ArticlePDF Available

Colour and size reveal hidden diversity of Necturus (Caudata: Proteidae) from the Gulf Coastal Plain of the United States

Authors:

Abstract and Figures

Recent data from the mitochondrial genome reveal six lineages of Gulf Coast Waterdogs traditionally classified as Necturus beyeri. Here, we use patterns of colour and body size, along with previously published data, to reveal a large, heavily spotted phenotype with an unstriped larva possessing numerous white spots; we re-describe N. beyeri to correspond to this phenotype. We also reveal a small, weakly spotted phenotype possessing an unstriped larva lacking numerous white spots. This phenotype characterises the Apalachicola and Escambia lineages, which current evidence suggests are paraphyletic. We reject taxonomies that place these two lineages in N. lodingi because the type specimen of this species appears to be a melanistic member of N. beyeri. Therefore, we describe the Apalachicola and Escambia lineages as independent new species.
Content may be subject to copyright.
Full Terms & Conditions of access and use can be found at
https://www.tandfonline.com/action/journalInformation?journalCode=tnah20
Journal of Natural History
ISSN: 0022-2933 (Print) 1464-5262 (Online) Journal homepage: https://www.tandfonline.com/loi/tnah20
Colour and size reveal hidden diversity of Necturus
(Caudata: Proteidae) from the Gulf Coastal Plain of
the United States
Craig Guyer, Christopher Murray, Henry L. Bart, Brian I. Crother, Ryan E.
Chabarria, Mark A. Bailey & Khorizon Dunn
To cite this article: Craig Guyer, Christopher Murray, Henry L. Bart, Brian I. Crother, Ryan E.
Chabarria, Mark A. Bailey & Khorizon Dunn (2020): Colour and size reveal hidden diversity of
Necturus (Caudata: Proteidae) from the Gulf Coastal Plain of the United States, Journal of Natural
History, DOI: 10.1080/00222933.2020.1736677
To link to this article: https://doi.org/10.1080/00222933.2020.1736677
View supplementary material
Published online: 19 Mar 2020.
Submit your article to this journal
Article views: 3
View related articles
View Crossmark data
Colour and size reveal hidden diversity of Necturus (Caudata:
Proteidae) from the Gulf Coastal Plain of the United States
Craig Guyer
a
, Christopher Murray
b
, Henry L. Bart
c
, Brian I. Crother
b
,
Ryan E. Chabarria
d
, Mark A. Bailey
e
and Khorizon Dunn
a
a
Department of Biological Sciences, Auburn University, Auburn, AL, USA;
b
Department of Biological Sciences,
Southeastern Louisiana University, Hammond, LA, USA;
c
Department of Ecology and Evolutionary Biology,
Tulane University, New Orleans, LA, USA;
d
Department of Biology, Lone Star College, The Woodlands, TX,
USA;
e
Conservation Southeast, Inc., Andalusia, AL, USA
ABSTRACT
Recent data from the mitochondrial genome reveal six lineages of Gulf
Coast Waterdogs traditionally classied as Necturus beyeri. Here, we use
patterns of colour and body size, along with previously published data,
to reveal a large, heavily spotted phenotype with an unstriped larva
possessing numerous white spots; we re-describe N. beyeri to corre-
spond to this phenotype. We also reveal a small, weakly spotted
phenotype possessing an unstriped larva lacking numerous white
spots. This phenotype characterises the Apalachicola and Escambia
lineages, which current evidence suggests are paraphyletic. We reject
taxonomies that place these two lineages in N. lodingi because the type
specimen of this species appears to be a melanistic member of
N. beyeri. Therefore, we describe the Apalachicola and Escambia
lineages as independent new species.
ARTICLE HISTORY
Received 14 May 2019
Accepted 25 February 2020
KEYWORDS
Taxonomy; morphology;
waterdogs; biogeography;
salamanders
Introduction
The genus Necturus traditionally contains ve species of paedomorphic salamanders
restricted to the eastern United States (Highton et al. 2017). Fossil and molecular evi-
dences indicate the genus to be ancient (originating at least 60 mya) but with all extant
species being derived within the southeastern United States about 5 mya (Bonett et al.
2013). Two species (N. lewisi and N. punctatus) inhabit drainages to the Atlantic Ocean and
all others are found largely in drainages of the Gulf of Mexico [hereafter, Gulf drainages,
with Gulf Coastal Plain referring to Southeastern Plains, Southern Coastal Plain, South
Central Plains, and Western Gulf Coastal Plain units of Level III ecoregions of the con-
tinental United States (U.S. Environmental Protection Agency 2013) associated with Gulf
drainages]. Waterdogs of the Gulf drainages currently include three named species,
N. alabamensis, N. beyeri, and N. maculosus (Highton et al. 2017). These three species are
inferred to share a common history, to the exclusion of those of the Atlantic drainages,
because they share derived karyotypes possessing 36 pairs of telocentric chromosomes
(Sessions and Wiley 1985), derived allelomorphs of glucose phosphate isomerase,
CONTACT Craig Guyer guyercr@auburn.edu
Supplemental data for this article can be accessed here
JOURNAL OF NATURAL HISTORY
https://doi.org/10.1080/00222933.2020.1736677
© 2020 Informa UK Limited, trading as Taylor & Francis Group
Published online 19 Mar 2020
isocitrate dehydrogenase, 1-lactate dehydrogenase, and malic enzyme (Guttman et al.
1990), and share sequence dierentiation of the mitochondrial genome (Chabarria et al.
2017). These evolutionary changes suggest an origin of Necturus in the Atlantic Coastal
Plain followed by dispersal of an ancestral population to the Gulf Coastal Plain that then
diverged into the currently recognised Gulf coastal species.
Clarication of waterdog evolution within the Gulf Coastal Plain has been hampered by
competing concepts of the content of N. beyeri, by faulty characterisation of the content
and geographic extent of N. alabamensis, and by confusion regarding the taxonomic
status of N. lodingi, an unspotted species described by Viosca (1937) based on the type
series from the Dog River of Mobile Bay, and of problematic status since (Bart et al. 1997).
Chabarria et al. (2017) recovered six monophyletic lineages of N. beyeri based on the
mitochondrial genome (Figure 1) and the concept of that species presented in Highton
et al. (2017), but this species was rendered paraphyletic by a sister-taxon relationship
between N. alabamensis (as diagnosed by Bart et al. 1997) and the Mobile lineage of
N. beyeri, and unresolved relationships of the Western and Pontchartrain lineages of
N. beyeri with N. maculosus. Additionally, the Apalachicola and Escambia lineages of
Chabarria et al. (2017) comprise waterdogs of the Gulf Coastal Plain that Hecht (1958)
considered to be N. punctatus lodingi, for unspotted and weakly spotted coastal speci-
mens, and to be N. beyeri alabamensis, for boldly spotted upland specimens. Likewise,
Neill (1963) considered the region occupied by the Apalachicola and Escambia lineages to
contain two taxa, a attened species living under submerged rocks and logs (incorrectly
placed in N. alabamensis) and a rounded species living in holes in stream banks (N. beyeri).
However, Mount (1975) considered all waterdogs from the Ochlockonee River westward
to belong to a single, variable species, N. beyeri. Recently, Dubois and Raaëlli (2012)
placed unspotted specimens from Mobile Bay eastward to the Ochlockonee River in
N. lodingi, a taxonomy rejected by Highton et al. (2017), who considered N. lodingi to be
a synonym for N. alabamensis. Thus, taxonomic confusion over the identity of waterdogs
of the Gulf Coastal Plain, especially for specimens from the Ochlockonee River to the
Mobile drainage, hinders study of these organisms.
Colour patterns, especially of larvae, play an important role in distinguishing species of
Necturus. Two colour phenotypes involve striped larvae. In N. lewisi, a single light middorsal
stripe is diagnostic of the species (Ashton and Braswell 1979). In N. alabamensis and
N. maculosus, bold light dorsolateral stripes on larvae are diagnostic (Hecht 1958;Bart
et al. 1997). All other forms, including all mitochondrial lineages of N. beyeri, possess an
unstriped larva, but the status of larvae from the type locality of N. lodingi has not been
described and variation of this phenotype among mitochondrial lineages of N. beyeri has not
been characterised. Based on the relationships recovered by Chabarria et al. (2017)and
parsimony, the unstriped phenotype is thought to have given rise, independently, to the two
striped phenotypes. Discovery of additional larval colour phenotypes would expand diag-
nostic character states that might clarify the taxonomy of Gulf Coastal Plain waterdogs.
Characterisation of species of Necturus based on colour phenotypes of adults is less
successful. Hecht (1958) used the size and number of dark dorsal, lateral, and ventral
spotting to characterise species of waterdogs. Unfortunately, these data were not sub-
jected to statistical analyses that diagnosed distinct populations. Thus, the degree to
which adult colour patterns dierentiate lineages deserves additional attention.
2C. GUYER ET AL.
Body size and shape also can distinguish species within Necturus. Bart et al. (1997) used
eight linear measurements of adult individuals to distinguish N. alabamensis from speci-
mens classied as N. beyeri. An elongate head and tail of N. alabamensis distinguish this
species from N. beyeri and likely represent adaptations of N. alabamensis to the rocky
habitat of the Upper Black Warrior River of north-central Alabama. Similarly, Bonett et al.
(2013) recovered body size as an important variable in the evolutionary divergence of
N. punctatus (small size) from N. lewisi (large size) and for divergence of N. maculosus (large
size) from N. beyeri and N. alabamensis (small size). Dubois and Raaëlli (2012) placed
N. beyeri, N. lodingi, and N. punctatus in a separate subgenus based largely on small body
size. Additional examination of patterns of body size, especially for the Apalachicola and
Escambia lineages relative to other lineages of N. beyeri, is important for understanding
speciation within the group.
Here, we use larval body colour and adult body colour, size, and shape to demonstrate
a dichotomy between the combined Mobile, Pearl, Pontchartrain, and Western lineages of
Chabarria et al. (2017) and the Apalachicola and Escambia lineages. We use this dichot-
omy to re-describe N. beyeri. We then use accumulated evidence to evaluate the status of
N. lodingi and to describe the Apalachicola and Escambia lineages as distinct species. We
take these actions to stabilise taxonomy for a region where competing taxonomies
currently yield confusion.
