Geoffroy's sideneck turtle (Phrynops geoffroanus)

Mohamed bin Zayed Species project number 1025921

The taxonomic history of the New World Pleurodira, particularly the family, Chelidae is in confusion, and muddled by few in depth scientific studies and those that have been published have not been subjected to scientific rigor or genetics, and based primarily on supostion and coloration  and from small sample sizes taken from museum collections for example, McCordet al. 2001.

 The Phrynops geoffroanus complex is a problematic group taxonomically (RhodineMittermeier, 1983; Rueda-Almonacidet al. 2007). This species has the greatest distribution geographica of all of the species in the family Chelidae, distributed widely in South America from the  Guianas, Venezuela, Colombia, Equador, Peru through Brazil to Bolívia, Paraguai, Uruguai and northen Argentina (Pritchard and Trebbau, 1984; Ernst and Barbour, 1989;Cisneros-Heredia, 2006; Baldo et al. 2007; Vogt, 2008). According to the distribution map of Iverson (1992) this species occurs in 6 distinct biomes: Floresta Amazônica, Cerrado, Caatinga, Pantanal, Mata Atlântica and Floresta de Araucária.

The lack of understanding of this species is related to its huge geographic distribution and the lack of systematic collections from populations of this species from thrughout its range, which did not permit the identication of characteristics of specific populations of this species within different river basins (Rueda-Almonacid, 2007), most likely  there are new species masked by the lumping of all individuals of this complex into one taxon.

 Because of this the detailed taxonomy and systematic relationships of genetic patterns of populations from different regions and habitas are necessary to discern the differences between populations and allow for identification of distinct populations (new species) such that conservation strategies can be undertaken where needed.

Molecular data have become the fundamental tool to provide the biological information necessary for separating sibling species or studies of bioconservation and systematics (Burks, 1997). Molecular studies have been extensively used not just to clarify problematic phylogenetic relationships which were impossible using other methods, but also to test the phylogenetic ideas that were historically  proposed without the quantitative data.

 Various studies havea demonstrated that mitochondrial genes present a strong tool to infer phylogenetic relationships in diverse groups of animals, including turtles (Georges et al., 1998; Noonan, 2000; Serbet al., 2001; Parhamet al., 2001). The control region, the largest region without mitocondrial DNA coding, has a mutation level much higher than the coded regions of this molecule. Because of this high variability, this region has been used widely in analysis of population structure and phylogeny of related species. In addition to the mitochondrial gene, Citocromo Oxidase I (COI) is the gene used in the studies using the protocol of genetic barcode. With the technology of DNA-Barcoding, or COI it has been shown to be an important tool for genetic analysis, permitting a good resolution for analyses involving identification and separation of sibling species (Hebertet al., 2004).

 

7. METHODS

 

Study Areas

 

We conducted 13 collecting expeditions from 2009- 2011 in Northern, Northeaster, Central-Western, Southern, and Southeastern Brazil specifically for the purpose of capturing a sample of 5-15 individual turtles from all of the major river basins where this species is known to occur.

 

Capture Methods

We captured turtles using for distinct methods: monofilament fishing line 0.50mm in diameter with nº 6 fish hooks baited with fresh pieces of bloody mammal meat, hoop traps with or without leads baited with chicken or fish, trammel nets and hand captures (Fig. 2). The methods varied with each locality depending on the size of the water bodies, current, and presence of human activity.

Figure 2 – 1) Trammel net; 2) Gill net n°16  for capturing fish; 3) Hooptrap; 4) Hoop trap with leads; 5) Dip net; 6) Hook and line. Photos: Vinícius Tadeu de Carvalho

 

Morphometric data

We took the following measurements to compare the specimens of P. geoffroanus collected at each locality: head length (CCB), head width (LCB), head height (ALC), interorbital width (LI), parieta width (LP); carapace length (CC), carapace width (LC), carapace height (AC), vertebral scute lengths (CV), vertebral scutes widths (LV), plastron length (CP), plastron width (CLP), width of anterior plastron lobe (LPAP), length of anterior plastral lobe (CPPP), width of posterior plastral lobe (LPPP), width of the gular/humeral seam (LPGH), with of the femoral/anal seam (LPFA),  intergular seam length (EIG), interhumeral seam length (EIH), inter pectoral seam length (EIP),  inter abdominal seam length (EIA),  inter femoral seam length (EIF), and inter anal seam length (EIAN).

