Elevational range extension of the Puna Mouse, Punomys (Cricetidae), with the first record of the genus from Chile

Abstract

We report an elevational record for the Andean sigmodontine Puna Mouse Punomys, which is also the first record of the genus in Chile. The record is based on a mummified specimen that we discovered at an elevation of 5,461 m (17,917 feet) in the caldera of Volcán Acamarachi, Región de Antofagasta, Chile. Results of a morphological assessment suggest that the specimen can be provisionally referred to the species P. lemminus. This new record also extends the known geographic distribution of the genus by 700 km to the south and brings the known Chilean mammal richness to a total of 170 living species and 88 genera. This finding highlights the need for increased survey efforts in more remote, high-elevation regions and demonstrates that there is still much to be learned about the mammal fauna of the Andean Altiplano.

We report a noteworthy elevational and latitudinal range extension for the puna mouse, genus Punomys, based on the finding of a mummified mouse at an elevation of 5,461 m in the caldera of Volcán Acamarachi, Región de Antofagasta, Chile.

The Linnaean shortfall refers to the discrepancy between known and actual numbers of species, and the Wallacean shortfall refers to the discrepancy between the known and actual range limits of species (Hortal et al. 2015). Reducing these knowledge gaps requires specimen collection in the field, especially in remote regions of the planet that are still comparatively underexplored by biologists (Patterson 2002). In spite of more than two centuries of zoological research in South America, mammal assemblages from large portions of the continent remain poorly characterized, and this is especially true for small mammal assemblages in the Andean Altiplano.

Puna mice of the genus Punomys, Osgood, 1943, represent a particularly enigmatic group of Andean sigmodontine rodents. The genus comprises two recognized species, P. lemminusOsgood, 1943, and P. kofordiPacheco and Patton, 1995. These medium-sized, stoutly built mice have a vole-like appearance with long and lax fur, a short tail, and relatively reduced ears (Patton 2015). All previously known Punomys specimens have been collected from a small number of localities in the Altiplano of southern Peru and northwestern Bolivia, at elevations between 4,550 and 4,770 m (see Map 44 from Patton et al. 2015). The two described species of Punomys differ in body proportions, pelage color (including dorsoventral contrast), and several craniodental traits, such as convergence of the zygomatic arches, narrowing of the zygomatic plate, breadth of the nasal region, and size of procingular conules of upper molars (see Pacheco and Patton 1995; Patton 2015). However, the most recent study of Punomys highlighted the difficulty of distinguishing the two nominal forms (Salazar-Bravo et al. 2011).

In recent years, high-elevation surveys in the Central Andes have revealed that we still have much yet to learn about the elevational range limits of small mammals. For example, mountaineering mammal surveys of multiple volcanoes in the Chilean Puna de Atacama have documented evidence of Andean leaf-eared mice, Phyllotis vaccarum (previously referred to as P. xanthopygus rupestris), living at elevations well above 6,000 m (Storz et al. 2020, 2023; Steppan et al. 2022), and mammal surveys in the Andean Altiplano of northwestern Argentina have documented surprisingly diverse rodent assemblages at elevations >4,000 m (Pearson 1951, 1982; Pearson and Ralph 1978; Urquizo et al. 2022). Thus, as part of a continuing effort to further redress Linnean and Wallacean shortfalls for Andean small mammals, here we report the first Chilean record of the genus Punomys, a record that also significantly extends the elevational and latitudinal range of the taxon.

Materials and Methods

Specimen collection, preparation, and morphological characterization.

During the course of a high-elevation mammal survey in the dry Puna ecoregion of the Central Andes of northern Chile, we discovered a mummified rodent carcass in the caldera of Volcán Acamarachi (also known as Cerro Pili), Region de Antofagasta (23°17ʹ33.88″S, 67°37ʹ4.48″W). We deposited the specimen (field number MQC 385) in the Colección de Mamíferos, Universidad Austral de Chile (UACH; catalog number UACH 8478). The mummy was completely intact, with the tail, forefeet, and hindfeet firmly attached to the body, and pelage in good condition. After taking several photographs to record the external appearance of the specimen, we removed the skull and mandible for the purpose of comparing craniodental traits with descriptions of the two named forms of Punomys (e.g., Osgood 1943; Pacheco and Patton 1995; Patton 2015). With the same purpose, we also recorded 13 cranial measurements with a digital caliper to the nearest 0.01 mm, following the protocol of Salazar-Bravo et al. (2011).

Tissue sampling and DNA sequencing.

