wi PHYTOLOGIA An international journal to expedite plant systematic, phytogeographical and ecological publication Vol. 76 January 1994 No. 1 CONTENTS Me REVEAL, J.L., New supraordinal names and recognition of five classes in PIA CAC at LNs ah amie RCA tetas weet e GR ak & WRK Ae 1 ea - TURNER, B.L., Taxonomic revision of Gentostemon (Gentianaceae). ....8 _- KING, R.M. & H. ROBINSON, New species of Adenostemma, Ayapana, and Brickellastrum from México, Panama, and Ecuador, and a new combi- nation in Grosvenorta (Eupatorieae: Asteraceae). .................. 14 ) ~ ROBINSON, H., New species of Ferreyranthus and Munnozia from Peru (Li- e plbeuer Macernmede rhe hiin cy Art le mal ye at ty ee a ee On 19 a -- ROBINSON, H., A new species of Oblivia and a new combination in Elaphan- A dra from Ecuador (Ecliptinae: Heliantheae: Asteraceae). ........... 24 ._. ROBINSON, H., New combinations in American Vernonieae (Asteraceae). CAEN TAAL Y CNB ER TT CA AAS NOP ARON CAVE A NER Bd AAS SF 27 \ .» MCMULLEN, C.K. & D.M. VIDERMAN, Comparative studies on the polli- nation biology of Darwiniothamnus tenutfolius (Asteraceae) and Plumbago scandens (Plumbaginaceae) on Pinta Island and Santa Cruz Island, Gala- TORR AAS RNA AIAN CAAA NMA ever SERIA ILN IA Uh CNL 30 -~ TURNER, B.L., Taxonomic study of the genus Synedrella (Asteraceae, He- TypriE ewe ie SL /a Whe Ra sta a alot a is tsla yaaa blolcle te einhot Yel hola aiglettareeaiert nie 39 a »~ TURNER, B.L., Taxonomic overview of Gilia, sect. Giliastrum (Polemoni- aneae) in: Merce wand! Mexteo: Fu Oy teh POA ube OQ AU UL 52 ..Contents continued on the inside cover. Published by Michael J. Warnock 185 Westridge Drive Huntsville, Texas 77340 U.S.A. PHYTOLOGIA is printed on acid free paper. (Contents continued) - TURNER, B.L., Two new gypsophilic species of Pinguicula (Lentibulari- aceae) from Nuevo Leén, Mexico. .......... see e cee e ee eee eee eeeees 69 LANDRY, P., A revised synopsis of the pines 5: The subgenera of Pinus, and their morphology and behavior. .........-.-seeee eee e rece seen ences 73 Books received. .c)6.cb46 did die tele oe lg wold dialed lntlel 9 alae lol ehahelln ial a ae 80. Index to reviewers in volume 75. ........ cc cere ccc escecsseeesteenssees 81 Disposition of manuscripts submitted for volumes 74 and 75. ........... 83 Publication dates for volume 75. .....626)..06 640050000 eles neieie ine p mleinaelag 84 PHYTOLOGIA (ISSN 00319430) is published monthly with two volumes per year — by Michael J. Warnock, 185 Westridge Drive, Huntsville, TX 77340. Second Class — postage at Huntsville, TX. Copyright ©1991 by PHYTOLOGIA. Annual domestic — individual subscription (12 issues): $36.00. Annual domestic institutional _ subscription (12 issues): $40.00. Foreign and/or airmail postage extra. Single copy > sales: Current issue and back issues volume 67 to present, $3.50; Back issues — (previous to volume 67), $3.00 (add $.50 per copy postage and handling US — [$1.00 per copy foreign]). Back issue sales by volume: $17.00 per volume rede (not all available as complete volumes); $21.00 per volume 67-present; add $2.00 — ‘per volume postage US ($4.00 per volume foreign). POSTMASTER: Send address changes to Phytologia, 185 Westridge Drive, Huntsville, TX 77340. FY Phytologta (January 1994) 76(1):1-7. NEW SUPRAORDINAL NAMES AND RECOGNITION OF FIVE CLASSES IN MAGNOLIOPHYTA James L. Reveal Department of Botany, University of Maryland, College Park, Maryland 20742-5815 U.S.A. ABSTRACT The flowering plants are divided into five rather than two classes, Magnoliopsida, Piperopsida, Liliopsida, Ranunculopsida, and Rosop- sida, corresponding to the magnolioides, the paleoherbs, the mono- cots, the ranunculids, and the eudicots. The subclasses Piperidae and Cornidae, and the superorders Piperanae, Capparanae, Cu- curbitanae, and Saxifraganae are proposed as new. A brief linear sequence of the higher taxa (above the rank of superorder) is presented for the flowering plants. KEY WORDS: Magnoliophyta, Magnoliopsida, Piperopsida, Lil- iopsida, Ranunculopsida, Rosopsida, Piperidae, Cornidae, Piper- anae, Capparanae, Cucurbitanae, Saxifraganae, classification, phy- logeny Ongoing research detailing the relationships among various families of an- giospermous plants (Magnoliophyta) is demonstrating the need for changes in the established system of classification (Cronquist 1981, 1988; Takhtajan 1987; Thorne 1992a, b) above the rank of order. Most of these recent efforts (Crane 1985; Chase et al. 1993; Doyle & Donoghue 1986, 1992; Donoghue & Doyle 1989; Loconte & Stevenson 1991; Qui et al. 1993) have shown that the traditional dicotyledonous group (Magnoliopsida) is not a monophyletic taxon although the monocotyledonous group (Liliopsida) is. To provide a work- able nomenclature for a classification reflective of the major highc: taxonomic trends within the flowering plants, namely the magnolioides, the paleoherbs, the ranunculids, the eudicots, and the monocots, it is suggested that each be recognized at the rank of class within Magnoliophyta as defined by Cronquist et al. (1966). 2 PHYTOL: GGA volume 76(1):1-7 January 1994 The use of names at a single rank for nested groups within a single taxon (Magnoliophyta) is a problem because not all members of the nested groups can be distinguished and the loss of phylogenetic information is exacerbated because the relationships of the nesting patterns are not expressed by single- rank names. The limitations of existing nomenclature preclude the recognition of numerous nested groups at different ranks as it is unrealistic to proliferate ranks to account for all the potentially recognizable groups. By using the rank of class for the five groups of flowering plants distinguished here one can at least recognize the major trends within Magnoliophyta, albeit with some still para-monophyletic. As for the formal taxonomic recognition of the unique groups for Ceratophyllum (a dicot: Loconte & Stevenson 1991; Chase et al. 1993) and for Acorus (a monocot: Duvall et al. 1993a, b), as distinct from the remainder of the dicots or monocots, I do not consider this worthy for nomenclatural and taxonomic reasons. Much remains to be done to confirm that these genera represent unique groups and that they are the most primitive of our modern dicots and monocots, respectively. The following linear sequence attempts to outline a system of classification for the flowering plants above the rank of superorder. Several superorders beyond those noted below remain to be published and until those names are available, a more detailed presentation is not appended. It should be noted that while such subclasses as Hamamelididae and Dillenidae are retained, their definitions differ greatly from the Cronquistian circumscriptions. Magnoliophyta Cronq., Takht., & Zimmerm., Taxon 15:134. 1966. A. Magnoliopsida Brongn., Enum. Pl. Mus. Paris xxvi, 95. 1843. 1. Magnoliidae Novak er Takht., Syst. Phylog. Magnolioph. 51. 1967. B. Piperopsida Bartl., Ord. Nat. Pl. 83. 1830. 2. Piperidae Reveal, Phytologia 76:3. 1994. C. Liliopsida Batsch, Regni Veg. 108. 1802. 3. Alismatidae Takht., Syst. Phylog. Magnolioph. 461. 1967. 4. Triurididae Takht. er Reveal, Novon 2:235. 1992. 5. Arecidae Takht., Syst. Phylog. Magnolioph. 425. 1967. 6. Liliidae Takht., Syst. Phylog. Magnolioph. 473. 1967. 7. Commelinidae Takht., Syst. Phylog. Magnolioph. 514. 1967. 8. Zingiberidae Cronq., Brittonia 30:505. 1978. D. Ranunculopsida Brongn., Enum. Pl. Mus. Paris xxvi, 96. 1843. 9. Ranunculidae Takht. er Reveal, Novon 2:235. 1992. Reveal: Five classes of Magnoliophyta 3 E. Rosopsida Batsch, Regnz Veg. 1. 1802. 10. Caryophyllidae Takht., Syst. Phylog. Magnolioph. 144. 1967. 11. Hamamelididae Takht., Syst. Phylog. Magnolioph. 113. 1967. 12. Dilleniidae Takht. er Reveal & Takht., Phytologia 74:171. 1993. 13. Rosidae Takht., Syst. Phylog. Magnolioph. 264. 1967. 14. Cornidae Frohne & Jensen ez Reveal, Phytologia 76:4. 1994. 15. Lamiidae Takht. ez Reveal, Novon 2:235. 1992. 16. Asteridae Takht., Syst. Phylog. Magnoltoph. 405. 1967. The following new names are required within Magnoliophyta. Piperidae Reveal, subcl. nov., validated by the Latin description associated with Piperopsida Bartl., Ord. Nat. Pl. 83. 1830 (as “Piperinae”). Piperanae Reveal, superord. nov., validated by the Latin description asso- ciated with Piperopsida Bartl., Ord. Nat. Pl. 83. 1830 (as “Piperinae”). By adopting Piperopsida for the paleoherbs it becomes necessary to pro- pose a name at the rank of class, and because the rank of superorder is now widely used (Takhtajan 1987; Dahlgren 1989a, b; Thorne 1992a, b; Reveal 1993)-albeit not recognized formally by the Code (Greuter et al. 1988), one at the rank of superorder is required as well. As defined here, Piperanae consists of three orders, Aristolochiales, Piperales, and Lactoridales, and four families Aristolochiaceae, Saururaceae, Piperaceae, and Lactoridaceae (Tucker et al. 1993). Chloranthaceae is referred to the Chloranthales and placed in Mag- noliopsida near Amborellaceae and Trimeniaceae in Illiciales (Endress 1987). The inclusion of Lactoridaceae is supported by their specialized flowers, but their anatropous ovules and follicular fruits are more typical of Magnoliales. The ultimate fate of this family remains to be ascertained. Capparanae Reveal, superord. nov., validated by the Latin description as- sociated with Order Capnanthemae Batsch, Regni Veg. 84. 