Male Burmese Pythons Follow Female Scent Trails and Show Sex- Specific Behaviors

University of Nebraska - Lincoln DigitalCommons@University of Nebraska - Lincoln

USDA National Wildlife Research Center - Staff U.S. Department of Agriculture: Animal and Publications Plant Health Inspection Service

Male Burmese pythons follow female scent trails and show sex- specific behaviors

Shannon A. Richard

Eric A. Tillman

John S. Humphrey

Michael L. Avery

M. Rockwell Parker

Follow this and additional works at: https://digitalcommons.unl.edu/icwdm_usdanwrc

Part of the Natural Resources and Conservation Commons, Natural Resources Management and Policy Commons, Other Environmental Sciences Commons, Other Veterinary Medicine Commons, Population Biology Commons, Terrestrial and Aquatic Ecology Commons, Veterinary Infectious Diseases Commons, Veterinary Microbiology and Immunobiology Commons, Veterinary Preventive Medicine, Epidemiology, and Public Health Commons, and the Zoology Commons

This Article is brought to you for free and open access by the U.S. Department of Agriculture: Animal and Plant Health Inspection Service at DigitalCommons@University of Nebraska - Lincoln. It has been accepted for inclusion in USDA National Wildlife Research Center - Staff Publications by an authorized administrator of DigitalCommons@University of Nebraska - Lincoln. Integrative Zoology 2019; 14: 460–469 doi:10.1111/1749-4877.12376

1 ORIGINAL ARTICLE 1 2 2 3 3 4 4 5 5 6 6 7 Male Burmese pythons follow female scent trails and show sex- 7 8 8 9 specific behaviors 9 10 10 11 11 12 Shannon A. RICHARD,1 Eric A. TILLMAN,2 John S. HUMPHREY,2 Michael L. AVERY2 and M. 12 13 13 1 14 Rockwell PARKER 14 15 1Department of Biology, James Madison University, Harrisonburg, Virginia, USA and 2National Wildlife Research Center, Florida 15 16 16 Field Station, U.S. Department of Agriculture, Gainesville, Florida, USA 17 17 18 18 19 19 20 Abstract 20 21 21 Animals communicate with potential mates using species-specific signals, and pheromones are powerful sexu- 22 22 al signals that modify conspecific behavior to facilitate mate location. Among the vertebrates, snakes are espe- 23 23 cially adept in mate searching via chemical trailing, which is particularly relevant given that many snake spe- 24 24 cies are invasive outside their native ranges. Chemical signals used in mate choice are, thus, potentially valuable 25 25 tools for management of invasive snake species. The Burmese python (Python bivittatus) is an invasive snake 26 26 in the Florida Everglades where it is negatively impacting native fauna. In this study, we sought to: (i) deter- 27 27 mine if males can follow conspecific chemical trails in a Y-maze; and (ii) describe the mate searching behav- 28 28 iors exhibited by males while trailing. All males consistently followed a single female scent trail in the maze, 29 29 but when only a male scent trail was present they did not discriminate between the male and blank arms. Rate 30 30 of tongue-flicking, a proxy for chemosensory sampling, was also marginally higher when males were following 31 31 female versus male scent trails. However, when both female and male scent trails were simultaneously present 32 32 in the Y-maze, males did not show a preference for the female arm, although the tongue-flick rate was higher in 33 33 the female-only trial compared to female versus male. Analyses of multiple male behaviors individually and us- 34 34 ing an ethogram revealed that behaviors were more frequent and complex in the female-only trials compared 35 35 to male-only trials. Additional behavioral trials are needed to determine if an effective pheromonal approach to 36 36 Burmese python management is possible. 37 37 38 Key words: chemical ecology, invasive species, pheromone, snake, trailing 38 39 39 40 40 41 41 42 42 43 INTRODUCTION specifics and relays valuable information about the 43 44 Animals use scent as a powerful form of sexual sig- signaler. For example, male field crickets (Gryllus bi- 44 45 naling that elicits robust, sex-specific behaviors in con- maculatus De Geer, 1773) can determine sex and age 45 46 from chemicals in the cuticle that prompt either court- 46 47 ship (female odor) or aggression (male odor) (Nagamo- 47 48 to et al. 2005). In vertebrates, a variety of sex-specific 48 49 Correspondence: M. Rockwell Parker, MSC 7801 Bioscience, behaviors have been documented in response to sexual 49 50 951 Carrier Drive, Harrisonburg, VA 22807, USA. chemical signals, such as the classic flehmen response in 50 51 Email: [email protected] male mammals exposed to female scent that also chang- 51

