Lectotypifications, synonymy, and a new name in Capsicum (Solanoideae, Solanaceae)

Abstract Considerable confusion exists within Capsicum (Solanaceae) regarding the status and typification of several names, in part due to misidentifications. Some types were destroyed in Berlin during the Second World War, some have not been found by modern systematics, while others exhibit uncertain locality data or contain material from more than one species. Fourteen lectotypes, synonyms, and a new name, Capsicum eshbaughii Barboza nom. nov.,are proposed here.

Th is genus has been known since the beginning of civilizations in the Western Hemisphere and has been part of the human diet since 6000-7500 BC (Basu and De 2003;Perry et al. 2007). After Columbus introduced Capsicum seeds into Spain, the crop was widely spread all over the world. Currently, fi ve domesticated species are cultivated in many countries for their great economic value as vegetables, pungent food additives, colourants, pharmaceuticals, and even as popular medicines (Bosland and Votava 2000;Reifschneider 2000;Buckenhüskes 2003;Th ampi 2003;Ravishankar et al. 2003;Barceloux 2008;Yamamoto and Nawata 2009). Th e fruits of most Capsicum species contain signifi cant quantities of a great variety of metabolites (vitamins, carotenoids, minerals, proteins, carbohydrates, fats, fi bre) but their importance is derived from their possession of the major pungent compounds capsaicin and dihydrocapsaicin (Bosland and Zewdie 2001;Pruthi 2003;Manirakiza et al. 2003) which accumulate in the secretory epidermis of the septum (Filippa and Bernardello 1992).
Th ere is currently no consensus classifi cation of Capsicum itself. Th e infrageneric taxa proposed by Kuntze (1891), Wettstein (1891), Bitter (1921) and Hunziker (1956) have later been recognized as the segregate genera: Witheringia, Brachistus, Saracha Ruiz & Pav., Tubocapsicum (Wettst.) Makino, Aureliana (Hunziker 2001). More recently, diff erent classical and molecular cytogenetic analyses, crossing experiments, enzymatic studies, and chloroplast and nuclear DNA sequence studies (see references in Moscone et al. 2007 andBarboza et al. 2010), have allowed considerable progress in the characterization of infrageneric groups in Capsicum. At present, there is no formal infrageneric classifi cation. Two attempts at grouping species were made based on cytogenetic studies (Moscone et al. 2007), and a combination of data from enzyme, crossing and molecular studies (Walsh and Hoot 2001). In both studies the informal classifi cation is still considered provisional despite more than 50% of the species having been analyzed.
Working towards a complete treatment of wild Capsicum species I became aware of several instances of confusion regarding erroneous or uncertain names. In addition, some type specimens have not been found or have been destroyed in Berlin during the Second World War while others exhibit uncertain locality data or are composed of material from more than one species.
In this paper, lectotypes are designated for 14 names, and these are synonymized under their accepted names in Capsicum. In addition, a new name in Capsicum is proposed. In each case below, the locality information given for the lectotype corresponds with the information found on the specimen itself.  (Rusby) Hunz. (Hunziker 1956).
Williams 634 at NY is the best and the most complete of the two cited Williams collections, with fl ower buds, fl owers and mature fruits; it is selected here as lectotype. Th e second collection (Williams 623: BM!, K!, NY!, US!) is predominantly in fruit. After describing B. coccineus, Rusby (1926) transferred his species to Lycianthes, a placement later accepted by Morton (1944). Th is was due to Rusby's inspiration by Bitter´s monograph on Lycianthes (Bitter 1919), which states that the possession of calyx teeth is a basic feature of Lycianthes. Hunziker (1956), on the other hand, emphasized the importance of the androecium (more so than the calyx) for the generic delimitation of Capsicum and retained C. coccineum. In relation to B. coccineus, Hunziker noted that the anthers are longitudinally dehiscent and the typical prominent stapet (base of the fi lament broadened and fused to the corolla tube, with lateral auriculate appendages) of Capsicum species present in both syntypes. Analysis of the original material suggests that Miers (1849) described B. hookerianus based mostly on the plant cultivated at Kew Gardens, now representated as a specimen at K. Th is specimen (with a duplicate at BM) has a small pubescent calyx with entire margin and 5 teeth, and a subcampanulate corolla as described in Miers' (1849) protologue. Th ree sheets of Jameson's fi eld collection from Ecuador are deposited in Kew as B. hookerianus. In two of them (K000585918!, K000585920!), the only label data present is "Guayaquil" whereas in the third sheet (K000585919!), the label information is exactly as in the protologue. In addition, this specimen and a fragment of it at US show a 10-toothed calyx and a subcampanulate corolla. My analysis of the cultivated specimens at K and BM cited by Miers indicates that they belong to C. rhomboideum (Dunal) Kuntze. Hence the original material of B. hookerianus belongs to more than one taxon. Jameson´s fi eld collected material matches current usage of the name C. hookerianum (calyx conspicuously 10-toothed, corolla brilliant yellow, subcampanulate to campanulate, mostly with simple hairs, and leaves strongly attenuate) and is selected here as the lectotype.  (Hunziker 1956).

