fish. J Anim Ecol 68:893–905. [DOI] [PubMed] [Google Scholar]
  12. Colinet H, Larvor V, Laparie M, Renault D. 2012. Exploring the plastic response to cold acclimation through metabolomics. Funct Ecol 26:711–22. [Google Scholar]
  13. Conover DO, Duffy TA, Hice LA. 2009. The covariance between genetic and environmental influences across ecological gradients. Ann N Y Acad Sci 1168:100–29. [DOI] [PubMed] [Google Scholar]
  14. Conover DO, Schultz ET. 1995. Phenotypic similarity and the evolutionary significance of countergradient variation. Trends Ecol Evol 10:248–52. [DOI] [PubMed] [Google Scholar]
  15. Crawley MJ. 2007. The R book. New York: Wiley. [Google Scholar]
  16. David JR, Gibert P, Moreteau B, Gilchrist GW, Huey RB. 2003. The fly that came in from the cold: Geographic variation of recovery time from low-temperature exposure in Drosophila subobscura. Funct Ecol 17:425–30. [Google Scholar]
  17. Findsen A, Andersen JL, Calderon S, Overgaard J. 2013. Rapid cold hardening improves recovery of ion homeostasis and chill coma recovery in the migratory locust Locusta migratoria. J Exp Biol 216:1630–7. [DOI] [PubMed] [Google Scholar]
  18. Fox HM. 1936. The activity and metabolism of poikilothermal animals in different latitudes.—I. Proc Zool Soc Lond 106:945–55. [Google Scholar]
  19. Gaston KJ, Chown SL, Calosi P, Bernardo J, Bilton DT, Clarke A, Clusella-Trullas S, Ghalambor CK, Konarzewski M, Peck LS, et al. 2009. Macrophysiology: A conceptual reunification. Am Nat 174:595–612. [DOI] [PubMed] [Google Scholar]
  20. Gibbs AG. 1999. Laboratory selection for the comparative physiologist. J Exp Biol 202:2709–18. [DOI] [PubMed] [Google Scholar]
  21. Glazier DS. 2015. Is metabolic rate a universal ‘pacemaker’ for biological processes? Biol Rev 90:377–407. [DOI] [PubMed] [Google Scholar]
  22. Gotthard K, Nylin S, Wiklund C. 2000. Individual state controls temperature dependence in a butterfly (Lasiommata maera). Proc Biol Sci 267:589–93. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Guppy M, Withers P. 1999. Metabolic depression in animals: Physiological perspectives and biochemical generalizations. Biol Rev Camb Philos Soc 74:1–40. [DOI] [PubMed] [Google Scholar]
  24. Koštál V, Korbelova J, Rozsypal J, Zahradníčková H, Cimlova J, Tomčala A, Šimek P. 2011. Long-term cold acclimation extends survival time at 0°C and modifies the metabolomic profiles of the larvae of the fruit fly Drosophila melanogaster. PLoS One 6:e25025.. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Lalouette L, Koštál V, Colinet H, Gagneul D, Renault D. 2007. Cold exposure and associated metabolic changes in adult tropical beetles exposed to fluctuating thermal regimes. Febs J 274:1759–67. [DOI] [PubMed] [Google Scholar]
  26. Lardies MA, Bacigalupe LD, Bozinovic F. 2004. Testing the metabolic cold adaptation hypothesis: An intraspecific latitudinal comparison in the common woodlouse. Evol Ecol Res 6:567–78. [Google Scholar]
  27. Laugen AT, Laurila A, Räsänen K, Merilä J. 2003. Latitudinal countergradient variation in the common frog (Rana temporaria) development rates—evidence for local adaptation. J Evol Biol 16:996–1005. [DOI] [PubMed] [Google Scholar]
  28. Levins R. 1968. Evolution in changing environments: Some theoretical explorations. Princeton (NJ: ): Princeton University Press. [Google Scholar]
  29. Lighton JRB. 2008. Measuring metabolic rates: A manual for scientists. Oxford, UK: Oxford University Press. [Google Scholar]
  30. MacKay TF, Richards S, Stone EA, Barbadilla A, Ayroles JF, Zhu D, Casillas S, Han Y, Magwire MM, Cridland JM, et al. 2012. The Drosophila melanogaster Genetic Reference Panel. Nature 482:173–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. MacMillan HA, Andersen JL, Davies SA, Overgaard J. 2015. The capacity to maintain ion and water homeostasis underlies interspecific variation in Drosophila cold tolerance. Sci Rep 5:18607.. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. MacMillan HA, Sinclair BJ. 2011. Mechanisms underlying insect chill-coma. J Insect Physiol 57:12–20. [DOI] [PubMed] [Google Scholar]
  33. MacMillan HA, Williams CM, Staples JF, Sinclair BJ. 2012. Reestablishment of ion homeostasis during chill-coma recovery in the cricket Gryllus pennsylvanicus. Proc Natl Acad Sci USA 109:20750–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Metcalfe NB, Monaghan P. 2001. Compensation for a bad start: Grow now, pay later? Trends Ecol Evol 16:254–60. [DOI] [PubMed] [Google Scholar]
  35. Michaud MR, Denlinger DL. 2007. Shifts in the carbohydrate, polyol, and amino acid pools during rapid cold-hardening and diapause-associated cold-hardening in flesh flies (Sarcophaga crassipalpis): A metabolomic comparison. J Comp Physiol B 177:753–63. [DOI] [PubMed] [Google Scholar]
