Diet breadth and biodiversity in the pre-hispanic South-Central Andes (Western South America) during the Holocene: An exploratory analysis and review

Abstract

Spanish

This paper presents an exploratory study on the taxonomic diversity of pre-Hispanic archaeofaunas in the South-Central Andes (SCA; western South America) from the Pleistocene-Holocene boundary to the Late-Holocene. The SCA is a complex of diverse environments and has undergone distinct climate events for the last 13,000 years, such as the occurrence of warmer and drier conditions in the Middle-Holocene. The South-Central Andean area was part of the larger Andes interaction area, which was a primary center for animal and plant domestication and the emergence of agro-pastoralist economies. Since subsistence was key to these processes, the SCA provides a relevant case study on the interactions among environment, foodways and sociocultural evolution. Taxonomic diversity was used here as a proxy for diet breadth. A total of 268 archaeofaunal assemblages were sampled from the zooarchaeological literature. Reviewed variables included the cultural chronology and spatial coordinates of the assemblages, as well as the presence and abundance of taxa at the family rank. Taxonomic diversity covered two dimensions: composition (families present in each assemblage) and structure (quantitative relationships among taxa), which was measured through richness (NTAXA), ubiquity and relative abundance (NISP based rank-order). Despite the uneven distribution of samples, the analyses revealed the following trends: (1) a moderate relationship between NTAXA and distance from coastline for most of the Holocene; (2) a potential decrease in assemblage richness for coastal ecoregions during the Late-Holocene; and (3) a generalized increase in the relative abundance of Camelidae.

Keywords

biodiversity diet breadth Holocene climate South-Central Andes zooarchaeology

Introduction

Feeding is a biological and socio-cultural function that can be investigated using different research strategies (Harris, 1987; Hintze et al., 1997; Messer, 1984; Mintz and Du Bois, 2002). Materialist and evolutionary anthropologies, in particular, have emphasized the correlation between feeding and environment, technique, demography, and social organization (Harris, 1987; Kelly, 1995; Richerson et al., 2009). This paper presents a preliminary quantitative synthesis on the animal component of pre-Hispanic diets across the South-Central Andes (SCA) for approximately the last 13,000 years. The SCA are located in a mostly arid region of western South America with pronounced abiotic and biotic heterogeneity, offering multiple potential niches for cultural adaptation.

This exploratory investigation primarily aimed to identify patterns of taxonomic diversity by examining a substantial collection of archaeofaunal assemblages. Results were subsequently analyzed within the context of current hypotheses concerning the evolution of pre-Hispanic subsistence under post-Pleistocene climate conditions. The study was guided by two research questions: (1) what animal taxa did pre-Hispanic populations use across various ecoregions and during significant climate events?; and (2) can modifications in average prey selection be traced to transformations in the environment, technology or demographics? These questions were approached within the framework of the Diet-Breadth Model of the Optimal Foraging Theory (Begon et al., 2006; Kelly, 1995; MacArthur and Pianka, 1966; Pyke, 1984; Smith et al., 1983). The model proposes that the acquisition of resources (i.e. prey) is a function of their net nutritional return and rates of encounter, which are influenced by environmental and socio-cultural factors.

First, I grouped together faunal assemblages into broad spatial and temporal units and calculated their descriptive statistics and then I used different non-parametric tests to test for significant cross-sectional and longitudinal differences. I obtained the datasets from the Zooarchaeological Database of the South-Central Andes (ZDBSCA) (Belotti López de Medina, 2019), an ongoing project that aims to store data from reports published between 1967 and 2020. The literature survey includes data until 2009 and therefore the findings reported here are provisional. A secondary objective of this paper was to identify flaws in the datasets to optimize the methodology applied in the ZDBSCA project.

Background

The South-Central Andes

The Andes represent an archeological unit located in western South America (González and Pérez Gollán, 1966; Isbell and Silverman, 2006; Lumbreras, 2008; Lumbreras, 1969). The region shows a high degree of environmental heterogeneity. First, topography and atmospheric circulation give rise to a series of distinct longitudinal zones that extend from cloud forests in the east to the highlands and the Pacific coast in the west (Aldenderfer, 1989; Brush, 1982; Grosjean et al., 2007; Olson et al., 2001). Second, these larger units exhibit also internal latitudinal, altitudinal and hydrological-ecological variations. Such heterogeneity in conjunction with Late Pleistocene and post-Pleistocene climate phases favored diverse lifeways and ecological complementarity (Aldenderfer, 1989; Browman, 1980; Concha Contreras, 1975; Dillehay and Núñez, 1988; Mayer, 2004; Murra, 1975, 1978). Indeed, the Andes are considered a major center for animal and plant domestication, the emergence of agropastoral economies and primary state formation (Fiedel, 1996; Gayo et al., 2015; Goldberg et al., 2016; Gómez-Carballa et al., 2018; Larson and Fuller, 2014; Lumbreras, 2008; Nakatsuka et al., 2020; Price and Bar-Yosef, 2011; Rothhammer et al., 2017; Spencer, 2010; Urban and Barbieri, 2020).

The study area of the ZDBSCA project and this paper is the South-Central Andes sub-region (SCA), stretching from southern Perú (Camaná-Majes basin) and Bolivia (Titicaca basin) to northwestern Argentina and northern Chile (15-30° S, Figure 1) (Aldenderfer, 1989; González and Pérez Gollán, 1966; Lumbreras, 1969). The SCA were originally differentiated from the Peruvian Central Andes by the distribution of certain Late-Holocene traits of material culture (González and Pérez Gollán, 1966; Lumbreras, 1969). The selection of SCA by the ZDBSCA project was initially motivated by pragmatic considerations because it exhibits many features typical of the entire Andes, has been the focus of a considerable amount of paleoenvironmental, archeological and zooarchaeological investigation, and represents the spatial unit of numerous research projects and synthetic studies.

Figure 1.

Study area with the distribution of reviewed faunal assemblages by terrestrial ecoregion, elevation and minimum geodesic distance from coastline. Map, plots and tallies were created with R and based on spatial data from the ZDBSCA and the following sources: Natural Earth 10 m coastline (https://www.naturalearthdata.com/downloads/10m-physical-vectors/10m-coastline/), SRTM Digital Elevation Data v4 (Jarvis et al., 2008, https://earthengine.google.com/) and Terrestrial Ecoregions of the World (Olson et al., 2001, https://www.worldwildlife.org/publications/terrestrial-ecoregions-of-the-world).

The timing and dynamics of the colonization of diverse habitats in the SCA is a subject of ongoing research (Aldenderfer, 1998; Borrero and Santoro, 2022; Capriles et al., 2016; Gómez-Carballa et al., 2018; Rothhammer et al., 2017; Santoro et al., 2017; Standen et al., 2004; Yacobaccio, 2017). The earliest radiocarbon dating suggests that colonization took place at the end of the Pleistocene (approximately 14,000–11,000 cal BP) (Gayo et al., 2015 supplementary data). There is very little evidence of interaction with extinct megafauna (Borrero and Santoro, 2022; Núñez and Grosjean, 2003). Consensus in the existing literature supports an early divergence between coastal and highland lifeways dating back to the beginning of the Archaic, or at least a persistent maritime specialization on the coast (Aldenderfer, 1989; DeFrance et al., 2009; Grosjean et al., 2007; Núñez and Grosjean, 2003). Borrero and Santoro (2022) propose a more complex scenario that includes metapopulations distributed in different habitats and sharing a pool of techniques suitable for different environments; a drastic reduction in water availability and the deterioration of ecological corridors at the beginning of the Early Holocene led to adaptive divergence across coastal, highland and lowland populations.

Early sites located on the present-day coastline were characterized by highly diverse faunal assemblages, with fish, shellfish, marine mammals and shore birds, as well as wild camelids and terrestrial birds at southern sites (Sterile Coast, 22°–27° S) (Aldenderfer, 1989; Grosjean et al., 2007). The inhabitants of the Fertile Coast (15°–22° S) would have exploited inland upstream oases on a seasonal basis to complement their diets (Aldenderfer, 1989; Grosjean et al., 2007). The SCA highlands -or puna- feature grasslands and wetlands that are suitable for large artiodactyls and smaller vertebrates, such as rodents and birds. The first faunal assemblages were predominantly composed of the families Camelidae (guanaco, Lama guanicoe and vicuña, Vicugna vicugna), Cervidae (taruca, Hippocamelus antisensis) and Chinchillidae (with large rodents such as the genera Lagidium and Chinchilla). Human settlements were drawn to bofedales in the northern dry puna, while early southward patterns of movement would have taken place from the Coastal Range to the high valleys beside the salt puna (Aldenderfer, 1989, 1998).

During the Holocene, Andean populations intensified subsistence activities in several ways. On the coast, this was evidenced by an increase in the diversity of marine genera and habitats exploited and by a relative specialization on some mollusk genera, as for example Mesodesma at the Ring site (Aldenderfer, 1989; Castro et al., 2016; Standen et al., 2004). Highland foragers became increasingly specialized in hunting camelids (Yacobaccio, 2006), which eventually resulted in the protection of wild herds and, by the Late-Holocene, in the domestication of llamas from guanaco populations (and alpacas from vicuñas in the wet Puna and northward areas) (Mengoni Goñalons and Yacobaccio, 2006; Yacobaccio and Vilá, 2013). Such specialization was likely to coincide with plant domestication and the development of extensive exchange networks (Yacobaccio, 2006).

