|Home | About | Journals | Submit | Contact Us | Français|
Territorial animals lay scent marks around their territories to broadcast their presence, but these olfactory signals can both attract and repel con-specifics. Attraction or aversion can have a profound impact in terms of space use and thereby influence an individual’s access to resources and mates. Here, we test the impact of chemical signals on the long-term space use and activity of receivers, comparing the response of males and females, territory holders, and temporary visitors in Sceloporus undulatus lizards in the field. We placed either male femoral gland secretions (chemical) or blank (control) cues on resident male landmarks, repeatedly over 5 d, while monitoring the activity and location of all lizards in the vicinity. We found that resident males and females, but not non-resident males, were active on more days near landmarks treated with chemical cues than landmarks treated with control cues. Non-resident males remained closer to chemical than control cues. These results suggest that territorial scent marks are attractive to conspecifics and impact space use, but that the specific effects depend on receiver sex and residency status. Such subtle or gradual changes in behavior may frequently be overlooked by short-term choice experiments. Future studies investigating the behavioral significance of a communicative signal should consider these finer details of behavior for a more comprehensive assessment.
In an apparent paradox, chemical signals used in territorial interactions both attract and repel conspecifics (see reviews by Stamps 1988; Mason & Parker 2010; Wyatt 2014). Conspecifics often form overlapping territories, and deposit odor signals on landmarks that mediate territorial and social interactions (see review by Heap et al. 2012). Although proposed to minimize costly fighting by reducing encounters between competitors, territorial signals can also attract conspecifics. These chemical signals contain information about signaler age (Amo et al. 2012), sex (Delbarco-Trillo et al. 2012), diet (Chouinard 2012), breeding condition (Crawford et al. 2011; Saveer et al. 2012), or size (Blaul & Ruther 2012; Ibáñez et al. 2012; Scott et al. 2013). Odor cues are then used by receivers to recognize familiar individuals (e.g., Whittaker et al. 2011; Kulahci et al. 2014) or gauge receiver physiology (e.g., Mathis 1990). Thus, the paradox arises in the specific behavioral response to an odor cue and whether that cue repels or attracts potential competitors, mates, or both. Attraction or aversion can have a profound impact on animal space use and thereby influence an individual’s access to resources and mates (e.g., Mourier et al. 2012; Green et al. 2015). Here, we test the impact of chemical signals on the space use and activity of receivers, comparing the long-term response of males and females, territory holders, and temporary visitors in Sceloporus undulatus lizards.
Olfactory signals differ from other forms of communication because they can continue to broadcast information for hours in the absence of the signaler, and can thus have a lasting impact on space use (Chivers et al. 2013; Van Buskirk et al. 2014; Campbell & Roberts 2015). Like other signal types, chemical cues attract (Durisko & Dukas 2013; Bett & Hinch 2015), repel (Goodale & Nieh 2012; Ibáñez et al. 2012), and immobilize conspecifics (Brechbühl et al. 2008; Fraker et al. 2009). However, chemical signals are more dynamic in that the complex mixture of components changes with time as more volatile compounds dissipate rapidly or exhibit time-released behavior (Wyatt 2014). In this way, the signal being emitted also changes, effectively transmitting new or different information that in turn may elicit a different response in the same receiver over time. Furthermore, the same odor signal can elicit context-dependent spatial responses in the same receiver (Saleh & Chittka 2006). For example, learning of particular odor cues in one context can have a significant impact on future behavior, as in salmon which use odor cues to return to spawning sites (Bett & Hinch 2015) or mice in which chemical cues facilitate learning of spatial preferences (Roberts et al. 2012). Thus, although choice tests that measure immediate responses to chemical cues in a laboratory context (e.g., Whittaker et al. 2011; Amo et al. 2012) provide important insights about the spatial ecology of an organism, they may also underestimate the impact of chemical cues on behavior in the wild. Here, we test the impact of chemical signals on the activity and space use of con-specifics after 24 h of exposure to the stimulus, and repeatedly over several days.
