Mental processing of environmental and social information and sensory input in salamanders.
There are approximately 8,000 species of known amphibians, about 500 of which are salamanders (Frost 2019). Amphibians arose about 360 million years ago with the oldest salamander fossil dating to the Miocene (Petranka 1998). Members of the class Amphibia are ectothermic tetrapods, lack scales, and have heavily glandular skin. Salamanders are unique from frogs and caecilians, most obviously, by always having a tail and forelimbs; a few salamander species lack hind limbs. Many species have biphasic life cycles where they spend the first portion of their lives in water and their adult life on land; however, there are some species that are solely terrestrial and solely aquatic (Petranka 1998). Modes of respiration include gills, lungs, and gas exchange through the skin (i.e., cutaneous gas exchange). Some groups and species (e.g., Proteidae, Eurycea spp.) are permanently aquatic, remaining in a somewhat larval form through life, and have conspicuous external gills. Ambystoma tigrinum heavily rely on gas exchange through lungs even as fully aquatic neonates. Adult A. tigrinum breathe through their nares or mouth using a two-stroke buccal pump (a method of ventilation where the animal moves the floor of its mouth in a rhythmic manner that forces air into the lungs). However, Plethodontidae lack lungs entirely and rely on cutaneous gas exchange.
Larval and adult-form (i.e., post-metamorphic) salamanders are carnivorous, with a diet consisting mostly of insects and other invertebrates. The largest species sometimes eat fish or small animals such as frogs and mice (Petranka 1998). Salamanders of the family Sirenidae (Siren, Pseudobranchus) also eat plant and algal material (Hill et al. 2015). Salamanders are prey to a variety of animals including fish, snakes, small mammals, birds, and invertebrates (Petranka 1998).
Cognitive research on all amphibians has been limited, especially compared to mammals and birds (Agrillo 2014; Kundey et al. 2016). Cognition is inarguably a broad field, so here we will mainly focus on how information from an organism’s environment may be used to modulate the organism’s behavior. For example, quantitative discrimination is often used as a measure of cognitive ability, in contrast to nonnumerical discrimination (Agrillo 2014). Nonnumerical mechanisms that organisms may be using to discriminate quantities include the surface area of an object, density, amount of movement, element configuration, and/or overall brightness. For example, salamanders are able to discriminate between groups of live prey by choosing the larger group, but performance dropped to the chance level when the movement of prey was controlled for (Krusche et al. 2010). Concepts of cognition also extend to include mating behavior, recognition of individuals, territoriality, and foraging success, among others. By experimentally testing these behaviors in the lab and in the field, researchers are better equipped to infer which behaviors are instilled in organisms by instinct and what may be the result of learning from their environment.
Herrick (1948) provided a detailed description of the tiger salamander brain (Ambystoma tigrinum), which is an appropriate model for general brain structure of salamanders. The brain is divided into three main parts: the striato-amygdaloid complex, the hippocampus, and a large olfactory bulb. While brain function is not well studied in salamanders, the large olfactory bulb clearly is associated with the strong influence of pheromonal and other olfactory information on virtually all aspects of salamander natural histories. Apparently in anticipation of spring season foraging and mating, chemosensory cells and cells in the telencephalon portion of the brain, which contains the olfactory bulb, proliferate in number during this season (Dawley et al. 2000).
Roth (1987), with special attention to the visual system, also provided morphological descriptions of the brain. He also found that the visual perception and recognition of prey could be a useful tool for studying cognition, as he found that the behavioral preferences of the salamanders for specific prey depended upon previous experience with those prey. He further generalized that movement by prey items is crucial to prey recognition in salamanders and coalesced these findings to posit a functional model of a neural “recognition module” concept.
Cognitive ecology encompasses the interface of cognitive phenomena with the natural and social environments and experiences of an animal. Cognitive ecology is not itself a cognitive phenomenon, like memory may be. Rather, cognitive ecology is a framework to interpret the basic natural history and behavioral functions of an animal, from which we may infer some essential cognitive abilities. By considering the effects of information processing on animal fitness, the specifics of a species’ cognitive ecology may be assessed. An example from feeding, which certainly is a direct influence upon fitness, prior to experience with prey appears to influence feeding behaviors. For example, naive Plethodon cinereus offered either dead or live fruit flies consumed more flies per minute on their second trial compared to their first, which is indicative of learning (David and Jaeger 1981). Additionally, adult Plethodon cinereus capture novel prey more quickly than do neonate or yearlings (Gibbons et al. 2003). Expanding beyond foraging behaviors, and more related to shelter and perhaps homing, salamanders (Ambystoma tigrinum) trained to either go left or right in a maze toward a reward of a moist microhabitat continued to go the direction in which they were trained even after the maze was rotated 180° (Kundey et al. 2016). This suggests that the salamanders preferentially used a learned turning response over visual cues.
