Costs and benefits of cold acclimation in field-released Drosophila
- Torsten N. Kristensen*,†,‡,§,
- Ary A. Hoffmann‡,
- Johannes Overgaard¶,
- Jesper G. Sørensen†,
- Rebecca Hallas‡, and
- Volker Loeschcke†
- *Department of Genetics and Biotechnology, University of Aarhus, P.O. Box 50, DK-8830 Tjele, Denmark;
- †Aarhus Centre for Environmental Stress Research, Department of Biological Sciences, University of Aarhus, Ny Munkegade, Building 1540, DK-8000 Aarhus, Denmark;
- ‡Center for Environmental Stress and Adaptation Research, Departments of Genetics and Zoology, Bio21 Institute, University of Melbourne, Melbourne VIC 3010, Australia; and
- ¶National Environmental Research Institute, Department of Terrestrial Ecology, University of Aarhus, Vejlsøvej 25, P.O. Box 314, DK-8600 Silkeborg, Denmark
-
Edited by David L. Denlinger, Ohio State University, Columbus, OH, and approved November 12, 2007 (received for review August 28, 2007)
Abstract
One way animals can counter the effects of climatic extremes is via physiological acclimation, but acclimating to one extreme might decrease performance under different conditions. Here, we use field releases of Drosophila melanogaster on two continents across a range of temperatures to test for costs and benefits of developmental or adult cold acclimation. Both types of cold acclimation had enormous benefits at low temperatures in the field; in the coldest releases only cold-acclimated flies were able to find a resource. However, this advantage came at a huge cost; flies that had not been cold-acclimated were up to 36 times more likely to find food than the cold-acclimated flies when temperatures were warm. Such costs and strong benefits were not evident in laboratory tests where we found no reduction in heat survival of the cold-acclimated flies. Field release studies, therefore, reveal costs of cold acclimation that standard laboratory assays do not detect. Thus, although physiological acclimation may dramatically improve fitness over a narrow set of thermal conditions, it may have the opposite effect once conditions extend outside this range, an increasingly likely scenario as temperature variability increases under global climate change.
As climatic conditions around the globe change rapidly, the distribution of many animal species is expected to be altered as a direct or indirect response to thermal conditions (1). Because evolutionary responses to climatic changes can take time and be difficult to achieve, the persistence of animal populations exposed to climatic stress will often depend on behavioral mechanisms and physiological plasticity rather than evolutionary responses (refs. 2 and 3, but see ref. 4). Animals can counter extreme thermal conditions by becoming physiologically acclimated, allowing them to survive and even reproduce under conditions that would otherwise be highly stressful (2, 5). Conspecific populations from different environments may vary substantially in stress resistance because they have become acclimated to local conditions (6, 7). However, acclimation is likely to be beneficial only when the acclimation regime correctly predicts the future regime (2, 5, 8). The overall balance between benefits and costs depends on the length and severity of the treatment, the environmental conditions that follow, and the time it takes to produce a more fit phenotype in response to an acclimation treatment (5, 9, 10). If temperatures fluctuate, organisms acclimated to cold or hot conditions could potentially suffer a decrease in fitness as temperatures move to the opposite extreme.
The costs and benefits of acclimation have been defined mainly in laboratory assays of fitness, making the ecological significance of acclimation unclear (5, 11, 12). Physiological changes need ideally to be defined under field conditions as laboratory experiments typically do not incorporate all factors that are of importance for fitness (13–15). For insects such as Drosophila melanogaster, field releases examining the ability to locate resources can be used as an efficient tool to assess a component of fitness in the wild. Releases have now been used to test associations between fitness in the laboratory and the field (13, 15–17), inbreeding effects (18) and effects of density and size traits on field performance (19, 20).
