Taxonomic Hypotheses Research Articles

Phylogenetic relationships among Palearctic and Nearctic burbot ( Lota lota): Pleistocene extinctions and recolonization

The burbot ( Lota lota Linnaeus, 1758) is the only freshwater species from the cod family. Various taxonomic hypotheses were tested against molecular data by sequencing the mitochondrial cytochrome b locus of 120 burbot from 41 populations together with the related species Molva molva (ling) and Brosme brosme (tusk), which represented the other Lotinae genera. Within the genus Lota two distinct phylogroups were observed: one in North America south of the Great Slave Lakes ( Lota lota maculosa) and one in Eurasia and the remainder of the Nearctic region ( Lota lota lota). The burbot lineage separated 10 Myr BP from the other Lotinae, while the genetic variation within burbot appeared to be approximately 1 Myr old. However, fossil evidence suggested that burbot already existed in the Early Pliocene in Europe, from were it probably colonized North America in the Early Pleistocene. While Nearctic burbot survived climatic oscillations and diverged in several refugia, the Eurasian form became extinct or was reduced to a very small population. In the Late Pleistocene the species recolonized the Palearctic region to establish its present distribution range.

Crossability relationships among the wild diploid potato species Solanum kurtzianum, S. chacoense and S. ruiz-lealii from Argentina

The wild tuber-bearing Solanumspecies represent a genetic pool of enormous diversity, embracing high variability for many agronomic traits. S. kurtzianum (ktz), S. chacoense(chc) and S. ruiz-lealii (rzl) are wild diploid self-incompatible relatives of the common potato, Solanum tuberosumssp. tuberosum, with resistance to adverse biotic and abiotic factors. Pollen-pistil/style compatibility relations and seed set in intra- and interspecific crosses among ktz, chc and rzl were investigated. Pre- and post-zygotic hybridization barriers were detected in the three species, with variable incidence because the percentage of genotypic combinations that produced seeds was15–30% and 7–28% in intra and interspecific crosses, respectively. In pollen-pistil/style incompatible combinations, either pollen grains did not germinate or pollen tube growth was arrested at different sites of the style, but mainly in its upper third. Unilateral incompatibility was detected in both intra-and interspecific crosses, as well as bilateral incompatibility in certain combinations. The incompatibility relations observed could be explained by a genetic model with dominant genes in pistils that interact with corresponding dominant genes in pollen, that are independent of the S-locus. The complex arrangement of compatibility relationships found could not be directly related to taxonomic hypothesis or geographic distribution. These results suggest that divergence between the three species has occurred recently, and that internal barriers of reproductive isolation have not been an important factor in their evolution.

Molecular systematics, hybridization, and phylogeography of the Bufo americanus complex in Eastern North America

We reconstruct phylogenetic relationships among a well-studied group of toads and find relationships that differ greatly from the current taxonomic understanding. We use mitochondrial sequences encoding ND1, tRNA Leu(UUR), and part of 16S to infer relationships among members of the Bufo americanus complex. Focusing on the four taxa that historically have been most problematic due to morphological similarity and hybridization in sympatry, we sample 150 individuals from multiple populations across each species' geographic range. Our evidence conflicts with previous taxonomic hypotheses that were based on ability to hybridize, geographic distribution, and call variation. First, sequences from B. fowleri do not comprise the sister clade to sequences of B. woodhousii; therefore the previous classifications of B. fowleri as sister species to, or eastern subspecies of, B. woodhousii are both called into question. Second, sequences from B. americanus are more closely related to those of B. woodhousii than to those of B. terrestris, indicating that similar advertisement call characteristics evolved independently. Third, sequences of B. fowleri are paraphyletic, with sequences of B. terrestris embedded within. Lastly, sequences from B. fowleri cluster into three distinct mitochondrial clades, with some divergences corresponding to greater than 2 mya. These clades are somewhat geographically structured, suggesting divergence in allopatry during the Pleistocene. These mitochondrial divergences are not accompanied by known phenotypic differences, however, suggesting either evolutionary stasis in morphology and behavior, cryptic phenotypic evolution, or that hybridization in secondary contact has homogenized phenotypic differences that may have arisen in allopatry.

