Teleostei: Cyprinidae) from Cytochrome B Sequences

Copeia, 2002(3), pp. 665–678

Evolutionary Relationships of the Plagopterins (Teleostei: Cyprinidae) from Cytochrome b Sequences

THOMAS E. DOWLING,C.ALANA TIBBETS,W.L.MINCKLEY, AND GERALD R. SMITH

Sequences of cytochrome b (cytb) were used to examine composition and phylogenetic relationships of cyprinid ﬁshes of the tribe Plagopterini, endemic to the Great Basin and Lower Colorado River in southwestern North America. The plagopterin genera, Lepidomeda, Meda, Plagopterus, and Snyderichthys, were most closely afﬁliated with the chubs Couesius and Margariscus of northern and eastern North America. As indicated by previous morphologic, allozymic, and mtDNA studies, Sny- derichthys is intimately related to Lepidomeda. The relationship is paraphyletic, however, according to our molecular data. Snyderichthys from the Snake and Bear River drainages are part of a clade that includes Lepidomeda mollispinis and Lepidomeda albivallis according to the cytb sequence, with Snyderichthys from the central and southern Bonneville basin more divergent. This paraphyly and the complex geographic relationships of mtDNA sequences indicate a complex history of the group and cast doubt on the validity of morphologically diagnosed Snyderichthys. Estimates of divergence time, based on a combination of fossil and molecular data, indicate that the plagopterins are an ancient clade, at least 17 million years old.

HE Cyprinidae or minnows are the most di- They hypothesized existence of three major T verse freshwater fish family in North Amer- groups: ‘‘shiner,’’ ‘‘chub,’’ and ‘‘western’’ ica, represented by more than 300 species in clades. Plagopterins fell within a Gila clade of approximately 50 genera (Burr and Mayden, western minnows, most closely related to a clade 1992). Although recent morphological studies containing Agosia, Moapa, and Rhinichthys.As provide hypotheses of relationships among gen- previously indicated by Uyeno (1960), Gila copei era, species groups, and species (Mayden, 1989; was considered distinct from the plagopterins. Cavender and Coburn, 1992; Coburn and Cav- Studies of molecular characters indicated dif- ender, 1992), analysis of molecular characters ferent plagopterin relationships. DeMarais sometimes produce conflicting results (e.g., Si- (1992) provided allozymic evidence that G. copei mons and Mayden, 1997, et seq.; Broughton was closely affiliated with Lepidomeda, possibly and Gold, 2000) and a different perspective of rendering the plagopterins paraphyletic. Fur- evolutionary relationships. ther taxonomic studies were deemed necessary The plagopterins are a monophyletic group to ensure proper phylogenetic placement of G. of North American cyprinids diagnosed by copei; however, its close affinities with Lepidomeda spinelike ossification of the first two rays of the led DeMarais (1992) to recommend resurrec- dorsal and pelvic fins. As defined by Miller and tion of Snyderichthys, a recommendation we fol- Hubbs (1960), the group consists of six species low. in three genera: Lepidomeda vittata, Lepidomeda Simons and Mayden (1997, et seq.) identified mollispinis, Lepidomeda albivallis, and Lepidomeda four major lineages (with sequences from 12S altivelis; Meda fulgida; and Plagopterus argentissi- ribosomal DNA) that differed from those iden- mus. These are (or were) endemic to the middle tified by Coburn and Cavender (1992). Both and lower Colorado River drainage. Relation- morphological and molecular data indicated a ships of this group to several other cyprinids of distinct western clade; however, there were ma- the region have been hypothesized, especially jor differences in composition and placement Snyderichthys copei (Miller, 1945) of the adjacent of taxa within the chubs and shiners. Creek Bonneville basin and upper Snake River drain- chubs and relatives (Couesius, Hemitremia, Mar- age. That species was placed in the monotypic gariscus, Semotilus) were identified as a distinct genus Snyderichthys by Miller (1945) but consid- lineage (‘‘creek chub clade’’) instead of being ered a monotypic subgenus of the genus Gila by members of the chub clade as hypothesized by Uyeno (1960). Coburn and Cavender (1992). Plagopterins (in- Coburn and Cavender (1992) completed an cluding Snyderichthys) were more closely related extensive analysis of phylogenetic relationships to this clade than to members of the genus Gila. among North American cyprinids based on Simons and Mayden (1997) included only two morphological and osteological characters. plagopterins (L. vittata, M. fulgida); therefore,

᭧ 2002 by the American Society of Ichthyologists and Herpetologists 666 COPEIA, 2002, NO. 3

gopterus argentissimus are found only in single drainages. Previous studies of variation in L. vittata indicated limited divergence among locali- ties (Tibbets, 1998; Tibbets et al., 2001); therefore, single individuals of Lepidomeda mollispinis mollispinis, L. vittata, and Plagopterus argentissimus were used to represent each. Previous studies of Meda fulgida (Anderson and Hendrickson, 1994; Tibbets and Dowling, 1996) identified divergent populations in the Gila and Verde Riv- ers; therefore each was represented. Because of the widespread distribution and potential paraphyly of Snyderichthys to the plagopterins, mul- tiple samples were also examined (Bonneville basin, Bear and Snake River drainages). Representatives of the three major clades (western, shiner, and chub) of Coburn and Cav- ender (1992) were also examined, with special emphasis on hypothesized sister taxa to plagopterins. These names will be used throughout manuscript. Chubs were Campostoma anomalum, Couesius plumbeus, Hybognathus hankinsoni, Mar- Fig. 1. Approximate distribution of species and gariscus margarita, Nocomis micropogon, and Semo- samples (indicated by lines and shading). Snyderichthys tilus atromaculatus. Shiners included Cyprinella copei, Goose Creek (SCGC), Sulphur Creek (SCSC), spiloptera, Luxilus chrysocephalus, and Pimephales Spanish Fork River (SCSF), Sevier River (SCSR); Lep- notatus. Representatives of the western clade idomeda albivallis (LA), Lepidomeda mollispinus mollispi- were Acrocheilus alutaceus, Agosia chrysogaster, Gila nus (LMM), Lepidomeda mollispinus pratensis (LMP); cypha, Moapa coriacea, Orthodon microlepidotum, Lepidomeda vitatta (LV); Plagopterus argentissimus (PA); and Rhinichthys atratulus. Based on relationships Meda fulgida, Verde River (MFV), Aravaipa Creek inferred from morphological characters (Cav- (MFA). The southwest range limits of the sister ender and Coburn, 1992; Coburn and Caven- groups, Couesius plumbeus and Margariscus margarita, are shown in the upper right. der, 1992), sequences from the leuciscins Abram- is brama (Briolay et al., 1998; GenBank accession nunber Y10441) and Notemigonus crysoleucas relationships within the group and effects of (Schmidt and Gold, 1995; GenBank accession taxonomic sampling on placement of Snyderi- number U01318) were used as outgroups. chthys could not be examined. We use sequences from the cytochrome b Sequencing.—Sequences were obtained from (cytb) gene from all extant plagopterins to ex- PCR products by either standard dideoxy se- amine relationships among members of this quencing or cycle-sequencing and deposited in group and to test the hypothesized placement GenBank (accession numbers in Material Ex- of these fishes relative to other cyprinids [as amined). In the standard method, templates proposed by Coburn and Cavender (1992) and were obtained through two rounds of amplifi- Simons and Mayden (1997)]. Given the unique cation. In the first, a double-stranded product morphology and geography of these fishes, was generated as follows: 1–5 ␮l (approximately knowledge of their phylogenetic relationships 1–5 ng) of genomic DNA; 0.25 units of Taq DNA to other minnow lineages provides important polymerase [from Promega Corp. or purified information for the evolution of morphological characters and the biogeography of several ma- from a clone as described by Pluthero (1993)]; jor drainages in western North America. 4 ␮l of dNTP mix (final concentration of 200 ␮M of each of the four dNTPs); 2.5 ␮l MgCl2 stock (final concentration of 2.5 mM); 2.5 ␮lof MATERIALS AND METHODS 10ϫ buffer stock [provided by the enzyme sup- Samples.—Most plagopterin taxa now occupy plier or as described in Pluthero (1993)]; 1.25 limited ranges (Fig. 1) in the lower Colorado ␮l of each primer (final concentration of 0.5 River system. Lepidomeda albivallis and Lepidome- ␮M of each); sterile-distilled water to a final vol- da mollispinis pratensis are restricted to single lo- ume of 25 ␮l. Amplification was performed by calities and L. m. mollispinis, L. vittata, and Pla- 20 cycles of denaturation at 94 C for 1 min, an- DOWLING ET AL.—RELATIONSHIPS OF PLAGOPTERINS 667

TABLE 1. SEQUENCES OF PRIMERS USED FOR AMPLIFICATION AND SEQUENCING OF CYTb.