After removing the skull, we preserved muscle tissue samples in ethanol as a source of genomic DNA. After extracting DNA, we PCR-amplified an 801-bp fragment of the mitochondrial gene, cytochrome b (Cytb), using primers MVZ 05 and MVZ 16 (da Silva and Patton 1993) in accordance with the protocol of Teta et al. (2013). Amplicons were purified and sequenced by Macrogen Inc., Korea. Since the genus Punomys was represented in GenBank by only two Cytb sequences (JQ434426 of 1,139 bp and KY754123 of 426 bp), which were generated from a single specimen of P. kofordi (VPT1890), we obtained additional tissue loans from a total of six Punomys specimens housed in the Museum of Vertebrate Zoology (MVZ), University of California–Berkeley (P. kofordi: MVZ 114757, MVZ 114758, MVZ 116190, MVZ 139589; P. lemminus: MVZ 115948, MVZ 116036). For each of these six specimens, we extracted genomic DNA from toe pads of museum skins. For these samples, we PCR-amplified two overlapping fragments of Cytb using the primer pairs MVZ 05+Oct349r and MVZ 16+Oct39f (Cadenillas and D’Elía 2021). PCR amplicons were purified and sequenced as described above. However, given the degraded state of the samples from museum skins, sequences of sufficient quality were recovered for only four specimens (MVZ 114757, MVZ 115948, MVZ 116036, and MVZ 139589), which were included in subsequent analyses. We deposited all new sequences in GenBank under accession numbers OQ571714–OQ571718 (see Supplementary Data SD1).

We created a Cytb sequence matrix that included all newly generated and previously generated Punomys sequences as well as sequences from representatives of the four main phylogroups of Andinomys edax (Jayat et al. 2017; KY608037, KY608044, KY608047, KY608056). The monotypic genus Andinomys is the sister group of the genus Punomys (Salazar-Bravo et al. 2013), and the Andinomys–Punomys clade is recognized as the tribe Andinomyini. A recent phylogenomic analysis Parada et al. (2021) found that Andinomyini is sister to a clade comprising by the tribes Abrotrichini, Euneomyini, and Phyllotini. We therefore used representative Cytb haplotypes from each of these tribes (MN275227 Abrothrix longipilis; AY95623 Phyllotis darwini; HM167872 Euneomys chinchilloides) to root the tree.

Phylogenetic analysis.

We aligned Cytb sequences using MAFFT v7 (Katoh et al. 2017) and used the E-INS-i strategy to infer character homology. We then visually inspected the alignment with AliView v1.26 (Larsson 2014) to check for the presence of internal stop codons and shifts in the reading frame. Using JModelTest2 (Darriba et al. 2012), we identified TIM2+F+G4 as the best-fit model of nucleotide substitution according to the Bayesian Information Criterion. We estimated a Cytb gene tree using a Bayesian inference phylogenetic analysis implemented in BEAST 2, version 2.6.2 (Bouckaert et al. 2014), running 100 × 10⁶ generations, and sampling produced trees every 1,000 generations. However, as this model cannot be implemented in BEAST2, we employed the GTR+G+I model, according to Ronquist and Huelsenbeck (2003). The first 10% of the total trees were discarded and the remaining trees were used to construct a maximum clade credibility tree with posterior probability (PP) values, employing TreeAnnotator version 2.6.0 (Rambaut and Drummond 2019). Finally, we estimated the percent of sequence divergence between haplotype pairs of Punomys using MEGA X 10.1.8 (Kumar et al. 2018) in the form of p-distances, ignoring sites with missing data.

Results

During a high-elevation mammal survey in the Puna de Atacama of northern Chile, we discovered the mummified carcass of a puna mouse (Punomys) at an elevation of 5,461 m (17,917 feet) asl in the caldera of Volcán Acamarachi, Comuna de San Pedro de Atacama, Antofagasta, Chile (coordinates: 23°17ʹ33.88″S, 67°37ʹ4.48″W; Fig. 1). The caldera in which we found the mummified carcass is devoid of vegetation, with a substrate of volcanic rock and sand, and is situated below a steep, rocky couloir that leads to the summit at 6,046 m (19,836 feet) asl.

Fig. 1.

