Synopsis and lectotypification of Solanum ( Solanaceae ) species endemic in the West Indies by

Solanum (Solanaceae) is one of the ten most speciesrich genera of angiosperms. With approximately 1,500 species distributed worldwide on all continents except Antarctica, Solanum encompasses a wide range of habitats, from deserts to tropical rain forests, and habits, from annual herbs to canopy trees. The genus, like the family Solanaceae, is most diverse in the Americas, and centres of species diversity in Solanum are found in the Andes of South America, Australia and East Africa. As part of the collaborative project “PBI Solanum: a world-wide treatment” (see Knapp & al., 2004; http://www.nhm.ac.uk/solanaceaesource), descriptions of all species of Solanum together with details of types and nomenclature are being provided via an on-line taxonomic resource, Solanaceae Source. One of the goals of the PBI Solanum project is to designate lectotypes for all Solanum names, helping to stabilise nomenclature and facilitate further taxonomic research. The West Indies are usually defined as the islands of the Caribbean comprising the Greater Antilles (Cuba, Hispaniola, Jamaica, Puerto Rico), the Bahamas, and the Lesser Antilles (Windward and Leeward Islands), but excluding the islands of Trinidad and Tobago, which are of continental origin, and the Dutch Antilles. The geological complexity of the West Indies has meant that the region is a natural laboratory for studies of evolution and diversification (Ricklefs & Bermingham, 2007). The region as a whole is of Resumen


Introduction
Solanum (Solanaceae) is one of the ten most speciesrich genera of angiosperms.With approximately 1,500 species distributed worldwide on all continents except Antarctica, Solanum encompasses a wide range of habitats, from deserts to tropical rain forests, and habits, from annual herbs to canopy trees.The genus, like the family Solanaceae, is most diverse in the Americas, and centres of species diversity in Solanum are found in the Andes of South America, Australia and East Africa.As part of the collaborative project "PBI Solanum: a world-wide treatment" (see Knapp & al., 2004; http://www.nhm.ac.uk/solanaceaesource), descriptions of all species of Solanum together with details of types and nomenclature are being provided via an on-line taxonomic resource, Solanaceae Source.One of the goals of the PBI Solanum project is to designate lectotypes for all Solanum names, helping to stabilise nomenclature and facilitate further taxonomic research.
The West Indies are usually defined as the islands of the Caribbean comprising the Greater Antilles (Cuba, Hispaniola, Jamaica, Puerto Rico), the Bahamas, and the Lesser Antilles (Windward and Leeward Islands), but excluding the islands of Trinidad and Tobago, which are of continental origin, and the Dutch Antilles.The geological complexity of the West Indies has meant that the region is a natural laboratory for studies of evolution and diversification (Ricklefs & Bermingham, 2007).The region as a whole is of composite character, with the Greater Antilles, Bahamas and Lesser Antilles each having different geological origins and histories (Iturralde-Vinent & MacPhee, 1999).The Bahamas have their origins in the North American tectonic plate, and are older and more stable than the rest of the Archipelago (Mullins & Lynts, 1977).The Greater Antilles are primarily of Central American crust origin (forming to the west of present day Central America and moving to the east on the Caribbean plate), while the Lesser Antilles are derived from a volcanic arc at the eastern edge of the Caribbean plate (Iturralde-Vinent & MacPhee, 1999;Ricklefs & Bermingham, 2007).The Caribbean region as a whole has its origin as islands no earlier than the middle Miocene (Iturralde-Vinent & MacPhee, 1999), and the Caribbean tectonic plate has many blocks giving many of the islands a composite geological character (e.g., Hispaniola, Jamaica).Iturralde-Vinent and MacPhee (1999) posit land connections to northern South America (the GAARlandia hypothesis) during the Eocene-Oligocene transition, while others (Hedges, 1996;Glor & al., 2005;Ricklefs & Bermingham, 2007) suggest dispersal accounts better for taxon distribution patterns observed, particularly for vertebrates.