Systematics of the tribe Podalyrieae (Fabaceae) based on DNA, morphological and chemical data

Phylogenetic relationships among nine genera and 28 species of the southern African tribe Podalyrieae were estimated from sequences of the internal transcribed spacer (ITS) of nuclear ribosomal DNA as well as morphological and chemical data. Morphological and ITS sequence data produced cladograms with similar topologies, both supporting the monophyly of Podalyrieae (excluding Hypocalyptus). The combined data sets indicate that subtribe Xiphothecinae are monophyletic, but embedded within Podalyriinae. The high degree of congruence between previous taxonomic hypotheses and those based on DNA data provides further evidence for the utility of ITS sequences in studying phylogeny. © 2002 The Linnean Society of London, Botanical Journal of the Linnean Society, 2002, 139, 159–170.

A cladistic analysis of Lithasia (Gastropoda: Pleuroceridae) using morphological characters

The classification of pleurocerid snails and other freshwater mollusks has historically been based on morphological characters. Despite years of taxonomic work on pleurocerids, no single work includes all recognized taxa from a given group and only a few systematic treatments of the family or individual genera exist. Modern methods of phylogenetic systematics have shown that some morphological traits do not support historically accepted mollusk classifications. If analyses of morphological characters do support current taxonomic hypotheses, then the classification of these groups can be considered as stable. If not, our approach regarding diagnostic characters for these groups must change. This paper uses 25 shell and radular characters in a cladistic analysis of the pleurocerid genus Lithasia, and compares the findings to previously suggested classifications of the group. Cladistic analyses do not support any current or historical classification of Lithasia. However, these morphological characters are found to roughly delineate Lithasia and other extant pleurocerid genera, challenging previous works that suggest such characters have limited utility.

Craniodental variation in Paranthropus boisei: a developmental and functional perspective.

What levels and patterns of craniodental variation among a fossil hypodigm are necessary to reject the null hypothesis that only a single species is sampled? We suggest how developmental and functional criteria can be used to predict where in the skeleton of fossil hominins we should expect more, or less, within-species variation. We present and test three hypotheses about the factors contributing to craniodental variation in extant primate taxa, and then apply these results to the interpretation of the P. boisei hypodigm. Within the comparative samples of extant Homo, Pan, Gorilla, Pongo, and Colobus, variables from the cranial base, neurocranium, and face that are not subject to high magnitudes of strain have consistently lower levels of intraspecific variation than variables from regions of the face subject to high levels of strain. Dental size variables are intermediate in terms of their reliability. P. boisei is found to have a low degree of variability relative to extant primates for variables shown to be generally useful for testing taxonomic hypotheses. Contrary to the claims of Suwa et al. ([1997] Nature 389:489-492), the recently discovered material from Konso falls within the range of variation of the "pre-Konso" hypodigm of P. boisei for available conventional metrical variables. Those aspects of the Konso material that appear to extend the range of the P. boisei hypodigm involve regions of the skull predicted to be prone to high levels of within-species variation. The approach used in this study focuses on craniodental data, but it is applicable to other regions of the skeleton.

CRANIAL MORPHOMETRIC AND EVOLUTIONARY RELATIONSHIPS IN THE NORTHERN RANGE OFOVIS CANADENSIS

Univariate and multivariate statistical methods were used to examine geographic variation in skull and horn characters of 694 bighorn sheep (Ovis canadensis) specimens from the Great Basin north to British Columbia and Alberta to test previous taxonomic hypotheses. Substantially more morphometric variation in skull and horn size and shape was found west of the Rocky Mountains than within the Rocky Mountains. Our results did not support the recognition of Audubon’s bighorn sheep ( O. c. auduboni) as a subspecies separate from Rocky Mountain bighorn sheep (O. c. canadensis). California bighorn sheep (O. c. californiana) from Washington and British Columbia were not distinguishable from Rocky Mountain bighorn sheep but differed notably from populations in the Sierra Nevada considered part of that subspecies. Extirpated native populations from northeastern California, Oregon, and southwestern Idaho, also considered to be O. c. californiana, shared with Nelson bighorn sheep (O. c. nelsoni) from the Great Basin desert a horn-related character that distinguished them from Rocky Mountain bighorn sheep. Bighorn sheep from the Sierra Nevada were found to be distinguishable from those of the adjacent Great Basin region. Our morphometric results were concordant in geographic patterns with mtDNA data. We synonymize O. c. auduboni with O. c. canadensis. We also assign extant and extinct native populations of O. c. californiana from British Columbia and Washington to O. c. canadensis. Finally, we assign the extinct native populations of O. c. californiana from Oregon, southwestern Idaho, northern Nevada, and northeastern California to the Great Basin Desert form of O. c. nelsoni, recognizing that some transition to Rocky Mountain bighorn sheep probably occurred along that northern boundary. With these taxonomic revisions, the range of O. c. californiana includes only the central and southern Sierra Nevada.