Primera Positionb Sequence LA-a 15248 5Ј GTG ACT TGA AAA ACC ACC GTT G 3Ј LC-s 15536 5Ј ATA CAT GCC AAC GGA GCA TC 3Ј HD-a 15894 5Ј GGG TTG TTT GAT CCT GTT TCG T 3Ј LD-a 15848 5Ј CCA TTC GTC ATC GCC GGT GC 3Ј

LDIB-a 15768 5Ј ACC CTT GTT CAA TGA ATC TG 3Ј LDLUX-a 15812 5Ј ACA TTA ACA CGA TTC TTC GC 3Ј LE-s 16084 5Ј CCC ACC ACA CAT TCA ACC 3Ј HA-a 16462 5Ј CAA CGA TCT CCG GTT TAC AAG AC 3Ј

a Primers are designated by strand (L ϭ light, H ϭ heavy) and application (a ϭ ampli®cation and sequencing, s ϭ sequencing only). b Position is identi®ed relative to the published sequence for carp, Cyprinus carpio, (Chang et al., 1994). nealing at 48 C for 1 min, and extension at 72 Swofford, Sinauer Associates, Sunderland, MA, C for 2 min. 1998, unpubl.). Topologies were obtained by Single-stranded DNA for sequencing was gen- heuristic search using the tree bisection-recon- erated by using 10 ␮l of 1:100 dilution of this nection (TBR) method with 50 random-addi- double-stranded PCR product. Conditions for tion sequences and uninformative characters amplification were identical except that one of excluded. Relative strength of nodes was deter- the primers was diluted (1:100) and annealed mined by completion of 1000 bootstrap repli- at 52 C. Single-stranded products were cleaned cates (Felsenstein, 1985; Hillis and Bull, 1993) by centrifugation (Millipore Corp.), as indicat- in PAUP, using the fast-search option. Neighbor- ed by the supplier, and sequenced using the Se- joining trees were also generated with PAUP, quenase 2.0 kit (U.S. Biochemical) as described using Tamura-Nei distances corrected for rate in Dowling and Naylor (1997). Alternatively se- variation using the alpha value (alpha ϭ 0.3) quences were obtained from cleaned, double- estimated by maximum likelihood from a larger stranded PCR products by cycle sequencing, fol- study of cytb variation in western cyprinids lowing recommendations of the manufacturer (Smith et al., in press). Relative strength of (Perkin-Elmer), or by automated sequencing of nodes was determined by completion of 1000 dye-labeled products (using an ABI377). bootstrap replicates in PAUP. Most primers (Table 1) were designed by T. R. Schmidt from a series of sequences from east- Estimation of divergence times.—Rates of molecu- ern North American cyprinids (described in lar evolution were estimated from match of Schmidt and Gold, 1993); however, it was nec- fossil to molecular data in the context of the essary to redesign some internal primers from molecular phylogeny (Table 2). This method sequences of several other minnows obtained permits calculation of estimated rates of diver- with external primers. External primers (LA gence, which then enable estimates of diver- and HA) were positioned in conserved flanking gence times for biogeographic units. Numera- tRNA genes. Because of its large size, sequences tors in the rate equation are pairwise Tamura- for the entire gene were obtained from two sep- Nei distances expressed as percentage sequence arate sets of primers (LA-HD and LD, LDIB,or divergence of cytb. Denominators in the rate LDLUX-HA, respectively, Table 1), producing equation are estimated times (in millions of fragments approximately 600 bp in length. The years ϭ Ma) of divergence for those taxa. The complete sequence of each fragment was ob- fossil record provides estimates of times of di- tained using two sets of primers, the limiting vergence as inferred from oldest fossil occur- external primer in the asymmetric amplification rences of diagnostic apomorphies of lineages reaction and an internal primer from the same (Smith et al., in press). Regression of pairwise strand (Table 1). percentage divergence against time of first appearance in the fossil record yields an estimated Data analysis.—Sequences were aligned by eye rate of divergence for cytb in cyprinids (Fig. 2). using the homologous sequence of Cyprinus car- When the intercept is constrained to pass pio (Chang et al., 1994) as reference. Although through the origin (Hillis et al., 1996), this es- this can be a problem for noncoding sequences, timate is 1.31% sequence divergence per pair- the alignment attained for the protein-coding wise comparison per Ma (or 0.66% per lineage sequence was unambiguous. Phylogenetic anal- per Ma), with an error of 10% to 25% (Smith ysis was performed using PAUP4.0b8w (D. L. et al., in press). This assumes no genetic diver- 668 COPEIA, 2002, NO. 3 % divergence 6 8.1 7 21.0 7 6.4 (Fig. 2). Diagnostic 15 27.0 STIMATION E ATE R tion, ID tion, ID UT UT SED IN Glenns Ferry forma- Truckee formation, NV 12Glenns Ferry forma- 15.0 Truckee formation, NV 12 11.5 Bonneville deposits, UT 6 6.2 Salt Lake formation, Truckee formation, NV 15 10.3 Salt Lake formation, U IVERGENCE D EQUENCE teeth geal teeth geal teeth geal teeth geal arch teeth geal arch S Molariform pharyngeal One row sharp pharyn- Long, slender pharyn- One row sharp pharyn- Narrow anterior frontal Lahontan deposits, NV 0.1 3.1 Short articular angular Bear Lake, UT, ID 0.03 0.9 Robust, curved pharyn- Hooked, non-slicing Robust, curved pharyn- STIMATES OF sp. Opposing molar teeth E AND Mylopharodon Siphateles Richardsonius Siphateles Siphateles b. obesus G. a. subspecies Gila atraria Klamathella milleri Gila turneri Mylocheilus (Ma) EARS Y ILLIONS OF M ID tion, ID mation, NV UT UT ID ID ID Chalk Hills formation, Glenns Ferry forma- Stewart Valley, NV Bonneville deposits, Salt Lake formation, Chalk Hills formation, Priest Rapids, Miocene, Priest Rapids, Miocene, OSSILS IN F PPEARANCE OF A teeth teeth ryngeal teeth ryngeal teeth geal arch geal teeth geal teeth geal teeth Striated pharyngeal Molariform pharyngeal Molar pharyngeal teeth Below Chalk Hills for- Two rows hooked pha- Broad anterior frontal Chewaucan Valley, OR Two rows hooked pha- Long, slender pharyn- Sharp, slicing pharyn- Conical canine pharyn- Conical canine pharyn- IRST F GE OF characters, stratigraphic, geographic, and age data from Smith et al., 1982; Gobalet and Negrini, 1992; Smith et al., 2000; and Smith et al., in press. 2. A Fossil taxon Diagonstic character Locality Sister taxon Diagnostic character Locality Time (Ma) ssp. ABLE T Lavinia Mylopharodon Mylocheilus inflexus Gila turneri Siphateles bicolor Gila a. atraria Klamathella milleri Acrocheilus Ptychocheilus Ptychocheilus DOWLING ET AL.—RELATIONSHIPS OF PLAGOPTERINS 669

