Effect of acute low body temperature on predatory behavior and prey-capture efficiency in a plethodontid salamander

    Effect of acute low body temperature on predatory behavior and prey-capture efficiency in a plethodontid salamander Glenn A. Marvin, Kayla Davis, Jacob Dawson PII: DOI: Reference:

S0031-9384(16)30082-8 doi: 10.1016/j.physbeh.2016.02.038 PHB 11233

To appear in:

Physiology & Behavior

Received date: Revised date: Accepted date:

7 January 2016 25 February 2016 26 February 2016

Please cite this article as: Marvin Glenn A., Davis Kayla, Dawson Jacob, Effect of acute low body temperature on predatory behavior and prey-capture efficiency in a plethodontid salamander, Physiology & Behavior (2016), doi: 10.1016/j.physbeh.2016.02.038

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT Marvin, Davis & Dawson 1

Effect of acute low body temperature on predatory behavior and prey-capture

PT

efficiency in a plethodontid salamander

SC

RI

Glenn A. Marvin*, Kayla Davis, and Jacob Dawson

MA

Florence, Alabama 35632-0002, USA

*Corresponding author.

AC CE P

TE

D

E-mail address: [email protected] Tel.: +12567654348

NU

Department of Biology, University of North Alabama, Box 5048, 1 Harrison Plaza,

ACCEPTED MANUSCRIPT Marvin, Davis & Dawson 2

Abstract. The low-temperature limit for feeding in some salamander species

PT

(Desmognathus, Plethodontidae) has been inferred from field studies of seasonal variation in salamander activity and gut contents, which could not determine whether

RI

feeding is more dependent on environmental conditions influencing salamander

SC

foraging behavior or prey availability and movement. We performed two controlled

NU

laboratory experiments to examine the effect of short-term (acute) low body temperature on predatory behavior and prey-capture efficiency in a semiaquatic

MA

plethodontid salamander (D. conanti). In the first experiment, we quantified variation in the feeding responses of cold salamanders (at 1, 3, 5 and 7°C) to a video recording

TE

D

of a walking, warm (15°C) cricket to determine the lower thermal limit for predatory behavior, independent of any temperature effect on movement of prey. Experimental-

AC CE P

group salamanders exhibited vigorous feeding responses at 5 and 7°C, large variation in feeding responses both among and within individuals (over time) at 3°C, and little to no feeding response at 1°C. Feeding responses at both 1 and 3°C were significantly less than at each higher temperature, whereas responses of control-group individuals at 15°C did not vary over time. In the second experiment, we quantified feeding by cold salamanders (at 3, 5, 7 and 11°C) on live, warm crickets to examine thermal effects on prey-capture ability. The mean feeding response to live crickets was significantly less at 3°C than at higher temperatures; however, 50% of salamanders captured and ingested prey with high efficiency at this temperature. We conclude that many individuals stalk and capture prey at very low temperatures (down to 3°C). Our results support a

ACCEPTED MANUSCRIPT Marvin, Davis & Dawson 3

growing body of data that indicate many plethodontid salamanders feed at

PT

temperatures only a few degrees above freezing.

SC

RI

Keywords: feed, forage, salamander, temperature, thermal, predatory behavior

NU

1. Introduction

Physiology and behavior of ectothermic animals can be profoundly affected by

MA

variation in environmental temperature. For example, the acquisition of energy in many species is strongly influenced by changes in body temperature (Tb). Thermal

TE

D

effects on movement, foraging behavior, prey capture, digestion, and absorption can alter food consumption and energy-assimilation rates. Consequently, the relationship

AC CE P

between Tb and performance for such physiological and behavioral processes has been investigated in a variety of species (e.g., [3,7,22,52,54,55]). Individuals of some species can diminish thermal effects on energy acquisition by altering behavior and/or physiology to maintain Tb within a range that is narrower than that of the environment. For example, many insects and reptiles utilize spatial and temporal heterogeneity in the thermal environment to optimize Tb and enhance foraging success and food-assimilation efficiency (e.g., [6,7,22,24,25]). Some amphibians behaviorally regulate Tb in adequately moist environments where temperature varies sufficiently among microhabitats [9,13,20,26,36,49]. However, regulation of Tb is not possible in many populations or species because the

ACCEPTED MANUSCRIPT Marvin, Davis & Dawson 4

environment does not have sufficient thermal diversity to allow thermoregulation [42]

PT

or moisture restrictions prevent individuals from choosing among thermally-diverse microhabitats [17,18,50]. Under these conditions, individual Tb often closely follows

RI

changes in environmental temperature, is negatively correlated with elevation, and

SC

experiences predictable seasonal variation [19]. Thus, the ability to minimize the effect

NU

of low temperature on energy acquisition would provide a selective advantage for individuals. This may partly explain why (1) the lower thermal limit for feeding is much

MA

less in many amphibians than in other ectothermic vertebrates and (2) the lowtemperature limits for feeding in plethodontid salamanders (family Plethodontidae) in

TE

D

temperate zone environments are similar to those of some invertebrate animals in polar environments (Table 1).

AC CE P

Various adaptations may allow plethodontid salamanders to have a relatively low thermal limit for feeding. Whereas some physiological rates are greatly diminished at low temperatures (e.g., Q10 > 2 for metabolism [57], Q10 > 3 for regeneration [39]), a relatively low thermal sensitivity for locomotor ability, and some capacity for thermal acclimation of locomotor performance, may allow individuals of some species to maintain activity at low environmental temperatures [16,37,38]. Although low temperature reduces performance of muscles involved in tongue movement, the quick release of energy stored in elastic tissue allows ballistic tongue projection for prey capture below 5°C in some plethodontid salamanders [2,15]. While foraging success and food-assimilation efficiency for many ectothermic animals (e.g., insects, lizards,

ACCEPTED MANUSCRIPT Marvin, Davis & Dawson 5

and snakes) are directly related to Tb [7,25,34], assimilation efficiency in some species

PT

of plethodontid salamanders is inversely related to temperature and can be ≥ 90% at Tb ≤ 10oC [10,27]. However, it is not known whether the ability to maintain energy

RI

acquisition at very low temperatures is prevalent among salamander populations or

SC

species that inhabit diverse thermal environments (e.g., at different elevations or

NU

latitudes) or those that exhibit disparity in temperature preference in the same environment [13,49].