Individuals can also differ in their response to chemical cues because of differences in their internal, or physiological, states (see review by Nathan et al. 2008). Territory holders fight to maintain access to specific landmarks and actively patrol to exclude temporary visitors from favored areas (Steingrímsson & Grant 2011; Heap et al. 2012). Different physiological constraints due to various reproductive requirements or strategies may partly explain space use (Bischof et al. 2012; Ophir et al. 2012) as selection should favor receivers that employ space use strategies complementary to individual reproductive or survival strategies. For example, in some species, residents able to hold territories gain reproductive advantages over non-residents (Bergman et al. 2007; Kresnik & Stutchbury 2014), and different levels of motivation among these competitors can lead to predictable contest outcomes (Bergman et al. 2010). Territorial dimensions can shift with the availability of landmarks (Heap et al. 2012; Suriyampola & Eason 2014), and threats produced by non-residents in more highly valued areas of a home range can be more likely to elicit resident responses relative to other areas (Furrer et al. 2011). Additionally, seasonal changes in receiver physiology alter sensory sensitivity (Saveer et al. 2012; Dey et al. 2015). While a territory holder may respond aggressively to a chemical signal perceived as coming from an intruder, that same signal may be attractive to the animal in breeding conditions. The same receiver may display a modified response due to altered gene expression following experience (Immonen & Ritchie 2012), or due to changes in the animal host’s gastrointestinal microbial community shown to influence brain processes linked to mate preferences (Ezenwa et al. 2012; Foster & McVey Neufeld 2013). Furthermore, sex differences in physiology can predict both time spent producing chemosensory behavior (Clapham et al. 2014) and odor preferences (Dantzer & Jaeger 2007; Whittaker et al. 2011, 2013; Baird et al. 2015). In this study, we compare the space use responses of territory holders to temporary visitors, and females to males, in response to conspecific odor cues.
Here, we conduct experiments with eastern fence lizards, S. undulatus, in the field. Small territory sizes, accessible and open habitat sites, and low vagility make lizards an excellent system for studying territorial space use (Fox et al. 2003). Lizards in the genus Sceloporus often have overlapping home ranges, and both males and females can defend territories with some exclusive use during the breeding season (Ferner 1974; Sheldahl & Martins 2000). As Sceloporus lizards are primarily sit-and-wait predators that do not actively forage, much of their movement patterns can be attributed to thermoregulatory and territorial behavior. During the breeding season, males boost production of chemical signals (femoral gland secretions; FGSs) that are secreted from specialized glands through pores on the inner thighs, which elicit aggressive territorial responses in conspecifics (Duvall 1979; Hews et al. 2011) and are used to scent-mark territories (Mayerl et al. 2015). S. undulatus exhibit site fidelity across years (Ferner 1974), associate with specific landmarks, and favor particular perching structures (Jones & Droge, 1980). Collectively, these characteristics make S. undulatus an excellent candidate for asking whether chemical signals are associated with territorial space use, and whether receiver sex or residency status matters. If territorial lizards are attracted to conspecific chemical signals over time, we predict that (1) lizards will be active on more days near landmarks treated with chemical cues (conspecific FGSs) than landmarks treated with control (plain) cues or (2) active lizards will perch closer to chemical cues than to control cues. However, if territorial lizards are repelled by conspecific chemical signals over time, we expect to find that lizards (1) will be active on fewer days near chemical landmarks compared to control landmarks or (2) will perch at greater distances from chemical cues relative to control cues. Alternatively, we could find a difference between (1) resident and non-resident males or (2) between males and females, suggesting that either residency status or sex, respectively, is associated with movement responses to conspecific odors.
We studied S. undulatus lizards on large rocks, the base of trees, and logs along the shoreline of Lake Monroe, Indiana, an artificial lake near a deciduous forest. This site has been used for other previous studies on chemical behavior in lizards (Hews et al. 2011). S. undulatus lizards form clumped spatial distributions, and most lizard activity takes place in and near rocky outcrops (Jones & Droge, 1980) between Apr. and Oct. in the northern ends of their range (Angilletta 2001). We conducted our study from late Jun. to early Aug. in 2013 on two adjacent field sites, stretching 3.3 km and 2.6 km along the southern shore of the lake. Cue donors were collected, and behavioral trials were conducted at both sites. During this time of year, lizards are at their most active and can be found easily from 9:00 to 18:00 EST, often with a break in activity during the hottest hours of the day. All lizards used for this study were adults.
We began the study by capturing (by handheld noose) and individually paint-marking all adult males in both study areas (total = 41 males) with unique combinations of red, green, or white dots on the dorsal surface (base of tail or between the shoulders) using Elmer’s Painters® non-toxic paint markers. We marked all females with a single white dot on the dorsum. We also recorded the location of each captured animal on a physical map of the area, recording any major landmarks (e.g., log, tree, large rock) near the capture site. For 6 wk, we repeatedly surveyed both sites, recording any repeated sightings of the captured animals, and identifying ‘residents’ as adult males that were observed near the same landmark on at least three occasions. Note that resident males were distinguished from non-resident males on the basis of site fidelity alone. Although most of these residents are likely to be territory holders, territoriality in Sceloporus lizards is complex and does not always imply exclusive use (Sheldahl & Martins 2000). Although male S. undulatus respond aggressively to territorial intruders (Quinn & Hews 2010), they also share over half of their home ranges with other males and increase movement during breeding season, whereas females share only 9.5% of their home ranges with other females and likely reduce movement during the breeding season (Ferner 1974).
