AR The Anatomical Record Advances in Integrative Anatomy and Evolutionary Biology

Transcription

1 A N O F F I C I A L P U B L I C AT I O N O F T H E A M E R I C A N A S S O C I AT I O N O F A N ATO M I S T S K U R T H. A L B E R T I N E E D I TO R - I N - C H I E F VOL. 301 NO. 5 MAY 2018 AR The Anatomical Record Advances in Integrative Anatomy and Evolutionary Biology elebrating COVER 100 YEARS ANATOMICAL PUBLICATION of E xcellence in

2 THE ANATOMICAL RECORD 00:00 00 (2018) The Lack of Nasolacrimal Ducts in Plethodontid Salamanders? DUSTIN S. SIEGEL, 1 * MICHAEL S. TAYLOR, 1 DAVID M. SEVER, 2 AND STANLEY E. TRAUTH 3 1 Department of Biology, Southeast Missouri State University, Cape Girardeau, Missouri 2 Department of Biological Sciences, Southeastern Louisiana University (Emeritus), Hammond, Louisiana 3 Department of Biological Sciences, Arkansas State University (Emeritus), State University, Arkansas ABSTRACT Nasolacrimal ducts are a terrestrial vertebrate adaptation and appear to have co-evolved with orbital glands. Although plethodontid salamanders possess orbital glands, a recent study concluded that plethodontid salamanders lack nasolacrimal ducts. Functionally, the absence of nasolacrimal ducts closes the route for orbital gland secretion passage into the nasal and vomeronasal organ cavities. Orbital glands have been implicated in enhancement of vomeronasal function so loss could have important implications for communication. Multiple older studies depict or discuss nasolacrimal ducts in plethodontid salamanders. Interestingly, the only consensus between recent and older literature is that Desmognathus lacks nasolacrimal ducts. To determine if plethodontid salamanders truly lack nasolacrimal ducts, we sectioned plethodontid salamander heads for general histological examination of species from the majority of the plethodontid tribes. From our representative sample, we found only two species that completely lacked nasolacrimal ducts (Desmognathus fuscus and Eurycea tynerensis) and one species that possessed nasolacrimal ducts that ended blindly before reaching the nasal cavities (E. spelaea). Bayesian ancestral state reconstruction resulted in the presence of nasolacrimal ducts on the branch leading to Plethodontidae and both subfamilies within Plethodontidae, with two independent losses in Desmognathus and Eurycea. Anat Rec, 00: , VC 2017 Wiley Periodicals, Inc. Key words: orbital gland; plethodontidae; eye; nasal cavity; vomeronasal organ Nasolacrimal ducts are epithelial tubes that drain excess secretions of orbital glands through the boney nasolacrimal canal and into the nasal cavities (Bertmar, 1969). These ducts are a common feature in the vast majority of tetrapods and probably co-evolved with glands associated with the eyes, and possibly the vomeronasal organ (for review see Hillenius and Rehorek, 2005). The evolutionary purpose of tearing into the nasal cavity has received little attention from researchers and, thus, the selective pressures that resulted in the evolution of nasolacrimal ducts are unknown; however, a passage for orbital gland fluids that potentially enhance vomeronasal organ function is a common utility addressed in the literature (for review see Hillenius, 2000). Additional Supporting Information may be found in the online version of this article. *Correspondence to: Dustin Siegel, Department of Biology, Southeast Missouri State University, One University Plaza MS6200, Cape Girardeau, MO Phone: dsiegel@semo.edu Received 21 August 2017; Revised 16 October 2017; Accepted 21 October DOI /ar Published online 00 Month 2017 in Wiley Online Library (wileyonlinelibrary.com). VC 2017 WILEY PERIODICALS, INC.

3 2 SIEGEL ET AL. Fig. 1.

