Journal of Chemical Ecology, Vol. 15, No. 3, 1989
FOLLOWING OF CONSPECIFIC AND AVOIDANCE OF PREDATOR CHEMICAL CUES BY PINE SNAKES
(Pituophis melanoleucus)
JOANNA BURGER Department of Biological Sciences Nelson Biological Laboratory Piscataway, New Jersey 08855-1059 (Received December 14, 1987; accepted March 7, 1988) Abstract--The ability of hatchling pine snakes (Pituophis melanoleucus) to follow or avoid the chemical trails of conspecifics and a king snake (Lampropeltis getulus) on paper substrates was investigated in Y-maze experiments. Hatchlings entered the arm with the adult conspecifictrail and avoided the arm containing the king snake trail at a frequency much greater than that due to chance. The data support the hypotheses that pine snakes follow the chemical trails of adult conspecifics and avoid the chemical trails of a predator. Key Words--Chemical cues, predator avoidance, pine snake, Pituophis melanoleucus, odor trails.
INTRODUCTION Trailing o f conspecifics or prey by chemical means has been demonstrated in controlled laboratory experiments for snakes (Brown and Maclean, 1983; Chiszar et al., 1986; Ford, 1982; Ford and O'Bleness, 1986; Ford and Schofield, 1984; Gehlbach et al., 1971; Heller and Halpern, 1981) and for lizards (Cooper and Vitt, 1986a). Further, reptiles often respond to odors or those on applicators by increased tongue flicks (Cooper and Vitt, 1984, 1986b; Cooper et al., 1986). In some cases pheromonal communication is suspected because adult male skinks follow the trails of adult conspecific females, but not those o f other males (e.g., Eumeces laticeps; Cooper and Vitt, 1986c). For reviews of the role of chemoreception in reptiles, see Burghardt (1980), Simon (1983), and Von Achen and Rakestraw (1984). 799 0098-0331/89/0300-0799506.00/0 9 1989 Plenum Publishing Corporation
800
BURGER
Earlier studies have shown the use of chemical abilities in several important contexts, including recognition of prey (Burghardt, 1973; Chiszar et al., 1986), detection of conspecifics (Cooper and Vitt, 1984), discrimination of male from female conspecifics (Cooper and Vitt, 1984), discrimination of conspecific and closely related syntopic congeners (Cooper and Vitt, 1986c), trailing conspecifics to find hibernacula (Brown and Maclean, 1983) or mates (Ford and O'Bleness, 1986), and differentiation of ophiophagous from nonophiophagous snakes (Weldon and Burghardt, 1979; Weldon, 1982). However, Weldon's (1982) studies dealt with changes in tongue-flick frequency. There has been no clear demonstration in snakes of discrimination of ophiophagous from nonophiophagous snakes. Snakes can distinguish conspecific from heterospecific odor trails (Ford and O'Bleness, 1986). Because some snakes are predators on other snakes, it would be advantageous for prey species to be able to detect the chemical trails of potential predators and avoid them. This would be particularly true for hatchlings that are more vulnerable to predators because of their small size. In this article I report experiments designed to examine choice discrimination by hatchling pine snakes (Pituophis m. melanoleucus) of the odor trail of conspecifics and of a heterospecific predator of pine snakes, the king snake (Lam-
propeltis getulus).