Brachistus pubescens
Th e transfer of B. pubescens to Capsicum necessitates a new epithet due to the earlier publication of C. pubescens Ruiz & Pav. (Ruiz and Pavón 1799). Both Hunziker (1956) and Heiser and Smith (1958) chose epithets alluding to its origin in the Galapagos Islands, with Hunziker´s name having priority. Erroneous type citations in Jørgensen and León-Yánez(1999) citing Stewart 3353 at CAS as the holotype, with an isotype at GH should be corrected to syntype and isosyntype respectively (see ICBN Art. 9.8;McNeill et al., 2006). Only two of the syntype collections have been critically examined (Stewart 3351: CAS!, GH!, NY!, MO!, US!; Stewart 3352: CAS!, GH!, US!); no duplicates of Stewart 3353 could be found at either CAS nor GH. Th e best preserved of these specimens is Stewart 3352 (CAS!) which includes a fl ower, numerous fl ower buds and fruits, and has label data in accordance with the protologue description; the other sheets of Stewart 3352 (GH!, US!) have mainly only fl ower buds. Capsicum galapagoense is rare and is the only endemic species of Capsicum in the Galapagos ( Chodat (1902) described this form using specimens from Paraguay he considered to belong to C. campylopodium, a Brazilian endemic. Th e type collections mentioned in the protologue (Hassler 1607 & 5134: A!, BM!, G!, K!, P!, S!, UC!, W!), and distributed in many herbaria constitute abundant and complete material, with buds, fl owers and fruits corresponding to the original description. Th e specimen Hassler 5134 in G is designated lectotype since it bears more fl owers. Th ese plants have nongeniculate pedicels, similar to the situation observed in C. fl exuosum, but not in C. campylopodium (Hunziker 1998;Barboza and Bianchetti 2005). Th e more abundant pubescence attributed to this form in the protologue is included within the variation of C. fl exuosum.
Capsicum hispidum Dunal var. glabriusculum Dunal,Prodr. 13 (1) Among the original material cited by Dunal (1852), only the fi rst widely distributed collection was successfully located. Dunal's protologue gives the collection locality as "In Mexico circa Bejar," now part of Texas (USA). After many searches for the correct name for the spontaneous variety of C. annuum (Shinners 1956;Heiser 1964;Heiser and Pickersgill 1969;Eshbaugh 1973, 1974), Heiser and Pickersgill (1975) confi rmed that C. hispidum var. glabriusculum is the correct basionym for the wild variety of C. annuum widely distributed in the Americas. baccatum (Hunziker 1998). Chodat and Hassler (1903) described this variety based on two very pubescent collections. Both specimens are complete and match the diagnosis, but the second one (Hassler 1926: BM!, G!, P!) has been found in fewer herbaria. Th us, I designate the widely distributed collection Hassler 6498 held in G as the lectotype. Th is variety was described as a xerophytic entity with a dense indumentum. Pubescence is a highly variable character in many Capsicum species (C. chacoense Hunz., C. rhomboideum (Dunal) Kuntze, C. parvifolium Sendtn., C. baccatum L., among others). After studying the original material of C. microcarpum var. tomentosum, Hunziker (1998) concluded that this taxon should be included in the range of variation existing in wild C. baccatum populations. Th is latter species is one of the few in Capsicum with a wide range of distribution in diff erent habitats where ecological conditions do not determine the degree of pubescence of the populations (Barboza, pers. obs.). For this reason, I agree with Hunziker's statement since the dense pubescence of Capsicum microcarpum var. tomentosum does not justify the validity of this taxon but fi ts very well under the wild C. baccatum var. Both syntypes are good quality specimens in agreement with the diagnosis. Berlandier 1907 at G is selected as the lectotype as its duplicates are broadly distributed. De la Sagra's specimen (nº 3, [año]1888) appears to be preserved only at G-DC! Berlandier´s collections held at G!, GH! (date on label: Oct. 1828), NY!, and P! are mainly in fruit whereas the ones at BM! and MO! also have fl ower buds or fl owers. Th e calyx shape and the fruit colour and shape are among the characters most useful for establishing the correct placement of this name.Th us, the small, cup-shaped and sub-dentate calyx and the globose-ovate red berry together with the presence of the solitary pedicels and white corolla (in sched. "corolla albida") clearly observed in these specimens are good matches for Capsicum annuum var. glabriusculum. Th e small size of the leaves is within the range of variation of this wild variety.  (Dunal) Kuntze (Jørgensen and León-Yánez 1999) Both syntypes cited in the original description are from the Andes of Ecuador. Th e specimen Spruce 5117 is deposited at K (K000201905!, K000201904!) and BM (BM000072741!) and apparently also at AWH. A note on sheet K000201904, states "Although Van Heurck quotes this number it does not agree with his description which answers to 5050 and is a Brachistus. N.E. Brown". According to Knapp (2002: 395), Spruce 5117 belongs to Solanum nudum Dunal. Th e collection Spruce 5050 has been widely distributed in diff erent herbaria, and is very complete. Two sheets of Spruce 5050 are held at Kew, where Spruce's original set is deposited; K000201915 is the better sheet with more complete data in the label, and is here chosen as lectotype. Spruce has handwritten "baccae nigrae" on this sheet, and the dense pubescence, the rhomboidal to elliptic leaves, 5-toothed calyx, and campanulate corolla are clearly visible and match the protologue. Th e calyx with clearly observable teeth and anthers with longitudinal dehiscence are typical of the genus Capsicum and the combination of characters cited above fi ts under C. rhomboideum. Th e mature fruits of this species are bright red more than black but often become purple or darker on dry specimens. Both syntypes were successfully located. Duplicates of Schott 5409 are housed at F!, GH!, and W!, and all of them are plants in fl ower. Th e Sellow collection has been distributed more widely; a photograph and a fragment (with fruit and seeds) of the Berlin specimen ("B 1543") destroyed during World War II are held at F; the P sheet of Sellow includes fl owering and fruiting material and has an original handwritten label assigning the number 6 to this collection. Of the two syntypes, the Sellow specimen at P is selected as lectotype since it is a complete one and it is possible to distinguish the most distinctive characters of C. campylopodium (toothless calyx, geniculate pedicels at anthesis, and black seeds with spine-like projections) (Barboza and Bianchetti 2005). Hunziker cited two syntypes. Th e fi rst collection (Hunziker 1907), is abundant consisting of 18 fruiting specimens held at CORD. Th e second one (Hunziker 7346) is here designated as lectotype; it originates from plants cultivated from seeds of the original collection and consists of two fl owering branches on a single sheet. A duplicate of this latter collection is a small specimen also in fl ower. Fruiting specimens of Capsicum are not only diffi cult to distinguish from other species of Capsicum but sometimes also from some species of Lycianthes as both genera share the similar calyx features. Th e fl owering collection is designated here as lectotype since the corolla is the best organ with which to characterize C. eximium (Hunziker 1950). Th e three syntypes were examined. Even though the second syntype, deposited at BR! ([Brazil] "Prope Rio de Janeiro, L.B. de Karwinski s.n., 1823"), matches the protologue, the Martius specimen at M is chosen here as the lectotype as it is more complete and better preserved. Both specimens show the characteristic features of C. mirabile such as the narrowly elliptic and glabrous leaves (young leaves with scarce short hairs on the margins), geniculate pedicels, glabrous calyx with 5 long teeth, stellate corolla, and black seeds with spine-like projections (Fig. 1) One of the two syntypes, the Sellow specimen at B (Sellow 79, in sched.), was destroyed and the only elements remaining of it are F neg. #2874 and an incomplete fragment at F! Th e other syntype (Schott 5416), here designated as the lectotype, is not a well-preserved specimen but the diagnostic characters of this variety (yellowish and long patent hairs on stem and pedicels, leaves densely pubescent in both surfaces, and angulate calyx with 5 short teeth), can be seen.  Fig. 2 Capsicum eximium var. tomentosum was described as an unusual glandular tomentose variety of C. eximium Hunz. Th e presence of this kind of pubescence densely covering the vegetative organs, pedicels, and calyx, and of a 5-10 toothed-calyx (Fig. 2  N) in specimens from a restricted area in south-central Bolivia (Dpto. Santa Cruz and Cochabamba) makes them quite diff erent from C. eximium. In fact, C. eximium has non-glandular pubescence (Fig. 3 M), a calyx with only 5 teeth (Fig. 3 C, E), and is distributed in a larger area (Bolivia: La Paz to Tarija and Argentina: Jujuy to Tucumán). Glandular hairs are rare in Capsicum. Th e dense indumentum of C. eshbaughii consists of diff erent types of glandular trichomes, some of them observed only in this species. Th e hair variations are: long simple hairs with multicellular and verrucate stalks and unicellular stipitate (Fig. 2 F) or not stipitate (Fig. 2 B) heads; short hairs with bicellular smooth stalks and multicellular heads (Fig. 2 J); and bifurcate hairs with both branches ending in a unicellular head (Fig. 2 L) or one branch non-glandular and the other longer and glandular (Fig. 2 H).

A new name for
Capsicum eshbaughii, together with C. eximium and C. cardenasii Heiser & P.G.Sm., is known as "ulupica" (Eshbaugh 1943 C); their very pungent fruits are very much appreciated as a spice or as pickles in the kitchens of Bolivia (Eshbaugh and Smith 1971).
Th is species is named in honor to Dr. W.H. Eshbaugh (Miami University) who fi rst recognised the distinctness of this species and devoted part of his research to the taxonomy of chili peppers.