Reopening the case of the enigmatic record hitherto referred to as the Mayan shrew, Cryptotis mayensis (Mammalia, Soricidae)
Lázaro Guevara1,* , Reinhard E. Matadamas2,3 , Stephanye Mata-González1,3 , Paola Zeferino1, and Joaquín Arroyo-Cabrales4 .
1Colección Nacional de Mamíferos, Departamento de Zoología, Instituto de Biología, Universidad Nacional Autónoma de México, Apartado Postal 70-153, CP 04510 Mexico City, Mexico. E-mail: matags_bio@ciencias.unam.mx (SMG); E-mail: paola-zeferino@ciencias.unam.mx (PZ)
2Museo de Zoología Alfonso L. Herrera, Facultad de Ciencias, Universidad Nacional Autónoma de México, Apartado Postal 70-153, CP 04510 Mexico City, Mexico. reinhardmata@ciencias.unam.mx (REM).
3Posgrado en Ciencias Biológicas, Universidad Nacional Autónoma de México.
4Instituto Nacional de Antropología e Historia, Subdirección de laboratorio y Apoyo Académico, Laboratorio de Arqueozoología “M. en C. Ticul Álvarez Solórzano”, Moneda Núm. 16, Col. Centro, 06060 Ciudad de México, México. E-mail: arromatu@hotmail.com (JAC)
*Corresponding author: llg@ib.unam.mx
The Mayan small-eared shrew, Cryptotis mayensis, is endemic to the Yucatán Peninsula, with records from México, Belize, and Guatemala. For half a century, skull fragments from dozens of individuals found in pellets of the barn owl (Tyto furcata) in Guerrero, Mexico, previously identified as the Mayan shrew, have intrigued taxonomists and biogeographers. A previous robust analysis of current and fossil C. mayensis dentary samples, including those from Guerrero, showed a high morphometric similarity between them. Given that these owl pellet remains are located 1,000 km from the known distribution of C. mayensis in the Yucatán Peninsula, this has raised the hypothesis that they are not an isolated population of C. mayensis but rather an as-yet-undescribed species. By integrating new specimens of C. mayensis from the Yucatán Península and Guerrero, as well as Cryptotis lacandonensis, the sister species of C. mayensis, we reanalyzed the morphological attributes. In addition, we used paleodistribution estimates to investigate the possible isolation of the records previously referred to as C. mayensis in Guerrero, Mexico. Multivariate analyses of morphological data from the cranium and dentary revealed high similarity in size among the three samples analyzed, especially between C. mayensis sensu stricto and Cryptotis from Guerrero. Paleodistribution models suggest that the population of Guerrero has remained in long-term isolation from the populations of the Yucatan Peninsula due to a very large area with unsuitable conditions for the connectivity of the shrew population during the last glacial-interglacial cycle. It is possible that the Guerrero population is an independent lineage, despite its morphological similarity to C. mayensis sensu stricto; however, genetic evidence to confirm this is essential.
Keywords: Eulipotyphla, morphometrics, Neotropics, niche modelling, taxonomy
La musaraña maya de orejas cortas, Cryptotis mayensis, es endémica de la Península de Yucatán, con registros en México, Belice y Guatemala. Durante medio siglo, fragmentos de cráneo de docenas de individuos, encontrados en egagrópilas de lechuza común (Tyto furcata) en Guerrero, México, que anteriormente se identificaron como la musaraña maya, han intrigado a taxónomos y biogeógrafos. Un exhaustivo análisis previo de muestras dentales actuales y fósiles de C. mayensis, incluidas las de Guerrero, mostró una alta similitud morfométrica entre ellas. Dado que estos restos de egagrópilas se encuentran a 1000 km de la distribución conocida de C. mayensis en la Península de Yucatán, esto ha planteado la hipótesis de que no se trata de una población aislada de C. mayensis, sino de una especie aún no descrita. Al integrar nuevos especímenes de C. mayensis de la Península de Yucatán y de Guerrero, así como de C. lacandonensis, especie hermana de C. mayensis, reanalizamos los atributos morfológicos. Además, utilizamos estimaciones de paleodistribución para investigar el posible aislamiento de los registros previamente atribuidos a C. mayensis en Guerrero, México. Los análisis multivariados de datos morfológicos del cráneo y del dentario revelaron una alta similitud en tamaño entre las tres muestras analizadas, especialmente entre C. mayensis sensu stricto y Cryptotis de Guerrero. Los modelos de paleodistribución sugieren que la población de Guerrero ha permanecido aislada durante largo tiempo de las poblaciones de la Península de Yucatán debido a una extensa área con condiciones desfavorables para la conectividad poblacional durante el último ciclo glacial-interglacial. Es posible que la población de Guerrero sea un linaje independiente, a pesar de su similitud morfológica con C. mayensis sensu stricto; sin embargo, es indispensable contar con evidencia genética que lo confirme.
Palabras clave: Eulipotyphla, modelado de nicho, morfometría, Neotrópico, taxonomía
© 2026 Asociación Mexicana de Mastozoología, www.mastozoologiamexicana.org
Shrews of the genus Cryptotis are small mammals found from southern Canada to northern South America. So far this century, the number of species within this genus has doubled due to intensive taxonomic research, making it the third most diverse genus among the 29 genera of shrews, with 55 species (Mammal Diversity Database 2025). Advances in taxonomic studies have refined previous proposals regarding the evolutionary relationships among several species and established species groups within the genus that may represent monophyletic groups (Guevara and Cervantes 2014; He et al. 2015; 2021). One such group is the Cryptotis nigrescens group, which includes eight species found from southern Mexico to northern Colombia (Woodman and Timm 2023): C. mayensis, C. merriami, C. lacandonensis, C. nigrescens, C. hondurensis, C. merus, C. brachyonyx, and C. colombianus.