4 NASOLACRIMAL DUCTS OF PLETHODONTIDAE 3 A recent study concluded that although orbital glands are present in salamanders of the Plethodontidae, plethodontid salamanders (of which there are currently 4501 extant species; AmphibiaWeb, 2017) lack nasolacrimal ducts, contrary to the findings in other families of salamanders; that is, In plethodontid salamanders, their nasolabial glands are substantial, the nasolacrimal ducts (NLDs) are absent, and secretions from the orbital glands thus have no direct route to the vomeronasal organ (Rehorek et al., 2013). These conclusions were based on examination of Desmognathus, Eurycea, Pseudotriton, and Plethodon, representing three tribes (Desmognathini, Plethodontini, and Spelerpini) within the two subfamilies (Hemidactyliinae and Plethodontinae) of Plethodontidae reconstructed by Shen et al. (2016). Wilder (1913) also discovered the lack of nasolacrimal ducts in one species of plethodontid, D. fuscus. This was an interesting conclusion of her study, as Hillenius (2000) reported that nasolacrimal ducts have never been reported absent when septomaxillae were reported as present. Wilder s (1913) previous work appears to contradict that conclusion because Desmognathus possesses septomaxillae, albeit, highly reduced (Wake, 1966). In accordance with the work by Rehorek et al. (2013) on plethodontids, many taxa possess septomaxillae but lack nasolacrimal ducts, if their conclusions were correct, as numerous plethodontids possess septomaxillae (Wake, 1966). Inexplicably, numerous previous studies indicated the presence of nasolacrimal ducts in plethodontid salamanders. For example, Higgins (1920) mentions nasolacrimal ducts from Plethodon erythronotus (P. cinereus) in his early comparative work on amphibian vomeronasal organ structures; nasolacrimal ducts are labeled in Wilder s (1925) famous treatise on amphibian metamorphosis using Eurycea bislineata as a model; while investigating the homology of the septomaxillary bone, LaPage (1928) mentions the opening of nasolacrimal ducts into the nasal cavities in at least P. cinereus; Dawley and Bass (1988, 1989) and Dawley et al. (2006) labeled nasolacrimal ducts in their studies on vomeronasal organs in P. cinereus; Wirsig-Wiechmann et al. (2002) used the opening of nasolacrimal ducts into the nasal cavities as structures to provide consistency in head measurements in their study on vomeronasal organs in Plethodon shermani; numerous osteological studies label openings (nasolacrimal canals) for nasolacrimal duct passage in skulls of plethodontid salamanders from multiple Bolitoglossa (Alberch, 1983) and Nototriton saslaya (Kohler, 2002), for example. Most conflicting to the recent interpretation that plethodontid salamanders do not possess nasolacrimal ducts (Rehorek et al., 2013) were descriptions of nasolacrimal duct passageways in every tribe of plethodontid salamander from Wake s (1966) monograph on comparative osteology of plethodontid salamanders. From the above, it appears that there are two competing observations with regard to the presence of nasolacrimal ducts in plethodontid salamanders; that is, plethodontid salamanders lack nasolacrimal ducts and that plethodontid salamanders possess nasolacrimal ducts. To address these two alternative interpretations, we collected an independent series of plethodontid salamanders from eight tribes (Aneidini, Batrachosepini, Bolitoglossini, Desmognathini, Ensatinini, Hemidactyliini, Plethdontini, and Spelerpini) of the two plethodontid subfamilies (Hemidactyliinae and Plethodontinae; Wake, 2012) to examine the evolution of nasolacrimal ducts in plethodontid salamanders. Two alternative predictions were examined in our study in reference to nasolacrimal ducts; (1) their presence is the ancestral condition for plethodontid salamanders and (2) their absence is the ancestral condition for plethodontid salamanders. MATERIALS AND METHODS Specimens Representatives from both subfamilies of Plethodontidae (Hemidactyliinae and Plethodontinae) and eight of nine tribes (Aneidini, Batrachosepini, Bolitoglossini, Desmognathini, Ensatinini, Hemidactyliini, Plethdontini, and Spelerpini) were obtained from natural history collections along with one species of Rhyacotritonidae for outgroup comparison (Appendix). Two males and two females were obtained for each species except Oedipina cyclocauda because of rarity in natural history collections. In that species, only one male and one female were obtained. All specimens were preserved in 70% ethanol but fixation method was unknown. For unknown reasons (potentially because of previous fixation/preservation methods employed by the original collector), one female Bolitoglossa subpalmata (FMNH ), one male Pseudoeurycea bellii (FMNH ), and one female Rhyacotriton olympicus (CAS 47651) became extremely brittle after tissue processing (see below) and could not be adequately sectioned for examination and were not used in any analysis. Tissue Processing Heads from every specimen were removed, rinsed in deionized water and then decalcified in RDO (Apex Engineering Products Corporation, Aurora, IL) for three hours. Heads were subsequently rinsed in deionized water and then dehydrated with ascending concentrations of ethanol, cleared in toluene, infiltrated with paraffin, and then embedded in embedding molds for microtomy. Heads from both a male and a female of each species were oriented for transverse sectioning and remaining heads were randomly embedded for frontal or sagittal sectioning. Serial sections of entire heads were obtained with a rotary microtome at 7 mm and affixed to Fig. 1. Representative sagittal sections through the head of a female Batrachoseps attenuatus (Hemidactyliinae, Batrachosepini; FMNH ; hematoxylin and eosin). The sections were ordered from most lateral (A) to most medial (E; scale bar mm). Low magnification images in the left column, a e, were used to provide orientation of internal features and correspond to A F (scale bar mm). The boxes in the low magnification images represent the approximate regions from where sections (A E) were isolated. (A and B) Sealing of the eyelids (white arrowheads) at the medial canthus of the eye. (C) Formation of the nasolacrimal duct (black arrowhead) at the medial canthus of the eye. (D E) Nasolacrimal duct migration (black arrowheads) and opening (gray arrowhead) into the nasal cavity (black arrow). Other symbol: grey arrow, tubules of orbital gland mass.

5 4 SIEGEL ET AL. Fig. 2.