MA

Variation in activity and gut contents for individuals in the field at different seasons and temperatures has been used to infer the low-temperature limit for

TE

D

feeding in some species of desmognathan plethodontid salamanders (i.e., Dusky Salamanders, genus Desmognathus) [4,29,47]. Dusky Salamanders are ambush,

AC CE P

generalist predators that primarily eat a variety of small, semiaquatic and terrestrial invertebrate animals [41]. Based on our observations, a typical predatory behavior pattern consists of visual orientation (by turning the head) toward moving prey, very slow locomotion toward the prey (i.e., stalking of the prey) until it is within capture distance (approximately 1-4 cm), and then capture of the prey with the tongue and jaws. Often during stalking, the salamander lunges (i.e., jumps [45]) toward the prey immediately before capture to quickly traverse the remaining distance. In the field, the effect of low temperature on Dusky Salamander foraging could be attributable to a combination of thermal effects on predator walking and jumping performance, predator tongue and jaw kinematics, prey activity, prey locomotor

ACCEPTED MANUSCRIPT Marvin, Davis & Dawson 6

performance, and prey availability (including possible variation in types and densities

PT

of various prey species among seasons). The low-temperature limit for feeding could also be influenced by both short-term (acute) and long-term thermal effects (i.e., via

RI

seasonal acclimatization in the field) on physiology and behavior of predator and prey.

SC

Because field studies could not determine whether salamander feeding is more

NU

dependent on thermal conditions influencing foraging behavior or prey availability and

controlled laboratory conditions.

MA

movement, we examined the effect of low Tb on salamander feeding behavior under

We performed two experiments with several low temperatures to assess acute

TE

D

thermal effects on predatory behavior and prey-capture ability in Spotted Dusky Salamanders (D. conanti), which inhabit seeps and streams of the south-central United

AC CE P

States [11,44,53]. Prior to experiments, we established that the predatory behavior exhibited by salamanders to either a live, walking cricket or a video recording of a walking cricket was the same. In the first experiment, we quantified the feeding responses by cold salamanders to a video recording of a walking, warm cricket to determine the (1) effect of acute low temperature on predatory behavior (independent of any thermal effect on prey movement), (2) low-temperature limit for feeding behavior, (3) degree of variability among individuals for feeding behavior near the low-temperature limit, and (4) temporal variability for feeding responses by individuals near the low-temperature limit. In the second experiment, we quantified the feeding responses by cold salamanders to live, warm crickets and observed prey

ACCEPTED MANUSCRIPT Marvin, Davis & Dawson 7

capture to determine whether (1) salamanders can capture prey near the low-

PT

temperature limit for feeding behavior, (2) prey-capture efficiency decreases near the low-temperature limit, and (3) prey-capture time increases near the low-temperature

RI

limit. We also used these data to evaluate whether the low-temperature limit for

SC

feeding in this species is similar to that proposed for other salamander species (e.g.,

NU

species of Desmognathus with geographic distributions that extend to higher

MA

latitudes).

2. Materials and methods

TE

D

2.1 Collection, measurement, and care of salamanders Individuals of D. conanti were collected during March of 2014 (n = 30 for

AC CE P

Experiment 1) and 2015 (n = 32 for Experiment 2) from Colbert County in northwestern Alabama (34o47.63′N, 087o37.85′W). After completion of experiments, animals were released at the collection locality. Body length (i.e., standard length measured from tip of snout to the posterior angle of the vent), body mass, and tail length were measured for each individual. Based on body size at sexual maturity in closely related species (35–40 mm in body length [28]), individuals were likely sexually mature. Sex of each salamander was inferred from sexual dimorphism in body size, tail width, and head width for Desmognathus species [8,12]. Individuals were paired (i.e., size-matched) based on similarity in body length, and then each individual from a sizematched pair was randomly assigned to either an experimental group or a control

ACCEPTED MANUSCRIPT Marvin, Davis & Dawson 8

group. To ensure that salamanders in the experimental group and control group were

PT

equivalent in sizes, we performed a t-test to compare the sizes (body lengths, tail lengths, and body masses) of individuals in the two groups for each experiment (Table

RI

2).

SC

Prior to each experiment, all individuals were acclimated to constant

NU

temperature (15oC) and photoperiod (LD 12:12 h) for four weeks in an environmental chamber. During each experiment, all salamanders were maintained at the LD 12:12 h

MA

photoperiod. Three 25-W fluorescent bulbs (Philips F25T8/TL741) provided light in each environmental chamber. Salamanders were housed individually in 14-cm long by

TE

D

14-cm wide by 5.5-cm high translucent, glass food-storage containers (Ziploc® brand). The bottom of each housing container was lined with several flattened layers of moist

AC CE P

paper towels (Experiment 1) or two flattened layers of moist filter paper (Experiment 2). To ensure that each individual could see the prey throughout a feeding trial, we did not allow sufficient space under or between the layers of paper for either the salamander or the live cricket (in Experiment 2) to hide. To ensure that the Tb of each experimental-group individual during a feeding trial was equivalent to the test temperature, salamanders were kept at the test temperature for 24 h prior to observation of feeding behavior. Three feeding trials were conducted for each individual at each test temperature. The second and third trials for each individual were conducted one week after the preceding trial. Except for the 24-h temperature-adjustment period prior to a trial, salamanders were maintained

ACCEPTED MANUSCRIPT Marvin, Davis & Dawson 9

at 15oC between trials. We used the same protocol for control-group individuals in

PT

each experiment, except that each trial was conducted with salamander Tb at 15oC. Each individual was given the opportunity to eat a cricket once a week during

RI

acclimation and the experiments. During Experiment 1, each individual was not fed for

SC

a week prior to a feeding trial but was given the opportunity to feed within an hour

NU

after a trial was completed. If an individual fed during a trial in Experiment 2, then it was not given the opportunity to feed again until the next trial. If an individual did not

MA

feed during a trial, then it was given an opportunity to feed within 24 h (at Tb of 15oC).

TE

D

2.2 Experiment 1: Effect of acute low body temperature on salamander feeding response to a video recording of walking, warm prey

AC CE P

After acclimation, the feeding responses of all experimental-group individuals were recorded sequentially at four test temperatures (i.e., acute body temperature of 7, 5, 3 or 1oC) in an environmental chamber at the same time period during photophase. To examine feeding behavior in a trial, each individual was presented with a video of a walking, warm (15oC) cricket which was previously recorded on an iPhone®. By using this video recording as the prey visual stimulus in each feeding trial for all salamanders, we kept the stimulus exactly the same for each trial. This technique enabled use to determine the effect of temperature on feeding behavior independent of any potential temperature effect on prey movement. In the video recording, the apparent body length of the cricket on the iPhone® screen was 8 mm

ACCEPTED MANUSCRIPT Marvin, Davis & Dawson 10

and its apparent rate of movement (i.e., walking without jumping) varied from about

PT

1.0 to 2.5 cm/sec. For a feeding trial, the iPhone® screen was placed against the outside of the salamander’s housing container so that the screen was approximately

RI

parallel to the direction that the salamander was facing. This screen orientation

SC

allowed us to determine when a salamander first noticed the prey movement because

NU

the salamander would turn its head to view the cricket. From behind a blind, we observed salamander behavior through a window in the environmental chamber.

