Journal of Thermal Biology 73 (2018) 41–49

Contents lists available at ScienceDirect

Journal of Thermal Biology

journal homepage: www.elsevier.com/locate/jtherbio

Geographic variation in responses of European yellow dung flies to thermal T stress

Stephanie S. Bauerfeinda, Jesper G. Sørensenb, Volker Loeschckeb, David Bergera,d, ⁎ E. Dale Brodera,c, Madeleine Geigera, Manuela Ferraria, Wolf U. Blanckenhorna, a Department of Evolutionary Biology and Environmental Studies, University of Zurich, Winterthurer Str. 190, CH-8057 Zurich, Switzerland b Department of Bioscience, Section for Genetics, and Evolution, University of Aarhus, DK-8000 Aarhus C, Denmark c Interdisciplinary Research Incubator for the Study of (in)Equality, University of Denver, Denver, CO 80208, USA d Evolutionary Biology Centre, University of Uppsala, Norbyvägen 18D, S-752 36 Uppsala, Sweden

ARTICLE INFO ABSTRACT

Keywords: Climatic conditions can be very heterogeneous even over small geographic scales, and are believed to be major Cold shock determinants of the abundance and distribution of species and populations. Organisms are expected to evolve in Egg size response to the frequency and magnitude of local thermal extremes, resulting in local adaptation. Using replicate Egg content yellow dung fly( stercoraria; Diptera: ) populations from cold (northern Europe) and Fecundity warm climates (southern Europe), we compared 1) responses to short-term heat and cold shocks in both sexes, 2) Heat shock protein heat shock protein (Hsp70) expression in adults and eggs, and 3) female reproductive traits when facing short- Latitude Reproductive investment term heat stress during egg maturation. Contrary to expectations, thermal traits showed minor geographic Survival differentiation, with weak evidence for greater heat resistance of southern flies but no differentiation in cold resistance. Hsp70 protein expression was little affected by heat stress, indicating systemic rather than induced regulation of the heat stress response, possibly related to this fly group's preference for cold climes. In contrast, sex differences were pronounced: males (which are larger) endured hot temperatures longer, while females featured higher Hsp70 expression. Heat stress negatively affected various female reproductive traits, reducing first clutch size, overall reproductive investment, egg lipid content, and subsequent larval hatching. These re- sponses varied little across latitude but somewhat among populations in terms of egg size, protein content, and larval hatching success. Several reproductive parameters, but not Hsp70 expression, exhibited heritable varia- tion among full-sib families. Rather than large-scale clinal geographic variation, our study suggests some local geographic population differentiation in the ability of yellow dung flies to buffer the impact of heat stress on reproductive performance.

1. Introduction Partridge, 2001; Sisodia and Singh, 2006; Gaston et al., 2008; Angilletta, 2009; Calosi et al., 2010), either by genetic adaptation or Temperature is one of the most important environmental factors acclimation (a typically reversible, plastic physiological response), affecting life histories of organisms. This is especially true for ecto- or which produce molecular and physiological variation in space (i.e. poikilothermic species for which nearly all physiological and beha- across the species’ geographical range) and time (i.e. daily and sea- vioural traits are sensitive to ambient temperatures (Hoffmann et al., sonal). 2003; Sinclair et al., 2003; Clarke, 2014). Environmental temperature The thermal environment is generally assumed to be the prime conditions can be highly heterogeneous even at small geographic scales, factor shaping large-scale geographic variation in many phenotypic including not only differences in mean temperatures but also in the traits. Prominent geographic or latitudinal life history patterns (e.g. frequency and magnitude of temperature extremes (Potter et al., 2013; Bergmann's rule) are known to be temperature mediated (e.g. Partridge Sørensen et al., 2016). Thus temperature is a major determinant of the and French, 1996; Chown and Gaston, 1999, 2010). Similarly, ther- abundance and distribution of species and populations (Robinson and motolerance varies geographically in response to the frequency and

⁎ Corresponding author. E-mail addresses: [email protected] (S.S. Bauerfeind), [email protected] (J.G. Sørensen), [email protected] (V. Loeschcke), [email protected] (D. Berger), [email protected] (E.D. Broder), [email protected] (M. Geiger), [email protected] (M. Ferrari), [email protected] (W.U. Blanckenhorn). https://doi.org/10.1016/j.jtherbio.2018.01.002 Received 17 November 2017; Received in revised form 29 December 2017; Accepted 19 January 2018 0306-4565/ © 2018 Elsevier Ltd. All rights reserved. S.S. Bauerfeind et al. Journal of Thermal Biology 73 (2018) 41–49

Table 1 Climate conditions at the sampling locations of our populations.

Temperature (°C) Rainfall (mm)

Population Latitude Longitude Altitude (m) Winter lengtha (d) Mean annual Mean annual range Max monthly Min monthly Mean annual

Sierra Nevada, Spain 37.20N − 3.20W 1290 91 12.6 18.3 30.9 − 0.1 558 Pyrenees, Spain 42.25N 1.90E 830 97 11.3 15.8 24.6 0.6 793 Locarno, Switzerland 46.26N 8.69E 300 98 11.6 18.5 26.0 − 0.9 1243 Grimstad, Norway 58.34N 8.59E 10 169 7.3 16.4 18.9 − 3.0 1108 Uppsala, Sweden 59.73N 17.71E 16 178 5.7 20.9 20.9 − 7.4 551 Tromsø, Norway 69.68N 18.72E 5 243 2.2 16.3 15.0 − 7.0 911

