Invasion and adaptation of a warm-adapted species to montane localities: effect of acclimation potential

Geographic distributions of species are constrained by several factors acting at different scales, with climate assumed to be a major determinant at broad extents. Environmental stresses such as desiccation, cold and heat are main barriers that restrict the distribution and abundance of Drosophila species (Kellermann et al., 2009; Kellermann et al., 2012). Recently, Drosophila ananassae has invaded successfully to the montane localities of the Western Himalayas and climate change has been considered as the primary reason for invasion of the cold-sensitive D. ananassae (Rajpurohit et al., 2008a; Rajpurohit et al., 2008b). Under the influence of changing climatic conditions, an organism can either adapt or shift the range or undergo extinction. Drosophila species from the warm humid tropics may be threatened if they cannot adapt to drier and colder conditions of montanes. Some studies have found evidence that desert insects are able to tolerate lower water levels or store more water (Hadley, 1994; Gibbs and Matzkin, 2001; Gibbs et al., 2003). Moreover, desiccation resistance in Drosophila can be increased by selection causing accumulation of more bulk water in selected lines than controls (Gibbs et al., 1997; Folk et al., 2001). An earlier study (Blows and Hoffmann, 1993) showed that selection resulted in reduced metabolic rates under desiccating conditions. The authors suggested that reduced respiratory water loss was responsible for increased desiccation resistance. Further, adaptations can be achieved by phenotypic plasticity (Bradshaw and Holzapfel, 2001). Desiccation resistance can be increased with phenotypic plasticity either of cuticular lipids (Parkash et al., 2008) or of melanisation (Parkash et al., 2009) under different environmental conditions. Presence of melanism thickens the cuticle and acts as a barrier to evaporation through the cuticle. In this way, melanism increases desiccation resistance by reducing cuticular water loss (Parkash et al., 2009). Further, the thickness provided by melanisation also insulates flies from cold (Parkash et al., 2010). However, in species like D. ananassae, there is no reaction norm (thermal-induced plasticity) for melanisation (Rajpurohit et al., 2008b), so alternative strategy of adaptation might be expected.

The acclimation ability or response to non lethal stress increases the resistance to desiccation stress (Hoffmann, 1990; Hoffmann, 1991) and thermal stresses (Hoffmann and Watson, 1993; Kristensen et al., 2008). Different acclimation treatments (developmental, gradual and rapid) had shown very beneficial effects on cold tolerance resulting in tolerant phenotypes of D. melanogaster (Colinet and Hoffmann, 2012). Many studies have examined the effects of developmental acclimation in Drosophila (Kristensen et al., 2008; Colinet and Hoffmann, 2012). Developmental acclimation has increased starvation and heat knockdown time in D. buzzatii and thus affecting clinal variation of stress resistance traits (Sarup and Loeschcke, 2010). The adult acclimation for cold resistance in D. melanogaster increases the lifespan, reduces mortality and recovery time after exposure to subzero temperatures (Rako and Hoffmann, 2006; Le Bourg, 2007). Further, the dark morph of D. ananassae is more resistant to cold and desiccating conditions as compared to the light morph; while the latter is more resistant to heat stress (Parkash et al., 2012). Thus, we may expect evolutionary responses to natural selection on traits related to desiccation and cold stress in subtropical populations of D. ananassae.

In the generalist Drosophila species, there is huge genetic variation for adaptations to spatially and temporally varying climatic conditions (Powell, 1997; Hoffmann and Weeks, 2007). By contrast, two rainforest Drosophila species of the montium species subgroup – D. serrata (Hallas et al., 2002) and D. birchii (Hoffmann et al., 2003), have low potential for climatic stress adaptations which match with their restricted distribution patterns. It is possible that absence of acclimation ability in D. birchii (Hoffmann, 1991) has resulted in its restricted distribution. However, desiccation and cold sensitive D. ananassae has extended its boundaries to adverse climatic conditions of the western Himalayas. But, it is not clear how it survives and proliferates under drier environments as there are seasonal and diurnal variations in thermal as well as humidity conditions in northern subtropics of the Indian subcontinent.

The objective of the present study is to explore the reason for successful adaptation of a sensitive stenothermal species to montane localities of the Western Himalayas. The questions addressed are: (1) has the frequency of D. ananassae body color morphs changed in past few years; (2) does the developmental and adult acclimation to stressful conditions increase stress tolerance of the morphs; (3) do the morphs differ in their acclimation responses for various stresses; and (4) do the number of acclimation treatments and duration of stress make any difference in stress tolerance?

