By continuing to browse this site you agree to us using cookies as described in About Cookies
Notice: Wiley Online Library will be unavailable on Saturday 7th Oct from 03.00 EDT / 08:00 BST / 12:30 IST / 15.00 SGT to 08.00 EDT / 13.00 BST / 17:30 IST / 20.00 SGT and Sunday 8th Oct from 03.00 EDT / 08:00 BST / 12:30 IST / 15.00 SGT to 06.00 EDT / 11.00 BST / 15:30 IST / 18.00 SGT for essential maintenance. Apologies for the inconvenience.
In the transition from aquatic to terrestrial life, the integument of tetrapods likely underwent significant evolutionary changes in order to accommodate the tasks and challenges associated with the new environment. Most of these changes remain only poorly understood due to inadequate preservation of skin in most fossil or early tetrapods (cf. Clack, 2012), and much of our knowledge instead derives from comparative approaches using extant taxa as the best available proximal models for the transitional forms. The integument of extant amphibians is highly specialized and embedded with numerous unicellular and multicellular glands, and its structure has been the subject of considerable study (cf. Duellman and Trueb, 1986; Fox, 1994). Many of its exocrine glandular structures are well documented (e.g., Brodi and Gibson, 1969; Mills and Prum, 1984; Warburg et al., 2000; Rigolo et al., 2008; Heiss et al., 2009). In most body regions, these glands are small, but regional specializations with substantially larger glands are common, especially in the head region. Specifically, the periorbital and snout regions often contain several such large compound multicellular glands. These includes glands found in the lateral nasal (nasolabial glands: Whipple, 1906), and orbital (anterior orbital and lacrimal glands: Piersol, 1887) regions.
The nasolabial gland has thus far only been described in selected plethodontid salamanders (Whipple, 1906). Considerably, more has been published on the tetrapod deep anterior orbital gland (DAOG: aka Harderian gland) which is found in the anterior (medial canthus) of the orbit of most tetrapod vertebrates. A diverse assortment of functions has been ascribed to this exocrine gland (see Payne, 1994, Hillenius and Rehorek, 2005; Rehorek et al., 2005 for reviews), several of which are mediated by the nasolacrimal duct (NLD), that drains the orbital fluid into the nasal region in the vicinity of the vomeronasal organ (VNO) (Hillenius and Rehorek, 2005). Likewise, there is variation in the relative development and presence of this gland both between major tetrapod taxa (Sakai, 1992; Payne, 1994; Chieffi et al., 1996; Rehorek, 1997) as well as within major groups (e.g., bats, primates, and rodents: Payne, 1994; Rehorek and Smith, 2006; Rehorek et al., 2010b). The vast majority of published research concerns the structure and function of the DAOG in mammals; however, considerably less is known about the nonmammalian lineages, and least of all the nonanuran amphibians. Because amphibians are placed as the basal group in virtually all published tetrapod phylogenetic trees (Laurin, 2002), examination of the amphibian orbital glands could shed considerable light on the evolutionary origins, and perhaps the primitive functions, of this intriguing glandular structure.
To date, studies of amphibian orbital glands have been largely restricted to frogs (for reviews, Chieffi et al., 1992, Santillo et al., 2011), with only occasional reference to caecilians (Wake, 1985). In contrast, there are very few studies available on salamanders, and most studies of these are well over 100 years old (e.g., Reichel, 1883; Piersol, 1887; Sardemann, 1887). This study was intended to reduce the gap at least somewhat. In these classic studies, it was postulated that at least the anterior portion of the single orbital gland is homologous to the DAOG of other tetrapods, and its posterior portion to the lacrimal gland. This classic interpretation has been cited and reiterated in most of the literature since (Weidersheim, 1898; Francis, 1934; Walls, 1942).
These earlier studies were largely restricted to European (Salamandridae: Piersol, 1887) and new world (Ambystomidae, Amphiumidae, and Plethodontidae: Piersol, 1887; Sardemann, 1887; Whipple, 1906) salamanders. Among these lineages, it appears that the Plethodontidae are one of the most derived families (see Fig. 1: Wiens et al., 2005). The purpose of this study was twofold: (1) to describe the orbitonasal glands of several North American species of salamanders (Salamandridae, Ambystomidae, and Plethodontidae). (2) To examine possible correlations of lifestyle (aquatic vs. biphasic) on relative orbitonasal gland development among salamander species.
