Descriptions of a new species and a new subspecies of freshwater mussels, Epioblasma ahlstedti and Epioblasma florentina aureola (Bivalvia: Unionidae), in the Tennessee River drainage, USA

2010·10 MB·English

by  Jess W Jones

THE NAUTILUS 124(2):77-92, 2010 Page 77
Descriptions of a new species and a new subspecies of freshwater
mussels, Epioblasma ahlstedti and Epioblasmaflorentina aureola
USA
(Bivalvia: Unionidae), in the Tennessee River drainage,
Jess W. Jones RichardJ. Neves
U.S. Fish andWildlife Service Department ofFisheries andWildlife Sciences
DepartmentofFisheries andWildlife Sciences VirginiaPolytechnicInstitute and State University
VirginiaPolytechnic Institute and State University Blacksburg,VA24061 USA
Blacksburg, VA24061 USA
[email protected]
ABSTRACT riversystem overthepast 100years have reducedmussel
diversity to approximately 80 extant species (Parmalee
A new species and a new subspecies of Epioblasma are de- and Bogan, 1998; Williams et al., 2008). The decline of
scribed from the Tennessee River drainage, USA. Epioblasma species belongingto the genus Epioblasma was the most
ahlstedti (Duck River Dartersnapper) currently is restricted severe (Johnson, 1978). Ofthe 18 species andsubspecies
to the Duck River in west-central Tennessee (TN). However, known from the system, only four remain, and of those
museum collections indicate that the species likely occurred in
the Buffalo River, TN, atributarytothe Duck River, and in the lost, most are considered extinct and one is extirpated
Tennessee River at Muscle Shoals, Alabama (AL), and lower but remains in a small isolated population in Ohio. Spe-
Shoal Creek, AL. The following diagnostic morphological char- cies in this genus have specialized reproductive traits,
acteristics of E. ahlstedti are based on the female: (1) pro- including species with shell morphologies and mantle-
nounced posterior-ventral shell expansion of the adult female lures that can attract and capture their fish hosts to
shell; (2) slate-gray to dark-purple mantle-pad; (3) spongy tex- facilitate infestation by the gloehidia (Jones et al., 2006;
ture ofthe mantle-pad; and (4) display ofa single, tan-colored Barnhart et al., 2008). Epioblasma shells are character-
micro-lurethat movesslowlyside-to-side. Epioblasmaflorentina ized by small to medium sizes (~30-70 mm) and sexual
aureola (Golden Riffleshell) currently is restricted to Indian dimorphism. The posterior-ventral end of the female
Creek, a tributary to the upper Clinch River, Virginia. Histori-
cally, the species occurred in numerous tributaries in the Ten- shell is expanded, to form a distinct protrusion, or an
nessee River drainage downstream at least to the Duck River. area of the shell herein termed, the shell expansion
The following diagnostic morphological characteristics of (Figures 2, 3, 5, 8). This distinctive feature of female
E.florentina aureola are based on the female: (1) gray mantle- Epioblasma was considered by Walker (1910) to be the
pad with a black mottled background; and (2) mantle-pad is “highest expression of unionid development. The shell
pustuled but the pustules are rounded. The genus Epioblasma expansion houses a modified portion of the mantle,
represents the mostendangeredgroup offreshwater mussels in known as the mantle lure that functions to attract host
North America; 18 ofthe recognized 25 species or subspecies fish. The focal species ofthe current studybelong to the
are alreadyextinct. Likewise,these newlydescribed species and Epioblasma subgenus Tondosa (Johnson, 1978), which
subspecies are critically endangered and despite being listed as have a prominent shell expansion and a mantle-lure that
endangeredunderthe Endangered Species Act remain in need
contains a mantle-pad and micro-lure (Figures 7, 9-14).
offocusedconservationtopreventtheirextinction.
The mantle-pad is afoldedand articulatedportionof the
Additional keywords: Endangered, molecular DNA markers, mantle, and the micro-lure is a modified and innervated
phylogeneticanalysis, Tennessee Riverbasin incurrent aperture papilla that moves to attract fish
hosts, seemingly mimicking aquatic insect larvae (Jones
et al., 2006; Barnhart, 2008).