METHODS AND MATERIALS
Under appropriate state permits, pine snake eggs were collected from the Pine Barrens of southern New Jersey (Ocean, Cumberland, and Monmouth counties), and 200 hatchlings were hatched and maintained in the laboratory in 1986. Date of hatching was noted, so the exact age of each snake at testing was known. Snakes were maintained individually in plastic (30 x 15 x 9 cm) cages containing paper for shelter. Snakes were given the opportunity to drink water daily and to eat young laboratory mice once a week. Not all snakes had eaten, but there were no significant differences in behavior as a function of this variable. Snakes were sexed by hemipenis eversion, a reliable rapid procedure for young snakes (Schaefer, 1934; Fitch, 1960; Gregory, 1983; Gutzke et al., 1985). To determine whether pine snakes could make directional responses to chemical trails, Y-maze experiments were conducted. The base arm of the Y maze was 1 m long and 15 cm wide, with 15-cm-high wooden sides. At the end of the base arm, two side arms (experimental arms, 45 angle from the base arm) were the same dimensions as the base arm. The floor of the maze was covered with paper that was changed after each trial. Plexiglas was placed over the maze to prevent the snakes from escaping from the center, choice point of o
PINE SNAKES R E S P O N D T O C H E M I C A L CUES
801
the maze. All ends of the maze were open during testing, allowing the snake to leave the apparatus at the end of the trial. There were four experiments: (1) control (N = 30, no chemical trail on either side), (2) hatchling conspecific versus no trail (N = 74, trail on one side), (3) adult conspecific (snout-vent length = 120 cm) versus no trail (N = 44, trail on one side), (4) adult king snake (snout-vent length = 86 cm) versus no trail (N = 117, trail on one side). Trials were conducted from September 25 to October 15 on hatchlings that were 25-35 days old. Each snake was tested once in either the control or predator experiment and once in either the hatchling or adult conspecific experiment. The order of the experiments was selected randomly each day. Snakes were assigned randomly to each experiment. Sample size varied because snakes were used only in one type of conspecific test. I was primarily interested in their responses to hatchlings (that they would encounter in nature) and the predator; thus I used more snakes in these two tests. On some trials a snake would not enter the test (they had recently eaten or were otherwise overactive), accounting for further unequal sample sizes. The temperature was maintained at 27~ during all experiments, which were conducted from 1100 to 1400 hr. The base ann contained no chemical trail, and there was no odor trail in one of the experimental arms. A chemical trail in the other experimental arm was produced by letting the stimulus (hatchlings, adult pine snake, or king snake) move freely over the paper and shavings for 10 rain while both ends of the experimental arm were closed. The hatchling trail was established by allowing three hatchlings from a brood not otherwise included in the sample to move freely over the experimental arm. Only one king and three pine snake adults were used for the odor trail because that is all that were available. I do not believe that these adults possessed an idiosyncratic odor different from other conspecifics. The chemical trail extended the full length of the experimental arm: the trail-providing snakes were seen to move the entire length of the arm. When handling snakes, paper, or shavings, the experimenters wore disposable surgical gloves to avoid imparting human or snake odors. Following each trial, the paper and shavings were changed, and the location of the experimental arm was changed. Whenever a snake contacted the sides of the maze, all walls were subsequently washed with soap and water. After every 10 trials the entire apparatus and floor were also cleaned with soap and water. Each snake was placed by hand into the base arm facing the opening with its head at the entrance and allowed to move freely up the arm to the intersection, where it usually moved into one of the experimental arms. The trial ended when the snake left the Ymaze, usually through the experimental arms, although some went back out the base ann after arriving at the Yjunction. The time each snake was in the base and experimental arm, time at the Y junction, and the
802
BURGER
number of tongue flicks when the snake was in each location were recorded. An assistant released the snake into the maze. The behavior of hatchlings in conspecific tests did not differ as a function of their choice; however, behavior did vary in the predator test, and these data are presented. Chi square tests were used to determine the significance of the responses frequency in the four experiments, and Kruskal-Wallis chi square tests were used to determine differences in central tendencies of time and tongue flicks under several conditions. Data are presented as means + 1 SD in the text. RESULTS
Most snakes moved rapidly to the Y junction, remained motionless for a few seconds (X" = 18.4 + 6 sec), explored the Y junction for 5-35 sec (X = 19.9 + 11.2 sec), and moved rapidly down one of the arms. When there was no chemical trail (control experiment), there was no significant difference in their choice (Table 1). Pine snake hatchlings did not follow the chemical trails of other conspecific hatchlings, but they did follow the trails of adult pine snakes (Table 1). In these conspecific trials, only one snake turned around and went back out the base arm, and none tried to go over the top of the maze. Hatchling pine snakes avoided the king snake chemical trail and chose the arm with no chemical trail (Table 1). In this experiment 100 of the snakes selected an arm, and 17 either turned around and went out the base arm (N = 8), tried to climb over the top displacing the Plexiglas (N = 4), or moved back
TABLE 1. RESPONSES OF
Pituophis melanoleucus TO CHEMICAL TRAILS IN Y
MAZE a
Arm contains
Number of times chosen
Control
No trail (right) No trail (left)
13 17
0.53 (NS)
Hatchling conspecific vs. no trail
Hatchling No trail
44 30
2.64 (NS)
Adult conspecifics vs, no trail
Adult No trail
38 6
23.3 (0.001)
Predator vs. no trail
King snake No trail
11 89
60.8 (0.001)
Experiment
X2 (P)
a Individual snakes tested in only one of the conspecific tests, and either the predator or the control test. NS = not significant.