Materials and methods
We examined 862 specimens attributed to N. beyeri in the Auburn University Museum,
Florida Museum of Natural History, Louisiana State Museum of Natural History, Mississippi
Museum of Natural Science, Southeastern Louisiana University Vertebrate Museum, and
US National Museum of Natural History herpetological collections (Appendix; all museum
acronyms from Sabaj 2016). These specimens represented the Apalachicola, Escambia,
and Mobile lineages and a composite group containing the Pearl, Pontchartrain, and
Western lineages of Chabarria et al. (2017;Figure 1). The latter three lineages were pooled
because genetic distances for pairs of specimens between lineages were short enough to
cloud phylogenetic processes (e.g. Ballard and Whitlock 2004). We retained the Mobile
lineage as a separate group in the analysis to examine the degree to which this lineage
might be dierentiated from other heavily spotted lineages west of Mobile Bay and to
clarify the relationship of the Mobile lineage to N. lodingi and to relatively unspotted
lineages east of Mobile Bay. For 232 larval specimens (31186 mm total length), we
recorded each individual as exhibiting one of the following patterns of colour (Figure
2): an unspotted form (0 = no light spots on sides and dorsum) or a spotted form
(1 = numerous small light spots on sides and dorsum). For larger individuals, we used
external morphological features to discover the smallest individual within each lineage
that was a mature male (swollen cloaca; spur-like tip on each side of cloacal opening).
Within each lineage, this value was used to create three age-sex categories: juvenile (<
SVL of smallest known male), adult male (male features present and smallest male), and
adult female (smallest male but lacking morphological features of a male).
For 179 adult individuals, we recorded snout-vent length (SVL; tip of snout to posterior
edge of vent) measured with a ruler. We then used photographs of the dorsal, lateral, and
ventral aspect of each specimen, along with a mm ruler for scale, to record the following
JOURNAL OF NATURAL HISTORY 3
eight linear measurements (Figure 3; distance measured in Photoshop 6.0): distance from
anterior tip of snout to anterior corner of right eye (SRE), width of right eye as distance
from anterior to posterior corners (REW), head width as distance between posterior
corners of right and left eyes (HWE), snout length as shortest distance from line uniting
anterior corner of right and left eyes to tip of snout (SL), postorbital length as distance
from posterior corner of right eye to base of insertion of right anterior-most gill (POL),
head width as distance between anterior insertion of anterior-most right and left gills
(HWG), distance from anterior tip of snout to anterior-most point along gular fold (GL),
and widest body width across contralateral costal folds (WBW). Five additional measure-
ments were derived by calculating missing edges associated with right triangles dening
Figure 1. Geographic distribution and phylogeny of six mitochondrial lineages (boldface) of Gulf Coast
waterdogs typically assigned to Necturus beyeri. Tree is rooted by N. punctatus and paraphyly of
N. beyeri is shown by inclusion of N. alabamensis and two lineages of N. maculosus. Phylogenetic
construction and lineage names are from Chabarria et al. (2017). Drainages for each lineage are
Apalachicola lineage Apalachicola, Chipola, Choctawhatchee, Econna, Ochlockonee, and Pea Rivers;
Escambia lineage Blackwater, Escambia (Conecuh in Alabama), Perdido, and Yellow Rivers; Mobile
lineage Dog, Escatawpa, Fish, Mobile, Pascagoula, and Biloxi Rivers; Pearl lineage Pearl and Wolf
Rivers; Pontchartrain lineage Amite, Bayou Bonfouca, Bayou Lacombe, Blind, Tangipahoa,
Tchefuncte, and Tickfaw Rivers; Western lineage Calcasieu, Cedar Bayou, Old, Neches, Sabine, San
Jacinto, and Trinity Rivers.
4C. GUYER ET AL.
snout, eye, and gill regions or dierences between head width measurements (Figure 3).
These were snout width (SW; missing edge of right triangle associated with SL and SRE),
right eye depth (ED; dierence between HWE/2 and SW), right eye length (EL; missing
edge of right triangle associated with WRE and ED), gill depth (GD; dierence between
HWG/2 and HWE/2), and length from posterior eye to level of gill insertion (GEL; missing
edge of right triangle associated with POL and GD). Finally, we recorded a count of costal
grooves, which were counted as recommended in Highton (1957). Our measurements
were selected because they were associated with shape variables revealed by geometric
morphometrics (Chabarria et al. 2017). We avoided use of other traditional measurements
used to characterise salamander species because Hecht (1958) failed to nd dierences
among recognised species with these measurements.
We also recorded colour patterns of adults from our images. Dorsally, adults were
rated for dark spotting on the snout (0 = spotting absent or 1 = present), dark dorsal
markings [0 = no dark markings, 1 = small dark spots (no larger than size of eye), or
Figure 2. Dorsal colour pattern of Necturus juveniles showing unspotted (a) and spotted (numerous
white spots; b) character states.
Figure 3. Measurements taken of adult specimens from dorsal (a; AUM 35138) and ventral (b; AUM
18533) views. Solid lines of panel A are measurements taken from photographs of specimens (HWE,
HWG, POL, SL, SRE, WRE; see text for description); dotted lines of panel A are measurements derived
from solid lines (ED, EL, GD, GEL, SW; see text for description). Solid lines of panel B are measurements
taken from photographs of specimens (GL, WBW; see text for description).
JOURNAL OF NATURAL HISTORY 5
2 = large dark spots (larger than size of eye)], and dark lateral markings (0 = no dark
markings, 1 = small dark spots, or 2 = large dark spots; Figure 4). Ventrally, individuals
were rated for spotting on the chin (0 = dark spots absent, 1 = spots present along the
mandibles, or 2 = dark spots across chin to level of gular fold) and dark ventral markings
(0 = belly immaculate, 1 = some dark spots ventrolaterally, or 2 = dark spots across
entire belly; Figure 5).
First, we examined patterns of larval colour by generating a contingency table of the four
groups (Apalachicola, Escambia, and Mobile lineages plus thecombined remaining lineages
of N. beyeri) crossed with the two categories of colouration (Figure 2). A chi-square test was
used to examine dierences in colour frequencies among groups. Next, we used canonical
discriminant analysis (PROC CANDISC of SAS; SAS Institute Inc. 2008) on the dataset of adult
individuals to generate multivariate axes maximising separation of the three lineages from
each other and from the combined group of remaining lineages. To remove the eect of
dierences in overall body size among individuals, each linear measurement was divided by
the geometric mean of all linear measurements taken on that specimen (Mosimann and
James 1979). Values from the total canonical structure were used to recover the variables
that most strongly separated the four groups. For linear variables that loaded heavily on
signicant canonical axes, we used two-way (sex and group), non-parametric ANOVA
(ARTool program in R; R Core Team 2016) of untransformed variables to reveal patterns
that might be useful in identifying the four groups in eld or museum settings. Non-
parametric tests were used because the untransformed variables displayed extensive
heteroscedasticity. Sex was included as a xed eect to reveal patterns of sexual dimorph-
ism within groups, a common feature of size and shape in salamanders. For colour variables
that loaded heavily on signicant canonical axes, contingency tables (groups crossed with
colour categories of Figures 4 or 5) were generated. Proportional colour dierences among
groups were tested with chi-square analyses, using exact probabilities to account for low
observed and expected cell counts (PROC FREQ of SAS).
Figure 4. Dorsal and lateral colour patterns of Necturus showing unspotted snout, dorsum, and sides
(a), unspotted snout with small dark dorsal and lateral spots (b), and spotted snout with large dark
dorsal and lateral spots (c).
6C. GUYER ET AL.
Because the multivariate analyses documented that adult colour patterns played
a strong role in separation among lineages, we performed a nal pair of canonical
discriminant function analyses. These tests used the adult colour variables to determine
the degree to which these characters can be used to distinguish the Apalachicola lineage
from the Escambia lineage; we then repeated the analysis to determine the degree to
which colour variables can distinguish the Mobile lineage from the pool of all other
remaining lineages of N. beyeri. Values of misclassication based on cross-validation
(PROC DISCRIM of SAS) were used to assess the utility of these variables in identifying
lineages in eld and museum settings.
Results
For larvae, we found signicant dierences in colour pattern among the Apalachicola,
Escambia, and Mobile lineages and the pooled remaining lineages of N. beyeri (Figure 6;
χ
2
= 204.2, df = 3, P< 0.0001), with the Apalachicola and Escambia lineages having
equivalent frequencies and a strong mode of unspotted juveniles (χ
2
= 1.9, df = 1,
P= 0.36). Larval colour patterns were similar for the Mobile lineage compared with
a composite of all remaining lineages of N. beyeri (χ
2
= 3.4, df = 1, P= 0.12), with both
sharing a high proportion of larvae possessing numerous white dorsal and lateral spots.
Figure 5. Ventral colour patterns of Necturus showing immaculate chin and belly (a), limited chin
spotting and narrow white mid-venter (b), and spotted chin and belly (c).
JOURNAL OF NATURAL HISTORY 7
Within the Apalachicola and Escambia lineage, 8% and 2%, respectively, of larvae exam-
ined were judged to possess white dorsal spotting, but these spots were not as numerous
as those characterising the other two groups. Within the Mobile lineage and the compo-
site of all other lineages of N. beyeri, 18% and 7% of specimens, respectively, lacked white
dorsal spotting.
Three signicant canonical variables were recovered from our multivariate analysis of
adult colour and morphology. Canonical variable 1 (F= 10.4; df = 57, 424; P<.0001;70%
explained variance) loaded heavily and positively on COLC, and COLV and heavily and
negatively on SVL. This axis separated the Mobile lineage (negative scores; larger size
associated with a lack of ventral spotting) from the Apalachicola and Escambia lineages
and a group containing all remaining lineages of N. beyeri (positive scores; smaller size or
large size associated with ventral spotting; Figure 7(a)). Canonical variable 2 (F= 5.7; df = 36,
286; P< .0001; 23% explained variance) loaded heavily and positively on COLL and COLD.
This axis separated the Escambia and Apalachicola lineages (negative scores; at most, small
spots on dorsum and sides) from the Mobile lineage and the pooled remaining lineages of
N. beyeri (positive scores; large dark spots on dorsum and sides; Figure 7(a)). Canonical
variable 3 (F= 3.0; df = 17, 144; P= .0002; 7% explained variance) loaded heavily and
negatively on GL. This axis further separated the Escambia lineage (positive scores; longer
size-free gular distance) from the other two lineages plus the pooled remaining lineages of
N. beyeri (shorter size-free gular distances; Figure 7(b)).
Examination of body size of adults revealed signicant dierences among the three
lineages and the pooled remaining lineages of N. beyeri (F= 11.7; df = 3, 142; P< 0.0001)
but not between sexes (F= 0.7; df = 1, 142; P= 0.39). There was no signicant interaction
between sex and lineage (F= 1.6; df = 3, 142; P= 0.19) indicating no size dimorphism
among groups (Figure 8). Nevertheless, the Mobile and pooled remaining lineages of
N. beyeri were larger than the Apalachicola and Escambia lineages.
Figure 6. Proportion of specimens of the Apalachicola, Escambia, and Mobile lineages, along with
a composite group of the remaining western lineages of Necturus beyericonforming to two juvenile
colour phenotypes (see text for explanation).
