Host Fishes and Reproductive Biolo K Y of 6 Freshwater Mussel Species From

J. N. Am. Benthol. Sm., 1997, 16(3):57&585 0 1997 by The North Amerxan Benthologxal ‘Society

Host fishes and reproductive biolo y of 6 freshwater mussel species from the MobiK e Basin, USA

W ENDELL R. HAAG AND M ELVIN L. WARREN, JR US Department of Agriculture, Forest Service, Southern Research Station, Forest Hydrology Laboratory, 1000 Front Street, Oxford, Mississippi 38655 USA

Abstract. Host fishes were identified for 6 species of freshwater mussels (Unionidae) from the Black Warrior River drainage, Mobile Basin, USA: Stropkitus subwxus, Pleurohemafurvum, Ptyckobran- thus greeni, Lampsilis yrrovnlis, Medionidus arutissimus, and Villosa nebulosn. Hosts were determined as those that produced juvenile mussels from glochidial infestations in the laboratory. The following mussel-fish-host relationships were established: Stropkitus subwxus with 10 species including Cy- prinidae, Catostomidae, Fundulidae, Centrarchidae, and Percidae; Pleurobemafurvum with Campostoma oligolepis, Cyprirlella callistia, C. uenusta, Srmotilus atromaculatus, and Fund&s olimceus; Ptyckobrunckus greeni with Etlleostoma brllator, E. doughs-i, Per&a nigroj&iata, and Percina sp. cf. caprodes; Lampsilis pen&is with Micropterus coosae, M. puuctulatus, and M. salmoides; Medionidus acutissimus with Fundulus olivuceus, Etkeostoma doughsi, E. wkipplei, Percina nigrqfusciata, and Percinn sp. cf. caprodes; and Villosa wbulosa with LqOnlis megul$is, Micropterus coosae, M. punctulutus, and M. sulmoides. Funduhs olhuceus served as host for 3 species and carried glochidia for long periods for 2 other species, suggesting that topminnows may serve as hosts for a wide variety of otherwise host-specialist mussel species. Host relationships for the species tested are similar to congeners. Methods of glochidial release, putative methods of host-fish attraction, and gravid periods are described for the 6 species. Kq rmrds: Mollusca, Bivalvia, Unionidae, mussels, host fish, glochidia, life history, conglutinates.

Freshwater bivalves of the superfamily assemblage (Mathiak 1979, Smith 1985). Con- Unionacea have a highly specialized life history servation efforts, whether centered on captive in which the modified veliger larvae (glochidia) propagation and reintroduction, relocation, or undergo a brief period as obligate ectoparasites habitat improvement, require comprehensive on the gills, fins, or other external parts of fish. host information. Glochidia are brooded in the gills of female Knowledge of host fishes is lacking or incom- mussels until mature, then released through the plete for many North American mussels. In par- siphons singly or in clusters called conglutinates ticular, there is a conspicuous dearth of infor- (Kat 1984). If the glochidia encounter a suitable mation for species endemic to Atlantic and Gulf host, they encyst for a few days to several Slope drainages, including the Mobile Basin. weeks, metamorphose into juvenile mussels, Most species endemic to the Mobile Basin have and then drop off the fish to assume a benthic experienced dramatic range reductions in the lifestyle. Glochidia encountering an unsuitable last 50 y (Stansbery 1976, Williams et al. 1992). host are rejected by the fish immune system, Several Mobile Basin species have become ex- usually within a few days. Host specificity var- tinct in this century, and 17 species are listed ies greatly among mussel species. Some species federally (USA) as endangered or threatened are able to parasitize a taxonomically wide va- (US FWS 1994). Host fish information is needed riety of fish species (Trdan and Hoeh 1982), but urgently for conservation and recovery of the re- others can use only a few, usually closely relat- maining Mobile Basin mussel fauna. ed, species (Zale and Neves 1982, Yeager and In this paper we describe determination of Saylor 1995). hosts, report observations on glochidial release Knowledge of fish hosts is essential in under- and putative host-attracting behavior, and re- standing patterns of distribution and abundance port periods of gravidity for Lampsilis perovalis, of mussels (Watters 1992) and in developing ef- Medionidus acutissimus, Pleurobema furvum, Pty- fective conservation programs. Lack of recruit- chobranchus greeni, Strophitus subvexus, and Vil- ment in some mussel populations has been at- losa nebulosa. With the exception of S. subvexus, tributed to unavailability of proper host fishes all are endemic to the Mobile Basin. Strophitus caused by human-induced changes in the fish subvexus occurs in Gulf Slope drainages from

