J. N. Am. Benthol. Soc., 2002, 21(1):76±88 q 2002 by The North American Benthological Society Life history and propagation of the endangered fanshell pearlymussel, Cyprogenia stegaria Ra®nesque (Bivalvia:Unionidae) JESS W. J ONES1 AND RICHARD J. NEVES Virginia Cooperative Fish and Wildlife Research Unit2, US Geological Survey, Department of Fisheries and Wildlife Sciences, Virginia Polytechnic Institute and State University, Blacksburg, Virginia 24061 USA Abstract. Aspects of the reproduction, age, growth, ®sh hosts, and culture of juveniles were de- termined for the endangered fanshell pearlymussel, Cyprogenia stegaria Ra®nesque, 1820, in the Clinch River, Tennessee. Glochidia of C. stegaria are contained in red, worm-like conglutinates that resemble oligochaetes. Conglutinates are 20 to 80 mm long and are released through the excurrent aperture. Estimated fecundity was 22,357 to 63,459 glochidia/mussel. Eighty-four valves of C. stegaria were thin-sectioned for aging; ages ranged from 6 to 26 y. Of 16 ®sh species tested, 9 hosts were identi®ed through induced infestations of glochidia: mottled sculpin (Cottus bairdi), banded sculpin (Cottus carolinae), greenside darter (Etheostoma blennioides), snubnose darter (Etheostoma simoterum), banded darter (Etheostoma zonale), tangerine darter (Percina aurantiaca), blotchside logperch (Percina burtoni), logperch (Percina caprodes), and Roanoke darter (Percina roanoka). Newly metamorphosed juveniles were cultured in recirculating and nonrecirculating aquaculture systems within dishes containing sediments of 300 to 500 mm diameter (sand) or ,105 mm diameter (silt), and fed either the green algae Neochloris oleoabundans or Scenedesmus quadricauda daily. Growth and survival of juvenile mussels were highest in the nonrecirculating aquaculture system, with a mean survival of 72% after 2 wk and 38% after 4 wk. Key words: fanshell pearlymussel (Cyprogenia stegaria), endangered, life history, ®sh hosts, repro- duction, propagation. The fanshell pearlymussel, Cyprogenia stegaria An estimated 760 km of rivers throughout the Ra®nesque, 1820, is endemic to the Ohio, Cum- United States currently contain populations of berland, and Tennessee river drainages (Ort- C. stegaria, which represents ,10% of its historic mann 1918, 1919). Because of severe declines in range (R. Biggins, USFWS, Asheville, North populations, the species was listed as endan- Carolina, unpublished data). River alterations gered by the US Fish and Wildlife Service by dams, channel dredging, sand and gravel (USFWS) in 1990. The fanshell inhabits medium mining, coal mining, and sewage wastes have to large rivers on shoals of coarse gravel and caused drastic declines in populations of C. ste- sand. At present, reproducing populations of garia in the 20th century. Even in reproducing this freshwater mussel are known from only 3 populations, such as the Clinch River, declines rivers: the upper Clinch River in Tennessee and of this species may still be occurring (Ahlstedt Virginia, and the Green and Licking rivers in 1991). Kentucky. A small reproducing population may This species was reported by Ortmann (1919) exist in the Tennessee River below Pickwick to be a long-term brooder; however, little else is Landing Dam (Parmalee and Bogan 1998). Oth- known of its life history. The purpose of our er presumed relict populations may persist in research was to provide needed biological data the Muskingum and Walhonding rivers in Ohio, speci®ed in the federal recovery plan for this Kanawha River in West Virginia, Wabash River species (USFWS 1991). Speci®cally, C. stegaria in Illinois and Indiana, Barren River and Ty- life history, propagation, and culture were stud- garts Creek in Kentucky, and the Tennessee and ied so this information could be applied to the Cumberland rivers in Tennessee (USFWS 1991). restoration and recovery of the species. 1 E-mail address: [email protected] Methods 2 The Unit is supported jointly by the US Geological Fecundity Survey, Virginia Department of Game and Inland Fisheries, Wildlife Management Institute, and Virginia We estimated fecundity of 4 female C. stegaria Polytechnic Institute and State University. from the number of mature glochidia in gravid 76 2002] LIFE HISTORY OF FANSHELL PEARLYMUSSEL 77 FIG. 1. Conglutinates of the fanshell pearlymussel, Cyprogenia stegaria. FIG. 2. A.ÐConglutinates of the fanshell pearlymussel, Cyprogenia stegaria, coiled and contained in the marsupium of the outer gill. B.ÐMarsupium attached to the base of the inner lamina of the outer gill. C.Ð Inner and outer gills (left). 