Full text
| |
| Author(s): |
Felipe de Mello Martins
Total Authors: 1
|
| Document type: | Doctoral Thesis |
| Press: | São Paulo. |
| Institution: | Universidade de São Paulo (USP). Instituto de Biociências (IBIOC/SB) |
| Defense date: | 2008-08-26 |
| Examining board members: |
Joao Stenghel Morgante;
Reinaldo Otavio Alvarenga Alves de Brito;
Alexandre Reis Percequillo;
Mario de Vivo;
Gabriel Henrique Marroig Zambonato
|
| Advisor: | Joao Stenghel Morgante |
| Abstract | |
The bat Desmodus rotundus is one of the three extant vampire bat species. It has a broad distribution, occurring from southern México until Argentina and Chile. Besides its unique feeding habit, this species is of particular interest for being the main vector of cattle rabies. Even with population control methods, studies have estimated in 33 million dollars per year the damage caused by this bat to cattle farming in Brazil. At the same time 200.000 specimens might have been killed in São Paulo state in the year 2000 using the population control methods. Besides the fact that this control did not diminish the number of rabies cases, the impact of this killing in the bats\' natural populations is unknown. Although this species has a broad distribution and recognized morphological variation, no effort was made thus far to understand how this species\' genetic variability is distributed geographically. This work is aimed at studying the common vapire bats\' phylogeographic pattern using a mitochondrial marker, two nuclear markers and skull morphometrics. The mitochondrial marker identified five monophiletic clades without shared haplotypes or contact zones. Each clade represents a distinct geographic region: South Atlantic Forest (SAF), North Atlantic Forest (NAF), Amazon and Cerrado (AMC), Central America (CA) and Pantanal (PAN). The Atlantic Forest clades form an Eastern monophiletic clade opposing the other clade that lies westwards. The nucleotide divergence between these clades is similar to the one described to congeneric species. The divergence times estimated by coalescent and non-coalescent methods point to a Pleistocene vicariant event. The neutrality tests also point to refugia allopatric fragmentation. The biogegraphic pattern described for D. rotundus has a parallel in many other organisms. The nuclear markers showed low variability and sharing of haplotypes among all localities, contrasting with the previous results. Coalescent simulations were carried with populational parameters estimated for the nuclear gene RAG2 and showed compatibility between the observed data and Pleistocene vicariance effect on a neutral nuclear marker. Skull morphometrics showed low differentiation throughout the bats\' distribution. Data on Fst, discriminant functions and canonic variables shows affinity between CA and AMC clades. These two clades together form the distribution of a subspecies previously described to this taxon, Desmodus rotundus murinus. The Mahalanobis distance analyses are also congruent with the results obtained withn the nuclear marker. The analysis done with the software treescan shows a statistic significant correlation between the mtDNA tree and the skull multivariate data. On the basis of the results presented, it is proposed that two lineages currently atributed to D. rotundus are to be recognized as different species: one to the east (Atlantic Forest) and one to the west. A detailed sampling of the Brazilian and South American country will determine the exact range of each species. (AU) | |