Premium
ANALYSIS OF DATA FROM BIOLOGICAL SURVEYS OF STREAMS: DIVERSITY AND SAMPLE SIZE 1
Author(s) -
Kaesler Roger L.,
Herricks Edwin E.
Publication year - 1977
Publication title -
jawra journal of the american water resources association
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.957
H-Index - 105
eISSN - 1752-1688
pISSN - 1093-474X
DOI - 10.1111/j.1752-1688.1977.tb01996.x
Subject(s) - diversity index , estimator , sample (material) , statistics , diversity (politics) , sample size determination , mathematics , species diversity , structural equation modeling , ecology , statistical physics , physics , biology , thermodynamics , species richness , sociology , anthropology
Brillouin's equation (H) for species diversity from information theory is to be preferred for the purposes of applied ecology over the equation of Shannon (H′) or the more commonly used approximate equation (H″). By its use, the difficult problem of delimiting the extent of the community being sampled in a stream survey can be avoided. Moreover, Brillouin's equation gives the exact diversity of the fully censused collection, whreas Shannon's diversity can only be approximated with a biased estimator. If we regard a sample as a message from the environment to the ecologist, Brillouin's equation is the proper one for computing its diversity. The product moment correlation coefficient between Brillouin's H for the total number of individuals from a group of samples and for randomly chosen subsets of 100 individuals from each sample was nearly as high as the correlation between H and H″ based on the total samples. This indicates that small sample sizes may give a useful diversity index. Replicated subsamples show that much smaller samples than are normally used can discriminate between communities from polluted and unpolluted environments. The use of smaller samples should reduce the cost of stream surveys.