Content uploaded by Will Selman
Author content
All content in this area was uploaded by Will Selman on May 06, 2025
Content may be subject to copyright.
117
Herpetological Conservation and Biology 20(1):117–130.
Submitted: 9 October 2023; Accepted: 30 January 2025; Published: 30 April 2025.
Copyright © 2025. Bailey Watkins
All Rights Reserved.
(
121,3
1Department of Biology, Millsaps College, 1701 North State Street, Jackson, Mississippi 39210, USA
2Mississippi Department of Wildlife, Fisheries, and Parks, Mississippi Museum of Natural Science,
2148 Riverside Drive, Jackson, Mississippi 39202, USA
3Corresponding author, e-mail: will.selman@millsaps.edu
Abstract.
Key Words.—amphibians; egg brooding; Hartselle Sandstone; imperiled species; Plethodontidae; rock outcrop; Species of
Greatest Conservation Need; visual encounter surveys
The southeastern U.S. is a biodiversity hotspot for
terrestrial and aquatic salamanders (Milanovich et
al. 2010) because of the many ecological niches and
long geologic history of the Appalachian Mountains
and Cumberland Plateau (Tilley 1980). This region
extends southwestward into northern Alabama and
extreme northeastern Mississippi (Tishomingo
County), providing the latter with habitats unavailable
elsewhere in the state. Consequently, those habitats
support many salamander species that are typically
associated with states to the northeast, such as Cave
SalamandersEurycea lucifuga), Spring Salamanders
(Gyrinophilus porphyriticus), and Eastern Hellbenders
(Cryptobranchus alleganiensis). One species whose
distribution is restricted to the northeastern corner
of Mississippi is the Green Salamander (Aneides
aeneus), which occurs patchily throughout much of
the Appalachian Plateau and the Ridge and Valley
provinces (Petranka 1998). The species is most
frequently observed in deep crevices of rock outcrops
that are surrounded by mature forest (Smith et al.
2017). They are highly adapted and specialized to
live and breed in and around rock outcrops (Walker
and Goodpaster 1941), and recent studies have shown
their seasonal dependence upon arboreal habitats in
the spring (Waldron and Humphries 2005). In all
13 of the states where A. aeneus occurs, it is ranked
as Vulnerable (S3), Imperiled (S2), or Critically
Imperiled (S1; https://explorer.natureserve.org/).
Aneides aeneus is also currently under review for
listing under the Endangered Species Act (U.S. Fish
and Wildlife Service 2015) and are listed as Near
Threatened in the Red List of Threatened Species
of the International Union for the Conservation of
Nature (Soto 2021).
Within Mississippi,A. aeneus has been reported
from nine distinct sites, and all but two of these are
located within Tishomingo State Park (TSP; Woods
1968; William Cliburn, unpubl. report). Furthermore,
all records are associated with Hartselle Sandstone
118
Watkins et al.—Status of Green Salamanders in Mississippi, USA.
rock formations, a geologic layer that exhibits large
rock wall formations (maximum height: about 12 m;
Merrill et al. 1988; Fig. 1). These rock formations
were deposited during the Carboniferous Period,
approximately 360–299 mya (Thompson 2011), and
they are more similar to the geology of the southern
Appalachians of Alabama, Tennessee, and Georgia
than to the geology elsewhere in Mississippi. Even
though Mississippi is on the periphery of their range,
the species has been consistently observed in TSP
since the 1960s with varying encounter rates reported
(Rauch et al. 2016). In Mississippi, however, they
are considered a Species of Greatest Conservation
Need (S1 species - Critically Imperiled; Mississippi
Museum of Natural Science 2015), are known from a
very small region in the state (about 5.3 km2), and are
very susceptible to extirpation (Mississippi Museum
of Natural Science 2015). Furthermore, there is no
recent information published about their population
status therein. Therefore, the primary objectives
of our study were to determine the distribution and
encounter rates (i.e., catch per unit eort, CPUE) of
A. aeneus in northeastern Mississippi. Secondary
objectives of our study were to describe A. aeneus
habitats in Mississippi, determine if A. aeneus is
extant at historical sites, report on reproductive
activity, and compare encounter rates across dierent
regions, property ownership types, and survey times.
In 2017 (March-November), 2018 (April-
October), and 2019 (April-December), we completed
93 surveys at 25 rock outcrop sites located on both
private and public lands to document the presence of
A. aeneus (Fig. 2; Appendix Table). During July and
August 2021–2023, we conducted 87 surveys at 32
rock outcrop sites also on private and public lands
(Fig. 2; Appendix Table). The distinction between
these two property types was made due to the dierent
land management activities on these properties that
may impact A. aeneus populations. For example,
timber harvesting occurs regularly on private lands,
and the loss of trees may impact arboreal habitat and/
or change rock outcrop microclimate for A. aeneus
(Pauley and Watson 2005; Newman et al. 2022;
Smith 2024).
Rock outcrop habitat varied, ranging from
continuous or semi-continuous rock walls (up to
15–20 m) at or near the point of original erosional
exposure (i.e., typically near the top of steep slopes)
to scattered boulders downslope from the original
erosional surface; the latter are often tilted dierently
from the original bedding plane. Most of the rock
outcrop locations were heavily forested with moderate
to steep topographic relief. We selected sites based on
the knowledge that A. aeneus has only been found in
Mississippi at Hartselle Sandstone outcrops, and we
consulted a detailed local geological map (Merrill et
al. 1988) that highlights the small area in Tishomingo
County that has this unique geological formation.
Along with the geological map, these outcrop sites
were also known from previous surveys (e.g., Woods
1968; Rauch et al. 2016), identied via Google Earth
(i.e., where exposed rock was visible during winter
aerial images), or located while walking/searching
in wooded areas. Because this is a relatively small
region in the state (< 30 km2), we believe that
most of the major Hartselle Sandstone outcrops in
Mississippi were identied using these methods.
