The table handed out with this exercise contains faunal data from site 1Pi61, a Woodland village on the Tombigbee River in west-central Alabama. Three distinct components can be recognized at this site: Late Miller II (AD 400-550), Early Miller III (AD 550-900), and Late Miller III (AD 900-1000). Do the data from this site support the hypothesis that there was an increase in the diversity (richness) of species exploited through time? (In considering this question, you may wish to discard the taxa that are not identified at least to the level of genus.)
Using Kintigh's diversity program, you should test the hypothesis in
two ways: (1) by generating a single expected diversity curve for all assemblages
combined, and (2) by "standardizing" the diversity to a constant sample
size for each phase separately (Baxter calls this technique “rarefaction”).
Which do you think is the more appropriate method in this case?
Here are some helpful hints:
1. Kintigh's diversity programs reside in the \TFQA folder on ASNT1\...\SHARE, and are also available on the course web site. The programs may be invoked from your own folder by opening a DOS window and issuing the commands DIVERS (for the diversity simulation) and DIVPLT (for the plotting program that displays the results).
2. In order to get a printed copy of the diversity graphs, answer "Y" when DIVPLT asks whether to produce plotter output, specify "prn" as the port to which the plotter is connected, and type "L" to select the laser printer with HPGL. Alternatively, you can specify “file” and then enter a file name with the extension “HPG” or “HPGL”; the resulting file can then be opened in either View Companion or WordPerfect and printed.
3. I suggest that you play with DIVERS and DIVPLT to familiarize yourself with the various options they offer before generating lots of printed output. Also note that it is easy to draw the expected diversity curve by hand or with SYSTAT, using the information in the .LST file.
4. Although the graphs drawn by the program are pretty and have
some heuristic value, the histograms in the .LST file are often far more
useful in assessing the relative diversity of an assemblage. Can
you see why?
Handouts, datasets, and program files for this exercise (right-click to download):