the opportunity for evaluation of environmental issues independently of food safety concerns. Herbicide resis-Creeping bentgrass (Agrostis stolonifera L.) is a commercially tance in particular is a potentially very useful objective important turfgrass used principally on golf courses. Weed control is one of the major problems encountered in golf course maintenance, of genetic engineering, but also one that raises environlargely because the major weed, Poa annua L., is another grass species mental concerns due to potential transfer of herbicide with herbicide responses similar to creeping bentgrass. Development resistance transgenes to related species through interof creeping bentgrass cultivars expressing one of the herbicide resisspecific hybridization (Dale, 1992; Ellstrand, 2001). tance genes would provide an effective solution. The prospects of Such hybrids may have the potential to become weeds commercialization of transgenic cultivars of creeping bentgrass have that cannot be controlled by specific herbicides, thus raised questions about the potential for pollen-mediated gene flow impacting on the effectiveness of current weed control to related Agrostis spp. In a field study we have measured the fremethods. Interspecific hybridization between a number quency of interspecific hybridization between transgenic creeping benof crop species and their wild relatives is known to occur tgrass and four related species, A. canina L., A. castellana Boiss. and (Hancock et al., 1996;Ellstrand et al., 1999). Reut., A. gigantea Roth, and A. capillaris L. Interspecific transgenic hybrids were recovered between creeping bentgrass and A. capillarisWe are currently investigating the potential of genetic and A. castellana at frequencies of 0.044 and 0.0015%, respectively, engineering to augment breeding efforts in the developwhich were considerably lower than intraspecific transgenic progeny ment of improved cultivars of creeping bentgrass. recovery in the same experimental plots (0.631%). No interspecific Creeping bentgrass is an outstanding cool season turftransgenic hybrids were recovered with A. gigantea or A. canina.grass whose fine leaves, prostrate habit, and tolerance to very low cutting heights have made it the principal species used for putting greens and fairways on temper-Edgar and Forde, 1991). Davies (1953) reported on hy-Pathology, and P.R. Day, Biotechnology Center for Agriculture and brids resulting from bagged crosses between creeping the Environment, Rutgers Univ., New Brunswick, NJ 08901-8520; T.R. bentgrass and A. capillaris (referred to A. tenuis Sibth.)

Some Agrostis spp., such as A. stolonifera L. (creeping bentgrass), A. capillaris L. (colonial bentgrass), and A. canina L. (velvet bentgrass), are commercially important turfgrass species which are used extensively on golf courses. Development of improved cultivars of these species is the focus of many commercial and academic breeding programs. Interspecific hybridization between Agrostis spp. has not yet been utilized in cultivar development. Here we have investigated the frequency of interspecific hybridization between transgenic creeping bentgrass and four related Agrostis spp. using transmission of a herbicide resistance gene as a marker to identify the hybrids. Interspecific hybrids were recovered with all four Agrostis spp. used, although the frequency was lower than the frequency of selfing. The hybrids were found to be fertile. Our results suggest that interspecific hybridization may be a useful approach in future Agrostis breeding, but will benefit from a screening method to distinguish the hybrids from the selfs.

The taxonomic classification of the genus Agrostis is one of the most complicated of the grass genera. Classification based upon morphological and anatomical characters is difficult and complicated by the presence of intermediate forms and the misapplication of names. Determining ploidy levels of new germplasm can assist in species determination and is necessary before initiating breeding or genetics studies. The objectives of this study were to (i) evaluate the use of laser flow cytometry as a quick, reliable tool to determine ploidy level and aid in Agrostis species determination, and (ii) identify morphological characters associated with DNA content or ploidy level. The six Agrostis species evaluated were A. canina L. subsp. canina, A. canina L. subsp. montana (Hartm.) Hartm., A. palustris Huds. [= A. stolonifera var. palustris (Huds.) Farw.], A. tenuis Sibth. (= A. capillaris L.), A. castellana Boiss. & Reut., and A. alba L. Ploidy level was determined by flow cytometry and root tip chromosome counts. Plant height, panicle height, flag leaf length, flag leaf width, and highest internode length of mature field-grown spaced plants were measured. Significant differences in 2C DNA content were found between species (P < 0.01) differing in ploidy level. Flow cytometry was effective in differentiating between diploid, tetraploid, and hexaploid species. Chromosome numbers previously reported and those observed in this study were positively correlated with 2C nuclear DNA content (r = 0.98, P < 0.01). Flag leaf length was the only morphological measurement taken that was significantly positively correlated to DNA content (r = 0.98, P < 0.001). The results of this study indicate that laser flow cytometry is a quick, reliable tool to determine ploidy levels and infer certain species of AGROSTIS: This technique will aid breeders to quickly and accurately determine ploidy levels of new germplasm collections.

The taxonomic classification of the genus Agrostis is one of the most complicated of the grass genera. Classification based upon morphological and anatomical characters is difficult and complicated by the presence of intermediate forms and the misapplication of names. Determining ploidy levels of new germplasm can assist in species determination and is necessary before initiating breeding or genetics studies. The objectives of this study were to (i) evaluate the use of laser flow cytometry as a quick, reliable tool to determine ploidy level and aid in Agrostis species determination, and (ii) identify morphological characters associated with DNA content or ploidy level. The six Agrostis species evaluated were A. canina L. subsp. canina, A. canina L. subsp. montana (Hartm.) Hartm., A. palustris Huds. [= A. stolonifera var. palustris (Huds.) Farw.], A tenuis Sibth. (= A. capillaris L.), A. castellana Boiss. & Reut., and A alba L. Ploidy level was determined by flow cytometry and root tip chromosome counts. Plant height, panicle height, flag leaf length, flag leaf width, and highest internode length of mature field‐grown spaced plants were measured. Significant differences in 2C DNA content were found between species (P < 0.01) differing in ploidy level. Flow cytometry was effective in differentiating between diploid, tetraploid, and hexaploid species. Chromosome numbers previously reported and those observed in this study were positively correlated with 2C nuclear DNA content (r = 0.98, P < 0.01). Flag leaf length was the only morphological measurement taken that was significantly positively correlated to DNA content (r = 0.98, P < 0.001). The results of this study indicate that laser flow cytometry is a quick, reliable tool to determine ploidy levels and infer certain species of Agrostis This technique will aid breeders to quickly and accurately determine ploidy levels of new germplasm collections.

Creeping bentgrass (Agrostis stolonifera L.) is a commercially important turfgrass used principally on golf courses. Weed control is one of the major problems encountered in golf course maintenance, largely because the major weed, Poa annua L., is another grass species with herbicide responses similar to creeping bentgrass. Development of creeping bentgrass cultivars expressing one of the herbicide resistance genes would provide an effective solution. The prospects of commercialization of transgenic cultivars of creeping bentgrass have raised questions about the potential for pollen‐mediated gene flow to related Agrostis spp. In a field study we have measured the frequency of interspecific hybridization between transgenic creeping bentgrass and four related species, A canina L., A castellana Boiss. and Reut, A gigantea Roth, and A capillaris L. Interspecific transgenic hybrids were recovered between creeping bentgrass and A capillaris and A castellana at frequencies of 0.044 and 0.0015%, respectively, which were considerably lower than intraspecific transgenic progeny recovery in the same experimental plots (0.631%). No interspecific transgenic hybrids were recovered with A gigantea or A canina