Weekly reading: Page (1988) on assumptions 0, 1, 2
Today we read Page (1988): Quantitative Cladistic Biogeography: Constructing and Comparing Area Cladograms. In this paper, the author presented algorithms that implement Assumptions 1 and 2 previously proposed by Nelson and Platnick (1981). Together with Assumption 0 (Zandee, 1987), these so-called ‘assumptions’ are used to resolve fundamental/individual area cladograms involving missing taxa, widespread taxa, and redundant distributions. They provide different sets of solutions for a given taxon-area cladogram with such conflicts, with Assumption 0 typically yielding the least and Assumption 2 the most area cladograms.
While these assumptions provide a means for resolving area cladograms, they say very little about the processes/events required to reach the solutions. This was an issue during our discussion. We will use Figure 3 in Page’s article as an example to illustrate this problem.
We will discuss each of the three assumptions.
1. Assumption 0 implies either dispersal that does not lead to speciation, or recent vicariance without speciation which, however, is in violation of the basic principle of cladistic biogeography; i.e. vicariance leads to speciation and speciation is caused by and only by vicariance.
2. Assumption 1 implies (1) dispersal, where areas with widespread taxa are considered sister areas (essentially the same as Assumption 0); or (2) two scenarios where the areas are paraphyletic: (i) a widespread taxon failed to speciate when the areas separated, however (and as above), this is inconsistent with the vicariance-speciation paradigm and is thus not legitimate, and (ii) the dispersal of a widespread taxon from a more ‘derived’ areas to the more ‘ancestral’ area. Though this also causes some confusion. How is it that there was no ancestral taxon present in the most ancestral area? Moreover, there must be an event that introduced an ancestral taxon to the ancestral area that would subsequently become a more derived area.
3. Assumption 2 implies a possibility of all solutions discussed for Assumptions 0 and 1. In addition, if the areas d & e occupied by a widespread taxon are polyphyletic, then this implies one dispersal and one extinction event; i.e. taxon 1 dispersed to area e, and the species formed by vicariance (with the derived area e and the ancestral area to a & b) has gone extinct in area e.
These issues were briefly touched on in Ronquist (1997), who introduced a new way of thinking about biogeography, termed ‘vicariance-dispersal analysis’. In that paper, Ronquist (1997) discussed the number of ‘events’ required to reconstruct ancestral areas using other existing methods (including methods based on Assumptions 0, 1, 2). All of these methods require more steps/events than his dispersal-vicariance analysis.