A New Living Species of Degu, Genus Octodon (Hystricomorpha: Octodontidae)

Home , Abrocoma, Aconaemys, Caviomorpha, Hystricomorpha, Octodon, Octodontidae

applyparastyle "fig//caption/p[1]" parastyle "FigCapt" applyparastyle "fig" parastyle "Figure"

Journal of Mammalogy, XX(X):1–16, 2020

DOI:10.1093/jmammal/gyaa143 Downloaded from https://academic.oup.com/jmammal/advance-article/doi/10.1093/jmammal/gyaa143/6029408 by Universidad Austral de Chile-Biblioteca Central user on 10 December 2020

A new living species of degu, genus Octodon (Hystricomorpha: Octodontidae)

Guillermo D’Elía,*, Pablo Teta, Diego H. Verzi, Richard Cadenillas, and James L. Patton Instituto de Ciencias Ambientales y Evolutivas, Facultad de Ciencias, Universidad Austral de Chile, campus Isla Teja s/n, Valdivia 5090000, Chile (GD) División Mastozoología, Museo Argentino de Ciencias Naturales “Bernardino Rivadavia,” Ciudad Autónoma de Buenos Aires, C1405DJR Argentina (PT) CONICET, Sección Mastozoología, Facultad de Ciencias Naturales y Museo de La Plata, Universidad Nacional de La Plata, Paseo del Bosque s/n, B1900FWA La Plata, Argentina (DHV) Doctorado en Cs. m. Ecología y Evolución, Facultad de Ciencias, Universidad Austral de Chile, campus Isla Teja s/n, Valdivia 5090000, Chile (RC) Museum of Vertebrate Zoology, University of California, Berkeley, CA, 94720-3160 USA (JLP) *Correspondent: [email protected]

We combine morphological (qualitative and quantitative data) and genetic (one mitochondrial and one nuclear gene) data from a large set of specimens of Octodon from the four currently recognized living species of the genus. The integration of the results (qualitative assessment, multivariate analysis of cranial measurements, and gene trees) allows us to state that 1) the current taxonomic scheme does not reflect the species diversity of Octodon; 2) in particular, as currently understood O. bridgesii likely is a complex of three species; 3) one of these, encompassing the southern populations of the genus, in the Araucanía Region (Chile) and Neuquén Province (Argentina), is named and described here as a new species; and 4) the mitochondrial gene tree departs from the nuclear gene tree with respect to O. pacificus and the new species here described.

Key words: Caviomorpha, Octodon bridgesii, Octodontoidea, Taxonomy

Combinamos datos morfológicos (datos cualitativos y cuantitativos) y genéticos (secuencias de un gen mitocondrial y uno nuclear) de un amplio conjunto de especímenes de Octodon de las cuatro especies actuales reconocidas del género. La integración de los resultados obtenidos (evaluación cualitativa, análisis multivariado de medidas craneales y árboles de genes) nos permite afirmar que 1) el esquema taxonómico actual no refleja la diversidad de especies de Octodon; 2) en particular, O. bridgesii, como está delimitada actualmente, es un complejo de probablemente tres especies; 3) una de estas, que comprende las poblaciones más australes del género, en la Región de la Araucanía (Chile) y la Provincia de Neuquén (Argentina), se nombra y describe en este trabajo como una nueva especie; y 4) el árbol del gen mitocondrial no refleja exactamente el árbol basado en genes nucleares con respecto a O. pacificus y la nueva especie aquí descrita.

Palabras clave: Caviomorpha, Octodon bridgesii, Octodontoidea, Taxonomía Nomenclatural statement: A Life Science Identifier (LSID) number was obtained for this publication: urn:lsid:zoobank. org:pub:C0509B1A-7AD5-4F24-8E3B-A9230D764A57

The family Octodontidae of caviomorph rodents is distrib- surface-dweller, Octodontomys Palmer 1903, Octodon uted widely in southern South America (Verzi et al. 2015). Bennett 1832, and Tympanoctomys Yepes 1942 (including It comprises six living genera of herbivorous rodents that Pipanacoctomys Mares, Braun, Barquez and Diaz 2000 and display a variety of habits; Octomys Thomas 1920 is a Salinoctomys Mares, Braun, Barquez and Diaz 2000) are

1 2 JOURNAL OF MAMMALOGY Downloaded from https://academic.oup.com/jmammal/advance-article/doi/10.1093/jmammal/gyaa143/6029408 by Universidad Austral de Chile-Biblioteca Central user on 10 December 2020 fossorial, Aconaemys Ameghino 1891 is semisubterranean, recovered from owl and eagle pellets, Massoia et al. (1991) and while Spalacopus Wagler 1832 is subterranean (Verzi et al. Podestá et al. (2000), respectively, reported a third and fourth 2015). Octomys and Tympanoctomys are restricted to open living population, enlarging the known Argentine distribution arid and semiarid areas of west-central Argentina; Spalacopus toward northern Neuquén. Finally, Sage et al. (2007) docu- is endemic to scrubland areas of central Chile; Octodon and mented three additional records for southwestern Neuquén, Aconaemys are present in open areas and forests of south- also referring them to O. bridgesii. This situation has remained western Argentina and central and south-central Chile; while unchanged since then, as no study has evaluated the degree Octodontomys is distributed in open areas of northwestern of differentiation among populations currently assigned to Argentina, southwestern Bolivia, and northern Chile (Verzi O. bridgesii. et al. 2015). Remarkably, the last taxonomic work focused on Octodon Octodontid genera show low species diversity, con- dates from more than two decades ago, when Hutteter (1994) taining one (Octomys, Octodontomys, and Spalacopus), three described a new species of degu from Isla Mocha and Pearson (Aconaemys), and four (Octodon and Tympanoctomys), recog- (1995) questioned the identity of Argentine populations re- nized living species. The diversification of crown octodontids ferred to O. bridgesii. Recently, framed in a study of octodontid was recently estimated to have occurred around 8.8 (7.3–10.4) relationships, Suárez-Villota et al. (2016) supported the spe- Ma during the Late Miocene (see Voloch et al. 2013; Upham cific distinction of O. pacificus based on analysis of sequences and Patterson 2015; Suarez-Villota et al. 2016; Álvarez et al. of the mitochondrial 12S rRNA and nuclear growth hormone 2017, for earlier, but nevertheless still Miocene, age estimates). receptor genes. In a subsequent study of sequences of the The taxonomic history of Octodon and its species is quite Cytochrome b (mitochondrial) gene, Vianna et al. (2017) found straightforward. The first described species wasO. degus , with that O. pacificus is very similar to O. bridgesii. As will be dis- type locality in “St. Jago” (= Santiago), central Chile (Molina cussed below, the difference in the results of these two studies, 1782). This is the only species of the genus that has several regarding the analysis of fragments of the mitochondrial ge- available names in its synonymy (i.e., clivorum Thomas 1927, nome, mostly is due to differences in taxonomic sampling. In cumingii Bennett 1832, getulinus Poeppig 1829, pallidus particular, O. bridgesii as currently understood represents a spe- Wagner 1845, peruana Tschudi 1845); however, no subspecies cies complex that is not monophyletic, and the referred studies currently are recognized (Díaz et al. 2015; see Thomas 1927 used different forms of this complex to represent O. bridgesii. for a scheme where two subspecies were recognized). In the Beyond that, these two studies are, so far, the only instances mid-19th century, Waterhouse (1845) described O. bridgesii, where taxonomy of Octodon has benefited from the analysis whose type locality was restricted by Thomas (1927) to “Río of DNA sequence data. However, species of Octodon have not Teno, Colchagua,” O’Higgins, central Chile. Almost a century been revised, although all but O. pacificus have large distri- later, Osgood (1943) described O. lunatus with the type lo- butions across environmentally and topographically complex cality at “Olmue, Province of Valparaiso, Chile.” Tamayo and areas. Species limits of Octodon have not been assessed with Frassinetti (1980) and Contreras et al. (1987) questioned the molecular data. distinction of O. lunatus, given its marked morphological sim- In light of the foregoing, here we evaluate the taxonomic ilarity with O. bridgesii. However, both forms have highly dis- status of Argentine and some Chilean populations of degus cur- tinct diploid numbers: that of O. brigdesii is 2n = 58 (Venegas rently allocated to O. bridgesii. Morphological and molecular 1975; Gallardo 1992), while O. lunatus has a diploid comple- evidence suggests that some of these populations correspond to ment of 2n = 78 (Spotorno et al. 1995). Finally, Hutterer (1994) a new species that we name and describe herein. described a fourth species, O. pacificus, from “Isla Mocha (38°22′S, 73°55′W), Arauco Province,” a small island ca. 31 km off the Pacific coast of south-central Chile. Materials and Methods For over a century, the genus Octodon was considered en- This study is based on molecular and morphological evidence. demic to Chile (e.g., Mann 1978; see also Thomas [1927], who Analyzed specimens, all of which were deposited in museum disregarded a putative Peruvian record of Octodon that is the collections, are listed in Appendix I. No fieldwork was done basis of O. cummingii var. peruana Tschudi 1845). The sit- in the course of this study; all studied specimens already were uation changed when Massoia (1979; see also Massoia et al. housed in museum collections. 1981) reported remains of Octodon from an archeological site Genetic and phylogenetic analyses.—Genetic comparisons at Neuquén Province, Argentina; Massoia (1979) suggested and phylogenetic analyses were based on DNA sequences of that these remains might belong to O. bridgesii. A decade later, the mitochondrial Cytochrome b (hereafter Cytb; the first 801 Verzi and Alcover (1990) reported two living populations from base pairs, bp) and the nuclear growth hormone receptor (here- Neuquén, assigning them to O. bridgesii, while Pearson and after GHR; 872 bp) genes. Analyses based on Cytb included Pearson (1993) reported additional subfossil remains from 27 sequences gathered from specimens of the four species of Holocene cave sediments of southern Neuquén. However, Octodon currently recognized (Díaz et al. 2015); this sampling Pearson (1995) pointed out differences between Argentinean includes topotypes of O. pacificus. Of the Cytb sequences ana- specimens and the holotype of O. bridgesii, casting doubt on lyzed, eight belong to a new species described below and were the identity of the Argentinean populations. Based on remains gathered from specimens (n = 6) collected at two Argentine D’ELÍA ET AL.—NEW SPECIES OF OCTODON (HYSTRICOMORPHA) 3 Downloaded from https://academic.oup.com/jmammal/advance-article/doi/10.1093/jmammal/gyaa143/6029408 by Universidad Austral de Chile-Biblioteca Central user on 10 December 2020