1 es based on female reproductive state (e.g. Rasmussen contribute to the success of Burmese pythons in the Ev- 1 2 et al. 1982). In leopard geckos [Eublepharis macular- erglades, notably their high reproductive potential, large 2 3 ius (Blyth , 1854)], males attack conspecifics of either size, and generalist diet (Reed et al. 2012). 3 4 sex when they produce masculine scent, which occurs Studying the chemical ecology of the Burmese py- 4 5 after treatment with male sex hormones (Rhen & Crews thon is a promising avenue for research given that other 5 6 2000). The sexual odors prompting such specific behav- species of pythons use chemical cues in mate tracking. 6 7 iors can be isolated and identified as pheromones, which Male carpet pythons (Morelia spilota imbricata Smith, 7 8 facilitates both basic and applied research. 1981) discriminate female scent trails (Bryant et al. 8 9 Reproductive chemical signals are useful to wild- 2011). Males of another python species (Liasis macklo- 9 10 life managers focused on invasive species because of ti Duméril & Bibron, 1844) trail conspecifics in a maze 10 11 the strong behavioral effects they stimulate. Many inva- and discriminate between subspecies (Carmichael et al. 11 12 sive insect species can be effectively controlled through 2007). Establishing that Burmese pythons use chemicals 12 13 pheromone-based mediation, and, more recently, pher- for communication is an important step in understand- 13 14 omones have shown significant promise in controlling ing the role of such signaling in this invasive species 14 15 vertebrate pest species (Witzgall et al. 2010; Takács et and will prove useful in future tests of python-derived 15 16 al. 2017). Furthermore, many other sexual signals are chemical cues for potential field applications. 16 17 important in mate searching and assessment and have 17 18 18 been manipulated by researchers, such as visual lures MATERIALS AND METHODS 19 for pest insects, acoustic traps for fish, and combination 19 20 lures using visual and acoustic cues for pest bird species Animals and husbandry 20 21 (Eriksson & Wallin 1986; Lelito et al. 2008; Moynan et 21 22 al. 2016). Across invasive animals, chemical cues have Male (n = 6) and female (n = 2) Burmese pythons 22 23 been the most successful type of signals adapted as tools (P. bivittatus) were caught in the Florida Everglades 23 24 for management (Witzgall et al. 2010). and transported to the U.S. Department of Agriculture 24 25 (USDA) National Wildlife Research Center (NWRC) 25 Reptiles rely heavily on reproductive chemical sig- 26 field station in Gainesville, Florida. We maintained 26 nals in mate tracking and assessment, and many species 27 snakes in individual outdoor pens (1.5 × 3 × 1.8 m) at 27 are currently invasive or have significant invasion po- 28 ambient conditions in accordance with guidelines estab- 28 tential. Of all groups of reptiles, chemical communica- 29 lished by the NWRC IACUC and the Florida Fish and 29 tion is most well-documented in snakes, especially mate 30 Wildlife Conservation Commission. One male died be- 30 searching and courtship (Mason 1992; Parker & Mason 31 fore all trials were completed, reducing the sample size 31 2011). Furthermore, the chemical cues used by snakes 32 to n = 5 for the male-only and male versus female trials. 32 in mate choice can be tested robustly in both the field 33 33 and the laboratory (Mason et al. 1998; Parker & Mason Y-maze trials 34 34 2011). Finally, male snakes use female scent trails to lo- 35 35 cate mates but also to ascertain multiple female quali- We used a Y-maze to determine how male Burmese 36 36 ties, such as size and condition (Ford 1986; Mason et al. pythons (n = 6 for female only trials; n = 5 for male 37 37 1989; Shine et al. 2003). only and male vs female trials) responded to conspecif- 38 ic scent trails (Fig. 1). The Y-maze had an initial 1.4-m 38 39 There is potential to develop chemical lures for man- passageway leading from the start box, ending at a 45° 39 40 agement of invasive snake species, yet the chemical Y-junction from which two 1.2 m passageways (desig- 40 41 ecology of just a single species, brown treesnake [Boiga nated North and South) led to collection boxes (plastic 41 42 irregularis (Merrem, 1802)], has been detailed (Greene storage bins). The maze was attached to a 2.4 × 2.4 m 42 43 & Mason 1998; Greene et al. 2001; Mathies et al. 2013). “Hardie board” base. The arms of the maze were made 43 44 The Burmese python (Python bivittatus Kuhl, 1820) is of 2.5 × 15.2-cm PVC side boards. The maze was se- 44 45 an invasive snake in the Florida Everglades that is seri- cured within a locked outdoor pen (6.1 × 3 × 1.8 m) to 45 46 ously impacting local ecological systems (Engeman et ensure no python escaped. We erected a canopy over the 46 47 al. 2011; Dorcas et al. 2018). These giant constrictors testing pen to shelter the maze and surveillance cameras 47 48 are directly implicated in significant reductions in local from rain and direct sun. We conducted trials between 48 49 mammal populations and cause concern for multiple en- 1300 hours and 1900 hours from 18 April 2017 to 24 49 50 dangered species (Dorcas et al. 2012; Green et al. 2007; August 2017. 50 51 Snow et al. 2007; McCleery et al. 2015). Several factors 51