Courtship and Reproduction
Courtship is highly variable among salamanders, with few commonalities. In many species, especially of ambystomatids, plethodontids, and salamandrids, courtship involves complex series of ritualized visual, physical, and pheromonal phenomena (Petranka 1998). For example, in some cases, the male deposits a spermatophore on the substrate and guides the female into the appropriate position so that she can take it up into the cloaca. In the best studied plethodontid salamanders, complex, coordinated, and species-specific courtship rituals include tail-straddle walks, physical contact, and body-stroking as part of the sequence to guide the female into position over the spermatophore. With no evidence to indicate learning, these displays and responses are assumed to be innate. Both males and females process spatial information, such as body size, for themselves and for potential mates during courtship displays. Such processing allows the salamanders to align the male’s spermatophore with the female’s cloaca. Roth (1987) reviewed experimental studies that demonstrated the importance of coloration in mate recognition and choice in aquatic newts (Triturus spp.) that have highly visual mating rituals.
Quantitative and Numerical Discrimination
Organisms should evolve foraging strategies that maximize caloric gain while reducing caloric output. Functionally, such strategies would act as caloric intake/output calculators using information from numerosity and/or mass discrimination. This suggests that they should be able to distinguish a number of prey items and not just a larger overall quantity (i.e., mass) of prey. Plethodon shermani and P. metcalfi were given a choice of 8 versus 16 and 8 versus 12 prey items (crickets: Gryllus bimaculatus). They significantly chose 16 over 8, but there was no significant difference in the 8 versus 12 trials (Krusche et al. 2010). Plethodon cinereus chose the larger numerical quantity of prey (Drosophila virilis) when choosing between one versus two flies and between two versus three flies (Uller et al. 2003); but there were no differences when the animals were offered 4 versus 6 and 3 versus 4.
All salamanders may rely heavily on olfactory cues in all aspects of their behaviors. Plethodontid salamanders have unique grooves in the skin of the snout that lead from the margin of the upper lip to the nostrils; the grooves encourage uptake of chemical cues from the substrate, or the body surface of other salamanders, directly into the nostrils and their sensory cells. Physical damage to these grooves, from external parasites or aggressive interactions with conspecifics, can reduce subsequent foraging abilities (Jaeger 1981). The importance of chemical sensation in salamanders is underscored by multiple examples of cave-dwelling species that are blind and swamp-dwelling species with very reduced eyes. Nevertheless, the eyes of salamanders are similar to those of all amphibians in having a unique visual sensory cell type known as green rods; these cells contribute to remarkable visual abilities in very low-light situations.
Salamanders may communicate sex and species identity information through pheromones. Jaeger and Gergits (1979) found that males and females of P. cinereus and P. shenandoah can discriminate between their own substrate versus a substrate upon which a conspecific had been housed. In addition, both males and females of P. cinereus avoided the substrates of conspecifics. However, female P. cinereus did not avoid substrates of heterospecific, male P. shenandoah. This suggests that salamanders exhibit pheromonal (conspecific) and allomonal (heterospecific) communication, and these chemicals may confer information about the species and sex of individuals.
While olfactory cues appear to be a primary source of perception in salamanders, Roth (1987) presented experimental and theoretical evidence supporting the importance of visual stimuli in salamanders, both in the context of feeding and mate selection. Plethodon cinereus, presented with dead prey items, used chemical cues but not visual cues were used to successfully locate, recognize, and ingest prey (David and Jaeger 1981). In the wild, immobile prey may comprise 5% of the diet of these salamanders, indicating foraging is mediated both chemically and visually (Roth 1987). Immobile preys typically are not dead individuals, but rather items such as chrysalises or nonmoving snails, etc.
