Interacting Effects of Temperature and Food on Early Growth of Two Plethodontid Salamanders
An amphibian’s life history includes growth, development, and reproduction, which are fundamental to their population dynamics. Yet for many amphibians, we lack knowledge about the environmental factors affecting early growth, which can affect size at metamorphosis, age and size at sexual maturity, and clutch sizes. In this study, we investigated the growth of young salamanders of two species of plethodontid salamanders with different life cycles, Plethodon cinereus (direct development) and Eurycea cirrigera (biphasic). To examine the interaction between body size and environmental conditions, we conducted a laboratory experiment on both species using a factorial design, with two food treatments crossed with two temperature treatments. Early growth patterns differed for both species of salamanders. Overall, both species grew at a slower rate when food was limited. Eurycea cirrigera displayed non-linear growth, which varied among treatments. In Plethodon cinereus, average growth in length was negligible, and mass declined at low temperatures. The results of this study suggest that effects of environmental change on salamander populations are complex and will vary according to species life history, food availability, and climate.
AMPHIBIAN body sizes depend on early growth, which depends on temperature, density, and predation risk. Like all ectotherms, amphibians rely on external heat sources for the regulation of their body temperatures (Huey and Stevenson, 1979; Pough, 1980; Duellman and Trueb, 1994), which fluctuate with the thermal range of their environment (Feder, 1983; Buckley et al., 2012). Individuals can adjust their thermal environment through habitat selection, occupying micro-environments with favorable temperatures and conditions. Nevertheless, both low and high environmental temperatures can significantly impact individual behavior and physiology by reducing foraging activities and modifying metabolic rates (Huey and Kingsolver, 1989; Angilletta et al., 2004). In particular, a warm environment can cause an amphibian’s metabolic rate to increase, potentially leading to a reduction in body size (Homyack et al., 2010; Sheridan and Bickford, 2011). For example, in two separate studies, Plethodon cinereus exhibited negative energy budgets when temperature was high, causing an increase in metabolic rate (Bobka et al., 1981; Homyack et al., 2011). Conversely, in cooler environments, ectotherms may enter a state of metabolic depression, reducing their growth (Zhu et al., 2022). Slower growth in cooler conditions, which leads to later maturation and ultimately larger body sizes, has been proposed to explain the phenomenon known as the temperature–size rule (TSR; Atkinson, 1994; Forster et al., 2012), whereby individuals within species are larger in cool climates and smaller in warm climates (Huey and Kingsolver, 1989; Angilletta et al., 2004). Projected increases in mean temperatures due to climate change may change growth, demography, and body size distributions of amphibian populations, threatening their long-term survival (Blaustein et al., 2010).
The phenomenon of the TSR has been documented for a range of vertebrate taxa, but evidence within amphibians, especially salamanders, in wild populations has been mixed (Adams and Church, 2008). One reason could be that observational and experimental studies of growth in salamanders have not addressed the interaction between temperature and the mode of early development. Among sympatric species in the salamander family Plethodontidae, species exhibit paedomorphosis (adults retain larval traits), direct development (young born as miniature adults), and an aquatic larval phase (a biphasic life history typical for most amphibians; Rabosky and Adams, 2012; Beachy et al., 2017). Plethodontid salamanders are also notable for being lungless, relying on cutaneous respiration for gas exchange (Whitford and Hutchison, 1965). As a result, the salamanders must remain moist to avoid desiccation due to their permeable skin (Feder, 1983; Marshall and Camp, 2006). How this affects their metabolic demands and foraging activity at increased temperatures is not well understood.
We conducted a laboratory experiment on plethodontid salamanders designed to disentangle the contributions of temperature and resource availability on early growth in plethodontid species with different early life histories. We compared the early growth of individually housed Eurycea cirrigera, the Southern Two-lined Salamander, and Plethodon cinereus, the Eastern Red-backed Salamander. Both species were collected from locations in southwest Virginia and maintained in laboratory conditions for over a year. We hypothesized that at higher levels of food availability, salamanders reared at lower temperatures will grow larger than salamanders reared at higher temperatures. We used a fully factorial experimental design, in which two thermal treatments were crossed with two levels of food availability, and monitored the growth of salamanders, housed individually, within each treatment. Our objective was to test for direct effects of temperature on variation in body size after one year, as well as test for interacting effects between temperature and resource (food) availability.
