The recent version of DIVERSI program package runs on IBM-PC compatible machines with DOS operating system or a buisness system which includes DOS client (such as WINDOWS NT, 2000, XP). All routines work with dbase data files. The program calculates diversity indices and their confidence intervals, the significance of diversity differences, index sensitivity and fit some abundance models to actual data. Quadratic entropy and members of the ESS and NESS similarity index family are also included. Hardware requirements, data file operations and examples are provided in a User's Guide. An example of calculating quadratic entropy is presented.
DIVERSI 2.2 is a new version of a program package for calculating diversity and related statistics. Shortcomings and errors in the earlier version have been corrected while the new version includes some new routines as well. This paper provides fundamental details together with numerical examples for the benefit of users.
Adult fly collections were made in May and August 2006, from the mud of a salty lake in the Kiskunság National Park, Hungary. We collected 21,600 individuals of 62 species in May and 18,400 individuals of 81 species in August, giving a total of 103 species (only 40 appearing in both seasons). Dominant species were different in the two collections. Five species are new for the Hungarian fauna. Diversity measured by several indices was higher in the May collection. We observed simultaneously a nearly lognormal distribution and a power-law like behaviour of the abundances.
It is an essential property of diversity indices that increases in the abundance or frequency of the most frequent species result in a decline in diversity, whereas increases in the abundance of the rarest species lead to an increase in diversity. At the same time, without resort to mathematical operations, it is difficult to determine the sign and measure of alteration in diversity when increasing an additional frequency while leaving all others unaltered. A more concrete task is to determine the index response to a partial alteration of fixed percentage in the frequencies of the multi-species community or collection. Plotting the observed responses or sensitivity values against the frequencies concerned makes possible a good overview of the sensitivity relations. The mathematical groundwork of sensitivity analysis with respect to diversity indices has already been elaborated. To date, however, the methodological possibilities engendered by such analyses have yet to be exploited. In the present work, sensitivity relations are discussed for apple-bait Drosophilidae collections and human faeces trap collections of flies inhabiting brook valleys in the low mountains of Hungary. Inspection of the results enables us to identify the range of frequencies at which significant increases or decreases in diversity will result. A relatively small increase of so-called nearly indifferent or quasineutral frequencies lying within that frequency range has a trivial influence on diversity values. While sensitivity is astonishingly sizeable with a few dominant case numbers, all other frequencies scarcely influence the index value.