- Email: [email protected]
Testing the thermal limits: Non-linear reaction norms drive disparate thermal acclimation responses in Drosophila melanogaster
Journal of Insect Physiology 118 (2019) 103946
Contents lists available at ScienceDirect
Journal of Insect Physiology journal homepage: www.elsevier.com/locate/jinsphys
Testing the thermal limits: Non-linear reaction norms drive disparate thermal acclimation responses in Drosophila melanogaster
T
⁎
Paul Vinu Salachan , Hélène Burgaud, Jesper Givskov Sørensen Department of Bioscience, Aarhus University, Ny Munkegade 116, 8000 Aarhus C, Denmark
A R T I C LE I N FO
A B S T R A C T
Keywords: Adult acclimation CTmax Time to knock-down Phenotypic plasticity Thermal tolerance
Critical thermal limits are important ecological parameters for studying thermal biology and for modelling species’ distributions under current and changing climatic conditions (including predicting the risk of extinction for species from future warming). However, estimates of the critical thermal limits are biased by the choice of assay and assay conditions, which differ among studies. Furthermore, estimates of the potential for phenotypic plasticity (thermal acclimation) to buffer against thermal variability are usually based on single assay conditions and (usually linear) extrapolation from a few acclimation temperatures. We produced high resolution estimates of adult acclimation capacity for upper tolerance limits at different assay conditions (ramping rates and knockdown temperatures) using CTmax (dynamic) and knock-down (static) thermal assays in the model species Drosophila melanogaster. We found the reaction norms to be highly dependent on assay conditions. We confirmed that progressively lower ramping rates or higher knock-down temperatures led to overall lower tolerance estimates. More surprisingly, extended assays (lower ramping rates or lower knock-down temperatures) also led to increasingly non-linear reaction norms for upper thermal tolerance across adult acclimation temperatures. Our results suggest that the magnitude (capacity) and direction (beneficial or detrimental) of acclimation responses are highly sensitive to assay conditions. The results offer a framework for comparison of acclimation responses between different assay conditions and a potential for explaining disparate acclimation capacity theories. We advocate cautious interpretation of acclimation capacities and careful consideration of assay conditions, which should represent realistic environmental conditions based on species’ ecological niches.
1. Introduction
short-term (hardening) and long-term (acclimation) temperature exposure (Hoffmann et al., 2003). This calls to question the role of adaptive phenotypic plasticity (thermal acclimation) for mitigating detrimental effects of stressful environmental temperatures (Gunderson and Stillman, 2015; Sgrò et al., 2016; Sørensen et al., 2016a). To reliably use estimates of thermal tolerance in general and acclimation capacity specifically, our measurements need to be accurate and ecologically relevant (i.e. relevant for the species and its thermal niche). Estimates of thermal limits (and of thermal plasticity) can be influenced by the method used (Rezende et al., 2014; Sørensen et al., 2016b). There are two primary methods of assessing upper thermal tolerance in insects: a static knock-down (KD) assay and a dynamic ramping (CTmax) assay (Lutterschmidt and Hutchison, 1997). In a static assay, individuals are kept at a constant stressful temperature and the time required to reach an end-point (usually the cessation of movement) is measured as the time to knock-down (TKD) at that temperature. In a dynamic assay, the temperature is increased at a predetermined rate from a benign temperature and the temperature at
Temperature can have a profound effect on the survival and fitness of organisms, prompting various strategies for adapting to variation in diel, seasonal, and annual mean temperatures. The thermal biology of organisms (e.g. adaptive differences among species and populations) can be studied via parameters of a thermal performance curve (Huey and Berrigan, 2001; Martin and Huey, 2008). Specifically, the lower and upper limits of the curve (thermal tolerance) as well as the peak and breadth (optimal temperature and thermal range) characterize the thermal performance curve (Huey and Kingsolver, 1989; Huey and Stevenson, 1979). For many ectotherms, lower- and upper critical thermal limits (CTmin and CTmax) have been extensively studied due to the relative ease of investigating these endpoints, and the estimates have been used as predictors of species distribution (Kellermann et al., 2012; Overgaard et al., 2014). However, the upper thermal limits are considered evolutionarily constrained (Araújo et al., 2013; Hoffmann et al., 2013; Kellermann et al., 2012), but does respond adaptively to
⁎
Corresponding author. E-mail address: [email protected] (P.V. Salachan).
https://doi.org/10.1016/j.jinsphys.2019.103946 Received 2 June 2019; Received in revised form 12 September 2019; Accepted 13 September 2019 Available online 13 September 2019 0022-1910/ © 2019 Elsevier Ltd. All rights reserved.
Journal of Insect Physiology 118 (2019) 103946
P.V. Salachan, et al.
2.2. Adult acclimation regimes
which the individual ceases movement is recorded as the CTmax. The potential for heat hardening (rapidly induced beneficial physiological adjustments) during such assays likely increases with assay duration and might explain some discrepancies among static assays (usually performed at relatively high temperatures) and more prolonged (dynamic) thermal assays (Sgrò et al., 2010). Furthermore, results are assay condition dependent with lower ramping rates or higher KD temperatures decreasing the thermal tolerance estimates (Allen et al., 2016; Machekano et al., 2018; Sørensen et al., 2013). The effect of methodology is most likely driven by the interaction between time and temperature leading to increased accumulation of heat damage in longer assays (Sørensen et al., 2013), although confounding effects of desiccation, starvation etc. have also been proposed (Santos et al., 2011). Consequently, while some argue that fast assays (high KD temperature or high ramping rates) are less likely to be compromised by confounding effects (Rezende et al., 2011), lower ramping rates (and lower KD temperatures) might be considered more ecologically relevant (Overgaard et al., 2012). Thermal traits likely harbor different levels of plasticity and reaction norms (Schulte et al., 2011; Teets and Hahn, 2018). Most studies do not verify the shape of the reaction norms but assume linear acclimation responses based on few acclimation temperatures (see e.g. Gunderson and Stillman, 2015). Studies on insect acclimation often distinguish between developmental and adult acclimation (Angilletta, 2009). For constant developmental acclimation temperatures, Drosophila species can have different (linear) reaction norms for thermal limits (Schou et al., 2016). To our knowledge, no study has systematically looked at the reaction norms for thermal tolerance over multiple adult acclimation temperatures and in connection with assay conditions in both ramping and static assays. Thus, no clear consensus exists as to how thermal tolerance assays affect the detection of ecologically relevant and adaptive acclimation responses (Overgaard et al., 2012; Rezende et al., 2011) and how this might affect predictions about acclimation capacity (Sørensen et al., 2016a). To investigate assay-dependent adult acclimation responses of D. melanogaster, we subjected flies to various thermal performance assays following adult acclimation to one of four different temperatures (15, 19, 25 or 29 °C). We assayed thermal tolerance using two different assays; a static KD assay and a dynamic ramping assay. To tease apart any differences associated with assay condition, we applied four different ramping rates (between 0 and 0.5 °C/min) and five different KD temperatures (between 36 and 40 °C) to cover a broad range of conditions and assay durations. We expect lower ramping rates or higher KD temperatures to result in lower estimates of thermal tolerance (Machekano et al., 2018; Sørensen et al., 2013). However, here we focus specifically on the assay condition and its interaction with the acclimation temperature, in order to compare patterns of acclimation responses in tolerance estimates, to generate a general framework for understanding insect acclimation capacity. Following the beneficial acclimation hypothesis (Leroi et al., 1994), and in accordance with the results of Schou et al. (2016), we expect heat tolerance to be positively and linearly related to acclimation temperature.
The collected vials from each batch of flies were randomly distributed among four constant acclimation temperatures (15, 19, 25 and 29 °C) and maintained there for seven days. We used data loggers (iButton) to verify the desired temperatures. Every second day, flies were transferred to fresh food vials.
2.3. Thermal assays We investigated the tolerance to high temperatures in acclimated adult flies (7 days old) using a dynamic or a static heat tolerance measurement. For both thermal assays, flies from all four acclimation treatments (from a single batch of collected eggs) were always assayed together. Thermal assays were conducted in a temperature-controlled water bath (glass aquarium) with a circulating pump to maintain the same temperature throughout the water bath. Flies were placed in small screw cap glass vials on a rack without the use of anesthesia and submerged in the water bath. For the dynamic assay, the temperature was increased from 25 °C at a rate of 0.02, 0.05, 0.2, or 0.5 °C/min until the flies were no longer responsive after mild tapping with a rod and stimulation with a flash light (i.e. they had reached the CTmax). For the static (KD) assay, flies were exposed to 36, 37, 38, 39 or 40 °C, until they were no longer responsive. For both types of assay, the flies were placed in the tube immediately prior to each assay and the maximal assay duration was on average around 12 h (for the slowest ramping rate of 0.02 °C/min) or 7.5 h (for the lowest KD temperature of 36 °C). At a 12-hour assay duration, we expect no or negligible confounding effects of starvation or desiccation on thermal tolerance for all assay temperatures (see Manenti et al., 2018). In each assay, ten females from each acclimation treatment were used for each experimental block. Three experimental blocks (for a total of 30 female flies) were assayed for each KD temperature or ramping rate, except for the ramping rate of 0.02 °C/min, where two blocks of 20 flies (for a total of 40 female flies) were assayed.
