Search

Summary. The monotypic genus Novaguinea (Compositae: Astereae: Lagenophorinae), is described and illustrated from Mount Jaya, Papua, Indonesia. The single species, Novaguinea rudalliae, is a small herb of alpine bogs. The genus clearly has affinities with Lagenocypsela but features of the capitula, involucres, receptacles, ray floret corollas, achenes and fertile disc florets, demonstrate its unique position. Generic affinities within the subtribe Lagenophorinae are discussed.

A colored painting from the Sesse and Mocifio expedition serves as the holotype for the validly published name Grindelia angustifolia DC. ex Dunal, 1819 (= Grindelia dunalii Sprengel, nom. nov., 1825; non Grindelia angustifolia Kunth, in HBK, 1818). The illustration represents the species heretofore identified as Keerlia linearifolia DC., 1836 (= Gutierrezia alamanii A. Gray, nom. nov., 1852; non Gutierrezia linearifolia Lagasca & Segura, 1816), which occurs in the south-central Mexican states of Mexico and Morelos. Gutierrezia dunalii (Sprengel) Nesom, comb. nov., is proposed as the earliest valid name for this species. Gutierrezia megalocephala (Fernald) Nesom is proposed for a closely related entity (Gutierrezia alamanii var. megalocephala (Fernald) Lane = Xanthocephalum megalocephalum Femald). The Sess6 and Mocifio expedition to New Spain (1787-1803) recorded many interesting observations, especially regarding the flora of Mexico. For a description of the expedition and its aims, travels, collections, paintings, and in particular its botanical accomplishments, see articles by McVaugh (1977, 1980, 1987, 1990, 1998) and a recent summary and update by Bartholomew and McVaugh (1997). Plant collections from the expedition are housed at a number of European herbaria as documentation for various early reports and names, but the paintings made in situ by expedition artists also provided the basic information for descriptions of new taxa. A. P. De Candolle and contemporaries proposed approximately 370 new specific names based wholly or in part on these paintings. Considerations regarding the formal typification of these taxa are outlined by Bartholomew and McVaugh (1997) and McVaugh (1998). The original set of paintings is now housed under the name of the Torner Collection of Sesse and Mocifio Biological Illustrations at the Hunt Institute for Botanical Documentation (Carnegie Mellon University, Pittsburgh). Some of these remarkably detailed paintings have remained without corroboration or reevaluation of their initial identifications, if such were ever made. We examined one of them (Torner No. 0884; Fig. 1), which includes on the same sheet two species of Asteraceae tribe Astereae, the left-hand one annotated by De Candolle as "Lemerya anthemoides" and the other as "Grindelia angustifolia." It is assumed that the plants were encountered by the expedition in Mexico, as surmised from the identities of the plants depicted. The Grindelia illustration served as the basis for a formal description of the species in a publication by Michel-Felix Dunal in 1819 (see nomenclature below). The left-hand plant is white-rayed, a species of Aphanostephus. Details at the top of the illustration, showing white ray flowers, disc corollas, and achenes, are from this plant. The slender taproot of annual duration, pinnatifid leaves, basally unexpanded disc corollas, and achenes with a conspicuous coronal pappus are features of A. ramosissima DC. var. ramosus (DC.) Turner & Birdsong (Turner, 1984), which occurs over a wide area of central Mexico, from Durango to San Luis Potosf and south to MichoacAn, Guerrero, Morelos, and Veracruz. The yellow-rayed plant (on the right-hand side) represents the species heretofore identified as Gutierrezia alamanii A. Gray, which is known from the south-central Mexican states of Mexico and Morelos (Lane, 1985). Salient features shown in the illustration are these: a basally ascending stem arising from a fibrous-rooted rhizome; leaves linear-oblanceolate without a well-defined petiolar portion, 1nerved, entire, primarily basally disposed but continuing up the stem; several relatively large heads (compared to other species of Gutierrezia) with conspicuous yellow rays (18, 21, and 23 rays, respecNovoN 10: 67-70. 2000. This content downloaded from 157.55.39.215 on Wed, 31 Aug 2016 04:26:17 UTC All use subject to http://about.jstor.org/terms

In the course of the preparation of the account of Conyza (Compositae -Astereae) for the Flora of Ethiopia, a number of specimens that could not be matched with any of the existing taxa in the genus from Africa were found among the material in the herbaria in Ethiopia (ETH), Nairobi (EA), Kew (K) and Uppsala (UPS). Some of the specimens in EA and at K had been determined as Conyza sp. E (Agnew 1974); these are here described as C. agnewi in honour of Dr A. D. Q. Agnew, who through consistent work in the field and in herbaria, produced 'Upland Kenya Wild Flowers', a flora which has also been useful to botanists and ecologists working on the flora of adjacent countries. An unmatched specimen from southern Ethiopia is also described as C. flabellata. Both species are illustrated. Conyza flabellata Mesfin sp. nov. frutex 30-45 cm altus; folia simplicia alterna vel in internodia abbreviata spiraliter disposita, flabellata vel obovata supra scabridohispida infra hispido-tomentosa sessilia, ad margine serrata usque 4 x 2 cm; capitula heterogama hemispherica 4 - 4 - 2 x 3- 5-4 mm; receptaculum planum; phyllaria c. 4-seriata incrassata vel firma, 2- 4- 3- 2 x 0- 5-0 -8 mm, exteriora dense pubescentia, interioria glabra; flosculi flavi, flosculi marginale pistillati c. 30-35 ligulati distincta; C. megensis F. G. Davies proxima sed habitu fruticosa (versus herbaceo perenni) capitula minora (4-4-2 x 3-5-4 mm versus 6-8 x 7-8 mm) et folia flabellata vel obovata (versus oblonga et ad basin cordata et subamplexicaules) differt. Typus: Ethiopia, Sidamo, Mega, 22 Oct. 1962, Mooney 9814 (holotypus ETH). Shrub, 30-45 cm high; stem striate, terete, puberulous in lower older parts, hispid-tomentose in upper parts and on branches, leafy. Leaves simple, alternate or spirally arranged on short internodes and thus leaves overlapping, flabellate to obovate, scabrid-hispid above, hispid-tomentose beneath, narrowed below into the sessile base, margins evenly serrate, teeth obtuse-apiculate, up to 4 x 2 cm, almost all leaves with tiny axillary shoots. Capitula heterogamous, radiate, hemispherical, 4 - 4 - 2 x 3 - 5-4 mm, in dense branched cymose clusters, peduncle

