sign in sign up
<<
Home , Catonyx, Macrauchenia, Quaternary extinction

Science Reviews 195 (2018) 21e31

Contents lists available at ScienceDirect

Quaternary Science Reviews

journal homepage: www.elsevier.com/locate/quascirev

The Casa del Diablo cave (Puno, ) and the late demise of in the Andean Altiplano

* Natalia A. Villavicencio a, c, , Lars Werdelin b a Departamento de Ecología, Facultad de Ciencias Biologicas, Pontificia Universidad Catolica de , Av. Libertador Bernardo O'Higgins 340, Santiago, Chile b Department of Paleobiology, Swedish Museum of Natural History, Box 50007, S-104 05, Stockholm, Sweden c Instituto de Ecología and Biodiversidad (IEB), Casilla, 653, Santiago, Chile article info abstract

Article history: During the Late Quaternary event lost ~83% of all its mega- Received 15 February 2018 faunal genera. As in other regions of the world, the debate about the possible drivers behind these ex- Received in revised form tinctions revolves around the role of arriving into the continent and on the effect of the climatic 18 June 2018 changes characteristic of the Pleistocene- transition. The availability of precise chronological Accepted 8 July 2018 information, in order to estimate the timing of extinction of the different taxa affected, is critical for Available online 14 July 2018 solving such debate. Here we present an updated study of the late Pleistocene mammalian deposits from Casa del Diablo Cave (CdD) in the Altiplano of Peru. The study includes an updated list of the mammalian Keywords: fi Megafaunal faunas found in the cave and 11 taxon-speci c XAD radiocarbon dates from extinct and extant mega- Late Pleistocene fauna bones. We compare this new chronology to the timing of major environmental changes and Peruvian altiplano arrival in the area, as well as with other megafaunal discoveries from the high . The radiocarbon Casa del Diablo cave dates from CdD fall in the time window between 23 and 12.8 cal kyr BP. Compared to other records of extinct megafauna in the high Andes, the one from CdD presents in general younger occurrences. No temporal overlap between humans and extinct megafauna emerges from comparing first dates of appearance of humans in the Altiplano, and last appearance dates of extinct megafauna from CdD. However, the possibility of temporal overlap among the records becomes evident when we compare confidence intervals calculated to estimate true times of human arrival and megafaunal local extinctions. © 2018 Elsevier Ltd. All rights reserved.

1. Introduction known to have inhabited the continent during the late Pleistocene (Brook and Barnosky, 2012). At the end of the Pleistocene South America was inhabited by a Globally, the debate about the causes behind these extinctions great diversity of endemic and North American immigrant large- revolves around the impact of modern humans migrating out of sized (>44 kg), many of which became extinct around into other continents, the role of environmental changes the transition to the Holocene Epoch in what is known as the Late proper to the last glacial- transition, and combinations Quaternary (LQE; Martin and Wright,1967, Martin, of these (Martin and Steadman, 1999; Cione et al., 2003). The 1990; Martin and Klein, 1984; Martin and Steadman, 1999; general picture at the continental level in South America suggests Barnosky et al., 2004; Koch and Barnosky, 2006). This worldwide that the timing of megafaunal extinctions approximately coincides event was marked by the global disappearance of 90 genera of with the major climatic changes of the Pleistocene-Holocene megafauna (Koch and Barnosky, 2006). Of all major regions of the transition between ~14 and 9 cal kyr BP, and also with the first world, South America was the most affected by megafauna ex- evidence of human presence dated at a continental level between tinctions during the LQE, losing ~83% of all genera of megafauna 18 and 11.5 cal kyr BP (Barnosky et al., 2004; Koch and Barnosky, 2006). Nevertheless, more specific analyses at regional scales sug- gest slightly different patterns of extinction in different eco-regions of the continent, with last appearances of megafauna occurring * Corresponding author. Departamento de Ecología, Facultad de Ciencias Bio- earlier (between 45 and 15 cal kyr BP) than human arrival logicas, Pontificia Universidad Catolica de Chile, Av. Libertador Bernardo O'Higgins (~11.5 cal kyr BP) in the northern regions of South America 340, casilla 114-D, Santiago. Chile. E-mail addresses: [email protected], [email protected] (N.A. Villavicencio). (Barnosky and Lindsey, 2010) and later at higher latitudes, where https://doi.org/10.1016/j.quascirev.2018.07.013 0277-3791/© 2018 Elsevier Ltd. All rights reserved. 22 N.A. Villavicencio, L. Werdelin / Quaternary Science Reviews 195 (2018) 21e31 the last dates on extinct megafauan fall between 12.5 and 10 cal kyr megafauna, as well as medium-sized and small rodents BP, after humans had arrived there (13.3e12.9 cal kyr BP) and when (See section 3). The non-mammalian fauna is limited to a diversity climatic changes were more severe (14.5e11 cal kyr BP) of bird bones. The collection from Casa del Diablo in the Swedish (Villavicencio et al., 2016; Barnosky and Lindsey, 2010). Museum of Natural History comprises 1134 numbered specimens, While the fossil record of late in South but boxes with dozens of rodent postcranial elements are collected America is abundant, especially for the subtropics (Farina~ et al., under single numbers, so that the actual number of specimens is 2013), the lack of chronological information on the fossil deposits much greater. is what limits accurate analyses of the extinction process. Several As it is common in cave deposits, the bone accumulations of studies have highlighted the usefulness of taxon-specific radio- Casa del Diablo showed signs of mixing among the different layers carbon dates in documenting the last appearances of some taxa of excavated, a situation that was originally noted by Nordenskiӧld in megafauna and their relation to changes in the environment and to his 1908 publication, in which he describes the occurrence of the timing of human arrival in the different regions of the continent modern sheep bones mixed with Onohippidium (now cf. (e.g., Prado et al., 2015; Villavicencio et al., 2016; Barnosky et al., devillei) in deeper layers of the excavation and also mentions the 2016). These synthetic works have also emphasized how some presence of an isolated (now ) tail bone on areas of the continent are severely lacking in available chronolog- the floor of the cave. The majority of the remains of this taxon were, ical information when compared to the areas with the most com- however, located 1 m below the surface of the cave. plete records. This is the case of the South American Altiplano. Nordenskiӧld proposed that small mammals and bird bones Referring strictly to the area defined as Altiplano (see section 5), were accumulated in the cave by raptors, a process that was still only a few sites containing extinct megafauna remains can be underway when he visited the cave. The skeletal remains of large assigned to the late Pleistocene (Fig. 1); from these no publication mammals could, on the other hand, have been dragged into the reports taxon-specific radiocarbon dates for extinct megafauna. cave by large predators, a hypothesis that is supported by the type The paleontological locality of Casa del Diablo (CdD) Cave, of skeletal remains found (mostly limbs), the stage of development located in the southern Peruvian Altiplano (Fig. 1), was first of many of the bones (young individuals are common in the described by the Swedish ethnographer Erland Nordenskiӧld collection), and the types of marks found on many of the elements, (1877e1932) who visited the area between 1904 and 1905 during which can be attributed to the action of carnivores rather than to one of several expeditions to South America. During his visit to the other sources. More recent studies identified in the cave the pres- cave, Nordenskiӧld collected numerous specimens of extant and ence of a peculiar climbing , Diabolotherium extinct mammals and birds. These are currently stored at the nordenskioldi (Pujos et al., 2007). Given the morphological charac- Department of Paleobiology, Swedish Museum of Natural History teristics of this taxa, is very possible it had used the cave as midden. (Naturhistoriska riksmuseet) in Stockholm, Sweden under spec- The preservation of the bone remains is remarkable (Fig. 2BeI). imen prefixes PAL-PZ M. One of the most remarkable features of The fresh looking appearance of the elements also caught Nor- these collections is the outstanding preservation of the bones, as desnkiӧld's attention during the excavations the cave, at which documented a few later (Nordenskiӧld, 1908). This publica- time he performed a test by dissolving an element of Onohippidium tion, together with a few revisions of particular faunal elements (Hippidion cf. devillei) in hydrochloric acid to see the state of from the cave (Sefve, 1910; Marshall et al., 1984; Pujos et al., 2007; mineralization of the bone (Nordenskiӧld, 1908). The good quality Pujos and Salas, 2004), represent the only scientific publications on of the bones is also reflected in the C/N ratio obtained as part of our the fossils found at this site until now. analyses (Table 1, section 4). Here we present an updated list of the mammalian fauna from Casa del Diablo, together with 12 radiocarbon dates from bones of 2.2. Evidence of humans in the cave? extinct and extant mammals from this collection. All of the dates were obtained from single amino acid samples extracted from bone Nordenskiӧld identified a layer bearing cultural evidence rep- collagen and represent the first published geochronological infor- resented by an obsidian arrowhead and a few potsherds. He mation on this site and the first taxon-specific radiocarbon dates on considered this evidence insubstantial and not associated with the extinct megafauna from the Altiplano. Finally, we discuss this new extinct fauna, as it lay above the layers containing Onohippidium chronological information in the context of environmental changes (Hippidion cf. devillei) and Catonyx. At the same time, he mentions that took place in the Altiplano at the Pleistocene-Holocene tran- the absence of anthropic marks on the bones (Nordenskiӧld, 1908). sition, a time when the megafauna from Casa del Diablo was dis- appearing from the area given the chronological information 3. Updated mammalian faunal list for Casa del Diablo. presented here. *Indetermined, but assignable only two one of the two taxa mentioned 2. Casa del Diablo Casa del Diablo cave lies at an altitude of >3800 masl (meters Megatherioidea above sea level) and is located north-west of Titicaca, near the Diabolotherium nordenskioldi town of Tirapata in the Puno Department of Peru (Nordenskiӧld, 1908, Fig. 1). Carved in limestone, the cave is 31 m long, 22 m sp. wide, with a greatest height of 7e8 m, and includes two main en- trances of which the largest is 8 m wide. The bulk of the bones Catonyx sp. collected were found in the left side of the cave and also among the Megatheriidae two major fallen blocks (Nordenskiӧld, 1908, Fig. 2A). sp. indet. Artiodactyla 2.1. About the bone deposits in the cave Cervidae Agalmaceros wingei The majority of the bones excavated from the site belong to mammals, among which there is both extinct and extant Lama guanicoe N.A. Villavicencio, L. Werdelin / Quaternary Science Reviews 195 (2018) 21e31 23