19971 H OST FISHES OF UNIONIDS

east Texas to Florida (Williams and Butler 1994). &dial infestations in laboratory trials and I Although Zale and Neves (1982) and Neves et itoring the rejection of glochidia or produ al. (1985) provided host-fish data for V. nebulosa of juvenile mussels (Zale and Neves 1982: from the Tennessee River system, these popu- each mussel species, replicate trials were ru lations are now referred to as V. iris (Williams ing glochidia from 2-3 gravid females. Glo et al. 1993, Yeager et al. 1994). Pleurobemafirvum ia from each species were exposed to 14-2( and P. greeni are listed as endangered, and L. species. Fish species were chosen to repri perovalis and M. acutissimus are listed as threat- most families and genera and all common ened by the US Fish and Wildlife Service (US ties present at sites inhabited by the muss1 FWS 1994). Strophitus qubvexus and V. nebulosa Study streams were searched for gravi are considered of special concern and threat- male mussels from February to late June, ened, respectively, by the American Fisheries and water temperature was recorded on Society (Williams et al. 1993). These species rep- sample date. Gravid mussels were broughl resent a broad cross-section of unionid diversity the laboratory and placed in individual be,

including the subfamily Anodontinae (S. subvex- at room temperature. Observations were I us), the tribe Pleurobemini (P:furvum), and the of the behavior of gravid females and the r lampsiline clade (L. perovalis, M. acutissimus, P. od of glochidial release. Lampsilis peroval, greeni, and V. nebulosg) (sensu Lydeard et al. furvum, and P. greeni released congluti 1996). within several days of exposure to room perature. Medionidus acutissimus, S. subwxus Methods V. nebulosa did not release glochidia in the oratory. Glochidia were obtained from Mussels and fishes were collected mainly species by flushing water through the gills from tributaries of the Black Warrior River in a hypodermic syringe. All females werl Lawrence and Winston counties, northwestern turned to their stream of origin after yie: Alabama. Mussels were collected from Brushy glochidia because of the threatened or er and Flannagin creeks and the Sipsey Fork Black gered status of some of the species. Reprl Warrior River. These streams are within William tative shells of each species are deposited 2 B. Bankhead National Forest and harbor diverse, Ohio State University Museum of Biologic; remnant examples of the Mobile Basin upland versity, Columbus, Ohio, and the Missis mussel fauna. Fishes were collected from Clear Museum of Natural History, Jackson, Missi and Hubbard creeks and the Sipsey Fork. Ad- pi. Glochidia were considered mature and ditional fishes were collected from small to parasitize fish if they snapped shut aftc streams in Lafayette and Panola counties, north- posure to synthetic aquarium salt. Immedi em Mississippi. Ictalurus punctatus and Microp- prior to trials, conglutinates or other agg tens salmoides were obtained from hatchery tions of glochidia were teased apart with 2 stock. probe and suspended in water in a small bc When possible, potential host fishes were col- Potential host fishes were anaesthetizel lected from streams without mussels or streams dividually in tricaine methanesulfonate an with low mussel densities in order to avoid us- spected for existing glochidial infestations. ing fish with existing glochidial infestations or 1 individual was found with a natural iru acquired immunity to glochidia (Zale and Neves tion (Cyprinella callistia with 1 glochidiu 1982). Some fish species could not be obtained Strophitus subvexus), and this fish was not from streams other than those with high mussel in trials. A suspension of glochidia (0.5-1.5 . densities (e.g., Etheostoma douglasi, Percina shu- was applied with a pipette directly onto mardi). These species were collected in early right gill of anaesthetized fishes until >lC February before natural infestations normally chidia could be seen attached to the outer occur in the study streams (W. R. Haag and M. face of the outer gill. After recovery fron L. Warren, unpublished data) and held at room aesthetic, all individuals of a fish species temperature (22-25°C) for at least 4 wk before placed in aquaria (6-25 L) with aeration being used in host trials to ensure that any glo- temperature-controlled environmental chal chidia from natural infestations had excysted. The first 2 trials for Lampsilis peroznlis and S Host fishes were determined by inducing glo- vexus were run at a mean temperature of 578 W. R. HAAG AND M. L. WARREN [Volume 16

