Geographic Distribution of FLCP:

Data relating to plant distributions are housed in herbaria, collections of plant specimens that are maintained by private and public institutions throughout the world. While this is a unique and extremely valuable resource, interpretation of herbarium data requires sensitivity to possible bias. Some plants are collected more often than others. Herbarium collections often house a relatively full assemblage of rare plants that are of interest to the field botanist and relatively easy to collect and process. Thus, genera and species of orchids are usually well represented in herbarium collections. On the other hand, common weeds that are difficult to identify, possibly not part of the native flora ('escaped' garden plant), and carrying awkward structures that are difficult to manage in a plant press (large, fragile flowers and gourds), are often neglected by the field botanist and, as a result, not present in herbarium collections. Given this potential bias, it is reasonable to assume the FLCP is under-collected and therefore under-represented in herbarium collections. The herbarium sample is also biased geographically in that areas near centers of botanical activity (universities, botanical gardens) are more intensively sampled. The distribution data presented here must be interpreted with these factors in mind.

The first effort to map the distribution of FLCP was made by Andres (1983). Subsequent work in this area has been reviewed by Smith (1992). The distribution presented here (Fig. 5) is based on data compiled by Asch and Sidell, (1992), Smith (1992) and Cowan and Smith (1993). This has been expanded by inclusion of data from Andres (1983), unpublished collection data from research collections at Texas A& M, and data taken from a request for information sent to 77 herbaria and 35 weed specialists from the eastern U.S. as part of the development of this report. While most of the county records depicted in Fig. 5 are vouchered by herbarium specimens, some represent sightings of the plant made by field botanists or weed specialists. The nature and source of these records are listed in Appendix I, which is an extension of Smith's table 4.1 (1992-chapter 4).

 The distribution depicted in Fig. 5 represents records of established, free-living populations. Reports that might represent recent escapes from cultivation were not mapped, although these were included in Appendix I.

Documented FLCP populations are recorded for 125 counties in 12 states. Given dates of collection (Appendix I), it is reasonable to assume that, with the possible exception of the extreme West Texas cluster (area 'A', fig. 5), this is a reasonably accurate depiction of the general range of distribution of extant populations. Clusters of county records from south-central Texas (Texas A& M), northwestern Arkansas (University of Arkansas), south-central Alabama (University of Alabama), and western Missouri (Missouri Botanical Garden) probably reflect institutional-proximity collection bias. This probably also explains relatively large areas in the central portion of the range of distribution, such as northern Alabama/Mississippi and western Tennessee, that lack FLCP records. Thus, a more dense pattern of county records would be expected from a fully accurate FLCP distribution map. This would, however, probably correspond to patterns of drainage in that the entire range of distribution is associated with river systems that lead from the central U.S. to the Gulf of Mexico. This corresponds to the floodplain or 'riverine' ecological adaptations of FLCP.

 The southeastern limit of distribution depicted in Fig. 5 is problematic. Duncan and Kartesz (1981) list C. pepo var. ovifera as present on the Georgia coastal plain, and there are records for C. pepo growing wild in the Carolinas (Asch and Sidell, 1992). However, contacts with a full complement of field botanists at the University of Georgia, floristic workers from northern Florida, and agricultural experiment station weed specialists from several areas of Georgia (S. Brown, pers. com.) produced no FLCP records for the states of Georgia and Florida. Northern limits are also uncertain. A fully accurate map might include more county records from Indiana and southeastern Kansas, and possibly some records for Ohio.

As indicated above, the genetic structure of FLCP within this range of distribution is complex and ill-defined at this point in time. It appears, however, that plants showing the classic 'wild type' morphology of free-living Cucurbita species, as depicted by L. H. Bailey (Fig. 3) can only be found along streams that lead to the Gulf of Mexico in Texas and the Ozark highlands of Arkansas, Missouri, and Oklahoma (areas 'B' in Fig. 5). While other populations are comparable in terms of vegetative morphology, patterns of variation in fruit color, shape, and the absence of bitterness, suggest genetic linkage - either ancient (Cowan and Smith, 1993) or recent (Asch and Sidell, 1992) - with domesticated forms of the species.