803
PINE SNAKES RESPOND TO CHEMICAL CUES
TABLE 2. BEHAVIOR OF NAIVE PINE SNAKES GIVEN A CHOICE OF PREDATOR TRAIL OR N O TRAIL a
Choice
Number of snakes Base arm Time (sec) Tongue flicks Choice siteb Motionless Time (sec) Flicks Explore Time (sec) Flicks Experimental arm Time (sec) Flicks
King snake trail
No chemical trail
15
89
4.8 _+ 0.9 18.0 + 0.7
4.4 _+ 0.3 17.4 + 0.2
0.36 (NS) 1.79 (NS)
24.0 + 11.7 13.2 _+ 4.2
13.4 _+ 2.6 18.7 + 1.1
3.47 (NS) 4.65 (0.09)
29.3 + 19.6 4.2 + 2.2
9.9 + 3.3 2.8 + 0.7
11.19 (0.003) 8.32 (0.01)
24.5 _+ 10.7 16.1 + .6
4.1 + 0.6 16.4 + 0.3
11.1 (0.003) 1.36 (NS)
x2(P)
~Given are means _ 1 SD. Tongue flicks are per 15 sec. NS = not significant. b
and forth between the two experimental arms (N = 5). It appeared that some hatchlings were responding to airborne odors by their obviously increased activity and attempts to get out of the Y maze. As might be expected, there were no differences in the time each snake was in the base arm as a function of its final choice (Table 2). However, snakes emitted significantly more tongue flicks when exploring at the Y junction and when they went down the king snake chemical trail compared to the arm with no chemical trail. Similarly, they explored longer when they selected the king snake arm and moved significantly more slowly down the king snake arm compared to the blank (Table 2). There were no significant differences as a function of age (15-25 days old) or sex of the hatchling, time of day, or location of the experimental arm (P
FOLLOWING OF CONSPECIFIC AND AVOIDANCE OF PREDATOR CHEMICAL CUES BY PINE SNAKES
(Pituophis melanoleucus)
JOANNA BURGER Department of Biological Sciences Nelson Biological Laboratory Piscataway, New Jersey 08855-1059 (Received December 14, 1987; accepted March 7, 1988) Abstract--The ability of hatchling pine snakes (Pituophis melanoleucus) to follow or avoid the chemical trails of conspecifics and a king snake (Lampropeltis getulus) on paper substrates was investigated in Y-maze experiments. Hatchlings entered the arm with the adult conspecifictrail and avoided the arm containing the king snake trail at a frequency much greater than that due to chance. The data support the hypotheses that pine snakes follow the chemical trails of adult conspecifics and avoid the chemical trails of a predator. Key Words--Chemical cues, predator avoidance, pine snake, Pituophis melanoleucus, odor trails.
INTRODUCTION Trailing o f conspecifics or prey by chemical means has been demonstrated in controlled laboratory experiments for snakes (Brown and Maclean, 1983; Chiszar et al., 1986; Ford, 1982; Ford and O'Bleness, 1986; Ford and Schofield, 1984; Gehlbach et al., 1971; Heller and Halpern, 1981) and for lizards (Cooper and Vitt, 1986a). Further, reptiles often respond to odors or those on applicators by increased tongue flicks (Cooper and Vitt, 1984, 1986b; Cooper et al., 1986). In some cases pheromonal communication is suspected because adult male skinks follow the trails of adult conspecific females, but not those o f other males (e.g., Eumeces laticeps; Cooper and Vitt, 1986c). For reviews of the role of chemoreception in reptiles, see Burghardt (1980), Simon (1983), and Von Achen and Rakestraw (1984). 799 0098-0331/89/0300-0799506.00/0 9 1989 Plenum Publishing Corporation
800
BURGER
Earlier studies have shown the use of chemical abilities in several important contexts, including recognition of prey (Burghardt, 1973; Chiszar et al., 1986), detection of conspecifics (Cooper and Vitt, 1984), discrimination of male from female conspecifics (Cooper and Vitt, 1984), discrimination of conspecific and closely related syntopic congeners (Cooper and Vitt, 1986c), trailing conspecifics to find hibernacula (Brown and Maclean, 1983) or mates (Ford and O'Bleness, 1986), and differentiation of ophiophagous from nonophiophagous snakes (Weldon and Burghardt, 1979; Weldon, 1982). However, Weldon's (1982) studies dealt with changes in tongue-flick frequency. There has been no clear demonstration in snakes of discrimination of ophiophagous from nonophiophagous snakes. Snakes can distinguish conspecific from heterospecific odor trails (Ford and O'Bleness, 1986). Because some snakes are predators on other snakes, it would be advantageous for prey species to be able to detect the chemical trails of potential predators and avoid them. This would be particularly true for hatchlings that are more vulnerable to predators because of their small size. In this article I report experiments designed to examine choice discrimination by hatchling pine snakes (Pituophis m. melanoleucus) of the odor trail of conspecifics and of a heterospecific predator of pine snakes, the king snake (Lam-
propeltis getulus).