1802, nom. wlleg. Takhtajan (1987) and Thorne (1992a, b) include Capparales within Vi- olanae, but evidence presented by Chase et al. (1993) shows the Capparales well removed from Violales. By placing Caricaceae in Caricales and associ- ating that order with the Capparales, as suggested by Rodman et al. (1993), the two groups become sufficiently distinct to require the recognition of a 4 PHY TOL@GTrAa volume 76(1):1-7 January 1994 new superorder. As here defined, the superorder includes Salvadorales (Sal- vadoraceae), Moringales (Moringaceae), Caricales (Caricaceae), Limnanthales (Limnanthaceae), Batales (Bataceae), Capparales (Pentadiplandraceae, To- variaceae, Resedaceae, Capparaceae, Brassicaceae) and Elaeocarpales (Elaeo- carpaceae). The family Gyrostemonaceae is also a member of the taxon, but the ordinal name remains to be validated. Cucurbitanae Reveal, superord. nov., validated by the Latin description associated with Order Cirrhatae Batsch, Regn: Veg. 220. 1802, nom. ulleg. The placement of Cucurbitales and its relatives is debatable. Cronquist (1981) and Thorne (1992a, b) retain the taxon in Violales while Takhtajan (1987) keeps the order in Violanae. The core membership of the group asso- ciated with Cucurbitaceae includes Datiscaceae and Begoniaceae, and Chase et al. (1993) have shown, based on rbcL data, that these families are not that closely related to Violales. In addition to the above three families (or four if one accepts Tetramelaceae), I have placed Coriariaceae in Cucurbitanae (in Cori- ariales), albeit provisionally, but well removed from the Ranunculales where placed by Cronquist (1981) and therefore closer to Rutanae (or Sapindanae) where aligned by Takhtajan and Thorne. The basis for this association is the preliminary data provided by Chase et al. Saxifraganae Reveal, superord. nov., validated by the Latin description as- sociated with Class Corniculatae Endl., Gen. Pl. 808. 1839, nom. illeg. The isolation of the Saxifragales from Rosales demonstrated by Morgan & Soltis (1993) also demands recognition of a separate superorder, distinct from. Celastranae, well removed from Rosanae, and situated more or less basally in Rosidae. Brexiales and Parnassiales are referred to Celastranae, with Hydrange- ales to Cornanae. As here defined, Saxifraganae consists of a series of orders (some not yet validated) which include such families as Greyi- aceae, Francoaceae, Crassulaceae, and Grossulariaceae. With the exception of Crossosomataceae, which is here tentatively included in Saxifraganae, all of these families are clearly related to Saxifragaceae. Cornidae Frohne & Jensen ez Reveal, subcl. nov., validated by the Latin description associated with Order Umbraculariae Batsch, Regni Veg. 40. 1802, nom. illeg. Reveal: Five classes of Magnoliophyta 5 The work published by Chase et al. (1993) and Xiang et al. (1993) shows Cornanae to constitute a group distinct from both Rosidae (where Hydrange- ales and Cornales have traditionally been placed) and Asteridae (where the Caprifoliales are typically assigned). As here defined, the subclass is composed of two superorders, Cornanae and Aralianae. The former is composed of four orders (Garryales, Aralidiales, Cornales, and Hydrangeales), the latter of six orders (Toricelliales, Pittosporales, Byblidales, Araliales, Viburnales, and Dip- sacales). Frohne & Jensen (1985, 1992) include Gentiananae in Cornidae, but this taxon is retained in Lamiidae. ACKNOWLEDGMENT Thanks are extended to B.E. Dutton and Dr. J.H. Wiersema for reviewing the manuscript. This is Scientific Article A-6541, Contribution No. 8752, of the Maryland Agricultural Experiment Station and Cooperative Extension Service. LITERATURE CITED Chase, M.W., D.E. Soltis, R.G. Olmstead, D. Morgan, D.H. Les, B.D. Mish- ler, M.R. Duvall, R.A. Price, H.G. Hills, Y.-L. Qiu, K.A. Kron, J.H. Ret- tig, E. Conti, J.D. Palmer, J.R. Manhart, K.J. Sytsma, H.J. Michaels, W.J. Kress, K.G. Karol, W.D. Clark, M. Hedren, B.S. Gaut, R.K. Jansen, K.-J. Kim, C.F. Wimpee, J.F. Smith, G.R. Furnier, S.H. Strauss, Q.-Y. Xiang, G.M. Plunkett, P.S. Soltis, S.M. Swensen, S.E. Williams, P.A. Gadek, C.J. Quinn, L.E. Equiarte, E. Golenberg, G.H. Learn, Jr., S.W. Graham, S.C.H. Barrett, S. Dayanandan, & V.A. Albert. 1993. Phylogenetics of seed plants: An analysis of nucleotide sequences from the plastid gene rbcL. Ann. Missouri Bot. Gard. 80:528-580. Crane, P.R. 1985. Phylogenetic analysis of seed plants and the origin of the angiosperms. Ann. Missouri Bot. Gard. 72:716-793. Cronquist, A. 1981. An Integrated System of Classification of Flowering Plants. Columbia University Press, New York, New York. . 1988. The Evolution and Classification of Flowering Plants, 2nd edit. The New York Botanical Garden, Bronx, New York. —_———, A. Takhtajan, & W. Zimmermann. 1966. On the higher taxa of Embryobionta. Taxon 15:129-134. PP N TT Osis © tre A volume 76(1):1-7 January 1994 Dahlgren, G. 1989a. “The last Dahlgrenogram: System of classification of the dicotyledons,” pp. 249-260. Jn: K. Tan, R.R. Mill, & T.S. Elias (eds.), Plant Tazonomy, Phytogeography and Related Subjects. Edin- burgh, Great Britain. . 1989b. An updated angiosperm classification. J. Linn. Soc., Bot. 100:197-203. Donoghue, M.J. & J.A. Doyle. 1989. “Phylogenetic studies of seed plants and angiosperms based on morphological characters,” pp. 181-193. Jn: B. Fernholm, K. Bremer, & J. Jornvall (eds.), The Hterarchy of Life: Molecules and Morphology in Phylogenetic Analysis. Amsterdam, The Netherlands. Doyle, J.A. & M.J. Donoghue. 1986. Seed plant phylogeny and the origin of the angiosperms: An experimental cladistic approach. Bot. Rev. 52:321- 431. . 1992. Fossils and seed plant phylogeny reanalyzed. Brittonia 44:89-106. Duvall, M.R., M.T. Clegg, M.W. Chase, W.D. Clark, W.J. Kress, H.G. Hills, L.E. Equiarte, J.F. Smith, B.S. Gaut, E.A. Zimmer, & G.H. Learn, Jr. 1993. Phylogenetic hypotheses for the monocotyledons constructed from rbcL sequence data. Ann. Missouri Bot. Gard. 80:607-619. , G.H. Learn, Jr., L.E. Equiarte, & M.T. Clegg. 1993. Phylogenetic analysis of rbcL sequences identifies Acorus calamus as the primal extant monocotyledon. Proc. Natl. Acad. Sci. USA 90:4641-4644. Endress, P.K. 1987. The Chloranthaceae: Reproductive structures and phy- logenetic position. Bot. Jahrb. Syst. 109:153-226. Frohne, D. & U. Jensen. 1985. Systematik des Planzenretchs, ed. 3. Stuttgart, Germany. . 1992. Systemattk des Planzenreichs, ed. 4. Stuttgart, Ger- many. Greuter, W., H.M. Burdet, W.G. Chaloner, V. Demoulin, R. Grolle, D.L. Hawksworth, D.H. Nicolson, P.C. Silva, F.A. Stafleu, E.G. Voss, & J. McNeill (editors.). 1988. International Code of Botanical Nomencla- ture adopted by the Fourteenth International Botanical Congress, Berlin, July-August 1987. Regnum Veg. 118. Loconte, H. & D.W. Stevenson. 1991. Cladistics of the Magnoliidae. Cladis- tics 7:267-296. Reveal: Five classes of Magnoliophyta 7 Morgan, D.R. & D.F. Soltis. 1993. Phylogenetic relationships among mem- bers of Saxifragaceae sensu lato based on rbcL sequence data. Ann. Missouri Bot. Gard. 80:631-660. Qiu, Y.-L., M.W. Chase, D.H. Les, & C.R. Parks. 1993. Molecular phyloge- netics of the Magnoliidae: Cladistic analyses of nucleotide sequences of the plastid gene rbcL. Ann. Missouri Bot. Gard. 80:587-606. Reveal, J.L. 1993. Automatically typified superordinal and ordinal names for the flowering plants (Magnoliophyta) as recognized by Thorne (1992) and arranged following the principles of priority, autonymy, and the sub- stitution of alternative names. Phytologia 74:203-263. Rodman, J., R.A. Price, K. Karol, E. Conti, K.J. Sytsma, & J.D. Palmer. 1993. Nucleotide sequences of the rbcL gene indicate monophyly of mus- tard oil plants. Ann. Missouri Bot. Gard. 80:686-699. Takhtajan, A.L. 1987. Systema Magnoltophytorum. Leningrad, U.S.S.R. Thorne, R.F. 1992a. An updated phylogenetic classification of the flowering plants. Aliso 13:365-389. . 1992b. Classification and geography of the flowering plants. Bot. Rev. 58:225-348. Tucker, S.C., A.W. Douglas, & H.-L. Liang. 1993. Utility of ontogenetic and conventional characters in determining phylogenetic relationships of Saururaceae and Piperaceae (Piperales). Syst. Bot. 18:614-641. Xiang, Q.-Y., D.E. Soltis, D.R. Morgan, & P.S. Soltis. 