Indeed, Aldridge et al. (2024) completed LIDAR
surveys for sheer rock faces in Tishomingo State
Park independent of this study, and all the locations
we surveyed were nearly identical to the rock outcrop
habitats that were remotely sensed. Nonetheless, we
acknowledge that future surveys could nd additional
small outcrops in Mississippi that might be suitable
habitat for A. aeneus.
For all surveys from 2017–2023, we used area-
constrained survey techniques to survey at each
rock outcrop. At each outcrop, we used hand-held
. Representative photograph of a Hartselle Sandstone rock
outcrop in Tishomingo State Park, Mississippi, USA. This location
is a previously unsearched rock outcrop, Z2, where we documented
ve new locations for Green Salamanders (Aneides aeneus). For
perspective, the photograph is taken facing northwest, with the
primary rock wall on the right (east), a secondary wall is on the left
(west), and a third wall is out of the photograph on the other side of the
boulder to the left. (Photographed by Will Selman)
119
Herpetological Conservation and Biology
ashlights and headlamps to search crevices and
openings in the rock formations during daytime and/
or nighttime hours. We searched all rock formations
completely between 0 and about 2.0 m height above
the ground, and this includes all rock surfaces and
crevices. Most rock outcrops could be surveyed in
their entirety during a single visit, although some
more extensive outcrops (e.g., Site Q) required
multiple visits to be surveyed completely. For every
individual salamander we found, we identied it to
species, and we counted all salamanders and other
lithophilic reptiles and anurans. For each A. aeneus,
we also recorded the age-class (juvenile/adult) and
the location with a handheld GPS; specic GPS
locations are withheld from this manuscript, but they
were deposited with the Mississippi Natural Heritage
Program at the Mississippi Museum of Natural
Science. For the 2021–2023 surveys, we recorded
the salamander height above ground, type of crevice
(e.g., horizontal, vertical, honeycomb, hole; Fig. 3),
whether the salamander was tending eggs (hereafter,
brooding females), and the person search hours
expended at each rock outcrop as a measure of eort.
We collected the latter information to calculate the
encounter rate (CPUE): the number of salamanders
observed at a rock outcrop divided by the total
person search hours expended during that rock
outcrop survey. We included the time to collect data
and tissue samples for a related study in the person
search hours. Waldron and Humphries (2005) found
signicant use of trees by Aneides, but we conducted
no formal surveys of trees in the proximity of outcrops
for Aneides. Thus, we acknowledge that our CPUE
calculation does not reect possible dierences
in local Aneides populations based on dierences
in use of trees among sites or the dierences in
forest composition (i.e., dierent tree species, stand
condition, and proximity to outcrops). Also, there
were many dierences across rock outcrops (e.g.,
rock crevice frequency, width, depth, and height
above the ground among sites), and these dierences
may have aected the detectability of salamanders.
To determine if encounter rates of A. aeneus
varied across the limited distribution of the species in
Mississippi, we divided the rock outcrops into three
regions, with sites west of Bear Creek in what we
called the West Region, sites east of Bear Creek and
west of Tishomingo County Road 993 in the Central
Region, and sites east of County Road 993 in the East
Region (Fig. 2). We designated these regions because
Bear Creek, a signicant geographical barrier,
separates the western and central sites, and a long
. Locations of Hartselle Sandstone rock outcrops and Green Salamander (Aneides aeneus) locations from 2017–2023 in northeastern
Mississippi, USA. Dark blue lines indicate the approximate location of the surveyed rock outcrops (letters match those in the Appendix Table),
while green points indicate locations where individual A. aeneus were documented. Also depicted on the map include Tishomingo State Park
(black boundary), the Natchez Trace Parkway (green boundary), Bear and Cedar creeks (light blue), and Tishomingo County Road 993 (red
line). The green dot on the inset map of Mississippi represents the location of the larger map.
120
Watkins et al.—Status of Green Salamanders in Mississippi, USA.
geographical distance of unsuitable habitat separates
the central and eastern sites. We then performed two
Kruskal-Wallis tests to determine if CPUE was the
same by region using all site CPUE data and rock
outcrop sites with only non-zero data. Similarly,
we used Wilcoxon Ranked Sums (two groups) or
Kruskal-Wallis (three groups) tests for independently
a priori hypotheses to determine if CPUE was the
same in day versus night surveys on private and
public lands, and by year (2021/2022/2023). For all
statistical analyses, we used the software JMP 12.2.0
(SAS Institute, Cary, North Carolina, USA), with
signicance of tests at α = 0.05.
.—We completed 182 surveys at 32
rock outcrop sites, and we observed A. aeneus at
25 of 32 sites (78.1%; Fig. 2). We conrmed their
presence at eight of the nine historical sites, and we
also documented A. aeneus from 17 new rock outcrop
sites from which the species had not been previously
documented. At most outcrop sites where the species
was observed, multiple discrete sightings were often
made along the length of the rock outcrop site. Of the
seven rock outcrop sites where we did not detect A.
aeneus, ve were on public land (B, B’, D’, E’, T), and
two were on private land (M, Y). Only one of these
sites had a historic record (M), and we surveyed this
site four times without detecting A. aeneus. Previous
records encompass a minimum convex polygon
area of approximately 5.3 km2, while the additional
records from this study increase the occupied area to
about 7.9 km2 (50.0% area increase; Fig. 4).