2.4. Statistical analysis To investigate the effect of acclimation temperature and assay condition on the thermal tolerance estimates, we performed ANOVA based on linear mixed effects modelling using the ‘lme4’ package (v.1.15) (Bates et al., 2015) in R (R Core Team, 2018). Q-Q and residual plots were used to verify the assumptions for normality of residuals and homogeneity of variances. Both acclimation temperature and ramping rate or KD temperature were considered as continuous variables. The data were transformed appropriately in cases where the assumptions were not met. Sequential model reduction was performed, and the Chisquare values were reported for the effect of each dropped term (using ‘anova’ function, R ‘stats’ package). For both estimates of thermal tolerance (CTmax and TKD), experimental block was included as a random factor. To analyze the effect of acclimation temperature and ramping rate on CTmax, we considered both predictor variables and their interaction in the model. The ramping rate was log transformed for the analysis. Acclimation and KD temperatures and their interaction were included as predictor variables in the analysis of log-transformed TKD. Inspection of the data revealed that the significant interactions between acclimation temperature and KD temperature/ramping rate in both full models were caused by different reaction norms among conditions (ramping rates or KD temperatures) within assays. Based on this we analyzed the data within each individual ramping rate or KD temperature and investigated reaction norms in each.
2. Materials and methods 2.1. Experimental animals We used a population of D. melanogaster collected in 2013, from Odder, Denmark (Schou et al., 2015). The population was maintained in the laboratory at 25 °C (12L:12D) on standard Drosophila oatmeal–sugar–yeast–agar medium. All experimental animals were raised under controlled density conditions by transferring approximately 40 ( ± 3) eggs into fresh 7 mL food vials. Upon emergence, flies were separated by sex under CO2 anesthesia and kept in vials of 20 virgin females with fresh 4 mL food to ensure experimental flies of equal treatment and age. 2
Journal of Insect Physiology 118 (2019) 103946
P.V. Salachan, et al.
Table 1 Effect of acclimation temperature on CTmax at each ramping rate. Assays were performed on 7 day acclimated adult female flies. Acclimation temperature corresponded to 15, 19, 25 and 29 °C. Model estimates were for either linear (single value, poly1) or second order polynomial fits (double value, poly1 & poly2) depending on the best fits for each ramping rate. ***P < 0.001, **P < 0.01, *P < 0.05.
Table 2 Effect of acclimation temperature on TKD at each KD temperature. Assays were performed on 7 day acclimated adult female flies. Acclimation temperature corresponded to 15, 19, 25 and 29 °C. Model estimates were for either linear (single value, poly1) or second order polynomial fits (double value, poly1 & poly2) depending on the best fits for each knock-down temperature. ***P < 0.001, **P < 0.01, *P < 0.05.
Ramping rate (°C/min)
Acclimation temperature (χ2) Model estimate (1poly1, 2 poly2)
Knock-down temperature (°C)
0.02
0.05
0.2
0.5
54.7***
20.93***
4.82*
1.59
−0.961, 0.0222
−0.391, 0.0092
0.0241
−0.0141
Acclimation temperature (χ2) Model Estimate (1poly1, 2 poly2)
36
37
38
39
40
121.8***
13.1***
40.2***
44.1***
24.7***
−149.61, 3.82
−29.11, 0.72
2.31
1.21
0.21
3. Results For TKD, the ANOVA analysis showed a significant negative effect of KD temperature on TKD (higher KD temperature yielded lower TKD, χ2 = 71.5, P < 0.001). However, we found significant effects of both acclimation temperature (χ2 = 112.3, P < 0.001) and the interaction between KD temperature and acclimation temperature on TKD (χ2 = 4.3, P < 0.05). To investigate the effects of acclimation temperature further, we analyzed the effects of acclimation temperature within each KD temperature. For the three highest KD temperatures (38, 39 and 40 °C), we found significant linear positive relationships between TKD and acclimation temperature (with decreasing slopes for increasing KD temperatures), while second order polynomial models provided superior fits to the data for the two lowest KD temperatures (36 and 37 °C) (Table 2, Fig. 2). For the KD temperature of 36 and 37 °C, TKD was non-linearly related to acclimation temperature with increasing TKD estimates at acclimation temperatures above or below 19 °C. With lower KD temperature the non-linearity of the acclimation response in TKD increased.
3.1. Thermal tolerance For CTmax, ANOVA analysis showed a significant positive effect of ramping rate on CTmax (faster rates lead to higher CTmax estimates, χ2 = 56.1, P < 0.001). ANOVA analysis also showed a significant (non-trivial) effect of acclimation temperature (χ2 = 15.4, P < 0.001), and a significant effect of the interaction between ramping rate and acclimation temperature on CTmax (χ2 = 5.8, P < 0.05). Due to the significant interaction and as we were particularly interested in the effect of acclimation temperature (patterns of responses rather than absolute differences), we tested the effect of acclimation temperature within each ramping rate. For all the ramping rates except for the fastest (0.5 °C/min), we found significant ramping rate specific effects of acclimation temperature (Table 1; Fig. 1). For the ramping rate of 0.2 °C/min CTmax increased linearly with acclimation temperature. For the ramping rates of 0.05 and 0.02 °C/min, CTmax was non-linearly related to acclimation temperature with increasing CTmax estimates at acclimation temperatures above or below 19 °C. The non-linearity of the acclimation response in CTmax increased as ramping rate decreased.
Fig. 2. Time to knock-down (TKD) for 7 day acclimated adult female flies at different knock-down temperatures. Symbols and error bars represent mean ± s.e.m TKD for the four acclimation temperatures; 15, 19, 25 and 29 °C. Grey lines show the reaction norms for linear fits (TKD at 38, 39 and 40 °C). Black lines show reaction norms for second order polynomial fits (TKD at 36 and 37 °C). N = 30 for all data points (N = 29 for 15 °C acclimation and knockdown at 40 °C). Post-hoc pair-wise differences are given in the electronic supplementary material, Table S1.
Fig. 1. Critical thermal maximum (CTmax) for 7 day acclimated adult female flies at different ramping rates. Symbols and error bars represent mean ± s.e.m CTmax for the four acclimation temperatures; 15, 19, 25 and 29 °C. Grey lines represent the reaction norms for linear fits (for ramping rates of 0.5 and 0.2 °C/ min) and black lines the reaction norms for second order polynomial fits (for ramping rates of 0.05 and 0.02 °C/min). N = 30 for all data points except the ramping rate of 0.02 °C/min where N = 40. Post-hoc pair-wise differences are given in the electronic supplementary material, Table S1. 3
Journal of Insect Physiology 118 (2019) 103946
P.V. Salachan, et al.
4. Discussion
However, under typical assay conditions, empirical work has dismissed the concerns regarding desiccation and starvation for particular temperatures and durations (Manenti et al., 2018; Overgaard et al., 2012). Differences in estimates of thermal plasticity may simply result from different assay conditions or treatment/sampling resolution as opposed to different underlying biological mechanisms. Even if assays do offer different scope for heat hardening, Sørensen et al. (2013) have shown little absolute effect of heat hardening on thermal tolerance during an assay. Furthermore, the results of this study do not suggest that the two types of thermal assays (static or dynamic) used reflect fundamentally different traits, e.g. the 29 and 15 °C acclimation treatments produced the most heat tolerant phenotypes for both assays (when low rates or low KD temperatures were used). Thus, according to this interpretation, the use of lower ramping rates or lower KD temperatures offers better resolution for detecting acclimation effects and should be the preferred approach. While this study does not allow a definite conclusion regarding the ‘correct’ estimates of acclimation capacity, we do suggest that the most ecologically relevant assay condition should reflect the conditions (temperatures and rates) organisms are exposed to in their natural habitats. A few studies have taken this into consideration while estimating thermal limits. For example, Mitchell et al. (2011) used a ramping rate of 0.06 °C/min which was typical of naturally occurring temperature changes in orchards around inland southern Australia. Similarly, Kelty and Lee (2001) modeled thermo-periods based on field recordings corresponding to rates of increase of 0.03 °C/min or decrease of 0.02 °C/min on average. Many studies estimating CTmax rely on a ramping rate of 0.1 °C/min (Overgaard et al., 2011) assumed to be ecologically relevant or a useful compromise between ecological relevance and preventing impractical long assays (see discussion in Sinclair et al., 2015).