1 Mating in Burmese pythons in south Florida occurs scent deposition in the natural environment and to min- 1 2 during December–April when breeding aggregations are imize stress on the stimulus animals that may affect the 2 3 observed (Smith et al. 2016). Egg-laying is generally in quality of the trail. In the male versus female trials, the 3 4 May and June (Harvey et al. 2008). Ambient tempera- male scent trail was deposited first, then the female; a 4 5 tures in Gainesville, Florida are too cool for pythons to partition along the length of the base arm separated the 5 6 be active as early in the calendar year as in south Flor- male and female trails as they were deposited. The arm 6 7 ida, so we initiated trials during April when average containing the scent was randomized for every trial via 7 8 temperatures resemble those during December–Janu- coin toss. To start each trial, the trailing male was accli- 8 9 ary in the Everglades (U.S. Climate Data: https://uscli- mated for 30 min in the holding box at the base of the 9 10 matedata.com/climate/gainesville/florida/united-states/ maze and then allowed to enter. The trial was complet- 10 11 usfl0163). Pythons were fed according to appetite and ed when the male’s head entered a holding box in either 11 12 activity levels as per our IACUC protocol. Food was of- arm. 12 13 fered once per week (March to May) or twice per week We tested each male twice within each trial type and 13 14 (June to August). Test pythons (stimulus and focal) were assigned male test order randomly. The length of the in- 14 15 not fed on the day prior to their trial to avoid post-pran- tertrial period per male ranged from 3 to 28 days. For 15 16 dial lethargy, and food was then offered after testing was the female-only and male versus female trials, each 16 17 completed. male was matched with each of the 2 females. In the 17 18 Initial trials were bias tests with no scent in the male-only trials, each male was randomly matched with 18 19 Y-maze to determine if snakes showed a preference 2 of the other males. Individual male responses were 19 20 for either arm of the maze (3/6 chose the N arm; P then averaged per trial type. The order for the trial types 20 21 = 0.5). The 3 experimental trials were: (i) male-only was prioritized to run female-only first at the time when 21 22 (male scent trail in base and 1 arm; other blank); (ii) fe- Burmese pythons in the Everglades are most reproduc- 22 23 male-only (female scent trail); and (iii) male versus fe- tively active (December–April; Smith et al. 2016). The 23 24 male (both present in maze). For each trial type, the female-only trials were run from 18 April to 19 May 24 25 stimulus python providing the scent trail moved free- 2017, male-only was from 22 May to 9 June 2017, and 25 26 ly through the base of the Y-maze and only 1 arm (the male versus female from 15 June to 18 July 2017. This 26 27 other was blocked with a partition then removed for the order was also prioritized because of the technical chal- 27 28 trial) at their own pace without our intervention. We lenges in running the stimulus python through the maze 28 29 let stimulus snakes create trails in this way to simulate to deposit a scent trail, especially in the female-only and 29 30 male versus female trials. 30 31 31 For each trial, we covered the floor of the maze with 32 32 a plastic sheet and then a piece of clean white paper. Be- 33 33 tween trials, we replaced the paper and sheeting and dis- 34 34 assembled and washed the maze components and hold- 35 35 ing boxes with soap (Micro laboratory cleaner) and 36 36 water to eliminate residual odor cues. After the stimulus 37 37 python completed its run, we washed and dried the start 38 38 box and the partitions used in the maze and did so prior 39 39 to placing the next snake in the box. 40 40 41 Data collection 41 42 42 We monitored each trial with digital surveillance 43 43 cameras at 3 different angles: one facing the base arm 44 44 and holding box and one facing down each arm from 45 45 the Y-junction to the holding boxes. Video was recorded 46 46 to DVR, and video files were analyzed at James Mad- 47 47 ison University in Harrisonburg, Virginia to quanti- 48 48 fy male python behaviors: arm choice, choice penalty 49 49 score, tongue-flick rate (tongue-flicks per minute), paus- 50 Figure 1 Y-maze located in Gainesville, Florida used in the tri- 50 es, head raises, head shakes and turns (Table 1). Choice 51 als. 51

1 Table 1 List of behaviors recorded when male Burmese pythons followed conspecific scent trails in aY-maze 1 2 2 Behavior Description 3 3 4 Pause Stops moving for longer than 3 s while trailing 4 5 Head shake Moves head laterally, side-to-side approximately 5 cm; usually occurs as a rapid series of movements 5 6 Head raise Lifts head above the substrate and cranes upward 6 7 Turn Change in the direction of movement by 90° to either the left or right 7 8 8 9 9 10 10 11 11 12 12 13 was determined when the male’s head entered a hold- male-only trial during 7–24 August 2017 and obtained 13 14 ing box in either arm. Choice penalty scores were as- a different result (7/10 chose the female arm; P = 0.17). 14 15 signed by dividing each arm into five 30-cm segments In those trials, males consistently followed 1 female’s 15 16 and giving 1 negative point for each segment the male scent trail (5/5 trials; P < 0.001), and she was the larger 16 17 entered in the non-target arm (blank arm in male-only of the 2 females. Choice penalty scores differed across 17

18 and female-only trials; male arm in the male–female tri- the trials (F2,8 = 4.84, P = 0.042) and were most nega- 18 19 als). Tongue-flick rate was recorded as tongue-flicks per tive in the male-only (−3.0 ± 0.93 SEM) and male ver- 19 20 minute when visible in the recordings. Behaviors were sus female trials (−3.20 ± 0.78) compared to the fe- 20 21 counted and time between sequential behaviors (sec) male-only trials (−0.50 ± 0.27; q = 3.65, P = 0.033; q = 21 22 was recorded. 3.94, P = 0.055, respectively) (Fig. 3). Therefore, males 22 explored the non-target arm the least in the female-only 23 Data analysis 23 24 trials (male-only vs male–female, q = 0.29, P = 0.84). In 24 25 For choice data, we used binomial tests but did not all but 1 of the trials, the stimulus python(s) providing 25 26 statistically compare choice across the trial types; this the scent trail never defecated in the maze; therefore, the 26 27 analysis approach is common in Y-maze experiments scent trails were produced passively as pythons moved 27 28 (Parker & Mason 2011). Tongue-flick rate, choice pen- 28 29 alty score and individual behaviors were analyzed using 29 30 one-way repeated measures ANOVA followed by Bon- 30 31 ferroni-corrected t-tests. To establish ethograms linking 31 32 individual behaviors, a single behavior was considered Blank 32 Scented 33 to follow another if they occurred within 20 s of one *** 33 34 other. The frequency of each sequence was recorded. 34 35 Head shakes would often occur in series, so we record- 35 36 ed each as a single incidence to follow the guidelines of 36 37 Slater (1973) for describing behavioral sequences. 37 38 38 39 RESULTS 39 40 40 Male Burmese pythons (n = 6 for female scent trials, 41 y 41 42 n = 5 for male scent trials) showed significant trailing 42