Diet breadth

Homo sapiens is one of the large omnivores (Root-Bernstein and Ladle, 2019) and has been described as a generalist-specialist by its ability to culturally adapt to a wide range of environments and to create specialized niches (Roberts and Stewart, 2018). It is worthy to mention that subsistence plasticity may be affected by different factors such as design constraints, metabolic costs associated with development and reproduction, inherited physiological and cognitive functions that guide and reinforce dietary choices. I analyzed the degree of generalization/specialization in faunal exploitation by ecoregion and time period under the framework of the Optimal Foraging Theory (OFT), a set of formal models used by ecologists to predict foraging behavior under different conditions. These models have been inspired by microeconomics and were originally applied to non-human predators under the assumption that natural selection favors fitness-enhancing behavior. Notably, they are neutral to the relative balance between inherited and learned behaviors because both can be optimized (Pyke, 1984). The use of the OFT in anthropology is justified on alternative theoretical rationales and, for example, it may be applied to account for foraging choices under current conditions and adaptive cultural patterns at different time scales (Harris, 2010; Kelly, 1995; Lupo, 2007; Smith et al., 1983).

From an anthropological perspective, adaptation designates two concepts (Morán, 2007): (1) a cultural adjustment to the environment that enhances biological survival, individual welfare and/or socio-cultural homeostasis (cultural ecology, cultural materialism); and (2) environmental selection of heritable traits that are transmitted either biologically or culturally (Darwinian evolution). At ecological time scales, the application of OFT models tends toward the first concept under the phenotypic gambit assumption, but behavior would remain constrained by inherited cultural adaptations lagging behind environmental changes. Historical and archeological time scales often deal with innovation and long-term environmental dynamics, and unequal success and imitation of alternative behavioral traits would fall within the scope of the Darwinian concept. In the latter case, optimization could be understood as a representation of selective pressures on hereditary constellations of cultural traits. Lastly, short-term optimization does not necessarily lead to invention, innovation or adaptation at larger temporal scales; innovation is a potential unintended and unforeseen consequence of invention (Blute, 2010; Mahner and Bunge, 2001).

The model used here was the Diet Breadth or Prey Choice model (Begon et al., 2006; MacArthur and Pianka, 1966). It predicts that a predator will consume the i-th encountered resource if and only if the post-encounter return is equal to or greater than its usual diet plus the additional time required to find its normal prey ( $E_{i} / h_{i} > = \bar{E} / (\bar{s} + \bar{h})$ ). Archeological applications proceed by building a ranking of animal taxa based on the average body weight of living adult prey. Taxonomic richness is used as a measure of the maximum diet breadth. An increase in the relative abundance of low-ranked taxa is often interpreted as a response to resource depression (Ugan, 2005).

Archeologists tend to focus on maximum diets (Grayson and Delpech, 1998), which are indicated by assemblages, and average diets (Lyman, 2003) derived from the aggregation of archaeofaunas over different spatio-temporal units. Both of these could be attributed to an accumulation of optimizing choices across different techno-environmental and demographic contexts. Archeological applications therefore deviate from the original goal of modeling behavioral plasticity on an ecological scale, requiring some degree of adjustment and reinterpretation¹ of models. Moreover, the application of the model is complicated by the fact that a faunal assemblage is a localized phenomenon, whereas humans may consume prey and discard the bones anywhere.

It should be emphasized that the Diet Breadth Model and its archeological applications have several limitations. First, it assumes that prey are substitutable (Begon et al., 2006), meaning that they are compositionally similar (same ratios of protein, fat, vitamins, and other nutrients) and that they differ only in their net energetic return. However, animals are just a subset of the diet for most human populations, especially for agriculturalists and low-latitude foragers, who must balance complementary resources (Kelly, 1995). In this regard, agricultural carbohydrates may even replace animals in terms of metabolic energy. In addition, this model does not consider animal husbandry having no energy cost of prey search, but with increased handling time related to herd care. Animal or meat trades pose a similar problem, with the main constraint being property of money or other tradable goods and services. Food acquisition choices may also be driven by optimization goals such as preventing poisoning and reducing subsistence risk due to random factors like climate or herd diseases. Additional assumptions of this and other ecological models may simplify the phenomena under study making it unrealistic even for non-human foragers (Begon et al., 2006; Kelly, 1995). The utility of the models relies on the testability of their predictions regarding the different factors that influence behavior and rejected hypotheses could prompt us to search for alternative explanations.

Materials and methods

Primary data

This study is based on the zooarchaeological literature reviewed to date for the South-Central Andes Zooarchaeological Database Project (ZDBSCA) (Belotti Lopez de Medina, 2019). The goal of this project is to create an integrated reference source and to provide a tool for synthetic analyses of faunal data. It was developed on the basis of two main activities. The first one is a systematic literature survey of zooarchaeological reports (including gray literature) written since 1967, containing at least a list of identified fauna. Another inclusion criterium is that faunal remains were analyzed by zooarchaeologists or biologists. Searching for reports was based on various strategies, ranging from online sources to citation tracking and consultation with experts. In addition, literature with relevant supplementary information on the assemblages, such as radiocarbon dating or geographic coordinates was also considered.

Information extracted from the reports included, whenever available, archeological (e.g. site, strata, site and context type, absolute and relative chronologies), zooarchaeological (e.g. taxa, anatomical elements, age groups), and bibliographical (e.g. title, year, authors, journal, book, or conference name, if applicable) data.

Then, data were manually loaded into a relational spatial database implemented in PostgreSQL using the Postgis extension (Juba et al., 2015; Marquez, 2015), which allows the storage and querying of spatial data, such as vector and raster layers. The server-client architecture permits access from different environments, such as QGIS and/or the Rstudio environment for the R programing language, both of which were used here. The database is currently hosted on an instance of Amazon Web Services. It should be highlighted that the database schema is normalized, implying that data are divided into several tables (called relations), which represent entities, such as sites or anatomical elements and variables (attributes). Relations are linked by foreign keys, thereby allowing for complex descriptions of data while retaining consistency. A former version of the database schema has been included in a previous paper (Belotti López de Medina, 2019), while the current version includes more relations (e.g. tables of radiocarbon dates). The supplemental data files provided here represent the results of SQL queries that join and select data from multiple tables.

To date, only 132 reports published between 1967 and 2009 were available for consultation among a total of 224 identified for the same period (Table 1) and 95% of these were published before 2004. Supplementary data file 1 shows the reports searched through citation backtracking as well as those actually reviewed. It is worthwhile to mention that the search and review of zooarchaeological literature for the database is still ongoing, with the publication period being extended to 2020.

Table 1.

Data entry progress.

Entity	N tuples
Sites	131
Assemblages	269
C¹⁴ Dates	419
Reports	224
Assemblage-Taxon	2166

The sampling unit was the faunal assemblage (also archaeofauna), defined for the purposes of the database as the aggregate of bone specimens described and quantified by the authors of the original report. The methods and resolution varied between reports, but approximate geographic coordinates, relative chronology, and a list of identified taxa were established for every assemblage. The focus was placed on these three items to maximize sample size, but the database covered a broader spectrum of entities and attributes. It should be pointed out that the number of assemblages varied between areas and periods due to different research traditions and that a few assemblages were dated to more than one archeological period. In addition, one site and one assemblage were excluded from present analyses because the former coordinates were not available (Pozo Cavado site, from Chile).

Units of analysis

I constructed broad spatio-temporal units to control for significant variation in the datasets. First, faunal assemblages were grouped by their location within the terrestrial ecoregions published by Olson et al. (2001) (Figure 2). These authors state that an ecoregion summarizes the general characteristics of the biome and the evolution of the modern biocenosis. Another advantage of this system is the availability of detailed vector layers covering the entire study area. Reported assemblages already loaded into the database came from seven ecoregions: Sechura Desert, Atacama Desert, Chilean Matorral, Central Andean wet Puna, Central Andean Puna, Central Andean dry Puna, and High Monte. No zooarchaeological data from the eastern Andean cloud forests (Yungas) have been found yet in the surveyed literature, but this may change as the database grows.

Figure 2.

Clustering of ecoregions based on Jaccard’s dissimilarity matrix for modern (Mollusca and Chordata) and archeological (all) occurrences of taxonomic families.

Ecoregions partially overlapped with the ecozones and other regional divisions used by archeologists (Table 2) (Aldenderfer, 1989; Grosjean et al., 2007). The main problem was that coastal ecoregions lumped the coast together with the Coastal Range and the basins of the western Andean slopes (“intermediate valleys” or “intermediate sierra” and “high valleys” or “high sierra”). Atacama included part of the “fertile coast” (15°−22° S) and part of the “sterile coast” (22°−27° S), the latter being characterized by a severe lack of fresh water, except for the Loa River. Published maps of these ecozones were schematic and not well suited for spatial analysis. To solve this problem, I included the ecozones west of the highlands for complementary analyses using SQL queries that combine the selection of several geographic statements. Finally, to avoid confusion, I must point out that the Central Andean Puna ecoregion is roughly equivalent to the dry puna used by archeologists and the Central Andean dry Puna is equivalent to the salt puna.

Table 2.

Equivalency between terrestrial ecoregions and other regional units used by archeologists.