8C. GUYER ET AL.
A composite score based on dorsal and lateral patterns of dark spotting of adults
diered signicantly among the three lineages and the pooled remaining lineages of
N. beyeri (χ
2
= 86.4, df = 12, P< .0001; Figure 9(a)). The Apalachicola and Escambia lineages
did not dier from each other in adult dorsal pattern (χ
2
= 4.8, df = 3, P= 0.16; modal
category of small dark dorsal and lateral spots) but the Mobile lineage and the combined
remaining lineages of N. beyeri diered in this colour pattern (χ
2
= 16.2, df = 3, P= 0.0002;
stronger mode of large dark dorsal and lateral spots in the remaining lineages of N. beyeri
than in the Mobile lineage). A composite score for the chin and venter also diered
signicantly among the three lineages and the composite of remaining lineages of
N. beyeri (χ
2
= 119.1, df = 12, P< 0.0001). Here, the Escambia and Mobile lineages were
not signicantly dierent, both sharing venters that lacked dark spotting (χ
2
= 1.1, df = 1,
P= 0.88). However, the remaining two groups diered in modal colour categories
(χ
2
= 21.3, df = 4, P= 0.0001; Figure 9(b); mode of chin with mandibular spotting and
Figure 7. Plots of canonical variables 1 and 2 (a) and canonical variables 1 and 3 (b). Solid square (blue
polygon) = Mobile lineage; solid circle (yellow polygon) = Escambia lineage; solid triangle (red
polygon) = Apalachicola lineage; open circles = composite of all western lineages of N. beyeri.
JOURNAL OF NATURAL HISTORY 9
ventrolateral spotting for Apalachicola lineage; mode of spotting across chin and venter in
the combined remaining lineages of N. beyeri).
When examined across all ve colour variables, a statistically signicant discriminant
function (F= 7.2; df = 6, 20; P= .0003) distinguished the Apalachicola lineage from the
Escambia lineage. This function was weighed most heavily by COLC and COLV.
Classication error rates based on cross-validation were 15% for the Apalachicola lineage
and 21% for the Escambia lineage. Similarly, a statistically signicant discriminant function
(F= 21.0; df = 6, 130; P< .0001) distinguished the Mobile lineage from the composite of
remaining lineages attributed to N. beyeri. This function was weighted most heavily by
COLC and COLV. Classication error rates based on cross-validation were 11% for the
Mobile lineage and 27% for the composite of remaining lineages.
Discussion
Overall, we recovered a clear separation of the Escambia and Apalachicola lineages from all
other lineages ascribed to N. beyeri. In particular, our data reveal two phenotypes of larvae
previously categorised as unstriped. One lacks evidence of striping and lacks numerous
small white spots. This phenotype characterises the Escambia and Apalachicola lineages as
well as N. punctatus. The association of this phenotype with these two lineages is strong,
with less than 10% of larvae from this region showing evidence of white spotting. Based on
phylogenetic relationships inferred by Sessions and Wiley (1985) and Guttman et al. (1990),
we assume this phenotype represents the ancestral condition for the genus. The second
Figure 8. Body sizes (snout-vent length) for the Apalachicola, Escambia, and Mobile lineages, along
with a composite group of the remaining western lineages of Necturus beyeri. Adult males (M) and
females (F) are shown separately within groups. Large dark dots are means; small dashes are individual
measurements; outline represents distribution of measurements.
10 C. GUYER ET AL.
larval phenotype is characterised by a lack of striping and presence of numerous white
spots, which we infer is a derived condition evolving rst in the common ancestor of the
Mobile, Pearl, Pontchartrain, and Western lineages of N. beyeri and then is retained by all
those lineages. Again, the association of this phenotype with these lineages is strong, with
less than 11% of larvae lacking evidence of white spotting. Given the utility of larval colour
Figure 9. Proportional representation of composite score for COLD and COLL (a; category 2 = no dark
dorsal or lateral spots; category 3 = small dark dorsal spots withno dark lateral spots or small dark lateral
spots with no dark dorsal spots; category 4 = small dark dorsal and lateral spots; category 5 = small dark
dorsal spots with large dark lateral spots or large dark dorsal spots with small dark lateral spots; category
6 = large dark dorsal and lateral spots) and composite score for COLC and COLV (b; category 2 = no dark
spots on chin or venter; category 3 = dark chin spots on mandibles and no dark spots on venter or no
dark spots on chin but small dark ventrolateral spots on venter; category 4 = dark chin spots on
mandibles and small dark ventrolateral spots on venter; category 5 = dark chin spots across entire chin
and small dark ventrolateral spots on venter or dark chin spots on mandibles and small dark spots across
venter; category 6 = dark spots across chin and venter). Colour categories are shown for Apalachicola,
Escambia, and Mobile lineages and a composite of all western lineages of N. beyeri.
JOURNAL OF NATURAL HISTORY 11
patterns in diagnosing other species of Necturus, we propose the white-spotted larva as an
autapomorphy diagnosing N. beyeri. Our analysis of adult body size indicates that this
revised concept of N. beyeri has larger average adult body sizes than the Apalachicola and
Escambia lineages. Additionally, adult body patterns indicate that the revised concept of
N. beyeri contains lineages that dier statistically in the size and distribution of spotting
relative to the Apalachicola and Escambia lineages.
Genetic distances between the Mobile, Pearl, Pontchartrain, and Western lineages are
short (Chabarria et al. 2017), corresponding to dierentiation among populations that
may not reect phylogenetic history (Ballard and Whitlock 2004). For this reason, addi-
tional study is needed to determine whether these mitochondrial lineages are popula-
tions within a single metapopulation lineage or are multiple cryptic species. Genetic
distances between these lineages and N. alabamensis and N. maculosus also are short.
However, N. alabamensis and N. maculosus share a derived striped larva, a feature that we
use to exclude these two species from the Mobile, Pearl, Pontchartrain, and Western
lineages of N. beyeri. This interpretation is bolstered by the observation that
N. alabamensis and the Mobile lineage of N. beyeri occupy the same major drainage but
show no evidence of hybridisation [narrow zone of sympatry with no evidence of inter-
mediate larval colouration (Bart et al. 1997); apparent morphological adaptation of
N. alabamensis to swift water with abundant slab-rock crevices (Bart et al. 1997); reciprocal
monophyly of the mitochondrial genome of specimens from above and below the zone of
contact (Chabarria et al. 2017)]. We interpret this evidence to support the hypothesis that
N. alabamensis and the Mobile lineage of N. beyeri are two biological species rather than
representing metapopulation structure within a single species. For these reasons, we
retain N. alabamensis and N. maculosus as separate species. This interpretation suggests
that the molecular clock within Necturus is ticking at a slower rate than is observed for
other salamander species of the Gulf Coastal Plain (e.g. Wray et al. 2017) and, therefore,
that the short divergence distances are not necessarily indicative of dierentiation among
populations within a single species. Our analysis of the Mobile lineage of N. beyeri
demonstrates it to have lower modal size, a tendency to possess smaller dark spotting
of the dorsum, and a tendency to have unspotted chin and belly regions, compared with
aggregated data from the Pearl, Pontchartrain, and Western lineages. Thus, our concept of
N. beyeri may mask further speciation within the group, but we refrain from recognising
additional taxa within our concept of N. beyeri until more precise measures of gene ow
among lineages are available and clearer evidence of diagnostic features emerge.
Our data for waterdogs from the region occupied by the Apalachicola and Escambia
lineages reveal that these salamanders are similar to N. punctatus in small body size,
limited number and size of spotting in adults, and lacking numerous white spots as larvae.
Based on our use of larval spotting to diagnose N. beyeri, the Apalachicola and Escambia
lineages must be excluded from that species. Similarly, the Apalachicola and Escambia
lineages are excluded from N. punctatus by derived numbers of telocentric chromosomes,
derived degree of heterochromatism of the Y chromosome, derived allelomorphs of
blood enzymes, and similarities of the mitochondrial genome that unite these two
lineages with the rest of the Gulf coastal waterdog species to the exclusion of
N. punctatus (Sessions and Wiley 1985; Guttman et al. 1990; Chabarria et al. 2017). This
evidence supports placement of the Apalachicola and Escambia lineages in some other
species than N. punctatus or our revised concept of N. beyeri, such as N. lodingi, or in newly
12 C. GUYER ET AL.
erected species. We note that adult Necturus are not known to cross overland barriers and
are intolerant of saltwater, isolating lineages within river basins that have limited oppor-
tunities to experience dispersal from other drainages, except for rare events such as
stream capture. These natural history observations provide further support for the
hypothesis that the Apalachicola and Escambia lineages are species with short sequence
divergence distances rather than representing divergent populations within a single
species.
Viosca (1937)describedN. lodingi for unspotted waterdogs collected from Eslava and
Halls Mill Creeks, major components of the Dog River, a small (230 km
2
) drainage in Mobile
County, AL. From the same locality, Viosca (1937) included specimens of a phenotype with
large dark spots in the type series of N. beyeri (Mobile lineage). Dubois and Raaëlli (2012)
broadened the concept of N. lodingi to include the entire range occupied by the
Apalachicola and Escambia lineages (lower Mobile drainage eastward to the Ochlockonee
River). However, this concept of N. lodingi is rendered paraphyletic by relationships of the
Escambia and Apalachicola lineages to each other and to other waterdog species of the Gulf
Coastal Plain (Figure 1). Additionally, despite eorts to conrm the presence of two
sympatric species in Eslava and Halls Mill Creeks, only specimens known to be of the dark-
spotted Mobile lineage have emerged and all larval specimens taken from these sites
display numerous white spots. Thus, no evidence conrms that an unspotted species with
an unspotted larva is present in the Mobile drainage and that it is referable to either the
Apalachicola or Escambia lineages. Therefore, we reject the concept of N. lodingi presented
by Dubois and Raaëlli (2012). Highton et al. (2017)considerN. lodingi to be a synonym of
N. alabamensis, but this designation appears to conictwithBartetal.(1997) who restricted
N. alabamensis to dorsoventrally attened, elongate waterdogs possessing striped larvae in
the Upper Black Warrior drainage. Thus, we reject the concept of N. lodingi presented in
Highton et al. (2017). Instead, we conclude that the Apalachicola and Escambia lineages
represent unnamed species that have phylogenetic anities with the heavily spotted
waterdogs to the west rather than with the phenotypically similar N. punctatus to the
east. The two lineages dier statistically in ventral colour pattern such that specimens
from the Apalachicola, Choctawhatchee, Pea, and Ochlockonee Rivers (Apalachicola line-
age) tend to have dull bellies created by invasion of dark spotting ventrolaterally. Specimens
from the Perdido, Escambia, and Yellow Rivers (Escambia lineage) have brilliant white bellies
lacking ventrolateral spotting. Our data indicate that, based on these colour features,
Apalachicola lineage specimens are misidentied approximately 15% of the time while
those of the Escambia lineage are misidentied approximately 21% of the time. Thus, the
frustration expressed by Mount (1975) in use of colour characters for diagnosing species
from the region occupied by the Apalachicola and Escambia lineages likely remains but is
perhaps tolerable given the challenge of eliminating uncertainty in the identity of speci-
mens from this region. We view the challenge of distinguishing the Apalachicola and
Escambia lineages to be no greater than that involved in distinguishing Desmognathus
apalachicolae from D. conanti or in distinguishing Eurycea hillisi and E. sphagnicola from
these same drainages (Means and Karlin 1989;Wrayetal.2017).