.—During July/August 2021–2023
(excluding 2017-2020 surveys), we completed 87
surveys of all 32 rock outcrop sites, and we searched
for 492.1 person hours. We documented 1,102
salamanders associated with rock faces and crevices
including 586 Slimy Salamanders (Plethodon
glutinosus) complex (53.0% relative abundance
[RA], 1.19/h), 254 A. aeneus (23.0%, 0.52/h), 205
Long-tailed Salamanders (Eurycea longicauda;
18.7%, 0.42/h), 42 Southern Two-lined Salamanders
(Eurycea cirrigera; 3.8%, 0.09/h), eight unknown
A comparison of Hartselle Sandstone microhabitat types where Green Salamanders (Aneides aeneus) were detected in Tishomingo
County, Mississippi, USA. A rock outcrop with (A) a long horizontal crevice and (B) with exposed rock face above. (C) Honeycomb habitat
in Hartselle Sandstone exhibited a series of scooped-out sections in the rock, and these sections were sometimes extensive. (D) Holes were
classied as singular, deep and narrow recesses that extended perpendicular into the rock face and (E) a vertical crevice extends above the
hole. (F) Overhang habitat was underrepresented along Hartselle Sandstone outcrops, but A. aeneus were overrepresented in these habitats.
(Photographed by Will Selman).
121
Herpetological Conservation and Biology
outcrops (Mean CPUE = 0.22/h ± 0.43; Z = ˗2.03, df
= 1, P = 0.042; Fig. 5).
.—We found gravid females in
2021 from 19 July until 28 July and in 2022 from 14
July until 9 August. Females initiated egg brooding
on 26 July 2021 and 3 August 2022. In 2023, we
began our surveys later, and we observed a brooding
female on the rst day of surveys on 24 July 2023.
Observations of brooding females extended until the
end of the study period in all years.
Of the 254 A. aneus individuals we found from
2021–2023, 39 (15.4%) were brooding females. We
observed brooding females at 13 sites (B, C, F, I, J, K,
O-, P, Q, S, W, X, and Z2), and the highest percentage
of brooding females was observed at site O- (four
of 11 individuals). Along with brooding location
variability, we noticed a dierence in the number of
brooding females observed by year. The percentage
of brooding females were fourfold higher in 2021
(21%; 16 of 77) and sixfold higher in 2023 (32%; 17
of 53) compared to 2022 (5%; six of 122). Along with
gravid and brooding females, we observed evidence
of reproduction in the form of small juveniles (< 2 cm
SVL) during our surveys. In 2021, we observed four
juveniles (5.1% of all individuals observed that year)
salamanders (< 1%), and seven Cave Salamanders
(Eurycea lucifuga; < 1%, 0.01/h; Appendix Table).
Mean A. aeneus CPUE of all surveys was 0.52/h ±
(standard deviation) 0.69 (range of values 0–4.0).
There was some variation by region (Central: mean
= 0.65/h ± 0.90, range of values 0–4.0; West: mean
= 0.42/h ± 0.61, range of values 0–2.4; East: mean
= 0.28/h ± 0.49, range of values 0–1.6), but there
was no signicant dierence by region (H = 2.91,
df = 2, P = 0.234; Fig. 5). A similar analysis with
all non-zero CPUE sites also indicated no regional
dierences in CPUE (H = 1.14, df = 2, P = 0.565).
Northeastern Mississippi experienced a wet year
due to persistent rainfall in 2021 (2021 CPUE Mean
= 0.46/h ± 0.82), while drought existed during the
spring and summer of 2022 and 2023 (2022 CPUE
mean = 0.49/h ± 0.63; 2023 CPUE mean = 0.40/h ±
0.44); however, there was no signicant dierences
in CPUE by year (H = 0.99, df = 2, P = 0.610; Fig.
5). Similarly, there was no signicant dierences
in the CPUE in surveys conducted during the day
(Mean CPUE = 0.51/h ± [0.65) compared to night
(Mean CPUE = 0.43/h ± 0.73; Z = 0.21, df = 1, P =
0.649; Fig. 5). CPUE on public land rock outcrops,
however, was signicantly higher (Mean CPUE
= 0.54/h ± 0.74) than CPUE on private land rock
A comparison of the known range of Green Salamanders (Aneides aeneus) prior to this study (blue polygon) and known range after
this study (yellow polygon) in Tishomingo County, Mississippi, USA. The red lines are approximations of rock outcrops.
122
Watkins et al.—Status of Green Salamanders in Mississippi, USA.
at two sites (K–1, L–3) and zero juveniles in 2023
(0%), but in 2022, we observed 40 juveniles (33%)
at eight sites (F–13, C–6, S–3, V–1, Q–7, N–1, X–8,
W–1).
Habitat.—Based on GPS points collected from
rock outcrop sites, we estimate that we searched about
15.7 linear km of rock outcrops and rock faces that
are potential habitat for A. aeneus. Of the 15.7 km,
about 9.9 km (63.1%) of rock outcrop habitat occurs
within protected lands of TSP and the Natchez Trace
Parkway (NTP). The remaining 36.9% is privately
owned, and parcels of three private landowners
contain the majority of rock outcrop habitat (88.1%
of private property rock outcrops).
Of the 254A. aeneus individuals we observed,
142 (55.9%) were in horizontal crevices, 50 (19.7%)
were in holes, 39 (15.4%) were on exposed rock
faces/ledges, 14 (5.5%) were in vertical crevices, and
9 (3.5%) were in honeycomb formations (i.e., large
series of scooped-out areas of the rock). The mean
height above ground level for individuals was 1.3 m
(standard deviation = 0.49, range of values 0.22–2.74),
but few observations could be completed above 2 m
due to inability to access higher crevices. Another
interesting observation was the strong tendency of
A. aeneus individuals to be found under large rock
overhangs (Fig. 3), with 10% of observations in these
locations even though their habitat availability was
about 5%.
.—We found A. aeneus at most rock
outcrop sites surveyed in northeastern Mississippi
and in every historical site except for one. We
found individuals at 17 previously unsurveyed rock
outcrops, and many of the rock outcrop sites had
multiple discrete locations that we documented
within each site. Our survey results increase the
potential area of occupation in the state by > 50%.
To the best of our knowledge, many of the new sites
we found were due to the extensive searching of
previously unsearched or unknown rock outcrops
within and outside TSP. Some of the areas were
remote and/or on private lands, and previous surveys
mostly focused on easily accessible locations on TSP.
