Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Open Access

Peer-reviewed

Research Article

Developmental and Evolutionary History Affect Survival in Stressful Environments

Abstract

The world is increasingly impacted by a variety of stressors that have the potential to differentially influence life history stages of organisms. Organisms have evolved to cope with some stressors, while with others they have little capacity. It is thus important to understand the effects of both developmental and evolutionary history on survival in stressful environments. We present evidence of the effects of both developmental and evolutionary history on survival of a freshwater vertebrate, the rough-skinned newt (Taricha granulosa) in an osmotically stressful environment. We compared the survival of larvae in either NaCl or MgCl2 that were exposed to salinity either as larvae only or as embryos as well. Embryonic exposure to salinity led to greater mortality of newt larvae than larval exposure alone, and this reduced survival probability was strongly linked to the carry-over effect of stunted embryonic growth in salts. Larval survival was also dependent on the type of salt (NaCl or MgCl2) the larvae were exposed to, and was lowest in MgCl2, a widely-used chemical deicer that, unlike NaCl, amphibian larvae do not have an evolutionary history of regulating at high levels. Both developmental and evolutionary history are critical factors in determining survival in this stressful environment, a pattern that may have widespread implications for the survival of animals increasingly impacted by substances with which they have little evolutionary history.

Citation: Hopkins GR, Brodie ED Jr, French SS (2014) Developmental and Evolutionary History Affect Survival in Stressful Environments. PLoS ONE 9(4): e95174. https://doi.org/10.1371/journal.pone.0095174

Editor: Daniele Canestrelli, Tuscia University, Italy

Received: December 13, 2013; Accepted: March 24, 2014; Published: April 18, 2014

Copyright: © 2014 Hopkins et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Funding: This research was supported financially by the Utah State University (USU) Department of Biology and the Ecology Center. The USU Merril-Cazier Library's Open Access Funding Initiative and the USU Ecology Center kindly assisted with publication costs. The Natural Sciences and Engineering Research Council of Canada (NSERC) provided additional financial support to GRH. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Competing interests: The authors have declared that no competing interests exist.

Introduction

Natural and anthropogenic stressors are commonplace throughout the environment. The ways in which stressors impact organisms, and their ability to successfully respond to these stressors is of paramount importance to our understanding of biological systems. For organisms with complex life cycles, the ability to respond to a given stressor may vary depending on life history stage, and there may be carry-over effects from one stage to the next [1] (see Table S1 in Supporting Information). However, organisms may or may not have an evolutionary history of regulating the stressor in question, and this may also affect their ability to effectively respond [2], [3]. We propose that both an organism's developmental history of exposure to a stressor (developmental history hypothesis) and its evolutionary history of regulating that stressor (evolutionary history hypothesis) play critical roles in the survival of organisms in stressful environments.

It has been suggested that the earlier in an organism's life history environmental stressors are experienced, the more severe the lasting consequences will be [4]–[6], and there is strong empirical evidence across animal taxa for this assertion (Table S1). This forms the basis of our developmental history hypothesis. In humans, for example, the environment of the womb can significantly affect an individual's chances of cardiac and other diseases later in life [5], [7], [8]. In birds, the temperature at which eggs are incubated can affect hatchling body composition, growth, immunocompetence and thermoregulatory ability [4], [9]. Developmental temperature also affects survival, growth and behavior of juvenile reptiles (e.g., [10]) (Table S1). Elevated CO2 as embryos results in decreased larval settlement success of sea urchins [11], and the ability of bryozoans to produce large, successful colonies is dependent on their embryonic experience and growth [12]. Thus, embryonic exposure to stressors can be critical to an animal's future fitness (Table S1).

Parsing critical life history stages, however, is not trivial, and many studies have given contradictory evidence for the developmental history hypothesis. For example, while multiple studies have shown that embryonic environment can significantly affect an individual's chances of success in later life (Table S1), others have shown that it is the larval or juvenile environment that has the greatest influence on survival, growth, or reproduction (e.g., [13], [14]). Still others have shown that while the embryonic environment has a significant role to play in later life, its effect may be dependent on the environment animals experience later in life (e.g., [15], [16], [17]). Experiments are often not designed to isolate the effects of environment on a specific life history stage from those of another (e.g., [18], [19]–[21]), and thus, consistent knowledge of the environmental and carry-over effects across multiple life history stages is lacking (but see [14], [17], [22], [23]).

While there is a strong empirical basis for the developmental history hypothesis (even with the conflicting evidence and limitations identified above), there is much less known regarding the evolutionary history hypothesis. Organisms in most habitats today face both natural stressors with which they have an evolutionary history, and thus evolved physiological mechanisms of regulating (e.g., CO2, temperature, NaCl), and novel stressors with which they do not have this same evolutionary history (e.g., pesticides, flame retardants, commercial non-NaCl-based deicing salts), and thus lack the physiological mechanisms to regulate. The effects of developmental history must therefore be placed in this environmental and evolutionary context. While many studies have documented the significant effects of unfamiliar substances such as pollutants on evolutionarily-naïve organisms (e.g., reviewed by [24] for amphibians), these cannot be directly compared to stressors with which the organism has an evolutionary history, and thus a means of regulating, as the nature of the two stressors is usually very different (i.e., comparing the effect of a herbicide with the effect of NaCl). At this point, we do not know how the potentially important effects of an organism's evolutionary history with a stressor may interact with its developmental history of exposure with the stressor.

