Within-species variation in grass weeds in Sweden : dormancy, herbicide response, genetic relationships

Sammanfattning: Variation within a weed species enables it to persist through varying conditions and is thus an important component of weediness. In this thesis, intra-specific variation in two agronomically important attributes - herbicide susceptibility and seed dormancy - are studied in Swedish Apera spica-venti (L.) Beauv. and Alopecurus myosuroides Huds., both serious annual weeds in winter cereals, and with many cases of herbicide resistance. Swedish Elymus repens (L.) Gould, a perennial, rhizomatous grass, is investigated for its genetic variability and variation in glyphosate response. The susceptibility to new and established herbicides in greenhouse studies in the two annual grasses ranged 0.5-4 orders of magnitude among populations, which mostly came from fields with no previous suspicion of resistance. The greatest variation was found in A. myosuroides in response to fenoxaprop-P-ethyl (an old herbicide in the sense that it had been used for a decade), with significant correlation with response to flupyrsulfuron-Na (a newly introduced herbicide). One fifth of the A. spica-venti populations were significantly less susceptible to isoproturon (old) than a susceptible reference population, but without correlation in response to sulfosulfuron (new). The findings indicate field evolved resistance at a scale of practical importance with possible cross-resistance to one new herbicide. Both A. spica-venti and A. myosuroides exhibited considerable primary and seasonally variable seed dormancy, with variation between seed collects. Light requirement in A. myosuroides was affected by stratification, and soil disturbance did not greatly enhance the low spring emergence occurring in this winter annual. Swedish E. repens seem to consist of one large, fairly homogeneous group, with a moderate geographic differentiation. Glyphosate dose-response patterns varied greatly between clones but were not linked to type of habitat or to genetic or geographic distance. The results are discussed in relation to methods for assessing herbicide susceptibility, for early detection of resistance, and for weed management.