Pacific Science (1979), vol. 33, nO. 1 © 1980 by The University Press of Hawaii. All rights reserved

Vicia menziesii Sprengel (Fabaceae) Rediscovered: Its Taxonomic Relationshipsl

J. STUART LASSETTER 2 and CHARLES R. GUNN 3

ABSTRACT: Taxonomic relationships among five taxa of Vicia, traditionally recognized at the species level, were studied. Data used included plant and seed morphology, seed amino acid content, palynology, and cytology. Two Old World species, V. dumetorum Linnaeus and V. pisij'ormis Linnaeus, are shown not to be closely related to the Hawaiian and American taxa. The Hawaiian species, V. menziesii Sprengel, is maintained as a species and is shown to be closely related to V. gigantea Hooker of North America and V. nigricans Hooker & Arnott ofSouth America. The latter two species are treated as subspecies under the name V. nigricans. The karyotype and chromosome number are reported for the first time for the endangered species, V. menziesii.

Vicia menziesii Sprengel, which is endemic to first three eophylls each bearing four leaflets the island of Hawaii and was last collected and a tendril. in 1915, has been regarded as extinct. It was A review of·New and Old World vetches rediscovered in 1973 on the same island by located four other species with diagnostic Wayne Gagne and Mae Mull (c. H. Lamou­ characters similar to those of Vicia menziesii: reux and D. Herbst, personal communica­ V. nigricans, V. gigantea, V. pisiformis, and tions). It is now considered an endangered V. dumetorum. Although Kupicha (1976) species (United States Department of the apparently combines V. nigricans Hooker & Interior 1978). This species was not men­ Arnott of Chile and Argentina and V. tioned by Kupicha (1976) when she presented gigantea Hooker of the Pacific Northwest of a general worldwide taxonomic treatment of North America, she neither formally pro­ the genus. posed this combination nor made any tax­ Vicia menziesii exhibits these diagnostic onomic judgment at the subspecific level. characters: (1) plants large, rank and darken­ These three New World taxa were originally ing or blackening with age or on drying; described as species: V. menziesii in 1826, (2) mature flowers rose-purple; (3) oblong V. nigricans in ·1830, and V. gigantea in 1831. standard; (4) compressed styles with encircl­ They traditionally have been· treated as ing apical hairs; (5) stipitate legumes without species in various floras or monographs of inner "woolly" parenchymatous tissue; (6) their respective geographical regions (Degen­ seeds with hilum occupying 75 percent of the er, Degener, and Gunn 1970, Hermann 1960, seed circumference and containing canava­ Reiche 1898, Welsh 1974). Vicia pisiformis nine; (7) six pairs of acrocentric chromo­ Linnaeus of middle Europe and V. dume­ somes, one pair with satellites, and one torum Linnaeus of Eurasia have been treated submetacentric pair; and (8) seedlings with as species in various floras or monographs of the .first eophylls at the fourth node and the their respective geographical regions (Ball 1968, Fedtschenko 1948), and both were studied by Kupicha. 1 Manuscript accepted 15 October 1978. 2 Eastern Kentucky University, Department of Bio­ logical Sciences, Richmond, Kentucky 40475. 3 Plant Taxonomy Laboratory, Science and Education MATERIALS AND METHODS Administration, United States Department of Agricul­ ture, Agricultural Research Center, Beltsville, Maryland Specimens from the following herbaria 20705. were examined: BAFC, BISH, CONC, F, 85 86 PACIFIC SCIENCE, Volume 33, January 1979