a Mean number of days below 6 °C (developmental threshold). magnitude of local thermal extremes (Angilletta, 2009; Addo-Bediako adapted (southern) populations would require less Hsp when compared et al., 2000; for some specific examples (mainly ) see to less heat adapted (northern) populations (Sørensen et al., 2003). e.g. Hoffmann et al., 2002; Castañeda et al., 2004; Chen and Kang, Moreover, (3) if the response to heat stress is costly and directly in- 2004; Kellermann et al., 2012; Schou et al., 2017). The ability to endure terferes with reproductive investment, all populations should suffer local temperature extremes is crucially important, as exposure to from reduced reproductive output when subjected to extreme tem- thermal extremes can impair individual performance and ultimately peratures; and (4) if populations are locally adapted to the regular fitness in various ways. For instance, heat shock has been shown to occurrence of temperature extremes, the costs of the heat stress re- interfere with ovarian development or vitellogenesis, and may cause sponse (i.e. the expected impairment in reproductive traits) should injury to oocytes, testes and (Wang et al., 2009). predictably differ between northern and southern populations (lower in While thermotolerance is mediated by various behavioural, phy- the south; cf. Scharf et al., 2010). To facilitate comparisons, we focused siological, or genetic responses (Hoffmann et al., 2002; Angilletta, on Hsp70, the major, highly conserved, and therefore most commonly 2009), one of the most important mechanisms is the induction of heat assessed Hsp gene family in the literature across species (Feder and shock proteins (Hsps) that protect the cellular machinery (Huang et al., Hofmann, 1999; Hoffmann et al., 2011). Our study goes well beyond 2007). Hsps usually act as molecular chaperones, ensuring and sup- others as we not only include egg number and size as estimates of fe- porting correct protein folding under stressful conditions (Feder et al., male reproductive performance, but also hatching success and egg 1992; Huang et al., 2007). While increased Hsp expression directly composition as assessed by egg lipid, glycogen and protein content. This improves resistance to various stressors, it usually results in modifica- is especially important because variation in egg size does not necessa- tion of protein synthesis within cells and often comes at costs that be- rily imply qualitative changes in egg composition, which can be eco- come apparent in cell division, development and (Feder logically and evolutionarily more important than variation in egg size and Hofmann, 1999; Huang et al., 2007). The impact of these costs is per se (Fox and Czesak, 2000; Giron and Casas, 2003). ameliorated by a number of regulatory mechanisms that not only en- sure the rapid expression of Hsps under stress but also minimize their 2. Materials and methods expression under favorable conditions (Feder et al., 1992). Ultimately, these costs become apparent in trade-offs with reproductive traits and/ 2.1. Population sampling and fly maintenance or individual performance that are mediated through redistribution of resources to the stress response machinery (Silbermann and Tatar, We examined flies originally collected from three sites (i.e. popu- 2000; Sørensen et al., 2003; Sisodia and Singh, 2006; Huang et al., lations) in northern Europe (Tromsø (T), northern Norway; Uppsala 2007). As geographic populations typically vary in their exposure and (U), central Sweden; Grimstad (G), southern Norway) and three sites ability to endure extreme hot and cold temperatures as well as high from southern Europe (Locarno, Switzerland (L); Pyrenees (P), northern (daily and seasonal) temperature fluctuations (Hoffmann et al., 2002; Spain; Sierra Nevada (S), southern Spain; Table 1). All populations had Schilthuizen and Kellermann, 2014; Stoks et al., 2014), we can expect been kept in the laboratory, at a minimum of 20–30 full-sibling families the costs of stress responses to correlate, such that populations facing per population and generation, for 3–5 generations prior to our ex- more extreme temperatures should feature more efficient Hsp-related periments (generation time ca. 2–3 months), long enough to eliminate protection to minimize the fitness costs associated with inducible stress carry-over maternal effects from the field, but not long enough for la- responses. boratory adaptation to erode naturally evolved patterns. Due to their We here investigate the effects of extreme temperatures on thermal association with pastures and their preference for cooler tem- traits (Hsp70 expression, heat and cold shock endurance), and the peratures, southern flies originate from higher altitudes (Table 1). Fe- consequences of heat stress for female reproductive traits by comparing male reproductive traits were assessed (in 2008) only for the four po- replicated yellow dung fly(; Diptera: pulations collected in 2007 (Tromsø, Uppsala, Pyrenees, Sierra Scathophagidae) populations from locations in northern vs. southern Nevada), whereas heat and cold shock responses and Hsp70 expression Europe that systematically vary in climate. This fly is widely distributed were assessed for males and females of all six populations (including across the northern hemisphere as far north as Iceland, Greenland or Grimstad and Locarno, collected in 2008; Table 1) in two different years Spitzbergen, showing a clear preference for cool temperatures (2008 and 2009). Sample sizes ranged from 49 to 177 individual flies (Sigurjónsdóttir and Snorrason, 1995; Blanckenhorn, 2009). Towards from 18 to 45 families per population and sex. the south of its distribution it appears to be limited by hot temperatures Prior to our experiments, flies were maintained in the laboratory for (Ward and Simmons, 1990; Blanckenhorn et al., 2001). We hypothe- several generations under optimal common garden holding conditions sized that (1) southern European populations should be more heat of 18 °C and 14 h light (Blanckenhorn, 2009). Any differences in po- tolerant and northern populations more cold tolerant (Hoffmann et al., pulation mean phenotypes therefore reflect underlying genetic rather 2002), and, rather than expecting higher baseline Hsp levels in organ- than environmental habitat differences. Separated by full-sib families, isms inhabiting hotter climates (which would be a reasonable hypoth- larvae were raised in groups of ca. 10 individuals in translucent plastic esis), we expected that (2) induction of Hsp70 in response to a heat containers with overabundant cow dung (> 2 g per ; challenge would be reflective of thermal adaptation such that more heat Blanckenhorn, 2009). Containers were checked daily for emerged flies.

42 S.S. Bauerfeind et al. Journal of Thermal Biology 73 (2018) 41–49

Upon adult eclosion, flies were kept individually in glass vials (100 ml) with the reproductive trait assessment (2.2 above), 10 of the first fe- and provided with water, sucrose and Drosophila virilis as prey in ample male eggs deposited were collected and frozen at − 80 °C to similarly supply. Blanckenhorn et al. (2010) describe rearing methods in more analyze the expression of Hsp70 in eggs. Only in the two populations detail. sampled in 2008 (Grimstad, Locarno), we further investigated the time course of Hsp70 expression by varying the time interval before freezing 2.2. Reproductive trait assessment the fly (i.e. fixation) after the heat shock in 0.5 h intervals, from 0 to 2h. Freshly eclosed female flies were randomly assigned either to the Thoraces and eggs were homogenized in 1 and 0.5 ml (respectively) control (18 °C, as above) or the heat stress treatment group. For the phosphate-buffered saline (PBS) containing 200 mM PEFAbloc and a latter, females were subjected to 33 °C for 1 h on day 2 after adult 1 vol% antiprotease cocktail (100 μl/ml pepstatin A, 50 μl/ml leu- eclosion before being exposed to a thermal extreme of 35 °C for 1 h on peptin, 10 mM benzamidine, 10 mM sodium metabisulfite), and then day 5. In yellow dung flies egg and sperm maturation commences only centrifuged for 30 min at 13,000 rpm (11,000g) at 4 °C. The supernatant after adult eclosion, which at the chosen rearing temperature takes ca. was divided into three replicate subsamples of 300 μl each (to ensure 10 days and requires protein nutrients from prey ( or in- extra samples if necessary), and frozen again at − 80 °C. Total protein come breeding; see Blanckenhorn and Henseler, 2005); the stress content was measured using the same BCA protein assay as above: to treatment was thus applied during the initial phase of egg maturation. standardize and facilitate direct comparisons of Hsp70 content, all

One half of the females (ntotal = 177, 13–31 individuals per population samples were diluted to a protein content of 30 μg/ml prior to mea- and treatment group) was then assigned to Hsp70 expression analyses surement. Hsp70 expression has not been investigated before in dung