MATERIALS AND METHODS
Two sources for salamanders were used in this study: animals were either (1) obtained from the Slippery Rock University herpetological collection (formalin fixed and stored in 70% ETOH) or (2) collected in the western Pennsylvania area (see Table 1). A total of eight different species, with 2–5 replicates per species were examined. Freshly captured animals were measured, weighed, and promptly sacrificed with an intraperitoneal injection of sodium pentobarbital (60–100 mg/kg of body weight) followed by decapitation, as per AVMA 2000 guidelines on euthanasia for amphibians. The lower jaws were removed and the heads were then placed in 10% phosphate-buffered formalin for at least a week. For the subsequent steps, all specimens (both recently captured and those from the existing herpetological collection) were treated the same. Either half-heads (Ambystoma spp.) or whole heads (remaining taxa) were then placed for 2–3 weeks in 10% EDTA in a 10% phosphate-buffered formalin solution, after which they were dehydrated through a series of ethyl alcohols, cleared in xylenes, and embedded in paraffin. Heads were then serially sectioned at 10 µm. Sections were then stained with either Hematoxylin and Eosin or with Mason's Trichrome.
Table 1. Salamander species examined in this study
Average SVL (mm)
These species were obtained from the SRU herpetological collection. These animals had been collected before the requirement of either collecting permits (collected as part of a field course) or IACUC approval.
The location and relative development of the orbital and nasolabial glands in the salamanders examined in this study is summarized in Table 2.
Table 2. List of all salamanders mentioned and source of data
Ecological designation determined from Petranka (1998): (1 = never leave water, 2 = forage on both land/water, 3 = forage mostly on land, and 4 = completely terrestrial). Summary of glands observed in salamanders. Orbital gland: L = large (palpebral and suborbital portions), R = row (palpebral only); nasolabial glands: − = absent, + = present but localized to small area of the lateral nasal region, ++ = present and fills much of the lateral nasal region; nasolacrimal duct: − = absent and + = present.
In general, two main multicellular exocrine glands may be observed in the salamanders examined in this study, namely in the periorbital (orbital gland) and lateral nasal (nasaolabial glands) regions (Fig. 2).
All species possess orbital glands, housed primarily in the lower eyelid (palpebra). The duct system in the orbital glands is not morphologically distinct from the body of the gland in any species. The orbital glands of the two largest species (Ambystoma sp., and Desmognathus monticola: Fig. 3a,b), and those of most of the Plethodontidae (Fig. 3c) are similar in shape. In these taxa, the orbital gland is large with a main suborbital portion. This orbital gland resides chiefly in the lower eyelid (palpebral), whose distinct anterior and posterior projections effectively cup the eyeball from below. There appear to be 1–2 ducts for this gland that open on the orbital surface of the lower palpebra. In contrast, the orbital gland in both Pseudotriton ruber (Plethodontidae) and Nothophthalmus viridescens (Salamandridae) is much smaller, more tubular in form, and resides wholly within the lower eyelid (Fig. 3d).
Nasolabial glands are present in all salamanders except for N. viridescens. In Amybstoma spp., D. monticola, and Pseudotriton ruber, these are small structures restricted to the nasal region (Fig. 4a,b). The nasolabial glands are much larger in both Eurycea spp. and Plethodon cinereus, and even extend to a large lobe over the orbital cavity (Fig. 4c). The rostral ducts of these glands are small and difficult to distinguish. They open inferior to the naris, in the nasolabial groove (when present) or in vicinity thereof.
A patent NLD is only found in N. viridescens and in both species of Ambystoma. The NLD connects the anterolateral portion of the lateral nasal diverticulum to the anterior orbital rim, superior to the lower eyelid. There is a single lacrimal punctum opening up in both N. viridescnes and A. tigrinum. There is both a superior and an inferior lacrimal punctum in A. maculatum opening 20 µm apart.
The observations of this study indicate that there is considerable variation in the relative development of the orbitonasal glands among salamanders. This implies that the concept of a common, representative morphology for salamander orbital glands may not be applicable. The results of this study largely agree with previously published descriptions, however new additional observations permit a different interpretation of these results. Although all species examined possessed at least one orbital gland, either the nasolabial glands (N. viridescens) or the NLDs (Plethodontidae) were absent in select taxa. Furthermore, the results of this study suggest that three broad themes are apparent with respect to the development of the orbitonasal glands in salamanders.