The percentage of sites where we found A. aeneus
in Mississippi (78.1%) was higher than what has
recently been reported in other states. In comparison,
the highest percentage of occupied sites previously
reported was in North Carolina (77.2%, 44 of 57 sites;
Williams et al. 2020). In Tennessee, Niemiller et al.
(2022) found A. aeneus at 47 of the 84 survey sites
(55.9%), while Newman et al. (2018) found 49.2%
of sites occupied in northwestern South Carolina.
John (2017) found only 15.5% of sites occupied at
Redstone Arsenal in northern Alabama. Because
most species typically exhibit rarity or scattered
populations on the edge of their range compared to
the core of their range (Channell 2004; Steen and
Barrett 2015), one would not expect higher rates of
sites occupied at peripheral sites (i.e., Mississippi)
compared to core areas of the range (i.e., Tennessee,
North Carolina, Alabama). Thus, comparative studies
in neighboring regions of northwestern Alabama
may provide additional insights to determine if this
is a local pattern we observed or if it extends more
broadly into neighboring areas.
We were unsuccessful in locating any individuals
at site M, a private property site previously reported
by Rauch et al. (2016). We are unsure why no
individuals were detected during our four surveys, but
it could possibly be due to historical events at the site
including rock mining that occurred at the site until
approximately 2015. Mining also occurred at a site
on TSP (Site A) prior to its establishment as a state
park in 1935, however, so mining may not entirely
explain their absence at site M. Indeed, Hinkle et
al. (2018) found that A. aeneus was observed in
70% of sites with evidence of extensive mining. So,
while mining may have had an adverse eect on the
Comparison of Green Salamander (Aneides aeneus) catch
per unit eort (CPUE) by (from left to right) region, year, time of
survey, and land ownership. Horizontal lines are the medians, the
boxes are the interquartile range (IQR; 25th -75th percentile), the
vertical lines represent the last point that is within 1.5×IQR, and
the dots are outlier points. The abbreviation NS = non-signicant
dierence and the asterisk (*) = signicant dierence.
123
Herpetological Conservation and Biology
population, it is not likely to be the only contributing
factor to our non-detection. Future eorts should
continue to survey sites like site M where we did not
nd A. aeneus. Further, it is possible that additional
rock outcrop habitat may occur in Mississippi, yet it
will likely be smaller and disjunct from the primary
areas we surveyed.
.—Our overall CPUE estimate
(0.52/search hour) was similar to densities found in
Tennessee (0.52/h; Niemiller et al. 2022). Unlike
Tennessee, however, our surveys were completed
entirely during the summer, and previous studies
have found that A. aeneus fall encounter rates in
Tennessee and North Carolina (Tennessee: 1.5–3.0/
h; North Carolina: 0.84–1.06/h) were greater than
summer encounter rates (Tennessee: 0.2–0.4/h;
North Carolina: 0.29–0.43/h; Niemiller et al. 2022,
Williams et al. 2020). Thus, using a similar seasonal
comparison, the summer encounter rates we observed
in Mississippi were higher than reported in either North
Carolina or Tennessee studies, both states considered
to be within the core range of A. aeneus. Previous
studies have found that detection and encounter rates
of A. aeneus can vary widely depending on numerous
environmental, seasonal, observer experience, and
physiological factors (Waldron and Humphries
2005; John 2017; Newman et al. 2018, Niemiller
et al. 2022), and summer, when our surveys were
completed, has not typically been considered a
season where high encounter rates would be expected
(Williams et al. 2020, Niemiller et al. 2022). Because
we did not survey in other seasons, it is not possible
to know how dierent seasons or environmental
conditions might have factored into our encounter
rates. Previous summer rock outcrop surveys at these
sites, however, yielded 10-fold dierences in CPUE
across years; for example, surveys in July 2017 at site
F had a CPUE of 13.3/h, while two years later in July
2019 the CPUE was 1.3/h (unpubl. data). The latter
CPUE is similar to the CPUE found in this study
for the same site. Therefore, we suspect some of
these CPUE dierences may also be associated with
dynamic yearly shifts in recruitment and survival
(i.e., Boom and Bust).
While the overall encounter rate of A. aeneus
in Mississippi is higher than other states, some
Mississippi sites appear to have experienced long-
term declines. For example, it seems that a marked
decline has occurred at site I (Cabin Clis) since the
1970s, especially considering that Woods (1968)
reported A. aeneus was extremely abundant at this
site. Further, Rauch et al. (2016) documented 13
individuals in ve surveys in 2008 and 2009, while
only two and one individuals were observed there
in 2017 and 2019, respectively (unpubl. data). In
2021–2023 CPUE surveys, we only detected three
individuals at this location, so it is unknown why
A. aeneus at this location have declined long-term.
First, other studies also found that similar declines
occurred for A. aeneus populations between the
1970s and 1980s, and climate change, habitat loss,
and epidemic disease were considered the most
likely scenarios to explain the declines (Snyder
1991; Corser 2001). These explanations do not
seem plausible for our study, however, as declines
have only been observed at a single site and not the
entirety of the Mississippi distribution. Second, it
could be due in part to scientic collection pressures
in the past (Corser 2001). Based on specimen
records from the Mississippi Museum of Natural
Science (Emily Field, pers. comm.) and in other
museums (VertNet.org), at least 63 specimens were
collected at Tishomingo State Park over a 12-y period
(between 1956–1967). It is unknown from what rock
outcrop(s) these individuals were collected, but a
portion were likely collected from Cabin Clis given
its prominence as a study site by Woods (1968). If
they were collected from one or a handful of sites, that
could lead to diculty in recovery/recolonization of
the site. Third, declines could be due to impacts from
nearby cabins and recreational practices. Rauch et al.
(2016) indicated that cabin renters dumped re ashes
and refuse over the clis, potentially impacting this
population. Lastly, we noticed that many crevices
were completely covered at this site with mosses in
2021–2023, obscuring some of the crevice habitat.