1993. Phylogenetic relationships of Cornus L. sensu lato and putative relatives inferred from rbcL sequence data. Ann. Missouri Bot. Gard. 80:723-734. Phytologia (January 1994) 76(1):8-13. TAXONOMIC REVISION OF GENIOSTEMON (GENTIANACEAE) Billie L. Turner Department of Botany, University of Texas, Austin, Texas 78713 U.S.A. ABSTRACT The genus Geniostemon, previously thought to have but two species, is revised as having four species: G. atarjanum B.L. Turner, spec. nov. (from easternmost Guanajuato, México); G. coultert (from Hi- dalgo, México); G. gypsophilum B.L. Turner, spec. nov. (from Nuevo Leon, México); and G. schaffneri (from San Luis Potosi, México). Re- vised and/or complete descriptions of each of the species are given, along with a key and distributional maps. KEY WORDS: Gentianaceae, Gentostemon, México Geniostemon was heretofore thought to be a small genus of only two species from northeastern México. It was first proposed by Engelmann & Gray in 1881 to accommodate G. coultert and G. schaffneri. The genus is largely distinguished from Centaurium by its 4-merous flowers and noncoiled anthers, and was maintained by Gilg (1895), who thought it not especially close to Centaurium, placing it between the genera Bisgoeppertia and Cicenda of his subtribe Erythriinae noting, however, that its pollen was similar to that of Neurotheca, a monotypic genus of South America. Except for its uncoiled anthers and 4-merous flowers, Geniostemon might fit comfortably into Centaurium, and I suggest here that a more intensive study might show its relationship to be within or near that large genus although, to my knowledge, this has not been suggested by yet other workers. Geniostemon Engelm. & A. Gray Erect or prostrate, annual or perennial (?) herbs to 15 cm high. Leaves opposite, small, linear-oblanceolate to ovate-lanceolate, the margins entire. Flowers terminal, one to a stem, or forming leafy terminal cymes by branching 8 Turner: Revision of Gentostemon 9 of the upper stems. Sepals 4, separate or nearly so. Petals 4, the tube some- what longer than the lobes, the latter pink or white. Stamens 4, inserted on the tube; filaments bearing numerous short glandular hairs at its mid-portions (rarely seemingly absent in Gentostemon gypsophilum B.L. Turner); anthers yellow, uncoiled. Pistil glabrous, the style fused throughout or nearly so, of- ten bifid at the apex, the stigmatic portion papillose. Capsule narrowly ovoid, dehiscent into halves, the seeds more or less ovoid or globose with an alveolate- like surface ornamentation. Type species: Gentostemon coulteri Engelm. & A. Gray Key to species 1. Stiffly erect or ascending annual herbs; corollas 4-8 mm long (tube-lobe length); San Luis Potosi and Nuevo Leon. 2. Corolla lobes mostly 4-5 mm long; anthers 0.8-1.2 mm long; foliage drying green; San. Luis Potosi. 2 a. Insect recorded for the first time as a flower visitor in the Galapagos Islands. ‘ Insect recorded for the first time as a flower visitor to this plant species. > Flowers recorded for the first time as visited by insects on Pinta Island. ® Insect recorded for the first time as a flower visitor on Pinta Island. McMullen & Viderman: Pollination biology in Galapagos Islands 35 moth (Lepidoptera: Gelechioidea). There were actually many more Tortrici- dae moths, T. crocker1, and D. marginalis present than could be accurately timed, so only a fraction of the activity that took place is recorded. Leptotes parrhasiotdes Wallengren (Lepidoptera: Lycaenidae) was the most common visitor to flowers of Plumbago scandens on Pinta (57 visits, 543 seconds). Other visitors included Cardtocondyla nuda Mayr (Hymenoptera: Formicidae), Lepidanthrarz tinctus, a species of Naucles (Coleoptera: Scrapti- idae), Ornebius erraticus Schudder (Orthoptera: Gryllidae), and a species of Paratrechina (Hymenoptera: Formicidae). The latter three were night visitors, and were not timed. On Santa Cruz, Plumbago scandens was visited most frequently by Phoe- bis sennae (Lepidoptera: Pieridae) (208 visits, 641 seconds). Also making visits were Leptotes parrhasioides, Urbanus dorantes galapagensis, Xylocopa darwini, and Wasmannia auropunctata (Hymenoptera: Formicidae). Flowers of this species were previously recorded as being visited on Santa Cruz by L. parrhasioides and P. sennae (McMullen 1986, 1990). DISCUSSION Darwiniothamnus tenuifolius is at least facultatively autogamous on both Pinta and Santa Cruz. With the level of insect activity observed at these plants, the open-pollinated inflorescences might be expected to produce more fruits than the bagged inflorescences. This was not the case, probably because the open-pollinated flowers were exposed to predation before being isolated with pollination bags. Finches were observed on several occasions feeding on the flowers and fruits of this species. In addition, due to differing rates of maturation within a corymbiform inflorescence, the fruits of some heads fell before the flowers of others had reached anthesis. Thus, some fruits were lost before the entire inflorescence could be bagged. From the timed observations, a species of Olcella and a species of Gonzozus appear to be most important as pollinators of Darwiniothamnus tenutfoltus on Pinta Island: However, it should be noted that Atteva hysginiella seemed to be much more common during casual observations than during the timed periods. Perhaps this insect was more aware of the observers, and avoided the plants at these times. Of the two visitors, Olcella and Gontozus, the latter may be more important in promoting cross-pollination since it spends less time on any one inflorescence. Three insect species may be important pollinators of D. tenuifolius on Santa Cruz. These are a Tortricidae moth, Tozomerus crockert, and Darwinysius marginalis. Based on the above logic, the latter two may be more useful as cross-pollinators. Plumbago scandens also produced fruits autogamously on both islands. Open-pollinated inflorescences showed a greater fruit set than bagged inflo- rescences. This would be expected given the insect activity during the timed 36 PH-YTOLOGIA volume 76(1):30-38 January 1994 observations, and suggests that the insect visitors may be important pollen vec- tors. However, crosses on Santa Cruz did not produce any fruits (McMullen 1987). Perhaps these insects are more important in effecting maximum selfing, while outcrossing is negligible. Darwiniothamnus tenutfolitus was newly reported as having its flowers vis- ited by insects in the Galapagos Islands. Plumbago scandens was recorded for the first time as having its flowers visited by insects on Pinta Island. In addition, D. tenutfohus and P. scandens, were newly reported as having their blossoms visited by Xylocopa darwint. Eleven of the insects in Table 2 were recorded for the first time as visiting flowers in the Galapagos Islands, while one was newly recorded as visiting flowers on Pinta Island. Two insects were newly reported as flower visitors to Darwiniothamnus tenutfolius, and three were newly reported as flower visitors to Plumbago scandens. Although pollen was not observed on all of the visitors, the high level of activity within flowers for many of them suggests that they may be functioning as pollinators to some extent. The fact that many of the visitors were lepidopterans is significant since these insects are common pollinators on the mainland. This adds support to their possible role as pollen vectors in the Galapagos Islands. In summary, the results of this study support the hypothesis that auto- gamous angiosperms were favored in the initial colonization of the Galapagos Islands (Rick 1966; McMullen 1987). This mode of reproduction may now be supplemented by nonautomatic selfing and cross-pollination due to the presence of various flower visitors. One of the more revealing discoveries of this research is the fact that, although few when compared to mainland standards, more insects visit flowers in this archipelago than previously believed. ACKNOWLEDGMENTS The authors wish to thank the following individuals: Alfredo Carrasco, Gonzalo Ceron, Gayle Davis-Merlen, Daniel Evans, Mary Lacey-Theisen, God- frey Merlen, Sandra J. Naranjo, Hugo Valdebenito, and Mr. and Mrs. George Viderman. Bryan E. Dutton and E. Gorton Linsley kindly reviewed an ear- lier version of this manuscript. The majority of insect identifications were made by the following individuals associated with the Systematic Entomol- ogy Laboratory, United States Department of Agriculture, Beltsville, Mary- land: T,.J. Henry (Hemiptera), R.W. Hodges (Lepidoptera), A.S. Menke (Hy- menoptera), D.A. Nickle (Orthoptera), J. Pakaluk (Coleoptera), D.R. Smith (Hymenoptera), and F.C. Thompson (Diptera). Additional identifications were made by N.L. Evenhuis (Diptera) of the Bernice P. Bishop Museum, Honolulu, Hawaii; and $.B. Peck (Coleoptera) and B.J. Sinclair (Diptera) of Carleton University, Ottawa, Canada. We also thank the staffs of the Charles McMullen & Viderman: Pollination biology in Galapagos Islands 37 Darwin Research Station and Galapagos National Park for their gracious as- sistance. This work was supported by National Geographic grant 4327-90 to CKM. LITERATURE CITED Aide, M. 1986. The influence of Xylocopa darwin: on floral evolution in the Galapagos. Charles Darwin Res. Sta. Annual Rep. 1983:19-21. Elisens, W.J. 1989. Genetic variation and evolution of the Galapagos shrub snapdragon. Natl. Geogr. Res. 5:98-110. Grant, B.R. & P.R. Grant. 1981. Exploitation of Opuntia cactus by birds on the Galapagos. Oecologia 49:179-187. Linsley, E.G. 1966. Pollinating insects of the Galapagos Islands. Pages 225-232. In: R.I. Bowman (ed.). The Galapagos. University California Press, Berkeley, California. —______, C.M. Rick, & S.G. Stephens. 1966. Observations on the floral relationships of the Galapagos carpenter bee. Pan-Pacific Entomol. 42:1- 18. McMullen, C.K. 1985. Observations on insect visitors to flowering plants of Isla Santa Cruz. I. The endemic carpenter bee. Not. Galapagos 42:24-25. 1986. Observations on insect visitors to flowering plants of Isla Santa Cruz. II. Butterflies, moths, ants, hover flies, and stilt bugs. Not. Galapagos 43:21-23. . 1987. Breeding systems of selected Galapagos Islands angiosperms. Amer. J. Bot. 74:1694-1705. . 1989. The Galapagos carpenter bee, just how important is it? Not. Galapagos 48:16-18. 1990. Reproductive biology of Galapagos Islands angiosperms. Monogr. Syst. Bot. Missouri Bot. Gard. 32:35-45. ———. 1993. Flower-visiting insects of the Galapagos Islands. Pan-Pacific Entomol. 69:95-106. ——— & D.D. Close. 1993. Wind pollination in the Galapagos Islands. Not. Galapagos 52:12-17. 