The West Indies is an area of considerable species richness in Solanum, but more interestingly, is home to many unusual and poorly known endemic species.Of the 62 species registered from the region (see Knapp, in press), 9 are introduced in cultivation, 28 are found elsewhere in the Americas and 25 are endemic.If we include Solanum bahamense L. as a near endemic (outside the West Indies it occurs only in Florida; it is included here, bringing the number of endemics to 26), then more than 1/3 of the species in the region are found only there.Of these, 19 are single island endemics (see Table 1), making the West Indies critically important for the conservation of Solanum diversity.The vast majority of these taxa have not been included in any phylogenetic analysis, and Whalen (1984), in his conspectus of the spiny solanums considered most of them to be of uncertain affinities.A tentative division into morphological assemblages can be made, however, this should be regarded as preliminary at best, and many of these taxa are isolated both morphologically and probably phylogenetically (see Table 2).Bohs (pers.comm.) had considerable difficulties aligning the plastid DNA sequences of S. polygamum, the unusual dioecious species from the Virgin Islands (see Knapp & al., 1998), so that it was not included in any analyses (Levin & al., 2006;Weese & Bohs, 2007).The West Indian endemic Solanum species are indeed morphologically unusual, and have combinations of characters not found elsewhere in the genus.It is apparent from morphology that radiations of Solanum have occurred within the West Indies, and that species present there will have their sister taxa in the region rather than outside.Phylogenetic analysis of these taxa will certainly contribute to our understanding of hypotheses of Caribbean biogeography, but as most of these taxa are rare and many have not been collected recently, exploration and re-collection is a priority.To this end, this synopsis is presented in order to highlight a promising area for future phylogenetic research.Taxa recorded as West Indian endemic Solanum but now regarded as synonyms of more widely distributed species are presented in Table 3.
Many of the Solanum species from the West Indies were described by O.E.Schulz, who worked in Berlin during the early part of the 20 th century.Schulz did not cite herbaria for any of the collections he cited, so it is generally assumed that he examined sets of material of these plants at B. The specimens he examined in B are no longer extant, necessitating the lectotypification of all of his names, including those he coined in collaboration with Eric Ekman.Fortunately duplicates of many of these collections are widely distributed, and for the names associated with Ekman collections, the top set of this material, annotated by Ekman, is kept at S. I have been unable to trace duplicates of some of the collections used by Schulz, but none of these species are accepted names.In these cases I have not neotypified these names here in the hope that future searches will uncover duplicates of the collections upon which these epithets are based Reasoning for the choice of all lectotypes is given, along with the distribution of each species and a representative specimen from each island where it occurs.Where lectotypes have been chosen previously full reference is given, but reasoning for the choice not repeated here.Descriptions of all these species, discussions of their putative relationships, full specimen citations and distribution maps can be found on Solanaceae Source (http://www.solanaceaesource.org).
Distribution.Endemic to Jamaica, where it is found in woods on limestone soils in the cloud forests of the John Crow Mountains from 100-700 m.
In describing Solanum fruticosum, Miller (1768) cited "Solanum fruticosum bacciferum, spinosum, flore caeruleo Sl. C. 108" and made reference to the collector William Houstoun.The specimen selected here as the lectotype for this name in BM is annotated with the same reference to Sloane, and bears a label in Houstoun's hand (Fig. 1a).It is also a good match for the description.
Solanum varginstonicum was named by Buc'hoz in honour of George Washington, with a spectacular misspelling of the name.No specimens have been traced for Buc'hoz's names, so the plate (Fig. 1b) is selected here as the lectotype.
In his description of  (Plumier, 1757).In the Prodromus (Dunal, 1852), Dunal cites several specimens, stating that the Burman plate was "mala".The Burman plate, however, is the only verifiable element associated with the original description (although Dunal may have seen the originals in Paris) and is thus here selected as the lectotype (Fig. 1c).Solanum igneum var.inerme was described by Dunal using two sheets in G-DC (Dunal, 1852), one without a collector from Cuba and the other a cultivated specimen from seeds send by Wydler from St. Thomas.He speculates that the un-armed nature of the plant is due to it being cultivated (Dunal, 1852), and places S. subarmatum tentatively in synonymy.The only unambiguously identifiable sheet cited is Wydler 90 from St. Thomas in the U.S. Virgin Islands; it is here selected as the lectotype (not illustrated).
