# Endangered and Threatened Wildlife and Plants; Determination of Endangered Status for the Rayed Bean and Snuffbox Mussels Throughout Their Ranges

> Briefs, arguments, decisions, and more.

URL: https://www.frixlaw.com/law-library/documents/fr%3A2012-2940

## Record

- **Collection:** Federal Register
- **Document type:** Rule
- **Published:** February 14, 2012
- **Citation:** 77 FR 8632

## Text

DEPARTMENT OF THE INTERIOR
Fish and Wildlife Service
50 CFR Part 17
[Docket No. FWS-R3-ES-2010-0019; 4500030113]
RIN 1018-AV96
Endangered and Threatened Wildlife and Plants; Determination of Endangered Status for the Rayed Bean and Snuffbox Mussels Throughout Their Ranges

AGENCY:

Fish and Wildlife Service, Interior.

ACTION:

Final rule.

SUMMARY:

We, the U.S. Fish and Wildlife Service (Service), determine endangered status for the rayed bean (
Villosa fabalis
) and snuffbox (
Epioblasma triquetra
) mussels throughout their ranges, under the Endangered Species Act of 1973, as amended (Act).

DATES:

This rule becomes effective on March 15, 2012.

ADDRESSES:

This final rule is available on the Internet at
http://www.regulations.gov
at Docket Number FWS-R3-ES-2010-0019. Comments and materials received, as well as supporting documentation used in preparing this final rule are available for public inspection, by appointment, during normal business hours, at the U.S. Fish and Wildlife Service, Columbus Ecological Services Field Office, 4625 Morse Road, Suite 104, Columbus, OH 43230; phone 614-416-8993; facsimile 614-416-8994.

FOR FURTHER INFORMATION CONTACT:

Angela Boyer, Endangered Species Coordinator, Columbus Ecological Services Field Office (see
ADDRESSES
). If you use a telecommunications devise for the deaf (TDD), call the Federal Information Relay Service (FIRS) at 800-877-8339.

SUPPLEMENTARY INFORMATION:

Background

This document is a final rule to list as endangered the rayed bean (
Villosa fabalis
) and snuffbox (
Epioblasma triquetra
).

Previous Federal Action

Federal actions for these species prior to November 2, 2010, are outlined in our proposed rule for these actions (75 FR 67552). Publication of the proposed rule opened a 60-day comment period, which ended on January 3, 2011.

Species Information

Rayed Bean

The rayed bean is a small mussel, usually less than 1.5 inches (in) (3.8 centimeters (cm)) in length (Cummings and Mayer 1992, p. 142; Parmalee and Bogan 1998, p. 244; West
et al.
2000, p. 248). The shell outline is elongate or ovate in males and elliptical in females, and moderately inflated in both sexes, but more so in females (Parmalee and Bogan 1998, p. 244). The valves are thick and solid. The anterior end is rounded in females and bluntly pointed in males (Cummings and Mayer 1992, p. 142). Females are generally smaller than males (Parmalee and Bogan 1998, p. 244). Dorsally, the shell margin is straight, while the ventral margin is straight to slightly curved (Cummings and Mayer 1992, p. 142). The beaks are slightly elevated above the hingeline (West
et al.
2000, p. 248), with sculpture consisting of double loops with some nodules (Parmalee and Bogan 1998, p. 244). No posterior ridge is evident. Surface texture is smooth and sub-shiny, and green, yellowish-green, or brown in color, with numerous, wavy, dark-green rays of various widths (sometimes obscure in older, blackened specimens) (Cummings and Mayer 1992, p. 142; West
et al.
2000, p. 248). Internally, the left valve has two pseudocardinal teeth (tooth-like structures along the hingeline of the internal portion of the shell) that are triangular, relatively heavy, and large, and two short, heavy lateral teeth (Cummings and Mayer 1992, p. 142). The right valve has a low, triangular pseudocardinal tooth, with possibly smaller secondary teeth anteriorly and posteriorly, and a short, heavy, and somewhat elevated lateral tooth (Parmalee and Bogan 1998, p. 244). The color of the nacre (mother-of-pearl) is silvery white or bluish and iridescent posteriorly. Key characters useful for distinguishing the rayed bean from other mussels are its small size, thick valves, unusually heavy teeth for a small mussel, and color pattern (Cummings and Mayer 1992, p. 142).

Snuffbox

The snuffbox is a small- to medium-sized mussel, with males reaching up to 2.8 in (7.0 cm) in length (Cummings and Mayer 1992, p. 162; Parmalee and Bogan 1998, p. 108). The maximum length of females is about 1.8 in (4.5 cm) (Parmalee and Bogan 1998, p. 108). The shape of the shell is somewhat triangular (females), oblong, or ovate (males), with the valves solid, thick, and very inflated. The beaks are located somewhat anterior of the middle, and are swollen, turned forward and inward, and extended above the hingeline (Cummings and Mayer 1992, p. 162). Beak sculpture consists of three or four faint, double-looped bars (Cummings and Mayer 1992, p. 162; Parmalee and Bogan 1998, p. 108). The anterior end of the shell is rounded, and the posterior end is truncated, highly so in females. The posterior ridge is prominent, being high and rounded, while the posterior slope is widely flattened. The posterior ridge and slope in females is covered with fine ridges and grooves, and the posterioventral shell edge is finely toothed (Cummings and Mayer 1992, p. 162). When females are viewed from a dorsal or ventral perspective, the convergence of the two valves on the posterior slope is nearly straight due to being highly inflated. This gives the female snuffbox a unique, broadly lanceolate or cordate perspective when viewed at the substrate and water column interface (Ortmann 1919, p. 329; van der Schalie 1932, p. 104). The ventral margin is slightly rounded in males and nearly straight in females. Females have recurved denticles (downward curved tooth-like structures) on the posterior shell margin that aid in holding host fish (Barnhart 2008, p. 1). The periostracum (external shell surface) is generally smooth and yellowish or yellowish-green in young individuals, becoming darker with age. Green, squarish, triangular, or chevron-shaped marks cover the umbone (the inflated area of the shell along the dorsal margin), but become poorly delineated stripes with age. Internally, the left valve has two high, thin, triangular, emarginate pseudocardinal teeth (the front tooth being thinner than the back tooth) and two short, strong, slightly curved, and finely striated lateral teeth. The right valve has a high, triangular pseudocardinal tooth with a single short, erect, and heavy lateral tooth. The interdentum (a flattened area between the pseudocardinal and lateral teeth) is absent, and the beak cavity is wide and deep. The color of the nacre is white, often with a silvery luster, and a gray-blue or gray-green tinge in the beak cavity. The soft anatomy was described by Oesch (1984, pp. 233-234) and Williams
et al.
(2008, p. 282). Key characters useful for distinguishing the snuffbox from other species include its unique color pattern, shape (especially in females), and high degree of inflation.

Taxonomy

The rayed bean is a member of the freshwater mussel family Unionidae and was originally described as
Unio fabalis
by Lea in 1831 (pp. 86-87). The type locality (the location of the first

identified specimen) is the Ohio River (Parmalee and Bogan 1998, p. 244), probably in the vicinity of Cincinnati, Ohio. Over the years, the rayed bean has been placed in the genera
Unio, Margarita, Margaron, Eurynia, Micromya,
and
Lemiox.
It was ultimately placed in the genus
Villosa
by Stein (1963, p. 19), where it remains today (Turgeon
et al.
1998, p. 33). We recognize
Unio capillus, U. lapillus,
and
U. donacopsis
as synonyms of
Villosa fabalis
.

The snuffbox is a member of the freshwater mussel family Unionidae and was described as
Truncilla triqueter
(Rafinesque 1820, p. 300). The species name was later changed to
triquetra
(Simpson 1900, p. 517), from the Latin
triquetrous
meaning “having three acute angles,” a reference to the general shape of the female. The type locality is the Falls of the Ohio (Ohio River, Louisville, Kentucky) (Parmalee and Bogan 1998, p. 108). The synonymy (scientific names used for the species) of the snuffbox was summarized by Johnson (1978, pp. 248-249), Parmalee and Bogan (1998, p. 108), and Roe (2004, p. 3). This species has also been considered a member of the genera
Unio, Dysnomia, Plagiola, Mya, Margarita, Margaron,
and
Epioblasma
at various times since its description. The monotypic subgenus
Truncillopsis
was created for this species (Ortmann and Walker 1922, p. 65). The genus
Epioblasma
was not in common usage until the 1970s (Stansbery 1973, p. 22; Stansbery 1976, p. 48;
contra
Johnson 1978, p. 248), where it currently remains (Turgeon
et al.
1998, p. 34).
Unio triqueter, U. triangularis,

U. triangularis longisculus, U. triangularis pergibosus, U. cuneatus,
and
U. formosus
are recognized as synonyms of
E. triquetra
. Tricorn pearly mussel is another common name for this species (Clarke 1981a, p. 354).

Life History

The general biology of the rayed bean and the snuffbox is similar to other bivalved mollusks belonging to the family Unionidae. Adults are suspension-feeders, spending their entire lives partially or completely buried within the substrate (Murray and Leonard 1962, p. 27). Adults feed on algae, bacteria, detritus, microscopic animals, and dissolved organic material (Silverman
et al.
1997, p. 1859; Nichols and Garling 2000, p. 873; Christian
et al.
2004, pp. 108-109; Strayer
et al.
2004, pp. 430-431). Recent evidence suggests that adult mussels may also deposit-feed on particles in the sediment (Raikow and Hamilton 2001, p. 520). For their first several months, juvenile mussels employ foot (pedal) feeding, consuming settled algae and detritus (Yeager
et al.
1994, p. 221). Unionids have an unusual mode of reproduction. Their life cycle includes a brief, obligatory parasitic stage on fish. Eggs develop into microscopic larvae called glochidia within special gill chambers of the female mussel. The female expels the mature glochidia, which must attach to the gills or the fins of an appropriate fish host to complete development. Host fish specificity varies among unionids. Some species appear to use a single host, while others can transform on several host species. Following successful infestation, glochidia encyst (enclose in a cyst-like structure) and drop off as newly transformed juveniles. For further information on freshwater mussels, see Gordon and Layzer (1989, pp. 1-17).

Mussel biologists know relatively little about the specific life-history requirements of the rayed bean and the snuffbox. Most mussels, including the rayed bean and snuffbox, have separate sexes. The age at sexual maturity, which is unknown for the rayed bean and snuffbox, is highly variable (0-9 years) among and within species (Haag and Staton 2003, pp. 2122-2123), and may be sex-dependent (Smith 1979, p. 382). Both species are thought to be long-term brooders; rayed bean females brood glochidia from May through October (Parmalee and Bogan 1998, p. 108; Ecological Specialists, Inc. (ESI) 2000, p. 5; Woolnough 2002, p. 23), and snuffbox brood glochidia from September to May (Ortmann 1912, p. 355; 1919, p. 327). Tippecanoe darter
(Etheostoma tippecanoe)
is the only verified host fish for the rayed bean (White
et al.
1996, p. 191). Other rayed bean hosts are thought to include the greenside darter
(E. blennioides),
rainbow darter
(E. caeruleum),
mottled sculpin
(Cottus bairdi),
and largemouth bass
(Micropterus salmoides)
(Woolnough 2002, p. 51). Based on inference of closely related species, additional hosts may be suitable, including other darter and sculpin species (Jones 2002, pers. comm.). Juvenile snuffbox have successfully transformed on logperch
(Percina caprodes),
blackside darter
(P. maculata),
rainbow darter, Iowa darter
(E. exile),
blackspotted topminnow
(Fundulus olivaceous),
mottled sculpin, banded sculpin
(C. carolinae),
Ozark sculpin
(C. hypselurus),
largemouth bass, and brook stickleback
(Culaea inconstans)
in laboratory tests (Sherman 1994, p. 17; Yeager and Saylor 1995, p. 3; Hillegass and Hove 1997, p. 25; Barnhart
et al.
1998, p. 34; Hove
et al.
2000, p. 30; Sherman Mulcrone 2004, pp. 100-103).

Habitat Characteristics

The rayed bean is generally known from smaller, headwater creeks, but occurrence records exist from larger rivers (Cummings and Mayer 1992, p. 142; Parmalee and Bogan 1998, p. 244). They are usually found in or near shoal or riffle (short, shallow length of stream where the stream flows more rapidly) areas, and in the shallow, wave-washed areas of glacial lakes, including Lake Erie (West
et al.
2000, p. 253). In Lake Erie, the species is generally associated with islands in the western portion of the lake. Preferred substrates typically include gravel and sand. The rayed bean is oftentimes found among vegetation (water willow (
Justicia americana
) and water milfoil (
Myriophyllum sp.
)) in and adjacent to riffles and shoals (Watters 1988b, p. 15; West
et al.
2000, p. 253). Specimens are typically buried among the roots of the vegetation (Parmalee and Bogan 1998, p. 245). Adults and juveniles appear to produce byssal threads (thin, protein-based fibers) (Woolnough 2002, pp. 99-100), apparently to attach themselves to substrate particles.

The snuffbox is found in small- to medium-sized creeks, to larger rivers, and in lakes (Cummings and Mayer 1992, p. 162; Parmalee and Bogan 1998, p. 108). The species occurs in swift currents of riffles and shoals and wave-washed shores of lakes over gravel and sand with occasional cobble and boulders. Individuals generally burrow deep into the substrate, except when spawning or attempting to attract a host (Parmalee and Bogan 1998, p. 108).

Strayer (1999a, pp. 471-472) demonstrated in field trials that mussels in streams occur chiefly in flow refuges, or relatively stable areas that display little movement of particles during flood events. Flow refuges conceivably allow relatively immobile mussels to remain in the same general location throughout their entire lives. Strayer thought that features commonly used in the past to explain the spatial patchiness of mussels (water depth, current speed, sediment grain size) were poor predictors of where mussels actually occur in streams.