38 PHY TOLOG IA volume 76(1):30-38 January 1994 Rick, C.M. 1966. Some plant-animal relations on the Galapagos Islands. Pages 215-224. Jn: R.I. Bowman (ed.). The Galapagos. University California Press, Berkeley, California. Usinger, R.L. 1972. Robert Leslie Usinger: autobiography of an entomolo- gist. The Pacific Coast Entomological Society, San Francisco, California. Phytologia (January 1994) 76(1):39-51. TAXONOMIC STUDY OF THE GENUS SYNEDRELLA (ASTERACEAE, HELIANTHEAE) Billie L. Turner Department of Botany, University of Texas, Austin, Texas 78713 U.S.A. ABSTRACT A taxonomic study of the monotypic genus Synedrella is rendered. It is represented by a single, highly variable, weedy species, S. nod:flora. Originally native to the New World tropics, it is now widely distributed in the tropical regions of the Old World and various island in between. An account is given of its taxonomic history, chromosome numbers, generic position, and worldwide distribution, including maps. KEY WORDS: Asteraceae, Heliantheae, Synedrella, tropical weed The genus Synedrella was established in 1791 by Gaertner, based upon Verbesina nodiflora L., which is typified by material collected in the West Indies (Jamaica) by Browne (Adams 1972; Howard 1989). According to most workers it is native to tropical America (perhaps the Caribbean Islands) but is now pantropical in distribution. Some authors (e.g., Adams 1972) have noted that the genus contains two species because of a second named taxon, Synedrella peduncularis Benth. from Ecuador, the latter transferred to the genus Schizoptera Turez. in 1916 by Blake, where it is properly positioned. Because of its weedy nature and pantropical distribution, Synedrella nodt- flora (L.) Gaertn. is accounted for in numerous regional or local floras. Never- theless, to my knowledge, infraspecific taxa within the species have not been proposed, nor would I suggest that such exist, there being considerable varia- tion of nearly all characters, both within and between populations. 39 40 PB OD OG A. volume 76(1):39-51 January 1994 CHROMOSOME NUMBERS There have been over 20 published articles in which chromosome counts for Synedrella are reported (cf. standard indices through 1989). It was first counted by Banerji & Pal (1959) as diploid with 2n = 40. Most subsequent workers have corroborated this number, except for a few workers who report counts of 2n = 32 (1 count), 2n = 36 (2 counts), and 2n = 38 (2 counts). Nirmala & Rao (1981, 1986) have reported the most recent anomalous counts for Synedrella (2n = 36 and 38, respectively), both counts from populations in India. Other workers, however, have reported yet other Indian populations as 2n = 40. In summary, the preponderance of counts suggest that the prevalent number in Synedrella is 2n = 40, suggesting an ancestral base number of z = 10. GENERIC RELATIONSHIPS Synedrella belongs to the tribe Heliantheae where it relates to a group of genera centering about the large pantropical genus Wedelia. Both Stuessy (1977) and Robinson (1981) positioned Synedrella in the subtribe Ecliptinae, the latter taxon mostly characterized by their compressed, often winged, car- bonized achenes; disk florets mostly hermaphroditic, having blackened anther thecae, and corollas with well-defined fiber sheaths. As treated by Robinson (1981) the subtribe contains 65 or more genera. Amongst these Synedrella seems closest to the poorly known South American genus Synedrellopsis Hi- eron. & Kuntze, with which it might comfortably be combined; at least | can find no significant characters to distinguish between them (the achenes of both Synedrella and Synedrellopsis are pictured in Robinson {1981]). Aside from Synedrellopsis, Synedrella appears closely related to Calyptocarpus Less. (cf. McVaugh & Smith 1967). According to restriction site analysis of chloro- plast DNA (Kim, K.-J., unpubl.), Calyptocarpus is clearly related to the genus Lastanthaea, a largely Mexican genus of the subtribe Ecliptinae. SYNEDRELLA Gaertn., nom. cons. Erect tap-rooted herbs. Leaves simple, opposite. Heads arranged 2-10 to a node, sessile or rarely pedunculate. Involucral bracts 2, herbaceous, tubes slender, as long or longer than the ligules; style branches slender and filiform with acute apices. Disk florets 4-merous; anther appendages plicate, wider than long; style branches linear-lanceolate, gradually acuminate. Achenes het- erocarpic, those of the ray tangentially flattened with stiff erose margins and erect pappus scales, those of the disk oblanceolate to clavate and somewhat Turner: Study of Synedrella 4] ‘ ‘ AN a —- Figure 1. North American distribution of Synedrella nodzflora. 42 Pop AYST OL O1GaiA volume 76(1):39-51 January 1994 “gS 5 SFG Figure 2. World distribution of Synedrella nodiflora. 43 Study of Synedrella Turner: 44 PHY TOLOGLA volume 76(1):39-51 January 1994 Figure 3. Mexican distribution of Synedrella nodiflora. Turner: Study of Synedrella 45 Figure 4. Central American distribution of Synedrella nodiflora (Guatemala to Panama). 46 PHYTOLOGIA volume 76(1):39-51 January 1994 4-sided, tangentially compressed, the pappus of 2-3 stiff divaricate awns. Base chromosome number, z = 20. Synedrella nodiflora (L.) Gaertn., Fruct. 2:456. pl. 171. 1791. BASIONYM: Verbesina nodtflora L., Cent. Pl. 1:28. 1755. Ucacou nodtflorum (L.) Hitchc., Rep. Missouri Bot. Gard. 4:100. 1893. According to Adams (1972), this name is typified by material collected in Jamaica by Browne. The illustration accompanying the original description and that of Adams (1972) leaves little doubt as to its correct application. Since the genus is monotypic the generic description provided here, as well as McVaugh’s excellent description and illustrations (1984) suffice to circum- scribe the species. As would be anticipated in a pantropical weedy species (Fig- ure 2) the taxon is exceedingly variable as to habit, foliage shape, and vestiture, but the floral characters are remarkable uniform throughout its range, except that the achenes are quite variable, especially as regards shape and degree of surface ornamentation of the disk achenes, these varying from tangentially flattened and without corky enations, to subquadrate and markedly corky, the 2-3 awns of the pappus varying from either stiffly erect to widely divergent. DISTRIBUTION - Synedrella nodiflora, as already noted, is a pantropical weed. It is likely that it was originally native to tropical America and subse- quently distributed elsewhere by modern man. The species is largely restricted to lowland regions from sea level to ca. 500 meters and is clearly weedy, most often found in disturbed habits, usually fallow fields. It flowers in all seasons, depending on rains and should be readily disseminated considering its small readily detached achenes with their lateral hooking devices. Because the plant is abundant in herbaria, and the fact that only a simple clearcut species is involved, I will let the distributional maps (Figures 1-4) document its occurrence as determined from sheets personally examined. I do provide below a brief listing of selected collections showing the location of appropriate voucher material and the earliest date of collection (shown in parentheses) within a region. NORTH AMERICA UNITED STATES. Hawaii: Honolulu, Degener 18308, NY (1925). Florida: Apalachicola, Chapman 6095, MO (1898; not collected since). MEXICO. Chiapas: Matuda 17229, NY (1947). Colima: McVaugh 24970, LL (1970). Guerrero: Hinton 10969, LL (1937). México: Hinton 5362, NY (1933). Michoacan: Hinton 7581, LL (1935). Quintana Roo: Lewis 6378 (1967). Tabasco: Cowan 1758, TEX (1978). Veracruz: Orcutt 3129, MO (1910). Yucatan: Millspaugh 185, F (1895). . Turner: Study of Synedrella 47 CENTRAL AMERICA GUATEMALA. Escuintla: Standley 64086, F (1939). Izabal: Ortiz 3555, F (1908). Jutiapa: Standley 75028, F (1940). Quezaltenango: Croat 32775, MO (1976). Retalhuleu: Molina 26967, F (1907). Santa Rosa: Standley 37907, F (1892). BELIZE. Gentle 108, F (1933). EL SALVADOR. Morazan: Tucker 493, F, NY (1941). HONDURAS: Choluteca: Standley 27758, F (1949). El Paraiso: Standley 16536, F (1949). Islas de la Bahia: Mohna 20742, F (1967). Morazan: Stan- dley 11658, F (1947). Ocotepeque: Molina 22540, F (1968). Yoro: Standley 53920, F (1927). NICARAGUA. Boaco: Stevens 14571, NY (1979). Chinandega: Grialva 597, F (1979). Esteli: Nelson 7704, NY (1968). Granada: Sandino 1281, F (1981). Jinotega: Stevens 15434, MO,NY (1979). Managua: Stevens 3431, F (1969). Masaya: Stevens 4244, F (1977). Nuevo Segovia: Moreno 13623, MO (1981). Rio San Juan: Sandino 1744, F,NY (1981). Zelaya: [poly 3546, MO (1979). COSTA RICA. Heredia: Hammel 7883, MO (1980). Limon: Davidson 6908, MO (1978). Puntarenas: Kernan 53 (1988). PANAMA. Bocas del Toro: Dunlap 314, F (1924). Chiriqui: Croat 21922, MO (1973). Cocle: Rodreguez 1173, MO (1947). Darien: Duke 4143, MO (1961). Herrera: Tyson 3136, MO (1966). Los Santos: Burch 1265, MO (1966). Panama: Macbride 2795, F (1923). San Blas: Croat 16964, MO (1971). Veraguas: Tyson 5170, MO (1968). W/O Locality: “Colon” O. Kuntze s.n., NY (1841). WEST INDIES BAHAMA ISLANDS. Abaco: Brace 1959, NY (1904). Crooked Island: Brace 1906, F,NY (1906). Eleuthra: Correll 48937, NY (1977). Great Exuma: Correll 44089, NY (1975). New Providence: Hitchcock s.n., F (1890). GREATER ANTILLES: CUBA. Camaguey: Shafer 120, NY (1909); Ha- vana: Rugel 1849, NY (1849); Matanzas: Britton 370, NY (1903). Oriente: Shafer 1581, F,NY (1909); Pinar del Rio: Alain 425, NY (1943); Santa Clara: Combs 478, F,NY (1895); Cuba: Wright s.n., NY (1856-57). DOMINICAN REPUBLIC. Taylor 764, F,NY (1909). GRAND CAYMAN. Armour 1276, F (1859). HAITI. Nash 1037, NY (1905). JAMAICA. Lloyd 1121, F,MO (1890). PUERTO RICO. Garber 27, NY (1880). TRINIDAD. Skousted 33, MO (1935). VIRGIN ISLANDS. Acevedo 1888, NY (1987); D’Arcy 5106, MO (1971); Saint Croix, Ricksecker 31 F,NY (1895); Saint Thomas, Kuntze 102, NY (1874). 