Free-living Cucurbita pepo in the United States
Viral Resistance, Gene Flow, and Risk Assessment

What is Cucurbita texana?

Cucurbita texana is a free-living system of populations that fall within the primary gene pool of Cucurbita pepo, a species that also includes fully domesticated elements and other free-living population systems.

The category 'species' is used here in the biological sense, i.e., to denote those population systems that share a common ancestor and lack full reproductive isolation. The term 'free-living' is assigned to plant populations that are able to survive, without direct human assistance, over the long term in competition with the native flora. This is a general ecological category that includes plants that colonize open, disturbed, prime habitat that is either under human control (weedy populations) or natural disturbed areas such as river banks and sand bars (wild populations). Domesticated populations show structural features that have resulted from human selection and, usually, an inability to maintain long term populations without human assistance.

A rational approach to any problem involving free-living C. pepo of the United States requires definition of the biological elements involved as well as the relative position of these elements within the broader biological context of the species. This involves two fundamental aspects of plant systematics: classification and phylogeny.

Classification: The biological connection between C. texana and C. pepo was not appreciated when the plant was first 'discovered' by Western Science. The first documented collections of the plant were taken from South Texas in 1835 by J. L. Berlandier, a botanist working for the U.S.-Mexico Boundary Commission. G. H. A. Scheele, examining collections of the plant made by F. J. Lindheimer along the upper Guadalupe River in 1845, decided that these specimens represented a genus of the Cucurbitaceae new to science, and provided the name Tristemon texanum in 1848. Asa Gray later studied the Lindheimer collections from Texas and noted that the 'Texas Gourd' was quite similar to forms of the domesticated ornamental gourds, then known as Cucurbita pepo var. ovifera. In an effort to insure that the classification system reflected this similarity, Gray changed the name from Tristemon texanum to Cucurbita texana in 1850. While the 'Texas Gourd' remained formally classified as a distinct species of Cucurbita until 1988, students of the genus recognized a close affinity between this free-living taxon and C. pepo. Debate centered on the position of C. texana as either a native element of the flora and possible progenitor of the domesticate, or a derivative, feral 'escape' from pre-historic cultivation of ornamental gourd-like plants that originated in Mexico (Bailey 1929;1943;Erwin, 1938). This debate has continued to the present (Heiser, 1985), although data generated in the 1980s has placed the problem of 'Texas Gourd' classification as one component of a broader problem that involves free-living types from Mexico and the eastern U.S.

Cucurbita fraterna, a putative relative of C. pepo was rediscovered at several sites in northeastern Mexico during the early 1980s (Nee, 1990). Subsequent comparative analyses indicate that C. fraterna will hybridize with C. pepo (Nee, 1990) and, in terms of both allozyme genetic identities (Wilson, 1989;Decker-Walters et al., 1993) and shared chloroplast DNA restriction site mutations (Wilson et al., 1992) placement of this taxon as a free-living element of C. pepo by Andres (1987) is generally accepted. Similar criteria, cross-compatibility and high indices of genetic similarity, were used by Decker (1985, 1986, 1988) to formally place the Texas Gourd as an intraspecific element of C. pepo.

In addition to justifying the placement of two free-living elements within C. pepo, recent studies of molecular genetics have also revealed a fundamental pattern of intraspecific variation, and the classification has been changed to reflect that perspective. As indicated by congruent patterns of variation among both from nuclear (Decker, 1985;Decker and Wilson, 1987;Wilson, 1989;1990) and plastid (Wilson, et al., 1992) molecular markers, the C. pepo lineage is composed of two fundamental elements, formally identified as subspecies ovifera and pepo. These, however, do not represent free- living vs. domesticated lineages. Subspecies pepo includes only domesticated elements. These include all Mexican landraces examined to date, all 'pumpkin' cultivars, all 'marrow' (zucchini, cocozelle, etc.) cultivars, and a subset of the ornamental gourd types (orange ball, miniature ball, warty hardhead) that usually carry reddish pigment in the fruit. Subspecies ovifera, on the other hand, contains a mixed group that is organized into three varieties. Variety ovifera is the domesticated element of the subspecies. It contains domesticated ornamental gourd types that are not included in ssp. pepo (spoon, egg, bicolor, crown of thorns, pear, etc.), as well as cultivars of acorn squash, crookneck squash, straightneck squash, and scallop or 'paddy-pan' squash. As most recently defined (Decker-Walters, et al., 1993), free-living gourds that occur in Texas are classified as C. pepo ssp. ovifera var. texana. Similar, free-living plants that occur in the U.S. beyond Texas (Illinois, Missouri, Arkansas, Oklahoma, and Louisiana) are placed in C. pepo ssp. ovifera var. ozarkana.

A summary of C. pepo classification, both 'traditional' (based primarily on plant structure) and 'current' (based on crossability and genetic structure), is presented here as Fig. 1. The 'current' alignment stands as a working hypothesis and a functional foundation for on-going studies of the species. While there is general agreement on fundamental aspects of this classification system, i.e., a basic, two-parted pattern of differentiation of the species indicated by the "Lineage Division" of Fig. 1, and conspecific placement of free- living and domesticated elements, other aspects of the classification are debated and under test at this point in time. However, primary areas of disagreement among those working with the C. pepo 'problem' do not center on the pattern of genetic differentiation that has produced the entities defined as subspecies and varieties in Fig. 1. These have been relatively well defined by a series of comparative studies that show considerable congruence. The 'Texas Gourd' has not been treated as a distinct species by specialists since the formal name change in 1988 (Decker, 1988;Decker-Walters, 1990;Decker-Walters et al., 1990;Decker-Walters et al., 1993;Harlan, 1992;Heiser, 1989;Smith, 1992;Cowan and Smith, 1993;Wilson, 1989;1990). There are, however, several competing hypotheses regarding phylogenetic linkage between both free-living and domesticated elements of C. pepo. As indicated below, these different notions relate to eastern North America as a possible center of New World agricultural origin and genetic differentiation under human selection. This, in turn, produces different perspectives on the biological/historical significance of free-living gourd populations currently inhabiting eastern North America.