78 J. W. JONES AND R. J. NEVES [Volume 21 females. We used the end of a hypodermic nee- fore, the average number of glochidia/mm of dle to unravel and expose the distal ends of in- conglutinate may be higher than average. Our dividual conglutinates. The red conglutinates of selection of samples from the tip, middle, and C. stegaria have the texture, shape, and color of end of the viable conglutinate probably under- oligochaetes (Fig. 1). Once a tip was exposed, estimated the number of glochidia contained in forceps were used to pull the conglutinate free larger conglutinates. For example, our method from the gill of the mussel. Each conglutinate sampled only 50% of a section of viable conglu- was examined with a dissecting microscope to tinate 60 mm in length. Furthermore, 20 mm of determine where mature glochidia were con- the sample came from sections of the congluti- tained. The distal end of a conglutinate was gen- nate containing considerably fewer glochidia erally comprised almost entirely of mature glo- than sections near the distal end or tip of the chidia, whereas the basal end (the end attached conglutinate. Although the most accurate esti- to the gill) was comprised almost entirely of mate would be obtained from counting all glo- eggs or developing embryos. Therefore, we vi- chidia in viable conglutinates, total counts were sually determined the point near the basal end not practical for this endangered species. of each conglutinate where the remaining ma- terial was ;90 to 100% embryos, usually ;60 Age and growth to 75% of the length from the distal end of a conglutinate. We considered the segment that Relic valves, fresh dead shells, and live C. ste- contained mostly mature glochidia to be the vi- garia were collected from various Clinch River able portion of the conglutinate. The total length kilometer (CRKM) locations between Horton and the viable length of each conglutinate were Ford (CRKM 321) and Swan Island (CRKM measured with calipers. Three conglutinates per 277). Live C. stegaria were measured for length mussel were selected; each conglutinate had a and either returned immediately to the site of length with viable glochidia $30 mm. The seg- collection or, if gravid, collected for juvenile ment of viable conglutinate was cut into three mussel propagation. Measurements of 16 live 10-mm long pieces (for ease of handling) from mussels were obtained from quadrat sampling the distal to the basal end. Under a microscope, conducted by Ahlstedt and Tuberville (1997), conglutinates were pinched and teased apart and 39 specimens from collecting trips from with tweezers to free the glochidia. This tech- 1998 to 2001. Relic valves and fresh dead shells nique damaged some glochidia but released of various lengths were collected to best repre- thousands unharmed. Released glochidia were sent the size range of the population in the river. enumerated to obtain a total count for each 10- We used relic valves if the periostracum and ex- mm long segment. Usually 100 to 300 glochidia ternal growth rings of collected specimens were could not be removed from the conglutinate, so intact. Valves of C. stegaria contain relatively dis- they were counted or visually estimated to add tinct external growth rings, which were enu- to the total of each 10-mm long segment. The merated visually. We thin-sectioned 84 valves, total number of glochidia from nine 10-mm long following procedures described by Clark (1980) segments of conglutinate per mussel was used and Neves and Moyer (1988), using a Buehler to obtain an average number of glochidia/seg- Isomet low-speed saw unit with a diamond-im- ment. We divided this number by 10 to obtain pregnated blade (Buehler Ltd., Evanston, Illi- the average number of glochidia per mm of con- nois). Valves were cut from the center of the glutinate. This number was multiplied by the to- umbo to the ventral shell margin. Cut valves tal viable length of conglutinate per female mus- were glued (2-Ton Clear Epoxy, Riviera Beach, sel to calculate the number of mature glochidia Florida) to petrographic microslides (27 3 46 in each gravid female. Mean number of glochid- mm), then vacuum-sealed into a petrographic ia per female mussel also was calculated. chuck attached to the cutting arm of the saw, Our estimates of fecundity were only approx- and sectioned at a thickness of 280 mm (Neves imate because we neither randomly selected and Moyer 1988). Because shells , age6were conglutinates, nor randomly selected 10-mm rare in the river, we obtained lengths at ages 1 sections of sampled conglutinates. We used the to 5 y by back-calculating length-at-age based larger conglutinates that contained more glo- on internal annuli of older valves (n 5 5) chidia/mm than smaller conglutinates; there- (Bruenderman and Neves 1993). Thin-sections 2002] LIFE HISTORY OF FANSHELL PEARLYMUSSEL 79 of valves were examined under 43 magni®ca- phoned every 1 to 2 d. The collection of mobile tion, and internal growth lines were considered juveniles indicated a ®sh species was capable of true annuli if they were continuous from the transforming glochidia to juveniles, and was a umbo region to the outer surface of the valve. potential host of C. stegaria. Thin-sections and external growth rings of valves were examined separately by 3 experi- Culture of juveniles enced technicians to obtain consensus on the mean number of internal and external growth Our recirculating aquaculture system for ju- lines per valve.