Studies on the Linetransect Census Method of the Woodland Bird Populations

IX. A Comparison with the Plot Census Method

Abstract

1. Although the line-transect census method is very simple and convenient for estimating bird densities, it has certain deficiencies; for example in over represents edge species, suffers from disturbance from the noise of walking and is inapplicability for small patches of forest. For such cases, it is desirable to use the plot census (point count) method. But up to the present, an estimation manner of a true density by this method has not been developed. The true density of birds can not be estimated by the Variable circular-plot method or Fixed circular-plot method, because these methods don't consider the potential effect of the conspicuousness of particular species. Some new methods for estimating the true density of bird communities or respective species by the plot census method compared with the line-transect census method were examined.
2. The recorded number of individual birds by the plot census method changed with the time of day, but the number did not show larger differences between 4:00 to 10:00 in any minutes spent on the plot.
3. The recorded number of individual birds showed a definite pattern of increase according with the time (minutes) extension staying on the plot. But the number does not increase so much comparing with the time increase on the plot.
4. The recorded number of individual birds showed no significant difference between trained observers.
5. The recorded number (P) of individual birds by the plot census method had a good correlation with the number recorded by the line-transect census in the same wood. Also, there is a correlation between the recorded number of individuals (n) of total bird species by the line-transect census (observing radius 50m, walking speed 1.5km per hour, one hour recording covering 15ha from 2 hours after sunrise, under good weather conditions) and the absolute density (D) of total bird species of that area. Then, D can be estimated from P. For example D=17.89 P₁₀+4.29. Where P₁₀ means ten minutes sample of the plot census with 50m observing radius.
6. The possibility of estimating the specific density (D_i) of i species by the plot census was suggested. For example, D_i=(7.31P_10i+6.29)•200/E_i. Where P_10i means the ten minutes sample in the plot census for i species. E_i means the census effectivity, the ratio (%) of the number of i species recorded in the standard line-transect census to the number of territories in that area. But, it was considered that a better way to estimate the specific density of i species by the plot census would be, from the first, to know the plot census effectivity for respective species.
7. The number of sample plots in the plot census needed to estimate the density of total bird species in 20-30ha woods with 95% confidence limit and 10% ±sample error against the average was regarded as 10-15. The desirable time to be spent in the plot was considered to be 2-5 minutes.