The advance of moss could be due to succession
via a lack of disturbances that have not happened at
this rock outcrop in a long time (i.e., 20+ y without
growing season res occurring on a longer rotation;
canopy openings from occasional blowdowns, and
desiccation of rock faces). Gordon (1952) indicated
that moss invasion into brooding crevices resulted in
A. aeneus females abandoning it over time, so this
may be an explanation into local declines at this site.
It is possible that two or more of these factors could
also be interacting at a local scale at this rock outcrop.
Future surveys to assess moss coverage and other
habitat changes through time would be valuable at
multiple sites.
With our extensive search of the areas within
and bordering TSP, we did not detect any A. aeneus
CPUE dierences between regions and surprisingly,
124
Watkins et al.—Status of Green Salamanders in Mississippi, USA.
we found no dierence in CPUE across years despite
dramatically dierent environmental conditions
between years. The only dierence found in CPUE
was the 2.7× higher encounter rates observed on
public compared to private land. While some
locations on private land held large tracts of extensive
rock faces and mature forests, others had adjacent
forests that had been disturbed by logging activities
within the last 10–20 y. For example, based on time
lapse imagery on Google Earth, the adjacent forest
downslope of site R (CPUE: 0.16/h) was harvested
in the late 1990s/early 2000s, while adjacent forest
along the Natchez Trace Parkway (Site S; CPUE:
1.68/h) remained intact (Fig. 6). Further, other sites
had timber operations that left a large buer around
the rock face (> 100 m; e.g., site X), and high CPUE
was found at this location. Because A. aeneus
can be seasonally highly arboreal, prefer larger
trees near rock outcrops, and are associated with a
greater density of trees near rock outcrops (Waldron
and Humphries 2005; Smith et al. 2017), timber
harvesting activities near rock outcrops could have
detrimental impacts on populations. Alternatively,
timber harvesting activities farther away may have
less to no impact on the population.
.—We observed about a two-week
dierence in the initiation of egg brooding in 2021
(19 July) and 2022 (8 August). In an earlier study
in TSP, Woods (1968) reported a similar initiation of
egg laying at TSP on 15 July, and that egg laying was
short lived and occurred simultaneously throughout
TSP. Because 2021 was a wet year, it seems possible
that this could have inuenced the early egg-brooding
initiation, whereas the drier conditions of 2022
may have delayed egg-brooding initiation. In West
Virginia, Canterbury and Pauley (1994) observed that
females initiated egg brooding in the second week of
June and tended eggs through mid-August, so our
initiation dates occurred later in the season than more
northerly latitude sites.
We also found that about 11% of all A. aeneus
observations were egg brooding females compared
to about 1.5% observed in Tennessee (ve of 329;
Niemiller et al. 2022). Because we completed all
our surveys in July/August and most of the surveys
by Niemiller et al. (2022) were performed between
April and June (outside of egg deposition season),
our higher percentage of egg brooding female
observations is expected. Future studies should
consider the timing of their surveys and how they
coincide with the egg brooding season to determine
the reproductive activity at each site.
One interesting nding of this study was the
alternation among a high number of brooding females
in one year (2021) immediately followed by a high
number of juveniles the following year (2022), while
there were dramatically fewer brooding females during
a dry year (2022) and zero juveniles the following
year (2023). It seems possible that wet years stimulate
A comparison of forest harvesting on private and public land surrounding rock outcrops (red line) and observed Green Salamander
(Aneides aeneus) locations (green pins) in Tishomingo County, Mississippi, USA. (A) 1996 aerial photograph of private property (below green
line) and Natchez Trace Parkway Property (above green line). (B) 2004 aerial photograph of same location depicting timber harvest (orange
polygon) that occurred between 1996 and 2004. Tishomingo State Park is in the bottom right corner of both images. (Images from Google Earth
Pro).
125
Herpetological Conservation and Biology
more females to reproduce, and subsequently, a
high number of juveniles the following year. This
pattern seems to suggest that there could be boom
and bust years for A. aeneus reproduction similar to
other salamanders (Whiteman and Wissinger 2005).
Alternatively, there could be other interacting factors,
such as female biennial reproduction (Snyder 1991;
Canterbury and Pauly 1994) or juvenile survival, that
may also be inuencing the patterns observed. Thus,
additional investigations into the potential for cyclic
reproductive patterns in A. aeneus should be a topic
of future long-term studies.
.—Our study is the rst to estimate the
approximate linear distance of habitat to A. aeneus in
Mississippi and the rst to measure the approximate
extent of occupied area in the state. Compared to
other states within the range of A. aeneus, the extent
of habitat in Mississippi is miniscule. While it is
small in scale, we have found that the approximate
area of occupied habitat is > 50% more than
previously documented in the state. Further, most
of the rock outcrop habitat in Mississippi occurs on
public lands, but some extensive rock outcrops that
hosted A. aeneus populations also occurred on private
land (e.g., R, U-X). Similarly, Hardman (2014)
also found that private lands can harbor A. aeneus
populations and in some states, private land holdings
with rock outcrops could be massive (e.g., Tennessee,
Alabama). Private lands, however, are more dicult
to access for surveys and are underrepresented in
most A. aeneus studies. Nonetheless, while private
properties may hold populations that are currently
unknown to biologists and managers, our study is
the rst to document lower encounter rates of A.
aeneus on private lands possibly due to dierences in
land management in comparison to public lands. To
identify private landowners that have high potential
for rock outcrop habitat suitable for A. aeneus, using
LIDAR data and slope analyses could be a worthwhile
venture in other states (Aldridge et al. 2024).
From a microhabitat perspective, Woods (1968)
indicated that rock overhangs shade potential
crevices, and thus, it seems likely that overhangs
may promote a more stable crevice environment.
Our data seem to corroborate this, as A. aeneus were
overrepresented in overhang areas compared to
their availability. Furthermore, the climate/rainfall
could have had an interacting eect on the rock
microhabitat chosen by these salamanders, because
the weather was very dierent in 2021 (wet) versus
2022 and 2023 (drought). Indeed, A. aeneus are more
commonly observed in crevice openings when there
has been less rain in the 24 h preceding the survey
(Smith et al. 2017).