Within the Cryptotis nigrescens group, C. mayensis (the Mayan shrew) is the most boreal in distribution and endemic to the Yucatán Peninsula, with records from Mexico, Belize, and Guatemala (Lorenzo et al. 2019; Woodman 2019). Cryptotis mayensis was described based on one specimen “from a Maya ruin at Chichén Itzá, Yucatan”, collected on February 5, 1901, by E. W. Nelson and E. A. Goldman (Merriam 1901; Woodman and Timm 1993), and its specific epithet “mayensis” refers to the predominant Mayan culture in the region (Carraway 2007; Guevara and Ramírez-Chaves 2025). It remained a little-known species until more specimens were collected decades later (Choate 1970). Nowadays, C. mayensis is distributed in lowland areas, usually at <100 m elevation, although a noteworthy record is from the Chiquibul Forest Reserve, Belize, at about 600 m elevation (Engilis et al. 2012). It inhabits lowland tropical forests, characterized by dry scrub, deciduous, and seasonally dry forests (Carraway 2007). Cryptotis mayensis has also been reported from Pleistocene fossils in the Loltún and Actun Spukil caves, Yucatán (Hatt 1953). The species most closely related to C. mayensis is C. lacandonensis (the Lacandona shrew), whose distribution is restricted to the Selva Maya on the border between Mexico and Guatemala, in humid tropical forests and seasonally flooded areas at elevations below 200 m (Guevara et al. 2014; Pérez et al. 2019).
For half a century, mandibles and incomplete skulls from almost 40 individuals found in pellets of the barn owl (Tyto furcata; previously T. alba) in Guerrero have intrigued taxonomists and biogeographers (López-Forment and Urbano 1977). The pellets were collected by William López-Forment in September 1969, within a stony and arid area surrounding the Cueva del Zopilote, located 13 km south of the Mezcala Bridge (Figure 1; Ramírez-Pulido and Sánchez-Hernández 1972). Based on the similarity of dental features and cranial and dental size, the preliminary assignment of the owl pellet was made to C. mayensis (Choate 1970; Woodman and Timm 1993). Their anatomical characteristics allowed these remains to be classified as belonging to the C. nigrescens group and not to the C. goldmani or C. parvus groups, which may also inhabit that region of Guerrero (Guevara et al. 2015). For example, the shape of the dentary, the location of the zygomatic plate, and the cusps on the lower and upper third molars are typical of what is observed in C. mayensis (Woodman and Timm 1993). Woodman (1995) analyzed current and fossil C. mayensis dentary samples, including those from Guerrero, to test the hypothesis that phenotypic changes may be a possible response to Quaternary climatic factors. He found a high degree of morphometric similarity between the modern samples from Yucatán and Guerrero, which differ slightly from the fossil samples. Given that these remains from owl pellets are located about 1,000 km from the known distribution of C. mayensis in the Yucatán Peninsula, this has raised the hypothesis that they are not from an isolated population of C. mayensis, but rather from an as-yet-undescribed species (Woodman and Timm 1993; Monroy-Gamboa 2021). The reason for this remarkable geographic isolation has been intuitively attributed to Quaternary environmental changes that have eliminated suitable habitat for this group of shrews between Guerrero and the Yucatán Peninsula, leaving a small population in Guerrero completely isolated (Choate 1970; Woodman and Timm 1993).
To our knowledge, no additional samples or complete specimens of these shrews have been obtained in Guerrero so far, and conducting fieldwork is currently unfeasible due to the high and growing insecurity in that region of the country (Carpio-Domínguez 2021), which has hindered the resolution of the enigma of C. mayensis records in Guerrero. In addition, genetic data remain limited for Neotropical Cryptotis, which complicates progress in resolving complex taxonomic questions. Over the last two decades, our team and collaborators have consulted with several natural history museums to refine databases for the genus Cryptotis, thereby obtaining high-quality morphological and geographic data (Guevara et al. 2024). This raises the possibility of reopening old taxonomic and biogeographical questions and addressing them with consideration of new evidence.
Hence, by integrating morphological and distributional analyses, we aim to reexamine the origin of the records referred to as Cryptotis mayensis in Guerrero, Mexico, in a more in-depth manner. To do so, (1) we collected more morphological information on populations of interest to assess morphological variation, and (2) we modeled the environmental requirements for C. mayensis in the Yucatán Peninsula to estimate its distributional pattern across the Late Quaternary. We predict that including more data will reveal significant differences between samples from the Yucatán Peninsula and Guerrero, suggesting long-term geographic isolation and divergence at a specific level.
Materials and methods
Specimens examined. All specimens examined in the present study are housed in the following collections: Colección Nacional de Mamíferos, Instituto de Biología, Universidad Nacional Autónoma de México, México City (CNMA); Colección Zoológica Regional Mammalia, Instituto de Historia Natural y Ecología (CZRMA); Colección Mastozoológica, El Colegio de la Frontera Sur, San Cristóbal de las Casas (ECO-SC-M); The University of Kansas Natural History Museum, Lawrence (KU); Museo de Zoología “Alfonso L. Herrera”, Facultad de Ciencias, Universidad Nacional Autónoma de México, México City (MZFC), Colección de Mamíferos de la Universidad, Autónoma de Yucatán, Mérida (UADY), the National Museum of Natural History, Smithsonian Institu-tion, Washington (USNM), and Colección de Mamíferos, Universidad Veracruzana, Xalapa (UV).
