Editorial Type: ARTICLES
 | 
Online Publication Date: 25 Apr 2024

Morphometric Variation in the Red-Cheeked Mud Turtle (Kinosternon cruentatum) and its Taxonomic Implications

and
Article Category: Research Article
Page Range: 13 – 22
DOI: 10.2744/CCB-1589.1
Save
Download PDF

Abstract

The Scorpion Mud Turtle (Kinosternon scorpioides) has been recognized by most recent authors to be a single species (3 subspecies) ranging from Tamaulipas, Mexico, to northern Argentina. However, recent molecular analyses have demonstrated that it is not monophyletic, but rather paraphyletic relative to other Neotropical Kinosternon. Based on extensive genetic sampling across the range of this species, a recent paper elevated the 3 subspecies (albogulare, cruentatum, and scorpioides) to species status, confirmed the divergence of Pacific and Atlantic versant populations of K. cruentatum, and recommended that both be given species status. However, the type locality of K. cruentatum is imprecise, making the allocation of the name problematic. Our study sought to determine the provenance of the type using morphometric analysis of specimens from across the range. That analysis demonstrates unequivocally that the type of K. cruentatum was collected on the Atlantic versant, likely from an eastern population. Furthermore, the analysis also revealed that the type of K. mexicanum, previously synonymized with K. cruentatum, was collected from along the Pacific versant. Hence, the name K. cruentatum should be restricted to populations in Atlantic drainages, and the name K. mexicanum should be restricted to Pacific drainages. Our analysis also indicated divergence among the 3 allopatric Atlantic versant populations, but future genetic work will be needed to determine whether they merit taxonomic recognition.

Keywords: Kinosternidae; Kinosternon mexicanum ; Kinosternon scorpioides: Mexico; taxonomy

Over more than 50 yrs, the Scorpion Mud Turtle (Kinosternon scorpioides) has been considered by most authors to be a single species ranging from Tamaulipas, Mexico, to northern Argentina, the greatest distribution of any living turtle (Turtle Taxonomy Working Group [TTWG] 2021). Fifteen synonyms of K. scorpioides (sensu lato [s.l.]) were subsequently described, but only 4 have been recognized as subspecies in the recent literature, based on Berry’s (1978) morphometric analyses: K. s. scorpioides from Panama to Argentina; K. s. albogulare from Panama to Guatemala; K. s. abaxillare from the Central Valley of Chiapas, Mexico, and western Guatemala; and K. s. cruentatum from northern El Salvador, Guatemala, and Belize to Oaxaca and Tamaulipas, Mexico (Berry and Iverson 2001; Legler and Vogt 2013; TTWG 2021; Fig. 1). However, some authors have recognized cruentatum and/or albogulare as full species (Schmidt 1941; Davis 1953; Duellman 1965; Wermuth and Mertens 1977; McCranie 2018), but without substantive morphometric or genetic data to support those actions.

Figure 1.Figure 1.Figure 1.
Figure 1. Distribution of Kinosternon cruentatum (sensu stricto; red shading) and Kinosternon mexicanum (blue shading) in Mexico and Central America, demonstrating the allopatry of the Pacific, eastern Atlantic, Veracruz, and Tampico embayment populations. Open circles are verified locality records (TTWG 2021). Green star is the restricted type locality of both K. mexicanum (and K. cruentatum sensu lato; see text); yellow star is our newly restricted type locality of K. cruentatum.

Citation: Chelonian Conservation and Biology: Celebrating 25 Years as the World’s Turtle and Tortoise Journal 23, 1; 10.2744/CCB-1589.1

Based on 3 mitochondrial and 3 nuclear DNA markers, Iverson et al. (2013) discovered that K. scorpioides (s.l.) was not monophyletic, and because K. s. abaxillare was allopatric relative to the other subspecies, as well as morphometrically and genetically distinct, they recommended its elevation to species level (as proposed earlier by Wermuth and Mertens 1961 and Alvarez del Toro 1973, 1982). That analysis also found the remaining 3 subspecies to be paraphyletic, but made no other taxonomic recommendations for the group pending further research. Two subsequent molecular phylogenetic analyses relying on 14 to 17 nuclear markers (Spinks et al. 2014; Thomson et al. 2021) did not include abaxillare among their samples, but also found K. scorpioides to be paraphyletic, although they also made no taxonomic recommendations.