Rayed Bean Historical Distribution

The rayed bean historically occurred in 115 streams, lakes, and some human-made canals in 10 States: Illinois, Indiana, Kentucky, Michigan, New York, Ohio, Pennsylvania, Tennessee, Virginia, and West Virginia; and Ontario, Canada. The mussel occurred in parts of the upper (Lake Michigan

drainage) and lower Great Lakes systems, and throughout most of the Ohio and Tennessee River systems. During historical times, the rayed bean was fairly widespread and locally common in many Ohio River system streams based on collections made over a several-decade period. The species was once fairly common in the Belle, South Branch Thames, Detroit, Scioto, Wabash, and Duck Rivers; several tributaries in the Scioto system (Olentangy River, and Big Darby and Alum Creeks); and Tippecanoe Lake, based on literature and museum records (Call 1900; Watters 1994, p. 105; West
et al.
2000, p. 251; Badra 2002, pers. comm.). The rayed bean was last reported from some streams several decades ago (North Branch Clinton, Auglaize, Ohio, West Fork, Beaver, Shenango, Mahoning, Mohican, Scioto, Green, Barren, Salamonie, White, Big Blue, Tennessee, Holston, South Fork Holston, Nolichucky, Clinch, North Fork Clinch, and Powell Rivers; Wolf, Conewango, Oil, Crooked, Pymatuning, Mill, Alum, Whetstone, Deer, Lick, and Richland Creeks; and Buckeye, Tippecanoe, Winona, and Pike Lakes). The rayed bean population in Lake Erie was once considerable (Ohio State University Museum of Biological Diversity (OSUM) collections), but has been eliminated by the zebra mussel.

Rayed Bean Current Distribution

Extant populations of the rayed bean are known from 31 streams and 1 lake in seven States and 1 Canadian province: Indiana (St. Joseph River (Fish Creek), Tippecanoe River (Lake Maxinkuckee, Sugar Creek)), Michigan (Black River (Mill Creek), Pine River, Belle River, Clinton River), New York (Allegheny River (Olean Creek, Cassadaga Creek)), Ohio (Swan Creek, Fish Creek, Blanchard River, Tymochtee Creek, Walhonding River, Mill Creek, Big Darby Creek, Scioto Brush Creek; Great Miami River, Little Miami River (East Fork Little Miami River), Stillwater River), Pennsylvania (Allegheny River (French Creek (Le Boeuf Creek, Muddy Creek, Cussewago Creek))), Tennessee (Duck River), and West Virginia (Elk River); and Ontario, Canada (Sydenham River, Thames River).

Rayed Bean Population Estimates and Status

Based on historical and current data, the rayed bean has declined significantly rangewide and is now known only from 31 streams and 1 lake (down from 115), a 73 percent decline (Table 1). This species has also been eliminated from long reaches of former habitat in hundreds of miles of the Maumee, Ohio, Wabash, and Tennessee Rivers and from numerous stream reaches and their tributaries. In addition, this species is no longer known from the States of Illinois, Kentucky, and Virginia. The rayed bean was also extirpated from West Virginia, until the 2006 reintroduction into the Elk River, and from Tennessee, until the 2008 reintroduction into the Duck River (Clayton 2007, pers. comm.; Urban 2010, pers. comm.; Moles and Layzer 2009, p. 2).

In this rule, mussel shell collection records have been classified according to the condition of shell material. Fresh dead shells still have flesh attached to the valves, they may or may not retain a luster to their nacre, and their periostracum is non-peeling, all indicating relatively recent death (generally less than 1 year) (Buchanan 1980, p. 4). Relic shells have lost the luster to their nacre, have peeling or absent periostracum, may be brittle or worn, and likely have been dead more than a year (Buchanan 1980, pp. 4-5; Zanatta
et al.
2002, p. 482). Generally, fresh dead shells indicate the continued presence of the species at a site (Metcalf 1980, p. 4). The presence of relic shells only, along with repeated failure to find live animals or fresh dead shells, likely signifies that a population is extirpated (Watters and Dunn 1993-94, pp. 253-254). Shells labeled R may originally have been reported by collectors as either weathered dead (or weathered dry) or subfossil. If no details on shell condition were provided for a record, the shell is simply referred to as dead. In this document, a population is considered viable if evidence of successful reproduction is documented and it has enough individuals to sustain the population at its current level for the foreseeable future.

Table 1—Rayed Bean Extant Stream Population Summary by Stream of Occurrence

Stream (state)

Last
observed

Recruiting

Potential
viability

Population size
Population trend

Black River (MI)
2001
Unknown
Low
Small
Unknown.

Mill Creek (MI)
2002
Unknown
Unknown
Small
Unknown.

Pine River (MI)
2002
Yes
High
Small
Declining.

Belle River (MI)
2010
Unknown
Unknown
Small
Unknown.

Clinton River (MI)
2009
Yes
Unknown
Small
Unknown.

Sydenham River (ON)
2010
Yes
High
Large
Stable.

Thames River (ON)
2008
Unknown
High
Large
Unknown.

Swan Creek (OH)
2010
Yes
High
Large
Stable.

St. Joseph River (IN)
1998
Unknown
Low
Small
Declining.

Fish Creek (IN, OH)
2009
Unknown
Low
Small
Declining.

Blanchard River (OH)
2010
Yes
High
Large
Unknown.

Tymochtee Creek (OH)
1996
Unknown
Unknown
Small
Unknown.

Allegheny River (PA, NY)
2010
Yes
High
Large
Stable.

Olean Creek (NY)
2000
Yes
High
Small
Unknown.

Cassadaga Creek (NY)
1994
Yes
Low
Small
Unknown.

French Creek (PA)
2005
Yes
High
Large
Stable.

Le Boeuf Creek (PA)
2006
Unknown
Unknown
Unknown
Unknown.

Muddy Creek (PA)
2006
Unknown
Unknown
Unknown
Unknown.

Cussewago Creek (PA)
1991
Unknown
Unknown
Small
Unknown.

Walhonding River (OH)
1991-95
Unknown
Low
Small
Declining.

Elk River (WV)
2010
Unknown (Reintroduced in 2006)
Unknown
Small
Unknown.

Mill Creek (OH)
2011
Unknown
Unknown
Unknown
Unknown.

Big Darby Creek (OH)
2008
Unknown
Unknown
Small
Declining.

Scioto Brush Creek (OH)
1987
Unknown
Unknown
Small
Unknown.

Great Miami River (OH)
2010
Unknown
Unknown
Unknown
Unknown.

Little Miami River (OH)
1990-91
Unknown
Unknown
Unknown
Unknown.

East Fork Little Miami River (OH)
1990-91
Unknown
Unknown
Unknown
Unknown.

Stillwater River (OH)
1987
Unknown
Unknown
Unknown
Unknown.

Tippecanoe River (IN)
1995
Unknown
Low
Unknown
Declining.

Lake Maxinkuckee (IN)
1997
Unknown
Unknown
Unknown
Declining.

Sugar Creek (IN)
1998
Unknown
Unknown
Unknown
Unknown.

Duck River (TN)
2008
Unknown (Reintroduced in 2008)
Unknown
Small
Unknown.

Upper Great Lakes Sub-Basin

The rayed bean was not known from the upper Great Lakes sub-basin until 1996, when relic specimens were documented from the Pigeon River, a tributary to the St. Joseph River that flows into Lake Michigan. No extant populations of the rayed bean are currently known from this system.

Lower Great Lakes Sub-Basin

Of the 115 water bodies from which the rayed bean was historically recorded, 27 are in the lower Great Lakes system. The species is thought to be extant in 12 streams, which are discussed below, but historically significant populations have been eliminated from Lake Erie and the Detroit River.

Black River
—A tributary of the St. Clair River, linking Lakes Huron and St. Clair, the Black River is located in southeastern Michigan. Hoeh and Trdan (1985, p. 115) surveyed 17 sites in the Black River system, including 12 mainstem sites over approximately 47 miles (75 km), but failed to find the rayed bean. The rayed bean was not discovered there until the summer of 2001, when a single live individual was found in the lower river in the Port Huron State Game Area (PHSGA) (Badra 2002, pers. comm.). A survey in 2003 failed to find any rayed bean, and two surveys in 2005 found only two valves (Badra 2008, pers. comm.). An additional survey was performed in 2005 at six sites, but no rayed bean were found (Badra 2008, pers. comm.). The status of this population cannot be accurately assessed at this time, but would appear to be small and of questionable viability (Butler 2002, p. 8).

Mill Creek
—Mill Creek is a tributary of the Black River, St. Clair County, in southeastern Michigan. The rayed bean was discovered in Mill Creek in August 2002. Five dead specimens were found approximately 0.5 miles (mi) (0.8 kilometers (km)) above its confluence with the Black River in the PHSGA (Badra 2002, pers. comm.). A Mill Creek site 0.25 mi (0.4 km) from the confluence of the Black River was surveyed in 2003 and 2004, with one rayed bean shell found during each survey (Badra 2008, pers. comm.). Similar to the population in the Black River, the status of this newly discovered population cannot be accurately assessed at this time.

Pine River
—Another tributary of the St. Clair River, the Pine River is located in southeastern Michigan. The rayed bean was apparently not collected in the Pine River until 1982, when specimens were found at three sites (Hoeh and Trdan 1985, p. 116). These collections included 5 live individuals and 23 fresh dead specimens (Badra 2002, pers. comm.). Hoeh and Trdan (1985, p. 116) considered it to be “rare,” semi-quantitatively defined as occurring at a rate of less than one specimen per person-hour sampling effort. In 1997, two live individuals were found. The last survey in the Pine River occurred in 2002 (Badra 2008, pers. comm.), and one live rayed bean was documented (Badra and Goforth 2003, p. 6). Comparing the historical and most recent survey resulting, it appears that the species may have declined significantly since the 1980s, but it is probably still viable in the Pine River.

Belle River
—The Belle River is a third tributary of the St. Clair River harboring an extant population of the rayed bean. This species was first collected from the Belle River in 1965, when 17 fresh dead specimens were collected (OSUM 1965:0106). The same site was revisited in 1978, but only one fresh dead shell is represented in OSUM 1978:0013. Since that time, live individuals or fresh dead specimens were found in 1983 and 1992, while only relic shells were found in 1994 (Badra 2008, pers. comm.). During summer 2002 sampling, single live specimens were found at two new sites in the Belle River, with four and two fresh dead specimens, respectively, also found at these sites (Badra 2008, pers. comm.). In 2010, five live individuals were found at the same two sites sampled in 2002 (Zanatta 2011, pers. comm.). These two sites are about 2 miles (3.2 km) apart in the lower portion of the river. The status of the Belle River population is still not well known, but appears to be small and restricted to a short reach in the lower river.

Clinton River
—The rayed bean was first recorded from the Clinton River in 1933 (Badra 2008, pers. comm.). The mussel fauna in the entire mainstem of the Clinton River downstream of Pontiac, Michigan, was apparently wiped out by pollution between 1933 and 1977 (Strayer 1980, p. 147). In 1992, Trdan and Hoeh (1993, p. 102) found 26 live individuals using a suction dredge from a bridge site slated for widening, where Strayer (1980, p. 146) previously found only relic shells. The rayed bean represented 1.2 percent relative abundance of the 10 species collected at the site (Trdan and Hoeh 1993, p. 102). The population in the Clinton River is probably viable but currently restricted to about 3 mi (4.8 km) of stream in the western suburbs of Pontiac (Butler 2002, p. 9). Zanatta (2011, pers. comm.) found one live rayed bean in 2009. The rayed bean's long-term viability appears to be precarious in the Clinton River.

Sydenham River
—The rayed bean in the Sydenham River represents one of the largest rayed bean populations remaining. West
et al.
(2000, pp. 252-253) presented a highly detailed collection history of the rayed bean in the Sydenham River. The rayed bean is currently thought to exist in an approximately 75-mi (120-km) reach of the middle Sydenham, from the general vicinity of Napier, Ontario, downstream to Dawn Mills. The species appears to be most abundant in the lower half of this river reach. Although the range has remained relatively consistent over time, abundance data at repeatedly sampled sites from the 1960s to the late 1990s indicate a general decline of the rayed bean. Based on the range of sizes and roughly equal number of specimens in various size classes of the live and fresh dead material they gathered, West
et al.
(2000, p. 256) considered the population to be “healthy” and “reproducing” (recruiting). Data from sampling in 2001 show evidence of

recruitment and variable size classes for both sexes from most of the sites (Woolnough 2002, p. 50). Based on this data, the rayed bean population in the Sydenham River is doing considerably better than West
et al.
(2000, pp. 252-253) suggested. Woolnough and Morris (2009, p. 19) estimate that there are 1.5 million mature rayed bean in the Sydenham River living in the 38-mile (61-km) stretch between Napier Road near Alvinston, Ontario, and Dawn Mills, Ontario.

Thames River
—The Thames River flows west through southwestern Ontario. The rayed bean was historically known from only the south branch until 2008, when it was discovered in the north branch. In July 2008, six gravid (gills full of glochidia) females were collected at two north branch sites (Woolnough 2008, pers. comm.). In September 2008, four live females and two live males were collected at two different north branch sites (Woolnough 2008, pers. comm.). All of these individuals were collected within a 4.5-mi (7.2-km) reach of the river (Woolnough 2008, pers. comm.). Woolnough and Morris (2009, p. 19) estimate that there are 4,300 mature rayed bean in the Thames River.

Maumee River System
—The Maumee River system, which flows into the western end of Lake Erie, was once a major center of distribution of the rayed bean. The species was historically known from eight streams in the system in addition to the mainstem Maumee. Further, an additional population was discovered in the system in 2005 in Swan Creek.

Swan Creek
—Swan Creek is a tributary of the lower Maumee River in northwestern Ohio. This population was discovered in 2005. Surveys conducted in 2006 and 2007 found that the Swan Creek population is limited to about 3 river mi (5 river km) between river mile (RM) 18.3 and 15.3 (Grabarkiewicz 2008, p. 11). The rayed bean was the fourth most abundant unionid present within the 2006-2008 sample area, reaching densities of eight individuals per square meter in some areas and comprising about 14.1 percent of the total mussel community (Grabarkiewicz 2008, p. 10). The rayed bean population in Swan Creek is viable and, although limited to a short reach, may be one of the most robust remaining populations.