(range 16-18°C). All other trials were run at a thetized and inspected for encysted glochidia mean temperature of 20°C (range 19-21°C). on an irregular basis until all fish were clean. For most trials, a subsample of individuals of Only fish that produced active juveniles were each fish species was anaesthetized l-2 d after considered hosts. Fishes are listed in tables in infestation to determine if rejection of glochidia phylogenetic sequence by family (Mayden et al. had occurred. If no glochidia were found on 1992) and alphabetically within families. gills of subsampled fish, all other individuals of that species were inspected, and the trial was Results terminated when all individuals had rejected all glochidia. If subsampled individuals carried Females of Medionidus acutissimus and Strophi- glochidia, they were revived, returned to the tus subwxus were found gravid with mature glo- aquaria, and the trial was continued. Fish spe- chidia from 26 February to 12 March 1996 (wa- cies that rejected all glochidia and produced no ter temperatures 8-13°C). Lampsilis perovalis and juvenile mussels were considered non-host spe- Villosa nebulosa were found gravid from 26 Feb- cies. For species that varied among individuals ruary to 2 April 1996 (water temperature 8- in time to rejection, a range of days to rejection 13°C). Ptychobranchus greeni were found gravid was established using the first day that clean on 16 April (water temperature 13”C), and Pleu- individuals were found and the day that the last robema firvum, from 18 June to 25 June 1996 clean individual was found. Forsome trials, fish (water temperature 25°C). were not inspected soon after infestation, and These species displayed striking diversity in the reported number of days to rejection for strategies used to infect host fishes. Glochidia of these trials probably overestimates the time to Strophitus subwxus were bound in a copious mu- rejection. cous matrix which, when released in the stream, After a species carried glochidia for ~10 d, may serve to tangle fish indiscriminately. Re- we inspected material on the bottom of the lease of glochidia in a loose mucous web has aquarium approximately every 3 d to search for been reported for at least 2 other anodontines, transformed juvenile mussels. Material was col- Anodonta cygnea (Wood 1974) and Anodontoides lected with a flexible hose attached to a vacuum ferussacianus (Hove et al. 1995). Pleurobema fur- pump, concentrated on a 100~km mesh sieve, vum and Ptychobranchus greeni released glochidia washed into a gridded petri dish, and examined in conglutinates that drift passively with the under a stereomicroscope. Juvenile mussels stream current, free from the gravid female, and were distinguished from rejected glochidia by may mimic food items of darters and minnows. the presence of a foot and mantle and were usu- Conglutinates of P. furvum were subcylindrical, ally crawling on the bottom of the petri dish. flattened, and peach-to-pink colored and similar Juvenile mussels were counted and preserved in to those described for species of Pkurobema and 70% EtOH. Inactive juveniles with a recogniza- Fusconaia (Bruenderman and Neves 1993, Hove ble foot and mantle were counted separately be- and Neves 1994). The conglutinates of P. greeni cause, although they had clearly undergone were round and pearl-colored with 2 black eye- some degree of metamorphosis as evidenced by spots and strongly resembled fertilized fish the presence of juvenile anatomical features, we eggs. This observation is in contrast to previous question if these individuals represented viable descriptions of Ptychobranchus conglutinates as offspring. From all confirmed host species, low elongate structures resembling newly hatched numbers of inactive juveniles were collected larval fish (Morrison 1973, species not given) or several days to a week before the appearance of dipteran larvae (Hartfield and Hartfield 1996, active juveniles, but we also collected inactive for P. greeni). Lampsilis perovalis released glochid- juveniles from species that later produced no ac- ia in a superconglutinate, a glochidial packet tive individuals. Hove and Neves (1994) de- pigmented and shaped like a small minnow and scribed empty valves identical to those of living tethered to the female, as previously described juveniles as premetamorphosed juveniles and (Haag et al. 1995). Medionidus acutissimus and similarly questioned whether or not these were Villosa nebulosa displayed modified mantle mar- indicative of a successful parasitization. If a fish gins with tentacles and flaps which may serve species carried glochidia for several weeks but to attract fishes to the gravid mussel. produced no juveniles, individuals were anaes- Glochidia of Strophitus subvexus metamor- 19971 H OST FISHES OF UNIONIDS 55

TABLE 1. Results of host trials for Strophitus subvexus. Letters AC represent replicate trials using glochid from 3 individual mussels. An asterisk denotes that all fish died before termination of the trial, and the numb in parentheses represents the maximum number of days before death that fish carried glochidia. (#) indicatl that only nonactive juveniles were produced from trial. A dash denotes that the fish species was not used the trial.

Mean number Number tested Days to transformation juveniles/fish

Fishes A B C A B C A B C Campostoma oligolepis 3 4 4 * (7) * (9) 14 0* o* <1 Cyprinella callistia 3 3 4 * (11) * (9) 0* 0* 1 Notropis asperifrons -2 - *:46) - 0” Semotilus atromuculatus 3 3 1 13* * (10) 14-24 1* 0” 4 Hypentelium etowanum 3 3 2 13-15* 7-15* 14-28 2* 1* 25 lctalurus punctatus 10 10 3 - * (10) 1&14# 0* O# Fund&s olivaceus 2 3 4 15* * (7) * (5) <:* o* o* Lepomis megalotis 4 $-- 13-21 13-21 3 2 - Micropterus coosae - 1 lo-l# - O# M. salmoides --5 5 4 *G * (7) 10-14 0* 0* 2 Etheostoma douglasi 5 4-13-15 13-28 1 10 - E. whipplei - 2 - 10 -

TABLE 2. Results of host trials for Pleurobema furvum. Letters A-C represent replicate trials using glochid from 3 individual mussels. An asterisk denotes that all fish died before termination of the trial. For non-hc species, a dagger (t) denotes that fish were not inspected before this day and rejection likely occurred befo the day indicated. A dash denotes that the fish species was not used in the trial.

Days to transformation (hosts) Mean number Number tested or rejection (non-hosts) juveniles / fish Fishes A B C A B C A B C Hosts Campostoma oligolqis 10 3 15 13-15 -

TABLE 3. Results of host trials for Ptychobranchus greeni. Letters A and B represent replicate trials using glochidia from 2 individual mussels. An asterisk denotes that all fish died before termination of the trial. For non-host fishes, a dagger (t) denotes that fish were not inspected before this day and rejection likely occurred before the day indicated. A dash denotes that the fish species was not used in the trial.