More recently, Hurtado-Gómez et al. (2024) undertook another molecular analysis of the genus Kinosternon using 3 mitochondrial and up to 17 nuclear markers, but with extensive geographic sampling within K. scorpioides (and Kinosternon leucostomum). That study clearly revealed that K. scorpioides (s.l.) was polytypic, and supported the recognition of K. cruentatum, albogulare, abaxillare, and scorpioides as full species. Hurtado-Gómez et al. (2024) also identified several additional lineages in South and Central America that may eventually merit taxonomic recognition.

Their analysis also revealed that populations of both K. cruentatum and K. albogulare on the Pacific versant were highly divergent from those on the Atlantic (e.g., 2.9% uncorrected p distance for the mitochondrial cytochrome b fragment for the K. cruentatum populations). Hurtado-Gómez et al. (2024) made no taxonomic recommendations regarding lineages within K. albogulare, but recommended that the Atlantic and Pacific populations of K. cruentatum were divergent enough to warrant species recognition. Interestingly, Legler and Vogt (2013) recently mentioned that K. s. cruentatum from Pacific populations have larger, and relatively wider and higher shells than those on the Atlantic. Nevertheless, if the recommendation of Hurtado-Gómez et al. (2024) to split Atlantic populations of K. cruentatum from the Pacific population is confirmed by morphometric analysis, this will result in a complication of nomenclature. Two of its current synonyms (K. mexicanum and K. triliratum) lack precise type localities, a situation made even more complex because Smith and Taylor (1950a, 1950b) restricted the type localities of all 3 taxa to San Mateo del Mar, Oaxaca, Mexico, without published justification. Hence, the appropriate names for the Pacific and Atlantic populations hinge on determining on which versant the types were collected.

The purposes of this study were 1) to examine morphometric variation within K. s cruentatum to determine whether the Pacific and Atlantic populations differed significantly (i.e., do these populations corroborate the genetic data of Hurtado-Gómez et al. 2024); and if so, 2) to identify taxonomic characters that diagnose the 2 populations (proposed species); and 3) to attempt to establish the versant origin of the types of K. cruentatum and K. mexicanum and thus restrict the correct names to the appropriate population(s).

METHODS

The following measurements for specimens of K. (s.) cruentatum from Berry’s (1978) doctoral dissertation were examined in this study: maximum carapace length, maximum plastron length, maximum carapace width, maximum shell height, bridge length, gular length, interhumeral seam length, interpectoral seam length, interabdominal seam length, interfemoral seam length, and interanal seam length. Most measurements were taken to the nearest millimeter by J.F.B. (see Berry 1978 for precise methods of measurement), although those for an additional 25 specimens from the Atlantic versant were measured by J.B.I. (listed in Supplemental Data S1; all supplemental material is available at http://dx.doi.org/10.2744/CCB-1589.1.s1) and included in our analyses. Our analysis did not include data from populations of the Rio Lempa basin in San Salvador, Honduras, and Guatemala that Berry (1978) considered as potential intergrades with albogulare. The full morphometric data sets used in this study are included in the Supplemental Data S1. The latter 10 variables were standardized for size by dividing each by carapace length. We are well aware of the controversy surrounding the use of ratios in statistical analyses (e.g., Atchley 1976; Atchley et al. 1975,), but concordance of the statistical results based on ratios vs. size-standardized residuals on these very data was established by Berry (1978).

Based on the lack of significant morphometric variation across Atlantic drainages revealed by Berry (1978), all Atlantic data were initially lumped into a single sample for our initial analyses. Similarly, all Pacific drainage data were lumped into a single sample. These preliminary analyses revealed substantial variation across the 3 allopatric Atlantic subpopulations of K. cruentatum (Fig. 1): the Yucatan region (from Belize across the Yucatan peninsula), the central Veracruz region, and the Tampico embayment. Our subsequent analyses examined morphometric variation among these 3 subpopulations.

Population comparisons of character ratios were made with t-tests and Mann-Whitney U-tests (i.e., Pacific vs. Atlantic) and employed the STATVIEWtm statistical package (Abacus Concepts, Berkeley, CA). These statistical analyses were used primarily to identify taxonomically useful characters that demonstrated statistically significant differences among populations. Mean ratios (± 1 SD) were calculated by sex and subpopulation. Hence, initial variables (character states) identified as differing between Pacific and Atlantic populations were recombined into new character-state ratios that were useful in diagnosing the 2 populations. Bivariate plots of these character states by sex were produced to clarify the population differences and facilitate the taxonomic diagnosis of future specimens.