If the Apalachicola and Escambia lineages are distinct species, then we are still left with
clarifying the status of N. lodingi. We see four hypotheses to explain our inability to
document this species from the type locality. One is that N. lodingi is a distinct, but rare
species. We nd this hypothesis to be implausible because it requires sympatry with
JOURNAL OF NATURAL HISTORY 13
N. beyeri of the Mobile lineage, a feature that is rare in waterdogs, and requires restriction of
the species to the type locality. The second hypothesis is that N. lodingi represents migrants
of the Escambia lineage. Mount (1975) hypothesised movements of musk turtles
(Sternotherus) between the Escambia and Mobile drainages, and this process might be
invoked to explain rare occurrences of unspotted Necturus in drainages of Mobile Bay.
However, colour patterns of the type specimen of N. lodingi are inconsistent with the colour
patterns we have documented for the Escambia lineage. In particular, the chin of the type
for N. lodingi shows extensive pigmentation along the mandibles and across the anterior
portion of the chin (Viosca 1937), features found in no specimen of the Escambia lineage.
Similarly, the type specimen of N. lodingi has ventrolateral pigmentation that nearly covers
the venter, a feature observed in no specimen of the Escambia lineage. Therefore, we nd it
unlikely that N. lodingi represents dispersing members of the Escambia lineage. A third
hypothesis posits that the type specimen of N. lodingi represents a hybrid caused by
dispersal of the Escambia lineage into Mobile Bay. Although this hypothesis does character-
ise gene ow within Sternotherus of the Gulf Coast (Scott et al. 2018), we nd no evidence of
clinal variation within Necturus between the Escambia River and Mobile Bay, especially of
the larval phenotype, a pattern expected of hybridisation. The nal hypothesis, that
unspotted specimens of Necturus from the Dog River are aberrant colour morphs of the
Mobile lineage, is judged to be plausible because the type specimen of N. lodingi is most
similar in colour pattern to MMNS 1261, a melanistic specimen of the Mobile lineage from
Black Creek, Lamar County, MS (Figure 10). These specimens share an immaculate dark
dorsum, extensive ventrolateral dark markings, and dark chin markings, leading us to
conclude that the type of N. lodingi is a melanistic member of the Mobile lineage. Bart
et al. (1997) also noted that the type of N. lodingi appears to be melanistic. Thus, we
conclude N. lodingi is a junior subjective synonym of N. beyeri.
Based on the conclusions above, the taxonomy of Necturus of the Gulf Coastal Plain
requires revision. Below, we re-describe N. beyeri and describe the Apalachicola and
Escambia lineages as new species.
Species descriptions
Necturus beyeri Viosca (1937)(Figures 1112)
Necturus lodingi Viosca 1937
Necturus maculosus beyeri Schmidt, 1953,in part
Necturus beyeri beyeri Hecht 1958
Necturus punctatus beyeri Brode, 1969, in part
Necturus (Parvurus)beyeri Dubois and Raaëlli, 2012
Holotype. USNM 102674 (Figure 11(a,b)), an adult female from the Upper Calcasieu River
at Oakdale (no exact locality given), Allen Parish, LA. This locality is within the range of the
Western lineage of Chabarria et al. (2017).
Paratypes. USNM 102676; MCZ 17732, 17733 are listed in the original species description
(Viosca 1937). To these we add representatives from within the ranges of the Mobile (AUM
14 C. GUYER ET AL.
40150), Pearl (MMNS 1288, LSUMZ 62933, SLU 744), and Pontchartrain (LSUMZ 99424, SLU
6211, 6213) lineages (Figure 11(ch))
Diagnosis. Membership of this species in the genus Necturus is demonstrated by reten-
tion of external gills in adults, presence of pigmented skin, and reduction of digits on hind
limbs to four toes. Our concept of N. beyeri contains specimens of the Gulf Coastal Plain
possessing a larva with numerous small white spots and lacking dark stripes (Figure 2(b)).
Comparisons. Juveniles of N. beyeri have small white punctations, an apparent autapo-
morphy that distinguishes this species from N. alabamensis and N. maculosus (larvae with
two light dorsolateral stripes), N. lewisi (larvae with light middorsal stripe), and
N. punctatus and the Apalachicola and Escambia lineages (larvae lacking stripes and
white spots). Adults of N. beyeri typically possess large, bold, dark dorsal and lateral
spots and retain the white punctations of the juveniles, creating a particularly colourful
phenotype that distinguishes N. beyeri from N. punctatus and the Escambia and
Apalachicola lineages. Along with N. maculosus, the Western lineage of N. beyeri has 6
pairs of telocentric chromosomes and an increased level of heterochromatin in the
Y chromosome of males that dierentiate these taxa from N. lewisi and N. punctatus
(reduced heterochromatin, no telocentric chromosomes), and from the Apalachicola
Figure 10. Type specimen of N. lodingi (USNM 61752; Mobile County, AL) dorsum (a) and venter (b;
photos by Jenna L. Welch, Division of Amphibians and Reptiles, National Museum of Natural History,
Smithsonian Institution; used with permission); melanistic Mobile lineage specimen (MMNS 1261,
Lamar County, MS) dorsum (c) and venter (d; photos by Craig Guyer).
JOURNAL OF NATURAL HISTORY 15
lineage (4 pairs of telocentric chromosomes, intermediate levels of heterochromatin in
Y chromosome; Sessions and Wiley 1985). Additional sampling of the Escambia, Mobile,
Pearl, and Pontchartrain lineages are needed to demonstrate the complete distribution of
these character states.
Emended description of holotype. In preservation, the type specimen is 211 mm TOT,
144 mm SVL, and 67 mm TL; number of costal grooves is 16. The dorsal ground colour in
preservation is slate brown with numerous white punctations across the entire body. The
head has round dark-brown spots that are small anteriorly (about the same size as eye) and
enlarged and bolder posteriorly (starting at insertion of gills). The head lacks a dark stripe
from the nostril through the eye. The enlarged, bold, dark brown spots continue along the
dorsum and sides of body to the tip of the tail. The ground colour of the venter is uniform
tan, lacking the white punctations along the mid-venter but with such punctations gradu-
ally appearing ventrolaterally. The chin has small brown spots on skin covering the
Figure 11. Type specimen of N. beyeri (USNM 102674; Allen Parish, LA) dorsum (a) and venter (b;
photos by Jenna L. Welch, Division of Amphibians and Reptiles, National Museum of Natural History,
Smithsonian Institution; used with permission); paratype of Mobile lineage (AUM 40150, Covington
County, AL) dorsum (c) and venter (d; photos by Craig Guyer); paratype of Pearl lineage (MMNS 1288,
Saint Tammany Parish, LA) dorsum (e) and venter (f; photos by Craig Guyer); paratype of Pontchartrain
lineage (LSUMZ 99424, Tangipahoa Parish, LA) dorsum (g) and venter (h; photos by Craig Guyer).
16 C. GUYER ET AL.
mandibles; spots are absent from the midventral skin of the chin. The venter has small
brown spots midventrally, becoming enlarged laterally. The holotype has the following
values for measured morphological features described above ED: 2.9 mm; EL: 1.8 mm; GD:
4.7 mm; GEL: 20.6 mm; GL: 26.1 mm; HWE: 18.8 mm; HWG: 28.2 mm; POL: 21.1 mm; REW:
3.4 mm; SL: 8.1 mm; SRE: 10.4 mm; SW: 6.5 mm; WBW: 21.8 mm.
Variation. Necturus beyeri achieves relatively large adult sizes, with a maximum male size
of 184.0 mm SVL (+68 mm TL) and a maximum female size of 177.0 mm SVL (+66 mm TL).
Mean male size is 134.2 mm SVL (n= 53) and mean female size is 120.8 mm SVL (n= 94). In
dorsal aspect, the modal adult lacks snout spotting (73% of specimens), has large dorsal
(66% of specimens) and lateral (74% of specimens) spotting, and lacks chin (61% of
specimens) and ventral (45% of specimens) spotting. The Pearl lineage diverges from
this modal pattern in having a high percentage of snout (63%), chin (54%) and ventral
(56%) spotting, and the Mobile lineage diverges from the mode in having a high percen-
tage of small dorsal (54%) spotting and no chin (93%) or ventral (84%) spotting. The
modal value for costal grooves is 17 (n = 147), with 35% of specimens possessing 16
grooves, and 2% possessing 18.
Adult males (n= 53) have the following mean (and range) values for measured
morphological features ED: 1.2 mm (0.15.3); EL: 1.8 mm (0.13.8); GD: 3.3 mm (0.96.2);
GEL: 16.1 mm (10.42.2); GL: 21.1 mm (13.134.1); HWE: 12.6 (8.023.7); HWG: 19.5 (12.-
836.1); POL: 16.5 mm (10.626.6); REW: 2.4 mm (1.54.2); SL: 6.9 mm (3.212.0); SRE:
8.7 mm (4.914.7); SW: 5.2 mm (2.09.7); WBW: 19.0 (12.037.4).
Adult females (n= 94) have the following mean (and range) values for measured
morphological features ED: 0.8 mm (0.14.3); EL: 1.9 mm (0.14.3); GD: 2.9 mm (1.46.8);
Figure 12. Distribution of localities of specimens assigned to N. beyeri. Open circle is approximate
location of type specimen.