Terrestrial ecoregions (Olson et al., 2001)	Environmental divisions used by archeologists (Aldenderfer, 1989; Grosjean et al., 2007)
Sechura Desert	Fertile coast Intermediate valleys/sierra (low transverse basins between the Coastal Range to the western Andean piedmont)
Atacama Desert	Fertile coast (north of 22º S) Sterile or desert coast (south of 22º S) Intermediate Valleys/Intermediate sierra
Chilean Matorral	Sterile or desert coast Semiarid coast Intermediate valleys / Intermediate sierra
Central Andean wet Puna	Wet puna High valleys/High sierra (high transverse basins at the western Andean ranges, beside the Puna)
Central Andean Puna	Dry puna High valleys/High sierra
Central Andean dry Puna	Dry puna Salt puna High valleys/High sierra
High Monte	Mesothermal valleys or Hills and valleys area of northwestern Argentina archeology

The Andean ecoregions covered different altitudinal zones, with median elevations ranging from 1208 to 1877 m a.s.l. for Sechura, Atacama and Matorral, 1875 m a.s.l. for High Monte, to 3953–4080 m a.s.l. for the Puna. However, all the ecoregions showed a very high difference between the maximum and minimum elevation or range of altitude (ALTr; 4674–6076 m) (Jiménez-Alfaro et al., 2016), which is a measure of abiotic heterogeneity of the habitats.

Olson et al. (2001) classified the Chilean Matorral as a mediterranean forest, the Puna and Monte ecoregions as “montane grasslands and shrublands” and the Atacama and Sechura deserts as belonging to the “desert and xeric shrublands” biome. The median and interquartile range of the Enhanced Vegetation Index (EVI; MOD13A2 VI, January 2022; retrieved from Google Earth Engine), were used as a proxy for the relative primary productivity of terrestrial ecosystems. EVI is a measure of surface photosynthesis based on satellite bands of reflected light and was calculated for each ecoregion for the Andean summer/rainy season of 2022. The lowest EVI median values (0.05–0.09) were observed in the Coastal ecoregions and the Central Andean dry Puna, while median values for highland and high monte ecoregions were 0.1 and 0.24, respectively. The interquartile range is a measure of dispersion and is also used as proxy for sub-ecoregional heterogeneity, with the lowest values (<0.1) being recorded for Sechura, Atacama, and Central Andean dry Puna. Finally, the aridity of the Puna increased southeastward (i.e. both the median and interquartile range decreased from wet to dry Puna). Previously, Grosjean et al. (2007) stated that terrestrial biomass is exceptionally low at the coast and in the intermediate valleys, and it increases in the Puna. The lowest coastal primary productivity was found in Atacama, which hosted the Sterile Coast.

Although these ecoregions were first aimed to contain distinct communities of terrestrial flora and fauna, information on freshwater and marine fauna was also included to provide an adequate baseline of potential resources. Thus, considering that coastal ecoregions supported a higher animal diversity than did puna and monte, modern taxonomic richness was calculated for the Andean terrestrial ecoregions and their adjacent continental shelf as a proxy for potential resources. Calculations were based on the occurrences of chordate and molluscan families retrieved from the GBIF (GBIF.org –20 October 2022- GBIF Occurrence Download https://doi.org/10.15468/dl.ub3dwx) and OBIS repositories (see Supplemental Material 2 file for references). NTAXA values ranged from 203 to 321 families for the coastal and from 73 to 154 families for the Puna and High Monte ecoregions, respectively. The coastal richness was mainly explained by fishes, mollusks and marine mammals. The taxonomic composition (represented by the relative occurrence of each family) was compared among ecoregions with a Jaccard’s dissimilarity matrix (Borcard et al., 2011). The clustering analysis revealed two main groups, one corresponding to the coastal, and the other to the Puna and Monte ecoregions (Figure 2, left).

The carrying capacity of the Andean habitats and the distribution of plant and animal species most probably changed during major Holocene climate events. Ecoregional boundaries and environmental parameters might have changed accordingly, except for gross relative differences due to elevation, hydrology, latitude, and continentality. Land resources would have suffered severe stress during the Middle-Holocene (see below) (Grosjean et al., 2007; Núñez and Grosjean, 2003), or since the Early Holocene in Atacama (Santoro et al., 2017). The actual mechanisms linking the diversity and distribution range of animals with vegetation and ultimately abiotic conditions are complex and influenced by several factors (Begon et al., 2006; Clarke and Gaston, 2006; Jiménez-Alfaro et al., 2016; Prugh et al., 2018). It is likely that terrestrial vertebrates concentrated in scattered favorable habitats of the Western Valleys and Highlands (Núñez and Grosjean, 2003). Changes in surface-water temperature of the Pacific Ocean, productivity of the upwelling system and its biodiversity, and body size of fish and shellfish (and their gross nutrient return) were also documented during the Holocene (Castro et al., 2016; Llagostera Martinez, 1979), as well as in the occurrence of the ENSO phenomenon (see below). Therefore, the variations in modern environmental conditions summarized above were considered here as ordinal differences between ecoregions and as a preliminary baseline.

On the other hand, I grouped assemblages into three temporal blocks that roughly coincided with major climatic phases that directly affected terrestrial primary productivity and water supply throughout the SCA. Relative dating based on Andean cultural sequences was used for the temporal classification of archaeofaunal assemblages because radiocarbon dates were only available for 153 (57%) of the 268 assemblages recorded in the database. The time span under study extended from the earliest human-fauna interactions from the end of the Pleistocene (ca. 11,000 rcbp, 11,000 cal BC) to the Late-Holocene (up to ca. the 16th century AD). This period was punctuated by the following major environmental, socio-cultural, and demographic transformations:

Block 1 (ca. 11.000–5000 BC). It included the Paleoindian and Early Archaic, which roughly correspond to the end of the Pleistocene, the Early Holocene, and the beginning of the Middle-Holocene. The unification of the Late Pleistocene and the Early Holocene is problematic for paleoenvironmental and archeological reasons, but evidence dated exclusively to the Paleoindian period consists of four assemblages from two sites and two ecoregions. The main environmental trends of this block would have been glacial retreat, increase in temperature and humidity, especially during the Second Central Andean Pluvial Event (13,000–10,000 BP), rise in lake and sea levels and, prior to the Early Holocene, the formation of localized oases, wetlands and riparian woodlands in Atacama (Baied and Wheeler, 1993; Santoro et al., 2017; Tchilinguirian et al., 2014; Tchilinguirian and Morales, 2013). More favorable conditions at the Pleistocene-Holocene transition probably opened ecological corridors connecting diverse habitats and populations (Borrero and Santoro, 2022; Santoro et al., 2017). This block begins with the earliest archeological signs of human colonization of the SCA, but this evidence likely reflects an advanced stage of this process. The earliest populations most likely belonged to metapopulations distributed over large megapatches and would have been equipped with cultural pools of subsistence techniques that could be selectively applied according to local circumstances (Borrero and Santoro, 2022). Part of the reviewed literature postulates a maritime adaptation in the coastal area, while highland strategies focused on terrestrial mammals, such as artiodactyls and large rodents, would have been implemented at least since the early Archaic (Aldenderfer, 1989; DeFrance et al., 2009; Grosjean et al., 2007; Standen et al., 2004). Pleistocene coastal subsistence would have been rather simple (e.g. small bag nets, collection from intertidal pools) and restricted to the shore (Llagostera, 1992; Santoro et al., 2019; Standen et al., 2004), while Early Archaic fishermen would have had a more specialized adaptation and more diverse fishing techniques (hooks and harpoons, among others). Béarez et al. argue for the possible presence of watercrafts and beach seines and gillnets near the southern limit of the SCA Pacific coast during the Early Archaic (2015). Fertile coast populations would have also exploited inland oases located 30–50 km upstream from the shore. Post-Pleistocene climate probably led to a redistribution of the population (e.g. people appear to have left the basins east of the Coastal Range toward either the coast or the highland (Santoro et al., 2017).

Block 2 (ca. 5000–3000 BC). It covered the Middle Archaic Period and most of the Middle-Holocene. It was characterized by a prolonged warm and dry event. The calibrated chronology of this multi-millennial drought varies between regions depending on the paleoenvironmental record (Grosjean et al., 2007), and may date back to the Early Holocene in the Atacama Desert (Santoro et al., 2017). The abandonment of the southwestern highlands (Chilean salt puna / Central Andean dry puna) and desert valleys or a reorganization of the settlement patterns around ecological refugia have been proposed during this period (Aldenderfer, 1989; Grosjean et al., 2007; Núñez A et al., 1999; Santoro et al., 2017). Sea level seemed to continue to rise until reaching modern levels ca. 4000 BC. Two relevant trends might have occurred during Block 2: a reduction in residential mobility and the intensification of subsistence activities (Aldenderfer, 1989). Coastal groups would have shown more diverse fishing and hunting toolkits and some bioindicators of navigation and diving (Castro et al., 2016; Standen et al., 2004). Among these techniques, bast fiber seine nets require significant time investment for manufacture/maintenance and are limited by the local availability of suitable plants (e.g. Asclepias), but allow for mass capture of fish (Beresford-Jones et al., 2018). Wild fibers could have been replaced by cultivated cotton at the end of this block. However, differences in the dates of some innovations (e.g. cactus spines, bone and shell hooks and harpoons) have been reported among archeological regions (Llagostera, 1992; Santoro et al., 2019), and inferences about navigation have been based solely on biological proxies. Evidence on the southernmost coast of Perú suggests that maritime specialization was complemented with foraging inland for wild resources and some cultivated plants and domestic animals toward the end of the Middle Archaic (Beresford-Jones et al., 2018). Innovations in hunting organization and techniques would also have occurred in the highlands, such as collaborative hunting and task division (beaters that drive game), use of blinds, changes in weapon systems (from atlatl to short-range throwing spear) (Moreno et al., 2021; Yacobaccio, 2006). Specialized big-game hunting eventually led to and continued along with domestication and early horticulture and pastoralism. The SCA have been proposed to be an independent center for domestication of the llama (Lama glama) from local stocks of guanaco (L. guanicoe cacsilensis) (Yacobaccio, 2001). The cultivation of edible cultivars from northwestern Argentina and northern Chile (Zea mays, Phaseolus, Solanum tuberosum), and the southern Peruvian coast limiting the SCA (e.g. Phaseolus, Canavalia, Psidium) are dated to this block (Pearsall, 2008).