We confirmed that lower ramping rates or higher KD temperatures result in decreased estimates of thermal tolerance (Machekano et al., 2018; Sørensen et al., 2013). Surprisingly, the result for individual ramping rates or KD temperatures revealed progressively non-linear reaction norms as ramping rates or KD temperatures decreased. The strength of the acclimation response (slope) as well as its non-linearity both increased with progressively slower ramping rates or lower KD temperatures (i.e. with longer assays). Thus, the strong linearity of developmental acclimation responses found across Drosophila species by Schou et al. (2016) might be restricted to developmental acclimation and/or to the assay condition of their study (ramping of 0.1 °C/min). In this study on D. melanogaster, a ramping rate of 0.5 °C/min produced a linear reaction norm with no benefit of acclimation, while a rate of 0.05 °C/min produced a much stronger, but non-linear acclimation response. Even if the absolute changes in thermal tolerance were limited (e.g. with a change in the CTmax of about 1 °C or less), it likely increases survival rates to biologically relevant high temperatures (e.g. increasing heat shock temperatures by 1 °C decreased survival rates from above 50% to nearly 0%; Kristensen et al., 2008). Many studies predict acclimation capacity based on two or three acclimation temperatures, a single assay-condition (ramping rate or KD temperature) and assume linearity (see e.g. data used in Gunderson and Stillman, 2015), these estimates might not be indicative of realized and ecologically relevant acclimation capacities. The ramping rate dependent shape of the reaction norms that we observe implies that alternate acclimation hypothesis may be supported depending on the ramping rate and acclimation temperatures selected. For example (at a ramp rate of 0.05 °C/min), one might conclude that “colder is better” if only comparing acclimation temperatures of 15 and 19 °C, but would conclude that “hotter is better” if comparing acclimation temperatures of 25 and 29 °C (see Deere and Chown, 2006, for a discussion of the different acclimation hypotheses). Further, they may conclude that acclimation has no effect if comparing acclimation temperatures of 15 and 25 °C. Finally, comparing acclimation temperatures of 15 and 29 °C might lead to observations of cross-tolerance in thermal performance (Sejerkilde et al., 2003). The choice of assay condition likewise affects the estimates of acclimation capacity. Using fast assays (high KD temperature or fast ramping rates) leads to low estimates of acclimation capacity (flat reaction norm slopes) as compared to longer assays. Thus, results from studies assuming linearity between two acclimation temperatures might be strongly influenced by the acclimation and assay conditions selected. This highlights that caution has to be exercised when extrapolating thermal tolerance reaction norms from a few acclimation temperatures. If the patterns of acclimation responses detected here are applicable across species, it has wide implications for the interpretation of acclimation capacity estimates and their ecological significance. Given the discrepancy in thermal tolerance estimates between various assay conditions, a natural question arises as to which condition represents the ‘correct’ acclimation response for thermal tolerance. Different ramping rates or KD temperatures likely measure different traits. Support for such an interpretation is found in the literature, as part of the inherent time-temperature interaction in thermal assays. Empirical studies suggest that discrepancies between ramping assays and static knock-down assays exist and might at least partly be explained by the higher potential for heat hardening (short term acclimation) during ramping assays (Overgaard et al., 2011; Overgaard et al., 2012; Sgrò et al., 2010; Terblanche et al., 2007). Such heat hardening potential would be increasingly pronounced at lower ramping rates, and likely also at lower KD temperatures (as both give rise to longer assays). Alternatively, Rezende et al. (2011) proposed that lower ramping rates offer increasing scope for factors such as desiccation and starvation to affect thermal tolerance estimates and should be avoided by using the highest rates or temperatures possible (see further discussion in Santos et al., 2011; Rezende, Tejedo, and Santos 2011).
5. Conclusion Here we investigated how adult acclimation to constant temperature modulates the reaction norms for high temperature thermal tolerance in Drosophila melanogaster. Our results highlight that the estimate of thermal tolerance and thermal acclimation capacity will typically reflect the assay condition and the history of the organisms assayed (Chown et al., 2009; Terblanche et al., 2007) and that acclimation capacity might be underestimated in many studies using relatively fast assays. Tolerance estimates based on high ramping rates or high KD temperatures might reflect partly different aspects of thermal tolerance as compared to low ramping rates or low KD temperatures. Although the tolerance estimates of dynamic and static assays yield results with different metric (temperature or time, respectively), both assays exhibit similar patterns of the acclimation responses and might be mechanistically related (Jørgensen et al., 2019). Strikingly, the lower ramping rates and lower KD temperatures led to increasingly non-linear reaction norms and stronger acclimation responses, presumably partly due to the better resolution offered at relatively 'mild' assay conditions. Our results and their consequences might not be restricted to drosophilids as they could apply more widely to ectotherms where thermal tolerances are estimated using similar dynamic or static assays. Future studies aiming at determining the limits for thermal tolerance should therefore carefully select assay conditions based on the species, it’s thermal niche and the rates of thermal changes that it experiences or that which might be ecologically relevant for that species. Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. 4
Journal of Insect Physiology 118 (2019) 103946
P.V. Salachan, et al.
Acknowledgements
Kristensen, T.N., Hoffmann, A.A., Overgaard, J., Sorensen, J.G., Hallas, R., Loeschcke, V., 2008. Costs and benefits of cold acclimation in field-released Drosophila. Proc. Natl. Acad. Sci. 105 (1), 216–221. https://doi.org/10.1073/pnas.0708074105. Leroi, A.M., Bennett, A.F., Lenski, R.E., 1994. Temperature acclimation and competitive fitness: an experimental test of the beneficial acclimation assumption. Proc. Natl. Acad. Sci. U.S.A. 91, 1917–1921. https://doi.org/10.1073/pnas.91.5.1917. Lutterschmidt, W.I., Hutchison, V.H., 1997. The critical thermal maximum: history and critique. Can. J. Zool. 75, 1561–1574. https://doi.org/10.1139/z97-783. Machekano, H., Mvumi, B.M., Nyamukondiwa, C., 2018. Loss of coevolved basal and plastic responses to temperature may underlie trophic level host-parasitoid interactions under global change. Biol. Control 118, 44–54. https://doi.org/10.1016/j. biocontrol.2017.12.005. Manenti, T., Cunha, T.R., Sørensen, J.G., Loeschcke, V., 2018. How much starvation, desiccation and oxygen depletion can Drosophila melanogaster tolerate before its upper thermal limits are affected? J. Insect Physiol. 111, 1–7. https://doi.org/10. 1016/j.jinsphys.2018.09.002. Martin, T.L., Huey, R.B., 2008. Why “suboptimal” is optimal: Jensen's inequality and ectotherm thermal preferences. Am. Nat. 171, 102–118. https://doi.org/10.1086/ 527502. Mitchell, K.A., Sgrò, C.M., Hoffmann, A.A., 2011. Phenotypic plasticity in upper thermal limits is weakly related to Drosophila species distributions. Funct. Ecol. 25, 661–670. https://doi.org/10.1111/j.1365-2435.2010.01821.x. Overgaard, J., Kearney, M.R., Hoffmann, A.A., 2014. Sensitivity to thermal extremes in Australian Drosophila implies similar impacts of climate change on the distribution of widespread and tropical species. Glob. Change Biol. 20, 1738–1750. https://doi.org/ 10.1111/gcb.12521. Overgaard, J., Kristensen, T.N., Mitchell, K.A., Hoffmann, A.A., 2011. Thermal tolerance in widespread and tropical Drosophila species: does phenotypic plasticity increase with latitude? Am. Natural. 