Fecal inspection is commonly observed in P. cinereus. Females that had been fed low-quality diets showed similar rates of squashing and inspection of fecal pellets of females and males that were fed a high-quality diet (Karuzas et al. 2004). The authors concluded that this behavior is used to select foraging areas. Males may take advantage of this behavior by defecating around their territory to attract females.
Larval and fully aquatic adult salamanders (e.g., Amphiuma spp.) have pressure-sensitive lateral line organs, as seen in fish and larval or fully aquatic frogs. The role of this sensory system in salamander perception has not been well studied.
Homing, Migration, and Orientation
Sensory systems involved in orientation appear to include olfaction, magnetoreception, and extraocular (i.e., pineal) photoreception. Many species of salamanders spend much of their adult life in terrestrial habitats, migrating to appropriate bodies of water to reproduce. Semlitsch (2008, p. 260) defined amphibian migration as movements toward and away from aquatic breeding sites. This definition distinguishes the behavior from various forms of simple dispersal. Although migrating salamanders tend to remain in or return to the area of hatching, dispersal by populations into areas does occur (Semlitsch 2008), even if the associated exploratory excursions rarely are observed.
Long-distance homing behaviors, which differ from strict migrations, have been well studied in the newt Taricha rivularis (Twitty et al. 1964) and less so in a few other salamanders. When T. rivularis were displaced, their movements were strongly oriented toward their home even when it was several kilometers distant from the release site. Such a displacement distance is far greater than would be typical of a breeding migration or individual dispersal.
Territoriality and Nesting
Jaeger and Forester (1983) reviewed various forms of territoriality and agonism in plethodontid salamanders – the best studied group by far – but other groups also have been surveyed (e.g., ambystomatids; Ducey 1989). Four characteristics of territoriality in salamanders are as follows: (1) the individual must have a “space” to be defended; (2) the individual must have the ability to defend this space through agonistic behavior; (3) individuals demarcate the space through pheromonal or visual displays; and (4) individuals are able to resist intrusion from invading salamanders.
Familiarity with a territory affects a resident salamander’s aggressiveness with intruders (Nunes and Jaeger 1989). Other factors that influence territoriality include body size of the resident, body size of the intruder, reproductive condition (i.e., courting season), number of intruders, condition of the tail of either salamander (e.g., complete vs. incomplete, amount of fat reserves), and food quantity/quality in the territory. Residents of a territory are more likely to initiate aggression than the unfamiliar salamander introduced to the territory (Jaeger 1981). In addition, there was more aggression observed when a resident was in contact with an unfamiliar salamander compared to a familiar salamander. When aggressive, the salamander’s bite is directed toward the tail or snout. Salamanders keep important fat reserves in their tail, and scar tissue on the snout negatively affects a salamander’s chemoreceptive abilities, leading to decreases in individual fitness.
Anthony et al. (1997) compared aggressive behavior in intra- and interspecific contexts in two species of sympatric salamanders: Plethodon ouachitae and P. albagula. These species have similar ecological requirements, but P. ouachitae is slightly smaller than P. albagula. Plethodon ouachitae were extremely territorial in both intra- and interspecific experiments, delivering bites 14 times more than in any previously studied Plethodon sp. Similar levels of aggression by P. ouachitae were observed when the salamander was an intruder into conspecific or heterospecific territories, indicating that elevated levels of aggression in this species also occur outside of the context of territorial defense.
The energy that male salamanders will expend defending their territory is proportional to the value of the territory. Adult male red-backed salamanders that have preferable prey (termites) are more likely to bite intruders than resident males that had less preferable prey (ants; Gabor and Jaeger 1995). Aggressively defending a territory sends honest signals to intruders that a territory is valuable. However, residents weigh these costs against the benefits of deterring the intruders. Female P. cinereus stay with their eggs while only occasionally leaving to forage, and her neonates stay with her for 2 weeks. Nest guarding behavior is known in many plethodontid species, as well as in Amphiumidae, Cryptobranchidae, and other examples. Kin recognition is consistent with the phenomena of territoriality and nest guarding and appears to occur in P. cinereus. Hungry females preferentially consume unrelated neonates compared to related neonates (Gibbons et al. 2003). Putative kin recognition and mother–offspring interactions have not been well studied beyond P. cinereus.