Phylogeny of Calostoma, the Gelatinous-Stalked Puffball, Based on Nuclear and Mitochondrial Ribosomal DNA Sequences

Nucleotide sequence of the nuclear small and large subunit and the mitochondrial small and large ribosomal genes of Calostoma were used to investigate evolutionary relationships with other Basidiomycetes. In separate analyses with each gene, Calostoma was placed within the Boletales and outside the clades containing most gilled mushrooms, puffballs, stalked puffballs, earthstars, stinkhorns and nonboletalean Gasteromycetes. This contradicted previous taxonomic hypotheses. Calostoma was most closely related to Gyroporus, Pisolithus, and Scleroderma based on maximum parsimony and maximum likelihood analysis of the mitochondrial large ribosomal gene database, which contained several representatives of most genera of the Boletales. In three of the four gene trees Calostoma shows an unusually high number of inferred nucleotide substitutions relative to other Boletales; this corresponds to its extreme morphological divergence.

Phylogeny of Calostoma, the gelatinous-stalked puffball, based on nuclear and mitochondrial ribosomal DNA sequences

Nucleotide sequence of the nuclear small and large subunit and the mitochondrial small and large ribosomal genes of Calostoma were used to investigate evolutionary relationships with other Basidiomycetes. In separate analyses with each gene, Calostoma was placed within the Boletales and outside the clades containing most gilled mushrooms, puffballs, stalked puffballs, earthstars, stinkhorns and nonboletalean Gasteromycetes. This contradicted previous taxonomic hypotheses. Calostoma was most closely related to Gyroporus, Pisolithus, and Scleroderma based on maximum parsimony and maximum likelihood analysis of the mitochondrial large ribosomal gene database, which contained several representatives of most genera of the Boletales. In three of the four gene trees Calostoma shows an unusually high number of inferred nucleotide substitutions relative to other Boletales; this corresponds to its extreme morphological divergence.

RAPD AND AFLP TECHNIQUES FOR THE ANALYSIS OF GENETIC RELATIONSHIPS IN TWO GENERA OF DECAPODA

RAPD was used fur analysing three (sub-)species of mitten crabs (Eriocheir sinensis, E. japonicus, and E. japonicus hepuensis) and three populations of E. sinensis. The results show that their relationships on DNA level are similar to the classical taxonomic hypotheses (Dai, 1991). No diagnostic RAPD marker could be found, but there were statistically significant genetic differences among these taxa (P < 0.001) or populations (P < 0.001). That is, the intraspecific similarities were larger than the interspecific similarities; the intrasubspecific similarities were larger than the intraspecific similarities; and the intrapopulational similarities were larger than the interpopulational similarities. In AFLP analysis, no significant genetic difference has been found between E. sinensis and E. japonicus, but AFLP markers among four species of Macrobrachium (M. rosenbergii. M. nipponense, M. hainanense, and M. asperulum) were found. The DNA similarities among these four species of Macrobrachium are in accordance with morphological similarities.