Geographic Distribution of Domesticated C. pepo within the FLCP range of distribution:

An accurate mapping of domesticated C. pepo populations during a given growing season would probably include nearly every county within the 12 state FLCP distribution presented in Fig. 5. Domesticated C. pepo, both ssp. pepo (zucchini) and ssp. ovifera (crookneck, acorn) are common elements of the U.S. home garden. With the exception of the trans-pecos and Edwards Plateau areas of western Texas, it is reasonable to assume that home gardens are common throughout the range and also that many of these support cultivated populations of C.pepo domesticates. While this aspect is difficult to quantify, a Gallop Poll census contracted by the U.S. Gardening Association (B. Butterfield, pers. com.) indicates that 40% of all households in the mid-western U.S. maintained a home garden during the 1991-1992 growing season. Of these 9,300,000 gardens, 18.3% grow 'summer squash' (mostly 'yellow crookneck' and 'zucchini'), 9.8% have 'winter squash' (mostly 'acorn squash') as part of the garden, and 13.5% cultivate 'pumpkins'. These data indicate that possible weed/crop hybridization events in the central U.S., with only 'summer squash' as the putative domesticated pollen source, involve 1,701,900 local, cultivated populations that are distributed throughout the FLCP range of distribution. The entire FLCP range is within the range of distribution of specialized pollen vectors, 'squash bee' species of the genera Xenoglossa and Peponapis (Hurd and Linsley, 1964). These large, solitary bees can transfer pollen between domesticated and free-living Cucurbita populations that are separated by at least 1300 m (Kirkpatrick and Wilson, 1988). Density of home gardens within the 12 state FLCP range is probably highest in fertile lands of floodplains and valleys; the habitat of FLCP. Thus, in tefrms of both population density and ecology, crop/weed genetic interaction within the FLCP range is most likely to involve domesticated elements inhabiting home gardens. If only 1% of home gardens carrying 'summer squash' fall within 'bee range' of a FLCP population, then over 17,000 sites would be expected to show some type of crop/weed genetic interaction during a given growing season.

In terms of crop-to-weed gene flow, large monocultures of domesticates that are typical of commercial plantings constitute a significant element in that these present a massive pollen source. The national census of agricultural production, produced by the U.S. Census Bureau, tracks national commercial production of 'squash' and 'pumpkins'. Data listed under these names are mostly linked to plants assignable to C. pepo cultivars. While commercial production of other taxa, such as C. argyrosperma (cushaw squash), C. moschata (butternut squash), and C.maxima (large pumpkins), are included in these figures, it is assumed here that their contribution is minimal.

The most recent data for 1987 indicate that 58,198 acres of squash were produced at 7,763 farms in the U.S. National production of pumpkins at 6,921 farms was drawn from 40,652 acres in 1987. Total squash production within the 10 'core' (excluding Kansas and Indiana) state FLCP range of distribution was, in terms of acres harvested, 13.7% of the national total. Commercial pumpkin crops were taken from 11,189 production acres within the FLCP range of distribution, which was 27.5% of the national total. A ranking of commercial production by state for both squash and pumpkin acreage combined, as a percentage of the national total, is presented in Fig. 6. This figure is based on data taken from the 1987 agricultural census that are present here as Appendix II. States within the 'core' FLCP range of distribution (Fig. 5) are indicated by an arrow in Fig. 6 and bold font in Appendix II.

 As indicated by these figures, significant levels of commercial C. pepo production occur in Illinois (#3), Texas, (#5), and Tennessee (#20). The southern center, Texas, features high levels of squash production (ssp. ovifera var. ovifera) whereas there is a shift to pumpkin production (ssp. pepo) at the northern portion of the FLCP range of distribution. Examination of state census data by county reveals that 41% of Illinois pumpkin production is centered along the upper Illinois River bottoms in Tazewell County (Fig. 5) and the state center for squash production (49% of the state total) is adjacent Mclean County. Of the 15 Illinois counties known to carry populations of FLCP, 3 are listed by census data as squash producers (8% of the state total acreage) and 4 report a total of 186 acres of commercial pumpkin production (3% of the state total). Squash production in Texas is centered along the lower Rio Grande Valley in Cameron County with 25% of total harvested acres. The set of 33 counties with documented records of FLCP populations in Texas includes 5 that report squash harvest of 97 acres (2% of the State total) to the census.

Is FLCP a weed? Is it a common weed or a serious weed in any localities? What weedy properties does it possess?

prior section	Table of Contents		next section