In describing Solanum bahamense var.lanceolatum, Grisebach unambiguously cites a plate from Sloane as part of the diagnosis, and additionally assigned some of the specimens cited (without herbarium attribution) to this variety.I feel that he is specifically referring to the Sloane plate as the element on which he is basing his name, and merely identifying the other collections as belonging to one or the other variety, thus these specimens are probably not type material.Two specimens at GOET annotated "? lanceolatum" in Grisebach's hand are potential original material, but the Sloane plate is the only element unambiguously linked to the variety.I therefore have chosen the plate in Sloane's A voyage to the islands Madera … Jamaica (1707-1725) labelled Solanum fruticosum bacciferum spinosum, flore caeruleo" as the lectotype of the variety (Fig. 1d).The epitype of this name with the same polynomial designation is found in the Sloane herbarium at BM (BM000589942, Fig. 1e).
Five collections were cited in the protologue of Solanum persicifolium var.belloi; Heller 4605, Sintenis 1260, Sintenis 1919 p.p., Eggers 35 and Benzoin s.n.Of these, Heller 4605 is the most widely distributed and has several sheets in very good condition and the sheet at F (Fig. 1f) is here selected as the lectotype.Representative specimen.Cuba, Camagüey, Ekman 8600 (S).

Solanum boldoense
In describing S. chamaeacanthum, Grisebach (1866) did not cite a specific Wright collection number.Wright's duplicates are not all numbered, and some distributed without numbers are clearly duplicates of numbered collections.A sheet at GOET (GOET0001970) is labelled "savanas Habanana" and is an un-numbered Wright collection, making it the clear choice for a lectotype (Fig. 2a).Another sheet (GOET0001971) was annotated lectotype by R.A.Howard in 1984, but as it does not have the locality associated with the protologue has not been selected here as the lectotype.Distribution.Known only from the montane cloud forests of Hispaniola (Haiti and the Dominican Republic), from 1000-1800 m elevation.
Although Lamarck's protologue indicates that the plant collected by Joseph Martin came from Martinique, Solanum crotonoides has never been collected there.A similar situation exists for S. pyrifolium (Dulcamaroid clade), also collected by Martin from "Martinique".I suspect that Martin did in fact also visit Hispaniola (the locality for his type collection of Aquartia microphylla is "S.Domingo", see below) or that Martin used Martinique as a broad locality covering most of the Caribbean region, and that both these plants are true Hispaniolan endemics.
In describing S. formonense, Schulz and Schmidt did not designate which of the two varieties was typical.Sheets of both syntypes are held at S; Ekman H.10365 (S04-2911, Fig. 2b) is in excellent condition and is here used to lectotypify both the species and the variety.Ekman H.7461 is the lectotype for var.grandifolium, but is not figured here.
The lectotype of S. hoplophorum (Fig. 2c) is a very robust, prickly male plant (Fuertes 1023), showing West Indian Solanum that combination of characters that Dunal (1814) noted in describing S. reticulatum as the "hermaphrodite" of S. crotonoides (which is a dioecious species, see Knapp & al., 1998) Distribution.Only known from eastern Cuba and from Haiti, in pine woods from 700-2225 m elevation.
The type collection of Solanum gundlachii (Wright 3030) is present in many herbaria, some of which have both flowers and fruit.The specimen at GH [GH00077553] has both flowers and mature fruits and is here selected as the lectotype.The sheet clearly represents two separate gatherings, as two labels bear different collection dates and appear to come from different localities (see Fig. 3b).It is clear from the protologue that Urban did not have at his disposal a specimen with a flower ("flores non vidi", Urban, 1925), and that he was describing this species from material in B that is no longer extant.The two rather scrappy sheets at S are both annotated "type", but only one (S04-2941) has reproductive parts (one flower and two fruits in the packet).This sheet (Fig. 3c) is chosen here as the lectotype.Distribution.In dry forests and xerophytic scrublands on Hispaniola and the Bahamas, usually on limestone soils at around sea level, but up to 200 m elevation.
The type specimen of Solanum squarrosum was cited as Schomburgk 51 and was at B, but is no longer extant.A rather scrappy duplicate of Schomburgk 51 at Leiden (L0403808, Fig. 3d) is labelled "typus" in Schulz's hand and is here selected as the lectotype.Distribution.Known only from a few collections in the Sierra de Moa in the eastern mountain ranges of Cuba.
Distribution.Endemic to the eastern part of the island of Hispaniola in the Dominican Republic in moist areas in montane pine forest from 1200-1400 m elevation.