METHODS AND MATERIALS
Under appropriate state permits, pine snake eggs were collected from the Pine Barrens of southern New Jersey (Ocean, Cumberland, and Monmouth counties), and 200 hatchlings were hatched and maintained in the laboratory in 1986. Date of hatching was noted, so the exact age of each snake at testing was known. Snakes were maintained individually in plastic (30 x 15 x 9 cm) cages containing paper for shelter. Snakes were given the opportunity to drink water daily and to eat young laboratory mice once a week. Not all snakes had eaten, but there were no significant differences in behavior as a function of this variable. Snakes were sexed by hemipenis eversion, a reliable rapid procedure for young snakes (Schaefer, 1934; Fitch, 1960; Gregory, 1983; Gutzke et al., 1985). To determine whether pine snakes could make directional responses to chemical trails, Y-maze experiments were conducted. The base arm of the Y maze was 1 m long and 15 cm wide, with 15-cm-high wooden sides. At the end of the base arm, two side arms (experimental arms, 45 angle from the base arm) were the same dimensions as the base arm. The floor of the maze was covered with paper that was changed after each trial. Plexiglas was placed over the maze to prevent the snakes from escaping from the center, choice point of o
PINE SNAKES R E S P O N D T O C H E M I C A L CUES
801
the maze. All ends of the maze were open during testing, allowing the snake to leave the apparatus at the end of the trial. There were four experiments: (1) control (N = 30, no chemical trail on either side), (2) hatchling conspecific versus no trail (N = 74, trail on one side), (3) adult conspecific (snout-vent length = 120 cm) versus no trail (N = 44, trail on one side), (4) adult king snake (snout-vent length = 86 cm) versus no trail (N = 117, trail on one side). Trials were conducted from September 25 to October 15 on hatchlings that were 25-35 days old. Each snake was tested once in either the control or predator experiment and once in either the hatchling or adult conspecific experiment. The order of the experiments was selected randomly each day. Snakes were assigned randomly to each experiment. Sample size varied because snakes were used only in one type of conspecific test. I was primarily interested in their responses to hatchlings (that they would encounter in nature) and the predator; thus I used more snakes in these two tests. On some trials a snake would not enter the test (they had recently eaten or were otherwise overactive), accounting for further unequal sample sizes. The temperature was maintained at 27~ during all experiments, which were conducted from 1100 to 1400 hr. The base ann contained no chemical trail, and there was no odor trail in one of the experimental arms. A chemical trail in the other experimental arm was produced by letting the stimulus (hatchlings, adult pine snake, or king snake) move freely over the paper and shavings for 10 rain while both ends of the experimental arm were closed. The hatchling trail was established by allowing three hatchlings from a brood not otherwise included in the sample to move freely over the experimental arm. Only one king and three pine snake adults were used for the odor trail because that is all that were available. I do not believe that these adults possessed an idiosyncratic odor different from other conspecifics. The chemical trail extended the full length of the experimental arm: the trail-providing snakes were seen to move the entire length of the arm. When handling snakes, paper, or shavings, the experimenters wore disposable surgical gloves to avoid imparting human or snake odors. Following each trial, the paper and shavings were changed, and the location of the experimental arm was changed. Whenever a snake contacted the sides of the maze, all walls were subsequently washed with soap and water. After every 10 trials the entire apparatus and floor were also cleaned with soap and water. Each snake was placed by hand into the base arm facing the opening with its head at the entrance and allowed to move freely up the arm to the intersection, where it usually moved into one of the experimental arms. The trial ended when the snake left the Ymaze, usually through the experimental arms, although some went back out the base ann after arriving at the Yjunction. The time each snake was in the base and experimental arm, time at the Y junction, and the