48 PAY TODOGTA volume 76(1):39-51 January 1994 LESSER ANTILLES: ANTIGUA. Boz 9386, F (1937). BARBADOS. Bovell 380 (1901). DOMINICA. Lloyd 465, NY (1903). GRENADA. Broad- way s.n., MO (1904); w/o data, F (1904). GUADELOUPE. Duss 2496, NY (1902). MARTINIQUE. Hahn 390, NY (1871). MONTSERRAT. Shafer 27, F,NY (1907). SAINT BARTHELEMY Forsstrom 7227, F (w/o date). SAINT KITTS. Britton 136, NY (1901). SAINT VINCINT. Smith 723, NY (w/o date). TOBAGO: Eggers 5527, NY (1889). CURACAO. Boldingh s.n., NY (1913). SOUTH AMERICA BOLIVIA. Gentry 44241, MO (1984). Beni: Solomon 6273, MO,NY (1981). Nor-Yungas: Solomon 7372, F,MO,NY (1982). Pando: Nee 31596, MO,NY (1980). COLOMBIA. Bolivar: Kilkp 14286, NY (1926). Caldas: Cuatrecasas 23104, F (1946). Choco: Gentry 17257, MO (1976). Cundinamarca: King 5888, F,NY (1965). Los Llanos: Cuatrecasas 1975, F (1938). Magdalena: Andre s.n., F (1875). Norte de Santander: Fabrega 850, F (1944). Tolima: Pennell 3376, MO (1917). Valle: Cuatrecasas 22596, F (1946). ECUADOR. Chimborazo: Camp E-3048, NY (1945). El Oro: Daly 114, NY (1978). Esmeraldas: Asplund 16443, NY (1955). Galapagos Islands: Snow 346, NY (1963). Guayas: Mille 1006, F (1937). Los Rios: Dodson 7051, F, MO (1978). Morona-Santiago: Brandbyge 32238, NY (1980). Napo: Bowbray 69927, MO (1969). Pichincha: Harling 9288, F,MO,NY (1968). w/o locality: Andre 483, NY (1975). PERU. Junin: Woytkowski 7862, MO (1962). Loreto: Croat 17989, MO (1972). Ucayali: MacRae 15, F (1981). VENEZUELA. Aragua: Pittter 148, MO (1927). Bolivar: Boom 6271, MO (1985). Delta Amacuro: Steyermark 115217, MO (1977). Fed. Distr.: Batley 206, MO (1921). Guarico: Davidse 4189, MO (1976). Isla Margarita: Miller 131, F;MO (1901). Miranda: Croat 21685, MO (1973). Monagas: Trujillo 9467, F (1969). Portuguesa: Smith 892, MO (1982). Sucre: Fernandez 3751, NY (1980). Tachira: Croat 54960, MO (1982). Zulia: Bunting 11339, NY (1982). BRITISH GUIANA. Persand 150, F (1923). FRENCH GUIANA. Feuillet 1695, F,MO (1985). SURINAM. Coulen 331, MO (1841). BRAZIL. Bahia: Chase 7879, MO (1924). Ceara: Drovet 2234, F (1935). Minas Geraes: Chase 9261, MO (1925). Para: Dahlgren 451, F (1929). Per- nambuco: Pickel 19, F (1930). w/o locality: Rusby 233, F,MO (1895). Turner: Study of Synedrella 49 ASIA AMBOINA. Robinson 1835, NY (1913). BANGLADESH. Khan 4312, MO (1976). CELEBES. Kauderns 258, NY (1917). CEYLON. Millspaugh 2498, F (1912). CHINA. Hainan: Chow 78475, MO,NY (1978). Kwong Tung: Levine 1899, MO (1917). HONG KONG. Millspaugh 2735, F (1911). INDIA. Madras, Pres. Madras Coll. s.n., NY (1929). JAPAN. Iwatsuki 682, MO (1976). MALAYSIA. Elmer 20709, MO (1922). PHILIPPINES. Merrill 36, NY (1902). SINGAPORE. w/o collector, MO (1904). SUMATRA. Toroes 3649, NY (1933). TAIWAN. Gressitt 485, NY (1934). THAILAND. King 5562, F (1963). VIETNAM. Squires 28, NY (1927). AFRICA BURUNDI. Lambinon 78/1383, MO (1978). CAMEROON. Gandoger s.n., MO (1906). GAMBIER. Chapin 935, NY (1934). GHANA. Ent: 69, NY (1971). HAUTE-VOLTA. Georges 15403, MO (1958). IVORY COAST. Amshoff 637, MO (1972). KENYA. Robertson 3319, MO (1982). KIVU. Alcool 8961, MO (1958). LIBERIA. Baldwin 6949, MO (1947). MALAWI. Pawek 8267, MO (1974). NIGERIA. Ekwuno 63767, MO (1971). REP. CENTRAFRICAINE. Leeuwenberg 6252, MO (1965). S. TOME. Viegas s.n., NY (1948). SENEGAL. Georges 582, MO (1948). SIERRA LEONE. Thomas 2955, MO (1914). TANZANIA. Tanner 1917, NY (1955). UGANDA. Dummer 3264, MO (1917). ZAIRE. Louis 11542, NY (1938). PACIFIC ISLANDS: CAROLINES. Carr 11093, NY (1935). COOK. Yuncker 9867, NY (1940). ELLICE. Chambers 46, MO (1974). FIJI. De- gener 13518, F (1940). FRENCH POLYNESIA. Florence 1982, NY (1952). 50 PHYTOLOGIA volume 76(1):39-51 January 1994 MARIANAS. Kanehira 2215, NY (1933). MARQUESAS. Chapin 762, NY (1934). NEW CALEDONIA. Baumann-B. 18744, NY (1951). SAMOA. Vau- pel 2, MO (1905). SEYCHELLES. Jeffrey 437, NY (1961). SOCIETY. Moore 36, MO (1926). SOLOMON. Riley 23, NY (1945). TONGA. Hurlimann 522, NY (1926). EXCLUDED NAMES Synedrella peduncularis Benth., Pl. Hartw. 119. 1843. Blake (1916, zn Hook. Ic. Pl. t. 8058) correctly positioned this name in the genus Schizoptera, S. peduncularis (Turez.) S.F. Blake, where it replaced the later name S. trichotoma Turez. (1851). ACKNOWLEDGMENTS This study is based upon the examination of approximately 866 sheets from the following herbaria (number on loan given in parenthesis): Field Museum, Chicago, F (213); C.L. Lundell, University of Texas, Austin, LL (60); Missouri Botanical Garden, MO (321); New York Botanical Garden, NY (232); Univer- sity of Texas, Austin, TEX (40). I am grateful to the directors concerned for these loans. Especial thanks to Guy Nesom and Piero Delprete for reviewing the manuscript. LITERATURE CITED Adams, C.D. 1972. Flowering Plants of Jamaica. Univ. of the West Indies, Mona, Jamaica. Banerji, I. & S. Pal. 1959. A contribution to the life history of Synedrella nodtflora Gaertn. J. Linn. Soc. London, Bot. 55:798-803. Howard, R.A. 1989. Synedrella, In Flora of the Lesser Antilles 6:600. McVaugh, R. 1984. Synedrella in Fl. Nova-Galictana 12:908-910. University of Michigan Press, Ann Arbor, Michigan & N.J. Smith. 1967. Calyptocarpus vialis and wendlandi (Composi- tae). Brittonia 19:268-272. Turner: Study of Synedrella 51 Nirmala, A. & P.N. Rao. 1981. In Chromosome number reports 70. Taxon 30:78. . 1986. Karyotypic studies in some Asteraceae. Kromosoma II 42:1311-1315. Robinson, H. 1981. A revision of the tribal and subtribal limits of the Heliantheae (Asteraceae). Smithsonian Contr. Bot. 51:1-102. Stuessy, T.F. 1977. Heliantheae - a systematic review. In Biol. Chem. Comp. 2:621-672. (V.H. Heywood et al., eds.; Academic Press, London, Great Britain). Phytologia (January 1994) 76(1):52-68. TAXONOMIC OVERVIEW OF GILIA, SECT. GILIASTRUM (POLEMONIACEAE) IN TEXAS AND MEXICO B.L. Turner Department of Botany, University of Texas, Austin, Texas 78713 U.S.A. ABSTRACT A taxonomic treatment of Gilia sect. Gtliastrum is rendered for Texas and México. Five species are recognized as occurring within Texas: G. inctsa, G. insignis, G. ludens, G. rigidula, and G. stewartit; and seven species occur in México: G. gypsophila, G. incisa, G. insig- nis, G. palmert, G. purpust (including G. platyloba), G. rigidula, and G. stewartit. One new species, G. gypsophila B.L. Turner sp. nov., is described and illustrated. Chromosomal data are reviewed for the sect. Giliastrum as a whole, and distribution maps of the Texas and Mexican taxa are provided. Complete synonymy is given for all of the taxa concerned, along with appropriate taxonomic comments. KEY WORDS: Polemoniaceae, Gilia, Giliastrum, México, Texas The sect. Gilastrum of Gilia was first erected by Brand (1907) to accom- modate Gila rigidula Benth. in DC. and its allies. He placed the section in his subgenus Benthamtophila, which in today’s nomenclature must be called subgenus Giulia since it contains the type element of Gilia. Brand recognized five species under his concept of Gilastrum: G. incisa Benth. in DC., G. foetida Gill. ex Benth., G. palmeri S. Wats., G. purpusit K. Brandegee (not to be confused with G. leptantha Parish subsp. purpusi: [Mlkn.] A. & V. Grant) and G. rigidula, the latter divided into two subspecies (rigidula and insignis Brand), one of these (rigidula) further divided into two varieties (acerosa A. Gray and rigidula). Grant & Grant (1956) provided a conspectus of the subgenus Giliain which they recognized four sections: Arachnion, Gilia, Gilmania, and Saltugilia. The sect. Gihastrum was not included in the subgenus at that time, but Grant (1959) subsequently modified their 1957 infrageneric treatment of Gilia to include five sections: Gila, Giliandra, Arachnion, Saltugilia, and Giliastrum 52 Turner: Taxonomy of Gila sect. Giltastrum 53 (which included the sect. Gilmanza of Grant & Grant 1957). So construed, Giliastrum was said to contain eleven North American species and at least one, and possibly more, South American species. Following Grant’s (1959) broad overview of sect. Giliastrum, two new species were added to the section by Shinners (1963), Gulia perennans Shin- ners, which I treat as synonymous with G. incisa, and G. ludens Shinners, which is recognized. Two Mexican taxa, G. insignis (Brand) Cory & Parks and G. stewartw I.M. Johnst., both of which occur in Texas, are recognized here as specifically distinct and G. platyloba I.M. Johnst. is treated as syn- onymous with the Mexican species, G. purpust. I have also added a newly described species G. gypsophila. Five species of the sect. Giliastrum are now recognized as occurring in Texas, and seven in México. Altogether, the sect. Gilastrum currently con- tains approximately seventeen species, most of these native to North America, but at least a few are confined to South America, as noted in the chromosomal account that follows. CHROMOSOME NUMBERS Chromosome numbers are available for only eight of the approximately seventeen species of the sect. Giliastrum, as noted in Table 1. Most of the species have chromosome numbers on a base of z = 9, with the exception of Gilia insignis (2n = 12) which apparently has an extant base number of z = 6. Indeed, if the latter were thought to be the ancestral base number for the section, one might view the remainder of the taxa as being either hexaploids (2n = 18) or dodecaploids (2n = 36). Grant (1959), because of the frequency and phyletic distribution of species having numbers on a base of z = 8 and 9, inferred that the base number of the family as a whole was z = 9; the relatively few taxa with z = 6 were thought to be derived via aneuploid reduction. This might be the case for Gilta insignis because, except for the anomalous count of 2n = 20 for G. rigidula var. acerosa, all of the taxa counted to date appear to be on a base of z = 9. Regardless, the base number of z = 6 found in G. insignis strongly suggests that it is best treated as a distinct species, the rank accorded the taxon by Shinners (1963), Wherry (1966) and yet others (Henrickson, in prep.) than as a subsp. of G. rigidula as originally proposed by Brand. KEY TO TEXAS AND MEXICAN SPECIES OF GILIA SECT. GILIASTRUM 1. Corollas mostly 4-7(-8) mm long; basal-most leaves frequently simple or MEGDIVESEDrabes © Viel Sic.ck cee aie eee he Se Ean ee ee G. incisa 54 PRY EOLO GTA volume 76(1):52-68 January 1994 1. Corollas mostly (7-)8-20 mm long; basal-most leaves variously dissected or pinnately. lobed... 