Study species.—
Eurycea cirrigera, the Southern Two-lined Salamander, is a common salamander found within the southeastern United States (Petranka, 1998; Brophy and Pauley, 2002; Guy et al., 2004). This biphasic species has aquatic larvae that metamorphose into semi-terrestrial juveniles and adults (Jakubanis et al., 2008). Larval E. cirrigera are fully aquatic and breathe through external gills. Eurycea cirrigera reaches sexual maturity after approximately two to four years (Pfingsten, 2013). Plethodon cinereus, the Eastern Red-backed Salamander, is a widely distributed plethodontid salamander ranging from Canada to the southeastern United States, including North Carolina and Tennessee (Nagel, 1977; Petranka, 1998). This species of salamander is a fully terrestrial woodland salamander whose eggs undergo direct development and hatch into miniature adults. Plethodon cinereus reaches sexual maturity after approximately two years (Bausmann and Whitaker, 1987).
MATERIALS AND METHODS
Collection and housing of experimental animals.—
We collected a total of 159 larval E. cirrigera from the Jefferson National Forest (Craig’s Creek; 518 m elevation; 37.33184, –80.33834) in Montgomery County, Virginia. We made four collections between late June and early August of 2020, using dipnets to capture larvae. We placed larval salamanders in buckets with a RESUN-160 oxygen bubbler. We then transported the salamanders back to the Virginia Tech campus in Blacksburg, Virginia. Upon arrival, we assigned each salamander an individual ID. We then individually housed each salamander in a Dart CH48DEF 48 oz Tamper-Resistant Clear Hinged Container filled with water filtered by reverse osmosis. After housing the salamander in the ATC60 Conviron Growth Chamber (Manitoba, Canada) for four months (the acclimation period), we assigned each individual to one of four temperature/food treatments (described below). After collection, individuals were maintained in similar conditions for a period to acclimatize to the laboratory environment. They were assigned to their treatments in November 2020, with approximately 40 individuals in each treatment.
Between June and early October in 2021, we collected a total of 122 young-of-the-year or juvenile P. cinereus from Minie Ball Hill (Salt Pond Mountain; 1,128 m elevation; 37.40501, –80.50979) in Giles County, Virginia. We measured each captured salamander to ensure they were not adults; only salamanders under 29 mm in total length were selected for this experiment (Sayler, 1966; Petranka, 1998). If salamanders were within the desired total length, we placed them in plastic bags with leaf litter in coolers and transported them to the Virginia Tech campus in Blacksburg, Virginia. Each salamander was assigned an individual ID and its snout–vent length (SVL), total length, and mass measured. Snout–vent length was the tip of salamander snout to its posterior vent. We then housed each salamander in a Dart CH48DEF 48 oz Tamper-Resistant Clear Hinged Container. To maintain a moist environment, we filled the plastic containers with moist Pacific Blue Basic Paper Towels. After an acclimation period in the growth chamber, individual P. cinereus were randomly assigned to one of our four temperature/food treatments (described below). There were approximately 30 individuals of P. cinereus in each treatment.
Experimental conditions.—
All salamanders of both species were housed in one of two of the ATC60 Conviron Growth Chambers, which maintained our experimental temperature regimes. The growth chambers were equipped with built-in thermostats as well as two backup thermometers (Ideal Sciences Temp Stick and a ThermoPro TP50 digital hygrometer/thermometer) to ensure accurate readings. Throughout the experiment, individual salamanders were monitored, and tanks were cleaned in accordance with animal care standards (VT-IACUC 20-094). Both E. cirrigera and P. cinereus were fed on the same day we cleaned their housing containers. Our experiment crossed two temperature regimes and two diets in a factorial design for a total of four treatments. We measured the growth of individual salamanders of E. cirrigera monthly from December 2020 to December 2022. To prevent any harm to the salamanders, which initially showed sensitivity to handling, we decided to not measure SVL. Due to low survival in some treatments, we analyzed data from only the first 15 months of the experiment (following acclimation), that is, December 2020 to March 2022. We monitored the growth of individuals of P. cinereus from November 2021 to October 2022.