Such morphological and life historyspecialization may
havecontributedtothe decline ofEpioblasma species, as
INTRODUCTION changes in environmental conditions over the 19 1 and
20thcenturies mayhavedisruptedthiscomplexlife cycle.
TheTennessee Riverandits tributaries support the most Unfortunately,thelossofnumeroustaxain thegenushas
species-rich mussel assemblage in North America (NA) prevented a more complete understanding of the life
(Figure 1), with 102 species known historically from the history and taxonomy of the group, based on modern
system (Parmalee and Bogan, 1998). Unfortunately, pol- diagnostic methods usingDNA and mantle-luredisplays.
NA
lution and hydrological modifications (e.g., dams) to the Since only 7 of the 25 species and subspecies in
Page 78 THE NAUTILUS, Vol. 124, No. 2
)
Figure 1. Distribution ofEpioblasma ahlstedti (current •, historic O) and E florentina aureola (current historic in the
Tennessee Riversystem, USA. AlsoshownisthedistributionofE. florentinawalkeri (current A historic A)intheCumberland River
system.
remain (Williams et ah, 1993), it is possible that
additionaltaxawere never identified before their extinc-
tion, while other described taxa may have been pheno-
typicvariants of the same species or subspecies.
The diverse phenotypic variation within Epioblasma
,
especiallythe varied shell morphologies and mantle-lure
displays of females, has allowed for a more comprehen-
sive approach to understanding the taxonomy of extant
species within the group, one that includes both tradi-
tional phenotypic characters and molecular markers.
The study conducted by Jones et al. (2006) showed that
the population ofEpioblasma capsaefonnis in the Duck
River, Tennessee (TN) was distinct from the population
in the Clinch River, TN and Virginia (VA), and that the
population of Epioblasma florentina walkeri in Indian
Creek, VA, a tributary to the upper Clinch River, was
distinct from the population in the Big South Fork
Cumberland River, TN and Kentucky (KY). Based on Figure 2. The height (axis-A) andlengthofthebase (axis-B)
oftheshellexpansionareshown. Thearrowsonaxis-Bpointto
extensive phenotypic data (e.g., shell morphology, the articulation points ofthe posterior-ventral shell expansion
m(ea.ngt.,lem-iltuorcehso,ndfrisihalhoDsNtAs,pencuicfilceiatry)DaNnAd mmioclroescautleallritdeast),a wBiutrhchth(e19m7a5i)nabnoddwyiotfhtpheermsiheslsli.oTnhferofmiguDrre.wJ.aBs.mBoudricfhi.ed from
these authors recommended the reclassification of the
Duck River and Indian Creekpopulations ofEpioblasma
,
respectively. Ilowever, thestudydidnotformallydescribe MATERIALS AND METHODS
and provide taxonomic recognition to these populations.
Thus, the purpose of this paper is to present formal de- Typespecimens,othershell material,andcollectionrecords
scriptions and provide scientific and common names for for Epioblasma ahlstedti E. capsaefonnis, E. florentina
,
the newspecies and new subspecies of freshwatermussel aureola E. florentina florentina and E. florentina walkeri
,
in the Tennessee Riversystem. were examined at the following museums: Academy of
J.W. Jones and R. Neves, 2010 Page 79
J.
Figures3-4. 3. Holotype (OSUM 68523) ofEpioblasmaahlsteclti (female),50.7mmlongand37.8mmhigh; 4. Paratype (OSUM
82238) ofEpioblasma ahlsteclti (male),45.8 mm iongand32.0mm high. Photosby G. ThomasWatters.
Natural Sciences of Philadelphia, Pennsylvania (ANSP); shell (Figure 2). Digital calipers were used to measure
Carnegie Museum, Pittsburgh, Pennsylvania (CM); shell dimensions to the nearest 0. mm. To test for
1
Florida Museum of Natural History, University of differences, the slopes of the fitted-lines of each
Florida(UF),Gainesville,Florida; MuseumofComparative regression equation were compared among populations
Zoology, Harvard University, Cambridge, Massachusetts using the homogeneity of regression coefficients test
(MCZ); Ohio State University Museum of Biological statistic.
Diversity, Columbus, Ohio (OSUM); and National Fecundity was estimated by counting the number of
Museum of Natural History, Smithsonian Institution, glochidia from each of 6-10 females per population.
