*For offprint requests: fax: (864)656-0435, tel: (864)656-3060, email: kalaou@clemson.edu;

SUMMARY

A potential risk of releasing bioengineered crops into the environment is gene flow via pollen from crops to related wild species. The cultivated strawberry (Fragaria ananassa) and its wild relative (F. virginiana) represent a particularly suitable system for investigating transgene escape and its potential ecological consequences as the two species are interfertile, their flowering times overlap, and they share common pollinators. Moreover, strawberries, unlike many crops, are perennials and clonal, and both these traits increase the risks of hybridization, introgression and persistence. In the eastern US, F. virginiana may occur near cultivated strawberry farms; for example, in the piedmont of South Carolina, seven of 10 strawberry farms had populations of F. virginiana within 200 m of the cultivar. In an initial test of potential gene flow, potted plants of F. virginiana were introduced into a single strawberry farm. Based on RAPD markers, most seeds produced by these potted plants were hybrids (i.e., seeds were sired by F. ananassa pollen).

Key words: Strawberry, Fragaria, gene escape, hybridization, RAPD markers.

INTRODUCTION

With increasing production of transgenic crops on a large scale, the risk of disseminating transformed genes among wild relatives of crops becomes a major ecological concern. The opportunity for gene escape via hybridization depends upon the presence of wild relatives capable of crossing with the crop under natural conditions (Ellstrand, 1988; National Research Council, 1989). In fact, almost every cultivated crop is capable of hybridizing with at least one wild species and gene flow between cultivated crops and their wild relatives has been reported in many crop systems (Arias and Rieseberg, 1994; Ellstrand and Marshall, 1985; Santoni and Bervillé, 1992). Moreover, sexually compatible crops and related wild species may co-occur in agroecosystems (Dale, 1994).

The most likely avenue of gene escape from crops to related wild species is gene flow by pollen (Kareiva et al., 1994; Ellstrand, 1992). The rate of gene flow by pollen may be influenced by a number of ecological and genetic factors. For example, floral traits such as self-incompatibility, high outcrossing rates, and biotic pollination will increase the probability of gene escape whereas self-compatibility, high selfing rates, and limited pollen dispersal will substantially reduce the potential for gene flow. Many crop plants are outcrossing and insect pollinated (Fryxell, 1957), further enhancing the likelihood of gene escape. Other factors that may influence gene flow by pollen include pollinator type and behavior (Peakall, 1989), the spatial structure and density of plant populations (Handel, 1983), the degree of overlap in flower phenology, and the compatibility level between the cultivar and its wild relative.

The cultivated strawberry (Fragaria ananassa) and its wild relative (F. virginiana) represent a particularly suitable system for investigating the potential for gene escape and its possible environmental consequences. The two species, F. ananassa and F. virginiana, are sexually cross-compatible such that fertile hybrid offspring are produced (Luby et al., 1991; Stahler, 1990). Moreover, their flowering times overlap, they have floral characteristics that promote outcrossing, and they share the same major pollinators (Scott and Lawrence 1975). Consequently, there exists a high potential for interspecific pollen flow between these two closely related species. While several authors have observed natural populations of apparent hybrids, based on morphological traits (Stahler, 1990), the incidence, rate and consequences of gene flow between the cultivated strawberry and its wild relative have not been documented.

The cultivated strawberry is a hybrid derived from two New World species, F. chiloensis and F. virginiana, in the mid-18th Century (Darrow, 1966). All three species are octoploids (2n = 8X = 56) with wide geographic and climatic distributions (Hancock and Luby, 1993). In the northwestern US, the cultivar is reported to be a serious weed in some perennial and no-till systems (Rissler and Mellon, 1993). Furthermore, transgenic strawberries have been developed (James et al., 1993) and like many important crop plants, both the cultivar and its wild relative are insect-pollinated and primarily outcrossing (Antonelli et al., 1988; Bagnara and Vincent, 1988).