Intergeneric and interspecific relationships of Psilactis were studied using data from restriction site analysis of chloroplast DNA and nuclear rDNA. Molecular evidence supports the removal of Psilactis from Machaeranthera and its alliance with Aster. Within the Aster lineage, chlo- roplast DNA data support an alliance composed of members of Psilactis that have n = 4 (P. tenuis, P. boltoniae) and those of Aster subg. Oxytripolium that are based on x = 5 (A. subulatus, A. tenuifolius), separating them from the n = 9 species of both Aster and Psilactis. Evidence from nuclear rDNA, in contrast, supports different alignments by combining n = 4 and n = 9 species of Psilactis and by allying the n = 5 and n = 10 species of Aster subg. Oxytripolium with other species of Aster. Comparison of cpDNA results with evidence from morphology, rDNA, and isozymes suggests that hybridization may account for some relationships supported by chloroplast DNA. Psilactis is re-established as a distinct genus of six species, with P. mexicana placed in synonymy under P. gentryi. The following new taxonomic combinations are made: Psilactis gentryi, Psilactis odysseus, and Psilactis hetero- carpa.

Chloroplast DNA restriction site analysis was carried out on 55 taxa of tribe Astereae, including 21 species of Machaeranthera and many additional putatively related taxa. Phylogenetic analysis delineates a well supported, monophyletic Machaeranthera alliance within Astereae that includes several segregate genera of Haplopappus and all taxa historically associated with Machaeranthera except sect. Psilactis, which is allied with Aster. The Machaeranthera alliance includes a large proportion of the taxa of tribe Astereae with x = 4, 5, and 6, and contains no taxa with base chromosome numbers greater than x = 6. Chloroplast DNA data indicate five well supported monophyletic groups within the Machaeranthera alliance. One of these groups includes Xanthocephalum, Isocoma, Prionopsis, and Grindelia (all with x = 6), but does not include the x = 6 genera Xylorhiza and Pyrrocoma. A second group includes M. sect. Sideranthus and the monotypic M. sect. Stenoloba. A third group consists of M. sect. Blepharodon and two species not normally associated with the section, M. viscida and M. gypsitherma. Comparison of cpDNA evidence with one rDNA site mutation suggests that chloroplast and nuclear genomes in M. viscida, M. gypsitherma, and M. heterophylla have different origins. A fourth group includes Pyrrocoma and M. sect. Arida. In the fifth group, cpDNA data suggest a close alliance between Oonopsis and M. sects. Machaeranthera and Hesperastrum. Molecular evidence, when considered with other data, is consistent with an hypothesis of aneuploid reduction in base chromosome number from x = 6, or ascending and descending aneuploidy from x = 5 in the Machaeranthera alliance. Machaeranthera Nees consists of mostly taprooted annual or perennial herbs that are distributed in western North America. As treated by Hartman (1976, 1990; table 1), and as here defined, the genus consists of 36 species. Except for three discoid species, all have both ray and disc florets. Ray color is variable within the genus, with both xanthic (yellow) and cyanic (white, blue, pink, purple) rays present. The composition of Machaeranthera, relationships among its sections, and its affinities to other genera have been the subject of much debate (i.e., Cronquist and Keck 1957; Hartman 1976, 1990; Shinners 1950; Turner and Horne 1964; Watson 1977). Large morphological and cytological data bases exist, but synthesis of these data with evidence from hybridization, flavonoid, and micromorphological studies has not resulted in a consensus concerning the systematics of Machaeranthera. Circumscription of Machaeranthera has been extremely difficult because the closest relatives of many species that have been placed in the genus may actually be in several different genera (i.e., Haplopappus Cass., Xylorhiza Nutt., Psilactis A. Gray, and Aster L.). The task of resolving relationships within Machaeranthera is further compounded when one considers that such taxa as Haplopappus and Aster may be even more poorly understood than Machaeranthera. Hall (1928) treated Haplopappus as a large genus of 21 sections, but most or all of the North American sections may eventually be treated as segregate genera or included in other genera. Haplopappus sect. Blepharodon is the group most commonly allied with Machaeranthera. The section has included cyanic-rayed, xanthic-rayed, and eradiate species, all of which Shinners (1950) combined with Machaeranthera. Cronquist and Keck (1957), on the other hand, included only the cyanic-rayed and eradiate species of the section in Machaeranthera (series Originales, table 2). Hartman (1976, 1990) placed the eradiate and cyanic-rayed species in Machaeranthera sect. Blepharodon and the xanthic-rayed species in Machaeranthera sects. Sideranthus, Stenoloba and Havardii (table 1). Xylorhiza has been variously treated as a section of Aster, a section of Machaeranthera, or as a distinct genus. Cronquist and Keck (1957) in