St. Joseph River
—The St. Joseph River is one of the two major headwater tributaries to the Maumee, with a drainage area in southeastern Michigan, northwestern Ohio, and northeastern Indiana. The mainstem flows in a southwesterly direction to its confluence with the St. Mary's River to form the Maumee in Ft. Wayne, Indiana. The rayed bean was historically known from numerous sites on the river, but now apparently persists only at a couple of sites in the lower St. Joseph River in Allen and DeKalb Counties, Indiana (Watters 1988b, p. 15; 1998, Appendix C); a few fresh dead specimens were found in both studies, but no live individuals were found. Grabarkiewicz and Crail (2008, p. 13) surveyed six sites on the West Branch St. Joseph River in 2007, but did not encounter any rayed bean.

Fish Creek
—A tributary of the St. Joseph River that begins in Ohio, Fish Creek flows west, then south through Indiana, then eventually east into Ohio before joining the St. Joseph River at Edgerton. The rayed bean persists in Williams County, Ohio, and possibly DeKalb County, Indiana. Based on the appearance of 2 live individuals and fresh dead shells, it inhabits the lower 10 mi (16.1 km) or less of the stream (Watters 1988b, p. 18; Grabarkiewicz 2009, pers. comm.). Watters (1988b, p. ii) considered Fish Creek to be “the most pristine tributary of the St. Joseph system.” A major diesel fuel spill from a ruptured pipeline in DeKalb County in 1993 resulted in a mussel kill in the lower portion of the stream (Sparks
et al.
1999, p. 12). It is not known if the rayed bean was affected by the spill. Surveys in 2004 (at 64 qualitative sites) and 2005 (at 11 quantitative sites) failed to detect the species (Brady
et al.
2004, p. 2; 2005, p. 3). However, Grabarkiewicz (2009, pers. comm.) reported finding two live and three fresh dead rayed bean in 2005, at the County Road 3 bridge in Ohio. In 2009, two fresh dead rayed bean were found in lower Fish Creek in Ohio (Boyer 2009, pers. obs.). The viability and status of this population are uncertain (Fisher 2008, pers. comm.).

Blanchard River
—The Blanchard River is a tributary of the Auglaize River in the Maumee River system, in northwestern Ohio. First discovered in 1946, this population is one of the largest of the rayed bean rangewide. The rayed bean in the Blanchard River is restricted to 25-30 river mi (40-48 river km) in the upper portion of the stream in Hardin and Hancock Counties upstream of Findley (Hoggarth
et al.
2000, p. 22). Hoggarth
et al.
(2000, p. 23) reported the rayed bean to be the fourth most common species in the drainage. Grabarkiewicz (2010, pers. comm.) found live individuals, including a juvenile, at six sites sampled in 2010. The population is considered to be viable.

Tymochtee Creek
—Tymochtee Creek is a tributary to the upper Sandusky River in north-central Ohio, which flows into the southwestern portion of Lake Erie. The rayed bean is known from three sites in a reach of stream in Wyandot County and was first collected in 1970. All collections of the rayed bean have been small, with not more than five fresh dead shells found in any one collection effort. The last record is for 1996, when a pair and three unpaired valves were collected. The condition of at least one of the valves indicated that the rayed bean is probably still extant in the stream, although no live individuals were observed (Athearn 2002, pers. comm.). The rayed bean status in Tymochtee Creek is, therefore, currently unknown.

Ohio River System

The rayed bean was historically known from the Ohio River in the vicinity of Cincinnati, Ohio, downstream to the Illinois portion of the river. It undoubtedly occurred elsewhere in the upper mainstem. Few historical records are known (mostly circa 1900), and no recent collections have been made, indicating that it became extirpated there decades ago. It was historically known from 74 streams, canals, and lakes in the system, representing roughly two-thirds of its total range. Ortmann (1925, p. 354) considered the rayed bean to be “abundant in small streams” in the Ohio River system. Currently, only 18 streams and a lake are thought to have extant rayed bean populations in the system.

Allegheny River System
—Nine streams and Chautauqua Lake historically harbored rayed bean populations in the Allegheny River system. Currently, the rayed bean is found in half of these water bodies, but in good numbers in two streams (Allegheny River and French Creek) in this drainage.

Allegheny River
—The Allegheny River drains northwestern Pennsylvania and western New York, joining the Monongahela River at Pittsburgh, Pennsylvania, to form the Ohio River. Ortmann (1909a, p. 179; 1919, p. 262) was the first to report the rayed bean from the Allegheny. The population once stretched from Cataraugus County, New York, to Armstrong County, Pennsylvania. Based on historical collections, it appears that the rayed bean is more abundant now than it was historically in the Allegheny River. This may indicate that the rayed bean population in the Allegheny has expanded in the past 100 years. Many streams in western Pennsylvania have improved water quality since Ortmann's time, when he reported on the

wholesale destruction of mussels in several streams (Ortmann 1909b, pp. 11-12). The species currently occurs in Pennsylvania downstream of Allegheny (Kinzua) Reservoir in Warren County to the pool of Lock and Dam 6 in northern Armstrong County, a distance of over 100 river mi (161 river km) (Villella Bumgardner 2008, pers. comm.). The Allegheny population is viable and one of the most important remaining rangewide today.

Olean Creek
—Olean Creek is a tributary of the Allegheny River in western New York. A small population of the rayed bean is known from the lower portions of the stream. Strayer
et al.
(1991, p. 67) reported the rayed bean from three sites during 1987-90 sampling, although just one live individual was located with relic shells from the other two sites. Only relic shells were found in Olean Creek in 1994, but three live individuals were found in 2000, at the proposed construction site of the City of Olean Water Treatment Plant (ESI 2000, p. 8). Collected only during their quantitative sampling effort, the rayed bean represented a relative abundance of 11.5 percent of the seven live species sampled. The rayed bean age distribution of these specimens also indicates recent recruitment into the population (ESI 2000, p. 9). Relic specimens are now known from an 8-mi (13-km) reach of stream, with live individuals known from less than 1.5 mi (2.4 km) of the lower creek. The Olean Creek population appears viable, but is small and tenuous (Butler 2008, pers. comm.)

Cassadaga Creek
—Cassadaga Creek is a tributary of Conewango Creek in the Allegheny River system, in western New York. A small population of the rayed bean is known from a single riffle (Ross Mills) in the lower creek north of Jamestown. Four live specimens were found in 1994 (Strayer 1995). Muskrat middens (a pile of shells) collected during the winter of 2002 produced 38 fresh dead specimens with a size range of 0.8-1.7 in (2.0-4.3 cm) (Clapsadl 2002, pers. comm.). Although the rayed bean is not known from other sites in the stream, it appears to be viable at this site.

French Creek
—French Creek is a major tributary of the middle Allegheny River, in western New York and northwestern Pennsylvania. One of the largest rayed bean populations known, it is found in much of the lower portions of the stream in four Pennsylvania counties (the species is not known from the New York portion of stream). Ortmann (1909a, p. 188; 1919, p. 264) reported the species from two counties, Crawford and Venango. Not until circa 1970 did the population become more thoroughly known, with museum lot sizes indicating sizable populations at several sites, particularly in the lower reaches of the stream. Recent collections indicate that population levels remain high with the rayed bean occurring throughout the mainstem (Villella Bumgardner 2002, pers. comm.; Smith and Crabtree 2005, pp. 15-17; Enviroscience 2006, p. 5).

Le Boeuf Creek
—Le Boeuf Creek is a small western tributary of upper French Creek, flowing in a southerly direction just west of West Branch French Creek in Erie County, Pennsylvania. A total of five live individuals were collected at two out of five sites during a 2006 survey (Smith
et al.
2009, pp. 68-76; Welte 2011, pers. comm.). No other information is available on the status of this population.

Muddy Creek
—Muddy Creek is an eastern tributary of upper French Creek in Crawford County, Pennsylvania. The rayed bean was not discovered until the summer of 2006. Live and fresh dead rayed bean were reported from 2 of 20 lower river sites (Mohler
et al.
2006, pp. 581-582). No live juveniles were found during the 2006 survey (Mohler
et al.
2006, p. 576). No other information is available on the status of this population.

Cussewago Creek
—Cussewago Creek is a tributary of lower French Creek, with its confluence at Meadville, Crawford County, Pennsylvania. A small population was reported in 1991 from Cussewago Creek (Proch 2001, pers. comm.). The rayed bean is thought to persist in the stream, but its current status is unknown.

Walhonding River
—The Walhonding River is a tributary of the upper Muskingum River system, in central Ohio, forming the latter river at its confluence with the Tuscarawas River at Coschocton. Small numbers of rayed bean shells are represented in OSUM collections from the 1960s and 1970s. During 1991-93, Hoggarth (1995-96, p. 161) discovered one live individual and one fresh dead specimen at one site, while four relic specimens were found at three other sites. A small rayed bean population is thought to remain in the Walhonding River; its status is unknown, but is deemed highly tenuous, given the small population size. The population is probably nearing extirpation (Hoggarth 2008a, pers. comm.).

Elk River
—The Elk River is a major 181-river-mi (291-river-km) tributary in the lower Kanawha River system draining central West Virginia and flowing west to the Kanawaha River at Charleston. The rayed bean was extirpated in the Elk River sometime in the 1990s. In 2006 and 2007, approximately 600 adults were reintroduced into the Elk River above Clendenin. In 2008, an effort was made to monitor the reintroduction. A 30-minute search yielded two live individuals, but efforts were discontinued due to high water and excessive habitat disturbance caused by the search effort (Clayton 2008, pers. comm.). In 2010, none of the individuals released in 2006 was found, but an additional 200 individuals were released (Clayton 2010, pers. comm.). The translocated adults are thought to persist in the stream, but it is unknown if this new population is reproducing.

Scioto River system
—The Scioto River system, in central and south-central Ohio, is a major northern tributary of the Ohio River. A historically large metapopulation of the rayed bean occupied at least 11 streams, the Ohio and Erie Canal, and Buckeye Lake. Sizable populations were noted in at least the Olentangy River, and Alum and Big Darby Creeks, based on OSUM collections primarily from the 1960s. A series of system reservoirs, mostly north of Columbus, reduced habitat and contributed to the elimination of some populations in several streams (Alum, Big Walnut, and Deer Creeks; Olentangy and Scioto Rivers). The location of the Columbus Metropolitan Area in the heart of the watershed has also taken a major toll on the species. The historical Scioto rayed bean metapopulation has since been decimated by anthropogenic factors. Currently, remnant populations are known only from Mill Creek, Big Darby Creek, and Scioto Brush Creek.

Mill Creek
—Mill Creek is a tributary of the Scioto River in central Ohio that joins the Scioto River at the O'Shaughnessy Reservoir northwest of the City of Columbus. In 2004, seven fresh dead specimens were found during a survey in the City of Marysville (Hoggarth 2005, p. 7). In 2007, Hoggarth (2007a, pp. 5-6) found two live rayed bean at the same site and one live individual at an additional site. No other information is available on the status of this population.

Big Darby Creek
—Big Darby Creek is one of the major tributaries draining the northwestern portion of the Scioto River system in central Ohio. A sizable rayed bean population was noted in Big Darby Creek from OSUM collections, primarily from the 1960s. Watters (1994, p. 105) reported finding a few fresh dead specimens in 1986, but none in 1990, and indicated that the rayed bean was probably extirpated from Big Darby Creek. In 2006, one live individual was

found at the U.S. Highway 42 bridge replacement project site (Hoggarth 2006, p. 6). This individual was relocated to a site upstream out of the impact zone of the bridge project, and nine additional live individuals were subsequently found at the relocation site (Hoggarth 2006, p. 6). In 2007, three live rayed bean were found at the relocation site (Hoggarth 2007b, p. 9). Hoggarth (2008b, pers. comm.) visited the same relocation site in 2008, and reported finding “numerous living specimens” of the rayed bean. The status of this population cannot be accurately assessed at this time, but would appear to be small and of questionable viability.

Scioto Brush Creek
—Scioto Brush Creek is a small western tributary of the lower Scioto River in Scioto County, south-central Ohio. Watters (1988a, p. 45) discovered the rayed bean in this stream in 1987, reporting two fresh dead and two relic specimens from a site, and a relic specimen from a second site among the 20 sites he collected. This population's current status is uncertain.

Great Miami River
- The Great Miami River is a major northern tributary of the Ohio River in southwestern Ohio that originates from Indian Lake in west-central Ohio and flows into the Ohio River west of Cincinnati. The occurrence of the rayed bean in the Great Miami River was discovered in August 2009, during a mussel survey for a bridge project in Logan County, Ohio. Only one individual was documented, a male approximately 7 to 8 years of age (Hoggarth 2009, pers. comm.). The following year, Hoggarth (2010, p. 5) found a juvenile rayed bean. The status of this newly discovered population is not known.

Little Miami River
—The Little Miami River is a northern tributary of the Ohio River in southwestern Ohio, flowing into the latter at the eastern fringe of the Cincinnati metropolitan area. Hoggarth (1992, p. 248) surveyed over 100 sites in the entire system. He found one live individual at a site in Warren County and possibly a subfossil shell at another site, although there is contradictory data in his paper (Butler 2002, p. 17). The latter site may have been the same as that reported for a pre-1863 record (Hoggarth 1992, p. 265). The rayed bean appears to be very rare in the Little Miami, having been found extant at only 1 of 46 mainstem sites. Hoggarth (1992, p. 267) highlighted the “fragile nature” of the extant mussel community in the system, while noting that localized reaches of the Little Miami were “severely impacted.” The species' status in the river is uncertain, but apparently very tenuous and probably headed toward extirpation (Butler 2002, p. 17).

East Fork Little Miami River
—The East Fork Little Miami River is an eastern tributary of the lower Little Miami River, with its confluence at the eastern fringe of the Cincinnati metropolitan area. According to OSUM records, eight fresh dead specimens were reported from a site in eastern Clermont County in 1973. Hoggarth (1992, p. 265) reported one live, three fresh dead, and one relic rayed bean from three sites in a 7-river-mi (11-river-km) stretch of the stream in western Clermont and adjacent Brown County (including the 1973 site). Harsha Reservoir on the East Fork destroyed several miles of potential stream habitat for the rayed bean a few miles downstream of the extant population. The status of the rayed bean in the river is uncertain, but probably of doubtful persistence (Butler 2002, p. 17).