To address these concerns, we tested the effects of developmental and evolutionary history on survival in stressful environments. We chose the rough-skinned newt (Taricha granulosa Skelton; Caudata: Salamandridae) as our model, an osmotically sensitive organism, and salinity as its stressor. Specifically, we tested the effects of both NaCl and MgCl2 on the post-hatching survival of newt larvae that had either been exposed to salt as both embryos and larvae or just as larvae. Salinity is an excellent stressor to use to test our two hypotheses, as it is a naturally occurring abiotic component of aquatic habitats, and is known to have significant carry-over effects from the embryonic to post-hatching life-stages in a variety of organisms, [15], [23], [25]–[27] (Table S1). We used salt concentrations that were within environmentally relevant limits of freshwater aquatic systems impacted by either natural (i.e., estuaries) or anthropogenic (i.e., road deicing salts) sources of salts [29], [30]. The two most common sources of salinity in North America today are two different salts, NaCl and MgCl2, only one of which most organisms have an evolutionary history of regulating. Sodium chloride (NaCl) is one of the most common osmolytes, and organisms have an evolutionary history of regulating this in a variety of habitats, whereas MgCl2 has not been identified as a common vertebrate osmolyte [31], and Mg2+ is not found in substantial concentrations in most freshwater habitats, nor the precipitation that feeds them (including in the newts' range) [32]. Therefore, animals do not have the same evolutionary history of physiological regulation of this ion. Nevertheless, MgCl2 is now the second most commonly used road deicer in North America (behind NaCl), and is used exclusively in some areas of the continent [33]. Thus, there is the potential that organisms will encounter MgCl2 in substantial quantities in their environment. We found that both salts caused significant developmental carry-over effects from the embryonic environment on larval survival, but that the salts differed in their effects on larval survival, according to the differential evolutionary history that amphibians have with regulating the two stressors. As more and more freshwater animals, mostly maladapted to salt, will be forced to cope with increasing salinization of their habitats due to the application of road deicing salts [34], [35]–[36], landscape modification and agricultural waste [37]–[40], and rising sea-levels [41]–[43], understanding the effects of both developmental and evolutionary history of salinity exposure will have important implications for both life history and evolutionary theory, as well as conservation efforts.

Materials and Methods

Ethics Statement

Adult rough-skinned newts (Taricha granulosa) (not an endangered or protected species) were collected by dip-net and hand from Soap Creek ponds (44°40′13.22″N, 123°16′39.65″W) under Oregon Department of Fish and Wildlife Scientific Taking Permit #062-11. Access to these ponds was granted by Joe Beatty, Oregon State University. The Utah State University Institutional Animal Care and Use Committee (IACUC) approved the collection and use of animals in this research, and all experimental protocols (approved protocol #1524). Animals were euthanized at the completion of experiments with MS-222, in accordance with the approved IACUC protocol (#1524).

Experimental Procedure

As reported in a previous study ([44] for detailed methods on habitat, field collection, rearing eggs and preparing salt solutions), we reared eggs from 16 different gravid wild-caught female rough-skinned newts (Taricha granulosa) from a single, salt-naïve population from Benton County, Oregon, in a laboratory environmental control chamber at 7°C. This population is truly salt-naïve [44], being highly philopatric to freshwater ponds that are separated by hundreds of meters from small county roads that are not salted (Kendal Weeks, Oregon Department of Transportation Road Maintenance, personal communication; Kent Mahler, Benton County Road Maintenance, personally communication). While MgCl2 is widely used in Oregon as its exclusive deicer, it is also not applied to the nearest stretch of highway to these ponds, located over 4 km away (Kendal Weeks, Oregon Department of Transportation Road Maintenance, personal communication). See [44] for additional details on this habitat. Eggs from wild-caught females were randomized to one of six different salt treatments, made with laboratory grade NaCl (Thermo Fisher Scientific, Fair Lawn, NJ, USA), MgCl2 (Acros Organics, Fair Lawn, NJ, USA) and distilled water (Low NaCl, Low MgCl = 1.0 g/l Cl−; Medium NaCl, Medium MgCl2 = 1.5 g/l Cl−; High NaCl, High MgCl2 = 2.0 g/l Cl−) and a control (20% Holtfreter's Solution = 0.7 g/l Cl− [45]). Those eggs that survived these treatments were used in the present experiment. At hatching, the size (total length) and developmental stage [46] of hatchlings were recorded (see [44] for full methods and results).

Eggs that were reared in a salt treatment remained in that salt treatment as larvae (Fig. 1). Approximately 7 times more control eggs were reared than salt treatment eggs, so that control eggs could be randomized to new larval treatments in the present experiment (similarly to [26]) (Fig. 1). Eggs were monitored daily and all larvae were transferred to their new treatment solution within 12 hours of hatching. This direct transfer, following a similar protocol of Petranka and Doyle [47], was meant to mimic the sharp spike in Cl− concentrations found in road-side environments that immediately occurs within hours of a deicing event or snowmelt [48]–[50], where minimal to no time is allowed for acclimation. While gradual acclimation of low salinity levels have led to increased tolerance in some amphibians (e.g., [51]) it has also led to increased susceptibility in others [52], and is less environmentally relevant to examining the sudden spikes of salinity seen in habitats due to road deicing salt application. In addition, while the salt concentrations used were typical for those immediately resulting from deicing events [48], [53], they were also well below recorded NaCl and MgCl2 LD-50 values for other amphibian larvae [37], [54], [55].

thumbnail
Download:
Figure 1. Outline of experimental design.

Embryonic and larval environments, salinity concentrations, treatment names, and sample sizes are shown. Newt eggs were reared in either a freshwater control, or one of six salt treatments. Upon hatching, embryos that were reared in salt stayed in that salt, whereas embryos reared in control either stayed in control or were randomized to one of the six salt treatments for the larval environment. The name of each treatment combination is listed, and sample sizes are given under each larval environment (numbers outside of parentheses indicate total number of individuals in the treatment, whereas numbers inside parentheses indicate number of containers in the treatment (up to five sibling larvae were reared in the same container, and individuals within containers were treated as nested subsamples. See Methods for more details).

https://doi.org/10.1371/journal.pone.0095174.g001

Larvae were housed in sibling groups of up to 5 individuals (keeping offspring from different female and treatment combinations separate) in 12.5 cm diameter, 10.5 cm deep round plastic containers, filled with 400 ml of solution. Each container was randomized to a location in a growth chamber set at 7°C, with a 12 h light: dark photoperiod. Containers were checked daily for larval survival, and dehydration. Dead larvae were noted and removed, and a small amount of distilled water was added to each container if necessary, to compensate for evaporation. Taricha granulosa larvae retain some embryonic yolk for up to approximately two weeks after hatching, and do not engage in feeding on prey before then. As we did not want to confound our survival results with possible negative effects of the salt treatments on larval prey, we only conducted this experiment for 15 days post-hatching; if a larva was alive at day fifteen, it was recorded as alive for the purposes of the analysis. A similar endpoint has also been used in a previous study on post-hatching survival of frog larvae in road deicing salt [47].