802
BURGER
number of tongue flicks when the snake was in each location were recorded. An assistant released the snake into the maze. The behavior of hatchlings in conspecific tests did not differ as a function of their choice; however, behavior did vary in the predator test, and these data are presented. Chi square tests were used to determine the significance of the responses frequency in the four experiments, and Kruskal-Wallis chi square tests were used to determine differences in central tendencies of time and tongue flicks under several conditions. Data are presented as means + 1 SD in the text. RESULTS
Most snakes moved rapidly to the Y junction, remained motionless for a few seconds (X" = 18.4 + 6 sec), explored the Y junction for 5-35 sec (X = 19.9 + 11.2 sec), and moved rapidly down one of the arms. When there was no chemical trail (control experiment), there was no significant difference in their choice (Table 1). Pine snake hatchlings did not follow the chemical trails of other conspecific hatchlings, but they did follow the trails of adult pine snakes (Table 1). In these conspecific trials, only one snake turned around and went back out the base arm, and none tried to go over the top of the maze. Hatchling pine snakes avoided the king snake chemical trail and chose the arm with no chemical trail (Table 1). In this experiment 100 of the snakes selected an arm, and 17 either turned around and went out the base arm (N = 8), tried to climb over the top displacing the Plexiglas (N = 4), or moved back
TABLE 1. RESPONSES OF
Pituophis melanoleucus TO CHEMICAL TRAILS IN Y
MAZE a
Arm contains
Number of times chosen
Control
No trail (right) No trail (left)
13 17
0.53 (NS)
Hatchling conspecific vs. no trail
Hatchling No trail
44 30
2.64 (NS)
Adult conspecifics vs, no trail
Adult No trail
38 6
23.3 (0.001)
Predator vs. no trail
King snake No trail
11 89
60.8 (0.001)
Experiment
X2 (P)
a Individual snakes tested in only one of the conspecific tests, and either the predator or the control test. NS = not significant.
803
PINE SNAKES RESPOND TO CHEMICAL CUES
TABLE 2. BEHAVIOR OF NAIVE PINE SNAKES GIVEN A CHOICE OF PREDATOR TRAIL OR N O TRAIL a
Choice
Number of snakes Base arm Time (sec) Tongue flicks Choice siteb Motionless Time (sec) Flicks Explore Time (sec) Flicks Experimental arm Time (sec) Flicks
King snake trail
No chemical trail
15
89
4.8 _+ 0.9 18.0 + 0.7
4.4 _+ 0.3 17.4 + 0.2
0.36 (NS) 1.79 (NS)
24.0 + 11.7 13.2 _+ 4.2
13.4 _+ 2.6 18.7 + 1.1
3.47 (NS) 4.65 (0.09)
29.3 + 19.6 4.2 + 2.2
9.9 + 3.3 2.8 + 0.7
11.19 (0.003) 8.32 (0.01)
24.5 _+ 10.7 16.1 + .6
4.1 + 0.6 16.4 + 0.3
11.1 (0.003) 1.36 (NS)
x2(P)
~Given are means _ 1 SD. Tongue flicks are per 15 sec. NS = not significant. b
and forth between the two experimental arms (N = 5). It appeared that some hatchlings were responding to airborne odors by their obviously increased activity and attempts to get out of the Y maze. As might be expected, there were no differences in the time each snake was in the base arm as a function of its final choice (Table 2). However, snakes emitted significantly more tongue flicks when exploring at the Y junction and when they went down the king snake chemical trail compared to the arm with no chemical trail. Similarly, they explored longer when they selected the king snake arm and moved significantly more slowly down the king snake arm compared to the blank (Table 2). There were no significant differences as a function of age (15-25 days old) or sex of the hatchling, time of day, or location of the experimental arm (P
