Hybridization of crops with nearby weeds may enable weeds to acquire traits we wish they didn't have, such as resistance to herbicides. Research (Kaiser, 2001b; http://www.osu.edu/researchnews/archive/radweed.htm) suggests that crop traits may escape from cultivation and persist for many years in wild populations. Genes that provide a competitive edge, such as resistance to viral disease, could benefit weed populations around a crop field (Kaiser, 2001b).

Gene flow from crops to weeds requires:

• the presence of sexually compatible wild or weedy relatives close to the crop,
• an overlap of flowering times between the crop and the wild relatives,
• the presence of a pollinating agent such as a bird or an insect unless pollination is accomplished by the wind.

Many cultivated crops have sexually compatible wild relatives with which they hybridize under favorable circumstances. The table below provides some of the documented instances of hybridization between crops and weeds.

crop species	pollination mode	area where studied	compatible relatives
alfalfa Medicago sativa	mostly cross-pollinated	USA	wild alfalfa Medicago sativa
asparagus Asparagus officinalis	mostly cross-pollinated	USA	wild asparagus Asparagus officinalis Wild asparagus. Photo: Kathleen Durbin, www.ediblewild.com
blueberry Vaccinium angustifolium Blueberries. Photo: www.noursefarms.com	mostly cross-pollinated	USA	wild blueberry Vaccinium angustofolium
burmuda grass Cynodon dactylon	mostly cross-pollinated	USA	wild burmuda grass Cynodon dactylon
carrot Daucus carota	mostly cross-pollinated	USA	wild carrot Daucus carota
celery Apium graveolens	self- and cross-pollinated	USA	wild celery Apium graveolens
chicory Chicorium intybus	mostly cross-pollinated	USA	wild chicory Chicorium intybus
clover Trifolium spp.	some cross-, some self-pollinated	USA	wild clover Trifolium spp.
corn Zea mays subsp. mays	mostly cross-pollinated	Mexico and Central America	wild relatives of corn Zea mays subsp. mexicana Zea mays subsp. parviglumis Zea mays subsp. huehuetenangensis Zea diploperennis Zea perennis Zea luxurians
cranberry Vaccinium macrocarpon Source: USDA	mostly cross-pollinated	USA	wild cranberry Vaccinium macrocarpon
foxtail millet Setaria italica	mostly self-pollinated	France	green foxtail Setaria viridis
lettuce Lactuca sativa	mostly self-pollinated	USA	wild lettuce Lactuca serriola
oats Avena sativa Source: Jim Manhart, Texas A&M Herbarium, www.csdl.tamu.edu/ FLORA/gallery.htm	mostly self-pollinated	USA	wild oats Avena fatua
oilseed rape, canola Brassica napus	mostly self-pollinated	France, USA	wild radish Raphanus raphanistrum wild brassicas Brassica napus Brassica campestris Brassica juncea
quinoa Chenopodium quinoa	mostly self-pollinated	USA	wild quinoa Chenopodium berlandieri
radish Raphanus sativus	mostly cross-pollinated	USA	wild radish Raphanus raphanistrum
rice Oryza sativa Source: USDA	mostly self-pollinated	USA	red rice Oryza sativa
tobacco Nicotiana tabacum	most self-pollinated	USA	tobacco escaped from cultivation Nicotiana tabacum
sorghum Sorghum bicolor	mostly self-pollinated	USA	Johnsongrass Sorghum halapense
squash Cucurbita pepo	mostly cross-pollinated	USA	wild squash Cucurbita texana
strawberry Fragaria X ananassa Source: USDA	self- and cross-pollinated	USA	wild strawberry Fragaria virginiana
sugar beets Beta vulgaris	mostly cross-pollinated	France	wild beets Beta vulgaris
sunflower Helianthus annuus Source: USDA	mostly cross-pollinated	USA	wild sunflower Helianthus annuus
walnut Juglans regia	self- and cross-pollinated	USA	California walnut Juglans hindsii
wheat Triticum aestivum Source: USDA	mostly self-pollinated	USA	jointed goatgrass Aegilops cylindrica Source: California Department of Food and Agriculture

Sources for table:
Arriola and Elstrand, 1996; Baranger et al., 1995; Desplanque et al., 1999; Doebley, 1990; Klinger et al., 1992; Saeglitz et al., 2000; Snow and Palma, 1997; Till-Bottraud et al, 1992; Zemetra et al., 1998.

The likelihood that transgenes will spread can be different for each crop-and-weed combination in each area of the world.

For example, there are no wild relatives of corn or soybean in the United States. If pollen grains from transgenic corn and soybean plants are released in these areas, they do not encounter any compatible weeds to pollinate, so there is no risk of gene flow. But wild relatives of corn do live in Mexico, so there is a risk that these plants could acquire foreign genes if transgenic corn is grown in Mexico.

Alfalfa pollen is carried by bees from one plant to another. U.S. government researchers report that bees can carry alfalfa pollen at least two-thirds of a mile (http://www.ars.usda.gov/is/AR/archive/oct01/pollen1001.htm). Thus, if transgenic alfalfa is grown in the future, bees might carry the pollen to wild alfalfa populations, resulting in wild plants that have whatever trait was transferred into the cultivated plants.

The risk of gene transfer via pollen is much smaller when crops and weeds are self-pollinating. Self-pollinating plants such as wheat and rice employ their own pollen to produce seeds, They make little use of pollen that might be available from nearby plants. Nevertheless, some exchange among plants does occur at a low rate.

The small risk of gene flow in certain situations must be balanced against the magnitude of the consequences if some gene flow does occur. Jointed goatgrass, a wild relative of wheat, is a serious weed problem in the United States. The consequences of jointed goatgrass acquiring herbicide resistance must be considered when decisions are made about whether to grow herbicide-resistant wheat.

Rice growers struggle to suppress a weed called red rice. A University of Arkansas researcher is tracking the escape of transgenes from cultivated rice to red rice. (http://pigtrail.uark.edu/pubs/Research_Frontiers/fall_2000/05_Feature1.4.html)

A report by the European Environment Agency assessing the potential for gene flow to the natural environment from six major crops--oilseed rape, sugar beet, potatoes, maize, wheat, and barley--is available at http://reports.eea.eu.int/environmental_issue_report_2002_28/en.

The proceedings from a conference on the risks of gene flow from crops to wild relatives are available at http://www.biosci.ohio-state.edu/~lspencer/gene_flow.htm.

The phenomenon of crop-to-weed gene flow is complex. Each crop must be evaluated individually for the risk of gene flow in the area where it will be grown.