CAURA RIVER BASIN (BOLÍVAR STATE) RECORDS New locality of occurrence Two adult living matamatas, a female and a male, were captured and measured in the remote Nichare River, an affluent of the Lower Caura River Basin (LCR) and part of The Caura National Park (Fig. 2). These individuals were sampled by chance during the dry season of 2000 (February-April) while conducting a three-year research study (1999-2001) on the nesting ecology of another native aquatic turtle (Podocnemis unifilis Troschel 1848) (Escalona 2003, Escalona et al. 2009a). The female was caught directly by hand from a boat after being spotted moving on the surface across a narrow section of the river. Although it tried to dive, it did so very slowly, allowing it to be approached and captured rapidly. The male, on the other hand, was caught with a fishing net used in a shallow flooded lagoon to catch P. unifilis (Fig. 3). Location details are shown in figure 2. On the map, the Nichare River locality is indicated by black circle 6. Within this locality, the specific site for the female was recorded at (06°20’7” N, 64°57’31” W), and the male at (06°19’16” N, 64°57’12” W). Habitat and Nichare River It is well recognized that matamatas are highly aquatic bottom-walkers that rarely swim or float and are never seen basking. Habitats of preference are rivers and streams, mainly in areas of slower water flow or still waters such as lagoons, wells, oxbow lakes and temporary overflow pools. It inhabits all types of water: white, clear, and black (Pritchard & Trebbau 1984, Barrio-Amorós & Narbaiza 2008, Morales-Betancourt et al. 2020, Cunha et al. 2021). The Nichare is a black water meandrous river about 5 to 20 meters wide that sustains habitats of similar characteristics and conditions to those described above. It contains lateral and meandric sand bars, shore complexes of pools, flooded lagoons, and oxbow lakes. It is considered a “Seasonally Flooded Riparian Ecosystem” in an “Evergreen Lowland Forest” (Rosales & Huber 1996, Rosales et al. 2003). Morphological characteristics (shell shape, plastron pigmentation, neck coloration, size) Shell shape (rectangular vs oval), plastron pigmentation (dark vs light) and neck underside coloration (bold black bands vs light or red-pink color) were used as features for species diagnostic. Shell characters were measured linearly using a 50 cm caliper. This included the maximum straight carapace length (SCLmax) and straight-line carapace width (SCW),as shown in figure4.Body mass was weighed using a 20 kg Pesola spring scale. No abnormalities were detected during measurements. After taking photographs and measurements, the matamatas were released into the same location of capture. Sex determination Although sexual dimorphism is not very marked in matamatas, the sex can be easily determined because females can reach larger sizes than males, and males have longer and thicker tails (Figs. 5 and 6) and seem to have a more concave plastron (Fig. 7) compared to females (Figs. 5 and 8). Differences, were also confirmed by the few available studies on matamata size, which have indicated that individuals over 40 cm in SCL tend to be females (see Pritchard & Trebbau 1984, Sánchez-Villagra et al. 1995, and see Figs. 3 and 4 of Morales-Betancourt et al. 2020). Female: This individual is characterized by an oval shell outline, such as in the Orinoco morphotype (see Sánchez-Villagra et al. 1995, Cadena et al. 2023). The plastron seems heavily stained rather than naturally pigmented, most likely due to the tannin-rich black waters typical of the Nichare ecosystem. However, further investigation is needed to confirm this observation. The underside neck coloration pattern was not possible to describe, due to lack of images (Fig. 5). This large female matamata turtle, measuring 48.8 cm in SCLmax (37.4 cm in SCW) and weighing 12.7 kg, is, to the best of my knowledge, the second-largest live matamata ever recorded in the Venezuelan Orinoquia. The largest so far was caught in Apure River and measured 52.6 cm in SCLmax with a weight of 17.7 kg (Barrio-Amorós & Manrique 2006). For Colombian Orinoquia (Bojonawi Natural Reserve -BNR), the maximum size reported is 47 cm SCLmax (37 cm in SCW) and a weight of 13.42 kg (Morales-Betancourt et al. 2020). Additional records of large female specimens are available in museum collections. For instance, a female with an oval shell shape from Raudales de Atures, Puerto Ayacucho (Amazonas State, Venezuela) was originally documented by Barrio-Amorós & Narbaiza (1999). This specimen, recently re-measured by T. Escalona, has an SCLmax of 48.7 cm and an SCW of 39 cm. It is deposited at the Estación Biológica Rancho Grande in Maracay, Venezuela, under the catalog number EBRG 3596. Other museum records indicating CL larger than 40 cm are described in Pritchard & Trebbau (1984) (Deposited in Peter CH Pritchard, personal collection, Oviedo, Florida, USA as PCHP 1209) and Sánchez-Villagra et al. (1995) (Deposited at Colección Estación Biológica Hato Piñero, Cojedes, Venezuela as CEBHP unnumbered). Male: This individual has an oval-shaped carapace similar to the Orinoco morphotype and a heavily stained plastron akin to that of the female described above. The ventral side of the neck is reddish-pink, while the dorsal side features a narrow, continuous dark line running along the midline, flanked by two thin dark lines on either side (Fig. 6). The size of this male, ranks among the largest ever reported. The SCLmax, SCW and weight are: 37.7 cm, 28.6 cm, and 5.7 kg. A similar maximum size but much heavier was registered by Morales-Betancourt et al. (2020) for Colombian Orinoquia (i.e., 37.7 cm in SCLmax, 30.3 cm in SCW, and 7.6 kg). During a sevenyear mark and recapture matamata population study, the latter authors found that most males examined (about 86%) had a SCLmax less than 33.1 cm, and the few larger ones (interval 36.1-39 cm) represented only 3%. The largest males ever reported in the literature include one individual living in captivity in Venezuela, measuring 39.7 cm in SCLmax (Pritchard & Trebbau 1984), and another specimen from Guyana reaching an SCLmax of 41.9 cm (Pritchard 2008). OTHER OCCURRENCES WITHIN THE LOWER CAURA BASIN Chelus was confirmed to inhabit the Mato (black circle 4) and Caura Rivers (black circle 5) (Fig. 2). It is important to point out that matamatas have previously been observed in both rivers, though they were referred to as C. fimbriata (Pritchard & Trebbau 1984, Rojas-Runjac et al. 2011). Mato River The observation was made at the locality known as Morrocoy (07°08’08” N, 65°12’43” W), where a male was captured for consumption by local hunters using an arrow and a long spear. Hunters were observed on top of trees near the river shoreline, waiting for any turtle to surface (Fig. 9). This live male, measured 36.6 cm in SCLmax, 28.6 cm in SCW, and weighted about 5 kg. Based on its oval shell shape, reddish-pink ventral neck coloration and probably stained plastron, this male appears to exhibit morphological characteristics typical of the Orinoco morphotype rather than Amazonian. Caura River Two males were caught with fishing nets by local fisherman near La Fortuna (06°48’04”N, 64°49’40”W) and were freed back into the river (Fig. 3). Sizes were as follows: 1) SCLmax: 32.1 cm, SCW: 23.9 cm, weight: 3.5 kg, and 2) SCLmax: 33.6 cm, SCW: 23.4 cm, weight: 3.5 kg (see Fig. 3 for fishing technique). Although no images were taken from these individuals, they both are described as having an oval-shell outline. Collectively, the data confirm the presence of Chelus in at least three river localities (Nichare, Mato, and Caura) within the Lower Caura River Basin. Additionally, the morphological characteristics of the observed individuals suggest that they are representatives of the Orinoco morphotype (C. orinocensis). Nonetheless, further genetic and morphological research is needed to substantiate these findings throughout the Caura drainage, especially since previous work on the Podocnemis unifilis turtle documented a complex geographical pattern of genetic variation heavily influenced by landscape features (Escalona et al. 2009b). The data also reinforces the notion by Pritchard & Trebbau (1984), Pritchard (2008), and studies by Morales-Betancourt et al. (2020) that matamata females can grow very large and attain larger sizes than males.Although there are exceptions to this pattern (see Pritchard 2008), most males are smaller than 40cm in SCLmax, indicating sexual size dimorphism. In matamatas, tail size differs between mature males and females of equal length, it remains to be explored whether there is variation in shell shape or other morphological traits (e.g., skull size)between individuals shorter than 40 cm. Plastron concavity as a reliable trait for sex determination requires further research, as some females show concavity and some males have a rather flat plastron (see Sánchez-Villagra et al. 1995, T. Escalona pers. obs.). Since shell shape is a key trait distinguishing the two Chelus lineages – C. orinocensis with its oval shape and C. fimbriata with its rectangular outline (Fig. 1)– future research utilizing geometric morphometrics is essential to enhance the identification of shape differences across various river systems. This work should also be complemented by genetic studies that prioritize comprehensive geographic sampling, especially in undersampled regions such as Venezuela. Such sampling will facilitate the testing of genetic distinctiveness within specific watersheds and the assessment of gene flow between peripheral and central localities within the species’ distributional range. It is important to note that field