Octodontomys, Octomys, Spalacopus, and Tympanoctomys) that were used to form the outgroup. Of the DNA sequences of Octodon, 13 were generated by us, while 26 additional ones were downloaded from GenBank. Cytb sequences gathered here were amplified using primers MVZ 05 and MVZ 16 as outlined in Cañon et al. (2014) and Chiquito et al. (2014). Fragments of the GHR gene were amplified using the primers GHR50F and GHREND (Adkins et al. 2001) and the protocol detailed by Rowe and Honeycutt (2002). Amplicons were sequenced using an external sequencing service (Macrogen, Inc., Seoul, Korea). DNA sequences were edited by eye using CodonCode (Codon-Code, Dedham, Massachusetts); all new sequences were deposited in GenBank (Appendix I; see also Fig. 2). Sequences were aligned with Clustal as implemented in MEGA 6 (Tamura et al. 2013) using the default parameter values. Observed genetic p-distances were calculated in MEGA 6. Phylogenetic relationships were inferred via Bayesian analyses (BA) using MrBayes 3.2 (Ronquist et al. 2012) by means of two independent runs, each with five heated and one cold Markov chains. Substitution models (Cytb: HKY+G; GHR: HKY+I) were selected using jModelTest (Darriba et al. 2012); all model parameters were estimated in MrBayes. Uniform- interval priors were assumed for all parameters except base composition and substitution model parameters, which assumed a Dirichlet prior. Runs were allowed to proceed for 20 million generations with trees sampled every 1,000 generations per chain. To check for convergence on a stable log-likelihood value, we plotted the log-likelihood values against generation time for each. The first 25% of the trees were discarded as burn-in and the remaining trees were used to compute a 50% majority rule consensus tree and obtain posterior probability (PP) estimates for each clade. Time estimates.—Dating of cladogenetic events was undertaken using the Cytb data set and BEAST v2.4.8 (Bouckaert et al. 2014). We used this matrix and not the GHR matrix, because we are interested in divergence times among species, but also on species crown ages and for some species the GHR matrix includes a single representative. To calibrate the molecular clock, we used the fossil record related to Octodon. As in Upham and Patterson (2015), we use †Pithanotomys, a Fig. 1.—Map of southern South America depicting the collection local- crown octodontid sister to Aconaemys of 5.0 Ma in age (Verzi ities of the specimens of Octodon analyzed in this study. Locality numbers correspond to those on Appendix I. Colors are as follow: golden et al. 2014), to calibrate the basal node of the clade formed dots, O. degus; orange dots, O. lunatus; light green dots, O. bridgesii by Octodon, Spalacopus, and Aconaemys. We set a lognormal s.s.; dark green dots, coastal O. “bridgesii”; blue dot, O. pacificus; red prior with 5.0 Ma as the minimum age and 11.8 Ma as a soft dots, O. ricardojeda n. sp. (= southern O. “bridgesii”). maximum age; the latter corresponds to the middle Miocene, an interval in which crown octodontids are no longer recorded (Verzi et al. 2014). The distribution has a mean of and one Chilean locality (n = 2; Fig. 1; Appendix I). Analyses 0.272 and SD of 1 (see calibration Q in Upham and Patterson based on GHR included sequences of 12 specimens belonging 2015:115). A birth–death model using an initial random tree to the four species of Octodon currently considered as valid was set as a prior. The employed substitution model, selected (Díaz et al. 2015). Of these sequences, four belong to the form using jModeltest, was HKY+G with empirical base frequen- described below as new and were gathered from specimens col- cies, and four G rate categories. Runs were undertaken under lected at two Argentine localities (n = 2) and one Chilean lo- a strict clock model. Three independent runs of 150 million cality (n = 2; Appendix I). Both matrices were completed with generations, sampled every 10,000 generations were carried sequences of the remaining octodontid genera (i.e., Aconaemys, out. Convergence to stable values was checked with Tracer 4 JOURNAL OF MAMMALOGY Downloaded from https://academic.oup.com/jmammal/advance-article/doi/10.1093/jmammal/gyaa143/6029408 by Universidad Austral de Chile-Biblioteca Central user on 10 December 2020

