ISSN 10630740, Russian Journal of Marine Biology, 2010, Vol. 36, No. 3, pp. 215–222. © Pleiades Publishing, Ltd., 2010.
Original Russian Text © V.V. Sukhanov, V.E. Zhukov, 2010, published in Biologiya Morya.
Most indices and comparative characteristics used
to describe communities involve the parameters of
their species diversity in a certain form. The major tar
get of our study was to compare methods used to deter
mine the degree of our knowledge about the species
composition in the community. This characteristic of
the community is among the most important ones, as
only biocenoses that are wellstudied can be com
The degree of our knowledge about a particular
community could be considered as the ratio of number
of actually distinguished species to the greatest
expected number. However, the maximum number of
species could hardly be determined directly in natural
communities. Usually, this would be an estimated
value that represents a parameter of a mathematical
model describing the species structure of the commu
We have taken the lognormal distribution as a basic
model describing species distribution in a community
using their diversity. This distribution is a direct off
spring of the widely known normal (Gaussian) distri
bution. The only difference is that we use not the spe
cies diversity itself, but its logarithm. Due to the rela
tively great number of parameters, the lognormal
distribution appears to be flexible enough to ade
quately describe almost any factual material. Other
The article was translated by the authors.
simpler models for the calculation of species distribu
tion by their abundance in a community that are pop
ular among ecologists could be considered as particu
lar asymptotic forms of the lognormal model [7, 16].
The creator of this model has already emphasized a
characteristic feature of this distribution, namely, that
the number of the rarest species in the collection could
not be smaller than one . On the other hand, the
classical lognormal distribution is determined along
the entire positive semiaxis, from zero to infinity.
Thus, the lognormal Preston’s distribution appears
truncated at an argument equal to 1.
The estimation of parameters for a truncated distri
bution is always a nontrivial problem. The best known
solution was proposed by Pearson (cited after ).
Since that time new methods to resolve this problem
have been elaborated. A comparative analysis of these
methods is the target of our study.
It would be worthwhile to perform such an analysis
using comprehensive factual materials. Such materials
were obtained in the course of yearround observa
tions of a community of coastal macrophytes.
MATERIAL AND METHODS
Macrophytes were collected in Vityaz’ Bay in the
Sea of Japan, from June 1979 to August 1980, in three
geomorphologically different intertidal and subtidal
METHODS OF INVESTIGATIONS
Biology of Communities
Estimation of Species Richness of a Community by its Species
Structure, as Illustrated by the Macrophytes of Vityaz Bay
in the Sea of Japan
V. V. Sukhanov and V. E. Zhukov
Institute of Marine Biology, Far East Division, Russian Academy of Sciences, Vladivostok, 690041 Russia
Received September 15, 2009
—Five methods of computing species richness in community are discussed using the example of a
wellstudied macrophytic taxocoen: (1) fitting an ogive for lognormal distribution; (2) model fitting of a rank
curve for this distribution; (3) iterative computation of statistical moments using Hald’s method; (4) compu
tation of these moments by explicit functions; and (5) model fitting of a “species–area” curve. According to
the repeatability of results and simplicity of computing, the best method is the computation of statistical
moments by explicit functions. The degree of our knowledge about a particular community could be consid
ered as the ratio of the number of actually distinguished species to the greatest expected number.
lognormal distribution, calculation of parameters, species biomass, species–area.