Fig. 1. Map of the Altiplano showing the location of Casa del Diablo and other sites mentioned in the text. Star: Casa del Diablo. Red dots: sites with taxon-specific dates on extinct megafauna (Appendix A, Supp. Table 1), 7: Pikimachay cave (MacNeish et al., 1970), 8: Cueva Rosello, 9: Trigo Jirka (Shockey et al., 2009). Yellow dots: Late Pleistocene paleon- tological sites with megafauna: 1: Tuina (Núnez~ et al., 2002), 2: Salar de Surire (Labarca, 2015), 3: Ulloma (Hoffstetter, 1986), 5: Quebrada de Cachimayu, locality No. 2 (Frailey and Campbell, 1980). Grey dots: Pleistocene paleontological sites with megafauna: a) Anzaldo, b) Ayo Ayo, c) Ayusbamba, d) Azangaro cave, e) Betanazos, f) Carguaicollo, g) Cerro Wimpillay, h) Charana,~ i) Cusco, j) Huanaro, k) Llalli, l) North-West of Oruro, m) Paria, n) Puchuni, n)~ San Sebastian, o) Yako (Eaton, 1914; Hoffstetter, 1952, 1986; Ramirez Pareja, 1958; Marshall et al., 1984; Marshall and Semper e, 1991; Alberdi et al., 2004; Pujos and Salas, 2004). Black dots: archaeological sites with evidence of humans between 12.7 and 8 cal kyr BP (See Appendix A, Supp. Table 2 for a list of the localities on the map). Blue dots: lake records, 4: Lake Titicaca, 6: Lake Pacucha. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

Vicugna vicugna Mustelidae Tayassuidae sp. indet. sp. indet. Rodentia Perissodactyla Cricetidae Punomys lemminus Hippidion cf. devillei Chinchillula sahamae Carnivora Neotomys ebriosus Canidae Akodon sp. “” peruanus (¼ Lycalopex culpaeus; Prevosti and Phyllotis sp. Forasiepi, 2018) Caviidae Galea musteloides onca or Puma concolor* Microcavia niata 24 N.A. Villavicencio, L. Werdelin / Quaternary Science Reviews 195 (2018) 21e31

Fig. 2. A) Plan of Casa Del Diablo cave (modified from Nordenskiold,€ 1908). B) right humerus* (M4286) of the megatherioid Diabolotherium nordenskioldi in anterior view. C) Antler (M4090) of the Agalmaceros wingei. D) right humerus (M4446) of the scelidotheriine Catonyx sp. in anterior view E) left upper toothrow (M1905b) of the south American horse Hippidion cf. devillei in occlusal view. F) Left metacarpal (no catalog number) of the camelid Lama guanicoe in anterior view. G) Left metacarpal (no catalog number) of the camelid Vicugna vicugna in anterior view. H) Right hemimandible* (M4098) of Hippidion cf. devillei in medial view. I) left astragalus* (M4290) of Megatherium sp. in anteromedial view. *Radiocarbon dated element.

Chinchillidae dremel 3000. A small portion of the sample was used to perform Lagidium sp. nov nitrogen content analyses. The nitrogen content (%N) in whole bone Echimyidae helps to assess the preservation and quality of the bone and gen et sp. nov. therefore its likelihood of providing usable radiocarbon dates (Stafford et al., 1988, Table 1). Prior to AMS radiocarbon dating, pieces of bone were decalcified to extract collagen and then puri- fied using XAD-2 resin according to the method described by 4. Radiocarbon dating of bones from Casa del Diablo Stafford et al. (1988). Detailed sample processing can be found in Appendix A (Supplemental Material). Small pieces of bone were cut and externally cleaned using a N.A. Villavicencio, L. Werdelin / Quaternary Science Reviews 195 (2018) 21e31 25

Table 1 Radiocarbon dates from Casa del Diablo. All are XAD and AMS dates (See section 4).

Collection Taxa Element 14C Lab Carbon 14C age Cal mean (cal yr Cal 2s range (cal %N whole C:N bone C:N whole number number (mg) BP) yr BP) bone collagen bone

M1910 Hippidion cf. devillei left CAMS 1090 10980 ± 90 12830 12700e13020 3.24 3.10 4.05 hemimandible 175037 M1917c Hippidion cf. devillei right CAMS 1090 11740 ± 100 13530 13330e13740 3.14 3.00 3.62 hemimandible 175035 M4103 Hippidion cf. devillei right III CAMS 840 11980 ± 100 13780 13550e14060 3.73 3.00 3.35 metacarpal 175039 M4093 Hippidion cf. devillei mandible CAMS 830 11980 ± 100 13780 13550e14060 3.68 3.07 3.50 175038 M4099 Hippidion cf. devillei right CAMS 830 12400 ± 80 14410 14070e14860 3.63 3.05 3.41 hemimandible 175734 M4098 Hippidion cf. devillei right CAMS 950 12480 ± 110 14580 14140e15070 3.55 3.08 3.60 hemimandible 175036 M4200 Vicugna vicugna right femur CAMS 960 12780 ± 90 15170 14770e15510 3.01 ND 3.49 175750 M1937a Lama guanicoe left femur CAMS 970 13260 ± 90 15880 15580e16180 3.13 ND 3.44 175751 M4445 Catonyx sp. right CAMS 1230 13690 ± 100 16460 16150e16840 3.92 2.97 3.36 hemimandible 175745 M4286 Diabolotherium right humerus CAMS 920 16300 ± 180 19620 19150e20080 3.46 3.00 3.43 nordenskioldi 175034 M5202 Punomys lemminus right CAMS 840 16870 ± 150 20300 19920e20680 ND ND ND hemimandible 175752 M4290 Megatherium sp. left astragalus CAMS 810 19180 ± 260 23060 22470e23660 3.06 3.00 3.66 175033