Statistical Analysis

For survival analyses, individual larvae were treated as subsamples within containers, which were treated as subsamples nested within individual female. Larval survival was analyzed using a binomial distribution, with a generalized linear mixed model blocking on individual female as a random effect. We first compared the survival of control newts (i.e., those reared in control as eggs and larvae) to newts in all other treatments for each salt, and then ran separate models to compare survival among salt treatments (minus control) for both larvae that were reared in control and those reared in salt as eggs, with Tukey-adjusted multiple comparisons among individual treatment levels, when an overall significant effect of treatment was found. We were, however, primarily interested in comparing and contrasting the effects of embryonic and larval environment on larval survival. As we did not have a complete factorial design in this study (e.g. embryonic low MgCl2 + larval high NaCl treatment combination), for this analysis, we analyzed the effects of the two different salt types separately, using embryonic and larval treatments as fixed effect factors in our models. We then analyzed the effect of embryonic versus larval environment on larval survival for each salt [56]. In these analyses, larval treatment had three levels, low, medium and high, and embryonic treatment had two levels, control and salt. This enabled a direct statistical comparison to be made of larval survival between animals that were reared as eggs in control or, for example, low MgCl2, for larvae that were reared in low MgCl2. We conducted Tukey-adjusted multiple comparisons, specifically comparing larval survival in each salt treatment level between eggs that were reared in either that salt treatment or control, for cases in which an overall significant effect of either embryonic treatment, larval treatment, or their interaction was found. Analyses were conducted using PROC GLIMMIX in SAS software version 9.3, with significance set at α = 0.05.

As embryonic exposure to salt affected the size and developmental stage at hatching of newts, as did differences among individual mothers (females) [44], we wanted to further assess the potential contribution of these variables, as well as embryonic and larval treatments in general, in explaining any overall effects of salt treatment in either embryonic or larval environments on larval survival. To do this, we conducted multivariate classification analyses, which measure variable importance in a model's ability to correctly classify larvae as having died or survived. As only one out of 778 newt larvae died after being reared in control as both an embryo and larvae (see Results), we restricted our analyses to larvae reared in salt post-hatching. We used three validated classification procedures [57], logistic regression, Classification Trees [58] and Random Forests [57], [59], and in each case assessed variable importance by examining the relative classification performance of models incorporating or not incorporating key variables.

Specifically, we assessed the ability of the models to correctly classify larvae as having died (sensitivity). For the full model, we included all larval and embryonic variables of potential interest, including: larval treatment, embryonic treatment, length at hatching, developmental stage at hatching, and female identity. We then withdrew the larval treatment variable, and reassessed the model's sensitivity, withdrew all embryonic variables (leaving only larval treatment and individual female) and again reassessed the model's sensitivity, to assess the potential relative contribution of larval environment in predicting larval mortality. As well as assessing variable importance in this manner, all three classification methods also provide separate indicators of variable importance [57]. This is achieved through a variable importance plot in Random Forests, a classification plot in Classification Trees, and the variable with the largest Wald Chi-Square value in logistic regression. We chose the most important variable identified in each of these methods from the original full model, and reinserted it back into the model including only larval treatment and female identity, and assessed whether the inclusion of this identified variable increased model performance. Classification analyses were completed in SAS (logistic regression) and R (R Development Core Team, 2008, www.R-project.org) (Classification Trees and Random Forests). Finally, as length at hatching was identified as a key variable of importance in predicting larval mortality (see Results), we compared the mean length at hatching of larvae that died versus survived in each treatment using t-tests in SAS software version 9.3, with significance set at α = 0.05.

Results

After 14 days, only one out of 778 larvae reared in control as both egg and larva (“control” treatment) died in this treatment, which was significantly fewer than in any other treatment (all p<0.001). The survival of the remaining larvae, all experiencing salts in their larval environment, was then compared. There was a significant effect of larval salt treatment on larval survival for both newts that were reared embryonically in salt (F5,74 = 16.54, p<0.0001) and control (F5,73 = 7.81, p<0.0001). For animals that were reared as eggs in salt and stayed in that salt as larvae, significantly more larvae died in low and medium MgCl2 than in those corresponding concentrations of NaCl (all Tukey adjusted multiple comparisons p<0.0001), with a similar percentage of larvae dying in high MgCl2 as high NaCl (Tukey adjusted p = 0.98). For animals that were reared as eggs in control and then transferred to salt as larvae, marginally more larvae died in low MgCl2 than low NaCl (Tukey adjusted p = 0.0698), and significantly more larvae died in medium and high MgCl2 than the corresponding concentrations of NaCl (Tukey adjusted p<0.02).

Increased salt concentration, in both the embryonic and larval environments, generally resulted in increased larval mortality (with the exception of high MgCl2) (Fig. 2). For both salts, larval survival was significantly affected by embryonic environment (Table 1). For the majority of treatment levels, larvae that were reared as eggs in control solution survived significantly better than larvae that were reared as eggs in salt treatments (Fig. 2). For NaCl, both embryonic and larval treatments significantly affected survival of larvae in this salt, but for MgCl2, only embryonic treatment significantly explained larval survival (Table 1). There were no significant interacting effects of embryonic and larval environments on larval survival (Table 1).

thumbnail
Download:
Figure 2. Percentage (mean ± SE) of larvae that died in each salt treatment.