The sheet of Fuertes 1890 at A (Fig. 4a) is the only duplicate of the type collection I have been able to trace, and thus I have chosen it as the lectotype of Solanum orthacanthum.Distribution.Endemic to Cuba, known only from the eastern part of the island from pine woods at ca. 700 m elevation.
Wright 382, the type collection of Solanum pachyneurum var.glabrescens and a syntype of S. pachyneurum, is a mixed collection of S. pachyneurum (e.g., the sheet at BM [BM000886750] is S. pachyneurum) and S. pachyneuroides (also see below).I have chosen the left-hand two stems of GH00112906 (Fig. 4b) to lectotypify the varietal name; the right-hand stem and a small stem in the lower left corner are S. pachyneurum.
Distribution.Endemic to eastern Cuba, in forests at about 700 m elevation.
Representative specimen.Cuba, Guantanamo, Herbario Flora de Cuba (HFC) 52625 (B).Schulz (1909) cited three collections in his original description of Solanum pachyneurum -Wright 382, Eggers 5197 and Linden 1944.Wright 382, original material of both Solanum pachyneurum and S. pachyneurum var.glabrescens (a synonym of S. pachyneuroides, see above), consists of a series of mixed sheets of S. pachyneurum and S. pachyneuroides.It is not clear whether the two taxa occur sympatrically, or whether Wright assigned his numbers to plants collected after the fact, common practice at the time (and very likely).The latter case is suggested by the labeling of the sheet of Wright 382 at GH (see Fig. 4b).The sheet that Schulz examined in B has been destroyed, and other sheets of Wright 382 are mostly of S. pachyneurum, but some are a mixture of S. pachyneurum and S. pachyneuroides (see above).For this reason, it is more sensible to choose one of the many other syntypes of S. pachyneurum as the lectotype to avoid confusion.A sheet of Linden 1944 was not found at B and is presumably destroyed, but well preserved duplicates have been located at BM and K. Neither the K nor the BM sheet has any locality data, but the sheet at K (Fig. 4c) is better preserved, has reproductive parts and has therefore been selected as the lectotype.Lam., Tabl. Encycl. 2: 23. 1794 Type: Cultivated in Paris from seeds from "St.Domingue" (Hispaniola), Anon.s.n.(holotype, P-LA [P00357671]).Solanum parviflorum Cav., Icon. 3: 19, tab. 236. 1795 Type: Cultivated in Paris, from Santo Domingo, 1786, Anon. s.n. (lectotype, MA [MA-206020], designated by Knapp, 2007: 198) Virgin Gorda, Little & al. 23829 (BM).

Solanum polyacanthos
Two syntypes were cited by Kuntze for Solanum polygamum var.thomae; Eggers 100 and Kuntze 36.Sheets of both of these were located at K; but the duplicate of Kuntze 36 at NY (Fig. 4d) is annotated by Kuntze and comes from his personal herbarium and is thus selected as the lectotype.Solanum inclusum var.albiflorum was described by Eggers in his Flora of St. Croix and the Virgin Islands (1879) without citing a specimen or a specific locality, but is clearly a synonym of S. polygamum.The description is of a female plant (S.polygamum is dioecious, see Knapp & al., 1998), and I prefer to continue to search for fruiting material annotated by Eggers S. polygamum before neotypifying this name.Distribution.Known only from the island of Hispaniola, primarily in the Dominican Republic, along forest edges on limestone and in cloud forests, from 120-1600 m elevation.
There has long been confusion over the correct name for this species.Dunal's (1813) re-naming of Jacquin's Aquartia aculeata as Solanum aquartia was illegitimate, but the name was used for many years.By the time Schulz (1909)  No specimen of Aquartia aculeata attributable to Jacquin has been found, despite intensive searches in many herbaria (principally BM, LINN and W) where such a sheet, if it had ever existed, might have been preserved.No original material was cited in the protologue (Jacquin, 1760), so any type selected must be a neotype.The illustration of A. aculeata in Jacquin's first illustrated edition of the Selectarum Stirpium Americanarum Historia (Jaqcuin, 1763), a copper engraving taken West Indian Solanum from original drawings done by Jacquin in the field (see Wiltshear, 1913) is very accurate, and I have chosen it here as the neotype of the species (Fig. 5c).
