Lethal Heat Exchange—Short-Term Thermoregulation in Two Triturus Species During Abrupt Changes in Living Media (Water vs. Air)

Abstract

As adults, some newts exhibit a multiphasic lifestyle, switching between aquatic and terrestrial habitats. Under laboratory conditions, we provided an experiment to expose newts, which were in their aquatic phase, to air as a surrounding medium. We studied how the body temperature changed when the ambient temperature increased. The thirteen specimens used in the study belonged to two species from the genus Triturus sp.—T. ivanbureschi and T. dobrogicus. For temperature measurements, we used a precise thermometer with a beaded thermocouple and an IR thermal camera. We started the experiment at 17 °C and increased the air temperature by approximately 1 °C every 10 min. The newts were exposed to the air until the first signs of physical exhaustion appeared. An increase of 1 °C in ambient temperature led to an average increase of 0.87 °C in the body temperature of the newts across the four experimental days. The measured body temperature showed a consistent increase during all experimental sessions, but it did not equalize with the environmental temperature. The body temperature in all specimens remained lower by an average of 2.24 (±0.02) °C.

1. Introduction

A multiphasic lifestyle is typical of some newt species and is characterized by periodic adaptation of adults for living in aquatic and terrestrial habitats [1,2,3]. These repeated transformations to the habitat pose a challenge for the organism and are often associated with major morphological, physiological, and behavioral changes to accommodate the different physical characteristics of water and air [4,5,6]. The adults of newts with a multiphasic way of life change their external habitus. Modifications in the form of the tail, the appendages, and the labial morphology occur [1,7,8] in the discrete aquatic and terrestrial morphotypes sensu [9]. Seasonal changes appear also in the oropharyngeal structures [9,10] and in the feeding modes of the newts [3,11]. Even the structure of the skin is different, as in the terrestrial phase, the corneum cell layer shows increased thickness [12].

The thermoregulation capacity of the newts differs in the aquatic and terrestrial phases, because of the different physical properties of the living media. In aquatic environments, most of the ectothermic vertebrates are not able to thermoregulate, and their temperature is directly dependent on the temperature of the water surrounding the body [13]. On land, spatial repositioning and other ethological mechanisms provide more options for little-sized ectotherm vertebrates to regulate their temperature [14]. In urodeles a rather complex “cutaneous water evaporation—water balance—body temperature” relation was exposed, but the investigations are at their initial phases [15,16]. The plethodontids are the only exception to that trend as more information was gathered in recent decades and the thermophysiology and behavior in this group were better understood [17]. For terrestrial salamanders, it is important to regularly rehydrate, as they dehydrate to some degree in relation to their life activities on land [18]. Water loss amounts are directly related to the air humidity, air temperature, and speed of the wind [19].

In the present investigation, we provided experiments with specimens of the Buresch’s crested newt (T. ivanbureschi) and the Danube crested newt (T. dobrogicus) during their aquatic annual phase. The aim of the present study was to track the body temperature dynamics induced by a sudden change in the living media and an increase in the temperature of the ambient air. We discuss the methods to obtain thermal data from such small, slender, and fragile vertebrates.

2. Materials and Methods

The Danube crested newt (Triturus dobrogicus) has a very narrow distribution area, occupying a very specific ecological niche along the Danube River. Ref. [20] described in detail the general areal of the species in the Danube lowlands. These newts occur in the following regions from west to east: lower Austria, southeastern Czech Republic, southern Slovakia, Hungary, Serbia, Bosnia and Herzegovina, Croatia, southern Romania, northern Bulgaria, Moldova, and Ukraine. The Danube crested newt is a slender-bodied newt with a total length (TL) of up to 160 mm [4]. This species has a biphasic lifestyle with an aquatic phase lasting 6 months or longer [19]. This is the representative of the genus Triturus with one of the longest aquatic phases. The animals winter on land [20].

The Buresch’s crested newt (Triturus ivanbureschi) is inhabiting still ponds with thick vegetation [11]. These newts were found across various habitats—they could live in regions both with long dry summers, as well in areas with longer water periods [21].

This study was reviewed and approved by the ethical committee of Ovidius University of Constanta and the Danube Delta Biosphere Reserve Administration (Number 2956). For our experiments, we used adult Danube crested newt specimens captured from Romania in 2020. The four Buresch’s crested newts were saved from destroyed habitat in Bulgaria. The animals were transported to a laboratory of the Faculty of Natural Sciences, “Konstantantin Preslavski” University of Shumen in Bulgaria. All newts were quarantined individually for one week and fed ad libitum with earthworms. In the experiment were included 5 males and 4 females from the species T. dobrogicus and 2 males and 2 females from the species T. ivanbureschi. After the experiment, all T. dobrogicus were released at the place of their capture, and the Buresch’s crested newts were released at suitable habitat.