MA

During a 5-min trial, the feeding behavior (response) of an individual was quantified as “0” if the salamander showed no orientation response for feeding (i.e.,

TE

D

did not turn its head to look at the cricket), “1” if the salamander turned its head to look at the cricket, “2” if the salamander partially approached the cricket (i.e., walked

AC CE P

toward but did not encounter the side of the test container), “3” if the salamander climbed the side of its housing container in an apparent attempt to encounter the cricket, or “4” if the salamander attempted to capture the cricket by lunging (i.e., jumping) and/or snapping its jaws against the side of its housing container. Except for one experimental-group individual, all salamanders in both groups exhibited a strong (“4”) feeding response during multiple trials in the experiment. The highest feeding response for this experimental-group individual was “2” (in only one of the 12 trials). Because the lack of feeding behavior over many weeks could indicate a health problem, data from that individual were not included in statistical analyses.

ACCEPTED MANUSCRIPT Marvin, Davis & Dawson 11

2.3 Experiment 2: Effect on acute low body temperature on salamander feeding

PT

behavior, prey-capture efficiency, and prey-capture time during feeding trials with live, warm prey

RI

After acclimation, we examined the feeding behavior, prey-capture efficiency,

SC

and prey-capture time for all experimental-group individuals at four test temperatures

NU

(i.e., acute body temperature of 11, 7, 5 or 3oC) in an environmental chamber at the same time period during photophase. In a feeding trial, a live cricket was introduced

MA

into the housing container of a salamander. Because the mobility of crickets may change with temperature, we used a warm cricket with Tb of 15oC in each trial to

TE

D

ensure that the prey was mobile enough to stimulate a feeding response by the salamander and that initial cricket mobility (i.e., at the beginning of each trial) would

AC CE P

not vary among test temperatures. To minimize the possible effect of prey size on salamander feeding behavior and prey-capture success, the body size of the cricket in each feeding trial corresponded to the body size of the salamander (i.e., cricket body length was approximately 15 to 20% of the salamander’s body length in each trial). Mean body mass of crickets was about 0.055 g. To minimize disturbance of the salamander during introduction of the cricket into the salamander’s housing container, a modified 12-ml syringe (i.e., with distal end cut off) was used to “inject” (introduce) the cricket through a 20-mm diameter hole centered in the top of the housing container. From behind a blind, we observed salamander behavior through a window in the environmental chamber.

ACCEPTED MANUSCRIPT Marvin, Davis & Dawson 12

During a 5-min trial, the feeding response of the salamander was quantified in

PT

the same manner as described for Experiment 1. However, because the prey was inside the test container for this experiment, the response was recorded as “2” if the

RI

salamander partially approached (i.e., walked toward but did not encounter) the

SC

cricket or “3” if the salamander continued to approach the cricket until it was close

NU

enough (i.e., within about 2 cm) to attempt prey capture. Prey-capture success (i.e., whether or not the cricket was captured and ingested) and prey-capture time (i.e.,

MA

time elapsed from the start of a trial until capture of prey) were also recorded for each trial. At each test temperature, we quantified the prey-capture efficiency for an

TE

D

individual as the mean prey-capture success during trials in which the salamander attempted to capture the cricket (e.g., if a salamander attempted to capture the

AC CE P

cricket in two of three trials but only captured one cricket, then mean capture success was 0.5 for the individual). Mean prey-capture time was also calculated for each individual.

2.4 Statistical analyses Data were transformed when necessary to meet assumptions of parametric tests. When assumptions were not met with data transformation, non-parametric tests were used. To compare the feeding behavior for experimental-group individuals among temperatures in each experiment, we performed a repeated-measures analysis of variance (ANOVA) on the mean of the two highest feeding responses (among the

ACCEPTED MANUSCRIPT Marvin, Davis & Dawson 13

three trials) for each individual at each test temperature. For control-group individuals

PT

in each experiment, a Friedman repeated-measures ANOVA on ranks was performed to determine whether feeding responses of salamanders changed during the

RI

experimental period irrespective of temperature variation.

SC

To examine the effect of temperature on individual feeding variability for the

NU

experimental-group salamanders in Experiment 1, we used repeated-measures ANOVA to compare among test temperatures the coefficient of variation for individual feeding

MA

response at each temperature (i.e., the standard deviation of the responses for an individual during three trials at the temperature divided by the mean of the

TE

D

responses). For control-group salamanders in Experiment 1, we used a Friedman repeated-measures ANOVA of the MADM (median absolute deviation from

AC CE P

median)/median (i.e., the non-parametric alternative to the coefficient of variation) for feeding responses to compare individual feeding variability over time. Because the experimental-group salamanders in each experiment exhibited significantly greater variability in feeding responses at 3oC, we analyzed data from this test temperature to determine if such variation was related to individual differences in sex, body size, or body condition. To examine whether feeding behavior differed between sexes, we compared the maximum feeding responses for males and females with a Mann-Whitney rank-sum test. To examine whether feeding behavior varied with individual body size, we performed Spearman rank-order correlations to examine the relationship between body sizes (i.e., body lengths or log-transformed body

ACCEPTED MANUSCRIPT Marvin, Davis & Dawson 14

masses) and the maximum feeding responses of individuals. To examine whether

PT

feeding behavior varied with individual body condition, we performed a Spearman rank-order correlation to examine the relationship between body conditions (i.e., the

RI

residuals from the linear regression of body mass on body length) and the maximum

SC

feeding responses of individuals.

NU

To examine variation in prey-capture efficiency in Experiment 2, we performed a Friedman repeated-measures ANOVA on ranks for both the experimental-group

MA

individuals and the control-group individuals. To examine variation in prey-capture times for both groups in Experiment 2, we performed a two-way repeated-measures

TE

AC CE P

3. Results

D

ANOVA on ranked data.

3.1 Experiment 1: Effect of acute low body temperature on salamander feeding response to a video recording of walking, warm prey For the experimental-group individuals, feeding responses were significantly different among the four test temperatures (repeated-measures ANOVA, F = 51.81, p < 0.001, df = 13 for subject, 3 for temperature, 39 for residual, 55 total). Feeding responses at 1°C were significantly less than feeding responses at 3, 5 or 7°C (HolmSidak multiple comparison method; t = 4.69, 9.79, and 11.13, respectively; p < 0.001 for each comparison; Fig. 1). Feeding responses at 3°C were significantly less than feeding responses at 5 or 7°C (Holm-Sidak multiple comparison method; t = 5.10 and

ACCEPTED MANUSCRIPT Marvin, Davis & Dawson 15

6.44, respectively; p < 0.001 for each comparison; Fig. 1). For control-group individuals

PT

tested four times at 15°C (i.e., at the same times when the experimental-group individuals were tested at 1, 3, 5 and 7°C), feeding responses did not significantly vary

RI

over time (Friedman repeated-measures ANOVA on ranks; χ2 = 1.82, p = 0.610, df = 3;

SC

Fig. 1).