Fig. 2.—Majority rule consensus trees obtained from the Bayesian analysis of 27 Cytb gene sequences (left) and 12 sequences of the GHR gene (right) of specimens of Octodon. Numbers indicate posterior probability values of adjacent nodes. Ages represented as mean node height (and their 95% highest posterior density credibility interval) are given for main nodes of the Cytb gene with small font. Terminal labels are as follow: specimen number (see Appendix I; Fig. 1) and GenBank accession numbers. Tree branches and species names are color-coded as in Fig. 1. v.1.7 (Rambaut et al. 2018) obtaining an effective sample Quantitative morphological analyses were carried out on size (ESS) greater than 200 for all parameters. Tree and log a set of 104 specimens with fully erupted M3 (age classes II files (40,503 trees after a 10% burn-in) were combined using and III sensu Verzi and Alcover 1990) of both sexes, including LogCombiner (Bouckaert et al. 2014). Resulting trees were representatives of the four known living species of Octodon compiled into a maximum clade credibility (MCC) tree with (O. bridgesii, O. degus, O. lunatus, and O. pacificus) and two TreeAnnotator (Bouckaert et al. 2014) displaying mean ages candidate species (see below). Among the assessed specimens and highest posterior density (HPD) intervals (95% upper and were the holotypes of O. lunatus and O. pacificus, as well as the lower) for each node. lectotype and paralectotype of O. bridgesii. The type locality Quantitative and qualitative analysis of morphological of O. bridgesii is imprecise. Waterhouse (1845) described the variation.—Descriptions of anatomical traits are based on species, establishing Chile as the provenance of the specimens the nomenclature of Hutterer (1994), Verzi et al. (2014), and on which the species was based. By selecting a lectotype and Olivares and Verzi (2015). Standard external measurements a paralectotype from Río Teno, Colchagua, Thomas (1927) (in mm) were recorded from specimen tags or field catalogs restricted the type locality of O. bridgesii to this area of cen- or calculated: total length (TL); head and body length (HB); tral Chile. Nevertheless, this a large area, as the Teno River is tail length (T); hindfoot length (HF); ear length (E); and mass about 100 km in length. Here, considering the relevance that (W, in g). We measured 13 craniodental variables with a dig- name-bearing specimens have in taxonomy, we further restrict ital caliper to the nearest 0.01 mm; of these, 10 were used the type locality of O. bridgesii to Maule, Curicó, Romeral, in multivariate analysis, see below. Definitions of measure- Río Teno (35°02′49″S, 70°35′54″W), locality 23 in Appendix I, ments follow Verzi and Alcover (1990), with modifications: an area that was on the route of T. Bridges (the collector of total length of the skull (TLS); condylo-incisive length (CIL); the specimens assessed by Waterhouse) and from where there basilar length (BL); nasal length (NL); nasal breadth (NB); are known records of O. bridgesii. Octodon degus lacks a type rostrum breadth (RB); zygomatic breadth (ZB); interorbital specimen, but its type locality, as indicated by Molina (1782), is breadth (IOB); frontal length (FL); upper diastema length Santiago (Chile), a general area from where we have included (UDL); palatilar length (PalL); upper toothrow length (UTR); several specimens (Appendix I). upper incisors breadth, taken at the base (UIB). Skull measure- Morphological comparisons among samples were guided ments for O. pacificus were those reported by Hutterer (1994) (i.e., composition of compared groups) by the results of the and Vianna et al. (2017) and/or estimated from the photographs analyses of molecular data (see below), geography, and current illustrating those works. taxonomy. Based on these considerations, the following groups D’ELÍA ET AL.—NEW SPECIES OF OCTODON (HYSTRICOMORPHA) 5 Downloaded from https://academic.oup.com/jmammal/advance-article/doi/10.1093/jmammal/gyaa143/6029408 by Universidad Austral de Chile-Biblioteca Central user on 10 December 2020 were constituted (Fig. 1; Appendix I): O. degus (localities the one hand, and on the other, to a clade (PP = 1) formed by 1–19); O. lunatus (localities 4, 6, 9, 13, 15, 20, 21); O. bridgesii the sequences of the other species. As in the Cytb topology, in s.s. (localities 22–24); coastal O. “bridgesii” (localities 25–29); the GHR tree, genetic variants recovered from O. bridgesii do O. pacificus (locality 30); southern O. “bridgesii” (localities not form a monophyletic group. However, an important depar- 31–38). Quantitative morphological variation was summarized ture from the Cytb topology is that in the GHR tree, the allele through principal component analysis (PCA). The PCA was of O. pacificus is not nested among the alleles of O. bridgesii executed based on the variance–covariance matrix of the log- from the piedmont of the Araucanía Region and from Neuquén transformed data set of craniodental measurements. All ana- Province (southern O. “bridgesii”). The allele of O. pacificus lyses were undertaken using the software InfoStat (Di Rienzo falls in the basal polytomy of a clade (PP = 0.87) with those of et al. 2008). O. lunatus, coastal O. “bridgesii,” and the clade of the alleles of O. bridgesii from the Araucanía and Neuquén. Divergence time.—Molecular clock analysis using the Results Cytb matrix estimated an age for crown Octodon of 3.84 Phylogenetic analyses.—The Cytb matrix had 235 variable Ma (95% HPD: 2.83–4.93). The stem age of the southern O. characters (127 sites if only sequences of Octodon are con- “bridgesii”–O. pacificus lineage is 1.97 Ma (1.37–2.66) and sidered). The tree topology resulting from MrBayes (Fig. 2) the crown age of this group is 0.90 Ma (0.50–1.32). Crown is similar to that generated using BEAST; as such we present age of the Neuquén–O. pacificus lineage is 0.06 Ma (0.01– the former in Fig. 2, with dates from the latter analysis. Minor 0.13). Crown ages for other clades of Octodon are provided in differences between topologies pertain to poorly supported in- Supplementary Data SD1, Table S2. traspecific relationships among haplotypes. Analyses show a Morphological analyses.—Qualitative morphological differ- monophyletic and highly supported Octodon (here and after- ences among main forms of Octodon are summarized below. wards posterior probability values are those gathered with The PCA shows that 12 of 13 variables (Table 1) were pos- MrBayes and BEAST, respectively; PP = 1; PP = 1). The basal itively correlated with the first principal component (PC1 dichotomy of the clade of Octodon leads to O. degus (PP = 1; 70.6%; Figs. 3A and 3B; Table 2). On PC1, UIB and UTR has PP = 1) on one hand and to a clade (PP = 1; PP = 1) formed the highest loadings among positive scores, while IOB has the by haplotypes of all other species of the genus. Haplotypes lowest loading (Table 2). Three principal groups can be identi- of O. bridgesii, as currently defined, do not form a mono- fied along PC1. Two of these are the samples of O. degus and phyletic group; some haplotypes of O. bridgesii are more O. pacificus, respectively, which are placed at the opposite ex- closely related to variants of O. lunatus or O. pacificus, than tremes of this PC; while the third group is composed by those to other haplotypes of O. bridgesii. The two haplotypes from samples labeled as O. bridgesii s.s., coastal O. “bridgesii,” the northern part of the range of O. bridgesii, on the Andean southern O. “bridgesii,” and O. lunatus (Figs. 3A and 3B). piedmont of the Chilean O’Higgins and Maule Regions, an PC2 and PC3 accounted for 11.6% and 5.7% of the total var- area within which the type locality (23) of O. bridgesii lies, iance; PC3 contributes to separate the samples of southern form a clade (PP = 1; PP = 1) that together with the haplo- O. “bridgesii” from those of O. bridgesii s.s., coastal O. type from coastal Maule O. bridgesii and the clade (PP = 1; “bridgesii,” and O. lunatus (Figs. 3A and 3B). Highest loading PP = 1) formed by the two haplotypes of O. lunatus forms a on PC2 and PC3 correspond to IOB and NB, respectively clade (PP = 0.97; PP = 0.99) to the exclusion of O. pacificus (Table 2). Specimens of O. lunatus were moderately to highly and southern O. bridgesii. Remarkably, Cytb haplotypes re- superimposed to those of coastal O. bridgesii and O. bridgesii covered from specimens of O. pacificus are very similar to s.s. (Figs. 3A and 3B). those from specimens of putative O. bridgesii from Neuquén, Argentina (localities 37 and 38), and as such fall in the same clade (PP = 0.93; PP = 1). This clade is sister (PP = 1; PP = 1) Discussion to a clade (PP = 1; PP = 1) formed by the two haplotypes Since the beginning of this century, species delimitation, the from specimens collected at the southern Chilean range of fundamental task of taxonomy, has been carried out within O. bridgesii on the Andean piedmont of the Araucanía Region the framework known as integrative taxonomy (Dayrat 2005; (locality 32). Haplotypes from Araucanía differ on average by Padial et al. 2009). This approach, not exempt of criticisms 2.9% from those from Neuquén and of O. pacificus. In turn, the (e.g., see Valdecasas et al. 2008 who claim that taxonomy southern O. “bridgesii”–O. pacificus lineage differs on average has been historically integrative; see for instance Patton et al. by 5.6% from the coastal-northern O. bridgesii–O. lunatus 2000), states that species delimitation will be maximally ro- clade. Divergence values for other Cytb comparisons are given bust when different lines of evidence are considered. Here we in Supplementary Data SD1, Table S1. generated morphologic and molecular data, which allow us to The GHR matrix has 42 variable sites (24 sites if sequences assess species boundaries within the genus Octodon. of Octodon only are considered). As in the Cytb topology, Under the unified concept, species are defined as independ- the Bayesian tree of GHR sequences (Fig. 2) shows a mon- ently evolving metapopulation lineages (de Queiroz 2007). As ophyletic Octodon (PP = 1); the dichotomy at its base leads such, species delimitation is based on the evaluation of the ac- to a clade (PP = 1) formed by all sequences of O. degus in quisition of properties, including morphological diagnosability, 6 JOURNAL OF MAMMALOGY Downloaded from https://academic.oup.com/jmammal/advance-article/doi/10.1093/jmammal/gyaa143/6029408 by Universidad Austral de Chile-Biblioteca Central user on 10 December 2020 23 ± 0.66, 21.5–24.07 Mean ± SD , Min–Max 5.72 ± 0.42, 5.2–6.25 8.03 ± 0.38, 7.39–8.66 8.16 ± 0.66, 6.98–9.04 9.74 ± 0.45, 9.02–10.23 10.8 ± 0.61, 9.75–12.07 3.77 ± 0.23, 3.34–4.16 43.19 ± 1.42, 40.7–45.24 38.79 ± 2.18, 35.29–41.29 34.32 ± 0.77, 32.37–35.02 16.17 ± 0.71, 14.96–17.03 13.53 ± 0.6, 12.56–14.64 15.78 ± 0.62, 14.88–16.63 n. sp. O . ricardojeda n 11 11 11 11 11 11 11 11 11 11 11 11 11 O. lunatus Mean ± SD , Min–Max 5.74 ± 0.33, 4.99–6.26 8.27 ± 0.51, 7.63–9.36 8.60 ± 0.41, 8.08–9.36 3.27 ± 0.33, 2.67–3.80 45.14 ± 1.89, 42.94–48.49 40.22 ± 2.84, 35.64–45.98 35.79 ± 2.75, 32.58–39.68 17.01 ± 1.32, 15.17–19.42 24.20 ± 1.19, 22.61–26.55 12.55 ± 0.57, 11.53–13.26 10.05 ± 0.81, 8.81–11.33 15.89 ± 1.29, 14.41–18.37 10.27 ± 0.68, 9.04–11.04 8 8 9 n 10 13 13 13 11 11 13 13 13 11 Mean ± SD , Min–Max ” O . “ bridgesi (coastal Chile) 6.11 ± 0.35, 5.63–6.90 8.26 ± 0.36, 7.61–8.75 8.34 ± 0.35, 7.62–8.82 3.38 ± 0.28, 2.86–3.75 45.50 ± 1.37, 43.30–47.53 41.57 ± 1.74, 38.26–43.97 35.21 ± 1.72, 31.68–37.45 17.63 ± 0.79, 16.38–19.13 23.59 ± 0.70, 22.48–24.62 13.49 ± 0.87, 12.27–15.16 10.16 ± 0.65, 9.02–11.06 16.02 ± 0.81, 14.15–17.51 10.09 ± 0.65, 8.70–11.21

n Fig. 3.—Specimen scores of adult individuals of Octodon (N = 104) 15 15 15 15 15 15 15 15 15 15 15 15 15 for principal components 1 and 2 (A) and 1 and 3 (B). Colors are as follow: light green dots, O. bridgesii s.s.; dark green dots, coastal O. “bridgesii”; red dots, O. ricardojeda n. sp. (= southern O. “bridgesii”); golden dots, O. degus; orange dots, O. lunatus; blue dot, O. pacificus.

reciprocal monophyly, differences in ecological niches, or reproductive isolation (Sites and Marshall 2003, 2004; de O . bridgesi Mean ± SD , Min–Max 6.06 ± 0.50, 5.62–6.61 8.29 ± 0.18, 8.11–8.47 9.19 ± 0.28, 8.89–9.44 9.93 ± 0.35, 9.52–10.16 3.68 ± 0.09, 3.62–3.78 45.49 ± 0.43, 45.08–45.93 41.93 ± 0.81, 41.36–42.86 35.34 ± 0.99, 34.36–36.34 17.63 ± 1.20, 16.65–18.97 24.69 ± 0.79, 24.01–25.56 13.46 ± 0.37, 13.11–13.84 16.06 ± 0.44, 15.56–16.41 10.47 ± 0.19, 10.27–10.65 Queiroz 2007). As speciation is a continuous process, lineages acquire their distinctive properties at different order and times; 3 3 n 3 3 3 3 3 3 3 3 3 3 3 even more, some property may even not be acquired at all (e.g., two species may be ecologically and morphologically indis- tinguishable and/or lack monophyly at a given locus). For this reason, there is a gray zone where different data set and delimitation criteria, eventually, provide conflicting schemes of species boundaries (de Queiroz 1998, 2007). Results suggest that the current taxonomic scheme of

O. degus Octodon, with four recognized monotypic species, does not Mean ± SD , Min–Max 5.30 ± 0.31, 4.63–6.08 7.63 ± 0.37, 6.83–8.81 9.48 ± 0.56, 8.17–10.53 9.40 ± 0.58, 8.10–10.72 9.12 ± 0.60, 7.90–10.37 2.90 ± 0.25, 2.37–3.47 reflect the true diversity within the genus. Our analyses sug- 40.70 ± 1.92, 36.93–44.96 36.51 ± 2.05, 31.81–40.79 32.34 ± 1.86, 28.16–36.37 15.03 ± 0.94, 13.04–17.25 22.27 ± 1.03, 19.92–24.32 12.74 ± 0.70, 10.95–14.08 14.36 ± 0.81, 12.50–16.29 gest that the taxon O. bridgesii encompasses more than one n 66 66 66 66 66 66 66 66 66 66 66 66 66 lineage that differ morphologically and genetically. Therefore, these forms are considered as independently evolving lineages of species level (de Queiroz 1998, 2007). In particular, popu- . See “Materials and Methods” for explanation 1.— Summary statistics (mean, SD , range [Minimum–Maximum]) of cranial measurements (in mm) adult samples ( n ) Octodon . See “Materials and Methods” for explanation Table TLS of the abbreviations. CIL