(A) Records of Punomys including the locality where the Punomys mummy specimen was discovered in the caldera of Volcán Acamarachi, Chile (pentagon with black star). Numbers correspond to localities listed in Supplementary Data SD1. Shading denotes Altiplano terrain > 4,000 m. (B) Aerial view of the locality where the mummy of Punomys was discovered at an elevation of 5,461 m in the caldera of Volcán Acamarachi, Región de Antofagasta, Chile (23°16.98ʹS, 67°37.46ʹW). The view is from the summit of Acamarachi (6,046 m [19,836 feet]) on the southern rim of the caldera. The mummy of Punomys was found at the base of a boulder near the edge of a small glacial pond (denoted by arrow). (C) Close-up view of the edge of the glacial pond and the boulder (denoted by arrow). (D) Mummy of Punomys at the site of discovery.

The external characters of the mummified specimen conformed to the diagnostic criteria that Patton (2015) described for Punomys. Hindfoot soles are proximally hairy and with well-defined pads; toes are of medium length with short claws. Not all external morphological characters (e.g., length of ears, legs, and body) could be properly measured on the mummy specimen due to its preservation condition. However, in spite of slight discoloration, the dorsoventral contrast of the general pelage is clear in this specimen and conforms to that of P. lemminus. This contrast is most conspicuous in the pelage of the cheeks and throat (Fig. 2).

Fig. 2.

External views of the mummified specimen of Punomys from Volcán Acamarachi, Chile (UACH 8478). From top to bottom, dorsal general appearance of pelage, and close-up of the right and left sides of the face, throat, and chest of the mummy, showing the dorsoventral contrast of fur coloration.

The skull of the Chilean specimen is robust and heavy, with a broad and short rostrum and a massive and squared braincase (Fig. 3). The most noteworthy cranial features include broad and anteriorly expanded nasals, narrow interorbital region, robust zygomatic arches, mesopterygoid fossa extending near to posterior edge of M3, and large and inflated tympanic bullae, which together are part of the diagnostic state characters of Punomys (Patton 2015). Measurements of each cranial character of the Chilean specimen (Table 1) fall within the range of values reported for the genus Punomys (see Pacheco and Patton 1995; Salazar-Bravo et al. 2011), but do not permit conclusive inference about species identity as some measurements were more similar to those reported for P. lemminus (e.g., incisive foramen length, palatal bridge length) and others were more similar to those reported for P. kofordi (e.g., greatest length of skull, condyloincisive length). Comparative examination of skull photographs of the holotype of P. kofordi (MVZ 139588) and reference material of P. lemminus (MVZ 115948) presented by Pacheco and Patton (1995) reveals that the convergent shape of the zygomatic arches and the presence of a deep zygomatic notch of the Chilean specimen conforms to the morphotype of P. lemminus (see Figs. 2 and 3 in Pacheco and Patton 1995). Dental characters of the Chilean specimen are also characteristic of Punomys (Pacheco and Patton 1995; Patton 2015): molars are complex, with the presence of a lingual style more developed than the labial style, M1 is curved and with the procingulum anteriorly divergent, and anterolabial and anterolingual conules with subequal sizes, and M3 has four well-defined cusps (Fig. 4).

Fig. 3.

Skull and jaw of the Punomys specimen from Volcán Acamarachi, Chile (UACH 8478). Scale bar = 5 mm.

Table 1.

Cranial measurements of specimens of Punomys kofordi and P. lemminus, the Bolivian specimen of P. kofordi reported by Salazar-Bravo et al. (2011; CBF 858), and the mummified specimen from Volcán Acamarachi, Chile, UACH 8478. Mean and range of the presented measures correspond to those described by Pacheco and Patton (1995) and by Salazar-Bravo et al. (2011), for Peruvian and Bolivian specimens. All measurements are in millimeters.

