. The Canadian field-naturalist. 1985 Dunn, Hussell and Siderius: The Great Blue Heron in Ontario 67 designed to locate new colonies turned up no new colonies in the southern or eastern census areas, but nine in the northern census area. Those results sug- gested that our inventory for the southern and eastern areas was essentially complete, but that the northern census area is likely to contain colonies as yet unknown to us. That area is relatively small, however, and has small colonies. If we guess arbitrarily that as many as % of the colonies in that area are still unknown to us, and that t
. The Canadian field-naturalist. 1985 Dunn, Hussell and Siderius: The Great Blue Heron in Ontario 67 designed to locate new colonies turned up no new colonies in the southern or eastern census areas, but nine in the northern census area. Those results sug- gested that our inventory for the southern and eastern areas was essentially complete, but that the northern census area is likely to contain colonies as yet unknown to us. That area is relatively small, however, and has small colonies. If we guess arbitrarily that as many as % of the colonies in that area are still unknown to us, and that they have the same average size as the known colonies, then the total number of nesting pairs in the three census areas would rise by only 9%. Most probably, underestimation from that source is less than 5%. We next considered the accuracy of the counts. Our only way of assessing the accuracy of nest counting in a colony was to compare counts made by different parties in the same year (Figure 3). Errors in nest counts within colonies may result from (a) omission of some of the nest trees from the count or, (b) miscounts of the number of nests in a tree, or both. Negative errors of both types are more likely to occur than positive errors because it is more likely that nest trees and nests will be missed than nonexistent ones counted. Our recommended procedure of tagging each tree as the nests in it are counted helps the cen- suser avoid type (a) errors, but in some colonies nest- ing trees are quite dispersed and our records of the numbers of nest trees in duplicate counts indicated that nest trees were sometimes missed. Additional data are needed to determine whether error variance changes with colony size; if we assume that it does not, we can pool the data from all dupli- cate counts to determine the average error. Based on the data in Figure 3, low counts averaged 71 % of the high counts. Stated another way, this means that the total of the low counts would have to be increas
Size: 1607px × 1555px
Photo credit: © Book Worm / Alamy / Afripics
License: Licensed
Model Released: No
Keywords: ., bookauthorottawafieldnaturalistsclub, bookcentury1900, bookcolle
