Solanum retroflexum Dunal

Solanum retroflexum is a species that shows great variation in its indumentum, varying from nearly glabrous to densely pubescent with either eglandular or glandular trichomes. AFLP studies have shown that specimens of S. retroflexum with different indumentum types are genetically highly similar (Manoko 2007, as S. hirsutum (Vahl.) Dunal, here recognised as a synonym of S. memphiticum). Developmental studies have shown that papillate glandular hairs are present on all young parts of Solanum plants (Seithe 1962, 1979; see Morphology above), suggesting that in the absence of other distinguishing features, glandular pubescence might not be taxonomically significant. We circumscribe S. retroflexum here to include both eglandular and glandular populations; this character is polymorphic in most species of morelloids in the Old World, perhaps due to genetic lability due to polyploidy.

The species can be distinguished from other Old World morelloids based on the combination of characters of few-flowered inflorescences with 3-7 flowers, relatively long filaments (1.2-1.5 mm long) compared to anther length (1.3-1.8(-2.0) mm), strongly reflexed calyx lobes in fruit, and matte purple-black berries that lack stone cells and drop without the pedicels. Leaf shape is generally rhomboidal, which is characteristic of the species and helpful in distinguishing it from closely related S. nigrum and S. villosum. Calyx lobes are generally longer than in S. villosum, S. scabrum or S. nigrum. It can be distinguished from the morphologically similar S. villosum based on its opaque purple-black fruit, its pedicels that remain behind after fruits drop, deeply stellate corolla with long corolla lobes, and the long filaments as compared to anther length.

Solanum retroflexum is cultivated commercially in South Africa in KwaZulu-Natal and Free State for jam production (Viljoen 2011), and farmers select individual plants with large berries to propagate for the next season. In one farm in KwaZulu-Natal 10 tons of berries harvested were used to make 20,000 bottles of “Umsobo” jam, most marketed in the Gauteng area. On other farms S. retroflexum is cultivated along with S. chenopodioides and S. americanum, and putative hybrid plants were found (Viljoen 2011). Leaves of S. retroflexum are also produced under irrigation in Limpopo province, for the commercial market in the local area (van Averbeke et al. 2007). The markets for both leaves and fruits are controlled by small scale rural traders in South Africa.
Solanum retroflexum is a tetraploid species with unclear parentage. Previous studies have suggested that one of the contributing parents of the tetraploid S. retroflexum is the diploid S. americanum (Edmonds 1977), supported by crossing studies showing vigorous hybrids between the autotetraploid S. villosum derived from S. americanum and S. retroflexum (Ganapathi 1987). Jacoby and Labuschagne (2006), however, reported that crosses of S. chenopodioides and S. retroflexum are much more successful than crosses between S. retroflexum and S. americanum. Ganapathi & Rao (1980) also reported weak chromosome pairing at low frequence in hybrids between S. retroflexum and S. americanum. Molecular studies suggest that either S. chenopodioides or S. nitidibaccatum are the likely parental species of S. retroflexum (Jacoby et al. 2003; van der Walt et al. 2008; Poczai and Hyvönen 2011). Further molecular studies are needed to fully establish the parental species of S. retroflexum, but the current data from molecular and crossing studies suggests that S. chenopodioides or S. nitidibaccatum are the likely best candidates.

Solanum retroflexum or one of its close relatives has been suggested to have contributed to the origin of the hexaploid S. nigrum (Ganapathi and Rao 1986c). Most current evidence suggests, however, that S. villosum and S. americanum were the parents of S. nigrum (Edmonds 1979a; Ganapathi and Rao 1986b; Poczai and Hyvönen 2011).

In describing S. retroflexum Dunal (1852) cited two elements; one from Saudi Arabia (“in Arabia circa Taifa”) and the other from South Africa (“caput Bona Spei, Drege”). He then used two different collections held at G-DC to describe the varieties. The specimen from Saudi Arabia corresponds to S. villosum (P00055172), so we have used the flowering collection (Drège 7864b) at G-DC (G00144331) that is also the holotype of var. angustifolium as the lectotype of S. retroflexum, making the two names homotypic.

Solanum burbankii was described from living material sent to Bitter ("colui ipso in horto Bremeno complures per annos e seminibus ab horto Hamburgensi acceptis" [grew in the Bremen garden for several years from seeds originally accepted from Hamburg]) and thought to have been from California (Bitter 1913b), and thus perhaps from Luther Burbank himself. Bitter (1913b) cites no herbarium specimens in the protologue; if they were in either Bremen or Berlin they were destroyed during the Second World War, and we have found no specimens in Goettingen where some small fragments of collections are found (Vorontsova and Knapp 2010). From the description of a somewhat pubescent plant (“utrinque sparsim breviter pilosa”) with pruinose berries with no stone cells, in which the calyx lobes are strongly reflexed at fruit maturity, we are certain Bitter is describing a sparsely pubescent form of S. retroflexum. We have selected a neotype for S. burbankii from specimens grown in St. Louis, Missouri that were obtained from the original distributor of Burbank’s “wonderberry” seeds, J.L. Childs of New York. These specimens are almost certainly among those seen by William Trelease during his part in the newspaper war that was waged over the identity of the “wonderberry” in the early part of the 19th century (see Heiser 1969) and match the description well. Several collections grown in Mr. Waldstein’s garden are held at MO, and we have chosen the sheet with the best fruiting material as the neotype.

The Swiss botanist Rudolf Probst worked extensively on the adventive flora found around woollen mills in Solothurn (Switzerland), and sent much of his Solanum material to the Hungarian botanist Sandor Polgár, apparently both as specimens and as seeds. In his various publications (Probst 1928, 1938) he quoted verbatim from Polgár’s letters, attributing both names and descriptions to Polgár for S. nigrum var. probstii and S. burbankii var. glabrescens. The numbers he cites in the protologues apparently refer to years of collection (e.g., “22, 24, 26, 27”) but we are not certain; because of the labels on specimens in BP seen and annotated by Polgár we are assuming Probst sent herbarium specimens rather than seeds. We are therefore considering this material as lectotypes for these two names, both of which correspond to S. retroflexum. The letter quoted verbatim in the protologue of S. nigrum var. probstii is mounted on a sheet with a collection date (29 Jul 1929) later than the description, so we select another Probst collection from 24 Aug 1926 and annotated by Polgár (BP acc. # 485285) as the lectotype for this name. The specimen collected by Probst on 20 Oct 1922 (BP acc. # 485327) has elements of the text repeated in the protologue attached to it, and we designate this as the lectotype for S. burbankii var. glabrescens.