15mm

V. DUMETORUM V. GIGANTEA V. MENZIESII V. NIGRICANS V. PISIFORMIS

FIGURE 1. Styles (tOP row) and standards (bottom row) of five species of Vicia. GH, NA, NY, R, SGO, SP, US [herbarium Pridham 1967). When lengths of flowers, abbreviations throughout this paper defined leaflets, legumes, seeds, hila, or stems are by Holmgren and Keuken (1974)]. Morpho­ considered, these five species are among the logical characters examined include those largest in the genus. used by Gunn (1968), Lassetter (1975,1978), Vicia nigricans and V. gigantea have and Kupicha (1976). External seed characters ochreoleucous to orange flowers, often with were compared, and seed amino acid con­ a tinge of rose-purple, Vicia menziesii flowers tents were evaluated. Pollen was examined, are ochreoleucous when young and rose­ and karyotypes were compared. No artificial purple when mature. Vicia dumetorum is hybridizations of these perennials could be blue-purplish flowered, and V. pisiformis is attempted due to lack of suitable growth yellow flowered. Although the significance of chambers that would provide cool tempera­ flower color may not have been adequately tures and high humidity required by the surveyed for the genus, Kupicha (1976) taxa. Greenhouse and field plot attempts to noted that it is of little value at the section grow Vicia gigantea and V. nigricans at level. Among the native North American Beltsville, Maryland, failed to produce flow­ species, V. gigantea is the only species with ering plants. No attempt was made to grow ochreoleucous to orange flowers tinged with V. menziesii to flower. rose-purple. If Hawaii is considered, V. menziesii is the only native New World­ Oceania species with a distinct rose-purple pigment in the flower. The rose-purplish RESULTS AND DISCUSSION pigment in both species becomes more Vicja menziesii, V. nigricans, and V. gigan­ pronounced as flowers age. Burkart (1966), tea, but not V. pisiformis and V. dumetorum, Moore and Scotter (1976), and Reiche (1898) usually darken or blacken with age or upon listed about 33 native species for South drying due to the presence of L-3, 4-dihy­ America. About 20 of the 33 species are droxyphenylalanine (L-dopa) (Andrews and valid and only 7 of these have ochreoleucous Vida menziesii Sprengel Rediscovered-LAssETTER AND GUNN 87 to orange or reddish flowers. Four of these similar study, Tschiersch and Hanelt (1967) represent forms of V. nigricans and should be also established three groups: (1) canavanine­ regarded as synonyms. Flowers of the other producing species; (2) species without cana­ native North and South American species vanine but with fJ-cyanoalanine and its y­ are bluish, white, lavender, or purple, but glutamyl derivative; and (3) species without never with a definite orangy or rose-purple canavanine but with high levels of arginine. color. Tschiersch and Hanelt (1967) placed V. The shape of the standard in all five dumetorum and V. pisiformis in their group 3. species is oblong (Kupicha 1976 and Figure Of the five species under consideration, only 1). V. dumetorum was included by Bell and The nature of stylar pubescence and com­ Tirimanna (1965) and Bell (1966) when they pression ofthe style is an important morpho­ surveyed the genus for seed amino acid con­ logical character in Vida (Gunn and Kluve tent. This species contains y-hydroxyarginine 1976, Kupicha 1976, Lassetter 1975). Gunn and an additional compound which is and Kluve found that about 40 percent of probably hydroxycitrulline (Bell 1966; Bell, the 106 species surveyed had apical hairs personal communication). Vida pisiformis, encircling the style. Members of several of analyzed at a later time, exhibited no amino Kupicha's sections, including Cassubicae acid pattern typical forthe genus. It contains with V. nigricans and Vicilla with V. pisi­ an unusual neutral amino acid and an unusual formis and V. dumetorum, have stylar pubes­ acidic amino acid (Bell; personal communica­ cence distributed evenly around the style. tion). Based on this biochemical compari­ Vida dumetorum is an exception in its section, son, V. dumetorum and V. pisiformis are because its stylar pubescence is densely con­ different. Tschiersch and Hanelt (1967) found centrated on the abaxial side (Figure 1). The that V. dumetorum and V. pisiformis were other four species iIi this study have evenly similar, differing only in amounts ofthreonine pubescent styles, the predominant condition and arginine and in presence or absence of in the subgenus Vicilla. Compression of alanine. These authors, however, may not styles is dorsal in all five species. Kupicha have analyzed for the neutral and acidic acids discussed the types of stylar compression. Bell found in V. pisiformis. Their data show The legumes ofthe five species are similar, that neither species is a canavanine producer. except that those of Vida menziesii may be Vida menziesii, V. nigricans, and V. gigantea much larger. These general legume characters contain appreciable amounts of canavanine are typical for the subgenus Vidlla (Kupicha with no trace of fJ-cyanoalanine or y­ 1976). glutamyl-fJ-cyanoalanine (Bell and B. V. All five species have a circumlinear hilum Charlwood, personal communications). This that occupies 75 percent of the seed circum­ would place these vetches in the canavanine­ ference. Only 15 percent of 100 species (± 150 containing group of Bell and Tschiersch and species in the genus) surveyed exhibited a Hanelt. Tschiersch and Hanelt (1967) stated circumlinear hilum (Gunn 1970). The seeds that biochemically defined species groupings of the five species have a similar external and taxonomically (they presumably mean morphology, except that Vicia menziesii seeds morphological data) defined groups are never average about 25 percent greater in diameter completely identical. Bell (1971) similarly (Figure 2). noted that his three main groups correspond On the basis of seed amino acid data, Bell in general to accepted subgeneric groups (1971) divided Vida into three groups: (1) based on morphology and cytology. These canavanine-producing species; (2) species not groups predate Kupicha's report, and in this producing canavanine but having high con­ instance the biochemical and morphological centrations ofy-glutamyl-fJ-cyanoalanine and data mesh. Kupicha (1976) placed V. pisi­ lower concentrations of free fJ-cyanoalanine; formis and V. dumetorum in a different and (3) species without canavanine and with­ section from V. nigricans. out compounds in the second group. In a It has been clearly established that karyo- U~I Y.MENZIESII A. .n. ,j :r\5·~ nn un Y.MENZIESII B. Jt j' n;(11 It " I 23'?4 5 6 7

V. DUMETORUM V. GIGANTEA V. MENZIESII V. NIGRICANS v. PISIFORMIS

;:l'pimm ,. .

50 IJ!m v' 1 ~'.

FIGURE 2. Idiograms of Vida menziesii Sprengel (top row) and seeds (middle row) and pollen (bottom row) of five species of Vida. . Vicia menziesii Sprengel Rediscovered-LASSETTER AND GUNN 89 typic differences are taxonomically significant Vicia seedlings. She was aware of our con­ at the species level (Kupich,i 1976, Lassetter current work on the taxonomic relationship 1975, Mettin and Hanelt )968, Shrivastava of V. nigricans and V. gigantea and treated 1963, Stankevich 1970, Sveshnikova 1927). them as a single species, but she had no V. The karyorype of Vicia pisiformis was re­ menziesii seedlings. Observations from our ported by Cineura (1962) and the karyotype chromosome vouchers show that eophylls of of V. dumetorum by Terziis1}i (1974) Cineura V. menziesii first appeared at the fourth node (1973) presented a morphological and cyto­ and that the first three eophylls on both logical comparison of the two species. plants each had four leaflets and a tendril. Terziiski and Cineura confirmed the earlier These characteristics place V. menziesii very work ofHeitz (1931), who presented chromo­ close to V. nigricans in Nozzolillo's keY. some numbers and karybtypes of both Because the leaflets are only two to three species. According to them, V. dumetorum is times as long as wide, V. menziesii would be 2n = 14, V. pis!formis is 4h = 12, and each placed with V. sylvatica, a species of section contains one satellited pa,i<,Cineura (1973) Vicilla according to Kupicha (1976). The determined that five chroriiosome pairs are wide leaflets of V. menziesii are typical of its essentially the same in sizeiind characteristics gigantism (see discussion under Vi(:ia men­ in both species. These incJude one metacen­ ziesii.) In general, V. menziesii seedlings are tric pair, two submetacentti¢ pairs, and two similar to those of V. nigricans. Vicia pisi­ acrocentric pairs (one ofthe~e is the satellited formis and V. dumetorum are alike according pair). There is a differen~r in the longest to Nozzolillo, but are in a different group of pair: In V. pisijormis it isianger and more species that have only two leaflets in the nearly metacentric. The 'third pair in V. first three eophylls. dumetorum (which is subirietacentric and A cursory study of Vicia pollen as part of equals pair F according _t6:'terziiski) is not a larger tribal survey (Clarke and Kupicha present in V. pisijormis. Citicura (1973) con­ 1976) did not reveal major intrageneric cluded that these vetches a.t¢.'. closely related variation (Kupicha 1977). The pollen is species, occupying similar;hltbitats, but that oblong, with small, heavily thickened endo­ they are morphologically <\.h4 karyologically apertures. Pollen of all five species in this definable and separable frqm one another. study is similar (Figure 2) and resembles Vicia menziesii has 2n = i 4 chromosomes pollen grains of the V. ludoviciana complex with one satellited pair (Figure 2). Pair 6 is (Lassetter 1972). Erdtman (1952) concluded submetacentric and all othersare acrocentric. that there are no striking differences among The karyotypes of V. gigantea and V. the pollen of the three subfamilies of nigricans are identical to each other (S. Fabaceae. Veerasethakul, unpublished data) but differ­ ent from V. menziesii. Ilithe former two species, the satellited pair of chromosomes is smaller and pair 6 is acrocentric, whereas TAXONOMIC TREATMENT pair 6 in V. menziesii is sUbmetacentric and Vicia menziesii is closely related to the New the satellited pair is mucl1,larger. Also, a World continental species V. gigantea and minor difference exists iii:. pair 5: In V. V. nigricans. Its isolated and endemic range, menziesii, pair 5 is less strongly acrocentric, coupled with its physical size, karyotype and in the other two species 1tis more so. The differences, larger rose-purple flowers, lack of other chromosome pairs, 1-'-4, are compa­ pellucid dots on leaflets, and possible pollina­ rable in size and charactetistics and are tors (see following discussion) make it a believed to be essentially, the same. Vicia valid species. The continental taxa, because menziesii voucher A was:cleposited in the of identical karyotypes, habitats, and other Eastern Kentucky University Herbarium and similarities, should best be viewed as dis­ voucher B in ISC. junct subspecies. These two taxa are com­ Nozzolillo (1977) presented a study of bined under the earlier of the two names, 90 PACIFIC SCIENCE, Volume 33, January 1979

