Cladistic biogeography: identity and place


Four papers published recently in this journal (Crisci, 2001; Brooks & Van Veller, 2003; Ebach et al., 2003; Van Veller et al., 2003) revised some concepts of historical biogeography. All of them deserve attention because they reflect the current state of science within the discipline. I believe, however, that some of their conclusions, particularly those referring to the scope of cladistic biogeography, may lead to misunderstandings. My objective is to comment on these contributions, in order to discuss cladistic biogeography alongside the other historical biogeographical approaches.

Crisci (2001) (see also Crisci et al., 2000, 2003) analysed some of the current forces shaping historical biogeography, defined the biogeographical processes that modify the spatial distribution of taxa, and presented a classification of the biogeographical approaches and techniques. The author recognized nine basic approaches to historical biogeography: centres of origin and dispersal, panbiogeography, phylogenetic biogeography, cladistic biogeography, phylogeography, parsimony analysis of endemicity, event-based methods, ancestral areas and experimental biogeography. Problems with this classification that have been discussed elsewhere (Morrone, 2004; Van Veller, 2004) include that ‘event-based’ methods, those that postulate explicit models of the processes that may have had an effect on geographical distributions, are excluded from cladistic biogeography, being treated as a different approach.

Brooks & Van Veller (2003) presented a critique of parsimony analysis of endemicity (PAE), an approach originally developed by Rosen (1988), which classifies the study areas according to their shared taxa by means of the most parsimonious solution (Morrone, 1994; Morrone & Crisci, 1995; Crisci et al., 2000, 2003; Escalante & Morrone, 2003). The authors rejected PAE as an appropriate a priori method for cladistic biogeography. This is erroneous, as the method does not include phylogenetic information about the taxa analysed, so it cannot be a cladistic biogeographical method (Morrone & Crisci, 1995; Humphries & Parenti, 1999).

Van Veller et al. (2003) (see also Brooks & Van Veller, 2003) questioned the usually accepted idea that several cladistic biogeographical methods have been developed to implement the same research programme. They considered that a priori methods – those that allow modification of the taxon–area cladograms to deal with dispersal, extinction or duplicated lineages, in order to obtain resolved area cladograms and provide the maximum fit to a general area cladogram – are intended to implement cladistic biogeography; whereas a posteriori methods – those that deal with dispersal, extinction or duplicated lineages after parsimony analysis of a data matrix based on the unmodified taxon–area cladograms – are intended to implement phylogenetic biogeography. Thus, they restricted the field of cladistic biogeography to encompass only component analysis, reconciled tree analysis, three-area statement analysis and paralogy-free subtree analysis, and assigned Brooks’ parsimony analysis and component compatibility to phylogenetic biogeography. I find this inappropriate, because ‘phylogenetic biogeography’ has been used for decades to refer to Hennig's (1966) and Brundin's (1966) cladistic implementation of the dispersalist approach (Morrone & Crisci, 1995; Humphries & Parenti, 1999). [In fact, Nelson (1969) initially tried to formalize Hennig's phylogenetic biogeography, but soon he (Nelson, 1974) rejected all a priori rules to resolve centres of origin and dispersal without reference to general patterns of vicariance.] In addition, the statement ‘all methods used in historical biogeographical analysis aim to obtain resolved area cladograms that represent historical relationships among areas’ (Van Veller et al., 2003: 319; see also Brooks & Van Veller, 2003) is not correct. Instead, it would be appropriate to refer to cladistic biogeography rather than to the whole field of historical biogeography, which also includes panbiogeography, that does not use area cladograms to represent relationships among areas but instead uses generalized tracks.

Ebach et al. (2003) noted the two meanings of ‘phylogenetic biogeography’, one, of Hennig (1966) and Brundin (1966), related to the search for centres of origin and the other more recent one, of Van Veller et al. (2003), related to a posteriori cladistic biogeographical methods. They accepted that phylogenetic biogeography sensuVan Veller et al. (2003) originates in Hennig's (1966)‘parasitological method’, and that it was inappropriate as a cladistic biogeographical approach. In their effort to respond to Van Veller et al. (2003), these authors have implicitly accepted their restriction of cladistic biogeography.

Crisci (2001), Van Veller et al. (2003) and Ebach et al. (2003) have restricted the field of cladistic biogeography, whereas Brooks & Van Veller (2003) have expanded it, by assigning PAE to it. So, what is cladistic biogeography? Instead of giving a formal definition, I will present a classification of historical biogeographical methods (Morrone, 2004) and discuss its basis (Table 1):

Table 1.  Taxonomy of the methods of historical biogeography
Basic approachesTechniques
DispersalismManual reconstruction (Matthew, 1915)
Phylogenetic biogeography (Brundin, 1966; Hennig, 1966)
Ancestral areas (Bremer, 1992)
Phylogeography (Avise, 2000)
Vicariance biogeography
 PanbiogeographyManual reconstruction (Croizat, 1964)
Connectivity and incidence matrices (Page, 1987)
Track compatibility (Craw, 1988)
Parsimony analysis of endemicity (Rosen, 1988)
 Cladistic biogeographyReduced area cladogram (Rosen, 1978)
Ancestral species map (Wiley, 1980)
Quantitative phylogenetic biogeography (Mickevich, 1981)
Component analysis (Nelson & Platnick, 1981)
Brooks' parsimony analysis (Wiley, 1988)
Component compatibility (Zandee & Roos, 1987)
Quantification of component analysis (Humphries et al., 1988)
Three-area statement analysis (Nelson & Ladiges, 1991)
Reconciled tree analysis (Page, 1994)
Paralogy-free subtree analysis (Nelson & Ladiges, 1996)
Vicariance event analysis (Hovenkamp, 1997)
Dispersal–vicariance analysis (Ronquist, 1997)
Comparative intraspecific phylogeography (Arbogast & Kenagy, 2001)