All sample preparation was done in the Barnosky laboratory in 5. The Altiplano: present and past the Department of Integrative Biology, University of California (UC) Berkeley, and at the Center for Accelerator Mass Spectrometry The Altiplano is the area resulting from the broadening of the (CAMS) in Lawrence Livermore National Laboratories, California, Andes above 3000 masl between 13 and 27S(Fornari et al., 2001; USA. Allmendinger et al., 1997). According to the hydrological setting of All radiocarbon dates were calibrated using Calib 7.02 (Stuiver the different basins in the plateau, it can be divided into the and Reimer, 1986e2014) and the calibration curve for the South- southern Altiplano, which includes the Uyuni, Coipasa, and Popoo ern Hemisphere SH13 (Hogg et al., 2013). All dates are discussed as lake basins, and the northern Altiplano, which consists of the calibrated thousands of years before present (cal kyr BP). Titicaca lake basin (Fig. 1).

5.1. Present day Altiplano 4.1. Radiocarbon dating results 5.1.1. Climate A total of 12 radiocarbon dates were obtained (Table 1), among The Altiplano experiences low temperatures, low air density, which there is one date for an extant rodent (Punomys lemminus) high radiative input, and extreme conditions of aridity most of the and two dates for the extant camelids Lama guanicoe and Vicugna with the exception of the austral summer (Aceituno, 1993). vicugna. The rest of the dates are for extinct megafauna. From November to March, intense convective storms bring signif- The values of total nitrogen content in whole bone (Table 1)are icant rainfall to the plateau (Garreaud et al., 2003), fed by moisture those of well conserved modern bone (Stafford et al., 1988) which coming from the South American Summer Monsoon system (Zhou proves the high level of bone preservation in this deposit. The good and Lau, 1998). These storms are related to the intense heating of preservation of the bones may be related to the cold and dry con- the land surface in the region and the consequent destabilization of ditions existing at high altitudes in the Andes. the local troposphere, coupled with the establishment of an upper- The calibrated dates fall into the time interval between 24 and level easterly wind flow that helps to transport moisture from the 12.5 cal kyr BP (Table 1, Fig. 3A), covering the time period from the interior part of the continent (Garreaud et al., 2003; Vuille et al., LGM to the end of the Pleistocene. Although there is no strati- 2003; Falvey and Garreaud, 2005). Dry mid-tropospheric westerly graphic information about the excavation from Casa del Diablo and flow during the rest of the year does not sustain any convective therefore we do not know if the bones analyzed cover the entire activity (Garreaud, 2009). There is a gradient in the amount of temporal range of Nordenskiold's€ excavation, the radiocarbon time summer precipitation along the Altiplano, with wetter conditions interval provided here gives minimum temporal bounds to bone in the northern portion of the plateau, where the Titicaca basin is deposition in the cave. located, and drier conditions in the southern part, which are re- The oldest date in our record corresponds to the megatheriid flected in the drier characteristics of the Uyuni and Coipasa salt flats Megatherium sp., one of the largest mammals found in the cave, in (Garreaud, 2009). The “El Nino~ Southern Oscillation” with an age of ~23 kyr BP. The youngest record is for a horse, (ENSO) affects the amount of summer rainfall over the Altiplano, by Hippidion cf. devillei. Worth noting is that the dates on extant modulating the intensity and direction of the tropical easterly megafauna (L. guanicoe and V. vicugna) and microfauna winds which are ultimately related to the variability of the sea (P. lemminus) fall among the dates for extinct fauna and show the surface temperatures over the tropical Pacific(Garreaud and presence of these taxa in this part of the Altiplano since at least the Aceituno, 2001). This results in El Nino~ years showing negative very late Pleistocene (Table 1). precipitation anomalies and La Nina~ years presenting positive ones 26 N.A. Villavicencio, L. Werdelin / Quaternary Science Reviews 195 (2018) 21e31

Fig. 3. A) Chronology of mammalian faunas from Casa del Diablo cave, including extant micro and megafauna and extinct megafauna. B) Chronology of human presence in the Altiplano based on dates from Supplementary Table 1, SA: Southern Altiplano, NA: Northern Altiplano. C) Taxon-specific radiocarbon dates from sites located in the surroundings of the Altiplano: * from Shockey et al. (2009), a from MacNeish et al.(1970). D) Summary of vegetation changes in the area around Lake Titicaca estimated from lake records, fire icons indicate the onset and increase of fire events estimated from charcoal particles (Paduano et al., 2003). E) Lake Titicaca level estimations from Baker et al. (2001), higher proportion of benthic diatoms ¼ drier conditions. F) Variations in oxygen stable isotopes (d1⁸O) from Sajama as an indicator of atmospheric temperature over the Altiplano, more negative values ¼ colder conditions (Thompson et al., 1998). G) Sporormiella record from Lake Pacucha, higher concentration ¼ more mega- present (Rozas-Davila et al., 2016). The GRIWM best-estimates of the time of of Hippidion cf. devillei (and of arrival of humans) is indicated by the grey boxes, with the 95% confidence band depicted by the grey shaded area around the mean (black dashed line ¼ most probable time of extinction).

(Vuille et al., 2000; Garreaud and Aceituno, 2001). There is also a dominated by the genera Plantago (Plantaginaceae), Ephedra tendency towards warmer temperatures in the Altiplano during El (Ephedraceae), members of the Caryophyllaceae family, the genera Nino~ years (Garreaud, 2009). Mean annual temperatures below Gynoxys and Baccharis (Asteraceae), Azorella (Apiaceae), and bunch 4000 masl vary between 7 and 10 C in the Altiplano (Paduano grasses like Stipa and Festuca () (Reese and Liu, 2005). The et al., 2003). small Andean tree Polylepis (Rosaceae), can also occur up to 5100 masl (Kessler, 1995; Braun, 1997). Going to lower elevation, the 5.1.2. Vegetation standard puna formation is found between 3900 and 4300 masl The physical setting and climatic characteristics of the Altiplano (Tosi, 1960). It is slightly warmer than the puna brava, exhibiting a underlie the existence of the Puna vegetation, a grassland mean annual temperature of 7 C and a mean annual precipitation e dominated by members of the Poaceae family, such as of 400 800 mm (Garreaud et al., 2003; Hansen et al., 1984; Reese the genera Stipa and Festuca (Hansen et al., 1984). The Puna can be and Liu, 2005). At this altitude, the grass cover is taller and there divided in three different vegetation belts, each of them occurring is a greater diversity of trees and shrubs, like Cupressus and at different altitudes between 3300 and 5000 masl (Reese and Liu, members of the Ericaceae family. Other families present are Bras- 2005). These are characterized by varying plant communities that sicacea, Asteraceae, Gentiacsea, and Valerianaceae family (Paduano are ultimately determined by the amount of precipitation received et al., 2003; Reese and Liu, 2005). Finally, the sub-puna occurs and the different temperatures experienced at different altitudes in between 3300 and 3900 masl, where Dodonaea (Sapindaceae), the Altiplano (Latorre et al., 2006). The highest plant community, Rapanea (Primulaceae), and Alnus (Betulaceae) are the most com- the superpuna or ‘puna brava’ (Troll, 1968), occurs between 4300 mon elements associated with grasses. The puna (puna-brava, and 5000 masl in the sub-nival level of the Andes (Graf, 1981). At standard puna, and sub-puna) is limited to the east by the yungas this elevation there is a mean annual temperature of 4 C(Reese vegetation below 3000 masl and to the west by the desert (shrub and Liu, 2005) and a mean annual precipitation of 500 mm in the and absolute deserts). The continuity of the puna vegetation in the ‘ ’ northern Altiplano (Hansen et al., 1984). The puna brava vegetation Altiplano is interrupted by marsh-like communities ( bofedales ) fl is characterized by a sparse plant cover (Paduano et al., 2003), that occur in areas where water accumulates or ows permanently N.A. Villavicencio, L. Werdelin / Quaternary Science Reviews 195 (2018) 21e31 27