We examined 25 Cryptotis mayensis sensu stricto specimens from the Yucatan Peninsula (CNMA 23796-23801; ECO-SC-M 2354-2356, 2358; KU 91463, 143892; UADY 833-840; USNM 170862, 108087, 173000; UV 3535, 3758), 37 skull remains recovered from the owl pellets in Guerrero (CNMA 11031-11043, 12666-12688, 12690), and two specimens of its sister species, C. lacandonensis (MZFC 7107, 7168) to gain a better understanding of the possible morphological variation within this clade. It is important to highlight that the material available for C. lacandonensis and C. merriami in Mexico is extremely scarce (Lorenzo et al. 2019; Guevara et al. 2024).
Morphometric analyses. Given the fragmentary nature of owl pellets and the difficulty of recovering complete material from all individuals, we performed separate morphometric analyses of the skull and mandibles. We obtained measurements of 20 skulls from the Yucatan Peninsula, 34 from Guerrero, and two from C. lacandonensis. For mandibles, we measured 20 right dentaries from the Yucatan Peninsula, seven right dentaries from Guerrero, and two right dentaries from C. lacandonensis.
We recorded four measurements of the cranium and five of the mandible (Figure 2) using a Mitutoyo electronic caliper, 500-171-20 model, at 0.1 mm precision under a stereomicroscope: interorbital breadth (IO), breadth of palate across second molars (M2B), breadth of zygomatic plate (ZP); length of molariform toothrow (MTR), height of coronoid process (HCP), the height of articular condyle (HAC), height of coronoid valley (HCV), lower toothrow length (TRD), distance from the upper border of the articular condyle to the posterior edge of the m3 (LAM). Tabular univariate statistics include mean ± SD and total range for each variable (Table 1).
To characterize and evaluate overall size variation across all specimens, we performed principal component analyses (PCA). To explore the cohesion of known species and samples from Guerrero and identify variables that distinguish them, we used discriminant function analyses (DFA). We performed multivariate analyses on correlation matrices of log10-transformed variables for the cranium and mandible separately, using Statistica software (StatSoft Inc. 2005).
Potential distribution over time. To investigate the possible long-term geographic isolation of Cryptotis from Guerrero, we modeled the climatic niche of C. mayensis sensu stricto. Based on the hypothesis of niche conservatism, we assume that the environmental requirements of the samples from Guerrero are similar to those of C. mayensis s.s., which will allow us to predict its distribution and, therefore, understand the possible dynamics in distribution during the last glacial-interglacial cycle. Hence, we estimated the potential distribution of C. mayensis s.s. through time, using the present conditions, the Mid-Holocene, and the Last Glacial Maximum (LGM), as climatic extremes during the Quaternary. Ecological niche models (ENMs) and cross-time projections were optimized using the ‘kuenm’ package (Cobos et al. 2019) and the Maxent modeling algorithm (Phillips et al. 2017) in R software version 4.4.2 (R Core Team, 2025).
Our initial database was obtained from Guevara et al. (2024), which contains 118 unique localities and represents the most up-to-date and refined database available for C. mayensis. To reduce bias introduced by field sampling strategies, we spatially filtered the full set of species occurrence records using the spThin package (Aiello-Lammens et al. 2015). We retained only records that were at least 5 km apart, resulting in a final data set of 22 unique locations for calibrating and evaluating niche models. For further analysis, we spatially partitioned our data (i.e, checkerboard) into 70% for model calibration and 30% for model evaluation.
As environmental predictors, we initially selected 19 bioclimatic variables from the WorldClim v2.1 portal (Fick and Hijmans 2017). These variables describe various aspects of temperature and precipitation and were downloaded at a 30-second resolution for the period from 1970 to 2000. All bioclimatic layers were downloaded from the terra package using WorldClim version 2.1 climate data for the period 1970-2000 at a 30-second resolution. Then, we computed Pearson’s correlation coefficient for each pairwise comparison to assess whether the variables are highly correlated and to decide which, if any, should be removed. Thus, we retained the variables BIO4 (temperature seasonality (standard deviation ×100)), BIO6 (minimum temperature of coldest month), BIO11 (mean temperature of coldest quarter), and BIO16 (precipitation of wettest quarter) because they showed lower correlations with the other variables in the calibration area (r ≤ 0.8).
For model calibration, we delimited an area by combining ecoregions (Dinerstein et al. 2017) and biogeographic provinces (Morrone 2014) that encompass our unique locations, excluding areas unlikely to be accessible to the species given their dispersal abilities (Barve et al. 2011). To obtain a high representation of the environments available in the study area, we randomly sampled ~195,000 pixels within the delimited area (50% of the total area; Guevara et al. 2018). To identify model settings approximating optimal levels of complexity, we constructed niche models with a wide variety of combinations of feature classes (Linear; Linear and Quadratic; Linear, Quadratic, and Product; Linear, Quadratic, Product, and Threshold; Linear, Quadratic, Product, Threshold, and Hinge) and regularization multipliers (0.4 – 6 at 0.25 intervals). This process yielded 115 unique FC-RM combinations. To select the optimal settings for building a final Maxent model, we examined the omission rate (OR5) at a 5% training omission rate threshold and the area under the curve (AUC) for testing points. These two metrics help evaluate performance with respect to overfitting and discrimination, respectively.
We transferred the final model onto the present, Mid-Holocene, and LGM climatic conditions to estimate potential distributions over the last interglacial–glacial cycle across Central America (i.e., from Mexico to Panama). We used unconstrained extrapolation for model projection, disabling the default ‘clamping’ option in Maxent (Guevara et al. 2018). To consider the variability in past climatic reconstructions (Guevara et al. 2019), we used three different General Circulation Models (GCMs): the Community Climate System Model (CCSM4), the Model for Interdisciplinary Research on Climate (MIROC-ESM), and the Max-Planck-Institute für Meteorologie (MPI). We downloaded past bioclimatic layers from the WorldClim database at a resolution of 2.5 arc-min. To identify uncertainty due to model extrapolation, we created a Multivariate Environmental Similarity Surface (MESS) map, which helps detect areas where non-analog conditions occur.