P. kofordi (n = 6)	P. kofordi (n = 3)	P. kofordi CBF 858	P. lemminus (n = 3)	P. lemminus UACH 8478
Pacheco and Patton (1995)	Salazar-Bravo et al. (2011)	Salazar-Bravo et al. (2011)	Pacheco and Patton (1995)	This study
Greatest length of skull
32.9 (31.4–33.9)	35.2 (35.1–35.4)	—	33.3 (32.5–34.1)	34.32
Condyloincisive length
30.9 (28.8–31.9)	33.7 (33.3–34.3)	—	31.8 (31.0–32.2)	33.40
Zygomatic breadth
17.6 (16.2–18.8)	18.4 (18.4–18.4)	—	18.9 (18.6–19.3)	20.22
Braincase breadth
14.7 (14.5–15.2)	13.8 (13.1–14.6)	13.6	15.2 (14.7–15.5)	15.27
Interorbital breadth
4.4 (4.3–4.5)	4.5 (4.4–4.7)	—	4.2 (4.0–4.3)	4.53
Diastema length
8.0 (7.4–8.4)	9.0 (8.7–9.3)	8.5	8.3 (8.0–8.7)	9.08
Maxillary tooth row length
6.9 (6.7–7.2)	6.7 (6.6–6.9)	6.7	7.0 (6.5–7.5)	7.36
Incisive foramen length
7.0 (6.9–7.2)	8.1 (7.8–8.5)	7.5	6.8 (6.7–6.9)	6.96
Palatal bridge length
6.6 (6.4–6.8)	—	—	7.4 (6.5–8.0)	7.76
Rostral breadth
6.9 (6.4–7.3)	6.6 (6.5–6.7)	6.6	7.2 (7.0–7.6)	6.41
Palatal bridge width
2.3 (1.8–2.9)	2.5 (2.48–2.58)	1.9	2.2 (2.1–2.3)	2.61
First upper molar breadth
2.1 (2.0–2.2)	—	—	2.1 (2.0–2.2)	2.16
Zygomatic plate breadth
3.2 (2.8–3.4)	3.6 (3.5–3.7)	2.9	3.6 (3.5–3.6)	3.53

Fig. 4.

Occlusal views of the molar series of the specimen of Punomys from Volcán Acamarachi, Chile (UACH 8478). U and L denote upper and lower rows, and L and R to left and right rows, respectively. Scale bar = 5 mm.

Phylogenetic analysis of Cytb sequences (Fig. 5) revealed that the Chilean specimen groups with the other sequences of Punomys in a well-supported clade (Bayesian posterior probability: PP = 1). Within this clade, haplotypes representing the nominal species P. kofordi and P. lemminus did not form reciprocally monophyletic groups and the Chilean specimen (UACH 8478) is recovered as sister to a haplotype of a specimen (MVZ 139589) assigned to P. kofordi and collected at a locality close to the type locality of this form (Fig. 1; Supplementary Data SD1). Pairwise genetic distances between the Chilean specimen and all other Punomys specimens ranged from 1.55% to 3.12%, while pairwise comparisons for the whole sample of Punomys ranged between 0.23% and 3.12% (Table 2).

Fig. 5.

Bayesian inference tree based on cytochrome b (Cytb) gene variation showing the placement of the haplotype recovered from the Punomys specimen from Volcán Acamarachi, Chile (UACH 8478). Numbers denote posterior probability values for nodes; only values for species clades and relationships among them are shown. GenBank accession numbers, museum catalog numbers, and collection localities (according to Fig. 1) are provided for representative haplotypes of Punomys.

Table 2.

Percentage of the observed genetic divergence (below diagonal) between mitochondrial DNA cytochrome b (Cytb) haplotype pairs retrieved from specimens of Punomys. Bootstrapped standard error estimate(s) are shown above the diagonal. No divergence estimate is reported between MVZ 114757 and MVZ 139589 due to lack of overlap between the sequenced portions of Cytb in these two specimens.

Cytb sequences	1	2	3	4	5	6
1. JQ434426 P. kofordi		0.0047	0.0070	0.0033	0.0037	0.0051
2. MVZ 114757 P. kofordi	0.8909		—	0.0023	0.0033	0.0095
3. MVZ 139589 P. kofordi	2.4845	—		0.0065	0.0055	0.0072
4. MVZ 115948 P. lemminus	0.7802	0.2278	1.8634		0.0019	0.0054
5. MVZ 116036 P. lemminus	1.0390	0.4525	1.5625	0.2621		0.0062
6. UACH 8478 Chilean Punomys	2.6217	3.1180	1.5528	2.3407	2.7273

Discussion

Analysis of morphological characters and Cytb sequences clearly demonstrate that the mummy specimen from Volcán Acamarachi (UACH 8478) is referable to Punomys, making it the first record for this genus from Chile. Given that we discovered the mummy at an elevation of 5,461 m, the Chilean specimen also represents an elevational record for the genus Punomys, as previous specimens in Peru and Bolivia were collected between 4,450 and 4,770 m asl (Pacheco and Patton 1995; Salazar-Bravo et al. 2011). With this new specimen-based elevational record, Punomys is second only to P. vaccarum as the mammal taxon with the highest known elevational range limit (Storz et al. 2020, 2023). In addition to the new elevational record for Punomys, our Chilean specimen also represents a ~700 km southward extension of the known latitudinal range of the genus. Our discovery of the Chilean specimen at such an extreme elevation in the Central Volcanic Zone suggests that the taxon could be continuously distributed along the upper reaches of the Andean Cordillera in northern Chile and bordering regions with Peru and Bolivia, and likely also Argentina (Fig. 1). This is a testable hypothesis that requires more survey efforts at extreme elevations.