BL NL NB RB ZB IOB FL UDL PalL UTR UIB lations from the southernmost part of the range (Araucanía D’ELÍA ET AL.—NEW SPECIES OF OCTODON (HYSTRICOMORPHA) 7 Downloaded from https://academic.oup.com/jmammal/advance-article/doi/10.1093/jmammal/gyaa143/6029408 by Universidad Austral de Chile-Biblioteca Central user on 10 December 2020

Table 2.—Results of principal components analyses (first, second, “bridgesii” diverge from that of O. pacificus by 0.54%. Given and third columns) and discriminant function analysis (fourth, fifth, the topology of the nuclear GHR tree, random sorting of neu- and sixth columns) performed on seven groups of adult specimens of tral variants seems the most likely explanation for the observed Octodon (N = 104). See “Materials and Methods” for explanation of pattern of Cytb variation. Future studies, with a larger spec- the abbreviations. imen and genetic sampling, would clarify if this hypothesis PC1 PC2 PC3 explains the similarity of the Cytb haplotypes of O. pacificus NL 0.3715 0.1879 0.2915 and the Octodon distributed in the southernmost part of the NB 0.2496 −0.1067 0.7652 genus range. As karyotypic data have been useful to differen- RB 0.2151 0.0419 0.2153 tiate O. lunatus from “coastal bridgesii” (Spotorno et al. 1995), ZB 0.2133 0.1968 0.0803 IOB −0.0907 0.8537 −0.1261 future studies may benefit from the inclusion of karyotypic data FL 0.2008 0.1692 −0.0641 that is missing for southern lineages of Octodon. Beyond that, UDL 0.2479 0.2889 0.0382 clearly the southernmost lineage of Octodon is not part of the PalL 0.3475 0.1206 −0.0818 UTR 0.3811 −0.1071 −0.2306 species represented by the name O. bridgesii, but more closely UIB 0.5759 −0.2143 −0.4418 related to O. pacificus, although both differ morphologically. Eigvalue 0.0097 0.0015 0.0008 As such, we considered their distinction needs to be taxonomi- %Variance 70.57 11.64 5.72 cally recognized. Here we considered the southernmost lineage of Octodon, currently considered as part of O. bridgesii, as a distinct species; as no name is available for it, it is described and Neuquén) of this genus hitherto assigned to O. bridgesii, below as a new species. differ genealogically (Fig. 2) and morphologically (Fig. 3) from all other populations of Octodon, including those from the area of typical O. bridgesii on the Andean piedmont of the Octodon ricardojeda, new species Maule Region. Ricardo Ojeda’s Degu Remarkably, the Cytb haplotype recovered from speci- (Fig. 4) mens of O. pacificus is shared with Argentinean specimens Holotype.—MLP 12.VII.88.2, adult female; skin, skull, and of the southernmost lineage of Octodon; in fact, haplotypes skeleton collected by A. Alcover, A. Carlini, and D. H. Verzi. of the southernmost lineage of O. bridgesii form a paraphy- Type locality.—Argentina: Neuquén, Huiliches, Parque letic group with respect to sequences of O. pacificus (Fig. 2). Nacional Lanin, neighborhood of río Currhué, less than 20 At least three alternative scenarios can explain this pattern. m from it, and 200–250 m from the eastern margin of Lago The first one pertains to retention of ancestral polymorphisms Curruhué Chico (−39.9050, −71.3256). (Ball et al. 1990; Hudson 1990); species-specific mutations Paratypes.—MVZ 164036 (field number OPP 6974), adult occurring over this ancestral variant would explain why both male, collected by O. P. Pearson at Neuquén, Huiliches, Arroyo species do not share the same haplotype but show highly sim- del Escorial (−39.81, −71.56); MVZ 201202 (field number ilar sequences. The second possibility involves introgression RDS 16623), adult male, collected by R. D. Sage at Neuquén, of the mitochondrial genome of one species into the nuclear Los Lagos, 0.8 km E, 3.3 km N of junction of Rta. 234 with background of the other species (e.g., Patton and Smith 1994; Río Pichi Traful (−40.48, −71.59); MLP 12.VII.88.1, collected Coyner et al. 2015). So far, no evidence of past hybridization by A. Alcover, A. Carlini, and D. H. Verzi at the type locality. between O. pacificus and Argentine Octodon, necessary for in- Measurements of paratypes are given in Supplementary Data trogression to occur, is known. We refer to past hybridization SD1, Table S3. A partial (801 bp) Cytb gene sequence gath- because O. pacificus is restricted to an offshore Pacific Island, ered from paratype MVZ 201202 was deposited in GenBank which naturally prevents current hybridization. Finally, natural with accession number MT461812; a partial (828 bp) GHR selection may in both species be favoring variants of similar gene sequence gathered from paratype MVZ 201202 was nucleotide sequence at the Cytb locus, whose coding protein deposited in GenBank with accession number MT461871. forms part of respiratory chain complex (Esposti et al. 1993). It These sequences are here considered as a para-genotypes of is important to note that the nuclear locus provides a partially O. ricardojeda n. sp. resolved genealogy with most nodes lacking significant sup- Diagnosis.—A species of the genus Octodon port. In fact, only three nodes of the GHR gene tree have sig- (Hystricomorpha, Octodontidae) characterized by the fol- nificant support. Notwithstanding, the allele of O. pacificus is lowing combination of characters: dorsal pelage brownish not nested within those of southernmost Octodon, which form mottled with black and yellow; venter yellowish gray, a weakly supported clade. The latter clade falls into a polytomy sometimes with a white spot in the inguinal region; skull ro- involving three other lineages, corresponding to the alleles of bustly built; zygomatic arch inserted high and descending; O. pacificus, O. lunatus, and coastal “O. bridgesii.” This lack of anterior portion of the sphenopalatine fissure absent; maxil- resolution is probably due to the fact that few characters are var- lary sheet dorsal to the sphenopalatine fissure wide, and no- iable in the GHR fragment we used. However, it is worth noting ticeably separating the foramen into lacrimal canal from the that on average alleles of southern O. “bridgesii” diverge by fissure; alveolar sheath of the M1 covered by the maxillary 0.08% (a value similar to the other available intraspecific value: sheet, not visible laterally; foramen ovale large, without an 0.06% for O. degus), while on average alleles of southern O. external alisphenoid margin; auditory bullae small; angular 8 JOURNAL OF MAMMALOGY Downloaded from https://academic.oup.com/jmammal/advance-article/doi/10.1093/jmammal/gyaa143/6029408 by Universidad Austral de Chile-Biblioteca Central user on 10 December 2020

Fig. 4.—Dorsal, ventral, and lateral view of the skull; lateral and dorsal view of the mandible of the holotype of O. ricardojeda n. sp. (MLP 12.VII.88.2). Scale = 5.0 mm. process of the mandible elongated; molar lobes obliquely Measurements of the holotype (in mm).—Total length, 326; elongated, the lingual end of the anterior lobe of m1-2 length of tail, 134; length of hind foot (with claw), 38; length expanded. of ear, 22; weight, 195 g; total length of the skull, 44.25; D’ELÍA ET AL.—NEW SPECIES OF OCTODON (HYSTRICOMORPHA) 9 Downloaded from https://academic.oup.com/jmammal/advance-article/doi/10.1093/jmammal/gyaa143/6029408 by Universidad Austral de Chile-Biblioteca Central user on 10 December 2020 condyle-incisive length, 41.04; basilar length, 35.02; nasal Sometimes there is a white spot in the inguinal region. In the head, length = 17.03; nasal breadth = 6.24; rostrum breadth = 7.92; including the cheeks and nose, the color is similar to the dorsum, zygomatic breadth = 23.29; interorbital breadth = 8.64; frontal but the overall look is more finely agouti. The vibrissae are long, length = 13.79; length of upper diastema = 10.19; palatilar reaching the middle portion of the ears; some vibrissae are white length, 15.46; upper toothrow length = 11.12; upper incisors while others are black. The ears are moderately long and are cov- breadth (taken at the base), 3.96. ered by brownish hairs. The legs are covered by silvery white hairs. Morphological description.—A large, dark, heavy-bodied rat The front claws are not enlarged. The ungueal tufts do not exceed with a moderately short tail. The dorsal coloration is brownish mot- the manual claws but are abundant and exceed the claws on the tled with black and yellow, with individual hairs dark (15–20 mm) hind legs. The plantar surface is dark brown and finely scutelated. gray at their bases and yellowish tips. Guard hairs are longer and al- The tail base is bicolor; brownish above, well separated from the most completely blackish. The coloration is paler toward the flanks, creamy white underneath; the distal 1/3 to 1/2 is dark brown, almost where there is no black hair and guard hairs are long and whitish. black, all around. Tail is not penicillated, with distal hairs extending The belly is yellowish gray, with gray hairs tipped with yellow. only 5–8 mm from the tip.