Days to transformation (hosts) Mean number Number tested or rejection (non-hosts) juveniles/fish

Fishes A B A B A B Hosts Etheostoma bell&or 1 30-34 - 5 E. douglasi 8 4 28-55 38-51 16 6 Percina nigrojasciata 6 4 26-55 32-40) 28 4* Percina sp.cf. caprodes - 1 48-54 - 9 Non-hosts Campostoma oligolepis 7 4 2-5 5t Cyprinella cullistia 8 ’ 2 2 5t Semotilus atromaculatus 5 3 2-5 5t Hypentelium etowanum . 5 - 1-2 - Ictalurus punctatus 3 5t Fundulus olivaceus 5 17* Lepomis macrochirus 4 2-5 - L. megalotis 2 2 2-5 7t Micropterus coosae 2 2-5 - M. punctulatus 2 - 2 - M. salmoides 5 5-6 Etheostoma whipplei 8 4 5-23 St Percina shumardi 3 >28 -

phosed into juvenile mussels on 10 species of phosed into juvenile mussels on 4 species of fish in 5 families, and results were equivocal for darters: Etheostoma bellator, E. douglasi, Percina ni- 3 other species including 1 additional family grofasciata, and Percina sp. cf. caprodes (Table 3). (Table 1). Hypentelium etouxmum and Etheostoma Two other darters, E. whipplei and P. shumardi, douglasi produced the highest number of juve- did not serve as hosts. One individual of E. niles. All other species produced low to mod- whipplei carried glochidia for 23 d, but all other erate numbers of juveniles, probably because individuals rejected glochidia in <8 d. Percina many fish died before completion of the trials. shumardi carried glochidia for at least 28 d, but Ictalurus punctatus and Micropterus coosae carried all individuals were clean after 53 d, and no ju- glochidia for an extended period and produced veniles were found. Similarly, one species of only inactive juveniles. Notropis asperifions car- topminnow, Fundulus olivaceus, carried glochidia ried glochidia for 6 d, but all fish died before for an extended period but died before termi- termination of the trial. nation of the trial and produced no juveniles. Glochidia of Pleurobema furvum metamor- Glochidia of Lampsilis perovalis metamorphosed into juvenile mussels on 1 species of phosed on 3 species of sunfishes (Centrarchi- topminnow, Fund&s olivaceus, and 4 species of dae): Micropterus coosae, M. punctulatus, and M. minnows: Campostoma oligolepis, Cyprinella callis- salmoides (Table 4). Two other sunfishes, Lepomis tia, C. venusta, and Semotilus atromaculatus (Table macrochirus and L. megalotis, and one topminnow 2). Campostoma oligolepis and F. olivaceus pro- (Fundulidae), Fundulus olizxzceus, carried glochid- duced low numbers of juveniles and C. oligolepis ia for extended periods but produced no juve- produced juveniles in less time than other host niles. One specimen of pickerel (Esocidae), Esox species. Three other cyprinids, Lythrurus bellus, niger, died after 4 d but carried a heavy infes- Notropis asperifrons, and N. stilbius, quickly re- tation at the time of death. jected glochidia. Glochidia of Medionidus acutissimus metamor- Glochidia of Ptychobrunchus areeni metamor- vhosed into iuvenile mussels on 1 sn~ci~s nf 19971 H OST FISHES OF UNIONIDS

TABLE 4. Results of host trials for Lampsilis peroznlis. Letters A-C represent replicate trials using glo& from 3 individual mussels. An asterisk denotes that all fish died before termination of the trial. For non- species, a dagger (t) denotes that fish were not inspected before this day and rejection likely occurred be the day indicated. A dash denotes that the fish species was not tested in the trial.

Days to transformation (hosts) Mean number Number tested or rejection (non-hosts) juveniles / fish

Fishes A B C A B C A B ( Hosts Micropterus coosae 2 46-53 38-59 2134 28 72 M. punctulatus 1 * 21-47 - * 3; M. salmoides 5 * * 2649 * * 1‘* Non-hosts Campos toma oligolepis 5 4 - 11t 13t Cyprinella callistia 6 5 4t 4t C. venusta 3 2 14t 13t Notropis asperifrons 5 4 3 2-3 Semotilus atromaculatus 6 5 1 6t 13t Hypentelium etoxanum 5 4 l-2 15t lctalurus punctatus 5 5 1st 17t Fund&s olivaceus 3 4 2 4-39t 19t Ambloplites ariommus 1 - Lepomis cyanellus - 2 - L. macrochirus 6 2 7-9 17-25 L. megalotis 4 3 7-20 12 Etheostoma douglasi 5 14t 16t E. whipplei 3 19t 1st Percina nigrofasciata 4 14t 16t l? shumardi 1 19t 16t Percina sp. cf. caprodes 1 19t 1st