records should not be forgotten, as they can still yield valuable insights despite the time since the observation. They can provide useful new knowledge, validate, and improve previous and recent records. This is particularly significant for elusive species with limited observations and remain poorly known, such as the matamata turtles. LAKE MARACAIBO BASIN (ZULIA STATE) RECORDS The occurrence of Chelus in the Lake Maracaibo Basin has remained uncertain. Previously, in this region, matamata turtle occurrences were only known from two observations dated more than 40 years ago (see Pritchard & Trebbau 1984). One of these observations has recently been validated by Gilson Rivas (Museo de Biología, Universidad del Zulia) after interviewing Professor Jose L. Lira from University of Zulia on April 18, 2024. Professor Lira recalls seeing on 1974, a live matamata (Chelus sp.) caught by a local fisherman in the Santa Ana River (Fig. 2; black triangle 2), a tributary on the west side of the lake (G. Rivas, pers. comm.). Furthermore, until recently (9 April 2024), museum specimens of matamata turtles from this region were not known. However, a matamata, preserved in alcohol, was discovered in the herpetological collection of Museo de Biología of Universidad Central de Venezuela (MBUCV), Caracas (deposited as MBUCV 7096, T. Escalona, pers. obs.) (Fig. 10). It is important to acknowledge, that although the hand-written catalogue provides the exact location of the specimen (i.e., northwestern side of the lake; Fig.2; black triangle 1), there is no mention of who brought it to the museum or even the date it was collected. It is only noted that it was found by local fisherman and identified by Oswaldo Fuentes in the mid-late 1990s as Chelus fimbriata. However, determining whether the specimen is truly a C. fimbriata or possibly a C. orinocensis or even a different species or subspecies remains uncertain, as it is a small juvenile (SCLmax: 6 cm) with an incompletely ossified shell that lacks a welldefined shape (Fig. 10). The true identity of this juvenile will require further morphological and genetic analysis. Recent museum records and the validated observation by Professor Lira, it suggests that Chelus sp. may inhabit Lake Maracaibo, warranting further investigation. If confirmed, matamata turtles in this area could represent an isolated breeding population that deserves attention. This is particularly significant because the Lake Maracaibo Basin is situated in a secluded geomorphological region, surrounded by substantial physical barriers (i.e., to the east, is separated from the Orinoco-Amazon system by the Andean mountain range, and to the west, from the Magdalena system by the Sierra de Perijá, and to the south by the union of these). Biogeographic evidence from freshwater fishes indicates that this relative isolation of the Lake acted as a Refugia, allowing the differentiation of fish populations into distinct biotas. This differentiation is notable when compared to neighboring hydrological systems like the Orinoco, Amazon, and Magdalena. The lake, today harbors ancient relictual lineages, new species that arrived via coastal dispersal, and species that evolved in seclusion, leading to high levels of endemism (RodriguesOlarte et al. 2011). SANTA ROSA RIVER–PUTUCUAL LAGOON SWAMP, GULF OF PARIA BASIN (SUCRE STATE) RECORD To date, there is only one known record of the matamata turtle from Sucre State. This specimen is catalogued at the KU Biodiversity Institute and Natural History Museum (Lawrence, KS) under the accession number KU 117344 (see Pritchard & Trebbau 1984). To validate this record, Ana Motta, the Herpetology Collection Manager at KU, confirmed the physical existence of the specimen. She also provided information from the hand-written catalogue, and took the carapace measurement (Curved Carapace Length - CCL: 35.5 cm), and photographs from both the dorsal and ventral views (Fig. 8). According to the catalogue, this specimen, identified as Chelus fimbriata, was collected by James D. Smith on February 26, 1967, in the vicinity of Santa Rosa River (10°24’55” N, 63°23’29” W), located by road 20 km southeast of Casanay Village (10°30’13” N, 63°25’03” W), likely from the Putucual Lagoon Swamp (Fig. 2; black star 3). However, after analyzing the photographs, it is apparent that this individual has an oval-shaped carapace with an unpigmented yellow plastron, and having two dark fine bands running along the light color of the underside of the neck, features diagnostic of the Orinoco morphotype. Consequently, based on these characteristics the specimen appears to be instead a C. orinocensis, pending further investigation (T. Escalona, pers. obs.). Interestingly, the Santa Rosa River is in the lowlands and flows southeast where it merges into Putucual. This swamp drains its water eastward