V. MENZIESII

FIGURE 3. Illustration of Vicia menziesii Sprengel.

.N' 459 lUll Iii& mJJ§iilii " ii: Vicia menziesii Sprengel Rediscovered-LASSETTER AND GUNN 91

Vida nigricans Hooker & Amott. Based on tricolporate, prolate, 50 x 32 micra; style data presented in this paper, V. dumetorum dorsally flattened, the upper portion encircled and V. pisijormis, a related species-pair,are by delicate hairs; legume glabrous, oblong, not closely related to the other three taxa. 90 to 100 mm long, 15 to 20 mm wide, In the following discussion, an exclamation becoming black at maturity, smooth within, mark indicates that the type specimen or its with more than four ovules, on 10-mm photograph was examined. The abbreviation stipe; seed spherical, smooth, monochrome F.M. Neg. indicates the negative number of blackish, 6 to 8 mm in diameter, containing the Field Museum's series ofnegatives taken canavanine; hilum color ofseed coat, circum­ in European herbaria by J. F. Macbride linear, occupying 75 percent of the seed from 1929 to 1939. circumference. FLOWERING: June. Chromosome number: Vida menziesii Sprengel, Car. Linn. Syst. 2n = 14, Hawaii Island, 15 km northwest of Veget. ed. 16,3:267.1826. Volcano, Keauhou Ranch, 1700 m. Seeds collected by C. J. Ralph and H. Sakai. Figure 3 DISCUSSION: This species was first named HOLOTYPE: "At the upper edge of the Vicia grandiflora by Smith. The holotype is forest on the mountain Mowna-rooa, in in the Smith Herbarium at LINN. However, Owhyee which is 6,000 feet high," A. Menzies Smith's V. grandiflora is an illegitimate 1803 (LINN!). homonym. Sprengellater proposed the name V. menziesii ISOTYPE (BM !): Vicia grandiflora Smith in in honor ofthe original collector, Rees, Cyclop. 37, no. 7.1819; non V. grandi­ A. Menzies. The Federal Register (United flora Scopoli, Fl. Cam. ed. 2, 2: 65. 1772. States Department of the Interior 1978) promulgated the poorly coined common DESCRIPTION: Coarse, strongly climbing, name "Hawaiian wild broad-bean." The perennial herb 15-90-(200) dm long; pubes­ term broad-bean is generally restricted to one cent when young, less so at maturity, darken­ species of Vicia, V. faba. A far better com­ ing with age or on drying; stems stout, 6 to 8 mon name is "Hawaiian vetch." mm thick, fistulose, conspicuously ridged; Vicia menziesii is the only vetch native in stipules large, 20 to 30 mm long, 10 to 20 mm Oceania (Figure 4). Its original range was wide, foliaceous, suborbicular, lacerate-den­ apparently restricted to forest slopes of tate with long subulate teeth; leaves 10 to 13 Mauna Loa and possibly Mauna Kea at cm long, ending in ramose tendril; leaflets about 2000 to 2500 m, as determined by alternate or subopposite, 8 to 12, entire, Skottsberg (1931), although he did not ovate lanceolate to ovate, 35 to 70 mm long, personally see living specimens. Hillebrand 15 to 30 mm wide, acute to rounded at (1888) also stated that it occurred on Mauna mucronulate apex, pubescent when young, Kea, but we have not personally seen a Mauna with few scattered fine hairs with age, neither Kea collection. Until 1973, when it was side bearing pellucid dots; flowering ped­ rediscovered by Wayne Gagne and Mae Mull uncles slightly shorter than the subtending (C. H. Lamoureux and D. Herbst, personal leaves, lengthening during fruiting, with six communications), the last known collection to nine loosely clustered racemose flowers, was by Forbes in 1915 (Rock 1920). pedicels 8to 10 mm long; corolla quite large, Since this initial rediscovery, five colonies 25 to 30 mm long, sharply curved, ochre­ with 15 plants have been located in the same oleucous when young and becoming rose­ general area. The major colony is 15 km purple with age; calyx gibbous, tube less than northwest of Volcano on private land at one-half length of corolla, upper pair of 1500 to 1700 m elevation (c. J. Ralph, teeth 3 to 5 mm long, lateral pair 5 to 6 mm, personal communication). Quick action was lowest tooth 6 to 7 mm and subequal to tube; taken by interested people in Hawaii, and androecial sheath oblique at apex; pollen the species is now protected as an endangered 22~

2L- ~~~ ."", L"••"S.'\...J 40 50; K"Il~lJ.