NU

For the experimental-group individuals, feeding variability was significantly different among the four test temperatures (repeated-measures ANOVA, F = 5.05, p =

MA

0.005, df = 13 for subject, 3 for temperature, 39 for residual, 55 total). Feeding variability at 3°C was significantly greater than feeding variability at 1, 5 or 7°C (Holm-

TE

D

Sidak multiple comparison method; t = 2.97, 2.66, and 3.60, respectively; p = 0.025, 0.044, and 0.005, respectively; Fig. 2). For control-group individuals, feeding variability

AC CE P

did not significantly change over time (Friedman repeated-measures ANOVA, χ2 = 1.857, p = 0.603, df = 3).

For the experimental group tested at 3°C, there was much variation among individuals in their maximum feeding responses. Fifty percent of individuals exhibited a “3” or “4” feeding response in one or more of the three trials while the other individuals showed a “1” or “2” feeding response. This variation could not be explained by individual differences in sex, body size, or body condition. Maximum feeding responses did not significantly differ between males and females (Mann-Whitney rank sum test, U = 18.5, T = 58.5, p = 0.456, n = 7 for both sexes). There was no significant relationship between the maximum feeding responses of individuals and their log-

ACCEPTED MANUSCRIPT Marvin, Davis & Dawson 16

transformed body masses (Spearman rank-order correlation, rS = 0.033, p = 0.904, n =

PT

14), body lengths (rS = –0.035, p = 0.892, n = 14), or body conditions (rS = –0.411, p =

RI

0.138, n = 14).

SC

3.2 Experiment 2: Effect of acute low body temperature on salamander feeding

NU

behavior, prey-capture efficiency, and prey-capture time during feeding trials with live, warm prey

MA

For the experimental-group individuals, feeding responses were significantly different among the four test temperatures (repeated-measures ANOVA, F = 6.81, p <

TE

D

0.001, df = 15 for subject, 3 for temperature, 45 for residual, 63 total). Feeding responses at 3°C were significantly less than feeding responses at 5, 7, or 11°C (Holm-

AC CE P

Sidak multiple comparison method; t = 2.75 and p = 0.028 for 5°C; t = 3.92 and p < 0.001 for 7°C; t = 5.09 and p < 0.001 for 11°C; Fig. 3). For control-group individuals tested four times at 15°C (i.e., at the same times when the experimental-group individuals were tested at 3, 5, 7, and 11°C), feeding responses did not significantly vary over time (Friedman repeated-measures ANOVA on ranks; χ2 = 1.98, p = 0.577, df = 3; Fig. 3). For the experimental-group individuals, prey-capture efficiency was not significantly different among the four test temperatures (Friedman repeated-measures ANOVA on ranks; χ2 = 2.00, p = 0.572, df = 3). For experimental-group individuals at all temperatures, prey-capture efficiency was 0.90 ± 0.23 SD (median = 1, range = 0–1,

ACCEPTED MANUSCRIPT Marvin, Davis & Dawson 17

25th percentile = 1, 75th percentile = 1, n = 64). For control-group individuals, prey-

PT

capture efficiency did not significantly vary over time (Friedman repeated-measures ANOVA on ranks; χ2 = 1.00, p = 0.801, df = 3). For control-group individuals at all times,

RI

prey-capture efficiency was 0.94 ± 0.18 SD (median = 1, range = 0–1, 25th percentile =

SC

1, 75th percentile = 1, n = 64).

NU

Prey-capture times did not significantly vary among test temperatures (twoway ANOVA on ranks; F = 0.43, p = 0.733, df = 1 for group, 30 for subject, 3 for

MA

temperature, 3 for group × temperature, 67 for residual, 104 total) but were significantly greater for experimental-group individuals than for control-group

TE

D

individuals (F = 15.04, p < 0.001) with no significant interaction between group and temperature (F = 0.737, p = 0.396; Fig. 4). Prey-capture times were significantly greater

AC CE P

for experimental-group individuals than for control-group individuals at 3 or 5°C (Holm-Sidak multiple comparison method; t = 3.78 and p < 0.001, t = 2.95 and p = 0.004, respectively) but not at 7 or 11°C (t = 0.721 and p =0.473, t = 1.65 and p = 0.102, respectively; Fig. 4).

For the experimental group tested at 3°C, there was much variation among individuals in their maximum feeding responses. Fifty percent of individuals exhibited a “4” feeding response in one or more of the three trials while the other individuals showed a “0” or “1” response. For all experimental-group individuals tested at 3°C from both experiments (n = 30), the variation in maximum feeding responses could not be explained by individual differences in sex, body size, or body condition. Maximum

ACCEPTED MANUSCRIPT Marvin, Davis & Dawson 18

feeding responses did not significantly differ between males and females (Mann-

PT

Whitney rank sum test, U = 109.0, T = 222.0, p = 0.911, n = 14 for males, n = 16 for females). There was no significant relationship between the maximum feeding

RI

responses of individuals and their log-transformed body masses (Spearman rank-order

SC

correlation, rS = –0.263, p = 0.159, n = 30), body lengths (rS = –0.143, p = 0.448, n = 30),

NU

or body conditions (rS = 0.076, p = 0.686, n = 30).

MA

4. Discussion

Previous research on the thermal limits for salamander foraging examined the

TE

D

effect of low temperature on tongue kinematics during feeding [2,15] or variation in salamander activity and gut contents in the field among seasons to infer changes in

AC CE P

feeding [4,29]. In contrast, we assessed how the low-temperature limit for salamander foraging is influenced by acute thermal effects on predatory behavior and preycapture efficiency. Our results show that the acute low-temperature limit (range) for predatory behavior, independent of any thermal effect on prey movement, is between 1 and 5oC for individuals of D. conanti in northwestern Alabama. Feeding responses by individuals at Tb of 1 or 3oC to a video of a walking, warm cricket were significantly less than at 5 or 7oC (Fig. 1). No individuals exhibited a strong feeding response at 1oC, but many individuals (50%) showed a strong feeding response at 3 oC (Fig. 1). Variation among individuals in their maximum feeding responses at 3oC was not significantly related to sex, body size, or body condition. However, additional data

ACCEPTED MANUSCRIPT Marvin, Davis & Dawson 19

are needed to determine whether individual variability in feeding at low temperature

PT

is influenced by sexual maturity, sex, body size, body condition, or heritable variation in behavior and physiology of individuals. Similarly, individual variation in the low-

RI

temperature limit for feeding occurs in some reptiles [7]. In addition to variation

SC

among individuals of D. conanti for predatory behavior at low Tb (3oC), behavior of