Phylogeny of Weigela and Diervilla (Caprifoliaceae) Based on Nuclear rDNA ITS Sequences: Biogeographic and Taxonomic Implications

and three species of Diervilla were inferred from nucleotide sequence variation in the internal transcribed spacer (ITS) regions of 18-26S nuclear ribosomal DNA. Phylogenetic trees were obtained using parsimony, distance, and maximum likelihood methods. The ITS phylogenies did not support the monophyly of Weigela, which comprises three major clades. The first clade includes a single species, W. middendorffiana, which is more closely related with American genus, Diervilla. The second clade consisted of a single species, W. maximowiczii, but its relationship to other Weigela species is equivocal. The third one is a strongly supported core clade which contains the remaining Weigela species. Within the core group, W. hortensis exhibits the most distinct ITS sequence type. The hybrid origin of the species from the putative parents, W. florida and W. subsessilis, is not supported. The ITS sequence evidence favors the taxonomic hypothesis proposed by Nakai who recognized four distinct evolutionary lines (Diervilla, Macrodiervilla, Weigelastrum and Weigela). However, recognition of the two sections, Calysphyrum and Weigela was not supported in the ITS phylogenies. Substantial morphological differences between Diervilla and Weigela middendorffiana do not support the morphological stasis concept which has been considered as a general evolutionary mode among the Asian and North American disjunct taxa.

Molecular evidence for the phylogenetic position ofHanabusaya asiatica Nakai (Campanulaceae), an endemic species in Korea

The phylogenetic relationship of the Korean endemic genus,Hanabusaya, to other campanulaceous genera has been controversial since it was described by Nakai in 1911. Three genera of Campanuloideae,Symphyandra, Adenophora, andCampanula, have been considered closely related by various taxonomists on the basis of anther shape, gross morphology, and pollen characters, respectively. We have tested these competing taxonomic hypotheses using the internal transcribed spacer (ITS) sequences of nuclear ribosomal DNA from 12 taxa representing 7 genera of Campanulaceae. The molecular phylogeny indicates strongly thatHanabusaya is more closely allied toAdenophora than toCampanula orSymphyandra. The phylogenetic affinity ofHanabusaya andAdenophora is supported by a 100% bootstrap value and a high decay index (13). The average sequence divergence value (Kimura’s 2-parameter method) betweenHanabusaya and theAdenophora species is 2.58. The value is significantly (about ten times) lower than the ones observed betweenHanabusaya and the species ofCampanula (average of 23.52) and betweenHanabusaya andSymphyandra (24.95). The ITS sequence phylogeny suggests that some morphological characters, such as fused anthers and corolla shape, are homoplastic in the Campanulaceae genera.

Morphometric differentiation and species status of the Allopatric fruit‐eating bats Artibeus Jamaicensis and A. Planirostris in Venezuela

Two competing taxonomic hypotheses have been proposed for the bats Artibeus jamaicensis and A. planirostris; one species with two intergrading populations or two allopatric sibling species. Morphometric analyses on specimens from the region of potential contact in Venezuela confirmed the two species hypothesis. The Orinoco River in combination with the adjacent savanna and rainforest may have acted as a geographical and ecological barrier. In Venezuela, 57 species of bats occur only on one side of the Orinoco River.

SEQUENTIAL RADIATIONS AND PATTERNS OF SPECIATION IN THE HAWAIIAN CRICKET GENUS LAUPALA INFERRED FROM DNA SEQUENCES

The tremendous diversity of endemic Hawaiian crickets is thought to have originated primarily through intraisland radiations, in contrast to an interisland mode of diversification in the native Hawaiian Drosophila. The Hawaiian cricket genus Laupala (family Gryllidae) is one of several native genera of flightless crickets found in rain-forest habitat across the Hawaiian archipelago. I examined the phylogenetic relationships among mitochondrial DNA (mtDNA) sequences sampled from 17 species of Laupala, including the 12S ribosomal RNA (rRNA), transfer RNA (RNA)val and 16S rRNA regions. The distribution of mtDNA variants suggests that species within Laupala are endemic to single islands. The phylogenetic estimate produced from both maximum likelihood and maximum parsimony supports the hypothesis that speciation in Laupala occurred mainly within islands. The inferred biogeographical history suggests that diversification in Laupala began on Kauai, the oldest rain-forested Hawaiian island. Subsequently, colonization to younger islands in the archipelago resulted in a radiation of considerable phylogenetic diversity. Phylogenetic patterns in mtDNA are not congruent with prior systematic or taxonomic hypotheses. Hypotheses that may explain the conflict between the phylogenetic patterns of mtDNA variation and the species taxonomy are discussed.