Temperature and light cycles in each treatment.—
Following each species’ acclimation period, we manipulated the temperature and lighting cycle to mimic seasonal variation in day length and thermal conditions. In late October 2020, growth chambers were set to 16°C and 11°C (respectively, high temperature and low temperature). The light:dark conditions were initially set to a 12:12 hour cycle. In November 2020, we lowered the HT chamber temperature to 15°C, as animals showed signs of stress in the warmer conditions. Initial environmental temperatures were chosen based on local temperatures at the collection site and were also consistent with published temperatures in similar experiments (Beachy, 2018). To represent the onset of winter, we gradually reduced the temperatures in the chambers so that by 22 January 2021, the growth chamber temperatures were 13°C in the high-temperature (HT) treatment and 9°C in the low-temperature (LT) treatment. On 7 May 2021, the growth chamber temperatures were lowered to 11°C (HT) and 7°C (LT). Beginning on 20 May 2021, we gradually increased the temperature of both growth chambers to 11°C (LT) and 15°C (HT) where they were maintained until the following autumn.
After collecting and acclimating P. cinereus in the summer of 2021, we added them to the experimental growth chambers in October 2021. Beginning on 1 October 2021, we incrementally lowered the chambers to 13°C (HT) and 9°C (LT), and on 8 November 2021, we decreased the temperatures to 12°C (HT) and 8°C (LT). On 8 December 2021, we lowered the temperatures to 11°C (HT) and 7°C (LT) and maintained them until spring. Chamber lighting was set to a 9:15 hour light:dark cycle during this period of cooler temperatures. Between 20 March 2022 and 26 March 2022, we changed the lighting cycle to a 12:12 light:dark hour cycle. Between 26 May 2022 and 4 June 2022, we incrementally increased the temperatures to 15°C (HT) and 11°C (LT) to mimic the onset of summer. In October 2022, we concluded the experiment with P. cinereus; we concluded the experiment with E. cirrigera in December 2022. At the end of each experiment, we humanely euthanized all the remaining salamanders in a buffered bath of MS-222.
Food treatments.—
For the duration of their time in experimental conditions, individuals of each species were assigned to either a high or a low quantity food treatment. The diet differed between aquatic (E. cirrigera) and terrestrial (P. cinereus) salamanders. We fed our aquatic larval E. cirrigera California Black Worms (Lumbriculus variegatus). These were obtained from Eastern Aquatics (Lancaster, Pennsylvania). We fed P. cinereus a wingless strain of Drosophila melanogaster obtained initially from Josh’s Frogs LLC and maintained in the lab. New cultures were made twice a month. Larval salamanders in the low-food treatment for E. cirrigera received one black worm per week, while the salamanders in the high-food treatment received two black worms per week. For P. cinereus, individuals in the high-food treatment received a total of 30 fruit flies per week, while the salamanders in the low-food treatment received 20 fruit flies per week.
Measurement methods.—
To measure total length (TL) of E. cirrigera, we used a light microscope (Leica M80–ACHRO 0.5X Lens) with a built-in camera and imaging computer. We used an analytical balance (U.S. Solid Model: USS-DBS15-1: accuracy 0.001 g/1 mg) for mass. We gently straightened the salamanders using our finger to measure TL. For P. cinereus, we monitored the growth of individual salamanders from October 2021 to October 2022, taking repeated measurements of mass (g), SVL (mm), and TL (mm) every month. To measure SVL and TL, we used an electric digital caliper (NEIKO 01407) after placing the individual in a Ziploc bag. Once inside the bag, we pressed the salamander against the edge and measured it in a straightened position. To measure mass, we used an analytical balance and weighed each individual in a weigh boat.
All data were entered into Excel spreadsheets and checked after entry by plotting their distribution and inspecting outliers to ensure they were not entered incorrectly. We analyzed the data in program R version 4.2.2 (R Core Team, 2023). To determine how environmental conditions at early life stages influence the growth of E. cirrigera and P. cinereus, body growth over time in each treatment was compared using a generalized additive mixed model (GAMM). Individual identity was set as a random effect (repeated samples), while food and temperature were fixed effects. Differences in body size (SVL in mm) and mass (g) at the end of the experiment were analyzed by analysis of variance (ANOVA). We used a Tukey post hoc test to evaluate significant differences between group means. For all statistical tests, we set α = 0.05.
RESULTS
Eurycea cirrigera.—
During preliminary inspection of the data, we noted that trends in total length were highly variable among individuals and were non-linear throughout the duration of the experiment. To account for this non-linearity, we used a generalized additive mixed model (GAMM) to describe average trends within each treatment, including a random effect of individual. We then analyzed the smoothed relationship between size of E. cirrigera and temperature and food (Fig. 1). An ANOVA evaluating the differences among treatments in body length after 15 months revealed a non-significant interaction between temperature and food on mass (F1,34 = 0.360, P = 0.552). The interacting effects of temperature and food on body length at the end of the experiment were statistically significant (F1,34 = 4.24, P = 0.049; Table 1).
![Line graph showing mean total length (mm) of E. cirrigera over 15 months across four treatments. The figure includes two panels representing food treatments (high food and low food). Each panel contains two lines representing temperature treatments (high temperature [HT] and low temperature [LT]).](/view/journals/cope/113/2/full-cope-113-2-380-f01.jpg)
![Line graph showing mean total length (mm) of E. cirrigera over 15 months across four treatments. The figure includes two panels representing food treatments (high food and low food). Each panel contains two lines representing temperature treatments (high temperature [HT] and low temperature [LT]).](/view/journals/cope/113/2/inline-cope-113-2-380-f01.jpg)
Citation: Ichthyology & Herpetology 113, 2; 10.1643/h2024041
The mixed-effects model of both body length and mass revealed that growth patterns over the course of the experiment differed according to both food and temperature. Similar horseshoe-shaped patterns in both length and mass were observed in all treatments, although the maximum sizes reached were greater in cool treatments (Figs. 1, 2). In the high-temperature treatments, the salamanders displayed a gradual increase in body size followed by a gradual decrease in body size. However, the timing of growth varied between treatments. The individuals exposed to the high-temperature treatments experienced decreases in mass after ten months, while the low-temperature treatment exhibited gradual increases in growth, followed by stabilization and a gradual increase in growth. The spike in growth (in Fig. 2) observed in the low-temperature/low-food treatment can be attributed to the low number of surviving animals in the treatment. To clarify this pattern, we conducted a sensitivity analysis of the results in Figure 2 using only the individuals that survived through the entire 15 months of these experiments (so that sample size across all time points is constant). We found similar patterns with this subset of individuals, suggesting that our conclusions are not driven by differential growth of survivors.
![Line graph showing mean mass (g) of E. cirrigera over 15 months across four treatments. The figure includes two panels representing food treatments (high food and low food). Each panel contains two lines representing temperature treatments (high temperature [HT] and low temperature [LT]).](/view/journals/cope/113/2/full-cope-113-2-380-f02.jpg)
![Line graph showing mean mass (g) of E. cirrigera over 15 months across four treatments. The figure includes two panels representing food treatments (high food and low food). Each panel contains two lines representing temperature treatments (high temperature [HT] and low temperature [LT]).](/view/journals/cope/113/2/inline-cope-113-2-380-f02.jpg)
Citation: Ichthyology & Herpetology 113, 2; 10.1643/h2024041
Plethodon cinereus.—
Preliminary inspection of the data on P. cinereus revealed nonlinear trends in length (SVL) and mass over the duration of the experiment, mirroring the results of E. cirrigera. However, variation in the data on SVL and TL between monthly measurements was highly variable and sometimes negative, suggesting that measurement error from the caliper could be masking differences in incremental growth among treatments. Therefore, only SVL measurements in October 2021, March 2022, and October 2022 were included in the smoothed analysis of body-size trends. The smoothed relationship between growth of P. cinereus and temperature and food produced with the GAMM shows that growth was minimal across all treatments (Fig. 3). Differences among treatments in body size after 12 months were analyzed using an ANOVA, which revealed a significant interaction between temperature and food on body size (F1,58 = 6.16; P = 0.0213; Table 2). A Tukey post hoc test suggested that the differences in length due to food levels were most significant at low temperatures (diff = –3.164, P adj = 0.013).
![Line graph showing mean SVL (mm) of P. cinereus after 13 months across four treatments. The figure includes two panels representing food treatments (high food and low food). Each panel contains two lines representing temperature treatments (high temperature [HT] and low temperature [LT]).](/view/journals/cope/113/2/full-cope-113-2-380-f03.jpg)
![Line graph showing mean SVL (mm) of P. cinereus after 13 months across four treatments. The figure includes two panels representing food treatments (high food and low food). Each panel contains two lines representing temperature treatments (high temperature [HT] and low temperature [LT]).](/view/journals/cope/113/2/inline-cope-113-2-380-f03.jpg)
Citation: Ichthyology & Herpetology 113, 2; 10.1643/h2024041
Figure 4 shows the smooth trends in mass (g) from the GAMM analysis over the duration of the experiment. All monthly mass measurements were included for this analysis. This GAMM model also supported an interacting effect of temperature and food on body mass (in grams) over the course of the experiment (Fig. 4). Individuals at high temperatures experienced increases in mass throughout the experiment at both food levels, but the effect was much greater for the high-food treatment. Individuals in the low-temperature treatment stayed the same or decreased in mass over the study, especially in the low-food treatment.
