No abstract

No abstract

In studying an ecological community, it is common practise to sweep herbage and to estimate the number of insects per unit of area from the results. A more legitimate use for the collections is to compare those made at different times and so determine population trends. The problem of determining how far sweeping can represent the population of an area and how much sweeping should be taken to represent a given area is considered in this paper. The extent to which the distribution of animals varies from a random dispersion is considered. An estimate of the number of collections necessary to secure a specified degree of accuracy is investigated. Finally, a few observations are made on the technique of sweeping. DESCRIPTION OF THE SITUATIONS IN WHICH CoLLECTIONS WERE MADEMost of the field collections for the study were made in a sixty-acre tract of woodland (University Woods) five miles northeast of Urbana, Illinois. In these woods, the dominant trees are hard maple, elm, and white oak. The sweeps were made in situations where there was an extensive, apparently homogeneous, stretch of herbaceous plants. The predominant herbage of the area in which the sweeps were taken is H ydrophyllum sp., Lobelia siphilitica L., Carex sp., and also a few plants of Aster cordifolius L. METHODS OF CoLLECTIONSweeping was done with an " American " net. This had a handle one yard long, a hoop 12.8 inches in diameter, and a bag 20 inches deep. In collecting, a long semicircular swing was taken. In an examination the sweeps were found to average 100 inches and to have a standard deviation of 9.5 inches. The insects taken in each sweep were stored separately.Comparative collections were made with a sweep net and by means of the cylinder method. The former were made during the day. The latter collections were made, as nearly as possible, on the preceding or succeeding night.Cylinder samples were taken with metal drums 16.8 inches in diameter and one foot deep. In these, one end was closed except that it had a small hole in the centre on which there was a threaded cap. In sampling, these cylinders were placed near an undisturbed region which was to be examined.

No abstract

No abstract