5.25.4.05 6200s 4424s ain. foc 6 ee ee = (2) 2. Calyx lobes united for 1/3 their length or less. ................ (3) 2. Calyx lobes united for 1/3 their length or more. ............... (4) 3. Corollas mostly (7-)8-9 mm long. ............... esse ee eee G. stewartu 3. Corollas. mostly 10-13 mm long. .........05.....00> anaes G. purpusi 4. Leaves simple or merely 3-parted, the blades or their divisions linear, when 3-parted the terminal lobe 3-5 times as long as the lateral lobes; Baja California; .............02--...0sa0- = eee G. palmerz 4. Leaves not as described in the above; not in Baja California. .. (5) 5. Corollas mostly 18-22 mm long, at anthesis broadly flaring, 22-32 mm across the expanded lobes... .25.0< 5: «52 sisms ose) eee G. insignis 5. Corollas mostly 10-18 mm long, at anthesis 15-20 mm across the expanded TOB€S.. os 2 keine eee ete bees niece a0 oa ance aloe nce ne ear (6) 6. Perennial, stiffly erect much-branched suffruticose herbs, the basal leaves scarcely persisting, the stem leaves numerous and usually acerose or pungently pointed. -.../..0-...ceeess eee G. rigidula 6. Annual or weakly perennial sparsely branched herbs, the basal leaves usually persisting, the stem leaves not especially acerose or pungent. 1 ca gain asia sn-ninn we obtains ys oe anf errr (7) 7. Plants with primary stems stiffly erect and somewhat zigzag; basal leaves coriaceous with deeply incised simple lobes, their apices acute; gypseous soils of Nuevo Leon, Mexico... 1... .-- «+0. 0o eRe G. gypsophila 7. Plants not as described in the above; basal leaves mostly pinnately lobed or dissected, the ultimate units mostly broadly ovate to oblanceolate in outline: southcentral: Texas. ...2...0...02.0 008 pee G. ludens Gilia gypsophila B.L. Turner, sp. nov., Figure 1. TYPE: MEXICO. Nuevo Leon: 14 km N of Rancho las Estacas, on road to Rancho Lechuguillal (26° 28’ 30” N, 100° 50’ W), gypsum flat, 680 m, 16 Mar 1973, M.C. Johnston, T.L. Wendt, & F. Chiang 10222 (HOLOTYPE: LL!). Turner: Taxonomy of Gilia sect. Giliastrum SG Figure 1. Gila gypsophila, from holotype. 55 56 PHY DOLO:GLA volume 76(1):52-68 January 1994 Giliae rigidulae Benth. in DC. var. acerosae A. Gray similis sed plantis majoribus strictisque rosulas persistentes basales parien- tibus, foliis midcaulis vix acerosis secus caules separatis internodiis plerumque 1-2 cm longis (vs. 0.3-1.0 cm) differt. Stifly erect perennial herbs 15-25 cm high. Stems 2-10 arising from a ligneous tap root, prominently glandular-pubescent throughout, the hairs 0.2 mm long or less. Basal leaves persistent as a well-defined rosette of coriaceous laciniately lobed leaves 3-5 cm long, 0.6-1.6 cm wide, minutely pubescent like the stems, the petioles 1.5-2.5 cm long; blades oblanceolate in outline, the lobes acute with callous apices. Stem leaves sessile or shortly petiolate, not at all acerose, well-spaced along the somewhat zig-zag stem, the internodes mostly 1-2 cm long. Flowers 3-10 per stem, arranged in terminal open cymes; the pedicels mostly 5-20 mm long. Flowering calyces ca. 5 mm long, the lobes linear-lanceolate, united for 1/2-3/5 their length, minutely glandular- pubescent dorsally. Corollas apparently lavender or bluish, rotate, ca. 16 mm long with lobes extended, the lobes ca. 12 mm long, apices broadly rounded. Anthers yellow, ca. 2 mm long. Capsule decidedly globoid, ca. 4-5 mm high, ca. 4.1 mm wide. Seeds (immature) oval, ca. 2 mm long. ADDITIONAL SPECIMEN EXAMINED: MEXICO. Nuevo Leon: Minas Manto Blanco y Sabana Blanca, just N of the Canon de Potrerillos (26° 04’ N, 100° 45’ W), gypsiferous clay loam, 950-1000 m, 17 Mar 1973, Johnston et al. 10250c (LL). This taxon is clearly related to Gilta rigidula but differs in having a persis- tent rosette of coriaceous leaves, numerous, rather zig-zag stems with elongate midstem internodes. It appears to stand somewhat between G. rigidula var. acerosa and G. ludens, as noted by J. Henrickson on the paratype. As shown in Figure 1, G. gypsophila occurs within the southern range of G. rigidula var. acerosa, but these apparently do not co-occur at a given site. Gila incisa Benth. in DC., Prodr. 9:312. 1845. Navarretia incisa (Benth.) Kuntze, Rev. Gen. Pl. 2:433. 1891. Polerontum inctsum (Benth.) Kuntze, Rev. Gen. 3:203. 1898. TYPE: U.S.A. Texas: Central Texas, 1835, T. Drummond 463 (HOLOTYPE: K; Isotypes: GH,TEX!). Gila lindhetmeriana Scheele, Linnaea 21:753. 1848. TYPE: U.S.A. Texas: Comal Co., vicinity of New Braunfels, 1844-1850, F. Lind- hetmer (holotype not located). Gila perennans Shinners, Sida 1:174. 1963. TYPE: U.S.A. Texas: Culberson Co., Guadalupe Mts., McKittrick Canyon, 18 Aug 1946, D.S. Correll 13958 (HOLOTYPE: SMU). Turner: Taxonomy of Giulia sect. Giltastrum Bit This is a widespread variable species but readily distinguished by its mostly simple unbranched stems, relatively weakly dissected thin leaves, and small corollas. It is sympatric with most of the Texas and Mexican taxa, except for Gilia ludens (Figure 2), preferring relatively moist shady habitats. Gilia insignis (Brand) Cory & Parks, Texas Agr. Exptl. Sta. Bull. 550:85. 1938. Gutha rigidula Benth. in DC. subsp. tnszgnis Brand in Engler, Pflanzenr. IV. 250:147. 1907. TYPE: MEXICO. Coahuila: Jimulco Station, ca. 25° 07’ N, 103° 20’ W, 16 May 1885, C.G. Pringle 248 (LEC- TOTYPE [selected here]: B, destroyed; Isolectotypes: GH!,UC!,VT!). This is a well-marked relatively widespread species, readily distinguished by its wiry stems, large corollas on elongate peduncles, and chromosome numbers on a base of z = 6 (Table 1). Nevertheless, Wherry (1966) retained it as a subspecies of Gulia rigidula. Gilia ludens Shinners, Sida 1:174. 1963. TYPE: U.S.A. Texas: Jim Wells Co.: 4 mi W of Alice, 10 Apr 1955, L.H. Shinners 19581 (HOLOTYPE: SMU). This is a relatively well-marked taxon what with its somewhat sprawling or recumbent habit with persistent basal leaves. It is superficially similar to Gila rigidula var. rigidula, but as indicated in figures 4 and 6, the two do not normally co-occur. Gilia palmert S. Wats., Proc. Amer. Acad. Arts 24:61. 1889. TYPE: MEX- ICO. Baja Calif. Norte: Stony ridges near Los Angeles Bay, 1887, E. Palmer 593 (HOLOTYPE: GH; Photoholotype: CAS!). Gilia palmeri S. Wats. subsp. spectabilis A. Day in Shreve & Wiggins, Veg. & Fl. Sonoran Desert 2:1164. 1964. TYPE: MEXICO. Baja Calif. Norte: Hills of Sierra del Volcan, 4 mi E of El Marmol, 13 Feb 1935, I. Wiggins 7570 (HOLOTYPE: DS!). A very distinct member of the sect. Giliastrum, not likely to be confused with another. Very large-flowered populations have been recognized as subsp. spectabilis, but there appears to be much variation in corolla size both within and among populations, to judge from collections at CAS. Gilta purpusi K. Brandegee, Zoe 5:179. 1904. TYPE: MEXICO. Coahuila: Viesca, ca. 25° 20’ N, 102° 26’ W, Mar 1904, C.A. Purpus 533 (HOLO- TYPE: CAS!). 58 PADDY TiO: 01Gb A volume 76(1):52-68 January 1994 Figure 2. Distribution of Gulia gypsophila (triangles), and G. ludens (circles). Turner: Taxonomy of Gilta sect. Giliastrum Table 1. Chromosome numbers in Giulia sect. Giltastrum.** Species G. campanulata A. Gray G. filiformis Parry Q G. incisa G. incisa G. incisa G. incisa G. insignis G. insignis G. latefoha S. Wats. G. rigidula var. rigidula G. rigidula var. rigidula G. rigidula var. rigidula . foetzda Gill. ex Benth. Source, reference, and/or voucher U.S.A. Nevada: Esmeralda Co. (Grant 1959) U.S.A. California: Inyo Co. (Grant 1959) ARGENTINA. Mendoza: Las Heras (Covas & Schnack 1946) MEXICO. San Luis Potosi: San Luis Potosi (Grant 1959) MEXICO. San Luis Potosi: Rio Verde (Grant 1959) U.S.A. Texas: Travis Co. (Grant 1959) U.S.A. Texas: Uvalde Co. (Grant 1959) MEXICO. Chihuahua: 20 mi NE Carmago (Weedin & Powell 1978) MEXICO. Nuevo Leon: 5 mi N of Espinoza (Weedin & Powell 1978) U.S.A. California: Riverside Co. (Grant 1959) Flory 1937; seed source unknown. U.S.A. Texas: Edwards Co. (Grant 1959) U.S.A. Texas: Hayes Co., (Grant 1959) 2n number 18* 18* 18* 18* 18* 18* 18* 6 prs 12 prs 36* 18* 36* 36* 60 PHYTOLOGIA volume 76(1):52-68 January 1994 Figure 3. Distribution of Gila incisa. Turner: Taxonomy of Gilia sect. Giliastrum 61 Figure 4. Distribution of Gila insignis (open circles) and G. palmerz (closed circles). 