(Baied and Wheeler, 1993). Bofedales, cushion-peat habitat, are conditions (Paduano et al., 2003). This variability continues all dominated by hydrophilous taxa including tussock grasses, yaretas throughout the Holocene with two prominent dry phases occurring (the Apiaceae Azorella yarita), Cyperaceae, Juncaceaea, and aquatic between 11.5 and 10 cal kyr BP and 7.9e3.1 cal kyr BP (Baker et al., plants like Isoetes (Isoetaceae) and Myriophyllum (Haloragaceae) 2001; Paduano et al., 2003). Regional warming and drying is (Reese and Liu, 2005). recorded in the Sajama ice record between 22 and 14.3 cal kyr BP which culminates with a cold reversal between 14 and 11.5 kyr BP 5.1.3. Megafauna (Thompson et al., 1998), cooling that is also recorded further north The current vegetation described above are capable in the Huascaran Ice record in Peru (Thompson et al., 1995). of sustaining six that fall into the commonly used definition of mammalian megafauna (>44 kg according to Martin and Klein, 5.2.2. Vegetation (Fig. 3D) 1984). Among these, there are the two species of South American Along with the changes in climatic conditions over the Alti- wild , Lama guanicoe and Vicugna vicugna, the two species of plano, there where major changes in the vegetation in the area as domestic camels, Lama glama and Vicugna pacos, the north Andean well. The sedimentary record from Lake Titicaca shows low deer (or Taruka), Hippocamelus antisensis, and the cat Puma con- concentration before 21.5 cal kyr BP suggesting the presence of a color. The camelid Vicugna vicugna is a grazer and the only ungulate sparsely vegetated area around the lake, which would have been exhibiting continuously growing incisors. Vicunas~ today have a closely related to the very cold conditions over the Altiplano around very restricted distribution inhabiting in the Altiplano from that time (Paduano et al., 2003). After 21.5 cal kyr BP puna brava southern Peru to northern Chile over 3500 masl; however, this vegetation was developed around Lake Titicaca, which changed to species had a much wider distribution during the late Pleistocene, standard puna vegetation at 17.7 cal kyr BP, when temperature was reaching as far as Tierra del Fuego on the southern tip of the warmer, around 2e3 C colder than today (Paduano et al., 2003). A continent (Weinstock et al., 2009). On the other hand, today the major shift in vegetation occurs around 13.2 cal kyr BP when the , Lama guanicoe shows a broader geographic distribution pollen record shows an increase in the sub puna pollen types, and a much more flexible diet, including bushes and grasses mixed with fewer elements attributed to puna brava and standard (Eisenberg and Redford, 1989). The cervid Hippocamelus antisensis puna vegetation belts (Paduano et al., 2003). The analysis of char- is found in the high Andes of , Bolivia, Chile, and Peru coal particles in the sedimentary record from Lake Titicaca places between 2000 and 4000 masl, inhabiting dry and bushy places in the onset of fire events in the area at 17.7 cal kyr BP with fires the high Andean , usually present in small valleys with becoming a regular feature of the landscape later, at 13.7 cal kyr BP running water (Iriarte, 2008). The only large predator present in (Paduano et al., 2003). The onset and increase in the frequency of this ecosystem is the felid Puma concolor. The puma is considered a fires around Titicaca has been related to the increase in fuel load as highly adaptable cat, inhabiting all sort of environments from the landscape was changing from a sparsely vegetated area to a one all through North, Central and South America to the dominated by puna elements (Paduano et al., 2003). southern tip of the continent (Sunquist and Sunquist, 2002), According to the record from Lake Titicaca, the transition from including in the high Andes up to 5800 masl (Sunquist and the Pleistocene to the Holocene seems to have been gradual at least Sunquist, 2002). for the vegetation (Paduano et al., 2003) however, an atmospheric and regional cooling event is observed in the Sajama and Huascaran 5.2. Late Pleistocene Altiplano (Fig. 3) ice records (Thompson et al., 1995, 1998, Fig. 3F). Present day vegetation was probably established as early as 13.2 cal kyr BP 5.2.1. Climate(Fig. 3E and F) (Paduano et al., 2003). The onset of deglaciation in the Altiplano is estimated around 21 cal kyr BP (Mercer, 1984; Seltzer, 1992; Baker et al., 2001). Prior 5.2.3. Megafauna to that time the area was in a full glacial state with extensive gla- Along with the finds at Casa del Diablo cave, there are a few ciations and a snowline as low as 4350 masl, 500 m lower than other paleontological sites in the Altiplano documenting the pres- present day limit (Seltzer, 1990). Temperatures during the Late ence of extinct megafauna during the late Pleistocene (Fig. 1), Glacial in the tropics were 5e8 C colder than present day mean although the chronological information for these sites requires (Ybert, 1988, 1992). The sedimentary record from Lake Titicaca, the further investigation. The presence of patachonica closest paleoclimate record to Casa del Diablo cave, shows evidence (), Megatherium sp. (Megatheriidae), Scelitothe- of high lake level during the Late Glacial and rium sp. (cf. Catonyx sp., Mylodontidae), and Mazama sp. (Cervidae) (LGM). This phenomenon began ~30 cal kyr BP in the plateau has been reported at Quebrada de Cachimayu Locality No. 2 (Frailey (Seltzer et al., 2002), suggesting wet conditions over the northern and Campbell, 1980, Ramirez Pareja, 1958), although some of these Altiplano during this time period (Baker et al., 2001). The same identifications are questionable. These deposits are assigned to the pattern of wetter conditions during the LGM is estimated from the late Pleistocene according to a wood fragment radiocarbon dated to high lake level recorded in Junin Lake (Peru) in the northern limit of 43.3 cal kyr BP and found in association with some of the faunal the Altiplano (Seltzer et al., 2002) and also from the ice core record remains (Frailey and Campbell, 1980). Further south, from localities from Sajama Mountain in the Bolivian Altiplano (Thompson et al., near Ulloma and Rio Desaguadero in Bolivia, the presence of Meg- 1998). Wet conditions over the Altiplano could be attributed to a atherium, Catonyx, , Macrauchenia, the gomphotheriid “La Nina~ ”-like climatic state since there is record of a 5 C cooling in , and Hippidion has been reported in assigned the sea surface temperatures over the equatorial Pacific during the to the Lujanian land age (<126 ka BP; Hoffstetter, 1986). A LGM (Hostetler and Mix, 1999). Palynological evidence from Lake very late Pleistocene Hippidion was reported recently from Salar del Titicaca suggests that wet conditions persisted until 17.7 cal kyr BP Surire in the Chilean Altiplano (Labarca, 2015). (Paduano et al., 2003). At the same time, humid conditions are Many other sites in the Altiplano bearing megafaunal remains recorded in the southern Altiplano between 18.1 and 14.1 cal kyr BP, have been assigned to the Pleistocene (Fig. 1) and they demonstrate when the Tauca paleolake shows its highest level in record in the the presence of the same taxa as listed above plus the addition of southern Altiplano (Placzek et al., 2013). Between 17.7 and Plaxhaplous (Glyptodontidae), (Mylodontidae), 15 cal kyr BP, Lake Titicaca exhibited high variability in its water Charitoceros cf. tarijensis (Cervidae), waringi (Gom- level, although never below overflow, but always suggesting drying photheriidae), and the extant Panthera onca (Felidae) (Eaton, 1914; 28 N.A. Villavicencio, L. Werdelin / Quaternary Science Reviews 195 (2018) 21e31