Block 3. (ca. 3000 BC–1600 AD). It included the Late Archaic and the subsequent Agro-pottery periods (from the earliest agricultural villages of the Formative or Initial Period to the Inka and Hispano-Indigenous periods), as well as the Late-Holocene. It showed the current climate, with colder and wetter conditions than those of the Middle-Holocene. The impact of El Niño Southern Oscillation (ENSO) on rainfall and marine productivity was probably stronger during this block (DeFrance et al., 2009; Williams et al., 2008). An increase in deep-water upwelling and marine productivity would have occurred since 4000 cal BP (Mohtadi et al., 2004). In the Late Archaic there were foraging societies living under modern climate conditions. Intensification would have led to domestication and food-production techniques. Societies in the Agro-pottery stage were sedentary, with subsistence strategies consisting of a combination of animal husbandry, agriculture, hunting and gathering. They were organized in polities ranging from egalitarian villages to imperial states. Other innovations included pottery, metallurgy and the diffusion of the bow and arrow. According to radiocarbon time series, this stage would have been marked by a demographic transition to exponential growth during the Late-Holocene for most of the SCA (Gayo et al., 2015; Goldberg et al., 2016), while the coast would have followed a serrated pattern throughout the Holocene (Gayo et al., 2015). Marine resources resulting from maritime specialization probably continued on the coast, but they would have been complemented by domestic terrestrial resources. Finally, coastal-inland interactions would have begun ca. 4000 cal BC (Santoro et al., 2019).

Measurement of taxonomic diversity and statistical analyses

Taxonomic diversity has two attributes (Lyman, 2008): (1) composition, or which taxa are part of the measured set; and (2) structure, or the frequencies of these taxa. The taxonomic level used in this study was family. Most of the measurements were calculated from nominal data, as many reports were limited to lists of identified taxa, and I chose to include the largest number of assemblages.

Previously, I generated R data frames from spatial SQL queries on the DBZACS. The data frames included the following data on the archaeofaunas: taxonomic family, relative chronology (archeological period and temporal block), terrestrial ecoregion (Olson et al., 2001), minimum geodesic distance from the current coastline (polygon layer 10m_coastline, https://www.naturalearthdata.com/), and current elevation above sea level (SRTM Digital Elevation Data v4, DEM layer generated on Google Earth Engine) (Jarvis et al., 2008). Most of the data are summarized in Supplemental Material 3 and 4.

The following measures and plots were generated with R (v. 4.1.3, all packages updated as of May 2, 2023):

• Taxonomic Richness (NTAXA): it is a count of the total number of different taxa of the same rank. I derived the basic statistics from NTAXA (e.g. sample size, median and interquartile range (IQR) and obtained the box plots for each terrestrial ecoregion, for the coastal and highland-monte ecoregions aggregates, and for each temporal block. To monitor trends at the sub-ecoregion scale, they were also tallied for the following ecozones: fertile coast and the adjacent intermediate and high valleys and for the sterile coast and its adjacent valleys (Grosjean et al., 2007). These statistics allow the longitudinal and cross-sectional study of trends in the data.

• Ubiquity: it is the number of analytical units (i.e. assemblages) in which a given taxon occurs (Lyman, 2008) and can be used to measure the taxonomic abundance or frequency for an encompassing spatio-temporal unit such as the ecoregion. Ubiquity was standardized as the percentage of assemblages within a given ecoregion and temporal block where a taxon was present. It was also calculated for ecozones and ecoregion aggregates.

• Rank-order abundance: it is the ordinal abundance of taxa for a given assemblage and is derived from frequency measures such as Number of Identified Specimens (NISP) or Minimum Number of Individuals (MNI). The former measure was used here, and the taxon with the highest NISP in each assemblage was assigned a value of 1, the taxon with the second highest NISP a value of 2 and so on. The four highest ranked taxa in each assemblage were later selected and used to tabulate the percentage of taxon/rank order by ecoregion aggregate and temporal block (e.g. percentage of assemblages having Camelidae with the first highest NISP value for Coastal Ecoregions during Block 1). NISP was reported for slightly more than half of the assemblages examined. This measure was restricted to ecoregion aggregates to maximize sample size.

Wilcoxon or paired Mann-Whitney U tests were used to detect significant longitudinal and cross-sectional differences in taxonomic richness (Carlson, 2017; Lyman, 2008; Wolverton et al., 2016). The correlation between NTAXA and the minimum geodesic distance from the current coastline for each temporal block was assessed by a Spearman $ρ$ correlation test. The significance level was p < 0.05 for all analyses. These tests follow the criterion of interpreting paleozoological measurements as ordinal scale units (Lyman, 2008; Wolverton et al., 2016).

The present study was conducted from a liberal approach (Surovell and Waguespack, 2009) and included all available data that met some basic requirements, thus raising some potential problems. Zooarchaeological measures are sensitive to sampling error (Lyman, 2008) and to recovery bias due to excavation techniques (e.g. Thomas, 1969). This is particularly true for taxonomic richness, as taxa with low accumulation rates may be missed entirely in small assemblages (Lyman, 2008). Moreover, erroneous conclusions may be drawn when comparing assemblages with different taphonomic histories related, for example, to the spatial organization of foodways, recurrence of activities over different time intervals, and post-depositional processes mediating between diets and their archaeofaunal proxies.

Unfortunately, most reports omit the information required to control sampling error (e.g. screening is almost never reported). Despite the common occurrence of the aforementioned problems in synthetic zooarchaeological studies, they will be probably mitigated by random variation among faunal assemblages excavated over decades by dozens of different teams (e.g. McKechnie and Moss, 2016). In a preliminary attempt to control sampling errors, I compared the distribution of NISP values among ecoregions, which are the main spatial unit of analysis in the present paper, and the distribution of NTAXA values between assemblages with and without NISP values using Wilcoxon tests. The guiding hypotheses were that: (1) the known and unknown (i.e. not reported) NISP values are similar and randomly distributed if both sets are sufficiently numerous and that the loss of marginal taxa due to sample size and the resulting NTAXA values are similar between sets; and (2) a similar distribution of NISP values between ecoregions implies a similar random loss of marginal taxa, and that differences in taxonomic richness can be partially explained by other factors such as prey choice. Neither test rejected the null hypothesis. A more accurate sampling error control and a best-evidence synthesis strategy will become feasible as the literature survey progresses.

Results

Cluster analysis based on Jaccard’s dissimilarity matrix separated the archeological families into two main groups, one for the coastal ecoregions and the other for puna and monte (Figure 2, right). Grouping families by higher taxonomic or phenetic categories, such as Pisces or Bivalvia, highlights the importance of marine animals in many assemblages from coastal ecoregions. In the Pacific coast ecoregions, terrestrial vertebrates (Tetrapoda) accounted for 14–32% of the families, while marine animals corresponded to 24–61%, 21–31%, and 5–12% of fish, mollusk and shellfish families, respectively. On the other hand, tetrapods made up as much as 70–100% of the families from the Puna and Monte ecoregions, while the remaining families belonged to fish (0–7%) and invertebrates (10–22%).

Box plots of the longitudinal temporal evolution of median family NTAXA for each studied ecoregion and ecoregion aggregate are shown in Figure 3. They exhibit considerable longitudinal and coeval variation for the measures of central tendency (median) and dispersion (Q1 and Q3, Q1 - 1.5 IQR, Q3 + 1.5 IQR and outliers). During Block 1, the median NTAXA for the Sechura Desert is lower than that for the Central Andean Puna, but it has higher maximum dispersion values (Q3 and upper whisker). These variations can be attributed to different causes, such as functional and environmental differences between sites within a same ecoregion and block or to sampling errors. Further studies, together with emerging evidence will help to identify the multiple factors that could be at play.

Figure 3.

Boxplots of taxonomic richness (family NTAXA) in faunal assemblages by terrestrial ecoregion and temporal block (b1: Block 1, Paleoindian to Early Archaic, Late Pleistocene to Early Holocene. b2: Block 2, Middle Archaic, Middle-Holocene; b3: Block 3, Late-Holocene, Late Archaic to Agro-Potter). Median: thick black line; upper and lower quartiles: box; 1.5× interquartile range: whiskers; outliers: circles.

The widest IQRs were obtained for Sechura - Block 1 (IQR 12.75) and Atacama - blocks 2 (IQR 12.5) and 3 (IQR 7), while the IQRs for the remaining ecoregions and blocks fall between zero and five. The IQRs for the highlands aggregate are seven, three and three for temporal blocks 1–3 respectively, and six, 10.5, and seven for the Coastal aggregate.

In the Paired Wilcoxon (Mann-Whitney U) test, the null hypothesis of no difference in median NTAXA between blocks by ecoregion was rejected in the following comparisons: (1) the median NTAXA was higher for Block 1 than for Block 3 in the Sechura Desert and the Central Andean Puna; (2) the median NTAXA was lower for Block 1 than for Block 3 in the Chilean Matorral.