Washington, D.C. (USNM) [see Jones (2004) for all Although not a diagnostic trait in this study, it helped to
examined lots]. Type specimens provided standard demonstrate quantitative differences amongspecies.
references for comparing shell material from various Live female mussels for fecundity analysis were col-
rivers, and collection records were used to construct lected from the following riverlocations: (1) E. ahlsteclti,
species distributions. Museum specimens ofE. ahlsteclti Duck River, Lillard Mill [River Kilometer (RKM) 2S7.7],
wereidentifiedusingonlyfemale shells. Marshall County (Co.), TN; (2) E. capsaeformis, Clinch
To assess differences in the shell expansion of adult River between Horton Ford (RKM 321) and Swan
female shells among populations, simple linear regres- Island (RKM 277), Hancock Co., TN; (3) E. florentina
sion equations of total length (x-axis) versus the height aureola Indian Creek, a tributary to the upper Clinch
((/-axis) of the shell expansion were computed for each River a,t RKM 518.2, Tazewell Co., Virginia (VA);
population. The height of the shell expansion was (4) E. f. walked, Big South Fork Cumberland River,
measured from its base, which is the length between Station Camp Creek, Scott Co., TN, downstream to
the two articulation points with the main body of the Bear Creek, McCreary Co., KY (Jones et al. 2006).
Page 80 THE NAUTILUS, Vol. 124, No. 2
Figures5-6. 5. Holotype (OSUM 82239) ofEpioblasmaflorentinaaureola (female),44.1 mmlongand34.8mmhigh; 6. Paratype
(OSUM 82240) ofEpioblasmaflorentinaaureola (male), 48.1 mmlongand33.4mm high. Photos byG. Thomas Watters.
Fecundity was compared using analysis of variance 2003), (3) cytochrome-b (Merritt et ah, 1998; Bowen and
(ANOVA). All statistical analyses ofshells and glochidia Richardson, 2000), and (4) ITS-1 (King et ah, 1999). All
were conducted in MINITAB 14 Statistical Software PCR products were sequenced with a Big Dye Termina-
(Minitab, Inc., State College, PA). tor Cycle Sequencing kit with AmpliVhq DNA Polymer-
Samples of mantle tissue from live mussels were ase (Applied Biosystems). Cycle sequence reactions
collected from the above riverlocations and, additionally, were purified using a Qiagen DNA Purification kit
individuals of E. torulosa rangiana were collected (Qiagene), and subjected to electrophoresis and
from the Allegheny River, Venango Co., Pennsylvania sequencing using an Applied Biosystems 3100 auto-
(PA) (Jones et ah, 2006). Other subspecies were not mated sequencer [detailed PCR methods are available
included in the studybecause theyarepresumedextinct, in Jones (2004) andJones et ah (2006)].
i.e., E. florentina florentina E. florentina curtisi and Phylogenetic analysis of DNA sequences and mor-
, ,
E. t. torulosa. A small piece ofmantle tissue (20-30 mg) phological characters were conducted to infer genealog-
was collected non-lethally from 8-20 live mussels from ical relationships among Epioblasma spp. Sequences
each population (Naimo etah, 1998). from mtDNA and nDNA were combined, and six mor-
Sequences of three regions of mitochondrial DNA phological characters (Table 1) were included in the
(mtDNA) and one region ofnuclear DNA (nDNA) were character matrix for a total evidence analysis (Kluge,
amplified by polymerase chain reaction (PCR) using 1989). DNA sequences were edited and aligned using
pRrNimAer(sLyadnedarcdondeittiaoh,ns1r9e9p6o)rt(e2)d NinD:1(,1)f1ir6sSt,sruibbuonsiotmaolf tChoedesprCoogrrpoarmatiSonE)Q,UanEdNpChHylEogRene(tviecrsainoanlysi3s.0,wasGepenre-
NADII dehydrogenase (Buhay et ah, 2002; Serb et ah. formed using PAUP* (version 4.0bl0, Swofford, 1998).