The present study focuses on adaptation of a desiccation and cold sensitive species, D. ananassae in the montane localities after successful invasion. We sampled D. ananassae from five mid-altitude localities (Dharamshala, Nauni, Kandaghat, Solan and Shoghi) of the Western Himalayas. We analysed how D. ananassae copes with wet/dry and high/low temperature conditions in spite of its lack of plasticity for melanisation and its sensitivity for dry environment. We investigated developmental and adult acclimation ability of two body color morphs for environmental stresses and the persistence of acclimation effect.

Drosophila ananassae Doleschall 1858 were collected twice a month from February to May and August to November 2010 from five mid-altitude localities of the Western Himalayas [Dharamshala (1219 m), Nauni (1300 m), Kandaghat (1432 m), Solan (1440 m) and Shoghi (1833 m)] with bait trap and net sweeping methods. Wild-caught flies showed dark versus light total body color dimorphism for all the six abdominal segments of both sexes. The wild-caught flies were examined for frequency of light and dark phenotypes and the frequencies thus obtained were compared with frequencies in 2000 (R.P., unpublished data). Both collections (2000 and 2010) were made in a similar manner.

All acclimation experiments were carried out on 10 homozygous (for phenotype) strains for the dark as well as light morphs (see Parkash et al., 2012). These strains were originally derived from true breeding (for body color) iso-female lines from Solan. These strains were grown at 25°C for six generations on cornmeal-yeast-agar medium under a 12 h:12 h light:dark photoperiod prior to experiments. Two types of acclimation treatments were conducted: (1) a developmental acclimation treatment, involving manipulation of the temperatures that pre-adult stages experience, but no manipulation of temperature for adult stages (i.e. kept at 25°C); and (2) an adult acclimation treatment, in which pre-adult stages developed at a common temperature, i.e. 25°C, but adults were exposed to different temperature and humidity conditions [partially following Hoffmann (Hoffmann, 1990)]. The experimental design is represented schematically in Fig. 1.

The 20 true breeding strains (10 for the dark and 10 for light body color) were allowed to lay eggs at 25°C in three replicates. One replicate (out of three) with eggs was transferred to one of three different growth temperatures (i.e. one at 20°C, one at 25°C and one at 30°C). All flies developed at three different temperatures were placed at 25°C after eclosion. After 6 days, assays were performed to check the effect of developmental acclimation on desiccation tolerance, cold recovery time (after 5 h cold stress at 0°C) and heat knockdown time.

Two different experiments (acclimation and persistence effect) were carried out to study adult acclimation benefits for stress-related traits. Controls were not acclimatized. Ten replicates of 10 flies each from homozygous strains were used for every treatment for both morphs.

Both morphs were given a prior desiccation stress of 3 h in three groups in 10 replicates each. Flies of the first, second and third groups were stressed once, twice and three times, respectively, with an interval of 12 h before they were left to recover for 12 h prior to measurement of desiccation resistance.

Ten replicates of 10 flies each from all strains (10 dark and 10 light) were subjected to desiccation treatment twice for 3 h each with an interval of 12 h and allowed to recover on food for 2, 4, 6, 8 and 10 days before they were tested for resistance to desiccation stress.

Ten replicates of five flies each for both morphs were given prior chilling of 1 h in three groups (acclimated once, twice and three times) as described above. Percent mortality of flies and recovery time after chill-coma stress (at 0°C) as a function of stress duration (1, 3 and 5 h) was recorded.

Ten replicates of five flies each from the 10 dark strains were given cold treatment of 1 h at 0°C (similarly in three groups as in desiccation) and then allowed to recover on cornmeal-yeast-agar food medium for 2, 4, 6, 8 and 10 days, respectively, before measuring percent mortality and recovery time after 5 h of cold stress at 0°C.

Flies were initially given 5 min heat stress at 39°C in three different groups for each of the three acclimation treatments, with 12 h interval between each treatment and before measuring the final effect in the form of knockdown time.

Persistence of the acclimation ability for heat was checked by placing the heat-treated flies (10 replicates of five flies each from the 10 light strains) on food for 2, 4, 6, 8 and 10 days before finally scoring the heat knockdown period at 39°C.