178 (Suppl 1), 80–96. https://doi.org/10.1086/661780. Overgaard, J., Kristensen, T.N., Sørensen, J.G., 2012. Validity of thermal ramping assays used to assess thermal tolerance in arthropods. PLoS One 7, 1–7. https://doi.org/10. 1371/journal.pone.0032758. R Core Team, 2018. R: A Language and Environment for Statistical Computing: R Foundation for Statistical Computing. Rezende, E.L., Castañeda, L.E., Santos, M., Fox, C., 2014. Tolerance landscapes in thermal ecology. Funct. Ecol. 28, 799–809. https://doi.org/10.1111/1365-2435.12268. Rezende, E.L., Tejedo, M., Santos, M., 2011. Estimating the adaptive potential of critical thermal limits: methodological problems and evolutionary implications. Funct. Ecol. 25, 111–121. https://doi.org/10.1111/j.1365-2435.2010.01778.x. Santos, M., Castañeda, L.E., Rezende, E.L., 2011. Making sense of heat tolerance estimates in ectotherms: lessons from Drosophila. Funct. Ecol. 25, 1169–1180. https://doi.org/ 10.1111/j.1365-2435.2011.01908.x. Schou, M.F., Loeschcke, V., Kristensen, T.N., 2015. Strong costs and benefits of winter acclimatization in Drosophila melanogaster. PLoS One 10, 1–13. https://doi.org/10. 1371/journal.pone.0130307. Schou, M.F., Mouridsen, M.B., Sørensen, J.G., Loeschcke, V., 2016. Linear reaction norms of thermal limits in Drosophila: predictable plasticity in cold but not in heat tolerance. Funct. Ecol. 31, 934–945. https://doi.org/10.1111/1365-2435.12782. Schulte, P.M., Healy, T.M., Fangue, N.A., 2011. Thermal performance curves, phenotypic plasticity, and the time scales of temperature exposure. Integr. Comp. Biol. 51, 691–702. https://doi.org/10.1093/icb/icr097. Sejerkilde, M., Sørensen, J.G., Loeschcke, V., 2003. Effects of cold- and heat hardening on thermal resistance in Drosophila melanogaster. J. Insect Physiol. 49, 719–726. https:// doi.org/10.1016/s0022-1910(03)00095-7. Sgrò, C.M., Overgaard, J., Kristensen, T.N., Mitchell, K.A., Cockerell, F.E., Hoffmann, A.A., 2010. A comprehensive assessment of geographic variation in heat tolerance and hardening capacity in populations of Drosophila melanogaster from eastern Australia. J. Evolution. Biol. 23, 2484–2493. https://doi.org/10.1111/j.1420-9101. 2010.02110.x. Sgrò, C.M., Terblanche, J.S., Hoffmann, A.A., 2016. What can plasticity contribute to insect responses to climate change? Annu. Rev. Entomol. 61, 433–451. https://doi. org/10.1146/annurev-ento-010715-023859. Sinclair, B.J., Coello Alvarado, L.E., Ferguson, L.V., 2015. An invitation to measure insect cold tolerance: methods, approaches, and workflow. J. Therm. Biol. 53, 180–197. https://doi.org/10.1016/j.jtherbio.2015.11.003. Sørensen, J.G., Kristensen, T.N., Overgaard, J., 2016a. Evolutionary and ecological patterns of thermal acclimation capacity in Drosophila: is it important for keeping up with climate change? Current Opinion in Insect. Science 17, 98–104. https://doi.org/ 10.1016/j.cois.2016.08.003. Sørensen, J.G., Loeschcke, V., Kristensen, T.N., 2013. Cellular damage as induced by high temperature is dependent on rate of temperature change – investigating consequences of ramping rates on molecular and organismal phenotypes in Drosophila melanogaster. J. Exp. Biol. 216, 809–814. https://doi.org/10.1242/jeb.076356. Sørensen, J.G., Schou, M.F., Kristensen, T.N., Loeschcke, V., 2016b. Thermal fluctuations affect the transcriptome through mechanisms independent of average temperature. Sci. Rep. 6, 1–11. https://doi.org/10.1038/srep30975. Teets, N.M., Hahn, D.A., 2018. Genetic variation in the shape of cold-survival curves in a single fly population suggests potential for selection from climate variability. J. Evol. Biol. 31, 543–555. https://doi.org/10.1111/jeb.13244. Terblanche, J.S., Deere, J.A., Clusella-Trullas, S., Janion, C., Chown, S.L., 2007. Critical thermal limits depend on methodological context. Proc. Roy. Soc. B: Biol. Sci. 274, 2935–2942. https://doi.org/10.1098/rspb.2007.0985.
We would like to thank Heidi J. MacLean and Mads F. Schou for fruitful discussions on data analysis and manuscript writing. Thanks to Daniel A. Hahn and anonymous referees for constructive criticism on previous versions of this manuscript. This work was supported by a Starting Grant from Aarhus Universitets Forskningsfond to JGS (AUFFE-2015-FLS-8-72). Author contributions JGS conceived the idea and designed the study, HB and PVS collected the data, PVS and JGS analyzed the data, PVS made the figures. PVS wrote the original draft, JGS edited and reviewed the draft. All authors read and approved the final manuscript. Data accessibility Data have been deposited in the Dryad digital repository: https:// doi.org/10.5061/dryad.5jn61nt. Appendix A. Supplementary data Supplementary data to this article can be found online at https:// doi.org/10.1016/j.jinsphys.2019.103946. References Allen, J.L., Chown, S.L., Janion-Scheepers, C., Clusella-Trullas, S., 2016. Interactions between rates of temperature change and acclimation affect latitudinal patterns of warming tolerance. Conserv. Physiol. 4, 1–14. https://doi.org/10.1093/conphys/ cow053. Angilletta, M.J., 2009. In: Thermal Adaptation: A Theoretical and Empirical Synthesis. Oxford University Press, Oxford, UK. https://doi.org/10.1093/acprof:oso/ 9780198570875.001.1. Araújo, M.B., Ferri-Yanez, F., Bozinovic, F., Marquet, P.A., Valladares, F., Chown, S.L., 2013. Heat freezes niche evolution. Ecol. Lett. 16, 1206–1219. https://doi.org/10. 1111/ele.12155. Bates, D., Mächler, M., Bolker, B., Walker, S., 2015. Fitting linear mixed-effects models using lme4. J. Statist. Software 67. https://doi.org/10.18637/jss.v067.i01. Chown, S.L., Jumbam, K.R., Sørensen, J.G., Terblanche, J.S., 2009. Phenotypic variance, plasticity and heritability estimates of critical thermal limits depend on methodological context. Funct. Ecol. 23, 133–140. https://doi.org/10.1111/j.1365-2435.2008. 01481.x. Deere, J.A., Chown, S.L., 2006. Testing the beneficial acclimation hypothesis and its alternatives for locomotor performance. Am. Nat. 168, 630–644. https://doi.org/10. 1086/508026. Gunderson, A.R., Stillman, J.H., 2015. Plasticity in thermal tolerance has limited potential to buffer ectotherms from global warming. Proc. Roy. Soc. B: Biol. Sci. 282, 1–8. https://doi.org/10.1098/rspb.2015.0401. Hoffmann, A.A., Chown, S.L., Clusella-Trullas, S., 2013. Upper thermal limits in terrestrial ectotherms: how constrained are they? Funct. Ecol. 27, 934–949. https://doi.org/10. 1111/j.1365-2435.2012.02036.x. Hoffmann, A.A., Sørensen, J.G., Loeschcke, V., 2003. Adaptation of Drosophila to temperature extremes: bringing together quantitative and molecular approaches. J. Therm. Biol 28, 175–216. https://doi.org/10.1016/s0306-4565(02)00057-8. Huey, R.B., Berrigan, D., 2001. Temperature, demography, and ectotherm fitness. Am. Nat. 158, 204–210. https://doi.org/10.1086/321314. Huey, R.B., Kingsolver, J.G., 1989. Evolution of thermal sensitivity of ectotherm performance. Trends Ecol. Evol. 4, 131–135. https://doi.org/10.1016/0169-5347(89) 90211-5. Huey, R.B., Stevenson, R.D., 1979. Integrating thermal physiology and ecology of ectotherms: a discussion of approaches. Am. Zool. 19, 357–366. https://doi.org/10. 1093/icb/19.1.357. Jørgensen, L.B., Malte, H., Overgaard, J., 2019. How to assess Drosophila heat tolerance: unifying static and dynamic tolerance assays to predict heat distribution limits. Funct. Ecol. 33, 629–642. https://doi.org/10.1111/1365-2435.13279. Kellermann, V., Overgaard, J., Hoffmann, A.A., Flojgaard, C., Svenning, J.C., Loeschcke, V., 2012. Upper thermal limits of Drosophila are linked to species distributions and strongly constrained phylogenetically. Proc. Natl. Acad. Sci. U.S.A. 109, 16228–16233. https://doi.org/10.1073/pnas.1207553109. Kelty, J.D., Lee, R.E., 2001. Rapid cold-hardening of Drosophila melanogaster (Diptera: Drosophilidae) during ecologically based thermoperiodic cycles. J. Experim. Biol. 204, 1659–1666.
