How to Perform the Dendogram Analysis.

1) Begin by entering the number of species for the analysis. For the Falls of the Ohio data set, you should enter "10" for your top 10 species. Hit the ENTER key after entering the number. The program will accept numbers from 3 to 10.

2) Now enter a descriptive name for each of the species as shown above. You should start with the most common species and work your way down to the least common. In case of ties, choose the one you think might be more important to the ecosystem as the first entry (maybe based on size or because you know they're common at many of the sites).

3) Next enter the number of sites you explored (five at the falls and two road cuts=7). The program will accept numbers between three and 10. Follow with a descriptive name for each as shown above.

4) Press the ENTER VALUES key when you have finished entering the names for the sites. Then enter the requested information. In the above figure, I saw 22 branching corals at site A. After hitting the ENTER key on your keyboard, it will request the next count. This will continue through all the sites and species until everyone is accounted for. Enter 0 if you didn't see any of a particular species at the requested site. When you have finished, press the RUN ANALYSIS button.

5) A dendogram similar to the above figure will be produced (yours may be very different- I made the data up). How is this interpreted?

• Based on the numbers and kinds of species found, sites D and E are most alike (they are joined by the shortest bracket). This makes sense since both sites are approximately the same age (both are at the river's edge)

• Road cut B has the next level of similarity and shares that with the D-E complex. Cut A is then tied to this complex. From this we see that the road cuts are similar to the sites located near the river, so they must be approximately the same age.

• Sites A and B are similar.

• Site C is way off in left field by itself.

6) If you click the SHOW DATA button you can see your data entries along with some statistics about the "distances" or similarities between nodes (scroll down). If you click the CLIPBOARD button you can copy this data into Excel or a word processor. Click the HIDE DATA button to return to the dendogram. Press the HIDE GRAPH button to return to the start window of the program.

7) We know from the dendogram that sites A and B are similar, that site C stands alone and that the road cuts are most like the C-D complex. This was based on 10 species. However, it might be that our 10th species (in this case crinoids) are not important to discriminate among sites. We can test this by pressing the LOAD LAST DATA button at the bottom of the start window. Then change the number of variables from 10 to 9. Now the program will only pay attention to the top 9 species when you press the RUN ANALYSIS button.

8 The above image shows the results as we take away one species after another. The first (V=10) represents our original data with all 10 species used in the calculations. When we drop one (the crinoids) for V=9, there is some minimal shifting, but the relationship between sites D and E or between the two road cuts doesn't really change (the order from top to bottom doesn't matter; just the branching and bracketing). As more species are removed (fungus coral for V=8 and cup corals for V=7) we still don't get any significant changes in the arrangement of the dendogram. Not until we remove the large brachiopods (V=6) does the order begin to get strange (now site D has more in common with road cut B than with site E). We therefore conclude that the first 7 species (including the large brachiopods) are necessary for describing the differences among these sites.