(A) NaCl, (B) MgCl2. Only 1 out of 778 larvae in Control died, and thus only results for mortality in salt treatments are shown. Direct comparisons are made between the mortality of larvae reared as embryos in salt (open squares) or control (closed circles). Asterisks indicate significant differences (Tukey-adjusted multiple comparisons) between the percentages of larvae died in each of these treatments (i.e., for the larval treatment Medium NaCl, significantly more larvae died when reared as eggs in that salt, than did larvae reared as eggs in control). “ns”  =  no significant difference between treatments.

https://doi.org/10.1371/journal.pone.0095174.g002

thumbnail
Download:
Table 1. Effects of embryonic environment, larval environment, and their interaction on larval survival in NaCl and MgCl2.

https://doi.org/10.1371/journal.pone.0095174.t001

Eggs that were reared in salt water resulted in smaller larvae at hatching than those reared in control [44]. Classification analyses with three different methods all revealed length at hatching as the consistently most important variable in determining larval survival (Table 2), further strengthening the evidence of the importance of embryonic environment on survival post-hatching. Although Classification Trees and Random Forests had better sensitivity than logistic regression (as was expected [57]), the ability of models, using any of the classification methods, to correctly classify larvae as having died declined dramatically with the exclusion of embryonic variables (i.e., larval treatment and female identity alone was a very poor classifier of larval survival), but recovered substantially with the re-inclusion of length at hatching as a predictor variable (Table 2), further identifying it as a critical variable for predicting larval survival. Larvae that survived, in each of the treatments, were significantly larger at hatching, on average, than larvae that died (Fig. 3; for all t-tests, p<0.01).

thumbnail
Download:
Figure 3. Mean (±SE) lengths at hatching (mm) of larvae that died (closed circles) or survived (open circles) in each salt treatment.

(A) NaCl, (B) MgCl2. In all treatments, larvae that survived averaged larger at hatching than those that died (all t-tests, p<0.01). See [44] for full results on length at hatching.

https://doi.org/10.1371/journal.pone.0095174.g003

thumbnail
Download:
Table 2. Classification analyses for predicting whether or not newt larvae died (“sensitivity”), for data excluding control data (i.e., only newts in salt as larvae).

https://doi.org/10.1371/journal.pone.0095174.t002

Discussion

Developmental and evolutionary history each significantly affected the survival of newt larvae in salts, and thus the importance of both hypotheses was supported. Eggs appear to be a critical life history stage for this amphibian in osmotically stressful environments. Animals that were exposed to salt as embryos and survived hatched at a smaller size than animals that did not experience embryonic salinity. Stunting of embryonic growth put amphibian larvae at greater risk for salt-induced mortality (Table 2). However, our results show that it is also important to understand the evolutionary history an organism has with a stressor. Even though there was no difference in egg mortality between embryos reared in NaCl or MgCl2 [44], more larvae died in MgCl2 than in NaCl (Fig. 4). While newt larvae have evolved with natural sources of NaCl in their environment, which they can osmoregulate, such common regulation of MgCl2 does not appear to have evolved. Understanding this evolutionary history, as well as parsing critical life history stages is imperative to understand the effects of stressors on the life history of an organism.

thumbnail
Download:
Figure 4. Mortality of newt eggs (black bars) and larvae (grey bars) in each salt treatment.

This figure shows only larvae that were reared as both eggs and larvae in salt. The percentage of individuals that survived in each treatment is indicated in white. All percentages are calculated based on the total number of eggs that started in each treatment (Control  = 2577, low NaCl  = 363, medium NaCl = 366, high NaCl = 345, low MgCl2 = 369, medium MgCl2 = 369, high MgCl2 = 354; [44]), some of which either died (black bars), or survived to hatching and were reared in salt, where they either died (grey bars) or survived (white bars).

https://doi.org/10.1371/journal.pone.0095174.g004

The majority of organisms have complex life cycles, and the experiences of one life stage can have profound impacts on those in subsequent stages [1] (Table S1). Embryonic salinity is known to affect the post-hatching survival, growth and development of marine and estuarine invertebrates, such as barnacles [23], crabs [15], [25], [27], horseshoe crabs [26] and tunicates [28]. While all life history stages of amphibians have, individually, repeatedly been found to be extremely sensitive to salt [21], [47], [54], [55], [60]–[68], with a few notable exceptions such as Fejervarya cancrivora [69], [70], the relative sensitivity of each life history stage, and potential down-stream effects of salinity from one stage to the next, have been less studied. In one of the only other studies on amphibians to examine embryonic carry-over effects of salinity, frog larvae (Lithobates sylvaticus) reared in salt water (NaCl-based) as eggs had reduced survival in salt compared to larvae that were reared in freshwater as eggs [47]. This study also found that growth and development of larvae that survived was depressed in those animals reared embryonically in salt, also suggesting carry-over effects of embryonic exposure to salt [47]. Snodgrass et al [21] also found that Bufo americanus toadlets exposed to stormwater pond sediment (which had an increased conductivity mainly due to road deicing salt) as embryos were smaller at metamorphosis than embryos and larvae exposed to freshwater (although the relative effects of embryonic vs. larval exposure were not separated). Other studies have also shown potential carry-over effects of larval salinity exposure on metamorphic traits important for adult fitness [37], [60]. These results all clearly show that Qiu and Qian's [23] statement regarding marine invertebrates, that “osmotic stress experienced in one life-stage can be passed over to the next life-stage”, can apply to freshwater vertebrates as well.

This pattern of decreased post-hatching survival as a consequence of embryonic exposure has also been found in amphibians in response to other stressors, such as nitrite [71] and pesticides [22]. Thus, studies that do not examine effects at each life history stage and do not consider the potential for cascading effects across stages may seriously underestimate the cumulative effects of exposure to stressors [19], [71]–[73].

One of the primary ways that osmotic stress affects the embryonic stage to influence post-hatching survival is through the retardation of growth and development. Newt eggs that were reared in salt water hatched sooner, smaller and less developed than newts reared in a freshwater control [44], and this resulting reduced length at hatching appears to be the single most important variable in predicting next-stage (larval) survival in salt water (Table 2). Size at hatching/birth is well known to have important implications on larval, juvenile, and adult health and survival in a wide variety of taxa, ranging from sea snails [74] and bryozoans [12], to birds [75] and humans [7]. Furthermore, this link between size and fitness has been identified as key to life history theory [12]. Among amphibians, hatching early, smaller and less developed, is known to affect larval survival, the onset of feeding competence, competitive and predatory interactions, and larval growth rate and timing of metamorphosis [6], [76]–[80]. Similar to our findings, small, less developed amphibian larvae are more susceptible to pollutants than are large larvae [81], [82]. Smaller larval rough-skinned newts are also more vulnerable to be injured and die in predatory encounters with dragonfly nymphs [83]. Thus, even if smaller hatchlings are able to survive short-term in osmotically stressful environments (which seems unlikely from our results (Fig. 3)), or even if compensatory growth occurred later in development, a host of other fitness consequences of this initial stunted embryonic growth and development are still likely later in life [84], further emphasizing the importance of the embryonic environment for life-time fitness.