J.W. Jones and R.J. Neves, 2010 Page 81
Figures 7-14. Mantle-paddisplaysoffemaleEpioblasmaahlstedti E.florentinaaureola andcongeners. 7. Mantle-padandmicro-
,
lure ofE. ahlstedti. Duck River, Marshall Co., TN. 8. Protruded shell expansion offemale E. ahlstedti. 9. E. capsaefonnis mantle-
pad,Clinch River, HancockCo.,TN. 10. Doublemicro-luredisplayofE. capsaefonnis 11. E. florentinaaureola mantle-pad.Clinch
;
River,Tazewell Co.,VA. 12. Micro-lureofE. florentinaaureola. 13.E. florentina walked mantle-padandmicro-lure, BigSouth Fork
Cumberland River, ScottCounty,TN. 14. Pustuled mantle surface ofE. florentina walked. Thegaps between shellvalves offemale
musselsareapproximately2cm. Thearrowsarepointingatmicrolures.This figurewaspreviouslypublishedinJonesetal. (2006)but
updatedinthisstudytoincludethenewspeciesandnewsubspeciesscientificnames;photograph in Figure9hasnotbeenpreviously
published. Photographs ofthe mantle-paddisplayofE. torulosa rangiana areavailable in Jones (2004).
Page 82 THE NAUTILUS, Vol. 124, No. 2
Table I. Matrix and coding for shell and mantle-lure (Epioblasma brevidens) and Snuffbox (Epioblasma
characters for Epioblasma species. Character states were triquetra) were designated as out-group taxa. The study
determined from direct obse—rvation and from those reported by Jones et al. (2006) showed that DNA sequences from
in Jonesetal. (2006). Agap( ’)indicatesthecharacterwasnot these two species are diverged from the in-group taxa
applicable to the species. by ~5%, and based on obvious differences in shell mor-
CHARACTER MATRIX phology, Johnson (1978) classified E. brevidens and
E. triquetra intodifferentEpioblasma subgenera, Plagiola
Species 1 2 3 4 5 6 7 8 and Truncillopsis respectively. Furthermore, previous
,
Epioblasma ahlstedti 1 i 0 0 0 - i 0 phylogenetic studies have demonstrated the monophyly
Epioblasmacapsaeformis 1 i 0 1 1 - 2 1 of the Epioblasma among North American unionids
Epioblasmaflorentina aureola 1 i 1 2 2 0 i 0 andthatE. brevidens and E. triquetra are basal to the in-
Epioblasmaflorentina walkeri 1 i 1 3 2 1 i 0 group taxa (Campbell et al., 2005; Zanatta and Murphy,
Epioblasma torulosa rangiana 1 0 2 4 i - 0 2 2006). Thus, available morphologicalandgeneticdatajus-
Epioblasmabrevidens 0 tifyuseofthesetwospeciesasappropriateout-grouptaxa.
Epioblasma triquetra 0 Photographs ofthe mantle-padand micro-lures oflive
CHARACTERS: female mussels were taken usinga NikonosVunderwater
(1) Prominent posterior-ventral shell expansion of adult camera with 28 or 35 mm macro-lenses and Kodak 200
female. 0=absentordiminutive; l=present. Ektachrome film. Female mussels were held in tempera-
(2) Denticulations present along margin ofposterior-ventral ture-controlled water in recirculating artificial streams
shellexpansion. 0=absent; 1=present. with gravel-filled bottoms. This setup allowed females to
(3) Periostracum color and ray pattern. 0=yellow-green
display their mantle-pad and behavioral observations of
periostraeum with irregularly spaced green rays;
l=honey-yellow periostracum with line green rays micro-lure movements to be recorded under controlled
evenly spaced over entire shell surface; 2=brown conditions [photographic methods were previously
periostracumwithirregularlyspacedgreen rays. publishedin |ones et al. (2006)].
(4) Mantle-padcolor. 0=darkpurpletoslategray; l=blue to Epioblasma ahlstedti newspecies
bluish-white; 2=gray with mottled black background;
3=brownwith mottledtanbackground; 4=white. Duck River Dartersnapper
(5) Mantle-padtexture. 0=spongy; l=smooth; 2—pustules. Figures 3, 4, 7, 8
(6) Mantle-padpustules. 0=rounded; l=pointed.