Stillwater River
—The Stillwater River is a western tributary of the middle Great Miami River in southwestern Ohio. The rayed bean is known from two specimens, one fresh dead and one relic, collected in 1987 at two sites spanning the Miami-Montgomery County line (OSUM records). Both sites occur in the footprint of Englewood Reservoir (constructed circa 1920), which serves as a retarding basin (a constructed empty lake used to absorb and contain flooding in periods of high rain) that is normally a free-flowing river except in times of flood, therefore continuing to provide riverine habitat that is normally destroyed by permanently impounded reservoirs. The rayed bean in the Stillwater River may be extant, but its status is currently unknown and considered highly imperiled (Butler 2002, p. 17).

Tippecanoe River
—The Tippecanoe River is a large northern tributary of the middle Wabash River in north-central Indiana. The first records for the rayed bean date to circa 1900 (Daniels 1903, p. 646). Historically, this species was known from numerous sites in six counties in the Tippecanoe River. A total of 12 fresh dead specimens from 5 of 30 sites were found when sampled in 1992. The rayed bean “is apparently on the decline” in the river (ESI 1993, p. 87). The Tippecanoe rayed bean population was thought to be recruiting by Fisher (2008, pers. comm.), but appears tenuous and its long-term viability is questionable.

Lake Maxinkuckee
—Lake Maxinkuckee is a glacial lake in the headwaters of the Tippecanoe River in north-central Indiana. The rayed bean has been known from the lake for more than a century (Blatchley 1901). A 1997 OSUM record included seven fresh dead specimens collected at its outlet to the Tippecanoe River. Fisher (2002, pers. comm.), who made the 1997 OSUM collection, noted that many native mussels had zebra mussels attached to their valves that were apparently contributing to their mortality. The status of the rayed bean in Lake Maxinkuckee is, therefore, highly tenuous, and its long-term persistence questionable.

Sugar Creek
—Sugar Creek is a tributary of the East Fork White River, in the lower Wabash River system in south-central Indiana. A rayed bean population was first reported there in 1930 (Butler 2002, p. 19). Harmon (1992, p. 33) sampled 27 mainstem and 16 tributary sites, finding fresh dead specimens at 3 mainstem sites and relic specimens from 2 other sites. The sites with fresh dead material were found in the lowermost 6 mi (9.7 km) of stream. The status and viability of this tenuous population is uncertain (Fisher 2008, pers. comm.).

Tennessee River System

Historically, the rayed bean was known from the Tennessee River and 12 of its tributary streams. Ortmann (1924, p. 55) reported that the rayed bean had a “rather irregular distribution”; however, museum lots show that it was fairly common in some streams (North Fork Clinch, Duck Rivers). The last live rayed bean records from the system, with the exception of the Duck River, were from the 1960s or earlier. The species persisted in the Duck until the early 1980s. Prior to the 2008 reintroduction into the Duck River, intensive sampling in the Duck watershed had failed to locate even a relic shell of the rayed bean (Ahlstedt
et al.
2004, p. 29). Tributaries in this system have been extensively sampled over the past 25 years.

Duck River
—The Duck River is the downstream-most large tributary of the Tennessee River draining south-central Tennessee and flowing 285 river miles (459 river km) west to its confluence near the head of Kentucky Reservoir. The rayed bean was considered to be extirpated from the river until a reintroduction took place in September 2008. A total of 969 adults were collected from the Allegheny River at East Brady, Pennsylvania, in 2008 (Welte 2011, pers. comm.). Following quarantine and retention of several individuals for propagation, a total of 681 rayed bean were translocated to the Duck River near Lillard Mill, Tennessee (Urban 2010, pers. comm.; Moles and Layzer 2009, pp. 2-3; Welte 2011, pers. comm.). Although the rayed bean was extirpated from the Duck River about 25 years ago, major improvements in water

quality and physical habitat conditions have occurred in the past 15 years. In response to these improvements, recruitment of nearly all extant mussel species has been documented and suggests that reintroduction of the rayed bean might be successful (Anderson 2008, pers. comm.). The status of the reintroduced population was assessed in 2009. Rayed bean survival rates from three plots and downstream dispersal areas ranged from 38 to 62 percent (Moles and Layzer 2009, pp. 4-7).

Summary of Rayed Bean Population Estimates and Status

The information presented in this final rule indicates that the rayed bean has experienced a significant reduction in range and most of its populations are disjunct, isolated, and, with few exceptions, appear to be declining (West
et al.
2000, p. 251). The extirpation of this species from over 80 streams and other water bodies within its historical range indicates that substantial population losses have occurred. Relatively few streams are thought to harbor sizable viable populations (Sydenham, Blanchard, and Allegheny Rivers, and French and Swan Creeks). Small population size and restricted stream reaches of current occurrence are a real threat to the rayed bean due to the negative genetic aspects associated with small, geographically isolated populations. This can be especially true for a species, like the rayed bean, that was historically widespread and had population connectivity among mainstem rivers and multiple tributaries. The current distribution, abundance, and trend information illustrates that the rayed bean is imperiled.

Snuffbox Historical Distribution

The snuffbox historically occurred in 210 streams and lakes in 18 States and 1 Canadian province: Alabama, Arkansas, Illinois, Indiana, Iowa, Kansas, Kentucky, Michigan, Minnesota, Mississippi, Missouri, New York, Ohio, Pennsylvania, Tennessee, Virginia, West Virginia, and Wisconsin; and Ontario, Canada. The major watersheds of historical streams and lakes of occurrence include the upper Great Lakes sub-basin (Lake Michigan drainage), lower Great Lakes sub-basin (Lakes Huron, Erie, and Ontario drainages), upper Mississippi River sub-basin, lower Missouri River system, Ohio River system, Cumberland River system, Tennessee River system, lower Mississippi River sub-basin, and White River system.

Snuffbox Current Distribution

Extant populations of the snuffbox are known from 79 streams in 14 States and 1 Canadian province: Alabama (Tennessee River, Paint Rock River, and Elk River), Arkansas (Buffalo River, Spring River, and Strawberry River), Illinois (Kankakee River and Embarras River), Indiana (Pigeon River, Salamonie River, Tippecanoe River, Sugar Creek, Buck Creek, Muscatatuck River, and Graham Creek), Kentucky (Tygarts Creek, Kinniconick Creek, Licking River, Slate Creek, Middle Fork Kentucky River, Red Bird River, Red River, Rolling Fork Salt River, Green River, and Buck Creek), Michigan (Grand River, Flat River, Maple River, Pine River, Belle River, Clinton River, Huron River, Davis Creek, South Ore Creek, and Portage River), Minnesota (Mississippi River, St. Croix River), Missouri (Meramec River, Bourbeuse River, St. Francis River, and Black River), Ohio (Grand River, Ohio River, Muskingum River, Walhonding River, Killbuck Creek, Olentangy River, Big Darby Creek, Little Darby Creek, Salt Creek, Scioto Brush Creek, South Fork Scioto Brush Creek, Little Miami River, and Stillwater River), Pennsylvania (Allegheny River, French Creek, West Branch French Creek, Le Boeuf Creek, Woodcock Creek, Muddy Creek, Conneaut Outlet, Little Mahoning Creek, Shenango River, and Little Shenango River), Tennessee (Clinch River, Powell River, Elk River, and Duck River), Virginia (Clinch River and Powell River), West Virginia (Ohio River, Middle Island Creek, McElroy Creek, Little Kanawha River, Hughes River, North Fork Hughes River, and Elk River), and Wisconsin (St. Croix River, Wolf River, Embarrass River, Little Wolf River, and Willow Creek); and Ontario, Canada (Ausable River and Sydenham River). It is probable that the species persists in some of the 132 streams or lakes where it is now considered extirpated (Butler 2007, p. 16); however, if extant, these populations are likely to be small and not viable.

Snuffbox Population Estimates and Status

Based on historical and current data, the snuffbox has declined significantly rangewide and is now known only from 79 streams (down from 210 historically), representing a 62 percent decline in occupied streams (Table 2). Because multiple streams may comprise a single snuffbox population (French Creek system), the actual number of extant populations is fewer than 79. Extant populations, with few exceptions, are highly fragmented and restricted to short reaches. Available records indicate that 25 of 79, or 32 percent, of streams considered to harbor extant populations of the snuffbox are represented by only one or two recent live or fresh dead individuals (Little Wolf, Maple, Pigeon, Kankakee, Meramec, Ohio, Muskingum, Olentangy, Stillwater, Hughes, Green, Powell, Duck, and Black Rivers; and Little Mahoning, Woodcock, McElroy, Big Darby, Little Darby, Salt, South Fork Scioto Brush, Slate, and Buck (Indiana), Graham, and Buck (Kentucky) Creeks.

Table 2—Snuffbox Extant Stream Population Summary by Stream of Occurrence

Stream (state)

Last
observed

Recruiting

Potential
viability

Population size
Population trend

Status
category

Wolf River (WI)
2010
Yes
High
Large
Declining
Stronghold.

Embarrass River (WI)
2006
Unknown
Unknown
Small
Unknown
Marginal.

Little Wolf River (WI)
2004
Unknown
Unknown
Small
Unknown
Marginal.

Willow Creek (WI)
2001
Unknown
Unknown
Small
Unknown
Marginal.

Grand River (MI)
2002
Yes
High
Medium
Unknown
Significant.

Flat River (MI)
2010
Yes
High
Medium
Unknown
Significant.

Maple River (MI)
2001
Unknown
Unknown
Small
Unknown
Marginal.

Pine River (MI)
2002
Unknown
Low
Small
Stable
Marginal.

Belle River (MI)
2010
Yes
High
Small
Unknown
Significant.

Clinton River (MI)
2009
Yes
High
Large
Declining
Significant.

Huron River (MI)
2008
Unknown
Low
Medium
Unknown
Significant.

Davis Creek (MI)
2008
Yes
High
Medium
Unknown
Significant.

South Ore Creek (MI)
1999
Yes
High
Small
Unknown
Significant.

Portage River (MI)
1998
Yes
High
Medium
Unknown
Significant.

Grand River (OH)
2006
Yes
High
Medium
Unknown
Significant.

Upper Mississippi River (MN)
2010
No
Unknown
Unknown
Unknown
Marginal.

St. Croix River (MN and WI)
2010
Yes
High
Large
Declining
Significant.

Kankakee River (IL)
1991
Unknown
Unknown
Small
Unknown
Marginal.

Meramec River (MO)
1997
Unknown
Unknown
Small
Declining
Marginal.

Bourbeuse River (MO)
2006
Yes
High
Large
Improving
Stronghold.

Ohio River (OH, WV)
2001
Unknown
Low
Small
Unknown
Marginal.

Muskingum River (OH)
2005
Unknown
Unknown
Small
Unknown
Marginal.

Walhonding River (OH)
1991
Unknown
Unknown
Small
Declining
Significant.

Killbuck Creek (OH)
2010
Unknown
Unknown
Small
Declining
Marginal.

Olentangy River (OH)
1989
Unknown
Unknown
Small
Declining
Marginal.

Big Darby Creek (OH)
2008
Unknown
Unknown
Small
Declining
Marginal.

Little Darby Creek (OH)
1999
Unknown
Unknown
Small
Declining
Marginal.

Salt Creek (OH)
1987
Unknown
Unknown
Small
Unknown
Marginal.

Scioto Brush Creek (OH)
1987
Unknown
Unknown
Small
Unknown
Marginal.

South Fork Scioto Brush Creek (OH)
1987
Unknown
Unknown
Small
Unknown
Marginal.

Little Miami River (OH)
1991
Unknown
Unknown
Small
Unknown
Marginal.

Stillwater River (OH)
1987
Unknown
Unknown
Small
Unknown
Marginal.

Pigeon River (IN)
1998
Unknown
Unknown
Small
Unknown
Marginal.

Salamonie River (IN)
2004
Yes
Low
Small
Unknown
Significant.

Tippecanoe River (IN)
2003
Unknown
Unknown
Small
Declining
Marginal.

Embarras River (IL)
2008
Yes
Low
Small
Declining
Significant.

Sugar Creek (IN)
1990
Unknown
Unknown
Small
Declining
Marginal.

Buck Creek (IN)
1990
Unknown
Unknown
Small
Unknown
Marginal.

Muscatatuck River (IN)
1988
Unknown
Unknown
Small
Unknown
Marginal.

Graham Creek (IN)
1990
Unknown
Unknown
Small
Declining
Marginal.

St. Francis River (MO)
2006
Yes
High
Medium
Stable
Significant.

Black River (MO)
2002
Yes
Low
Small
Unknown
Significant.

Tygarts Creek (KY)
1995
Unknown
Unknown
Small
Declining
Marginal.

Kinniconick Creek (KY)
2005
Unknown
Low
Small
Declining
Marginal.

Licking River (KY)
2006
Unknown
Low
Small
Unknown
Marginal.

Slate Creek (KY)
1992
Unknown
Unknown
Small
Declining
Marginal.

Middle Fork Kentucky River (KY)
1997
Unknown
Unknown
Small
Unknown
Marginal.

Red Bird River (KY)
1995
Unknown
Unknown
Small
Unknown
Marginal.

Red River (KY)
~2002
Unknown
Unknown
Small
Unknown
Significant.

Rolling Fork Salt River (KY)
~2005
Unknown
Unknown
Small
Unknown
Marginal.

Green River (KY)
1989
Unknown
Unknown
Small
Declining
Marginal.

Buck Creek (KY)
1987-90
Unknown
Unknown
Small
Declining
Marginal.

Clinch River (TN and VA)
2006
Yes
High
Large
Stable or Declining
Stronghold.

Powell River (TN and VA)
2008
Unknown
Unknown
Small
Declining
Marginal.

Tennessee River (AL)
2006
Unknown
Unknown
Small
Unknown
Marginal.

Paint Rock River (AL)
2008
Yes
High
Large
Improving
Stronghold.

Elk River (TN and AL)
2007
Yes
Low
Small
Stable
Significant.

Duck River (TN)
2001
Unknown
Unknown
Small
Unknown
Marginal.

Buffalo River (AR)
2006
Unknown
Unknown
Small
Unknown
Marginal.

Spring River (AR)
2005
Unknown
Low
Medium
Unknown
Significant.

Strawberry River (AR)
1997
Unknown
Unknown
Small
Unknown
Marginal.

Allegheny River (PA)
2001
Unknown
Unknown
Small
Unknown
Marginal.