As treated for the forthcoming revision of the Jepson Manual of the Flowering Plants of California, there are 14 species of Lessingia Cham. in California. Benitoa Keck and Corethrogyne DC. are considered to be congeneric with Lessingia. A forthcoming publication will provide more supporting data for the transfer of Benitoa occidentalis (H. M. Hall) Keck, Corethrogyne filaginifolia (Hook. & Arn.) Nutt., and C. californica DC. into Lessingia Cham., and discuss the treatment of Lessingia by Spence (1963). However, new combinations to accommodate these three taxa in Lessingia are needed for the revision of the Jepson Manual of the Flowering Plants of California. A brief justification for these transfers follows. Benitoa Keck has an annual habit, turbinate involucres, and disk corollas that are funnelform, yellow tinged with deep purple-red, and deeply lobed, and style-branch appendages that are linear; it shares these characters with several undisputed species of Lessingia (Howell, 1929; Spence, 1963). Benitoa is monotypic, and is found on serpentine-derived soils in southern San Benito, southwestern Fresno, and southeastern Monterey counties (Keck, 1956). Though Benitoa can be a much larger plant than most lessingias, in branching pattern, leaf morphology, pubescence, and glandularity it is very similar to them. Corethrogyne DC. taxa have a perennial herbaceous habit, turbinate to hemispheric involucres, disk corollas that are linear, yellow, and shallowly lobed, and style-branch appendages that are tufted; they share these characters with several other undisputed species of Lessingia (Spence, 1963). Corethrogyne has been treated as having various numbers of taxa (Canby, 1927; Howell, 1929; Keck, 1959), because it is extremely variable both within and between populations. Here it is treated as having a single species with two highly polymorphic varieties, based on an unpublished revision of Corethrogyne by J. P. Saroyan, D. R. Parnell, and J. L. Strother. The range of variation in vestiture of the foliage, the capitulescence, and the shape of the involucres of Corethrogyne taxa is all encompassed by that of Lessingia. The three genera have been maintained in the past because taxa of Lessingia sensu stricto are all eradiate and the others have ray florets, and because the rays of Benitoa are yellow while those of Corethrogyne are purple. The presence or absence of ray florets is probably only a single-gene character (Jackson & Dimas, 1981; Gottlieb, 1984). Given that species of Lessingia have yellow, white, pink, or purple disk corollas, it does not seem logical to use cor lla color as a generic marker. Also, the members of all three genera have phyllaries with resin-canals a d (usually) gland-tipped apices, obconic, trinerved, mottled achenes with a pappus of several (late-)deciduous bristles, and chromosome numbers of n = 5. In addition, they share at least 20 unique restriction site mutations (and differ among themselves in none) found among 63 taxa of Astereae in a chloroplast DNA study using 17 enzymes (M. Lane, unpublished data). Lessingia filaginifolia (Hook. & Arn.) M. A. Lane, comb. nov. Basionym: Aster ? filaginifolius Hook. & Arn., Bot. Beech. Voy. 146. 1833. Corethrogyne californica DC. [var.]filaginifolia (Hook. & Arn.) Kuntze, Rev. Gen. P1. 1: 330. 1891, illegit. TYPE: U.S.A. [California]: presumably Lay and/or Collie s.n. (Beechey voyage) (holotype, E fide Saroyan et al.). Kuntze's combination is illegitimate because its basionym, Asterfilaginifolius, is older than Corethrogyne californica. Lessingia filaginifolia (Hook. & Arn.) M. A. Lane var. californica (DC.) M. A. Lane, comb. et stat. nov. Basionym: Corethrogyne californica DC., Prodr. 5: 215, 1836. TYPE: "Nova California," Douglas, 1833 (holotype, G-DC fide Saroyan et al.; isotypes, BM, K fide Saroyan et al.; microfiche, KANU). Lessingia occidentalis (H. M. Hall) M. A. Lane, comb. nov. Basionym: Haplopappus occidentalis H. M. Hall, Carnegie Inst. Wash. Publ. 389: 214. 1928. Benitoa occidentalis (H. M. Hall) Keck, Leafl. W. Bot. 8: 25-40. 1956. NOVON 2: 213-214. 1992. This content downloaded from 207.46.13.85 on Tue, 23 Aug 2016 05:38:24 UTC All use subject to http://about.jstor.org/terms