(7) Number of micro-lures prominently displayed. 0=0; Diagnosis: The following diagnostic morphological
1=1; 2=2. characteristics of Epioblasma ahlstedti are based on the
(8) Mantle-padisinvaginatedwhereit meetsincurrentaper- female and are summarized in Table 2: (1) pronounced
ture. 0=yes; l=no; 2=incomplete. posterior-ventral shell expansion of the adult shell;
(2) slate-gray to dark purple mantle-pad; (3) spongy tex-
ture of the mantle-pad; and (4) display of a single,
tan-colored micro-lure that moves slowly side-to-side. In
Phylogenetic trees were evaluated using the maximum young individuals, the base ofthe shell expansion is con-
parsimony (MP) criterion because the extent of stricted, appearingnarrowbut distinctlyprotruded. How-
sequence divergence was low among in-group taxa (Nei ever, as the female shell grows, the shell expansion
and Kumar, 2000; Felsenstein, 2004). Characters were becomes extremelyprotrudedandenlarged, comparedto
treated as unordered andof equalweight forthe analysis the main body of the shell. The shell expansion of
due to in-group taxa being closely related (Nei and E. ahlstedti is distinguishable fromthatofE. capsaeformis
TtKhuBemRabrro,patn2ic0ohn0s-0;)a.nidnT-sbheroetuinotdnrseemaesntedharodcdehlewtwiiatoshnscAwoCenrdCeuTcRttreAedaNtuesdainnadgs usfisroionnmgbta~hs3ee5-fio4nl5lyoowmuimnng,g cirtniydtpieivrciiaadl:ulay(l1s)apl(pe~en3ag-rts5h oyc)fo,ntsrhtaernisgchtieenldgl,einbxepsaiinnz-ge
missingdata. Bootstrap analyses (10,000 replicates) were ~5-10 mm narrower basally than those of female
conducted using the FAST step-wise addition option of E. capsaeformis of similar age and size, and (2) mean
PAUP* to assess support fortheindividual nodes ofeach height (9.6 mm) and maximum height (24.6 mm) ofthe
phylogenetictree (Felsenstein, 1985). An additionalphy- shell expansion of adult females is significantly greater
logenetic analysis was conducted using Bayesian infer- thanthatoffemale E. capsaeformis (Table3; Figure 15).
ence in MrBayes v3.0b4 (Huelsenbeek and Ronquist,
DNA
2001), where the and morphological data sets Description: Length of the female shell can reach
were combined followingthe approach ofNylander et al. 60 to 70 mm, with mean length ~42 mm (Table 3). Male
(2004). MrBayes was run for I million generations, shell lengths are similar. The shell outline of males is
sampling trees every 100 generations and posteriorprob- typically elliptical, appearing pointed at the posterior
abilities were computed after a burn-in of 40,000 genera- end (Figure 4), whereas the shell outline of females is
tions. In-group taxa were E. ahlstedti, E. capsaeformis, more sub-oval and rounded, with a veiy enlarged and
E.florentinaaureola E.florentinawalkeri and E. torulosa protrudedposterior-ventral shellexpansion (Figures 3, 8).
, ,
rangiana. Because ofsignificant differences in morphol- The mean height ofthe shell expansion ofthe female is
ogy and DNA sequences, the Cumberland Combshell ~10 mm or ~23% ofshell length, but maximum height is
1
J.W. Jones and R.J. Neves, 2010 Page S3
Table 2. Diagnostic morphological and molecular genetic characters for Epioblasma alilstedti, E.florentina aureola and closely
related taxa. Data are summarized from Jones et al. (2006).