6 NASOLACRIMAL DUCTS OF PLETHODONTIDAE 5 albumenized slides. Slides were stained with hematoxylin and eosin for general histological examination. Histological procedures followed Kiernan (1990). Phylogenetic Analysis Fifteen representative taxa were analyzed with sequences used for both paedomorphic and metamorphic Eurycea tynerensis. Paedomorphic and metamorphic E. tynerensis were used for the analysis to account for potential differences in their morphological features because of their different life histories (respectively, reproductive with larval features versus loss of larval features through metamorphosis; Bonett and Chippindale, 2004). For each taxon, a nuclear recombinationactivating gene 1 (Rag1) sequence was obtained from GenBank (Supporting Information, Table S1). Rhyacotriton was included as the outgroup taxon. Surrogate taxa were used for Bolitoglossa subpalmata (B. helmrichi), Oedipina cyclocauda (O. taylori), Pseudoeurycea bellii (P. rex), and R. olympicus (R. cascadae) because Rag1 sequences were not available from GenBank. Sequences were aligned using an internet implementation of Clustal Omega ( The final sequence length used for the analysis was 1,080 bp with no gaps. A general time reversible model with a proportion of invariant sites and gamma-distributed rate heterogeneity was chosen using MrModelTest 2.25 (Nylander, 2004). Phylogenetic relationships were inferred with MrBayes (Ronquist et al., 2011) by Markov Chain Monte Carlo (MCMC) sampling for 30 million generations. Two independent runs of four Metropolis-coupled chains were run simultaneously using uniform prior probabilities and appropriate model parameters estimated on a randomly generated starting phylogeny. Trees were sampled from the posterior-probability distribution once every 100 generations. Stationarity was reached within the first three million generations, but the first 25% of samples (75,000 trees) from each run was conservatively discarded as burn-in. The remaining 225,000 trees from each run were combined to determine final posterior probabilities. Nasolacrimal ducts were coded as 0 for present, 1 for absent, and 2 for present but not patent (does not provide complete passage to the nasal cavity). Eurycea tynerensis was split into two terminals, paedomorphic individuals and metamorphic individuals; however, they possessed identical presence/absence for nasolacrimal duct state and a second analysis was conducted with both condensed into one terminal. The results were practically identical (see Supporting Information, Figure S1 for analysis with only the metamorphic E. tynerensis used as a terminal). Ancestral state reconstructions were conducted using BayesTraits v3.0 ( Traits.html) with a reverse jump MCMC model and multistate reconstruction (Pagel et al., 2004). The analysis was performed on 100,000 trees by combining the final 50,000 trees from each of the independent MrBayes runs. Hyper uniform priors from 0 to 100 were used for analysis of each character. Each analysis was run for 50 million generations, with samples taken every 1,000 generations. The first 50% of the sampled generations were discarded as burn-in, leaving 25,000 samples for analysis. RESULTS All salamanders examined possessed orbital glands (for example see Figs. 1A E and 2A E) except for paedomorphic individuals of Eurycea tynerensis. Eurycea spelaea possessed extremely small orbital gland tubules but no terminal dilation of individual tubules as observed in every other taxon. Similarly, paedomorphic individuals of E. tynerensis did not possess nasolacrimal ducts, and E. spelaea individuals possessed nasolacrimal ducts but the ducts did not make contact with the nasal cavities (non-patent condition). Only two taxa were found to possess fully developed orbital glands but lacked nasolacrimal ducts; that is, individuals of Desmognathus fuscus (Figs. 2A E) and metamorphic individuals of E. tynerensis. The morphology of nasolacrimal ducts was similar between all taxa of plethodontids. Left and right nasolacrimal ducts formed at the medial canthus where the superior eyelids and lower eyelids sealed their respective superior and inferior conjunctiva from the external environment (Figs. 1A C and 3A,B). In taxa lacking nasolacrimal ducts, no formation of a duct was observed at the medial canthus where the eyelids join (Fig. 2A C), nor at any other region of the head. The nasolacrimal ducts traveled rostral and slightly ventro-medial through connective tissue of the rostral canthus (dorsal to cranial bones and ventral to the epidermis) toward the nasolacrimal canals (Figs. 1C E and 3B F) in all taxa except Ensatina, where the nasolacrimal ducts primarily project laterally to the nasal cavities. The nasolacrimal ducts curved perpendicular to the horizontal plane of the salamanders at the nasolacrimal canals, passed through the nasolacrimal canals, curved caudally, and emptied into the nasal cavities at their junctions with the vomeronasal organs (Figs. 1C E and 3C F). The nasolacrimal canals were formed from foramina or evacuations of cranial bones (for a detailed gross analysis of the association of cranial elements with the nasolacrimal ducts in plethodontids, see Wake, 1966) and openings in the cartilagines oblique, which resided immediately beneath the cranial bones (Fig. 3E,F). Small papillae were typically observed through which the nasolacrimal ducts emptied into the nasal cavities but the presence of papillae varied intraspecifically. Bending as the nasolacrimal ducts approached the nasal cavities was highly variable among taxa. In Eurycea, Batrachoseps, and Plethodon the nasolacrimal ducts made a short lateral bend before bending back medially Fig. 2. Representative sagittal sections through the head of a female Desmognathus fuscus (Plethodontidae, Desmognathini; FMNH 86979; hematoxylin and eosin). The sections were ordered from most lateral (A) to most medial (E; scale bar mm). Low magnification images in the left column, a through e, were used to provide orientation of internal features and correspond to A F (scale bar mm). The boxes in the low magnification images represent the approximate regions from where sections (A E) were isolated. (A C) Sealing of the eyelids (white arrowheads) at the medial canthus of the eye. (D and E) The lack of formation of a nasolacrimal duct that opens into the nasal cavity (black arrow). Other symbols: gray arrow, tubules of orbital glands; white arrow, external nasal glands.

