Sixteen pea taxa representing both wild populations and cultivated accessions of the genus Pisum are scored for morphological characters, allozymes, RAPDs, ISSRs and ITS sequences. Tests of skewness and partition homogeneity indicate that each of these individual data sets contains strong phylogenetic signal and sufficient congruence to support their combination in a single data set. Using both the maximum parsimony and bootstrap methods from PAUP, the small number of morphological characters precisely organizes the species into traditional taxonomic groupings, perhaps in part reflecting the role morphology plays historically in pea classification. Individual and combined molecular data sets support several of these same groupings. Decay indices and other branch values for the data-rich ISSR and RAPD trees reveal particularly strong support for three nodes that: 1) separate P. fulvum from the P. humile, P. elatius and P. sativum ingroup; 2) establish the northern humile, elatius and sativum clade; and 3) define sativum as a monophyletic group. The southern humile populations form a clade that is distinct from elatius, sativum and even northern humile in the combined molecular tree; although, this node is supported only tenuously by the individual molecular trees. The RAPD trees indicate that elatius is the sister taxon to the cultivated sativum, a relationship that is observed in the combined molecular tree. By comparison, the ISSR tree presents only modest support for northern humile as the single closest relative of the domesticated pea. A small number of polymorphic ITS sites actually places northern humile further from sativum and underscores the close affinity among all the non-fulvum peas. Species fulvum is clearly the most distinct of the pea taxa in every analysis.

Key words: ISSRs, ITS sequences, morphological characters, phylogenetics, Pisum, RAPDs