Fig. 2. Scatter of pairwise sequence divergences against time since initial divergence [millions of years (Ma)], as estimated from first appearance of diagnostic apomorphy in the fossil record (from Table 2). Lines identify regressions unconstrained (solid, R2 ϭ 0.48) and forced through the origin (dashed, R2 ϭ 0.44), yielding estimated divergence rates of 0.53% and 0.66%/lineage/ma, respectively. Fig. 3. Results from parsimony analysis of cytb sequences. Unweighted and weighted analyses pro- gence at the time of branching and unbiased duced similar results; therefore, only the more re- estimates of the denominator. This method solved topology (weighted) is presented. This topology represents a strict consensus of six most-parsi- probably overestimates the rates of molecular monious trees with transversions given twice the evolution because the first appearance of fossil weight of transitions. Numbers above and below taxa are almost surely underestimated but has branches identify bootstrap values (percent) from un- the advantage of being based on a sample of weighted and weighted analyses, respectively. Boot- values for numerators and denominators, rather strap values not provided are less than 60%. than a single assumption. If, however, estimates of ages of ancestors are biased toward under- estimation, the regression should not be con- number (76) in the first position. Heuristic strained, producing a line with a positive inter- search with all characters weighted equally re- cept and shallower slope of 1.05% per million covered 15 most-parsimonious trees of 2040 years (or 0.53% per lineage per million years). steps (CI ϭ 0.34, RI ϭ 0.54). A strict consensus Constancy of evolutionary rates was examined of these (not shown) identified a polytomy of using the relative rates test as implemented in three major lineages: (1) western clade (Acroch- the program PHYLTEST (vers. 2.0, S. Kumar, eilus, Gila, Moapa, and Orthodon); (2) an unre- Pennsylvania State University, University Park, solved group of chubs and shiners (Campostoma, PA, 1996). Estimates of divergence time were Cyprinella, Hybognathus, Luxilus, Nocomis, and Pi- corrected for variation in rates of evolution mephales) that included the western clade mem- among lineages by calculating branch lengths bers Agosia and Rhinichthys; and (3) a well-re- for each lineage using MEGA2 (S. Kumar, K. solved clade containing members of the western Tamura, I. B. Jakobsen, and M. Nei, Arizona and chub clades (plagopterins, Snyderichthys, State University, Tempe, AZ, 2001). Because it Couesius, Margariscus, and Semotilus) and consis- is difficult to determine whether constrained or tent with the creek-chub clade ϩ plagopterin unconstrained estimates are more appropriate, grouping identified by Simons and Mayden minimum and maximum divergence estimates (1997). Further evaluation by bootstrap analysis were derived using the constrained (0.66% per (Fig. 3) indicated monophyly of the western Ma) and unconstrained calibrations (0.53% per and chub ϩ shiner clades (including Agosia and Ma). Rhinichthys), with these groups occurring in 95% and 91% of replicates, respectively. The re- RESULTS lationship of Couesius, Margariscus,orSemotilus to the plagopterins was not strongly supported, Of the 1140 bp of cytb, 425 were informative. because they formed a monophyletic group in Patterns of variation were typical, with most var- less than 50% of replicates. iable characters (335) in the third position, few- Relationships among the plagopterins were est in the second (14), and an intermediate generally well resolved (Fig. 3). This monophy- 670 COPEIA, 2002, NO. 3 letic group (80% of bootstrap replicates) was di- terin ϩ creek-chub lineage sensu Simons and vided into a southern clade containing Meda Mayden, western minnows were more divergent, and Plagopterus (99% of replicates) and a and Semotilus was the most divergent lineage. northern lineage including all Lepidomeda and Bootstrap values supported monophyly of the Snyderichthys (94% of replicates). The northern western, chub ϩ shiner, and plagopterin (in- lineage was further divided into four groups including Snyderichthys) groups (100%, 99%, and consistent with current taxonomy and biogeog- 87% of replicates, respectively), but relation- raphy (Fig. 1). Two forms of L. mollispinis and ships among the major lineages and Couesius, L. albivallis formed an unresolved trichotomy Margariscus, and Semotilus were not well re- (100% of bootstrap replicates) that was most solved. Taxonomic groupings and levels of boot- closely related to Snyderichthys from the Snake strap support for plagopterins provided by the and Bear River drainages (94% of bootstrap neighbor-joining analysis were similar to those replicates). Lepidomeda vittata was the sister to found in parsimony analyses. Plagopterins (plus this clade (63% of bootstrap replicates), and Snyderichthys) formed a monophyletic clade in Snyderichthys from the central and southern 93% of bootstrap replicates, and relationships Bonneville basin was the most basal lineage. among Snyderichthys, Lepidomeda, Meda, and Pla- The impact of differential weighting was test- gopterus were identical to unweighted and ed by giving transversions twice the weight of weighted parsimony results, with all bootstrap transitions. This weighting scheme was selected values higher than 70%. because it reflects the minimumum transition: transversion ratio for this group of taxa (calcu- DISCUSSION lated using PAUP). Heuristic search recovered six most-parsimonious trees (2494 steps) similar Phylogenetic position of the Plagopterini.—Sequenc- to those obtained in the unweighted analysis, es of the cytb gene indicate that Snyderichthys is with the western clade sister to a chub ϩ shiner a member of Plagopterini as suggested by clade and the creek chub ϩ plagopterin clade DeMarais (1992) and Simons and Mayden of Simons and Mayden (Fig. 3). Unlike the pre- (1997). Monophyly was strongly supported, with vious analysis, shiners and chubs (including Ago- bootstrap values Ն 80% in all analyses. sia as a shiner and Rhinichthys as a chub) were These results are also consistent with the con- sister groups (found in 82% and 51% of boot- clusion of Simons and Mayden (1997) that pla- strap replicates, respectively), with only place- gopterins are related to Margariscus, Couesius, ment of the chub Hybognathus discordant with and perhaps Semotilus. When using unweighted previously hypothesized relationships. Bootstrap and weighted parsimony, plagopterins fall with- support for the western and chub ϩ shiner in the monophyletic creek chub clade, with clades was strong (95% and 97% of replicates, Margariscus their closest relative. In the neigh- respectively). Monophyly of the creek chub ϩ bor-joining topology, a group consisting of plagopterin clade occurred in more bootstrap Couesius and Margariscus was the sister lineage replicates (60%); however, relationships of to plagopterins; however, the creek chub clade Couesius, Margariscus,orSemotilus to plagopter- (sensu Simons and Mayden) was not monophy- ins were again poorly supported. letic as Semotilus was the most divergent North Relationships and levels of support among American cyprinid. The three results are simi- plagopterins were similar to those found in the lar, however, and consistent with results of Si- unweighted analysis when transversions were mons and Mayden (1997) using 12S rDNA. weighted twice as much as transitions. That is, Unfortunately, neither study provided strong plagopterins (including Snyderichtys) were bootstrap support, indicating that additional monophyletic in 90% of bootstrap replicates. data are necessary to further test these relation- Remaining relationships and bootstrap values ships. A close relationship of the plagopterins were essentially unchanged. with Margariscus and Couesius is supported by Pairwise estimates of sequence divergence biogeography. The present distribution of these were clustered by neighbor-joining, using gam- northern genera complements that of the pla- ma-corrected Tamura-Nei distances to correct gopterins near the Continental Divide in west- for differences in levels of variation among sites ern Wyoming (Fig. 1). (Table 3). Standard errors on divergence estimates were generally small, typically 15–20% Relationships within the Plagopterini.—Studies of within the plagopterin lineage. This analysis re- allozyme variation (DeMarais, 1992) supported covered groupings similar to those found a close relationship of Snyderichthys to plagop- through parsimony analysis (Fig. 4). The chub terins, rendering them polyphyletic. The ex- ϩ shiner group was most similar to the plagop- amination by Simons and Mayden (1997) indi- DOWLING ET AL.