7 6 SIEGEL ET AL. Fig. 3.

8 NASOLACRIMAL DUCTS OF PLETHODONTIDAE 7 just before passage through the nasolacrimal canals. In Aneides, a sharp medial bend was observed as the nasolacrimal ducts pass through the nasolacrimal canals with a sharp bend laterally before emptying into the nasal cavities. The opposite was true of the nasolacrimal ducts in Bolitoglossa, Hemidactylium, and Oedipina; that is, a lateral bend after passage through the nasolacrimal canals with a bend back medially before insertion into the nasal cavities. A lateral bend of the nasolacrimal ducts was observed in Ensatina and Pseudoeurycea where the nasolacrimal ducts passed through the nasolacrimal canals with no medial bend. Nasolacrimal ducts of Rhyacotriton were similar when compared to plethodontids with only a few exceptions. The nasolacrimal ducts communicated with the nasal cavities far more rostral in Rhyacotriton (near the opening of the nostrils). These openings were still in close proximity to where the nasal cavities and vomeronasal organs joined, as the vomeronasal organs were also more rostrally located in Rhyacotriton. In all plethodontids, the nasolacrimal ducts traversed rostral to the nasolacrimal canals before curving caudally to insert into the nasal cavities after passage through the nasolacrimal canals. In Rhyacotrtion, the nasolacrimal ducts passed through the nasolacrimal canals and inserted into the nasal cavities without caudal curving. Histologically, the nasolacrimal ducts were lined by a bistratified epithelium in all species examined (Fig. 4). The principle cell layer of the epithelium (the most apical layer) was always columnar with basal nuclei. The cytoplasm of the principle cells was eosinophilic, and no cilia, microvilli, or other apical modifications to the epithelial linings could be found at the level of light microscopy (Fig. 4). Basal cells were cuboidal to squamous in shape and possessed nuclei that filled the majority of the cytoplasm (Fig. 4). Although the taxon sampling was inadequate to perform detailed statistics, there did not appear to be a trend between nasolacrimal duct diameter (measured with a reticle as the mean duct diameter for each species at its origin and where it passed through the nasolacrimal canal) and mean snout-vent length (SVL; r ). This was highlighted by one of the largest species examined, Pseudotriton ruber, possessing relatively small diameter nasolacrimal ducts (mean mm, SD mm; mean SVL mm) versus a relatively small species, H. scutatum, possessing relatively large diameter nasolacrimal ducts (mean mm, SD mm; mean SVL mm). Phylogenetic reconstruction using genera examined histologically provided a topology similar to the recent large-scale phylogenetic reconstruction of Pyron and Fig. 4. High magnification sagittal section through the nasolacrimal duct of a female Aneides aeneus (Plethodontinae, Aneidini; FMNH ; hematoxylin and eosin; scale bar 5 50 mm). The epithelium of the nasolacrimal ducts of plethodontids is bistratified. Principle cells (Pc; apical cell layer) are columnar with basal nuclei and an eosinophilic cytoplasm. Basal cells (Bc) are more cuboidal with nuclei filling the vast majority of the cytoplasmic space. Other abbreviations: asterisk, nasolacrimal duct lumen; Mgl, mucous gland; Nld, nasolacrimal duct. Wiens (2011; Fig. 5). The recovery of Hemidactylium as the sister lineage to Bolitoglossini 1 Batrachosepini (with low support) instead of sister to Bolitoglossini 1 Batrachosepini 1 Spelerpini (Pyron and Wiens, 2011) was the only major inconsistency (Fig. 5), a variation that was also recovered by Shen et al. (2016) with great support. Ancestral state reconstruction recovered the presence of nasolacrimal ducts on the branch leading to plethodontids and all lineages within Plethodontidae with great support, except Eurycea spelaea 1 E. tynerensis and Desmognathus (Fig. 5). Thus, the analysis recovered loss of nasolacrimal ducts twice within Plethodontidae (Fig. 5). Loss of nasolacrimal ducts on the branch leading to E. spelaea 1 E. tynerensis resulted in the non-patent condition observed in E. spelaea evolving from ancestors that lacked nasolacrimal ducts; however, the loss of nasolacrimal ducts on the branch leading to E. spelaea and E. tynerensis was not accompanied by great support (Fig. 5). DISCUSSION Examination of nasolacrimal ducts at the histological level revealed that the majority of taxa inspected Fig. 3. Representative transverse sections through the head of a female Eurycea longicauda (Hemidactyliinae, Spelerpini; DSS 0094; hematoxylin and eosin). The sections were ordered from most caudal (A) to most rostral (G; scale bar mm). Low magnification sagittal (a f; left column; scale bar 5 2,000 mm; from a non-catalogued specimen of E. longicauda in the private collection of DSS, used in a recent study on orbital gland sexual dimorphism in E. longicauda; Siegel et al., 2017) and transverse sections (a 1 f 1 ; center column; scale bar 5 1,000 mm) were