Detached 3-week-old leaves of Agrostis palustris, Lolium perenne, Poa annua, and Poa pratensis were inoculated with conidial suspensions of two isolates of Colletotrichum graminicola obtained from A. palustris. Inoculated leaves were incubated at 23 degrees C under high relative humidity (>95%). The infection process was investigated by light microscopy from 2 to 168 h after inoculation (AI). Spore germination was observed within 2 h AI, appressoria within 6 h AI, and penetration pores within 8 h AI on all four hosts. Infection hyphae were observed inside epidermal cells within 24 h AI on all four hosts, but significantly greater infection was observed in A. palustris and P. annua than in L. perenne or P. pratensis at both 96 and 120 h AI. Acervuli appeared on leaves of A. palustris at 72 h AI and on L. perenne at 96 h AI but were not found on either P. annua or P. pratensis during the first 168 h AI. The infection process was similar to that reported for C. graminicola from other hosts; however, disease development of the two isolates of C. graminicola from A. palustris was faster or fungal growth more extensive on detached leaf tissue of A. palustris than on other turfgrass species tested.

Eighty-seven isolates of Colletotrichum graminicola, mostly from Agrostis palustris, were collected in grass fields, most of which were in Ontario, Canada. Specific primers were designed to amplify the mating-type (MAT) genes and, among 35 isolates tested, all yielded a band of the expected size for MAT2. For six isolates, the MAT2 PCR products were sequenced and found to be similar to that reported for MAT2 of C. graminicola from maize. Based on 119 polymorphic bands from 10 random amplified polymorphic DNA primers, analyses of genetic distances were found to generally cluster isolates by host and geographic origin. Among 42 isolates from a grass field in Ontario, significant spatial autocorrelation was found to occur within a 20-m distance, implying that this is the effective propagule dispersal distance. Although clonal propagation was observed in the 87 isolates with 67 unique genotypes, the extent of genetic variation in local populations implies some occurrence of sexual or asexual recombination.Key words: anthracnose, turfgrass, RAPD, MAT, reproduction, fungi.Résumé : Quatre-vingt-sept isolats de Colletotrichum graminicola, la plupart du temps à partir d'Agrostis palustris, ont été récoltés dans des champs d'herbes, principalement en Ontario, Canada. Des amorces spécifiques ont été contruites pour amplifier les gènes des différents types de croisement (MAT) et, parmi 35 isolats vérifiés, tous produisaient une bande dont la taille correspondait à celle de MAT2. Chez six isolats, les produits PCR MAT2 ont été séquencés et se sont révélés semblables à ceux connus pour MAT2 de C. graminicola du maïs. En se basant sur 119 bandes de polymorphisme obtenues avec 10 amorces d'ADN polymorphe amplifié au hasard, la mesure des distances parcourues par les gènes a permis de regrouper de façon générale les isolats selon l'hôte et la région géographique. Avec 42 isolats provenant d'un même champ, nous avons observé une autocorrélation spatiale significative dans les limites de 20 m, ce qui serait donc la distance maximale de la dispersion des propagules. Même si une propagation des clones a été observée chez les 87 isolats qui portaient 67 génotypes uniques, l'importance des variations génétiques dans les populations locales suggère la présence de phénomènes de recombinaison sexuée ou asexuée.

Hybrid identification of 16 sunflower hybrids was confirmed using simple-sequence repeat methodology. Of 20 specific simple-sequence repeat primers, 18 authenticated the purity of these hybrids; the remaining two specific primer pairs gave ambiguous DNA fragments. The results indicate that simple-sequence repeat analysis for the identification of hybrids derived from the crossing of different inbred sunflower lines can improve the accuracy of selection, save time and reduce cost.