5
Contents lists available at ScienceDirect
Journal of Insect Physiology journal homepage: www.elsevier.com/locate/jinsphys
Testing the thermal limits: Non-linear reaction norms drive disparate thermal acclimation responses in Drosophila melanogaster
T
⁎
Paul Vinu Salachan , Hélène Burgaud, Jesper Givskov Sørensen Department of Bioscience, Aarhus University, Ny Munkegade 116, 8000 Aarhus C, Denmark
A R T I C LE I N FO
A B S T R A C T
Keywords: Adult acclimation CTmax Time to knock-down Phenotypic plasticity Thermal tolerance
Critical thermal limits are important ecological parameters for studying thermal biology and for modelling species’ distributions under current and changing climatic conditions (including predicting the risk of extinction for species from future warming). However, estimates of the critical thermal limits are biased by the choice of assay and assay conditions, which differ among studies. Furthermore, estimates of the potential for phenotypic plasticity (thermal acclimation) to buffer against thermal variability are usually based on single assay conditions and (usually linear) extrapolation from a few acclimation temperatures. We produced high resolution estimates of adult acclimation capacity for upper tolerance limits at different assay conditions (ramping rates and knockdown temperatures) using CTmax (dynamic) and knock-down (static) thermal assays in the model species Drosophila melanogaster. We found the reaction norms to be highly dependent on assay conditions. We confirmed that progressively lower ramping rates or higher knock-down temperatures led to overall lower tolerance estimates. More surprisingly, extended assays (lower ramping rates or lower knock-down temperatures) also led to increasingly non-linear reaction norms for upper thermal tolerance across adult acclimation temperatures. Our results suggest that the magnitude (capacity) and direction (beneficial or detrimental) of acclimation responses are highly sensitive to assay conditions. The results offer a framework for comparison of acclimation responses between different assay conditions and a potential for explaining disparate acclimation capacity theories. We advocate cautious interpretation of acclimation capacities and careful consideration of assay conditions, which should represent realistic environmental conditions based on species’ ecological niches.
1. Introduction
short-term (hardening) and long-term (acclimation) temperature exposure (Hoffmann et al., 2003). This calls to question the role of adaptive phenotypic plasticity (thermal acclimation) for mitigating detrimental effects of stressful environmental temperatures (Gunderson and Stillman, 2015; Sgrò et al., 2016; Sørensen et al., 2016a). To reliably use estimates of thermal tolerance in general and acclimation capacity specifically, our measurements need to be accurate and ecologically relevant (i.e. relevant for the species and its thermal niche). Estimates of thermal limits (and of thermal plasticity) can be influenced by the method used (Rezende et al., 2014; Sørensen et al., 2016b). There are two primary methods of assessing upper thermal tolerance in insects: a static knock-down (KD) assay and a dynamic ramping (CTmax) assay (Lutterschmidt and Hutchison, 1997). In a static assay, individuals are kept at a constant stressful temperature and the time required to reach an end-point (usually the cessation of movement) is measured as the time to knock-down (TKD) at that temperature. In a dynamic assay, the temperature is increased at a predetermined rate from a benign temperature and the temperature at
Temperature can have a profound effect on the survival and fitness of organisms, prompting various strategies for adapting to variation in diel, seasonal, and annual mean temperatures. The thermal biology of organisms (e.g. adaptive differences among species and populations) can be studied via parameters of a thermal performance curve (Huey and Berrigan, 2001; Martin and Huey, 2008). Specifically, the lower and upper limits of the curve (thermal tolerance) as well as the peak and breadth (optimal temperature and thermal range) characterize the thermal performance curve (Huey and Kingsolver, 1989; Huey and Stevenson, 1979). For many ectotherms, lower- and upper critical thermal limits (CTmin and CTmax) have been extensively studied due to the relative ease of investigating these endpoints, and the estimates have been used as predictors of species distribution (Kellermann et al., 2012; Overgaard et al., 2014). However, the upper thermal limits are considered evolutionarily constrained (Araújo et al., 2013; Hoffmann et al., 2013; Kellermann et al., 2012), but does respond adaptively to
⁎
Corresponding author. E-mail address: [email protected] (P.V. Salachan).
https://doi.org/10.1016/j.jinsphys.2019.103946 Received 2 June 2019; Received in revised form 12 September 2019; Accepted 13 September 2019 Available online 13 September 2019 0022-1910/ © 2019 Elsevier Ltd. All rights reserved.
Journal of Insect Physiology 118 (2019) 103946
P.V. Salachan, et al.
2.2. Adult acclimation regimes
which the individual ceases movement is recorded as the CTmax. The potential for heat hardening (rapidly induced beneficial physiological adjustments) during such assays likely increases with assay duration and might explain some discrepancies among static assays (usually performed at relatively high temperatures) and more prolonged (dynamic) thermal assays (Sgrò et al., 2010). Furthermore, results are assay condition dependent with lower ramping rates or higher KD temperatures decreasing the thermal tolerance estimates (Allen et al., 2016; Machekano et al., 2018; Sørensen et al., 2013). The effect of methodology is most likely driven by the interaction between time and temperature leading to increased accumulation of heat damage in longer assays (Sørensen et al., 2013), although confounding effects of desiccation, starvation etc. have also been proposed (Santos et al., 2011). Consequently, while some argue that fast assays (high KD temperature or high ramping rates) are less likely to be compromised by confounding effects (Rezende et al., 2011), lower ramping rates (and lower KD temperatures) might be considered more ecologically relevant (Overgaard et al., 2012). Thermal traits likely harbor different levels of plasticity and reaction norms (Schulte et al., 2011; Teets and Hahn, 2018). Most studies do not verify the shape of the reaction norms but assume linear acclimation responses based on few acclimation temperatures (see e.g. Gunderson and Stillman, 2015). Studies on insect acclimation often distinguish between developmental and adult acclimation (Angilletta, 2009). For constant developmental acclimation temperatures, Drosophila species can have different (linear) reaction norms for thermal limits (Schou et al., 2016). To our knowledge, no study has systematically looked at the reaction norms for thermal tolerance over multiple adult acclimation temperatures and in connection with assay conditions in both ramping and static assays. Thus, no clear consensus exists as to how thermal tolerance assays affect the detection of ecologically relevant and adaptive acclimation responses (Overgaard et al., 2012; Rezende et al., 2011) and how this might affect predictions about acclimation capacity (Sørensen et al., 2016a). To investigate assay-dependent adult acclimation responses of D. melanogaster, we subjected flies to various thermal performance assays following adult acclimation to one of four different temperatures (15, 19, 25 or 29 °C). We assayed thermal tolerance using two different assays; a static KD assay and a dynamic ramping assay. To tease apart any differences associated with assay condition, we applied four different ramping rates (between 0 and 0.5 °C/min) and five different KD temperatures (between 36 and 40 °C) to cover a broad range of conditions and assay durations. We expect lower ramping rates or higher KD temperatures to result in lower estimates of thermal tolerance (Machekano et al., 2018; Sørensen et al., 2013). However, here we focus specifically on the assay condition and its interaction with the acclimation temperature, in order to compare patterns of acclimation responses in tolerance estimates, to generate a general framework for understanding insect acclimation capacity. Following the beneficial acclimation hypothesis (Leroi et al., 1994), and in accordance with the results of Schou et al. (2016), we expect heat tolerance to be positively and linearly related to acclimation temperature.
The collected vials from each batch of flies were randomly distributed among four constant acclimation temperatures (15, 19, 25 and 29 °C) and maintained there for seven days. We used data loggers (iButton) to verify the desired temperatures. Every second day, flies were transferred to fresh food vials.
2.3. Thermal assays We investigated the tolerance to high temperatures in acclimated adult flies (7 days old) using a dynamic or a static heat tolerance measurement. For both thermal assays, flies from all four acclimation treatments (from a single batch of collected eggs) were always assayed together. Thermal assays were conducted in a temperature-controlled water bath (glass aquarium) with a circulating pump to maintain the same temperature throughout the water bath. Flies were placed in small screw cap glass vials on a rack without the use of anesthesia and submerged in the water bath. For the dynamic assay, the temperature was increased from 25 °C at a rate of 0.02, 0.05, 0.2, or 0.5 °C/min until the flies were no longer responsive after mild tapping with a rod and stimulation with a flash light (i.e. they had reached the CTmax). For the static (KD) assay, flies were exposed to 36, 37, 38, 39 or 40 °C, until they were no longer responsive. For both types of assay, the flies were placed in the tube immediately prior to each assay and the maximal assay duration was on average around 12 h (for the slowest ramping rate of 0.02 °C/min) or 7.5 h (for the lowest KD temperature of 36 °C). At a 12-hour assay duration, we expect no or negligible confounding effects of starvation or desiccation on thermal tolerance for all assay temperatures (see Manenti et al., 2018). In each assay, ten females from each acclimation treatment were used for each experimental block. Three experimental blocks (for a total of 30 female flies) were assayed for each KD temperature or ramping rate, except for the ramping rate of 0.02 °C/min, where two blocks of 20 flies (for a total of 40 female flies) were assayed.