Hoffstetter, 1952, 1986; Ramirez Pareja, 1958; Marshall et al., 1984; megafauna and humans in the Altiplano. Furthermore, there are Marshall and Semper e, 1991; Alberdi et al., 2004; Pujos and Salas, 100 years between the last record of Hippidion cf. devillei 2004). (12.8 cal kyr BP) and the earliest record of humans from Cueva Bautista in Bolivia (12.7 cal kyr BP). 5.2.4. Human arrival (Fig. 3B) Possible temporal coexistence between humans and Hippidion The earliest evidence of human presence in the Altiplano comes cf. devillei emerges when we compare confidence intervals (CI, grey from Cueva Bautista (Bolivia) in the southern Altiplano, dated to boxes) calculated using a Gaussian-resampled, inverse-weighted 12.7 cal ka BP (Capriles et al., 2016). Referring strictly to the record method (GRIWM) (Bradshaw et al., 2012; Saltre et al., 2015, from the northern Altiplano, the earliest evidence of humans comes Fig. 3AeB). This approach is used here to estimate the true time of from the site of Cuncaicha in Peru and is dated around 12.1 cal kyr local extinction of horses and the true time of human arrival into BP (Rademaker et al., 2014). The colonization of a harsh and the different sections of the Altiplano (north and south). The extreme environment such as the Andean Altiplano is challenging, advantage of using GRIWM is that it deals with the non-random but nevertheless, the radiocarbon record documenting the pres- fossilization process by progressively up-weighting the gap sizes ence of humans over 3000 masl after 12.7 cal kyr BP is abundant the closer they are to the time of disappearance (appearance in the and suggests that human presence in the area was not uncommon case of humans) from the fossil record. It also takes into account the from the late Pleistocene onward. See Supplemental Table 2 for a uncertainties associated with the radiocarbon dates, providing a list of radiocarbon dates from archaeological contexts used to build 95% confidence band around the estimated time of extinction. All Fig. 3B. the analyses were run in the R programing environment (R Core Team) using the R code published in Saltre et al. (2015). 6. Discussion Taking into account the CIs (shaded boxes Fig. 3AeB) there is ~3 kyr of overlap between human presence in the northern Altiplano 6.1. Comparison with other radiocarbon dated megafauna and the horses from Casa del Diablo and ~2.5 kyr between horses (Fig. 3AeC, Supp. Table 1) and evidence of humans in the southern Altiplano. If we compare only the estimated timing of human arrival and demise of horses While no other taxon-specific radiocarbon dates on extinct (dashed line in grey boxes in Fig. 3AeB), the time overlap is reduced megafauna have been reported for the Altiplano (as defined in to 0.9 kyr between humans in the northern Altiplano and horses section 5), work at a few sites located in the surroundings of the and 1.1 kyr between horses and humans present in the southern plateau (>2500 masl) has yielded chronological information that part of the Altiplano. can be compared to the new information presented here. Evidence of stratigraphic association between extinct mega- The closest of these sites to Casa del Diablo is Pikimachay cave fauna and early hunter gatherers in the Altiplano is scarce. At the (Fig. 1). This site was studied in the late 1960s by MacNeish and site of Tuina 5 in the southern Altiplano of Chile, an innominate collaborators (MacNeish et al., 1970) who were able to retrieve fragment identified as Equidae indet. occurs in stratigraphic asso- several radiocarbon dates on bones generally assigned to mega- ciation with evidencen of early Holocene humans (Núnez~ et al., theriidae. Given that these dates were obtained during the early 2002). Capriles et al. (2016) also mentions the possible presence stages of development of radiocarbon dating techniques, using of extinct megafauna in late Pleistocene deposits at Cueva Bautista; standard methods that are obsolete at present, they should be this suggestion has not been confirmed, however. Moreover, there considered with caution. Moreover all the dates from Pikimachay are no signs of human manipulation of the megafaunal remains at but one include a large error terms, reflecting the inaccuracy of this this site. At the Cuncaicha site, there is evidence of the use by particular record (MacNeish et al., 1970). When we compare this humans of extant megafauna such as vicuna~ (Vicugna vicugna), record to our new set of dates (Fig. 3), despite the uncertainty guanaco (Lama guanicoe), and taruka (Hippocamelus antisensis), but around the dates, it is possible to note that the temporal range of no remains of extinct megafauna have been reported for the site the record from Pikimachay overlaps the one from Casa del Diablo, (Rademaker et al., 2014). differing only by being older at both ends, with a youngest date of It is important to recall that, while archaeological evidence ~17 cal kyr BP (calibrated median probability) and an oldest date of showing the late Pleistocene presence of humans in the Altiplano is ~24 cal kyr BP (calibrated median probability). abundant (Fig. 3B, Supp Table 2), it is of general agreenment among More recent work in two other caves from northern Peru, Cueva authors that this archaeological data shows ephemeral and Rosello and Trigo Jirka (Fig. 1), resulted in important findings of late exploratory occupations rather than more stable settlements Pleistocene faunas (Shockey et al., 2009). From these, taxon-specific (Capriles et al., 2016; Haas et al., 2017). Moreover, the onset of radiocarbon dates have been reported for Hippidion, Vicugna, and permanent occupation in areas at high altitudes in the Central Diabolotherium. Compared to these dates, the record from Casa del Andes is dated between 9 and 7 cal kyr BP (Haas et al., 2017). The Diablo shows in general younger occurrences (Fig. 3). Dia- later Holocene stablishment of humans in the Altiplano, plus the bolotherium from Trigo Jirka and Hippidion from Cueva Rosselloare lack of evidence proving the use of extinct megafauna by early ~13 kyr older than dated specimens belonging to the same genera hunter gatherers in the area, makes of the role of humans a less from Casa del Diablo. Vicugna from Cueva Rossello is ~11.5 kyr older probable driver of megafaunal extinctions in the High Andes. than the radiocarbon dated finding of the same taxon in Casa del Diablo. 6.3. Fauna from Casa del Diablo cave in relation to vegetation and As shown in Fig. 3, extinct megafauna persisted in the sur- climate change (Fig. 3A, D) roundings of Casa del Diablo until 12.8 cal kyr BP, the last appear- ance datum for Hippidion cf. devillei. As described in section 5.2, multiple changes in vegetation and climate took place in the environs of Casa del Diablo cave during the 6.2. Extinct megafauna from Casa del Diablo in relation to human time when the excavated fauna was deposited. The oldest date in arrival (Fig. 3AeB) the record is the megathere dated to 23 cal kyr BP. This would have lived near Casa del Diablo cave prior to the onset of the The record of megafauna presence in Casa del Diablo shows no deglaciation (~21 cal kyr BP), at a time when temperature was ~8 C definitive evidence of temporal coexistence between extinct colder than at present and conditions were overall wet. Vegetation N.A. Villavicencio, L. Werdelin / Quaternary Science Reviews 195 (2018) 21e31 29 cover during this period is estimated to have been scattered and del Diablo cave, is the Sporormiella record representing mainly the scarce; however, judging from the evidence, it