In combination with our discovery of the Punomys mummy at 5,461 m, the discovery of mummies of leaf-eared mice (P. vaccarum) on the summits of multiple >6,000 m Atacama volcanoes (Storz et al. 2023) and live-capture and photographic records of Phyllotis at elevations 6,205–6,739 m (Storz et al. 2020; Steppan et al. 2022) demonstrate that Andean rodents exist at elevations that far surpass previously assumed limits. In comparison to highland rodent taxa like Phyllotis, which tend to have fairly broad elevational ranges, the notable feature of Punomys is that it appears to be restricted to puna habitats >4,400 m (Pearson 1951).

Although specimen UACH 8478 is clearly referable to the genus Punomys, it is difficult to make a species-level determination, largely because diagnostic differences between P. kofordi and P. lemminus are not clearly defined (Salazar-Bravo et al. 2011). The lack of reciprocal monophyly between Cytb haplotypes of the two nominal species further highlights the need to assess species boundaries of Punomys. Thus, operating under the assumption that these nominal forms represent distinct species, three morphological characters appear to be especially informative with regard to the identity of the Chilean specimen: the dorsoventral contrast of fur coloration, convergent shape of the zygomatic arches, and the deepth of the zygomatic notch (Pacheco and Patton 1995). These three characters provide a basis for provisionally recognizing the Chilean Punomys specimen as P. lemminus. Although the haplotype of the Chilean specimen appears as sister to a haplotype of P. kofordi, this relationship lacks significant support (PP = 0.81; Fig. 5). The phylogenetic analysis confirms the identity of the Chilean specimen as Punomys, but it does not help determine its identity as either P. lemminus or P. kofordi since those two nominal forms are not reciprocally monophyletic.

The provisional designation of Chilean Punomys as P. lemminus makes sense in terms of biogeography, given that Volcán Acamarachi and the Peruvian collection localities for P. lemminus are situated in the western slopes of the Andes, and are therefore connected by a continuous tract of highland terrain, whereas the collection localities for P. kofordi lie much further to the north on the eastern side of the Andes (Fig. 1A). However, biogeographic plausibility by itself does not provide a reliable means of taxonomic identification. In summary, a clearer picture of the taxonomic identity of the Chilean specimen requires an assessment of the species limits within Punomys, which in turn requires the collection of additional specimens.

Our finding highlights the value of conducting surveys in inaccessible, poorly explored regions of South America, and suggests that the Andean Altiplano is a region where increased survey work can yield important biodiversity information and surprising new discoveries for native small mammals (Storz et al. 2020, 2023; Rengifo et al. 2022). With our new record of Punomys, the number of living mammal species and genera in Chile now stands at 170 and 88, respectively (D’Elía et al. 2020; Rodríguez-San Pedro et al. 2022, 2023; Quiroga-Carmona et al. 2023). Recent additions to the Chilean mastofauna have been based on descriptions of new species (e.g., Abrothrix manni and Dromiciops bozinovici and D. mondaca; D’Elía et al. 2015, 2016a), as well as range extensions for previously described species (e.g., Eumops perotis, Histiotus laephotis, Notiomys edwardsii, Nyctinomops aurispinosus, Oligoryzomys flavescens; Ossa et al. 2015; D’Elía et al. 2016b; Rodríguez-San Pedro et al. 2022, 2023; Quiroga-Carmona et al. 2023). More field-based specimen collection is needed to further redress Linnean and Wallacean shortfalls in our knowledge of biodiversity in South America. The required efforts would be facilitated by simplifying permitting procedures to conduct biological surveys (D’Elía et al. 2019; Alexander et al. 2021; Teta 2021).

Funding

NIH Grant R01 HL159061, NSF Grants IOS-2114465 and OIA-1736249, National Geographic Explorer Grant NGS-68495R-20, FONDECYT Grant 1221115.

Author Contributions

MQ-C and JFS performed fieldwork. MQ-C and GD generated and analyzed data and wrote the manuscript. All authors have read, commented, and agreed to the final draft of the manuscript.

Conflict of Interest

The authors declare no conflicts of interest regarding this article.

Supplementary Data

Supplementary data are available at Journal of Mammalogy online.

Supplementary Data SD1.—Localities where specimens of Punomys have been collected and material housed in biological collections.