DNAregions andbase-pairsitepositions
Species orsubspecies Moqihologicalcharacters mtDNA nuclear DNA
Epioblasmaahlstedti • shell expansion offemale: largeand • 16S: 2 • ITS-1: 152, 153
protruded inadults • cytochrome-b: 43, 208,
• mantle-padcolor: darkpurpleto 211, 214
slate-gray • ND1: 281,564
• mantle-padtexture: spongy
• micro-lure display: 1 lureprominent;
rotatesclockwise sweepingside-to-side
Epioblasma capsaefonnis • mantle-padcolor: bluetobluish-white • 16S 179 • ITS-h 467, 468
:
• mantle-padtexture: smooth • cytochrome-b: 13, 87,
• micro-lure display: 2together; theleft 148, 184, 327
lurerotates clockwise,whilethe right • ND1: 282, 444
lurerotates counter-clockwise
• posteriorportion ofthe mantle-pad
is notinvaginatedwhereitmeets
incurrentaperture
Epioblasmaflorentina • mantle-padcolor: lightgraywith • cytochrome-b: 252
aureola mottled blackbackground • ND1 405
:
• mantle-padtexture: roundedpustules
• micro-lures: 1 lureprominent; rotates
clockwise sweepingside-to-side
Epioblasmaflorentina • mantle-padcolor: brownwith mottled • cytochrome-b: 52, • ITS-h 215
walkeri tan background 96
• mantle-padtexture: pointedpustules • ND1: 249, 423
• micro-lures: 1 lureprominent; rotates
clockwise sweepingside-to-side
Epioblasma tomlosa • denticulations: absent from posterior- • 16S: 167, 452 • ITS-1:44, 127,385,
rangiana ventralshell expansion • cytochrome-b: 99, 386, 387, 388, 51
• mantle-padcolor: white 172
• micro-lures: absent • ND1: 30
Table 3. Mean shell length, mean height ofshell expansion, and linear regression equations ofshell expansion height (y-axis) to
total length (x-axis) of adult female mussels. Pairwise comparisons of regression equation slopes were significantly different
(p<0.001), except equations A vs D, and C vs D. *The p-value indicates the significance level of the slope for each regression
equation.
Mean lengthofshell Meanheightofshell
Mussel species N (mm) (range) expansion (mm) (range) Regressionequations A-D R2 p-value*
Epioblasma ahlstedti 62 41.9 (31.0-68.9) 9.6 (3.1-24.6) A-y =-19.8 + 0.701x 0.76 p<0.001
EEppiioobbllaassmmaafclaoprseanetfionnanaiusreola 6535 4400..70 ((3300..47--5486..41)) 58..73 ((21..80--1140..62)) CB.-ly/ ==--41.69.27 ++00..536214.xx 00,.8554 pp<<00..000013
Epioblasmaflorentina walkeri 20 40.5 (30.3-45.2) 8.5 (2.0-14.1) D.y :-19.7 + 0.696x 0.51 p<0.007
~25 mm (Table 3), and it is dark green, sometimes The color of the foot and gills of E. ahlstedti is dull
appearing almost black. Denticulations occur along the white. In females, only the two outer gills are marsupial
margin of the shell expansion and are typically ~0.5 to (i.e., the outer water-tubes contain and brood glochidia
1 mm long and spaced ~0.5 to 1 mm apart. The whengravid). Locatedatthe distalendofeach marsupial
periostracum of adults isyellowish green, becomingmore water-tube is apore,which allows forrelease ofglochidia.
yellowish at the anterior end. The shell surface contains The mantle-pad is slate-gray to dark purple and has a
distinct broad to fine green rays that are typically irregu- spongy texture, and the micro-lure is tan (Figure 7). The
larly spaced. The male shell is short and high with a posterior portion ofthe mantle-pad is invaginated where
shallow sulcus. Nacre color is white, but hues of blue it meets the incurrent aperture, so the attachment points
and salmon may be present, especially near the beak of the micro-lures cannot be seen when the female is
cavities. displaying(Jones et al., 2006).
Page 84 THE NAUTILUS, Vol. 124, No. 2
30
03035404550566065
70
Figure 15. Relationship of posterior-ventral shell expansion height versus shell length of female Epioblasma ahlstedti and
E. capsaeformis, and offemale E. florentina aureola and E. florentina walked. Pairwise comparison ofregression equation slopes
was significantly different (p<0.001) between Epioblasma ahlstedti and E. capsaeformis but not significantly different between
E florentina aureola and E. florentina walked. However, the slope of each regression equation is significant lor all four taxa
(seeTable3).
Type Material: Holotype: Designated herein as Type Locality: Duck River, Lillard Mill, MarshallCo.,
OSUM 68523, collectedby 11.15. Athearn, from type local- TN, 35°35'09.08" N; 86°47'14.07" W.
