Imperfect isolation: factors and filters shaping Madagascar's extant vertebrate fauna

Abstract

Analyses of phylogenetic topology and estimates of divergence timing have facilitated a reconstruction of Madagascar's colonization events by vertebrate animals, but that information alone does not reveal the major factors shaping the island's biogeographic history. Here, we examine profiles of Malagasy vertebrate clades through time within the context of the island's paleogeographical evolution to determine how particular events influenced the arrival of the island's extant groups. First we compare vertebrate profiles on Madagascar before and after selected events; then we compare tetrapod profiles on Madagascar to contemporary tetrapod compositions globally. We show that changes from the Mesozoic to the Cenozoic in the proportions of Madagascar's tetrapod clades (particularly its increase in the representation of birds and mammals) are tied to changes in their relative proportions elsewhere on the globe. Differences in the representation of vertebrate classes from the Mesozoic to the Cenozoic reflect the effects of extinction (i.e., the non-random susceptibility of the different vertebrate clades to purported catastrophic global events 65 million years ago), and new evolutionary opportunities for a subset of vertebrates with the relatively high potential for transoceanic dispersal potential. In comparison, changes in vertebrate class representation during the Cenozoic are minor. Despite the fact that the island's isolation has resulted in high vertebrate endemism and a unique and taxonomically imbalanced extant vertebrate assemblage (both hailed as testimony to its long isolation), that isolation was never complete. Indeed, Madagascar's extant tetrapod fauna owes more to colonization during the Cenozoic than to earlier arrivals. Madagascar's unusual vertebrate assemblage needs to be understood with reference to the basal character of clades originating prior to the K-T extinction, as well as to the differential transoceanic dispersal advantage of other, more recently arriving clades. Thus, the composition of Madagascar's endemic vertebrate assemblage itself provides evidence of the island's paleogeographic history.

Figure 1. Simplified bathymetric map of the western and central Indian Ocean, and adjacent Southern Ocean.

Note that the 0–200 m bathymetric interval around Antarctica is not portrayed due to a lack of detailed information in the areas adjacent to the continent. Based on the GEBCO chart.

Figure 2. Simplified bathymetric map of the Mozambique Channel area.

The relative positions of Madagascar, Mozambique, and nearby islands are shown in (a), while (b) shows a bathymetric cross-section along the Davie Ridge. Red dots on (a) indicate the two ends of the cross-section. Even if elements of the ridge were subaerial, the deep, broad troughs widely separating the peaks would have posed formidable barriers to obligatorily terrestrial animals.

Figure 3. Timetable summarizing major paleogeographical and paleoclimatic events relevant to the biogeographic history of Madagascar.

Table includes a non-exhaustive summary of vertebrate taxa that arrived or were present on the island during four major intervals (as in ; herein, taxa from the second and third period were merged for analysis). The box for the Jurassic-Cretaceous shows extinct fossil taxa that were present in Madagascar in the Late Cretaceous, plus extant taxa reconstructed to have existed during this time. Subsequent boxes show taxa that are estimated to have arrived on the island during the respective interval, with each image representing one (or rarely two) endemic Malagasy clades. Extinct taxa are marked with † (in parentheses if taxon is extinct on Madagascar but surviving elsewhere). Red font marks taxa that might be of younger origin according to some molecular estimates. Maps show changing landmass configurations and patterns of vertebrate appearance in or colonization of Madagascar by time slice and proportion of dominant colonizer types (unadjusted frequencies in percent, after [13]).

Figure 4. Dendrogram based on the correlations between assemblages or the averages of assemblages, joining clusters for which the average similarity between members is the greatest.

Each group is a vector comprised of 4 numbers (the percentages of amphibian, reptile, bird, and mammal families, or in the case of the Cenozoic for Madagascar, the percentages of clades belonging to each class known or inferred to have arrived independently).