In the laboratory the newts were placed in a room kept at 17 °C, ambient temperature, maintained constant for the period from 4 to 12 June 2020, with natural photoperiod. The air humidity was kept in the range of 36–42% during the days of the experiment (measured with a hygrometer Eterim ThermoPro (ThermoPro Ltd, Fuzhou, Fujian, China). The newts were kept in 2 glass tanks filled with dichloride water. T. dobrogicus were kept in water with temperatures of 15.4 °C to 17.1 °C [±0,7]. The specimens of T. ivanbureschi were kept in water temperatures from 16.6 to 16.7 °C [±0.5]. All of the newts were in a period of aquatic phase and were not prepared for their annual transition into terrestrial phase. The experiment was designed in order to expose the newts to surrounding air without the opportunity to hide in a humid place.

The experiments were conducted in a closed room for four days: three successive days, 4, 5, 6 June, and after a six-day interval on 12 June 2020. For each of the days, the newts were removed from the water and placed in 4 dry plastic containers (25 cm × 15.8 cm × 15 cm (l × w × h)). Four T. ivanbureschi individuals were housed in a single container, while T. dobrogicus individuals were distributed across three containers, with three individuals in each. At the end of the experiments, the newts were returned to the water with the initial temperature to rehydrate.

At the beginning of the experiment for each of the days, the ambient temperature was gradually increased, always starting from 17 °C (a constant initial value of the ambient temperature for the 4 days of the experiment). We stopped the experiments by reaching an average maximum ambient temperature of 26.4 °C [±0.4]. The temperature increase was at intervals of an average of 12.07 [±3] min. The degrees were increased in an interval of 0.9 to 1.2 [±0.1] °C on average for each set per day. The total temperature amplitude was an average of 9.5 °C [±0.5] between the minimum and maximum temperature for the day. Depending on the condition of the newts, between 8 and 10 steps of increasing the ambient temperature were performed. The condition of the newts was monitored during the experiment, and all of the specimens were tested for their response to lateral tipping with a cotton swab. The body surface was gently touched with filter paper for testing the presence of moisture. The experiment was stopped when we detected the lack of reaction or drying of the skin.

During the experiment, we monitored the ambient temperature using a thermometer “Therma Elite 221-061” (ETI Ltd., Easting Close, Worthing, West Sussex, UK). The accuracy of the unit is ±0.4 °C, or ±0.1% in the range of −99.9 to 299.9 °C. As a K-thermocouple for the thermometer, we used a “Fluke Electronics 80PK-1 K-Type Thermocouple Bead Probe” (Fluke Corporation, Everett, WA, USA). The thermocouple was used for contact measurement of the skin of the newts. We performed thermal profiles of the newts in dorsal projection using a thermal camera “FLIR C2” (the accuracy of the item is ±1.5 °C), with an MSX Thermosystem (FLIR^® Systems, Inc., Wilsonville, OR, USA). The emissivity settings were adjusted for the measurement according to [22]. The radiometric images were analyzed using “FLIR Tools 6.X” software (FLIR^® Systems, Inc., Wilsonville, OR, USA). The images were used for the calculation of the minimal, maximal, and average temperatures measured right beside the head (see Figure 1).

To assess the effect of the ambient temperature changes on the body temperature of the newts, a linear multiple regression was performed with fixed factors (dummy variables) corresponding to each of the four experimental days. Analyses were conducted in Microsoft Excel 365 (Microsoft Corporation, Redmond, WA, USA) using the Data Analysis Toolpak. The overall model significance was evaluated using an F-test, considering the degrees of freedom for regression and residuals, the F-value, and the associated p-values. The coefficient of determination (R²) was used to quantify the proportion of variance in the dependent variable explained by the model, while individual β coefficients and their p-values indicated the magnitude, direction, and significance of each predictor’s effect.

To compare body temperatures during the first three days (D1–D3) with those on day 4 (D4), a one-way ANOVA with “day” as a fixed factor (α = 0.05) was applied. Due to the lack of an automated Tukey post hoc test in Excel, pairwise t-tests between D4 and each of D1–D3 were performed manually, with a Bonferroni correction for multiple comparisons (α = 0.0167). For each individual, the change Δ = D4—mean(D1–D3) was calculated, and group-wise means and standard deviations of Δ were reported. Statistical significance was set at p < 0.05.

