Uninformative characters and apparent conflict between molecules and morphology

Abstract

Incongruence between different data sets remains one of the central issues in systematics. Numerous tree- based and character-based tests to identify incongruence have been developed, and each has strengths and weaknesses (e.g., for reviews, see Cunningham 1997a; Johnson and Soltis 1998 ). Currently, one of the most widely used methods for evaluating incongruence within a parsimony framework is the homogeneity test of Farris et al. (1995) , usually termed the incongruence length difference (ILD) test (Cunningham 1997a; but see Johnson and Soltis 1998 ) or the partition homogeneity Test (Swofford 2000 ). The test has been argued to produce more accurate results than other tests (Cunningham 1997a ) and is also easy to implement using PAUP* (Swofford 2000 ).