To measure desiccation resistance, 10 individuals from each of the 10 dark and 10 light strains from both acclimation treatments (developmental and adult) were isolated in a dry plastic vial, which contained 2 g of silica gel at the bottom of each vial, and were covered with a foam disc. The vials were then placed in a desiccation chamber (Secador electronic desiccator cabinet, Tarsons Products, Kolkata, West Bengal, India) that maintained 1–2% relative humidity at 25°C. The vials were inspected every hour and the number of dead flies (completely immobile) was recorded.

For thermoresistance traits, 7-day-old flies from both acclimation treatments (developmental and adult) were isolated in separate vials after mild anesthesia with vapors of diethyl ether for a maximum of 30 s followed by a 1 day recovery period at 25°C. Heat knockdown was measured individually on 10 homozygous strains each of the dark as well as the light morph (five flies × 10 strains). For the heat knockdown assay, individual flies were placed in 5 ml glass vials submerged into a water bath at a constant temperature of 39°C. Flies were scored for the time (min) taken to be knocked down.

Further, for chill-coma recovery, 10 groups of 10 flies either for the dark or light morph from both acclimation treatments were transferred without anesthesia into empty 5 ml glass vials. These vials were set in thermocool boxes (24×13×10 cm) containing ice flakes (made with an ice flaking machine, Aicil Lab Instruments, Ambala, Haryana, India) and were kept at 0°C in the refrigerator. The vials were removed after 5 h (for developmental acclimation treatment flies) and after variable time periods, i.e. 1, 3 or 5 h (for adult acclimation treatment flies); flies were transferred to Petri dishes (9 cm diameter) in a temperature-controlled room at 25°C; the cold recovery period (min) for each fly was recorded.

Data were checked for normality and homoscedasticity using a normal probability plot and a Kolmogorov–Smirnov test, respectively. All the variables showed a normal distribution, so parametric statistics were used. Two-way ANOVA was applied to determine the effect of morph, developmental acclimation treatment and their interaction on various stress-related traits. One-way ANOVA was used to compare the effects of multiple adult acclimation treatments on chill-coma resistance in dark and light morphs. Comparison of stress tolerance between multiple acclimation treatments for desiccation and heat knockdown was carried out using one-way ANOVA for the dark and light morphs separately. Further, for testing for differences in the acclimation treatments between the two morphs, factorial ANOVA was applied to determine the interaction effect for various stress assays (data shown in supplementary material Tables S1, S2 and S3). The persistence of acclimation effect for the traits was compared using a Newman–Keuls post hoc test. STATISTICA software (releases 5.0 and 7.0, StatSoft, Tulsa, OK, USA) was used for calculations as well as illustrations.

The 10 year collection record (2000–2010) documented onset and increase in frequency of the dark morph (from 0 to 0.18) and decrease in the light morph (1.0 to 0.61) at the five mid-altitude localities of the Western Himalayas. In 2000, only the light morph was prevalent, whereas in 2010, the dark morph was also present at a considerable frequency (supplementary material Fig. S1). An intermediate morph (F₁ of dark and light morph) was also abundant in collections from 2010 (supplementary material Fig. S1), but in the present study we only analyzed two contrasting pure body color morphs (the dark and light).

The homozygous laboratory strains of dark and light morphs lack developmental thermal acclimation (supplementary material Fig. S2) when tested at three temperatures (20, 25 and 30°C). ANOVA results (Table 1) showed significant differences between morphs for all three stresses. However, no effect of developmental acclimation temperature was found on desiccation resistance, chill-coma recovery time or heat knockdown time in D. ananassae (Table 1). Further, no significant interaction effect between morphs and developmental acclimation treatments was observed (Table 1).

One-way ANOVA (Table 2) showed a significant effect of adult acclimation on the dark morph when controls were included (F=168.06, P<0.001), but this was not significant when controls were excluded (F=1.62, P=0.121). There was no difference between desiccation resistance of flies after one, two or three acclimation treatments (Table 2), which is why the data were pooled and are shown as a single bar (Fig. 2A). The light morph showed no acclimation potential for desiccation resistance (Table 2). Further, there was a significant interaction effect between morphs and multiple desiccation acclimation treatments (supplementary material Table S1). Acclimated dark flies had significantly higher desiccation resistance even after 10 days of prior acclimation compared with controls (Newman–Keuls post hoc test, P<0.001, d.f.=54, MS error=0.793; Fig. 2B). However, the desiccation resistance of flies did not differ significantly between the second and fourth days (P=0.47), the second and sixth days (P=0.11) or the second and eighth days (P=0.12) after acclimation, showing a high persistence of acclimation effect up to the eighth day according to a Newman–Keuls post hoc test. But the resistance differed significantly between the eighth and the tenth days of acclimation (Newman–Keuls post hoc test, P=0.001; Fig. 2B), indicating that the effect of acclimation had declined after 8 days.

