Effect of border rows and trap/donor ratios on pollen-mediated gene movement

One of the concerns associated with the field testing and agronomic-scale release of transgenic crops is the potential for the pollen-mediated escape of engineered genes into naturally occurring populations of wild relatives. While border rows have been used frequently to restrict the pollen-mediated escape of engineered genes from field trials, the efficacy of this approach has been little studied. To test the effectiveness of border plantings, isogenic lines of cucumber (Cucumis sativus L.) differing for the seedling marker trait blunt leaf apex (bla) were planted in four designs: 1 m² of wild-type donor plot surrounded by a 399-m² border of bla recipients; 1 m² of donor plot/99 m² border; 4 m² of donor plot/96 m² border; and 1-m² of donor plot with no border. Each planting was encircled by eight 1.4-m² satellite plots 50 m from the plot center. Progeny of plants from the satellites and borders were screened to determine the percentage of gene movement as measured by the occurrence of the dominant phenotype. Gene movement within the plot borders followed a leptokurtic distribution; there was greater movement from the 4-m² donor plot than from the 1-m² plot. Long-distance movement to the satellites significantly increased as the trap-to-donor ratio decreased. Although movement to individual satellites was generally consistent within a treatment, there was one instance of unusually high outcrossing to a single satellite, indicating that the effectiveness of borders was influenced by both the relative numbers of donor plants and by environmental variables.