—RELATIONSHIPS OF PLAGOPTERINS 671 cated Snyderichthys was more closely related to Snyderichthys and Plagopterus ϩ Meda) exhibiting Lepidomeda vittata than to Meda fulgida. These significant deviation from equal rates. Although relationships, however, were not strongly sup- variation in rates among lineages may cause es- ported by bootstrap analysis of weighted char- timates of divergence from the mean and stan- acters as the Snyderichthys ϩ L. vittata node oc- dard errors to be misleading, the shortest and curred in 64% of bootstrap replicates while longest branches from the common ancestor monophyly of plagopterins was indicated in were used to provide minimum and maximum 58% of replicates. estimates of divergence time, respectively (S. In this analysis, sequence variation at cytb pro- Kumar, pers. comm.). duced highly resolved relationships supported The Meda ϩ Plagopterus, Lepidomeda, and Sny- by high bootstrap values. In all analyses, Lepi- derichthys clades last shared a common ancestor domeda [L. m. mollispinis and L. m. pratensis from 17 to 40 Ma (Table 4), corresponding to or pre- the Virgin River drainage and L. albivallis from dating the initial formation of the southern Col- the adjacent pluvial White River, a Pleistocene- orado Plateau margin 22 to 25 Ma (Potochnik, Holocene tributary to Virgin River (Fig. 1)] 1989) and the beginning of extensional faulting formed a trichotomy most closely related to of adjacent terrains Ͼ 20 Ma (Dokka and Ross, Snake River samples of Snyderichthys. The sister 1995) to form the Basin-and-Range Physio- species to this group is L. vittata of the Little graphic Province. Divergence of Meda and Pla- Colorado River. Snyderichthys from the central gopterus also was early, 19 to 34 Ma (Table 4), and southern Bonneville basin are the most di- and may have occurred when the Gila and Col- vergent. orado paleorivers (Howard, 1996) were isolated Although previous studies found Snyderichthys from one another, or associated with any one of closely affiliated with Lepidomeda, none had suf- the numerous closed basins that were formed ficient geographic samples to identify polyphyly south of the Plateau (Nations et al., 1982, 1985; of the species and existence of two forms of Sny- Peirce, 1987). Divergence of Meda populations derichthys. As reported here, analyses of several in the Verde and upper Gila basins occurred 0.9 additional populations from several drainages to 3.4 Ma. This separation might be attributable ( Johnson and Jordan, 2000) showed differenti- to habitat isolation because of a proclivity of the ation of northern and southern Snyderichthys. genus for hard-bottomed streams, separated by Previous studies of life-history variation ( John- the sand-bottomed, erosive nature of the main- son et al., 1995, and references therein) iden- stem lower Gila River. Alternatively, it might also tified differences in spawning time between have resulted from isolation in endorheic drain- Main Creek (Bonneville basin) and Sulphur age basins (e.g., Douglas et al., 1999). Creek, Wyoming (Bear River), providing fur- Differentiation of the southern and central ther evidence for differentiation. Given this, Bonneville basin Snyderichthys and L. vittata Snyderichthys should be considered a possible (presently in only the Little Colorado River ba- synonym of the genus Lepidomeda and possibly sin) from a common progenitor, 7.8 to 11.6 Ma divided into two species. As yet, however, we and 6.3 to 8.6 Ma, respectively, was also likely have not found diagnostic morphological char- related to isolation through Plateau uplift and acters to differentiate S. copei of the upper extensional collapse of Basin-and-Range. These Snake and Bear Rivers (type locality ‘‘tributary time estimates are bracketed by estimated pla- of Bear River at Evanston, Wyoming,’’ Jordan teau-margin development as recent as approxi- and Gilbert, 1881) and what may be S. aliciae mately 6.0 Ma on its extreme southwest (Lu- ( Jouy, 1881; type locality Provo River at Utah chitta, 1979, 1984) and other Bonneville Basin-rim Lake). structures like the Wasatch Front (30 Ma or before; Anderson and Mehnert, 1979). Biogeographic implications.—The broader rela- A limited time of differentiation (0.0–1.3 Ma) tionships hypothesized here corroborate pat- exists for Lepidomeda taxa of the Virgin River ba- terns derived from allozymes, morphology, and sin [L. m. mollispinis of the Virgin River, L. m. karyology (DeMarais, 1992; Miller and Hubbs, pratensis from Meadow Valley Wash (ϭ pluvial 1960; Uyeno and Miller, 1973). Divergence Carpenter River, a recent tributary of the Vir- times were estimated with the method of Smith gin), and L. albivallis of pluvial White River]. et al. (in press) using the neighbor-joining tree Lepidomeda altivelis, an extinct White River spe- (Fig. 4), allowing comparison of biogeographic cies was in this clade as well. Similar divergence hypotheses and geologic events. Tests for varia- times between our middle Snake and Bear river tion in rates of evolution among nodes of inter- samples (0.0–0.4 Ma) and Central and Southern est were conducted, with only the deepest node Bonneville samples (0.6–0.9 Ma) of Snyderichthys within the plagopterin lineage (Lepidomeda ϩ from north and south of the ‘‘Old River Bed’’ 672 COPEIA, 2002, NO. 3 0.463 0.483 0.506 0.522 0.485 0.474 0.456 0.480 0.588 0.428 0.444 0.367 0.438 0.449 0.047 0.059 0.038 0.055 0.052 0.039 0.047 0.062 0.047 0.056 0.069 0.056 0.064 0.426 0.438 0.564 0.555 0.461 0.450 0.436 0.436 0.455 0.401 0.439 0.431 0.393 0.404 0.049 0.042 0.281 0.055 0.051 0.022 0.034 0.041 0.053 0.046 0.067 0.064 0.064 0.417 0.427 0.582 0.557 0.435 0.445 0.408 0.421 0.538 0.404 0.462 0.425 0.447 0.443 0.038 0.326 0.429 0.049 0.059 0.042 0.048 0.040 0.064 0.053 0.063 0.059 0.061 0.387 0.410 0.435 0.437 0.483 0.471 0.440 0.404 0.465 0.429 0.505 0.387 0.415 0.431 0.291 0.353 0.344 0.057 0.059 0.035 0.044 0.038 0.057 0.056 0.057 0.062 0.060 0.438 0.451 0.524 0.507 0.488 0.488 0.470 0.444 0.499 0.400 0.401 0.425 0.446 0.443 0.276 0.301 0.304 0.432 0.060 0.054 0.033 0.048 0.055 0.059 0.065 0.057 0.062 ).Populations are abbreviated as in Figure 1. 0.412 0.422 0.558 0.519 0.459 0.459 0.448 0.416 0.500 0.458 0.468 0.460 0.473 0.469 0.314 0.345 0.249 0.341 0.060 0.045 0.029 0.346 0.044 0.048 0.044 0.055 0.063 ESPECTIVELY ,R 0.395 0.399 0.592 0.560 0.444 0.443 0.442 0.392 0.483 0.372 0.480 0.386 0.375 0.383 0.260 0.321 0.157 0.287 0.061 0.040 0.210 0.245 0.054 0.053 0.057 0.060 0.058 IAGONAL D 0.454 0.468 0.539 0.548 0.478 0.489 0.433 0.451 0.554 0.435 0.507 0.388 0.489 0.485 0.419 0.430 0.344 0.380 0.054 0.287 0.327 0.394 0.069 0.065 0.061 0.059 0.062 0.344 0.366 0.578 0.528 0.413 0.408 0.351 0.360 0.415 0.296 0.331 0.300 0.340 0.318 0.396 0.359 0.392 0.405 0.398 0.401 0.402 0.414 0.043 0.039 0.026 0.011 0.016 BELOW AND ABOVE THE ( 0.344 0.356 0.617 0.569 0.407 0.403 0.352 0.359 0.452 0.283 0.390 0.366 0.357 0.342 0.421 0.440 0.440 0.460 0.103 0.443 0.393 0.423 0.431 0.052 0.049 0.027 0.011 RRORS E 0.347 0.367 0.523 0.510 0.378 0.383 0.355 0.361 0.413 0.277 0.310 0.306 0.342 0.326 0.422 0.422 0.426 0.413 0.069 0.420 0.393 0.370 0.394 0.044 0.040 0.023 0.072 TANDARD S 0.390 0.401 0.555 0.537 0.417 0.411 0.388 0.395 0.460 0.286 0.327 0.357 0.377 0.364 0.398 0.450 0.459 0.490 0.187 0.438 0.396 0.322 0.445 0.039 0.037 0.162 0.191 HEIR T 0.336 0.337 0.541 0.514 0.391 0.395 0.362 0.341 0.451 0.328 0.279 0.321 0.327 0.329 0.402 0.391 0.352 0.411 0.265 0.470 0.374 0.349 0.407 0.028 0.257 0.267 0.320 ISTANCES AND 1234567891011121314 D 0.311 0.329 0.498 0.479 0.372 0.368 0.350 0.324 0.450 0.337 0.318 0.351 0.352 0.357 0.418 0.447 0.390 0.368 0.309 0.495 0.392 0.336 0.402 0.192 0.289 0.296 0.356 EI -N AMURA AR UV 3. T GC SC SF SR ABLE T L. m. pratensis L. m. mollispinus M. fulgida M. fulgida S. copei S. copei L. vittata L. albivallis Plagopterus Margariscus Semotilus Couesius S. copei S. copei Rhinichthys Campostoma Hybognathus Agosia Nocomis Acrocheilus Cyprinella Luxilus Pimephales Notemigonus Abramis Orthodon Gila Moapa 6. 7. 8. 9. 1. 2. 3. 4. 5. 25. 26. 27. 28. 20. 21. 22. 23. 24. 15. 16. 17. 18. 19. 11. 12. 13. 14. 10. DOWLING ET AL.—RELATIONSHIPS OF PLAGOPTERINS 673 0.050 0.052 0.005 0.059 0.059 0.057 0.049 0.051 0.072 0.083 0.063 0.054 0.055 0.058 0.078 0.079 0.069 0.077 0.077 0.079 0.069 0.070 0.067 0.079 0.076 0.073 0.089 0.049 0.050 0.022 0.059 0.059 0.059 0.050 0.047 0.073 0.078 0.060 0.054 0.056 0.058 0.083 0.079 0.067 0.088 0.075 0.084 0.075 0.074 0.072 0.085 0.079 0.075 0.103 0.002 0.319 0.323 0.010 0.010 0.016 0.002 0.043 0.036 0.044 0.035 0.019 0.020 0.056 0.057 0.061 0.067 0.050 0.065 0.057 0.060 0.063 0.044 0.046 0.057 0.051 0.048 0.006 0.314 0.318 0.009 0.010 0.015 0.002 0.040 0.034 0.043 0.034 0.018 0.019 0.053 0.055 0.060 0.064 0.046 0.063 0.057 0.059 0.061 0.041 0.047 0.056 0.048 0.047 0.285 0.295 0.299 0.311 0.051 0.050 0.045 0.043 0.052 0.062 0.054 0.044 0.043 0.067 0.077 0.064 0.091 0.057 0.080 0.069 0.070 0.071 0.064 0.066 0.066 0.057 0.065 0.006 0.006 0.319 0.317 0.009 0.009 0.015 0.295 0.035 0.043 0.034 0.019 0.020 0.055 0.056 0.061 0.066 0.049 0.062 0.056 0.058 0.062 0.043 0.047 0.056 0.050 0.049 0.096 0.100 0.363 0.354 0.014 0.014 0.098 0.298 0.032 0.043 0.036 0.017 0.017 0.061 0.054 0.059 0.063 0.048 0.058 0.063 0.060 0.065 0.047 0.049 0.054 0.049 0.048 0.049 0.050 0.365 0.357 0.001 0.088 0.048 0.327 0.038 0.048 0.040 0.018 0.018 0.067 0.060 0.062 0.065 0.057 0.067 0.063 0.063 0.068 0.049 0.055 0.058 0.053 0.056 ONTINUED 3. C 0.049 0.049 0.363 0.355 0.002 0.087 0.048 0.330 0.039 0.049 0.040 0.018 0.019 0.069 0.058 0.064 0.067 0.057 0.066 0.063 0.063 0.068 0.049 0.055 0.059 0.052 0.057 ABLE T 0.125 0.131 0.359 0.357 0.121 0.120 0.111 0.131 0.288 0.031 0.043 0.041 0.003 0.059 0.060 0.055 0.062 0.042 0.067 0.055 0.066 0.060 0.049 0.044 0.050 0.044 0.046 0.122 0.127 0.356 0.351 0.118 0.116 0.112 0.127 0.298 0.032 0.043 0.040 0.008 0.057 0.060 0.053 0.060 0.045 0.067 0.053 0.066 0.060 0.048 0.044 0.053 0.047 0.048 0.255 0.258 0.418 0.424 0.296 0.293 0.273 0.251 0.352 0.032 0.045 0.291 0.296 0.053 0.058 0.059 0.049 0.041 0.052 0.054 0.070 0.062 0.048 0.045 0.052 0.043 0.052 0.313 0.322 0.536 0.559 0.356 0.352 0.320 0.317 0.445 0.048 0.320 0.317 0.318 0.076 0.065 0.064 0.061 0.047 0.070 0.070 0.072 0.056 0.042 0.037 0.048 0.043 0.056 15 16 17 18 19 20 21 22 23 24 25 26 27 28 0.256 0.269 0.470 0.457 0.289 0.286 0.249 0.265 0.349 0.337 0.238 0.241 0.233 0.060 0.054 0.055 0.056 0.040 0.059 0.050 0.063 0.056 0.045 0.045 0.039 0.039 0.038 AR UV GC SC SF SR L. m. pratensis L. m. mollispinus M. fulgida M. fulgida S. copei S. copei L. vittata L. albivallis Plagopterus Margariscus Semotilus Couesius S. copei S. copei Rhinichthys Campostoma Hybognathus Agosia Nocomis Acrocheilus Cyprinella Luxilus Pimephales Notemigonus Abramis Orthodon Gila Moapa 6. 7. 8. 9. 1. 2. 3. 4. 5. 25. 26. 27. 28. 20. 21. 22. 23. 24. 15. 16. 17. 18. 19. 11. 12. 13. 14. 10. 674 COPEIA, 2002, NO. 3