Fig. 5.—Lateral view of the rostrum (A–E) and orbital region (F–J), ventral view of the basitemporal (K–O) and basicranial (P–T) regions. (A and P) O. ricardojeda n. sp. MLP 12.VII.88.2 (holotype); (F and K) O. ricardojeda n. sp. MLP 12.VII.88.1; (B) O. pacificus ZFMK 92.383 (paratype); (G, L, and Q) O. pacificus MZUC-UCCC 44277; (C, H, M, and R) O. degus UACH 6198; (D, I, N, and S) O. lunatus UACH 7050; (E, J, O, and T) coastal O. “bridgesii” UACH 5214. al, alisphenoid; asp, anterior sphenopalatine fissure; et, ethmoidal foramen; fo, foramen ovale; ju, jugular foramen; ms, maxillary sheet; M1, alveolar sheath of the first upper molar; nl, foramen into nasolacrimal canal; sp, sphenopalatine fissure; st, styloid process. White dotted line denotes the position of the nasal with respect to the anterior face of incisor (the molar series being horizontal); arrow indicates the external alisphenoid margin; left-right arrow denotes the basioccipital width. Out of scale. 10 JOURNAL OF MAMMALOGY Downloaded from https://academic.oup.com/jmammal/advance-article/doi/10.1093/jmammal/gyaa143/6029408 by Universidad Austral de Chile-Biblioteca Central user on 10 December 2020

The skull is broadly built, with a long diastema, descending the alveolus of the M1 and markedly separating the fissure from zygomatic arch, proportionally small auditory bullae, and a the foramen that opens into the nasolacrimal canal (Fig. 5). The nearly square braincase (Fig. 4). The interorbital constriction alisphenoid has a slender anterior apophysis expanded on the is broad, with its borders slightly divergent anteriorly and the maxilla and a frequently combined masticatory–buccinator fo- lateral ridges relatively well marked, forming a thin, translu- ramen (Figs. 5 and 6). The incisive foramina are short because cent edge. The nasals are extended only slightly in front of the they are strongly restricted by the maxilla (Fig. 6). The pre- anterior face of the incisors and bowed toward the tip in lat- maxillary septum separating the foramina is short and wide. eral profile (Fig. 5). The lacrimal bones are small and nearly The palatal bridge is short, not extended beyond the posterior square in outline. The frontoparietal suture is widely opened lobe of the M1 (Fig. 6). The mesopterygoid fossa is deeply ex- and “V”-shaped. The parietals are inflated and nearly smooth. cavated and “V”-shaped. The basicranial bones are wide, espe- The gnathic process of the premaxilla is moderate in size, not cially the basioccipital. The foramen ovale is large and lacks exceeding the anteriormost point of nasals. The zygomatic arch an external alisphenoid margin. The auditory bullae are little is inserted high, so that there is a considerable distance be- inflated, hence the jugular foramen is broad; they are short tween the lower border of the zygoma and the alveolar border and moderately expanded at their anterior portion and have a of the upper molars. Consequently, the infraorbital foramen long styloid process (Fig. 5). The mastoid process is reduced is proportionally low. The antorbital zygomatic bar is nearly (Fig. 5). vertical to slightly inclined forwards. The jugal-maxillary su- The mandible is large and heavy. The diastema is short and ture is acuminated. The jugal lateral fossa is high, occupying moderately excavated. The plane defined by the occlusal sur- most of this bone. The paraorbital process is low in lateral view face of molar series is slightly above the incisor alveolus. The and formed by the squamosal. In the orbit, the sphenopalatine notch for the insertion of the anterior medial masseter muscle fissure lacks the portion anterior to the alveolar sheath of M1 is moderately expressed and lies just below the dp4-m1. The (Fig. 5); the maxillary sheet dorsal to the fissure is wide, hiding masseteric crests are moderately demarcated. The capsular

Fig. 6.—Incisive foramina (A–E), lateral view of the basitemporal region (F–J) and the posterior portion of the mandible (K–O). (A, F, and K) O. ricardojeda n. sp. MLP 12.VII.88.2 (holotype); (B, G, and L) O. pacificus ZFMK 92.384 (holotype); (C, H, and M) O. degus UACH 6198; (D, I, and N) O. lunatus UACH 7050; (E, J, and O) coastal O. “bridgesii” UACH 5214. al, alisphenoid; ap, angular process; m, maxilla; M3, third upper molar; pt, pterygoid. Arrow indicates the anterior apophysis of the alisphenoid; white dotted line denotes the position of the angular process with respect to the postcondyloid process (the molar series being horizontal). Out of scale. D’ELÍA ET AL.—NEW SPECIES OF OCTODON (HYSTRICOMORPHA) 11 Downloaded from https://academic.oup.com/jmammal/advance-article/doi/10.1093/jmammal/gyaa143/6029408 by Universidad Austral de Chile-Biblioteca Central user on 10 December 2020 projection is well expressed and lies below the sigmoid notch. species (e.g., Massoia et al. 1981; Tammone et al. 2020), sug- The coronoid process is small and turned backward (Fig. 6). gesting that this rodent had a slightly wider distribution in the The condyloid process is broad in lateral view and located past 10,000 years. above of the coronoid process (Fig. 6). The angular process is Etymology.—The name is constructed as a noun in appo- styliform and flattened, with a dorsal groove for insertion of the sition by the union and contraction of the first and family m. masseter lateralis; it extends posteriorly beyond the condy- names of Ricardo Ojeda as an homage to him. Ricardo is an loid process (Fig. 6). The semilunar notch is well excavated and Argentine mammalogist with an extensive production cen- nearly “C”-shaped (Fig. 6). tered on the ecology and evolutionary biology of many South The upper incisors are opisthodont and short, with the American rodents, including caviomorphs. He is an active bottom of their alveolar sheath dorsal to the ventral zygomatic member of the Argentinean community of mammalogists and root, at level of the anterior portion of the DP4. The bottom of one of the founding members of the Sociedad Argentina para the alveolar sheath of the lower incisors is located behind of el Estudio de los Mamíferos (SAREM), well known for his the m3. leadership toward a more active and productive interaction Both upper and lower molars have their two lobes obliquely among the communities of mammalogists of Latin American elongated, in the anterolabial–posterolingual direction. The countries. hypoflexus is deep in occlusal view, and approaches or even Natural history.—In Parque Nacional Lanín, Ricardo reaches the labial side of the molars. The lower molars have Ojeda’s Degu was captured in semiopen areas of wet forests elongated and narrow lobes and a deep mesoflexid. The ante- dominated by Nothofagus, where animals were found in dens rior lobe is expanded at the tip, especially in m1-2. In m1, the excavated in sandy hillocks. It remains to be determined if bottom of the mesoflexid reaches the level of the bottom of the O. ricardojeda n. sp. excavates the burrows it inhabits, or only hypoflexid in occlusal view and is anterior to the latter (Fig. 7). uses those dug by other animals (Verzi and Alcover 1990). Morphological comparisons.—In its overall external ap- This species also is found on sandy hillocks within patches pearance, O. ricardojeda n. sp. looks more uniform, less of Chusquea spp. surrounded by bushes of Baccharis spp., agouti in color, and with a finer pelage, than the other species Berberis buxifolia, and Milunum spinosum, and trees of the of the genus. Cranially, O. ricardojeda n. sp. differs from the Antarctic beech Nothofagus antarctica, or within clear- remaining species of the genus by having a descending zygo- ings of coigue (Nothofagus dombeyi) forests. At Estancia la matic arch, shorter incisive foramina, no anterior division of Querencia, O. ricardojeda n. sp. inhabits a shrub steppe com- the sphenopalatine fissure, the foramen into the nasolacrimal posed by Stipa spp., neneo (M. spinosum), the spiny shrub canal markedly separated from the sphenopalatine fissure by a Colletia spinosissima, calafate (Berberis microphylla), and wider maxillary sheet, the alveolar sheath of the M1 not visible molle (Schinus molle). At Malalcahuello, Araucanía, Chile, laterally, and the foramen ovale without an external alisphe- O. ricardojeda n. sp. was captured in a semiopen area with noid margin (Figs. 5 and 6). The nasal bones are only slightly Nothofagus, growths of Chusquea, and the monkey puzzle extended in front of the anterior face of the incisors (Fig. 5). In tree (Araucaria araucana). addition, its auditory bullae are proportionally smaller, the jug- While Ricardo Ojeda’s Degu is primarily terrestrial it does ular foramina are broader, the basioccipital is wider (Fig. 5), the climb, as evidenced by droppings having been found on plat- angular process of the mandible is more extended and with a forms formed from bent reeds (Verzi and Alcover 1990). distinct notch for insertion of the masseter lateralis (Fig. 6), and Pearson (1995) reported that captive animals ate grass, clover, the lingual tip of the anterior lobe of m1-2 is expanded (Fig. 7). apple, and bread. At Estancia La Querencia, O. ricardojeda The qualitative craniodental morphology of O. ricardojeda n. sp. is the prey of black-chested buzzard-eagle Geranoaetus n. sp. is more similar to O. pacificus than to that of the other melanoleucus (Podesta et al. 2000), while at La Lipela was Octodon species (see Figs. 5–7); these two species share the found among the prey items of the Magellanic horned owl highly inserted zygomatic arch, the small anterior apophysis Bubo magellanicus (Massoia et al. 1991). of the alisphenoid, and the obliquely elongated molars lobes Other small mammals collected at the localities inhabited by (Fig. 7). However, they differ in the abovementioned traits, Ricardo Ojeda’s Degu were Abrothrix hirta, Irenomys tarsalis, which supports the taxonomic status of the new species. Loxodontomys micropus, Oligoryzomys longicaudatus, Distribution.—Extant populations of this species are known Aconaemys porteri, and Dromiciops bozinovici (Verzi and at only eight localities in western Neuquén Province, Argentina, Alcover 1990; D’Elía et al. 2016). at an elevation between ca. 500 and 1,100 m, and from the area Genetic variation.—At the Cytb locus Ricardo Ojeda’s of Reserva Nacional Malalcauello in the northeastern Araucanía Degu shows phylogeographic structure. Haplotypes from Region, Chile (Fig. 1; Appendix I). Fossil remains referable to Araucanía form a monophyletic group sister to the clade con- O. ricardojeda n. sp. were detected in several Holocene archae- taining all other haplotypes of the species (and of O. pacificus). ological sites of Neuquén and Río Negro, Argentina (Massoia Both clades show a 2.9% of divergence at the Cytb gene and et al. 1981; Pearson and Pearson 1993; Massoia and Silveira would have split at about 0.90 Ma (0.50–1.32), which is sim- 1996; Podesta et al. 2000; Tammone et al. 2020). Some of these ilar to the estimate of the split between O. lunatus and coastal records are extralimital to the current geographical range of the O. bridgesii (Fig. 2; Supplementary Data SD1, Table S2). 12 JOURNAL OF MAMMALOGY Downloaded from https://academic.oup.com/jmammal/advance-article/doi/10.1093/jmammal/gyaa143/6029408 by Universidad Austral de Chile-Biblioteca Central user on 10 December 2020

Fig. 7.—Upper right (above) and lower left (below) molars of (A) O. ricardojeda n. sp. MLP 12.VII.88.2 (holotype); (B) O. pacificus ZFMK 92.384 (holotype; both series reversed); (C) O. degus UACH 6198; (D) O. lunatus UACH 7050; (E) coastal O. “bridgesii” UACH 5214. Scale = 1.0 mm.