(as explained above), while the other half (ntotal = 236, 12–43) was flies, and consequently no specific antibodies were known. Western assigned to the reproductive assays. For all flies (egg-to-adult) devel- blotting confirmed measuring a protein of the predicted size (70 kDa) opment times and body sizes (hind tibia length, as commonly used in for the antibody used (results not shown). Hsp70 expression level of this species; Blanckenhorn, 2009) were also recorded. females was measured using an enzyme-linked immunosorbent assay For mating, which took place after application of heat stress (> 7 d (ELISA) following Dahlgaard et al. (1998), using an Hsp70 specific after adult eclosion), female flies were transferred to a small glass vial. monoclonal antibody (clone 5A5, mouse anti-rabbit, 1:2000, ABR) and Then a randomly chosen male fly (from the same population, but a a horseradish-peroxidase-conjugated secondary antibody (polyclonal different family) and a smear of cow dung for egg deposition were rabbit anti-mouse, 1:2000, DAKO). The primary antibody used detects added. After mating, males were removed and females allowed to de- both the constitutive and induced Hsp70 family members (referred to posit eggs. If mating did not take place, the procedure was repeated the Hsp70 throughout). The color reaction was measured by a spectro- following day using a different male. We recorded egg number (first photometric microplate reader at 562 nm. Results are presented as the clutch size), volume, content (glycogen, lipid, protein), and viability average signal of the tested sample, i.e. the mean of four replicate (i.e. larval hatching success = eggs hatched/eggs laid). Hsp70 expres- measures per sample corrected by the unspecific color reaction of two sion in the eggs was also assessed (cf. 2.4. below). replicate blanks per sample. One reference sample was replicated on all For egg size estimation, length (l) and width (w) of 3–5 eggs per plates, thus allowing for a control of inter-plate differences. clutch were measured using a digital camera connected to a binocular microscope (Leica DFC 490). The resulting images were analyzed using 2.5. Heat knockdown and chill coma recovery time assessment ImageJ 1.38× (NIH, USA; National Institutes of Health, 2007). Mean egg volume was calculated assuming the form of an ellipsoid: (1/ Resistance to heat stress was assessed by a knockdown assay, and 6)*π*l*w2 (cf. Blanckenhorn and Heyland, 2004). cold resistance by a chill coma recovery assay (cf. Angilletta, 2009), using male and female flies from all six populations. Though ultimately 2.3. Physiological analyses of egg content arbitrary, our choice of temperature extremes was based on preliminary assays (for cold and heat shock) and previous studies (for heat shock: Egg glycogen and lipid content were analyzed using a modified Blanckenhorn et al., 2014). Freezing temperatures (potentially down to protocol based on standard methods by Van Handel (1985a, 1985b); − 20 °C) are regularly experienced by yellow dung fly adults at the end Van Handel and Day (1988; consider Foray et al., 2012 for an update). of the season in November/December, when they are still active on Groups of 10 eggs from the same clutch were homogenized in 200 μl of pastures, and/or in the midst of winter in cases when adults emerge

2% Na2SO4, which absorbs and precipitates the glycogen. Glycogen was during warm spells (Blanckenhorn, 2009). Hot temperatures in the mid- then quantified photometrically (spectrophotometric microplate thirties regularly occur on summer pastures in bright sunshine. reader, Spectra Max 340 PC) by anthrone reaction (using 0.1% glucose were individually placed in small, sealed plastic vials (15 ml), in 25% EtOH as standard). After a (1:1) chloroform:methanol extrac- with a moist piece of cotton in the bottom in order to prevent de- tion, the lipid content was measured using a vanillin-phosphoric acid siccation. For the heat shock test, 8–10 vials at a time were then sub- reaction (with 0.1% soybean oil in chloroform as standard). Egg protein merged in a water bath kept at constant 37 ± 1 °C. The flies were content (of 10 different eggs) was measured with the BCA protein assay continuously monitored, and heat knock-down time (defined as the kit of Thermoscientific© according to the manufacturers specification, time until a fly was no longer able to stand upright) for each individual which uses a colorimetric method based on the reduction of Cu2+ to was recorded. Cold shock was induced by submerging the vials in a Cu1+ by bicinchoninic acid in an alkaline medium, with bovine serum water-ethanol mix at − 20 °C for 15–20 min, again in groups of 8–10 albumin as standard. The energetic values of all components were individuals at a time. Thereafter the chilled flies were placed into larger converted to calories with 1 cal corresponding to 250 μg of carbohy- glass vials at room temperature of ca. 22 °C to record their chill coma drates or protein and 110 μg of lipids. recovery time (defined as the time after exposure to freezing tem- peratures until the flies spontaneously stood up and moved again). Data 2.4. Hsp analysis on flies that died due to treatment, which occasionally happened, were excluded from analyses. To measure Hsp70 levels, female flies were frozen at − 80 °C 2 h after the heat stress treatment to allow the expression of the stress re- 2.6. Statistical analysis sponse to accumulate (only thoraces were used for analysis; cf. Fig. S1). Therefore it was not possible to determine both Hsp70 expression level Thermal traits (Hsp70 expression in the fly thorax, heat stress and reproductive traits for the same female. Additionally, in connection knockdown time, chill coma recovery time) were separately analyzed

43 S.S. Bauerfeind et al. Journal of Thermal Biology 73 (2018) 41–49 using General Linear Models (normal error distribution) with geo- graphic region (north/south), sex, and heat stress treatment (Hsp70 only) as fixed factors, population nested within geographic region and family nested within population and region as random factors, and body size (hind tibia length) as covariate. Interaction terms were first included but removed if non-significant. The female reproductive traits were similarly analyzed separately using Generalized Linear Models (normal error distribution) with geo- graphic region (north/south) and stress treatment as fixed factors, po- pulation nested within geographic region and family nested within population and region as random factors. Body size and egg size were added as covariates if appropriate, and excluded from the final minimum adequate model if not significant. For larval hatching success models with binomial error distribution were used (hatched/not hat- ched), initially with egg size and number and body size included as covariates. All reproductive trait analyses were additionally run using family means as the level of replication (weighted for the number of individuals within each family) to estimate the impact of Hsp70 ex- pression on all traits. This was necessary as it was not possible to obtain data on reproductive traits and Hsp70 expression from the same in- dividuals. Means are given ± 1 SE throughout. Unless otherwise stated, egg content data refer to measures for 10 eggs. Data transformation (to meet the statistical assumption of normal distribution of residuals) was not necessary for any trait.

3. Results

3.1. Heat knock-down and cold shock recovery times

Chill coma recovery time did not vary among regions or popula- tions, and was marginally lower in females than males (Table 2; Fig. 1a). However, heritable variation in the response to cold shock was detected (family effect in Table 2). Heat stress knockdown time was longer for males than females, and showed a slight albeit non-sig- nificant trend to be longer in the southern populations (Table 2; Fig. 1b). When the non-significant region factor was removed, the six populations varied significantly in heat shock knockdown time (F5,167 = 2.56, P = 0.039; cf. Table 2). Heritable family effects were again

Table 2 Analyses of variance testing for effects of heat stress (Hsp70), sex, geographic region (R, northern vs. southern Europe), population (P, nested within geographic region; random factor), and family (nested within geographic region and population, random factor) on thermal traits in yellow dung flies from three northern (Grimstad, Uppsala, Tromsø) and three southern (Sierra Nevada, Pyrenees, Locarno) European populations. Non-significant covariates and interaction terms removed; varying df due to separate data sets; P < 0.05 Fig. 1. Mean ± SE (a) chill coma recovery time, (b) heat knockdown time, and (c) Hsp70 in bold, P < 0.1 in italics. expression for male (dark grey) and female (light grey) yellow dung flies (Scathophaga stercoraria) from three southern (left) and three northern (right) European populations (N Factor MS df F P =69–153 individuals per sex and population; heat stress did not significantly elevate Hsp70). Two individual columns differ significantly roughly when their SE bars do not Chill coma recovery Sex 2.236 1,808 3.26 0.071 overlap. * P < 0.05; (*) P < 0.1. time Region 0.380 1,4 0.33 0.566 Population[R] 0.571 4,162 0.50 0.684 evident (Table 2). Body size (within the sexes) did not influence heat or Family[R,P] 1.374 162,808 1.98 < 0.001 cold shock responses of individuals (covariate hence removed from Error 0.695 808 Table 2). Heat stress knockdown Sex 382.33 1,974 5.54 0.019 time Region 834.56 1,4 2.68 0.183 3.2. Hsp70 expression in flies Population[R] 336.07 4,167 2.15 0.095 Family[R,P] 200.34 167,974 2.91 < 0.001 Error 68.98 974 Interestingly, Hsp70 expression was much more pronounced in fe- males, and also higher in the northern than the southern populations Hsp70 expression Heat stress 0.021 1,419 1.82 0.178 ff Sex 0.093 1,419 8.25 0.004 (region e ect in Table 2: P = 0.021; region by sex interaction mar- Region 0.039 1,4 34.70 0.021 ginally significant as well: P = 0.082). Contrary to expectation, how- Region*Sex 0.034 1,419 3.03 0.082 ever, Hsp70 did not differ between heat stressed and unstressed flies, Population[R] 0.001 4,182 0.14 0.934 and no significant interaction between heat stress and geographic re- Family[R,P] 0.008 182,419 0.69 0.998 Error 0.011 419 gion or population origin was apparent (removed in Table 2). Supplementary Fig. S1 further shows no significant variation in Hsp70

