‐flying vertebrates: Italian Journal of
Zoology, v. 71, p. 217-228.
Vogt, C., 1852, Classification des poissons ganöides: Ann. Sci. Natur., v. 4, p. Paris.
Wade, R., 1932, Preliminary note on
, representing a new
family of chondrostean fishes the Pholidopleuridae: Annual Magazine of Natural
History, v. 10, p. 473-475.
White, E., 1932, On a new Triassic fish from North-East Madagascar: Annual
Magazine of Natural History, v. 10, p. 80-83.
—, 1933, New Triassic Palaeoniscides from Madagascar: Annual Magazine of Natural
History, v. 10, p. 118-128.
Woodward, A., 1910, On some Permo-Carboniferous fishes from Madagascar: Annual
Magazine of Natural History, v. 8, p. 1-6; London.
Yoder, A., Burns, M., Zehr, S., Delefosse, T., Veron, G., Goodman, S., and Flynn, J.,
2003, Single Origin of Malagasy Carnivora from an African ancestor: Nature, v.
421, p. 734-737.
Yoder, A., and Nowak, M., 2006, Has Vicariance or Dispersal been the Predominant
Biogeographic Force in Madagascar? Only time will tell: Annual Review of
Ecology and Evolutionary Systematics, v. 37, p. 405-431.

This study shows that the fish from Late Cretaceous deposits of Madagascar
provide insight into the biogeography of Madagascar, including the level of endemicity
in Recent fauna. This study also affects the understanding of biogeography of
particular groups of teleosts. Conclusions from the previous chapters are summarized
and synthesized below.

The fossil fishes of the Late Cretaceous of Madagascar are one of the last
remaining groups to be described and analyzed from the Maevarano Formation. The
terrestrial vertebrate fauna from this time has lead to important analyses and
hypotheses of the biogeography of Madagascar and how fauna have changed over
time. The fish fauna adds another piece to this puzzle.
Our understanding of fossil fishes during the Mesozoic is continuously growing,
but is still far from being complete. The adaptive radiation of teleostean fishes makes
it difficult to ascertain phylogenetic relationships. Diversification is more rapid and
more complex than the lithologic record can preserve. The identification of fish taxa
new to the Late Cretaceous of Madagascar supports two major themes:
1. Our idea of “Laurasian” vs. “Gondwanan” fauna after the break-up of Pangea is
not applicable to all faunal groups. Some of the fishes identified here were
before only known from Laurasian deposits, but it appears that there is a more
cosmopolitan nature to the epicontinental seaway fish fauna during the
fractionation of Pangea, and subsequently Laurasia and Gondwana. A broader
scale regional endemicity or cosmopolitanism plays a role.

160

2. Our understanding of teleost groups, especially Ostariophysans, and their
diversification during the Mesozoic is still limited. Even with such a poor fossil
record of ostariophysans, the group is the subject of continual hypotheses. Too
much emphasis is placed upon the current distribution of species to construct
distribution diagrams for 70 million years ago, which is unrealistic for migratory
fauna. As more of these depauperate faunas are found and identified within
fossil assemblages, we will be able to better assess the true distributions,
paleoecology and evolutionary history of these groups.

From the Triassic to the Late Cretaceous and then to the Recent, there have
been major faunal turnovers of the fishes on Madagascar. Throughout this geologic
window, Madagascar became progressively more isolated from other landmasses.
Dispersal is clearly a significant part of the history of the fauna on Madagascar, and is
likely to have overwritten older vicariance events after Madagascar became isolated.
The endemicity of species on Madagascar during the Late Cretaceous is difficult to
quantify (as is with most fossil assemblages), but based on the newly identified taxa,
there appears to have been less endemicity during the Late Cretaceous than there is
today. Most of the taxa represented in the Late Cretaceous of Madagascar identified
to genus level, are now extinct genera (
), or even extinct at the family level (Enchodontidae and Phyllodontidae), but
were widespread during the Late Cretaceous.

161

If marine dispersal has had a large impact on the fauna of Madagascar and its
endemicity, then it is natural to ask where these fauna came from and how. Currently,
researchers are trying to determine how terrestrial fauna can move between islands,
or from the mainland to an island. Extant taxa have been observed rafting on material
between landmasses (e.g. iguanas (Censky et al., 1998) and invertebrates (Thiel and
Gutow, 2005)). Thiel and Gutow (2005), also noted occurrences of fish populations
following rafting material. Samonds et al. (2012) recently evaluated the probability of
rafting animals based on ocean currents through the Mesozoic and Tertiary, and
argued for the probability that fauna successfully traversed the Mozambique Channel
(including rafting and swimming) to reach Madagascar from Africa. Paleoceanography
and paleontology were synthesized to explain how and when populations could be
established in novel environments. This synthesis concluded that African groups were
able to traverse the Mozambique Channel by rafting during opportune times of oceanic
currents (Samonds et al., 2012), which could be referred to as “opportunistic
endemicity” as seen on Madagascar.

This project is a step towards reevaluating and better understanding the highly
complex nature of the teleost fossil record. Teleost fishes are often considered a
nightmare among paleoichthyologists, however, their fossil history can answer many
questions, if properly analyzed and understood. The rapidity of teleost adaptation and
diversification allows for analyses on the scale of faunal responses to global change
(climatic and tectonic), but it is difficult to view in the fossil record. Often fossil fishes

162

within faunal assemblages are the last to be described, and it is often very difficult
material to identify and work with. Our understanding of the fish fossil record could
start becoming more complete by looking through museum collections and identifying
fishes that have remained unidentified. Even when identification is at a cruder
taxonomic level, it is still significant, especially within Mesozoic assemblages when
teleosts noticeably started diversifying. Large Gondwanan landmasses with
Cretaceous fossil records such as Africa, South America, Antarctica, and Australia, still
have relatively poor fish records, so filling in these gaps will lead to a greater
representation of global diversity instead of the reliance upon Laurasian assemblages,
for reconstructing ancient global distributions.
I plan on remaining involved in the Mahajanga Basin Project due to the plethora
of material, though it is difficult to work with. As of right now, there is a lack of
understanding of Cretaceous teleosts. The poor Cretaceous fossil fish record and
frustrating teleost osteology, does not leave many described and comparative
collections that can be used for precise taxonomy of disarticulated material. Being
able to compare the Maevarano fauna more directly with other Gondwanan fauna will
be imperative to unraveling the evolutionary history of teleosts.

163

164

Censky, E., Hodge, K., and Dudley, J., 1998, Over-water dispersal of lizards due to
hurricanes: Nature, v. 395, p. 556.
Samonds, K., Godfrey, L., Ali, J., Goodman, S., Vences, M., Sutherland, M., Irwin, M.,
and Krause, D., 2012, Spatial and temporal arrival patterns of Madagascar's
vertebrate fauna explained by distance, ocean currents, and ancestor type:
Proceedings of the National Academy of Sciences.
Thiel, M., and Gutow, L., 2005, The ecology of rafting in the marine environment. II.
The rafting organisms and community, Gibson, R., Atkinson, R., and Gordon,
J., eds., Oceanography and Marine Biology: An Annual Review, Volume 43,
Taylor & Francis, p. 279-418.

165