We found that A. aeneus occupied crevice heights
similar to the observations by Cliburn and Porter
(1987; mean height = 1.25 ± 0.59 m, range of values
0.3–3.96); however, many of the rock walls extended
well above our maximum survey height (about
2 m), and rock walls sometimes reached 15–20
m. Therefore, it seems very likely that numerous
individuals went unobserved in higher crevices.
Future surveys should focus on surveying heights
that are typically out-of-reach by most surveys (i.e.,
above 2 m).
One interesting aspect for further exploration
would be to quantify how Hartselle sandstone habitat
in Mississippi and northern Alabama compares to
other rock habitat types throughout the range of A.
aeneus. Qualitatively, Hartselle Sandstone appears
to have fewer crevices and is not as tall/extensive in
comparison to other rock habitat used by A. aeneus
in their range. Further, it may be that the Hartselle
Sandstone outcrops have dierent microclimates or
microhabitats compared to other rock geologies used
by A. aeneus in other parts of their range. If there
are rock morphology or structural dierences, those
dierences may explain why Hartselle rock outcrop
sites in Mississippi were occupied at a high rate and
had higher summer encounter rates/CPUEs compared
to other parts of the range. Furthermore, extensive
surveys for A. aeneus in the Hartselle Sandstone
in northern Alabama are lacking yet warranted. If
completed, an Alabama study would provide a
suitable and interesting comparison to this study.
—Because the
majority of our survey sites are within public owner-
ship (TSP and NTP), they are protected from some
of the documented threats to A. aeneus (e.g., mining,
clear-cut logging, etc.). Sites on private property,
however, are still vulnerable to management changes
that could negatively aect this species. For exam-
ple, some sites were bordered by large agricultural
areas, and others (sites R, T, and the north end of X)
had recently been logged up to the rock faces. To en-
sure high quality forest habitat around rock outcrops,
landowners should be contacted by and work with
state biologists regarding with property registration
conservation programs for these special habitats; de-
pending on the details of the agreement, forest buers
should be considered (Soto 2021). For sites that are
degraded and logged, it would also be benecial to
126
Watkins et al.—Status of Green Salamanders in Mississippi, USA.
plant trees in the areas that have favorable rock habi-
tat (Soto et al. 2021).
In summary, A. aeneus still occurs in a small area
of Mississippi, but it is found in many more locations
than previously documented. Further, encounter
rate data indicate Mississippi is similar to or above
other states within the range of the species. Our data
could be helpful for state managers when evaluating
the conservation status of the species and serve as a
baseline for future surveys. We also recommend that
a subset of these rock outcrops be regularly assessed
as part of a long-term monitoring program for the
species in the state.
Acknowledgments.—This project was funded
by a Competitive State Wildlife Grant through
the Southeastern Association of Fish and Wildlife
Agencies (SE-U2-F16AP00113) and a State Wildlife
Grant via the Mississippi Department of Wildlife,
Fisheries, and Parks and the U.S Fish and Wildlife
Service Division of Federal Aid (#F21AF03948).
Along with State Wildlife Grant funding, Millsaps
College Science Fellowship Awards also supported
this project, and we are grateful to Tim Ward and
Leah Babb for administering these funds. Field
activities for BW and WS were approved by
Mississippi Department of Wildlife, Fisheries, and
Parks through scientic research and collecting
permit numbers 0513211, 0510221, and 0510231.
The Animal Care and Use Committee of Millsaps
College also approved the project (WS041717). We
would like to thank Kaitlin Cross, Sheena Feist, and
Debora Mann for assisting with surveys, as well as
the many Millsaps College undergraduate research
students who helped (Jack Welsh, Thomas Weber,
Bryce King, Tyler Hamby, Peyton Parker, Braden
Robinson, Connor Ladner, Greyson Hewitt, Katie
Williams, and Lucas Rutherford). We are also
grateful to AirBNB homeowners Lisa and Arthur in
Belmont, Brooke in Iuka, and Wesley and Brittany
in Iuka for being such great and hospitable hosts.
We also appreciate the willingness of several private
landowners for allowing us to survey for salamanders
on their properties, and for Ruth Elsey for reviewing
an earlier version of the manuscript.
Aldridge, C.A., A. Moczulski, K.T. Wilson, and
E.C. Boone. 2024. Detecting sheer rock outcrops
using digital elevation models for ecological
conservation. IEEE Journal of Selected Topics in
Applied Earth Observations and Remote Sensing
17:6712–6720.
Canterbury, R.A., and T.K. Pauley. 1994. Time of mating
and egg deposition of West Virginia populations
of the salamander Aneides aeneus. Journal of
Herpetology 28:431–434.
Channell, R. 2004. The conservation value of peripheral
populations: the supporting science. Pp. 1–17 In
Proceedings of the Species at Risk 2004 Pathways
to Recovery Conference. Hooper, T.D. (Ed.). Species
at Risk 2004 Pathways to Recovery Conference
Organizing Committee, Victoria, British Columbia,
Canada.
Cliburn, J.W., and A.B. Porter. 1987. Vertical
stratication of the salamanders Aneides aeneus and
Plethodon glutinosus (Caudata: Plethodontidae).
Journal of the Alabama Academy of Science 58:18–
22.
Corser, J.D. 2001. Decline of disjunct Green Salamander
(Aneides aeneus) populations in the southern
Appalachians. Biological Conservation 97:119–126.
Gordon, R.E. 1952. A contribution to the life history
and ecology of the plethodontid salamander Aneides
aeneus (Cope and Packard). American Midland
Naturalist 47:666–701.
Hardman, R.H. 2014. Modeling occurrence of the
Green Salamander, Aneides aeneus, in the Blue
Ridge Escarpment. M.Sc. Thesis, Western Carolina
University, Cullowhee, North Carolina, USA. 42 p.