Cross-sectional comparisons of median NTAXA between ecoregions by temporal block showed that the null hypothesis was rejected for the Sechura Desert during Block 3, with a median NTAXA lower than those in the High Monte, Central Andean Puna, Central Andean wet Puna and Chilean Matorral, as well as in the Central Andean wet Puna. The median NTAXA in the latter ecoregion was significantly higher than those in the Sechura and Puna ecoregion. No significant longitudinal or cross-sectional differences in median NTAXA were found between ecoregional aggregates.

In brief, descriptive statistics showed differences among ecoregions and blocks but statistical tests did not reveal any clear difference in subsistence strategies between the coastal and highland ecoregions, or in their historical trends toward intensification, failing to support two of the main hypotheses postulated by the reviewed authors. This may be partially due to pooling together assemblages close to the current coastline with others coming from inland sites within the coastal ecoregions. For example, some sites in the Atacama Desert ecoregion are located at a geodesic distance of up to 170 km from the coastline and at elevations above 2800 m a.s.l. (i.e. in the high valley) where a highland-associated diet would have been expected.

Therefore, I used the aforementioned statistics and tests to search for significant differences in median NTAXA among the western ecozones described by Grosjean et al. (2007): the fertile and sterile coasts and their respective Intermediate and high valleys areas (Figure 4). The median NTAXA in the sterile and fertile coasts ranged from 5 to 17, while the values in the Intermediate and high valleys ranged from one to five. The interquartile ranges were highly variable for both the coasts and valleys. The Mann-Whitney test did not reject the null hypothesis between blocks by ecozone. On the other hand, during Block 1 the median NTAXA in the Fertile and Sterile coasts were significantly higher than those in the fertile high valleys and the median NTAXA in the Sterile coast was significantly higher than that in the high valleys.

Figure 4.

Boxplots of taxonomic richness (family NTAXA) in faunal assemblages by ecozone, except for highlands and monte (Grosjean et al., 2007), and temporal block. See Figure 3 for description of temporal blocks.

The Spearman $ρ$ correlation analysis between geodesic distance from the current coastline and NTAXA is an alternative test to evaluate the hypothesis of a geographic divergence of foodways (Table 3). Results indicated a moderate and significantly negative correlation (i.e. NTAXA increased with decreasing distance of sites from the coastline) during blocks 1 (r −0.47, p < 0.01) and 2 (r = −0.57, p < 0.05). No correlation was found during block 3 (r = 0.09, p > 0.05), possibly due to a combination of an actual taxonomic diversity reduction in coastal assemblages and the addition of assemblages from the Central Andean wet Puna in this block, with higher family NTAXA values than those from other Puna ecoregions.

Table 3.

Spearman’s $ρ$ (Rho) correlation coefficient and p-value for family NTAXA versus geodesic minimum distance from current coastline by temporal block. See Figure 3 for description of temporal blocks.

	Rho	p
b1	−0.4726285	0.0002689
b2	−0.5696054	0.0029572
b3	−0.0939025	0.1928008

To provide a first approach to the relative contribution of different families to the zooarchaeological record, the percentages of ubiquity of five families with the highest values were plotted for each ecoregion or ecoregion aggregate and temporal block (Figure 5). Camelidae was the most ubiquitous family in the highlands and Monte ecoregions throughout the whole temporal sequence. This family reached standardized values close to 100% of the assemblages during Block 1, followed by Cervidae and Chinchillidae (86%). Camelidae maintained high values (83–100%) in the subsequent temporal blocks, while the other taxa fell below 50%. A notable exception was observed in the Central Andean wet Puna, where Cervidae, Caviidae (e.g. domestic guinea pigs) and Cyprinodontidae (i.e. fishes of the genus Orestias from Lake Titicaca) had values above 50% during Block 3. It should be kept in mind that Ctenomyidae and Abrocomidae, which comprise species of small fossorial rodents, are usually considered intrusive in archeological contexts. In summary, it could be argued that highland assemblages are characteristically low in diversity.

Figure 5.

Percentage of ubiquity by terrestrial ecoregion and temporal block, considering the five most ubiquitous families. See Figure 3 for description of temporal blocks.

In coastal ecoregions Camelidae showed a higher variation in the percentage of ubiquity during blocks 1 and 2 (36–100%). Other taxa such as mollusks (Fissurellidae, Mytilidae, Muricidae, Chitonidae, 30–70%), marine carnivores (Otariidae 86%) and crustaceans (Carangidae 57%) showed high but variable values. The mass collection of invertebrates and fishes can rival and even surpass the net nutritional gain obtainable from large terrestrial herbivores (Ugan, 2005). For example, mussels (Mytillidae) can form dense banks in intertidal and subtidal zones. Camelidae became clearly dominant during Block 2 in Atacama and Sechura deserts (80–85%), while the other taxa fell below 25–50%.

Camelids showed an increasing trend in both coastal and highland ecoregion aggregates. This could be related to domestication, the development of pastoral lifestyles and an increase in exchange or trade between groups with different ecological or economic specializations. These trends were previously identified by Yacobaccio (2006), who sampled 20 published highland sites from northwestern Argentina and northern Chile and found that Camelidae became the most abundant taxon, increasingly dominating the assemblages.

Figure 6 shows the percentage of ubiquity of different families for each ecozone and temporal block. Camelid ubiquity ranged from 0 to 100% in the Sterile and fertile coasts and from 50 to 100% in the valleys and non-camelid families had higher values on the coast compared to inland valleys. Marine Families were dominant on the coast and occurred in the Intermediate (Otariidae, Cancridae) and high valleys (Mytilidae). Although no significant diachronic trend could be established, the ubiquity values in Intermediate and high valleys were more similar to those in the highlands, except for the occasional presence of some marine taxon.

Figure 6.

Percentage of ubiquity by ecozone and temporal block considering the five most ubiquitous families. See Figure 3 for description of temporal blocks.

Figures 7 and 8 show the percentages of rank order abundances (first-fourth highest NISP values per assemblage) for ecoregion aggregates through time. Camelidae is the most abundant family in most assemblages and shows an increasing trend through the Holocene. However, the percentage of assemblages where Camelidae is the most abundant family (highest NISP) is higher in the highlands (50–91%) than in the coastal ecoregions (33–65%).

Figure 7.

Frequency of rank order abundance for the five highest NISP values of each assemblage (family rank) by temporal block (highland and monte ecoregions). See Figure 3 for description of temporal blocks.

Figure 8.

Frequency of rank order abundance for the five highest NISP values of each assemblage (family rank) by temporal block (Coastal terrestrial ecoregions). See Figure 3 for description of temporal blocks.

Discussion

Changes in subsistence are mainly explained by demographic, environmental, and technical factors. The combination of the former two may drive the intensification of existing practices or may act as selective forces on alternative strategies and techniques.² Technical factors primarily refer to innovations, defined as widely adopted novel techniques³ that have an impact on foodways. Tools and facilities can improve the efficiency of food acquisition through a number of ways, but their production and maintenance increase handling time, and investments are subject to diminishing returns above an optimal threshold that depends on techniques and targeted resources (Bettinger et al., 2006; Ugan et al., 2003). Here I attempted to explore these topics within the broader context of Andean archeology, with taxonomic diversity serving as a proxy for past diets. My aim was to provide possible interpretations of the preliminary results obtained from the ongoing systematic review of SCA zooarchaeology.

The early diversification of lifeways and subsistence strategies is a common topic in the literature. Diversification was particularly evident between the coast and the highlands (Aldenderfer, 1989; Borrero and Santoro, 2022; DeFrance et al., 2009; Grosjean et al., 2007). The former supported broad-spectrum diets related to the high marine productivity and biodiversity, while foraging in inland valleys and the highland was restricted to terrestrial resources. Subsistence intensification was a major trend during the Middle and Late-Holocene. Intensification may be achieved through different pathways, which can be complementary: diversification, specialization, and increased investment (i.e. proper intensification) (Betts and Friesen, 2004). Specialization and intensification are relative categories that can be applied independently to different aspects of subsistence, such as prey taxa, resource patches or technology. In ecological research, specialization refers to either taxa or anatomical parts (Begon et al., 2006). In the SCA, intensification led to the diversification of tools and facilities involving food acquisition, changes in diet breadth and ecological complementarity, the domestication of plants and animals, and the diffusion of exotic domesticates (Aldenderfer, 1989; Grosjean et al., 2007; Standen et al., 2004; Yacobaccio, 2006).

Late Pleistocene marine collectors on the Pacific coast used low-investment techniques to exploit the intertidal zone and shallow waters (Block 1). The Archaic Period (Block 1 - early Block 3) was marked by more intensive maritime exploitation, as evidenced by more diverse industries. In Sechura and Atacama, fertile inland areas (i.e. floodplains, oases, lomas) provided terrestrial resources, including large herbivores. Núñez Atencio proposed a seasonal transhumant circuit that traversed the seaside and the northern intermediate valleys during the Archaic period (Aldenderfer, 1989; Núñez, 1983). On the other hand, Schiappacasse and Niemeyer (1984) argue for a more circumstantial occupation of inland patches. The Middle-Holocene / Middle Archaic (Block 2) is characterized by the diversification of fishing, hunting, and processing tools, the intensification of netting and an increase in exploited taxa and biotopes, and potential evidence of navigation has been also proposed for this block (Aldenderfer, 1989; Beresford-Jones et al., 2018; Castro et al., 2016; Santoro et al., 2019; Standen et al., 2004).