used to provide orientation of internal features. The dotted lines a f represent the approximate plane of section for corresponding transverse sections A F. The boxes in the low magnification images (a 1 f 1 ) represent the approximate regions from where sections (A f) were isolated. (A and B) Formation of the nasolacrimal duct (black arrowhead) just rostral to the eye from the sealing of the eyelids at the medial canthus. (C E). Rostral projection of the superior portion of the nasolacrimal duct (black arrowhead) that originates at the medial canthus and caudal projection of the inferior portion of the nasolacrimal duct (white arrowhead) that empties into the nasal cavity. (F) Joining of the superior (black arrowhead) and inferior (white arrowhead) portions of the nasolacrimal duct at the caudal bend of the nasolacrimal duct. Other symbols: black arrow, nasal opening; black asterisks, prefrontal bone; gray arrow, choana; grey arrowhead, cartilago obliqua; gray asterisks, nasal bone; white arrow, vomeronasal organ opening; white asterisks, maxilla.

9 8 SIEGEL ET AL. Fig. 5. Bayesian reconstruction of nasolacrimal ducts within Plethodontidae. Three states were considered: present (white), absent (black), and present but not patent (gray). Pie charts at each node provide the probability of each state. Numerical values indicate posterior probability of the phylogenetic reconstruction and the highest probable state at each node. The phylogeny was reconstructed using Bayesian analysis of Rag1 sequences (see Materials and Methods). Abbreviations: m, metamorphic; p, paedomorphic; Nld, nasolacrimal duct. possessed nasolacrimal ducts. This included members of both subfamilies of Plethodontidae recognized by Wake (2012; i.e., Hemidactyliinae and Plethodontinae) and every tribe within except Desmognathini. Only two taxa lacked nasolacrimal ducts (Desmognathus fuscus and Eurycea tynerensis) and one taxon (E. spelaea) possessed ducts that did not come into contact with the nasal cavities. Not surprisingly, Bayesian ancestral state reconstruction on a novel phylogeny reconstructed from the genera/taxa used for histological examination revealed that the ancestor to Plethodontidae possessed nasolacrimal ducts. These ducts were subsequently lost within Spelerpini and somewhere within Desmognathini or potentially the branch leading to Desmognathini. The results presented here coincide well with the osteological descriptions from Wake (1966). Using gross examination Wake (1966) described variation in the cranial osteology along the routes of nasolacrimal ducts in all subfamilies of plethodontids and grouped different genera based on similarities. While Wake s (1966) osteological descriptions were detailed, it is difficult to ascertain specifics about the actual nasolacrimal ducts. Our only discrepancy with his work was that Wake (1966) does not note the non-patent condition we uncovered in Eurycea (then Typhlotriton) spelaea, a troglodytic taxon (Stejneger, 1892); however, again, Wake s (1966) descriptions were more focused on cranial osteology, and he described similar nasolacrimal duct routes of E. spelaea over the cranial bones as observed in Gyrinophilus and Stereochilus. Although Wake (1966) noted that the nasolacrimal ducts open in the nasal cavities in Gyrinophilus, he only stated that the ducts of E. spelaea followed similar routes. It is unclear if Wake (1966) followed the nasolacrimal ducts of E. spelaea all the way to the nasal cavities. Although Wake (1966) examined E. tynerensis, it is doubtful that this species was examined for nasolacrimal duct routes as only one specimen was available to Wake (1966). We found no evidence of nasolacrimal ducts in this species as a paedomorphic adult or metamorphic adult, and Wake (1966) does not mention this taxon in his section on the route of nasolacrimal ducts. Considering orbital glands and nasolacrimal ducts do not develop until metamorphosis (Wilder, 1925) the lack of nasolacrimal ducts in paedomorphic E. tynerenesis is not surprising. At present, we have no sound hypothesis as to why these ducts are also lacking in E. tynerensis that complete metamorphosis; however, as Wake (1966) noted, the metamorphic variability within Eurycea is prominent, and it is possible that any variation observed in this genus is due to this variability. Conversely, our results, and those of others, are inconsistent with those presented by Rehorek et al. (2013). In their Table 2, Rehorek et al. (2013) listed the nasolacrimal ducts as absent in every plethodontid examined in their study (Desmognathus monticola, Eurycea bislineata, E. longicauda, Pseudotriton ruber, Plethodon cinereus, and P. glutinosus) and one examined by Wilder (1913; Desmognathus fuscus). Two of the species examined in the current study were examined by Rehorek et al. (2013), E. longicauda and P. ruber, and these taxa clearly possessed nasolacrimal ducts (for complete path of the nasolacrimal ducts in E. longicauda, see Fig. 3).