Our objectives for this preliminary study were to (1) survey representative strawberry farms in the piedmont of South Carolina for nearby populations of wild strawberry and, (2) use random amplified polymorphic DNA (RAPD) as molecular genetic markers to determine if there is a detectable incidence of hybridization between cultivated strawberry and its wild relative F. virginiana, by introducing potted F. virginiana into a strawberry farm.

MATERIALS AND METHODS

Farm survey. In spring 1994 and 1995, we surveyed a total of 10 strawberry farms in four counties within the piedmont of South Carolina, USA, for nearby populations of wild strawberry (F. virginiana). For each of these farms, the cultivars currently used and the number of years planted was recorded as was the distance, estimated population size and habitat type of nearby populations of wild strawberry.

Plant material. A total of 24 potted F. virginiana plants (mostly females) were interspersed with F. ananassa at the Wood's Farm, Oconee County, South Carolina to maximize the chance of hybridization between cultivars and wild plants. Potted plants were interspersed with the cultivar in different areas of the field such that six potted plants were interspersed with each of four F. ananassa cultivars (Cardinal, Titan, Atlas and Apollo). After 18 days, potted F. virginiana plants were returned to the greenhouse where fruits were allowed to mature (and unopened flowers were removed). Fruits matured after about four weeks after which seeds were removed and scarified to promote germination. For DNA isolation and RAPD analysis, young unexpanded leaf material was collected from seedlings of these potted plants and from the four cultivars.

DNA isolation and amplification. Extraction of DNA was carried out using a CTAB miniprep protocol (Davis et al. 1995). DNA amplifications for RAPD experiments were done in a volume of 25 µl containing 25 ng of genomic DNA as template, 0.2 µM of oligonucleotide primers (Operon Technologies, Almeda, CA), 0.75 unit of Taq DNA polymerase (Perkin-Elmer/Cetus) in reaction buffer [10 mM Tris-HCl (pH 8.3), 50 mM KCl and 3 mM MgCl₂] containing 200 M of dNTP's. The reaction mix was incubated in a thermal cycler (Perkin-Elmer/Cetus 480) using the following conditions: samples were preheated at 94C for 3 min. for DNA denaturation and subjected to 35 repeats of the following thermal cycle: 1 min. at 92C; 1 min. at 37C and 1 min. 30 sec at 72C, as in Sossey-Alaoui et al. (1994). Amplified fragments were size-fractionated by electrophoresis in 1.2% agarose gels in TAE 1X buffer (40 mM Tris, 5 mM NaOAc, 0.77 mM EDTA, pH 8 with glacial acetic acid). The amplified DNA was visualized by ethidium bromide staining, and gels were photographed to document fragment polymorphisms.

RESULTS

Field observations. Seven of the 10 strawberry farms examined had nearby wild strawberry (F. virginiana) populations . Wild strawberries were found at distances from 10 to 800 meters from the cultivar; undoubtedly, additional populations of the wild species occur within several km of each farm. Populations of the wild species ranged from about 50 to 500 plants, and habitats included roadside areas, abandoned farms, disturbed fields, and woodland borders. These sites demonstrate that populations of the wild species often occur near the cultivar in the piedmont region of South Carolina, and provide an opportunity to look for evidence of past introgression.

Preliminary experiments developing molecular genetic markers. To obtain an estimate of the level of detectable polymorphism using RAPD analysis, a set of ten-mer primers was screened on DNA samples from plants of four commercial cultivars and 11 plants of the wild species (F. virginiana), collected from an early successional field near the Simpson Experimental Station in Pickens County, South Carolina. We were interested in RAPD markers which distinguish cultivars from wild plants and in particular, those which are present in the cultivars and absent in wild plants. Nine primers out of ten tested (90%) showed polymorphism between F. ananassa cultivars and F. virginiana plants. Amplification with three primers detected polymorphism among the cultivars and six primers showed polymorphism within the wild F. virginiana plants.