JOURNAL OF NATURAL HISTORY 17
GEL: 14.3 mm (7.127.9); GL: 19.1 mm (10.834.4); HWE: 11.2 mm (11.220.9); HWG:
17.0 mm (8.934.5); POL: 14.6 mm (7.328.5); REW: 2.3 mm (1.24.3); SL: 5.7 mm (2.712.4);
SRE: 7.6 mm (3.514.6); SW: 4.9 mm (2.010.2); WBW: 16.7 mm (8.036.8).
Etymology. The specic epithet is a patronym for George E. Beyer, a naturalist at Tulane
University early in the 1900s who provided summaries of the herpetofauna of the state of
Louisiana. The recommended English common name is Western Waterdog.
Distribution and natural history. Our re-description of Necturus beyeri restricts each
lineage to the following drainages: Mobile lineage Mobile (AL) to Biloxi (MS); Pearl lineage
Wolf (MS) to Pearl (LA); Pontchartrain lineage Bayou Bonfouca to Blind River (LA); Western
lineage Calcasieu (LA) to West Fork of the San Jacinto River (TX; Figure 12).
No life history study has been performed on the Mobile lineage. Specimens are known
from drainages entering Mobile Bay, where these salamanders can be found in dipnet
samples of leaf beds. Farther north, we have sampled them from rocky reaches of the
Tallapoosa River and from leaf packs under bridges of the Coosa River. The species is
found in the same drainage as N. alabamensis, but with the Mobile lineage occupying
slower-moving waters of the Coastal Plain, and N. alabamensis occupying swifter waters of
the terminus of the Sand Mountain formation. The two species are both present in the
Black Warrior River just east of Northport Lock and Dam, Yellow Creek at Lake Nichol Road
crossing, and the North River near Samantha, all in Tuscaloosa County, AL (Bart et al.
1997). These represent the only cases of sympatry involving Gulf Coast waterdogs. We
examined no specimen that appeared to represent a morphological hybrid. Gunter and
Brode (1964) collected specimens from the Biloxi and Pascagoula drainages, mostly on
hooks baited with earthworms. Undercut banks or overhanging stumps and tree trunks
associated with deep pools are key habitat features associated with this lineage (Gunter
and Brode 1964). Use of litter bags might increase detection of the Mobile lineage in
occupied streams. From this sampling technique, Eurycea cirrigera, E. quadridigitata, and
Desmognathus conanti are known associates (Lamb and Qualls 2013).
Within the Pearl lineage, males have swollen cloacal openings and motile sperm in
December and January, and detection of gravid females increases during November and
December, when mating likely occurs (Shoop 1965). Undercut banks or overhanging
stumps and tree trunks associated with deep pools are key habitat features associated
with detection of females and nests. Clutch size averages 57 eggs. Eggs are retained by
females and clutches are deposited in April or May in deep water under rocks, logs or
other sunken objects. Females attend the nests and can store sperm for at least six
months after mating (Sever and Bart 1996). Juveniles and subadults eat isopods, midges,
and mayies while adults feed on mayies and caddisies; prey consumption decreases in
warm months when leaf litter used by detritivorous prey is decreased (Bart and Holzenthal
1985). Specimens from coastal areas may occasionally be eaten by blue crabs (Gunter and
Brode 1964). The Pearl lineage often is infested with acanthocephalan parasites, for which
waterdogs are a denitive host. These salamander hosts likely become infected via
consumption of isopods, the intermediate host (Bart and Holzenthal 1985).
No life history study has been performed on the Pontchartrain lineage. However,
recent study of a dense population in Bayou Lacombe has documented the presence of
18 C. GUYER ET AL.
Batrachochytrium dendrobatidis and B. salamandrivorans in populations of this lineage
(Glorioso et al. 2017).
Brenes and Ford (2006)providetheonlypublishedeld study of the Western lineage of
N. beyeri. They examined the species in Gilley and Hill Creeks of Smith County, TX where
minnow traps placed along stream banks were used to capture waterdogs. The lineage was
detected more frequently during November through January; none was detected during
May, June, July, and August. This pattern was associated with a negative relationship
between captures and water temperature, with no captures occurring when water tem-
peratures exceeded 18°C. Reduction of activity at this temperature is similar to that reported
for N. lewisi by Braswell and Ashton (1985), suggesting similar thermal tolerance within the
entire genus. Recaptured individuals typically were found within 20 m of the initial capture,
but distances of up to 230 m were observed between captures. Males retained in aquaria
showed behaviours consistent with defence of underwater refugia during January and
February, the presumed breeding period for free-ranging individuals. Animals captured in
the eld were associated with sandy- or gravel-bottom regions, typically with logjams.
Remarks. The collection locality of the holotype likely is where LA highway 10 crosses
the Calcasieu River west of Oakdale (30.822529 -92.684595; WGS84).
Necturus mounti sp. nov.(Figures 1314)
Necturus punctatus lodingi Hecht, 1958,in part
Necturus beyeri alabamensis Hecht, 1958,in part
Necturus alabamensis Neill 1963,in part and misidentication
Necturus beyeri Mount 1975,in part
Necturus cf beyeri Bart et al. 1997,in part
Necturus (Parvurus)lodingi, Dubois and Raaëlli, 2012,in part
Necturus cf lodingi Escambia lineage, Chabarria et al. 2017
Holotype. AUM 40698 (Figure 13), a female collected 14 February 2014 at Camp Creek
where it crosses Conecuh National Forest Road 332 (31.1642786.53378; WGS84),
Covington County, AL, by David Laurencio.
Paratypes. AUM 37483 (a female), FLMNH 68839 (a male), and MMNS 1248 (a male).
Diagnosis. Membership of this species in the genus Necturus is demonstrated by reten-
tion of external gills in adults, presence of pigmented skin, presence of four well-
developed limbs, and reduction of digits on hind limbs to four toes. This species is unique
in possessing a larval stage lacking numerous small white spots and an adult stage that is
small in size, possesses dark dorsal and lateral spotting that is no larger than the size of the
eye, and lacks evidence of spotting on the chin or belly. Ten sequence autapomorphies
diagnose N. mounti, one of which is unambiguous (Chabarria 2008). The species is sister to
all other Necturus of the Gulf Coastal Plain (Figure 1).
Comparisons. Larvae of N. mounti are uniform pinkish grey in life, typically lacking any
small white spots, features that distinguish this species from the Mobile, Pearl,
Pontchartrain and Western lineages of N. beyeri (numerous small white spots);
JOURNAL OF NATURAL HISTORY 19
N. alabamensis (light dorsolateral stripes); N. lewisi (middorsal light stripe); and
N. maculosus (light dorsolateral stripes). Larvae of N. mounti are indistinguishable from
those of the Apalachicola lineage and N. punctatus. Adults of N. mounti typically possess
Figure 14. Distribution of localities of specimens assigned to N. mounti. Open circle is approximate
location of type specimen.
Figure 13. Photo of dorsum (a) and venter (b) of type specimen of N. mounti (AUM 40698; Covington
County, AL; photos by Craig Guyer).
20 C. GUYER ET AL.
small (no larger than size of eye) dark dorsal and lateral spots, features that distinguish this
lineage from the Pearl, Pontchartrain and Western lineages of N. beyeri and from N. lewisi
(large dark dorsal spots); adults of North Carolina populations of N. punctatus are similar in
adult dorsal colour pattern to N. mounti while South Carolina and Georgia populations
typically lack dark spots. Adults of N. alabamensis and N. maculosus frequently have large
dark spotting on the dorsal and lateral surfaces (rarely seen in N. mounti) but may retain
faded evidence of the dorsolateral light stripe of the larval stage (never seen in N. mounti).
The belly of N. mounti, in life, is white, changing abruptly to the dark lateral colouration,
and the chin also is immaculate. These features dier in the Apalachicola, Pearl,
Pontchartrain, and Western lineages of N. beyeri (ventrolateral dark spots; spotting on
mandibles of chin, occasionally to level of gular fold).
Description of holotype. In preservation, the type specimen is 143 mm TOT, 103.0 mm
SVL, and 40 mm TL (but tail cut for tissue sample); number of costal grooves is 16. The
dorsal ground colour in preservation is uniform dark grey brown and the dorsum has
distinct round dark spots that are approximately the size of the eye. This colour pattern
also characterises the dorsal surface of the tail. The head is uniformly dark slate brown,
lacking a dark stripe from the nostril through the eye. The dark dorsal colouration extends
onto the lateral surface of the venter; changing abruptly to immaculate white at mid-
venter; the edge of this transition zone is scalloped. The chin is immaculate white. The
holotype has the following values for measured morphological features ED: 1.3 mm; EL:
2.0 mm; GD: 2.3 mm; GEL: 15.7 mm; GL: 17.2 mm; HWE: 11.4 mm; HWG: 16.1 mm; POL:
15.9 mm; REW: 2.4 mm; SL: 4.6 mm; SRE: 6.4 mm; SW: 4.4 mm; WBW: 18.1 mm.
Variation. Necturus mounti achieves relatively small adult sizes, with a maximum male
size of 116 mm SVL (+49 mm TL) and a maximum female size of 110 mm SVL (+53 mm TL).
Mean male size is 108.4 mm SVL (n= 9) and mean female size is 94.8 mm SVL (n= 11).
Approximately 80% of specimens conform to the dorsal colour categories described
above. About 10% of specimens lack dark dorsal and lateral spots and about 10% have
large dark dorsal spots. Ventrally, about 5% of specimens have spots along the mandibles
with all others being spotless. Similarly, about 15% of specimens have ventrolateral spots
on the belly with the rest being spotless. Modal value for costal grooves is 17 (n= 17), with
47% of specimens having 16.
Adult males (n= 7) have the following mean (and range) values for measured mor-
phological features ED: 1.4 mm (0.12.9); EL: 1.9 mm (0.72.9); GD: 3.2 mm (2.23.9); GEL:
15.6 mm (12.618.4); GL: 18.2 mm (15.220.9); HWE: 11.9 mm (9.414.3); HWG: 18.4 mm
(13.921.5); POL: 16.9 mm (11.021.8); REW: 2.7 mm (2.33.0); SL: 6.5 mm (5.37.8); SRE:
8.0 mm (5.99.2); SW: 4.6 mm (2.45.8); WBW: 16.9 (12.818.9).
Adult females (n= 10) have the following mean (and range) values for measured
morphological features ED: 0.7 mm (0.022.4); EL: 1.9 mm (0.32.6); GD: 3.0 mm (2.33.9);
GEL: 14.8 mm (8.418.8); GL: 16.6 mm (12.121.2); HWE: 10.8 mm (9.213.3); HWG:
16.8 mm (14.719.3); POL: 14.4 mm (8.020.9); REW: 2.2 mm (1.52.6); SL: 5.2 mm (4.46.4);
SRE: 7.0 mm (5.68.9); SW: 4.7 mm (3.36.3); WBW: 15.9 mm (12.618.1).