62 PHY TOL0 GTA volume 76(1):52-68 January 1994 Figure 5. Distribution of Gilia purpusii. Typical collections from calcareous soils at lower elevations shown as closed circles; atypical collections from pine forests in gypseous soils at higher elevations shown as open circles. 63 Turner: Taxonomy of Gulia sect. Giltastrum Figure 6. Distribution of Gilia rigidula var. acerosa (open disks); and var. rigidula (closed disks). January 1994 volume 76(1):52-68 PHY: TOLGGIA 64 Figure 7. Distribution of Gilia stewartit. Turner: Taxonomy of Gila sect. Giliastrum 65 Table 1 (continued). Species Source, reference, and/or voucher 2n number G. rigidula U.S.A. Texas: Travis 36* var. rigidula Co. (Grant 1959) G. rigidula U.S.A. Texas: Val Verde 36* var. rigidula Co. (Grant 1959) G. rigidula U.S.A. Texas: Brewster 10 prs var. acerosa Co. (Weedin & Powell 1978) G. rigidula U.S.A. New Mexico: Dona 9 prs var. acerosa Ana Co. (Ward & Spellenberg 1986) G. purpusi MEXICO. Nuevo Leon: Mpio. 9 prs Galeana, Turner 93-154 (TEX) Mitotic counts. * As modified by Grant (1959). Gilia platyloba 1.M. Johnst., J. Arnold Arb. 24:95. 1943. TYPE: MEXICO. Coahuila: Saltillo, 1898, E. Palmer 799 (HOLOTYPE: GH!). This is a widespread highly variable taxon largely distinguished from Gilta stewartii by its larger corollas and leaves which tend to have broader ultimate divisions. Type material of G. purpusi has pinnately divided leaves with very broad ultimate divisions, while type material of what has been called G. platyloba has leaves with relatively narrow ultimate divisions. Between these extremes occur a large number of intermediates scattered over much of the range of the species. Selected examples of such intermediates follow: Coahuila: 1 mi SW of Las Delicias, Henrickson 6056 (TEX); 39 mi (air) NE of Tlahualilo, Henrickson 12192 (TEX); ca. 67 mi (air) SW of Cuatro Cienegas, Henrickson 12475 (TEX); 27 mi SE of Torreon, Henrickson 13226 (TEX); near Cuatro Cienegas, Bacon et al. 1089 (TEX); ca. 15 mi S of Cuatro Cienegas, Johnston et al. 10335 (LL); a few mi W of Las Delicias, Stewart 2817 (GH). Durango: Cerro de San Ignacio, Purpus 4595 (GH). Most of the specimens of this species which I examined tend to have leaves with relatively narrow segments; otherwise these differ little, if at all, from specimens with broader leaf segments. I have included under the fabric of Gilia purpusi, ten specimens from pine forests of Nuevo Leon, mostly in the vicinity of Galeana, which were collected 66 PEY TOLOGIA volume 76(1):52-68 January 1994 on gypseous soils and which I think belong to a distinct taxon (Figure 5). Such plants have generally more pinnately divided lower leaves with smaller ultimate divisions and their corollas, at least in the field, are decidedly blue, vs. pale blue or lavender as in the more typical elements of G. purpusit, the latter occurring at usually lower, more xeric habitats (most often with Agave lechugilla and associates). I have counted chromosomes of individuals of the Galeana populations and these are diploid with n=9 pairs (Turner 93-154, TEX). Additional study of this complex in the field is needed before formal recognition is tendered. Giha rigidula Benth. in DC., Prodr. 9:312. 1845. Two well-marked regional varieties (Figure 6) are recognized under this taxon, as follows: 1. Leaves with their divisions mostly stiffly linear, acerose; widespread in western Texas, closely adjacent states and México. ........ var. acerosa 2. Leaves with most or many of their divisions broader, scarcely acerose; central'and southern Texas... ... 2... 4/..0005.00-:oe see oe var. rigidula Gilia rigidula Benth. in DC. var. rigidula. Giliastrum rigidulum (Benth.) Rydb., Fl. Rocky Mts. 1066. 1917. Gila glandulosa Scheele, Linnaea 21:753. 1848. TYPE: U.S.A. Texas: Comal Co.: vicinity of New Braunfels, 1846, F. Roemer s.n. (HOLO- TYPE: B, probably destroyed). This taxon is readily distinguished from the following and is largely al- lopatric with it. Occasional intermediates occur in the regions of overlap. Giha rigidula Benth. in DC. var. acerosa A. Gray, Proc. Amer. Acad. Arts 8:280. 1870. Gulia acerosa (A. Gray) Britt., Man. Bot. N.E. St. 761. 1901. Gihastrum acerosum (A. Gray) Rydb., Fl. Rocky Mts. 699. 1917. TYPE: U.S.A. Texas: Pecos Co., Escondido Creek, 30 Jun 1852, C. Wright 551, exsic. no. 1645 (LECTOTYPE [selected here]: GH!; Isolec- totypes: [2] GH!). As noted in the above, a distinctive taxon well worthy of recognition. Brand (1907) recognized both acerosa and rigidula as varieties, but several workers have treated these as distinct species. Phytologra (January 1994) 76(1):69-72. TWO NEW GYPSOPHILIC SPECIES OF PINGUICULA (LENTIBULARIACEAE) FROM NUEVO LEON, MEXICO Billie L. Turner Department of Botany, University of Texas, Austin, Texas 78713 U.S.A. ABSTRACT Two new species of Pingutcula are described from southern Nuevo Leon, México: P. jorgehintonii B.L. Turner and P. hintoniorum B.L. Turner. Both are believed to be gypsophilous endemics and relate to the recently described P. essertana B. Kirchner, a calciphile from Tamaulipas, Mexico. KEY WORDS: Lentibulariaceae, Pzngutcula, México, Nuevo Leon Identification of gypsophilous species from Nuevo Leon, México has re- vealed the following two novelties, both known only by collections of the Hin- ton family. Pinguicula jorgehintonii B.L. Turner, sp. nov. TYPE: MEXICO. Nuevo Leon: Mpio. Aramberri, N of Aramberri, 1145 m, 23 Nov 1993, G.B. Hinton et al. 24000 (HOLOTYPE: TEX!). Pingutculae essertanae B. Kirchner similis sed plantis minoribus (4-6 cm altis vs. 8-11 cm), pedicellis glabris (vs. conspicue pubescen- tibus), et corollis plus minusve regularibus lobis plerumque brev- ioribus (5-15 mm longis vs. 15-20 mm) etiam calcaribus brevioribus (5-12 mm longis vs. 15-20 mm). Annual (?) herbs 4-6 cm high. Leaves of the early rosettes oblanceo- late to spatulate, 5-10 mm long, 1.0-1.5 mm wide; leaves of flowering rosettes thin, broadly obovate, 1.2-1.8 cm long, 6-8 mm wide, sparsely pubescent with multiseptate hairs. Pedicels mostly 3-6 cm long, glabrous. Calyces zygomor- phic, the lobes acute, 1.0-1.5 mm long, minutely sparsely glandular-pubescent. 69 66 PHYTOL OGiIA volume 76(1):52-68 January 1994 on gypseous soils and which I think belong to a distinct taxon (Figure 5). Such plants have generally more pinnately divided lower leaves with smaller ultimate divisions and their corollas, at least in the field, are decidedly blue, vs. pale blue or lavender as in the more typical elements of G. purpuszz, the latter occurring at usually lower, more xeric habitats (most often with Agave lechugilla and associates). I have counted chromosomes of individuals of the Galeana populations and these are diploid with n=9 pairs (Turner 93-154, TEX). Additional study of this complex in the field is needed before formal recognition is tendered. Gilia rigidula Benth. in DC., Prodr. 9:312. 1845. Two well-marked regional varieties (Figure 6) are recognized under this taxon, as follows: 1. Leaves with their divisions mostly stiffly linear, acerose; widespread in western Texas, closely adjacent states and México. ........ var. acerosa 2. Leaves with most or many of their divisions broader, scarcely acerose; central:and southern Texas. 2... 35.0. ...<%2« s)emtoegereie ene var. rigidula Gilia rigidula Benth. in DC. var. rigidula. Guiliastrum rigidulum (Benth.) Rydb., Fl. Rocky Mts. 1066. 1917. Gilia glandulosa Scheele, Linnaea 21:753. 1848. TYPE: U.S.A. Texas: Comal Co.: vicinity of New Braunfels, 1846, F. Roemer s.n. (HOLO- TYPE: B, probably destroyed). This taxon is readily distinguished from the following and is largely al- lopatric with it. Occasional intermediates occur in the regions of overlap. Gika rigidula Benth. in DC. var. acerosa A. Gray, Proc. Amer. Acad. Arts 8:280. 1870. Gulia acerosa (A. Gray) Britt., Man. Bot. N.E. St. 761. 1901. Giliastrum acerosum (A. Gray) Rydb., Fl. Rocky Mts. 699. 1917. TYPE: U.S.A. Texas: Pecos Co., Escondido Creek, 30 Jun 1852, C. Wright 551, exsic. no. 1645 (LECTOTYPE (selected here]: GH!; Isolec- totypes: {2] GH!). As noted in the above, a distinctive taxon well worthy of recognition. Brand (1907) recognized both acerosa and rigidula as varieties, but several workers have treated these as distinct species. Phytologra (January 1994) 76(1):69-72. TWO NEW GYPSOPHILIC SPECIES OF PINGUICULA (LENTIBULARIACEAE) FROM NUEVO LEON, MEXICO Billie L. Turner Department of Botany, University of Texas, Austin, Texas 78713 U.S.A. ABSTRACT Two new species of Pinguicula are described from southern Nuevo Leon, México: P. jorgehintonii B.L. Turner and P. hintoniorum B.L. Turner. Both are believed to be gypsophilous endemics and relate to the recently described P. essertana B. Kirchner, a calciphile from Tamaulipas, Mexico. KEY WORDS: Lentibulariaceae, Pinguicula, México, Nuevo Leon Identification of gypsophilous species from Nuevo Leon, Mexico has re- vealed the following two novelties, both known only by collections of the Hin- ton family. Pinguicula jorgehintonii B.L. Turner, sp. nov. TYPE: MEXICO. Nuevo Leon: Mpio. Aramberri, N of Aramberri, 1145 m, 23 Nov 1993, G.B. Hinton et al. 24000 (HOLOTYPE: TEX!). Pingutculae essertanae B. Kirchner similis sed plantis minoribus (4-6 cm altis vs. 8-11 cm), pedicellis glabris (vs. conspicue pubescen- tibus), et corollis plus minusve regularibus lobis plerumque brev- ioribus (5-15 mm longis vs. 15-20 mm) etiam calcaribus brevioribus (5-12 mm longis vs. 15-20 mm). Annual (?) herbs 4-6 cm high. Leaves of the early rosettes oblanceo- late to spatulate, 5-10 mm long, 1.0-1.5 mm wide; leaves of flowering rosettes thin, broadly obovate, 1.2-1.8 cm long, 6-8 mm wide, sparsely pubescent with multiseptate hairs. Pedicels mostly 3-6 cm long, glabrous. Calyces zygomor- phic, the lobes acute, 1.0-1.5 mm long, minutely sparsely glandular- pubescent. 