After most of the animals in both study areas had been marked, we began behavioral trials. Each day, we began by capturing an adult male cue donor from one of the two study areas. We gently rubbed the femoral pores of each cue donor along a 2.5 cm by 2.5 cm piece of black construction paper, using fresh nitrile gloves to handle the cue and placing the cue in a fresh plastic bag for temporary storage. Male cue donors ranged in size from 57 to 70 mm snout-to-vent length. At the same time, we created and placed a matching control cue (construction paper with no lizard scent, also created while wearing fresh nitrile gloves) in a second plastic bag. To minimize familiarity effects, we then moved at least 1 km from the location the cue donor was captured before placing the cue.
We began behavioral trials by identifying an active resident male that had not been recently captured (within the previous 48 h). We placed a chemical cue at the landmark (e.g., rock, log, base of tree) that had been used to define that male as a resident, wedging the cue between rocks or pieces of bark to hold it in place. The next subject received a control cue, and so on, in a counterbalanced design. In total, we tested the responses of 22 resident male lizards. Eleven of these received chemical cues, and 11 received control cues. As the act of approaching lizards to place treatment cues always caused the focal lizards to move or hide away, we did not conduct further observations for at least 1 d after placing the cues.
Each trial proceeded for 5 d, consecutive when weather permitted. Sceloporus lizards at this site are often inactive, and we chose 5 d as a minimum estimate of the time needed to ensure that every resident lizard was active on at least 1 d. On each of those days (starting 24 h after placing the cue), two observers slowly circled the landmark of each resident male, moving in opposite directions along the perimeter of a home range centered on that landmark (approximately 18 m from the landmark: Ferner 1974). While circling, the two observers scanned the area with binoculars and recorded the identity and sex of all active lizards in the area, as well as the distance of each lizard from the cue at the moment at which it was first spotted using a retractable tape measure. Distance was recorded as the distance from the center of the cue paper to the location of the lizard’s head (where chemosensory receptors are located). Lizards that were not observed during these surveys may have been inactive (e.g., buried under the substrate) or wandering far from the area in which they were originally sighted. After confirming the identities of any present and active lizards, we replaced the cue with a fresh stimulus (chemical or control) placed in the same location. Note that we obtained chemical cues from different lizards on each day, such that the chemical treatment at each landmark consisted of secretions from five different males over the 5-d period. No cue donor had behavior recorded until at least 48 h following handling.
To determine whether chemical signals attract or deter conspecifics, we used chi-squared tests to compare the total number of (1) resident male lizards that we observed near the landmark on at least one of the 5 d during the behavioral test in chemical or control treatments, as well as the total number of days on which we observed resident males. In addition, we calculated the number of days in which each resident male was near the landmark and used an unpaired t-test to compare males presented with chemical vs. control treatments. We then conducted similar analyses to compare the activity of (2) females and (3) nonresident males. Next, we calculated the median distance of each category of observed lizard from the cue and then used nonparametric Wilcoxon rank-sum tests to compare the mean distance of (1) resident males, (2) females, and (3) non-resident males for chemical vs. control cues. We obtained very similar results if we considered instead the mean or minimum distance from the cue, or the distance from the cue on the first sighting of each lizard. We conducted all statistical analyses using the base functions of R (R Development Core Team 2014).
Resident male lizards were more likely to be present and active near landmarks treated with chemical cues than in control treatments. Residents in the chemical cue treatments were active on more days (up to 4 d, total of 20 of 55 d summed across 11 males) than were residents in the control group (up to 3 d, total of 8 of 55 d summed across 11 males), a difference that was statistically significant (χ21 = 6.90, p = 0.009; Fig. 1a). Only three of the 11 resident males in the chemical treatment group were not observed at all during the 5-day trials, as compared to six of the 11 residents in the control group, although this difference was not statistically significant (χ21 = 1.69, p = 0.2). Resident males presented with chemical cues were active on approximately twice as many days as were males presented with control cues (one-tailed t = 2.0, df = 18, p = 0.03).