NU

individuals at low temperature also varied over time. Individuals would often exhibit a strong feeding response in one or two trials but no feeding response in the other

MA

trial(s) at low temperature. Thus, variability in feeding behavior at 3oC was significantly greater than at either lower or higher temperature (Fig. 2). Variation in the feeding

TE

D

behavior of an individual over time might be partly explained by an increase in the duration of satiety when feeding at low temperature. Because metabolic rate in

AC CE P

salamanders decreases significantly at low temperature (with Q10 > 2) [57], the frequency of feeding may change correspondingly if digestion efficiency is maintained. Also, variation among individuals in predatory behavior at low temperature may be associated with individual variation in metabolic rate [33]. However, further research is needed to determine whether the frequency of feeding varies with metabolic rate in salamanders. Results from feeding trials with live crickets (Fig. 3) demonstrate that many salamanders (50%) exhibit typical predatory behavior at 3oC and capture prey with high efficiency (i.e., about 90% success rate). Thus, results from both experiments indicate that the low-temperature limit for feeding in this species (i.e., “Group B”

ACCEPTED MANUSCRIPT Marvin, Davis & Dawson 20

lineage of D. conanti [11]), which has a relatively lower-latitude geographic distribution

PT

(with northern limit about 37oN) is very similar to that proposed for closely related species with relatively higher-latitude distributions (with northern limits about 46oN).

RI

The relationship between temperature, salamander activity, and gut contents in the

SC

field has been studied for individuals of D. fuscus and D. ochrophaeus from localities

NU

near the northwestern limits of the species’ respective geographic ranges in Ohio. Individuals of D. fuscus (from “Group A” lineage [11]) begin to move into subterranean

MA

(sub-surface), winter retreats in Ohio when stream temperatures drop below 7oC, but because subterranean retreats remain between 1 and 6oC when stream temperatures

TE

D

drop to 0–4oC, salamanders may continue to feed in subterranean retreats throughout the winter [4]. Captive individuals of D. ochrophaeus from northeastern Ohio cease to

AC CE P

feed at 5oC [21]. Similarly, Keen [29] concluded that individuals of D. ochrophaeus near the species’ northwestern range limit in Ohio are less active and ingest fewer prey items at 0 to 5oC.

Field studies that inferred the low-temperature limit for salamander feeding could not determine whether feeding is more dependent on environmental conditions influencing salamander foraging behavior or prey availability and movement since such conditions could affect both salamanders and their prey. Our results indicate that the thermal effect on predatory behavior determines the low-temperature limit for feeding in Spotted Dusky Salamanders regardless of any thermal effect on prey availability and movement. However, although many individuals continue to feed

ACCEPTED MANUSCRIPT Marvin, Davis & Dawson 21

below 5oC, digestive processes and food-assimilation efficiency may not be maintained

PT

at such low temperatures. While an ability to maintain an assimilation efficiency ≥ 90% at temperatures ≤ 10oC has been demonstrated for some plethodontid salamanders in

RI

the genus Plethodon [10,27], the thermal effect on digestion for Desmognathus below

SC

15oC is not known [21].

NU

If digestive processes and food passage rates are more thermally sensitive than predatory behavior, then our results may not be indicative of Dusky Salamander

MA

feeding behavior during winter. Our experiments more closely mimicked thermal conditions during spring or fall (i.e., overall moderate temperatures with short-term

TE

D

fluctuations to low temperatures) than thermal conditions during winter (with prolonged periods of low temperatures). Salamanders may feed during short-term

AC CE P

exposure to low temperatures when overall long-term, moderate temperatures allow digestion of the meal. In contrast, feeding during a prolonged exposure to low temperatures may be harmful or lethal if digestive processes and food passage rates are greatly reduced [24,46].

A low thermal sensitivity for tongue-projection distance in other plethodontid salamander species (Eurycea and Hydromantes) indicates that their low-temperature limits for feeding are similar to those for species of Desmognathus. Elastic-powered tongue-projection distances in E. guttolineata and H. platycephalus are maintained at low temperatures (down to 4.9 and 2oC, respectively, in the two species) to a much higher degree than muscle-powered tongue retraction, and this ability allows prey

ACCEPTED MANUSCRIPT Marvin, Davis & Dawson 22

capture at low temperatures [2,15]. Although current data suggest that the ability to

PT

feed at very low temperatures is ubiquitous among plethodontid salamander species (Table 1), additional comparisons among populations or species with disparate thermal

RI

ecology (e.g., from different latitudes and elevations) or thermal preferences [13,49],

SC

particularly via studies that utilize similar methods, are needed.

NU

Moreover, comparisons among salamander populations and species for the possible effect of low temperature on prey-capture ability would help to provide a

MA

more complete understanding of thermal effects on salamander foraging. In the present study, low temperature had no effect on prey-capture efficiency for

TE

D

individuals when prey were always within 14 cm of the salamander and the prey could not hide. However, prey-capture times for experimental-group individuals at very low

AC CE P

temperatures (3 and 5oC) were longer than prey-capture times for control-group individuals at 15oC (Fig. 4), which indicates that low temperature might affect preycapture efficiency in the field, if an increase in prey-capture time allows greater opportunity for prey to escape or find shelter. Alternatively, prey-capture time might not increase at low temperatures in the field when the Tb of salamander and prey are equivalent throughout the predatory encounter. In other words, the increase in preycapture time at very low temperatures in the lab may have been caused by the initial 10 to 12oC difference in Tb of the salamander and prey during a feeding trial, if the Tb disparity hindered the ability of the salamander to stalk close enough for prey capture.

ACCEPTED MANUSCRIPT Marvin, Davis & Dawson 23

Our results contribute to a growing body of data that indicate many

PT

plethodontid salamanders, unlike many other ectothermic animals (Table 1), continue to feed and maintain energy consumption at temperatures only a few degrees above

RI

freezing. This ability would be advantageous for individuals in populations or species

SC

that inhabit thermally constrained environments which do not allow individuals to

NU

elevate Tb to enhance foraging and digestive abilities. For example, an ability to feed at low temperatures could allow Fire Salamander larvae (Salamandra salamandra), which

MA

often inhabit environments where temperatures prevent individuals from reaching the preferred Tb range [42], to maintain energy intake for development at low

TE

D

temperatures. For salamander species that inhabit cool, high-elevation environments (e.g., H. platycephalus), thermal independence of prey capture is likely advantageous

AC CE P

for individuals foraging at near-freezing temperatures [15]. The ability to feed and maintain energy intake at low temperature would also be advantageous for individuals that inhabit environments with thermally-diverse microhabitats, but which prevent behavioral regulation of Tb because such behavior incurs excessive costs such as increased evaporative water loss, energy expenditure, predation risk, or competition with other species.