  36. Morgan TJ, MacKay TF. 2006. Quantitative trait loci for thermotolerance phenotypes in Drosophila melanogaster. Heredity 96:232–42. [DOI] [PubMed] [Google Scholar]
  37. Norry xFM, Gomez FH, Loeschcke V. 2007. Knockdown resistance to heat stress and slow recovery from chill coma are genetically associated in a quantitative trait locus region of chromosome 2 in Drosophila melanogaster. Mol Ecol 16:3274–84. [DOI] [PubMed] [Google Scholar]
  38. Overgaard J, Malmendal A, Sørensen JG, Bundy JG, Loeschcke V, Nielsen NC, Holmstrup M. 2007. Metabolomic profiling of rapid cold hardening and cold shock in Drosophila melanogaster. J Insect Physiol 53:1218.. [DOI] [PubMed] [Google Scholar]
  39. Schultz ET, Reynolds KE, Conover DO. 1996. Countergradient variation in growth among newly hatched Fundulus heteroclitus: Geographic differences revealed by common-environment experiments. Funct Ecol 10:366–74. [Google Scholar]
  40. Sinclair BJ, Williams CM, Terblanche JS. 2012. Variation in thermal performance among insect populations. Physiol Biochem Zool 85:594–606. [DOI] [PubMed] [Google Scholar]
  41. Sokolova IM, Pörtner HO. 2007. The role of thermal environment in stress adaptation and tolerance: Integration of multiple stressors. J Therm Biol 32:117. [Google Scholar]
  42. Steffensen JF. 2002. Metabolic cold adaptation of polar fish based on measurements of aerobic oxygen consumption: Fact or artefact? Artefact!. Comp Biochem Physiol A Mol Integr Physiol 132:789–95. [DOI] [PubMed] [Google Scholar]
  43. Stevens MM, Jackson S, Bester SA, Terblanche JS, Chown SL. 2010. Oxygen limitation and thermal tolerance in two terrestrial arthropod species. J Exp Biol 213:2209–18. [DOI] [PubMed] [Google Scholar]
  44. Stoks R, De Block M, McPeek MA. 2006. Physiological costs of compensatory growth in a damselfly. Ecology 87:1566–74. [DOI] [PubMed] [Google Scholar]
  45. Storey KB, Storey JM. 1990. Metabolic rate depression and biochemical adaptation in anaerobiosis, hibernation and estivation. Quart Rev Biol 65:145–74. [DOI] [PubMed] [Google Scholar]
  46. Sunday JM, Bates AE, Dulvy NK. 2011. Global analysis of thermal tolerance and latitude in ectotherms. Proc R Soc Lond B Biol Sci 278:1823–30. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Team RC. 2015. R: A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing. [Google Scholar]
  48. Terblanche JS, Clusella-Trullas S, Deere JA, Van Vuuren BJ, Chown SL. 2009. Directional evolution of the slope of the metabolic rate–temperature relationship is correlated with climate. Physiol Biochem Zool 82:495–503. [DOI] [PubMed] [Google Scholar]
  49. Terblanche JS, Jaco Klok C, Chown SL. 2005. Temperature-dependence of metabolic rate in Glossina morsitans morsitans (Diptera, Glossinidae) does not vary with gender, age, feeding, pregnancy or acclimation. J Insect Physiol 51:861–70. [DOI] [PubMed] [Google Scholar]
  50. Todgham AE, Stillman JH. 2013. Physiological responses to shifts in multiple environmental stressors: Relevance in a changing world. Integr Comp Biol 53:539–44. [DOI] [PubMed] [Google Scholar]
  51. Torres JJ, Somero GN. 1988. Metabolism, enzymic activities and cold adaptation in Antarctic mesopelagic fishes. Mar Biol 98:169–80. [Google Scholar]
  52. White CR, Alton LA, Frappell PB. 2012. Metabolic cold adaptation in fishes occurs at the level of whole animal, mitochondria and enzyme. Proc Biol Sci 279:1740–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. White CR, Kearney MR. 2013. Determinants of inter-specific variation in basal metabolic rate. J Comp Physiol B 183:1–26. [DOI] [PubMed] [Google Scholar]
  54. Williams CM, Watanabe M, Guarracino MR, Ferraro MB, Edison AS, Morgan TJ, Boroujerdi AFB, Hahn DA. 2014. Cold adaptation shapes the robustness of metabolic networks in Drosophila melanogaster. Evolution 68:3505–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Wohlschlag DE. 1960. Metabolism of an Antarctic fish and the phenomenon of cold adaptation. Ecology 41:287–92. [Google Scholar]
  56. Zeileis A, Grothendieck G. 2005. zoo: S3 infrastructure for regular and irregular time series. J Stat Softw 14:1–27. [Google Scholar]
  57. Zera AJ, Harshman LG. 2001. The physiology of life history trade-offs in animals. Annu Rev Ecol Syst 32:95. [Google Scholar]
  58. Zuur AF, Ieno EN, Elphick CS. 2010. A protocol for data exploration to avoid common statistical problems. Methods Ecol Evol 1:3. [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Data
supp_56_1_62__index.html (1.3KB, html)
supp_icw009_icb-2016-0057-File008.tif (2MB, tif)
supp_icw009_icb-2016-0057-File005.tif (1.7MB, tif)
supp_icw009_icb-2016-0057-File006.tif (3.4MB, tif)
supp_icw009_icb-2016-0057-File007.tif (3.4MB, tif)
supp_icw009_icb-2016-0057-File009.docx (25.1KB, docx)

Articles from Integrative and Comparative Biology are provided here courtesy of Oxford University Press

RESOURCES