Other lines of evidence strongly support the continuation of marine diets along the coast until the end of Block 3 (Correa et al., 2018; Díaz-Zorita Bonilla et al., 2016; Moragas Wachtendorff and Mendez-Quiros Aranda, 2022; Tomczak, 2003). According to Tomczak (2003), specialized groups with particular subsistence bases and who were subject to common polities inhabited the area between the seaside and the Coastal Range, near the northern limit of the fertile coast. Food production was introduced later in the far south of the SCA coast (30° S) (Pascual et al., 2019; Troncoso et al., 2016). Additionally, unlike other Andean regions, the coast does not demonstrate a Late-Holocene demographic transition (Gayo et al., 2015).

The results of this study indicate that there was an inverse correlation between distance from the coastline and taxonomic richness prior to the Late- Holocene. Taxonomic richness at coastal sites was likely accounted for by fish, shellfish, marine mammals and shorebirds. The ubiquity of Camelidae differed between coastal ecoregions and temporal blocks but showed a rising trend in both Atacama and Sechura ecoregions since the Middle-Holocene. The stratification of Atacama and Sechura assemblages into ecozones allowed a more detailed scenario. The assemblages close to the coast remained highly diverse throughout the Holocene, with median NTAXA peaks during Block 2. Camelids had high ubiquity values in the Fertile coast during blocks 1 and 2 and in the Sterile coast during Block 3. Shellfish were more ubiquitous than fish before Block 3. Finally, remains of otariids, fish and gastropods were present in the Intermediate and high valleys parallel to the fertile coast.

Acquisition and processing of fish and invertebrates as aggregates can yield high net returns (Madsen and Schmitt, 1998; Ugan, 2005). Early Archaic foragers (Block 1) possibly used small bag nets (Standen et al., 2004) as well as larger coastal seine nets or gill nets (Béarez et al., 2015), the latter facilitating mass collection. Seine nets, which require significant labor and result in delayed returns were likely employed by cooperative groups with some level of hierarchization. The devices were utilized to collect schools of fish of varying body size composed of the families Sciaenidae, Carangidae, Mugilidae and Engraulidae (Beresford-Jones et al., 2018; Disspain et al., 2017; Santoro et al., 2019). Some authors have suggested that these families were also captured with fishing line. Net production appears to have increased by the end of Block 2 (Beresford-Jones et al., 2018). Offshore navigation and fishing for large fish families (e.g. Xiphidae) most likely required an adequate watercraft and the cooperation of large groups (Castro et al., 2016; Salazar et al., 2020).

In addition to the indications of mass collection and open-sea fishing for certain taxa, other tool types suggest individual acquisition techniques such as fishing line and the gathering of mollusks from intertidal banks. Hunting of sea mammals and exploitation of land resources (e.g. camelids) also continued during the Archaic, with the inclusion of cultivars since blocks 2 (North) and 3 (South). Coastal populations invested in various techniques and instruments, thus covering a wide range of resources with different net returns. The diet breadth model assumes that diversification maximizes the net caloric return if high-ranked prey are relatively scarce. However, diversification could have been favored by both seasonal and unpredictable fluctuations of resources (e.g. random short- or long-term climatic events). The Pacific coast of the SCA was characterized by weakened upwelling, reduced productivity and less intense El Niño activity for most of the Early and Middle-Holocene (Mohtadi et al., 2004), and enhanced El Niño events for the Late-Holocene (Mohtadi et al., 2004; Moy et al., 2002; Riedinger et al., 2002). Changes in sea and land productivity at different time scales could have triggered episodes of resource stress. The summed probabilities of radiocarbon dates indicate millennial-scale cycles of population growth and decline on the coast during the Holocene (Gayo et al., 2015), suggesting that fluctuations in atmospheric and oceanic conditions posed a severe constraint on maritime specialization and population. It is important to note that prey grouped at higher taxonomic levels (i.e. Mammalia, Mollusca, Pisces) vary in nutritional composition and are not wholly interchangeable, and diversification would have been partially directed to meet the minimal requirements for multiple nutrients (Haws and Hockett, 2004; Kelly, 1995).

According to the reviewed literature, inland and highland populations depended on terrestrial resources. Wild camelids would have been a valuable prey since the late Pleistocene, showing an increasing relative contribution during the Holocene. This development involved specialized hunting, the implementation of new hunting techniques and the subsequent domestication of llamas. Intensification can be dated back at least to the middle-Holocene (Núñez and Grosjean, 2003; Yacobaccio, 2006). The development of specialized hunting, llama husbandry, plant domestication, agriculture and trade (diversification) occurred during blocks 2 and 3. Time-series analyses of radiocarbon dates for South America (Goldberg et al., 2016) and the SCA (Gayo et al., 2015) indicate a distinct demographic transition in the highlands during the late-Holocene (Block 3), probably associated with the establishment of food production.

Results partially agree with the aforementioned scenario for the intermediate and high valleys and the highlands. Cross-sectional comparisons of median richness between coastal and inland/highland ecozones, ecoregions, and ecoregion aggregates failed to show statistically significant differences in most cases. However, there was less variability in taxonomic richness revealed for the intermediate/high valleys and highland assemblages and correlation analyses revealed that it was significantly lower in the inland, highland and monte than in the coast until the Middle-Holocene. Camelidae was the most represented family based on ubiquity values and percentage of rank order abundances, gaining importance since the Middle-Holocene, in agreement with previous studies (Olivera, 1997; Yacobaccio, 2006). Such increase in frequency occurred simultaneously with the appearance of new tools, including diversified weapon systems and hunting facilities (e.g. hunting blinds) dated to ca. 7200 BP in the puna. These tools suggest the formation of large hunting parties with specialized tasks (e.g. beaters driving the animals, killers) in some cases.

During Block 1 in the highlands, Cervidae and Chinchillidae exhibited high ubiquity values and percentage of rank order abundances. Cervids (H. antisensis) are large herbivores (high-ranking prey) showing a mean biomass/km² comparable to that of camelids, but with a higher variance (Yacobaccio, 1994). Likewise, chinchillids can reach similar biomass values as camelids in the “Nutrient Concentration Zones” of puna (Yacobaccio, 1994). The high frequencies observed in early assemblages have been attributed to the exploitation of near-site resources (Yacobaccio, 1994). Ugan (2005) posits that the high costs of processing small terrestrial vertebrates result in low returns. In contrast, Jones (2006) argues that the mass collection of some small mammals partially violates the body-size proxy. The relative abundance of Chinchillidae and Cervidae decreased since Block 2 (Middle-Holocene), supporting the hypothesis of camelid specialization (Olivera, 1997; Yacobaccio, 2001; Yacobaccio and Vilá, 2013). It is possible that large rodents and cervids were more susceptible to the effects of Middle-Holocene desertification.

So far, the reviewed literature concerning archaeofaunas from the wet Puna and High Monte reports sites dated to the Late-Holocene only. The patterns of the assemblages from High Monte were similar to those from Central and Central dry Puna of the same period, suggesting that the respective Preceramic diets could have been comparable as well. Vicuña’s current distribution is excluded from the High Monte. On the other hand, guanacos would have been abundant in this ecoregion given their high behavioral flexibility and current wide latitudinal and altitudinal ranges (Vilá, 2012). Within the Central Andean wet Puna, Cyprinodontidae (fishes of Orestias spp.), Cervidae and Caviidae had remarkably high frequencies. Cyprinodontidae were particularly concentrated in the Lake Titicaca basin. Caviidae was represented by records of guinea pig (Cavia porcelus), a rodent presumed to have been domesticated in the Peruvian Central Andes and later imported to the SCA. Post-Hispanic families (e.g. Bovidae) have been also documented in Block 3 due to temporally transgressive assemblages.

Camelid specialization persisted during the Late-Holocene despite the establishment of agropastoral economies that supported the exponential population growth. Richerson et al. (2009) postulated that the rates of technical and social innovation are the main limiting factors for population growth. During the Holocene, several technical innovations allowed the reduction of hunting distance (prey interception and driving), development of projectile weaponry and the simultaneous capture of multiple prey (Aschero and Martínez, 2001; Moreno et al., 2021; Yacobaccio, 2006). The spread of the bow and arrow occurred during Block 3. Arrows would have been more effective than previous weapons against fast and elusive prey such as vicuñas (Tomka, 2013). Resource depletion could have been first prevented by alternating between different ravines and, later, through wild herd management, semi-captive breeding and domestication. On the other hand, meat drying and the subsequent appearance of ceramic cooking vessels contributed to an increase in the gross return of animal parts.

Despite the undeniable importance of camelid pastoralism for the Andean societies, it had some drawbacks worth considering, such as the constant care required by herds. This was compensated for by the year-round availability of animals and the advantages of selective breeding for valuable traits. In addition, environmental engineering techniques were used to enhance and expand wetlands for camelid pasture (Lane et al., 2022). Domestication, animal husbandry and agriculture were mutually beneficial through the utilization of crop stubble (e.g. maize) as feed for domestic herds and llama dung as fertilizer (Langlie and Capriles, 2021). On the other hand, risks associated with pastoralism included high mortality rates of young animals due to predation and diseases like enterotoxemia. In this regard, some authors suggested that hunting remained important until very late in pre-Hispanic times partly because it helped to reduce pressure on domestic herds in the context of mono-specific husbandry (Escola, 2002; Göbel, 1994; Yacobaccio et al., 1997), this being analogous to the multi-species (e.g. cattle, goats and llamas) husbandry practices adopted by post-Hispanic Highland pastoralists.

