Thirty morphological characters were measured for each of 63 rodent species. All characters were standardized to a mean of zero (O) and standard deviation equal to one (1). From this assemblage, subsets were chosen to represent ten (10) naturally occurring rodent communities of New Mexico and Arizona. Numerical taxonomy techniques of distance, correlation and principal components analysis were used to analyze the data.
Results reveal three distinct community 'types'. The presence or absence of morphologically divergent species is responsible for these differences. The appearance of Castor canadensis (the most morpho-logically divergent species) is dependent upon the addition of a new resource axis: water.
The distribution of values (distance to nearest neighbor) demonstrates that most community members are morphologically similar and that only a few species are morphologically distinct.
Principal components analysis reveals size as being by far the most important factor explaining the variation observed among the rodents. Other important factors are the height of the coronoid and condyloid processes, auditory bullae dimensions, and tail length.
Physiographic factors correlated with increasing average community distance values are: increasing altitude, decreasing annual average temperature, and increasing precipitation.
The morphological information gained is used as the basis for ecological predictions.
Level of Degree
UNM Biology Department
First Committee Member (Chair)
James S. Findley
Second Committee Member
Joe Scott Altenbach
Third Committee Member
J. David Ligon
May, Terrence Morgan. "Community Structure Of Ten Southwestern Rodent Faunas." (1977). https://digitalrepository.unm.edu/biol_etds/500