Finally, to display the final maps for the present and the paleoclimatic scenarios, we classified the continuous model output into suitable and unsuitable conditions using the 10th percentile training presence threshold. This threshold excludes the lowest 10% of suitability scores from locations where the species is known to occur (Merow et al. 2013). We retained the suitability values in the original continuous format above this threshold to show the suitability gradient, and we also converted the original model output into a binary prediction of suitable-unsuitable areas. Using binary maps (0 = unsuitable; 1 = suitable), we summed the present and past potential distribution maps (for each of the three GCMs) to identify regions that have remained climatically suitable for C. mayensis.
Results
Considering the fragmentary origin of the owl pellet remains from Guerrero and the limitations this entails, they are very similar to Cryptotis mayensis sensu stricto. The zygomatic plate is quite broad (ZP = 2.35 ± 0.14 mm); the anterior border of the zygomatic plate runs from the parastyle-mesostyle valley to the mesostyle of M1, and the posterior border runs from the metastyle of M2 to the middle of M3. The mandible is moderately long (LAM = 4.69 ± 0.29 mm), and the coronoid process of the mandible is high (HCP = 5.34 ± 0.23 mm); the anterior border of the mandible joins the horizontal ramus at approximately a right angle.
Principal component analysis. For the analysis of the skulls, the principal components 1 and 2 explained 78.6% of the variation in the data (PC 1 = 60.6% and PC 2 = 18.0%) and were primarily correlated with the length of molariform toothrow (MTR) and the breadth of zygomatic plate (ZP), respectively. In the plot of 1st PCs from the PCA, the specimens of Lacandona shrew tended to plot among the largest in the sample (Figure 3a). Regarding the analysis of the dentary, the principal components 1 and 2 explained 70.6% of the variation in the data (PC 1 = 55.6% and PC 2 = 15.0%) and were primarily correlated with the height of articular condyle (HAC) and distance from the upper border of the articular condyle to the posterior edge of the m3 (LAM), respectively. A plot of the 1st PCs from the PCA highlights again the difference in size of C. lacandonensis with respect to C. mayensis sensu stricto and the samples from Guerrero, which tend to show a very similar pattern of variation in size (Figure 3). In general, the PCA analyses indicate that the specimens of C. mayensis sensu stricto and Cryptotis from Guerrero exhibit a similar pattern of morphometric variation and differ from C. lacandonensis.
Discriminant function analysis. For the analysis of the crania, the 1st function accounts for over 80.7% of the explained variance (Table 2), which is weighted most heavily by the length of molariform toothrow (MTR), contrasted with the interorbital breadth (IO). A plot of the two canonical axes failed to show a clear separation of the samples in multivariate space. The DFA based on the four cranial variables did not efficiently discriminate the Guerrero specimens from the other species (Wilks’ Lambda = 0.652; F8,66 = 1.9626; P < 0.0651). In addition, the DFA of the four cranial variables produced moderate to null correct classification rates of 75%, 53%, and 0% for C. mayensis sensu stricto (n = 20), Cryptotis from Guerrero (n = 17), and C. lacandonensis (n = 2), respectively, demonstrating the high morphometric similarity between the samples. In general, the length of molariform toothrow (MTR) is the measure that most helps distinguish samples, but only between the Lacandona shrew from the other two groups.
On the other hand, for the analysis of the mandibles, the 1st function accounts for over 85% of the explained variance (Table 2), which is weighted most heavily by the lower toothrow length (TRD), contrasted with the height of the articular condyle (HAC). A plot of the two canonical axes shows that the specimens of the Lacandona shrew are slightly separated on the first axis due to their larger size. In contrast, the second axis does not allow a clear separation between the three samples (Figure 3d). The DFA based on the five variables showed better performance than the one based only on the skull for distinguishing the a priori groups (Wilks’ Lambda = 0.423; F10,44 = 2.3621, P < 0.0246). In addition, the DFA produced correct classification rates of 100%, 40%, and 50% for C. mayensis sensu stricto (n = 22), Cryptotis from Guerrero (n = 5), and C. lacandonensis (n = 2), respectively, suggesting that this structure may have a greater contribution to distinguishing between taxa. In general, lower tooth length (TRD) in isolation is the measure that most helps distinguish samples, but again, only between the Lacandona shrew and the other two taxa (C. lacandonensis: 5.73 ± 0.16 mm; C. mayensis s.s.: 5.37 ± 0.19 mm; Guerrero: 5.36 ± 0.27 mm).
In sum, considering the small sample size of C. lacandonensis, multivariate analyses of the cranium and dentary indicate high similarity in size across the three samples analyzed, especially between C. mayensis sensu stricto and Cryptotis from Guerrero. Cryptotis lacandonensis is distinguishable from the other two by its larger size, which can be more easily detected by measuring both the upper and lower tooth rows (Table 1).
Biogeographic analyses. Of the 115 niche models generated for C. mayensis sensu stricto, four stood out for their better performance based on significance, omission rate, and AICc. Although these four models did not show significant differences, we present the output of the model that performed slightly better. This model was implemented using the following feature classes and regularization multiplier: Linear and Quadratic, with a value of 1.2 (AUC: 0.829; mean AUC ratio: 1.8392; omission rate at 5%: <0%). Of the four variables used to build the final model, precipitation of the wettest quarter (BIO 16) was the most important, indicating a steep decline in climatic suitability when precipitation exceeds 400 mm. Accordingly, the model suggests a current potential distribution that is closely linked to the distribution of warm-humid and warm-dry forests of the Yucatan Peninsula and the Pacific slope, including the area surrounding the Cueva del Zopilote, Guerrero, where the owl pellets were collected. Between the estimated suitable conditions for C. mayensis s.s. on the Yucatan Peninsula and those towards the Pacific slope, there are more than 500 km of areas with unsuitable habitat for this shrew, where more humid and temperate climates predominate across several mountain ranges (Figure 4).