44 S.S. Bauerfeind et al. Journal of Thermal Biology 73 (2018) 41–49

Table 3 Table 3 (continued) Analyses of variance testing for effects of geographic region (R, northern vs. southern Europe), population (P, nested within geographic region; random factor), heat stress (S) χ2 and family (nested within geographic region and population, random factor) on female reproductive traits in yellow dung flies from two northern (Uppsala, Tromsø) and two p(eggs hatched)a Region 1,2 10.64 0.001 southern (Sierra Nevada, Pyrenees) European populations. Egg size and body size were Population[R] 2,171 11.33 0.004 added as covariates as appropriate. Non-significant covariates and interaction terms re- Stress 1,171 4.04 0.045 moved; varying error df due to missing values; P < 0.05 in bold, P < 0.1 in italics. Family[R,P] 86,171 160.20 < 0.001 Egg size 1,171 3.89 0.049 Factor MS df F P Egg protein content 1,171 4.41 0.037 Egg glycogen content 1,171 16.62 < 0.001 Egg number Region 98.7 1,2 0.42 0.594 (first clutch size) Population[R] 261.9 2,188 1.00 0.367 a Binary variable. Stress 1551.4 1,171 6.04 0.015 Family[R,P] 261.1 102,171 1.53 0.028 expression as a function of time of freezing (i.e. fixation) over the first Error 257.7 171 two hours after heat shock. Results thus indicate that Hsp70 expression Egg size Region 0.0010 1,2 1.49 0.336 in yellow dung flies is not induced but systemic. Also, variation in Population[R] 0.0007 2,160 4.13 0.018 Hsp70 expression does not appear to be heritable (non-significant fa- Stress 0.0003 1,150 2.80 0.098 Family[R,P] 0.0002 102,150 1.79 0.008 mily effect in Table 2). Error 0.0001 150

Reproductive Region 6.0 1,2 5.37 0.090 investment 3.3. Reproductive traits in response to heat stress (egg size * clutch size) Population[R] 0.9 2,164 0.22 0.781 Stress 29.9 1,146 10.56 0.001 Heat stress significantly reduced the number of eggs females laid Family[R,P] 4.0 102,146 1.44 0.026 (control > stress treatment: 65.7 ± 1.3 > 60.1 ± 1.4 eggs), and also Error 2.8 146 tended to reduce egg size (control > stress treatment: Egg glycogen content Region 4.8 1,2 0.18 0.697 0.112 ± 0.001 > 0.109 ± 0.001 mm³; Table 3; Fig. 2). While first clutch Population[R] 24.2 2,102 0.78 0.458 Stress 2.3 1,76 0.05 0.825 Family[R,P] 24.8 88,76 0.53 0.998

Egg size rpartial = 330.6 1,76 7.06 0.010 0.23

Body size rpartial = 261.7 1,76 5.59 0.021 0.20 Error 46.9 76

Egg lipid content Region 41.9 1,2 0.23 0.733 Population[R] 299.0 2,101 2.61 0.078 Stress 436.4 1,76 8.33 0.005 Family[R,P] 132.9 82,76 2.55 < 0.001 Error 52.9 76

Egg protein content Region 401.4 1,2 0.13 0.741 Population[R] 3241.9 2,156 4.19 0.017 Stress 372.3 1,120 0.58 0.454 Family[R,P] 818.4 94,120 1.24 0.134

Egg size rpartial = 78,232. 1,120 118.4 < 0.001 0.72 Error 660.6 120

Energy content of 10 Region 0.006 1,3 0.18 0.702 eggs Population[R] 0.038 2,98 1.45 0.240 Stress 0.001 1,55 0.04 0.836 Family[R,P] 0.030 74,55 1.88 0.008

Egg size rpartial = 0.744 1,55 46.96 < 0.001 0.68

Body size rpartial = 0.059 1,55 3.71 0.059 0.25 Error 0.016 55

Energy content of Region 0.3 1,2 0.09 0.791 clutch Population[R] 3.9 2,90 1.72 0.195 Stress 1.4 1,55 1.38 0.234 Family[R,P] 2.8 74,55 2.93 < 0.001

Egg size rpartial = 41.4 1,55 43.63 < 0.001 0.66

Body size rpartial = 7.8 1,55 8.25 0.006 0.36 Error 0.9 55

Hsp70 expression Region 0.03 1,2 0.84 0.451 eggs Population[R] 0.3 2,140 0.89 0.391 Stress 0.12 1,89 3.31 0.072 Fig. 2. Mean ( ± SE) effects of population origin and heat stress during egg maturation in Family[R,P] 0.04 89,89 1.01 0.442 the yellow dung fly on: (a) clutch size, (b) egg volume, (c) reproductive investment, (d) Error 0.04 89 energy content of the first clutch. Populations are aligned from south to north. Two in- dividual columns differ significantly roughly when their SE bars do not overlap. * P < 0.05; (*) P < 0.1.

45 S.S. Bauerfeind et al. Journal of Thermal Biology 73 (2018) 41–49 size did not vary across geographic regions or populations, egg size dif- fered significantly among populations (SierraS = PyreneesS≤ 3 3 TromsøN < UppsalaN: 0.105 ± 0.001 mm = 0.109 ± 0.002 mm ≤ 0.111 ± 0.002 mm3 <0.116±0.001mm3; Tukey HSD after ANOVA; Table 3). Consequently, overall reproductive investment (egg size * clutch size, i.e. clutch volume) was also significantly reduced after heat stress (7.5 ± 0.2 mm3 >6.7±0.2mm3; Table 3), with regional differences suggesting a tendency towards greater investment in northern than southern populations (P = 0.090; 7.3 ± 0.1 mm3 >6.9±0.2mm3; Table 3; Fig. 2). However, when reproductive investment was expressed as the total energy content of the first clutch (cf. energy analysis below), the pattern changed: neither heat stress nor geographic region showed sig- nificant effects (Table 3). Hsp70 expression in the eggs was only mar- ginally increased after heat shock (Table 3).Themother'sbodysizeonly significantly influenced the energy content of the eggs and the total first clutch positively (Table 3;allotherP >0.2). We analogously tested for possible effects of the mother's Hsp70 expression on reproductive traits using family means (weighted by the number of individuals per family) as the unit of replication (see methods 2.6). These analyses revealed a slight positive relationship between Hsp70 expression and clutch size (partial r = 0.16, F1,160 = 3.9, P = 0.049) as well as between Hsp70 and overall reproductive investment (partial r = 0.25, F1,140 = 9.0, P = 0.003). There further was a trend for a negative correlation between Hsp70 and egg protein content (partial r = − 0.17, F1,117 = 3.3, P = 0.073). Family origin significantly influenced clutch size, egg size, and re- productive investment (Table 3), explaining 11.1%, 15.2% and 18.2% of the detected variation, respectively.