2.4. Statistical analysis To investigate the effect of acclimation temperature and assay condition on the thermal tolerance estimates, we performed ANOVA based on linear mixed effects modelling using the ‘lme4’ package (v.1.15) (Bates et al., 2015) in R (R Core Team, 2018). Q-Q and residual plots were used to verify the assumptions for normality of residuals and homogeneity of variances. Both acclimation temperature and ramping rate or KD temperature were considered as continuous variables. The data were transformed appropriately in cases where the assumptions were not met. Sequential model reduction was performed, and the Chisquare values were reported for the effect of each dropped term (using ‘anova’ function, R ‘stats’ package). For both estimates of thermal tolerance (CTmax and TKD), experimental block was included as a random factor. To analyze the effect of acclimation temperature and ramping rate on CTmax, we considered both predictor variables and their interaction in the model. The ramping rate was log transformed for the analysis. Acclimation and KD temperatures and their interaction were included as predictor variables in the analysis of log-transformed TKD. Inspection of the data revealed that the significant interactions between acclimation temperature and KD temperature/ramping rate in both full models were caused by different reaction norms among conditions (ramping rates or KD temperatures) within assays. Based on this we analyzed the data within each individual ramping rate or KD temperature and investigated reaction norms in each.
2. Materials and methods 2.1. Experimental animals We used a population of D. melanogaster collected in 2013, from Odder, Denmark (Schou et al., 2015). The population was maintained in the laboratory at 25 °C (12L:12D) on standard Drosophila oatmeal–sugar–yeast–agar medium. All experimental animals were raised under controlled density conditions by transferring approximately 40 ( ± 3) eggs into fresh 7 mL food vials. Upon emergence, flies were separated by sex under CO2 anesthesia and kept in vials of 20 virgin females with fresh 4 mL food to ensure experimental flies of equal treatment and age. 2
Journal of Insect Physiology 118 (2019) 103946
P.V. Salachan, et al.
Table 1 Effect of acclimation temperature on CTmax at each ramping rate. Assays were performed on 7 day acclimated adult female flies. Acclimation temperature corresponded to 15, 19, 25 and 29 °C. Model estimates were for either linear (single value, poly1) or second order polynomial fits (double value, poly1 & poly2) depending on the best fits for each ramping rate. ***P < 0.001, **P < 0.01, *P < 0.05.
Table 2 Effect of acclimation temperature on TKD at each KD temperature. Assays were performed on 7 day acclimated adult female flies. Acclimation temperature corresponded to 15, 19, 25 and 29 °C. Model estimates were for either linear (single value, poly1) or second order polynomial fits (double value, poly1 & poly2) depending on the best fits for each knock-down temperature. ***P < 0.001, **P < 0.01, *P < 0.05.
Ramping rate (°C/min)
Acclimation temperature (χ2) Model estimate (1poly1, 2 poly2)
Knock-down temperature (°C)
0.02
0.05
0.2
0.5
54.7***
20.93***
4.82*
1.59
−0.961, 0.0222
−0.391, 0.0092
0.0241
−0.0141
Acclimation temperature (χ2) Model Estimate (1poly1, 2 poly2)
36
37
38
39
40
121.8***
13.1***
40.2***
44.1***
24.7***
−149.61, 3.82
−29.11, 0.72
2.31
1.21
0.21
3. Results For TKD, the ANOVA analysis showed a significant negative effect of KD temperature on TKD (higher KD temperature yielded lower TKD, χ2 = 71.5, P < 0.001). However, we found significant effects of both acclimation temperature (χ2 = 112.3, P < 0.001) and the interaction between KD temperature and acclimation temperature on TKD (χ2 = 4.3, P < 0.05). To investigate the effects of acclimation temperature further, we analyzed the effects of acclimation temperature within each KD temperature. For the three highest KD temperatures (38, 39 and 40 °C), we found significant linear positive relationships between TKD and acclimation temperature (with decreasing slopes for increasing KD temperatures), while second order polynomial models provided superior fits to the data for the two lowest KD temperatures (36 and 37 °C) (Table 2, Fig. 2). For the KD temperature of 36 and 37 °C, TKD was non-linearly related to acclimation temperature with increasing TKD estimates at acclimation temperatures above or below 19 °C. With lower KD temperature the non-linearity of the acclimation response in TKD increased.
3.1. Thermal tolerance For CTmax, ANOVA analysis showed a significant positive effect of ramping rate on CTmax (faster rates lead to higher CTmax estimates, χ2 = 56.1, P < 0.001). ANOVA analysis also showed a significant (non-trivial) effect of acclimation temperature (χ2 = 15.4, P < 0.001), and a significant effect of the interaction between ramping rate and acclimation temperature on CTmax (χ2 = 5.8, P < 0.05). Due to the significant interaction and as we were particularly interested in the effect of acclimation temperature (patterns of responses rather than absolute differences), we tested the effect of acclimation temperature within each ramping rate. For all the ramping rates except for the fastest (0.5 °C/min), we found significant ramping rate specific effects of acclimation temperature (Table 1; Fig. 1). For the ramping rate of 0.2 °C/min CTmax increased linearly with acclimation temperature. For the ramping rates of 0.05 and 0.02 °C/min, CTmax was non-linearly related to acclimation temperature with increasing CTmax estimates at acclimation temperatures above or below 19 °C. The non-linearity of the acclimation response in CTmax increased as ramping rate decreased.
Fig. 2. Time to knock-down (TKD) for 7 day acclimated adult female flies at different knock-down temperatures. Symbols and error bars represent mean ± s.e.m TKD for the four acclimation temperatures; 15, 19, 25 and 29 °C. Grey lines show the reaction norms for linear fits (TKD at 38, 39 and 40 °C). Black lines show reaction norms for second order polynomial fits (TKD at 36 and 37 °C). N = 30 for all data points (N = 29 for 15 °C acclimation and knockdown at 40 °C). Post-hoc pair-wise differences are given in the electronic supplementary material, Table S1.
Fig. 1. Critical thermal maximum (CTmax) for 7 day acclimated adult female flies at different ramping rates. Symbols and error bars represent mean ± s.e.m CTmax for the four acclimation temperatures; 15, 19, 25 and 29 °C. Grey lines represent the reaction norms for linear fits (for ramping rates of 0.5 and 0.2 °C/ min) and black lines the reaction norms for second order polynomial fits (for ramping rates of 0.05 and 0.02 °C/min). N = 30 for all data points except the ramping rate of 0.02 °C/min where N = 40. Post-hoc pair-wise differences are given in the electronic supplementary material, Table S1. 3
Journal of Insect Physiology 118 (2019) 103946
P.V. Salachan, et al.
4. Discussion
However, under typical assay conditions, empirical work has dismissed the concerns regarding desiccation and starvation for particular temperatures and durations (Manenti et al., 2018; Overgaard et al., 2012). Differences in estimates of thermal plasticity may simply result from different assay conditions or treatment/sampling resolution as opposed to different underlying biological mechanisms. Even if assays do offer different scope for heat hardening, Sørensen et al. (2013) have shown little absolute effect of heat hardening on thermal tolerance during an assay. Furthermore, the results of this study do not suggest that the two types of thermal assays (static or dynamic) used reflect fundamentally different traits, e.g. the 29 and 15 °C acclimation treatments produced the most heat tolerant phenotypes for both assays (when low rates or low KD temperatures were used). Thus, according to this interpretation, the use of lower ramping rates or lower KD temperatures offers better resolution for detecting acclimation effects and should be the preferred approach. While this study does not allow a definite conclusion regarding the ‘correct’ estimates of acclimation capacity, we do suggest that the most ecologically relevant assay condition should reflect the conditions (temperatures and rates) organisms are exposed to in their natural habitats. A few studies have taken this into consideration while estimating thermal limits. For example, Mitchell et al. (2011) used a ramping rate of 0.06 °C/min which was typical of naturally occurring temperature changes in orchards around inland southern Australia. Similarly, Kelty and Lee (2001) modeled thermo-periods based on field recordings corresponding to rates of increase of 0.03 °C/min or decrease of 0.02 °C/min on average. Many studies estimating CTmax rely on a ramping rate of 0.1 °C/min (Overgaard et al., 2011) assumed to be ecologically relevant or a useful compromise between ecological relevance and preventing impractical long assays (see discussion in Sinclair et al., 2015).