French Creek (PA)
2008
Yes
High
Large
Stable
Stronghold.

West Branch French Creek (PA)
2008
Unknown
Unknown
Small
Unknown
Marginal.

Le Boeuf Creek (PA)
2006
Yes
Low
Small
Unknown
Marginal.

Woodcock Creek (PA)
2007
Unknown
Unknown
Small
Unknown
Marginal.

Muddy Creek (PA)
2008
Yes
Low
Medium
Unknown
Significant.

Conneaut Outlet (PA)
1997
Unknown
Unknown
Small
Unknown
Marginal.

Little Mahoning Creek (PA)
1991
Unknown
Unknown
Small
Unknown
Marginal.

Shenango River (PA)
2010
Yes
Unknown
Small
Unknown
Significant.

Little Shenango River (PA)
2002
Unknown
Unknown
Small
Unknown
Significant.

Middle Island Creek (WV)
2009
Unknown
Unknown
Small
Declining
Marginal.

McElroy Creek (WV)
2010
Unknown
Unknown
Unknown
Unknown
Marginal.

Little Kanawha River (WV)
2010
Yes
Unknown
Unknown
Unknown
Significant.

Hughes River (WV)
2008
Unknown
Unknown
Unknown
Unknown
Marginal.

North Fork Hughes River (WV)
2001
Unknown
Low
Small
Declining
Significant.

Elk River (WV)
2010
Unknown
Low
Medium
Improving
Significant.

Ausable River (ON)
2008
Yes
High
Large
Unknown
Stronghold.

Sydenham River (ON)
2010
Yes
High
Large
Unknown
Stronghold.

Butler (2007, pp. 70-71) categorized the extant populations into three groups based on population size: general distribution, evidence of recent recruitment, and assessment of current viability. Stronghold populations were

described as having sizable populations; generally distributed over a significant, and more or less contiguous, length of stream (30 or more river mi (48 or more river km)), with ample evidence of recent recruitment; and currently considered viable. Significant populations were defined as small, generally restricted populations with limited recent recruitment and viability. Many significant populations are susceptible to extirpation, but this category has a broad range of quality. The third category, marginal populations, are defined as those which are very small and highly restricted, with no evidence of recent recruitment, of questionable viability, and that may be on the verge of extirpation in the immediate future. Following this criteria, there are 7 stronghold populations, 24 significant populations, and 48 marginal populations of snuffbox.

A population is considered extant if live individuals or fresh dead specimens have been located since approximately 1985. A population is considered to be recruiting if there was recent (within approximately 10 years) evidence of subadults (generally, individuals less than or equal to 1.5 in (3.8 cm) long or less than or equal to 4 years). Table 2 provides information on the 79 streams thought to harbor extant populations. Butler (2007, pp. 160-200) provides the complete distributional history of the snuffbox, including streams where the snuffbox is thought to be extirpated.

Upper Great Lakes Sub-Basin

The snuffbox was formerly known from 15 streams and lakes in the upper Great Lakes sub-basin. The Fox River system in Wisconsin, particularly its major tributary, the Wolf River (and its tributaries), had a widespread and locally abundant population. The species is thought to be extant in eight sub-basin streams; however, all but the Wolf and Grand Rivers have populations that are considered marginal.

Wolf River
—The Wolf River is the major tributary of the Fox River draining a large portion of northeastern Wisconsin and flowing southward to join the Fox River at Lake Butte Des Morts, near Oshkosh. Snuffbox records are known from Shawano, Waupaca, and Outagamie Counties. The snuffbox is known from a 30-river-mi (48-river-km) reach of the Wolf River (Butler 2007, p. 21). It is one of the few stronghold populations, but appears to exhibit a low level of recruitment. Only 4 of 257 individuals collected in the mid-1990s were less than 6 years old (Butler 2007, p. 21). A bridge replacement project on the south side of Shawano, scheduled to begin in 2010, may adversely impact the large snuffbox bed located just downstream (ESI 2006, p. 10). The zebra mussel occurs in this river, with a 0.7 percent infestation rate on unionids sampled in 2006 (ESI 2006, p. 6). This large population continues to be viable but appears to be in decline (Butler 2008, pers. comm.).

Embarrass River
—A western tributary of the lower Wolf River, the Embarrass River parallels the western bank of the Wolf River before joining it at New London, Wisconsin. A population of the snuffbox is located in the headwaters below a small dam at Pella, Wisconsin. Records exist for three live individuals and two dead specimens during 1987-1988 and a single dead specimen in 1995 (Butler 2007, p. 22). Its current status is unknown.

Little Wolf River
—The Little Wolf River is a western tributary of the lower Wolf River in Waupaca County, Wisconsin. The snuffbox is known from a single live individual collected in 1988 at RM 14, below the Mill Pond dam at Manawa (Butler 2007, p. 22). Five dead specimens were found during 1999 at RM 2, where shells were abundant in a muskrat midden (Butler 2007, p. 22). Nothing else is known regarding this population.

Willow Creek
—Willow Creek flows eastward into Lake Poygan, a large flow-through lake of the Wolf River system, in Waushara County, Wisconsin. The snuffbox is known from a single observation of two live females in 2001 (Butler 2007, p. 22). No other information is available on the status of this population.

Grand River
—The Grand River, a major Lake Michigan tributary, represents the largest lotic (moving water) watershed in Michigan and is located in the southwestern portion of the State. The snuffbox is sporadically distributed in approximately 25 river mi (40 river km) of the middle Grand River, approximately between the confluences of the Flat and Maple Rivers. The medium-sized population appears to be viable, with recruitment noted in 1999 (Badra 2008, pers. comm.; Zanatta 2011, pers. comm.).

Flat River
—The Flat River is a tributary to the Grand River. Zanatta (2011, pers. comm.) found 32 live snuffbox in the Flat River immediately upstream of the confluence with the Grand River in 2009. The snuffbox only occurs in the lower Flat River for approximately 0.5 river mi (0.75 river km) from the mouth upstream to the dam at State Route 21 (Zanatta 2011, pers. comm.).

Maple River
—The Maple River is a northeastern tributary of the Grand River draining south-central Michigan. A single snuffbox record (one live individual) is known from 2001 in southern Gratiot County, approximately 20 river mi (32 river km) upstream of the Grand River (Badra 2008, pers. comm.). Portions of the Maple River and several tributaries have been channelized, but the suitability of these channelized areas for the snuffbox is unknown (Badra 2010, pers. comm.). The current status of this small population is unknown.

Pigeon River
—The Pigeon River is a headwater tributary of the St. Joseph River system of Lake Michigan, flowing westward across northern-most Indiana, crossing the State border to its confluence in southwestern Michigan. One very large fresh dead specimen was found in 1998, among thousands of shells in LaGrange County, Indiana (Butler 2007, p. 24). The same site was sampled in 1996 without evidence of this species, and relic shells were found at three of nine sites sampled in 2004 (Butler 2007, p. 24). The snuffbox's occupied reach historically covered more than 10 river mi (16.1 river km) in north-central LaGrange County. The species is very rare in this river, and its viability is unknown.

Lower Great Lakes Sub-Basin

Of all the water bodies from which the snuffbox was historically recorded, 32 are in the lower Great Lakes sub-basin, including several chains-of-lakes, springs, and channels in some systems (Clinton, Huron Rivers). Historically, sizable populations occurred in some streams (Lake Erie; Belle, Clinton, Huron, Portage, and Niagara Rivers), but the species had become “characteristically uncommon” by the 1970s (Strayer 1980, p. 147). A pre-zebra-mussel decline of unionids in Lake Erie was noted (Mackie
et al.
1980, p. 101), and the snuffbox appeared extirpated there by the late 1960s. The Lake St. Clair population of snuffbox persisted until around 1983 (Nalepa and Gauvin 1988, p. 414; Nalepa 1994, p. 2231; Nalepa
et al.
1996, p. 361), which was the year the zebra mussel is thought to have invaded (Schloesser
et al.
1998, p. 70). Observations of live and fresh dead snuffbox from the Detroit River were made until 1994, but the mussel fauna has since been devastated by zebra mussels, and the snuffbox is now considered to be extirpated (Schloesser
et al.
1998 p. 69; Butler 2007, p. 25). Other snuffbox populations in the sub-basin may also have suffered from zebra mussel invasions, but not those in the Ausable and Sydenham Rivers in

Ontario. The lack of impounded area on these streams has likely prevented the introduction or the establishment of zebra mussels (Dextrase
et al.
2000, p. 10; Ausable River Recovery Team 2005, p. 12). The snuffbox is considered extant in 10 streams of the lower Great Lakes sub-basin, including stronghold populations in the Sydenham and Ausable Rivers and sizable but reach-limited populations in the Clinton River and Davis Creek. A single fresh dead valve was reported in 1998, from among 24 sites sampled in the Thames River, but no evidence of the snuffbox was found at 16 Thames sites in 2004 (McGoldrick 2005, pers. comm.). Currently, the species is considered extant in Canada only in the Ausable and Sydenham Rivers (Morris and Burridge 2006, p. 9). Both of these populations are viable.

Ausable River
—The Ausable River is a southeastern tributary of Lake Huron, draining southwestern Ontario, Canada. A survey conducted in 2008 found that a sizable population of snuffbox occurs in the lower portion of the stream in over 36 river mi (59 river km) (Zanatta 2011, pers. comm.). The size range of individuals found in the 2008 survey indicates recent recruitment in the viable population (Zanatta 2011, pers. comm.).

Pine River
—A tributary of the St. Clair River, the Pine River flows south and is located in St. Clair County, in southeastern Michigan. Although apparently stable, the snuffbox population is small, very restricted in range, and has a low potential for viability (Badra 2002, pers. comm.; Badra and Goforth 2003, p. 23).

Belle River
—The Belle River is another tributary of the St. Clair River in St. Clair County, flowing in a southeasterly direction. Records for the snuffbox date to the early 1960s, but all live and fresh dead records over the past 40 years have been from the same lower mainstem site. Historically, a sizable population was found in the Belle (65 specimens, 1965). In 2010, Zanatta (2010, pers. comm.) found four live individuals at one site and one fresh dead at another site. The Belle is located in a primarily agricultural watershed (Hoeh and Trdan 1985, p. 115), and is impacted by sedimentation and runoff. The population has declined to the point of being small, but shows evidence of recruitment and viability (Badra 2002, pers. comm.; Badra and Goforth 2003, p. 24; Sherman 2005, pers. comm.).

Clinton River
—The Clinton River is an eastward flowing chain-of-lakes tributary of Lake St. Clair in southeastern Michigan. The snuffbox population in the Clinton River is limited to around 10 river mi (16.2 river km) and lakeshore in the western suburbs of Pontiac, primarily between Cass and Loon Lakes. This population appears to be recruiting (Sherman Mulcrone 2004, p. 64; Zanatta 2011, pers. comm.) and viable, although apparently in decline since the early 1990s (Badra 2002, pers. comm.; Butler 2007, p. 27).

Sydenham River
—The Sydenham River is a large, southeasterly flowing, eastern tributary of Lake St. Clair in extreme southwestern Ontario. The snuffbox was reported in the mid-1960s and early 1970s, but was overlooked during surveys in 1985 (except dead shells) and 1991 (Butler 2007, p. 28). During the 1997-1999 sampling, a total of 10 live and fresh dead individuals were found from 4 of 12 sites, including the 3 1960s sites (Metcalfe-Smith
et al.
2003, p. 41). The snuffbox was recorded at a rate of 0.22 per hour of effort during 1997-1998 (Metcalfe-Smith
et al.
2000, p. 728). More recent sampling found 57 live and fresh dead individuals from 21 collection events (some individuals may have been counted multiple times) at six sites during 2000-2002. The increase in numbers relative to historical collections may be attributed to more intensive sampling methods rather than to improving population size (Metcalfe-Smith
et al.
2003, p. 46), thus making population trend assessments difficult (Morris and Burridge 2006, p. 12). This stronghold population is recruiting (Butler 2007, p. 28), viable, and is currently known from approximately 30 river miles (48 km) of the middle Sydenham.

Huron River
—The Huron River is a major tributary of western Lake Erie draining a significant portion of southeastern Michigan. It is a complex system of flow-through chains-of-lakes and tributaries. The snuffbox is considered extant in two disjunct upper mainstem reaches. Individuals in the middle Huron River reach and in Davis Creek are considered a single population segment (Marangelo 2005a, pers. comm.).

Zebra mussels invaded the Huron River system in the early 1990s. Zebra mussel densities on individual mussels increased from less than 1 in spring 1995 to 245 in winter 1998 (Nichols
et al.
2000, p. 72). Despite the increasing presence of zebra mussels, the Huron population is probably recruiting and viable (Butler 2007, p. 29).

Davis Creek
—Davis Creek is a chain-of-lakes in the upper Huron River system, primarily in southeastern Livingston County, Michigan. The snuffbox appears to be limited to the lower 3 river mi (4.8 river km), comprising a single population with one of the extant Huron River population segments in this area. This viable population appears to be sizable and is experiencing recent recruitment (Marangelo 2005a, pers. comm.; Zanatta 2005, pers. comm.).

South Ore Creek
—South Ore Creek is a northern tributary of the Huron River, forming a southward flowing chain-of-lakes draining southeastern Livingston County, Michigan. The snuffbox was discovered in 1999, just upstream of Ore Lake, which is near the Huron River confluence (Butler 2007, p. 31). Three subadult snuffbox (two age 2, one age 3-4) were recorded. Despite the lack of additional information, the small population appears to be viable, based on recent recruitment.

Portage River
—The Portage River is a chain-of-lakes in the northwestern portion of the Huron River system. Two University of Michigan Museum of Zoology (UMMZ) records suggest historical abundance (Badra 2002, pers. comm.). The species was reported as “rare” in the lower river during 1976-78 (Strayer 1979, p. 94). At least 22 live, young (age 4 and younger) individuals were identified in 1998, at one of three sites upstream of Little Portage Lake and Portage Lake (Butler 2007, p. 31). The localized population appears to be medium-sized and viable.