While the effects of the two salt types were not significantly different on embryonic survival [44], there were differences in the larval stage, whereby MgCl2 had relatively greater effects on survival (Fig. 4). This is in spite of the fact that embryos actually hatched slightly larger at MgCl2 than at NaCl [44]. Although most amphibian eggs, like those of many other aquatic organisms [85], have little means of osmoregulating at the salt concentrations used in this study [37], [63], [86], and thus the effects of NaCl and MgCl2 at this life history stage are equally destructive (any affect of evolutionary history is minimized in the absence of regulatory ability), amphibian larvae have evolved to osmoregulate Na+ and Cl− ion concentrations in their body through the use of integumental and gill Na+ pumps [31], [87]–[90]. Larvae have not evolved this same ability to regulate Mg2+ ions, however, and thus larvae in NaCl were able to attempt osmoregulation to survive in this solution whereas larvae in MgCl2 were not. In addition to lacking this evolutionary history of osmoregulation, Mg2+ has also been shown to be inhibitory to important osmoregulatory skin ion pump functioning in other amphibian larvae [91]. Whereas the effects of NaCl on larvae act in a typical dose-response fashion (Fig. 2a), it appears that any concentration of MgCl2 is detrimental to larvae (Fig. 2b), as they have less means to regulate it. This may explain why larval salinity concentration significantly influences larval survival for animals in NaCl, but not those in MgCl2 (Table 1). The fact that Mg High had lower mortality than Mg Low or Medium (Fig. 2b) may be due to a number of possible reasons, including hormesis [92]. In the only other studies on the effects of MgCl2 on amphibian larvae to date, both Dougherty and Smith [62] and Harless et al [54] also found that this emerging deicing salt was more toxic to frog tadpoles than NaCl. Magnesium chloride may in fact, be more toxic than NaCl to life in general, as studies have found that otherwise salt-tolerant plants [93], [94] and archaea [95] are often intolerant of MgCl2, and the threshold for biological processes in MgCl2 is lower than other salts, including NaCl [95]. These results make sense from an evolutionary perspective, given the small quantities of Mg2+ generally found in most aquatic ecosystems, relative to the higher quantities of Na+ found in precipitation and the ocean [32], and thus many organisms may not have an evolutionary history of regulating Mg2+ in high concentrations in their environment.

Vulnerability of a particular life history stage can be described as the ability of that life history stage to regulate the stressor in question. Using this criterion, it appears that eggs are the most vulnerable life history stage to salts overall in amphibians (this study; [61], [65], [66], [96]) and effects on embryonic development at this stage have profound survival consequences in later life history stages, even possibly affecting population viability indirectly through influencing post-embryonic (larval) mortality [97]. Similarly, amphibian larvae cannot successfully osmoregulate in MgCl2, and thus all life history stages are particularly vulnerable to this evolutionarily novel but emerging deicing agent, which is now the second most commonly used road deicer in North America [33].

Conclusions

Understanding the evolutionary history of an organism with its stressor, and the differential sensitivity of life history stages to that stressor are critical in assessing the vulnerability of organisms to stressful environments. It is now apparent that embryonic exposure to a stressor can have profound implications on the post-hatching survival and fitness of organisms in practically all animal taxa (Table S1), through influencing growth and development in this critical life history stage. In post-hatching individuals, however, even the largest, best-developed organism can only successfully deal with stressors that they have evolved to regulate. As the world of these organisms becomes increasingly impacted by anthropogenic factors, understanding this evolutionary history and its survival implications at and across different life history stages will be critical for the future conservation of animals in increasingly stressful environments.

Supporting Information

Table S1.

Animal phyla where components of the embryonic environment have been demonstrated to have significant carry-over effects post-hatching. This list is not exhaustive, but is representative of the diversity and breadth of this phenomenon throughout the animal kingdom.

https://doi.org/10.1371/journal.pone.0095174.s001

(DOCX)

Acknowledgments

We thank B. Parrish for substantial help in the laboratory, as well as L. Neuman-Lee, N. Kiriazis, B. Gall, A. Stokes, T. Stokes, and B. Rowland for assistance. Thanks also to S. Durham and Z. Stopher for valuable help with data analysis, and J. Beatty and Oregon State University for access to their research ponds. B. and E. Gall kindly collected newts under Oregon Department of Fish and Wildlife permit # 062-11. We also thank G. Smith, Z. Stopher, and M. Baker for constructive feedback on an earlier version of the manuscript. Data Accessibility: Raw experimental data is publicly available at Dryad doi:10.5061/dryad.jg0j5.

Author Contributions

Conceived and designed the experiments: GRH EDB. Performed the experiments: GRH. Analyzed the data: GRH SSF. Contributed reagents/materials/analysis tools: EDB SSF. Wrote the paper: GRH EDB SSF. Designed the analysis: SSF.

