Skip to main content
ARS Home » Research » Publications at this Location » Publication #307385

Title: Modeling feral alfalfa (Medicago sativa subsp. sativa L.) occurrence using topographical and environmental variables

Author
item KESOJU, SANDYA - Washington State University
item Greene, Stephanie
item Martin, Ruth
item Boydston, Rick

Submitted to: Meeting Abstract
Publication Type: Abstract Only
Publication Acceptance Date: 5/5/2014
Publication Date: 7/9/2014
Citation: Kesoju, S., Greene, S.L., Martin, R.C., Boydston, R.A. 2014. Modeling feral alfalfa (Medicago sativa subsp. sativa L.) occurrence using topographical and environmental variables. Meeting Abstract. North American Alfalfa Improvement Conference, July 9-10, 2014, Lethbridge, Canada..

Interpretive Summary: Alfalfa is cross pollinated by bees and can establish along roadsides and weedy area. Because of this, there is concern that feral plants can serve as reservoirs and conduits in the movement of genetically engineered traits. The objective of this study was to survey feral alfalfa in alfalfa seed production areas in California, Idaho and Washington to see if we could find any plants that tested positive for the Round Up Ready® (RRA) trait. Because the survey occurred 4-6 months after the February 2011 deregulation of RRA, occurrence of the GE trait in feral plants could be attributed to escape during the first deregulation (2006-2007). Feral-RRA populations were detected at each study location which suggested that the RRA transgene can persist in the environment. In California and Idaho, out of 812 and 700 sites visited, 130 (Fresno) and 87 (Canyon) sites had feral plants, and out of those sites, 53 (Fresno) and 12 (Canyon) sites were positive for the transgene. In Washington, although 127 sites had feral plants out of 786 sites visited, only 10 sites were positive. Seed- mediated gene flow may also be significant since feral-RRA plants occurred along arterial roads. Feral-RRA sites were also located away from seed production areas, suggesting hay production may be a source of feral-RRA escapes. The logistic regression model showed no significant relationship between elevation, slope, and feral plant occurrence, but a weak relationship between aspect and feral plant occurrence in Idaho.

Technical Abstract: Because alfalfa is a perennial species cross pollinated by bees and can establish along roadsides and ruderal areas, there is concern that feral plants can serve as reservoirs and conduits for transgenic genes. The objective of this study was to survey feral alfalfa in alfalfa seed production areas in Fresno Co., CA, Canyon Co., ID, and Walla Walla Co., and test for the occurrence of the Round Up Ready® (RRA) transgene. Because the survey occurred 4-6 months after the February 2011 deregulation of RRA, occurrence of the transgene in feral plants could be attributed to escape during the first deregulation (2006-2007). Feral plant occurrence was modeled using topographical, environmental, and crop occurrence variables. Survey sites were randomly selected along rural roads using the Spatially Balanced Sample Design tool from ARC GIS 10. Leaf and seed samples were collected from alfalfa plants growing at the site. Leaf and seed were tested for the RRA transgene with RUR immunotest strips and positive tests confirmed with PCR. At each site data on topography, climate, cropping pattern, roadside weed management and habitat were obtained. Chi square tests and logistic regression models were used to test the influence of variables on feral plant occurrence. Feral-RRA populations were detected at each study location which suggested that the RRA transgene can persist in the environment. In Fresno and Canyon Counties, out of 812 and 700 sites visited, 130 (Fresno) and 87 (Canyon) sites had feral plants, and out of those sites, 53 (Fresno) and 12 (Canyon) sites were positives. In Walla Walla Co., although 127 sites had feral plants out of 786 sites visited, only 10 sites were positives. Frequency of feral-RRA plants was higher in Fresno Co. followed by Canyon Co., but lower in Walla Walla Co. Seed- mediated gene flow may also be significant since feral-RRA plants occurred along arterial roads. Feral-RRA sites were also located away from seed production areas, suggesting hay production may be a source of feral-RRA escapes. The logistic regression model showed no significant relationship between elevation, slope, and feral plant occurrence, but a weak relationship between aspect and feral plant occurrence in Canyon Co.