Sub-Antarctic Freshwater Invertebrate Thermal Tolerances: An Assessment of Critical Thermal Limits and Behavioral Responses

Physiological thermal limits of organisms are linked to their geographic distribution. The assessment of such limits can provide valuable insights when monitoring for environmental thermal alterations. Using the dynamic critical thermal method (CTM), we assessed the upper (CTmax) and lower (CTmin) thermal limits of three freshwater macroinvertebrate taxa with restricted low elevation distribution (20 m a.s.l.) and three taxa restricted to upper elevations (480 and 700 m a.s.l.) in the Magellanic sub-Antarctic ecoregion of southern Chile. In general terms, macroinvertebrates restricted to lower altitudinal ranges possess a broader thermal tolerance than those restricted to higher elevations. Upper and lower thermal limits are significantly different between taxa throughout the altitudinal gradient. Data presented here suggest that freshwater macroinvertebrates restricted to upper altitudinal ranges may be useful indicators of thermal alteration in their habitats, due to their relatively low tolerance to increasing temperatures and the ease with which behavioral responses can be detected.


Introduction
Understanding the mechanisms by which species distributions change along geographical gradients has been a key tool in biogeography, ecology, and evolution [1], and in the development of macroecological theories. Latitudinal and altitudinal gradients generally encompass several environmentally associated variables and can be considered ecological and evolutionary equivalents in terms of their influences on species adaptations [2]. The variations in and expression of species physiological traits are a direct response to the environment [2,3]. The "climate variability hypothesis" (CVH) predicts that more variable climates may select for organisms with broader thermal tolerances, whereas less variable (stable) climates select for narrower thermal tolerances [4][5][6][7][8]. High latitude environments are often characterized by broad thermal variation while low latitude or tropical world's cleanest rainwater, as it is located to the south of the typical tracks of industrial-polluted wind currents [30][31][32]. Navarino Island lies in the southern part of this ecoregion. Freshwater invertebrates (amphipods, diving beetles, midges, stoneflies, water boatmen, planarians) were collected from lagoons located within the Róbalo River watershed, which provides domestic water supply to Puerto Williams, the southernmost town in the world (Figure 1).
Insects 2020, 11, x FOR PEER REVIEW 3 of 13 Freshwater invertebrates (amphipods, diving beetles, midges, stoneflies, water boatmen, planarians) were collected from lagoons located within the Róbalo River watershed, which provides domestic water supply to Puerto Williams, the southernmost town in the world (Figure 1).

Site Description
At the coastline, the average air annual temperature is 5.7 °C, while above the tree line (~500-600 m a.s.l.) it is 0 °C [33]. The average annual temperature of one of the rivers within Navarino Island is 5.7 °C at 120 m a.s.l. and only 1 °C at 586 m a.s.l. [34]. The temperature decreases by approximately 1 °C per 100 m of elevation increase, compared to the global average decrease of 0.6 °C per 100 m of elevation [35].

Water Temperature
Temperature loggers (HOBO ® model U22 Water Temp Pro Version 2) were installed at a 20 to 30 cm depth in the littoral areas of each lagoon (Castor, el Salto and Bandera, 20, 480 and 700 m. a.s.l., respectively, Figure 1) from which macroinvertebrates were collected over the entire study period (from February 2015 to March 2016). These recorded the water temperature every 4 h. The monthly and annual maximum, minimum, and mean temperatures were calculated from these data.

Collection of Study Organisms
Sampling was carried out in autumn 2015 and 2016 (March and April) to avoid immature stages and allow the collection of identifiable adults, advanced larval stages, or both. Organisms inhabiting medium and high elevation were collected in March because of accessibility to sampling sites while a lower elevation site was surveyed in April, this difference being driven by. The main reason of this sampling lapse is the accessibility to sites and prevalent harsh weather conditions. We followed key considerations for the selection of taxa: organisms collected from the same location (to avoid variability due to the collection site), available in sufficient numbers (relatively abundant and ease of collection), practicability of identification, and suitability for accurately measuring the point of thermal reactivity (PTR) and thermal critical end point (CTE). The PTR is defined as the point at

Site Description
At the coastline, the average air annual temperature is 5.7 • C, while above the tree line (~500-600 m a.s.l.) it is 0 • C [33]. The average annual temperature of one of the rivers within Navarino Island is 5.7 • C at 120 m a.s.l. and only 1 • C at 586 m a.s.l. [34]. The temperature decreases by approximately 1 • C per 100 m of elevation increase, compared to the global average decrease of 0.6 • C per 100 m of elevation [35].