References

  1. 1. Pechenik JA, Wendt DE, Jarrett JN (1998) Metamorphosis is not a new beginning. BioScience 48: 901–910.
  2. 2. Calow P (1991) Physiological costs of combating chemical toxicants: ecological implications. Comparative and Biochemical Physiology 100C: 3–6.
  3. 3. Badyaev AV (2005) Stress-induced variation in evolution: from behavioural plasticity to genetic assimilation. Proceedings of the Royal Society B 272: 877–886.
  4. 4. DuRant SE, Hopkins WA, Wilson AF, Hepp GR (2011) Incubation temperature affects the metabolic cost of thermoregulation in a young precocial bird. Functional Ecology 26: 416–422.
  5. 5. Lindström J (1999) Early development and fitness in birds and mammals. Trends in Ecology and Evolution 14: 343–348.
  6. 6. Orizaola G, Dahl E, Laurila A (2010) Compensating for delayed hatching across consecutive life-history stages in an amphibian. Oikos 119: 980–987.
  7. 7. Barker DJP (2006) Adult consequences of fetal growth restriction. Clinical obstetrics and gynecology 49: 270–283.
  8. 8. Gluckman PD, Hanson MA, Beedle AS (2007) Early life events and their consequences for later disease: a life history and evolutionary perspective. American Journal of Human Biology 19: 1–19.
  9. 9. DuRant SE, Hopkins WA, Hawley DM, Hepp GR (2011) Incubation temperature affects multiple measures of immunocompetence in young wood ducks (Aix sponsa). Biology Letters 8: 108–111.
  10. 10. Webb GJW, Cooper-Preston H (1989) Effects of incubation temperature on crocodiles and the evolution of reptilian oviparity. American Zoologist 29: 953–971.
  11. 11. Dupont S, Dorey N, Stumpp M, Melzner F, Thorndyke M (2013) Long-term and trans-life-cycle effects of exposure to ocean acidification in the green sea urchin Strongylocentrotus droebachiensis. Marine Biology 160: 1835–1843.
  12. 12. Marshall DJ, Bolton TF, Keough MJ (2003) Offspring size affects the post-metamorphic performance of a colonial marine invertebrate. Ecology 84: 3131–3137.
  13. 13. Mandrillon A-L, Saglio P (2007) Effects of embryonic exposure to conspecific chemical cues on hatching and larval traits in the common frog (Rana temporaria). Chemoecology 17: 169–175.
  14. 14. Räsänen K, Laurila A, Merilä J (2002) Carry-over effects of embryonic acid conditions on development and growth of Rana temporaria tadpoles. Freshwater Biology 47: 19–30.
  15. 15. Giménez L, Anger K (2003) Larval performance in an estuarine crab, Chasmagnathus granulata, is a consequence of both larval and embryonic experience. Marine Ecology Progress Series 249: 251–264.
  16. 16. McKnight CM, Gutzke WHN (1993) Effects of the embryonic environment and of hatchling housing conditions on growth of young snapping turtles. Copeia 1993: 475–482.
  17. 17. Watkins TB, Vraspir J (2006) Both incubation temperature and posthatching temperature affect swimming performance and morphology of wood frog tadpoles (Rana sylvatica). Physiological and Biochemical Zoology 79: 140–149.
  18. 18. Kusch RC, Krone PH, Chivers DP (2007) Chronic exposure to low concentartions of waterborne cadmium during embryonic and larval development results in the long-term hindrance of antipredator behavior in zebrafish. Environmental Toxicology and Chemistry 27: 705–710.
  19. 19. Rohr JR, Palmer BD (2005) Aquatic herbicide exposure increases salamander desiccation risk eight months later in a terrestrial environment. Environmental Toxicology and Chemistry 24: 1253–1258.
  20. 20. Rohr JR, Sager T, Sesterhenn TM, Palmer BD (2006) Exposure, postexposure, and density-mediated effects of atrazine on amphibians: breaking down net effects into their parts. Environmental Health Perspectives 114: 46–50.
  21. 21. Snodgrass JW, Casey RE, Joseph D, Simon JA (2008) Microcosm investigations of stormwater pond sediment toxicity to embryonic and larval amphibians: variation in sensitivity among species. Environmental Pollution 154: 291–297.
  22. 22. Greulich K, Pflugmacher S (2003) Differences in susceptibility of various life stages of amphibians to pesticide exposure. Aquatic Toxicology 65: 329–336.
  23. 23. Qiu J-W, Qian P-Y (1999) Tolerance of the barnacle Balanus amphitrite amphitrite to salinity and temperature stress: effects of previous experience. Marine Ecology Progress Series 188: 123–132.
  24. 24. Egea-Serrano A, Relyea RA, Tejedo M, Torralva M (2012) Understanding of the impact of chemicals on amphibians: a meta-analytic review. Ecology and Evolution 2: 1382–1397.
  25. 25. Charmantier G, Giménez L, Charmantier-Daures M, Anger K (2002) Ontogeny of osmoregulation, physiological plasticity and larval export strategy in the grapsid crab Chasmagnathus granulata (Crustacea, Decapoda). Marine Ecology Progress Series 229: 185–194.
  26. 26. Ehlinger GS, Tankersley RA (2004) Survival and development of horsehoe crab (Limulus polyphemus) embryos and larvae in hypersaline conditions. Biological Bulletin 206: 87–94.
  27. 27. Giménez L (2002) Effects of prehatching salinity and initial larval biomass on survival and duration of development in the zoea 1 of the estuarine crab, Chasmagnathus granulata, under nutritional stress. Journal of Experimental Marine Biology and Ecology 270: 93–110.
  28. 28. Thiyagarajan V, Qian P-Y (2003) Effect of temperature, salinity and delayed attachment on development of the solitary ascidian Styela plicata (Lesueur). Journal of Experimental Marine Biology and Ecology 290: 133–146.
  29. 29. Environment Canada (2001) Priority Substances List Assessment Report: Road Salts. Environment Canada, Ottawa pp. 283.
    • 30. Vernberg FJ, Vernberg WB (2001) The coastal zone: past, present, and future. Columbia, South Carolina: University of South Carolina Press.
      • 31. Shoemaker VH, Nagy KA (1977) Osmoregulation in amphibians and reptiles. Annual Reviews of Physiology 39: 449–471.
      • 32. Drever JI (1997) The geochemistry of natural waters: surface and groundwater environments. Upper Saddle River, NJ: Prentice Hall.
