Short Description:

The computer program COMM has been developed to assist ecologists in the analyses of medium-sized species-station- tables commonly generated in ecological field investigations. The program features a menu and window environment, and - optionally - mouse support (COMMM.BAT). It runs on IBM PCs and compatibles with 640K of RAM memory. Supported output devices are an Epson-compatible lineprinter and a HP-compatible plotter.

Species-station-tables with up to 12,000 cells (species in rows, stations in columns, in addition species names and station labels) can be entered, loaded/saved from/to an ASCII- file, edited and printed.

For each station several diversity parameters can be computed (e.g. after Shannon-Weaver, Margalef, Simpson, Pielou, Heip). In addition, up to six external station-specific parameters (such as depths, temperatures, etc.) may be entered. The entire set of station parameters can be printed and/or saved to an ASCII-file for interfacing other programmes.

For each species a set of parameters can be computed (minimum, median, mean, and maximum of scores, constancy, dominance, Biological Index after McCloskey (1970)). The whole set of species parameters can be printed and/or saved to an ASCII- file for interfacing other programmes.

Species and/or stations may be sorted in ascending or descending order by species-specific and station-specific parameters. Data may be standardized and/or transformed (percentages, Z-values, ranks, logarithms, roots, abundance codes, dominance codes) prior to further analyses.

Cluster analyses may be performed to delimitate faunistic zones (Q-analysis: classification of stations) or species assemblages (R-analysis: classification of species). Eight resemblance measures are available (Jaccard, Sorenson, Simple Matching, Percentage Similarity, Canberra, Bray-Curtis, Euclidean Distance, Pearson correlation, Spearman rank correlation). The symmetric resemblance matrix can be printed and/or saved to an ASCII-file for further analyses like multidimensional scaling being performed by other programmes. For object classification, various agglomerative clustering methods are available (e.g. single linkage, complete linkage, average linkage, flexible strategy). A dendrogram as illustration of the resemblance structure can be displayed at the screen and may also be printed and/or plotted.

Subsets of the species and/or stations may be selected, either manually or automatically by various criteria. All analyses may be performed with any table subset. In case of station selection, two measures of fidelity (degree of association) are computed for each species to assist in the identification of indicator species of the station group selected.

Reference:

Piepenburg D, Piatkowski U (1992) A program for computer-aided analyses of ecological field data. CABIOS 8(6): 587-590.

Dieter Piepenburg
Institut fuer Polaroekologie
Universitaet Kiel
Wischhofstr. 1-3, Geb. 12
D-24148 Kiel Germany

ph: +49 431 72087 64
fax: +49 431 72087 20
e-mail:npf32@rz.uni-kiel.d400.de