DIVA: Reconstructing the Ancestral Range Distribution

Since the beginning of the semester, we have been learned that the geographic distributions of organisms are determined by historical and ecological mechanisms. We also have learned some of the methodology that has been applied in historical biogeography studies. My last two posts were dedicated to historical biogeography, its pattern-based methodology, and assumptions. But I have not yet described an event-based methodology…


In event-based approaches, the processes responsible for taxa distribution are analyzed and the events are weight modeled.  One example of event-based approach is the Dispersal-vicariance analysis –DIVA. It reconstructs the ancestral range distribution of a particular group of organisms, minimizing the dispersal events needed for explaining the distributions. In this approach, the vicariance events have no cost, while the dispersals and extinctions cost one per area unit added to the distribution (Gomez & Lobo, 2006). This method was first applied to the theory of coevolutionary inference, and it uses the computer program DIVA (Ronquist, 1997).

Event-based methods have been criticized because the accuracy of the result depends on the validity of the model, and is necessary to specify a relative cost for each event in the model. Based on the model, the program will do a parsimony analysis and will provide the minimum-cost reconstruction.

To run the analysis, a three-dimension step matrix is needed (due to one ancestor can have two direct descents). The input information comprehends of the taxon relationship cladogram, the area cladogram, and the studied areas. Dispersal-vicariance, however, does not take general area relationships into account. It is just related to taxa cladogram and their current distribution. Thus, it is possible to use DIVA in taxon biogeography studies, even when no general area cladogram is available. 

Since the Ronquist (1997), many studies have used DIVA to reconstruct taxa ancestral areas. In Alberti et al. (2007), for example, proboscids (Mammalia) were studied applying dispersion-vicariance analysis. They used DIVA 1.1 (Ronquist, 1996), applying an exact search according to the default program. The genera of Gomphotheriidae ancestral distributions were given in an exact solution requiring 15 dispersal events. They discussed each processes (e. g. vicariance, dispersal, extinction) and each node, concluding that the ancestral distribution for the group included Africa-Europe-Asia-North America. The distribution could be achieved during earliest Miocene, a time of low sea-levels and low temperatures (Alberdi et al. 2007).

Gomes & Lobo (2006) studied Iguania using Dispersal-vicariance, Fitch Optimisation, and Weighted Ancestral Area Analysis. All their results were congruent. The ancestral area analyses agreed in an Andean-Patagonian origin for the studies clade, and were congruent with previous hypotheses and paleontological data.

One problem that I see in DIVA is that we do not have probabilities to base on. Each node received a hypothesis of ancestral area and can have many ambiguous results, which difficult our interpretation. The researcher or specialist interprets and proposes historic factors and processes for the under studied taxa distribution. Fortunately, other programs and approaches have been developed based on maximum likelihood and Bayesian analyses that given statistical results.




Alberdi, M.T.; Prado, J.L.; Ortiz-Jaureguizar, E.; Posadas, P. & Donato, M. (2007). Historical Biogeography of Trilophodont Gomphotheres (Mammalia, Proboscidea) reconstructed applying dispersion-vicariance analysis. Cuadernos del Museo Geominero (8), 4th European Meeting on the Palaeontology and Stratigraphy of Latin America. Instituto Geológico y Minero de España, Madrid.

Gómez, J. M. D. & Lobo, F. (2006). Historical Biogeography of a clade of Liolaemus (Iguania: Liolaemidae) based on ancestral areas and dispersal-vicariance analyis (DIVA). Papéis Avulsos de Zoologia (São Paulo) 46 (24): 261-274.



Other references:


DIVA v1.1 User’s manual




  1. Juanita Said:

    Cissa I agree with you. The difficult part comes when you get an output with 20 possible ancestral ranges for a node and there is no way of choosing the best, because all (for the program) are equally possible.

  2. Jayani Nimanthika Said:

    This is not actually a comment, I need a help from youi. I am a Sri Lankan and currently doing a research on biogeography of Sri Lankan Calophyllum species. I got a problem while analysis using DIVA. If you are able to help me please reply me. I’m looking for a person who is very familier with DIVA software in windows version. Thank you.

    • cwaichert Said:

      Sorry, but I’m not using this blog anymore, so I didn’t see your comment. And I have used DIVA just once, I’m not very familiar with it.

{ RSS feed for comments on this post} · { TrackBack URI }

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: