ABSTRACT
Depending on the objectives of sampling, density is considered in one of two ways, absolute and relative. Southwood grouped Morris' estimate types into three convenient categories, absolute and related estimates, relative estimates, and population indices. Absolute estimates present information on numbers of arthropods usually per unit of land surface area in the habitat or provide data that can be easily converted to such. Relative estimates differ greatly from absolute and related estimates because they have no direct relationship to land surface area. A statistical transformation can be achieved by making paired relative and absolute estimates from a series of populations, followed by analyses that regress the relative estimates on absolute. Experimental populations represent one of the few methods available to substantiate efficiency of methods designed to give absolute estimates. The sampling-unit concept is most easily explained with absolute estimates, where total counts are taken from a unit-area of land surface. Consequently, sampling units and samples are taken within sampling universe.