Discriminant function analyses (DFA; SPSS BASE v.10.0) were performed (males and females separately) on the 10 size-standardized ratio variables at 3 levels. First, the Pacific population was compared to the 3 Atlantic populations collectively; second, all 4 allopatric subpopulations identified above were compared, and third, only the 3 Atlantic versant subpopulations were compared. Data from the holotypes of K. mexicanum and K. cruentatum (measured by J.B.I.) were included in all sex-appropriate analyses, but without a drainage designation, for a posteriori classification. The type of K. triliratum has been lost (Malnate 1971), and hence could not be included in this study.

RESULTS

Character ratio analyses (Tables 1–5) demonstrated considerable difference between the Pacific and Atlantic populations. Turtles from Pacific slope populations have, on average, longer carapace and plastron lengths, greater carapace widths and height, and shorter interabdominal and interanal seams than turtles from Atlantic slopes (Tables 1–5; Figs. 2–4). Furthermore, this analysis indicated that the female type of K. cruentatum was nested among Atlantic coast females (Table 5; Figs. 2–3), and that the male type of K. mexicanum was nested among Pacific coast males (Table 4; Fig. 2–3).

Figure 2.Figure 2.Figure 2.
Figure 2. Bivariate plot of character ratios useful in discriminating Pacific from Atlantic populations of Kinosternon cruentatum (males above; females below). Abbreviations are IAN (interanal seam length), SH (shell height), CW (carapace width), and PL (maximum plastral length). Male type (0.72, 0.62) is Kinosternon mexicanum (see also Table 4); female type (0.68, 0.65) is K. cruentatum (see also Table 5).

Citation: Chelonian Conservation and Biology: Celebrating 25 Years as the World’s Turtle and Tortoise Journal 23, 1; 10.2744/CCB-1589.1

Figure 3.Figure 3.Figure 3.
Figure 3. Bivariate plot of character ratios useful in discriminating Pacific from Atlantic populations of Kinosternon cruentatum (males above; females below). Abbreviations are IAB (interabdominal seam length), SH (shell height), CW (carapace width), and PL (maximum plastral length). Male type (0.72, 0.56) is Kinosternon mexicanum (see also Table 4); female type (0.68, 0.60) is K. cruentatum (see also Table 5).

Citation: Chelonian Conservation and Biology: Celebrating 25 Years as the World’s Turtle and Tortoise Journal 23, 1; 10.2744/CCB-1589.1

Figure 4.Figure 4.Figure 4.
Figure 4. Photographs of the plastron of a typical female Kinosternon cruentatum from the Atlantic versant (Manuel, Tamaulipas; upper left; photo by J.B.I.), that of the holotype of K. cruentatum (MNHN 1759; upper right; courtesy of Roger Bour), that of a typical male Kinosternon mexicanum (Suciate, Chiapas; lower left; courtesy of Eduardo Reyes-Grajales), and that of the type of K. mexicanum (ANSP 90; lower right; photo by J.B.I.). Note the relatively longer interabdominal and interanal seams of K. cruentatum (see text).

Citation: Chelonian Conservation and Biology: Celebrating 25 Years as the World’s Turtle and Tortoise Journal 23, 1; 10.2744/CCB-1589.1

Table 1. Comparisons of body size (carapace length [CL]) by sex for subpopulations of Kinosternon cruentatum (sensu lato) based on data from Berry (1978) and additional specimens measured by Iverson (J.B.I., unpubl. data).
Table 1.
Table 2. Character comparisons of Pacific and Atlantic populations of female Kinosternon cruentatum.
Table 2.
Table 3. Character comparisons of Pacific and Atlantic populations of male Kinosternon cruentatum. See Table 2 footnote for character abbreviations.
Table 3.
Table 4. Character ratios most useful in discriminating Pacific versus Atlantic populations of male Kinosternon cruentatum, along with measurements of the male holotype of K. mexicanum. Comparisons made with Student’s t-test (t) and Mann-Whitney U test (Z). See Table 2 footnote for character abbreviations.
Table 4.
Table 5. Character ratios most useful in discriminating Pacific vs. Atlantic populations of female Kinosternon cruentatum, along with measurements of the female holotype of K. cruentatum. Comparisons made with Student’s t-test (t) and Mann-Whitney U-test (Z). See Table 2 footnote for character abbreviations.
Table 5.