20~

HAWAII 40. • ISlAlO 50

19- I I I 130 157 156 155 80

V. NIGRICANS V. NIGRICANS V. MENZIESII SUBSP. NIGRICANS SUBSP. GIGANTEA

FIGURE 4. Distribution ofVida menziesii Sprengel, V. nigricans Hooker & Arnott subsp. nigricans, and. V. nigricans Hooker & Arnott subsp. gigantea (Hooker) Lassetter & Gunn. Vida menziesii Sprengel Rediscovered-LAssETTER AND GUNN 93 species (Fosberg and Herbst 1975, United which we are aware. Hawaii Island: C. N. States Department ofthe Interior 1978). Forbes 938-H (BISH, NY, two duplicates of According to C. J. Ralph (personal com­ this collection are at BM but were not seen); munication), the main stem of Vicia men­ M. J. Remy 656 (GH, this sheet is ex P, ziesii may grow upward into the subcanopy of duplicates may exist there); F. R. Warshauer M etrosideros collina ssp. polymorpha and then & G. Murakami 1240 (US); A. Menzies often branches laterally. Stems up to 20 m (BM photograph). Rock (1920) states that in length have been observed. An individual Macrae also collected V. menziesii; we have plant can contain as many as 20 flower not seen these specimens, but judging from clusters and about 90 flowers. Mann (1866), specimens might be at CGE, Examples ofgigantism are common among G, or K. Approximately the same number of oceanic island species. In Hawaii, several specimens we examined was used in the legumes as well as other plants show a trend treatment of this species by Degener et al. toward larger fruits and/or seeds from smal­ (1970). ler, presumably more transportable ancestral forms. Larger seeds are believed better for Vicia nigricans Hooker & Arnott subsp. competing and surviving on the terrestrial nigricans areas of the islands (Carlquist 1974). Vicia menziesii has the largest seeds of the three Figure 5 taxa, and indeed, has the ninth largest seed Vicia nigricans Hooker & Arnott, Bot. of the 100 species included by Gunn (1970). Capt. Beechey's Voy. 20. 1830 No extensive listing of legume size for the Ervum nigricans (Hooker & Arnott) genus was found, but a comparison with Alefeld in Oesterr. Bot. Zeit. 9(11): 364. species included in Gunn (1971) and Her­ 1859 mann (1960) showed that V. menziesii has larger legumes than all vetches listed except HOLOTYPE: Concepcion, Chile. Beechey V. faba, V. narbonensis, and V. sativa (all (K !). Lathyrus macraei Hooker & Arnott, human-influenced species). Leaflets are also Bot. Capt. Beechey's Voy. 21. 1830. V. among the largest in the genus. macraei Hooker & Arnott, Bot. Misc. 3: 195. Carlquist (1974) stated that most flowers 1833 (HOLOTYPE K !). V. magnifolia Clos in of oceanic island species are small, with red Gay, Fl. Chile 2: 135. 1847. V. macraei var. being the least frequent color. These charac­ valdiviana Philippi, Linnaea 33: 51. 1864­ ters were related to small insects available to 1865. V. lanceolate Philippi I.c.51. V. ley­ perform pollination. Vicia menziesii, how­ boldii Philippi I.c.51 (HOLOTYPE: SGO!, ever, has large flowers that turn rose-purple ISOTYPE: F.M. Neg. 2341 ex B!). V. andina on maturation, an exception to the general Philippi, Anal. Univ. Chile 41:692. 1872 correlation of Carlquist. (HOLOTYPE: SGO!, ISOTYPE: F.M. Neg. 2336 C. J. Ralph (personal communication) ex B!). V. fodinarum Philippi I.c.692. V. reported that two species of native birds feed moorei Philippi, Anal. Univ. Chile 84:270. at Vicia menziesii and possibly pollinate it. 1894 (HOLOTYPE : SGO !). V. apiculata Philippi The 'I'iwi (Vestiara coccinea) has a down­ l.c.274 (HOLOTYPE: SGO!). V. coxii Philippi curved bill that fits the curve of the corolla, l.c.275 (HOLOTYPE: SGO i). V. commutata and the birds turn the flowers upside down Philippi l.c.276 (HOLOTYPE: SGO!). V. darap­ to feed. 'Amakihi (Loxops virens) has a skyana Philippi l.c.276 (HOLOTYPE: SGO i). similar but smaller bill, and somewhat V. speciosa Philippi l.c.277 (HOLOTYPE: damages flowers when it feeds. This bird­ SGO!). F.M. Neg. 2334 ex B! is of a Poeppig flower relationship may explain why mature specimen, the name for which has not been flowers are large and rose-purple. validly published.

SPECIMENS: Apparently very few specimens DESCRIPTION: Coarse, strongly climbing of Vicia menziesii exist. We have listed all of perennial herb 20 to 30 dm tall; pubescent 94 PACIFIC SCIENCE, Volume 33, January 1979