Similarly, females were more likely to be present and active near landmarks receiving chemical rather than control treatments (Fig. 1b). Females in the chemical cue treatments were also present and active on more days (up to 4 d, total of 18 of 55 d) than were females in the control group (up to 3 d, total of 8 of 55 d) (χ21 = 5.04, p = 0.02; Fig. 1b). Females were never present at only two of the 11 landmarks treated with chemical cues, as compared to seven of the 11 landmarks in the control group (χ21 = 4.7, p = 0.03).
In contrast, non-resident males did not differ in activity near chemical as opposed to control cues (Fig. 1c). At least one non-resident male was active in the vicinity of the landmark on 7 of 55 d for chemical and 8 of 55 for control trials (χ21 = 0.08, p = 0.78). There were never non-resident males at six of the 11 landmarks that were treated with chemical cues, and at six of the 11 landmarks that received control cues (χ21 = 0, p = 1; Fig. 1c).
Multiple lizards were observed near the same landmark on the same day on a total of 12 d. For the chemical treatment group, we observed multiple lizards near four distinct landmarks on a total of 8 d. On six of those days, we observed a resident male with a single female. On the remaining 2 d, we observed non-resident males with females (once with a single female and once with two females). We never observed a resident male and a non-resident male near the same chemically treated landmark on the same day. In contrast, we observed multiple lizards near three distinct control landmarks on a total of 4 d, and the most common occurrence near control landmarks was a non-resident male with a female (3 d). We also observed a non-resident male with a resident male once. We never observed resident males near the same control landmark as a female.
All categories of lizards remained closer to landmarks with chemical cues than to control landmarks (Fig. 2). We found resident males (Fig. 2a) approximately 0.7 m from the landmark in chemical trials, but nearly three times as far from the landmark (2.1 m) in control trials. Lizards were highly variable, though, such that this difference was not statistically significant when tested using Wilcoxon rank-sum test (W = 19, p = 0.88).
Females (Fig. 2b) remained somewhat closer to the landmark than did males, especially when it had a chemical cue (1.2 m) than when it was a control (2.9 m). However, again, this difference between treatment groups was not statistically significant when tested using Wilcoxon rank-sum test (W = 13; p = 0.44).
In contrast, non-resident males (Fig. 2c) ranged widely from the control landmarks (8.9 m), although they remained roughly the same distance from the chemical cues (0.6 m) as did resident males. This difference between chemical and control group distance was statistically significant for non-resident males using Wilcoxon rank-sum test (W = 8, p = 0.04).
Although we never observed lizards interacting directly with the treatment cues in our study, our results suggest that territorial S. undulatus lizards are attracted to conspecific chemical cues and that this attraction has a real impact on long-term space use. We found that resident males and females were present and active more often near landmarks treated with chemical cues vs. control cues and that non-resident males remained closer to scented landmarks than to control landmarks. Thus, our results further support the paradoxical notion that even solitary territorial animals are drawn to conspecific chemical signals, displaying semigregarious tendencies (Stamps 1988; Arnold et al. 2011; Mourier et al. 2012).
Our study also highlights the long-term impact of chemical cues on behavior. Although short-term playback experiments have been invaluable in advancing our understanding of animal communication, lack of a response in those experiments may not always indicate an inability to discriminate or to respond appropriately over a longer time frame. As in tests of mate preference (see reviews by Edward 2014; Dougherty & Shuker 2015; Reinhold & Schielzeth 2015), response to short-term stimulus presentations can differ depending on time frame, social context, and many other elements of the experimental design. Long-term and indirect effects may be especially important with chemical signals which can retain some effect hours after deposition (e.g., Martín & López 2013a). Sceloporus graciosus lizards replenish chemical secretions weekly (Martins et al. 2006), suggesting that these signals may persist for even longer time periods. Here, we found profound changes in space use by S. undulatus males presented with chemical cues near the same site in which Hews et al. (2011) found no effect of chemical cues on behavior during the 5-min field trials. Collectively, these results emphasize that even when attraction to chemical signals is not detected in immediate behavioral measures, it can manifest gradually over time and in more subtle ways.