Conclusion

This study is part of the ongoing ZDBSCA project. Although the literature survey is still in progress, the current sample stands out as the most exhaustive dataset of its kind available to date. It provides a characterization of the taxonomic diversity over large spatio-temporal units, revealing trends that allow for interesting preliminary insights into current hypotheses of pre-Hispanic subsistence strategies.

The analysis of archaeofaunal taxonomic diversity indicated a contrast between the trajectories of the coastal ecoregion and those of the inland and highland ecoregions. The Coast displayed a greater diversity of strategies in terms of prey taxa, foraging techniques, and complementary exploitation of marine and terrestrial habitats. One possible explanation for this diversification relies on the adaptation of coastal groups to the limited availability of large terrestrial mammals, resulting in a wider diet breadth aimed to maximize net nutritional return. Other factors contributing to the diversity in coastal diets would be related to risk-aversion under conditions involving unpredictable resource availability and the varying nutritional composition of fish and shellfish. Different optimization goals were likely at play. Inland strategies ended up in camelid specialization, which led to diversification and intensification in terms of domestication (i.e. llama) and, at the technical level, to multiple hunting and food processing techniques, animal husbandry and landscape engineering. Plant domestication and eventually agriculture also contributed to further diversification of subsistence. There may have been reasons other than caloric return for animal exploitation under agropastoral economies. Climate fluctuations likely favored intensification and innovation during the Holocene. The establishment of agro-pastoralist economies during the Late-Holocene in inland and highland regions led to exponential population growth. It is worth noting that socio-political factors such as gender labor division, emerging hierarchies and class systems might have also influenced subsistence and diet in terms of food acquisition and distribution, but these topics are beyond the scope of the present study.

The present exploratory investigation was approached within the framework of the Diet-Breadth Model, which shows some limitations, as is the case for other zooarchaeological applications. First, it only relies on one type of evidence and involves a single subcomponent of real diets. For example, the potential contribution of animals in terms of protein requirements is a topic worthy of exploration, and both wild and domestic plants should be included if the focus is on the metabolic energy (Haws and Hockett, 2004). Second, the likelihood of sampling errors may have been increased by the fact that the sample size of the assemblages varied among ecoregions and blocks, data from two ecoregions were restricted to the Agro-potter periods, and there were no data for the Yungas ecoregion. Third, the taxonomy at the genus and species levels (particularly for fish and shellfish) was omitted here. Fourth, the coastal ecoregions encompassed more than one environmental zone and data were pooled together, mixing up ecological and functional evidence. Finally, other potential issues may arise from using a simplified optimization model to explain evidence that spans complex systems which evolved across different habitats.

The ZDBSCA project covers up to 2020 and the interval following the present study (from 2009 onward) has witnessed a significant expansion of the zooarchaeological literature within SCA, which I am currently reviewing and incorporating into the database. After its completion, the findings and conclusions drawn from this study will need to be revised for a more thorough examination involving the connection between human-animal interactions and the ecological and sociocultural dynamics of the pre-Hispanic period.

Supplemental Material

sj-csv-1-hol-10.1177_09596836241231446 – Supplemental material for Diet breadth and biodiversity in the pre-hispanic South-Central Andes (Western South America) during the Holocene: An exploratory analysis and review

Supplemental material, sj-csv-1-hol-10.1177_09596836241231446 for Diet breadth and biodiversity in the pre-hispanic South-Central Andes (Western South America) during the Holocene: An exploratory analysis and review by Carlos R Belotti López de Medina in The Holocene

Supplemental Material

sj-csv-2-hol-10.1177_09596836241231446 – Supplemental material for Diet breadth and biodiversity in the pre-hispanic South-Central Andes (Western South America) during the Holocene: An exploratory analysis and review

Supplemental material, sj-csv-2-hol-10.1177_09596836241231446 for Diet breadth and biodiversity in the pre-hispanic South-Central Andes (Western South America) during the Holocene: An exploratory analysis and review by Carlos R Belotti López de Medina in The Holocene

Supplemental Material

sj-csv-3-hol-10.1177_09596836241231446 – Supplemental material for Diet breadth and biodiversity in the pre-hispanic South-Central Andes (Western South America) during the Holocene: An exploratory analysis and review

Supplemental material, sj-csv-3-hol-10.1177_09596836241231446 for Diet breadth and biodiversity in the pre-hispanic South-Central Andes (Western South America) during the Holocene: An exploratory analysis and review by Carlos R Belotti López de Medina in The Holocene

Footnotes

Acknowledgements

Supplemental spatial and environmental data utilized for constructing thematic maps, queries and analyses were acquired from the following sources and repositories: World Wildlife Fund (Olson et al., 2001) terrestrial ecoregions (www.worldwildlife.org); Natural Earth Data for continental coastline vector layer (https://www.naturalearthdata.com/); the University of Texas at Austin GeoData Portal for LSIB South America vector layer from the USA Department of State (https://geodata.lib.utexas.edu/), Google Earth Engine (https://earthengine.google.com/) was used to create raster layers from SRTM Elevation v4 and MOD13A2 VI datasets; the Ocean Biogeographic Information System (OBIS) and the Global Biodiversity Information Facility (www.gbif.org) repositories were used to generate reports on modern occurrences of animal taxa. For the former, Supplementary Data 2 was generated by OBIS and details the original references, it is included here as required by the repository, while a stable DOI was generated by GBIF for consultation ().

I especially thank the experts across the four countries within SCA whose work served as the foundation for my research. The list of all references used to construct the complete zooarchaeological database (up to 2020) is presented in Supplementary Data 1. I am also grateful to Andrés Izeta, Cristina Scattolin, Luis Borrero, Luis Coll, María Eugenia de Feo, Daniel Olivera, Jennifer Grant, and the Digital Archeology Argentinian Network (RadAR), among others, for their encouragement and support. Finally, I would like to express my gratitude to the organizers and the session coordinators of the Sixth National Congress of Argentinian Zooarchaeology, as well as to the editors and the anonymous reviewers for their valuable insights that greatly improved the manuscript.

Funding

The author disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was supported by Agencia Nacional de Promoción Científica y Tecnológica (ANPCyT) Argentina (PICT 2019-01341) and the Wenner-Gren Foundation through a postdoctoral grant (Grant 9968) for the South-Central Andes Zooarchaeological Database Project. In particular, the latter played a key role in the literature survey and the development of the ZDBSCA during the first 2 years of the COVID-19 pandemic. The research was conducted at the Instituto de Las Culturas (Universidad de Buenos Aires - CONICET).

ORCID iD

Carlos R Belotti López de Medina

Supplemental material

Supplemental material for this article is available online.

Notes

References

Aldenderfer

(1989) The archaic period in the South-Central Andes. Journal of World Prehistory 3(2): 117–158.

Aldenderfer

(1998) Montane Foragers: Asana and the South-Central Andean Archaic. Iowa City, IA: University of Iowa Press.

Aschero

Martínez

(2001) Técnicas de caza en Antofagasta de la Sierra, Puna meridional argentina. Relaciones de la Sociedad Argentina de Antropología 26: 215–241.

Baied

Wheeler

(1993) Evolution of high Andean Puna ecosystems: environment, climate, and culture change over the last 12,000 years in the Central Andes. Mountain Research and Development 13: 145–156.

Béarez

Jackson

Mollaret

(2015) Early Archaic fishing (12,600–9,200 cal yr BP) in the Semiarid North Coast of Chile. Journal of Island and Coastal Archaeology 10(1): 133–148.

Begon

Townsend

Harper

(eds) (2006) Ecology. From Individuals to Ecosystems, 4th edn. Hoboken, NJ: Blackwell Publishing.

Belotti López de Medina

(2019) Base de datos zooarqueológica de los Andes Centro-Sur. Cuadernos del Instituto Nacional de Antropología y Pensamiento Latinoamericano - Series Especiales 7(2): en prensa.

Beresford-Jones

Pullen

Chauca

, et al. (2018) Refining the maritime foundations of Andean civilization: How plant fiber technology drove social complexity during the preceramic period. Journal of Archaeological Method and Theory 25(2): 393–425.

Bettinger

Winterhalder

McElreath

(2006) A simple model of technological intensification. Journal of Archaeological Science 33(4): 538–545.

10.

Betts

Friesen

(2004) Quantifying hunter–gatherer intensification: A zooarchaeological case study from Arctic Canada. Journal of Anthropological Archaeology 23(4): 357–384.

11.

Blute

(2010) Darwinian Sociocultural Evolution: Solutions to Dilemmas in Cultural and Social Theory. Cambridge: Cambridge University Press.

12.

Borcard

Gillet

Legendre

(2011) Numerical Ecology With R. New York: Springer.

13.

Borrero

Santoro

(2022) Metapopulation processes in the long-term colonization of the Andean highlands in South America. Journal of World Prehistory 35(2): 135–162.

14.

Browman

(1980) Tiwanaku expansion and altiplano economic patterns. Estudios Arqueológicos 5: 107–120.

15.

Brush

(1982) The Natural and human environment of the Central Andes. Mountain Research and Development 2: 19.

16.

Bunge

(1999) Buscar La Filosofía En Las Ciencias Sociales. México D. F: Siglo XXI editores.

17.

Bunge

(2009a) Tratado de Filosofía. Semántica I: sentido y referencia. Barcelona: Gedisa SA.

18.

Bunge

(2009b) Tratado de Filosofía. Semántica II: interpretación y verdad. Barcelona: Gedisa SA.

19.