For the initial Pacific vs. Atlantic DFA analyses of both males and females, 100% of the variance was explained by the first canonical axis. The canonical plots clearly demonstrated that populations of K. cruentatum from the Pacific and Atlantic versants differed significantly in morphometric hyperspace (Fig. 5). The DFA for males grouped the male type of K. mexicanum with turtles from the Pacific slope with a probability of 100%, and the DFA for females grouped the female type of K. cruentatum with turtles from the Atlantic slope with the same probability.

Figure 5.Figure 5.Figure 5.
Figure 5. Distribution of canonical discriminant function (DF) scores on the first canonical axis (representing 100% of the variance) for female (above) and male (below) specimens of Kinosternon cruentatum from the Pacific and Atlantic drainages of Mesoamerica, including the holotype of K. cruentatum among the males, and the holotype of Kinosternon mexicanum among the females (both in green). Holotypes were unassigned a priori and classified a posteriori. DF equation for males: Score = 3 × PL/CL + 13 × GL/CL + 8 × IH/CL + 27 × IP/CL + 30 × IAB/CL + 15 × IF/CL + 47 × IAN/CL − 20 × SH/CL + 17 × BL/CL − 24 × CW/CL − 10.605. DF equation for females: Score = −24 × PL/CL + 23 × GL/CL + 25 × IH/CL + 25 × IP/CL − 12 × IAB/CL + 1 × IF/CL − 18 × IAN/CL + 15 × SH/CL − 4 × BL/CL + 36 × CW/CL − 6.707. CL = maximum carapace length; PL = maximum plastron length; CW = maximum carapace width; SH = maximum shell height; BL = bridge length; GL = gular length; IH = interhumeral seam length; IP = interpectoral seam length; IAB = interabdominal seam length; IF = interfemoral seam length; IAN = interanal seam length.

Citation: Chelonian Conservation and Biology: Celebrating 25 Years as the World’s Turtle and Tortoise Journal 23, 1; 10.2744/CCB-1589.1

The second set of DFA analyses (comparing the 4 allopatric populations) again demonstrated the morphometric divergence between Pacific and Atlantic populations, but also revealed substantial variation across the Atlantic versant populations (Figs. 6–7). With 100% probability the DFA again placed the male type of K. mexicanum with Pacific coast turtles, and the female type of K. cruentatum with the Atlantic coast turtles, specifically within the eastern Atlantic subpopulation. Finally, the third set of DFA analyses of just the 3 allopatric Atlantic subpopulations revealed that the eastern Atlantic subpopulation was the most divergent of the 3 (especially for females; Figs. 8–9), and again placed the holotype of K. cruentatum among the eastern Atlantic specimens.

Figure 6.Figure 6.Figure 6.
Figure 6. Distribution of canonical discriminant function scores on the first 2 canonical axes (representing 83.8% and 13.5% of total variation, respectively) for male Kinosternon cruentatum from the Pacific (1; red), eastern Atlantic (2; green), Veracruz (3; dark blue), and Tampico drainages (4; magenta). Group centroids (large yellow squares) are numbered. The male type of Kinosternon mexicanum (light blue square) was ungrouped and classified a posteriori with the Pacific turtles with 100% probability.

Citation: Chelonian Conservation and Biology: Celebrating 25 Years as the World’s Turtle and Tortoise Journal 23, 1; 10.2744/CCB-1589.1

Figure 7.Figure 7.Figure 7.
Figure 7. Distribution of canonical discriminant function scores on the first 2 canonical axes (representing 66.3% and 31.8% of total variation, respectively) for female Kinosternon cruentatum from the Pacific (1; red)), eastern Atlantic (2; green), Veracruz (3; dark blue), and Tampico drainages (4; magenta). Group centroids (large yellow squares) are numbered. The female type of K. cruentatum (light blue) was ungrouped and classified a posteriori with the Pacific turtles with 100% probability (and falls on the plot with cluster 2; coordinates, −0.92, 0.72).

Citation: Chelonian Conservation and Biology: Celebrating 25 Years as the World’s Turtle and Tortoise Journal 23, 1; 10.2744/CCB-1589.1

Figure 8.Figure 8.Figure 8.
Figure 8. Distribution of canonical discriminant function scores on the first 2 canonical axes (representing 88.7% and 11.3% of total variation, respectively) for male Kinosternon cruentatum from the eastern Atlantic (2; red), Veracruz (3; green), and Tampico drainages (4; dark blue). Group centroids (large magenta squares) are numbered.