Etymology. The specic epithet is a noun in the genitive case honouring Robert
H. Mount, the curator of Herpetology at Auburn University for many productive years
JOURNAL OF NATURAL HISTORY 21
and the person who pointed out the taxonomic challenge that was represented by the
waterdogs of Alabama. The recommended English common name is Escambia Waterdog.
Distribution and natural history. Necturus mounti occurs in the Blackwater, Escambia
(Conecuh in Alabama), Perdido, and Yellow River drainages of Alabama and the western
Panhandle of Florida (Figure 14). No published study has documented the life history of
N. mounti. But this taxon is common at the type locality where, for the past 30 years, we have
made haphazard samples because the site contains a large, permanent leaf pack from which
we have rarely failed to detect N. mounti in dipnet samples of those leaves. Juveniles are
detected year-round, with detections of adults being concentrated in January and February.
Siren intermedia, Desmognathus conanti, Ichthyomyzon gagei,dragony naiads, and small
craysh are frequent associates of N. mounti when they occupy these leaf packs. Samples
from the surface of the leaf pack rarely yield specimens. Instead, samples from greater than
two feet below the water surface, which are associated with cooler waters, yield these
salamanders. Heavy silt appears to reduce the chances of detection as does heavily decayed
leaves. We infer from these observations that adults migrate to leaf packs to mate. We
suspect that nesting takes place in the leaf packs. However, nesting might take place
elsewhere, with juveniles and adults migrating to leaf packs, as has been observed for the
Pearl lineage of N. beyeri (Shoop 1965;SeverandBart1996). Nevertheless, juveniles of
N. mounti appear to remain in leaf packs as they grow to adulthood.
Necturus moleri sp. nov.(Figures 1516)
Necturus punctatus lodingi Hecht, 1958,in part
Necturus beyeri alabamensis Hecht, 1958,in part
Necturus alabamensis Neill 1963,in part and misidentication
Necturus beyeri Neill 1963,in part
Necturus cf beyeri Bart et al. 1997,in part
Necturus (Parvurus)lodingi Dubois and Raaëlli, 2012,in part
Necturus cf beyeri Apalachicola lineage Chabarria et al. 2017
Holotype. AUM 35617 (Figure 15), a female collected 1 March 2002 in Big Creek near
Cottonwood (31.01916 85.35017; WGS84), Houston County, AL, by Michael Buntin.
Paratypes. FLMNH 164246 (female); FLMNH 177187 (male).
Diagnosis. Membership of this species in the genus Necturus is demonstrated by reten-
tion of external gills in adults, presence of pigmented skin, presence of four well-
developed limbs, and reduction of digits on hind limbs to four toes. This species is unique
in possessing a larval stage lacking numerous small white spots and an adult stage that is
small in size, possesses dark dorsal and lateral spotting that is no larger than the size of the
eye, and evidence of spotting on the mandible of the chin and lateral portion of the belly.
Five sequence autapomorphies distinguish this species, all of which are ambiguous
(Chabarria 2008). This species is sister to the Mobile lineage + N. alabamensis + Pearl
lineage + N. beyeri +N. maculosus (Figure 1).
22 C. GUYER ET AL.
Figure 15. Photo of dorsum (a) and venter (b) of type specimen of N. moleri (AUM 35617; Houston
County, AL; photos by Craig Guyer).
Figure 16. Distribution of localities of specimens assigned to N. moleri. Open circle is approximate
location of type specimen.
JOURNAL OF NATURAL HISTORY 23
Comparisons. Necturus moleri has a juvenile stage that is pinkish grey and lacks white
spotting, features that distinguish this species from the Mobile, Pearl, Pontchartrain and
Western lineages of N. beyeri (numerous small white spots); N. alabamensis (light dorsolat-
eral stripes); N. lewisi (middorsal light stripe); and N. maculosus (light dorsolateral stripes).
Larvae of N. moleri are indistinguishable from those of N. mounti and N. punctatus.Asadults,
N. moleri typically possesses small (no larger than size of eye) dark spots, but might lack
spots altogether, features that distinguish this species from the Pearl, Pontchartrain and
Western lineages of N. beyeri and from N. lewisi (large dark dorsal spots). Adults of North
Carolina populations of N. punctatus are similar in adult dorsal colour pattern to N. moleri
while South Carolina and Georgia populations typically lack dark spots. Adults of
N. alabamensis and N. maculosus frequently have large dark spotting on the dorsal and
lateral surfaces (rarely seen in N. moleri) but may retain faded evidence of the dorsolateral
light stripe of the larval stage (never seen in N. moleri). In life, the chin of N. moleri typically
has small dark spots on the mandibles and the belly is dull white with a gradual change
along the border between the immaculate white mid-venter and the light brown sides of
the body. These features distinguish this species from N. mounti (typically with bright white
chin and belly with sharp border between venter and sides of body).
Necturus moleri may possess a karyotype that includes 3 (Econna Creek, Jackson Co FL,
Choctawhatchee drainage; Guttman et al. 1990), 4 (Black Creek near Bruce, Walton Co FL,
Choctawhatchee drainage; Sessions and Wiley 1985), or 6 (Juniper Creek, Marion Co GA,
Chattahoochee drainage; Sessions and Wiley 1985) pairs of telocentric chromosomes.
These characteristics of the karyotype represent a transition from an ancestral state
lacking telocentric chromosomes (N. punctatus and N. lewisi) to a derived state of 6
pairs of telocentric chromosomes (N. maculosus and the Western lineage of N. beyeri;
Sessions and Wiley 1985). Similarly, the Y chromosome changes from an ancestral condi-
tion with reduced heterochromatism (N. lewisi and N. punctatus) to a derived condition
possessing extensive heterochromatism (N. maculosus and the Western lineage of
N. beyeri), with N. moleri displaying intermediate levels of heterochromatism (Sessions
and Wiley 1985). Additional karyotypic data are needed from N. mounti and the Mobile,
Pearl and Pontchartrain lineages to determine whether the above modications of the
presumed primitive Necturus karyotype are autapomorphies of N. moleri or are synapo-
morphies of all Gulf Coast Necturus.
Description of holotype. The type specimen measures 197 mm TOT, 134 mm SVL, and
63 mm TL; number of costal grooves is 16. The dorsum is a uniform dark greyish brown.
The top of the head has numerous small dark brown punctations. The sides of the body
are o-white, shading gradually to the dark dorsal colouration. The chin is light brown
with a few dark brown punctations, especially along the mandibles. The gular fold is
immaculate white, and the venter of the body is light brown with a few dark spots
anteriorly, becoming immaculate shiny white at mid-venter, and shading gradually to
a light smoky brown along the sides of the body. The holotype has the following values
for measured morphological features ED: 2.0 mm; EL: 0.9 mm; GD: 5.8 mm; GEL:
22.5 mm; GL: 24.9 mm; HWE: 17.2 mm; HWG: 28.9 mm; POL: 23.2 mm; REW: 2.2 mm; SL:
6.8 mm; SRE: 9.5 mm; SW: 6.6 mm; WBW: 27.5 mm.
24 C. GUYER ET AL.
Variation. Necturus moleri is of small size, with a largest male of 173 mm SVL (+73 mm
TL) and a largest female of 134 mm SVL (+63 mm TL). Adult males average 126.4 mm SVL
(n= 11) and females average 102.3 mm SVL (n= 15). Unlike the type specimen, most
individuals of N. moleri are light brown to tan, often with a pinkish cast, and possess
numerous small dark dorsal (75% of specimens) and lateral spots (71% of specimens).
However, the species might lack spots (11% of specimens) or have bold dark spots (4% of
specimens). In ventral colouration, most specimens (71%) have dark spots along the
mandibles of the chin and the ventrolateral aspect of the belly but might have spots
across the entire chin and belly (25% of specimens) or might be immaculate white (4% of
specimens). Thus, the colouration of N. moleri is more variable than N. punctatus or
N. mounti, the two species with the most similar size and colour pattern. Juveniles of
N. moleri are uniform in colour at the smallest sizes but develop dark dorsal and lateral
spotting at sizes above ca. 45 mm SVL. The modal value for costal grooves is 16 (n= 15),
with 33% of specimens possessing 17 grooves.
Adult males (n= 6) have the following mean (and range) values for measured mor-
phological features ED: 1.4 mm (0.32.9); EL: 1.8 mm (0.13.8); GD: 3.8 mm (2.06.4); GEL:
19.1 mm (14.224.4); GL: 24.7 mm (18.932.1); HWE: 14.5 mm (10.517.6); HWG: 22.2 mm
(17.529.7); POL: 18.7 mm (14.123.5); REW: 2.6 mm (1.93.8); SL: 7.2 mm (4.78.2); SRE:
9.3 mm (6.011.2); SW: 5.8 mm (3.68.5); WBW: 23.7 mm (17.035.5).
Adult females (n= 9) have the following mean (and range) values for measured
morphological features ED: 1.2 mm (0.12.9); EL: 1.4 mm (0.12.2); GD: 3.2 mm (2.25.8);
GEL: 14.4 mm (7.923.2); GL: 19.4 mm (10.824.9); HWE: 11.3 mm (7.117.2); HWG:
17.6 mm (11.428.9); POL: 14.0 mm (7.422.5); REW: 2.1 mm (1.62.7); SL: 5.3 mm (3.16.8);
SRE: 7.1 mm (3.79.5); SW: 4.6 mm (2.06.6); WBW: 16.7 mm (10.627.5).
Etymology. The specic epithet is a noun in the genitive case honouring Paul E. Moler for
his tireless devotion to eld herpetology of the southeastern United States and for
recognising the need to improve our understanding of species richness within Necturus
of the Gulf Coastal Plain. The recommended English common name is Apalachicola
Waterdog.
Distribution and natural history. This species is restricted to the Apalachicola, Chipola,
Choctawhatchee/Pea, Econna, and Ochlockonee drainages of Alabama, Florida, and
Georgia (Figure 16). No life history study has been performed on this taxon, but features
are likely to be similar to those of N. mounti.
Acknowledgements
We thank Paul E. Moler for insightful discussions about waterdog evolution and ecology.