3. Results

All newts included in our experiments showed similar behavior when they were removed from the water and placed in plastic containers. The animals remained calm and showed limited locomotor activities. At the beginning of every set of measurements, the newts reacted to the lateral touching—when stimulated, they made several steps. We continued the experiment until one of the newts stopped reacting to the stimulus.

The experimental sets during the four days lasted between 1:35:05 and 2:13:44. The analysis showed a statistically significant linear relationship between the body temperature of newts in both species, which was more than 92% influenced by the change in environmental temperature. The correlation coefficient for all combinations was above 0.92, varying within the limits of 0.93–0.99, which indicated a strong dependence of the newt’s body temperature on the change in environmental temperature. Despite the relatively consistent increase in the body temperature of the newts during all experimental sessions, it did not equalize with the environmental temperature. The body temperature of all specimens remained lower by an average of 2.24 (±0.02) °C. In T. dobrogicus, the difference ranged from 2.06 to 2.50 (±0.19) °C, and in T. ivanbureschi, from 2.02 to 2.63 (±0.28) °C (

Table 1. Mean daily ambient and body temperatures of all specimens: t A—ambient temperature; t TI—temperature of the newts Triturus ivanbureschi; t TD—temperature of the newts Triturus dobrogicus; t TD (TC)—T. dobrogicus—average body temperature measured by thermometer with thermocouple; t TI (TC)—T. ivanbureschi—average body temperature measured by thermometer with thermocouple; n—number of animals; t A—t TD (TC) difference between mean ambient temperature and mean body temperature of T. dobrogicus (measured with a thermocouple); t A -t TI (TC) difference between mean ambient temperature and mean body temperature of T. ivanbureschi (measured with a thermocouple); all temperature values were reported in °C.

Day	t A	t TD (TC)	t TI (TC)	t A—t TD (TC)	t A—t TI (TC)
mean	23.19	21.13	21.22	2.06	2.06
min	17.00	15.58	15.58	1.42	1.43
max	25.90	23.71	23.90	2.38	2.40
SD	3.08	2.84	2.84	0.33	0.39
n		9	4
number of sets	9	9	9	9	9
4 June 2020
mean	23.36	21.10	21.34	2.26	2.02
min	17.00	16.31	16.35	0.69	0.65
max	26.30	24.48	24.60	4.10	3.38
SD	3.03	3.07	3.02	0.95	0.78
n		9	4
number of sets	10	10	9	10	9
5 June 2020
mean	23.38	21.20	21.18	2.18	2.19
min	17.00	16.66	16.73	0.34	0.27
max	26.80	24.53	24.35	2.80	2.78
SD	3.34	2.86	2.66	0.76	0.81
n		9	4
number of sets	8	8	8	8	8
6 June 2020
mean	22.98	20.48	20.35	2.50	2.63
min	16.60	16.04	16.43	−0.33	0.18
max	26.60	23.49	23.43	4.50	4.38
SD	3.34	2.69	2.60	1.22	1.17
n		9	4
number of sets	10	10	10	10	10
12 June 2020

). The body temperature of newts of both species on the 4th day (after an interval of 6 days) was measured slightly lower.

A multiple linear regression analysis revealed that the model significantly predicted the outcome variable on all four days. On Day 1, the model explained a very large proportion of the variance, F (3, 113) = 2378, p < 0.001, R² = 0.98. Similar patterns were observed on Days 2–4, with F-values consistently high (F (3, 126) = 408; F (3, 100) = 818; F (3, 126) = 317) and p-values < 0.001, indicating that the predictors had strong and stable effects across all measurement days (

Table 2. Multiple linear regression with fixed factors. Dependent variable: body temp of two species, T. dobrogicus and T. ivanbureschi; independent variables: ambient temperature, group dummy, and interaction: ambient temperature*group dummy. p < 0.05. Notes: β = standardized regression coefficient, SE = standard error, t = t-statistic, p = significance level, CI = confidence interval, R² = coefficient of determination. Effect indicates direction of the effect (+ = positive, − = negative; only significant predictors receive a direction (+/–) in the effect column). The model row was italicized to visually separate model statistics from individual predictors, n.s. = not significant, “*” indicates the interaction between ambient temperature and group dummy.