Two-way ANOVA for cold acclimation showed a significant effect of acclimation treatment for the dark morph for percent cold mortality (Fig. 3A, Table 3) and chill-coma recovery (Fig. 3B, Table 3), but this was not significant for the light morph (Table 3). However, stress duration showed a significant effect for both morphs in measures of cold resistance (Table 3). The interaction effects between morph, acclimation treatment and duration of stress were highly significant (supplementary material Table S2). The differences between different acclimation treatments were more distinct under 5 h cold stress compared with 1 or 3 h for both measures of cold resistance (Fig. 3A, percent mortality; Fig. 3B, recovery time). The flies of the dark morph differed significantly between control and 10 days after acclimation for mortality (P<0.001, d.f.=180, MS error=0.139; Fig. 3C) and for recovery time (P<0.001, d.f.=180, MS error=0.139; Fig. 3D) in all three acclimation treatments, when means were compared with a Newman–Keuls post hoc test, showing the effect of cold acclimation. However, the percent mortality of cold-acclimated dark flies did not differ among the second, fourth and sixth days after acclimation for three treatments (Newman–Keuls post hoc test, 1.00<P>0.05; Fig. 3C). Further, the percent cold mortality differed significantly between the sixth and eighth days of recovery on food for all three acclimation treatments (Newman–Keuls post hoc test, P<0.001; Fig. 3C), indicating a decrease in the effect of cold acclimation after 6 days. The cold recovery time did not differ between the second and fourth days after three cold acclimation treatments (Newman–Keuls post hoc test, P=0.75), but differed significantly between the fourth and sixth days (P<0.001), indicating that the acclimation effect persisted up to 4 days and declined after that (Fig. 3D). However, flies of the dark morph acclimated once and twice differed among all days after acclimation in terms of persistence (Newman–Keuls post hoc test, P<0.001), showing a continuous decline in acclimation effect (Fig. 3D).

One-way ANOVA showed significant effects of acclimation/hardening treatments on the light morph in the three groups, regardless of whether controls were included (F=2701.3, P<0.001) or excluded (F=1579.9, P<0.001; Table 4, Fig. 4A), but effects were not significant for the dark morph (Table 4, Fig. 4A). Further, both morphs differed significantly in their interaction with subsequent acclimation treatments (supplementary material Table S3). The heat knockdown time between the second and sixth days after heat acclimation treatment did not differ significantly (Newman–Keuls post hoc test, 1.00<P>0.05, d.f.=180, MS error=0.302; Fig. 4B) for the three heat acclimation treatments, showing strong persistence of heat acclimation up to 6 days after treatment. However, after the sixth day the effect of acclimation decreased (sixth versus eight day of recovery, Newman–Keuls post hoc test, P<0.001; Fig. 4B). The difference in heat resistance between control versus acclimated flies had completely vanished on the 10th day of recovery (Newman–Keuls post hoc test, P=0.32) in flies that were acclimated once, whereas for flies acclimated two or three times, the differences remained (Newman–Keuls post hoc test, P<0.001).

Insects, particularly drosophilids, are known as good biological indicators for climate change as they respond to minor fluctuations in temperature either by changing their adaptive behavior or by shifting their boundaries. Climatic suitability depends upon three factors: deviation from optimum conditions, extent of variation among years and extent of variation within a year (Danks, 1999; Danks, 2007). Further, spatial complexity of habitats is one template for the evolution of seasonal adaptations. Most organisms experience high thermal variations in the environment and this poses substantial challenges for key elements of fitness such as survival and reproduction (Dahlhoff and Rank, 2007); therefore, temperature is considered to be an important selective agent (Hoffmann et al., 2003). In the present study, an onset and increase in the frequency of dark morph of D. ananassae was observed at mid-altitude localities over the past 10 years. The adaptation of individual species in a particular habitat depends on two main factors: (1) constraint, i.e. adversity, which in the present case may be a seasonal cycle with changes in temperature and humidity; and (2) need, i.e. suitability, which may be the increased temperature as a result of global warming creating optimum conditions for the species in montane habitats. The first factor may have led to the evolution and increase in the frequency of dark morph and the second factor may have helped the species to invade montane habitats.