Numerous friends, colleagues, and students participated in the many eld excursions that
provided eld observation associated with this study. Activities requiring use of live specimens
were covered under Auburn University IACUC protocols 2010-1827, and 2013-2386 and collec-
tions were authorized under AL scientic collecting permits 2010000058268680,
2013000068668680, 2014049802068680 to Craig Guyer. Data presented in this contribution
were generated largely from museum specimens. Chris Austin (LSUMZ), Kenny Krysko (FLMNH),
Bob Jones (MMNS), Max Nickerson (FLMNH), and Ken Tighe (USNM) are thanked for access to
materials under their care.
JOURNAL OF NATURAL HISTORY 25
Disclosure statement
No nancial interest or benet has arisen from the direct application of this research.
Funding
This work was supported by the Jenkins and Jordan Funds of the Auburn University Museum.
ORCID
Craig Guyer http://orcid.org/0000-0003-4909-2690
References
Ashton RE, Braswell AL. 1979. Nest and larvae of the Neuse River Waterdog, Necturus lewisi (Brimley)
(Amphibia: Proteidae). Brimleyana. 1:1522.
Ballard JWO, Whitlock MC. 2004. The incomplete natural history of mitochondria. Mol Ecol.
13:729744. doi:10.1046/j.1365-294X.2003.02063.x.
Bart HL, Bailey MA, Ashton RE Jr., Moler PE. 1997. Taxonomic and nomenclatural status of the upper
Black Warrior River Waterdog. J Herpetol. 31:192201. doi:10.2307/1565387.
Bart HL, Holzenthal RW. 1985. Feeding ecology of Necturus beyeri in Louisiana. Journal of
Herpetology. 19:402410.
Bonett RM, Trujano-Alvarez AL, Williams MJ, Timpe EK. 2013. Biogeography and body size shuing
of aquatic salamander communities on a shifting refuge. Proc R Soc B. 280:20130200.
doi:10.1098/rspb.2013.0200.
Braswell AL, Ashton RE. 1985. Distribution, ecology, and feeding habits of Necturus lewisi (Brimley).
Brimleyana. 10:1335.
Brenes R, Ford NB. 2006. Seasonality and movements of the Gulf Coast Waterdog (Necturus beyeri)in
eastern Texas. Southwestern Naturalist. 51:152156.
Chabarria RE. 2008. The historical biogeography and phylogenetics of Necturus beyeri (Proteidae:
Caudata) [M.S. thesis]. Hammond (LA): Southeastern Louisiana University.
Chabarria RE, Murray CM, Moler PE, Bart HL, Crother BI, Guyer C. 2017. Evolutionary insights into the
North American Necturus beyeri complex (Amphibia: Caudata) based on molecular genetic and
morphological analysis. J Zool Syst Evol Res. 2017:112.
Dubois A, Raaëlli J. 2012. A new ergotaxonomy of the order Urodela Duméril. 1805 (Amphibia,
Batrachia). Alytes. 28:77161.
Glorioso BM, Waddle JH, Richards-Zawacki CL. 2017. Prevalence of Batrachochytrium dendrobatidis
and B. salamandrivorans in the Gulf Coast Waterdog, Necturus beyeri, from southeast Louisiana,
USA. Herpetological Review. 48:114126.
Gunter G, Brode WE. 1964.Necturus in the state of Mississippi, with notes on adjacent areas.
Herpetologica. 20:114126.
Guttman SI, Weigt IA, Moler PE, Ashton RE Jr., Mansell BW, Peavy J. 1990. An electrophoretic analysis
of Necturus from the southeastern United States. J Herpetol. 24:163175. doi:10.2307/1564224.
Hecht MK. 1958. A synopsis of the mud puppies of eastern North America. Proc Staten Island Inst
Arts Sci. 21:138.
Highton R. 1957. Correlating costal grooves with trunk vertebrae in salamanders. Copeia.
1957:107109. doi:10.2307/1439396.
Highton R, Bonett RM, Jockusch EL. 2017. Caudata salamanders. In: Crother BI, editor. Scientic and
standard English common names of amphibians and reptiles of North America north of Mexico,
with comments regarding condence in our understanding. Shoreview (MN): Society for the
Study of Amphibians and Reptiles Herpetological Circular No. 43; p. 2537.
26 C. GUYER ET AL.
Lamb JY, Qualls CP. 2013.Necturus beyeri (Gulf Coast Waterdog). Detection by leaf litter bag.
Herpetol Rev. 44:491.
Means DB, Karlin AA. 1989. A new species of Desmognathus from the eastern Gulf Coastal Plain.
Herpetologica. 45:3746.
Mosimann JE, James FC. 1979. New statistical methods for allometry with application to Florida
red-winged blackbirds. Evolution. 33:444459. doi:10.1111/evo.1979.33.issue-1Part2.
Mount RH. 1975. The reptiles and amphibians of Alabama. Auburn (AL): Auburn University Alabama
Agricultural Experiment Station.
Neill WT. 1963. Notes on the Alabama waterdog, Necturus alabamensis Viosca. Herpetologica.
19:166174.
Nosil P. 2012. Ecological speciation. New York (NY): Oxford University Press.
R Core Team. 2016. R: a language and environment for statistical computing. Vienna (Austria):
R Foundation for Statistical Computing. https://www.R-project.org/.
Sabaj MH, 2016. Standard symbolic codes for institutional resource collections in herpetology and
ichthyology: an online reference Version 6.5 (16, August 2016). Washington (DC): American
Society of Ichthyologists and Herpetologists. Electronically accessible at http://www.asih.org/.
SAS Institute Inc. 2008. SAS/STAT® 9.2 users guide. Cary (NC): SAS Institute Inc.
Schmidt KP. 1953. A checklist of North American amphibians and reptiles. 6th ed. Chicago (IL):
University of Chicago Press.
Scott PA, Glenn TC, Rissler LJ. 2018. Resolving taxonomic turbulence and uncovering cryptic
diversity in the musk turtle (Sternotherus) using robust demographic modeling. Mol Phylogenet
Evol. 120:115. doi:10.1016/j.ympev.2017.11.008.
Sessions SK, Wiley JE. 1985. Chromosome evolution in salamanders of the genus Necturus.
Brimleyana. 10:3752.
Sever DM, Bart HL Jr. 1996. Ultrastructure of the spermathecae of Necturus beyeri (Amphibia:
Proteidae) in relation to its breeding season. Copeia. 1996:927937. doi:10.2307/1447655.
Shoop CR. 1965. Aspects of reproduction in Louisiana Necturus populations. Am Midland Nat.
74:357367. doi:10.2307/2423266.
U.S. Environmental Protection Agency. 2013. Level III ecoregions of the continental United States.
Corvallis (Oregon): U.S. EPA National Health and Environment Eects Research Laboratory. Map
scale 1:7,500,000. http://www.epa.gov/eco-research/level-iii-and-iv-ecoregions-continental-
united-state.
Viosca P Jr. 1937. A tentative revision of the genus Necturus with descriptions of three new species
from the southern Gulf drainage area. Copeia. 1937:120138. doi:10.2307/1436953.
Wray KP, Means DB, Steppan SJ. 2017. Revision of Eurycea quadridigitata (Holbrook 1842) complex of
dwarf salamanders (Caudata: Plethodontidae: Hemidactyliinae) with description of two new
species. Herpetol Monogr. 31:1846. doi:10.1655/HERPMONOGRAPHS-D-16-00011.
JOURNAL OF NATURAL HISTORY 27
... Gulf Coast Waterdogs occur in two disjunct areas separated by the Mississippi River, namely, east-central Texas into westcentral Louisiana and the Florida Parishes of Louisiana east through Mississippi to the Mobile drainage in Alabama (Guyer et al., 2020). Guyer et al. (2020) restricts each lineage to the following drainages: Mobile lineage, Mobile (AL) to Biloxi (MS); Pearl lineage, Wolf (MS) to Pearl (LA); Pontchartrain lineage, Bayou Bonfouca to Blind (LA); and Western lineage, Calcasieu (LA) to West Fork of the San Jacinto, TX. ...
... Gulf Coast Waterdogs occur in two disjunct areas separated by the Mississippi River, namely, east-central Texas into westcentral Louisiana and the Florida Parishes of Louisiana east through Mississippi to the Mobile drainage in Alabama (Guyer et al., 2020). Guyer et al. (2020) restricts each lineage to the following drainages: Mobile lineage, Mobile (AL) to Biloxi (MS); Pearl lineage, Wolf (MS) to Pearl (LA); Pontchartrain lineage, Bayou Bonfouca to Blind (LA); and Western lineage, Calcasieu (LA) to West Fork of the San Jacinto, TX. All ecological studies of Gulf Coast Waterdogs in Louisiana have occurred in Talisheek Creek, which is part of the Pearl lineage (Shoop, 1965;Shoop and Gunning, 1967;Bart and Holzenthal, 1985). ...
... This study took place along Bayou Lacombe, which is part of the Pontchartrain lineage of Gulf Coast Waterdogs described by Guyer et al. (2020). Here, we captured Gulf Coast Waterdogs by using unbaited minnow traps along a 13.4-km stretch of the bayou, beginning at its headwaters, to better understand factors that may influence their distribution and abundance. ...
Article
Full-text available
Few ecological studies have been conducted on Gulf Coast Waterdogs (Necturus beyeri), and published studies have focused on relatively small stream sections of 125 m to 1.75 km. In 2015, we sampled 25 sites along a 13.4-km stretch of Bayou Lacombe (Saint Tammany Parish, Louisiana, USA) to better understand factors that may influence the distribution of Gulf Coast Waterdogs within streams. We checked 250 unbaited traps once per week for 3 weeks, capturing 170 Gulf Coast Waterdogs at 18 of 25 sites. We used hierarchical models of abundance to estimate abundance at each site, as a function of site covariates including pH, turbidity, and distance from headwaters. The abundance of Gulf Coast Waterdogs within Bayou Lacombe was highest toward the center of the sampled stream segment, but we found no evidence that pH or turbidity affected abundance within our study area. Site-level abundance estimates of Gulf Coast Waterdogs ranged from 0 to 82, and we estimated there were 767 (95% Bayesian credible interval [CRI]: 266–983) Gulf Coast Waterdogs summed across all 25 sampling sites. We derived an estimate of 6,321 (95% CRI: 2,139–15,922) Gulf Coast Waterdogs for the entire 13.4-km stream section, which includes our 25 sites and the adjoining stream reaches between sites. Our results suggest that Gulf Coast Waterdogs may be uncommon or absent in the headwaters, possibly because of shallow water and swift currents with limited preferred habitats. Gulf Coast Waterdogs favor the middle stream reaches with adequate depth and abundant preferred microhabitats.