3.4. Egg composition

Only egg lipid content was significantly reduced by heat stress (45.9 ± 1.1 μg > 44.1 ± 1.1 μg; Table 3), while the other components (egg glycogen and protein content, total energy content per 10 eggs) remained unaffected. Furthermore, none of the measured egg compo- nents was significantly affected by geographic region, although egg Fig. 3. Mean ( ± SE) effects of population origin and heat stress during egg maturation in fl fi protein content varied significantly across populations, Pyr- the yellow dung y on: (a) egg glycogen, (b) lipid, and (c) protein content of the rst clutch, (d) total energy content of 10 eggs. Populations are aligned from south to north. eneesS < SierraS = TromsøN=UppsalaN; 143.1 ± 5.6 μg < 155.6 ± Two individual columns roughly differ significantly when their SE bars do not overlap. * 4.8 μg = 161.1 ± 3.6 μg = 164.9 ± 4.4 μg; Tukey HSD after ANOVA; P < 0.05; (*) P < 0.1. Table 3; Fig. 3). Note that population differences in egg lipid content almost reached significance (P = 0.078; P = 0.051 when dropping the non-significant region factor). All egg components except lipid content correlated positively with egg size (Table 3). Additionally, egg glycogen content was positively related with body size (Table 3). Family origin significantly influenced egg lipid content and total energy content, explaining 42.2% and 33.7% of the variation (respec- tively), but not egg glycogen or protein content (Table 3).

3.5. Larval hatching success

Northern populations showed higher larval hatching success than Fig. 4. Mean ( ± SE) effects of population origin and heat stress during egg maturation in southern populations (83.4 ± 3.6% vs. 80.7 ± 4.9%), although the the yellow dung fly on larval hatching success. Populations are aligned from south to Spanish populations exhibited the highest and lowest hatching success north. Two individual columns differ significantly when their SE bars do not overlap. * (SierraS < TromsøN < UppsalaN < PyreneesS: P < 0.05; (*) P < 0.1. 75.0 ± 5.0% < 82.3 ± 3.5% < 84.4 ± 3.6% < 86.5 ± 4.9%; Fig. 4; Table 3). As expected, hatching success was significantly reduced by 4. Discussion heat stress overall (85.2 ± 2.6 vs. 77.5 ± 3.3%; P = 0.045; Table 3). However, while the southern- and northernmost populations (Sierra Contrary to our expectations, responses of yellow dung flies to Nevada: − 1.0%, Tromsø: − 1.7%) were barely affected by heat stress, short-term extreme heat and cold exposure showed little differentiation hatching success was strongly reduced in the populations from the over a wide latitudinal range from southern Spain to northern Norway Pyrenees and Uppsala (− 15.9% and − 16.7%, respectively; Fig. 4). (Fig. 1). While females, which are smaller in this species, recovered Larval hatching success was negatively related to egg protein content (r slightly faster after cold shock, the larger males persisted longer at hot = − 0.16), but positively related to egg size (r = 0.13) and egg gly- temperatures (37 °C, which should regularly occur on pastures in bright cogen content (r = 0.27; Table 3).

46 S.S. Bauerfeind et al. Journal of Thermal Biology 73 (2018) 41–49 summer sunshine), with only weak indications of southern flies being significantly increase after exposure to heat stress, and it also did not more heat resistant. Hsp70 levels of the flies were not affected by depend on time since the heat shock (Fig. S1). Our results thus contrast temperature stress, though slightly higher levels were evident in po- with the situation in (Velazquez et al., 1983; pulations from higher latitudes. This indicates that Hsp70 expression in Feder et al., 1992; Hoffmann et al., 2002, 2003) in that Hsp70 is not yellow dung flies is not stress-induced but systemic (or constitutive: cf. stress induced but systemically present in unstressed flies at high con- Zatsepina et al., 2016). Interestingly, despite being smaller, females stitutive levels, adding to mounting evidence to that effect in other showed considerably higher Hsp70 levels than males (Fig. 1). Yet even dipterans and perhaps many other (Zatsepina et al., 2016). It is brief periods of heat stress during the egg maturation period after adult possible that this relates to this entire family of flies (Scathophagidae: eclosion significantly reduced the reproductive output of yellow dung Gorodkov, 1984) being common in sub-arctic habitats and therefore fly females in terms of clutch size, overall reproductive investment, egg necessarily very cold-resistant, such that permanent, rather than plastic lipid content, and larval hatching success. Reproductive performance Hsp expression has evolved. Alternatively, other heat shock proteins was primarily diminished via egg number and subsequent larval (such as Hsp90: e.g. Sgrò et al., 2010b) could play a more important hatching success rather than egg size. Again, systematic latitudinal role in yellow dung flies, which may respond in very different ways but variation in these responses was minor relative to (micro)geographic were not assessed here. It is well established by now that upper, but variation among populations likely reflecting adaptation to local (ra- likely also lower thermal limits of organisms depend substantially on ther than global) climate (see Potter et al. (2013) for a discussion). As the assessment methods and traits used (Sgrò et al., 2010a; Santos et al., all these effects were found in a laboratory common garden, they reflect 2011; Hoffmann et al., 2011). Whereas our common garden study as- genetic differentiation, and most investigated traits (but not Hsp70 sessed heritable population differentiation, extrapolations on how the expression) indeed revealed substantial heritable components. laboratory results documented here actually translate to the natural situation must remain limited, however, not least because of the sub- 4.1. Responses to heat and cold shock and Hsp70 expression stantial role of in thermal responses (Hoffmann et al., 2011; e.g. Carbonell et al., 2017). Responses of yellow dung flies to extremely cold (here estimated in We also expected heat shock protein expression to differ between terms of chill coma recovery time after brief exposure to − 20 °C) or hot cold- and warm-adapted populations (Potter et al., 2013). However, temperatures (here estimated as heat knockdown resistance after brief Hsp70 expression levels were only marginally higher in flies from exposure to 37 °C) did not vary strongly among populations spanning northern European populations, thus showing little geographic speci- roughly 30 degrees latitude in Europe (Table 1), as had been expected ficity. Instead, Hsp70 levels were much higher in females than males, from work on other invertebrates (e.g. Sgrò et al., 2010a; Hoffmann but independent of body size within the sexes. Males also had longer et al., 2002, 2011; Schilthuizen and Kellermann, 2014; Stoks et al., heat knockdown times (Fig. 1; see Esperk et al., 2016, for contrary 2014). This result does not match differentiation patterns of life history results in the black scavenger fly punctum). Taken together, these traits in yellow dung flies, which at times show pronounced latitudinal results are somewhat contradictory. As yellow dung fly males are reg- trends (e.g. in diapause induction or growth rate; Blanckenhorn, 2009; ularly exposed to direct sunshine for extended time periods when Scharf et al., 2010; Berger et al., 2011), perhaps because these are often waiting and fighting for females on pastures (Blanckenhorn, 2009), attributed to seasonal time constraints rather than adaptation to tem- they should require greater heat protection on average; this would perature per se. We nevertheless do not believe that the particular explain their longer heat endurance in the laboratory, which however temperature or light conditions used for our rearing (18 °C and 14 h cannot possibly be directly mediated by their lower Hsp70 levels. Fe- light, representing optimal conditions for this fly) and experiments, males, in contrast, generally invest more in reproduction than males, which were informed by prior experience but otherwise admittedly which could explain their generally higher Hsp70 levels in the face of arbitrary, strongly affected our results, though in the end we cannot substantial heat-related reproductive decrements (Figs. 2–4). prove this because environmental effects were not varied here to de- If expression of Hsp70 is indeed costly, as can be assumed termine extents of thermal plasticity (cf. Blanckenhorn, 2009). (Silbermann and Tatar, 2000; Sisodia and Singh, 2006; Huang et al., The widely distributed yellow dung fly is well known to be unu- 2007), a trade-off between reproductive investment (or other life his- sually cold-adapted (Sigurjónsdóttir and Snorrason, 1995; tory traits, such as e.g. longevity) and Hsp70 expression was expected, Blanckenhorn, 2009) and very sensitive to heat (Ward and Simmons, either (1) due to a reallocation of resources from reproduction towards 1990; Blanckenhorn et al., 2001). Towards the end of the season flies the stress response, or (2) due to a general cessation of protein synthesis are regularly active, mating and ovipositing on pastures at temperatures (other than stress proteins) during times of stress (see Introduction). close to freezing, as they can only successfully overwinter in the pupal However, no such pattern was observed here: heat stress reduced stage (but not as adults). At the same time, continued exposure of adult overall female reproductive investment both via egg number (strongly) flies to temperatures of 25 °C (Reim et al., 2006), 27 °C (Ward and and egg size (weakly), so there was no evidence for egg size-number Simmons, 1990)or31°C(Blanckenhorn et al., 2014)definitely reduces trade-offs(Smith and Fretwell, 1974; which has been previously survival (or “starvation resistance”). We therefore expected some de- documented for yellow dung flies by Blanckenhorn and Heyland, gree of local adaptation to the climate conditions of their habitat, which 2004). Note that it was not possible here to measure reproductive in- however turned out not to be the case for the thermal traits examined vestment and Hsp70 expression level in the same individuals, so these here. Rather, flies from all populations were similarly capable of phy- effects were indirectly derived from family means as the level of re- siologically responding to brief heat (37 °C) and cold shocks (− 20 °C), plication. At minimum, our results suggest that the applied heat stress irrespective of their origin. As shown e.g. for stream invertebrates (Shah treatment was within a tolerable range for the flies so as to not show et al., 2017), it is thus possible that the seasonal (or even daily) range of immediate costs in some traits, for instance the total energy invested temperatures these experience in nature, which for the six sites into the first clutch (see also below). This of course does not exclude the investigated here does not differ so much, is more important for such possibility that stronger or longer periods of stress might result in the adaptation than the overall mean temperature or the temperature ex- expected reduction in these fitness components, nor that a trade-off tremes experienced in their habitat (which do differ strongly; Table 1). may appear later during reproduction, which was not assessed here. After all, heat spells also occur regularly at intermediate and high la- titude sites, albeit more rarely (Hoffmann et al., 2011). 4.2. Heat stress effects on reproductive investment and egg quality We expected heat shock protein expression to be up-regulated in response to heat stress (cf. Feder and Hofmann, 1999). However, in While there was no indication of a trade-off between reproductive yellow dung flies expression of the heat shock protein Hsp70 did not parameters and Hsp70 expression in yellow dung flies, even brief