Sexual Differences in Territoriality
Salamanders show sex differences in choice of cover, searching unoccupied territories, and in mate guarding behaviors. Mathis (1990) showed that choices of cover objects, for both males and females, are positively correlated with body size. After removing salamanders from a section of the forest floor, females invaded unoccupied territories more often than did the males. Females may be more competitive because not only does she have to brood her eggs there, but she also must have sufficient prey (energy) in order to adequately provision eggs with yolk.
Males in a mated pair spend significantly more time threatening male intruders than do the females in the same mated pair (Lang and Jaeger 2000). However, when the intruding salamander is female, the female in the mated pair acts more aggressive than their male counterpart. Pairs co-defend territories, but not cooperatively because they show disproportional aggression toward same-sex intruders. In addition, both males and females aggressively attack their partners if the partner is anointed with pheromones from other opposite-sex individuals, which would indicate that they have been involved in social polyamory (Prosen et al. 2004).
There also are sexual differences in cost-benefit considerations of energy. For females, the cost-benefit ratio of territory defense depends on the number and sex of intruders, as Kohn et al. (2005) found that female salamanders defended their territory when one male was present, but not when there were two males present. In this case, it appears that the energy cost of defense was acceptable in the context of one male but unacceptable in the context of two.
Evidence suggests that females prefer substrates of partner males over those of unfamiliar males (Gillette et al. 2000). Social monogamy is preferential to both males and females because the female can take advantage of the male’s resources and the male can insure that she will mate with him. However, males showed no preference between their partner’s pheromones and an unpaired female. This suggests that adults display social polygamy when there is an unpaired female available. Additionally, interactions that display social monogamy are exclusively seen during the courtship season (Gillette 2003).
Dear Enemy Recognition and Mating Systems
Dear enemy recognition (Jaeger 1981) is a phenomenon in which salamanders have territories in close proximity to one another, where there will likely be an aggressive interaction between the residents. Regardless of the outcome, a second aggressive encounter between residents is less likely – evidently because the salamanders recognize their former enemies. Plethodon cinereus is able to distinguish between familiar and unfamiliar conspecifics. McGavin (1978) found that if a resident salamander was presented with either an unfamiliar salamander or familiar salamander, the resident bites the unfamiliar individual significantly more frequently. Similar results were found despite the sex of the resident.
Jaeger et al. (1995) found that juvenile P. cinereus are attracted to territorial pheromones of male and female adult P. cinereus. Adults were more tolerant of sharing their territory with a juvenile than an adult and more tolerant of a familiar juvenile compared to an unfamiliar juvenile.
Much remains unknown about Caudata, but recent research suggests that salamanders are more cognitively complex than previously thought. There is evidence of intricate inter- and intraspecific dynamics both in the field and the lab. Clearly, cognitive abilities are manifest in many areas of an organism’s life including searching for a territory, choosing a mate, and foraging, among others. Being the most basal group in the phylogeny of tetrapods, amphibians form the comparative basis for all other cognitive studies with tetrapods, as well as with fishes.
- Agrillo, C. (2014). Numerical and arithmetic abilities in non-primate species. Oxford: Oxford University Press.Google Scholar
- Ducey, P. K. (1989). Agonistic behavior and biting during intraspecific encounters in Ambystoma salamanders. Herpetologica, 1989, 155–160.Google Scholar
- Frost, D. R. (2019). Amphibian species of the world: An online reference. Version 6.0. New York: American Museum of Natural History. Retrieved from http://research.amnh.org/herpetology/amphibia/index.html.Google Scholar
- Gillette, J. R. (2003). Population ecology, Social behavior, and intersexual differences in a natural population of red-backed salamanders: A long-term field study. Unpublished doctoral dissertation. University of Louisiana at Lafayette, LafayetteGoogle Scholar
- Gillette, J. R., Kolb, S. E., Smith, J. A., & Jaeger, R. G. (2000). Pheromonal attractions to particular males by female redback salamanders (Plethodon cinereus). In R. Bruce, R. Jaeger, & L. Houck (Eds.), The Biology of plethodontid salamanders (pp. 431–440). New York: Springer Science and Business Media.CrossRefGoogle Scholar
- Petranka, J. W. (1998). Salamanders of the US and Canada. Washington, DC: Smithsonian Institution Press.Google Scholar