 

Analysis of morphometric data

Estamos gerando matrizes de dissimilaridades Bray-Curtis entre as medidas biométricas: (CCB), (LCB), (ALC), (LI), (LP), (CC), (LC), (AC), (CV) (LV), (CP), (CLP), (LPAP), (CPPP), (LPPP), (LPGH), (LPFA), (EIG), (EIH), escama inter peitoral (EIP), (EIA), (EIF), (EIAN). Estas matrizes serão ordenadas e as dimensionalidades reduzidas por escalonamento não-métrico multidimensional (NMDS).

Será gerado um modelo de NMDS para cada localidade, de modo que a variação biométrica entre populações poderá ser demonstrada em duas dimensões (Legendre & Legendre, 1998). Os escores produzidos serão considerados fatores de quantificação das diferenças biométricas entre indivíduos provenientes de localidades diferentes.Os dados biométricos serão agrupados hierarquicamente para determinar relações de similaridade entre populações provenientes de diferentes localidades. 

 

8. RESULTS

 

Turtles were collected in the following localities:Rio Anhanduinzinho, município de Campo Grande – MS; Córrego do São Gonçalo, município de Cuiabá – MT; Estação Ecológica Serra das Araras, município de Porto Estrela – MT; Rio Piracicaba, município de Piracicaba – SP; Rio Uberabinha, município de Uberlândia – MG; Estação Biológica de Caratinga, município de Caratinga – MG; Lago Paranoá, Brasília – DF; Ribeirão Lindóia, município de Londrina – PR; Reserva Biológica Saltinho, município de Tamandaré – PE; Lago Itans, município de Caicó – RN; Rio Palmeiras, município de Juazeiro do Norte – CE; Reserva Biológica Jaru, município de Vale do Paraíso – RO e Rio Cauamé, município de Boa Vista – RR (Figure 3 and 4).

The results of the Project show advances in the known geographic distribution of the species and the relative abundance of the species at these collecting sites. The morphological and osteological aspects are being analyzed at the moment, to resolve the confusion of the taxonomic features of this species in different river basins with the certainty of identifying new species within this material collected in eleven Brazilian states.

 

 

Figure 3 – Habitat types where we collected Phrynops geoffroanus. 1) Rio Piracicaba, SP; 2) Rio Anahanduizinho, MS; 3) Rio Paquevira, PE; 4) Lago Paranoá, DF; 5) Lago Itans, RN; 6) Rio Uberabinha, MG; 7) Córrego São Gonçalo, MT; 8) Rio Machado, RO; 9) Rio Palmeiras, CE; 10) Ribeirão Lindóia, PR. Photos: Vinícius Tadeu de Carvalho

 

Figure 4 – Represention of areas sampled. Red dots indicate the areas sampled in the states of: Rondônia, Mato Grosso, Mato Grosso do Sul, Minas Gerais, São Paulo, Paraná, Rio Grande do Norte, Ceará, Pernambuco and Distrito Federal.

 

Preliminary molecular results and discussion

 

A total of 89 samples of P. geoffroanus were sequenced from the following localities: s(08 specimens from Pernambuco; 08 specimens from Minas Gerais; 07 specimens from Mato Grosso do Sul; 15 specimens from Rio Grande do Norte; 10 specimens from São Paulo; 07 specimens from  Paraná; 09 specimens from Ceará; 07 specimens from  Distrito Federal; 09 specimens from Minas Gerais-II; 02 specimens from Rondônia; 03 do estado do Mato Grosso; 02 specimens from Roraima; 02 specimens from Corrientes na Argentina.

The raw sequences have 570 base pairs. The sequences referred to as the outgroup are species povided by GenBank and can be visualized in the phylogenetic tree, the sequences utilized in the final analysis have 530 base pairs.

The sequences obtained were edited with the program Bioedit (Hall, 1999), with the help of the tool Clustal W (Thompson et al., 1996) forming a base for the construction of a matrix of the data containing all of the sequenced nucleotides.

After the alignement of the sequences, phylogenetic reconstructions were made using the method of Maximum liklehood, using the evolutionary model of  Kimura-2-parameters (Kimura, 1980) with the program MEGA 5 (Tamura et al. 2011) to estimate the number of cladosmonophyletics  existing in the sample. The support of the branches was evaluated through boot strap analysis, with 1000 réplicas (Felsenstein, 1985).

 

The resulting tree shows the species Phrynops viridis with a value of support above 90% of  bootstrap (Fig. 5), it is a well defined species and not part of the P. geoffroanus complex and needs to be elevated to species status and taken out of the synonomy of P. geoffroanus.  Phrynops tuberosus is also a distinct species.  More sequencing needs to be undertaken using other base pairs without these two species to look for finer separations within the material collect. 

 

At the present time we are comparing the morphological characteristics of these  two species, incluiding descriptions of coloration and patterns of the carapace, plastron, and soft parts.  The ranges of these two new species are also being defined. 

 

T