47 S.S. Bauerfeind et al. Journal of Thermal Biology 73 (2018) 41–49 exposure to heat stress during egg maturation reduced female egg rapidly adjust egg composition to buffer the adverse effects of heat numbers by an average of 8.6% across all populations (cf. Sisodia and stress, or that adjusting egg composition may not be the means of Sing, 2006; Huang et al., 2007). This may be caused by direct inter- choice to mitigate effects of thermal stress. Instead, females might ference with egg production if egg maturation ceased during stress simply adjust their egg laying behavior (Blanckenhorn and Henseler, exposure (unlikely, because it was short), or an indirect effect mediated 2005; Blanckenhorn et al., 2014). by behavioural changes (such as adult feeding) or other physiological processes (including potential involvement of Hsps; Huang et al., 2007; 4.4. Conclusion Blanckenhorn et al., 2014). Blanckenhorn and Henseler (2005) de- monstrated negative effects of elevated temperatures (22–26 °C) on egg While yellow dung flies of all regions appear well equipped to size and number, and Blanckenhorn et al. (2014) previously showed physiologically resist both cold and heat shocks, possibly related to the that the egg stage is less susceptible to heat stress than subsequent ju- permanently high constituent levels of Hsp70 of these cold-adapted venile and adult stages in yellow dung flies. flies, even relatively minor single or double bouts of environmental By contrast, exposure to heat stress had limited impact on egg (here heat) stress during egg maturation nonetheless can profoundly quality, affecting only egg lipid content (reduction of ca. 4%). Other impair reproductive fitness components (Sisodia and Singh, 2006; this measures of egg quality (egg protein and glycogen levels) were un- study). Whereas Hsp70 expression changed little in response to heat affected by heat stress, but were instead strongly associated with egg stress in yellow dung flies, egg output was instead adjusted by way of size and female body size, indicating a strong dependence on female some other maternal effects. Perhaps surprisingly, expected clinal dif- condition (see also Scharf et al., 2010, and below). Nevertheless, there ferentiation patterns were not found (Hoffmann et al., 2002, 2003; Sgrò was some variation among populations in egg protein and, to a lesser et al., 2010a; Schilthuizen and Kellermann, 2014; Stoks et al., 2014). degree, egg lipid content. However again, no large scale geographic (i.e. The ability to buffer the fitness consequences of environmental stress latitudinal) pattern emerged, likely reflecting local rather than sys- seems to confer some costs at benign conditions (apparent in local re- tematic latitudinal adaptation (cf. Leishnam et al., 2008; Potter et al., sponse variation in larval hatching success), suggesting that exposure to 2013; Fig. 3). Thus, the various egg energy components appear differ- heat stress not only produces direct costs, but also that local adaptation entially and independently affected by environmental stress, and egg to stress itself may be costly (in cases an evolutionary response is size, which was only slightly affected by heat stress, does not necessa- mounted). rily reflect changes in egg quality (Bernardo, 1996; Fox and Czesak, 2000; Karl et al., 2007). Acknowledgments