Xl%

V. NIGRICANS SUBSf.> NIGRICANS

FIGURE 50 Illustration of Vida nigricans Hooker & Arnott subspo nigricanso

• i_fda Mm ;;g;;£;;&£ & & w Vicia menziesii Sprengel Rediscovered-LASSETTER AND GUNN 95

when young, becoming less so at. maturity, observations were made on sexual cells, but turning black with age or on drymg, some­ the drawing is clearly of mitotic chromo­ times remaining green; stems stout, 1to 7 mm somes. 2n = 14, S. Veerasethakul (unpub­ thick, fistulose, usually conspicuously ridged; lished data). stipules moderately large (up to 14 mm l.ong and 7 mm wide), semisagittate, or occasion­ DISCUSSION: Hooker and Amott described ally orbicular and foliaceous, but upper Vicia nigricans in 1830 and V. macraei in ~ 833 portion usually not foliaceou~, lower portIOn when the two were compared. Only mmor entire or deeply angled and wIth two to three differences between the two species were teeth; leaves 5 to 23 cm long, ending in a listed. All vegetative characteristics they ramose tendril; leaflets alternate or subop­ mentioned are now known to intergrade, and posite, 6-14-(18), entire, linear-~a~ceolate, we believe that such variation is best viewed lanceolate to ovate and widely elhptIc, 11 to as expected phenotypic plasticity not war­ 50 mm long, 5 to 20 mm wide, acute, obtuse, ranting a species-level distinction. The holo­ leafle~s rounded, or truncate to emarginate at the type of V. macraei (6 to 9 from the mucronulate apex, pubescent when young, description) has several leaves wIth 15 leaf­ mature leaflets usually somewhat strigose and lets and one with 17. Hooker and Arnott obscurely pellucid dotted below, usually also stated that V. macraei discolors but glabrous with prominent pe.lluci~ dots above; does not blacken; this characteristic also is flowering peduncles averagmg shghtly longer variable. Vicia faba, which is not related to than subtending leaves but often shorter or these taxa, also often turns blackish upon up to twice the leaflength, lengthening during drying, but not every individual does. so, ~nd fruiting, with 4 to 25 closely packed racemose the degree of darkening varies. LIkeWIse, flowers; pedicels 1.5 to 7 mm long; corolla Philippi at various times (see .synon.omy 13 to 24 mm long, ochreoleuCOl,lS to orange, citations) named many phenotypIc varIants often tinged with rose-purple; calyx gibbous, as species. An example is his V. fordi~arur:z, tube about one-fourth to one-third length of which is distinguished from V. macraez by Its corolla, upper pair of teeth 0.3 to 1.4 mm large stipular teeth. long, lateral pair 0.4 to 5 mm long, lowest Vicia nigricans subsp. nigricans, the largest tooth 1.3 to 6.2 mm long, much shorter to known South American vetch, is restricted to slightly longer than tube, averaging ab~ut the mountains of central Chile from Valpar­ three-fourths length of tube; androecml aiso Provincia to Aysen Provincia and to the sheath oblique at apex; pollen tricolporate, mountains of extreme western Argentina in prolate, 45 x 32 micra; style dorsally flat­ the· National Territories of Neuquen, Rio tened, upper portion encircled by rather stout Negro, and Chubut (Figure 4). This taxon hairs; legume glabrous, oblong, 37 to 46 mm follows the distribution of the austral Notho­ associat~d long, 8 to 10 mm wide, becomin.g black at fagus forest and is usually found maturity, smooth within, four to Olne ovuled, with it. Masses of plants may be found m on a (1.5)-4-6 mm stipe; seed usually three open to semiopen areas within the forest or to four per legume, spherical, smooth, mono­ along the edges of the granite slopes of the chrome reddish-ochre to darker brown, 5 to Andes from 850 to 1250 m. 6 mm in diameter, containing canavanine; Judging from the material of ~. nilfricans hilum color of seed coat, circumlinear, subsp. nigricans we have seen, It dIsplays occupying 75 percent of the seed circum­ considerable phenotypic variation. The ference. diverse habitats and the inland extension of this taxon may be related to its phenotypic Vicia nigricans gigante.a FLOWERING: October through March. variability. subsp. Chromosome number: n = 7, Quebrada manifests fewer phenotypic extremes and IS Macul, Santiago, Chile (Cortazar 1948). The more restricted in habitat. voucher specimen is Vicia macraei Pisano SELECTED SPECIMENS: Argentina: Chubut 1577, 1948 (SGO!). Cortazar stated that the -A. A. Beetle & D. Bignoli 224 (US); 96 PACIFIC SCIENCE, Volume 33, January 1979

V. NIGRICANS SUBSF? GIGANTEA

FIGURE 6. Il1ustration of Vida nigricans Hooker & Arnott subsp. gigantea (Hooker) Lassetter & Gunn.

-M I! 'MilibMl 'iM liii .;;g;;;;MJtzaa Vicia menziesii Sprengel Rediscovered-LASSETTER AND GUNN 97