Because specimens from the Araucanía are morphologically structured and allocate both samples to O. ricardojeda n. sp. similar to those of Neuquén, we interpret the divergence at This taxonomic scenario should be further tested assessing the Cytb locus as intraspecific variation that is geographically more specimens and loci. D’ELÍA ET AL.—NEW SPECIES OF OCTODON (HYSTRICOMORPHA) 13 Downloaded from https://academic.oup.com/jmammal/advance-article/doi/10.1093/jmammal/gyaa143/6029408 by Universidad Austral de Chile-Biblioteca Central user on 10 December 2020

Conservation.—As currently known, Ricardo Ojeda’s We note that most other octodontid genera have not been Degu is protected at Parque Nacional Lanín (Argentina) reviewed with a comprehensive approach (i.e., one inte- and Reserva Nacional Malalcahuello (Chile) and may be grating morphologic and genetic variation within and among found in other protected areas across its distributional range. populations) and framed in the current view of species as Based on its apparently small distribution (less than 20,000 metapopulation lineages that evolve independently (de Queiroz square kilometers), and given that is known from fewer than 2007). As such, and as is common for the whole order Rodentia 10 localities, this species should be regarded as Vulnerable (D’Elía et al. 2019), the identification and description of new (VU B1ab[ii,iii]). Habitat loss and forest fragmentation are living octodontid species is likely. among the main threats for this rodent species (cf. Sobrero Finally, we expect that the morphological data generated and Tammone 2019), which had a larger distribution in the here be incorporated in future studies focused on the morpho- near past. logic evolution of caviomorph rodents. The inclusion of this data set onto such studies seems feasible given the renewed in- Final Considerations terest into comparative studies in this group (Verde Arregoitia et al. 2020). Analyses unequivocally indicate that the genus Octodon needs to be reviewed, because additional candidate species were identified. In particular, O. bridgesii, even Acknowledgments after removing O. ricardojeda n. sp., appears to encom- pass more than one species. For instance, populations We are grateful to the colleagues at different institutions that from coastal BioBio and Maule regions, here referred facilitated our study of several specimens: Bruce Patterson to as coastal O. “bridgesii,” are genetically and mor- (FMNH), Fredy Mondaca (UACh). Juliana Vianna and Daniel phologically distinct from those of typical O. bridgesii Gonzalez-Acuña provided samples and photographs of (Fig. 3). In addition, the coastal form of O. “bridgesii” O. pacificus. Andrés Parada helped us with the molecular clock cannot be assigned to O. lunatus, its sister species. analysis. The late Oliver P. Pearson early on emphasized to one Both species are genetically (Supplementary Data of us (DHV) the uniqueness of the Octodon populations from SD1, Table S1) and morphologically (e.g., O. lunatus Argentina. Two anonymous reviewers made comments on an presents a distinctive M3; Fig. 7) distinct. In addi- earlier version of this study. Financial support was provided tion, both species are chromosomally different; coastal by Fondo Nacional de Desarrollo Científico y Tecnológico O. bridgesii presents a diploid chromosomal comple- (FONDECYT) 11800366. ment of 2n = 58 (Gallardo 1992), which complement is widespread among octodontids (e.g., Reig et al. 1972; Supplementary Data Gallardo 1992), while O. lunatus has a unique 2n = 78 (Spotorno et al. 1995). We note that if the distinction Supplementary data are available at Journal of Mammalogy of coastal O. bridgesii at the species level is corrobor- online. ated, no name is available for it. Supplementary Data SD1.—Supplementary tables (ob- Our results help explain the different topologies obtained by served genetic p-distances, divergence times, and measure- Suarez-Villota et al. (2016) and Vianna et al. (2017) by the ments of the paratypes). fact that both studies differ in the lineage used to represent O. bridgesii. In fact, neither of these studies included represent- Literature Cited atives of O. bridgesii s.s.; while Suarez-Villota et al. (2016) in- Adkins, R. M., E. L. Gelke, D. Rowe, and R. L. Honeycutt. 2001. cluded specimens of coastal O. bridgesii, Vianna et al. (2017) Molecular phylogeny and divergence time estimates for major ro- used a sample of O. ricardojeda n. sp. As such, in the former dent groups: evidence from multiple genes. Molecular Biology and study what was labeled as O. bridgesii was recovered sister to Evolution 18:777–791. O. lunatus and in the later as sister to O. pacificus. Adler, G. H., and R. Levins. 1994. The island syndrome in rodent New studies aimed to assess species boundaries within populations. The Quarterly Review of Biology 69:473–490. Octodon should include the analysis of additional nuclear loci, Álvarez, A., R. L. Moyers Arévalo, and D. H. Verzi. 2017. and also may benefit from the inclusion of karyotypic data. Diversification patterns and size evolution in caviomorph rodents. Such data sets would allow us to clarify the causes (e.g., intro- Biological Journal of the Linnean Society 121:907–922. gression of the mitochondrial genome of one species into the Álvarez-Castañeda, S. T., and L. A. Nájera-Cortazar. 2019. other, fixation of ancestral polymorphisms or natural selection) Do island populations differ in size and shape compared to main- for why mitochondrial haplotypes of O. ricardojeda n. sp. and land counterparts? Journal of Mammalogy 101:373–385. O. pacificus are quite similar and closely related. By the same Ameghino, F. 1891. Mamíferos y aves fósiles Argentinas. Especies nuevas, adiciones y correcciones. Revista Argentina de Historia token, a larger sample of O. pacificus, a species known from Natural 1:240–259. only three adult specimens (two juveniles and four embryos of Angerbjorn, A. 1986. Gigantism in island populations of wood one adult female) is needed to assess the extent of an insularity mice (Apodemus) in Europe. Oikos 47:47–56. effect (Adler and Levins 1994) on its morphological distinction Ball, R. M., Jr., J. E. Neigel, and J. C. Avise. 1990. Gene geneal- (e.g., Case 1978; Angerborgn 1986; Álvarez-Castañeda and ogies within the organismal pedigrees of random-mating popula- Nájera-Cortazar 2019). tions. Evolution 44:360–370. 14 JOURNAL OF MAMMALOGY Downloaded from https://academic.oup.com/jmammal/advance-article/doi/10.1093/jmammal/gyaa143/6029408 by Universidad Austral de Chile-Biblioteca Central user on 10 December 2020

