PiscesLogoSmallerStill Truncated log-normal

Top  Previous  Next

 

Many have claimed that the majority of communities display a log normal distribution. However, rarer species will not be fully represented in a finite sample so that in practice we lose the left-hand tail of the distribution. The veil line represents the distance from the right-hand edge of the distribution at which species becomes too rare to occur in the sample.

 

The output is displayed in a tabbed window. The O/E tab displays the observed and expected abundances of the species. These are arranged in abundance classes and the upper column gives the upper bound of each class. The Chart tab shows a plot of the observed and expected frequency distributions arranged by class. The observed are plotted as a histogram and the expected as a green line. The Param/Chi tab presents the estimated values for the following parameters:

the observed mean of the logged observations,

the observed variance of the logged observations,

the estimated mean of the log normal distribution,

the estimated variance of the log normal distribution,

the predicted total number of species in the community,

the observed number of species,

the species beyond the veil line, in other words the species number missing from the sample,

lambda, the diversity statistic (the estimated species number divided by the standard deviation).

 

truncln

 

The results of a Chi-Squared test of the observed and expected observations are also given. If the value of p is < 0.05 then the distributions are significantly different at the 5% level. The goodness of fit results are displayed in a rich text window. To move around the displayed text click on the text and then use the arrow keys on the keyboard. These results can be copied and pasted in to other applications by selecting Copy from the Edit menu in the normal Windows manner.

 

It has been suggested that SHE analysis is one of the best ways of deciding if a log series, log normal or broken stick model gives the best fit to the observed data.