69 70 PAY T OL O'GPA volume 76(1):69-72 January 1994 Sea Len >. @ aS ee Figure 1. Ptngutcula jorgehintonit, from holotype. Turner: New gypsophilic Pinguicula from Nuevo Leon 71 Corollas pink to purplish-white, + zygomorphic to nearly regular, the lobes 5-8 mm long, 5-8 mm wide, the throat (fully developed) ca. 8 mm long, ca. 6 mm across, the spurs mostly 5-8 mm long, glabrous or nearly so, the orifice of the throat with numerous slender glandular trichomes. Fruiting material not available. ADDITIONAL COLLECTION EXAMINED: MEXICO. Nuevo Leon: Mpio. Galeana, road from Agua Blanca to La Purisima, pine-oak woodland, 3 Mar 1992, G.B. Hinton et al. 21812 (TEX). This taxon is similar to Pingutcula esseriana, both species having relatively reduced thin leaves and relatively short pedicels. Pingutcula jorgehintoni can be distinguished from the latter by its nearly regular corollas, glabrous, shorter pedicels (3-6 cm long vs. 7-10 cm) and shorter glabrous spurs (5-8 mm long vs. 15-20 mm). The type is accompanied by numerous excellent close up photographs of the plant. Living specimens were collected in the field and taken to the home of Jaime and George Hinton and maintained in their garden until flowering at which time the photos were taken. It is a pleasure to name this delicate species for George Hinton, grandson of the well known Mexican collector, G.B. Hinton. The collection from Mpio. Galeana (cited above) has larger flowers with longer spurs and may represent a different taxon. The above formal description is taken from type material. The Galeana plants have spurs up to 15 mm long and corolla lobes up to 15 mm long. Pinguicula hintoniorum B.L. Turner, sp. nov. TYPE: MEXICO. Nuevo Leon: Mpio. Zaragoza, Dr. Arroyo, below Puerto Pino, gypsum hillsides, 1975 m, 16 Mar 1993, G.B. Hinton et al. 22661(HOLOTYPE: TEX). Pinguiculae esserianae Kirchner similis sed corollis malvinis vel purpuratis calcaribus glabris longioribusque (2-3 cm long vs. 1.5- 2.0 cm) et pedicellis glabris differt. Annual herbs 5-12 cm high. Leaves of the early rosettes oblanceolate, mostly ca. 1 cm long, 2-5 mm wide; leaves of flowering rosettes thin, broadly obovate, mostly ca. 2 cm long, 0.4-0.6 cm wide, sparsely pubescent with mul- tiseptate hairs. Pedicels mostly 4-10 cm long, glabrous or nearly so. Ca- lyces zygomorphic, 6-8 mm across, the lobes acute, 3-4 mm long, minutely glandular-pubescent, especially along the margins. Corollas reportedly “pur- ple” or “mauve”, strongly zygomorphic, the lobes oblanceolate, 8-16 mm long, 4-9 mm wide, the throat (fully developed) 3-6 mm long, 2-4 mm across; spurs 2-3 cm long, glabrous or nearly so, the orifice of the throat sparsely pubescent with relatively few multiseptate hairs. 72 PHY. POwk OGiaA volume 76(1):69-72 January 1994 ADDITIONAL COLLECTION EXAMINED: MEXICO. Nuevo Leén: Mpio. Zaragoza, below Puerto Pino, gypsum hillside, 1840 m, 8 Feb 1989, G.B. Hinton et al. 19312 (TEX). This taxon superficially resembles Pinguicula essertana but the pedicels are glabrous (vs. densely pubescent) and the corollas are a deeper purple with longer glabrous spurs. It differs from P. jorgehintonz in having strongly zygo- morphic purple corollas with shorter tubes and much longer spurs. I take pleasure in naming this taxon for the remarkable Hinton family, which includes George, eponymized above, who has taken a special interest in the gypsophilous pinguiculas of southern Nuevo Leon. ACKNOWLEDGMENTS I am grateful to Guy Nesom for the Latin diagnoses, and him and Mark Mayfield for reviewing the manuscript. Especial thanks to D. Kearns of MO for providing literature of recently described species of Pinguicula from México. Piero Delprete provided the illustration. Phytologia (January 1994) 76(1):73-79. A REVISED SYNOPSIS OF THE PINES 5: THE SUBGENERA OF PJNUS, AND THEIR MORPHOLOGY AND BEHAVIOR Pierre Landry 50 Dussault, suite 306, Hull, Québec, J8X 4G2 CANADA ABSTRACT Seven subgenera of Pinus are proposed on the basis of morpholog- ical and behavioral characters. The taxonomic separations are made by detaching special species or groups of species from the two older subgenera Pinus and Strobus. From subgenus Pinus, three “new” sub- genera are detached: Pinea, Sabinia, and Tamaulipasa. From subgenus Strobus, two “new” subgenera are detached: Gerardia and Balfourta. Clear keys are given: one morphological, the other proposing behav- ioral differences. KEY WORDS: Pinus, Pinaceae, systematics INTRODUCTION Starting with a brief history of the nomenclature, we recall that Koehne (1893, pp. 28, 30) keyed the pines according to only two basic groups: The Diplozylon or “Hard Pines” with needles having two vascular bundles, and the Haplozrylon or “Soft Pines” having only one vascular bundle in their needles. Then Shaw (1914, pp. 1, 24, 25) accepted totally Koehne’s taxonomy, but he preferred to set his key otherwise: instead of mentioning the vascular bundles, he mentioned that the said two subgenera “are even more accurately charac- terized” by the fact that in Diplorylon the bases of the “fascicle-bracts” that subtend the sheaths wrapping the lower parts of the needles are decurrent on the shoots, while in Haplozylon, the said bases are not decurrent on the shoots. But recently Carvajal & McVaugh (1992, pp. 38, 50, 95) reported that that system does not hold very well. In some of the Haplorylon species (with only one vascular bundle), namely the Pinon Pines and Pinus rzedowski Madrigal & Caballero, the situation differs. By the end of the current growing season 73 74 PHYO LOGLiLA volume 76(1):73-79 January 1994 and during the following autumn and winter, “it is usually evident that the said bases are indeed decurrent”. Nevertheless, in accordance with well established custom, our key presented here below retains the same basic division, while specifying that 1) the dis- tinction between the the number of vascular bundles is discarded, 2) the easy to observe distinction of the decurrence is retained, 3) but we must also look at the recently elongated section of the twig as it appears at the end of the growth season and during the next autumn (dead season). More precisely we recall that the said “recently elongated section of the twig” is also termed “young shoot” or “current season’s twig” in modern botan- ical literature. It must also be understood that the end of the growth season varies according to latitude and altitude. However, it can be witnessed by the state of development of the winter buds situated at the very tip of the twig. Fully grown buds, no longer increasing in size, indicate that the growth season is over. Many manuals, including that of Krussmann (1985, p. 207), show a drawing of both striped (“furrowed”) shoots due to the decurrent bases, and smooth shoots due to the non decurrent bases. Photos are also helpful, and here we present Photo 1 of the striped shoot of Pinus sylvestris L., and Photo 2 of the smooth shoot of Pinus strobus L. Then, within each basic division, we key further subgenera, which we dis- tinguish according to the easy-to-see characters of the cones and seeds. Yes, the mature cones are generally easy to observe and the fundamental seed types (wing present or absent, long or short, retained or not) can be found readily when one looks at the interior side of the cone scale, where the seed marks and the wing marks (if they exist) are evident. A behavioral key is furnished to justify the morphological one. It con- sists of an attempt to show the interaction of both time and efficiency with morphology. Logic-wise it is systematically more practical to divide a genus of about 98 species into more subgenera than the two previously recognized ones. We are thus proposing seven. Each “new” subgenus includes extraordinary species or groups of species, well and clearly defined. MORPHOLOGICAL KEY A. Twigs (aged one year older than current season’s) striped (photo 1). ..B B. Cones with double concentric umbos. ............. subgenus Pinea BB. Cones. with-simple umbos. |. : ....... s/: . sis isin eee C C. Umbos all around the middle of the cones are terminal, long (at least 5 mm) and stout. .......0.:sneeee subgenus Sabinia Revised synopsis of the pines 5 75 Zed + < y “ ; ii # cae see i Sn “Wire ns 7 PG” ag Leigh pa ff wT f ~ ; 4 7 A W ey Me, . 1% % Figure 1. Pinus sylvestris, one-year old twig showing the striped rough surface constituted by the decurrent needle sheaths. 76 PHY TOUOGTA volume 76(1):73-79 January 1994 Figure 2. Pinus strobus, showing the smoother surface resulting from the non-decurrent needle sheaths. Even the twig’s two-year old section is smooth except that the small bumps are bigger. Shaw (1914, p. 1) cautions the reader that the difference between those two characters is better seen on long, vigorous twigs because the intervals between the needle sheaths (or fascicles) are wider. Landry: Revised synopsis of the pines 5 (ue CC. Umbos at least on the interior face of the cones are dorsal, Shor (atmost 5 mm)iand small: |... 23,505.64 40.yes-00 sno n