4.3. Larval hatching success We are grateful to U. Briegel for technical assistance, and thank the German Research Council (BA3760 to S.S.B.), the Swiss National Heat stress during the egg maturation stage of the mother ultimately Science Foundation (3100AO-111775 to W.U.B.), the European Science reduced larval hatching success, demonstrating that the egg maturation Foundation (ThermAdapt) and the Zoological Museum, University of period is sensitive to even short bouts of high temperatures (Wang and Zurich, for funding. Several reviewers improved the manuscript. Kang, 2005). This seems peculiar because the eggs – once deposited in the dung patch – should be and indeed are more resistant to high Appendix A. Supplementary material temperatures than the adult flies themselves (Blanckenhorn et al., 2014), as the dung pats into which the eggs are deposited are usually Supplementary data associated with this article can be found in the found to be fully sun-exposed. The lower larval hatching success found online version at http://dx.doi.org/10.1016/j.jtherbio.2018.01.002. here must therefore be a maternal effect. It is possible that heat shock disturbs egg fertilization processes in the female tract (Ward, 2000; References Demont et al., 2011). The adverse effect of heat stress on larval hatching depended Addo-Bediako, A., Chown, S.L., Gaston, K.J., 2000. Thermal tolerance: climatic variability somewhat on population origin (rather than latitude), as the (geo- and latitude. Proc. R. Soc. Lond. B 267, 739–745. Angilletta Jr., M.J., 2009. Thermal Adaptation: A Theoretical And Empirical Synthesis. graphically marginal) northern- and southernmost populations did not Oxford University Press, UK. su ffer as great a reduction in hatching success as the more central po- Berger, D., Bauerfeind, S.S., Blanckenhorn, W.U., Schäfer, M.A., 2011. High temperatures pulations; yet at the same time under favorable conditions their reveal cryptic genetic variation in a polymorphic organ. Evolution 65, 2830–2842. hatching success was lower (Fig. 4). This would either suggest that Bernardo, J., 1996. The particular maternal effect of propagule size, especially egg size: marginal populations show lower egg hatchability overall but are at the patterns, models, quality of evidence and interpretations. Am. Zool. 36, 216–236. same time better at buffering the effects of heat stress, or that the lower Blanckenhorn, W.U., 2009. Causes and consequences of phenotypic plasticity in body fl hatching success in the marginal populations is a direct consequence or size: the case of the yellow dung y Scathophaga stercoraria (Diptera: Scathophagidae). In: Whitman, D.W., Ananthakrishnan, T.N. (Eds.), Phenotypic cost of the ability to buffer the effects of heat stress on egg hatchability. Plasticity of Insects: Mechanism and Consequences. Science Publishers, Enfield, NH, However, beyond documenting some geographic variation in stress USA, pp. 369–422. Blanckenhorn, W.U., Henseler, C., 2005. Temperature-dependent ovariole and testis resistance independent of latitude, perhaps related to the experienced fl – ff maturation in the yellow dung y. Entomol. Exp. Appl. 116, 159 165. temperature range (Table 1)ordi erences in insolation regimes not Blanckenhorn, W.U., Heyland, A., 2004. The quantitative genetics of two life history treated here (cf. Busso and Blanckenhorn, 2018), conclusions here must trade-offs in the yellow dung fly in abundant and limited food environments. Evol. – remain limited. Ecol. 18, 385 402. Blanckenhorn, W.U., Henseler, C., Burkhard, D.U., Briegel, H., 2001. Summer decline in Hatching success also depended on egg composition (i.e. quality). populations of the yellow dung fly: diapause or quiescence? Physiol. Entomol. 26, Larger eggs, and eggs with less protein but more glycogen were more 260–265. likely to hatch. These findings reiterate that egg size and composition Blanckenhorn, W.U., Gautier, R., Nick, M., Schäfer, M.A., 2014. Stage- and sex-specific heat tolerance in the yellow dung fly. J. Therm. Biol. 46, 1–9. (i.e. quality) are not synonymous, as both vary independently from Blanckenhorn, W.U., Pemberton, A.J., Bussière, L.F., Römbke, J., Floate, K.D., 2010. each other and affect egg survival and hatching in different ways (Fox Natural history and laboratory culture of the yellow dung fly, Scathophaga stercoraria and Czesak, 2000; Giron and Casas, 2003). It would be interesting to (L.; Diptera: Scathophagidae). J. Insect Sci. 10, 1–17. Busso, J.P., Blanckenhorn, W.U., 2018. Climatic factors shaping plastic trade-offs in the further investigate the quantitative temporal changes of egg resources polyphenic black scavenger fly (Diptera: ). J. Biogeogr. during embryonic development. Together with the low sensitivity of http://dx.doi.org/10.1111/jbi.13140. in press. egg constituents to thermal stress, the adverse effect of heat stress on Carbonell, J.A., Bilton, D., Calosi, P., Millán, A., Stewart, A., Velasco, J., 2017. Metabolic larval hatching success suggests that females are either not able to and reproductive plasticity of core and marginal populations of the eurythermic

