The subfamily Ambleminae is a diverse group of fresh-water bivalve mollusks found throughout North America east of the continental divide. They are also highly imperiled due to habitat specificity, limited ranges, and complex life history (Strayer et al. 2004). Many species require free-flowing medium to large rivers, a habitat type that is extensively impacted by human activity. In addition, the need for suitable host fish for the parasitic larval stage (the glochidium) contributes to mussels' vulnerability to habitat disturbance (Lydeard et al. 2004). As a result, 12.6% of the North American species and subspecies of Ambleminae are believed to be extinct, 22.9% are federally endangered or threatened, and many others are locally or globally rare (Turgeon et al. 1998).
Conservation efforts are hampered, however, by our limited understanding of their systematics. High levels of ecophenotypic plasticity may result in intraspecific variation in shell form exceeding intergeneric differences (Davis 1983). Thus, Roe (2000) used molecular sequences to demonstrate that several putative specimens of highly imperiled Ptychobranchus species actually represented other genera. Likewise, morphological identification of glochidia is often difficult, whereas molecular data readily identify them (White et al. 1996). This morphological variability has made both taxonomic definition and identification problematic, but both are essential to establishing conservation needs.
Well-supported phylogenetic hypotheses can provide valuable information for conservation. Many aspects of the biology of rare mussel species, such as host fish choice, habitat preferences, and breeding cycles, are often inferred based on better-studied common species (e.g., many species accounts in Parmalee & Bogan  have only tentative suggestions for these features, based on other species in the genus). The closest relatives of the species of interest are likely to be the best basis for such inference. If the species is misclassified, the inferences are less likely to be accurate. Also, incorrect synonymization may lead to the neglect of phylogenetically distinctive taxa. Incorrect assumptions may lead to inappropriate research or management approaches, with potentially disastrous results for the rarest species (Minton & Lydeard 2003). Likewise, an accurate general phylogenetic framework is necessary for detailed phylogenetic studies to correctly identify relevant taxa that should be included. Ecological studies may also be misled by incorrectly grouping unrelated taxa. Such issues, along with the development of new techniques, have prompted a renewed interest in the systematics of unionids (Roe & Hoeh 2003; Strayer et al. 2004). Nevertheless, molecular data for many genera and species are still lacking. Previous analyses have generally focused either on several species from a few genera or on a broad sampling of genera represented by one or two species a piece. They also have relied on only one or two genes. The present analyses double to triple the taxonomic coverage of previous studies, incorporating all presently recognized genera of North American Ambleminae.
Both molecular and morphological data have influenced the most recent classifications of the Ambleminae. Molecular studies indicate that all studied North American unionoids fall into three categories, corresponding to the family Margaritiferidae and the unionid subfamilies Ambleminae and Unioninae (Graf 2002); however, classifications in the past few decades have varied widely in detail (see the discussion below). The present paper treats the subfamily Anodontinae, referred to by earlier authors, as a tribe within the subfamily Unioninae. Molecular data (Hoeh et al. 2001, 2002a; Graf 2002; Huang et al. 2002; Roe & Hoeh 2003) and the presence of hooked, subtriangular glochidia (Nagel et al. 1998; Hoeh et al. 2001; Roe & Hoeh 2003) suggest that Anodonta belongs in the same subfamily as Unio. Ambleminae includes ∼80% of North American species and ∼75% of the genera. Some East Asian taxa appear closely related to Ambleminae (Huang et al. 2002); however, other Asian taxa once assigned to Ambleminae appear to be only distantly related (Graf 2002). A few European species have also been assigned to Ambleminae (Nagel et al. 1998). However, the apparent similarities may reflect convergent shell form or plesiomorphic features rather than true affinity. The present study included data for the Asian Hyriopsis and Inversidens, and the European Psilunio, to represent Old World “amblemines.”Gonidea, from northwestern North America, is exceptionally problematic in its affinities. If it is an amblemine, it is the only one in Pacific drainages of North America. Existing molecular data suggest that it may be a sister taxon to the remaining North American Ambleminae (Graf 2002). Amblema, the type genus of Ambleminae, and all species that are undisputed close relatives of it occur in the Atlantic and Gulf of Mexico drainages of North America. Within Ambleminae, the North American species have been divided among the tribes Amblemini, Pleurobemini, and Lampsilini, plus Gonideini, with some authors also recognizing Quadrulini or other taxa (Graf 2002).