As treated for the forthcoming revision of the Jepson Manual of the Flowering Plants of California, there are six species of Grindelia in California. Hybridization among them is common, and is apparently the reason for proliferation of epithets in the group. Thorough study of over 6,000 specimens (including the types for all epithets in- volved) from 34 herbaria has necessitated the rec- ognition of several new combinations, and synony- mization of many others. Grindelia Willd. (Compositae: Astereae) is even more highly variable in California than elsewhere in its range in the western United States, Mexico, and South America. Several of the taxa apparently hy- bridize freely when they are in contact; this is the primary source of the taxonomic confusion that has led to the erection of more than 60 epithets for taxa of California alone. Though a more thorough dis- cussion of the systematics of this group will be pub- lished elsewhere, it is necessary now to provide new combinations that may be used in a treatment of Grindelia for the forthcoming revision of the Jepson Manual of the Flowering Plants of California. Based on examination of the types for all the bas- ionyms involved and over 6,000 specimens from 34 herbaria, I have come to the conclusion that it is most reasonable to recognize six species for the state. One of these, Grindelia squarrosa (Pursh) Dunal, represented in California by G. squarrosa var. ser- rulata (Rydberg) Steyermark, is probably adventive, and occurs in only a few localities. It is not a major player in the hybridization reticula, and is easily distinguished from other California specimens by its bluntly dentate-serrulate leaf margins. Another species, Grindeliafraxino-pratensis Re- veal & Beatley, is known only from the Ash Mead- ows area of Nye Co., Nevada, and Inyo Co., Cali- fornia, and is not involved in hybridizations with other taxa. The involucres of G. fraxino-pratensis are broadly turbinate; all other grindelias in Cali- fornia have broadly campanulate to hemispheric heads. Grindelia nana Nutt. is found in northeastern

The Asteraceae are generally divided into two large subfamilies, the Cichorioideae (syn. Lactucoideae; tribes Mutisieae, Cardueae, Lactuceae, Vernonieae, Liabeae, Arctoteae) and the Asteroideae (Inuleae, Astereae, Anthemideae, Senecioneae, Calenduleae, Heliantheae, Eupatorieae). Recent phylogenetic analyses based on morphological and chloroplast DNA data show that the Mutisieae-Barnadesiinae are the sister group to the rest of the family. The Mutisieae-Barnadesiinae are here excluded from the Mutisieae and elevated to the new subfamily Barnadesioidae

Summary. Six new species are described and two new combinations are made in Psiadia which were found necessary during the preparation of the account of the Compositae-Astereae for the 'Flore des Mascareignes'. Notes are also provided for the seventeen other species of Psiadia and for the endemic genera Sarcanthemum and Rhamphogyne. The Astereae are represented by eight genera in the Mascarenes, but only three of these are native; these are Psiadia with twenty-five species, all endemic, and two endemic genera, Sarcanthemum and Rhamphogyne. The following new species and combinations are required for the account of the Astereae for the 'Flore des Mascareignes'. Full synonymy, keys and descriptions of all species will be given in the Flora account and are not given here to avoid repetition. All specimens have been seen unless otherwise stated. The taxonomy of the Mascarene species of Psiadia has previously been very confused. Baker (1877) recognised only three species, P. trinervia (Lam.) Willd., P. penninervia DC., and P. rodriguesiana Balf.f. Baker recognised seven varieties within P. trinervia; I have not been able to identify all of these, but several represent different growth stages of the heterophyllous P. arguta (Pers.) Voigt. Baker's var. macrodon from Rodrigues I identify as Sarcanthemum coronopus (Lam.) Cass. However, Baker did correctly recognise that P. glutinosa (Lam.) Jacq. was a Madagascan species and was not native on Mauritius.

Recent collections of Aster blepharophyllus A. Gray, the first since its discovery in 1851, prompted a reevaluation of its identity. It is properly placed in the genus Machaeranthera as M. gypsitherma Nesom, Vorobik, & Hartman. Within that genus, it shares characters with species of sect. Blepharodon as well as those of sect. Arida (both groups sensu Hartman), but we place it tentatively in the latter group. Of particular significance is its chromosome number of n = 5 pairs, reported for the first time, which is shared with species of sect. Arida.

Restriction site mutations analysis using 18 restriction endonucleases for chloroplast DNA's from North American Gutierrezia has led to a generic circumscription which is somewhat different from those previously proposed. Molecular data agree with a recent taxonomic treatment of Lane (1982, 1985), which transferred the x = 4 species from Xanthocephalum and submerged Greenella into Gutierrezia, but they do not support the submersion of the monotypic genus Thurovia. Thurovia is closely related as a sister group to Amphiachyris and should be resegregated from Gu- tierrezia. The study indicated that Xanthocephalum with x = 6 is very distantly related to Gutierrezia, even though Xanthocephalum has often been treated as part of the Xanthocephalum-Gutierrezia complex. Xanthocephalum is, in contrast, very closely related to Grindelia, which also has a base chromosome number of x = 6. Traditional species of Gutierrezia, G. californica, G. grandis, G. microcephala, G. sarothrae, G. serotina, and G. texana, constitute a closely related cluster and are well-separated from other purported members of the genus. Gutierrezia californica s. str. and G. bracteata should be separated again as distinct species. The herbaceous species, G. sericocarpa and G. sphaerocephala, appear to be the most primitive in the genus.