When the model was transferred to Mid-Holocene and LGM conditions, the predictions showed an apparent reduction in suitable areas across the three GCMs, with the reduction even more pronounced during the LGM (Figure 4). During the Mid-Holocene, approximately 6,000 years ago, suitable conditions for C. mayensis appear to have persisted across much of the Yucatan Peninsula; however, this is less extensive if we consider only the continental crust at the surface or currently exposed. On the Pacific slope, conditions may have occurred in a more restricted area and farther from the area surrounding Cueva del Zopilote. During the LGM, around 20,000 years ago, suitable conditions for C. mayensis declined considerably on the Yucatán Peninsula and the Pa-cific coast, although less drastically in the latter region. As in the Mid-Holocene, favorable climatic conditions for Cryptotis mayensis may have been restricted to a smaller area, farther from Cueva del Zopilote.
The MESS analysis indicates that the climatic conditions of the various paleoclimatic scenarios closely resemble those of the calibration area, particularly in the Yucatan Peninsula and along the Pacific slope. This similarity in climatic conditions is consistent across the three analyzed Global Climate Models (GCMs), with positive values observed in these regions. In contrast, areas indicating non-analogous climatic conditions are primarily located in northern Mexico. Specific scenarios, such as the MIROC model during the mid-Holocene, also reveal regions with non-analogous climates that seem to correspond with mountainous areas in Chiapas and central Mexico. However, these areas of high uncertainty do not impact the predicted distribution of C. mayensis. Overall, the MESS analysis reinforces our model projections, showing that climatic conditions in the Pacific region remained within the range used to train the model.
In sum, none of the three-time-slice analyses support the hypothesis of a corridor of suitable conditions for the connectivity of shrew populations with requirements similar to those of C. mayensis during the last glacial-interglacial cycle or in analogous climatic phases.
Discussion
Acknowledgments
We dedicate this work to Dr. Livia León Paniagua, whose distinguished career and inspiring leadership have significantly advanced Mexican mammalogy. Her dedication, integrity, mentorship, and passion for science and education have shaped generations of new scientists.
We thank the following curators and collection managers for the facilities they provided for the examination of specimens in their collections: F. A. Cervantes, J. Vargas, and Y. Hortelano-Moncada (CNMA); A. Riechers (CZRMA); C. Lorenzo and J. Bolaños Citalán (ECO-SC-M); Robert M. Timm (KU); L. León Paniagua (MZFC), S. Hernández Betancourt (UADY), Neal Woodman and S. Peurach (USNM), and C. A. Delfín Alfonso and N. V. Rodríguez Santiago (UV). This work was supported by UNAM-PAPIIT (IN218125).
Declaration of Artificial Intelligence use
ChatGPT Plus was used to generate Figure 2, with a scientific photograph of a museum specimen as input. Grammarly Pro, an AI writing assistant, was used to review the spelling and structure of English.
Author contributions
Lazaro Guevara: Conceptualization (lead); Data curation (supporting); Formal analysis (lead); Writing – review and editing (lead). Reinhard E. Matadamas: Data curation (lead); Formal analysis (lead); Writing – review and editing (supporting). Stephanye Mata-González: Formal analysis (lead); Writing – review and editing (supporting). Paola Zeferino: Conceptualization (supporting); Data curation (supporting); Joaquín Arroyo-Cabrales: Conceptualization (supporting); Writing – review and editing (supporting).
Literature cited
Aiello‐Lammens ME, Boria RA, Radosavljevic A, Vilela B, and Anderson RP. 2015. spThin: an R package for spatial thinning of species occurrence records for use in ecological niche models. Ecography 38:541–545. https://doi.org/10.1111/ecog.01132
Avenant N. 2005. Barn owl pellets: a useful tool for monitoring small mammal communities. Belgian Journal of Zoology 135:39–43.
Baker RJ, and Bradley RD. 2006. Speciation in mammals and the genetic species concept. Journal of Mammalogy 87:643–662. https://doi.org/10.1644/06-MAMM-F-038R2.1
Barnosky AD. 2005. Effects of Quaternary climatic change on speciation in mammals. Journal of Mammalian Evolution 12:247–264. https://doi.org/10.1007/s10914-005-4858-8
Barve N, Barve V, Jiménez-Valverde A, Lira-Noriega A, Maher SP, Peterson AT, et al. 2011. The crucial role of the accessible area in ecological niche modeling and species distribution modeling. Ecological Modelling 222:1810–1819. https://doi.org/10.1016/j.ecolmodel.2011.02.011
Biedma L, Román J, Godoy JA, and Calzada J. 2019. Using owl pellets to infer habitat associations and clarify the regional distribution of a cryptic shrew. Journal of Zoology 308:139–148. https://doi.org/10.1111/jzo.12660
Bonilla-Sánchez A, Sartor CC, Fox-Rosales LA, Feijó A, Ramírez-Fernández JD, Brenes-Mora E, et al. 2024. Ecological niche modeling of the Leopardus tigrinus complex sheds light on its elusive evolutionary history. Journal of Mammalogy 105:953–964. https://doi.org/10.1093/jmammal/gyae074
Carpio-Domínguez JL. 2021. Crimen organizado (narcotráfico) y conservación ambiental: el tema pendiente de la seguridad pública en México. CS 33:237–274. https://doi.org/10.18046/recs.i33.4076
Carraway LN. 2007. Shrews (Eulypotyphla: Soricidae) of México. Monographs of the Western North American Naturalist 3:1–91.