Water Temperature
Temperature loggers (HOBO ® model U22 Water Temp Pro Version 2) were installed at a 20 to 30 cm depth in the littoral areas of each lagoon (Castor, el Salto and Bandera, 20, 480 and 700 m. a.s.l., respectively, Figure 1) from which macroinvertebrates were collected over the entire study period (from February 2015 to March 2016). These recorded the water temperature every 4 h. The monthly and annual maximum, minimum, and mean temperatures were calculated from these data.

Collection of Study Organisms
Sampling was carried out in autumn 2015 and 2016 (March and April) to avoid immature stages and allow the collection of identifiable adults, advanced larval stages, or both. Organisms inhabiting medium and high elevation were collected in March because of accessibility to sampling sites while a lower elevation site was surveyed in April, this difference being driven by. The main reason of this sampling lapse is the accessibility to sites and prevalent harsh weather conditions. We followed Insects 2020, 11, 102 4 of 13 key considerations for the selection of taxa: organisms collected from the same location (to avoid variability due to the collection site), available in sufficient numbers (relatively abundant and ease of collection), practicability of identification, and suitability for accurately measuring the point of thermal reactivity (PTR) and thermal critical end point (CTE). The PTR is defined as the point at which an organism exhibits signs of thermal stress (i.e., obvious change in body movement, swimming, or crawling capability) while, at the CTE, locomotor functions and activity becomes disorganized to the extent that the organism loses the ability to escape from the conditions, leading to death [17,36]. This last step was achieved through an initial test to observe and identify the PTR and the CTE. We selected six macroinvertebrate taxa inhabiting lagoons from the watershed, three restricted to low elevation (20 m a.s.l.), one to a medium elevation lagoon (480 m a.s.l.), and two to high elevation (700 m a.s.l.) ( Table 1). The invertebrates were collected with a D-frame net (150 µm mesh), in addition to hand collection with soft forceps to avoid injury to the organisms. Collected specimens were transferred to a cool box containing site water and transported to the laboratory where they were held in aquaria with an air pump. Before experiments, invertebrates were kept at 5 ± 1 • C (approximate water temperature at collection site) for 24 h. No mortality was recorded during this 24 h period undercontrol conditions, suggesting all collected invertebrates were 'healthy' and would have continued to survive this control phase for the duration of the CT min and CT max experiments, which lasted a maximum of 6 h. Temperature was also recorded in situ with a multimeter sensor (Conductivity pH TDS Hanna Tester HI98130). After completion of the experiments, the invertebrates were returned to their collection locations.

Critical Thermal Limits
For CTmax assessment, invertebrates were placed in an experimental chamber (15 × 10 × 7 cm) and submerged in a thermoregulated digital water bath (Lab. Companion RW-0525G, Figure 2). Two to six individuals of each taxon were assessed per trial (n = 3-4 per taxa), all in the same chamber. An air pump was placed in each experimental chamber to maintain the oxygen saturation above 65% during trials, due to its relevance for thermal tolerance [37]. The first 60 min in the chamber was a control phase, with water at the temperature of the place of collection (5 • C). Subsequently, the temperature was increased at a constant rate (0.14 • C min −1 ) through the experimental phase. The rate of increase aimed to be rapid enough to avoid acclimatization but slow enough to ensure that any reaction observed was a response to the rise in temperature. The reaction of the organisms was evaluated by observing their behavioral responses [36]. Observation of each individual's reaction took between 30 s to 1 min (i.e., detection of antennal movement, gill, leg, or body movement). When an individual exhibited signs of reaching its CTE (e.g., loss of swimming capacity), the temperature was recorded and it was removed from the experimental chamber and returned to the aquarium at the starting temperature. Only individuals that recovered from the experiment were included in the results (recovery time for CT max trials was 1 h). When 50% (CT 50 ) of the individuals showed a response, the trial was terminated [37]. Recovery was assessed by placing each individual in an oxygenated aquarium.
was subsequently lowered to 0 °C at a constant rate (0.03 °C min −1 ). When 0 °C was reached, the temperature was held for 1 h, and then lowered to −1 °C (0.03 °C min −1 ). Once this target temperature was reached, it was held for 1 h. Ice formation in the chamber was recorded at −1 °C, but the chamber was never frozen completely. After the experiment, individuals were placed in Petri dishes with water at 5 °C (collection site temperature), and allowed to recover for 24 h. The recovered organisms were then placed in the experimental chamber, where the temperature was reduced by 1 °C from the previous test temperature. Only individuals that recovered and exhibited normal locomotive functions (antennal, gill, leg, or body movement) were included in the analyses. When 50% (CT50) of the individuals failed to recover, the experiment was terminated. Both experiments were monitored using a 4-channel HOBO ® data logging thermocouple (UX120-014M) attached to the thermoregulated bath. Experimental procedures were performed under Bioethics resolution n° 80/CEC/2018 from the University of Magallanes Bioethics committee.