Acknowledgements Scientific collecting permits (license numbers 2014024543868680 and 2015065974868680) were issued by the Alabama Department of Conservation and

ACCEPTED MANUSCRIPT Marvin, Davis & Dawson 24

Natural Resources. This study was approved by the Institutional Animal Care and Use

PT

Committee at the University of North Alabama (UNA). The research was funded by the Biology Department at UNA. GM conceived the study, worked together with KD and JD

RI

to complete experimental design, and completed data analyses and preparation of the

NU

SC

manuscript. KD performed experiment 1 and JD performed experiment 2.

References

MA

[1] Anderson, C.V., Deban, S.M., 2010. Ballistic tongue projection in chameleons maintains high performance at low temperature. Proceedings of the National

TE

D

Academy of Science 107, 5495–5499. [2] Anderson, C.V., Larghi, N.P., Deban, S.M., 2014. Thermal effects on the

AC CE P

performance, motor control and muscle dynamics of ballistic feeding in the salamander Eurycea guttolineata. Journal of Experimental Biology 217, 3146–3158. [3] Angilletta, M.J., 2001. Thermal and physiological constraints on energy assimilation in a widespread lizard (Sceloporus undulatus). Ecology 82, 3044–2056. [4] Ashton, R.E., 1975. A study of movement, home range, and winter behavior of Desmognathus fuscus. Journal of Herpetology 9, 85–91. [5] Avery, R.A., 1971. Estimates of food consumption by the lizard Lacerta vivipara Jacquin. Journal of Animal Ecology 40, 351–365. [6] Avery, R.A., Bedford, J.D., Newcombe, C.P., 1982. The role of thermoregulation in lizard biology: predatory efficiency in a temperate diurnal basker. Behavioural

ACCEPTED MANUSCRIPT Marvin, Davis & Dawson 25

Ecology and Sociobiology 11, 261–267.

PT

[7] Ayers, D.Y., Shine, R., 1997. Thermal influences on foraging ability: body size, posture and cooling rate of an ambush predator, the python Morelia spilota.

RI

Functional Ecology 11, 342–347.

SC

[8] Bakkegard, K.A., Rhea, R.A., 2012. Tail length and sexual size dimorphism (SSD) in

NU

desmognathan salamanders. Journal of Herpetology 46, 304–311. [9] Balogová, M., Gvoždík, L., 2015. Can newts cope with the heat? Disparate

MA

thermoregulatory strategies of two sympatric species in water. PLoS ONE 10(5): e0128155 doi:10.1371/journal.pone.0128155

TE

D

[10] Bobka, M.S., Jaeger, R.G., McNaught, D.C., 1981. Temperature dependent assimilation efficiencies of two species of terrestrial salamanders. Copeia 1981,

AC CE P

417–421.

[11] Bonnett, R.M., 2002. Analysis of the contact zone between the Dusky Salamanders Desmognathus fuscus fuscus and Desmognathus fuscus conanti (Caudata: Plethodontidae). Copeia 2002, 344–355. [12] Bruce, R.C., 1993. Sexual size dimorphism in desmognathine salamanders. Copeia 1993, 313–318. [13] Camp, C.D., Wooten, J.A., Jensen, J.B., Bartek, D.F. 2013. Role of temperature in determining relative abundance in cave twilight zones by two species of lungless salamander (family Plethodontidae). Canadian Journal of Zoology 92, 119–127. [14] Deban, S.M., Lappin, A.K., 2011. Thermal effects on the dynamics and motor

ACCEPTED MANUSCRIPT Marvin, Davis & Dawson 26

control of ballistic prey capture in toads: maintaining high performance at low

PT

temperature. Journal of Experimental Biology 214, 1333–1346. [15] Deban, S.M., Richardson, J.C., 2011. Cold-blooded snipers: thermal independence

SC

Journal of Experimental Zoology A 325, 618–630.

RI

of ballistic tongue projection in the salamander Hydromantes platycephalus.

NU

[16] Else, P.L., Bennett, A.F. 1987. The thermal dependence of locomotor performance and muscle contractile function in the salamander Ambystoma tigrinum

MA

nebulosum. Journal of Experimental Biology 128, 219–233. [17] Feder, M.E., 1982. Thermal ecology of neotropical lungless salamanders

TE

D

(Amphibia: Plethodontidae): environmental temperature and behavioral responses. Ecology 63, 1665–1674.

AC CE P

[18] Feder, M.E., 1983. Integrating the ecology and physiology of plethodontid salamanders. Herpetologica 39, 291–310. [19] Feder, M.E., Lynch J.F., 1982. Effects of latitude, season, elevation, and microhabitat on field body temperatures of neotropical and temperate zone salamanders. Ecology 63, 1657–1664. [20] Feder, M.E., Pough, F.H., 1975. Temperature selection in the Red-backed Salamander, Plethodon c. cinereus (Green) (Caudata: Plethodontidae). Comparative Biochemistry and Physiology 50A, 91–98. [21] Fitzpatrick, L.C., 1973. Energy allocation in the Allegheny Mountain Salamander, Desmognathus ochrophaeus. Ecological Monographs 43, 43–58.

ACCEPTED MANUSCRIPT Marvin, Davis & Dawson 27

[22] Greenwald, O.E., 1974. Thermal dependence of striking and prey capture by

PT

Gopher Snakes. Copeia 1974, 141–148. [23] Hanks, J.E., 1957. The rate of feeding of the common oyster drill, Urosalpinx

RI

cinerea (Say), at controlled water temperatures. Biological Bulletin 112, 330–335.

SC

[24] Harlow, H.J., Hillman, S.S., Hoffman, M., 1976. The effect of temperature on

Comparative Physiology B 111, 1–6.

NU

digestive efficiency in the herbivorous lizard, Dipsosaurus dorsalis. Journal of

MA

[25] Harrison, J.F., Fewell, J.H., 1995. Thermal effects on feeding behavior and net energy intake in a grasshopper experiencing large diurnal fluctuations in body

TE

D

temperature. Physiological Zoology 68, 453–473. [26] Hutchison, V.H., Spriestersbach, K.K., 1986. Diel and seasonal cycles of activity and

AC CE P

behavioral thermoregulation in the salamander Necturus maculosus. Copeia 1986, 612–618.