Grand River
—The Grand River is a 99-river-mi (159-river-km) tributary of Lake Erie, flowing north, then west to its confluence northeast of Cleveland, Ohio. Several museum snuffbox records date back to the 1800s. Dozens of fresh dead snuffbox were found washed up on the banks in the vicinity of the Interstate 90 crossing in Lake County, Ohio, following a major flood in 2006 (Butler 2007, p. 32). The species is known from approximately 12 river mi (19.3 river km) downstream of Harpersfield Dam (Huehner
et al.
2005, p. 59; Zimmerman 2008a, pers. comm.). The sizable population was considered recruiting, based on the 1995 Huehner
et al.
(2005, p. 59) survey.

Upper Mississippi River Sub-Basin

The snuffbox was historically known from 17 streams in the upper Mississippi River sub-basin. Records exist for Mississippi River Pools (MRPs) 3-4, 5a-6, and 14-16 (Kelner 2003, p. 6), with early surveys summarized by van der Schalie and van der Schalie (1950, p. 456). The snuffbox was considered to be extirpated from the mainstem of the Mississippi River until 2010, when it was reintroduced (Havlik and Sauer 2000, p. 4; Davis and Pletta 2010, p. 2). Only 5 of 17 historical

populations remain, but they include two of the largest rangewide (St. Croix and Bourbeuse Rivers). Three populations, including the St. Croix, appear to be declining.

Upper Mississippi River
—The Upper Mississippi River is the portion of the Mississippi River upstream of Cairo, Illinois. From the headwaters at Lake Itasca, Minnesota, the river flows approximately 1,250 miles (2,000 km) to Cairo, where it is joined by the Ohio River to form the Lower Mississippi River. The snuffbox was reported live in the upper river in the 1920s (Grier 1922, p. 15; Grier 1926, p. 119), but not from subsequent surveys (254 sites upstream of the Ohio River during 1930-1931 (UMMZ, Ellis 1931, pp. 1-10), MRPs 5-7 and 9 in 1965 (Finke 1966, Table 2; Thiel 1981, p. 16), MRPs 3-11 during 1977-79 (Thiel 1981, p. 16)). A reintroduction effort into the Mississippi River was initiated in 2010, when 200 logperch inoculated with snuffbox glochidia were placed into cages in Upper Pool 2 (Davis and Pletta 2010, p. 2: Delphey 2011, pers. comm.). It is not yet known if this reintroduction effort was successful.

St. Croix River
—The St. Croix River is a major south-flowing tributary of the upper Mississippi River and forms the border between southeastern Minnesota and northwestern Wisconsin. Densities of juvenile snuffbox declined at eight sites between 1992 and 2002 (Hornbach
et al.
2003, p. 344). Snuffbox density at Interstate Park declined significantly between 1988 and 2004 (WIDNR 2004). A flood in 2001 may have contributed to these declines in mussel density, but post-flood recruitment was also surprisingly low (WIDNR 2004). The St. Croix snuffbox population occurs from the Northern States Power Dam, at RM 54.2 to RM 36.8 (Heath 2005, pers. comm.); represents the species' northernmost occurrence; and despite recent observed declines, remains one of the most significant populations rangewide.

Kankakee River
—The Kankakee River is a major, westward-flowing, upper Illinois River tributary with its headwaters in northwest Indiana and northeast Illinois. The snuffbox was reported over a century ago (Baker 1906, p. 63), but surveys in 1911 (43 sites; Wilson and Clark 1913, pp. 41-50), 1978 (13 sites; Suloway 1981, p. 236), 1975-2000 (18 samples from an unknown number of sites in Will County, Illinois; Sietman
et al.
2001, p. 279), and 1999 (4 sites, Stinson
et al.
2000, Appendix C) failed to find it. It was considered extirpated from the Kankakee by Cummings
et al.
(1988, p. 16), but single fresh dead specimens in Illinois (Will County in 1988, Kankakee County in 1991) were subsequently found. Only relic shells have been found since 1991. The Kankakee River population, if extant, appears small, localized, and of doubtful viability.

Meramec River
—The Meramec River is a 236-mi (380-km) tributary that flows northeasterly into the Mississippi River downstream of St. Louis and drains the northeastern slope of the Ozark Plateaus in east-central Missouri. Early species lists failed to report the snuffbox (Grier 1916, p. 518; Utterback 1917, p. 28). Buchanan (1980, p. 63) found fresh dead specimens at three sites and relic shells at two other sites sampled in 1977-78. Roberts and Bruenderman (2000, p. 85) sampled 42 sites in 1997, including 26 of Buchanan's (1980, p. 5) sites, and found fresh dead specimens at RM 33.5, 48.8, and 59.8; and one live individual at RM 39.8. The live individual (2.4 in (6.1 cm), approximately 6 years old) was reported from a reach where a die-off, perhaps attributable to disease, was reported in 1978 (Buchanan 1986, p. 44). There was an obvious decline of mussels in the system based on catch-per-unit-effort data over the 20-year period (Roberts and Bruenderman 2000, p. 8). The Meramec snuffbox population is rare, sporadically distributed over approximately 26 river mi (41.8 river km), and of unknown viability.

Bourbeuse River
—The Bourbeuse River is a 149-mi (240-km), northeasterly flowing, northern tributary of the Meramec River, joining it at RM 68. The snuffbox is currently distributed over about 60 river mi (96.6 river km) upstream of RM 16, plus a disjunct site at the mouth of the river. Although it was considered to have “greatly declined” by the late 1990s (Roberts and Bruenderman 2000, p. 15), post-2000 sampling indicates that the population is recruiting, viable, and improving (McMurray 2006, pers. comm.). The Bourbeuse, one of the few stronghold snuffbox populations rangewide, has been augmented with laboratory propagated juveniles since 2002 (McMurray 2006, pers. comm.).

Lower Missouri River System

The snuffbox was historically known from four streams in this system. The highly disjunct occurrences suggest that it was more widespread historically. All populations in the system are considered extirpated (Butler 2007, p. 36).

Ohio River System

Half of the water body occurrences for the snuffbox rangewide are known from the Ohio River system. The Ohio River system once represented the largest block of available habitat for this species prior to the initiation of the navigational improvements in 1830 (Butler 2007, p. 36). Nearly the entire Ohio River mainstem is now impounded with a series of locks and dams (Butler 2007, p. 37). Sizable populations historically occurred in at least a dozen streams in the system. Today, only French Creek is considered to have a stronghold population, although nine others are also significant. Currently, the species is known from 45 of the 107 streams of historical occurrence.

Ohio River
—The Ohio River is the largest eastern tributary of the Mississippi, with its confluence marking the divide between the upper and lower portions of the latter system. Numerous historical records are known from throughout the River. Recently, single fresh dead and live specimens have been reported from just below Belleville Lock and Dam, Ohio and West Virginia, in 1995 and 2001, respectively (ESI 2002, p. 27). Having persisted in this highly modified river may indicate that the small population exhibits at least a low level of viability.

Allegheny River
—The 325-mi (523-km) Allegheny River drains northwestern Pennsylvania and a small portion of adjacent New York flowing south before joining the Monongahela River at Pittsburgh to form the Ohio River. Snuffbox collections are sporadically known since around 1900 in Pennsylvania from Forest County downstream to Armstrong County. The snuffbox is currently known from three disjunct sites over a 42-river-mi (67.6-river-km) reach centered in Venango County (Butler 2007, p. 37). Its occurrence in the lower Allegheny River and lower French Creek could be considered a single population segment. The viability status of the small population is unknown.

French Creek
—French Creek is a major tributary of the middle Allegheny River with its headwaters in western New York and flowing south into northwestern Pennsylvania. The snuffbox is known from the length of the stream in Pennsylvania in Erie, Crawford, Mercer, and Venango Counties. Most records date since approximately 1970 (Dennis 1971, p. 97). Snuffbox collections made during 2002-2004 were summarized by Smith (2005, p. 3-9). Live and fresh dead specimens were found at 19 sites throughout the stream. The size of the L individuals indicated that multiple year classes were represented, including subadults. The species stretches for approximately 80 river mi (128.7 river km) from around RM 10, upstream. The

population encompasses several of its tributary population segments as well, making it relatively more secure when compared to most of the other stronghold populations that are linearly distributed and, thus, more susceptible to stochastic events (Sydenham, Bourbeuse, and Clinch Rivers). The French Creek snuffbox population is considered large and viable (Evans 2003a, pers. comm.; Zimmerman 2008c, pers. comm.), appears stable, and may represent the best stronghold population rangewide.

West Branch French Creek
—West Branch of French Creek follows a southerly course to its parent stream in Erie County, Pennsylvania. The only record for the snuffbox dates from 1993, but the number of specimens and shell condition are unknown (Evans 2003b, pers. comm.). Union City Lake isolates the upper French Creek and West Branch French Creek population segment from the main French Creek population. The snuffbox was not found at three sites sampled in 2006 (Smith 2006, pers. comm.). Zimmerman (2008c, pers. comm.) documented 38 live individuals at a site near Wattsburg, Pennsylvania. This population appears to be small and of unknown viability.

Le Boeuf Creek
—Le Boeuf Creek is a small western tributary of upper French Creek flowing in a southerly direction just west of West Branch French Creek in Erie County. The first snuffbox collections in this creek were made 100 years ago (Ortmann 1909a, p. 188). Two fresh dead and 6 relic shells were reported in 1988 (Evans 2003b, pers. comm.), and 1 live, 16 fresh dead, and 8 relic specimens were found in 1991 (Butler 2007, p. 40). Three live individuals were found at a site in 2006 (Smith 2006, pers. comm.; Smith
et al.
2009, p. 69). The snuffbox population has recently recruited and exhibits some level of viability, but appears to be very limited in extent.

Woodcock Creek
—Woodcock Creek is an eastern tributary of upper French Creek in Crawford County, Pennsylvania. Until recently, the snuffbox was thought to be extirpated from this stream. In 2007, one live male was found at one of three sites sampled (Smith
et al.
2009, pp. 84-85). Viability is unknown.

Muddy Creek
—Muddy Creek is an eastern tributary of upper French Creek in Crawford County, Pennsylvania. The snuffbox was not discovered until the summer of 2003. Forty-two L individuals were reported from 11 of 20 lower river sites (Morrison 2005, pers. comm.; Mohler
et al.
2006, pp. 581-582). Low numbers were found at most sites, but 18 live individuals were collected from a site near the mouth of the river. This occurrence is considered to be part of the more extensive French Creek snuffbox population. Zimmerman (2008c, pers. comm.) documented one live female in 2008. The population is medium-sized, occurs along 8 river mi (12.9 river km) of the lower mainstem, and is recruiting, as recent juveniles were recorded (Morrison 2005, pers. comm.; Mohler
et al.
2006, p. 576).

Conneaut Outlet
—This stream forms the outlet to Conneaut Lake, flowing in a southeasterly direction until its confluence with middle French Creek, Crawford County. The snuffbox was first reported by Ortmann (1909a, p. 188), and was rediscovered live in 1997, but without collection details (Butler 2007, p. 40). No specimens were found at a site sampled in 2006 (Smith 2006, pers. comm.). The snuffbox is considered rare in this stream and its viability is unknown.

Little Mahoning Creek
—Little Mahoning Creek is a tributary of Mahoning Creek, a lower eastern tributary of the Allegheny River northeast of Pittsburgh. The snuffbox was discovered in 1991, when sampling produced two FD and one R specimen at 1 of 12 sites in the system (Butler 2007, p. 41). The lower 10 miles (16 km) of Little Mahoning Creek are subject to periodic inundation by a reservoir on Mahoning Creek (Butler 2010, pers. comm.). However, the impact of this periodic flooding on the snuffbox is not known. A 2007 survey failed to find any live or fresh dead snuffbox (Chapman and Smith 2008, p. 166). Viability is unknown.

Shenango River
—The Shenango is a large tributary in the Beaver River system, a northern tributary of the upper Ohio River in west-central Pennsylvania. The snuffbox was reported from four sites on the Shenango in 1908 (Ortmann 1919, p. 328). Six live individuals were collected from three sites sampled in 2001-2002 between Jamestown and New Hamburg (about 25 river mi (40.2 river km)). Nelson and Villella (2010, p. 17) found 45 L individuals in 2010. The upper reach is considered the best habitat in the Shenango River. The population is small and has declined, although some recent reproduction is evident (Zimmerman 2008b, pers. comm.; Nelson and Villella 2010, p. 17).

Little Shenango River
—The Little Shenango River is a small tributary of the upper Shenango River, Mercer County, Pennsylvania. This population was not located during limited surveys (Dennis 1971, p. 97; Bursey 1987, p. 42), but a single fresh dead museum record from 1991 exists. The species was reported to be relatively abundant and reproducing in the lower portion in 2002 (Zimmerman 2008b, pers. comm.). Viability of the small population is unknown.

Middle Island Creek
—Middle Island Creek is a small tributary of the Ohio River in northwestern West Virginia. The first snuffbox records were made at six sites in 1969, when the species was locally common in Doddridge, Tyler, and Pleasants Counties (Taylor and Spurlock 1981, p. 157). The snuffbox was later found at two sites in Tyler County in 1980, and the overall mussel population was considered to be “thriving” (Taylor and Spurlock 1981, p. 157). The most recent records are for three live individuals in 2009 at two sites and four live individuals in 2010 at three sites (Clayton 2011, pers. comm.). This snuffbox population has declined, is currently rare, and has questionable viability (Zimmerman 2008b, pers. comm.).

McElroy Creek
—McElroy Creek is a tributary to Middle Island Creek in West Virginia. There are no historical records for the snuffbox in McElroy Creek. Clayton (2011, pers. comm.) reported finding one live individual in 2010 in Tyler County. The status of this snuffbox population is unknown.

Muskingum River
—The Muskingum River is a large, southerly flowing, northern tributary of the upper Ohio River draining a significant portion of east-central Ohio. The snuffbox, which has a long collection history dating to the early 1800s, occurred along the entire mainstem and was locally abundant. Two live individuals and two fresh dead shells were found in 1979, but no live or fresh dead snuffbox were found in surveys conducted in 1979-81 (Stansbery and King 1983) and in 1992-93 (Watters and Dunn 1993-94, p. 241). A single live male was located during sampling for a construction project in 2005 near Dresden, Ohio (Jones
et al.
2005, p. 30). Viability of this population is unknown.