In a previous study on D. buzzatii, flies developed at fluctuating temperatures had a higher resistance to stress than flies developed at a constant temperature, thus affecting clinal variation and indicating the role of developmental acclimation in stress resistance (Sarup and Loeschcke, 2010). Further, field releases of D. melanogaster on two continents across a range of temperatures showed enormous benefits of developmental as well as adult acclimation at low temperature in the field (Kristensen et al., 2008). However, in the present study, no benefit of developmental temperature acclimation was found on stress resistance of body color morphs in D. ananassae. This is in accordance with the fact that D. ananassae also lack developmental plasticity for body color and melanisation at range of temperatures (Rajpurohit et al., 2008b; Parkash et al., 2012). One reason may be that the flies were kept at a common temperature after eclosion, which did not allow for adult acclimation. In contrast, D. melanogaster flies were kept at the same temperature as adults and during development in a previous study (Kristensen et al., 2008), which possibly resulted in adult acclimation. Further, a study on the tsetse fly Glossina pallidipes demonstrated that the stage at which acclimation occurs has significant effects on adult physiological traits (Terblanche and Chown, 2006). The study on G. pallidipes has shown that adult acclimation had a larger effect on critical thermal minima than developmental acclimation. Furthermore, Colinet and Hoffmann (Colinet and Hoffmann, 2012) reported that gradual thermal acclimation of a few days was beneficial for cold recovery time in D. melanogaster, while no such result was found for developmental acclimation. Similarly, in the present study there was no increase in stress resistance as a result of developmental acclimation in D. ananassae. Thus, we can say that developmental temperature does not show any significant contribution towards adaptation of D. ananassae when adults face a different temperature than pre-adult stages.

Adult acclimation is the other form of plasticity used to counter environmental stresses (Hoffmann, 1990; Hoffmann, 1991; Hoffmann and Watson, 1993). Acclimation to a particular environment enhances performance, which proves beneficial for adaptation (Leroi et al., 1994). Tropical and rainforest species inhabiting high-humidity conditions have been found to be more desiccation sensitive than temperate or arid species (Parsons, 1983), suggesting adaptive significance of the acclimation responses. The variations between species and geography have been examined for acclimation potential for heat resistance (Levins, 1969) and for desiccation tolerance (Hoffmann, 1990; Hoffmann, 1991). However, to the best of our knowledge, the acclimation potential of morphs of any Drosophila species had not been analysed until now.

Acclimation responses were indicated after different types of thermal treatments in populations of D. melanogaster and D. simulans (Hoffmann and Watson, 1993; Watson and Hoffmann, 1996) and D. buzzatii (Krebs and Loeschcke, 1996). Long-term cold survival of D. melanogaster was increased significantly after low-temperature acclimation (Overgaard et al., 2008). Apart from temperature acclimation, geographical variation in acclimation responses for desiccation stress in D. melanogaster and D. simulanswere also found (Hoffmann, 1990). Acclimation to low humidity can significantly increase desiccation tolerance. Some studies have shown strong responses of insect desiccation rate to changes in relative humidity (Gibbs et al., 2003; Hoffmann, 1990; Hoffmann, 1991). Further, in the present study the two morphs differed in their acclimation response to desiccation stress. Desiccation tolerance after acclimation increased in the dark morph, but there was no effect on tolerance in the light morph.

The majority of research on cold tolerance in insects has focused on seasonal acclimation, which is a result of cold hardiness acquired during overwintering. Cold hardening is described as a process induced by exposure to lower temperature for a short time (Lee et al., 1987) and, after some time, protection against otherwise lethal temperatures is gained by the insect. Further, a combined cold acclimation treatment had a strong impact on field stress resistance in D. melanogaster (Colinet and Hoffmann, 2012). In present study, the dark morph of D. ananassae gained some protection against low temperature and low humidity by acclimation, while the light morph remained sensitive. The dark morph of D. ananassae is more desiccation and cold tolerant compared with the light morph, while the latter is more tolerant to heat (Parkash et al., 2012). In spite of the above fact (respective high resistance of the morphs), the dark morph has high acclimation ability for desiccation and cold stress while the light morph shows high acclimation ability for heat stress. From this observation we may postulate that these morphs are still under divergent selection for stress-resistant traits in nature and that selection will be complete when there is no response to further acclimation or prior stress, as shown in desiccation-selected D. melanogaster, which did not respond to prior desiccation treatment or acclimation (Hoffmann, 1990).