Remya (tribe Astereae) is a genus of three species endemic to the Hawaiian Islands. Until 1985, only two species, R. mauiensis and R. kauaiensis, were known. They were both thought to be extinct. Although rediscovered in 1971 and 1983, respectively, the plants are very rare and the populations endangered. A new species in the genus is described: R. montgomeryi sp. nov., known only from cliffs near the rim of Kalalau Valley, Kaua'i. The most diagnostic features distin- guishing it within the genus include complete lack of pubescence; long, thin peduncles; heads with a larger, hemispherical involucre; acute to obtuse phyllaries; and more numerous disk florets. Examination of material of all three species suggests that the heads are monoecious; the ray florets are pistillate and disk florets functionally staminate. Remya kauaiensis and R. montgomeryi differ from R. mauiensis in having dimorphic disk florets; the outer ones are smaller, have funnelform corollas, and apparently are neutral or pistillate whereas the inner ones are considerably larger, have cam- panulate corollas, and appear to be functionally staminate. Remya may be related to the Australasian genus Olearia. Remya is a small genus in the tribe Astereae. Three species can be distinguished, all endemic to the Hawaiian Islands. Two of them, R. kauaiensis Hillebrand and R. mauiensis Hille- brand, were described in 1888. Apparently nei

Tetramolopium is a genus of shrubby to suffrutescent perennial Compositae known only from New Guinea and the Hawaiian Islands. Evidence from biogeography suggests the genus was introduced into Hawaii from New Guinea recently, and perhaps originated from a single introduction. The morphological and ecological diversity of Tetramolopium in the Hawaiian Islands suggest there has been rapid divergence associated with adaptive radiation. Despite the diversity, all taxa are completely interfertile with fertile F1, F2, and F3 hybrids having been obtained. An electrophoretic study of allozymes revealed little divergence at 22 genes coding soluble enzymes. The mean genetic identity for pairwise comparisons of 19 populations from seven species is 0.95, a very high value normally obtained for conspecific plant populations. The results for Tetramolopium contrast sharply with other instances of high genetic identity between species, where taxa are similar morphologically and usually reproductively isolated. The enzyme data also support the hypothesis of recent, rapid divergence in Hawaiian Tetramolopium.

Amphiachyris is a North American genus of two species, A. dracunculoides and A. amoena. Related to Gutierrezia, it has in the past been placed in synonymy under that genus and also under Xanthocephalum. The generic status of A mphiachyris, as recognized by Solbrig, is accepted. Chromosome counts for both species and two species of Gutierrezia are reported. Counts of n = 4 for A. amoena and A. dracun- culoides and n 5 and n = 8 for G. texana are new. Preliminary chromatographic results indicate the presence of sulfated flavonoids in Amphiachyris. The genus Amphiachyris is composed of two annual species, one of which is restricted to the limestone-derived clay soils of the Edward's Plateau region of central Texas. The other, a very weedy species, is distributed widely on black prairie soils in the midwestern United States with scat- tered localities known as far east as South Carolina (Fig. 1). The distri- butions of Amphiachyris and of the annual species of Gutierrezia overlap throughout much of their ranges; these taxa are also quite similar in gross morphology. It is not surprising, therefore, that the taxonomic placement of Amphiachyris has been altered several times by various work- ers. The purpose of the present study is to evaluate the status of the taxon, using morphological, cytological and chemical methods.

Bradburia hirtella is a small annual endemic to eastern Texas and Louisiana. The species includes two cytotypes n = 3 and n = 4, the latter known from only one plant. Under cultivation an n = 4 individual existed as a short-lived perennial blooming several times over the course of 2.5 years. The two cytotypes were crossed and the 2n = 7 progeny survived and bloomed. Meiosis was studied in 2n = 6, 7, and 8 individuals. Three bivalents were typically observed in late prophase and metaphase I in pollen parent cells of 2n = 6 plants. Four bivalents were typical in the 2n = 8 plant. In meiosis in the 2n = 7 plants, bivalents and multivalents were observed. The mitotic meta- phase karyotypes were also compared. Both the 2n = 6 and the 2n = 8 plants had one large acro- centric pair and a smaller acrocentric satellite chromosome with a short arm with subequal satellite and proximal portions. The 2n = 8 plant had two small metacentric pairs. The 2n = 6 plants had only one large submetacentric pair. The presumed closest relatives are in Chrysopsis. Efforts to cross Bradburia with several species of Chrysopsis were unsuccessful. Flavonoid patterns in the two genera were similar but did not confirm or refute a close relationship. On the basis of karyotypes the n = 4 race of Bradburia appears to be intermediate between the n = 4 perennial C. mariana and the more specialized n = 4 annual C. pilosa. While B. hirtella could be transferred to Chrysopsis, the distinctive staminate disc florets of the former are considered sufficient to separate the two at the generic level.

Gutierrezia comprises sixteen North American species; two have two varieties each. Thurovia and Greenella are congeners of Gutierrezia, five taxa formerly of Xanthocephalum are mem- bers of Gutierrezia, and G. argyrocarpa is a distinct species. Characteristics shared by all taxa include alveolate receptacles that are pubescent with uniseriate trichomes that are swollen medially and long-attenuate apically, disc corollas that are clavate or funnelform and glabrous (except in one species), disc style-branch appendages that are linear with penicillate collecting hairs, achenes that are clavate or cylindrical and pubescent, and a pappus of scales that are united and/or greatly reduced in some taxa. In North America, the genus is distributed from the area of Mexico City northward throughout the Sierras Madres Occidental and Oriental, the central plateau of Mexico, and the central and western United States northward to western Canada. Habitats are well-drained, high montane slopes, mesic or semi-arid grasslands, and deserts.