Walhonding River
—The Walhonding River is a short (23.3 river mi (37.5 river km)), east flowing tributary of the Muskingum River in central Ohio, forming the latter river at its confluence with the Tuscarawas River, and formed by the confluence of the Mohican and Kokosing Rivers. The snuffbox historically occurred throughout the river. The extant snuffbox reach (RM 1.8-6.8) is downstream from Killbuck Creek. The population had apparently declined in range and size by the early 1990s, and possibly further since. A once productive site about 0.25 mi (0.40 km) downstream of the Killbuck Creek confluence yielded only a few mussels

of very common species in 2006, but no snuffbox (Butler 2007, p. 44). The Walhonding River population is considered small and of unknown viability.

Killbuck Creek
—Killbuck Creek is a large tributary of the lower Walhonding River, flowing south from southern Medina County to Coshocton County and entering the latter at approximately RM 7. Live and fresh dead snuffbox were found by Hoggarth (1997, p. 33) at eight sites from RM 15 to the mouth. Its occurrence has become more sporadic in the last 10 years. In spring 2006, 4 live adults were found at 2 sites approximately 3 river mi (4.8 river km) apart, while 9 large live individuals and a single fresh dead specimen were collected near RM 13 during fall 2006 (Ahlstedt 2007, pers. comm.; Butler 2007, p. 45). Two large live males were collected in 2010 (Ahlstedt 2010, pers. comm.). A shrinking distribution, declining population size, and lack of evidence of recent recruitment suggest that the population may be losing viability and trending towards extirpation.

Little Kanawha River
—The Little Kanawha River is a 169-mi (269-km) long tributary of the Ohio River in western West Virginia. Schmitt
et al.
(1983, p. 137) reported snuffbox from three sites during a 1981-82 survey. Snuffbox were not documented again in the Little Kanawha River until 2010, when four live individuals, including at least one young mussel, were found at a site in Gilmer County, West Virginia (Clayton 2011, pers. comm.). Additionally, two fresh dead specimens were found in 2010, below Wells Dam near Elizabeth, Wirt County, West Virginia (Clayton 2011, pers. comm.). The current status of this snuffbox population is unknown.

Hughes River
—The Hughes River is an 18-mile (29-km) long tributary of the Little Kanawha River in western West Virginia. Schmitt
et al.
(1983, p. 137) reported snuffbox during a 1981-82 survey. No additional snuffbox were found in the Hughes River until 2008, when one fresh dead specimen was found in Wirt County (Clayton 2011, pers. comm.). The current status of this snuffbox population is unknown.

North Fork Hughes River
—The North Fork Hughes River is a westerly flowing tributary of the Hughes River in the lower Little Kanawha River system in northwestern West Virginia. The snuffbox was found at one of six North Fork sites sampled during a 1981-1982 survey of the Little Kanawha River system (Schmidt
et al.
1983). A total of 41 live adult individuals (23 reported as gravid) were reported at 5 sites located over a 1.5-mi (2.4-km) reach in North Fork State Park, Richie County, in 1993 (Butler 2007, p. 46). At least 10 live individuals were found at a site in the park in 1997 (Butler 2007, p. 46), and a single fresh dead specimen was collected at an additional site downstream in 2001 (Butler 2007, p. 46). This small snuffbox population is declining and currently restricted to less than 4 river mi (6.4 river km), but may be viable.

Elk River
—The Elk River is a major, 181-mi (291-km) tributary in the lower Kanawha River system draining central West Virginia flowing west to the Kanawha at Charleston. The snuffbox went undetected in a 1920s survey (Butler 2007, p. 46). Ten live individuals were collected during 1991-1995, the smallest being about 5 years old (Butler 2007, pp. 46-47). Collectively, 16 live individuals were identified at 8 sites in a 13-river-mi (20.9-river-km) reach in Kanawha County in 2002, and 4 live individuals were found at 4 sites in 2004 over a 16.8-river-mi (27-river-km) reach farther upstream (Douglas 2005, pers. comm.). This medium-sized population extends over 30 river mi (48.3 river km), is viable, and may have improved since the 1970s.

Tygarts Creek
—Tygarts Creek is a small, north-flowing, southern tributary of the Ohio River in northeastern Kentucky. Thirteen snuffbox were reported from one of five sites sampled in 1977 (Taylor 1980, p. 90). Fresh dead specimens are also known from 1981 and 1987 (Cicerello 2003, pers. comm.). Nine live (Butler 2007, p. 47) and 36 fresh dead specimens were found at 2 sites, respectively, in 1988, while 1 live and 2 fresh dead were reported from at least 2 sites in 1995 (Cicerello 2003, pers. comm.). The overall mussel population appeared “healthy” in 1977 (Taylor 1980), but the small snuffbox population has recently declined, and its viability is unknown.

Scioto River System
—The Scioto River system in central and south-central Ohio is a major northern tributary of the upper Ohio River. The system was one of the most routinely sampled watersheds for mussels (mostly OSUM records), and historically harbored a large and thoroughly dispersed snuffbox population in the mainstem and 16 tributaries. The system was either exceptional for its snuffbox population, or it provided a general historical perspective of what researchers may have found if other systems had been as thoroughly sampled. Sizable populations were noted in at least the Olentangy River, Big Darby Creek, and Big Walnut Creek. Development associated with the Columbus metropolitan area has taken a major toll on the aquatic fauna. Pollutants from the 1800s included wastes from sawmills, breweries, and slaughterhouses (Butler 2007, p. 48). Only a few fish species were found in the Scioto River 100 years ago (Trautman 1981, p. 33). Currently, 90 to 95 percent of the normal summer-fall flow in the river consists of wastewater treatment plant discharges (Yoder
et al.
2005, p. 410). Museum records indicate that the snuffbox had completely disappeared from the mainstem by the 1970s. A series of reservoirs around Columbus fragmented habitat and eliminated or reduced populations (Olentangy and Scioto Rivers; Alum, Big Walnut and Deer Creeks). Currently, remnant populations remain in six streams, making the snuffbox precariously close to extirpation throughout this once rich system.

Olentangy River
—The Olentangy River is a major headwater tributary of the Scioto River, draining central Ohio and flowing south to its confluence in Franklin County. OSUM snuffbox records date to the 1870s, although most are from the 1950s and 1960s. The snuffbox was reported from 15 of 31 mainstem sites collected during a 1960-1961 survey, when it appeared “fairly common” in the lower river (Stein 1963, p. 138). A single live individual in southern Delaware County and two fresh dead specimens in eastern Marion County were found among 30 sites in 1989, with relic shells at 7 other sites (Hoggarth 1990, pp. 20-27). The small population has declined (Hoggarth 1990, p. 14), and its viability is unknown.

Big Darby Creek
—Big Darby Creek is one of the major tributaries draining the northwestern portion of the Scioto River system in central Ohio. Dozens of large OSUM lots of snuffbox date to the late 1950s; six Pickaway County collections in 1962 alone had 250 live and fresh dead specimens. Watters (1990, p. 4; 1994, p. 100) surveyed 42 mainstem sites in 1986 and 49 sites in 1990. Combining the data from both years, 80 live and fresh dead snuffbox were collected at 22 sites (Watters 1994, p. 101). The population in 1990 occurred in a reach from approximately RM 11.5 to RM 42.5. The snuffbox was recruiting (Watters 1994, p. 101); four individuals during both 1986 and 1990 were 2 to 5 years of age. The overall population trend over the past 40 years has been downward. Between 1986 and 1990, the number of live and fresh dead specimens was reduced from 54 to 16, and the population's distribution declined from 17 to 8 sites. Two fresh

dead specimens were found at sites in Franklin (1996) and Pickaway (2000) Counties, and three other sites produced only relic specimens (OSUM records). This historically large snuffbox population has declined to marginal status, and its viability is questionable.

Little Darby Creek
—Little Darby Creek is the major tributary in the Big Darby Creek system, flowing in a southeasterly direction to its confluence in southwestern Franklin County, Ohio. The 25 OSUM lots for this species are small (fewer than 5 specimens per lot), date to the early 1960s, and represent lower mainstem sites in Madison County. Single fresh dead and relic specimens were collected in 1999, from a Union County site (OSUM 66740), where live individuals were collected in 1964 (Stein 1966, p. 23). This site yielded only relic specimens in 1990 (Watters 1990, Appendix A.11; 1994, p. 102). Overall, the snuffbox was historically known from 35 river mi (56 river km). The well-documented OSUM collection history illustrates the steady decline of a snuffbox population nearing extirpation.

Salt Creek
—Salt Creek is an eastern tributary in the Scioto River system, south-central Ohio. All records (OSUM) were collected in the lower mainstem (Ross County) beginning in 1958. A single live individual from 1987 represents the last known record. The mussels in this system “have been heavily impacted, apparently by the towns of Adelphi and Laurelville” (Watters 1992, p. 78). The current status of this snuffbox population is unknown.

Scioto Brush Creek
—Scioto Brush Creek is a small, western tributary of the lower Scioto River in Scioto County, south-central Ohio. The snuffbox was discovered here in the 1960s (Watters 1988a, p. 45). Three live and fresh dead specimens from 2 sites and relic shells from 2 other sites were collected during a 1987 survey covering 11 sites (Watters 1988a, pp. 210-220). The snuffbox population, collectively known from five fragmented sites along the lower two-thirds of stream, is small, and its viability is unknown.

South Fork Scioto Brush Creek
—South Fork Scioto Brush Creek is a small tributary of Scioto Brush Creek, in the lower Scioto River system. A single snuffbox was found during a survey of five sites in 1987 (Watters 1988a, pp. 210-220). The South Fork and Scioto Brush Creek populations can be considered a single population unit; the viability of this unit is uncertain.

Kinniconick Creek
—Kinniconick Creek is a small, southern tributary of the Ohio River in northeastern Kentucky. Snuffbox were reported live from 4 of 15 sites sampled in 1982, with relic shells from an additional 2 sites (Warren
et al.
1984, pp. 48-49). Single fresh dead and live snuffbox were collected in 2001 and 2004, respectively, from sampling efforts at several sites (Butler 2007, p. 51), and a single fresh dead specimen was found while resurveying four sites in 2005 (Butler 2007, p. 51). The snuffbox declined in the past few decades, it is considered rare, and its viability is uncertain.

Little Miami River
—The Little Miami River is a northern tributary of the Ohio River in southwestern Ohio, flowing south into the latter at the eastern fringe of the Cincinnati metropolitan area. Snuffbox records from the Little Miami date to the mid-1800s, but most collections are from the past several decades. Seven fresh dead specimens were found at 4 of 46 mainstem sites surveyed during 1990-1991, with 10 relic shells at 6 other sites (Hoggarth 1992, p. 265). The fresh dead specimens were found in approximately 20 river mi (32.2 river km), mostly in Warren County. Current viability of this small population is unknown.

Licking River
—The Licking River is a southern tributary of the Ohio River in northeastern Kentucky, flowing in a northwesterly direction to its confluence across from Cincinnati. The snuffbox occurred at 13 of 60 historical mainstem sites below Cave Run Reservoir (Laudermilk 1993, p. 45) and a preimpoundment site in the reservoir footprint (Clinger 1974, p. 52). The population extended approximately 50 river mi (80.5 river km). All collections of snuffbox are small in number (Butler 2007, p. 52). A single live individual and a fresh dead specimen were found at 2 sites, and relic shells were reported from 7 other sites among 49 sites sampled in 1991 (Laudermilk 1993, p. 45). Single live and fresh dead snuffbox were collected in 1999 (Cicerello 2003, pers. comm.), and a single live individual was found in 2006 (Butler 2007, p. 53). At this location, the snuffbox has become very rare and sporadic in occurrence, and its viability is questionable.

Slate Creek
—Slate Creek is a southern tributary of the Licking River below Cave Run Dam in east-central Kentucky. Historically, the snuffbox was considered “extremely abundant throughout the stream” (Taylor and Spurlock 1983) and collectively known from six sites (Laudermilk 1993, p. 45). Seventeen dead specimens were recorded from a site in 1987 (Cicerello 2003, pers. comm.). A single fresh dead and seven relic specimens were found at three sites sampled in 1991 (Butler 2007, p. 53), when it was considered “occasional” in distribution (Laudermilk 1993, p. 45). Twelve live individuals were found in 1992 (Cicerello 2003, pers. comm.). Subsequent sampling has produced no additional snuffbox; two sites and four sites yielded only relic specimens in 2001 and 2002, respectively (Cicerello 2005, pers. comm.). If extant, the population is marginal at best, with unlikely viability.

Stillwater River
—The Stillwater River is a 67-mi (108-km), western tributary of the Great Miami River draining southwestern Ohio. The species was collectively known from eight sites throughout the river (Watters 1988a, pp. 59-71; OSUM records). One fresh dead specimen below Englewood Dam in Montgomery County was found among 18 sites surveyed in 1987, with relic shells from 5 other sites (Watters 1988a, pp. 59-71). No other information on the small population is available, and its viability is unknown.

Middle Fork Kentucky River
—The Middle Fork is one of three headwater tributaries (with the North and South Forks) forming the Kentucky River, flowing in a northerly then westerly direction and draining a portion of southeastern Kentucky. The snuffbox was first reported in 1966. Three live individuals and a relic shell were found at three sites in 1996, and a single live individual was collected from another site in 1997 (Cicerello 2003, pers. comm.). All sites occur within a 10-river-mi (16-river-km) reach above Buckhorn Reservoir in Leslie County. This small population has unknown viability.

Red Bird River
—The Red Bird River is a north-flowing headwater tributary of the South Fork Kentucky River in Clay County, southeastern Kentucky, forming the latter at its confluence with Goose Creek. Ten fresh dead specimens were recorded from two sites in 1988, and three live and one fresh dead snuffbox were collected from four sites in 1995 (Cicerello 2003, pers. comm.). This small population occurs sporadically in the lower 20 river mi (32 river km), and viability is unknown (Cicerello 2003, pers. comm.; 2006, pers. comm.).