[27] Jaeger, R.G., 1980. Fluctuations in prey availability and food limitation for a terrestrial salamander. Oecologia 44, 335–341. [28] Jones, R.L., 1986. Reproductive biology of Desmognathus fuscus and Desmognathus santeetlah in the Unicoi Mountains. Herpetologica 42, 323–334. [29] Keen, W.H., 1979. Feeding and activity patterns in the salamander Desmognathus ochrophaeus (Amphibia, Urodela, Plethodontidae). Journal of Herpetology 13, 461–467. [30] Kimball, M.E., Miller, J.M., Whitfield, P.E., Hare, J.A., 2004. Thermal tolerance and

ACCEPTED MANUSCRIPT Marvin, Davis & Dawson 28

potential distribution of invasive lionfish (Pterois volitans/miles complex) on the

PT

east coast of the United States. Marine Ecology Progress Series 283, 269–278. [31] Kingsolver, J.G., Woods, H.A., 1997. Thermal sensitivity of growth and feeding in

RI

Manduca sexta caterpillars. Physiological Biochemistry and Zoology 70, 631–638.

NU

Limnology and Oceanography 16, 580–581.

SC

[32] Kirby, H.V., 1971. Effect of temperature on the feeding behavior of Daphnia rosea.

[33] Kristín, P., Gvoždík, L., 2014. Aquatic‐to‐terrestrial habitat shift reduces energy

MA

expenditure in newts. Journal of Experimental Zoology 321A, 183–188. [34] Kukal, O., Dawson, T.E., 1989. Temperature and food quality influences feeding

TE

D

behavior, assimilation efficiency and growth rate of arctic woolly-bear caterpillars. Oecologia 79, 526–532.

AC CE P

[35] Marchand, F., Magnan, P., Boisclair, D., 2002. Water temperature, light intensity and zooplankton density and the feeding activity of juvenile brook charr (Salvelinus fontinalis). Freshwater Biology 47, 2153–2162. [36] Marek, V., Gvoždík, L., 2012. The insensitivity of thermal preferences to various thermal gradient profiles in newts. Journal of Ethology 30, 35–41. [37] Marvin, G.A., 2003. Aquatic and terrestrial locomotor performance in a semiaquatic plethodontid salamander (Pseudotriton ruber): influence of acute temperature, thermal acclimation, and body size. Copeia 2003, 704–713. [38] Marvin, G.A., 2003. Effects of acute temperature and thermal acclimation on aquatic and terrestrial locomotor performance of the Three-lined Salamander,

ACCEPTED MANUSCRIPT Marvin, Davis & Dawson 29

Eurycea guttolineata. Journal of Thermal Biology 28, 251–259.

PT

[39] Marvin, G.A., Lewis, M., 2013. Effect of temperature, photoperiod, and feeding on the rate of tail regeneration in a semiaquatic plethodontid salamander. Journal of

RI

Thermal Biology 38, 548–552.

SC

[40] Peck, L.S., Webb, K.E., Miller, A., Clark, M.S., Hill, T., 2008. Temperature limits to

NU

activity, feeding and metabolism in the Antarctic starfish Odontaster validus. Marine Ecology Progress Series 358, 181–189.

Institution Press, Washington.

MA

[41] Petranka, J.W., 1998. Salamanders of the United States and Canada. Smithsonian

TE

D

[42] Piasecna, K., Poncova, A., Tejedo, M., Gvoždík, L., 2015. Thermoregulatory strategies in an aquatic ectotherm from thermally-constrained habitats: an

AC CE P

evaluation of current approaches. Journal of Thermal Biology 52, 97–107. [43] Randolph, K.N., Clemens, H.P., 1976. Some factors influencing the feeding behavior of channel catfish in culture ponds. Transactions of the American Fisheries Society 105, 718–724. [44] Rossman, D.A., 1958. A new race of Desmognathus fuscus from the south-central United States. Herpetologica 11, 158–160. [45] Ryerson, W.G., 2013. Jumping in the salamander Desmognathus ocoee. Copeia 2013, 512–516.

ACCEPTED MANUSCRIPT Marvin, Davis & Dawson 30

[46] Secor, M.S., Boehm, M., 2006. Specific dynamic action of ambystomatid

Physiological and Biochemical Zoology 79, 720–735.

PT

salamanders and the effects of meal size, meal type, and body temperature.

RI

[47] Shealy, R.M., 1975. Factors influencing activity in the salamanders Desmognathus

SC

ochrophaeus and D. monticola (Plethodontidae). Herpetologica 31, 94–102.

NU

[48] Sheng, J., Lin, Q., Chen, Q., Gao, Y., Shen, L., Lu, J., 2006. Effects of food, temperature and light intensity on the feeding behavior of three-spot juvenile

MA

seahorses Hippocampus trimaculatus Leach. Aquaculture 256, 596–607. [49] Sievert, L.M., Andreadis, P.T., 2002. Differing diel patterns of temperature

TE

D

selection in two sympatric Desmognathus. Copeia 2002, 62–66. [50] Spotila, J.R., 1972. Role of temperature and water in the ecology of lungless

AC CE P

salamanders. Ecological Monographs 42, 94–125. [51] Stevens, E.D., 1988. Feeding performance of toads at different acclimation temperatures. Canadian Journal of Zoology 66, 537–539. [52] Stevenson, R.D., Peterson, C.R., Tsuji, J.S., 1985. The thermal dependence of locomotion, tongue flicking, digestion, and oxygen consumption in the wandering garter snake. Physiological Zoology 58, 46–57. [53] Titus, T.A., Larson, A., 1996. Molecular phylogenetics of desmognathine salamanders (Caudata: Plethodontidae): a reevaluation of evolution in ecology, life history, and morphology. Systematic Biology 45, 451–472

ACCEPTED MANUSCRIPT Marvin, Davis & Dawson 31

[54] Van Damme, R., Bauwens, D., Verheyen, R.F., 1991. The thermal dependence of

PT

feeding behavior, food consumption and gut-passage time in the lizard Lacerta vivipara Jacquin. Functional Ecology 5, 507–517.

RI

[55] Waldschmidt, S.R., Jones, S.M., Porter, W.P., 1986. The effect of body

SC

temperature and feeding regime on activity, passage time, and digestive

NU

coefficient in the lizard Uta stansburiana. Physiological Zoology 59, 376–383. [56] Warkentin, K.M., 1992. Effects of temperature and illumination on feeding rates

MA

of Green Frog tadpoles (Rana clamitans). Copeia 1992, 735–730. [57] White, C.R., Phillips, N.F., Seymour, R.S., 2006. The scaling and temperature

AC CE P

TE

D

dependence of vertebrate metabolism. Biology Letters 2, 125–127.

Figure Captions

Fig. 1. Effect of salamander body temperature on feeding behavior for individuals of Desmognathus conanti during feeding trials with a video recording of a walking, warm (15oC) cricket. Control-group individuals (n = 15) were tested at 15oC when experimental-group individuals (n = 14) were tested at each acute low temperature. Feeding response to video was quantified as “0” if salamander showed no orientation response, “1” if salamander turned head toward cricket, “2” if salamander partially approached cricket, “3” if salamander fully approached cricket, or “4” if salamander attempted prey capture (see text for more description). *p < 0.001 (repeated-

ACCEPTED MANUSCRIPT Marvin, Davis & Dawson 32

measures ANOVA; Holm-Sidak multiple comparison). Box plots show min, max,

PT

median, and percentiles (10th, 25th, 75th, and 90th).