Neuquen-A. Burkart 1293 (BAFC), R. gose with age and obscurely pellucid dotted Leal & V. Roig 18.212 (BAFC), J. Vallerini below, glabrous with prominently pellucid 220 (BAFC), T. M. Pedersen 1517 (US); Rio dots above; flowering peduncles averaging Negro-O. Boelcke & M. N. Correa 5855 about same length as subtending leaves but (BAFC), L. R. Parodi 11540 (GH), H. A. shorter or longer, lengthening during fruiting Senn 4274 (US), F. W. Hunnewell 16,033 with 6 to 19 closely packed racemose flowers; (GH). Chile: Colchaqua-G. Montero O. pedicels 2 to 3 mm long; corolla 10 to 15 mm 49A (GH); Concepcion-C. Junge K. 16.248 long, ochreoleucous to orange, often tinged (US); Curico-Bro. Claude-Joseph 5637 with rose-purple; calyx gibbous, tube one­ (US); Cautin-E. Barros 7601 (BAFC); half length of corolla, upper pair of teeth 0.5 Malleco-H. Gunckel 16.046 (BAFC); Nue­ to 1.2 mm long, lateral pair 1.3 to 3.2 mm ben-F. Philippi & A. Borchers (F); O'Hig­ long, lowest tooth 5.4 to 8.7 mrp. long and gins-C. Jiles 4646 (BAFC); Santiago-G. subequal to the tube; androecial sheath Montero O. 49 (GH); Valdivia-O. Buchtien oblique at the apex; pollen tricolporate, (GH, US); Valparaiso-J. L. Morisson & R. prolate, 46 x 32 micra; style dorsally flat­ Wagenknecht 17155 (GH), W. J. Eyerdam tened, upper portion encircled by delicate 10084 (F, US). hairs; legume glabrous, oblong, 33 to 55 mm long, 10 to 20 mm wide, becoming black at maturity, smooth within, four to eight ovuled, Vicia nigricans Hooker & Arnott subsp. on a 4-mm stipe; seed usually three to four gigantea (Hooker) Lassetter & Gunn comb. per legume, ,Spherical, smooth, ochre to nov. reddish-ochre, mbderately to densely mottled Figure 6 and pointed with'darker brown, occasionally with markings so dense as to make seeds Vicia gigantea Hooker, Fl. Bor. Am. appear monochrome dark-brown orpurplish­ I: 157. 1831. black to velvety black, 5 to 6 mm in diameter, HOLOTYPE: In open wood, common North­ containing caflavanine; hilum color of seed west America, Douglas (K !). V. sitchensis coat, circumlinear, occupying 75 percent of Bong., Mem. Acad. Sci. St. Peters., ser. 6, the seed circumference. 129-130. 1833. V. hookeriana Walpers, Rep. FLOWERING: May through July. Chromo­ Bot. Sys. 1:715-716. 1842. Lathyrus cinctus some number:n = 7, Graham Island between S. Wats. Proc. Amer. Acad. Arts Sci. 23: 263. Skidegate and Skidegate Village. The voucher 1888. Vicia semicincta Greene Erythea 3: 17. (in DAO) is Calder, Savile, and Taylor 34665 1895. (Taylor and Mulligan 1968). 2n = 14 (S. Veerasethakul; unpublished data). DESCRIPTION: Coarse, strongly climbing, perennial herb, 6 to 20 dm tall; pubescent DISCUSSION: Vicia nigricans subsp. gigan­ when young, becoming less so at maturity, tea, the largest North American vetch, is turning black with age or on drying, rarely restricted to the coast of Pacific North remaining green; stems stout, 3 to 7 mm America from the environs of Sitka, Alaska, thick, fistulose, conspicuously ridged; stipules to San Luis Obispo County, California large (up to 25 mm long and 15 mm wide), (Figure 4). [Anderson (1959) and Welsh upper portion almost foliaceous, lower por­ (1974) stated that it is reported from Cook tion smaller, lacerate-dentate; leaves 9 to 23 Inlet, which is much further north. This is cm long ending in a ramose tendril; leaflets based on a herbarium specimen collected by alternate or subopposite, 16 to 29, entire, W. H. Osgood in 1900 (US). The location on narrowly lanceolate to ovate or elliptic to the specimen is incorrect; it was actually oblong or oblong-Ianceolate, 20 to 50 mm collected in the Queen Charlotte Islands long, 5 to 17 mm wide, rounded, obtuse, or (Calder and Taylor 1968 and personal com­ truncate at the mucronulate apex, pubescent munication).] In the northern portion of its when young, mature leaflets somewhat stri- range, V. nigricans subsp. gigantea usually is 98 PACIFIC SCIENCE, Volume 33, January 1979