(Minckley et al., 1986; Drummond et al., 1993) contain the Bear River watershed that now flows into the Bonneville basin from Wyoming, Ida- ho, and Utah. Bear Lake basin also hosted a succession of Miocene to Recent lakes (Peale 1879; Hubbs and Miller, 1948; Miller, 1965). Three lake-adapted, endemic whitefishes (Sny- der, 1919) and an endemic Cottus attest to both its antiquity and drainage relations. Exchange of Bear River between the Snake and Bonneville basins caused by block faulting and volcanism (Morrison, 1965; Stokes, 1979) are also corrob- orated by fossil fishes. A relative of the Bear Lake whitefishes occurs in Mio-Pliocene deposits of the Snake River Plain, southern Idaho (Smith, 1975), and two others were in the Bon- neville system in Pleistocene (Smith, 1981); an endemic Cottus in Utah Lake (also Bonneville) is a possible Pleistocene relative of the Bear Lake endemic (Smith et al., 1968); and Bright Fig. 4. Neighbor-joining tree generated by cluster- (1967) reported a fossil lakesucker (Chasmistes) ing gamma corrected Tamura-Nei distances (alpha ϭ related to taxa in the upper Snake or Bonneville 0.3). Numbers by branches identify bootstrap values (Utah Lake) from Thatcher Basin, along Bear (percent). Only bootstrap values Ն 70% are shown. River’s path during the most recent, Late Pleis- Scale represents number of substitutions/site. tocene diversion from the (Bear River into Lake Bonneville). are consistent with geological history (Oviatt et The sister relationship (1.3–7.4 Ma; Table 4) al., 1992). between Snyderichthys of the Bear River-Snake Affinities of the Bear and Snake River popu- complex and L. mollispinis (plus L. albivallis and lations to each other rather than to the central presumably L. altivelis) of the Virgin-pluvial and southern Bonneville populations (Bear Riv- White system is more difficult to rationalize. It er is now tributary to the Bonneville basin) must corresponds in part with a past hypothesis of reflect Middle to Late Pleistocene divergence, Snake-Bonneville-Virgin River drainage connec- before Bear River’s most recent connection tions by Keyes (1917, 1918), who proposed the (ϳ34,000 years ago; Bright, 1963) with the Bon- upper Snake River as a former Bonneville-sys- neville system. It will be of interest to examine tem tributary that ultimately flowed into the Snyderichthys from Wood River, Idaho. Miller lower Colorado River. Geologists did not take (1945) found morphological differences; since Keyes seriously (Stokes, 1979). Hubbs and Mill- Wood River enters the middle Snake, molecular er (1948:31) rejected his proposal and provided characters are predicted to resemble the middle biogeographic evidence against it in a later Snake River form. monograph (Hubbs et al., 1974) on fishes of Highlands dating to Early Miocene or before western Nevada. Ives (1948) identified a poten-