1. The first dichotomy divides historical biogeography into dispersalism and vicariance biogeography. The former seeks to locate centres of origin/ancestral areas and then use dispersal out of them to explain the biogeographical histories of particular taxa. Phylogenetic biogeography (sensuHennig, 1966 and Brundin, 1966), ancestral areas (Bremer, 1992) and intraspecific phylogeography (Avise, 2000) may be ascribed to the dispersalist approach because they basically concur in explaining distributions in terms of narrative dispersal explanations, although they eventually accept vicariance explanations. Vicariance biogeography seeks to uncover patterns of biotas by comparing the distributional patterns of different plant and animal taxa. The ontology of this approach lies in biogeographical homology (Craw et al., 1999; Humphries & Parenti, 1999; Morrone, 2001, 2004).

2. The second dichotomy divides vicariance biogeography into panbiogeography and cladistic biogeography. Panbiogeography (Croizat, 1964; Craw et al., 1999) reconstructs ancestral biotic distributions by drawing individual tracks on maps (connecting localities of taxa by lines approximating to minimum spanning trees) and looking for coincidence among unrelated taxa, which leads to the recognition of generalized tracks. Morrone (2001, 2004) considered that panbiogeography – including PAE as a quantitative technique of this approach (Smith, 1992; Craw et al., 1999; Luna et al., 2000) – allows us to identify primary biogeographical homology, which represents a conjecture on a common biogeographical history. Cladistic biogeography (Nelson & Platnick, 1981; Humphries & Parenti, 1999; Ebach & Humphries, 2002) assumes a correspondence between the phylogenetic relationships of the taxa under study and the relationships among the areas that they inhabit. Comparative phylogeography (Arbogast & Kenagy, 2001) may be ascribed to cladistic biogeography because its aim is to compare phylogeographical patterns of multiple co-distributed taxa, in order to recognize vicariant patterns. A cladistic biogeographical analysis basically comprises three main steps (Morrone & Carpenter, 1994): construction of taxon–area cladograms from taxon cladograms by replacing their terminal taxa with the areas of endemism that they inhabit; conversion of taxon–area cladograms into resolved area cladograms (if necessary); and derivation of general area cladogram(s). Morrone (2001, 2004) considered that cladistic biogeography deals with secondary biogeographical homology, namely the cladistic test of the primary biogeographical homology formerly recognized. Thus, panbiogeography and cladistic biogeography may be applied as two successive steps of a biogeographical analysis (Morrone & Crisci, 1995). [A similar distinction between these two steps has been made by Donoghue et al. (2001) and Hausdorf & Hennig (2003).]

3. Within cladistic biogeography, I was unable to find a satisfactory way to classify the different techniques. Several features may be employed as demarcation criteria: pattern-based vs. event-based methods (Ronquist & Nylin, 1990), parsimony vs. tree manipulation methods (Humphries & Parenti, 1999), a priori vs. a posteriori methods (Van Veller et al., 2000, 2002; Van Veller & Brooks, 2001), taxon vs. area biogeography (Crisci et al., 2003) or cladogenetic vs. event-based methods (Van Veller, 2004). I believe that these criteria are inadequate because they do not consider all the complexities of this range of techniques. Furthermore, I consider the more pressing need at this moment to be for an evaluation of them in order to provide practitioners the means by which to chose the most appropriate technique.

With the taxonomy discussed herein I do not pretend to describe the ‘progress’ of historical biogeography from dispersal biogeography to cladistic biogeography. What I have attempted to do is to provide some bases to characterize the three basic research programmes in the field, each of which asks different questions. I fully agree that ‘the present plethora of techniques reflect a lack of scientific debate and agreement as to what constitutes the ontology (specification of conceptualization) of biogeography’ (Ebach & Humphries, 2003, p. 959). I hope this contribution furthers an understanding of the current approaches in historical biogeography and their techniques, in order to provide practitioners and others with a clearer perspective of this discipline.


I thank Malte Ebach, Chris Humphries and Gary Nelson for their critical comments. Economic support by grants CONACyT 36488 and DGAPA IN206202 is acknowledged.


Juan J. Morrone is a Professor of biogeography, systematics, and comparative biology at the Facultad de Ciencias, Universidad Nacional Autónoma de México (UNAM). His main interests are in biogeography and phylogenetic systematics of weevils (Coleoptera: Curculionoidea). He has published more than 150 scientific articles and 14 books. His latest book is Homología biogeográfica: El componente espacial de la evolución (Cuadernos del Instituto de Biología 37, Instituto de Biología, UNAM, México, DF, 2004).

Editor: Chris Humphries