found along the upper margin of sand or Ancestral V. menziesii may have been trans­ shingle beaches in the driftwood zone, on ported from the Pacific Northwest to Hawaii rocky headlands, and in open woods near the by ocean currents or by migratory birds. coast (Calder and Taylor 1968). In the Although ocean currents may transport buoy­ southern portion of its range, it is found ant, viable seeds and seed-bearing fruits over along the coast in moist, usually wooded or thousands of kilometers, there is no direct shaded places. The only inland incursion is transport current from the Pacific Northwest along the Columbia River to the Willamette to Hawaii (Gunn, Dennis, and Paradine Valley, Oregon. 1976). In an analysis of the Hawaiian flora, Carlquist (1974) found examples of 52 suc­ SELECTED SPECIMENS: Canada: British Col­ cessful arrivals from American sources. He umbia-J. Macoun (US), W. Spreadbor­ concluded that repetitive in-seawater drifting ough 70.434 (NY), C. F. Newcombe 19 (F), was unlikely from the Americas. Rare or W. H. Osgood (US), J. Macoun 79.704 (NY). infrequent rafting by logs also was unlikely, United States: Alaska-M. W. Gorman 55 though six logs bearing Pacific Northwest (NY, US), Mr. & Mrs. E. P. Walker 989 brands have been stranded on Hawaiian (F, NY, US), J. P. Anderson 833 (NY); beaches (Strong and Skolmen 1963). Unfor­ California-AlamedaCounty, V. K. Chesnut tunately, the authors neither related these (US); Contra Costa County, W. C. Blasdale strandings to ocean currents, storms, or (US); Humboldt County, C. R. Gunn 2691 tsunamis, nor to where these logs entered the (US); Mendocino County, J. McMurphy 157 Pacific Ocean. (NY, US); San Mateo County, C. F. Baker Carlquist (1974) believes, as do many 506 (NY, US); Santa Cruz County, C. R. others, that dispersal of seeds and small Gunn 2654 (US); Oregon-J. T. Howell fruits by migratory birds is probably the (F, NY, US); Multnomah County, E. P. method by which most groups ofangiosperms Sheldon, S. 12033 (F); Washington County, arrived on the high islands ofthe Pacific. The A. Kellogg & W. G. W. Harford 174 (US); presence of Vicia menziesii on Hawaii Island Washington-Clallam County, C. R. Gunn is best explained, Carlquist said, by internal & C. E. Smith, Jr., 3670 (NA, US); Jefferson transport by birds, and we agree. County, C. R. Gunn & C. E. Smith, Jr., 3671 Raven (1963) compiled information con­ (US); Pacific County, W. W. Eggelston 11478 cerning temperate North and South Ameri­ (F, US); Whatcom County, W. N. Suksdorf can disjuncts. Although Vicia nigricans subsp. 964 (F, NY, US). gigantea and V. nigricans subsp. nigricans were not included, they are an additional example similar to those he did include. DISCUSSION OF DISTRIBUTION Raven believes that long-distance bird dis­ A distribution pattern similar to that of persal is the most logical explanation ofmost Vicia menziesii and V. nigricans subsp. ofthe disjuncts. Another possibility would be nigricans and subsp. gigantea is exhibited by migration over a land bridge or continuous Fragaria chiloensis (Linnaeus) Duchesne mountain chains that would have allowed (Darrow 1966). In fact, the distribution is direct movement through the tropics to the almost identical with disjunct varietal popu­ temperate zones. He believes this method less lations of F. chiloensis in coastal and moun­ supportable, because of a lack of corre­ tain areas of Chile and western Argentina, sponding animal disjuncts that would be along the Pacific coast from California to the expected were a continuous route available. Aleutian Islands, and above 1800 m in the Lincoln (1935) stated that the Pacific Flyway mountains of Hawaii. No explanation of the is the longest migration route in the Americas Fragaria distribution was offered by Darrow. and that a number of species migrate from Although there are several possible expla­ Alaska to southern South America along nations for the distribution ofthe Vicia taxa, this route. Lincoln (1952) and Dorst (1962) two appear to be more likely than the others. also cite several species of migrating birds Vicia menziesii Sprengel Rediscovered-LASSETTER AND GUNN 99 that travel from Alaska to Hawaii and Washington, D. C. We are especially thankful beyond. to C. J. Ralph, Science and Education Cruden (1966) believes that long-distance Administration, United States Department of migration dispersal may not be as attractive Agriculture, Honolulu, for the viable seeds an explanation as previously believed. A and recent field observations from Hawaii. major point he makes is that internally They were invaluable to this study. Special transported seeds (these are more likely to appreciation is also extended to E. A. Bell be successfully carried great distances as and B. V. Charlwood of King's College, opposed to those carried externally) do not London, for their amino acid analyses. The remain in the digestive tract long enough. senior author extends special appreciation to However, Proctor (1968) and Vlaming and the C. K. Becks for their support. Proctor (1968) present data that demonstrate that shore birds can retain seeds long enough to transport them the required distances, even to the most remote oceanic islands. LITERATURE CITED They found that larger seeds (2-6 mm in diameter) with a hard impervious seed coat ANDERSON, J. P. 1959. Flora of Alaska and were best retained, and that some shore adjacent parts of Canada. Iowa State bird species tend to retain seeds better than University Press, Ames. other species they studied. Seeds of the Vicia ANDREWS, K. S., and J. B. PRIDHAM. 1967. spp. in this study are 5 to 8 mm in diameter Melanins from dopa-containing plants. and are hard-seeded. Phytochemistry 6: 13-18. BALL, P. W. 1968. Vicia. Pages 129-136 in T. G. Tutin, et aI., eds. Flora Europaea. Vol. 2. Cambridge University Press, Cam­ ACKNOWLEDGMENTS bridge. We are grateful to the curators of the cited BELL, E. A. 1966. Amino acids and related herbaria for lending specimens. Encourage­ compounds. Pages 195-209 in T. Swain, ment and critical comments from the late ed. Comparative Phytochemistry. Aca­ Arturo Burkart, Universidadde Buenos demic Press, New York. Aires, and dataobtained from Melica Munoz, ---. 1971. Comparative biochemistry of Museo Nacional de Historia Natural, San­ non-protein amino acids. Pages 179-206 tiago, Chile, are most appreciated. Raquel in J. B. Harborne, D. Boulter, and B. L. Giangualani, Ministeriode Agricultura y Turner, eds. Chemotaxonomy ofthe Legu­ Ganaderia, furnished us with additional minosae. Academic Press, New York. South American distribution data, especially BELL, E. A., and A. S. L. TIRIMANNA. 1965. the southern limit (Aysen Provincia, Chile) Associations of amino acids and related for Vicia nigricans subsp. nigricans. Regina compounds in the seeds of forty-seven O. Hughes prepared the habit, style, and four species of Vicia: Their taxonomic and of the seed illustrations, and F. A. Uecker nutritional significance. Biochem. J. 97: photographed the pollen grains. The late 104-111. Elmar E. Leppik assisted in translation of BURKART, A. 1966. Las leguminosas syl­ some German text. Certain data were tabu­ vestres y cultivadas. Acme Agency, Buenos lated by Josette Williams and Lydia Poole. Aires. We are grateful for the use of unpublished CALDER, J. A., and R. L. TAYLOR. 1968. Flora karyotype data ofSudaratana Veerasethakul. of the Queen Charlotte Islands. Part 1. We also extend special appreciation to Queen's Printer, Ottawa. Charles H. Lamoureux, Derral Herbst, Wil­ CARLQUIST, S. 1974. Island biology. Col­ liam H. Sager, and Carolyn Corn for the aid umbia University Press, New York. they extended from Hawaii, and to Roger CINCURA, F. 1962. Poznakmy k cyto16gii McManus, Office of Endangered Species, druhov rodu Vicia L. zo slovenskych 100 PACIFIC SCIENCE, Volume 33, January 1979