TABLE 4. MINIMUM AND MAXIMUM DIVERGENCE TIME ESTIMATES (IN MILLIONS OF YEARS) FOR PLAGOPTERIN TAXA. Abbreviations for each group are provided in parentheses following ﬁrst use.

Divergence time for: Minimum Maximum L. albivallis (LA), L. m. mollispinus (LMM), and L. m. pratensis (LMP) 0.0 1.3 Snyderichthys Goose (SCGC) and Sulphur Creeks (SCSC) 0.0 0.4 LA ϩ LMM ϩ LMP and SCGC ϩ SCSC 1.3 7.4 LA ϩ LMM ϩ LMP ϩ SCGC ϩ SCSC and L. vittata (LV) 6.3 8.6 Synderichthys Spanish Fork R. (SCSF) and Sevier R. (SCSR) 0.6 0.9 LA ϩ LMM ϩ LMP ϩ SCGC ϩ SCSC ϩ LV and SCSF ϩ SCSR 7.8 11.6 Meda Aravaipa Cr. (MFA) and Verde R. (MFV) 0.9 3.4 MFA ϩ MFV and Plagopterus (PL) 19.0 33.7 MFA ϩ MFV ϩ PL and LA ϩ LMM ϩ LMP ϩ SCGC ϩ SCSC ϩ LV ϩ SCSF ϩ SCSR 16.6 40.0 DOWLING ET AL.—RELATIONSHIPS OF PLAGOPTERINS 675 tial connection from the Escalante Basin of mtDNA among lineages. Extensive transfer of Lake Bonneville southward through Meadow mtDNA among species is common in cyprinids Valley Wash (hence to Virgin River). Smith (Dowling et al., 1989; Dowling and Hoeh, 1991) (1966) and Smith and Koehn (1971) found including species in these basins (Dowling and morphologic and genetic evidence for such DeMarais, 1993; Gerber et al., 2001). Broader communication. Taylor (1985) proposed a Late sampling and analysis of additional indepen- Miocene or Pliocene connection based on mol- dent loci is necessary to evaluate these alterna- luscan distributions from upper Snake River tives. through the Bonneville area and pluvial White River to the lower Colorado basin. Material examined.—General locality informa- Molecular similarity between Virgin River Lep- tion for ingroup sequences used in this study. idomeda and Bear-Upper Snake River Snyderi- Precise information available upon request to chthys suggests connections between these drain- TED. Sequences generated for this study were ages about 5 to 6 Ma. At this time, the Salt Lake deposited in Genbank under the accession Formation in northern Utah and southern Ida- numbers AF452072–AF452094. Genbank num- ho was separate from the Snake River drainage bers for sequences from other studies are pro- on the Western Snake River Plain (Taylor, vided in parentheses. Acrocheilus alutaceus, OR, 1985). The Snake River was connected to the Maries River drainage, Benton County; Agosia Sacramento (Wheeler and Cook, 1954; K. R. chrysogaster, AZ, Gila River drainage, Greenlee Fecht, S. P. Reidel, and A. M. Talklman, 1982, County; Campostoma anomalum, IN, Illinois River unpubl.) and Klamath drainages (Taylor, 1960; drainage, LaPorte County; Couesius plumbeus, Smith et al., 2000). Based on molecular data AK, Yukon River drainage, North Star Borough; linking northern Snyderichthys to Virgin River Cyprinella spiloptera, NY, Susquehanna River Lepidomeda, we hypothesize connections follow- drainage, Tioga County (U66605); Gila cypha, ing structural alignments of basin-and-range to- AZ, Colorado River drainage, Coconino County; pography in the vicinity of the Bonneville Basin, Hybognathus hankinsoni, MI, Lake Michigan which was not lacustrine at the time (G. K. Gil- drainage, Charlevoix County; Lepidomeda albi- bert, Lake Bonneville, U.S. Geological Survey V, vallis, NV, White River drainage, Nye County; 1890, unpubl.). Lepidomeda mollispinis mollispinis, AZ, Virgin Riv- If this pattern is the consequence of vicari- er drainage, Mohave County; Lepidomeda mollis- ance, the progenitor of Lepidomeda and Snyderi- pinis pratensis, NV, Meadow Valley Wash, Lincoln chthys was distributed throughout the Lower County; Lepidomeda vittata, AZ, Chevelon Creek, Colorado and Upper Snake Rivers around 12 Little Colorado River drainage, Coconino Ma and survived to differentiate as relicts of tec- County; Luxilus chrysocephalus, TN, Tennessee tonism disrupting a continuous distribution. River drainage, Blount or Hancock Counties; The populations now represented in the Bon- (U66595); Margariscus margarita, MI, Charlevoix neville (Snyderichthys) and Little Colorado (L. County, Lake Michigan drainage; Meda fulgida, vittata) were connected until 6.3 to 8.6 Ma. AZ, Gila River drainage, Graham Co., and AZ, More recently diverged forms (e.g., L. m. mollis- Verde River drainage, Yavapai County; Moapa pinus, L. m. pratensis, L. albivallis, L. altivelis) like- coriacea, NV, Moapa River drainage, Clark Coun- ly resulted from disruption of water connections ty; Nocomis micropogon, MI, Raisin River drainage, through recent aridification of the region (Ax- Washtenaw County; Orthodon microlepidotum, CA, elrod, 1979; Winograd et al., 1992). These tem- King’s River drainage, Fresno County; Pimephal- poral relationships demonstrate the importance es notatus, MI, Raisin River drainage, Washtenaw of tectonism and reliction in speciation of west- County (U66606); Plagopterus argentissimus,UT, ern fishes. Virgin River drainage, Washington County; If the vicariance hypothesis is correct, Snyder- Rhinichthys atratulus, MI, Rouge River drainage, ichthys will become a synonym of Lepidomeda and Macomb County; Semotilus atromaculatus, MI, the southern form will be recognized as a sep- Raisin River drainage, Washtenaw Co.; Snyderi- arate species, possibly L. aliciae ( Jouy). At this chthys copei, ID, Goose Creek drainage, Cassia time, however, we must also consider two other County; WY, Bear River drainage, Uinta County; possibilities. Polyphyly of Snyderichthys mtDNA UT, Sevier River drainage, Garfield County; and lineages could also result from stochastic pro- UT, Spanish Fork River drainage, Utah County. cesses associated with the sorting of different lineages (Neigel and Avise, 1986); however, the ACKNOWLEDGMENTS high levels of sequence divergence among them would make this alternative less likely. A possi- Wendell L. (Mink) Minckley, our coauthor, ble alternative could be introgressive transfer of passed away on 22 June 2001 while this paper 676 COPEIA, 2002, NO. 3 was in review. His knowledge of western fishes and North American freshwater fishes, R. L. May- and his long-time dedication to their study is den (ed.). Stanford Univ. Press, Stanford, CA. reflected in this and many other publications. CHANG, Y., F. HUANG, AND T. LO. 1994. The complete His friendship and guidance will be sorely nucleotide sequence and gene organization of carp (Cyprinus carpio) mitochondrial genome. J. Mol. missed. We thank R. Bailey, W. Barber, D. Buth, Evol. 38:138–155. M. Belk, C. Latta, and P. Marsh for providing COBURN, M. M., AND CAVENDER, T. M. 1992. Interre- specimens. S. Kumar provided helpful discus- lationships of North American cyprinid fishes, p. sion of molecular clocks. Federally listed fishes 328–373. In: Systematics, historical ecology, and were collected or otherwise obtained under fed- North American freshwater fishes, R. L. Mayden eral endangered-species permits from US Fish (ed.). Stanford Univ. Press, Stanford, CA. and Wildlife Service (Region 1, Portland, OR; DEMARAIS, B. D. 1992. Genetic relationships among Region 2, Albuquerque, NM; and Region 6, fishes allied to the genus Gila (Teleostei: Cyprini- Denver, CO), and state collecting permits for dae) from the American southwest. Unpubl. Ph.D. Arizona, Nevada, and Utah. Research was con- diss., Arizona State Univ., Tempe. DOKKA,R.K.,AND T. M. ROSS. 1995. Collapse of south- ducted under animal care protocol 98–454R to western North America and the evolution of Early TED. We thank these agencies for these and Miocene detachment faults, metamorphic core other courtesies, colleagues and associates for complexes, the Sierra Nevada Orocline, and the help in obtaining specimens and other assis- San Andreas Fault system. Geology 23:1075–1078. tance, and a number of organizations, primarily DOUGLAS, M. E., W. L. MINCKLEY, AND B. D. DEMARAIS. Arizona State University (TED, CAT, WLM), Na- 1999. Did vicariance mold phenotypes of western tional Science Foundation (to TED, DEB- North American fishes? Evidence from Gila River 9220683), and US Fish and Wildlife Service, Re- cyprinids. Evolution 53:238–246. gion 2 (TED, WLM) for monetary support. DOWLING,T.E.,AND B. D. DEMARAIS. 1993. Evolution- ary significance of introgressive hybridization in cyprinid fishes. Nature 362:444–446. LITERATURE CITED ———, AND W. R. HOEH. 1991. The extent of introgression outside the hybrid zone between Notropis ANDERSON,A.A.,AND D. A. HENDRICKSON. 1994. Geo- cornutus and Notropis chrysocephalus (Teleostei: Cy- graphic variation in Meda fulgida (Pisces: Cyprini- prinidae). Evolution 45:944–956. dae), Arizona and New Mexico. Southwest. Nat. 39: ———, AND G. J. P. NAYLOR. 1997. Evolutionary rela- 148–155. tionships of minnows in the genus Luxilus (Teleos- ANDERSON,A.A.,AND D. A. MEHNERT. 1979. Reinter- tei: Cyprinidae) as determined from cytochrome b pretation of the Hurricane Fault in Utah, p. 143– sequences. Copeia 1997:758–765. 165. In: Basin and range symposium and Great Ba- ———, G. R. SMITH, AND W. M. BROWN. 1989. Repro- sin field conference, G. W. Newman and H. D. Goo- ductive isolation and introgression between Notropis de (eds.). Rocky Mtn. Assoc., Geol., Denver, CO. cornutus and Notropis chrysocephalus (family Cyprini- AXELROD, D. I. 1979. Age and origin of Sonoran de- dae): comparison of morphology, allozymes, and sert vegetation. Calif. Acad. Sci. Occ. Pap. 132:1– mitochondrial DNA. Evolution 43:620–634. 74. DRUMMOND, C.N., B. H. WILKINSON,K.C.LOHMANN, BRIGHT, R. C. 1963. Late-Pleistocene stratigraphy in AND G. R. SMITH. 1993. Effect of regional topotra- Thatcher Basin, southeastern Idaho. Tebiwa 10:1–7. phy and hydrology on the lacustrine isotopic record ———. 1967. Pleistocene lakes Thatcher and Bon- of Miocene paleoclimate in the Rocky Mountains. neville, southeastern Idaho. Unpubl. Ph.D. diss., Palaeogeogr. Palaeoclimat. Palaeoecol. 101:67–79. Univ. of Minnesota, Minneapolis. FELSENSTEIN, J. 1985. Confidence limits on phyloge- BRIOLAY, J., N. GALTIER,R.M.BRITO, AND Y. BOUVET. nies: An approach using the bootstrap. Evolution 1998. Molecular phylogeny of Cyprinidae inferred 39:783–791. from cytochrome b DNA sequences. Mol. Physiol. GERBER, A. S., C. A. TIBBETS, AND T. E. DOWLING. 2001. Evol. 9:100–108. The role of introgressive hybridization in the evo- BROUGHTON,R.E.,AND J. R. GOLD. 2000. Phylogenetic lution of the Gila robusta complex (Teleostei: Cy- relationships in the North American cyprinid genus prinidae). Ibid. 55:2028–2039. Cyprinella (Actinopterygii: Cyprinidae) based on se- GOBELET,K.W.,AND R. M. NEGRINI. 1992. Evidence quences of the mitochondrial ND2 and ND4L for endemism in fossil tui chub, Gila bicolor, from genes. Copeia 2000:1–10. Pleistocene Lake Chewaucan, Oregon. Copeia BURR, B. M., AND R. L. MAYDEN. 1992. Phylogenetics 1992:539–544. and North American freshwater fishes, p. 18–75. In: HILLIS,D.M.,AND J. J. BULL. 1993. An empirical test Systematics, historical ecology, and North American of bootstrapping as a method for assessing confi- freshwater fishes, R. L. Mayden (ed.). Stanford dence in phylogenetic analysis. Syst. Biol. 42:182– Univ. Press, Stanford, CA. 192. CAVENDER,T.M.,AND M. M. COBURN. 1992. Phyloge- ———, B. K. MABLE, AND C. MORITZ. 1996. Applica- netic relationships of North American cyprinid fish- tions of molecular systematics, p. 515–43. In: Mo- es, p. 293–327. In: Systematics, historical ecology, lecular systematics. 2d ed., D. M. Hillis, C. Moritz, DOWLING ET AL.—RELATIONSHIPS OF PLAGOPTERINS 677