Cost and benefits associated with cold acclimation have never been investigated directly in the field or compared simultaneously in the field and laboratory. Here, we test the effect of two types of cold acclimation on the ability to locate resources in the field at low, intermediate, and high ambient temperatures on two continents. Flies were cold-acclimated by rearing at 15°C throughout development (developmental cold acclimation) or exposing newly hatched flies (developed at 25°C) to 11°C for 5 days (adult cold acclimation). These acclimated flies were tested against flies kept constantly at 25°C (referred to as controls). In the developmental cold acclimation releases we contrasted 5-day-old cold-acclimated flies and 5-day-old control flies. In the adult cold acclimation releases we contrasted 5-day-old adult cold-acclimated flies with 2- and 5-day-old control flies (different age classes were included because physiological aging is expected to be slower at 11°C). At the same time, the costs and benefits of cold acclimation for cold and heat resistance were tested in the laboratory with mortality assays. Finally, we contrasted the effects of acclimation to cold temperatures across the thermal range with effects from heat acclimation.
Results
Developmental Cold Acclimation.
To test the effects of developmental cold acclimation at 15°C on the ability of flies (D. melanogaster) to locate resources in the field, we performed eight releases (developmental cold acclimation 1–8) in Denmark at cold temperatures and in Australia at warm temperatures (Table 1). A total of 2,000–3,000 flies were released per treatment group (flies cold-acclimated at 15°C or kept at 25°C throughout development). The percentage of flies captured in these eight releases varied between 2.2% and 39.3% (Table 1). Logit analyses (Table 2) showed a significant effect of sex in all eight releases where females were more often caught than males. In addition, three of eight experiments showed a significant interaction between sex and treatment (Table 2).
Description of releases of nonacclimated or cold-acclimated (developmental cold acclimation and adult cold acclimation) flies
Logit models testing the effects of sex and developmental cold acclimation on the probability of capture under cold (DK, Denmark) and warm (AU, Australia) field temperatures
Relative capture success was calculated separately for males and females because of significant sex and/or treatment by sex effects. In developmental cold-acclimation releases 1–3 we only caught three flies reared at 25°C, whereas 630 flies reared at 15°C were caught, yielding a strong treatment effect in both sexes (Table 2). Average capture success could not be calculated as flies reared at 25°C were not caught in two of these releases. In developmental cold-acclimation releases 4–6 we also caught significantly more flies reared at 15°C compared with flies reared at 25°C except for males in release 4 (Table 2). Males and females reared at 15°C were caught in numbers 24% and 75% higher than those reared at 25°C in releases 4–6. In the two developmental cold-acclimation releases performed at warm temperatures (releases 7 and 8) we caught significantly more of the control flies reared at 25°C (Table 2). Both males and females reared at 25°C were >2,000% more likely to be caught than flies reared at 15°C, showing that the benefits at low temperatures come at an enormous cost at high temperatures.
There was a tendency for flies reared at 25°C to have a higher probability of being caught farther away from the release point than flies reared at 15°C in the two releases performed in Australia at warm temperatures. In the same two releases there was a tendency for male flies reared at 15°C to be caught later [supporting information (SI) Table 4].
Adult Cold Acclimation.
We also tested potential costs and benefits associated with cold acclimation at 11°C for 5 days after emergence in field releases under cold and warm temperatures in Denmark and Australia. This test was done in nine releases (adult cold acclimation 1–9). In all releases cold-acclimated flies were released together with 2- and 5-day-old control flies.
Logit analyses (Table 3) showed that the interaction between treatment and sex was significant in seven of the nine releases. The effect of sex was significant in all releases, with females being caught more frequently in most releases (Table 3). Relative capture success was calculated separately for the sexes, and in all nine releases we found a significant effect of adult cold acclimation (Table 3). The male cold-acclimated flies were captured with a significantly higher likelihood than both 2- and 5-day-old control flies in five of six releases performed at cool conditions in Denmark (Table 3). In the three adult acclimation releases performed at warm temperatures in Australia, cold-acclimated flies of both sexes had a lower capture success than both the 2- and 5-day-old control flies (Table 3). In four of six releases performed at cool conditions in Denmark, female cold-acclimated flies had higher capture success than 2-day-old control flies, whereas no consistent pattern emerged in the comparison between flies acclimated at 11°C and 5-day-old control flies (Table 3).
Logit models testing the effects of sex and adult cold acclimation on the probability of capture under cold (DK, Denmark) and warm (AU, Australia) field temperatures
On average across the six releases performed under cool conditions, males acclimated at 11°C were caught with 56% and 21.2% higher likelihood relative to 2- and 5-day-old control flies. In females the same numbers were 14.3% and 0%, respectively. In the three warm releases, cold-acclimated males were 50% and 58% less likely to be caught relative to 2- and 5-day-old control males, respectively, and for the females the equivalent numbers were 57% and 55%. Adult cold acclimation therefore increased capture success under cool conditions (particularly in males) but decreased it sharply under warm conditions.