43 preference for female scent trails (11/11 trials; P < 0.001) Male-onl vs female 43 Female-only 44 but did not follow male scent (4/10 trials; P = 0.37) in Male 44 the Y-maze (Fig. 2). In the male-only trials, 3/5 males 45 Figure 2 Arm choice in a Y-maze by male Burmese pythons. 45 chose the male arm, with 1 male choosing the male arm 46 Males (n = 5 in male-only and male vs female; n = 6 in fe- 46 47 in both trials. When given the simultaneous choice be- male-only) showed a significant preference for female scent 47 48 tween male and female scent trails, males did not prefer trails (P < 0.001) but not male scent trails compared to the un- 48 49 female scent (4/10 trials; P = 0.37). In those trials, 3/5 scented arm. Each male was run twice in each trial. When giv- 49 50 males chose the female scent trail, with 1 male choos- en the choice between male and female scented arms, the py- 50 51 ing the female trail both times. We also re-ran the fe- thons showed no preference (4/10 chose the female arm). 51

1 through the maze, as is common in all other snake spe- compared to male-only (4.7 ± 1.03; q = 4.65, P = 0.027) 1 2 cies tested in conspecific trailing (Parker and Mason and male versus female (5.4 ± 0.92; q = 3.66, P = 0.032) 2 3 2011). (male-only vs male-female, q = 0.98, P = 0.50). There 3 4 When trailing, males exhibited a higher rate of were no significant differences in turns F( 2,8 = 1.71, P = 4 5 tongue-flicking (RTF, tongue-flicks per min) to female 0.24) or head shakes (F2,8 = 0.91, P = 0.43). 5 6 scent in the female-only and male vs female trials com- Trailing behaviors occurred in predictable sequenc- 6 7 7 pared to male-only (F2,8 = 4.76, P = 0.043) (Fig. 4). es, and the dominant sequence was head shake to pause 8 Male RTF was higher in female-only (66.52 TFs/min ± to head raise as depicted in an ethogram (Fig. 6). Rela- 8 9 4.51 SEM) compared to male vs female (55.1 ± 5.48; q tive to male-only trials, additional behavior sequences 9 10 = 3.71, P = 0.031) and marginally higher than male-on- were seen in female-only and male vs female trials, with 10 11 ly (54.74 ± 5.61; q = 3.84, P = 0.061) (male-only vs many sequences increasing in frequency (Table 2). 11 12 male-female, q = 0.134, P = 0.92). Trailing times did not 12 13 13 differ across the trials (F2,8 = 2.23, P = 0.16; male-on- DISCUSSION 14 ly = 2.6 min ± 0.27; female-only = 4.23 ± 0.63; male-fe- 14 15 male = 4.7 ± 0.98). Our results show that male Burmese pythons recog- 15 16 16 Males displayed distinct, quantifiable behaviors (Ta- nize and follow scent trails from females in a Y-maze. 17 17 ble 1). Specifically, males showed significant variation Chemical and tactile information are of central impor- 18 18 in the number of head raises (F = 13.07, P = 0.003) tance to snakes, and their general reliance on other sen- 19 2,8 19 and pauses (F = 6.01, P = 0.025) (Fig. 5). Head raises sory information is believed to be limited (Ford 1995). 20 2,8 20 were most frequent in the male versus female trials (9.2 Male pythons showed increased chemosensory sampling 21 21 ± 1.83) compared to the male-only (3.4 ± 0.92; q = 7.22, of female scent trails via tongue-flicking which strongly 22 22 P = 0.002) and female-only trials (6.5 ± 1.51; q = 3.36, suggests that chemical communication is the mechanism 23 23 P = 0.045). Head raises were also higher in female-on- by which these snakes locate potential mates in the en- 24 24 ly compared to male-only (q = 3.86, P = 0.026). Males vironment, as is true of virtually all snake species stud- 25 25 paused most frequently in female-only (8.0 ± 0.79) ied (Ford 1986; Mason 1992; Mason & Parker 2010; 26 26 27 27 28 28 29 29 30 30