might have been presence of larger herbivores such as giant ground or gom- enough to sustain an animal of the size of a megathere. The pos- photheres rather than the presence of smaller herbivores like sibility that the occurrence of a megathere in Casa del Diablo is just Hippidion cf. devillei? More information about the extinct faunas a one-time event and does not represent a portion of the realized living in the Altiplano is needed in order to approach all of the niche of this taxon at this times cannot be ruled out. However, the above. late Pleistocene presence Megatherium (along with Catonyx) has been reported for the high altitude locality of Tres Ventanas Cave in 7. Conclusions Peru (Engel, 1970; Pujos and Salas, 2004), north of the Altiplano as defined in this work. The chronology of this site is poorly con- The record from Casa del Diablo cave is the first taxon-specific strained and only represented by a minimum 14C age of 40,000 yr set of dates documenting the presence of extinct megafauna and BP (Engel, 1970). providing some clues to its demise in the Altiplano. Even though we After the transition to puna brava vegetation, at ~21 cal kyr BP, compare and discuss this chronology in the light of environmental the only dated megafauna taxon on record is Diabolotherium nor- and anthropic changes occurring over the plateau during the denskioldi, dated around 19.6 cal kyr BP. This taxon has been Pleistocene-Holocene transition, it is important to recall that this is described as a possible rock climber, given its anatomy (Pujos et al., only a small piece of information in a debate of great magnitude 2007) and the fact that its remains have been found in Andean and complexity dealing with the LQE. caves from northern Peru to southern Chile (~45 S) (Shockey et al., It seems worth noting the presence of fauna at this high-altitude 2009; Lopez-Mendoza and Mena-Larraín, 2011). site from the LGM onwards, despite the extreme conditions expe- Once the temperature rose to only 2e3 C colder than the rienced in the Altiplano during this time period. Interesting is as present day, standard puna vegetation was established in the area well the increase in taxa diversity observed at the site once climatic and the extant Lama guanicoe and Vicugna vicugna are recorded in conditions became less cold and vegetation became more abundant Casa del Diablo cave at 15.8 and 15.1 cal kyr BP respectively. In the compared to LGM and deglacial states, suggesting the possibility of same environmental context, Catonyx sp. is dated around environmental constrains for some of the taxa on record prior to 16.4 cal kyr BP, while Hippidion cf. devillei appears in the record for that time. the first time around 14.5 cal kyr BP. This suggest that perhaps, as No dated record of megafauna species (extinct and extant) exists conditions over the northern Altiplano became warmer and more for Casa del Diablo cave after the cold reversal recorded in the vegetated, a greater variety of taxa (n ¼ 4) was able to inhabit the Sajama ice core. Along with the cold reversal, there is a drying area. tendency observed in the record from Lake Titicaca. This observa- The dates for Hippidion cf. devillei (n ¼ 6) are clustered between tion opens the possibility that this cold event and the drying con- 14.5 and 13 cal kyr BP, which means this taxon inhabited the area ditions over the plateau could have had a negative effect over the during the transition to present day vegetation (13.2 cal kyr BP). It is fauna living in the surroundings of Casa del Diablo at the time. the last megafauna taxon on record, with last occurrences during According to the Casa del Diablo record, there is no temporal the cold reversal recorded in the Sajama ice core and right before coexistence between extinct fauna and humans in the Altiplano, conditions became drier in this part of the Altiplano as inferred although the calculation of CIs using GRIWM opens the possibility from the record from Lake Titicaca (Fig. 3). for some temporal interaction. No temporal coexistence is in A record of note is the presence of the small extant rodent agreement with the lack of archaeological evidence of interaction Punomys lemminus at 20.3 cal kyr BP, when the vegetation in the between extinct megafauna and humans. area was probably puna brava and conditions over the altiplano The observations presented here for the Altiplano are in general were rather cold (Thompson et al., 1998, Fig. 3F). While the present opposed of what is reported for other areas in South America. In day ecology and habitat of P. lemminus is not well known, it has southern for example, practically all of the extinct been only reported for localities over 4300 masl, at the lower megafauna seems to disappear from the area during a warming modern limit of puna brava (Patterson and Zeballos, 2008), which period after the cold reversal experienced there (Metcalf et al., suggests a conserved habitat preference of this taxon since the late 2016; Villavicencio et al., 2016). While this warming could have Pleistocene and a range contraction following postglacial warming. represented more favorable conditions for megafauna and humans in the area, it appears to have led to major changes in vegetation 6.4. More evidence of megafaunal demise in the Altiplano (Fig. 3G) which were adverse for most of the megafauna taxa present in the area (Villavicencio et al., 2016). In the Pampas region, in the At the northern limit of the Altiplano, 340 km from Casa del coast of the continent, most of the last appearances of extinct Diablo cave, a Sporormiella record from Lake Pachuca (Figs. 1 and megafauna are also dated after the cooling reported 3G) shows the history of the demise of megaherbivores in this for the area, during a period of warmer conditions (Prado et al., part of the plateau. A two-step drop in the concentration of the 2015). On the other hand, there is an overlap between the re- coprofagous fungus Sporormiella is interpreted as the functional cords of the first humans arriving to southern Patagonia and the decline and extinction of megaherbivores in the area. Compared to extinct megafauna living there (Metcalf et al., 2016; Villavicencio our new record, the of megafauna is recorded et al., 2016). This overlap is of ~2000 years for the particular case at Lake Pacucha 3e4 kyr before the latest record of Hippidion cf. of (Mylodontidae) (Villavicencio et al., 2016). Temporal devillei at Casa del Diablo cave, proving the persistence of extinct overlap between extinct megafauna and evidence of humans is also faunas much later than the time of functional extinction estimated reported for the Pampas (Prado et al., 2015). Both for southern from the lake record. Patagonia and for the Pampas, it has been proposed the possibility This comparison leaves several open questions: were the horses of synergistic effects of anthropogenic factors and changes in the living in the surroundings of Casa del Diablo cave the last remnants environment in driving megafaunal extinctions (Metcalf et al., of a long-declining functional group (megaherbivores)? Was the 2016; Villavicencio et al., 2016; Prado et al., 2015). According to area around Casa del Diablo cave a refuge for the last representa- the information presented in this work, this would not be the case tives of extinct megafauna, once they had disappeared further for the Altiplano but more data is needed in order to reach robust north? Given that only horses appear late in the record from Casa conclusions. 30 N.A. Villavicencio, L. Werdelin / Quaternary Science Reviews 195 (2018) 21e31