ity, 30 Sep. 1956. Paratypes: CM 61.11669, Duck River,
Lillard Mill, MarshallCo.,TN,collectedbyA.E. Ortmann, Comparisonwith SimilarSpecies: The shell ofadult
25 Aug. 1923; MCZ 236214, Duck River, Lillard Mill, Epioblasma capsaeformis (.sensu stricto) in the Clinch
Marshall Co., TN, collected by H.D. Athearn, 30 Sep. River is of small to medium length (~30-50 mm). The
1956; OSUM 52509, Duck River, Lillard Mill, Marshall shell surface contains distinct broad to line green rays
Co., TN, collected by S.A. Ahlstedt, Oct. 1982; OSUM that are irregularly spaced, and veiy similar to those of
I
82238, Duck River, Lillard Mill, Marshall Co., TN, col- E. ahlstedti. However, the followingdiagnostic, morphol-
lected by S.A. Ahlstedt, Oct. 1999 (see Appendix for ogical characteristics of female E. capsaeformis distin-
I
other material examined); OSUM 82241, Duck River, guish it from E. ahlstedti: (1) bluish-white mantle-pad
Lillard Mill, MarshallCo.,TN, collectedbyH.D. Athearn, (Figure 9), (2) smooth texture of the mantle-pad, and
30 Sep. 1956. (3) simultaneous display ol two micro-lures that move
J.W. Jones and R.J. Neves, 2010 Page 85
Table 4. Fecundity estimates offemale Epioblasma.
Mean numberol
Mussel species Numberoffemales (N) Meanlength (mm) Length range (mm) glochidia/female Range
(
Epioblasmaahlstedti 6 45.4 35.8-56.44 18,757* 6,668-38,716
Epioblasmacapsaeformis 10 41.5 36.7-46.4 13,008 7,780-16,876
Epioblasmaflorentinaaureola 7 42.2 40.6-45.7 7,602 3,261-12,558
Epioblasmaflorentina walkeri 6 42.8 40.5-45.2 9,606 1,828-16,921
*Mean fecundityforEpioblasmaahlstedti issignificantlydifferent(p<0.05) fromtheothertaxa.
synchronously in a circular motion; the left micro-lure showsE. ahlstedti asthebasalmemberofthein-grouptaxa,
moves clockwise, and the right micro-lure moves coun- whereas the MP tree (Figure 16) reported in this study
terclockwise (Figure 10). The dorsal margin of the man- shows E. tomlosa rangiana as basal. However, severalinte-
tle-pad is black, forming a discrete uniform band rior nodes in both MP trees are weakly supported and
~2—3 mm wide. The posterior portion ofthe mantle-pad collapse in the respective consensus trees (not shown). A
is not invaginated where it meets the incurrent aperture, 50% majority-rule Bayesian consensus tree depicting a
so the attachment points of the micro-lures can be seen more conservativetopologyis given in Figure 17.
when the female is displaying. The dentieulations along
the margin of the shell expansion of E. capsaeformis Distribution: Epioblasma ahlstedti is currently
are typically finer and more closely spaced than those restricted to 48.3 RKM in the Duck River from Lillard
ofE. ahlstedti. The shell expansion is greenish, but not as Mill (RKM 286.5) downstream to the backwaters ofthe
darkas thatofE. ahlstedti. Old Columbia Dam reservoir (RKM 238.2) in Marshall
Co. and Maury Co., west-central Tennessee (Figure 1).
Life History: The typical habitat of E. ahlstedti is Basedon shells, the species likelyoccurredhistoricallyin
gravel shoals in medium to large rivers. It is a long-term the Buffalo River, TN (Parmalee and Bogan, 1998), a
brooder, gravidfrom late summerto the followingspring tributary to the Duck River, and the Tennessee River at
and early summer. In the Duck River, at least some Muscle Shoals and Shoal Creek, Lauderdale Co., AL
females will emerge from the substrate in early spring (Jones, 2004).
(e.g., Mareh-April) to display their mantle-pad lure to
attract host fishes, while others emerge later in the Conservation Status: The oyster mussel (Epioblasma
spring and summer (e.g., May-July). Mean fecundity of capsaeformis) was listedas endangeredunderthe Endan-
females was 18,757 glochidia, and ranged from 6,668 to gered Species Act (ESA) in 1997, to include the Duck
38,716 (Table4). Thelargestfemaleswere notexamined; Riverpopulation. NowthatE. ahlstedtihasbeendesignated
hence maximum fecundityis unknownbut likelyexceeds a separate species, it is restricted to onlythe Duck River
50,000 glochidia. Known fish hosts for E. ahlstedti population. Beinglinearly distributed to 48 river kilome-
ters and susceptible to a stochastic impact, this species
include three darter species, the greenside darter
(Etheostoma hlennioides), fantail darter Etheostoma should be considered criticallyendangered andcontinue
flabellare), and redline darter (Etheostoma r(ufilineatum) to receive full protection underthe ESA.