Assessment of Suitability and Thermal Sensitivity Ranks
Suitability ranks (SRs) and the thermal sensitivity ranks (TSRs) were assigned to each of the taxa to compare their behavioral responses to increasing temperature. Both ranks are an adaptation of the method proposed by Dallas and Rivers-Moore (2012). The suitability rank for each taxon was evaluated using the following scale: 1 = very adequate, 2 = adequate with limitations, and 3 = inadequate. Suitability is based on the ease with which the behavioral responses of PTR and CTE are identified, size and age of organisms, and availability in nature. The rank of thermal sensitivity was based on the average of the maximum critical thermal maximum of each taxon, where 1 = very sensitive (≤25 °C), 2 = moderately sensitive (≤30 °C), and 3 = not very sensitive (≥30 °C).

Statistical Analyses
The data were tested for normality (Shapiro-Wilk's test), and homogeneity of variances (Hartley´s Fmax ratio). As the data were not normally distributed, a Kruskal-Wallis rank analysis for unbalanced designs was performed. When statistical significance was obtained, further nonparametric comparisons were made using the Wilcoxon method adjusted for multiple comparisons with Holm correction. Regression analyses with linear fit were used to examine the relationship In a similar procedure, the CT min assessment was carried out in experimental chambers submerged in the thermoregulated digital water bath, with an air pump ( Figure 2). Each trial had a control phase of 60 min, then a lapse of 15 to 20 min of a temperature decrease (1 • C per trial), and 60 min of exposure to the experimental temperature. Organisms exposed to cooling had a 24 h time frame for recovery. The invertebrates were initially exposed to the collection site temperature, which was subsequently lowered to 0 • C at a constant rate (0.03 • C min −1 ). When 0 • C was reached, the temperature was held for 1 h, and then lowered to −1 • C (0.03 • C min −1 ). Once this target temperature was reached, it was held for 1 h. Ice formation in the chamber was recorded at −1 • C, but the chamber was never frozen completely. After the experiment, individuals were placed in Petri dishes with water at 5 • C (collection site temperature), and allowed to recover for 24 h. The recovered organisms were then placed in the experimental chamber, where the temperature was reduced by 1 • C from the previous test temperature. Only individuals that recovered and exhibited normal locomotive functions (antennal, gill, leg, or body movement) were included in the analyses. When 50% (CT 50 ) of the individuals failed to recover, the experiment was terminated. Both experiments were monitored using a 4-channel HOBO ® data logging thermocouple (UX120-014M) attached to the thermoregulated bath. Experimental procedures were performed under Bioethics resolution n • 80/CEC/2018 from the University of Magallanes Bioethics committee.

Assessment of Suitability and Thermal Sensitivity Ranks
Suitability ranks (SRs) and the thermal sensitivity ranks (TSRs) were assigned to each of the taxa to compare their behavioral responses to increasing temperature. Both ranks are an adaptation of the method proposed by Dallas and Rivers-Moore (2012). The suitability rank for each taxon was evaluated using the following scale: 1 = very adequate, 2 = adequate with limitations, and 3 = inadequate. Suitability is based on the ease with which the behavioral responses of PTR and CTE are identified, size and age of organisms, and availability in nature. The rank of thermal sensitivity was based on the average of the maximum critical thermal maximum of each taxon, where 1 = very sensitive (≤25 • C), 2 = moderately sensitive (≤30 • C), and 3 = not very sensitive (≥30 • C).

Statistical Analyses
The data were tested for normality (Shapiro-Wilk's test), and homogeneity of variances (Hartley´s F max ratio). As the data were not normally distributed, a Kruskal-Wallis rank analysis for unbalanced designs was performed. When statistical significance was obtained, further non-parametric comparisons were made using the Wilcoxon method adjusted for multiple comparisons with Holm correction. Regression analyses with linear fit were used to examine the relationship between CT max , CT min , and thermal breath with elevation. The significance level was set at α = 0.05. Analyses were performed using R version 3.6.1. (R Development Core Team 2019).