nalezisk II. Acta Fac. Rerum Nat. Univ. GUNN, C. R., J. V. DENNIS, and P. J. PARA­ Comenianae Bot. 7: 349-388. DINE. 1976. World guide to tropical drift ---.1973. Angaben tiber die Zahl und die seeds and fruits. Quadrangle/New York Morphologie der Chromosomen von Vida Times, New York. dumetorum L. und deren Vergleich mit HEITZ, E. 1931. Nukleolen und Chromoso­ Vida pisiformis L. Acta Fac. Rerum Nat. men in der Gattung Vicia L. Planta 15: Univ. Comenianae Bot. 21: 103-107. 495-505. CLARKE, G. C. S., and F. K. KUPICHA. 1976. HERMANN, F. J. 1960. Vetches of the United The relationship of the genus Cicer L. States-native, naturalized and cultivated. (Leguminosae): The evidence from pollen U.S. Dept. Agr. Agr. Handb. 168. U.S. morphology. Bot. J. Linn. Soc. 72: 35-44. Govt. Printing Office, Washington, D.C. CORTAZAR, C. S. DE. 1948. Observaciones HILLEBRAND, W. 1888. Flora ofthe Hawaiian chromosomaces en seis especies Chilenas. Islands. Williams and Norgate, London. Agri. Tecn. (Chile) 8 :28-36. HOLMGREN, P., and W. KEUKEN. 1974. Index CRUDEN, R. W. 1966. Birds as agents oflong­ herbariorum. Part 1. The herbaria of the distance dispersal for disjunct plant groups world. 6th ed. Regnum Veg. 92: 1-397. of the temperate western hemisphere. Evo­ KUPICHA, F. K. 1976. The infrageneric struc­ lution 20: 517-532. ture of Vida. Notes Roy. Bot. Gard. DARROW, G. M. 1966. The strawberry. Holt, Edinburgh 34:287-326. Rinehart, and Winston, New York. ---. 1977. The delimitation of the tribe DEGENER, 0., 1. DEGENER, and C. R. GUNN. Vicieae (Leguminosae) and the relation­ 1970. Vicia menziesii Spreng. Unnumbered ships of Cicer L. Bot. J. Linn. Soc. 74: supplemental page for family 169 in O. 131-162. Degener and 1. Degener, eds. New illus­ LASSETTER, J. S. 1972. A biosystematic study trated flora of the Hawaiian Islands. of the Vida ludoviciana complex (Legu­ Published by the authors, Volcano, Hawaii. minosae). Microfilm 73-9458. University DORST, J. 1962. The migrations of birds. Microfilms, Ann Arbor, Mich. Houghton Mifflin, Boston. . ---. 1975. Taxonomic status of Vicia ERDTMAN, G. 1952. Pollen morphology and hassei (Leguminosae). Madrono 23: 73-78. plant taxonomy. Angiosperms. Chronica ---.1978. Seeds ofsome native American Botanica, Waltham, Mass. vetches. Sida Contrib. Bot. 7: 255-263. FEDTSCHENKO, B. A. 1948. Vicia. Pages 406­ LINCOLN, F. C. 1935. Ancestral highways of 475 in V. L. Komarov, Flora U.S.S.R. the sky. Amer. Forests 41: 157-159, 196. Vol. 13. Akad. Nauk S.S.S.R., Moscow. ---.1952. Migration ofbirds. Doubleday, FOSBERG, F. R., and D. HERBST. 1975. Rare New York. and endangered species of Hawaiian vas­ MANN, H. 1866. Enumeration of Hawaiian cular plants. Allertonia 1: 1-72. plants. Proc. Amer. Acad. Arts Sci. 7: 143­ GUNN, C. R. 1968. The Vicia americana 235. complex (Leguminosae). Iowa State J. Sci. METTlN, D., and P. HANELT. 1968. Bemerken 42: 171-214. zur Karyologie und Systematik einiger --. 1970. A key and diagrams for the Sipper der Gattung Vicia L. Feddes Re­ seeds of one hundred species of Vida pert. Z. Bot. Taxon. Geobot. 77: 11-30. (Leguminosae). Proc. Internat. Seed Test. MOORE, D. M., and B. SCOTTER, 1976. Vicia Assoc. 35: 773-790. L. in Tierra del Fuego. Darwiniana 20: ---. 1971. Seeds of native and naturalized 371-386. vetches of North America. U.S. Dept. NOZZOLILLO, C. N. 1977. Identification of Agr. Agr. Handb. 392. U.S. Govt. Printing Vida seedlings. Can. J. Bot. 55: 2439-2462. Office, Washington, D.C. PROCTOR, V. W. 1968. Long-distance dis­ GUNN, C. R., and J. KLUVE. 1976. Andro­ persal of seeds by retention in digestive ecium and pistil characters for tribe Vicieae tract of birds. Science 160:321-322. (Fabaceae). Taxon 25: 563-575. RAVEN, P. H. 1963. Amphitropica1 relation-

MO, tn... .. ,_ • _ ." " _"'" "'", ~ ,'" , Vicia menziesii Sprengel Rediscovered-LASSETTER AND GUNN 101

ships in the floras of North and South TAYLOR, R. L., and G. A. MULLIGAN. 1968. America. Quart. Rev. BioI. 38: 151-177. Flora of the Queen Charlotte Islands. REICHE, C. 1898. Flora de Chile. Volume II. Part 2. Queen's Printer, Ottawa. Imprenta Cervantes, Santiago. TERZIISKI, D. P. 1974. Cytotaxonomic studies RocK, J. F. C. 1920. The leguminous plants on Bulgarian leguminous species: The of Hawaii. Unnumbered publication by karyotype of Vida dumetorum L. C. R. Exp. St. Hawaii Sugar Plant. Assoc., Acad. Bulg. Sci. 27: 1267-1270. Honolulu. TSCHlERSCH, V. B., and P. HANELT. 1967. Die SHRIVASTAVA, L. M. 1963. Cytogenetical freien Aminosiiuren der Samen von Vicia studies in certain species of Vida. Cytolo­ L. und die Systematische Gliederung der gia 28 : 154-169. Gattung. Flora (Jena) Abt. A. Physiol. SKOTTSBERG, C. 1931. Remarks on the flora Biochem. 157: 389-406. of the high Hawaiian volcanoes. Acta UNITED STATES DEPARTMENT OF THE IN­ Horti Gotob. 6:47-65. TERIOR. 1978. Determination that various STANKEVICH, A. K. 1970. On clarification of plant taxa are endangered or threatened the Vida L. genus systematics. Tr. Prikl. species. Federal Register 43(81): 17910­ Bot. Genet. Sel. 43: 110-125. 17916. STRONG, C. c., and R. G. SKOLMEN. 1963. VLAMING, V. L. DE, and V. W. PROCTOR. 1968. Origin of drift-logs on the beaches of Dispersal of aquatic organisms: Viability Hawaii. Nature 197: 890. of seeds recovered from the droppings of SVESHNIKOVA, I. N. 1927. Karyological captive killdeer and mallard ducks. Amer. studies in certain species of Vida. Tr. Prikl. J. Bot. 55: 20-26. Bot. Genet. Sel. 17: 37-72. WELSH, S. L. 1974. Flora of Alaska. Brigham Young University Press, Provo.