        • 33. National Transportation Research Board (2007) National Cooperative Highway Research Program Report 577: Guidelines for the Selection of Snow and Ice Control Materials to Mitigate Environmental Impacts. Transportation Research Board of the National Academies, Washington, D.C.: National Cooperative Highway Research Program.
          • 34. Kaushal SS, Groffman PM, Likens GE, Belt KT, Stack WP, et al. (2005) Increased salinization of fresh water in the northeastern United States. Proceedings of the National Academy of Sciences of the United States of Amerca 102: 13517–13520.
          • 35. Thunqvist E (2004) Regional increase of mean chloride concentration in water due to the application of deicing salt. Science of the Total Environment 325: 29–37.
          • 36. Cañedo-Argülles M, Kefford BJ, Piscart C, Prat N, Schäfer RB, et al. (2013) Salinization of rivers: an urgent ecological issue. Environmental Pollution 173: 157–167.
          • 37. Chinathamby K, Reina RD, Bailey PCE, Lees BK (2006) Effects of salinity on the survival, growth and development of tadpoles of the brown tree frog, Litoria ewingii. Australian Journal of Zoology 54: 97–105.
          • 38. Christy MT, Dickman CR (2002) Effects of salinity on tadpoles of the green and golden bell frog (Litoria aurea). Amphibia-Reptilia 23: 1–11.
          • 39. Kearney BD, Byrne PG, Reina RD (2012) Larval tolerance to salinity in three species of Australian anuran: an indication of saline specialisation in Litoria aurea. PLoS One 7: e43427.
          • 40. Williams WD (2001) Anthropogenic salinisation of inland waters. Hydrobiologia 466: 329–337.
          • 41. Gornitz V (1995) Sea-level rise: a review of recent past and near-future trends. Earth Surface Processes and Landforms 20: 7–20.
          • 42. Nicholls RJ, Hoozemans FMJ, Marchand M (1999) Increasing flood risk and wetland losses due to global sea-level rise: regional and global analyses. Global Environmental Change 9: S69–S87.
          • 43. Purcell KM, Hitch AT, Klerks PL, Leberg PL (2008) Adaptation as a potential response to sea-level rise: a genetic basis for salinity tolerance in populations of a coastal marsh fish. Evolutionary Applications 1: 155–160.
          • 44. Hopkins GR, French SS, Brodie Jr ED (2013) Potential for local adaptation in response to an anthropogenic agent of selection: effects of road deicing salts on amphibian embryonic survival and development. Evolutionary Applications 6: 384–392.
          • 45. Armstrong JB, Duhon ST, Matacinski GM (1989) Raising the axolotl in captivity. In: Armstrong JB, Malacinski GM, editors. Developmental Biology of the Axolotl. New York: Oxford University Press. pp. 220–227.
            • 46. Harrison RG (1969) Harrison stages and description of the normal development of the spotted salamander, Amblystoma punctatum (Linn.). In: Harrison RG, editor. Organization and Development of the Embryo. New Haven, CT.: Yale University Press. pp. 44–66.
              • 47. Petranka JW, Doyle EJ (2010) Effects of road salts on the composition of seasonal pond communities: can the use of road salts enhance mosquito recruitment? Aquatic Ecology 44: 155–166.
              • 48. Borst M, Brown RA (2013) Chloride released from three permeable pavement surfaces after winter salt application. Journal of the American Water Resources Association 50: 29–41.
              • 49. Whitfield PH, Wade NL (1992) Monitoring transient water quality events electronically. Water Resources Bulletin 28: 703–711.
              • 50. Whitfield PH, Wade NL (1996) Transient water quality events in British Columbia coastal streams. Water Science and Technology 33: 151–161.
              • 51. Wu C-S, Yang W-K, Lee T-H, Gomez-Mestre I, Kam Y-C (2014) Salinity acclimation enhances salinity tolerance in tadpoles living in brackish water through increased Na+, K+-ATPase expression. Journal of Experimental Zoology 321A: 57–64.
              • 52. Hua J, Pierce BA (2013) Lethal and sublethal effects of salinity on three common Texas amphibians. Copeia 2013: 562–566.
              • 53. Hoffman RW, Goldman CR, Paulson S, Winters GR (1981) Aquatic impacts of deicing salts in the central Sierra Nevada mountains, California. Journal of the American Water Resources Association 17: 280–285.
              • 54. Harless ML, Huckins CJ, Grant JB, Pypker TG (2011) Effects of six chemical deicers on larval wood frogs (Rana sylvatica). Environmental Toxicology and Chemistry 30: 1637–1641.
              • 55. Langhans M, Peterson B, Walker A, Smith GR, Rettig JE (2009) Effects of Salinity on survivorship of wood frog (Rana sylvatica) tadpoles. Journal of Freshwater Ecology 24: 335–336.
              • 56. Able KW, Palmer RE (1988) Salinity effects on fertilization success and larval mortality of Fundulus heteroclitus. Copeia 1988: 345–350.
              • 57. Cutler DR, Edwards TC, Beard KH, Cutler A, Hess KT, et al. (2007) Random forests for classification in ecology. Ecology 88: 2783–2792.
              • 58. Breiman L, Friedman JH, Olshen RA, Stone CJ (1984) Classification and regression trees. Monterey, California: Wadsworth and Brooks/Cole.
                • 59. Breiman L (2001) Random forests. Machine Learning 45: 15–32.
                • 60. Alexander LG, Lailvaux SP, Pechmann JHK, De Vries PJ (2012) Effects of salinity on early life stages of the Gulf Coast toad, Incilius nebulifer (Anura: Bufonidae). Copeia 2012: 106–114.
                • 61. Brand A, Snodgrass JW, Gallagher MT, Casey RE, Van Meter R (2010) Lethal and sublethal effects of embryonic and larval exposure of Hyla versicolor to stormwater pond sediments. Archives of Environmental Contamination and Toxicology 58: 325–331.
                • 62. Dougherty CK, Smith GR (2006) Acute effects of road de-icers on the tadpoles of three anurans. Applied Herpetology 3: 87–93.
                • 63. Karraker NE, Gibbs JP (2011) Road deicing salt irreversibly disrupts osmoregulation of salamander egg clutches. Environmental Pollution 159: 833–835.
                • 64. Karraker NE, Gibbs JP, Vonesh JR (2008) Impacts of road deicing salt on the demography of vernal pool-breeding amphibians. Ecological Applications 18: 724–734.