On days 1 and 2 of capture, there was a tendency toward females acclimated at 11°C to be caught with a higher likelihood than flies reared at 25°C farther away from the release point (SI Table 5). In none of the releases did we observe any effect of treatment on the within day likelihood of capture at later capture rounds (SI Table 5). The likelihood of capture on the second day (in cool releases) was higher for nonacclimated 2-day-old flies in males in five of six releases (SI Table 6).
Overall Patterns.
The main conclusion emerging from the releases is that flies cold-acclimated throughout development or at the adult stage have large benefits when tested at low temperatures but have equally large costs when tested at warm temperatures (Fig. 1). Flies cold-acclimated for 5 days at 11°C (adult cold acclimation) before release have lower benefits relative to flies reared at 15°C throughout development (developmental cold acclimation) when released at low temperatures but also lower costs when tested at warm temperatures. Therefore, developmental cold acclimation is both more costly and more beneficial. We can compare the relative costs and benefits of acclimation at opposite thermal extremes by comparing our results to those of Loeschcke and Hoffmann (16), who considered capture success in heat-acclimated flies, following a similar design used for Trichogramma brassicae by Thomson et al. (21). Beneficial effects of heat acclimation in D. melanogaster are observed only at very high temperatures (Fig. 1), and acclimation temperatures have to exceed 34°C to have a positive effect. Benefits can be extremely large although relative to cold acclimation of adult flies the costs appear to be smaller (Fig. 1).
Number of captured acclimated female (A) and male (B) flies relative to the number of captured nonacclimated flies (± SEM). We contrast our results on cold acclimation with data on heat-acclimated flies (16). Results from releases performed at similar temperatures within the two datasets were pooled, and SEM represents standard errors across releases within temperature. In releases involving flies acclimated at 11°C at the adult stage, proportions captured are presented relative to the average of the two control age classes tested. In the study by Loeschcke and Hoffmann (16) flies were acclimated at 36°C for 1 h and released 6 h after the end of the heat acclimation. Control flies were the same age and treated similarly to the acclimated flies except for acclimation. Data points are connected for ease of interpretation.
Laboratory Assays.
Because acclimation effects are normally assessed under laboratory conditions, we tested for both benefits and costs of cold acclimation under commonly used laboratory assays and obtained results that did not match those from the field assays (results for females and males are presented in Fig. 2 and SI Fig. 3, respectively). As expected, male and female flies acclimated at 11°C at the adult stage had higher cold resistance relative to both control age class flies. Two-day-old control flies had higher cold resistance relative to 5-day-old control flies. However, no costs of adult cold acclimation were detected in these laboratory assays. Male flies did not differ in level of heat resistance relative to 2- and 5-day-old control flies, whereas female flies acclimated at 11°C had slightly higher heat resistance relative to 5-day-old control flies.
Mean cold and heat resistance (± SEM) of female flies acclimated at 15°C or nonacclimated control flies (25°C) (A) or acclimated at 11°C (11°C 5d) or nonacclimated control flies (25°C 2 or 5 d) (B). Cold and heat resistance were estimated based on mortality assays. Mortality was scored after exposure of 5 replicates each consisting of 10 flies per treatment (cold acclimated or control) at cold or warm temperatures for 1 h. χ2 tests were performed to test whether resistance differed between treatment, and probabilities from these tests are presented.
For developmental cold acclimation, both females and males developed at 15°C had higher cold resistance relative to control flies (Fig. 2 and SI Fig. 3). However, again there was no evidence for a cost; female and male flies developed at 15°C did not differ in heat resistance relative to control flies. These results from the laboratory are in agreement with other studies and support the beneficial acclimation hypothesis. But the absence of costs stands in sharp contrast to our results obtained in the field and emphasizes that results from the laboratory often do not reliably predict outcomes in the wild.