... However, the Coastal Plain rivals the Appalachians for overall lineage diversity. There have been numerous recent descriptions of new or cryptic species from several families in the Coastal Plain, including Ambystomatidae (Pauly et al., 2006), Plethodontidae Wray et al., 2017), Proteidae (Guyer et al., 2020), and Sirenidae (Graham et al., 2018). These descriptions and revisions total hundreds of millions of years of evolutionary history (see Jetz and Pyron, 2018). ...
Article
We describe a new, narrowly endemic species of swamp-dwelling dusky salamander (Plethodontidae: Desmognathus pascagoula sp. nov.) from the Gulf Coastal Plain of southeastern Mississippi and southwestern Alabama based on linear morphometrics, mitochondrial DNA, and single nucleotide polymorphisms from 881 loci produced using genotype-by-sequencing. Some populations of the new species were historically referred to as D. auriculatus, a polyphyletic assemblage of at least three species in the Atlantic and Gulf Coastal Plain from Texas to North Carolina. Populations of D. auriculatus from the Gulf Coastal Plain in Louisiana and Mississippi were recently described as D. valentinei. The new species includes populations that were tentatively referred to D. valentinei, but we find it is morphologically, genetically, and geographically distinct. It is smaller, has a more defined dorsal color pattern, more irregular whitish “portholes” in up to three rows on the lateral surfaces of the body and tail, and a brighter orange or yellowish orange postocular stripe. At present, the new species is known from only six extant populations in the lower Pascagoula, Escatawpa, and Mobile drainages. The latter represents a distinct phylogeographic lineage. We also refer a historical collection from the northeastern side of the Mobile-Tensaw River Delta to this species, suggesting a much broader range in the past. We suspect that more populations remain to be discovered in the area, and their potential species-level distinctiveness should be tested further. This discovery increases knowledge of the biodiversity in the southeastern United States Coastal Plain, a candidate region meeting the global criteria for a “biodiversity hotspot,” and underscores the amount of cryptic diversity likely remaining to be discovered and described in Nearctic salamanders.
... In this special issue, eleven new species are described (nine frogs and two salamanders) and several revisionary actions are taken. There are contributions dealing with taxa from Africa (including Tanzania and Madagascar, Greenwood et al. 2020;Rancilhac et al. 2020), Asia (Mahony et al. 2020;Nguyen et al. 2020a;Nguyen et al. 2020b;Stuart et al. 2020), North America (Guyer et al. 2020), and South America (Melo-Sampaio et al. 2020). Interestingly, more than half of these studies describe new species from countries with recent high activity in alpha taxonomy research: Brazil, India, Madagascar, and Vietnam (Table 3). ...
Article
Full-text available
The Eurycea quadridigitata complex is currently composed of the nominate species and E. chamberlaini, with no other species recognized. However, recent molecular studies have revealed at least five genetic lineages within this species complex, with one lineage more closely related to the neotenic Eurycea species of central Texas and E. chamberlaini nested within E. quadridigitata sensu lato. We use large scale geographic sampling in combination with a multilocus species delineation method and morphology to test whether these genetic lineages represent distinct species under the general lineage concept of species. We describe two new species of salamander from this complex, resurrect and elevate a former subspecies to full species status, add to the diagnosis of E. chamberlaini, and redefine E. quadridigitata in the context of this revision. All five species are diagnosable from one another through a number of meristic, morphometric, molecular, and ecological criteria.
Article
Necturus beyeri (Caudata: Proteidae), as conceived by some, contains paedomorphic salamanders distributed from the Ochlockonee drainage of Florida to the Angelina drainage of Texas. Because these salamanders differ in color pattern and karyotype across their geographic range, we performed a phylogeographic analysis that included representatives from all major drainages as well as of all congeners. The mitochondrially encoded ND2 gene was used to infer phylogenetic relationships using Bayesian inference. Morphometrics of head shape were analyzed and included as an independent data set. Our work suggests that Necturus comprises 11 lineages. A basal split within the genus separates an ancestor of two Atlantic Coastal Plain species (Necturus lewisi and Necturus punctatus) from the ancestor of nine distinct Gulf Coastal Plain lineages. One lineage is consistent with Necturus alabamensis, a species currently recognized in the Black Warrior drainage of Alabama. Two lineages comprise Necturus maculosus, as historically recognized, and six lineages comprise N. beyeri, as recognized by some, each of which occupies a unique drainage. Both of these species are demonstrated to be paraphyletic. Head morphometrics show the same patterns as the mtDNA. Overall, lineages within Necturus exhibit an east-to-west progression of appearance on the phylogenetic trees. This pattern corroborates biogeographic hypotheses based on previous karyological work. Within N. beyeri, this progression separates a pattern class of two eastern lineages lacking bold spotting and possessing relatively small mean body lengths from a pattern class of four western lineages possessing bold spotting and larger mean body sizes. Thus, the two eastern lineages of N. beyeri are similar in color pattern and body size to N. punctatus either through retention of the ancestral color pattern and size for the genus or through convergent selection in eastern streams of the Gulf Coastal Plain.
Article
Accurate and consistent delimitation of species and their relationships provides a necessary framework for comparative studies, understanding evolutionary relationships, and informing conservation management. Despite the ever-increasing availability of genomic data, evolutionary dynamics can still render some relationships exceedingly difficult to resolve, including underlying speciation events that are rapid, recent, or confounded by post-speciation introgression. Here we present an empirical study of musk turtles (Sternotherus) which illustrates approaches to resolve difficult nodes in the Tree of Life that robust species-tree methods fail to resolve. We sequence 4430 RAD-loci from 205 individuals. Independent coalescent-based analyses, corroborated with morphology and geography, strongly support the recognition of cryptic species within Sternotherus, but with conflicting or weak support for some intraspecific relationships. To resolve species-tree conflict, we use a likelihood-based approach to test support for alternative demographic models behind alternative speciation scenarios and argue that demographic model testing has an important role for resolving systematic relationships in recent, rapid radiations. Species-tree and demographic modeling strongly support the elevation of two nominal subspecies in Sternotherus to species and the recognition of a previously cryptic species (S. intermedius sp. nov.) described within. The evolutionary and taxonomic history of Sternotherus is discussed in the context of these new species and novel and well-supported systematic hypotheses.
Article
Describes Desmognathus apalachicolae n. sp. (Plethodontidae). Its populations are distributed among deep, moist ravines south of the Fall Line in the Ochlockonee, Apalachicola-Flint-Chattahoochee, and upper Choctawhatchee river systems in Alabama, Georgia, and Florida. -from Authors
Article
The origin of biological diversity, via the formation of new species, can be inextricably linked to adaptation to the ecological environment. Specifically, ecological processes are central to the formation of new species when barriers to gene flow (reproductive isolation) evolve between populations as a result of ecologically based divergent natural selection. This process of 'ecological speciation' has seen a large body of focused research in the last ten-fifteen years, and a review and synthesis of the theoretical and empirical literature is now timely. The book begins by clarifying what ecological speciation is, its alternatives, and the predictions that can be used to test for it. It then reviews the three components of ecological speciation and discusses the geography and genomic basis of the process. A final chapter highlights future research directions, describing the approaches and experiments which might be used to conduct that future work. The ecological and genetic literature is integrated throughout the text with the goal of shedding new insight into the speciation process, particularly when the empirical data is then further integrated with theory.
Article
The feeding ecology of the Gulf Coast Waterdog, Necturus beyeri, is described based on (1) analysis of gut contents from 91 specimens collected between February and December 1980 in Talisheek Creek, Louisiana, and (2) information on the habits, hatitats, seasonal availability and size ranges of N. beyeri and its prey. Waterdogs and their prey were closely associated with leaf litter (detritus) deposits in the stream. Necturus beyeri were collected almost exclusively from leaf litter beds. Sixteen of the 18 prey taxa recorded inhabit sediments and detritus in depositional areas and 12 of these function as detritivores. Declining abundance of N. beyeri in leaf litter beds and decreased food consumption during warmer months was associated with declining invertebrate richness and abundance. It is hypothesized that N. beyeri burrow during warm periods of the year and that this behavior is related to low food availability. Young and adult N. beyeri were trophically segregated by prey type and size, and by feeding habitat to a degree. Juveniles and subadults fed predominantly on isopods, midges and small mayflies. Adults fed predominantly on large mayflies and caddisflies. The high intensity of acanthocephalan parasitism observed for N. beyeri was linked to predation on isopods, the intermediate hosts of these parasites. The potential impact of acanthocephalan parasitism and isopod consumption on N. beyeri ecology is discussed.
Article
Aspects of the reproductive biology of Necturus maculosus louisianensis and N. b. beyeri were studied in Big Creek and Talisheek Creek, Louisiana, respectively. Collecting techniques included electrical shocking, seining, dip netting, and baited set lines. There was unusual success of female N. m. louisianensis captures on set lines. Morphological changes related to reproduction in the Louisiana forms are very similar to those of N. m. maculosus. The number of ovarian eggs in N. m. louisianensis (overline x = 54.2) and N. b. beyeri (overline x = 56.7) is less than recorded for N. m. maculosus. Oviposition occurs in May in Louisiana at water temperatures of 18-20 C. Guardian adults were not observed in nests, but water turbidity may have prevented their detection. Nest structure and eggs with capsules are like those of N. m. maculosus. Pigment patterns of embryos and hatchlings of N. m. louisianensis and N. b. beyeri differ.
Article
In 1937 Percy Viosca described Necturus alabamensis as a flattened form from the southwestern extremity of the Cumberland [=Appalachian] Plateau. Necturus beyeri was described as a more cylindrical-bodied form occurring over a wide area of the lower Gulf Coastal Plain, but sympatric with N. alabamensis in the Black Warrior River near Tuscaloosa, Alabama. Subsequent workers considered N. alabamensis to be allied to or synonymous with N. beyeri, ultimately applying the name alabamensis to populations throughout the Mobile Bay drainage, and to lowland populations from southern Mississippi and eastern Louisiana to southern Georgia and panhandle Florida. We present pigmentation, morphometric, and distributional evidence that: (a) the distinctive waterdog in the upper Black Warrior River drainage is in fact the form Viosca described as Necturus alabamensis; therefore, the epithet alabamensis applies to that form; and (b) N. alabamensis and the lowland form Viosca referred to as N. beyeri are in fact syntopic on the Appalachian Plateau just above the Fall Line near Tuscaloosa. The juvenile pigmentation pattern of the upper Black Warrior waterdog (unknown to Viosca) is described and related to the distinctive adult pigmentation pattern. Past taxonomic and nomenclatural treatments of eastern Gulf Slope Necturus are summarized, and recommendations are offered for revising the taxonomy and nomenclature of these populations/species based on available evidence.