Water Temperature Variability
Using the data retrieved from the loggers, monthly temperature profiles were created for each  between CTmax, CTmin, and thermal breath with elevation. The significance level was set at α = 0.05. Analyses were performed using R version 3.6.1. (R Development Core Team 2019).

Water Temperature Variability
Using the data retrieved from the loggers, monthly temperature profiles were created for each

Critical Thermal Maxima and Minima of Macroinvertebrates Restricted to Low Elevation
All macroinvertebrate taxa generally showed a relatively wide thermal tolerance range ( Figure  4), with CTmax being broader than the lagoon temperature range. All species had some tolerance of sub-zero temperatures. The resulting thermal range for the dytiscid Lancetes angusticollis was 41.7 °C, between 37.6 (CTmax) and −4.1 °C (CTmin), the widest tolerance range of amongst the studied taxa. The water boatman Sigara thermal range was 38.6 °C (34.7 to −3.9 °C). Hyalellid amphipods (Hyalella sp.) had a range of 35.8 °C (32.4 to −3.4 °C) ( Table 2). All organisms (including those that did not recover from experimental exposure) survived sub-zero temperatures and freezing conditions during the experiment.

Critical Thermal Maxima and Minima of Macroinvertebrates Restricted to High Elevation
High-elevation-restricted macroinvertebrate taxa showed narrower thermal ranges compared to those from low elevation. The thermal tolerance range was 28.9 °C for dugesiid planarians, 32.0 °C for the stonefly Aubertoperla kuscheli, and 25.2 °C for tanypodinae midge larvae. The latter had the narrowest thermal range of the studied taxa ( Figure 4, Table 2).

Critical Thermal Maxima and Minima of Macroinvertebrates Restricted to Low Elevation
All macroinvertebrate taxa generally showed a relatively wide thermal tolerance range (Figure 4), with CT max being broader than the lagoon temperature range. All species had some tolerance of sub-zero temperatures. The resulting thermal range for the dytiscid Lancetes angusticollis was 41.7 • C, between 37.6 (CT max ) and −4.1 • C (CT min ), the widest tolerance range of amongst the studied taxa. The water boatman Sigara thermal range was 38.6 • C (34.7 to −3.9 • C). Hyalellid amphipods (Hyalella sp.) had a range of 35.8 • C (32.4 to −3.4 • C) ( Table 2). All organisms (including those that did not recover from experimental exposure) survived sub-zero temperatures and freezing conditions during the experiment.   The results of the Kruskal-Wallis analyses and linear regression fit showed significant differences in the CTmax of the studied taxa (H = 114.09, p < 0.00001, R 2 = 0.6917), indicating a marked decrease associated with an elevation increase. Differences were also found in taxa CTmin (H = 56.272, p < 0.0043, R 2 = 0.3257, Table 3), particularly between higher elevation taxa (at 700 m. a.s.l. the stonefly A. kuscheli and the tanypodine midge) with the other four taxa (at 20 and 480 m. a.s.l.) (Figure 5a), yet not detected by the Wilcoxon test adjusted with Holm correction (Supplementary Material). Thermal range differences were also detected between taxa restricted to low and high elevation (H = 92.736, p < 0.0001, R 2 = 0.7337, Figure 5b, Table 3). In some cases, the thermal range was twice the temperature range experienced in the environment (Figure 6).

Critical Thermal Maxima and Minima of Macroinvertebrates Restricted to High Elevation
High-elevation-restricted macroinvertebrate taxa showed narrower thermal ranges compared to those from low elevation. The thermal tolerance range was 28.9 • C for dugesiid planarians, 32.0 • C for the stonefly Aubertoperla kuscheli, and 25.2 • C for tanypodinae midge larvae. The latter had the narrowest thermal range of the studied taxa ( Figure 4, Table 2).
The results of the Kruskal-Wallis analyses and linear regression fit showed significant differences in the CT max of the studied taxa (H = 114.09, p < 0.00001, R 2 = 0.6917), indicating a marked decrease associated with an elevation increase. Differences were also found in taxa CT min (H = 56.272, p < 0.0043, R 2 = 0.3257, Table 3), particularly between higher elevation taxa (at 700 m. a.s.l. the stonefly A. kuscheli and the tanypodine midge) with the other four taxa (at 20 and 480 m. a.s.l.) (Figure 5a), yet not detected by the Wilcoxon test adjusted with Holm correction (Supplementary Materials). Thermal range differences were also detected between taxa restricted to low and high elevation (H = 92.736, p < 0.0001, R 2 = 0.7337, Figure 5b, Table 3). In some cases, the thermal range was twice the temperature range experienced in the environment (Figure 6).