10 NASOLACRIMAL DUCTS OF PLETHODONTIDAE 9 Fig. 6. Sagittal section through the head of Eurycea bislineata (Hemidactyliinae, Spelerpini) from a slide used in Wilder (1925; sex unknown; appears to be stained with hematoxylin and eosin; scale unknown), digitally scanned by DMS. A nasolacrimal duct is clearly observed projecting rostral, bending ventral, and then opening into the nasal cavity after a slight caudal bend (black arrow). Although we did not examine P. cinereus and P. glutinosus, we examined another species from this genus (P. serratus), and this species also possessed nasolacrimal ducts. Furthermore, Wirsig-Wiechmann et al. (2002), Dawley et al. (2006) and Dawley and Bass (1988, 1989) clearly labeled the nasolacrimal ducts emptying into the nasal cavities near the vomeronasal organ in P. shermani and P. cinereus. Two other members of Plethodontinae that we examined (Aneides aeneus and Ensatina eschscholtzii) also possessed nasolacrimal ducts. Although we did not section heads from E. bislineata, one of us (DMS), digitally scanned all of Wilder s (1925) sections of E. bislineata from her monograph on amphibian metamorphosis. From these images, we concluded that metamorphosed individuals absolutely possessed nasolacrimal ducts. In addition, these ducts were clearly labeled in Wilder s (1925) hand-drawings of salamander head sections (see our Fig. 6, which corresponds to the hand drawing on Plate II, Figure 4b in Wilder, 1925). A consistency between Rehorek et al. (2013), Wake (1966), and the present study was the lack of nasolacrimal ducts within Desmognathini, which was originally noted in D. fuscus by Wilder (1913). Wake (1966) and Rehorek et al. (2013) added to the list of Desmognathini that lack nasolacrimal ducts through inspection of D. monticola (Rehorek et al. 2013) and D. marmoratus (Leurognathus in Wake, 1966), and also Phaeognathus (Wake, 1966). Considering there is good support for the inclusion of these taxa within a monophyletic group (Desmognathini) with Phaeognathus at the base of the lineage and the Desmognathus taxa examined embedded within the lineage at multiple levels (Pyron and Wiens, 2011), it is conceivable to hypothesize that nasolacrimal ducts were lost on the branch leading to Desmognathini and that no species within possess nasolacrimal ducts; however, this interpretation should be treated as nothing more than conjecture until a more complete sampling of Desmognathini has been accomplished. Recent

11 10 SIEGEL ET AL. phylogenetic analyses recovered the closest relative of Desmognathini as Aneides (Pyron and Wiens, 2011; Shen et al., 2016). Aneides aeneus (present study) undeniably possesses nasolacrimal ducts, and other studies have demonstrated the pathway of nasolacrimal ducts from the medial canthus to the nasal cavities in species from all tribes of the Plethodontinae; e.g., Ensatina (Wake, 1966; this study), Hydromantes (Wake, 1966), Karsenia (Buckley et al., 2010), and Plethodon (Higgins, 1920; Wake, 1966; Dawley and Bass, 1988, 1989; Wirsig- Wiechmann et al., 2002; Dawley et al., 2006; this study). These data provide definitive support that nasolacrimal ducts were lost on the branch leading to Desmognathini and not in a more distant ancestor. While understanding the evolution of nasolacrimal ducts through presence/absence optimization onto a phylogeny is interesting, of more importance is what is the benefit or lack thereof for either state in different plethodontid salamander habitats; that is, does this suggest anything about the natural history of plethodontid salamanders? Unfortunately, the answer to this question is no, as too few data are available that have actually tested functional hypotheses on nasolacrimal ducts. It is clear, however, that most terrestrial vertebrates possess nasolacrimal ducts (for review see Hillenius and Rehorek, 2005), that nasolacrimal ducts transfer fluids secreted by orbital glands to the nasal cavities, and that the nasolacrimal ducts (if present) also carry orbital gland fluids to the vomeronasal organs (if present) because of their close association with the nasal cavities, as demonstrated in at least frogs (Hillenius et al., 2001; Nowack and W ohrmann-repenning, 2010), caecilians (Schmidt and Wake, 1990), and snakes (Rehorek et al., 2000a,b). A common functional hypothesis for nasolacrimal ducts is that orbital gland secretions pass through these ducts and enhance vomeronasal organ function (Hillenius, 2000), potentially through solubilizing water insoluble pheromones, which was effectively demonstrated