Communities: Use of cluster and ordination techniques

Background:

Community ecologists use cluster and ordination techniques to visualize relationships amongst different communities. Both are multivariate techniques, but the way in which an ecologist analyzes the outcomes are slightly different as the cluster analysis results in associations of communities in a classificatory sense (similarities) and ordination relates communites on the basis of their dissimilarities. Basic to both techniques is the construction of a mattrix that includes all the communities as a two dimensional array. For clustering, the array is obtained by doing a distance measure (euclidian, mean euclidian, centroid, etc.). For ordination the array is obtained by measuring the dissimilarity amongst the communities.

The Assignment:

Half of the class will do a cluster diagram, the other half an ordination diagram. Check the list below as to your assignment:

Chan Sigmund cluster
Currier Seth R. cluster
Figueroa Benjamin Harriman cluster
Franz, Carabeth cluster
Gibson Christopher Warren cluster
Goulette Jason Bartholomew cluster
Hilton Alicia Caroline cluster
Mello Sarah Amy cluster
Mickiewicz Jeff M. ordination
Olisky Jason R.ordination
Pesta Meredith Ellen ordination
Ring Jennifer M. ordination
Scagliarini Matthew Roger ordination
Scrafford Elizabeth Jane ordination
Solberg Erik Otis ordination
Stewart Elizabeth Ann ordination
Turner Dean Alan ordination

The data:

You are to use the data you collected from Mount Morgan, Pemi Riverside Park, and past classes. For Mount Morgan you have four different elevations (thus perhaps four different communities). You can use either the data collected by point quarter techniques or by quadrat techniques. The Pemi Riverside Park data is located at the following location:

Pemi Riverside Park Data

The Fox Park Data is located at

http://oz.plymouth.edu/~natsci/studiescenter/Fox_Park/foxpark.html

You can use either the 1996 or 1998 studies.

The Langdon Park data is located at:

http://oz.plymouth.edu/~natsci/studiescenter/Langdon_Park/langdon.html

Again, you can use either the 1996 or 1998 data.

Use any of the abundance measures present in the data (abundance, density, coverage). You must use the same measure for all communities, i.e., you can't switch from density to coverage between the communities you are comparing. For the cluster technique, I suggest you use the euclidian distance techniques we did a while back. For the ordination, use the Bray-Curtis similarity measure.

What you need to supply me:

1. A matrix that shows the species in the far left column and seven columns to the right that have the orignal abundance values (the relevant example here is the southern Wisconsin forest data I sent you via e-mail)
2. The distance matrix or the dissimilarity matrix.
3. The cluster diagram or the ordination diagram
4. A discussion section where you try to relate the outcomes to ecological factors associated with the communites. In this case, one main feature might be the elevation, but other features might include differences in soil, differences in aspect (east facing slopes versus south facing slopes), soil moisture content, etc.

Note: All the work could be done by computer, but it is not a requirement of this exercise to do so. Excel actually has a number of functions that allow you to work with data in a matrix, but the learniing curve to do would far exceed the gain. Likewise, you could graph the cluster diagram or ordination diagram in Excel, but it is probably just as easy to hand draw the diagrams on graph paper.