The great biological diversity of North American amblemines was first recognized in the early 1800s, by workers such as Say, Rafinesque, Lea, and Conrad. However, this work focused primarily on describing species. A few authors recognized distinct genera within what is now the Ambleminae, but others placed them, along with most of the rest of the global unionid fauna, into the single genus Unio.Rafinesque (1820) named the Ambleminae (as “Amblemidia”), the only suprageneric group proposed specifically for amblemines before 1900. The 19th century classifications were based almost entirely on shell characters. Simpson (1891, p. 86) observed that existing classifications were largely artificial but left the work of revising them “to some capable student of the future.” In fact, Simpson himself took up this challenge.
Beginning in the late 1800s, many workers recognized the distinctiveness of the American amblemines, adding both new genera and higher taxa. Greater emphasis on anatomical characters contributed to this development. Simpson (1900) provided the first thorough anatomy-based classification of amblemines and other Unionoida, but did not use formal subfamilial nomenclature, introducing vernacular names instead. He also grouped several taxa with disparate but relatively simple patterns of gill brooding (present Anodontini, Pleurobemini, and Margaritiferidae). Many authors have overlooked previous names and made redundant ones (e.g., 7 of the 11 families and subfamilies Modell  used for North American Unionidae are junior synonyms). Different authors treat the same taxon as a tribe, subfamily, or family. Thus, the suprageneric nomenclature is confused. For consistency, the present paper treats Unioninae and Ambleminae as subfamilies of Unionidae, with both divided into tribes. Among the most widely used higher taxa in Ambleminae, albeit often under junior synonyms, are von Ihering's (1901) Quadrulini and Lampsilini, Hannibal's (1912) Pleurobemini, and Ortmann's (1916) Gonideini (all proposed as subfamilies). Additional tribes or subfamilies were proposed by several workers, especially Modell (1964 and references therein), Starobogatov (1970), and Heard & Guckert (1971), but these higher taxa have been synonymized or ignored by other workers. The proposed relationships of amblemine tribes to true Unionini (characterized by the European Unio, although the name has been widely misapplied to North American Pleurobemini), Anodontini, and Alasmidontini (now considered a synonym of Anodontini) also varied greatly, with many authors suggesting that Ambleminae is paraphyletic or polyphyletic. Also, authors have varied in including or excluding Eurasian and African genera in the amblemine tribes with North American types. Table 1 compares the assignments of North American amblemine genera to higher taxa in several recent classifications.
|Genus||Modell (1964)||Haas (1969)||Heard & Guckert (1971)|
|Actinonaias||Elliptionidae, Lampsilinae||Unionidae, Lampsilinae||Unionidae, Lampsilinae, heterogenae|
|Amblema||Elliptionidae, Ambleminae||Unionidae, Quadrulinae||Amblemidae, Ambleminae|
|Cyclonaias||Unionidae, Quadrulinae||Unionidae, Quadrulinae||Unionidae, Pleurobeminae|
|Cyprogenia||Elliptionidae, Lampsilinae||Unionidae, Lampsilinae||Unionidae, Lampsilinae, mesogenae|
|Cyrtonaias||Elliptionidae, Lampsilinae||Unionidae, Lampsilinae||Unionidae, Popenaiadinae|
|Dromus||Elliptionidae, Lampsilinae||Unionidae, Lampsilinae||Unionidae, Lampsilinae, eschatigenae|