Citation: Chelonian Conservation and Biology: Celebrating 25 Years as the World’s Turtle and Tortoise Journal 23, 1; 10.2744/CCB-1589.1

Figure 9.Figure 9.Figure 9.
Figure 9. Distribution of canonical discriminant function scores on the first 2 canonical axes (representing 91.1% and 8.9% of total variation, respectively) for female Kinosternon cruentatum from the eastern Atlantic (2; red), Veracruz (3; green), and Tampico drainages (4; dark blue). The female type of K. cruentatum was ungrouped and classified a posteriori with the eastern Atlantic turtles with 100% probability (coordinates, −0.92, 0.72). Group centroids (large light blue squares) are numbered.

Citation: Chelonian Conservation and Biology: Celebrating 25 Years as the World’s Turtle and Tortoise Journal 23, 1; 10.2744/CCB-1589.1

Bivariate plots of the most discriminating charter ratios (Fig. 10–11) also suggest that the Veracruz and Tampico embayment subpopulations are morphometrically similar (despite being separated by the eastern terminus of the Transverse Volcanic Arc in north-central Veracruz), but that they are distinct from the eastern Atlantic populations.

Figure 10.Figure 10.Figure 10.
Figure 10. Bivariate plot of character ratios useful in discriminating among Atlantic populations of male Kinosternon cruentatum. Abbreviations are IAB (interabdominal seam length), IAN (interabdominal seam length), SH (shell height), CW (carapace width), and PL (maximum plastral length).

Citation: Chelonian Conservation and Biology: Celebrating 25 Years as the World’s Turtle and Tortoise Journal 23, 1; 10.2744/CCB-1589.1

Figure 11.Figure 11.Figure 11.
Figure 11. Bivariate plot of character ratios useful in discriminating Atlantic populations of female Kinosternon cruentatum. Abbreviations are IAB (interabdominal seam length), IAN (interabdominal seam length), SH (shell height), CW (carapace width), and PL (maximum plastral length). Female holotype of K. cruentatum falls in the east Atlantic cluster (see also Table 5).

Citation: Chelonian Conservation and Biology: Celebrating 25 Years as the World’s Turtle and Tortoise Journal 23, 1; 10.2744/CCB-1589.1

DISCUSSION

Our morphometric results support the recognition of separate species for the Pacific and Atlantic populations previously referred to under the name K. cruentatum, as recently proposed by Hurtado-Gómez et al. (2024). In addition, our analyses also identified the female holotype of K. cruentatum as originating from the Atlantic versant and that of the male K. mexicanum as originating from Pacific drainages. Given the genetic data provided by Hurtado-Gómez et al. (2024) showing the reciprocal monophyly of turtles from those 2 versants, an estimated divergence time of 2.4 million years, and an uncorrected p distance of 2.9% for the cytochrome b marker, the species recognition of K. cruentatum from the Atlantic and K. mexicanum from the Pacific is strongly supported. Future phylogenetic work will be necessary to evaluate whether K. cruentatum (sensu stricto) exhibits subspecific variation in the Atlantic versant. However, because Veracruz has been a major trade center at least since it was founded by Hernán Cortez in 1519, market specimens should not be included in those analyses (as some were in this study).

The collector and date of collection of the type of K. cruentatum (Muséum national d’Histoire naturelle, Paris, MNHN-RA-0.1759, a female) are unknown, and its type locality was reported as “Amérique septentrionale” (northern America, presumably in contrast to South America; Berry and Iverson 2001; Bour 2004). Hence, as indicated above, the name cruentatum should be restricted to Atlantic populations (the Atlantic Red-cheeked Mud Turtle). This action invalidates the type locality restriction of cruentatum to San Mateo del Mar, Oaxaca, by Smith and Taylor (1950a, 1950b). We propose that the type locality of K. cruentatum be restricted to Pueblo Nuevo X-Can, Quintana Roo, Mexico (20°52′9″N, 87°36′11″W), from which at least 20 specimens are known (University of Utah 9563–80, 11874, and American Museum of Natural History 93244).

The type of K. mexicanum (Academy of Natural Sciences Philadelphia 90, a male) was collected by William S. Pease on an unspecified date in “Mexico” (LeConte 1854). However, it was likely collected or purchased in 1847, when, on behalf of the Philadelphia Academy of Natural Sciences, Pease accompanied the US military march from Veracruz to Jalapa and on to Mexico City during the Mexican–American War (Cassin 1848–1849; Pease 1848–1849, 1849). There is no evidence that the expedition traveled to the Pacific coast, but the holotype could have been purchased in a market. Nevertheless, based on our morphometric data, and pending genetic data from both holotypes, we recommend recognizing Pacific populations as K. mexicanum (the Pacific Red-cheeked Mud Turtle), and retaining the restricted type locality designated by Smith and Taylor (1950a, 1950b).

