On the Number of Individuals Per Occupation in a Human Society
1964
Clark, Philip J. | Eckstrom, Philip T. | Linden, Larry C.
The number of individuals per occupation in the city of Ann Arbor fits the logarithmic series at least as well as do the data from most plant and animal communities. It is likely that departures from the logarithmic series in samples from natural communities are largely attributable to nonrandomness in the sampling procedure. In view of the relative ease of working with the logarithmic series it may continue to serve a useful purpose in community ecology even in competition with the truncated log—normal distribution, provided the sampling is done in a sufficiently random manner. The hypothesis is proposed that the reason the number of individuals per species in random samples from natural communities tends to follow the logarithmic series and the truncated log—normal distribution is that the factors affecting abundance tend to combine multiplicatively. To test this hypothesis a synthetic community consisting of 220 species and 26,000 individuals was created by making the abundance of each species proportional to the product of 10 random normal numbers having a mean of 2 and a standard deviation of 1. The synthetic community of thus formed does, in fact, follow the log—normal distribution approximately. The number of individuals per species in a random sample of 1,083 adults from the synthetic community followed the logarithmic series almost perfectly. It is concluded that a multiplicative combination of factors affecting abundance is capable of producing both the log—normal distribution and the logarithmic series.
显示更多 [+] 显示较少 [-]