Bennett, E. T. 1832. Characters of a new genus of rodent Hutterer, R. 1994. Island rodents: a new species of Octodon from Mammalia, from Chili, presented by Mr. Cuming. Proceedings of Isla Mocha, Chile (Mammalia: Octodontidae). Zeitschrift für the Committee of Science and Correspondence of the Zoological Säugetierkunde 59:27–41. Society of London 1832, part 2:46–48. Mann, G. 1978. Los pequeños mamíferos de Chile, marsupiales, Bouckaert, R., et al. 2014. BEAST 2: a software platform for quirópteros, edentados y roedores. Gayana 40:1–342. Bayesian evolutionary analysis. PLoS Computational Biology Mares, M. A., J. K. Braun, R. M. Barquez, and M. M. Díaz. 10:e1003537. 2000. Two new genera and species of halophytic desert mammals Cañón, C., D. Mir, U. F. J. Pardiñas, E. P. Lessa, and G. D’Elía. from isolated salt flats in Argentina. Occasional Papers, Museum of 2014. A multilocus perspective on the phylogenetic relationships Texas Tech University 203:i+1–27. and diversification of rodents of the tribe Abrotrichini (Cricetidae: Massoia, E. 1979. El género Octodon en la Argentina. Neotropica Sigmodontinae). Zoologica Scripta 43:443–454. 25:36. Case, T. J. 1978. A general explanation for insular body size trends in Massoia, E., J. J. Pereiro, and C. Reboledo. 1991. Análisis de terrestrial vertebrates. Ecology 59:1–18. regurgitados de Bubo virginianus en La Lipela, departamento Chiquito, E. A., G. D’Elía, and A. R. Percequillo. 2014. Los Lagos, provincia de Neuquén. APRONA, Boletín Científico Taxonomic review of genus Sooretamys (Rodentia: Cricetidae: 19:53–57. Sigmodontinae): an integrative approach. Zoological Journal of the Massoia, E., S. Renard de Coquet, and J. Fernández. 1981. Lama Linnean Society 171:842–877. guanicoe en la economía primitiva, según registros arqueológicos Contreras, L. C., J. Torres-Mura, and J. Yáñez. 1987. verificados en la excavación de Chenque Haichol, Neuquén. INTA, Biogeography of octodontid rodents: an eco-evolutionary hypoth- IDIA 389–390:79–82. esis. Fieldiana Zoology, New Series 39:401–411. Massoia, E., and M. J. Silveira. 1996. Los micromamíferos del Coyner, B. S., P. J. Murphy, and M. D. Matocq. 2015. sitio Alero los Cipreses (Departamento de Los Lagos, provincia de Hybridization and asymmetric introgression across a narrow Neuquén, República Argentina). Revista del Museo Argentino de zone of contact between Neotoma fuscipes and N. macrotis Ciencias Naturales “Bernardino Rivadavia” 131:1–4. (Rodentia: Cricetidae). Biological Journal of the Linnean Molina, G. I. 1782. Saggio sulla storia naturale del Chili. 1st ed. Society 115:162–172. Stamperia di S. Tommmaso d’Aquino. Bologna, Italy. Darriba, D., G. L. Taboada, R. Doallo, and D. Posada. 2012. Olivares, A. I., and D. H. Verzi. 2015. Systematics, phylogeny jModelTest 2: more models, new heuristics and parallel computing. and evolutionary pattern of the hystricognath rodent Eumysops Nature Methods 9:772. (Echimyidae) from the Plio-Pleistocene of southern South America. Dayrat, B. 2005. Towards integrative taxonomy. Biological Journal Historical Biology 7:1042–1061. of the Linnean Society 87:407–415. Osgood, W. H. 1943. The mammals of Chile. Field Museum of D’Elía, G., P.-H. Fabre, and E. P. Lessa. 2019. Rodent systematics Natural History, Zoological Series 30:1–268. in an age of discovery: recent advances and prospects. Journal of Padial J. M., S. Castroviejo-Fisher, J. Köler, C. Vilà, Mammalogy 100:852–871. J. C. Chaparro, and I. De La Riva. 2009. Deciphering the prod- D’Elía, G., N. Hurtado, and A. D’Anatro. 2016. Alpha taxonomy ucts of evolution at the species level: the need for an integrative of Dromiciops (Microbiotheriidae) with the description of 2 new taxonomy. Zoologica Scripta 38:331–447. species of monito del monte. Journal of Mammalogy 97:1136–1152. Palmer, T. S. 1903. Some new generic names of mammals. Science de Queiroz, K. 1998. The general lineage concept of species, spe- 17:873. cies criteria, and the process of speciation: a conceptual unification Patton, J. L., M. N. F. da Silva, and J. R. Malcolm. 2000. and terminological recommendations. Pp. 57–75 in Endless forms: Mammals of the Rio Juruá and the evolutionary and ecological di- species and speciation (D. J. Howard and S. H. Berlocher, eds.). versification of Amazonia. Bulletin of the American Museum of Oxford University Press. New York. Natural History 244:1–306. de Queiroz, K. 2007. Species concepts and species delimitation. Patton, J. L., and M. F. Smith. 1994. Paraphyly, polyphyly, and the Systematic Biology 56:879–886. nature of species boundaries in pocket gophers (genus Thomomys). Díaz, M. M., R. M. Barquez, and D. H. Verzi. 2015. Genus Systematic Biology 43:11–26. Octodon Bennett, 1832. Pp. 1029–1035 in Mammals of South Pearson, O. P. 1995. Annotated keys for identifying small mammals America. 2. Rodents (J. L. Patton, U. F. J. Pardiñas, and G. D’Elía, living in or near Nahuel Huapi National Park or Lanin National eds.). University of Chicago Press. Chicago, Illinois. Park, southern Argentina. Mastozoología Neotropical 2:99–148. Di Rienzo, J., F. Casanoves, M. Balzarini, L. Gonzalez, Pearson, O. P., and A. K. Pearson. 1993. La fauna de mamíferos M. Tablada, and C. Robledo. 2008. InfoStat versión 2008. pequeños de cueva Traful I (provincia del Neuquén, República Grupo InfoStat, FCA. UNC. www.infostat.com.ar. Accessed 1 July Argentina). Præhistoria 1:211–224. 2017. Podestá, D. H., S. Cirignoli, and U. F. J. Pardinas. 2000. Nuevos Esposti, M. D., S. De Vries, M. Crimi, A. Ghelli, T. Patarnello, datos sobre la distribución de Octodon bridgesii (Mammalia: and A. Meyer. 1993. Mitochondrial cytochrome b: evolu- Rodentia) en la Argentina. Neotrópica 46:75–77. tion and structure of the protein. Biochimica et Biophysica Acta Poeppig, E. F. 1829. In L. F. Froriep, Schreiben des jesst in Chile 1143:243–271. reifenden hrn. Dr. Pöppig. Notizen aus dem Gebiete der Natur - Gallardo, M. H. 1992. Karyotypic evolution in octodontid rodents und Heilkunde 502:273–288. based on C-band analysis. Journal of Mammalogy 73:89–98. Rambaut, A., A. J. Drummond, D. Xie, G. Baele, and Hudson, R. R. 1990. Gene genealogies and the coalescent M. A. Suchard. 2018. Posterior summarization in Bayesian process. Pp. 1–44 in Oxford surveys in evolutionary biology phylogenetics using Tracer 1.7. Systematic Biology 67:901–904. 7 (D. J. Futuyma and J. Antonovics, eds.). Oxford University Reig, O. A., A. Spotorno, and R. Fernández. 1972. A pre- Press. New York. liminary survey of chromosomes in populations of the Chilean D’ELÍA ET AL.—NEW SPECIES OF OCTODON (HYSTRICOMORPHA) 15 Downloaded from https://academic.oup.com/jmammal/advance-article/doi/10.1093/jmammal/gyaa143/6029408 by Universidad Austral de Chile-Biblioteca Central user on 10 December 2020

burrowing octodont rodent Spalacopus cyanus Molina Valdecasas, A. G., D. Williams, and Q. D. Wheeler. 2008. (Caviomorpha, Octodontidae). Biological Journal of the Linnean ‘Integrative taxonomy’ then and now: a response to Dayrat (2005). Society 4:29–38. Biological Journal of the Linnean Society 93:211–216. Ronquist, F., et al. 2012. MrBayes 3.2: efficient Bayesian phylo- Venegas, W. 1975. Los cromosomas somáticos de Octodon bridgesii genetic inference and model choice across a large model space. Waterhouse (Rodentia, Octodontidae). Boletín de la Sociedad de Systematic Biology 61:539–542. Biología de Concepción 44:7–15. Rowe, D. L., and R. L. Honeycutt. 2002. Phylogenetic rela- Verde Arregoitia, L. D., P. Teta, and G. D’Elía. 2020. Patterns tionships, ecological correlates, and molecular evolution within in research and data sharing for the study of form and function in the Cavioidea (Mammalia, Rodentia). Molecular Biology and caviomorph rodents. Journal of Mammalogy 101:604–612. Evolution 19:263–277. Verzi, D. H., and A. Alcover. 1990. Octodon bridgesii Waterhouse, Sage, R. D., O. P. Pearson, J. Sanguinetti, and A. K. Pearson. 1844 (Rodentia: Octodontidae) in the Argentinian living mamma- 2007. Ratada 2001: a rodent outbreak following the flowering of lian fauna. Mammalia 54:61–67. bamboo (Chusquea culeou) in southwestern Argentina. Pp. 177– Verzi, D. H., M. Díaz, and R. M. Barquez. 2015. Family 224 in The quintessential naturalist: honoring the life and legacy Octodontidae. Pp. 1023–1048 in Mammals of South America. of Oliver P. Pearson (D. A. Kelt, E. P. Lessa, J. Salazar-Bravo, and 2. Rodents (J. L. Patton, U. F. J. Pardiñas, and G. D’Elía, eds.). J. L. Patton, eds.). University of California Publications in Zoology University of Chicago Press. Chicago, Illinois. 134, University of California Press. Berkeley. Verzi, D. H., A. I. Olivares, and C. C. Morgan. 2014. Phylogeny Sites, J. W., and J. C. Marshall.Jr., 2003. Delimiting species: and evolutionary patterns of South American octodontoid rodents. a Renaissance issue in systematic biology. Trends in Ecology & Acta Palaeontologica Polonica 59:757–769. Evolution 18:462–470. Vianna, J. A., et al. 2017. Record of an alleged extinct rodent: Sites, J. W., and J. C. Marshall. Jr.,2004. Operational criteria molecular analyses of the endemic Octodon pacificus from Chile. for delimiting species. Annual Review of Ecology, Evolution, and Journal of Mammalogy 98:456–463. Systematics 35:199–227. Voloch, C. M., J. F. Vilela, L. Loss-Oliveira, and C. G. Schrago. Sobrero, R., and M. N. Tammone. 2019. Octodon bridgesi. In 2013. Phylogeny and chronology of the major lineages of New Categorización 2019 de los mamíferos de Argentina según su World hystricognath rodents: insights on the biogeography of the riesgo de extinción (SAyDS–SAREM, eds.). Lista Roja de los Eocene/Oligocene arrival of mammals in South America. BMC mamíferos de Argentina. Versión digital: http://cma.sarem.org.ar. Research Notes 6:160. Accessed 23 May 2020. Wagler, J. 1832. Neue Sippen und Gattungen der Säugethiere und Spotorno, A. E., et al. 1995. Chromosome divergence of Octodon vögel. I. Säugthiere. Isis von Oken 25 11:1218–1221. lunatus and Abrocoma bennetti and the origins of Octodontoidea Wagner, J. A. 1845. Bericht Uuberdie Leistungen in der (Rodentia: Histricognathi). Revista Chilena de Historia Natural Naturgeschichte der Säugthiere während des Jahres 1844. Archiv 68:227–239. für Naturgeschichte 11:1–43. Suárez-Villota, E. Y., C. A. González-Wevar, M. H. Gallardo, Waterhouse, G. R. 1845. On the various skins of Mammalia from R. A. Vásquez, and E. Poulin. 2016. Filling phylogenetic Chile, with notes relating to them by Mr. Bridges. Proceedings of gaps and the biogeographic relationships of the Octodontidae the Zoological Society of London 12:153–157. (Mammalia: Hystricognathi). Molecular Phylogenetics and Yepes, J. 1942. Zoogeografía de los roedores octodóntidos de Evolution 105:96–101. Argentina y descripción de un género nuevo. Revista Argentina de Tamayo, M., and D. Frassinetti. 1980. Catálogo de los mamíferos Zoogeografía 2:69–81. fósiles y vivientes de Chile. Boletín del Museo Nacional de Historia Natural, Chile 37:323–399. Submitted 26 June 2020. Accepted 15 October 2020. Tammone, M. N., E. A. Lacey, and U. F. J. Pardiñas. 2020. Dramatic recent changes in small mammal assemblages from Associate Editor was Kevin Rowe. Northern Patagonia: a caution for paleoenvironmental reconstruc- tions. The Holocene 30:1579–1590. Tamura, K., G. Stecher, D. Peterson, A. Filipski, and S. Kumar. 2013. MEGA6: molecular evolutionary genetics analysis version Appendix I 6.0. Molecular Biology and Evolution 30:2725–2729. List of specimens of Octodon included in the genetic- and morpho- Thomas, O. 1920. On mammals from near Tinogasta, Catamarca, logic- (indicated with an m) based analyses. Species allocation follows collected by Sr. Budin. Annals and Magazine of Natural History, the taxonomic scheme here proposed (see text). For each specimen 9th Series 6:116–120. of Octodon we provide locality information, catalog number, and Thomas, O. 1927. The Octodon of the highlands near Santiago. GenBank accession number. Sequences gathered here are indicated Annals and Magazine of Natural History, 9th Series 19:556–57. with an * next to GenBank accession numbers (if two numbers are Tschudi, J. J. von. 1845. Untersuchungen über die Fauna Peruana. given, these are as follows: left of the dash: Cytb; right of the dash: Therologie, [part 3:77–132; part 4:133–188; part 5:189–244]. GHR). Acronyms are as follow: BM, The Natural History Museum Scheitlin und Zollikofer. St. Gallen, Switzerland. (London, United Kingdom); FMNH, Field Museum of Natural History Upham, N. S., and B. D. Patterson. 2015. Phylogeny and evolu- (Chicago, United States); LCM, Laboratorio de Citogenética, Faculdad tion of caviomorph rodents: a complete timetree for living genera. de Medicina, Universidad de Chile (Santiago, Chile); MLP, Museo Pp. 63–120 in Biology of caviomorph rodents: diversity and evolu- de La Plata (Buenos Aires, Argentina); MVZ, Museum of Vertebrate tion (A. I. Vassallo and D. Antenucci, eds.). SAREM Series A, 1. Zoology (Berkeley, United States); UACH, Universidad Austral de Buenos Aires, Argentina. Chile (Valdivia, Chile); ZFMK, Alexander Koenig Research Museum 16 JOURNAL OF MAMMALOGY Downloaded from https://academic.oup.com/jmammal/advance-article/doi/10.1093/jmammal/gyaa143/6029408 by Universidad Austral de Chile-Biblioteca Central user on 10 December 2020