topminnow, Fund&s olizxxceus, and 4 species of odontinae) was a host generalist, using a t; darters: Etheostoma douglasi, E. whipplei, Per&a nomically wide array of fish species. The o nigrofasciata, and Percina sp. cf. caprodes (Table 5). 5 species were host specialists for which 1 Fundulus oliznzceus produced low numbers of ju- use largely followed generic or familial li veniles in less time than on other hosts. Pleurobema furvum (tribe Pleurobemini) 1: Glochidia of Villosa nebulosa metamorphosed minnows (Cyprinidae) and topminnows. Y on 4 species of sunfishes: Lepomis megalotis, Mi- species in the lampsiline clade, Lampsilis pew cropterus coosae, M. punctulatus, and M. salmoides and Villosa nebulosa, used sunfishes as h (Table 6). Lepomis megalotis produced fewer ju- (Centrarchidae, primarily basses, Micropt veniles in less time than Micropterus in one trial, spp.). The chain pickerel (Esox niger) may but rejected all glochidia in the other. Two other serve as a host for L. perovalis; specimens sunfishes, Lepomis cyanellus and L. macrochirus, infected in the wild (W. R. Haag and M. L. \ carried glochidia for an extended period, but L. ren, unpublished data), but we were unabl cyanellus produced no juveniles, and L. macro- confirm this species as host due to death of chirus only inactive juveniles. One species of fishes. Two other species in the lampsiline cl topminnow, Fundulus olivaceus, carried glochidia Medionidus acutissimus and Ptychobranchus grc for 14 d but produced no juveniles. used darters (Percidae) and topminnows (I dulidae). Discussion Patterns of host-fish use for these 6 species consistent with host information for conge: The 6 species of mussels in this study varied from other drainages. Strophitus undulatus, in host use. Strophitus subvexus (subfamily An- tive to the Mississippi River drainage and 580 W. R. HAAG AND M. L. WARREN [Volume 16

Days to transformation (hosts) Mean number Number tested or rejection (non-hosts) juveniles/fish

Fishes A B A B A B Hosts Etheostoma bell&or 1 30-34 - 5 E. douglasi 8 4 28-55 38-51 16 6 Percina nigrofasciata 6 4 26-55 32-40) 28 4* Percinasp. cf. caprodes - 1 48-54 - 9 Non-hosts Campostoma oligolepis 7 4 2-5 5t Cyprinella cullistia 8 ’ 2 2 5t Semotilus atromaculatus 5 3 2-5 5t Hypentelium etowanum -. 5 - 1-2 - lctalurus punctatus 3 5t Fundulus olivaceus 5 17* Lepomis macrochirus 4 2-5 - L. megalotis 2 2 2-5 7t Micropterus coosae 2 2-5 - M. punctulatus 2 - 2 - M. salmoides 5 5-6 Etheostoma whipplei 8 4 5-23 St Percinn shumardi 3 >28 -

phosed into juvenile mussels on 10 species of phosed into juvenile mussels on 4 species of fish in 5 families, and results were equivocal for darters: Etheostoma bellator, E. douglasi, Percina ni- 3 other species including 1 additional family grofasciata, and Percina sp. cf. cuprodes (Table 3). (Table 1). Hypentelium etowanum and Etheostoma Two other darters, E. whipplei and P. shumardi, douglasi produced the highest number of juve- did not serve as hosts. One individual of E. niles. All other species produced low to mod- whipplei carried glochidia for 23 d, but all other erate numbers of juveniles, probably because individuals rejected glochidia in <8 d. Percina many fish died before completion of the trials. shumardi carried glochidia for at least 28 d, but Ictalurus punctatus and Micropterus coosae carried all individuals were clean after 53 d, and no ju- glochidia for an extended period and produced veniles were found. Similarly, one species of only inactive juveniles. Notropis asperifions car- topminnow, Fundulus olivaceus, carried glochidia ried glochidia for 6 d, but all fish died before for an extended period but died before termi- termination of the trial. nation of the trial and produced no juveniles. Glochidia of Pleurobema furvum metamor- Glochidia of Lampsilis perovalis metamorphosed into juvenile mussels on 1 species of phosed on 3 species of sunfishes (Centrarchi- topminnow, Fundulus olivaceus, and 4 species of dae): Micropterus coosae, M. punctulatus, and M. minnows: Campostoma oligolepis, Cyprinella callis- salmoides (Table 4). Two other sunfishes, Lepomis tia, C. venusta, and Semotilus atromaculatus (Table macrochirus and L. megalotis, and one topminnow 2). Campostoma oligolepis and F. olivaceus pro- (Fundulidae), Fundulus olivaceus, carried glochid- duced low numbers of juveniles and C. oligolepis ia for extended periods but produced no juve- produced juveniles in less time than other host niles. One specimen of pickerel (Esocidae), Esox species. Three other cyprinids, Lythrurus bellus, niger, died after 4 d but carried a heavy infes- Notropis asperifrons, and N. stilbius, quickly re- tation at the time of death. jected glochidia. Glochidia of Medionidus acutissimus metamor- Glochidia of Ptuchobrunchus areeni metamor- vhosed into iuvenile mussels on 1 sn~ci~s nf 19971 H OST FISHES OF UNIONIDS

TABLE 6. Results of host trials for Villosa nebulosa. Letters A and B represent replicate trials using glod from 2 individual mussels. An asterisk denotes that all fish died before termination of trial. (#) indicates only inactive juveniles were produced from the trial. A dash denotes that the fish species was not used ir trial.

