Theor Appl Genet (1994) 89:655-660 9 Springer-Verlag 1994 D. M. Arias 9 L. H. Rieseberg Gene flow between cultivated and wild sunflowers Received: 16 December 1993 / Accepted: 18 April 1994 Abstract With the development oftransgenic crops, con- cern has been expressed regarding the possible escape of genetically-engineered genes via hybridization with wild relatives. This is a potential hazard for sunflowers because wild sunflowers occur as weeds in fields where cultivated sunflowers are grown and hybridization between them has been reported. In order to quantify the potential for gene escape, two experimental stands of sunflower cultivars were planted at two sites with different rainfall and alti- tude profiles. Populations of wild plants were planted at different distances from each cultivar stand. An allele ho- mozygous in the cultivar (6Pgd-3-a), but absent in the wild populations, was used as a molecular marker to document the incidence and rate of gene escape from the cultivar into the wild populations of sunflowers. Three-thousand achenes were surveyed to determine the amount of gene flow from the cultivated to the wild populations, The mar- ginal wild populations (3 m from the cultivar) showed the highest percentage (27%) of gene flow. Gene flow was found to decrease with distance; however, gene flow oc- curred up to distances of 1000 m from the source popula- tion. These data suggest that physical distance alone will be unlikely to prevent gene flow between cultivated and wild populations of sunflowers. Key words Hybridization 9 Gene flow 9 Sunflowers Transgenes Introduction Plant breeders have long been concerned with gene flow between crop plants and their wild relatives (Anderson Communicatedby J. Mac Key D. M. Arias Centro de Investigaci6nen Biotecnologfa,UniversidadAut6noma del Estado de Morelos, Cuernavaca,Morelos,Mexico 62210 L. H. Rieseberg ([]) Department of Biology,IndianaUniversity,Bloomington, Indiana 47405, USA 1949; Harlan 1965; De Wet 1975; Barrett 1983; Ellstrand 1988). Early studies mainly focused on gene flow into crop strains because of fears that seed lots would be contami- nated by foreign germplasm (Sprague 1938; Crane and Mather 1943; Haskell 1943; Bateman, 1947 a, b, c; Jones 1948; Hutchcroft 1955; Nieuwhof 1963). More recently, concern has been expressed regarding gene flow from crop plants into their wild relatives via hybridization. In partic- ular, it has been suggested that genetically-engineered genes (transgenes) may be transferred into natural or weed populations through hybridization, potentially creating in- vasive weeds or increasing the difficulty of weed control (Ellstrand 1988, 1992; U. S. National Research Council 1989; Crawley 1990; Keeler and Turner 1990; Manasse 1992). Nonetheless, there are few examples where the extent of gene flow between cultivated and weedy populations have been quantified (Kirkpatrick and Wilson 1988; Lan- gevin et al. 1990; Klinger et al. 1991; Till-Bottraud et al. 1992; Wilson and Manhart 1993). The lack of quantitative data documenting crop-weed gene flow can be traced to a number of possible causes including the facts that: (1) the potential environmental consequences of crop-weed hy- bridization have only recently been widely publicized; (2) most of the earlier work employed characters whose ge- netic basis was unknown, making it difficult to determine whether the observed patterns of variation were due to hy- bridization and introgression, convergent evolution, or plasticity; (3) because of the close relationships between many crops and their wild relatives, it has often been dif- ficult to find genetically-based markers that are exclusive to the crop plant. The potential for gene exchange within the crop/weed "complex depends on successful gene flow by pollen (Ell- strand and Hoffman 1990; Kareiva et al. 1991). Factors that increase the likelihood of gene exchange include self- incompatibility, high outcrossing rates, and generalist pol- lination mechanisms. Crop and weedy relatives must also have overlapping flowering periods, occur sympatrically, and be compatible (Keeler and Turner 1990). Many culti- vated plants have weedy relatives growing along the field