Ceballos G, Arroyo-Cabrales J, and Ponce E. 2010. Effects of Pleistocene environmental changes on the distribution and community structure of the mammalian fauna of Mexico. Quaternary Research 73:464–473. https://doi.org/10.1016/j.yqres.2010.02.006
Choate J. 1970. Systematics and Zoogeography of Middle American shrews of the genus Cryptotis. University of Kansas Publications, Museum of Natural History 19:195–317. https://doi.org/10.5962/bhl.part.15450
Cobos ME, Peterson AT, Barve N, and Osorio-Olvera L. 2019. kuenm: an R package for detailed development of ecological niche models using Maxent. PeerJ 7:e6281. https://doi.org/10.7717/peerj.6281
Dinerstein E, Olson D, Joshi A, Vynne C, Burgess ND, Wikramanayake E, et al. 2017. An ecoregion-based approach to protecting half the terrestrial realm. BioScience 67:534–545. https://doi.org/10.1093/biosci/bix014
Engilis A, Cole RE, and Caro TM, editors. 2012. Small mammal survey of Chiquibul forest reserve, Maya mountains, Belize, 2001. Lubbock, TX (EEUU): Museum of Texas Tech University.
Espinosa-Martínez DV, Ríos-Muñoz CA, Rosales Nanduca H, Arroyo-Cabrales J, and León-Paniagua L. 2017. Mamíferos de Guerrero. Revista Mexicana de Mastozoología ٧:٣٨–٦٧. https://doi.org/١٠.٢٢٢٠١/ie.٢٠٠٧٤٤٨٤e.٢٠١٧.١.٢.٢٤٧
Fick SE, and Hijmans RJ. 2017. WorldClim 2: new 1km spatial resolution climate surfaces for global land areas. International Journal of Climatology 37:4302–4315. https://doi.org/10.1002/joc.5086
Guevara L, and Cervantes FA. 2014. Molecular systematics of small-eared shrews (Soricomorpha, Mammalia) within Cryptotis mexicanus species group from Mesoamérica. Acta theriologica 59:233–242. https://doi.org/10.1007/s13364-013-0165-6
Guevara L, Cervantes FA, and Sánchez-Cordero V. 2015. Riqueza, distribución y conservación de los topos y las musarañas (Mammalia, Eulipotyphla) de México. Therya 6:43–68. https://doi.org/10.12933/therya-15-211
Guevara L, Gerstner BE, Kass JM, and Anderson RP. 2018. Toward ecologically realistic predictions of species distributions: A cross‐time example from tropical montane cloud forests. Global Change Biology 24:1511–1522. https://doi.org/10.1111/gcb.13992
Guevara L, Morrone JJ, and León‐Paniagua L. 2019. Spatial variability in species’ potential distributions during the last glacial maximum under different global circulation models: Relevance in evolutionary biology. Journal of Zoological Systematics and Evolutionary Research 57:113–126. https://doi.org/10.1111/jzs.12238
Guevara L, and Ramírez-Chaves HE. 2025. Cryptotis nigrescens Species Group (C. brachyonyx, C. colombianus, C. hondurensis, C. lacandonensis, C. mayensis, C. merriami, C. merus, and C. nigrescens). In Guevara L, and León-Tapia MA, editors. Mammals of Middle and South America: Eulipotyphla. Cham: Springer Nature Switzerland; p. 1–4. https://doi.org/10.1007/978-3-031-24765-1
Guevara L, and Sánchez-Cordero V. 2018. Patterns of morphological and ecological similarities of small-eared shrews (Soricidae, Cryptotis) in tropical montane cloud forests from Mesoamerica. Systematics and Biodiversity 16:551–564. https://doi.org/10.1080/14772000.2018.1470582
Guevara L, Sánchez-Cordero V, León-Paniagua L, and Woodman N. 2014. A new species of small-eared shrew (Mammalia, Eulipotyphla, Cryptotis) from the Lacandona rain forest, Mexico. Journal of Mammalogy 95:739–753. https://doi.org/10.1644/14-MAMM-A-018
Guevara L, Vargas-Cuenca J, Hortelano-Moncada Y, and Cervantes FA. 2024. A specimen-based database of small-eared shrews (Mammalia, Eulipotyphla, Cryptotis) in the Neotropical Region. Biodiversity Data Journal 12:e135180. https://doi.org/10.3897/BDJ.12.e135180
Guimaraes S, Fernandez‐Jalvo Y, Stoetzel E, Gorgé O, Bennett EA, Denys C, et al. (2016). Owl pellets: a wise DNA source for small mammal genetics. Journal of Zoology 298:64–74. https://doi.org/10.1111/jzo.12285
Hatt RT. 1953. The mammals. In: R.T. Hatt, H.I. Fisher, D.A. Langebartel, and G.W. Brainerd, editors. Faunal and archeological researches in Yucatan Caves. Cranbrook Institute of Science Bulletin; p. 45–77. https://doi.org/10.1093/aibsbulletin/3.3.12-f
He K, Chen X, Qiu YB, Liu Z, Wang WZ, Woodman N, Maldonado JE, and Pan X. 2021. Mitogenome and phylogenetic analyses support rapid diversification among species groups of small-eared shrews genus Cryptotis (Mammalia: Eulipotyphla: Soricidae). Zoological Research 42:739−745. https://doi.org/10.24272/j.issn.2095-8137.2021.199
He K, Woodman N, Boaglio S, Roberts M, Supekar S, and Maldonado JE. 2015. Molecular phylogeny supports repeated adaptation to burrowing within small-eared shrews genus of Cryptotis (Eulipotyphla, Soricidae). PloS One 10:e0140280. https://doi.org/10.1371/journal.pone.0140280
Lawlor TE. 1982. Ototylomys phyllotis. Mammalian Species 181:1–3. https://doi.org/10.2307/3503849
León-Carreño M, Correa-Metrio A, Ramos-Fabiel MA, Escobar J, Curtis JH, Franco-Gaviria F., et al. 2025. The environmental signal of lake sediments in Mesoamerica: Modern pollen and geochemical composition. The Holocene 35:600–615. https://doi.org/10.1177/09596836251320342
López-Forment W, and Urbano G. 1977. Restos de pequeños mamíferos recuperados en regurgitaciones de lechuza, Tyto alba en México. Anales del Instituto de Biología, Universidad Nacional Autónoma de México, Serie Zoología 48:231–242.
