Aiming to estimate the number of samples, and the diameter of soil sampler ideal for red rice infestation level, a irrigated rice field survey has been conducted in Santa Maria, Rio Grande do Sul (RS), Brazil, The experiment was conducted in three areas with 400m² (20x20m) with different red rice infestation levels considering low, medium and high infestation levels with 71, 282 and 498 red rice seeds per m², respectively, sample with 5, 10 and 15cm core diameter. The results showed that with larger core diameter the samples coefficient variation among samples was smaller. The core of 10 cm is more functional, been necessary 280, 55 and 31 (D=20%), 33, 9 and 6 (D=60%) and 13, 4 and 3 (D=100%) soil samples for the infestation levels 71, 282 and 498 red rice seeds per m² respectively. The number of soil samples necessary to estimate the red rice seed bank is variable with the infestation level and the desirable precision.
confidence interval; infestation level; Oryza sativa; phytosociological; variability