Hinkle, M.P., L.A. Gardner, White, E., W.H. Smith,
and R.D. VanGundy. 2018. Remnant habitat
patches support Green Salamanders (Aneides
aeneus) on active and former Appalachian
surface mines. Herpetological Conservation and
Biology 13:634–641.
John, R.R. 2017. Movement, occupancy, and
detectability of Green Salamanders (Aneides
aeneus) in northern Alabama. M.Sc. Thesis, Auburn
University, Auburn, Alabama, USA. 73 p.
Merrill, R.K., D.E. Gann, and S.P. Jennings. 1988.
Tishomingo County geology and mineral resources.
Bulletin 127, Mississippi Department of Natural
Resources, Bureau of Geology, Jackson, Mississippi,
USA. 178 p. plus color plates.
Milanovich, J.R., W.E. Peterman, N.P. Nibbelink,
and J.C. Maerz. 2010. Project loss of a salamander
diversity hotspot as a consequence of projected global
climate change. PLoS ONE 5: e12189. https://doi.
org/10.1371/journal.pone.0012189.
Mississippi Museum of Natural Science. 2015.
127
Herpetological Conservation and Biology
Mississippi State Wildlife Action Plan. Mississippi
Department of Wildlife, Fisheries, and Parks,
Mississippi Museum of Natural Science, Jackson,
Mississippi, USA. 692 p.
Newman, J.C., K. Barrett, and J.W. Dillman. 2018.
Green salamander estimated abundance and
environmental associations in South Carolina. Journal
of Herpetology 52:437–443.
Newman, J.C., E.A. Riddell, L.A. Williams, M.W.
Sears, and K. Barrett. 2022. Integrating physiology
into correlative models can alter projections of habitat
suitability under climate change for a threatened
amphibian. Ecography 2022: e06082. https://doi.
org/10.1111/ecog.06082.
Niemiller, M.L., R. Hardman, D. Thames, D. Istvanko,
M.A. Davis, C. Ogle, K.D.K. Niemiller, K.E.
Dooley, and T.M. Clark. 2022. The distribution
and conservation status of the Green Salamander
(Aneides aeneus) in Tennessee, USA. Herpetological
Conservation and Biology 17:249–265.
Pauley, T.K., and M.B. Watson. 2005. Aneides
aeneus. Pp. 656–658 In Amphibian Declines: The
Conservation Status of United States Species.
Lannoo, M. (Ed.). University of California Press,
Berkeley, California, USA.
Petranka, J.M. 1998. Salamanders of the United
States and Canada. Smithsonian Institution Press,
Washington, D.C., USA.
Rauch, T.J., III, J.R. Lee, and A. Lawler. 2016. Green
Salamander (Aneides aeneus) populations in
Mississippi: a preliminary update. Journal of the
Mississippi Academy of Sciences 61:222–226.
Smith, W.H. 2024. Woody understory vegetation
removal alters the microclimate proles of rock
crevices used by Green Salamanders (Aneides
aeneus) in the Cumberland Mountains of Virginia,
USA. Herpetological Conservation and Biology
19:23–32.
Smith, W.H., S.L. Slemp, C.D. Stanley, M.N. Blackburn,
and J. Wayland. 2017. Rock crevice morphology and
forest contexts drive microhabitat preferences in
the Green Salamander (Aneides aeneus). Canadian
Journal of Zoology 95:353–358.
Snyder, D.H. 1991. The Green Salamander (Aneides
aeneus) in Tennessee and Kentucky, with comments
on the Carolina’s Blue Ridge populations. Journal of
the Tennessee Academy of Science 66:165–169.
Soto, K.M., R.K. McKee, and J.C. Newman. 2021.
Conservation Action Plan: Green Salamander
(Aneides aeneus) Species Complex. Southeast
Partners in Amphibian and Reptile Conservation,
Amphibian and Reptile Conservancy, Inc., Louisville,
Kentucky, USA. 15 p.
Steen, D.A., and K. Barrett. 2015. Should states in the
USA value species at the edge of their geographic
range? Journal of Wildlife Management 79:872–876.
Thompson, D.E. 2011. Geologic map of Mississippi.
Mississippi Oce of Geology, Jackson, Mississippi,
USA.
Tilley, S.G. 1980. Life histories and comparative
demography of two salamander populations. Copeia
1980:806–821.
U.S. Fish and Wildlife Service. 2015. Endangered and
threatened wildlife and plants; 90-day ndings on 31
petitions. Federal Register 80:37568–37579.
Waldron, J.L., and W.J. Humphries. 2005. Arboreal
habitat use by the Green Salamander, Aneides aeneus,
in South Carolina. Journal of Herpetology 39:486–
492.
Walker, K.L., and W. Goodpaster. 1941. The Green
Salamander, Aneides aeneus, in Ohio. Copeia
1941:178.
Whiteman, H.W., and S.A. Wissinger. 2005. Amphibian
population cycles and long-term data sets. Pp. 177–
184 In Amphibian Declines: The Conservation Status
of United States Species. Lanoo, M. (Ed.). University
of California Press, Berkley, California, USA.
Williams, L., K.A. Parker, Jr., C.R. Lawson, and A.D.
Cameron. 2020. Variation in seasonal detection
probability and site abundance of populations of
Aneides aeneus and A. caryaensis in North Carolina,
USA. Herpetological Conservation and Biology
15:238–252.
Woods, J.E. 1968. The ecology and natural history
of Mississippi populations of Aneides aeneus and
associated salamanders. Ph.D. dissertation, University
of Southern Mississippi, Hattiesburg, Mississippi,
USA. 89 p.