Interspecific studies on heat acclimation have shown significant differences between D. melanogaster and D. simulans (Levins, 1969). Hardening treatments at high temperature resulted in an increase in thermo-tolerance for eight Drosophila species (Kellett et al., 2005) and in codling moths, Cydia pomonella (Chidawanyika and Terblanche, 2011). However, until now, there has been a gap in the research of acclimation potential for heat in morphs of a species. In present study, the light morph of D. ananassae showed higher acclimation potential for heat stress compared with the dark morph, with increasing heat tolerance after subsequent heat treatments. The heat knockdown time of the dark morph was unaffected by heat hardening treatments.

The frequency and duration of stress experienced are major factors that strongly influence and define the acclimation ability of a species. However, the number of times of exposure (acclimation) to desiccation stress did not affect the level of increase in desiccation resistance in the dark morph in the present study. But with increasing frequency of cold exposures, the cold acclimation potential of the dark morph increased; thus percent mortality and recovery time during cold stress decreased. Further, duration of cold stress also affected the mortality and recovery of morphs (Fig. 3A,B). However, the dark morph was not acclimated to heat, but the heat tolerance (in terms of knockdown time) of the light morph increased with every acclimation treatment.

The acquired tolerance to stress persists for several days after initial stress exposure. The desiccation acclimation effect is known to persist for up to 24 h in D. melanogaster (Hoffmann, 1990). In the present study, the acquired resistance persisted for longer durations (10 days; Fig. 2B). The effect of cold acclimation also persisted for up to 10 days; however, with every advancing day the acclimation effect became less effective (Fig. 3C,D). In the light morph also, the hardening effect persisted in the form of high heat knockdown time in hardened flies in comparison to control flies (Fig. 4B).

According to the present study, adult acclimation ability has increased the stress tolerance, which has possibly led the species to adaptation. The increased adult acclimation ability of morphs of D. ananassae may be a result of the absence of developmental plasticity in these genotypes, as some costs have been postulated as a way of accounting for the absence of genotypes with a very high degree of plasticity (Heslop-Harrison, 1964; Bradshaw, 1965). Drosophila ananassae have responded towards stress in different ways, showing inter-morph variation in acclimation potential, which suggests that two morphs of the same species may use different adaptive strategies according to their need. For instance, the dark morph has better acclimation potential under desiccating conditions compared with the light morph, while the light morph has more acclimation potential for heat resistance, which may help the species to survive during hot summer days. The high acclimation potential of the dark morph for cold and desiccating conditions has helped this sensitive species to survive during adverse dry climatic conditions during seasonal change, and thus supports invasion. The respective acclimation potential of the two morphs can also account for the changing frequency of morphs during two different seasons in the Western Himalayas. According to a recent study (Colinet and Hoffmann, 2012), stress acclimation experienced by adults just before chronic stress is more influential than acclimation experienced during development. The present study supports the results of Colinet and Hoffmann (Colinet and Hoffmann, 2012), as developmental acclimation did not result in any increase in stress resistance, whereas adult acclimation showed enormous benefits.

In conclusion, increased frequency of the dark morph in the past few years may have enabled D. ananassae to invade and adapt successfully in seasonally varying montane localities. This suggests that natural selection and evolution are more firmly associated with gradual changes in the environment of any organism, giving rise to a graded response. Further, it seems like acclimation to one extreme decreases the capacity to acclimate under the opposite extreme, as the dark morph can acclimate to colder and desiccating conditions but not to high temperatures, while the reverse is true for light morph. Also, the number of acclimation treatments increases the tolerance for thermal stresses and this increase in stress tolerance persists for longer in D. ananassae. Thus, the different acclimation abilities of the two morphs have enabled this sensitive species to increase some tolerance to stress. Further, field testing should be carried out to determine whether the high acclimation potential of D. ananassae enables this species to adapt under the adverse conditions of montane localities.

We are indebted to the two anonymous reviewers for constructive comments towards improvement of the manuscript.