Lorenzo C, Bolaños-Citalán J, Navarrete-Gutiérrez D, Pérez-López JA, and Guevara L. 2019. In search of shrews of Chiapas: analysis of their distribution and conservation. Therya 10:121–129. https://doi.org/10.12933/therya-19-717
Mammal Diversity Database. 2025. Mammal Diversity Database (Version 2.3) [Soricidae]. Zenodo. https://doi.org/10.5281/zenodo.17033774
Merow C, Smith MJ, and Silander Jr JA. 2013. A practical guide to MaxEnt for modeling species’ distributions: what it does, and why inputs and settings matter. Ecography 36:1058–1069. https://doi.org/10.1111/j.1600-0587.2013.07872.x
Merriam CH. 1901. Seven new mammals from Mexico, including a new genus of rodents. Proceedings of the Washington Academy of Sciences 3:559–563.
Monroy-Gamboa AG. 2021. The ghost mammals from Mexico and their implications. Therya 12:477–486. https://doi.org/10.12933/therya-21-1186
Morrone JJ. 2014. Biogeographical regionalisation of the Neotropical region. Zootaxa 3782:1–110. https://doi.org/10.11646/zootaxa.3782.1.1
Pérez SG, Jolón MR, Mérida JE, and Andino-Madrid AJ. 2019. First record of the shrew Cryptotis lacandonensis (Eulipotyphla: Soricidae) for Guatemala. Therya 10:187–193. https://doi.org/10.12933/therya-19-766
Peterson AT, Soberón J, Pearson RG, Anderson RP, Martínez-Meyer E, Nakamura M, et al. 2011. Ecological niches and geographic distributions. Princeton (EEUU): Princeton University Press. https://doi.org/10.1515/9781400840670
Phillips SJ, Anderson RP, Dudík M, Schapire RE, and Blair ME. 2017. Opening the black box: An open‐source release of Maxent. Ecography 40:887–893. https://doi.org/10.1111/ecog.03049
R Core Team. 2025. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria. https://www.R-project.org/
Ramírez-Pulido J, and Sánchez-Hernández C. 1972. Regurgitaciones de lechuza, procedentes de la cueva del Cañón del Zopilote, Guerrero, México. Revista de la Sociedad Mexicana de Historia Natural 33:107–112.
Raxworthy CJ, Martinez-Meyer E, Horning N, Nussbaum RA, Schneider GE, Ortega-Huerta MA, et al. 2003. Predicting distributions of known and unknown reptile species in Madagascar. Nature 426:837–841. https://doi.org/10.1038/nature02205
StatSoft Inc. 2005. STATISTICA (data analysis software system). Version 7.1. www.statsoft.com.
Woodman N. 1995. Morphological variation between Pleistocene and recent samples of Cryptotis (Insectivora: Soricidae) from the Yucatan Peninsula, Mexico. Journal of Mammalogy 76:223–231. https://doi.org/10.2307/1382330
Woodman N. 2005. Size evolution in Goodwin’s small-eared shrew, Cryptotis goodwini (Mammalia: Soricomorpha: Soricidae). Special Publication of the International Society of Shrew Biologists. 125–138.
Woodman N. 2019. Three new species of small-eared shrews, genus Cryptotis, from El Salvador, Guatemala, and Honduras (Mammalia: Eulipotyphla: Soricidae). Special Publications of the Museum of Texas Tech University.
Woodman N, and Gaffney SA. 2014. Can they dig it? Functional morphology and semifossoriality among small‐eared shrews, genus Cryptotis (Mammalia, Soricidae). Journal of Morphology 275:745–759. https://doi.org/10.1002/jmor.20254
Woodman N, and Morgan J.J. 2005. Skeletal morphology of the forefoot in shrews (Mammalia: Soricidae) of the genus Cryptotis, as revealed by digital X‐rays. Journal of Morphology 266:60–73. https://doi.org/10.1002/jmor.10367
Woodman N, and Timm RM. 1993. Intraspecific and interspecific variation in the Cryptotis nigrescens species complex of small-eared shrews (Insectivora: Soricidae), with the description of a new species from Colombia. Fieldiana Zoology 74:1–30. https://doi.org/10.5962/bhl.title.3574
Woodman N, and Timm RM. 1999. Geographic variation and evolutionary relationships among broad-clawed shrews of the Cryptotis goldmani-group (Mammalia: Insectivora: Soricidae). Fieldiana Zoology 91:1–35. https://doi.org/10.5962/bhl.title.2669
Woodman N, and Timm RM. 2023. Cryptotis nigrescens (Eulipotyphla: Soricidae). Mammalian Species 55(1035):1–11. https://doi.org/10.1093/mspecies/sead011
Associated editors: Giovani Hernández Canchola and Pablo Colunga Salas
Submitted: October 31, 2025; Reviewed: December 12, 2025
Accepted: March 4, 2026; Published online: May 29, 2026
THERYA, 2026, Vol. 17(2):131-142
DOI: 10.12933/therya.2026.6248 ISSN 2007-3364
Figure 1. Occurrence records of small-eared shrews of the Cryptotis nigrescens group north of Mesoamerica (Guevara et al. 2015): Mayan shrew, Cryptotis mayensis (yellow dots); Lacandona shrew, C. lacandonensis (green dots), and Merriam’s shrew, C. merriami (red dots). The arrow indicates the approximate location of owl pellet records obtained in 1969.