Behavioral Responses, Suitability, and Thermal Sensitivity Ranks
Behavioral responses to the increased temperature varied slightly among the studied taxa. In general, responses included "normal" mobility in the early stages (i.e., swimming, crawling), followed by an abrupt increase in mobility (PTR) and then increasing immobility influencing swimming ability when approaching the final stages of the experiment (CTE). Detailed information on behavioral  Table 4. Macroinvertebrates restricted to high elevation had a narrower thermal tolerance range, so they were assigned to the suitable and thermally sensitive category (Table 4).  The results of the Kruskal-Wallis analyses and linear regression fit showed significant differences in the CTmax of the studied taxa (H = 114.09, p < 0.00001, R 2 = 0.6917), indicating a marked decrease associated with an elevation increase. Differences were also found in taxa CTmin (H = 56.272, p < 0.0043, R 2 = 0.3257, Table 3), particularly between higher elevation taxa (at 700 m. a.s.l. the stonefly A. kuscheli and the tanypodine midge) with the other four taxa (at 20 and 480 m. a.s.l.) (Figure 5a), yet not detected by the Wilcoxon test adjusted with Holm correction (Supplementary Material). Thermal range differences were also detected between taxa restricted to low and high elevation (H = 92.736, p < 0.0001, R 2 = 0.7337, Figure 5b, Table 3). In some cases, the thermal range was twice the temperature range experienced in the environment (Figure 6).   Table 4. Behavioral responses to water temperature increase (point of thermal reactivity, PTR, and critical thermal endpoint, CTE) of sub-Antarctic macroinvertebrate taxa. Suitability ranks SR: 1 = very suitable, 2 = suitable with limitations, and 3 = not suitable, and thermal sensitivity rank TSR: 1 = very sensitive (≤25 • C), 2 = moderately sensitive (≤30 • C), and 3 = less sensitive (≥30 • C) (adapted from Dallas and Rivers-Moore 2012).
Before temperature increase, individuals swam intermittently in the experimental chambers, moving from bottom to top, and vice-versa. The PTR was apparent as a substantial decrease in swimming speed. The CTE was identified when individuals remained at the bottom of the chamber and stopped swimming. The only detectable movement was then from the antennae and hind legs.

3
Coleptera: Dytiscidae Lancetes angusticollis Before temperature increase, individuals swam intermittently in the experimental chambers, moving from bottom to top, and vice-versa. The PTR was apparent as a substantial increase in movement, constantly grabbing air bubbles with difficulty in maintaining them. The CTE was identified as inactivity at the bottom of the chamber, or floating near the surface.