![Line graph showing mean mass (g) of P. cinereus after 13 months across four treatments. The figure includes two panels representing food treatments (high food and low food). Each panel contains two lines representing temperature treatments (high temperature [HT] and low temperature [LT]).](/view/journals/cope/113/2/full-cope-113-2-380-f04.jpg)
![Line graph showing mean mass (g) of P. cinereus after 13 months across four treatments. The figure includes two panels representing food treatments (high food and low food). Each panel contains two lines representing temperature treatments (high temperature [HT] and low temperature [LT]).](/view/journals/cope/113/2/inline-cope-113-2-380-f04.jpg)
Citation: Ichthyology & Herpetology 113, 2; 10.1643/h2024041
DISCUSSION
Early growth of salamanders can influence size at key life history transitions, including metamorphosis and maturation, which in turn are likely to affect reproductive traits and fitness (Semlitsch et al., 1988). To understand environmental factors contributing to early growth, we investigated the interacting effects of food availability and temperature on two species with different life cycles. We compared growth responses of a species with an aquatic larval phase (Eurycea cirrigera) to a species with direct development (Plethodon cinereus). We found differences in growth between species which could indicate that their populations will respond to changes in future climate conditions differently, even though they are sympatric and have a similar morphology as adults.
While we did find clear differences between species, some of our results may be an artifact of experimental conditions. First, the clearest effect of the factorial study design for E. cirrigera was a difference in survival, which may have been due to poor water quality in the high-temperature, high-food treatment. The larval salamanders in this treatment had much higher mortality than the other three treatments. Second, we expected a higher proportion of our experimental animals to reach metamorphosis before the experiment’s end, based on similar studies on related species (Beachy, 1995). The high-food, high-temperature treatment had two salamanders undergo metamorphosis (21 June 2021 and 15 May 2022). One salamander metamorphosed in the low-food, high-temperature treatment (5 July 2021) and in the high-food, low-temperature treatment (29 December 2021). While it is interesting that most metamorphic animals were in high-food treatments, the numbers are too low to say anything conclusive about the effect of treatment on metamorphosis.
The decline in body sizes for the larval salamanders of E. cirrigera (Fig. 1) was surprising. Our measurements of these salamanders were taken with a microscope and analytical balance, both of which had less measurement error than the methods used for P. cinereus. Declines in average body length and mass were also robust to a sensitivity analysis where we considered only individuals surviving the entirety of the experimental period. One possibility is that declines may have been caused by pre-metamorphic shrinkage due to reduced tail fins (Parichy, 1998; Alcobendas et al., 2004). The GAMM shows that average mass declined in all the high-temperature treatments before average mass declined in low-temperature treatments (Fig. 2). This could mean that under more natural conditions, salamanders in warmer conditions metamorphose earlier than those in cool conditions.
For P. cinereus, survivorship was much higher, but again, growth patterns did not match our expectations. In the warm-temperature treatments, there was a gradual increase in both length and mass. Conversely, the individuals subjected to the cool-temperature treatments experienced a decrease in mass, especially at low-food levels; length in these treatments appeared to be constant (Figs. 3, 4). The high-temperature, high-food treatment had the greatest increase in mass, while the low-temperature, low-food treatment had the greatest decrease in mass (Fig. 4). In all the treatments, salamanders did not necessarily consume all prey each week, suggesting that food limitation is not responsible for declines in mass. Decreasing mass of salamanders exposed to lower temperatures may have been caused by decreases in metabolic activity in addition to reduced growth rates (Arendt, 2011).