Effect of Phylogeny on Nasolabial Gland and Nasolacrimal Duct
Both of these structures appear to correlate with phylogenetic association, but in a mutually exclusive pattern. Both may serve a common functional role as accessory structures for the VNO. In plethodontid salamanders, their nasolabial glands are substantial, the NLDs are absent, and secretions from the orbital glands thus have no direct route to the VNO. It is possible that in this taxon the hypertrophied nasolabial glands may have evolved as an alternative source for VNO lubricant, with shallow nasolabial grooves to provide access to the VNO via characteristic nose-tapping behavior (Brown, 1968; Arnold, 1976; Dawley and Bass, 1989; Graves, 1994).
Nonplethodontid salamanders (Salamandridae and Ambystomidae), however, show an opposite pattern: they possess NLDs with either greatly reduced or absent nasolabial glands. The absence of large nasolabial glands and nasolabial grooves in ambystomids is consistent with their putatively VNO-associated behavior. These salamanders stick their noses closer to the substrate (Eisthen and Park, 2005) than their plethodontid counterparts.
This anatomical arrangement conforms to a pattern seen broadly in tetrapods, in which there is a (possibly ancient) correlation between the VNO, NLD and orbital glands, such that the orbital gland secretions may pass to the VNO by means of the NLD (see Hillenius and Rehorek, 2005). Unfortunately, there is currently no experimental data concerning the flow of orbital gland secretion in nonplethodontid salamanders that might corroborate the anatomical correlation observed.
Although nasolabial glands have not been described in other tetrapods, the presence of the NLD is clearly documented. It is universally present in the other two extant orders of amphibians (frogs and caecilians), as well as in leipodosaur reptiles, archosaurs, and various mammals (Hillenius and Rehorek, 2005). However, the presence of the NLD is variable within some mammalian taxa (i.e., primates and bats: Rehorek and Smith, 2006; Rehorek et al., 2010b). Thus, the presence of the NLD cannot be used as a defining feature for taxonomic purposes in certain taxa as its presence may instead be due to other factors.
Effect of Lifestyle on Orbitonasal Gland Development
Salamanders exhibit variation in their relative dependence on water, ranging from fully aquatic (never leave water) to fully terrestrial (live in damp places). Petranka (1998) categorized these in four ecological designations, from 1 (fully aquatic) to 4 (fully terrestrial). Smaller nasolabial and orbital glands are observed in the aquatic species (N. viridescens and P. ruber), whereas in the smaller terrestrial plethodontid salamanders both the nasolabial and orbital glands are large. The remaining three larger terrestrial salamanders uniformly have large orbital glands and small nasolabial glands (see Table 2 for summary). Thus, the presence and relative development of orbital glands, and potentially the nasolabial glands (each with a potential functional correlation to the VNO) may correlate with adult terrestrial phases. This is supported by observations from other aquatic salamanders in whom the orbital glands (Amphiuma: Piersol, 1887) and NLD (Amphiuma, Siren, and proteid salamanders: see Hillenius and Rehorek, 2005 for review) are absent, and the VNO is reduced.
Previously, no orbital glands were observed in the aquatic larval stages of Triturus spp., Lissotriton spp., and Salamandra salamandra (Sardemann, 1887) and there are also no published descriptions of nasolabial glands in aquatic larvae. Further studies are needed to determine the inception and rate of development of these glands in salamanders. In comparison to frogs, it has been found that the orbital gland develops later during metamorphosis in the fully aquatic Xenopus in comparison with that of the other more terrestrial frogs (Shirama et al., 1982). Aquatic lifestyles appear to have no effect on relative development of amniote orbital glands (Saint Girons, 1988; Burns, 1992; Sakai, 1992; Rehorek et al., 2005).
Nomenclature of the Orbital Gland
In tetrapods, three distinct glandular regions in the orbit can generally be recognized (Fig. 5a): an anterior region that encompasses the deep anterior orbital/Harderian (DAOG), superficial anterior orbital/nictitans (SAOG), and anterior lacrimal glands (ALG); a lower palpebral region that contains the palpebral gland; and a posterior region that encompasses the posterior lacrimal gland. The question here is to which of these, if any, the orbital glands of salamanders corresponds.