|Ellipsaria||Elliptionidae, Lampsilinae||Unionidae, Lampsilinae||Unionidae, Lampsilinae, heterogenae|
|Elliptio||Elliptionidae, Elliptioninae||Unionidae, Unioninae||Unionidae, Pleurobeminae|
|Elliptoideus||Unionidae, Quadrulinae||Unionidae, Unioninae||Amblemidae, Ambleminae|
|Epioblasma||Elliptionidae, Lampsilinae||Unionidae, Lampsilinae||Unionidae, Lampsilinae, heterogenae|
|Fusconaia||Elliptionidae, Pleurobeminae||Unionidae, Quadrulinae||Amblemidae, Ambleminae|
|Glebula||Elliptionidae, Lampsilinae||Unionidae, Lampsilinae||Unionidae, Lampsilinae, heterogenae|
|Gonidea||Margaritiferidae, Pseudodontinae||Unionidae, Unioninae||Amblemidae, Gonideinae|
|Hemistena||Elliptionidae, Alasmidontinae||Unionidae, Unioninae||Unionidae, Pleurobeminae|
|Lampsilis||Elliptionidae, Lampsilinae||Unionidae, Lampsilinae||Unionidae, Lampsilinae, heterogenae|
|Lemiox||Elliptionidae, Lampsilinae||Unionidae, Lampsilinae||Unionidae, Lampsilinae, heterogenae|
|Leptodea||Elliptionidae, Lampsilinae||Unionidae, Lampsilinae||Unionidae, Lampsilinae, heterogenae|
|Lexingtonia||Elliptionidae, Pleurobeminae||Unionidae, Unioninae||Unionidae, Pleurobeminae|
|Ligumia||Elliptionidae, Lampsilinae||Unionidae, Lampsilinae||Unionidae, Lampsilinae, heterogenae|
|Medionidus||Elliptionidae, Lampsilinae||Unionidae, Lampsilinae||Unionidae, Lampsilinae, heterogenae|
|Megalonaias||Unionidae, Quadrulinae||Unionidae, Quadrulinae||Amblemidae, Megalonaiadinae|
|Obliquaria||Elliptionidae, Lampsilinae||Unionidae, Lampsilinae||Unionidae, Lampsilinae, mesogenae|
|Obovaria||Elliptionidae, Lampsilinae||Unionidae, Lampsilinae||Unionidae, Lampsilinae, heterogenae|
|Plectomerus||Unionidae, Quadrulinae||Unionidae, Quadrulinae||Amblemidae, Ambleminae|
|Plethobasus||Elliptionidae, Pleurobeminae||Unionidae, Unioninae||Unionidae, Pleurobeminae|
|Pleurobema||Elliptionidae, Pleurobeminae||Unionidae, Unioninae||Unionidae, Pleurobeminae|
|Popenaias||Not mentioned||Unionidae, Unioninae||Unionidae, Popenaiadinae|
|Potamilus||Elliptionidae, Lampsilinae||Unionidae, Lampsilinae||Unionidae, Lampsilinae, heterogenae|
|Ptychobranchus||Elliptionidae, Lampsilinae||Unionidae, Lampsilinae||Unionidae, Lampsilinae, ptychogenae|
|Quadrula||Unionidae, Quadrulinae||Unionidae, Quadrulinae||Amblemidae, Ambleminae|
|Quincuncina||Unionidae, Quadrulinae||Unionidae, Quadrulinae||Amblemidae, Ambleminae|
|Toxolasma||Elliptionidae, Lampsilinae||Unionidae, Lampsilinae||Unionidae, Lampsilinae, heterogenae|
|Tritogonia||Unionidae, Quadrulinae||Unionidae, Quadrulinae||Amblemidae, Ambleminae|
|Truncilla||Elliptionidae, Lampsilinae||Unionidae, Lampsilinae||Unionidae, Lampsilinae, heterogenae|
|Uniomerus||Elliptionidae, Elliptioninae||Unionidae, Unioninae||Unionidae, Pleurobeminae|
|Venustaconcha||Elliptionidae, Lampsilinae||Unionidae, Lampsilinae||Not mentioned|
|Villosa||Elliptionidae, Lampsilinae||Unionidae, Lampsilinae||Unionidae, Lampsilinae, heterogenae|
|Genus||Starobogatov (1970)||Davis & Fuller (1981) (Unionidae)||Previous molecular studies (Unionidae)||Present study (Unionidae)|
|Actinonaias||Lampsilidae, Lampsilinae, Lampsilini||Ambleminae, Lampsilini||Ambleminae, Lampsilini||Ambleminae, Lampsilinia|
|Amblema||Amblemidae, Ambleminae||Ambleminae, Amblemini||Ambleminae, Amblemini||Ambleminae, Amblemini|
|Cyclonaias||Lampsilidae, Pleurobeminae||Ambleminae, Amblemini||Ambleminae, Lampsilini?||Ambleminae, Quadrulini|
|Cyprogenia||Lampsilidae, Cyprogeniinae||Not sampled||Ambleminae, Lampsilini||Ambleminae, Lampsilini|
