A revised cline theory that can be used for quantified analyses of evolutionary processes without parapatric speciation

Aim Classical cline theory interprets clines as speciation processes in two stages: (1) smooth clines, typical of species with no marked subspecies, and (2) stepped clines, characteristic of species with marked subspecies, separated by hybrid zones, where the clines are steeper than within the areas of the subspecies. (These steeping clines eventually become vertical clines—a situation, where the taxa turn into separate species, termed `parapatric speciation'.) This theory, dating from the late 1930s, still has a number of active supporters. But, because results from population genetics suggest that barriers are necessary for speciation to take place, most authors nowadays reject both parapatric speciation and classical cline theory, the former appearing to be the logical consequence of the latter. The theory presented here does not involve parapatric speciation, but provides a revised cline theory including a third stage, the broken cline, which corresponds to the stage at which speciation finally takes place. I suggest that a cline theory without the prerequisite of parapatric speciation can be very useful, for it provides both an evolutionary interpretation of a given geographical distribution and also a method of quantitative analysis. Location An example is taken within a species complex of the chiffchaff, Phylloscopus collybita‐ibericus (Aves, Sylviidae), using populations ranging from south Iberia to north Sweden for one variable and to south Belgium for three other variables. Methods I used regression analyses to interpret clinal variations, and Student's t‐tests to evaluate the broken or continuous nature of the clines. Results In this example, the cline is broken between the adjacent populations of P. ibericus and P. c. collybita, and is continuous and gradual elsewhere. Main conclusions (1) Phylloscopus ibericus and P. c. collybita are good examples of taxa undergoing current speciation; (2) the southern and northern populations of P. ibericus on one hand, and of P. c. collybita and of P. c. abietinus on the other hand, show cline patterns illustrating pairs of vicariant and isolated `weak' subspecies in the sense of Deuve (1994 ).