and B. K. Mable (eds.). Sinauer Associates, Sunder- Point, Utah. U.S. Geol. Surv. Prof. Pap. 525C:10– land, MA. 119. HOWARD, J. L. 1996. Paleocene to Holocene paleo- NATIONS, D., J. J. LANDYE, AND R. H. HEVLY. 1982. Lo- deltas of ancestral Colorado River offset by the San cation and chronology of Tertiary sedimentary de- Andreas Fault system, southern California. Geology posits in Arizona: a review, p. 102–121. In: Cenozoic 24:783–786. nonmarine deposits of California and Arizona, R. HUBBS,C.L.,AND R. R. MILLER. 1948. Correlation be- V. Ingersoll and M. O. Woodburne (eds.). Pacific tween fish distribution and hydrographic history in Section, Society of Ecology, Paleontology, and Min- the desert basins of western United States. Bull. eralogy, Los Angeles. Univ. UT 38, Biol. Ser. 10:17–144. ———, J. C. WILT, AND R. H. HEVLY. 1985. Cenozoic ———, ———, AND L. C. HUBBS. 1974. Hydrographic paleogeography of Arizona, p. 335–355. In: Ceno- history and relict fishes of the north-central Great zoic paleogeography of the west-central United Basin. Calif. Acad. Sci. Mem. 7:1–259. States, R. M. Flores and S. S. Kaplan (eds.). Rocky IVES, R. L. 1948. The outlet of Lake Bonneville. Sci. Mountain Paleogeography Symposium 3. Society of Monthly 67:415–426. Ecology, Paleontology, and Mineralogy, Denver, JOHNSON, J. B., AND S. JORDAN. 2000. Phylogenetic di- CO. vergence in leatherside chub (Gila copei) inferred NEIGEL J., AND J. C. AVISE. 1986. Phylogenetic relation- from mitochondrial cytochrome b sequences. Mol. ships of mitochondrial DNA under various demo- Ecol. 9:1029–1035. graphic models of speciation, p. 515–534. In: Evo- ———, M. C. BELK, AND D. K. SHIOZAWA. 1995. Age, lutionary processes and theory, S. Karlin and E. growth, and reproduction of leatherside chub (Gila Nevo (eds.). Academic Press, New York. copei). Great Basin Nat. 55:183–187. OVIATT, C. G., D. R. CURRY, AND D. SACK. 1992. Ra- JORDAN,D.S.,AND C. H. GILBERT. 1881. Notes on a diocarbon chronology of Lake Bonneville, eastern collection of fishes from Utah Lake. Proc. U.S.. Nat. Great Basin, USA. Paleaogeogr. Palaeoclimat. Pa- Mus. 3:459–465. leaoecol. 99:224–241. JOUY, P. L. 1881. Description of a new species of Squal- PEALE, A. C. 1879. Report of A. C. Peale, M.D., geol- ius (Squalius aliciae) from Utah Lake. Proc. U. S. ogist of the Green River Division. U.S. Geol. Surv. Natl. Mus. 4:19. Terr. (Hayden Surv.) Annu. Rept. 11:509–646. KEYES, C. R. 1917. Orographic origin of ancient Lake PEIRCE, H. W. 1987. An ancestral Colorado Plateau Bonneville. Geol. Soc. Am. Bull. 28:371–374. edge: Fossil Creek Canyon, Arizona, p. 41–42. In: ———. 1918. Lacustral record of past climates. Geological Society American centennial field guide. Monthly Weath. Rev. 46:277–280. Vol. 1, M. L. Hill (ed.). Cordilleran Section Geolog- LUCCHITTA, I. 1979. Late Cenozoic uplift of the south- ical Society of America, Boulder, CO. western Colorado Plateau and adjacent lower Col- PLUTHERO, F. G. 1993. Rapid purification of high-ac- orado River region. Tectonophysics 61:63–95. tivity Taq DNA polymerase. Nucl. Acids Res. 21: ———. 1984. Development of landscape in northwest 4850–4851. Arizona: The country of plateaus and canyons, p. POTOCHNIK, A. R. 1989. Depositional style and tecton- 269–301. In: Landscapes of Arizona, T. L. Smiley, J. ic implications of the Mogollon Rim formation (Eo- D. Nations, T. L. Pe´we´, and J. P. Schafer (eds.). cene), east-central Arizona, p. 107–118. In: New Univ. Press Am., Lanham, MD. Mexico Geological Society Guidebook, 40th Field MAYDEN, R. L. 1989. Phylogenetic studies of North Conference of the Southeastern Colorado Plateau, American minnows, with emphasis on the genus Cy- Albuquerque, NM. prinella (Teleostei: Cypriniformes). Univ. Kans. SCHMIDT,T.R.,AND J. R. GOLD. 1993. Complete se- Mus. Nat. Hist. Misc. Publ. 80:1–189. quence of the mitochondrial cytochrome b gene in MILLER, R. R. 1945. Snyderichthys, a new generic name the cherryfin shiner, Lythrurus roseipinnis (Teleostei: for the leatherside chub of the Bonneville and up- Cyprinidae). Copeia 1993:880–883. per Snake drainages in western United States. J. ———, AND ———. 1995. Systematic affinities of No- Wash. Acad. Sci. 35:28. tropis topeka (Topeka shiner) inferred from sequenc- ———. 1965. Quaternary freshwater fishes of North es of the cytochrome b gene. Ibid. 1995:199–204. America, p. 569–581. In: The Quaternary of the SIMONS,A.M.,AND R. L. MAYDEN. 1997. The phylo- United States, H. E. Wright Jr. and D. G. Frey genetic relationships of the creek chubs and spi- (eds.). Princeton Univ. Press, Princeton, NJ. nefins: an enigmatic group of North American cyp- ———, AND C. L. HUBBS. 1960. The spiny-rayed cypri- rinid fishes (Actinopterygii: Cyprinidae). Cladistics nid fishes (Plagopterini) of the Colorado River sys- 13:187–205. tem. Misc. Publ. Mus. Zool. Univ. Mich. 115:1–39. SMITH, G. R. 1966. Distribution and evolution of MINCKLEY, W. L., D. A. HENDRICKSON, AND C. L. BOND. North American catostomid fishes of the subgenus 1986. Geography of western North American fresh- Pantosteus, genus Catostomus. Univ. Mich. Mus. Zool. water fishes: description and relationships to intra- Misc. Publ. 129:1–132. continental tectonism, p. 519–614. In: The zooge- ———. 1975. Fish of the Pliocene Glenns Ferry For- ography of North American freshwater fishes, C. H. mation, southwest Idaho. Univ. Mich. Pap. Paleon- Hocutt and E. O. Wiley (eds.). Wiley-Intersci. Publ., tol. 14:1–68. New York. ———. 1981. Late Cenozoic freshwater fishes of MORRISON, R. B. 1965. New evidence on Lake Bon- North America. Annu. Rev. Ecol. Syst. 12:163–193. neville stratigraphy from southern Promotory ———, AND R. K. KOEHN. 1971. Phenetic and cladistic 678 COPEIA, 2002, NO. 3

