Solanum cheesmaniae

Solanum cheesmaniae is morphologically quite variable, but can be distinguished from its sympatric close relative S. galapagense by its less divided leaves, less numerous interjected leaflets, and shorter calyx lobes that do not enclose the ripe fruit. In the morphological analyses of Darwin et al. (2003) plants of Solanum cheesmaniae form a cohesive group despite considerable variation, but two extreme morphs show differences in leaflet shape, margin, leaf division and pubescence. The type specimen of S. cheesmaniae represents the ‘typical’ morphology, and has very small leaves and leaflets, with entire to regularly dentate margins and dense pubescence in all parts of the plant, and bearing short glandular trichomes on the adaxial surface of the leaflets. These characters are consistently present in specimens collected from the north coast of Santa Cruz and Baltra, San Cristóbal, Santa Fé, Pinzón and coastal Isabela. The other extreme morph of S. cheesmaniae, called by Darwin et al. (2003) the ‘Academy Bay’ morph, has leaves to three times the size of the ‘typical’ morph, irregularly dentate leaflet margins (see Figure 4 in Darwin et al. 2003), and the plants are altogether less pubescent; the sparseness of trichomes is especially notable on both leaflet faces. The ‘Academy Bay’ morph has been collected from near areas of human habitation on southern Santa Cruz (Academy Bay = Puerto Ayora). Other specimens that we here consider to fall within the ‘Academy Bay’ morph show intermediate leaf morphology compared to the two extremes. These plants have a velvety pubescence composed of short trichomes of similar length; this pubescence is more apparent on the abaxial face of the leaflets. These intermediates have been collected in Isabela, Fernandina and Santa Cruz, and also possibly occur on San Cristóbal and Española. Specimens of the ‘Academy Bay’ morph have been collected mostly from the southern sides of the islands or at high altitudes (areas with maximum precipitation). Many of these specimens were collected during El Niño years (see Quinn & Neal 1992 for a list of El Niño dates).

Hooker (1847) recognized three different species of tomatoes from the Galápagos, all based on the specimens collected by Charles Darwin. We found that one of these specimens documented as having been collected from San Cristóbal, and identify by Hooker as “L. pimpinellifolium”, belongs to the ‘Academy Bay’ morph. Some of the specimens cited by Müller (1940) as “Lycopersicon pimpinellifolium” are also assigned here to S. cheesmaniae ‘Academy Bay’ morph. Rick (1956, 1963) referred to three different tomato variants in Galápagos including a Galápagos “L. pimpinellifolium type” (TGRC accession number LA166), which had orange fruit but flowers with corolla divided 2/3 the way to the base (for illustration see Rick 1956). In his later work, Rick (1971) re-classified the Galápagos accessions of “L. pimpinellifolium” with orange fruits under “L. cheesmanii”. He considered the red-fruited tomatoes in Galápagos to be introduced species (C.M. Rick in litt. 1998). The leaf morphology of some specimens of S. cheesmaniae ‘Academy Bay’ morph from Isabela is similar to that of individuals of S. pimpinellifolium found in Galápagos and on mainland South America (coastal Perú and Ecuador).

The morphological variation in S. cheesmaniae is indeed complex. Rick (1963) also recognized this and pointed out that populations from the type locality on the northern side of Santa Cruz are intermediate in terms of pubescence density between S. galapagense and what we here define as the ‘Academy Bay’ morph of S. cheesmaniae. He also observed that ‘typical’ S. cheesmaniae shared morphological similarities with the ‘Academy Bay’ morph but resembled S. galapagense with respect to its shorter internodes. Rick (1963) felt that the typical S. cheesmaniae had leaves that were less divided than in the two other forms and with orbicular lateral segments (see leaf variation in Figures 4 & 5 in Darwin et al. 2003). Our observations are consistent with these morphological differences noted by Rick (1963). Using AFLP polymorphisms, Nuez et al. (2004) identified Galápagos tomatoes as “Lycopersicon cheesmanii”, and designated some of the TGRC accessions we identify as S. cheesmaniae ‘Academy Bay’ as “L. esculentum” and others as “L. cheesmanii long”. Our AFLP data from the S. peruvianum complex (Spooner et al. 2005) indicate that if hybridization is occurring or has occurred, then AFLP polymorphisms are not useful for distinguishing species that may be hybridizing. We know that S. cheesmaniae and S. pimpinellifolium hybridize in the Galápagos (S.C. Darwin, unpublished observations), complicating patterns of molecular relationships.

A comparison between herbarium specimens and greenhouse-grown progeny collected from the same plants showed that the greenhouse grown-plants had larger leaf dimensions than their field-collected parents, as is common in wild tomatoes. Size difference, however, did not wholly account for the differences between the typical S. cheesmaniae and the ‘Academy Bay’ morphs. Further investigations being undertaken by S. C. Darwin may resolve the relationships and taxonomic status of the ‘Academy Bay’ morph and to establish the reasons for these different morphologies found within S. cheesmaniae. Variation in S. cheesmaniae is potentially due to: (1) plants here recognized as the ‘Academy Bay’ morph could be an ecotype of S. cheesmaniae, (2) plants are potentially of hybrid origin involving S. pimpinellifolium, or (3) plants could be morphologically aberrant due to increased soil humidity found in the southern parts of the islands and during El Niño years.

Results from allozyme electrophoresis (Rick & Fobes 1975) suggested that the populations of S. pimpinellifolium from the coastal region of Motupe-Olmos (ca. 6ºS latitude) in the northern Peruvian department of Lambayeque were most closely related to the Galápagos tomatoes. They also suggested that the observation that the Humboldt Current flows from about 6ºS latitude to the Galápagos at certain times of year also lent support to this origin for S. pimpinellifolium in the Galapagos (Rick & Fobes 1975).

The spelling of the specific epithet has been corrected from cheesmanii to cheesmaniae; the species was named in honor of Lucy Evelyn Cheesman (1881–1969), who collected the type (see also Fosberg 1987; Spooner et al. 1993; McNeill et al. 2006). Although Fosberg (1987) pointed this out, spelling has not changed in the plant breeding literature, but floristic studies have consistently used the correct feminine ending (Jørgensen & León Yanez 1999).

The publication date of Hooker’s “Enumeration of Galápagos Plants”, in which he described both Lycopersicon peruanum var. parviflorum (=Solanum cheesmaniae) and Lycopersicon esculentum var. minor (=Solanum galapagense) is given on the frontispiece of Volume 20 of the Transactions of the Linnean Society of London as 1851, but Part II, in which the paper appeared (read on 4 May, 6 May and 16 December 1845) was available as a separate on 11 December 1847 (Raphael 1970). Thus the correct date of publication of the names published therein is 1847, not 1851 as it appears in most bound library copies of the Transactions.

The combination Solanum cheesmaniae (L. Riley) Fosberg (1987) postdates the epithet Solanum cheesemanii Gerasimenko, coined to honor the New Zealand botanist Thomas H. Cheeseman (a synonym of S. aviculare G. Forst.). The two names are very similar and the rejection of Gerasimenko’s epithet in favor of the commonly and widely used combination for this tomato species has been proposed (Knapp & Darwin 2006).