|Cyrtonaias||Lampsilidae, Lampsilinae, Lampsilini||Not sampled||Ambleminae, Lampsilini?||Ambleminae, Lampsilini?|
|Dromus||Lampsilidae, Medionidinae, Dromini||Not sampled||Ambleminae, Lampsilini||Ambleminae, Lampsilini|
|Ellipsaria||Lampsilidae, Lampsilinae, Glebulini||Not sampled||Ambleminae, Lampsilini||Ambleminae, Lampsilini|
|Elliptio||Lampsilidae, Elliptioninae||Ambleminae, Pleurobemini||Ambleminae, Pleurobemini||Ambleminae, Pleurobeminia|
|Elliptoideus||Amblemidae, Ambleminae||Not sampled||Ambleminae, Pleurobemini||Ambleminae, Pleurobemini?|
|Epioblasma||Lampsilidae, Lampsilinae, Pilaeini||Not sampled||Ambleminae, Lampsilini||Ambleminae, Lampsilini|
|Fusconaia||Amblemidae, Quadrulinae, Quadrulini||Ambleminae, Pleurobeminia||Ambleminae, Pleurobeminia||Ambleminae, Pleurobeminia|
|Glebula||Lampsilidae, Lampsilinae, Glebulini||Ambleminae, Lampsilini||Ambleminae, Lampsilini?||Ambleminae, Lampsilini|
|Gonidea||Margaritiferidae, Pseudodontinae, Pseudodontini||Ambleminae, Gonideini||Ambleminae, Gonideini||Ambleminae, Gonideini|
|Hemistena||Lampsilidae, Medionidinae, Medionidini||Not sampled||Ambleminae, Pleurobemini||Ambleminae, Pleurobemini|
|Lampsilis||Lampsilidae, Lampsilinae, Lampsilini||Ambleminae, Lampsilini||Ambleminae, Lampsilinia||Ambleminae, Lampsilinia|
|Lemiox||Lampsilidae, Lampsilinae, Lampsilini||Not sampled||Not sampled||Ambleminae, Lampsilini|
|Leptodea||Lampsilidae, Lampsilinae, Lampsilini||Ambleminae, Lampsilini||Ambleminae, Lampsilini||Ambleminae, Lampsilini|
|Lexingtonia||Lampsilidae, Pleurobeminae||Not sampled||Not sampled||Ambleminae, Pleurobemini|
|Ligumia||Lampsilidae, Lampsilinae, Lampsilini||Ambleminae, Lampsilini||Ambleminae, Lampsilinia||Ambleminae, Lampsilinia|
|Medionidus||Lampsilidae, Medionidinae, Medionidini||Not sampled||Ambleminae, Lampsilini||Ambleminae, Lampsilini|
|Megalonaias||Amblemidae, Ambleminae||Ambleminae, Amblemini||Ambleminae, Quadrulini||Ambleminae, Quadrulini|
|Obliquaria||Lampsilidae, Cyprogeniinae||Not sampled||Ambleminae, Lampsilini?||Ambleminae, Lampsilini|
|Obovaria||Lampsilidae, Lampsilinae, Glebulini||Not sampled||Ambleminae, Lampsilinia||Ambleminae, Lampsilinia|
|Plectomerus||Amblemidae, Ambleminae||Ambleminae, Amblemini||Ambleminae, Amblemini?||Ambleminae, Lampsilini?|
|Plethobasus||Lampsilidae, Pleurobeminae||Not sampled||Not sampled||Ambleminae, Pleurobemini|
|Pleurobema||Lampsilidae, Pleurobeminae||Ambleminae, Pleurobemini||Ambleminae, Pleurobemini||Ambleminae, Pleurobeminia|
|Popenaias||Lampsilidae, Elliptioninae||Not sampled||Not sampled||Ambleminae, Amblemini|
|Potamilus||Lampsilidae, Lampsilinae, Lampsilini||Ambleminae, Lampsilini||Ambleminae, Lampsilini||Ambleminae, Lampsilini|
|Ptychobranchus||Lampsilidae, Ptychobranchinae||Ambleminae, Lampsilini||Ambleminae, Lampsilini||Ambleminae, Lampsilini|
|Quadrula||Amblemidae, Quadrulinae, Quadrulini||Ambleminae, Amblemini||Ambleminae, Quadrulinia||Ambleminae, Quadrulinia|
|Quincuncina||Amblemidae, Quadrulinae, Quadrulini||Ambleminae, Amblemini||Ambleminae, Pleurobeminia||Ambleminae, Pleurobeminia|
|Toxolasma||Lampsilidae, Lampsilinae, Lampsilini||Ambleminae, Lampsilini||Ambleminae, Lampsilini?||Ambleminae, Lampsilini?|
|Tritogonia||Amblemidae, Quadrulinae, Quadrulini||Ambleminae, Amblemini||Ambleminae, Quadrulini||Ambleminae, Quadrulini|
|Truncilla||Lampsilidae, Lampsilinae, Glebulini||Not sampled||Ambleminae, Lampsilini||Ambleminae, Lampsilini|
|Uniomerus||Lampsilidae, Elliptioninae||Ambleminae, Pleurobemini||Ambleminae, Pleurobemini||Ambleminae, Quadrulini|