Discussion
Evolutionary physiologists have debated the importance of acclimation for surviving harsh environments for years. Some of this controversy has arisen because laboratory experiments can produce misleading results for interpreting field performance (13, 15, 17). Although some theoretical and empirical considerations suggest that acclimation will have benefits under a range of environmental conditions (22), these may rarely be realized in nature if conditions leading to acclimation do not predict future conditions (10, 14, 23).
In laboratory studies, cold acclimation of D. melanogaster adults increases lifespan and reduces mortality and chill coma recovery time after exposure to subzero temperatures (24, 25). Developmental acclimation at low temperatures also influences many aspects of cold resistance, including chill coma recovery time, mortality, fecundity, and longevity in diverse insect groups including drosophilids (26–28). However, plastic responses to cold acclimation are complex; they depend on rearing temperature, are often sex specific, and may extend across generations (24, 29). In laboratory studies on cold acclimation it is common to maintain most external stimuli constant while only temperature and (in some cases) the light regime is altered. This approach has many benefits for elucidating and isolating underlying physiological mechanisms, but fails to portray the often much more complex situation that insects and other ectothermic animals encounter in nature. In the present study, we addressed this central question by testing the cost and benefits of cold acclimation on cold and heat tolerance under both field and laboratory conditions. In the laboratory we used simple mortality assays to test the ability of D. melanogaster to acclimate for improved survival after cold and heat shock. In the field test we used the ability of the flies to find a resource as a proxy of fitness. With this approach we tested how physiological adaptation through cold acclimation changes the individuals' ability to perform in a situation where many external stimuli such as humidity, predators, or competitors are present in addition to thermal stimuli.
The results obtained from the field study were clear cut; both types of cold acclimation (adult and developmental cold acclimation) were beneficial in the field when temperatures were low but this came at a huge cost when flies were tested at warm temperatures (Fig. 1). Such costs and strong benefits were not evident in laboratory tests where we found no reduction in heat survival in the cold acclimated flies (Fig. 2 and SI Fig. 3). This result from the laboratory is in accordance with most previous investigations in the laboratory of either developmental or adult cold acclimation in ectotherms (see references in refs. 22 and 30). Thus, field release studies reveal costs associated with acclimation that laboratory-based mortality assays do not pick up, emphasizing that laboratory results can be misleading (13, 15, 17, 18).
One feature of our work is that we were able to distinguish between developmental and adult cold acclimation, whereas many studies have not distinguished between these two types of acclimation or focused on only one type (22, 31). The acclimation temperatures used for adult and developmental cold acclimation in our study were different to reflect the colder conditions experienced by overwintering adults, but nevertheless suggest that both the benefits and costs of cold acclimation are much more severe for developmental acclimation when compared with adult acclimation. Different mechanisms may be involved in acclimation at immature and adult life stages (28, 32), with potentially quite different effects on field performance.
Although laboratory tests almost universally have indicated benefits of acclimation, only a few studies have indicated costs (2, 3, 22), in marked contrast to our field results. It is therefore important to evaluate cost and benefits of acclimation and other plastic responses under field conditions by using assays relevant to field performance. Loeschcke and Hoffmann (16) previously undertook field releases with D. melanogaster to examine capture success in heat-acclimated flies and also found that field results differed from patterns established in the laboratory. They showed that beneficial effects were observed only at very high temperatures exceeding 34°C. At those high temperatures, however, the benefits were extremely large but countered by costs not previously detected in laboratory assays (Fig. 1). As in our study, this result shows that benefits can be extremely large, although, relative to cold acclimation during development or at the adult stage, the costs appear to be smaller.
It has been suggested that the ability to acclimate for future cold or warm conditions is likely to be adaptive in temperate species exposed to seasonally variable conditions (2). However, costs associated with cold and heat acclimation in the field suggest that the net benefits of acclimation will depend on levels of climatic variability experienced by organisms, which will change as climatic conditions become increasingly variable in the future (33, 34).
Materials and Methods
Experimental Population.
A genetically diverse mass bred population of D. melanogaster was used in the experiments. It was held at census sizes of 1,000–1,500 individuals per generation at standard laboratory conditions (25 ± 0.2°C, 50% relative humidity, 12/12-h light/dark cycle) since being established in 2002 (for details see ref. 35).
Flies Cold Acclimated Throughout Development or Kept at 25°C (Developmental Cold Acclimation).