(Jones et ah, 2006). However, tested hosts were not col-
lected from the Duck River, but from the North Fork Etymology: The species name is given in honor ol
Holston River, VA, in the upper Tennessee River drain- biologist Steven A. Ahlstedt, U.S. Geological Survey
age. Although allthreedarters arewidelydistributedand (retired), Knoxville, TN, who has dedicated over 30
common in Duck River, the native Duck River hosts years of sendee to freshwater mussel conservation in
remain uncertain. the United States. The common name denotes the
snapping behavior of many female Epioblasma species
Molecular DNA Markers anti Phylogenetic Analysis: (i.e., displaying females wall quickly close their shells
The Duck River population of E. ahlstedti contains pre- when touched, which can capture a host fish to facili-
sumably diagnostic nucleotides at three mitochondrial tate infestation with glochidia on the host) (Jones et al.,
DNA gene regions [16S (n=1), cytochrome-b (n—4), and 2006).
ND1 (n=2)] and at one non-coding nuclear DNA region Epioblasmaflorentina aureola newsubspecies
[ITS-] (n=2)] (Table 2). Rased on analysis of 10 nuclear Golden Riffleshell
DNA
microsatellite loci, the population is moderately Figures 5, 6, 11, 12
diverged (FST=0.12) from the Clinch River population of
E. capsaeformis (Jones et ah, 2006). The MP and Rayesian Diagnosis: The following diagnostic morphological
phylogenetic analyses showed high statistical support lor characteristics ofE. florentina aureola are based on the
the E. ahlstedti clade (Figures 16, 17); a findingpreviously female and are summarized in Table 2: (1) gray mantle-
demonstrated by Jones et al. (2006) using only the DNA pad with a black mottled background, and (2) mantle-
sequences. Furthermore,die MPtreein Jones etal. (2006) pad has rounded pustules.
Page 86 THE NAUTILUS, Vol. 124, No. 2
E. capsaeformis 1 (A/=4)
E. capsaeformis 2 (AM)
E. capsaeformis 5 (A/=1)
E. capsaeformis 6 (AM)
E. capsaeformis 3 (N=2)
E. capsaeformis 4 (AM)
E. fiorentina aureola 1 (N=5)
E. fiorentina aureola 2 (A/=1)
E. fiorentina walkeri 1 (AMO)
E. fiorentina walkeri*
E. ahlstedti 1 (A/=9)
E. ahlstedti2 (AM)
E. torulosa rangiana 1 (AM)
E. torulosa rangiana 2 (A/=3)
E. torulosa rangiana 3 (A/=2)
E. triquetra (AM)
E. brevidens (AM)
Figure 16. Cladogram showing phylogenetic relationships among Epioblasma ahlstedti E. fiorentina aureola and congeners,
inferred from the combined mitochondrial DNA regions of16'S (468 bp), cytochrome-b (36,0 bp), A’/JI (568 bp), the nuclear DNA
region ITS-1 (515 bp) and eight morphological characters (see Table 1) using maximum parsimony (MP) (31 equallyparsimonious
trees were resolved; length=195 steps; 0=0.92; RI=0.89). Numbers abovethebranches (MP) representbootstrapsupport(10,000
replicates); only values >50% are shown. All E.fiorentina walkeri were identical; however, to demonstrate the monophyly of this
population, an additional sequencewas addedtothe analysis. Out-grouptaxaareE. triquetra andE. brevidens.
Description: Length of the female shell can reach height can reach ~10 mm (Table 3; Figure 15). Dentic-
mm
46 in Indian Creekandthe upperClinch River, with ulations occur along the margin of the shell expansion
mean length of40 mm (Table 3). Shell lengths ofmales and are fine and narrowly spaced ~0.5 mm apart. The
are similar. The shell outline of females is sub-oval and shell outline of males is elliptical, appearing pointed at
rounded with a moderately protruded shell expansion the posterior end (Figure 6), and may have a shallow
(Figure 5). The mean height ofthe female shell expan- sulcus, especially older individuals. The periostracum of
sion is 5.7 mm or ~14% of shell length, but maximum adults can range from golden honey-yellow, to tan and