Résumé :
|
Sequential decision plans based on aphid counts and binomial counts of infested plants (presence or absence of aphids) were developed to guide chemical control decisions for the potato aphid, Macrosiphum euphorbiae, on two growth stages of oilseed flax in western Canada. The plans were derived from studies of aphid dispersion among plants in field plots at two locations over 4 years, and verified in samples from 51 commercial fields, in Manitoba. The relationship between variance (s(2)) and mean aphid density ((x) over bar) per plant was log(e)s(2) = 0.790 +/- 0.050 + (1.649 +/- 0.031) (n = 69, r(2) = 0.98), for both crop growth stages. Neither sweep samples nor pan samples produced reliable estimates of the number of aphids per plant and, therefore, these sampling tools could not replace aphid counts on individual plants. Aphid counts and the binomial method gave similar control decisions with similar amounts of effort, but the aphid counting method required fewer plants to reach a decision. The same decisions were reached in 85-95% of fields by counting aphids on a minimum sample of 25 plants when the crop was in full bloom, or 20 plants at the green boil stage, as with samples of 50-100 plants.
|