) 0 31 31 32 32 33 −2 33 34 MFM vs f 34

35 −4 35 36 36

37 Choice penalty (0 to −5 37 38 −6 38 39 39 40 y 40 y 41 41 Male-onl Female-only 42 Male-onl 42 Male vs female 43 Female-only 43 Figure 3 Male Burmese pythons (n = 5) explored the unscent- Male vs female 44 44 ed arm of the maze more thoroughly in the male-only tri- Figure 4 Male Burmese pythons (n = 5) had higher rates of 45 als than female-only; therefore, they received a more nega- chemosensory sampling (rates of tongue-flicking) when fol- 45 46 tive choice penalty score. When both male and female scent lowing female scent trails alone than when they were paired 46 47 trails were present, the males explored the male-scented arm with male scent trails. Inset: Trailing times did not differ across 47 48 (non-target) more extensively than in female-only. Bars repre- the trials. Bars represent means. Positive error is SEM, nega- 48 49 sent means. Positive error is SEM, negative is 95% confidence tive is 95% confidence interval. Uppercase letters are signif- 49 50 interval. Uppercase letters are significant differences P( < 0.05); icant differences (P < 0.05); lowercase letters are marginally 50 51 lowercase letters are marginally significant (0.05

1 1 2 2 3 3 4 4 5 5 6 6 7 7 8 8 9 Figure 5 Male Burmese pythons (n = 9 10 5) showed a variety of behaviors in the 10 11 Y-maze across all trials. Only pausing and 11 12 head raising showed significant variation 12 13 between trial type. Males paused most of- 13 14 ten when only female trails were present. 14 15 Head raising behavior was most frequent 15 16 when males discriminated between female 16 17 and male scent trails simultaneously. Bars 17 y y 18 represent means. Positive error is SEM, 18 negative is 95% confidence interval. Let- female 19 Male-onl Male-onl 19 ters are significant differences P( < 0.05). Female-only Female-only 20 Male vs female Male vs 20 21 21 22 Male-only Female-only Male vs female 22 23 23 24 24 25 Pause Shake Pause Shake Pause Shake 25 26 26 27 27 28 28 29 29 30 30 31 31 Head raise Turn Head raise Turn Head raise Turn 32 32 33 33 34 34 35 Figure 6 Kinematic diagram illustrating behavioral sequences of male pythons (n = 6) in different conspecific scent trailing scenar- 35 36 ios. Numbers indicate the frequency of a behavior following another (arrow points to subsequent behavior; size of arrow indicates 36 37 magnitude of frequency). Black arrows are behavioral sequences that increased in frequency compared to the male-only trials; light- 37 38 er gray arrows decreased. 38 39 39 40 Table 2 Differences in the frequencies of behavioral sequences from male Burmese pythons (n = 5) relative to male-only Y-maze 40 41 trials 41 42 2nd behavior 42 43 1st behavior Pause Head shake Head raise Turn 43 44 Male-only Male vs Male vs Male vs Male vs 44 Female-only Female-only Female-only Female-only 45 (reference) female female female female 45 46 Pause 0 0 −0.054 −0.13 0.001 0.049 0.053 0.081 46 47 Head shake −0.036 0.1 0.041 −0.05 −0.005 −0.05 0 0 47 48 Head raise 0.029 −0.2 −0.084 0.017 0.06 0.183 −0.005 0 48 49 Turn 0.143 0 −0.5 −0.389 0.071 0.167 0.286 0.222 49 50 The first column indicates first behavior, and each subsequent column represents the sequential behavior. Gray cells are behavior -se 50 51 quences that decreased relative to male-only. 51