We confirmed that lower ramping rates or higher KD temperatures result in decreased estimates of thermal tolerance (Machekano et al., 2018; Sørensen et al., 2013). Surprisingly, the result for individual ramping rates or KD temperatures revealed progressively non-linear reaction norms as ramping rates or KD temperatures decreased. The strength of the acclimation response (slope) as well as its non-linearity both increased with progressively slower ramping rates or lower KD temperatures (i.e. with longer assays). Thus, the strong linearity of developmental acclimation responses found across Drosophila species by Schou et al. (2016) might be restricted to developmental acclimation and/or to the assay condition of their study (ramping of 0.1 °C/min). In this study on D. melanogaster, a ramping rate of 0.5 °C/min produced a linear reaction norm with no benefit of acclimation, while a rate of 0.05 °C/min produced a much stronger, but non-linear acclimation response. Even if the absolute changes in thermal tolerance were limited (e.g. with a change in the CTmax of about 1 °C or less), it likely increases survival rates to biologically relevant high temperatures (e.g. increasing heat shock temperatures by 1 °C decreased survival rates from above 50% to nearly 0%; Kristensen et al., 2008). Many studies predict acclimation capacity based on two or three acclimation temperatures, a single assay-condition (ramping rate or KD temperature) and assume linearity (see e.g. data used in Gunderson and Stillman, 2015), these estimates might not be indicative of realized and ecologically relevant acclimation capacities. The ramping rate dependent shape of the reaction norms that we observe implies that alternate acclimation hypothesis may be supported depending on the ramping rate and acclimation temperatures selected. For example (at a ramp rate of 0.05 °C/min), one might conclude that “colder is better” if only comparing acclimation temperatures of 15 and 19 °C, but would conclude that “hotter is better” if comparing acclimation temperatures of 25 and 29 °C (see Deere and Chown, 2006, for a discussion of the different acclimation hypotheses). Further, they may conclude that acclimation has no effect if comparing acclimation temperatures of 15 and 25 °C. Finally, comparing acclimation temperatures of 15 and 29 °C might lead to observations of cross-tolerance in thermal performance (Sejerkilde et al., 2003). The choice of assay condition likewise affects the estimates of acclimation capacity. Using fast assays (high KD temperature or fast ramping rates) leads to low estimates of acclimation capacity (flat reaction norm slopes) as compared to longer assays. Thus, results from studies assuming linearity between two acclimation temperatures might be strongly influenced by the acclimation and assay conditions selected. This highlights that caution has to be exercised when extrapolating thermal tolerance reaction norms from a few acclimation temperatures. If the patterns of acclimation responses detected here are applicable across species, it has wide implications for the interpretation of acclimation capacity estimates and their ecological significance. Given the discrepancy in thermal tolerance estimates between various assay conditions, a natural question arises as to which condition represents the ‘correct’ acclimation response for thermal tolerance. Different ramping rates or KD temperatures likely measure different traits. Support for such an interpretation is found in the literature, as part of the inherent time-temperature interaction in thermal assays. Empirical studies suggest that discrepancies between ramping assays and static knock-down assays exist and might at least partly be explained by the higher potential for heat hardening (short term acclimation) during ramping assays (Overgaard et al., 2011; Overgaard et al., 2012; Sgrò et al., 2010; Terblanche et al., 2007). Such heat hardening potential would be increasingly pronounced at lower ramping rates, and likely also at lower KD temperatures (as both give rise to longer assays). Alternatively, Rezende et al. (2011) proposed that lower ramping rates offer increasing scope for factors such as desiccation and starvation to affect thermal tolerance estimates and should be avoided by using the highest rates or temperatures possible (see further discussion in Santos et al., 2011; Rezende, Tejedo, and Santos 2011).
5. Conclusion Here we investigated how adult acclimation to constant temperature modulates the reaction norms for high temperature thermal tolerance in Drosophila melanogaster. Our results highlight that the estimate of thermal tolerance and thermal acclimation capacity will typically reflect the assay condition and the history of the organisms assayed (Chown et al., 2009; Terblanche et al., 2007) and that acclimation capacity might be underestimated in many studies using relatively fast assays. Tolerance estimates based on high ramping rates or high KD temperatures might reflect partly different aspects of thermal tolerance as compared to low ramping rates or low KD temperatures. Although the tolerance estimates of dynamic and static assays yield results with different metric (temperature or time, respectively), both assays exhibit similar patterns of the acclimation responses and might be mechanistically related (Jørgensen et al., 2019). Strikingly, the lower ramping rates and lower KD temperatures led to increasingly non-linear reaction norms and stronger acclimation responses, presumably partly due to the better resolution offered at relatively 'mild' assay conditions. Our results and their consequences might not be restricted to drosophilids as they could apply more widely to ectotherms where thermal tolerances are estimated using similar dynamic or static assays. Future studies aiming at determining the limits for thermal tolerance should therefore carefully select assay conditions based on the species, it’s thermal niche and the rates of thermal changes that it experiences or that which might be ecologically relevant for that species. Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. 4
Journal of Insect Physiology 118 (2019) 103946
P.V. Salachan, et al.
Acknowledgements
Kristensen, T.N., Hoffmann, A.A., Overgaard, J., Sorensen, J.G., Hallas, R., Loeschcke, V., 2008. Costs and benefits of cold acclimation in field-released Drosophila. Proc. Natl. Acad. Sci. 105 (1), 216–221. https://doi.org/10.1073/pnas.0708074105. Leroi, A.M., Bennett, A.F., Lenski, R.E., 1994. Temperature acclimation and competitive fitness: an experimental test of the beneficial acclimation assumption. Proc. Natl. Acad. Sci. U.S.A. 91, 1917–1921. https://doi.org/10.1073/pnas.91.5.1917. Lutterschmidt, W.I., Hutchison, V.H., 1997. The critical thermal maximum: history and critique. Can. J. Zool. 75, 1561–1574. https://doi.org/10.1139/z97-783. Machekano, H., Mvumi, B.M., Nyamukondiwa, C., 2018. Loss of coevolved basal and plastic responses to temperature may underlie trophic level host-parasitoid interactions under global change. Biol. Control 118, 44–54. https://doi.org/10.1016/j. biocontrol.2017.12.005. Manenti, T., Cunha, T.R., Sørensen, J.G., Loeschcke, V., 2018. How much starvation, desiccation and oxygen depletion can Drosophila melanogaster tolerate before its upper thermal limits are affected? J. Insect Physiol. 111, 1–7. https://doi.org/10. 1016/j.jinsphys.2018.09.002. Martin, T.L., Huey, R.B., 2008. Why “suboptimal” is optimal: Jensen's inequality and ectotherm thermal preferences. Am. Nat. 171, 102–118. https://doi.org/10.1086/ 527502. Mitchell, K.A., Sgrò, C.M., Hoffmann, A.A., 2011. Phenotypic plasticity in upper thermal limits is weakly related to Drosophila species distributions. Funct. Ecol. 25, 661–670. https://doi.org/10.1111/j.1365-2435.2010.01821.x. Overgaard, J., Kearney, M.R., Hoffmann, A.A., 2014. Sensitivity to thermal extremes in Australian Drosophila implies similar impacts of climate change on the distribution of widespread and tropical species. Glob. Change Biol. 20, 1738–1750. https://doi.org/ 10.1111/gcb.12521. Overgaard, J., Kristensen, T.N., Mitchell, K.A., Hoffmann, A.A., 2011. Thermal tolerance in widespread and tropical Drosophila species: does phenotypic plasticity increase with latitude? Am. Natural. 