As noted by Hurtado-Gómez et al, (2024), the biogeographic history of the subgenus Kinosternon (including albogulare, cruentatum, oaxacae, chimalhuaca, and integrum) has been primarily along the Pacific versant of Central America and Mexico. This suggests that the common ancestor of K. cruentatum and K. mexicanum likely dispersed from the Pacific to the Atlantic versant, probably across the relatively low Isthmus of Tehuantepec. From there it apparently migrated northward to the Tampico embayment and eastward as far as the Yucatan peninsula and Belize. That extensive Atlantic range was apparently subsequently fragmented (Fig. 1), perhaps by the eastward orogeny of the Transverse Volcanic Arc in north central Veracruz, and the Tabasco lowlands, where it appears to be absent.

Acknowledgments

Iverson’s research was funded by the American Philosophical Society, the American Museum of Natural History, the Joseph Moore Museum of Earlham College, and the National Science Foundation. The Paris Museum and the Academy of Natural Sciences of Philadelphia generously loaned J.B.I. the type specimens over 40 yrs ago. Berry’s research was funded by the University of Utah and Elmhurst College. Curators of 19 herpetological collections provided access to the material used in this study (full details in Berry 1978). Anders Rhodin generously produced the map of the current range of these populations, updated from TTWG (2021). Eduardo Reyes Grajales and Roger Bour provided photographs.