Days to transformation (hosts) Mean number Number tested or rejection (non-hosts) juveniles/fish

Fishes A B A B A B Hosts Lepomis megalot is 3 7 21-23 2 0 Micropterus coosae 2 1 23-39 22-37 16 7 M. punctulatus 1 2 25-32 22-31” 5 6* M. salmoides 3 2246 - 178 Non-hosts Campostoma oligolepis 4 6 2-7 2-7 Cyprinella callistia 4 2 - C. camura 5 2 Notropis asperifions 2 2 2 1 Semotilus atromaculatus ._ 3 1 3-7 2 Hypentelium etouxnum 3 3 Fund&s olivaceus 6 14 Lcpomis cyunellus - 1 - 14 L. macrochirus 3# 2 7-16 24 L. megalotis 3 7 2-14 Etheostoma bell&or 1 3 E. douglusi 4 3-7 - E. whipplei 4 5 7 4 Percina nigrofasciata 4 5 7 P7 Percinasp. cf. cuprodes 1 1 3 2 serve as hosts for many, unrelated mussel spe- work is needed to determine the significant cies. Fishes of the families Fundulidae and Poe- the wild of these marginally suitable hosts. ciliidae may act as “universal hosts” for many Most fishes identified as hosts gave consir otherwise host-specific mussel species. Howev- results among trials and produced modera er, in at least some cases, these species may be high numbers of juveniles. Further, in I only marginally suitable hosts, and in nature, cases, host suitability for a mussel species the surface-feeding mode of most cyprinodon- similar among congeneric fish species. Ar tiform fishes may render them less vulnerable ception involved the use of darters as host to infestation by host-attracting mechanisms tar- Ptychobranchus greeni. Two species each of geted at fishes in benthic or mid-water feeding eostoma and Percina produced juvenile mus guilds. but E. whipplei and P. shumardi failed to proc Other fishes identified as hosts in this study juveniles, (Table 4). Per&a shumardi carried gave equivocal results indicating that these spe- chidia for >4 wk but produced no juveniles cies also may be only marginally suitable hosts. may serve as a marginal host for P. greeni. These species produced consistently low num- eostoma whipplei rejected all glochidia in <8 bers of juveniles, gave inconsistent results 2 trials with the exception of 1 individual among replicate trials, produced only inactive carried for 23 d. This species is clearly n juveniles, and/ or produced juveniles in less suitable host for P. greeni. In contrast, glocl time than other hosts. The sunfish Lepomis mega- of Medionidus acutissimus metamorphosed o lotis and possibly L. macrochirus appear to be species of darters tested. marginal hosts for Lampsilis peroualis and Villosa One of the most important niche dimem nebulosa; Campostoma oligolepis appears margin- for freshwater mussels may be host-fish USE ally suitable for Pleurobema furvum. Further host-specific mussel species, methods for in 582 W. R. HAAG AND M. L. W ARREN [Volume 16

TABLE 5. Results of host trials for Medionidus acutissimus. Letters A and B represent replicate trials using glochidia from 2 individual mussels. An asterisk denotes that all fish died before termination of the trial, and the number in parentheses represents the maximum number of days before death that fish carried glochidia. For non-host fishes, a dagger (t) denotes that fish were not inspected before this day and rejection likely occurred before the day indicated. A dash denotes that the fish species was not tested in the trial.

Days to transformation (hosts) Mean number Number tested or rejection (non-hosts) juveniles/fish

Fishes A B A B A B Hosts Fundulus olivaceus 5 3 * (8) 12-14 * 1 Etheostoma douglasi 5 36-53 5 E. whipplei 7 3 29-53 25-39 30 12 Percina nigrofasciata 7 4 19-53 25-35 4 6 Per&a sp.cf. caprodes 2 - 4043 5 Non-hosts Campostoma oligokpis 4 * 4t 1 Cyprinella callistia 4 4t 1 Notropis asperifrons 2 4t 1 Semotilus atromaculatus 5 4t Hypentelium etavanum 3, 4t 3 lctalurus punctatus 3 4t - Lepomis macrochirus 2 - 4t L. megalotis 4 2 4t 4-8 Micropterus coosae 1 1 4t 4 M. salmoides 5 1 4-8 3

era1 Atlantic Slope drainages, can use as hosts ceus produced few or no juvenile mussels/fish at least 11 species of fishes representing four and in less time than other hosts, it may be only families (Hove et al. 1997). Several other species a marginally suitable host for these mussels. in the subfamily Anodontinae are also host gen- However, other members of the order Cyprino- eralists (Trdan and Hoeh 1982, Watters 1994, dontiformes apparently serve as hosts for a va- Barnhart et al. 1997) although the subfamily riety of mussel species. Fundulus diuphanus is im- does contain some specialists (Watters 1994, Mi- plicated as host for 6 genera (Watters 1994). chaelson and Neves 1995). The majority of hosts Gambusia uffinis (Poeciliidae) served as host for for Pleurobema (Pleurobemini) are minnows Villosa iris (Neves et al. 1985, as V. nebulosa), 3 (Yokely 1972, Weaver et al. 1991, Hove and Ne- species of Anodontinae (Watters 1994), and car- ves 1994). Most species in the lampsiline clade ried glochidia of Pleurobema collina for 45 d but use either sunfishes or darters as hosts. All spe- produced no juveniles (Hove and Neves 1994). cies of Lampsilis for which hosts are identified Gambusia holbrooki serves as host for at least 9 can use sunfishes, especially basses (Micropterus mussel species in 6 genera in Atlantic and Gulf spp.), although some species also can use large Slope drainages (C. O’Brien, National Biological piscivorous percids (Perca and Stizostedion) or Service, personal communication). Poecilia reti- esocids (Esox) (Zale and Neves 1982, Waller and culata (as Lebistes reticulata, Poeciliidae), a cy- Holland-Bartels 1988). Other species of Medion- prinidontiform not native to North America, idus and Ptychobranchus use darters as hosts served as host for Lasmigona compressa (Tompa (Zale and Neves 1982, Barnhart et al. 1997). 1979) and serves as host for at least 5 species in Fundulus olizwceus served as host for 3 species 5 genera in Atlantic and Gulf Slope drainages (Medionidus acutissimus, Pleurobema fu~vum, and (C. O’Brien, personal communication). Al- Strophitus subwxus) and carried glochidia of 2 though it is well established that some mussel others (Lampsilis perovalis and Ptychobranchus species are generalists in their host-fish use, lit- greeni) for an extended period. Because F. olive- tle is known about fishes that may be able to 19971 H OST FISHES OF UNIONIDS LE