Progress toward a phylogenetic classification of the Polyporaceae through parsimony analysis of mitochondrial ribosomal DNA sequences

We used sequence data from mitochondrial small-subunit ribosomal DNA to infer phylogenetic relationships of the Polyporaceae. We examined 62 species representing 14 families of Aphyllophorales and Agaricales. Parsimony analyses of these sequences suggest that the Polyporaceae are polyphyletic. Higher order relationships are poorly resolved, but seven groups of species are generally well supported (as measured by bootstrapping) or are congruent with previous taxonomic hypotheses. Group 1 includes Polyporus s.str., seven other genera of Polyporaceae, Lentinus, and Ganoderma. Because this clade contains the type species of Polyporus, it may serve as the core for a future recircumscription of the Polyporaceae. Group 2 is morphologically and ecologically diverse, but all members have amyloid, ornamented spores (with the possible exception of Heterobasidion). This group includes Bondarzewia, Heterobasidion, Hericium, Echinodontium, Lentinellus, Auriscalpium, and Russula. Group 3 includes five exemplars of the Hymenochaetaceae, as well as Oxyporus and Trichaptum. Trichaptum and members of the Hymenochaetaceae are unusual among the holobasidiomycetes in their possession of imperforate parenthosomes. Group 4 represents the Boletaceae and includes Boletus and Suillus. Group 5 includes Bjerkandera, which is a polypore, and Pulcherricium and Phanerochaete, which are corticioid. In Group 5 the corticioid habit may have been derived by reduction. Group 6 includes Fomitopsis, Piptoporus, and Daedalea, all of which are brown rot polypores with bipolar mating systems. Group 7 includes Laetiporus and Phaeolus. Laetiporus is classified in Polyporaceae, but Phaeolus has been placed in both the Polyporaceae and Hymenochaetaceae. In general, our results suggest that macromorphology is evolutionarily flexible, but that certain anatomical and physiological characters, while not free from homoplasy, contain clues to higher order relationships of polypores. Key words: Aphyllophorales, basidiomycetes, classification, phylogeny, Polyporaceae, ribosomal DNA.

Distribution Patterns and Zoogeography of Atlapetes Brush-Finches (Emberizinae) of the Andes

We attempted to determine why the distribution of Atlapetes rufinucha (Rufous- naped Brush-finch; Emberizinae) is so patchy. This common, sedentary species is found in several discrete areas of the humid Andes separated by distances of hundreds of kilometers, yet the gaps contain seemingly suitable habitat. Mapping of 906 specimen localities by both latitude and elevation shows that these gaps are filled by populations of other Atlapetes species, especially A. tricolor and A. schistaceus (a gray-plumaged species currently thought to be only distantly related to rufinucha, a species with yellow-and-green plumage), which in turn also show complementary, patchy distributions. Where rufinucha is the only species found, it occupies the entire elevational gradient. Where two or more species occur, their elevational distributions are restricted and often complementary. We attribute these patterns to interspecific competition. Hypothetical reconstructions of the sequence of historical events that would generate such a checkerboard distribution pattern in these sedentary taxa are complex. A novel hypothesis that would greatly simplify historical reconstructions is that rufinucha, schistaceus, and perhaps tricolor refer only to recurring color patterns characterized by differing amounts of pigment in the feathers and, therefore, adjacent populations (re- gardless of current taxonomic designation) are more closely related than either is to more distant populations of the same species. Therefore, rufinucha and schistaceus populations would be merely allopatric forms of the same lineage that alternate in color pattern (yellow or gray) between adjacent populations, as known for three other lineages of Andean birds. We found limited support for such a hypothesis. Even if rufinucha and schistaceus are distinct species, we predict that they are much more closely related than currently believed and that they differ primarily in pigment concentration. The dramatic differences in phenotype created by differences in pigment concentration in the plumage may frequently cause problems for phenotype-based taxonomic hypotheses. Received 22 December 1992, accepted 12 May 1993.