Solanum inclusum was described without specific collection or locality but said to be widespread across the Caribbean and in northern South America (Grisebach, 1861).Various subsequent authors have suggested that Grisebach's S. inclusum included both S. tetramerum and S. polygamum.No specimens have been identified at GOET that can be directly linked to Grisebach's naming of S. inclusum.In his Cuban catalogue Grisebach (1866) cited specimens as S. inclusum that are now identified as S. tetramerum (Linden 1986(Linden , 2000)).Specimens indicated as type material of S. inclusum in some herbaria (i.e., Wright 1350 in GH) are in fact S. torvum.Since Grisebach (1861) did not indicate any specimens or specific localities for his name, these sheets cannot be considered type material.I neotypify Grisebach's Solanum inclusum here with the BM duplicate (Fig. 5d) of Linden 1986 in order to stabilise usage (synonymy) of this ambiguous name.
No duplicates of Ekman H.7289, the type of Solanum miragonae, are extant at B, and the only sheet of this collection encountered is at S, so this is the clear choice as the lectotype (Fig. 6a).It may be that the S sheet is the holotype, as S. miragonae was described in a Swedish journal along with other taxa whose types are in S, but as there is no evidence of this, I prefer to lectotypify the name here.

Solanum troyanum
Distribution.Endemic to Jamaica, in the Blue Mountains in forests and forest edges at middle to high elevations.
Schulz did not cite a herbarium when describing Solanum lobulatum, and no sheet of Smith & Smith 1575 is extant at B. The K duplicate of this collection is here designated the lectotype (Fig. 6d) as no other duplicates have been found, but are likely to exist.Solanum lobulatum was described from sterile material, Schulz (1909) stated that the specimen he examined was annotated "we have only seen one root"; from the morphology of the specimen at K it is clear that this is a juvenile individual with a single stem.Howard, J. Arnold Arb. 47: 138. 1966 Type

Table 2 .
Preliminary groupings for West Indian endemic Solanum.

Table 3 .
Species of Solanum recorded as endemic to the West Indies now regarded as synonyms of widespread taxa (for those taxa now recognised as members of the genus Lycianthes see Knapp, in press and http://botany.si.edu/Antilles/WestIndies/ catalog.htm).
Distribution.Endemic to Cuba, where it is found in forests and forest edges at low to middle elevations.Representative specimen.Cuba, Pinar del Río, Ekman 17544 (S).Solanum boldoense has occasionally been confused with the Mexican and Central American species S. dulcamaroides Dunal, to which it is probably closely related.Distribution.Endemic to Cuba, in low elevation swampy areas, common in the northwestern part of the island around Havana.
. Both varieties of S. formonense are female plants.Endemic to the Samaná Peninsula of the Dominican Republic in northeastern Hispaniola, in matorrales and forests on limestone derived from ancient corals from sea level to 350 m elevation.Representative specimen.Dominican Republic, Prov.Samaná, Peguero & Veloz 549 (JBSD, MO).Of the four syntypes cited bySchulz (Ekman  H.14847, Ekman H.15042, Ekman H.15102 and Abbott 2922), only one (Ekman H15102) is fertile.The sheet in S (Fig.2d) has been chosen as the lectotype because Eric Ekman's top set of specimens is housed in Stockholm, all have been annotated by Ekman, and S04-2902 is in excellent condition.Of these, Sintenis 2374 is represented in many herbaria, and is composed a number of sheets in good condition with both flowers and fruit.I have selected the BM sheet as the lectotype (Fig.3a) because it has all reproductive stages and is in good condition.
Distribution.Endemic to Cuba, known only from the type.Representative specimen.Cuba, Oriente, Ekman 14456 (S).
Distribution.Endemic to Jamaica, where it is found in forests and forest edges at high elevations.Representative specimen.Jamaica, St. Thomas Parish, Stearn 99 (BM, K, S).
made the correct combination Solanum aculeatum, the name Solanum aculeatum St.-Lag.had been validly (but illegitimately) published as a replacement for Solanum aculeiger Dunal (now recognised as a synonym of the Brazilian species S. thomasiifolium Sendtn., see Solanaceae Source, http://www.solanaceaesource.org).The oldest available name for this species is S. tetramerum.