RI

Fig. 2. Effect of salamander body temperature on the variability of feeding behavior for

SC

experimental-group individuals (n = 14) of Desmognathus conanti during three feeding

NU

trials with a video recording of a walking, warm (15oC) cricket. Feeding variability was determined by coefficient of variation (CV) for individual feeding responses at each

MA

temperature. *p < 0.01 (repeated-measures ANOVA); p < 0.05 (Holm-Sidak multiple comparison). Box plots show min, max, median, and percentiles (10th, 25th, 75th, and

TE

D

90th).

AC CE P

Fig. 3. Effect of salamander body temperature on feeding behavior for individuals of Desmognathus conanti during feeding trials with live, warm (15oC) crickets. Controlgroup individuals (n = 16) were tested at 15oC when experimental-group individuals (n = 16) were tested at each acute low temperature. Feeding response to cricket was quantified as “0” if salamander showed no orientation response, “1” if salamander turned head toward cricket, “2” if salamander partially approached cricket, “3” if salamander fully approached cricket, or “4” if salamander attempted prey capture (see text for more description). *p < 0.001 (repeated-measures ANOVA; Holm-Sidak multiple comparison). Box plots show min, max, median, and percentiles (10th, 25th, 75th, and 90th).

ACCEPTED MANUSCRIPT Marvin, Davis & Dawson 33

PT

Fig. 4. Effect of salamander body temperature on prey-capture time for individuals of Desmognathus conanti during feeding trials with live, warm (15oC) crickets. Control-

RI

group individuals (n = 16) were tested at 15oC when experimental-group individuals (n

SC

= 16) were tested at each acute low temperature. Prey-capture times did not

NU

significantly differ among temperatures for experimental-group individuals, but preycapture times were significantly greater for experimental-group individuals at 3 and

MA

5°C (ANOVA on ranks; Holm-Sidak multiple comparison; *p < 0.001 and *p = 0.004, respectively) than for control-group individuals at 15°C. Box plots show min, max,

AC CE P

TE

D

median, and percentiles (10th, 25th, 75th, and 90th).

ACCEPTED MANUSCRIPT

AC CE P

TE

D

MA

NU

SC

RI

PT

Marvin, Davis & Dawson 34

Figure 1

ACCEPTED MANUSCRIPT

Figure 2

AC CE P

TE

D

MA

NU

SC

RI

PT

Marvin, Davis & Dawson 35

ACCEPTED MANUSCRIPT

AC CE P

TE

D

MA

NU

SC

RI

PT

Marvin, Davis & Dawson 36

Figure 3

ACCEPTED MANUSCRIPT

Figure 4

AC CE P

TE

D

MA

NU

SC

RI

PT

Marvin, Davis & Dawson 37

ACCEPTED MANUSCRIPT Marvin, Davis & Dawson 38

Table 1. Examples of low-temperature limits for feeding in ectothermic animals. Taxonomic Common name group Echinodermata Antarctic Starfish

Scientific name

Mollusca

Common Oyster Drill

Urosalpinx cinerea

Crustacea

Daphnia

Insecta

Two-striped Grasshopper Arctic Woolly-bear Caterpillar Tobacco Hornworm Caterpillar

[23]

Daphnia rosea

<5

[32]

Melanoplus bivittatus Gynaephora groenlandica

11 <5

[25] [34]

Manduca sexta

14

[31]

Pterois volitans/miles Ictalurus punctatus Salvelinus fontinalis Hippocampus trimaculatus

16 12 <17 19

[30] [43] [35] [48]

Lionfish Channel Catfish Brook Charr (juvenile) Three-spot Seahorse

Amphibia

American Toad Southern Toad Green Frog (tadpole) Red-backed Salamander Spotted Dusky Salamander

Bufo americanus Bufo terrestris Rana clamitans Plethodon cinereus Desmognathus conanti

5 <11 <20 <5 3

Northern Dusky Salamander Mountain Dusky Salamander Three-lined Salamander Mount Lyell Salamander

Desmognathus fuscus Desmognathus ochrophaeus Eurycea guttolineata Hydromantes platycephalus

<6 <5

[51] [14] [56] [27] Present study [4] [21,29]

<5 2

[2] [15]

Diamond Python Gopher Snake Chuckwalla Lizard Common Lizard Veiled Chameleon

Morelia spilota Pituophis catenifer affinis Dipsosaurus dorsalis Lacerta vivipara Chamaeleo calyptratus

10 <18 28 10 15

[7] [22] [24] [5] [1]

AC CE P

D

MA

Osteichthyes

TE

NU

RI

6

SC

PT

Odontaster validus

Limit Reference (oC) <0 [40]

Reptilia

ACCEPTED MANUSCRIPT Marvin, Davis & Dawson 39

Table 2. Body size comparisons for individuals of Desmognathus conanti in control and experimental groups for two experiments on the effect of acute low body

PT

temperature on predatory behavior and prey-capture efficiency. Body size is presented

Control

51.1 ± 5.8

46.8 ± 6.6

2.9 ± 1.0

1

(n = 15)

(40.9–63.1)

(31.4–59.4)

(1.6–5.0)

50.2 ± 6.7

45.0 ± 7.2

3.1 ± 1.0

(41.0–64.8)

(29.8–57.2)

(1.7–5.3)

t = 0.415

t = 0.692

t = 0.409

p = 0.681

p = 0.495

p = 0.686

48.5 ± 7.0

48.8 ± 8.2

3.0 ± 1.1

(38.1–61.7)

(34.2–62.4)

(1.6–5.5)

48.4 ± 7.0

48.7 ± 7.4

2.9 ± 0.9

(36.7–66.8)

(34.9–66.5)

(1.6–4.9)

t-test

t = 0.025

t = 0.038

t = 0.546

(df = 30)

p = 0.980

p = 0.893

p = 0.589

NU

Experiment

Body mass (g)

MA

(mm)

Tail length (mm)

SC

Body length

RI

as mean ± SD (range).

Experimental (n = 14) t-test

Control

2

(n = 16)

TE

Experiment

D

(df = 27)

AC CE P

Experimental (n = 16)

ACCEPTED MANUSCRIPT Marvin, Davis & Dawson 40

Highlights Foraging by ectotherms is influenced by thermal effects on predator and prey.



Thermal effect on salamander predator, independent of prey, is not known.



Feeding behavior of cold salamanders was quantified in lab with warm prey.



Many individuals (50%) exhibited predatory behavior at 3oC.



Individuals captured prey with high efficiency at 3oC.

AC CE P

TE

D

MA

NU

SC

RI

PT