studies of biochemical and morphological charac- TIBBETS, C. A. 1998. Evolutionary genetics of Agosia teristics of Catostomus. Syst. Zool. 29:282–297. chrysogaster. Unpubl. Ph.D. diss., Arizona State ———, W. L. STOKES, AND K. F. HORN. 1968. Some Univ., Tempe. Late Pleistocene fishes of Lake Bonneville. Copeia ———, AND T. E. DOWLING. 1996. Effects of intrinsic 1968:807–816. and extrinsic factors on population fragmentation ———, K. SWIRYDCZUK,P.G.KIMMEL, AND B. H. WIL- in three North American minnows (Teleostei: Cy- KINSON. 1984. Fish biostratigraphy of late Miocene prinidae). Evolution 50:1280–1292. to Pleistocene sediments of the western Snake River ———, A. C. WEIBEL, AND T. E. DOWLING. 2001. Mi- plain, Idaho, p. 519–541. In: Cenozoic geology of tochondrial DNA variation within and among pop- Idaho, B. Bonnichsen and R. M. Breckenridge ulations of the Little Colorado spinedace, Lepido- (eds.). Idaho Bureau of Mines and Geology Bull. meda vittata (Teleostei: Cyprinidae). Copeia 2001: 26, Moscow, ID. 813–819. ———, N. MORGAN, AND E. GUSTAFSON. 2000. Fishes UYENO, T. 1960. Osteology and phylogeny of Ameri- of the Mio-Pliocene Ringold formation, Washing- can cyprinid fishes allied to the genus Gila. Unpubl. ton: Pliocene capture of the Snake River by the Co- Ph.D. diss., Univ. of Michigan, Ann Arbor. lumbia River. Univ. Mich. Mus. Paleont. Pap. Pa- ———, AND R. R. MILLER. 1973. Chromosomes and leont. 32:1–47. the evolution of the plagopterin fishes (Cyprini- ———, T. E. DOWLING,K.W.GOBALET,T.LUGASKI, dae) of the Colorado River system. Copeia 1973: D. K. SHIOZAWA, AND R. P. EVANS. In press. Bioge- 776–782. ography and timing of evolutionary events among WHEELER,H.E.,AND E. F. COOK. 1954. Structural and Great Basin fishes. In: The Great Basin: Cenozoic stratigraphic significance of the Snake River cap- geology and biogeography, R. Hershler and D. Cur- ture, Idaho-Oregon. J. Geol. 62:525–536. ry (eds.). Smithsonian Institution, Washington, DC. WINOGRAD, I. J., T. B. COPLEN,J.M.LANDWEHR,A.C. SNYDER, J. O. 1919. Three new whitefishes from Bear RIGGS,K.R.LUDWIG,B.J.SZABO,P.T.KOLESAR, AND Lake, Idaho and Utah. Bull. U.S.. Bur. Fish. 36:1– K. M. REVESZ. 1992. Continuous 500,000-year cli- 9. matic record from vein calcite in Devils Hole, Ne- STOKES, W. L. 1979. Paleohydrographic history of the vada. Science 258:255–260. Great Basin region, p. 345–351. In: Basin and range symposium and Great Basin field conference, G. W. Newman and H. D. Goode (eds.). Rocky Mountain (TED, CAT, WLM) DEPARTMENT OF BIOLOGY, Association of Geologists, Denver, CO. ARIZONA STATE UNIVERSITY,TEMPE,ARIZONA TAYLOR, D. W. 1960. Distribution of the freshwater 85287-1501; AND (GRS) MUSEUM OF ZOOLOGY, clam Pisidium ultramontanum: a zoogeographic in- UNIVERSITY OF MICHIGAN,ANN ARBOR,MICHI- quiry. Am. J. Sci. (Bradley Vol.) 258-A:325–334. GAN 48109. WLM DECEASED 22 JUNE 2001. E- ———. 1985. Evolution of freshwater drainages and molluscs in western North America, p. 265–322. In: mail: (TED) [email protected]. Send Late Cenozoic history of the Pacific Northwest, C. reprint requests to TED. Submitted: 4 June J. Smiley (ed.). American Association of Advanced 2001. Accepted: 13 Feb. 2002. Section editor: Science, San Francisco, CA. J. D. McEachran.