Day	Predictor	β	SE	t	p	95% CI	Effect
4 June2020	Ambient temperature	0.92	0.01	70.29	p < 0.001	0.89–0.94	+
4 June2020	Group dummy	0.11	0.11	0.2	n.s.	−0.98–1.20
4 June2020	Ambient temperature*Group dummy	<0.001	0.02	−0.04	n.s.	−0.05–0.05
	Model			F(3, 113) = 2378		p < 0.001	R2= 0.98
5 June 2020	Ambient temperature	0.96	0.03	29.09	p < 0.001	0.9–1.03	+
5 June 2020	Group dummy	0.21	1.41	0.15	n.s.	2.99–−2.57
5 June 2020	Ambient temperature*Group dummy	<0.001	0.06	0.02	n.s.	−0.12–0.12
	Model			F(3, 126) = 408		p < 0.001	R2= 0.91
6 June 2020	Ambient temperature	0.84	0.02	42.01	p < 0.001	0.8–0.88	+
6 June 2020	Group dummy	1.21	0.85	1.42	n.s.	−0.48–2.89
6 June 2020	Ambient temperature*Group dummy	−0.05	0.04	−1.45	n.s.	−0.12–0.02
	Model			F(3, 100) = 818		p < 0.001	R2= 0.96
12 June 2020	Ambient temperature	0.76	0.03	25.8	p < 0.001	0.7–0.82	+
12 June 2020	Group dummy	0.22	1.23	0.18	n.s.	−2.21–2.66
12 June 2020	Ambient temperature*Group dummy	−0.02	0.05	−0.29	n.s.	−0.12–0.09
	Model			F(3, 126) = 317		p < 0.001	R2= 0.88

). These results demonstrate that the predictors reliably influence the dependent variable, with the model capturing nearly all variability in the response.

An increase of 1 °C in ambient temperature led to an average increase of 0.87 °C in the body temperature of the newts across the four experimental days (β = 0.92, 0.96, 0.84, 0.76), while other environmental conditions were kept constant. The effect of the group (experimental groups T. dobrogicus: n = 9 and T. ivanbureschi: n = 4) was not statistically significant, nor was the interaction between group and ambient temperature (p > 0.05). This indicates that animals from both groups responded similarly to changes in ambient temperature. Although the unbalanced experimental design may have caused minor deviations, these did not substantially affect the main conclusions.

Body temperatures of the animals during the first three days (D1–D3) and on day 4 were analyzed using a one-way ANOVA with “day” as a factor. The ANOVA revealed a statistically significant difference among the days (F = 78.02, p < 0.001). To determine which days differed from day 4, manual t-tests were performed between day 4 and each of the first three days, with a Bonferroni correction applied for multiple comparisons (α = 0.0167). All comparisons were highly statistically significant. Day 1 vs. day 4: t = 13.33, p < 0.001; day 2 vs. day 4: t = 10.45, p < 0.001; day 3 vs. day 4: t = 12.27, p < 0.001 (

Table 3. Mean daily body temperatures of studied animals by group: animal ID; group—group affiliation; D1—4 June 2020; D2—5 June 2020; D3—6 June 2020; D4—12 June 2020; mean 1_3—average body temperature for 1–3 days; Δ—day 4—mean of 1–3 days. Notes: TD—T. dobrogicus (1–9); TI—T. ivanbureschi (1–4).

Animal ID	Group	D1	D2	D3	D4	Mean 1_3	Δ
TD1	A	21.28	21.4	21.29	20.46	21.32	−0.86
TD2	A	21.22	21	21.00	20.26	21.06	−0.80
TD3	A	21.16	21	21.20	20.81	21.13	−0.32
TD4	A	21.07	21.1	20.90	20.42	21.02	−0.60
TD5	A	21.01	21	21.18	20.27	21.08	−0.81
TD6	A	21.08	21.1	21.29	20.35	21.15	−0.80
TD7	A	21.10	20.9	21.26	20.63	21.08	−0.45
TD8	A	21.07	21.2	21.26	20.47	21.16	−0.69
TD9	A	21.22	21.3	21.40	20.66	21.29	−0.63
TI1	B	21.17	21.1	21.31	20.27	21.21	−0.94
TI2	B	21.21	21.3	21.21	20.45	21.25	−0.80
TI3	B	21.26	21.4	21.11	20.45	21.26	−0.81
TI4	B	21.26	21.5	21.10	20.24	21.27	−1.03

).

The mean change Δ (day 4—mean of days 1–3) indicates a systematic decrease in body temperature. These results demonstrate that the drop on day 4 was statistically significant and consistent across both groups of animals (

Table 4. Mean change (Δ) and one-sample t-test results by group. Notes: group—experimental group of animals; n—number of animals in the group; mean (Δ)—mean change (Δ = day 4—mean of 1–3 days); SD (Δ)—standard deviation of Δ; t—t statistic from one-sample t-test; p-value—significance level (two-tailed test against H₀: Δ = 0).