freshwater mussels. Bulletin of the American Reproductive biology and fish hosts of the 1 Malacological Union, Inc. 1973:4546. nessee clubshell Pleurobemu omforme (Mollu:

NEVES, R. J., L. R. WEAVER, AND A. V ZALE. 1985. An Unionidae) in Virginia. American Midland i’ evaluation of host fish suitability for glochidia of uralist 126:82-89. Villosa vanuxemi and V. nebulosu (Pelecypoda: W ILLIAMS, J. D., AND R. S. BUTLER. 1994. Freshw, Unionidae). American Midland Naturalist 113:13- bivalves. Pages 53-128 in M. D. Deyrup and 19. Franz (editors). Endangered biota of Florida, SMITH, D. G. 1985. Recent range expansion of the IV Invertebrates. University Press of Flor freshwater mussel Anodonta implicata and its re- Gainesville, Florida. lationship to clupeid fish restoration in the Con- W ILLIAMS, J. D., S. L. H. FULLER, AND R. GRACE. 15 necticut River system. Freshwater Invertebrate Bi- Effects of impoundment on freshwater mus! ology 4:105-108. (Mollusca:Bivalvia:Unionidae) in the main ch STANSBERY, D. H. 1976. Naiad mollusks. Pages 42-52 nel of the Black Warrior and Tombigbee river! in H. Boschung (editor). Endangered and threat- western Alabama. Bulletin of the Alabama b ened plants and animals of Alabama. Bulletin of seum of Natural History 13:1-10. the Alabama Museum of Natural History 2:1-92. WILLIAMS, J. D., M. L. WARREN, K. S. CUMMINGS, J TOMPA, A. S. 1979. Life cycle completion of the fresh- HARRIS, AND R. J. NEVES. 1993. Conservation ! water clam Lawnigonu compressa ‘in an experimen- tus of freshwater mussels of the United States : tal host, Lebistes reticulutus. The Veliger 22:188- Canada. Fisheries 18(9):6-22. 190. -_ WOOD, E. M. 1974. Some mechanisms involved in h recognition and attachment of the glochidium TRDAN, R. J., AND W. R. HOEH. 1982. Eurytopic host use by two congeneric species of freshwater mus- va of Anodontu cygneu (Mollusca:Bivalvia). Jour of Zoology, London 173:15-30. sel (l’elecypoda:Unionidae:Anodonta). American YEAGER, B. L., AND C. E SAYLOR. 1995. Fish hosts Midland Naturalist 108:381-388. four species of freshwater mussels (Pelecypo US FWS (US FISH AND W ILDLIFE SERVICE). 1994. En- Unionidae) in the upper Tennessee River dr; dangered and threatened wildlife and plants, 50 age. American Midland Naturalist 133:1-6. CFR 17.11 & 17.12. Division of Endangered Spe- YEAGER, M. M., D. S. CHERRY, AND R. J. NEVES. 19 cies, US Fish and Wildlife Service, Washington, Feeding and burrowing behaviors of juvei DC. rainbow mussels, Villosu iris (Bivalvia:Unionid, W ALLER, D. L., AND L. E. HOLLAND-BARTELS. 1988. Journal of the North American Benthological Fish hosts for glochidia of the endangered fresh- ciety 13:217-222. water mussel Lumpsilis higginsi Lea (Bivalvia: YOKELY, I’. 1972. Life history of Pleurobemu cordut Unionidae). Malacological Review 21:119-122. (Rafinesque 1820) (Bivalvia:Unionacea). Mala WATERS, G. T. 1992. Unionids and the species-area logia 11:351-364. curve. Journal of Biogeography 19:481490. ZALE, A. V., AND R. J. NEVES. 1982. Fish hosts of f( WATERS, G. T. 1994. An annotated bibliography of species of lampsiline mussels (Mollusca:Unic the reproduction and propagation of the Union- dae) in Big Moccasin Creek, Virginia. Canad oidea. Ohio Biological Survey Miscellaneous Con- Journal of Zoology 60:2535-2542. tributions No. 1. Received: 11 September 199f W EAVER, L. R., G. B. PARDUE, AND R. J. NEVES. 1991. Accepted: 26 December 199C 584 W. R. HAAG AND M. L. WARREN [Volume 16 ing hosts apparently have evolved to exploit unionids from the Ozark uplifts. Pages 16-20 in predator-prey relationships and feeding guilds K. S. Cummings, A. C. Buchanan, C. A. Mayer, of hosts. In contrast, methods of glochidial re- and T. Naimo (editors). Conservation and man- lease in host-generalist mussel species may agement of freshwater mussels II: initiatives for the future. Upper Mississippi River Conservation function to increase the likelihood of chance en- Committee, Rock Island, Illinois. counters with a wide variety of fishes, irrespec- BRUENDERMAN, S. A., AND R. J. NEVES. 