Among the salamanders examined in this study, the orbital gland is always found in the lower palpebral region (Fig. 5b). The ducts of the gland open onto the orbital surface of the lower eyelid, whereas the body of the gland extends downward below the eye into the ventral aspect of the orbit. This condition was also observed in European salamanders (Piersol, 1887; Sardemann, 1887). Sardemann (1887) described two separate glandular masses in the lower eyelid of Salamandra salamandra (Salamandridae) one anterior and the other more posterior, which he labeled as the Harderian and posterior lacrimal (or simply lacrimal) glands, respectively. This nomenclature has been since been cited by subsequent authors, including Wiedersheim (1898) and Walls (1942).
Sardemann's (1887) use of the terms Harderian and lacrimal for the anterior and posterior glandular masses implies homology with the Harderian (or DAOG) and lacrimal glands, respectively, of other tetrapods. However, the accuracy of this assertion is far from clear.
In other tetrapods, the term lacrimal gland refers to a gland situated in the posterior aspect (outer canthus) of the orbit. It is usually only a mucous secreting gland, although in turtles it appears to have some salt-secreting ability (Chieffi Baccari et al., 1992). Although the lacrimal gland may occur in several different locations in the posterior orbital (outer canthus) region in tetrapods (e.g., Saint Girons, 1988; Rehorek et al., 2010b), there is never more than one such gland at a time. Because there are no other glands that can confuse the situation, the lacrimal gland passes the test for conjunction (cf. Patterson, 1982, 1988), and thus does not conflict with Sardemann's (1887) interpretation of salamander orbital glands: the orbital gland of salamanders could be homologous to the lacrimal gland of other tetrapods.
The situation at the anterior orbital region (medial canthus) in tetrapods is more complicated. The main anterior orbital gland complex (AOGC: variously compromised of DAOG and/or SAOG: see Rehorek et al., 2010a) is found in the medial canthus of the orbit of most terrestrial vertebrates. In amniotes, this glandular complex is usually associated with the nictitating membrane, and its ducts open to the inner aspect (corneal side) of the nictitating membrane. However, two other distinct glands may be present alongside the AOGC in the same organism: (1) an ALG, which may be found on the outer (conjunctival) aspect of the nictitating membrane in some larger species of lizards (Saint Girons, 1988; Rehorek et al., 2006, 2009); (2) palpebral glands, which are contained in the lower eyelid and open onto the ocular surface. Among amniotes, the latter are observed only in alligators (Rehorek et al., 2005).
Of these anterior orbital glandular structures, the AOGC and ALG are associated with the nictitating membrane, whereas the palpebral gland resides wholly within the anterior portion of the lower eyelid. The latter condition is broadly similar to the orbital gland of salamanders (Fig. 5b). Moreover, the ducts of the orbital glands of salamanders open directly in the lower eyelid (like the palpebral gland of alligators), unlike the AOGC and ALG whose ducts open in the medial canthus, in close association with the nictitating membrane. Thus, an alternative to Sardemann's (1887) view may therefore be that the orbital glands of salamanders represent a palpebral gland, either as an independent evolutionary development in salamanders, or in some manner homologous to the crocodilian palpebral glands. This interpretation somewhat complicates Wiedersheim's (1898) premise that the DAOG (Harderian gland) evolved as a lubricatory structure in ancestral tetrapod vertebrates, in that this gland may have no homologue in salamanders.
Wiedersheim (1898) went on to state that although in many urodeles there is first a single band of glands in the lower eyelid, the middle parts soon disappears to leave two separate glandular regions. Whipple (1906) described one large orbital gland in the lower eyelid of Desmognathus fusca, but declined to name it. The salamanders examined in this study also appeared to have only one glandular mass in the lower eyelid. In the larger species (Ambystoma spp. and D. monticola) and most plethodontids, anterior and posterior masses could be discerned in addition. However, these are merely subunits of one gland whose ducts opened up onto the conjunctival aspect of the lower palpebra. The previous authors (Piersol, 1887; Wiedersheim, 1898; Whipple, 1907; Sardemann, 1887) did not apply a name for the single gland in salamanders. Based on its close association with the lower eyelid, we suggest that this entire structure be considered the medial palpebral gland. In those salamanders in which there are two distinct glands (Salamandra salamandra: Sardemann, 1887, and not reexamined in this study), we further suggest they be called the anterior palpebral and posterior palpebral glands, respectively. Additional studies, especially on the development of these glandular structures, are needed to determine whether these conditions represent truly distinct glands among salamanders or merely heterogenous development of what is otherwise essentially a single, continuous structure.