Phylogeny: As indicated by the structural difference between an 'egg' gourd cultivar and a world record pumpkin that might weigh several hundred pounds, C. pepo shows remarkable polymorphisms in plant structure. It also shows a worldwide pattern of distribution within a wide range of habitats, both cultivated and natural. The extant pattern of distribution and variation represents the end points of a lineage or 'family tree' that evolved through time. Phylogenetic analysis is an attempt to define this evolutionary history.

 It is generally assumed that, prior to the development of New World agriculture, C. pepo existed as a single, wild-growing entity. This is based on comparative analyses of crossing relationships (Whitaker and Bemis, 1964;1975;1976), pollen vectors (Hurd et al., 1971), and chloroplast mutations (Wilson et al., 1992) among species of Cucurbita. All elements of the species, as depicted in Fig. 1, therefore share a common ancestor, i.e. the species is monophyletic. The common ancestor of extant members of the C. pepo lineage was probably an annual, gourd-producing plant adapted to colonization of disturbed ground in riverine habitats. The archaeological record indicates that this free-living ancestral type initially became involved as a fundamental element of the New World agricultural revolution in the highlands of Oaxaca, Mexico more than 10,000 years ago. (Smith, 1986). Subsequent evolution under both human and natural selection, coupled with human-mediated and 'natural' means of dispersal, has produced the remarkable suite of structural forms that now comprise the species and now show a global distribution.

The archaeological record shows domesticated C. pepo in central (Tehuacan) and northeastern (Tamaulipas) Mexico at about 7,000 years ago (McClung de Tapia, 1992). The plant is found among the first archaeological indications of agriculture in the northern desert borderlands of the Mexican agricultural center (southwestern U.S.) by about 3,000 years before present (Minnis, 1992). This sequence of dates establishes a relatively uncomplicated picture of origin and diffusion, with the fraterna types from northeastern Mexico or texana types from Texas representing extant descendants of the ancestral progenitor type of C. pepo (Nee, 1990). It appears, however, that this is not the full picture of domestication and differentiation of C. pepo.

Centers of origin are usually centers of genetic diversity. However, in terms of molecular and structural differentiation, landraces now under cultivation in traditional milpas throughout Mexico are remarkably uniform (Decker, 1985;Wilson, 1990). The produce section of most U.S. supermarkets carries a more diverse sample of C. pepo than could be obtained from a full sample of modern field collections taken from throughout Mexico. While native C. pepo cultivar diversity in Mexico could have been extirpated during European colonization, it is also quite possible that one arm of the C. pepo lineage - diverse elements associated with ssp. ovifera (Fig 1) - evolved under human selection in the eastern North America. (Whitaker and Carter, 1946;Decker, 1988;Smith, 1992;Cowan and Smith, 1993).

Recent archaeological work has demonstrated that C. pepo was present in the eastern U.S. by 7,000 years ago. It is difficult to determine with certainty if these early remains represent a domesticated form (Smith, 1987), although there appears to be a consensus that fully domesticated types were present in eastern North America at least 3,000 years ago (Heiser, 1989;Smith, 1992). Perspectives on the phylogenetic significance of C. pepo in eastern North America are reviewed by Smith (1992- chapter 4), Cowan and Smith (1993), Decker-Walters et al. (1993), and Asch and Sidell (1992). Data relating to this issue are a complex mix of biochemical genetics, archaeology, history, plant population biology, and personal inclinations. Assessment of the situation and interpretation of the data are difficult for the non-specialist as well as those actively engaged in the issue. Of various phylogenetic scenarios presented by Smith (1992), two are presented here (redrawn and modified) to depict opposing hypotheses. The first (diagram A) firmly roots the wild progenitor of C. pepo domesticates in Mexico, with human-mediated dispersal from Mexico into eastern North America whereas the second (diagram B) places the ancestral wild type with a broader distribution that includes eastern North America, and subsequent independent domestication in both Mexico and eastern North America.

Depictions of C. pepo evolution present as panels A and B in Fig. 2 are consistent with available data, both botanical and archaeological. In terms of gaining a better understanding of the genetic structure and phylogenetic history of C. pepo, the difference between these two notions is critical. However, for the purposes of this report, both scenarios contain common elements that speak directly to the issue of free-living gourds of the eastern U.S. and risk assessment:

Consideration of both classification and phylogeny reveals that the biological entity defined by the name 'C. texana' constitutes only a portion of a broader system of free-living C. pepo populations that inhabit eastern North America. Plants carrying this name are not biologically distinct or reproductively isolated from other elements of C. pepo, both domesticated and free-living. The U.S. populations show levels of genetic differentiation that allow classification as C. pepo ssp. ovifera var. texana (Texas) and C. pepo ssp. ovifera var. ozarkana (Illinois, Missouri, Arkansas, Oklahoma, and Louisiana). Populations from other areas of the U.S. distribution are difficult to classify, possibly because of hybridization or other connections to domesticated elements of the species (Asch and Sidell, 1992;Decker-Walters et al., 1993). Since the focus of this report is not restricted to the traditional 'texana' types that are limited to Texas, plants under consideration here will be referred to as free-living C. pepo of the U.S (excluding the Mexican ssp. fraterna), or FLCP.

What is known about the geographic distribution of FLCP, especially with respect to the areas where cucurbits are cultivated commercially?

Geographic Distribution of FLCP:

prior section  Table of Contents  next section