1993. Life his- tive of feeding guild. Two extreme examples of tory of the endangered fine-rayed pigtoe Fuscon- this are illustrated by species in this study. aia cuneolus (Bivalvia:Unionidae) in the Clinch Lampsilis perowlis, a host specialist restricted to River, Virginia. American Malacological Bulletin use of basses (Micropterus spp.) and perhaps 10:83-91. pickerel (Esox spp.), releases its glochidia in a HAAG, W. R., R. S. BUTLER, AND I? D. HARTFIELD. 1995. minnow-shaped mass that may attract piscivo- An extraordinary reproductive strategy in fresh- rous fish and be too large to be ingested by water bivalves: prey mimicry to facilitate larval most fishes but bass or pickerel (Haag et al. dispersal. Freshwater Biology 34:471476. HARTFIELD, l?, AND E. HARTFIELD. 1996. Observations 1995). In contrast, glochidia of Strophitus subvex- on the conglutinates of Ptychbrunchus greeni us, a host generalist, are bound in a mucous ma- (Conrad 1834) (Mollusca:Bivalvia:Unionoidea). trix which may be released as a loose web to American Midland Naturalist 135:37&375. ensnare fish indiscriminate of feeding guild or HOVE, M. C. AND 6 OTHERS. 1997. Suitable fish hosts size. for glochidia of four freshwater mussels. Pages Knowledge of host fishes is crucial to under- 21-25 in K. S. Cummings, A. C. Buchanan, C. A. standing the ecology of freshwatkr mussels. Mayer, and T. Naimo (editors). Conservation and However, related factors such as the relationship management of freshwater mussels II: initiatives among modes of glochidial release, natural his- for the future. Upper Mississippi River Conser- tory attributes of host fishes, and host-fish den- vation Committee, Rock Island, Illinois. HOVE, M. C., R. A. E NGELKING, M. E. PETELER, E. M. sity may have important effects on mussel re- PETERSON. 1995. Anodontoides ferussaciunus and productive success. Further study of these re- Anodonta imbecillis host suitability tests. Triannual lationships may provide insight in explaining Unionid Report 6:22. Available from: US Fish and distribution and abundance patterns of mussels Wildlife Service, Asheville, North Carolina and in understanding causes for the precipitous H OVE, M. C., AND R. J. NEVES. 1994. Life history of decline of many mussel species during the past the endangered James spinymussel Pleurobema 50 years. collinu (Conrad, 1837) (Mollusca:Unionidae). American Malacological Bulletin 11:2940. Acknowledgements KAT, I’. 1984. Parasitism and the Unionicea (Bivalvia). Biological Reviews 59:189-207. We thank the following people for their help LYDEARD, C, M. MULVEY, AND G. M. DAVIS. 1996. Mo- in this project. Mahala Shillingsford gave tre- lecular systematics and evolution of reproductive mendous assistance in the laboratory. Mark traits of North American freshwater unionacean Hove and Richard Neves provided suggestions mussels (Mollusca:Bivalvia) as inferred from 16s and advice during the initial phases of the pro- rRNA gene sequences. Philosophical Transactions of ject. Gordon McWhirter helped with collection the Royal Society (London), Series B 351:1593-1603. MATHIAK, H. A. 1979. A river survey of the unionid of fishes. Paul Hartfield assisted in collection of mussels of Wisconsin, 1973-1977. Sand Shell gravid female mussels and coordinated admin- Press, Horicon, Wisconsin. istrative details with the US Fish and Wildlife MAYDEN, R. L., B. M. BURR, L. M. PAGE, AND R. R. Service. Jim Bartlett provided us with channel MILLER. 1992. The native freshwater fishes of catfish. Cathy Jenkins and Francis McEwen gave North America. Pages 827-863 in R. L. Mayden logistical support. This study was supported by (editor). Systematics, historical ecology, and the Southern Research Station and National North American freshwater fishes. Stanford Uni- Aquatic Monitoring Center, US Department of versity Press, Stanford, California. Agriculture, Forest Service. MICHAELSON, D. L., AND R. J. NEVES. 1995. Life history and habitat of the endangered dwarf wedgemus- se1 Alasmidontu heterodon (Bivalvia:Unionidae). Literature Cited Journal of the North American Benthological So-

BARNHART, M. C., A. D. ROBERTS, AND A. P. FARNS- ciety 14:324-340. WORTH. 1997. Reproduction and fish hosts of MORRISON, J. I? E. 1973. The families of the pearly