This change in orbital gland nomenclature challenges the assumed significance of the salamander as retaining features of a putative tetrapod ancestor. Based upon ductal openings, salamander orbital glands do not appear to be homologous to either anterior (AOGC or ALG) or possibly posterior (LG) orbital glands of other tetrapods. In the case of these anterior glands, whose ducts all open onto a nictitating membrane or in the anterior conjunctiva adjacent to the cornea, there is no case to homologize any of these glands specifically to the anterior most palpebral glands of the salamander. In the case of the posterior gland (LG), whose ducts open onto the conjunctiva in the posterior aspect of the orbit, the case for homology with the salamander orbital gland is circumstantial, but stronger than that for the anterior glands. Given the distribution of the orbital glands in salamanders, it seems most parsimonious to consider the salamanders as exhibiting a derived, not an ancestral, condition. Thus, the salamander orbital glands may not be representative of the ancestral tetrapod condition.
However, at a wider level, examination of all tetrapod orbital glands suggests a general pattern of orbital gland location. There appear to be three areas where orbital glands regularly develop: anterior (anterior orbital and anterior lacrimal), lower eyelid (palpebral), and posterior (lacrimal) regions of the orbit. Tissues in these three areas in the orbit apparently are especially prone to glandular induction. The difference between glands in these three areas is location and type of secretant produced (when known). The lacrimal (e.g., Saint Girons, 1988; Rehorek et al., 2006, 2009, 2010b) and palpebral (Rehorek et al., 2005) glands are generally mucus-secreting structures. The anterior ocular gland complex, however, produces many different secretions (Payne, 1994; Rehorek, 1997; Rehorek et al., 2005).
Thus, while all tetrapods have some sort of orbital gland, there are three sets of unresolved questions: (1) What triggers, in terms of ontogeny, the initial development of the orbital glands? For example: the DAOG develops from the orbital extension of the NLD in snakes (Rehorek, 1998). However, in other tetrapods, though the DAOG is associated with the orbital extension of the NLD, there is no direct developmental connection (Rehorek et al., 2011). (2) Once triggered, what then determines the relative development (size) of the gland? For example, in humans, the AOGC is present, but it fails to develop beyond two small solid cords in the fetal stage (Arends and Schramm, 2004). In contrast, an AOGC is present in adults of several lower primates (Rehorek and Smith, 2006). (3) What determines the nature of the secretant? Why is there so much variation in the nature of the secretant of the anterior orbital gland among tetrapod taxa?
These three questions all appear to revolve around molecular signaling events. If the inception, relative development and ultimate function of the orbital glands involve the same signaling events, then the question becomes: what is the nature of these signaling events? Could these signaling events be homologous across tetrapods? If that were to be the case, this would be an example of genetic homology (cf. Hall, 2012). Alternatively, the events may be examples of independent signaling, converging upon a common phenotype. If this is the case here, this would be an example of convergent evolution.
However, if there is a commonality in developmental pathway, this would be an example of developmental homology. In this case, there may be no direct structural homology, but there may be potential genetic and developmental homology (cf. Hall, 2012). To test these ideas, studies need to be conducted at the molecular level.
In conclusion, there appears to be no such thing as a “typical” salamander, when it comes to orbital and nasolabial gland structures. The variable presence of a NLD and nasolabial glands could be correlated to a functional shift in chemosensory behavior: from relatively simple nose waving in Ambystoma spp. to complex nose-tapping in Plethodon spp. All salamanders possess either one (medial palpebral) or two (anterior and posterior palpebral) orbital glands. This is unlike the condition observed in other extant amphibians (anurans and gymnophionans) in whom there is a deep anterior orbital (Harderian) gland located in the medial canthus. Thus, salamanders may exhibit a unique, and probably derived, condition. Whether there is a functional similarity between the anterior orbital gland and the palpebral gland remains to be elucidated. To fully understand the significance, nomenclature, and homology of the orbital glandular region in tetrapods, further research is required to unravel the both the associated genetic processes and the ontogeny of salamander palpebral glands.
The authors thank Sara Woodley, Nikki Xenakis, and Trent Davis for technical assistance during the course of this project. Animals were captured in accordance with the regulations of the Commonwealth of Pennsylvania Fish and Boat Commission (License # 006–061-808). Animals were kept in captivity for a brief time in accordance with the guidelines set by the Slippery Rock University Institutional Animal Care and Use Committee (protocol number 2008-01).