1 Parker & Mason 2011). Based on our ethograms, we intuit that the increased 1 2 Male Burmese pythons following female scent interactions between behaviors in female-only trials re- 2 3 showed specific changes in their searching behaviors flects stronger male response to female scent. Behavior- 3 4 (head raises, pauses). We believe head raising is a gen- al sequence diagrams for other species reliably inform 4 5 eralized searching pattern where males are looking for researchers about qualities of an unknown chemical 5 6 either additional chemical signals or other potential trail, such as sex, species, female receptivity and male 6 7 cues from females (e.g. visual cues and movement). In fitness (Ford 1995). Sequences of courtship behavior 7 8 the wild, vertical exploration is possible for male Bur- have been described for 3 different rat snake species 8 9 mese pythons, and this species climbs trees and basks (genus Elaphe). The courtship behaviors were described 9 10 in arboreal habitats (Reed & Rodda 2009). Captive In- triphasically with the behaviors varying slightly across 10 11 dian rock pythons [Python sebae (Gmelin, 1788)] dis- each. Courtship sequences of the brown tree snake, the 11 12 played a “head up posture” when reproductively active, Indian python and the Burmese python have also been 12 13 with males lifting their heads 30–60 cm above the sub- similarly described in 3 phases, but behavioral sequenc- 13 14 strate (Walsh & Murphy 2003). It may be that the head es have not been established (Gillingham & Chambers 14 15 raises we observed are the same. Breeding males of oth- 1982; Greene & Mason 2000; Walsh & Murphy 2003). 15 16 er snake species exhibit periscoping behavior to search When male and female scent were present in the 16 17 for females, and in this behavior trailing males quickly trailing environment, male Burmese pythons performed 17 18 raise their heads and scan the horizon to possibly hone worse than if only female scent was present. Males in- 18 19 searching (garter snakes, Shine et al. 2005). It has also creased searching in the non-target arm (choice penal- 19 20 been suggested that head raises in Indian pythons [Py- ty score), did not discriminate between trails (lack of 20 21 thon molurus (Linnaeus, 1758)] function as sexual sig- choice and reduced tongue-flicking) and decreased their 21 22 nals (Barker et al. 1979). processing time (pausing). Either conflicting chemical 22 23 23 In contrast, pausing behavior in the pythons may stimuli from conspecifics impede proper mate trailing 24 24 serve to allow processing time at the neural level for as- or, more likely, our male versus female trials occurred 25 25 sessing chemical cues. In countless examples, snakes once males were no longer responsive to female scent. 26 26 pause when encountering a new chemical trail before In other snakes (e.g. garter snakes), males respond to fe- 27 27 exhibiting additional behaviors (reviewed in Mason male pheromones only during the breeding season then 28 28 1992). In an elegant example, male garter snakes deter- switch exclusively to prey cue detection (O’Donnell 29 29 mined a female’s movement vector by assessing the side et al. 2004). Males only become responsive to female 30 30 of a wooden post she used to propel herself in a maze. pheromones again after prolonged low-temperature dor- 31 31 The males rapidly tongue-flicked the sides of the post, mancy (e.g. Garstka et al. 1982). Conflicting stimuli 32 32 then propelled from the female side to follow the fe- significantly alter receiver detection and preference for a 33 33 male’s trajectory (Ford & Low 1984). signal (e.g. Thompson et al. 2008), but our experimental 34 timeline precludes us from further inference. 34 35 Head shakes and turns by male Burmese pythons are 35 Our current study is the first on Burmese python 36 not scent-trail specific, and we propose that these be- 36 chemical ecology; however, this species is not the only 37 haviors are general components of the searching reper- 37 invasive snake in the USA. In Florida alone, there are 38 toire of Burmese pythons. Head shakes may be useful 38 multiple established invasive snake species, including 39 for chemosensation by disrupting the chemical environ- 39 Boa constrictor Linnaeus, 1758 and P. sebae (Krysko 40 ment of the substrate to waft additional cues for detec- 40 et al. 2011). Reticulated pythons [Malayopython reticu- 41 tion and assessment. In Walsh and Murphy (2003), male 41 latus (Schneider, 1801)] and green anacondas (Eunect- 42 pythons (P. sebae) exhibited “jerky side-to-side mo- 42 es murinus Linnaeus, 1758) have also been recorded in 43 tion” along the females’ bodies during courtship, which 43 Florida but are not established. Boa constrictors are in- 44 are observed in many species of snakes during court- 44 vasive in Puerto Rico and several Caribbean islands also 45 ship (Carpenter 1977; Greene & Mason 2000; Senter 45 (Reed & Rodda 2011; Reynolds et al. 2013). In addition 46 et al. 2014). We observed head shakes regardless of the 46 to large constrictors, smaller snakes are invasive, such 47 scent present, so we cannot ascribe an alternate func- 47 as king snakes in the Canary Islands (Cabrera-Pérez et 48 tion. Turns were rarely observed compared to the oth- 48 al. 2012) and northern water snakes in California (Rose 49 er behaviors but were recorded based on previous work 49 et al. 2013). Ultimately, the threat posed by invasive 50 in rattlesnakes that used turning behavior as a proxy for 50 snakes is their propensity to shift diets rapidly in new 51 certainty while trailing (Parker & Kardong 2006; 2017). 51

