Regional Indicator Species Selection

This analysis focused on selecting indicator species using the results of the regional fish assemblages. We utilized the FishTraits database of species life history, trophic ecology, habitat, and temperatrature tolerance traits to group similar species.

NMDS vs Hierarchical Clustering

To group species based on similarity between traits, we used both hierarchical clustering and nMDS. Both method use distance matrices that indicate similarities between observations. The distance matrix is then used to ordinate or cluster.

Results:

NMDS vs hierarchical clustering for Central Valley Species:

In the above, the height of the fusion, provided on the vertical axis, indicates the similarity between two observations. The higher the height of the fusion, the less similar the observations are.

NMDS vs hierarchical clustering for Great Basin Species:

NMDS vs hierarchical clustering for North Coast Species:

NMDS vs hierarchical clustering for South Coast Species:

NMDS vs hierarchical clustering for ALL Species:

This is a lot to take in…

To break this up a bit, drawing 12 groups:

Notes:

-The same distance matrices were used for both methods and euclidean distance was used to create them.

NMDS: used metaMDS from vegan package on distance matrix. Did not specify number of dimensions in final results, but experimented with changing the number of dimensions and it didn’t change the results significantly.

Hierarchical clustering:: used hclust on distance matrix. Experimented with different agglomeration methods (wards, complete, and average) but clustering results did not change significantly. Ended up using “complete” as the agglomeration method, which computes all pairwise dissimilarities between the elements in cluster 1 and the elements in cluster 2, and considers the largest value (i.e., maximum value) of these dissimilarities as the distance between the two clusters.