in Red-sided Gartersnakes (Thamnophis sirtalis; Huang et al., 2006). Current analysis of pheromones in salamanders would tend to reject that orbital gland secretions function similarly to dissolve water insoluble pheromones, as plethodontid pheromones are water soluble (e.g., Feldhoff et al., 1999; Rollmann et al., 1999). If orbital gland secretions do function to enhance vomeronasal organ function in some other manner, it is possible that Desmognathini and a select few Spelerpini rely less on pheromonal communication because of the lack of nasolacrimal ducts. Less reliance on pheromonal communication could potentially be a convergent feature because of similar environmental variables, but this hypothesis requires testing. To our knowledge, only primates (Rehorek and Smith, 2006) and bats (Rehorek et al., 2010) also variably possess nasolacrimal ducts, and turtles lack nasolacrimal ducts altogether (Hillenius and Rehorek, 2005). In conclusion, we have presented data that may draw interest to a variable characteristic found in plethodontid salamanders. Although plethodontid salamanders have many unique and interesting characteristics, the complete lack of nasolacrimal ducts is not one of them. Curiously, while most plethodontid lineages possess nasolacrimal ducts, nasolacrimal ducts have been lost at least twice during their diversification. This makes plethodontids a fruitful model for studying the evolution/ function of nasolacrimal ducts in the future. ACKNOWLEDGMENTS We thank are institutions of employment for continued support of our research. KW Conway reviewed an earlier version of this manuscript and provided above average suggestions. RD Aldridge and MR Parker provided useful conversation on construction of the discussion. C Lee, an undergraduate student at Southeast Missouri State University, stained heads from select taxa. We thank SG Tilley and Smith College for access to the IW Wilder histological slide collection. LITERATURE CITED Alberch P Morphological variation in the neotropical salamander genus Bolitoglossa. Evol 37: AmphibiaWeb University of California, Berkeley, CA, USA. Accessed 18 Aug Bertmar G The vertebrate nose, remarks on its structural and functional adaptation and evolution. Evol 23: Bonett RM, Chippindale PT Speciation, phylogeography and evolution of life history and morphology in plethodontid salamanders of the Eurycea multiplicata complex. Mol Ecol 13: Buckley D, Wake MH, Wake DB Comparative skull osteology of Karsenia koreana (Amphibia, Caudata, Plethodontidae). J Morphol 271: Dawley EM, Bass AH Organization of the vomeronasal organ in a plethodontid salamander. J Morphol 198: Dawley EM, Bass AH Chemical access to the vomeronasal organs of a plethodontid salamander. J Morphol 200: Dawley EM, Nelsen M, Lopata A, Schwartz J, Bierly A Cell birth and survival following seasonal periods of cell proliferation in the chemosensory epithelia of red-backed salamanders, Plethodon cinereus. Brain Behav Evol 68: Feldhoff RC, Rollmann SM, Houck LD Chemical analyses of courtship pheromones in a plethodontid salamander. In: Johnson RE, M uller-schwarze D, Sorenson PW, editors. Chemical signals in vertebrates volume 10. New York, NY: Springer Science1Business Media, Inc., p Higgins GM The nasal organ in Amphibia. Ill Biol Mono VI: Hillenius WJ Septomaxilla of nonmammalian synapsids: Soft-tissue correlates and a new functional interpretation. J Morphol 245: Hillenius WJ, Rehorek SJ From the eye to the nose: Ancient orbital to vomeronasal communication in tetrapods. In: Mason RT, LeMaster MP, M uller-schwarze D, editors. Chemical signals in vertebrates volume 10. New York, NY: Springer Science1Business Media, Inc., p Hillenius WJ, Watrobski LK, Rehorek SJ Passage of tear duct fluids through the nasal cavity of frogs. J Herp 35: Huang G-Z, Zhang J-J, Wang D, Mason RT, Halpern M Female snake sex pheromone induces membrane responses in vomeronasal sensory neurons of male snakes. Chem Senses 31: Kiernan JA Histological and histochemical methods: Theory and practice. New York, NY: Pergamon Press, Inc. p 425. Kohler G A new species of salamander of the genus Nototriton from Nicaragua (Amphibia: Caudata: Plethodontidae). Herpetologica 58: LaPage EL The septomaxillary. I. In the Amphibia Urodela. J Morphol Physiol 45: Nowack C, W ohrmann-repenning A The nasolacrimal duct of anuran amphibians: Suggestions on its functional role in vomeronasal perception. J Anat 216: Nylander JAA MrModeltest v2. Program distributed by the author. Evolutionary Biology Centre, Uppsala University.