                • 65. Karraker NE, Ruthig GR (2009) Effect of road deicing salt on the susceptibility of amphibian embryos to infection by water molds. Environmental Research 109: 40–45.
                • 66. Padhye AD, Ghate HV (1992) Sodium chloride and potassium chloride tolerance of different stages of the frog, Microhyla ornata. Herpetological Journal 2: 18–23.
                • 67. Viertel B (1999) Salt tolerance of Rana temporaria: spawning site selection and survival during embryonic development (Amphibia, Anura). Amphibia-Reptilia 20: 161–171.
                • 68. Collins SJ, Russell RW (2009) Toxicity of road salt to Nova Scotia amphibians. Environmental Pollution 157: 320–324.
                • 69. Dunson WA (1977) Tolerance to high temperature and salinity by tadpoles of the Philippine Frog, Rana cancrivora. Copeia 1977: 375–378.
                • 70. Gordon MS, Schmidt-Nielsen K, Kelly HM (1961) Osmotic regulation in the crab-eating frog (Rana cancrivora). Journal of Experimental Biology 38: 659–678.
                • 71. Griffis-Kyle KL (2005) Ontogenetic delays in effects of nitrite exposure on tiger salamanders (Ambystoma tigrinum) and wood frogs (Rana sylvatica). Environmental Toxicology and Chemistry 24: 1523–1527.
                • 72. Pahkala M, Laurila A, Merilä J (2001) Carry-over effects of ultraviolet-B radiation on larval fitness in Rana temporaria. Proceedings of the Royal Society B 268: 1699–1725.
                • 73. Smith GR, Waters MA, Rettig JE (2000) Consequences of embryonic UV-B exposure for embryos and tadpoles of the plains leopard frog. Conservation Biology 14: 1903–1907.
                • 74. Moran AL, Emlet RB (2001) Offspring size and performance in variable environments: field studies on a marine snail. Ecology 82: 1597–1612.
                • 75. Gorman HE, Nager RG (2004) Prenatal developmental conditions have long-term effects on offpsring fecundity. Proceedings of the Royal Society B 271: 1923–1928.
                • 76. Boone MD, Scott DE, Niewiarowski PH (2002) Effects of hatching time for larval Ambystomatid salamanders. Copeia 2002: 511–517.
                • 77. Touchon JC, Warkentin KM (2010) Short- and long-term effects of the abiotic egg environment on viability, development and vulnerability to predators of a Neotropical anuran. Functional Ecology 24: 566–575.
                • 78. Touchon JC, Warkentin KM (2013) Effects of plastic hatching timing carry over through metamorphosis in red-eyed treefrogs. Ecology 94: 850–860.
                • 79. Warkentin KM (1995) Adaptive plasticity in hatching age: a response to predation risk trade-offs. Proceedings of the National Academy of Sciences of the United States of Amerca 92: 3507–3510.
                • 80. Warkentin KM (1999) Effects of hatching age on development and hatchling morphology in the red-eyed treefrog, Agalychnis callidryas. Biological Journal of the Linnean Society 68: 443–470.
                • 81. Beebee TJC (1986) Acid tolerance of natterjack toad (Bufo calamita) development. Herpetological Journal 1: 78–81.
                • 82. Cooke AS (1979) The incluence of rearing density on the subsequent response to DDT dosing for tadpoles of the frog Rana temporaria. Bulletin of Environmental Contamination and Toxicology 21: 837–841.
                • 83. Gall BG, Stokes AN, French SS, Schlepphorst EA, Brodie III ED, et al. (2011) Tetrodotoxin levels in larval and metamorphosed newts (Taricha granulosa) and palatability to predatory dragonflies. Toxicon 57: 978–983.
                • 84. Metcalfe NB, Monaghan P (2001) Compensation for a bad start: grow now, pay later? Trends in Ecology and Evolution 16: 254–260.
                • 85. Charmantier G (2010) Ontogeny of osmoregulation in crustaceans: a review. Invertebrate Reproduction and Development 33: 177–190.
                • 86. Gosner KL, Black IH (1957) The effects of acidity on the development and hatching of New Jersey frogs. Ecology 38: 256–262.
                • 87. Alvarado RH, Dietz TH (1970) Effect of salt depletion on hydrominearal balance in larval Ambystoma gracile - I. Ionic composition. Comparative and Biochemical Physiology 33: 85–92.
                • 88. Alvarado RH, Dietz TH (1970) Effect of salt depletion on hydromineral balance in larval Ambystoma gracile - II. Kinetics of ion exchange. Comparative and Biochemical Physiology 33: 93–110.
                • 89. Alvarado RH, Moody A (1970) Sodium and chloride transport in tadpoles of the bullfrog Rana catesbeiana. American Journal of Physiology 218: 1510–1516.
                • 90. Bernabò I, Bonacci A, Coscarelli F, Tripepi M, Brunelli E (2013) Effects of salinity stress on Bufo balearicus and Bufo bufo tadpoles: tolerance, morphological gill alterations and Na+/K-ATPase localization. Aquatic Toxicology 132–133: 119–133.
                • 91. Alvarado RH, Cox TC (1985) Action of polyvalent cations on sodium transport across skin of larval and adult Rana catesbeiana. Journal of Experimental Zoology 236: 127–136.
                • 92. Calabrese EJ (2003) Hormesis: the dose-response revolution. Annual Reviews of Pharmacology and Toxicology 43: 175–197.
                • 93. Ashraf M, McNeilly T, Bradshaw AD (1989) The potential for evolution of tolerance to sodium chloride, calcium chloride, magnesium chloride and seawater in four grass species. New Phytologist 112: 245–254.
                • 94. Wu L (1981) The potential for evolution of salinity tolerance in Agrostis stolonifera L. and Agrostis tenuis Sibth. New Phytologist 89: 471–486.
                • 95. Hallsworth JE, Yakimov MM, Golyshin PN, Gillion JLM, D'Auria G, et al. (2007) Limits of life in MgCl2-containing environments: chaotropicity defines the window. Environmental Microbiology 9: 801–813.
                • 96. Beebee TJC (1985) Salt tolerances of natterjack toad (Bufo calamita) eggs and larvae from coastal and inland populations in Britain. Herpetological Journal 1: 14–16.
                • 97. Vonesh JR, De la Cruz O (2002) Complex life cycles and density dependence: assessing the contribution of egg mortality to amphibian declines. Oecologia 133: 325–333.