1 environments and function as second-order if not apex snakes. American Zoologist 17, 217–23. 1 2 predators in novel habitats. Considerable research effort Dorcas M, Willson JD, Reed RN et al. (2012). Severe 2 3 is thus warranted to decipher the mechanisms that en- mammal declines coincide with proliferation of inva- 3 4 able snakes to thrive in their non-native habitats, espe- sive Burmese pythons in Everglades National Park. 4 5 cially the ways in which invasive snakes locate and se- Proceedings of the National Academy of Science USA 5 6 lect their mates. 109, 2418–22. 6 7 7 Dorcas M, Pittman S, Willson JD (2018). Burmese py- 8 8 ACKNOWLEDGMENTS thons. In: Pitt W, Beaseley J, Witmer G, eds. Ecology 9 9 and Management of Terrestrial Vertebrate Invasive 10 K. Keacher and W. Bruce provided animal care at the 10 Species in the United States. CRC Press, Boca Raton, 11 National Wildlife Research Center. Colleagues at the 11 FL, pp. 135–62. 12 Florida Fish and Wildlife Conservation Commission, 12 13 Miami-Dade Fire Rescue Venom Response Program, Engeman R, Jacobson E, Avery ML, Meshaka W (2011). 13 14 National Park Service, and University of Florida assist- The aggressive invasion of exotic reptiles in Florida 14 15 ed in obtaining the wild-caught pythons. K. Rush assist- with a focus on prominent species: A review. Current 15 16 ed with identifying and describing behaviors. All pro- Zoology 57, 599–612. 16 17 cedures involving the use of pythons was approved by Eriksson D, Wallin L (1986). Male bird song attracts fe- 17 18 the IACUC of the National Wildlife Research Center. males—A field experiment. Behavioral Ecology and 18 19 This research was supported through cooperative agree- Sociobiology 19, 297–9. 19 20 ments 14-7412-1061-CA, 15-7412-1155-CA, and 16- Ford N (1986). The role of pheromone trails in the so- 20 21 7412-1269-CA between James Madison University and ciobiology of snakes. In: Duvall D, Müller-Schwarze 21 22 the U.S. Department of Agriculture’s Animal and Plant D, Silverstein RM, eds. Chemical Signals in Verte- 22 23 Health Inspection Service (APHIS). brates, vol. 4. Springer, Boston, pp. 261–78. 23 24 Conflict of Interest Statement: The authors declare no Ford N (1995). Experimental design in studies of snake 24 25 competing interests. behavior. Herpetological Monographs 9, 130–39. 25 26 Ford N, Low J (1984). Sex pheromone source location 26 27 REFERENCES by garter snakes: A mechanism for detection of direc- 27 28 tion in nonvolatile trails. Journal of Chemical Ecolo- 28 Barker D, Murphy J, Smith K (1979). Social behavior 29 gy 10, 1193–9. 29 30 in a captive group of Indian pythons, Python molurus 30 Garstka W, Camazine B, Crews D (1982). Interactions 31 (Serpentes, Boidae) with formation of a linear social 31 of behavior and physiology during the annual repro- 32 hierarchy. Copeia 3, 466–71. 32 ductive cycle of the red-sided garter snake (Thamno- 33 Bryant G, Bateman P, Fleming P (2011). Tantalising 33 phis sirtalis parietalis). Herpetologica 38, 104–23. 34 tongues: Male carpet pythons use chemoreception to 34 35 differentiate among females. Australian Journal of Gillingham J, Chambers J (1982). Courtship and pel- 35 36 Zoology 59, 42–8. vic spur use in the Burmese python, Python molurus 36 bivittatus. Copeia 1982, 193–6. 37 Cabrera-Pérez M, Gallo-Barneto R, Esteve I, Pa- 37 38 tiño-Martínez C, López-Jurado L (2012). The man- Greene M, Mason RT (1998). Chemically mediated sex- 38 39 agement and control of the California kingsnake in ual behavior of the brown tree snake, Boiga irregu- 39 40 Gran Canaria (Canary Islands): Project LIFE+ Lam- laris. Ecoscience 5, 405–9. 40 41 propeltis. Aliens: The Invasive Species Bulletin 32, Greene M, Mason RT (2000). Courtship, mating, and 41 42 20–8. male combat of the brown tree snake, Boiga irregu- 42 43 43 Carmichael C, Kreiser B, Barker D, Barker T, Gilling- laris. Herpetologica 56, 166–75. 44 44 ham J (2007). Geographic variation in pheromone Greene M, Stark S, Mason RT (2001). Pheromone trail- 45 45 trailing behaviors of the Indonesian water python (Li- ing behavior of the brown tree snake, Boiga irregu- 46 46 asis mackloti) of Indonesia’s Lesser Sunda Archipel- laris. Journal of Chemical Ecology 27, 2193–201. 47 47 ago. In: Henderson R, Powell R, eds. Biology of the Harvey R, Brien M, Cherkiss M et al. (2008). Burmese 48 48 Boas and Pythons. Eagle Mountain Publishing, Eagle pythons in South Florida, scientific support for inva- 49 49 Mountain, Utah, pp. 227–40. sive species management. University of Florida IFAS 50 50 Publication Number, WEC-242, 8–17. 51 Carpenter C (1977). Communication and displays of 51