Eggs producing flies to be released were laid at 25°C in 200-ml bottles with 35 ml of a sugar-agar-dead yeast medium. Density was partly controlled in bottles by restricting the laying period of 50–60 parental adults to 1 day. Eggs developing at 15°C were laid 17 days before eggs developing at 25°C because of slower development. Between 300 and 500 offspring emerged per bottle. These were collected over a 24-h period and distributed into bottles with fresh medium with 200 flies per bottle. Flies were unsexed (the sex ratio of the flies did not differ significantly from 1:1; results not shown). We had two experimental treatments, one constituted by flies developed at 25°C and another by flies developed at 15°C. After emergence flies were aged for 5 days before release at the temperatures at which they developed (15°C or 25°C). Flies from each of the two experimental groups were also collected to be used for testing heat and cold resistance in the laboratory. These flies were reared under controlled density (20 larvae per vial with 7 ml of medium) at either 15°C or 25°C.
Flies Cold Acclimated as Adults or Kept at 25°C (Adult Cold Acclimation).
All flies were reared at 25°C throughout development. After hatching two groups of flies were aged 2 or 5 days at 25°C and one group of flies was aged for 5 days at 11°C before release. The rationale for including the 2-day group was to have control flies with roughly the same physiological age as the flies cold acclimated for 5 days at 11°C. A colder temperature was used for adult acclimation compared with developmental acclimation because D. melanogaster overwinters at the adult stage. Flies from each of the three experimental groups were also collected for testing heat and cold resistance in the laboratory.
Release of Flies Reared at 15°C or 25°C During Development (Developmental Cold Acclimation).
Six releases were undertaken at cool temperatures in Denmark (designated “developmental cold acclimation 1 DK” to “developmental cold acclimation 6 DK”). Capture points in these releases were 5 m apart starting 5 m from the release site and extending up to 25 m in opposite directions from the release point. Two releases were undertaken at warm temperatures in Victoria, Australia (designated “developmental cold acclimation 7 AU” and “developmental cold acclimation 8 AU”). In these releases capture points were 10 m apart starting 10 m from the release site and extending up to 50 m in opposite directions from the release point. During all releases temperatures were recorded with data loggers (Tinytalk II, Chichester, U.K.) positioned close to the release site. Temperatures are summarized in Table 1. All releases took place in woodland, consisting of pine and oak trees at the locations in Denmark and eucalyptus with flowering shrubs at the locations in Australia.
Flies were transported by car to the release sites in insulated styrofoam boxes kept at 24–26°C. Just before the release, groups of 200 flies were transferred into vials with 0.0015 (± 0.0005) g of fluorescent micronized dust. Dust colors were randomly assigned to the different treatments and changed between releases. The number of flies released per treatment and the times at which the flies were released are presented in Table 1. To release flies, vials with flies from the different lines were randomly arranged in a container and foam stoppers were removed from vials. After release flies were captured from resources (buckets with mashed banana) by netting and/or aspirating, starting 1 h after release and every hour subsequently until at least four rounds of captures were performed.
Captured flies were held on ice in the field to knock them out so that transfer of dust color between the flies was minimized. They were then transported to the laboratory and stored in a freezer until colors were scored under UV light. Few captured flies (≪1%) had no color and were discarded. We never observed flies with two colors.
Release and Recapture of Flies Exposed to 11°C or 25°C Before Releases (Adult Cold Acclimation).
Six releases testing effects of adult cold acclimation were undertaken at cool temperatures in Denmark (designated “adult cold acclimation 1 DK” to “adult cold acclimation 6 DK”). Capture points in these releases were 5 m apart starting 5 m from the release site and extending up to 30 m in opposite directions from the release point. Three releases were undertaken at warm temperatures in Victoria, Australia (designated “adult cold acclimation 7 AU” to “adult cold acclimation 9 AU”). In these releases capture points were 10 m apart starting 10 m from the release site and extending up to 50 m in opposite directions from the release point. Flies were released in the morning or around noon (Table 1) and captured from resources, starting 1 h after release and every hour subsequently until at least four rounds of captures were performed on the first day of capture. In six releases (adult cold acclimation 1 DK to adult cold acclimation 6 DK) we also performed three rounds of captures on the second day. All aspects related to marking flies, temperature logging, releasing, and recapturing of flies were as described above.