128
Watkins et al.—Status of Green Salamanders in Mississippi, USA.
completed research on Green Salamanders as an Honor’s Student in Biology at Millsaps
College in Jackson, Mississippi, USA. He graduated with a B.S. in Biology in May 2023, and he is currently
serving as a Scientist 3 at Allen Engineering and Science in Mobile, Alabama, USA. (Photographed by Will
Selman).
has been a Zoologist with the Mississippi Natural Heritage Program at the Mississippi
Museum of Natural Science, Mississippi Department of Wildlife, Fisheries, and Parks, in Jackson, Mississippi,
USA, since 1990. Tom received his B.S. degree at the University of Miami (Florida, USA) in 1972, and his
M.S. degree at Florida Atlantic University, Boca Raton, Florida, USA, in 1977, where he studied the impact
of developed shoreline in southeast Florida on nesting and hatchling sea turtles. As Natural Heritage Program
Zoologist, Tom has been involved in more general surveys throughout the Scala Naturae but has also undertaken
focused surveys on the distribution and abundance of Gopher Tortoises (Gopherus polyphemus), Diamond-
backed Terrapins (Malaclemys terrapene pileata), Gulf Saltmarsh Watersnakes (Nerodia clarkii), the Alabama
Red-bellied Cooter (Pseudemys alabamensis), and Florida Harvester Ants (Pogonomyrmex badius). More
recently, in collaboration with Dr. Debora Mann, Tom has studied previously unknown migratory behavior in
Webster’s Salamander (Plethodon websteri), and morphological variation in this species throughout its range.
(Photographed by Katelin Cross).
is an Associate Professor of Biology at Millsaps College in Jackson, Mississippi, USA. He
received his B.S. from Millsaps College and his Ph.D. from the University of Southern Mississippi. Recent
research includes the distribution and call phenology of Crawsh Frogs (Rana areolata) in central Mississippi,
the distribution and abundance of Graptemys turtles in the Tombigbee River system of northeastern Mississippi,
and surveys for rare salamanders in northeastern Mississippi; many of these studies emphasize the involvement
of undergraduate students. He is also currently working on a book, The Amphibians and Reptiles of Mississippi.
(Photographed by Will Selman).
Site Rock Region Total #S
Public/
Private Aa Present?
CPUE
# S Eort Aa
Aa
CPUE Ec Elo Elu Pg
Rock Quarry Branch A West 4 Public X 3 24.45 4 0.12 8 9 6
Gatehouse Rock B West 12 Public X 7 8.82 1* 0.08 1 12 19
Disgo Rock B West 4 Public - 3 8.28 0.00 11 3
Saddleback Ridge B’ West 1 Public - 1 4.92 0.00 2
Classic Rock C West 6 Public X 4 20.28 21* 1.18 1 3 2
Caution Rock D West 4 Public X 1 4.25 2 0.47 1
Long Tail Rock D’ West 1 Public - 1 3.33 0.00 8
Castle Rock A E West 5 Public X 2 3.34 1 0.30 1 6
Kenobi Rock E’ West 1 Public - 1 1.58 0.00 1
Spring Hill F West 39 Public X 5 38.68 42* 1.19 315 98
Drip Rock G West 8 Public X* 3 6.15 0.00 6 10
Dining Hall Rock H West 4 Public X 3 2.92 3 1.03 1
Cabin Cli I West 5 Public X 3 21.92 3* 0.08 234
Pavilion to Cabin J West 3 Public X 2 11.65 4* 0.47 13
Pavilion K West 8 Public X 4 33.02 8* 0.19 1 5 32
CC Camp L West 3 Public X 2 7.07 5* 0.71 1 4
Old Quarry M West 4 Private - 2 9.1 0.00 2 6
Rock N N Central 4 Public X 3 5.32 4* 0.75 3 3
O Trail O Central 5 Public X 3 10.03 4* 0.40 3 5
O negative Rock O- Central 1 Public X 1 2.75 11* 4.00 2 3
Swinging Bridge Far
East P Central 3 Public X 2 18.0 13* 0.83 4 21
Swinging Bridge East Q Central 8 Public X 5 52.54 39* 0.75 1 31 52
Surveys for Green SalamanderAneides aeneus) at rock outcrop sites in Tishomingo County, Mississippi, USA, 2017–2023. The column labeled Rock is the rock
outcrop location as depicted in Figure 2, and Eort refers to the number of person hours of survey eort at each site. Abbreviations are Total #S = number of observation surveys from
2017–2023, * = presence of reproducing A. aeneus, CPUE #S = number of encounter rate surveys from 2021–2023, Aa = Aneides aeneus (Green Salamander), Aa CPUE = Aneides
aeneus Catch Per Unit Eort, Ec = Eurycea cirrigera (Two-lined Salamander), Elo = Eurycea longicauda (Long-tailed Salamander), Elu = Eurycea lucifuga (Cave Salamander), and Pg
= Plethodon glutinosus complex (Slimy Salamander).
129
Site Rock Region Total #S
Public/
Private Aa Present?
CPUE
# S Eort Aa
Aa
CPUE Ec Elo Elu Pg
Swinging Bridge West R Central 6 Private X 3 25.43 4* 0.16 4 31 27
Rock S S Central 5 Public X 2 25.55 42* 1.68 1 7 45
TNT Rock T Central 5 Public - 3 9.72 0.00 15 2
Bloody Springs South U East 7 Private X 2 14.65 1* 0.07 10 2 38
Bloody Springs West V East 10 Private X 5 35.23 2 0.09 59 2 70
Bloody Springs East W East 5 Private X 4 26.30 12* 0.68 4 37
Rock X north X East 4 Private X 2 25.08 20* 0.80 1 7 5 30
Rock Y Y East 4 Private - 1 8.00 0.00 1 1 10
Rock Z1 Z1 Central 2 Public X 2 5.52 0.00 4 4
Rock Z2 Z2 Central 2 Public X 2 18.18 8* 0.43 12 2
Total 182 25 of 32 87 482.9 254 42 205 7 586
CPUE 0.53 0.09 0.43 0.01 1.21
%RA 0.22 0.04 0.18 0.01 0.55
, continued
130