1 Krysko K, Burgess J, Rochford M et al. (2011). Veri- and substrate cues from non-envenomated mice 1 2 fied non-indigenous amphibians and reptiles in Flor- during rattlesnake (Crotalus oreganus) post-strike 2 3 ida from 1863 through 2010: Outlining the invasion trailing. Herpetologica 62, 349–56. 3 4 process and identifying invasion pathways and stag- Parker MR, Kardong KV (2017). Airborne chemical in- 4 5 es. Zootaxa 3028, 1–64. formation and context-dependent post-strike foraging 5 6 Lelito J, Fraser I, Mastro V, Tumlinson J, Baker T (2008). behavior in Pacific rattlesnakes Crotalus( oreganus). 6 7 Novel visual-cue-based sticky traps for monitoring Copeia 105, 649–56. 7 8 8 of emerald ash borers, Agrilus planipennis (Col., Bu- Parker MR, Mason RT (2011). Pheromones in snakes: 9 9 prestidae). Journal of Applied Entomology 132, 668– History, patterns and future research directions. In: 10 10 74. Sever D, Aldridge R, eds. Reproductive Biology and 11 11 Mason RT (1992). Reptilian pheromones. In: Gans C, Phylogeny of Snakes. CRC Press, Boca Raton, FL, 12 12 Crews D, eds. Biology of the Reptilia, vol 18. Univer- pp. 551–72. 13 13 sity of Chicago Press, Chicago, pp. 115–206. 14 Rasmussen LE, Schmidt M, Henneous R, Groves D, 14 15 Mason RT, Parker MR (2010). Social behavior and Daves G (1982). Asian bull elephants: Flehmen-like 15 16 pheromonal communication in reptiles. Journal of responses to extractable components in female ele- 16 17 Comparative Physiology A, Neuroethology, Sensory, phant estrous urine. Science 217, 159–62. 17 18 Neural, and Behavioral Physiology 196, 729–49. Reed RN, Rodda G (2009). Giant constrictors: Biolog- 18 19 Mason RT, Fales H, Jones T, Pannell L, Chinn J, Crews ical and management profiles and an establishment 19 20 D (1989). Sex pheromones in snakes. Science 245, risk assessment for nine large species of pythons, an- 20 21 290–93. acondas, and the boa constrictor, U.S. Geological 21 22 Mason RT, Chivers D, Mathis A, Blaustein A (1998). Survey Open-File Report 2009-1202. U.S. Geologi- 22 23 Bioassay methods for amphibians and reptiles. In: cal Survey, Reston, VA. 23 24 Haynes KF, Millar JG eds. Bioassay Methods, vol 2. Reed RN, Rodda G (2011). Burmese pythons and oth- 24 25 Springer, Boston, pp. 71–325. er giant constrictors. In: Encyclopedia of Biological 25 26 Mathies T, Levine B, Engeman R, Savidge J (2013). Invasions. University of California Press, Berkeley, 26 27 Pheromonal control of the invasive brown treesnake: CA, pp. 85–91. 27 28 Potency of female sexual attractiveness pheromone Reed RN, Willson JD, Rodda G, Dorcas M (2012). Eco- 28 29 varies with ovarian state. International Journal of logical correlates of invasion impact for Burmese py- 29 30 Pest Management 59, 141–9. thons in Florida. Integrative Zoology 7, 254–70. 30 31 McCleery R, Sovie A, Reed RN, Cunningham MW, Reynolds G, Puente-Rolón A, Reed RN, Revell L (2013). 31 32 Hunter M, Hart KM (2015). Marsh rabbit mortalities Genetic analysis of a novel invasion of Puerto Rico 32 33 tie pythons to the precipitous decline of mammals in by an exotic constricting snake. Biological Invasions 33 34 the Everglades. Proceedings. Biological sciences/The 15, 953–9. 34 35 35 Royal Society 282, 20150120. Rhen T, Crews D (2000). Organization and activation of 36 36 Moynan C, Neumann C, Welsh C (2016). The effect of sexual and agonistic behavior in the leopard gecko, 37 37 gender, tone, and sound location on the response be- Eublepharis macularius. Neuroendocrinology 71, 38 38 havior of Neogobius melanostomus (Round goby) 252–61. 39 39 and the possibility of future trapping of this invasive 40 Rose J, Miano O, Todd B (2013). Trapping efficiency, 40 species in Lake Superior. Zebrafish 13, 287–92. 41 demography, and density of an introduced population 41 42 Nagamoto J, Aonuma H, Hisada M (2005). Discrimi- of Northern watersnakes, Nerodia sipedon, in Cali- 42 43 nation of conspecific individuals via cuticular phero- fornia. Journal of Herpetology 47, 421–7. 43 44 mones by males of the cricket Gryllus bimaculatus. Senter P, Harris S, Kent D (2014). Phylogeny of court- 44 45 Zoological Science 22, 1079–88. ship and male-male combat behavior in snakes. PLoS 45 46 O’Donnell R, Shine R, Mason RT (2004). Seasonal an- ONE 9, e107528. 46 47 orexia in the male red-sided garter snake, Thamno- Shine R, Phillips B, Waye H, LeMaster M, Mason RT 47 48 phis sirtalis parietalis. Behavioral Ecology and So- (2003). Chemosensory cues allow courting male gar- 48 49 ciobiology 56, 413–9. ter snakes to assess body length and body condition 49 50 Parker MR, Kardong KV (2006). The role of airborne of potential mates. Behavioral Ecology and Sociobi- 50 51 51

1 ology 5, 162–6. Florida. Herpetological Bulletin 101, 5–7. 1 2 Shine R, O’Donnell R, Langkilde T, Wall M, Mason RT Takács S, Gries R, Gries G (2017). Sex hormones func- 2 3 (2005). Snakes in search of sex: the relationship be- tion as sex attractant pheromones in house mice and 3 4 tween mate-locating ability and mating success in brown rats. ChemBioChem 18, 1391–5. 4 5 5 male garter snakes. Animal Behavior 69, 1251–8. Thompson J, Bissell A, Martins E (2008). Inhibitory in- 6 6 Slater P (1973). Describing sequences of behavior. In: teractions between multimodal behavioural responses 7 7 Bateson P, Klopfer P eds. Perspectives in Ethology. may influence the evolution of complex signals. Ani- 8 8 Springer, Boston, pp. 131–53. mal Behavior 76, 113–21. 9 9 10 Smith B, Cherkiss M, Hart K et al. (2016). Betrayal: ra- Walsh T, Murphy J (2003). Observations on the hus- 10 11 dio-tagged Burmese pythons reveal locations of con- bandry, breeding and behaviour of the Indian python. 11 12 specifics in Everglades National Park. Biological In- International Zoo Yearbook 38, 145–52. 12 13 vasions 18, 3239–50. Witzgall P, Kirsch P, Cork A (2010). Sex pheromones 13 14 Snow R, Brien M, Cherkiss M, Wilkins L, Mazzotti F and their impact on pest management. Journal of 14 15 (2007). Dietary habits of Burmese python, Python Chemical Ecology 36, 80–100. 15 16 molurus bivittatus, from Everglades National Park, 16 17 17 18 18 19 Cite this article as: 19

20 20 21 Richard SA, Tillman EA, Humphrey JS, Avery ML, Parker MR (2019). Male Burmese pythons follow female 21 22 scent trails and show sex-specific behaviors.Integrative Zoology 14, 460–9. 22 23 23 24 24 25 25 26 26 27 27 28 28 29 29 30 30 31 31 32 32 33 33 34 34 35 35 36 36 37 37 38 38 39 39 40 40 41 41 42 42 43 43 44 44 45 45 46 46 47 47 48 48 49 49 50 50 51 51