This study is one of only a few studies of growth of larval plethodontid salamanders in response to temperature and food availability and is the first to examine the effects of temperature and food availability on a direct-developing plethodontid salamander. Previous studies have examined similar factors in different species with similar development to E. cirrigera: Chattooga Dusky Salamander (Desmognathus perlapsus; metamorphosis; Beachy, 1995), Four-toed Salamander (Hemidactylium scutatum; postembryonic metamorphosis; O’Laughlin and Harris, 2000), Blue Ridge Black-bellied Salamander (D. mavrokoilius; metamorphosis; Hickerson et al., 2005), and Blue Ridge Two-lined Salamander (Eurycea wilderae; metamorphosis; Beachy, 2018). As in Beachy (2018), we observed that both E. cirrigera and P. cinereus with limited food availability grew at a slower rate compared to the salamanders with higher food availability. In a study examining the growth of the D. perlapsus (Beachy, 1995), individuals in high-temperature treatments appeared to be affected more by the food regimes. Surprisingly, we found the opposite relationship between food and temperature for E. cirrigera. In our study of E. cirrigera, high-food levels had a greater impact in low-temperature treatments, allowing the salamanders to grow larger in mass. The results are more similar to those reported by Hickerson et al. (2005), in which high-food treatments had bigger impacts on the individual growth of D. mavrokoilius in lower temperatures.
Our results suggest that resources can interact with temperature to determine responses of salamanders to environmental gradients and could explain mixed evidence for the temperature–size rule (TSR) in observational studies of salamanders, which have shown an intricate relationship between body size and temperature. In a large-scale study of plethodontid salamanders collected across naturally occurring environmental gradients in temperature, Adams and Church (2008) found that only three out of 40 species of plethodontid salamanders exhibited a negative correlation between body size and temperature. Additionally, seven out of 40 species exhibited a positive relation between body size and temperature, the inverse of the TSR (Adams and Church, 2008). By contrast, in a separate study, Caruso et al. (2014) reported that, when comparing museum specimens with current size distributions in matching populations, six species of Plethodon displayed a reduction in body size over decades between collections, presumably due to environmental warming. Studies of amphibian populations that span an elevational gradient, including plethodontid salamanders in the Appalachian Mountains, have demonstrated that elevation is correlated with shifts in growth, development, and reproduction (Morrison and Hero, 2003). For example, amphibians at higher altitudes have extended larval periods and take longer to reach minimum size for sexual maturity (Morrison and Hero, 2003; Liles et al., 2017).
The results of these studies and ours highlight the need for additional research addressing the complex relationship between temperature, size, and the ecological and physiological mechanisms that may be impacting salamander growth. Our results on E. cirrigera and P. cinereus suggest that resource availability will mediate effects of temperature on salamander growth, and that growth responses to temperature and food depend on the species’ the life cycle. Future studies under more natural conditions could address how temperature affects individual metabolic demands and foraging activity of plethodontids, as these mechanisms likely determine the timing and size at key life history transitions and will affect species’ responses to environmental change.
DATA ACCESSIBILITY
Unless an alternative copyright or statement noting that a figure is reprinted from a previous source is noted in a figure caption, the published images and illustrations in this article are licensed by the American Society of Ichthyologists and Herpetologists for use if the use includes a citation to the original source (American Society of Ichthyologists and Herpetologists, the DOI of the Ichthyology & Herpetology article, and any individual image credits listed in the figure caption) in accordance with the Creative Commons Attribution CC BY License.
AI STATEMENT
The authors declare that AI (ChatGPT) was used only for grammar purposes during the preparation of this article.
Mean total length (mm) of Eurycea cirrigera in two food treatments, crossed with two temperature treatments (high temperature [HT], high food; high temperature, low food; low temperature [LT], high food; low temperature, low food). The blue line represents the low-temperature treatment, and the red line represents the high-temperature treatment. Shaded areas represent 95% confidence intervals around the GAMM trend parameters.
Mean mass (g) of Eurycea cirrigera in two food treatments, crossed with two temperature treatments (high temperature [HT], high food; high temperature, low food; low temperature [LT], high food; low temperature, low food). The blue line represents the low-temperature treatment, and the red line represents the high-temperature treatment. Shaded areas represent 95% confidence intervals around the GAMM trend parameters.
Mean SVL (mm) of Plethodon cinereus in two food treatments, crossed with two temperature treatments (high temperature [HT], high food; high temperature, low food; low temperature [LT], high food; low temperature, low food). The blue line represents the low-temperature treatment, and the red line represents the high-temperature treatment. Shaded areas represent 95% confidence intervals around the GAMM trend parameters.
Mean mass (g) of Plethodon cinereus in two food treatments, crossed with two temperature treatments (high temperature [HT], high food; high temperature, low food; low temperature [LT], high food; low temperature, low food). The blue line represents the low-temperature treatment, and the red line represents the high-temperature treatment. Shaded areas represent 95% confidence intervals around the GAMM trend parameters.
Contributor Notes
Associate Editor: J. M. Davenport.