Experimental Protocol for Testing Cold and Heat Resistance in the Laboratory.
The three experimental groups of flies from the experiment testing effects of adult cold acclimation (25°C 2-day-old, 25°C 5-day-old, and 11°C 5-day-old flies) and the two experimental groups of flies from the experiment testing effects of developmental cold acclimation (15°C 2-day-old and 25°C 5-day-old flies) were tested for cold and heat shock tolerance in the laboratory. Cold shock resistance was tested by transferring 5-day-old (separated in sexes without the use of CO2) flies from each of the experimental groups to empty glass vials that were acutely exposed to 1-h cold shocks at −1°C, −3°C, −5°C, −7°C, −9°C, or −11°C. Five vials with 10 flies were exposed per sex, experimental group, and temperature. Immediately after the cold shock flies were returned to vials with fresh medium and transferred to 25°C, and mortality was scored 72 h later. Heat shock resistance was assessed by exposing flies (separated in sexes without the use of CO2) acutely to a 1-h heat shock at 37°C, 37.5°C, 38°C, 38.5°C, or 39°C. The flies were heat-shocked in empty glass vials by using five vials with 10 flies per sex, experimental group, and temperature. After the heat shock, flies were transferred to fresh food vials and mortality was scored 24 h later.
Statistical Analysis.
First, we tested whether the total number of flies caught differed between treatments and sexes. Capture data were initially treated as categorical and analyzed separately for each release. The difference between treatments and sexes in capture success was assessed by using Logit models (presented in SI Table 7). The sample of flies that was not captured was estimated by assuming a 1:1 sex ratio. This analysis assumes that each fly represents an independent data point for a treatment and sex. The assumption of independence is likely to be justified because flies were pooled across multiple culture bottles (to minimize culture effects) and releases in a randomized position at the field site (to minimize any point of release effects). Apart from using a Logit analysis, we also undertook a simpler analysis where capture rates were compared directly between the treatment groups in each release and tested by using the χ2 statistic against an expectation of equal numbers of individuals being captured from each treatment group in a release (presented in SI Tables 8 and 9).
We then considered only the capture data to assess capture success at various time points after release and distances from the release point. The proportion of flies of each sex caught later and farther away from the release point was calculated for all treatments and releases. To look for significant interactions between treatment and time of capture and capture point, G tests were performed (SI Tables 4 and 5).
To indicate the relative capture success of the cold-acclimated or cold-reared flies relative to nonacclimated control flies, we computed relative capture rates comparing the likelihood of capture of males vs. females or cold-acclimated vs. control flies (males and females tested separately because of sex and/or treatment by sex effects in the Logit models). For this measure of relative capture success, values >1 indicate that more flies from the cold-acclimated groups than flies from the control groups were caught (or more males than females were caught).
In all releases, flies caught at the same distance from the release point were combined when testing for effects of treatment on how far out we caught flies. Flies caught at capture round 1 were tested against flies caught at later captures when testing the effects of treatment on time of capture.
To test the effects of treatment on cold and heat resistance estimated in the laboratory we calculated the areas below graphs and used conservative χ2 tests to look for differences between treatments.
Acknowledgments
We thank D. Andersen, K. V. Kristensen, and T. S. Thomsen for technical help; C. Pertoldi for advice on statistical analyses; R. Huey and one anonymous referee for constructive input on the manuscript; and the Danish and Australian Research Councils for financial support with center grants (A.A.H. and V.L.), postdoc positions (T.N.K., J.O., and J.G.S.), and a Federation Fellowship (A.A.H.).
Footnotes
- §To whom correspondence should be addressed. E-mail: torsten.nygaard{at}agrsci.dk
-
Author contributions: T.N.K. and V.L. designed research; T.N.K., A.A.H., J.O., J.G.S., R.H., and V.L. performed research; T.N.K. and A.A.H. analyzed data; and T.N.K., A.A.H., J.O., and V.L. wrote the paper.
-
The authors declare no conflict of interest.
-
This article is a PNAS Direct Submission.
-
This article contains supporting information online at www.pnas.org/cgi/content/full/0708074105/DC1.
- © 2007 by The National Academy of Sciences of the USA