48 S.S. Bauerfeind et al. Journal of Thermal Biology 73 (2018) 41–49

saline water bug Sigara selecta (Hemiptera: Corixidae) in a climate change context. J. Scathophagidae). Evol. Ecol. Res. 8, 1215–1234. Insect Physiol. 98, 59–66. Robinson, S.J.W., Partridge, L., 2001. Temperature and clinal variation in larval growth Calosi, P., Bilton, D.T., Spicer, J.I., et al., 2010. What determines a species' geographical efficiency in Drosophila melanogaster. J. Evol. Biol. 14, 14–21. range? Thermal biology and latitudinal range size relationships in European diving Santos, M., Castaneda, L.E., Rezende, E.L., 2011. Making sense of heat tolerance estimates (Coleoptera: Dytiscidae). J. Anim. Ecol. 79, 194–204. in ectotherms: lessons from Drosophila. Funct. Ecol. 25, 1169–1180. Castañeda, L.E., Lardies, M.A., Bozinovic, F., 2004. Adaptive latitudinal shifts in the Scharf, I., Bauerfeind, S.S., Blanckenhorn, W.U., Schäfer, M.A., 2010. Effects of maternal thermal physiology of a terrestrial isopod. Evol. Ecol. Res. 6, 579–593. and offspring environmental conditions on growth, development and diapause in Chen, B., Kang, L., 2004. Variation in cold hardiness of Liriomyza huidobrensis (Diptera: latitudinal yellow dung fly populations. Clim. Res. 43, 115–125. Agromyzidae) along latitudinal gradients. Environ. Entomol. 33, 155–164. Schilthuizen, M., Kellermann, V., 2014. Contemporary climate change and terrestrial Chown, S.L., Gaston, K.J., 1999. Exploring links between physiology and ecology at invertebrates: evolutionary versus plastic changes. Evol. Appl. 7, 56–67. macro-scales: the role of respiratory metabolism in insects. Bio Rev. 74, 87–120. Schou, M.F., Mouridsen, M.B., Sørensen, J.G., Loeschcke, V., 2017. Linear reaction norms Chown, S.L., Gaston, K.J., 2010. Body size variation in insects: a macroecological per- of thermal limits in Drosophila: predictable plasticity in cold but not in heat tolerance. spective. Biol. Rev. 85, 139–169. Funct. Ecol. 31, 934–945. Clarke, A., 2014. The thermal limits to life on Earth. Int. J. Astrobiol. 13, 141–154. Sgrò, C.M., Overgaard, J., Kristensen, T.N., Mitchell, K.A., Cockerell, F.E., Hoffmann, Dahlgaard, J., Loeschcke, V., Michalak, P., Justesen, J., 1998. Induced thermotolerance A.A., 2010a. A comprehensive assessment of geo- graphic variation in heat tolerance and associated expression of the heat-shock protein HSP70 in adult Drosophila mel- and hardening capacity in populations of Drosophila melanogaster from eastern anogaster. Funct. Ecol. 12, 786–793. Australia. J. Evol. Biol. 23, 2484–2493. Demont, M., Buser, C.C., Martin, O.Y., Bussière, L.F., 2011. Natural levels of polyandry: Sgrò, C.M., Wegener, B., Hoffmann, A.A., 2010b. A naturally occurring variant of Hsp90 differential sperm storage and temporal changes in intensity in that is associated with decanalization. Proc. R. Soc. Lond. B 277, 2049–2057. wild yellow dung flies. Funct. Ecol. 25, 1079–1090. Shah, A.A., Gill, B.A., Encalada, A.C., Flecker, A.S., Funk, W.C., Guayasamin, J.M., Esperk, T., Kjaersgaard, A., Walters, R.J., Berger, D., Blanckenhorn, W.U., 2016. Plastic Kondratieff, B.C., Poff, N.L., Thomas, S.A., Zamudio, K.R., Ghalambor, C.K., 2017. and evolutionary responses to heat stress in a temperate dung fly: negative correla- Climate variability predicts thermal limits of aquatic insects across elevation and tion between basal and induced heat tolerance? J. Evol. Biol. 29, 900–915. latitude. Funct. Ecol. 31, 2118–2127. Feder, J.H., Rossi, J.M., Solomon, J., et al., 1992. The consequences of expressing hsp70 Sigurjónsdóttir, H., Snorrason, S.S., 1995. Distribution of male yellow dung flies around in Drosophila cells at normal temperatures. Genes Dev. 6, 1402–1413. oviposition sites: the effect of body size. Ecol. Entomol. 20, 84–90. Feder, M.E., Hofmann, G.E., 1999. Heat shock proteins, molecular chaperones, and the Silbermann, R., Tatar, M., 2000. Reproductive costs of heat shock protein in transgenic stress response: evolutionary and ecological physiology. Annu. Rev. Physiol. 61, Drosophila melanogaster. Evolution 54, 2038–2045. 243–282. Sinclair, B.J., Vernon, P., Klok, C.J., Chown, S.L., 2003. Insects at low temperatures: an Foray, V., Pelisson, P.F., Bel-Venner, M.C., Desouhant, E., Venner, S., Menu, F., Giron, D., ecological perspective. TREE 18, 257–262. Rey, B., 2012. A handbook for uncovering the complete energetic budget in insects: Sisodia, S., Singh, B.N., 2006. Effect of exposure to short-term heat stress on survival and the van Handel's method (1985) revisited. Physiol. Entomol. 37, 295–302. fecundity in Drosophila ananassae. Can. J. Zool. 84, 895–899. Fox, C.W., Czesak, M.E., 2000. Evolutionary ecology of progeny size in . Ann. Smith, C.C., Fretwell, S.D., 1974. The optimal balance between size and number of off- Rev. Entomol. 45, 341–369. spring. Am. Nat. 108, 499–506. Gaston, K.J., Chown, S.L., Evans, K.L., 2008. Ecogeographical rules: elements of a Sørensen, J.G., Kristensen, T.N., Loeschcke, V., 2003. The evolutionary and ecological synthesis. J. Biogeogr. 35, 483–500. role of heat shock proteins. Ecol. Lett. 6, 1025–1037. Giron, D., Casas, J., 2003. Mothers reduce egg provisioning with age. Ecol. Lett. 6, Sørensen, J.G., Kristensen, T.N., Overgaard, J., 2016. Evolutionary and ecological pat- 273–277. terns of thermal acclimation capacity in Drosophila: is it important for keeping up Gorodkov, K.B., 1984. In: Soós, A., Papp, L. (Eds.), Scathophagidae. Catalogue of with climate change? Curr. Opin. Insect Sci. 17, 98–104. Palaearctic Diptera. Elsevier, Amsterdam, The Netherlands, pp. 11–41. Stoks, R., Geerts, A.N., de Meester, L., 2014. Evolutionary and plastic responses of Hoffmann, A.A., Anderson, A., Hallas, R., 2002. Opposing clines for high and low tem- freshwater invertebrates to climate change: realized patterns and future potential. perature resistance in Drosophila melanogaster. Ecol. Lett. 5, 614–618. Evol. Appl. 7, 42–55. Hoffmann, A.A., Sørensen, J.G., Loeschcke, V., 2003. Adaptation of Drosophila to tem- Van Handel, E., 1985a. Rapid determination of total lipids in mosquitoes. J. Am. Mosq. perature extremes: bringing together quantitative and molecular approaches. J. Control Assoc. 1, 299–301. Therm. Biol. 28, 175–216. Van Handel, E., 1985b. Rapid determination of glycogen and sugars in mosquitoes. J. Am. Hoffmann, A.A., Chown, S.L., Clusella-Trullas, S., 2011. Upper thermal limits in terrestrial Mosq. Control Assoc. 1, 302–303. ectotherms: how constrained are they? Funct. Ecol. 27, 934–949. Van Handel, E., Day, J.F., 1988. Assay of lipids, glycogen and sugars in individual mos- Huang, L.-H., Chen, B., Kang, L., 2007. Impact of mild temperature hardening on ther- quitoes: correlations with wing length in field-collected Aedes vexans. J. Am. Mosq. motolerance, fecundity, and hsp gene expression in Liriomyza huidobrensis. J. Insect Control Assoc. 4, 549–550. Physiol. 53, 1199–1205. Velazquez, J.M., Sonoda, S., Bugaisky, G., Lindquist, S., 1983. Is the major Drosophila Karl, I., Lorenz, M.W., Fischer, K., 2007. Energetics of reproduction: consequences of heat shock protein present in cells that have not been heat shocked? J. Cell Biol. 96, divergent selection on egg size, food limitation and female age for egg composition 286–290. and reproductive effort in a butterfly. Biol. J. Linn. Soc. 91, 403–418. Wang, X.H., Kang, L., 2005. Differences in egg thermotolerance between tropical and Kellermann, V., Overgaard, J., Hoffmann, A.A., Fløjgaard, C., Svenning, J.C., Loeschcke, temperate populations of the migratory locust Locusta migratoria (Orthoptera: V., 2012. Upper thermal limits of Drosophila are linked to species distributions and Acridiidae). J. Insect Physiol. 51, 1277–1285. strongly constrained phylogenetically. Proc. Natl. Acad. Sci. USA 109, 16228–16233. Wang, X.G., Johnson, M.W., Daane, K.M., Nadel, H., 2009. High summer temperatures Leishnam, P.T., Sala, L.M., Juliano, S.A., 2008. Geographic variation in adult survival and affect the survival and reproduction of olive fruit fly (Diptera: Tephritidae). Physiol. reproductive tactics of the mosquito Aedes albopictus. J. Med. Entomol. 45, 210–221. Entomol. 38, 1496–1504. National Institutes of Health, 2007. ImageJ 1.38x. USA. Ward, P.I., 2000. Cryptic female choice in the yellow dung fly Scathophaga stercoraria. Partridge, L., French, V., 1996. Thermal evolution of ectotherm body size: why get big in Evolution 54, 1680–1686. the old? In: Johnston, I.A., Bennett, A.F. (Eds.), Phenotypic and Evolutionary Ward, P.I., Simmons, L.W., 1990. Short-term changes in numbers of the yellow dung fly Adaptation. Cambridge University Press, Cambridge, pp. 265–292. Scathophaga stercoraria. Ecol. Entomol. 15, 115–118. Potter, K.A., Arthur Woods, H., Pincebourde, S., 2013. Microclimatic challenges in global Zatsepina, O.G., Przhiboro, A.A., Yushenova, I.A., Shilova, V., Zelentsova, E.S., Shostak, change biology. Glob. Change Biol. 19, 2932–2939. N.G., Evgeniev, M.B., Garbuz, D.G., 2016. A Drosophila heat shock response re- Reim, C., Teuschl, Y., Blanckenhorn, W.U., 2006. Size-dependent effects of temperature presents an exception rather than a rule amongst Diptera species. Insect Mol. Biol. 25, and food stress on energy stores and survival in yellow dung flies (Diptera: 431–449.

49