Figure 2. AI-generated image from scientific photographs of a Cryptotis mayensis specimen to illustrate the nine qualitative characters measured in the skull and mandible. The lines represent the region of the skull that is generally destroyed or severely damaged in specimens from the owl pellets collected in Guerrero. Scale bar: 5 mm.
Table 1. Cranial and mandibular measurements of three samples of small-eared shrews (Mammalia, Cryptotis) from the Cryptotis nigrescens group. Statistics are mean ± SD and range, and the number of samples examined by variable.
|
C. mayensis s.s. |
Guerrero |
C. lacandonensis |
|
|
IO |
4.51 ± 0.19 |
4.42 ± 0.19 |
4.75 ± 0.01 |
|
4.88 - 4.22 |
4.77 - 4.14 |
4.76 - 4.74 |
|
|
n = 20 |
n = 27 |
n = 2 |
|
|
ZP |
2.33 ± 0.23 |
2.35 ± 0.14 |
2.53 ± 0.23 |
|
2.67 – 1.68 |
2.60 – 2.05 |
2.69 – 2.36 |
|
|
n = 23 |
n = 28 |
n = 2 |
|
|
MTR |
4.90 ± 0.16 |
4.99 ± 0.14 |
5.28 ± 0.07 |
|
5.17 – 4.57 |
5.30 – 4.45 |
5.33 – 5.23 |
|
|
n = 23 |
n = 25 |
n = 2 |
|
|
M2B |
5.11 ± 0.22 |
5.21 ± 0.14 |
5.21 ± 0.14 |
|
5.63 - 4.49 |
5.50 – 5.02 |
5.50 – 5.02 |
|
|
n = 23 |
n = 21 |
n = 2 |
|
|
HCP |
5.19 ± 0.26 |
5.34 ± 0.23 |
5.43 ± 0.11 |
|
5.54 - 4.7 |
5.83 – 5.18 |
5.50 – 5.35 |
|
|
n = 22 |
n = 7 |
n = 2 |
|
|
HAC |
3.73 ± 0.18 |
3.73 ± 0.15 |
3.61 ± 0.09 |
|
4.07 – 3.4 |
3.94 – 3.53 |
3.67 – 3.54 |
|
|
n = 22 |
n = 7 |
n = 2 |
|
|
HCV |
2.82 ± 0.19 |
2.86 ± 0.14 |
2.75 ± 0.07 |
|
3.13 – 2.3 |
3.03 – 2.59 |
2.80 – 2.70 |
|
|
n = 22 |
n = 7 |
n = 2 |
|
|
LAM |
4.77 ± 0.21 |
4.69 ± 0.29 |
5.00 ± 0.01 |
|
5.08 - 4.24 |
4.97 – 5.54 |
5.00 – 4.99 |
|
|
n = 22 |
n = 7 |
n = 2 |
|
|
TRD |
5.37 ± 0.19 |
5.36 ± 0.27 |
5.73 ± 0.16 |
|
5.68 – 5.02 |
5.54 - 4.9 |
5.84 – 5.62 |
|
|
n = 22 |
n = 5 |
n = 2 |
Table 2. Standardized coefficients of canonical variables based on four and five log-transformed cranial and mandibular variables, respectively, of three samples of small-eared shrews (Mammalia, Cryptotis) from the Cryptotis nigrescens group.
|
Cranial |
Mandibular |
||||
|
Root 1 |
Root 2 |
Root 1 |
Root 2 |
||
|
IO |
-0.324328 |
-1.03949 |
HCP |
0.527026 |
-0.675966 |
|
PZ |
-0.368533 |
-0.43909 |
HAC |
0.531030 |
0.912683 |
|
MTR |
0.959846 |
0.05730 |
HCV |
0.559793 |
-0.500879 |
|
M2B |
0.530850 |
0.51862 |
LAM |
-0.776315 |
0.529173 |
|
- |
- |
- |
TRD |
-0.948245 |
-0.673339 |
|
Eigenval |
0.399078 |
0.09528 |
1.000187 |
0.180837 |
|
|
Explained variance (%) |
80.7% |
19.2% |
84.7% |
15.3% |
|
Figure 3. Plot of scores on first two axes from principal components (PCA) and discriminant function (DFA) analyses of the cranium and mandibles of specimens from Cryptotis mayensis s.s., C. lacandonensis, and samples from Guerrero, México. A, PCA for cranium; B, PCA for mandibles; C, DFA for cranium; D, DFA for mandibles.
Figure 4. Potential distribution of the Mayan shrew, Cryptotis mayensis, estimated for the present, the Mid-Holocene, and the Last Glacial Maximum using ecological niche models. The estimation of Mid-Holocene (A), and Last Glacial Maximum (B) is based on the consensus of three general circulation models, CCSM4, MIROC-ESM, and MPI-ESM-P. The colors on the map indicate the three temporal scenarios considered: Present (blue), mid-Holocene or Last Glacial Maximum (green), and overlap zones between them (red). The bar chart shows the number of pixels estimated to contain suitable conditions for C. mayensis in each period.