RAPD amplification with primer L17 (Figure 1) generated a fragment of 1.4 kb which was present in all four cultivars and absent in all but two wild F. virginiana plants, designated F. v. #2 and F. v. #4. It is possible that these two plants are already hybrids as F. ananassa was once grown in a small experimental plot within 5 km of our source population of F. virginiana. This L17-1.4 kb fragment was useful in testing the progeny of other plants, for possible hybridization.

Among the F. virginiana potted plants placed within rows of F. ananassa cultivars, we analyzed seedlings from only three plants. The seedlings of these three plants were examined for evidence of hybrid seed formation, using the RAPD marker described above (e.g., L17-1.4 kb). Here we present the results obtained with only one F. virginiana plant and its seedlings. Marker L17-1.4 kb was present in all eight seedlings of F. virginiana #6 (Figure 2). This suggests that all these seedlings are hybrids.

DISCUSSION

Few studies have demonstrated the frequent occurrence of a related wild species in close proximity to a crop plant. The large number of "pick your own" strawberry farms in the US, and the presence of wild strawberry (F. virginiana) near these farms (in the southeastern US as shown in this study and in the northeastern US; Davis T., per. com.) are likely to increase the risk of transgene escape in strawberry . While previous studies on potential gene flow from crops to wild species have largely focused on annuals, wild strawberry is a perennial, and because individuals reproduce sexually by seed and vegetatively by stolons, the risk of transgene escape and persistence is further magnified. Moreover, F. virginiana is weedy in that it often occurs in disturbed areas and has a number of characteristics often associated with weeds (sensu Baker, 1965; 1974). For example, it has a short pre-reproductive period (less than one year), its flowers are self-compatible, individual plants may produce large numbers of seeds which may be dispersed long distances (e.g., by birds) and, once established, plants may spread vegetatively via vegetative propagation. Given its weedy predisposition, the transfer of certain transgenes into wild strawberry could potentially result in an adaptive change whereby it becomes a problem weed of agricultural and/or natural communities.

By placing potted F. virginiana plants within rows of cultivated F. ananassa and using RAPD markers specific to the cultivar, we were able to show that hybrid seeds are formed. These results demonstrate that there are no barriers to potential gene flow from cultivated to wild strawberry populations, and that the RAPD technique provides a reliable approach for detection of such events. F. virginiana, like many of the 18 species of wild strawberry, is dioecious (Dorrow, 1966; Scott and Lawrence, 1975), further increasing the risk of gene flow from the cultivar to wild populations. Furthermore, the frequency of such events is underestimated when a few dominant markers, such as RAPDs, are utilized since for heterozygous loci not all alleles are transmitted to each progeny. Therefore, to estimate the frequency of hybridization events, it is important to know the allelic composition of markers.

Based on this study, it is clear that amplification profiles from only a few different primers are necessary for discriminating cultivated strawberry from its wild relative F. virginiana and that numerous RAPD markers should be available for screening wild populations for previous introgression.

The results of this study suggest that hybridization can occur between cultivated strawberry and F. virginiana when potted plants of the latter species are introduced into a strawberry farm. Currently, we are estimating levels of past introgression of crop genes into the wild species by examining populations of F. virginiana that occur at varying distances from the cultivar. The results of this research should provide baseline information for developing guidelines for the management of transgenic strawberry.

REFERENCES

Antonelli AL, Mayer DF, Burgett DM, Sjulin T. 1988. Pollinating insects and strawberry yields in the Pacific Northwest. Amer. Bee Journal, 128:618- 620.

Arias D, Rieseberg LH. 1994. Gene flow between cultivated and wild sunflowerError! Bookmark not defined.s. Theoretical and Applied Genetics, 89:655-660.

Arnold ML. 1994. Natural hybridization and Louisiana Irises. Defining a major factor in plant evolution. BioScience, 44 (3):141-147.

Bagnara D, Vincent C. 1988. The role of insect pollination and plant genotype in strawberry fruit set and fertility. Journal of Horticultural Science, 63:69-79. related species: general considerations. Molecular Ecology 3:32-36.