(Bonn, Germany). Localities are arranged from north to south and FMNH 23212m [20]); Valparaiso, Papudo, Papudo (FMNH 23210m their numbers (between brackets) correspond to those in Fig. 1. [9]), Valparaiso, Quillota (FMNH 23903m [21]); Valparaiso, Olmué Octodon degus (n = 75): Chile: no locality specified BM 727 (FMNH 23203, FMNH 23204 [holotype], FMNH 23205m, FMNH AF422914; Atacama, Vallenar, Vallenar (FMNH 23200m, FMNH 23206m, FMNH 23207m [13]); Valparaiso, Villa Alemana, Reserva 23201m, FMNH 23202m, FMNH 24051m [1]); Coquimbo, La Serena, Nacional Lago Peñuelas (UACH 7049m, UACH 7050m, UACH 7051m Romero (FMNH 23196m, FMNH 23197m [2]); Coquimbo, Tongoy [15]). (UACH 8147m MT461733*/MT877235* [3]); Coquimbo, Ovalle, Octodon bridgesii (n = 5): Chile: O’Higgins, San Fernando, La Parque Nacional Fray Jorge (KX298478, T1052 -/AM407928, UACH Polcura (UACH 8146m MT461771*/MT877238* [22]); Maule, 1763m, UACH 2745m, UACH 2746m [4]); Coquimbo, Ovalle, Parque Curicó, Romeral, Río Teno (BM 43.20.7.5m [paralectotype], BM Nacional Fray Jorge, 3 km NE headquarters (FMNH 169655m, 55.12.24.126m [lectotype], UACH 3882m [23]); Maule, Colbún, El FMNH 169658m, FMNH 169659m, FMNH 169660m [5]); Coquimbo, Melado (UACH 8145m MT461770* [24]). Ovalle, Parque Nacional Fray Jorge, Quebrada de las Vacas (UACH Octodon “bridgesii” (coast) (n = 16): Chile: Maule, Pelluhue, 2738m, UACH 2739m, UACH 2740m, UACH 2741m, UACH 2742m, Reserva Nacional Los Queules (LCM 2433 AF405350 [25]); Maule, UACH 2743m [6]); Coquimbo, Illapel, Reserva Nacional Chinchillas Cauquenes, Fundo Unihue (UACH 5212m, UACH 5214m, UACH (KX298477 [7]); Coquimbo, Salamanca, San Agustín (UACH 4488m 5215m [26]); Ñuble, Quirihue, Las Eras (UACH 3876m, UACH 3877m, [8]); Valparaiso, Papudo, Papudo (FMNH 23186m, FMNH 23187m UACH 3878m, UACH 3879m, UACH 4485m, UACH 4486m -/AF20646, [9]); Valparaiso, Los Andes, sin localidad precisa (UACH 3890m [10]); UACH 4487m [27]); Ñuble, Quirihue, Los Remates (UACH 3881m, Valparaiso, La Calera, La Calera (FMNH 23905m [11]); Valparaiso, UACH 4328m [28]); Ñuble, Coelemu, Burca, Fundo La Madera Olmué, Parque Nacional La Campana-Ocoa (UACH 5216m, UACH (UACH 3145m, UACH 3501m, UACH 3503m [29]). 5219m, UACH 5225m, UACH 5436m [12]); Valparaiso, Olmué Octodon pacificus (n = 4): Chile, Bío Bío, Isla Mocha (MZUC44277 (FMNH 22249m, FMNH 23173m, FMNH 23174m, FMNH 23176m, KX298475/MT461872*; UACH 8148m MT461814*; ZFMK 92.383m; FMNH 23177m, FMNH 23178m, FMNH 23179m, FMNH 23181m, ZFMK 92.384m [holotype] [30]). FMNH 23182m, FMNH 23183m [13]); Valparaíso, Tiltil (KX298481, Octodon ricardojeda n. sp. (n = 20): Chile: Araucanía, Curacautín, KX298482 [14]); Valparaíso, Villa Alemana, Reserva Nacional Río Colorado (UACH 3885m [31]); Araucanía, Caracautín, Lago Peñuelas (KX298479, KX298480, UACH 6193m, UACH Entrada a la Reserva Nacional Malalcahuello (UACH 8143m 6194m, UACH 6195m -/AF520647, UACH 6196m, UACH 6197m, MT461800*/MT461863*, UACH 8144m MT461801/MT877237* UACH 6198m, UACH 6199m, UACH 6200m, UACH 6201m [15]); [32]). Argentina: Neuquén, Aluminé, Estancia La Querencia (MLP Región Metropolitana, Santiago (K61 -/AF520648 [16]); Región 22-II-00-1m, MLP 22-II-00-2m, MLP 22-II-00-3m, MLP 22-II-00-4m, Metropolitana, Santiago, 2.5 km NE Cerro Manquehue, Trappist MLP 22-II-00-5m [33]); Neuquén, Huiliches, Arroyo del Escorial Monastery (FMNH 119611m, FMNH 119613m, FMNH 119614m, (MVZ 164034m, MVZ 164036m [34]); Neuquén, Huiliches, track FMNH 119616m, FMNH 119625m, FMNH 119626m, FMNH 119627m, to Huanquihué Volcano, Laguna Verde (MLP 12-VII-88-6m, MLP FMNH 119628m, FMNH 119629m, FMNH 119642m, FMNH 119643m, 12-VII-88-7m [35]); Neuquén, Huiliches, Parque Nacional Lanin, FMNH 119645m, FMNH 119646m, FMNH 119647m, FMNH 119648m, Lake Curruhué Chico (MLP 12-VII-88-1m, MLP 12-VII-88-2m [hol- FMNH 35902m [17]); Región Metropolitana, Santiago, San Cristóbal otype], MLP 12-VII-88-3m, MLP 12-VII-88-4m, MLP 12-VII-88-5m (FMNH 35902m, FMNH 35904m [18]); Región Metropolitana, Maipú, [36]); Neuquén, Los Lagos, 0.8 km E, 3.3 km N of junction of Route Quebrada de La Plata (UACH 6191m, UACH 6190m [19]). 234 with Rio Pichi Traful (MVZ 201202m MT461812*/MT461871* Octodon lunatus (n = 16): Chile: Coquimbo, Ovalle, Parque [37]); Neuquén, Los Lagos, Refugio 3.5 km N and 1.5 km E Estancia Nacional Fray Jorge (KX298476, LCM 2315 AF2277514, UACH Paso Coihue (MVZ 184958 KJ742651/MT461868*, MVZ 184959 2747m [4]); Coquimbo, Ovalle, Parque Nacional Fray Jorge, Quebrada MT461808*, MVZ 184960 MT461809*, MVZ 184961 MT461810*, de las Vacas (UACH 2747m [6]); Valparaiso, La Ligua (FMNH 23211m, MVZ 184963 MT461811* [38]).