178 (Suppl 1), 80–96. https://doi.org/10.1086/661780. Overgaard, J., Kristensen, T.N., Sørensen, J.G., 2012. Validity of thermal ramping assays used to assess thermal tolerance in arthropods. PLoS One 7, 1–7. https://doi.org/10. 1371/journal.pone.0032758. R Core Team, 2018. R: A Language and Environment for Statistical Computing: R Foundation for Statistical Computing. Rezende, E.L., Castañeda, L.E., Santos, M., Fox, C., 2014. Tolerance landscapes in thermal ecology. Funct. Ecol. 28, 799–809. https://doi.org/10.1111/1365-2435.12268. Rezende, E.L., Tejedo, M., Santos, M., 2011. Estimating the adaptive potential of critical thermal limits: methodological problems and evolutionary implications. Funct. Ecol. 25, 111–121. https://doi.org/10.1111/j.1365-2435.2010.01778.x. Santos, M., Castañeda, L.E., Rezende, E.L., 2011. Making sense of heat tolerance estimates in ectotherms: lessons from Drosophila. Funct. Ecol. 25, 1169–1180. https://doi.org/ 10.1111/j.1365-2435.2011.01908.x. Schou, M.F., Loeschcke, V., Kristensen, T.N., 2015. Strong costs and benefits of winter acclimatization in Drosophila melanogaster. PLoS One 10, 1–13. https://doi.org/10. 1371/journal.pone.0130307. Schou, M.F., Mouridsen, M.B., Sørensen, J.G., Loeschcke, V., 2016. Linear reaction norms of thermal limits in Drosophila: predictable plasticity in cold but not in heat tolerance. Funct. Ecol. 31, 934–945. https://doi.org/10.1111/1365-2435.12782. Schulte, P.M., Healy, T.M., Fangue, N.A., 2011. Thermal performance curves, phenotypic plasticity, and the time scales of temperature exposure. Integr. Comp. Biol. 51, 691–702. https://doi.org/10.1093/icb/icr097. Sejerkilde, M., Sørensen, J.G., Loeschcke, V., 2003. Effects of cold- and heat hardening on thermal resistance in Drosophila melanogaster. J. Insect Physiol. 49, 719–726. https:// doi.org/10.1016/s0022-1910(03)00095-7. Sgrò, C.M., Overgaard, J., Kristensen, T.N., Mitchell, K.A., Cockerell, F.E., Hoffmann, A.A., 2010. A comprehensive assessment of geographic variation in heat tolerance and hardening capacity in populations of Drosophila melanogaster from eastern Australia. J. Evolution. Biol. 23, 2484–2493. https://doi.org/10.1111/j.1420-9101. 2010.02110.x. Sgrò, C.M., Terblanche, J.S., Hoffmann, A.A., 2016. What can plasticity contribute to insect responses to climate change? Annu. Rev. Entomol. 61, 433–451. https://doi. org/10.1146/annurev-ento-010715-023859. Sinclair, B.J., Coello Alvarado, L.E., Ferguson, L.V., 2015. An invitation to measure insect cold tolerance: methods, approaches, and workflow. J. Therm. Biol. 53, 180–197. https://doi.org/10.1016/j.jtherbio.2015.11.003. Sørensen, J.G., Kristensen, T.N., Overgaard, J., 2016a. Evolutionary and ecological patterns of thermal acclimation capacity in Drosophila: is it important for keeping up with climate change? Current Opinion in Insect. Science 17, 98–104. https://doi.org/ 10.1016/j.cois.2016.08.003. Sørensen, J.G., Loeschcke, V., Kristensen, T.N., 2013. Cellular damage as induced by high temperature is dependent on rate of temperature change – investigating consequences of ramping rates on molecular and organismal phenotypes in Drosophila melanogaster. J. Exp. Biol. 216, 809–814. https://doi.org/10.1242/jeb.076356. Sørensen, J.G., Schou, M.F., Kristensen, T.N., Loeschcke, V., 2016b. Thermal fluctuations affect the transcriptome through mechanisms independent of average temperature. Sci. Rep. 6, 1–11. https://doi.org/10.1038/srep30975. Teets, N.M., Hahn, D.A., 2018. Genetic variation in the shape of cold-survival curves in a single fly population suggests potential for selection from climate variability. J. Evol. Biol. 31, 543–555. https://doi.org/10.1111/jeb.13244. Terblanche, J.S., Deere, J.A., Clusella-Trullas, S., Janion, C., Chown, S.L., 2007. Critical thermal limits depend on methodological context. Proc. Roy. Soc. B: Biol. Sci. 274, 2935–2942. https://doi.org/10.1098/rspb.2007.0985.
We would like to thank Heidi J. MacLean and Mads F. Schou for fruitful discussions on data analysis and manuscript writing. Thanks to Daniel A. Hahn and anonymous referees for constructive criticism on previous versions of this manuscript. This work was supported by a Starting Grant from Aarhus Universitets Forskningsfond to JGS (AUFFE-2015-FLS-8-72). Author contributions JGS conceived the idea and designed the study, HB and PVS collected the data, PVS and JGS analyzed the data, PVS made the figures. PVS wrote the original draft, JGS edited and reviewed the draft. All authors read and approved the final manuscript. Data accessibility Data have been deposited in the Dryad digital repository: https:// doi.org/10.5061/dryad.5jn61nt. Appendix A. Supplementary data Supplementary data to this article can be found online at https:// doi.org/10.1016/j.jinsphys.2019.103946. References Allen, J.L., Chown, S.L., Janion-Scheepers, C., Clusella-Trullas, S., 2016. Interactions between rates of temperature change and acclimation affect latitudinal patterns of warming tolerance. Conserv. Physiol. 4, 1–14. https://doi.org/10.1093/conphys/ cow053. Angilletta, M.J., 2009. In: Thermal Adaptation: A Theoretical and Empirical Synthesis. Oxford University Press, Oxford, UK. https://doi.org/10.1093/acprof:oso/ 9780198570875.001.1. Araújo, M.B., Ferri-Yanez, F., Bozinovic, F., Marquet, P.A., Valladares, F., Chown, S.L., 2013. Heat freezes niche evolution. Ecol. Lett. 16, 1206–1219. https://doi.org/10. 1111/ele.12155. Bates, D., Mächler, M., Bolker, B., Walker, S., 2015. Fitting linear mixed-effects models using lme4. J. Statist. Software 67. https://doi.org/10.18637/jss.v067.i01. Chown, S.L., Jumbam, K.R., Sørensen, J.G., Terblanche, J.S., 2009. Phenotypic variance, plasticity and heritability estimates of critical thermal limits depend on methodological context. Funct. Ecol. 23, 133–140. https://doi.org/10.1111/j.1365-2435.2008. 01481.x. Deere, J.A., Chown, S.L., 2006. Testing the beneficial acclimation hypothesis and its alternatives for locomotor performance. Am. Nat. 168, 630–644. https://doi.org/10. 1086/508026. Gunderson, A.R., Stillman, J.H., 2015. Plasticity in thermal tolerance has limited potential to buffer ectotherms from global warming. Proc. Roy. Soc. B: Biol. Sci. 282, 1–8. https://doi.org/10.1098/rspb.2015.0401. Hoffmann, A.A., Chown, S.L., Clusella-Trullas, S., 2013. Upper thermal limits in terrestrial ectotherms: how constrained are they? Funct. Ecol. 27, 934–949. https://doi.org/10. 1111/j.1365-2435.2012.02036.x. Hoffmann, A.A., Sørensen, J.G., Loeschcke, V., 2003. Adaptation of Drosophila to temperature extremes: bringing together quantitative and molecular approaches. J. Therm. Biol 28, 175–216. https://doi.org/10.1016/s0306-4565(02)00057-8. Huey, R.B., Berrigan, D., 2001. Temperature, demography, and ectotherm fitness. Am. Nat. 158, 204–210. https://doi.org/10.1086/321314. Huey, R.B., Kingsolver, J.G., 1989. Evolution of thermal sensitivity of ectotherm performance. Trends Ecol. Evol. 4, 131–135. https://doi.org/10.1016/0169-5347(89) 90211-5. Huey, R.B., Stevenson, R.D., 1979. Integrating thermal physiology and ecology of ectotherms: a discussion of approaches. Am. Zool. 19, 357–366. https://doi.org/10. 1093/icb/19.1.357. Jørgensen, L.B., Malte, H., Overgaard, J., 2019. How to assess Drosophila heat tolerance: unifying static and dynamic tolerance assays to predict heat distribution limits. Funct. Ecol. 33, 629–642. https://doi.org/10.1111/1365-2435.13279. Kellermann, V., Overgaard, J., Hoffmann, A.A., Flojgaard, C., Svenning, J.C., Loeschcke, V., 2012. Upper thermal limits of Drosophila are linked to species distributions and strongly constrained phylogenetically. Proc. Natl. Acad. Sci. U.S.A. 109, 16228–16233. https://doi.org/10.1073/pnas.1207553109. Kelty, J.D., Lee, R.E., 2001. Rapid cold-hardening of Drosophila melanogaster (Diptera: Drosophilidae) during ecologically based thermoperiodic cycles. J. Experim. Biol. 204, 1659–1666.
5