Chaetopappa elegans, a narrow endemic from the White Mountains of south-central New Mexico, is unusual in Chaetopappa because of its double pappus of numerous bristles and narrow scales, comparatively long, lanceolate stylar appendages, crinkled stem pubescence, alveolate receptacle, faint cypsela nerves, long ligules, and occurrence on granitic rock. Its chromosome number is 2n = 9, In habit it resembles C. hersheyi from the nearby Guadalupe Mountains, but characteristics of the head are more like those of C. bellioides and C. pulchella from north-central Mexico and the Big Bend region of Texas, and Leucelene ericoides, widespread in arid southwestern North America. In its alveolate receptacle it resembles some species of Aster and the West Coast genus Pentachaeta, whereas its double pappus is reminiscent of some species of Erigeron and Aster (sect. Ionactis). The crinkly stem hairs are like those of some Erigeron species. A portion of the front ranges of southcentral New Mexico, the White and Sacramento mountains, form a continuous north-south oriented chain about 130 km long. The northern, primarily igneous portion, the White Mountains, rises to about 3660 m on Sierra Blanca, the southernmost glaciated peak in North America. The southern portion of the chain, the Sacramento Mountains, are primarily limestone and are considerably lower, barely exceeding 2740 m. This small mountain chain is isolated from other mountain ranges in the Southwest by fairly broad low gaps of arid grassland or desert, gaps that serve as barriers to migration of montane vegetation. To the north and south these gaps are relatively small and not particularly low (cf. map on p. 34 in Patterson 1980). During elevational fluctuation of vegetation in the late Pleistocene (Van Devender and Spaulding 1979) these gaps were narrowed even more and were, therefore, even less of a barrier to vegetation migration. On the one hand, the isolation of these mountains has produced an area of comparatively high endemism for New Mexico, while on the other the isolation has not been so complete that the vegetation has sharply differentiated. Plants from the higher elevations are mostly conspecific with those to the north; plants from lower elevations find their congeners mostly to the south. Most endemics in these mountains apparently have diverged but slightly from ancestral stock and show patterns of relationship similar to that of the general composition of the vegetation with which they are associated. The new species described herein adds one more endemic element to the White Mountains, one that is associated with a small genus of more southern affinity. Species of Chaetopappa are found in the mountains, hills, and plains in and surrounding the arid grassland regions of the northern portion of the Mexican Plateau, extending onto the southern Great Plains as far north as Kansas. This new Chaetopappa is unusual in that it occurs on granitic rocks at relatively high elevations, whereas its closest relatives all occur well to the south and then on limestone. It is also unusual in a number of morphological features. Shinners (1946a) redefined Chaetopappa to include a number of smaller genera previously distinguished by pappus differences. The publication of C. plomoensis B. Turner (Turner 1977) added a new kind of pappus to the genus. Our new species, with its pappus of numerous bristles and scales, further adds to pappus types known in the genus. In addition, it has stylar appendages, stem hairs, and a receptacle unique in Chaetopappa but similar in these characteristics to at least some species in Pentachaeta, Aster, and Erigeron. The faint nerves of the cypsela of C. elegans contrast to the prominent nerves on all other Chaetopappa. Chaetopappa elegans Soreng & Spellenberg, sp. nov. (fig. 1).-TYPE: U.S.A., New Mexico

Reports suggesting that the basic chromosome numberfor Aster section Heterophylli is x = 9 have been cited to support the hypothesis that x = 9 is the ancestral basic number for the whole genus. The present report gives chromosome counts for 35 populations in ten species, nine of which are traditionally placed in section Heterophylli. The chromosome numbers in 33 populations present a polyploid series of n = 8, 16, 24, and 32. Reinterpretation of photomicrographs by Avers (1954) lends additional support to the finding that the basic number for the section is x = 8 rather than x = 9. If counts for the three species not yet re-examined agree with these results, the balance between species of the x = 8 and x = 9 groups will be significantly shifted, and phylogenetic considerations based on the number of species belonging in each of the three "modes" (x = 5, 8, and 9) in the genus must be reassessed. Inasmuch as all Old World species of Aster exhibit x = 9, this group remains the largest on a worldwide basis, but with the inclusion of section Heterophylli, the x = 8 group comprises the largest number of species in North America, the generally acknowledged center of origin for the genus. The presentfindings are interpreted as supporting the hypothesis that evolution proceeded from the low basic chromosome number of x = 4 or 5 rather than the high one of x = 9. Aster is one of the relatively few members of the tribe Astereae that exhibits more than one mode of chromosome numbers, namely x = 5, 8, and 9. Several publications dealing with the phylogeny of the tribe are concerned with the question of which of the basic chromosome numbers is primitive, the high x = 9 or the low x = 4 or 5. Authors in favor of the hypothesis that the high number is ancestral (Huziwara, 1959, 1967; Raven et al., 1960; Solbrig, 1967; Solbrig et al., 1964, 1969) point out that the majority of species have the base number of x = 9 whereas x = 5 is found in a much smaller number of rather heterogeneous species adapted to arid regions and saline habitats. They reason further that the high base number is usually correlated with some other characteristics considered primitive in the Asteraceae, such as perennial and often somewhat woody habit and a high degree of polyploidy. In the view of these authors, Aster species with x = 5 chromosomes have been derived through successive aneuploid loss and subsequent catastrophic extinction of the less successful intermediate types-i.e., taxa with x = 6 or 7. The opposite stand is taken by Turner et al. (1961), who reason that 1I thank Christine A. Newell for checking some of my chromosome counts. 2 Botany, University of Illinois at Urbana-Champaign, Urbana IL 61801.