Acknowledgements Haas, R., Stefanescu, I.C., Garcia-Putnam, A., Aldenderfer, M.S., Clementz, M.T., Murphy, M.S., et al., 2017. Humans permanently occupied the Andean highlands by at least 7 ka. R. Soc. Open Sci. 4 (6), 170331. This research was funded by the NSF grant EAR 1148181, a Hansen, B.C., Wright Jr., H.E., Bradbury, J.P., 1984. Pollen studies in the Junin area, BECAS CHILE PhD fellowship, by the University of California central Peruvian Andes. Geol. Soc. Am. Bull. 95 (12), 1454e1465. Museum of (UCMP, summer research award) and by Hoffstetter, R., 1952. Les mammiferes pleistocenes de la Republique de l'Equateur. Memoirs de la Societ eG eologique de 31 (66), 391 p. the Department of Integrative Biology at UC Berkeley (summer Hoffstetter, R., 1986. High andean mammalian faunas during the Plio-Pleistocene. research grant). We want to thank to Dr. Matias Frugone for his help In: Vuilleumier, F., Monasterio, M. (Eds.), High Altitude Tropical , with the map and especially to the reviewers, Ross MacPhee and 9. Oxford University Press, Oxford, , pp. 219e245. Hogg, A.G., Hua, Q., Blackwell, P.G., Niu, M., Buck, C.E., Guilderson, T.P., et al., 2013. François Pujos, for their valuable comments that made of this SHCal13 Southern Hemisphere calibration, 0e50,000 years cal BP. Radiocarbon manuscript a better contribution. Lastly, we would like to thank T. 55 (4), 1889e1903. Stafford, E.L. Lindsey, T. Guilderson and A. Hedgpeth for their help Hostetler, S.W., Mix, A.C., 1999. Reassessment of ice-age cooling of the tropical ocean and atmosphere. Nature 399 (6737), 673. and guide in the radiocarbon dating analyses of our samples. Iriarte, A., 2008. Mamiferos de Chile. Ediciones, Barcelona, . Kessler, M., 1995. The Polylepis (Rosaceae) in Bolivia. Candollea 50 (1), Appendix A. Supplementary data 131e171. Koch, P.L., Barnosky, A.D., 2006. Late Quaternary extinctions: state of the debate. Annu. Rev. Ecol. Evol. Syst. 37. Supplementary data related to this article can be found at Labarca, R.L., 2015. La meso y megafauna terrestre extinta del Pleistoceno de Chile. https://doi.org/10.1016/j.quascirev.2018.07.013. Publicacion Ocas. Mueso Nac. Hist. Nat. Chile 63, 401e465. Latorre, C., Betancourt, J.L., Arroyo, M.T., 2006. Late Quaternary vegetation and climate history of a perennial river canyon in the Río Salado basin (22 S) of References Northern Chile. Quat. Res. 65 (3), 450e466. Lopez-Mendoza, P., Mena-Larraín, F., 2011. Extinct dermal bones and ~ Aceituno, P., 1993. Aspectos generales del clima en el altiplano sudamericano. In: II their role in the taphonomic research of caves: the case of Bano Nuevo-1 Simposio internacional de estudios altiplanicos. Arica, Chile. (Andean Central Patagonia, Chile). Rev. Mex. Ciencias Geol. 28 (3). Alberdi, M.T., Prado, J.L., Salas, R., 2004. The Pleistocene Gomphotheriidae (Probo- MacNeish, R.S., Berger, R., Protscha, R., 1970. Megafauna and man from Ayacucho, scidea) from Peru. Neues Jahrbuch Geol. Palaontol. Abhand. 231 (3), 423e452. highland Peru. Science 168 (3934), 975e977. Allmendinger, R.W., Jordan, T.E., Kay, S.M., Isacks, B.L., 1997. The evolution of the Marshall, L.G., Sempere, T., 1991. The to Pleistocene vertebrates of Bolivia altiplano-puna plateau of the central Andes. Annu. Rev. Earth Planet Sci. 25 (1), and their stratigraphic context: a review. Rev. Tec. Yacimientos Pet. Fisc. Boliv. 139e174. 12 (3e4), 631e652. Baied, C.A., Wheeler, J.C., 1993. Evolution of high Andean puna ecosystems: envi- Marshall, L.G., Berta, A., Hoffstter, R., Pascual, R., Bombin, M., Mones, A., 1984. ronment, climate, and culture change over the last 12,000 years in the Central Mammals and stratigraphy: geochronology of the continental mammal-bearing Andes. Mt. Res. Dev. 145e156. Quaternary of South America. Paleovertebrata, Memoire Extraordinaire 1e76, Baker, P.A., Seltzer, G.O., Fritz, S.C., Dunbar, R.B., Grove, M.J., Tapia, P.M., et al., 2001. 1984. The history of South American tropical precipitation for the past 25,000 years. Martin, P.S., 1990. 40,000 years of extinctions on the “planet of doom”. Palaeogeogr. Science 291 (5504), 640e643. Palaeoclimatol. Palaeoecol. 82 (1e2), 187e201. Barnosky, A.D., Lindsey, E.L., 2010. Timing of Quaternary megafaunal extinction in Martin, P.S., Klein, R.G. (Eds.), 1984. Quaternary Extinctions: a Prehistoric Revolu- South America in relation to human arrival and climate change. Quat. Int. 217 tion. Press. (1e2), 10e29. Martin, P.S., Steadman, D.W., 1999. Prehistoric extinctions on islands and conti- Barnosky, A.D., Koch, P.L., Feranec, R.S., Wing, S.L., Shabel, A.B., 2004. Assessing the nents. In: Extinctions in Near Time. Springer, Boston, MA, pp. 17e55. causes of Late Pleistocene extinctions on the continents. Science 306 (5693), Martin, P.S., Wright, H.E., 1967. Pleistocene Extinctions: the Search for a Cause. Yale 70e75. University Press, New Haven, p. 453. Barnosky, A.D., Lindsey, E.L., Villavicencio, N.A., Bostelmann, E., Hadly, E.A., Mercer, J.H., 1984. Late Cainozoic glacial variations in South America south of the Wanket, J., Marshall, C.R., 2016. Variable impact of late-Quaternary megafaunal . In: Vogel, J.C. (Ed.), Late Cainozoic Palaeoclimates of the Southern extinction in causing ecological state shifts in North and South America. Proc. Hemisphere. Balkema, Rotterdam, pp. 45e58. Natl. Acad. Sci. Unit. States Am. 113 (4), 856e861. Metcalf, J.L., Turney, C., Barnett, R., Martin, F., Bray, S.C., Vilstrup, J.T., et al., 2016. Bradshaw, C.J.A., Cooper, A., Turney, C.S.M., Brook, B.W., 2012. Robust estimates of Synergistic roles of climate warming and human occupation in Patagonian extinction time in the geological record. Quat. Sci. Rev. 33, 14e19. megafaunal extinctions during the Last Deglaciation. Sci. Adv. 2 (6) e1501682. Braun, G., 1997. The use of digital methods in assessing patterns in an Andean Nordenskiold,€ E., 1908. Ein neuer fundort für Saugetierfossilien€ in Peru. Ark. Zool. 4 environment: the Polylepis example. Mt. Res. Dev. 253e262. (11), 13e21. ~ Brook, B.W., Barnosky, A.D., 2012. Quaternary extinctions and their link to climate Núnez, L., Grosjean, M., Cartajena, I., 2002. Human occupations and climate change change. In: Saving a Million Species. Island Press/Center for Resource Eco- in the Puna de Atacama, Chile. Science 298 (5594), 821e824. nomics, pp. 179e198. Paduano, G.M., Bush, M.B., Baker, P.A., Fritz, S.C., Seltzer, G.O., 2003. A vegetation Capriles, J.M., Albarracin-Jordan, J., Lombardo, U., Osorio, D., Maley, B., and fire history of Lake Titicaca since the last glacial maximum. Palaeogeogr. Goldstein, S.T., et al., 2016. High-altitude adaptation and late Pleistocene Palaeoclimatol. Palaeoecol. 194 (1e3), 259e279. foraging in the Bolivian Andes. J. Archaeol. Sci.: Report 6, 463e474. Patterson, B., Zeballos, H., 2008. Punomys lemminus. In: The IUCN Red List of Cione, A.L., Tonni, E.P., Soibelzon, L.H., 2003. The broken zig-zag: late Cenozoic large Threatened Species, 2008. mammal and tortoise extintion in South America. Rev. Mus. Argent. Ciencias Placzek, C.J., Quade, J., Patchett, P.J., 2013. A 130 ka reconstruction of rainfall on the Nat. 5. Bolivian Altiplano. Earth Planet Sci. Lett. 363, 97e108. Eaton, G.F., 1914. Vertebrate fossils from Ayusbamba, Peru. Am. J. Sci. 218, 141e154. Prado, J.L., Martinez-Maza, C., Alberdi, M.T., 2015. Megafauna extinction in South Eisenberg, J.F., Redford, K.H., 1989. Mammals of the Neotropics, Volume 2: the America: a new chronology for the Argentine Pampas. Palaeogeogr. Palae- Southern Cone: Chile, Argentina, , , vol. 2. University of Chi- oclimatol. Palaeoecol. 425, 41e49. cago Press. Prevosti, F.J., Forasiepi, A.M., 2018. Evolution of South American Mammalian Engel, F., 1970. La grotte du Megatherium et les ecologies du Haut- Peruvien. Predators during the Cenozoic: Paleobiogeographic and Paleoenvironmental Semiotica 1, 413e436. Contingencies. Springer Geology, Cham, Switzerland, 196 pp. Falvey, M., Garreaud, R.D., 2005. Moisture variability over the South American Al- Pujos, F., Salas, R., 2004. A systematic reassessment and paleogeographic review of tiplano during the South American low level jet experiment (SALLJEX) fossil Xenarthra from Peru. Bulletin de l'Institut français d'etudes andines 33 observing season. J. Geophys. Res.: Atmosphere 110 (D22). (2), 331e377. Farina,~ R.A., Vizcaíno, S.F., De Iuliis, G., 2013. Megafauna: Giant Beasts of Pleistocene Pujos, F., De Iuliis, G., Argot, C., Werdelin, L., 2007. A peculiar climbing Mega- South America. Indiana University Press. lonychidae from the Pleistocene of Peru and its implication for sloth history. Fornari, M., Risacher, F., Feraud, G., 2001. Dating of paleolakes in the central Alti- Zool. J. Linn. Soc. 149 (2), 179e235. plano of Bolivia. Palaeogeogr. Palaeoclimatol. Palaeoecol. 172 (3e4), 269e282. Rademaker, K., Hodgins, G., Moore, K., Zarrillo, S., Miller, C., Bromley, G.R., et al., Frailey, D., Campbell, K.E., 1980. A Pleistocene record of Bufo spinulosus from Peru. 2014. Paleoindian settlement of the high-altitude Peruvian Andes. Science 346 Copeia 1980 (4), 935e936. (6208), 466e469. Garreaud, R.D., 2009. The Andes climate and weather. Adv. Geosci. 22, 3. Ramirez Pareja, J.A., 1958. Mamíferos fosiles del Dpto. Del Cuzco. Doctoral disser- Garreaud, R., Aceituno, P., 2001. Interannual rainfall variability over the south tation. Universidad de Cuzco, Cuzco, Peru. American altiplano. J. Clim. 14 (12), 2779e2789. Reese, C.A., Liu, K.B., 2005. A modern pollen rain study from the central Andes Garreaud, R., Vuille, M., Clement, A.C., 2003. The climate of the Altiplano: observed region of South America. J. Biogeogr. 32 (4), 709e718. current conditions and mechanisms of past changes. Palaeogeogr. Palae- Rozas-Davila, A., Valencia, B.G., Bush, M.B., 2016. The functional extinction of An- oclimatol. Palaeoecol. 194 (1e3), 5e22. dean megafauna. Ecology 97 (10), 2533e2539. Graf, K., 1981. Palynological investigations of two post-glacial peat near the Saltre, F., Brook, B.W., Rodríguez-Rey, M., Cooper, A., Johnson, C.N., Turney, C.S., boundary of Bolivia and Peru. J. Biogeogr. 353e368. Bradshaw, C.J., 2015. Uncertainties in dating constrain model choice for N.A. Villavicencio, L. Werdelin / Quaternary Science Reviews 195 (2018) 21e31 31