Discussion
The critical thermal method is an effective tool to assess the relative thermal tolerance of freshwater macroinvertebrates. This method also offers the possibility of monitoring behavioral responses, which, in turn, are useful to identify potential bioindicators of thermal alterations [17]. The results obtained here showed significant differences between the critical thermal limits (CT max and CT min ) of different macroinvertebrate taxa inhabiting the same watershed at different elevations. In particular, the CT max decreased for upper elevation macroinvertebrate taxa when compared to those from lower elevations. While CT min did not show large variation, it was significantly lower for higher elevation macroinvertebrates (Figures 4 and 5, Tables 2 and 3). Despite the current study lacking closely related invertebrate taxa to compare across the gradient, we identified a consistent underlying pattern of broader thermal tolerance limits in taxa from more variable environments. These results are consistent with other studies [8], and also with Janzen´s climate variability hypotheses predictions [6], more variable climates may select for organisms with broader thermal tolerances, whereas less variable climates select for narrower thermal tolerances [4]. Nonetheless, comparison among families or genera with phylogenetic or physiological affinities, acclimation effects, intensity, and duration of thermal stress, and robust analyses are necessary to improve our understanding on thermal ecology [21].
Midges (Chironomidae: Tanypodinae) had the lowest thermal tolerance range and also the lowest thermal limits while the diving beetle Lancetes angusticollis had the broadest thermal tolerance (Figures 4 and 5, Tables 2 and 3). This species, along with the other two taxa inhabiting low elevation lagoons, can be observed throughout the year, even when ice is present. Additionally, the broader thermal breadth of the studied taxa when compared to the environmental temperature range suggests that, in some cases, these organisms are able to tolerate temperatures well above the maxima recorded for their habitats. These relatively high tolerances might be an adaptation to the annual environmental conditions (long periods of ice cover and sunlight in winter, and warm exposure with an increased photoperiod in summer). In particular, we found that the stonefly Aubertoperla kuscheli had a thermal range of 32.1 • C, which is double the thermal range of Bandera Lagoon (700 m a.s.l.). It is also notable that this is the only stonefly present in lagoons in the Róbalo watershed [28].
During the experimental observations, we recorded that stoneflies and planarians respond with stirred movements to ice formation into the chamber, probably because of ice-relationinjuries. These behavioral responses to cold in organisms restricted to high elevations, in addition to their common habitat structure (mosses and submerged vegetation), suggests that they attempt to avoid contact with ice rather than tolerating it. Suren [38] and Suren and Winterbourn [39] noted the role of bryophytes as a refuge and nursery for freshwater macroinvertebrates in habitats subject to harsh environmental conditions. Aquatic and/or semi-submerged bryophytes can provide shelter for the development of immature stages, protection against predators, and may themselves be a food source [39].
The critical limits identified, in addition to the behavioral responses, suggest that temperature may act as a distribution barrier. However, other factors that can be of biotic (e.g., predation, exclusion by competition, etc.) or abiotic origin (e.g., dissolved oxygen, resource availability, etc.), might confound these observations. Studies focused on assessing the long-term thermal regimes in addition to other abiotic variables are necessary to understand how variation in these factors, and their combined effects, influences the biology of freshwater fauna [19]. Shah et al. [8] compared the thermal ranges and tolerances of tropical and temperate aquatic insects, both at latitudinal and altitudinal scales. The CTmax of aquatic insects from temperate regions (Rocky Mountains) were similar to those measured in the sub-Antarctic invertebrates assessed in the present work.
There is a considerable literature on methodological protocols and sources of variation in the determination of critical thermal limits [19,40]. The main sources of variation are acclimation and the rate of temperature change. Recording collection site water temperature is helpful to avoid distortion in thermal limit assessments [17]. The effect of different acclimation temperatures on thermal limits has been addressed in South African amphipods [17,41], different families of Coleoptera [19], and also in sub-Antarctic spiders [25]. Different rates of temperature change can increase or decrease the measured upper and lower thermal limits during experiments, as shown for Tenebrio molitor (Tenebrionidae) and Cyrtobagous salviniae (Curculionidae) [19]. A further source of variation is the size/age of the organisms examined, but with more complex outcomes, with positive [42], negative [43], or no effects [36] being reported. In the current study, the potential for this source of error was minimized by collecting only adult or late juvenile stage individuals for each test.

Conclusions
Several considerations should be taken into account when addressing physiological tolerances, including the use of control and acclimation trials, exposure intensity and duration, etc. in an attempt to provide stronger evidence that the observed effects are due to treatments. The predictions of the CVH [6] are consistent with our data, but a robust test of this hypothesis will require further study and analyses. Comparisons between families or genera, further trials, and different treatments are needed in order to improve support for the CVH predictions. Additionally, the potential sources of variation in the thermal ecology studies must be considered, including a temporal dimension and the methodological context on which limits are estimated and then extrapolated into natural settings [21]. New questions arise regarding the determination of thermal limits: Do tolerance ranges of a species inhabiting different watersheds differ? How influential are the thermal regimes of other aquatic systems (including streams)? How sensitive are high elevation organisms to longer periods of exposure to thermal alteration? Addressing such questions will be required to give insight into how this remote southern biota will respond under future global environmental change scenarios.
Supplementary Materials: The following are available online at http://www.mdpi.com/2075-4450/11/2/102/s1. Author Contributions: J.R.-C. conceived and designed the experiments, analyzed the data, prepared figures and tables, led the drafting of the paper, and approved the final draft. T.C. contributed to experimental design, reviewed drafts of the paper, and approved the final draft. P.C. and J.K. contributed to the analyses, critically reviewed drafts of the paper, and approved the final draft. All authors agree to be accountable for all aspects of the work, ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. All authors have read and agreed to the published version of the manuscript.