The eight taxa of Xanthocephalum, including X. durangense, X. gymnospermoides var. intermedium, and X. gymnospermoides var. eradiatum, are characterized by glabrous recep- tactes. goblet-shaped, pubescent disc corollas, deltate style branch appendages with short, rounded collecting hairs, and essentially glabrous achenes with very little or no pappus. The taxa are dis- tributed in damp, generally montane valleys from central Mexico northwest through the Sierra Madre Occidental to southern Arizona and western Texas. The chromosome number of all species is n = 6. Two natural interspecific hybrids are known. Xanthocephalum has affinities with other entities of Astereae that have x = 6, particularly Grindelia and Olivaea. Xanthocephalum has recently been reduced in number of species via the transfer of five taxa

Aster breweri is a rayless, high montane Californian species first described as Chrysopsis breweri and placed in C. subg. Ammodia (=Heterotheca sect. Ammodia). It has always been considered to be related to the goldenaster Heterotheca oregona (synonym: C. oregona) because both lack the well developed short outer pappus whorl and yellow rays typical for Heterotheca. However, Aster breweri has the diagnostic hair and phyllary traits of Aster sect. Eucephalus, not those of Heterotheca. Critical are its smooth-surfaced woolly hairs, lacking the tuberculae typical of Heterotheca. Chrysopsis breweri var. multibracteata is placed in synonymy under Aster brickellioides (including A. siskiyouensis). New combination: Aster breweri. Brewer's aster (basionym: Chrysopsis breweri A. Gray (1866); synonym: Heterotheca breweri (A. Gray) Shinners (1951)) was found to be unlike the true goldenasters (Compositae: Astereae) of the genus Heterotheca in that it lacked knobs on the osteolate cells of the uniseriate hairs (Sem

Xanthocephalum, Gutierrezia, Greenella, Gymnosperma, Amphiachyris, and Thurovia have been variously treated with regard to their generic limits by several authors in the last 3 decades. The generic delimitation presented here is based on a thorough re-evaluation of all 26 known North American members of these genera. The data obtained in the study strongly support the following disposition of the taxa, which differs from all previous treatments. Xanthocephalum (x = 6), a genus of montane areas in western Mexico and the southwestern United States, contains five species (one with three varieties) and is more closely related to Grindelia and Olivaea than to Gutierrezia and its allies. Gutierrezia (x = 4) encompasses seventeen specific and infraspecific taxa in North America, including not only the eight taxa recognized for it by Solbrig, but also five taxa formerly of Xanthocephalum, and the genera Greenella and Thurovia (the last resulting in Gutierrezia triflora, comb. nov.). The genus thus constituted ranges from central Mexico north and west through the central and western United States to Saskatchewan. Amphiachyris (n = 4 or 5) with two grassland species of Texas and central United States, and the monotypic Gymnosperma (x = 8), which ranges from northern Guatemala to Texas and westward to Arizona, are closely allied to Gutierrezia. The taxonomic problem presented by the group of genera that unfortunately have been called the "Xanthocephalum complex" is one of generic rather than specific circumscription. The genera concerned, besides Xanthocephalum, are Amphiachyris, Amphipappus, Greenella, Gutierrezia, Gymnosperma, and Thurovia. Taxonomic histories have been provided by Solbrig (1960a, 1960b, 196 1a, 196 1b, 1966), Ruffin (1974), Porter (1943), and Lane (1979, 1980b). A very brief summary of the several evaluations is as follows. A number of workers, including Bentham (1873) and Gray (1852, 1853, 1880a, 1880b, 1883), recognized the affinity of most of these genera, but for the most part maintained them as distinct entities. Shinners (1950) united all except Amphipappus and Thurovia under Xanthocephalum. Solbrig (1960b) again segregated them, while Ruffin (1974, 1977a) recommended that Gutierrezia and Greenella be submerged in Xanthocephalum and that Amphiachyris, Amphipappus, Gymnosperma, and Thurovia be considered distinct. It might seem unnecessary to reconsider this taxonomic problem in view of the several treatments already in the literature. However, data that have accumulated even since the most recent of these suggest that such a reconsideration is in order. None of the previous studies included every species of each genus; further, the various authors changed only the taxonomic rankings of whole, historically constituted genera. I wished

Monographic studies of Aphanostephus and Achaetogeron have made it necessary to investigate some of the related genera, particularly as to the value of certain morphological characters which are used in generic differentiation. The genera are quite similar in habit but the pappus, which furnishes important diagnostic characters, varies greatly within the group and is on the whole inconspicuous. For this reason it is necessary to make microscopic studies of the minute achenial and pappus characters upon which the differentiation of genera is primarily dependent. In this connection it was desirable to study Astranthium and Keerlia in detail; and it seems advisable to put on record the results which have been obtained relative to these genera. Astranthium, a genus of the Compositae belonging to the tribe Astereae, was described by Nuttall in 1841.1 It contained but one species, A. integrifolium, based on Bellis integrifolia Michaux.2 The 'Flora of North America,' published by Torrey and Gray in 1842, relegated Astranthium to synonymy. Since that time American and Mexican species have been merged with the genus Bellis3 which is indigenous to the Old World and especially to Europe. Three South American species have been described by Vellozo4 as occurring in Brazil. Two of these, Bellis campestris Vell. and B. pedunculata Vell., may well be members of the genus Spilanthes. The relationship of the third, Bellis scandens Vell., is unknown to me. Bellis perennis L. is an attractive plant, and for this reason it has been cultivated in the north Atlantic states where it has become naturalized. The American species which have been referred hitherto to the genus Bellis are so strikingly different that it seems strange the two generic elements should have been regarded

