Ascospore gradients of Gibberella zeae from overwintered inoculum in wheat fields

Ascospore gradients of Gibberella zeae, causal agent of fusarium head blight of wheat, were measured over a 3-month period in 1999 at two sites in Quebec, Canada: Macdonald Campus (MAC), McGill University, Sainte-Anne-de-Bellevue, and L'Acadie Experimental Farm (LAC), Agriculture and Agri-food Canada, L'Acadie. Spores were sampled daily with rotorod-type samplers along transects at intervals of 10 m up to 50 m away from plots of wheat inoculated at anthesis the previous summer and allowed to overwinter. At MAC, ascospore concentrations > 100 m–3 were detected on 38 nights from 1 June to 20 August, in 16 separate release events that lasted from 1 to 3 nights. Most of the ascospore releases occurred 1–6 days after an initiating rainfall, and the ascospore concentration decreased over time in evenings subsequent to the initiating rainfall. Spore gradients were modeled with the inverse-power model of Gregory, the negative exponential model of Kiyosawa and Shiyomi, and the general model of Lambert. The general model outperformed the other two models at both sites, with an adjusted R 2 = 0.88 and 0.79 for MAC and LAC, respectively. The inverse-power model outperformed the negative exponential model at MAC, and vice versa at LAC. Based on gradients that fit the negative exponential model, ascospore concentrations declined by 50% within 18 m of the source, and by 90% within 60 m of the source at MAC. This is the first detailed study of ascospore gradients of G. zeae from naturally overwintered inoculum on a field scale. Although ascospore concentrations declined considerably within the 50-m transect, long-distance transport of spores beyond the field may occur.