|Venustaconcha||Lampsilidae, Lampsilinae, Lampsilini||Not sampled||Not sampled||Ambleminae, Lampsilinia|
|Villosa||Lampsilidae, Lampsilinae, Lampsilini||Ambleminae, Lampsilini||Ambleminae, Lampsilini||Ambleminae, Lampsilinia|
The many morphological studies have provided much additional data, but relatively few characters have shaped most classifications. Particularly important anatomical features include the structure of the gills and their modification for brooding. Other studies focused on hinge and shell features. Heard & Guckert (1971) summarized the traditional distinguishing suprageneric characters in Ambleminae from many previous sources. Lampsilini was distinguished by the distinctive modifications of the female posterior outer gills for brooding and by the sexually dimorphic shells. Quadrulini (used interchangeably with Amblemini) was based on the frequently sculptured shells and the use of all 4 gills for brooding (“tetrageny”). Pleurobemini was characterized by the generally smooth shells and the use of only the outer gills for brooding (“ectobranchy”), without the gill specializations of Lampsilini. Gonideini was based on the lack of hinge teeth and distinctive gill anatomy. Even within this rather short list, most authors emphasized only one or two characters. For example, Modell (1964) relied heavily on beak sculpture, and Haas (1969) emphasized shell form. In contrast, Heard & Guckert (1971, p. 337) “selectively elected to ignore one entire array of characters,” namely shell features, and instead emphasized reproductive features such as larval brooding periods and brooding structures of the gills.
This focus on character states for selected features produced an emphasis on grades rather than clades. Putatively primitive characters, as well as putative synapomorphies, were frequently used in defining genera and higher taxa. Thus, Hannibal (1912) proposed that Pleurobemini evolved from Quadrulini (including the present Amblemini), and that Unioninae (including the present Gonideini) and Lampsilini evolved from Pleurobemini. Heard & Guckert (1971) devised a very similar system, except that Gonideini was seen as the sister taxon to Amblemini. Such paraphyletic taxa are likely to be a poor guide to evolutionary relationships. Likewise, basing higher taxa on one or a few characters runs the risk of possible homoplasy in that character (Roe & Hoeh 2003). Currently used morphological characters may not provide enough data to resolve relationships within Ambleminae (Graf & Ó Foighil 2000a). Recent studies performing cladistic analyses of morphological features have found little or no resolution of relationships of Ambleminae within Unionidae, apart from the distinctive gill structures of Lampsilini (Hoeh et al. 2001; Roe & Hoeh 2003). This paucity of morphological characters has resulted in relatively few morphology-based hypotheses about the phylogenetic interrelationships of individual genera. Apart from general agreement about the monophyly and constituent genera of Lampsilini, there have been many differences between classifications. Several species have also varied in their generic assignment from author to author. Also, the use of paraphyletic taxa in non-cladistic classifications makes it unsafe to assume that previous authors thought that the genera and higher taxa that they used were monophyletic. This pattern of classification prevailed until the advent of molecular data and cladistic methodologies prompted thorough re-examination of the taxonomy.