Chromosome numbers are reported for 85 collections of Compositae, representing 26 genera and 68 species, from Colombia and Venezuela. Most of the counts are previously unreported, including four genera, Blakiella (n = 9), Hinterhubera (n = 9), Oyedaea (n = ca. 28), and Paragynoxys (n = ca. 40), and about 30 species of Espeletia (n = 19). Chromosome counts for 85 collections of Colombian and Venezuelan Compositae are listed in Table 1. Bud material and the specimens were obtained by the junior author during his 1969 trips for collecting Compositae. The vials of buds were airmailed to the Department of Botany, Smithsonian Institution, and then relayed to Alpine, Texas, where counts were obtained by the senior author. Identifications were made by the junior author. Voucher specimens are deposited at US and COL or MER; duplicates will be distributed later. We are grateful to Robert M. King, Smithsonian Institution, for information regarding the genera of Eupatorieae. RESULTS AND DISCUSSION Eupatorieae-The chromosome numbers listed for Adenostemma (n = 10), Fleischmannia (n = 20), and Lourteigia (n = 10) correspond with previous records, although it should be noted that the species of Fleischmannia and Lourteigia were reported as Eupatorium (Cave, 1956-64; Powell & King, 1969). The count for Critonia paezense (n = 10) agrees with those obtained for five other species of this large genus, the others being reported as Eupatorium (R. M. King, personal communication). Astereae-Counts listed for the species of Baccharis (n = 9) and Conyza (n = 27) are consistent with well established base numbers for the genera. The report for Diplostephium glandulosum (n = ca. 9; Table 1) represents only the third species of this large genus for which chromosome numbers are available, with the other counts, for Peruvian species, also begin n = 9 (Turner, Powell & Cuatrecasas, 1967). Chromosome counts for Blakiella (n = 9) and Hinterhubera (n = 9) are first reports. The type species of Blakiella (Table 1) was first published by Blake as Podocoma, a very different genus from southern South America and Australia (Cuatrecasas, 1969). A single species of Podocoma from Brazil, P. hirsuta (n = 9), has been reported by Coleman (1968). According to the thesis advanced by Solbrig et al. (1964), the new generic counts would seem to offer a small measure of support for the idea that x = 9 is the base number for Astereae. 'Supported in part by NSF Grants GB-5806 and GB-6095, to J. Cuatrecasas. 21 acknowledge Sul Ross State University for the use of some facilities. 3Box 6064, Alpine, Texas 79830. 4Department of Botany, Smithsonian Institution, Washington, D. C. 20560. This content downloaded from 157.55.39.180 on Sat, 27 Aug 2016 06:22:23 UTC All use subject to http://about.jstor.org/terms

The overlooked genus Llerasia Triana (Compositae), which was synonymized with Vernonia by Bentham, is here reinstated and included in the Astereae. Its identity with the section Diplostephioides (Benth. & Hook.) Bl. of genus Haplopappus and its separation from this genus are discussed. The name Neosyris Greene is also a synonym. A key is given for the eleven species of Llerasia, which are trees, large shrubs, or woody lianas of the Andean forests. The intertropical distribution of the genus is emphasized. The distribution of the true remaining H[aplopappus species is split into two areas, one North American, the other Chilean-Argentinian. Eight new specific combinations are proposed and a new species (Llerasia hutchisonii) is described. HAVING HAD THE opportunity to compare the Andean species attributed to Haplopappus from Colombia, Ecuador, Peru, and Bolivia with many Haplopappus specimens from North America, Chile, and Argentina, I realize that they in fact do not belong to Haplopappus. The Andean species are large shrubs, small trees, or woody lianas with broad, coriaceous leaves; large corymbiform, paniculate, terminal inflorescences; homogam heads with cylindric, strongly imbricate, broad, obtuse or subacute phyllaries; a small number of hermaphrodite flowers (3-15); tubular corollas deeply cleft with long linear lobes; and pappus with 2-3 series of bristles. This set of characters is unique and defines as a genus the group of species with which we are dealing. Several botanists, in the past, called attention to the distinctiveness of this group. Bentham and Hooker (1873) ,' on the occasion of their treatment of Chrysothamnus, said of two species, Haplopappus hypoleucus and Baccharis fuliginea: "sectionem formant (Diplostephioidem) a caeteris habitu distinctissimam. Frutices sunt vel arbusculae. Folia lanceolata, subtus tomentosa vel villosa. Capitula more Diplostephii dense corymbosa. Involucra oblonga, bracteis obtusis. Corollae limbus profunde 5-fidus." In 1895, Greene made a new genus for these two species (Haplopappus hypoleucus and Baccharis fuliginea) which he called Neosyris, saying that they "are clearly distinct from this Chrysothamnus series (Bigelovia) by having broad revolute-margined coriaceous leaves that are conspicuously veiny, the veins almost pinnately divergent from the conspicuous midrib; by their thinnish, obtuse, closely appressed and regularly imbricated involucral bracts; their deeply-cleft corollas with linear and spreading segments; their short and lanceolate style-append