Density, speies richness and diversity of breeding bird communities in nineteen types of vegetation in Japanese woodland were estimated based on 190 censuses. Two types of density value were calculated. The first, relative density (
n), which is the total number of individuals of all species recorded during a one hour line transect census made under standard conditions: viz, conducted around 150 minutes after sunrise, 50 meters observing radius, 1.5km walking speed per hour and in favourable weather. The second, is absolute density per 15 hectares (
D), which was estimated from the regression formula of correlation between and the absolute density of study the areas. Two types of species richness value were also calculated. One, the number of species (
S) recorded during a one hour line transect census made under standard conditions, was estimated using the correlation between actual
S, in several study areas, and values gained by Rarefaction method (James & Wamer 1982). The other is the number of species (S
T) which is considered to be actually dwelling in 15 hectares, and was estimated using correlations between actual
ST and
S in several study areas. Moreover, the maximum number of species (
S'
max) thought to be dwelling in each type of vegetation were estimated, and the number of species dwelling in different areas were estimated using the law of truncated negative binominal distribution. Two types of bird species diversity (BSD) value were calculated using the Shannon-Weaver function: viz. BSD=-PiΣlog
2 P
i. One is rude BSD (
H') which was calculated from rude census data in each vegetation type. The other is the estimated value of BSD (
H') which will be thought to be attained during a one hour line transect census made under the standard conditions, and was calculated by
J'•log
2 S, where
J' is a mean of evenness derived from rude census data in each vegetation type.
There was a positive correlation between
D and
ST. Also,
D and
H', showed a positive correlation with FLD (forest layers diversity), which is calculated by (4Σa
i_??_a
j=1a
i•a
j)
1/2, where a
i and a
j take relative values checked by foliage cover in four layers woodlot: viz, no cover O, sparse 1/3, medium 2/3, dense 3/3. BSD was considered to be determined primarily by FLD or FHD (foliage height diversity; Macarthur & MacArthur 1961) and the degree of the mixture of broad-leaved and needle-leaed trees in the crown layer. Moreover, it was thought that conditions of geographical or topographical loci and continuity or patchiness of vegetation would be important factors affecting BSD. Finally, methods for utilizing the data in this report, and for collecting or standardizing techniques of collecting data for the purposes of comparison, were suggested.
View full abstract