| dc.description.abstract | Field studies were carried out in order to assess the populations of thrips on cotton leaves, by developing a sampling program to estimate the density with an acceptable level of precision. Thus, in a cotton field (1.1 ha) in central Greece, 45 leaves were collected (one leaf per cotton plant) at 10-day intervals from May to September, for four consecutive growing seasons (1995-98). Five species of thrips were found: Frankliniella intonsa was the most abundant species followed by Thrips angusticeps , Thrips tabaci , Frankliniella occidentalis and Aeolothrips intermedius . Although considerable differences were observed in the population fluctuations among species, the highest population densities, for all species found, were recorded in July and August. As indicated by Taylor's power law estimates, all species presented aggregated distribution among sampling units. As this type of spatial pattern indicates, the accuracy obtained in estimating mean population density increased with the increase of the mean. Furthermore, the increase of the mean caused an exponential decrease in sample size. However, the precision level is acceptable only in high mean values, while at the same time the benefit from an increase in sample size is of no practical value. | en |