A Chloroplast DNA Inversion as a Subtribal Character in the Phaseoleae (Leguminosae)

A 78 kb DNA inversion encompassing most of the large single copy region of the chloroplast genome marks a monophyletic group that comprises most genera in the subtribe Phaseo- linae of the legume tribe Phaseoleae. The inversion, whose endpoints map next to and which includes the infA and psbA genes, originally was characterized from the chloroplast genomes of Phaseolus and Vigna (subtribe Phaseolinae), but was known to be absent from that of Glycine (subtribe Glycininae). We report here a survey of 50 genera of Phaseoleae and related tribes screened using small chloroplast DNA probes, whose hybridization patterns are diagnostic for the presence or absence of the inversion. The taxonomic distribution of the inversion is consistent with morpho- logical and other data, and supports recent subtribal classifications. The disputed inclusion of Macrotyloma in the Phaseolinae is supported by the presence of the inversion in the genus. In contrast, other genera whose placement in the Phaseolinae recently has been questioned, lack the inversion (Pachyrhizus, Clitoria, Centrosema, Psophocarpus). Although still possibly related to Phaseolin- ae, the lack of the 78 kb chloroplast DNA inversion in these genera supports taxonomic hypotheses that place them elsewhere.

Testing Nominal Species Boundaries Using Gene Flow Statistics: The Taxonomy of Two Hybridizing Admiral Butterflies (Limenitis: Nymphalidae)

Gene flow estimates between groups of populations (NmGT) derived from Wright's hierarchical F-statistics can be used to test taxonomic hypotheses about the location of biological species boundaries. The interpretation of NmGT is limited by the assumptions of the island model: balancing selection, time to equilibrium, and the limits to detectability of alleles are biases acting in the same direction as gene flow to cause similarity among populations. However, confidence limits of 0.01 ≤ NmGT ≤ 0.15 between sympatric sibling species suggest that these biases are weak relative to gene flow in causing genetic similarity. NmGT also compares favorably to gene flow estimated from a mark-recapture study, and NmGT values across a well-studied grasshopper hybrid zone suggest that gene flow is limited but not stopped by the barrier. The nominal species Limenitis lorquini and L. weidemeyerii hybridize along their contact areas in western North America. Their hybrid zone was mapped using wing pattern characters, and the same populations were analyzed electrophoretically at 18 enzyme loci. Gene flow is strong within these taxa (1.56 ≤ Nm ≤ 15.23). Across the hybrid zone, 0.57 ≤ NmGT ≤ 1.97, suggesting that although there is a partial barrier to genetic exchange, these taxa share significant portions of their gene pools. Small population sizes and other circumstantial evidence argue that much of the similarity observed among populations is actually due to gene flow, rather than balancing selection, although undetected allelic differences may also be partially responsible. L. weidemeyerii and its “subspecies” are best considered as univoltine subspecies of lorquini, and all should be treated as a single genetic unit when cladistic analyses are performed. The ease with which these results were obtained suggests that systematists should pay more attention to the genetic population structure of taxonomic boundaries.

EXPERIMENTAL HYBRIDIZATION, CHROMOSOMAL DIVERSITY, AND PHYLOGENY WITHIN GAURA (ONAGRACEAE)