Darrow GM. 1966. The strawberry: history, breeding and physiology. Holt, Rinehart and Winston, NY.

Davis TM, Yu H, Haigis KM, Mcgowan PJ. 1995. Template mixing: a method of inhancing detection and interpretation of codominant RAPD markers. Theoretical and Applied Genetics, 91:582-588.

Ellstrand NC. 1988. Pollen as a vehicle for the escape of engineered genes? Trends in Ecology and Evolution, 3:30-32.

Ellstrand NC. 1992. Gene flow by pollen: implications for plant conservation genetics. Oikos, 63:77-86.

Ellstrand NC, Marshall DL. 1985. Interpopulation gene flow by pollen in wild radish Raphanus sativus. The American Naturalist, 126:606-616.

Hancock JF, Luby JJ. 1993. Genetic resources at our doorstep: the wild strawberries. Bioscience, 43:141-147.

James DJ, Passey AJ, Webster AD, Barbara DJ, Dandekar AM, Uratsu SL, Viss P. 1993. Transgenic apples and strawberries: advances in transformation, introduction of genes for insect resistance and field studies of tissue cultured plants. Acta Horticulturae, 336:179-182.

Kareiva P, Morris W, Jacobi CM. 1994. Studying and managing the risk of cross-fertilization between transgenic crops and wild relatives. Molecular Ecology, 3:14-21.

Luby JJ, Hancock JF, Cameron JC. 1991. Expansion of the strawberry germplasm base in North America. In: Dale, A. and J. Luby (ed), The strawberry into the 21^st Century, 66-75. Timber Press, Portland, OR.

National Research Council (1989) Field testing genetically modified organisms: framework for decisions. National Academy Press, Washington, D.C.

Peakall R. 1989. A new technique for monitoring pollen flow in orchids. Oecologia, 79:361-365.

Rajapakse S, Belthoff LE, He G, Estager AE, Scorza R, Verdae I, Ballard RE, Baird WV, Callahan A, Monet R, Abbott AG. 1995. Genetic linkage mapping in peach using morphological, RFLP, and RAPD markers. Theoretical and Applied Genetics. 90:503-510.

Rissler J, Mellon M. 1993. Perils amidst the promise: ecological risks of transgenic crops in a global market. Union of Concerned Scientists, Cambridge, MA.

Santoni S, Bervillé A. 1992. Evidence for gene exchanges between sugar beet (Beta vulgaris L.) and wild beets: consequences for transgenic sugar beets. Plant Molecular Biology, 20:578-580.

Schmitt J. 1980. Pollinator foraging behavior and gene dispersal in Senecio. Evolution, 34:934-943.

Scott DH, Lawrence FJ. 1975. Strawberries. In: Advances in Fruit Breedings, 71-96. Janick, J. and J. N. Moor (ed). Purdue Univ. Press.

Sossey-Alaoui K, Serieys H, Lamber P, Zaharieva M, Belhassan E, Bervillé A. 1994. The use of RAPD markers to discriminate sections and species belonging to Helianthus. Biotechnology and Biotechnological Equipment, 7 (4):21-31.

Stahler MM. 1990. Variation for morphological and horticultural traits in Minnesota/Wisconsin Fragaria virginiana germplasm. Ph.D. thesis, University of Minnesota, St. Paul, MN.

Staudt G. 1962. Taxonomic studies in the genus Fragaria: typification of Fragaria species known at the time of Linnaeus. Canadian Journal of Botany, 40:869-886.

Figure 1. Amplification of DNAs from cultivated and wild strawberry species, with primer L17. RAPD fragment L17-1.4Kb is present in all five cultivars, and absent in all F. virginiana plants, except F. v. #2 and #4, which may be hybrids.

Figure 2. Amplification of DNAs from four strawberry cultivars, F. virginiana #6 and its seedlings with primers L17. Fragments L17-1.4Kb is amplified from all four cultivars and absent in F. virginiana #6. This marker is also present in all seedlings, which suggests that all seedlings of F. v. #6 are hybrids.