Group	n	mean Δ (℃)	SD Δ	t	p
T. dobrogicus	9	−0.66	0.18	−10.99	p < 0.001
T. ivanbureschi	4	−0.90	0.11	−16.35	p < 0.001

).

Both newt species exhibited lower mean body temperatures on day 4 of the experiment, which was conducted six days after the initial three consecutive days. In group T. dobrogicus, the mean difference compared to the average body temperature of days 1–3 was –0.66 °C, while in group T. ivanbureschi it was –0.9 °C.

The comparison between the two methods of measuring body temperature indicated that the results obtained using both systems were rather close (Figure 2).

4. Discussion

4.1. Physiological Remarks

Thermoregulation in ectothermic species under terrestrial conditions has long challenged researchers [14,23,24,25]. In aquatic newts and salamanders, studies have predominantly focused on water-related mechanisms, whereas their terrestrial strategies remain largely unexplored [26,27,28]. The limited knowledge of habitat preferences and the secretive lifestyle of the two studied newt species further contribute to this gap [20,29]. In the context of climate change and rising air temperatures, advancing our understanding of terrestrial thermoregulation is particularly important for small-sized ectotherms [30,31].

Our investigation focused on two newt species, tracking their body temperature under systematically increasing ambient temperatures in a laboratory simulation. The results demonstrated a rapid rise in body temperature following the increase in air temperature, with both species maintaining consistently lower values than the surrounding environment. However, after approximately 120 min of exposure to warm air, the newts exhibited clear signs of stress and fatigue, prompting termination of the trial to prevent harm. Two significant symptoms were observed: the skin of the newts became rather dry at the end of the sessions, and the animals remained almost motionless after stimulation. Despite the limited sample size, this passive behavioral response was consistently detected in all individuals from both species. To establish the critical temperature threshold at which newts exhibit terrestrial “thermoconformity”, further studies are required, ideally under conditions closely reflecting their natural habitats.

Besides ethological mechanisms, the thermoregulation in small-sized salamanders is related mainly to water loss [16,18,32]. The water loss represents a threatening factor for amphibians and may be even more harmful than temperature increase [33,34]. The animals used in our experiments were not adapted to enter their annual terrestrial phase and showed low potential to retain water—after about two hours they showed severe signs of dehydration. In the wild, both investigated species have different ecologies, as T. ivanbureschi is predominantly terrestrial and T. dobrogicus is more related to water [19,20,21,35].

According to our results, both species demonstrated vulnerability to sudden exposure to increased air temperatures. Despite lack of direct comparison of physiological parameters in this study, the results of the reaction tests in both species indicated that after about two hours the animals were not willing for locomotor activities.

The findings of this study indicate a marginal decline in newt body temperature on the fourth day, after a six-day interval from the initial measurements. The present design does not permit conclusions regarding early terrestrial adaptation. In order to confirm this trend, further studies with larger samples, more repetitions, and longer experimental periods are required.

4.2. Methodological Considerations

We used two approaches to obtain thermal information from the investigated newt specimens. Both used temperature measurement techniques demonstrated that they are reliable and deliver comparable results. The used thermocouple system demanded limited contact with the newt’s skin. This way the need for catching and handling the animals was avoided. We needed about 2 s to obtain results from the thermometer and virtually no recovery time between successive measurements. However, in field surveys this technique may be used with some limitations, because it requires direct touching of the investigated animal.

Our comparison under laboratory conditions demonstrated that the use of IR thermal imaging gear provided reliable results. The working parameters of the camera have to be adjusted strictly in accordance with the physical properties of the amphibian’s skin, see [22,36]. In their investigation on thermal ecology in frogs, Ref. [37] successfully used only distant IRT imaging by avoiding disturbance of the investigated anurans. Luna and Font [38] tested a combined approach based on the use of thermocouples and an IR thermal camera and stated that the infrared technology delivers reliable results without inflicting stress when used on small-bodied herpetofauna. In our case, the temperature of the entire dorsal surface of the newts was evenly distributed, and no remarkable thermal gradient was detected—contrary to data from some anurans [36,39]. With the precise fine-tuning of the device (for an overview see [37]), the IR measuring systems deliver completely reliable data (at least at temperature intervals from positive 15 to 30 °C). The direct comparison of the measurements, obtained with the use of thermocouple and thermal imaging, demonstrated that the differences were within the acceptable tolerance for biological research. For small newts, we recommend the use of IR thermal systems, which provide data in the form of thermal profiles. The use of IR-based “point” thermometers may be inappropriate because of the small size and the slender form of the newt’s body [40].