LITERATURE CITED

  • Alvarez del Toro, M. 1973. Los Reptiles de Chiapas.
    Second edition
    .
    Tuxtla Gutierrez, Mexico
    :
    Gobierno del Estado de Chiapas
    .
  • Alvarez del Toro, M. 1982. Los Reptiles de Chiapas.
    Third edition
    .
    Tuxtla Gutierrez, Chiapas, Mexico
    :
    Publicaciones del Instituto de Historia Natural
    .
  • Atchley, W.R. 1976. Statistical properties of ratios. I. Empirical results. Systematic Zoology25:137–148.
  • Atchley, W.R., Gaskins, C.T., and Anderson,D. 1975. Statistical properties of ratios of biological data. American Zoologist15:289.
  • Berry, J.F. 1978. Variation and systematics in the Kinosternon scorpioides and K. leucostomum complexes (Reptilia: Testudines: Kinosternidae) of Mexico and Central America. PhD Dissertation,
    University of Utah
    ,
    Salt Lake City
    . 326 pp.
  • Berry, J.F. and Iverson,J.B. 2001. Kinosternon scorpioides (Linnaeus 1766). Scorpion Mud Turtle. Catalogue of American Amphibians and Reptiles725:1–11.
  • Bour, R. 2004. Holotype descriptions: Kinosternon cruentatum Duméril et Bibron, 1851. Emys11:32–35.
  • Cassin, J. 1848–1849. Catalogue of birds collected by Mr. Wm. S. Pease, during the march of the Army of the United States from Vera Cruz to the City of Mexico. Proceedings of the Academy of Natural Sciences of Philadelphia4:87–91.
  • Davis, W.B. 1953. The turtle Kinosternon cruentatum cruentatum in Tamaulipas. Copeia1953:65.
  • Duellman, W.E. 1965. Amphibians and reptiles of the Yucatan Peninsula, México. University of Kansas Publications, Museum of Natural History15:577–614.
  • Hurtado-Gómez, J.-P., Vargas-Ramírez, M., Iverson, J.B., Joyce, W., McCranie, J.R., and Fritz,U. 2024. Diversity and biogeography of South American Mud Turtles elucidated by multilocus DNA sequencing (Testudines: Kinosternidae). Molecular Phylogenetics and Evolution. In final review.
  • Iverson, J.B., Le, M., and Ingram,C.M. 2013. Molecular phylogenetics of the mud turtles of the family Kinosternidae. Molecular Phylogenetics and Evolution69:929–939.
  • LeConte, J. 1854. Description of four new species of Kinosternum. Proceedings of the Academy of Natural Sciences of Philadelphia7:180–190.
  • Legler, J.M. and Vogt,R.C. 2013. The Turtles of Mexico: Land and Freshwater Forms.
    Berkeley
    :
    University of California Press
    . 402 pp.
  • Malnate, E.V. 1971. A catalog of primary types in the herpetological collections of the Academy of Natural Sciences, Philadelphia (ANSP). Proceedings of the Academy of Natural Sciences of Philadelphia123(
    9
    ):345–375.
  • McCranie, J.R. 2018. The lizards, crocodiles, and turtles of Honduras: systematics, distribution, and conservation. Bulletin of the Museum of Comparative Zoology15:1–129.
  • Pease, W.S. 1848-1849. Observations on the geology and natural history of Mexico. Proceedings of the Academy of Natural Sciences of Philadelphia4:91–94.
  • Pease, W.S. 1849. Observations on the geology and natural history of Mexico. Annals and Magazine of Natural History Series3(
    17
    ):427–431.
  • Schmidt, K.P. 1941. The amphibians and reptiles of British Honduras. Zoological Series Field Museum of Natural History22:475–510.
  • Smith, H.M. and TaylorE.H. 1950a. Type localities of Mexican reptiles and amphibians. Kansas University Science Bulletin33:313–380.
  • Smith, H.M. and TaylorE.H. 1950b. An annotated check list and key to the reptiles of Mexico exclusive of the snakes. Bulletin of the United States National Museum199:1–253.
  • Spinks, P.Q., Thomson, R.C., Gidiş, M., and Shaffer,H.B. 2014. Multilocus phylogeny of the New-World mud turtles (Kinosternidae) supports the traditional classification of the group. Molecular Phylogenetics and Evolution76:254–260.
  • Thomson, R.C., Spinks, P.Q., and Shaffer,H.B. 2021. A global phylogeny of turtles reveals a burst of climate associated diversification on continental margins. Proceedings of the National Academy of Sciences118:e2012215118.
  • Turtle Taxonomy Working Group (TTWG), (Rhodin,A.G.J.,Iverson,J.B.,Bour,R.,Fritz,U.,Georges,A.,Shaffer,H.B., and van Dijk,P.P.). 2021.
    Turtles of the world: Annotated checklist and atlas of taxonomy, synonymy, distribution, and conservation status
    (
    9th edition
    ). In: Rhodin,A. G. J.,Iverson,J.B.,van Dijk,P.P.,Stanford,C.B.,Goode,E.V.,Buhlmann,K.A., and Mittermeier,R.A. (Eds.). Conservation Biology of Freshwater Turtles and Tortoises.
    Chelonian Research Monographs
    8:1–472.
  • Wermuth, H. and Mertens,R. 1961. Schildkröten, Krokodile, Bruchenechsen.
    Jena
    :
    Gustav Fischer Verlag
    .
  • Wermuth, H. and Mertens,R. 1977. Liste der rezenten amphibien und reptilien: Testudines, Crocodylia, Rhynchocephalia. Das Tierreich100:1–174.
Copyright: © 2024 Chelonian Research Foundation 2024
word
Figure 1.
Figure 1.

Distribution of Kinosternon cruentatum (sensu stricto; red shading) and Kinosternon mexicanum (blue shading) in Mexico and Central America, demonstrating the allopatry of the Pacific, eastern Atlantic, Veracruz, and Tampico embayment populations. Open circles are verified locality records (TTWG 2021). Green star is the restricted type locality of both K. mexicanum (and K. cruentatum sensu lato; see text); yellow star is our newly restricted type locality of K. cruentatum.

Figure 2.
Figure 2.

Bivariate plot of character ratios useful in discriminating Pacific from Atlantic populations of Kinosternon cruentatum (males above; females below). Abbreviations are IAN (interanal seam length), SH (shell height), CW (carapace width), and PL (maximum plastral length). Male type (0.72, 0.62) is Kinosternon mexicanum (see also Table 4); female type (0.68, 0.65) is K. cruentatum (see also Table 5).

Figure 3.
Figure 3.

Bivariate plot of character ratios useful in discriminating Pacific from Atlantic populations of Kinosternon cruentatum (males above; females below). Abbreviations are IAB (interabdominal seam length), SH (shell height), CW (carapace width), and PL (maximum plastral length). Male type (0.72, 0.56) is Kinosternon mexicanum (see also Table 4); female type (0.68, 0.60) is K. cruentatum (see also Table 5).

Figure 4.
Figure 4.