12 NASOLACRIMAL DUCTS OF PLETHODONTIDAE 11 Pagel M, Meade A, Barker D Bayesian estimation of ancestral character states on phylogenies. Syst Biol 53: Pyron RA, Wiens JJ A large-scale phylogeny of Amphibia including over 2800 species, and a revised classification of exant frogs, salamanders, and caecilians. Mol Phylo Evol 61: Rehorek SJ, Firth BT, Hutchinson MN. 2000a. The structure of the nasal chemosensory system in squamate reptiles. 2. Lubricatory capacity of the vomeronasal organ. J Biosci 25: Rehorek SJ, Hillenius WJ, Quan W, Halpern M. 2000b. Passage of the Harderian gland secretions to the vomeronasal organ of the snake. Thamnophis siratlis. Can J Zool 78: Rehorek SJ, Smith TD, Bhatnagar KP The orbitofacial glands of bats: An investigation of the potential correlation of gland structure with social organization. Anat Rec 293: Rehorek SJ, Smith TD The primate Harderian gland: Does it really exist? Ann Anat 188: Rehorek SJ, Grand-Pierre AE, Cummings JR, Jewell B, Constantine J, Hillenius WJ A re-examination and re-evaluation of salamander orbital glands. Anat Rec 296: Rollmann SM, Houck LD, Feldhoff RC Proteinaceous pheromone affecting female receptivity in a terrestrial salamander. Science 285: Ronquist F, Teslenko M, van der Mark P, Ayres D, Darling A, H ohna S, Larget B, Liu L, Suchard MA, Huelsenbeck JP MrBayes 3.2: Efficient Bayesian phylogenetic inference and model choice across a large model space. Syst Biol 61: Schmidt A, Wake MH Olfactory and vomeronasal systems of caecilians (Amphibia:Gymnophiona). J Morphol 205: Shen XX, Lian D, Chen MY, Mao RL, Wake DB, Zhang P Enlarged multilocus data set provides surprisingly younger time for origin for the Plethodontidae, the largest family of salamanders. Syst Biol 65: Siegel DS, Al-baghdadi YH, Frangello M, Herrboldt M, Alvino S, Trauth SE, Sever DM Sexual tears in a salamander? Zoomorphol 136: Stejneger L Preliminary description of a new genus and species of blind cave salamander from North America. Proc US Natl Mus 15: Wake DB Comparative osteology and evolution of the lungless salamanders, family Plethodontidae. Mem South Cal Acad Sci 4: Wake DB Taxonomy of salamanders of the family Plethodontidae (Amphibia: Caudata). Zootaxa 3484: Wilder IW The life history of Desmognathus fusca. Biol Bull 24: Wilder IW The morphology of amphibian metamorphosis. Northampton, MA: Smith College. p 161. Wirsig-Wiechmann CR, Houck L, Feldhoff PW, Feldhoff RC Pheromonal activation of vomeronasal neurons in plethodontid salamanders. Brain Res 952: APPENDIX: SPECIMENS EXAMINED Abbreviations: ASUMZ, Arkansas State University Museum of Zoology; CAS, California Academy of Sciences; DSS, personal collection of DS Siegel; FMNH, Field Museum of Natural History. Hemidactyliinae Batrachosepini: Batrachoseps attenuatus (FMNH , FMNH , FMNH , FMNH ); Bolitoglossini: Bolitoglossa subpalmata (FMNH , FMNH , FMNH , FMNH ), Oedipina cyclocauda (FMNH , FMNH ), Pseudoeurycea bellii (FMNH 65600, FMNH 65605, FMNH 65609, FMNH ); Spelerpini: Eurycea longicauda (DSS 0037, DSS 0039, DSS 0082, DSS 0094), Eurycea multiplicata (ASUMZ 4202, ASUMZ 5487, ASUMZ 5495, ASUMZ 6864), Eurycea spelaea (ASUMZ 4533, ASUMZ 8158, ASUMZ 9356, ASUMZ 9357), Eurycea tynerensis [metamorphs] (ASUMZ 18103, ASUMZ 18423, ASUMZ 18424, ASUMZ 18426), Eurycea tynerensis [paedomorphs] (ASUMZ 15339, ASUMZ 15359, ASUMZ 22946, ASUMZ 22947), Pseudotriton ruber (FMNH , FMNH , FMNH , FMNH ); Hemidactyliini: Hemidactylium scutatum (ASUMZ 4928, ASUMZ 4930, ASUMZ 5256, ASUMZ 5368). Plethodontinae Aneidini: Aneides aeneus (FMNH , FMNH , FMNH , FMNH ), Desmognathini: Desmognathus fuscus (FMNH 86913, FMNH 86979, FMNH 86986, FMNH 87102), Ensatinini: Ensatina eschscholtzii (CAS 13843, CAS 26996, CAS 26997, CAS ), Plethodontini: Plethodon serratus (ASUMZ 5547, ASUMZ 6787, ASUMZ 6842, ASUMZ 29943); Rhyacotritonidae Rhyacotriton olympicus (CAS 47622, CAS 47651, CAS 47679, CAS 47684).