into the San Juan River, which eventually flows into the Gulf of Paria (Melfran Herrera, pers. comm.). This hydrographic system is part of the Gulf of Paria Basin, where some tributaries can connect with those of the Orinoco Delta during flooding. These connections may facilitate the exchange of water and aquatic species between the Orinoco and the Gulf of Paria, influencing the biodiversity and ecological dynamics of the entire region. Given this context, the presence of matamata turtles in Sucre State is plausible. Nevertheless, our current understanding of the chelonian fauna in this region remains largely unknown, highlighting the necessity for further research to explore turtle diversity in this area. CASIQUIARE RIVER CANAL (AMAZONAS STATE) RECORD Morphological shell shape data from museum records indicate the occurrence of both matamata lineages in the Upper Orinoco and Upper Rio Negro Basins within the Amazonas State in southern Venezuela. However, it remains unclear where along this hydrographic gradient the two lineages coexist (Pritchard & Trebbau 1984, Sánchez-Villagra et al. 1995, Barrio-Amorós & Narbaiza 1999, Pritchard 2008). Here, I report about one record from the Casiquiare drainage, the only contact zone between the Orinoco and Rio Negro Basins. A large-sized specimen, with burned carapace and missing plastron is documented, measuring 48.4 cm in SCLmax and 36.9 cm in SCW (Fig. 11). This incomplete shell was discovered in the community of San Carlos de Rio Negro (Fig. 2; black square7)after being consumed by locals.The locals reported capturing the turtle along the Casiquiare River, though no specific collection site was provided (C. Alvarado. pers. comm.). This matamata is housed at MBUCV with catalogue number 7278 and identified as C. orinocensis. A closer examination and using shell shape outline as a diagnostic character for taxon identification, indicates that the specimen has a more rectangular shape rather than oval (Fig. 10). Based on this feature, the specimen is identified tentatively as an Amazonian morphotype (C. fimbriata) (A. Rhodin and T. Escalona, pers. obs.) until further morphological investigation (Fig. 10). This finding is particularly intriguing, as current genetic studies restrict C. fimbriata to the Amazon River Basin and eastern Guianas (Vargas-Ramirez et al. 2020). However, before making any conclusion, caution is warranted as this specimen may also represent an interspecific hybrid or intergrade, considering it originates from a locality where both matamata lineages are suggested to co-occur. Indeed, previous investigations have mentioned that some specimens from this region exhibit a blend of traits from the two Chelus morphotypes, making them difficult to identify, and implying a subspecific relationship or intermediate morphotype between C. orinocensis and C. fimbriata (Sánchez-Villagra et al. 1995, Pritchard 2008). Considering all these factors, it is reasonable to suggest that the Casiquiare River Canal may serve as a dispersal corridor for matamata turtles between the Orinoco and Rio Negro Basins and vice versa, facilitating habitat sharing and coexistence among Chelus lineages along this contact zone. The Casiquiare River’s role as an important biogeographic corridor in shaping species distribution, genetic flow, and speciation processes is well-documented for diverse aquatic fauna, such as in Amazon river dolphins, Podocnemididae turtles (Emmons & Feer 1999, Pearse et al. 2006, Escalona et al. 2009b), particularly for fish species of the genus Cichla (Winemiller & Willis 2011). Additionally, evidence of interspecific hybridization, as seen in various fish studies (Crampton et al. 2003, Willis et al. 2010, 2012, Thomaz et al. 2017), underscore the river’s function in facilitating genetic mixing between the Amazon and Orinoco Basins. However, the extent to which this hybridization or intergrade has occurred in other aquatic organisms, such as matamata turtles, remains understudied. It is important to emphasize, that hybridization among chelonian species have been previously reported in other hydrographic regions (e.g., Lovich et al. 1990, Arantes et al. 2020). If this is the case of Chelus sp., it demands rigorous investigation in order to help mitigate taxonomic ambiguity, as evidenced by similar issues documented in Australian chelids (Cann & Legler 1994, Spinks et al. 2015). Lastly, expanding the discovery of new matamata specimens is crucial for comprehensively understanding the distribution, genetic diversity, and morphological variation of Chelus lineages within and among these river basins.

Published as part of Escalona, Tibisay, 2024, New and unusual field records of Chelus spp. in Venezuela (Testudines: Chelidae), pp. 52-63 in Anartia (Oxford, England) (Oxford, England) 38 on pages 53-62, DOI: 10.5281/zenodo.13755882