Island Biogeography unit area The mean area per tree is obtained by squaring the mean distance This method works well under certain restrictions Two methods are briefly outlined The Point-Centered Quarter Method A random starting point is established in the forest to be measured Four distances are measured from this point to the nearest tree in four quarters, forming each a 901 exclusion angle over the point The sampling point therefore is used like a rectangular cross, and the distance from the point to the nearest tree is measured in each quarter When repeated at 20 sampling points, established at random or along a compass line, one obtains 80 distances The mean distance squared then gives an estimate of the mean area per tree The number of trees per acre ( ¼ 4000 m2) or hectare (10,000 m2) is obtained by dividing such reference area by the mean area/tree The pointcentered quarter method is relatively simple to apply, and it compares well to the count-plot method However, two restrictions apply: the trees should be randomly distributed, and no tree should be measured twice The second requirement is usually easy to follow as it requires that the sampling points are far enough apart so that no tree is measured twice The Wandering Quarter Method This second recommended distance method is often less affected by the distribution pattern of trees (random versus clumped) as it traverses for the same number of distance measures through a larger area The method begins with establishing a random sampling point in the forest segment to be sampled A 901 exclusion angle is established in a certain compass direction, and the nearest tree to the sampling point is measured That nearest tree then becomes the next sampling point over which a 901 exclusion angle is established in the same compass direction From this tree the distance is measured to the next tree in that exclusion angle The procedure is continued for 20 distances in that same direction, moving from tree to tree and always using the same 901 exclusion angle This results in a wandering movement or zigzag line depending on where the next tree occurs within the 901 exclusion angle After 20 distances, the direction may be changed following a right angle for another 20 distances, then once more at a right angle in a parallel reverse direction to the first zigzag line, and finally, 20 distances in direction of the starting point In this case, the sampling also includes 80 distances, but the area traversed is larger than in the point-centered quarter method adjusted to a standard reference area, such as a hectare In the distance methods, the number of individuals is obtained from squaring the mean distance of all trees and dividing this measure into the reference area The number of individuals per species is then obtained from the proportion of species among the total number of individuals ‘‘Cover’’ as stem cover per species is obtained from totaling the individual basal area measurements in the sample of count plots or points and by relating this measure to the reference area For examples, see Mueller-Dombois and Ellenberg (1974) Another aspect relates to the tree-size stratification in countplot and distance sampling In addition to sampling mature trees/species, it is ecologically desirable in biodiversity inventories to also establish the density of seedlings and saplings First, mature trees have to be defined This is often done by setting height and/or diameter limits ‘‘Mature trees,’’ for example, may be all individuals m and taller, thereby forming a broad size class ‘‘Saplings’’ may be defined as individuals from 0.5 to m tall and ‘‘seedlings’’ as those from 0.1 to 0.5 m tall Smaller individuals may be considered as ‘‘germinants’’ (i.e not yet well established or ephemeral individuals) Smaller subplots are often required for seedling counts because of their smaller size and frequently greater density In the distance methods, usually only mature trees are measured because of the random pattern requirement for accuracy In such cases, seedlings and saplings may be counted in subplots of appropriate sizes at predetermined sampling points Measures Obtained in Count-Plot and Distance Methods The two types of methods provide for the same three parameters desired often in quantitative vegetation surveys These are frequency, density, and cover ‘‘Frequency’’ is simply the occurrence of a species in any number of plots in relation to the total number of plots in the sample In the distance methods, frequency is the occurrence of a species at a sample point out of the total number of sample points Frequency is a mixed measure of abundance and dispersion or distribution It is a relative measure as it depends on the plot size used in sampling ‘‘Density’’ is the actual count of individuals per species in the total number of count plots The sample plot area is then 401 Figure Comparative sizes of sample plots Reproduced from Mueller-Dombois D and Ellenberg H (1974) Aims and Methods of Vegetation Ecology New York: John Wiley & Sons