Patterns of interfertility for intersectional hybridizations in Gaura are generally consistent with current taxonomic hypotheses regarding the sectional division of the genus. Crossability is generally lower for intersectional crosses than for intrasectional, interspecific crosses. Average intersectional cross-success ranged from 0-14% among the eight sections, and a total of 54 intersectional hybrids were produced from 1,872 intersectional pollinations. Low rates of capsule-set and no intersectional hybrids were produced by G. mutabilis (section Gauridium), an early offshoot of a primitive ancestor of Gaura. Two species of section Xerogaura that are regarded as relicts of the primitive ancestor of all other Gaura showed low averages for crosssuccess, but did produce hybrids in combination with three relatively advanced sections (Campogaura, Stipogaura, Pterogaura). Low capsule-set and no intersectional hybrids were produced by G. parviflora (section Schizocarya). Extreme morphological specialization for autogamous reproduction has been accompanied by reproductive isolation of G. parviflora. Only three hybrids resulted from crosses using G. coccinea (section Campogaura), a polyploid complex. Each successful cross paired G. coccinea with a putative progenitor, either a species of section Xerogaura or G. villosa of section Stipogaura. Hybrids of G. coccinea with odd ploidy number showed reduced pollen staining. Species of Stipogaura produced only three hybrids, one each in combination with species of sections Xerogaura, Campogaura, and Pterogaura. Gaura drummondii (section Xenogaura) registered the highest average cross-success despite being a tetraploid. All 25 hybrids of G. drummondii resulted from crosses with G. suffulta, and were triploids that showed reduced pollen staining. Intersectional crosses fail to support the hypothesized relationships of G. drummondii either to G. coccinea or to species of section Stipogaura. Section Gaura produced moderate averages of cross-success, but all 20 hybrids resulted from crosses with G. suffulta. Fifty of the 54 intersectional hybrids were descended from a species of section Pterogaura. Crosses that used G. suffulta as one parent produced 49 of these hybrids. Four hybrids were from crosses with G. macrocarpa (section Xerogaura), and the remaining 45 were due to the recurrent success of crosses that paired G. suffulta as maternal parent with either G. drummondii or a species of section Gaura. ARTIFICIAL HYBRIDIZATION experiments have proven extremely valuable within Onagraceae by helping to characterize relationships between species (Lewis and Lewis, 1955; Cleland, 1972; Dietrich, 1977; Seavey and Raven, 1977; Towner, 1977; Raven, 1980; Berry, 1982; Boufford, 1982; Seavey and Solomon, 1984). Estimates of interspecific fertility can be useful for evaluating the potential for hybridization and introgression between wild populations. The production of artificial hybrids, their viI Received for publication 16 February 1987; revision accepted 26 June 1987. This work was supported by a series of NSF research grants to Raven, an NIH Fellowship to Tai, and an NIH Population Biology Training Grant to Stanford University and to the Department of Plant Biology, Carnegie Institution of Washington. ability and fertility, provide information about the mechanisms that establish and maintain reproductive isolation between species (Grant, 1963; Levin, 1971; Dobzhansky, 1974). Meiotic behavior, chromosome end arrangements, and pollen fertilities of experimental hybrids can be useful as markers of the effectiveness and timing of reproductive barriers, and can also contribute to the synthesis of an appropriate interpretation of relatedness among species. Gaura (Onagraceae) is a genus of weedy herbs that ranges widely across the United States and Mexico, and reaches southern Canada. Fifteen of the 2 1 species occur within or near the center of distribution in Texas. Gaura may be distinguished among the 11 genera of Onagraceae tribe Onagreae by the divided sporogenous tissue of its anthers; the hard, indehiscent, nutlike

EXPERIMENTAL HYBRIDIZATION, CHROMOSOMAL DIVERSITY, AND PHYLOGENY WITHIN GAURA SECTION STIPOGAURA (ONAGRACEAE)

Experimental hybridization data generally agree with patterns of morphological, ecogeographic, and chromosomal variation in support of current taxonomic hypotheses for Gaura section Stipogaura. Average rates of capsule‐set for intersectional crosses are significantly lower than for interspecific crosses within Stipogaura (14% vs. 63% respectively, P &lt; 0.001). Post‐mating reproductive isolation within the section is not well‐developed, as indicated by comparable averages of cross‐success for interspecific (19%) and intraspecific (17%) crosses. Within Stipogaura, lowered rates of capsule‐set (for G. filipes crosses) and germination (seed from G. calcicola, G. filipes, and G. sinuata crosses) segregate the five species into four fertility groups. Gaura calcicola (n = 7) and G. sinuata (n = 14) differ by lower pollen staining in the hybrids of G. sinuata, a result of odd‐ploid hybridity. The order of average, interspecific cross‐success is G. mckelveyae (35%), G. villosa (25%), G. calcicola (14%), G. sinuata (10%), and G. filipes (2%). Patterns of interspecific fertility do not precisely parallel phylogenetic hypotheses based upon morphology. Three diploid taxa that closely resemble one another in growth habit (G. mckelveyae, G. calcicola, G. filipes) differ widely in rates of crossing success. Gaura villosa is relatively specialized morphologically but retains a high degree of interspecific fertility. Gaura sinuata is also specialized, but readily produces hybrids in crosses with three of the diploids.