Red River
—The Red (or North Fork Red) River is a westerly flowing tributary of the upper Kentucky River in eastern Kentucky. No live snuffbox were found in surveys of the 9-river-mi (15-river-km) reach of the Wild River section during surveys of 1980, 1986, and 1991 (Houp 1980, p. 56; 1993, p. 96), but two fresh dead and one live snuffbox were found at three sites in 1988, while five live individuals were found in 1996 (Cicerello 2006, pers.

comm.). Mostly males have been found since 2002, and they are being held in captivity for future culture efforts (Butler 2007, p. 55). A small population persists over a 10-river-mi (16-river-km) reach in the lower section of the Red River Gorge Geological Area of the Daniel Boone National Forest in Menifee, Wolfe, and Powell Counties (Cicerello 2006, pers. comm.). Viability of this population is unknown.

Rolling Fork Salt River
—The Rolling Fork is a major southern tributary of the Salt River in central Kentucky, flowing in a northwesterly direction to join the Salt near its mouth. The snuffbox was first reported in 1958 (Rosewater 1959, p. 62). Seven fresh dead specimens and a single live subadult were collected in 1988, from four sites in Larue, Marion, and Nelson Counties (Cicerello 2003, pers. comm.; Haag 2006, pers. comm.). A survey of 12 mainstem and 30 tributary sites in the Rolling Fork system in 1998-1999 yielded no evidence of the snuffbox, prompting an investigator to consider it extirpated (Akers 2000, p. 13), but occasional specimens may still be found (Butler 2007, p. 55). The species is sporadically distributed over 40 river miles of the upper river (Cicerello 2006, pers. comm.). If it is still extant, the viability of this small population is unknown.

Green River
—A major southern tributary of the lower Ohio River, the Green River flows in a westerly direction and drains west-central Kentucky. Ortmann (1926, p. 182) considered the snuffbox to be well-distributed over the system, but not abundant. Large museum collections of snuffbox were taken from Munfordville during 1961-1966, but only six relic shells were reported there in 1967. The snuffbox has been rare since. Five live and fresh dead snuffbox were collected at 4 of 42 sites during 1987-1989 sampling in Mammoth Cave National Park (Cicerello and Hannan 1990, pp. 16-17). Three live and six fresh dead snuffbox were reported in the upper Green River from 1984-1990 (Cicerello 2003, pers. comm.). A single live individual was collected in Taylor County in 1989 (Layzer 2009, pers. comm.), but no evidence of the snuffbox was reported at numerous other sites in 1999, 2000, 2001, and 2003 (Cicerello 2006, pers. comm.). Once abundant and occurring over 200 river mi (322 river km), the species has become exceedingly rare since the 1960s. Current snuffbox viability is unknown, and it may be nearing extirpation from the entire Green River system, where it was formerly known from eight tributaries.

Wabash River system
—The Wabash River is the second largest sub-basin within the Ohio River system, the watershed of the 350-mi (563-km) river encompassing much of Indiana, west-central Ohio, and southeastern Illinois. The mainstem and at least 27 streams had one of the largest snuffbox population clusters. The species persists today as seven small populations in the system; the viability of these populations is unknown (Butler 2007, p. 57).

Salamonie River
—The Salamonie River is a southern tributary of the upper Wabash River, flowing in a northwesterly direction and draining east-central Indiana. Two historical museum records were found. Nine sites were surveyed during 1993-1994, without finding any evidence of the snuffbox (ESI 1995, p. 19). The snuffbox was rediscovered in 2004, above Salamonie Reservoir, where two live individuals at one site and fresh dead shells, including a very small juvenile, were found at another site 2 mi (3 km) away (Fisher 2005, pers. comm.). The small population is considered to be recruiting and viable at some level.

Tippecanoe River
—The largest tributary of the upper Wabash River system, the Tippecanoe River drains north-central Indiana and flows westerly, then southerly before joining the Wabash near Lafayette. Nearly all records of the snuffbox were made in the past 20 years. Two weathered shells were found in the lower mainstem among 16 sites sampled in 1987 (Cummings
et al.
1987, p. 25; Cummings and Berlocher 1990, p. 93) and 30 sites in 1991-1992 (ESI 1993, p. 68). One L individual and over 32 fresh dead specimens were found at a site at the upper end of Freeman Reservoir during a 1993 drawdown that may have contributed to their demise (Fisher 2003, pers. comm.). A single fresh dead specimen was found below Shafer Reservoir among 13 sites sampled in 2003 (ESI 2003, p. 9). The viability of this declining population is unknown, but it appears close to extirpation (Fisher 2003, pers. comm.).

Embarras River
—The Embarras River is a southerly flowing, western tributary of the lower Wabash River in southeastern Illinois. Museum lots represent collections dating to 1956 and contain snuffbox from nine mainstem and two tributary sites. A total of 9 live and 15 fresh dead specimens were collected at four sites in 1986, in Coles and Douglas Counties (Cummings
et al.
1988, p. 8). Although overall mussel abundance at the 21 sites sampled in both 1956 and 1986 dropped 86 percent, the snuffbox was one of only five species that showed relatively stable population size over the 30-year period (Cummings
et al.
1988, p. 9). Additional L and FD snuffbox from museum collections were recorded from single sites in 1988. Three L and eight FD snuffbox were found at two sites in 1992, and one live and three fresh dead were found at three of six sites surveyed during 2001-2002. Since 1986, the small snuffbox population has occurred sporadically at six sites over 50 river mi (80 river km) of the upper river. The species was reported as significant and viable by Butler (2007 pers. comm.), but it has declined to some extent. Recent surveys, however, documented only one live individual in 2005 and 5 live adult males in 2008, indicating that the Embarras River population may be closer to a marginal population than a significant one (Tiemann 2010, p. 53).

Sugar Creek
—Sugar Creek is a tributary in the upper East Fork White River system, draining central Indiana east and south of Indianapolis. A single live individual from one site, fresh dead specimens from seven sites, and relic shells from an additional eight sites were reported in 1990 (Harmon 1992, pp. 40-41 1998). The snuffbox population occurred sporadically over 35 river mi (56 km) to near the mouth. Only relic shells were found while resampling some historical sites in 1995, 1998, and 2001 (Butler 2007, p. 59). It is questionable whether the population remains extant.

Buck Creek
—Buck Creek is a southerly flowing, western tributary of Sugar Creek in the upper East Fork White River system east of Indianapolis. A fresh dead snuffbox was found near the mouth and relic specimens at an upstream site in 1990 (Harmon 1992, p. 41). Similar to the parent stream population in Sugar Creek, the snuffbox may already be extirpated in Buck Creek (Fisher 2003, pers. comm.).

Muscatatuck River
—The Muscatatuck River is a large, westerly flowing tributary of the upper East Fork White River in southeastern Indiana. The snuffbox was first reported from the stream by Daniels (1903, p. 646). Fresh dead specimens (unknown number) were recorded at a site downstream from Graham Creek that was sampled in 1988 (Harmon 1989, p. 118). Status and viability of snuffbox in the Muscatatuck River are unknown.

Graham Creek
—Graham Creek flows southwesterly to join Big Creek in forming the Muscatatuck River in the East Fork White River system in southeastern Indiana. The species was found fresh dead (numbers unknown) at six sites over 10 river mi (16 river km) of the lower stream in Jennings County

in 1988 (Harmon 1989, p. 117), and a single fresh dead specimen was found in 1990 (Harmon 1998). Viability of these small populations is unknown.

Cumberland River System
—Snuffbox populations are known from the mainstem Cumberland River and 6 of its tributaries. With few exceptions, most mainstem records were made prior to the 1920s, when the species was locally common (Wilson and Clark 1914, p. 45). The snuffbox is considered extirpated from the mainstem. Currently, a single tributary population may be extant, but is considered not viable. The species is likely to become extirpated from the entire river system in the foreseeable future.

Buck Creek
—Buck Creek is a southerly flowing, northern tributary of the upper Cumberland River below Cumberland Falls in southeastern Kentucky. One dead valve was found at a site in 1981 (Clarke 1981b, Appendix), and two live and one fresh dead snuffbox were reported from three sites during 1983-1984 (Schuster
et al.
1989, p. 82). The species was also reported live from a lower mainstem site among seven sites sampled from 1987-1990 (Layzer and Anderson 1992, p. 16). A recent survey found only relic shells at 3 of 23 sites (Hagman 2000, p. 21). If extant, the declining snuffbox population in Buck Creek is likely to become extirpated in the foreseeable future.

Tennessee River System

The Tennessee River is the largest tributary of the Ohio River, draining seven southeastern States and joining the Ohio near its mouth in western Kentucky. The snuffbox originally was known from throughout all but the lower section of river and 17 of its tributaries. Hundreds of miles of large river habitat on the mainstem have been lost under nine reservoirs, with additional dams on several tributaries (Clinch, Holston, and Elk Rivers) (Tennessee Valley Authority (TVA) 1971, p. 4). The loss of mussel resources has been substantial (Watters 2000, p. 262). Muscle Shoals, the 53-river-mi (85-river-km) reach in northwestern Alabama, historically harbored 69 mussel species, the most diverse mussel fauna ever known (Garner and McGregor 2001, p. 155). The construction of three dams (Wilson in 1925, Wheeler in 1930, and Pickwick Landing in 1940) inundated most of the mussel beds. No live snuffbox have been reported at Muscle Shoals for around 100 years (Garner and McGregor 2001, p. 162). The snuffbox may persist in the mainstem at a very low density and in only five tributaries. The Clinch River maintains a stronghold population, but highly restricted populations persist in the other streams.

Clinch River
—The 350-mi (563-km) Clinch River is a major tributary of the upper Tennessee River originating in southwestern Virginia, and flowing in a southwesterly direction to its confluence near Kingston in eastern Tennessee. No other river in North America has extant populations of more federally endangered (15) species of mussels than does the upper Clinch River above Norris Reservoir. The snuffbox was reported from nine sites by Ortmann (1918, pp. 601-606). Museum records from Hancock County, Tennessee, during 1965-1971 documented a very large population of snuffbox. The snuffbox is generally distributed from RM 170 to RM 195 in Hancock County, but is sporadic in Virginia (RM 213-235), where it has recently declined (Butler 2007, p. 62). The snuffbox population is recruiting, viable, and currently stable, although decreased in size and range from 40 years ago. The Clinch River ranks among the six stronghold snuffbox populations rangewide.

Powell River
—The Powell River is the major tributary of the upper Clinch River flowing in a southwesterly direction parallel to and northwest of the Clinch River in southwestern Virginia and northeastern Tennessee. The snuffbox was reported at three sites by Ortmann (1918, pp. 597-598), five sites during 1973-1978 by Dennis (1981, p. 3), four sites from 1975-1978 by Ahlstedt and Brown (1979, p. 42), and four Virginia sites in 1988-1989 by Wolcott and Neves (1994, p. 7). Large collections attest to its former abundance. The species was found live and fresh dead in the Powell River, Tennessee, during 1989-1990 (Hubbs
et al.
1991, Appendix A). Of twenty two sites sampled in the Powell River, Johnson (2010) collected seven L individuals among three sites between RM 80.4 and 95.3. The population has declined, viability is questionable, and its extirpation may be imminent (Butler 2007, p. 63).

Tennessee River
—The snuffbox originally was known from all but the lower section of the river. Butler (2007, p. 61) reported the snuffbox as “believed to be extirpated from the entire Tennessee River.” However, Yokley (2002, p. 1) collected a single fresh dead male in 2002 at the U.S. 231 Bridge, Madison and Morgan Counties, Alabama. In 2006, one live female was found at the same location, although it was the only snuffbox out of 8,978 mussels collected at the site (Yokley 2006, p. 1). Nothing further is known about the status of the snuffbox in the Tennessee River mainstem.

Paint Rock River
—The Paint Rock River is a southerly flowing, northern tributary of the southern bend of the Tennessee River in northeastern Alabama and adjacent Tennessee. The snuffbox was first reported from one of six mainstem sites by Ortmann (1925, p. 359). No evidence of snuffbox was found in two surveys during 1965-1967 (Isom and Yokley 1973, p. 444) and a 1980 survey (Butler 2007, p. 64). Twelve live and fresh dead snuffbox were found at four sites between RMs 13 and 21 (Ahlstedt 1995-1996, p. 70). The species was again absent from 10 upper mainstem sites surveyed in 2002 (Godwin 2002, p. 9). Four fresh dead specimens of varying sizes were found at lower river sites in 2002 (Fraley 2003, pers. comm.; Smith 2005, pers. comm.) and 2003-2006 (Freeman 2006, pers. comm.). One live and 11 fresh dead specimens were found at RM 21 in 2005, and 2 live and 16 fresh dead were collected at RM 31 in 2007 (Gangloff 2007, pers. comm.). In July 2008, Freeman (2008, pers. comm.) observed multiple age classes (sizes) of fresh dead snuffbox in middens between RM 34.7 and 32.5. Fobian
et al.
(2008, p. 14) collected 21 live snuffbox at 7 sites and fresh dead specimens at 8 sites between RM 46.7 and 13.1. A stronghold snuffbox population exists between RMs 13 and 44, and is recruiting, viable, and has clearly improved since 1980.

Elk River
—The Elk River is a large, northern tributary flowing 200 river mi (322 river km) in a southwesterly direction in the southern bend of the Tennessee River in south-central Tennessee and north-central Alabama. Snuffbox collections have been sporadic. The species was found at 2 sites in the mid-1960s (Isom
et al.
1973, p. 440), and a single live individual was found among 108 sites sampled in 1980 (Ahlstedt 1983, p. 47). Single specimens were also reported from 4 sites sampled in the lower river in 1997 (Madison and Layzer 1998, Table 6) and 16 sites sampled in 1999 (Service 1999, p. 3). A very large fresh dead specimen was found at RM 51 among 4 sites sampled in 2001 (Hubbs 2002, p. 5; Butler 2007, p. 65). A single live and a fresh dead snuffbox were found at a site in Giles County during qualitative sampling events at five sites in 2005 (Ahlstedt
et al.
2006). Ford (2008, pers. comm.) reported collecting a fresh dead specimen at Stairstep Shoals in Giles County, Tennessee, in July 2007. The small snuffbox population has recently recruited and exhibits some level of viability, and its numbers appear relatively stable in recent history.

Duck River
—The Duck River is the downstream-most large tributary of the Tennessee River draining south-central Tennessee and flowing

[Text truncated at 120,000 characters. The full text is on the page linked above.]

---

Source: Frix Law Library, https://www.frixlaw.com/law-library/documents/fr%3A2012-2940. Public record. Not legal advice.