inferring extinction time from fossil records. Quat. Sci. Rev. 112, 128e137. ecologico del Perú, vol. 5. IICA Biblioteca . Sefve, I., 1910. Hyperhippidium, eine neue Süd-americanische pferdegattung. K. - Troll, C., 1968. The cordilleras of the tropical , aspects of climatic, phyto- Sven. Vetenskapsakademiens Handl. 46 (2), 1e43 (Stockholm). geographical and agrarian ecology. In: Colloquium Geographicum (Univ. Bonn), Seltzer, G.O., 1990. Recent glacial history and paleoclimate of the Peruvian-Bolivian vol. 9, pp. 15e56. Andes. Quat. Sci. Rev. 9 (2e3), 137e152. Villavicencio, N.A., Lindsey, E.L., Martin, F.M., Borrero, L.A., Moreno, P.I., Seltzer, G.O., 1992. Late quaternary glaciation of the cordillera real, Bolivia. J. Quat. Marshall, C.R., Barnosky, A.D., 2016. Combination of humans, climate, and Sci. 7 (2), 87e98. vegetation change triggered Late Quaternary megafauna extinction in the Úl- Seltzer, G.O., Rodbell, D.T., Baker, P.A., Fritz, S.C., Tapia, P.M., Rowe, H.D., Dunbar, R.B., tima Esperanza region, southern Patagonia, Chile. Ecography 39 (2), 125e140. 2002. Early warming of tropical South America at the last glacial-interglacial Vuille, M., Bradley, R.S., Keimig, F., 2000. Interannual climate variability in the transition. Science 296 (5573), 1685e1686. Central Andes and its relation to tropical Pacific and Atlantic forcing. J. Geophys. Shockey, B.J., Salas-Gismondi, R., Baby, P., Guyot, J.L., Baltazar, M.C., Huaman, L., Res.: Atmosphere 105 (D10), 12447e12460. Clack, A., Stucchi, M., Pujos, F., Emerson, J.M., Flynn, J.J., 2009. New Pleistocene Vuille, M., Bradley, R.S., Werner, M., Healy, R., Keimig, F., 2003. Modeling d18O in cave faunas of the Andes of central Perú: radiocarbon ages and the survival of precipitation over the tropical Americas: 1. Interannual variability and climatic low latitude Pleistocene DNA. Palaeontol. Electron. 12 (3), 1e15. controls. J. Geophys. Res.: Atmosphere 108 (D6). Stafford Jr., T.W., Brendel, K., Duhamel, R.C., 1988. Radiocarbon, 13C and 15N analysis Weinstock, J., Shapiro, B., Prieto, A., Marín, J.C., Gonzalez, B.A., Gilbert, M.T.P., of fossil bone: removal of humates with XAD-2 resin. Geochem. Cosmochim. Willerslev, E., 2009. The Late Pleistocene distribution of vicunas~ (Vicugna Acta 52 (9), 2257e2267. vicugna) and the “extinction” of the gracile (“Lama gracilis”): new mo- Stuiver, M., Reimer, P.J., 1986e2014. Calib 7.01. Calibration Radiocarbon Program. lecular data. Quat. Sci. Rev. 28 (15e16), 1369e1373. Sunquist, M., Sunquist, F., 2002. Wild Cats of the World. University of Press. Ybert, J.P., 1988. Apports de la palynologie a la connaissance de l’histoire du lac Thompson, L.G., Mosley-Thompson, E., Davis, M.E., Lin, P.N., Henderson, K.A., Cole- Titicaca (Bolivie-Perou) au cours du Quaternaire recent. Travaux de la Section Dai, J., et al., 1995. Late glacial stage and Holocene tropical ice core records from Scientifique et Technique, Institute Français de Pondichery 25, 139e150. Huascaran, Peru. Science 269 (5220), 46e50. Ybert, J.P., 1992. Ancient lake environments as deduced from pollen analysis. In: Thompson, L.G., Davis, M.E., Mosley-Thompson, E., Sowers, T.A., Henderson, K.A., DeJoux, C., Iltis, A. (Eds.), Lake Titicaca. A Synthesis of Limnological Knowledge. Zagorodnov, V.S., et al., 1998. A 25,000-year tropical climate history from Kluwer Academic, Boston, MA, pp. 49e62. Bolivian ice cores. Science 282 (5395), 1858e1864. Zhou, J., Lau, K.M., 1998. Does a monsoon climate exist over South America? J. Clim. Tosi, J.A., 1960. Zonas de vida natural en el Perú: memoria explicativa sobre el mapa 11 (5), 1020e1040.