Flora of North America species comparison

Subspecies 2 (1 in the flora).

Archaeological and molecular-genetic research, especially data from mitochondrial DNA and RAPD studies (O. I Sanjur et al. 2002; D. S. Decker et al. 2002b) and earlier isozymic and chloroplast DNA studies (for example, Decker et al. 1993), indicates that Cucurbita pepo in the broad sense includes two lineages: (1) C. pepo in the strict sense, a Mexican lineage of domesticates that differs from plants generally identified previously as C. pepo subsp. ovifera (here as C. melopepo) by a derived molecular feature (a difference in three adjacent base pairs) that occurs also in the C. moschata and C. sororia L. H. Bailey and C. argyrosperma Huber groups, and was shared presumably by the wild ancestor of C. pepo, which is unknown and possibly extinct; and (2) C. melopepo, a lineage of northeastern Mexico and the eastern Unites States in which the three wild varieties (var. fraterna, var. ozarkana, var. texana) and the domesticated variety (var. ovifera) share identical mitochondrial DNA sequences (Sanjur et al.) as well as similarities in isozymes and other kinds of DNA. Domesticates of C. pepo and C. melopepo are independently derived lineages.

Cucurbita pepo subsp. gumala Teppner comprises domesticates from Guatemala and adjacent southern Mexico and apparently is native there (H. Teppner 2000, 2004). The plants have depressed-globose pepos 13–20 cm in diameter with extremely thick rind, ripening orange-yellow, and with orange flesh. Teppner observed that the fruits of subsp. gumala are similar to the ancient ones from Guilá Naquitz cave in Oaxaca.

Cultivars of Cucurbita pepo with edible pepos have been divided into eight groups (H. S. Paris 1986, 1989; see also E. F. Castetter 1925), based mainly on pepo morphology. Pepos of cultivated forms differ from those of their wild ancestors in their larger size and more variable shape, less durable and more varicolored rinds, and less fibrous, nonbitter flesh.

Plants of Cucurbita pepo are likely to be found as non-persistent waifs all over the world, wherever they can be grown in temperate and upland tropical areas.

(Discussion copyrighted by Flora of North America; reprinted with permission.)

Subspecies 2 (1 in the flora).

Cucurbita okeechobeensis originally was common in extensive pond-apple (or custard-apple, Annona glabra) forests on the southern shore of Lake Okeechobee (J. K. Small 1922). By the time it was described in 1930, most of these forests had been destroyed. Currently C. okeechobeensis is known only in two population systems Decline and current rarity of Cucurbita okeechobeensis has resulted mostly from conversion of swamp forests to agriculture and from water-level management in Lake Okeechobee. Fluctuations in lake level are necessary as high levels facilitate dispersal, and the inundation and withdrawal create open habitats for quick growth of the seedlings. Exotic vegetation around the edges of the lake, especially Melaleuca, also is a threat.

Cucurbita okeechobeensis subsp. martinezii (L. H. Bailey) Andres & Nabhan ex T. Walters & D. S. Decker (C. martinezii L. H. Bailey) is widespread in a narrow strip of eastern Mexico (Chiapas, Oaxaca, Puebla, San Luis Potosí, Tamaulipas, Veracruz), often in riparian vegetation and as a weed in coffee and citrus plantations. Subspecies okeechobeensis is morphologically distinct only in its densely pubescent staminate hypanthia and pubescent pistillate ovaries. There is no sterility barrier between the two, and F1 hybrids are fertile (R. W. Robinson and J. T. Puchalski 1980). There is considerable allelic variation within subsp. martinezii (T. C. Andres and G. P. Nabhan 1988) but little or none in subsp. okeechobeensis.

R. W. Robinson and J. T. Puchalski (1980) found through isozyme analysis a single allelic difference between the Okeechobee and Martinez gourds. T. W. Walters and D. S. Decker (1991) confirmed the minor allelic difference and estimated the time of divergence as

about 450,000 years, with the conclusion that the two were appropriately treated as conspecific. A population of Cucurbita okeechobeensis was recently discovered in the Dominican Republic (B. Peguero and F. Jiménez 2005); the authors were unable to assign the plants to either subspecies and it is unclear whether the population is natural or the result of introduction within the last century (M. Nee, pers. comm.).

The closest relative of Cucurbita okeechobeensis and C. martinezii appears to be C. lundelliana L. H. Bailey (T. C. Andres and G. P. Nabhan 1988), which is geographically isolated in the Yucatán Peninsula but can form artificial hybrids with them. Cucurbita lundelliana differs in its orange corollas, less lignified exocarp, and seeds with crenulate margins.

(Discussion copyrighted by Flora of North America; reprinted with permission.)