The fossil record provides limited help in resolving the relationships of Ambleminae, due to the sporadic nature of their fossil record and the problems of convergence in shell form. Probable amblemines occur in the Cretaceous faunas of central North America (Hartman 1998), suggesting that the North American amblemines have a long evolutionary history separate from other unionoids. However, these might represent unrelated, morphologically convergent taxa (Watters 2001).
The development of molecular techniques provided a novel source of information on the systematics of the Ambleminae. Some traditional classifications were supported, whereas others were called into question and new ideas were suggested. In the first molecular study to include many amblemines, Davis & Fuller (1981) used immunoelectrophoresis to assess the relationships of North American unionoidean genera (Table 1). This study demonstrated the distinctiveness of Ambleminae from “Anodontinae” (Unioninae). The early immunological work was followed by numerous studies using DNA sequencing, as well as a few other genetic techniques, e.g., RFLP analysis (White et al. 1996). Rosenberg et al. (1994) found little variation in the D6 region of the 28S gene within the Unionidae, but subsequent studies have identified more variable genes (16S: Lydeard et al. 1996; COI: Roe & Lydeard 1998; ITS: King et al. 1999; D2 region of 28S: Graf & Ó Foighil 2000b; Graf 2002; ND1: Buhay et al. 2002; Serb et al. 2003; male mitotype COI: Hoeh et al. 1996, 2002b; COII: Curole & Kocher 2002; cytB: Mock et al. 2004). Studies using DNA sequencing have generally supported the higher taxa recognized by Davis & Fuller (1981), with the exception of their Amblemini, which appears to be a polyphyletic group sharing plesiomorphic features (Lydeard et al. 1996). In general, traditional species-level classification has been upheld, but genera and higher taxonomic categories often appear polyphyletic (Lydeard et al. 2000; Roe et al. 2001; Serb et al. 2003; Huff et al. 2004). However, sampling issues remain a problem, with many genera and type species unavailable to previous studies. The frequent polyphyly of genera makes data for type species especially important; otherwise, it is unclear which group of species actually belongs in the genus. Although these studies provide explicit hypotheses about the phylogenetic relationships of the included taxa, the patchy taxonomic coverage makes it impossible to extrapolate phylogenetic relationships for the Ambleminae as a whole.
The modern concept of Ambleminae as a monophyletic group has only arisen with the advent of molecular and cladistic studies. Most workers before Davis & Fuller (1981) placed Pleurobemini as closely related to, if not synonymous with, the Old World Unionini. However, Davis & Fuller (1981) found that Unioninae (as Anodontinae) were genetically very distinct from the Ambleminae (including Pleurobemini), a conclusion substantiated by all subsequent molecular studies. In these studies (e.g., Lydeard et al. 1996; Bogan & Hoeh 2000; Graf 2002; Graf & Ó Foighil 2000a; Hoeh et al. 2002a, b; Krebs et al. 2003; Roe & Hoeh 2003), the relationships between the tribes in Ambleminae have varied. Not all studies included Gonideini or Quadrulini. Those including Gonideini placed it basal to the other tribes or outside of Ambleminae. The interrelationships of Amblemini, Quadrulini, Pleurobemini, and Lampsilini differed from analysis to analysis, and Lampsilini was not consistently monophyletic. Often, these groupings did not have strong bootstrap support, and some, if not all, of the tribes were represented by only a few taxa in each analysis.
Thus, Simpson's (1891) challenge remains a problem: what are natural groups in the Unionidae? We sought to answer this for the Ambleminae by addressing three main questions: (1) Are the North American Ambleminae a monophyletic group? (2) What are the relationships among the North American genera of Ambleminae? (3) Are these genera, as currently recognized, natural entities?