Photographs of the plastron of a typical female Kinosternon cruentatum from the Atlantic versant (Manuel, Tamaulipas; upper left; photo by J.B.I.), that of the holotype of K. cruentatum (MNHN 1759; upper right; courtesy of Roger Bour), that of a typical male Kinosternon mexicanum (Suciate, Chiapas; lower left; courtesy of Eduardo Reyes-Grajales), and that of the type of K. mexicanum (ANSP 90; lower right; photo by J.B.I.). Note the relatively longer interabdominal and interanal seams of K. cruentatum (see text).

Figure 5.
Figure 5.

Distribution of canonical discriminant function (DF) scores on the first canonical axis (representing 100% of the variance) for female (above) and male (below) specimens of Kinosternon cruentatum from the Pacific and Atlantic drainages of Mesoamerica, including the holotype of K. cruentatum among the males, and the holotype of Kinosternon mexicanum among the females (both in green). Holotypes were unassigned a priori and classified a posteriori. DF equation for males: Score = 3 × PL/CL + 13 × GL/CL + 8 × IH/CL + 27 × IP/CL + 30 × IAB/CL + 15 × IF/CL + 47 × IAN/CL − 20 × SH/CL + 17 × BL/CL − 24 × CW/CL − 10.605. DF equation for females: Score = −24 × PL/CL + 23 × GL/CL + 25 × IH/CL + 25 × IP/CL − 12 × IAB/CL + 1 × IF/CL − 18 × IAN/CL + 15 × SH/CL − 4 × BL/CL + 36 × CW/CL − 6.707. CL = maximum carapace length; PL = maximum plastron length; CW = maximum carapace width; SH = maximum shell height; BL = bridge length; GL = gular length; IH = interhumeral seam length; IP = interpectoral seam length; IAB = interabdominal seam length; IF = interfemoral seam length; IAN = interanal seam length.

Figure 6.
Figure 6.

Distribution of canonical discriminant function scores on the first 2 canonical axes (representing 83.8% and 13.5% of total variation, respectively) for male Kinosternon cruentatum from the Pacific (1; red), eastern Atlantic (2; green), Veracruz (3; dark blue), and Tampico drainages (4; magenta). Group centroids (large yellow squares) are numbered. The male type of Kinosternon mexicanum (light blue square) was ungrouped and classified a posteriori with the Pacific turtles with 100% probability.

Figure 7.
Figure 7.

Distribution of canonical discriminant function scores on the first 2 canonical axes (representing 66.3% and 31.8% of total variation, respectively) for female Kinosternon cruentatum from the Pacific (1; red)), eastern Atlantic (2; green), Veracruz (3; dark blue), and Tampico drainages (4; magenta). Group centroids (large yellow squares) are numbered. The female type of K. cruentatum (light blue) was ungrouped and classified a posteriori with the Pacific turtles with 100% probability (and falls on the plot with cluster 2; coordinates, −0.92, 0.72).

Figure 8.
Figure 8.

Distribution of canonical discriminant function scores on the first 2 canonical axes (representing 88.7% and 11.3% of total variation, respectively) for male Kinosternon cruentatum from the eastern Atlantic (2; red), Veracruz (3; green), and Tampico drainages (4; dark blue). Group centroids (large magenta squares) are numbered.

Figure 9.
Figure 9.

Distribution of canonical discriminant function scores on the first 2 canonical axes (representing 91.1% and 8.9% of total variation, respectively) for female Kinosternon cruentatum from the eastern Atlantic (2; red), Veracruz (3; green), and Tampico drainages (4; dark blue). The female type of K. cruentatum was ungrouped and classified a posteriori with the eastern Atlantic turtles with 100% probability (coordinates, −0.92, 0.72). Group centroids (large light blue squares) are numbered.

Figure 10.
Figure 10.

Bivariate plot of character ratios useful in discriminating among Atlantic populations of male Kinosternon cruentatum. Abbreviations are IAB (interabdominal seam length), IAN (interabdominal seam length), SH (shell height), CW (carapace width), and PL (maximum plastral length).

Figure 11.
Figure 11.

Bivariate plot of character ratios useful in discriminating Atlantic populations of female Kinosternon cruentatum. Abbreviations are IAB (interabdominal seam length), IAN (interabdominal seam length), SH (shell height), CW (carapace width), and PL (maximum plastral length). Female holotype of K. cruentatum falls in the east Atlantic cluster (see also Table 5).

Contributor Notes

*Corresponding author

Handling Editor: Jeffrey E. Lovich

Received: 27 Apr 2023
Accepted: 23 Oct 2023
  • Download PDF