Although all lizards were attracted to chemical signals in our study, the form of that attraction depended on receiver identity. Resident males were more likely to remain active near landmarks scented with chemicals from a novel male, perhaps in an attempt to increase the likelihood of a direct encounter with that male. Previous studies show that territorial responses to intruder chemical signals likely depend on the competitive quality or familiarity of the signaler relative to and as assessed by the receiver (Aragón et al. 2001; Martín & López 2007). Here, however, we find that resident lizards were attracted to chemical signals from different males deposited on consecutive days, regardless of individual differences in signal composition. Since we did not include a pungent control in this study, it is possible that this response may simply be to a new stimulus rather than lizard-specific odors. However, in the present study day-old control cues were replaced with new control cues just as day-old chemical cues were replaced with new chemical cues. Furthermore, other species of Sceloporus respond differently to conspecific chemical signals than to cologne or distilled water controls (Duvall 1981). While possible that lizards had interacted previously, we minimized familiarity between cue donors and residents by maximizing the distance between landmark and cue donor locations. Thus, a high degree of familiarity between signalers and receivers is unlikely. As shown by Stamps & Krishnan (1995), a transfer of territorial space between residents and non-residents transpires when one individual is persistent in its pursuit of that space. Thus, residents may be ensuring their territorial boundaries by remaining near the contested location. Males may also be increasing mating opportunities because females were also more likely to be active near the scented landmarks.
There is little evidence of female choice shaping the evolution of male signals in lizards (e.g., Ord et al. 2015), and it seems likely that female lizards choose among territories rather than male phenotypes (but see Swierk et al. 2012). Female attraction to the chemical cues in our study may reflect female attraction to novel males or to the combined signals of more than one male, as has been shown for Iberian rock lizards (Martín & López 2013b). Simply remaining active in the presence of male odor cues may also aid females by providing an opportunity to accelerate reproduction by increasing the amount of courtship they receive (Kelso & Martins 2008; Ruiz et al. 2010). Alternatively, similar spatial responses to odor cues in females as in resident males, and in contrast to non-resident males, may be because females were also territorial residents. Female lizards may be defending territories of their own or in concert with resident males, as they do in other Sceloporus species (Martins 1993; Sheldahl & Martins 2000).
While we did not find a difference in activity between chemical and control cues for non-resident males, those non-resident males that were active remained much closer to chemical than control cues. One potential explanation for non-residents to remain near chemical cues is for scent-matching purposes (Gosling & McKay 1990). Little is known about the duration of Sceloporus FGSs or the efficacy of these signals over time (but see Martins et al. 2006). Smaller, lighter volatile compounds are detected with olfaction, potentially from some distance, while heavier information-containing compounds detected via vomerolfaction require that lizards ‘taste’ the chemical deposit or surrounding air (Vitt et al. 2003). Thus, it is likely that Sceloporus lizards must approach FGSs to assess composition thoroughly and to detect heavier proteins or lipids. Non-resident satellite or competitor male receivers could compare chemical information to visual information gathered from an approaching male, for example, allowing for assessment of residency status and informing decisions about how to spatially respond. In the absence of a resident, non-residents may stick around and attempt to confiscate the landmark. Additionally, we never observed a non-resident and resident male active simultaneously near a chemical cue and only once near a control cue, and non-residents were equally likely to show up near chemical as control cues. Collectively, these results suggest non-residents may have been passing over the landmarks and stopped because they did not see a resident male, and then, non-residents were drawn toward chemical cues. Alternatively, non-residents observed in this study may share home range space with residents or otherwise defend neighboring territories. If non-residents recognized the cue donor scent as unfamiliar, remaining nearby the cue could function as a form of vigilance to reduce the potential costs of re-establishing territorial boundaries with new neighbors. Similar behavior has been demonstrated in other species (Detto et al. 2010). Regardless of the true identity of non-residents, our results demonstrate a gradual physical attraction to male FGS deposits.
We restrict our understanding of animal communication by neglecting to measure more gradual behavioral nuances that are only apparent after an extended period or in the finer details of animal behavior. Gradual or subtle adjustments to behavior may be frequently overlooked in common choice tests, despite having important consequences for the behavioral ecology of the animal. Thus, we underestimate the role of these social signals in regulating or influencing behavior. Future studies investigating the behavioral significance of social signals must consider both immediate and gradual changes in behavior.
We thank Christian King for assistance in the field, as well as Jake Pruett and Diana Hews for helpful discussions. Johanel Caceres, Jesualdo Fuentes-G, Jay Goldberg, Alison Ossip-Klein, Delia Shelton, Piyumika Suriyampola, Delawrence Sykes, and Jaime Zúñiga-Vega provided useful comments on earlier versions of this manuscript. This material is based on work conducted by EPM while serving at the National Science Foundation and supported by the National Science Foundation through grant IOS-1050274 (to EPM).