Bunge

(2011) Tratado de Filosofía. Ontología I: el moblaje del mundo. Barcelona: Gedisa SA.

20.

Capriles

Albarracin-Jordan

Lombardo

, et al. (2016) High-altitude adaptation and late Pleistocene foraging in the Bolivian Andes. Journal of Archaeological Science Reports 6: 463–474.

21.

Carlson

(2017) Quantitative Methods in Archaeology Using R. Cambridge: Cambridge University Press.

22.

Castro

Aldunate

Varela

, et al. (2016) Ocupaciones arcaicas y probables evidencias de navegación temprana en la Costa Arreica de Antofagasta, Chile. Chungará (Arica) 48(4): 503–530.

23.

Clarke

Gaston

(2006) Climate, energy and diversity. Proceedings of the Royal Society B Biological Sciences 273(1599): 2257–2266.

24.

Concha Contreras

JDD

(1975) Relación entre pastores y agricultores. Allpanchis 7(8): 67–101.

25.

Correa

Carrasco

Ballester

, et al. (2018) Efectos colaterales de la transición al Formativo: una nueva culinaria entre los cazadores-recolectores marinos del desierto de Atacama. Chungará (Arica) 50(1): 87–106.

26.

DeFrance

Grayson

Wise

(2009) Documenting 12,000 years of coastal occupation on the Osmore littoral, Peru. Journal of Field Archaeology 34(3): 227–246.

27.

Díaz-Zorita Bonilla

Drucker

Richardin

, et al. (2016) Marine food consumption in coastal northern Chilean (Atacama Desert) populations during the formative period: Implications of isotopic evidence (C, N, S) for the Neolithic process in south central Andes. Journal of Archaeological Science Reports 6: 768–776.

28.

Dillehay

Núñez

(1988) Camelids, caravans and complex societies in the Soth-Central Andes. In: Saunders

de Montmollin

(eds) Recent Studies in Pre-Columbian Archaeology. Oxford: BAR International Series, pp.603–634.

29.

Disspain

MCF

Ulm

Santoro

, et al. (2017) Pre-Columbian fishing on the coast of the Atacama Desert, northern Chile: An investigation of fish size and species distribution using otoliths from Camarones Punta Norte and Caleta Vitor. Journal of Island and Coastal Archaeology 12(3): 428–450.

30.

Escola

(2002) Caza y pastoralismo: un reaseguro para la subsistencia. Relaciones de la Sociedad Argentina de Antropología 27: 233–245.

31.

Fiedel

(1996) Prehistoria de América. Barcelona: Crítica.

32.

Gayo

Latorre

Santoro

(2015) Timing of occupation and regional settlement patterns revealed by time-series analyses of an archaeological radiocarbon database for the South-Central Andes (16°–25°S). Quaternary International 356: 4–14.

33.

Göbel

(1994) El Manejo del Riesgo en la Economía Pastoril de Susques. Zooarqueología de Camelidos 1: 43–56.

34.

Goldberg

Mychajliw

Hadly

(2016) Post-invasion demography of prehistoric humans in South America. Nature 532(7598): 232–235.

35.

Gómez-Carballa

Pardo-Seco

Brandini

, et al. (2018) The peopling of South America and the trans-Andean gene flow of the first settlers. Genome Research 28(6): 767–779.

36.

González

Pérez Gollán

(1966) El Área Andina Meridional. In: Alcina Franch

(ed.) XXXVI Congreso Internacional de Americanistas. Sevilla, Vol. 1, pp.241–265.

37.

Grayson

Delpech

(1998) Changing diet breadth in the early upper Palaeolithic of Southwestern France. Journal of Archaeological Science 25(11): 1119–1129.

38.

Grosjean

Santoro

Thompson

, et al. (2007) Mid-Holocene climate and culture change in the South Central Andes. In: Anderson

Maasch

Sandweiss

(eds) Climate Change and Cultural Dynamics. London: Academic Press, Elsevier, pp.51–115.

39.

Harris

(1987) Foodways: Historical Overview and Theoretical Prolegomenon. In: Harris

Ross

(eds) Food and Evolution. Toward a Theory of Human Food Habits. Philadelphia, PA: Temple University Press, pp.57–92.

40.

Harris

(2010) Bueno Para Comer: Enigmas de Alimentación y Cultura. Madrid: Alianza Editorial.

41.

Haws

Hockett

(2004) Theoretical perspectives on the dietary role of small animals in human evolution. In: Brugal JP and Desse J (eds) Petits animaux et sociétés humaines. Du complément alimentaire aux ressources utilitaires. Antibes: Editions APDCA, pp.173–184.

42.

Hintze

Álvarez

Pinotti

(1997) Apuntes para un abordaje multidisciplinario del problema alimentario. In: Álvarez

Pinotti (eds) Procesos socioculturales y alimentación. Buenos Aires: Ediciones del Sol, pp.11–31.

43.

Isbell

Silverman

(2006) Regional Patterns. In: Isbell

Silverman

(eds) Andean Archaeology III: North and South. Boston, MA: Springer, pp.3–19.

44.

Jarvis

Reuter

Nelson

, et al. (2008) Hole-filled SRTM for the globe Version 4. Available from the CGIAR-CSI SRTM 90m Database. 15(25–54): 5. Available at: http://srtm.csi.cgiar.org

45.

Jiménez-Alfaro

Chytrý

Mucina

, et al. (2016) Disentangling vegetation diversity from climate-energy and habitat heterogeneity for explaining animal geographic patterns. Ecology and Evolution 6(5): 1515–1526.

46.

Jones

(2006) Prey choice, mass collecting, and the wild European rabbit (Oryctolagus cuniculus). Journal of Anthropological Archaeology 25(3): 275–289.

47.

Juba

Vannahme

Volkov

(2015) Learning PostgreSQL. Birmingham: Packt Publishing Ltd. Available at: https://www.packtpub.com/product/learning-postgresql/9781783989188

48.

Kelly

(1995) The Foraging Spectrum. Diversity in Hunter-Gatherer Lifeways. Washington: Smithsonian Institution Press.

49.

Lane

Beresford Jones

Coll

, et al. (2022) Pre-Hispanic anthropogenic wetlands in the upper Ica drainage, south-central Andes: Dating and context. Antiquity 96: 1251–1271.

50.

Langlie

Capriles

(2021) Paleoethnobotanical evidence points to agricultural mutualism among early camelid pastoralists of the Andean central Altiplano. Archaeological and Anthropological Sciences 13(7): 107.

51.

Larson

Fuller

(2014) The evolution of animal domestication. Annual Review of Ecology Evolution and Systematics 45: 115–136.

52.

Llagostera

(1992) Early occupations and the emergence of fishermen on the Pacific Coast of South America. Andean Past 3(1): 9.

53.

Llagostera Martinez

(1979) 9,700 Years of Maritime Subsistence on the Pacific: An Analysis by Means of Bioindicators in the North of Chile, vol. 44. Washington, DC: American Antiquity, pp.309–324.

54.

Lumbreras

(2008) Los orígenes de la sociedad andina. In: Contreras

(ed.) Economía Prehispánica. Compendio de historia económica de Perú. Lima: BCRP, IEP, pp.23–136. Available at: https://www.bcrp.gob.pe/docs/Publicaciones/libros/historia/economia/1-economia-prehispanica.pdf

55.

Lumbreras

S LG

(1969) El área cotradicional meridional andina. Boletín del Museo Nacional de Historia Natural 30: 65–79.

56.

Lupo

(2007) Evolutionary foraging models in zooarchaeological analysis: Recent applications and future challenges. Journal of Archaeological Research 15: 143–189.

57.

Lyman

(2003) The influence of time averaging and space averaging on the application of foraging theory in zooarchaeology. Journal of Archaeological Science 30: 595–610.

58.

Lyman

(2008) Quantitative Paleozoology. Cambridge Manuals in Archaeology. Cambridge; New York: Cambridge University Press.

59.

MacArthur

Pianka

(1966) On optimal use of a patchy environment. American Naturalist 100(916): 603–609.

60.

Madsen

Schmitt

(1998) Mass collecting and the diet breadth model: A great basin example. Journal of Archaeological Science 25(5): 445–455.

61.

Mahner

Bunge

(2001) Function and functionalism: A synthetic perspective. Philosophy of Science 68: 75–94.

62.

Marquez

(2015) PostGIS Essentials. Birmingham: Packt Publishing Ltd.

63.

Mayer

(2004) Casa, Chacra y Dinero. Economías Domésticas y Ecología En Los Andes. Lima: Instituto de Estudios Peruanos.

64.

McKechnie

Moss

(2016) Meta-analysis in zooarchaeology expands perspectives on Indigenous fisheries of the Northwest Coast of North America. Journal of Archaeological Science Reports 8: 470–485.

65.

Mengoni Goñalons

Yacobaccio

(2006) The domestication of South American camelids. In: Zeder

(ed.) Documenting Domestication–New Genetic and Archaeological Paradigms. Berkeley, CA: University of California Press, pp.228–244.

66.

Messer

(1984) Anthropological perspectives on diet. Annual Review of Anthropology 13: 205–249.

67.

Mintz

Du Bois

(2002) The anthropology of food and eating. Annual Review of Anthropology 31(1): 99–119.

68.

Mohtadi

Romero

Hebbeln

(2004) Changing marine productivity off northern Chile during the past 19 000 years: A multivariable approach. Journal of Quaternary Science 19(4): 347–360.

69.

Moragas Wachtendorff

Mendez-Quiros Aranda

(2022) La secuencia cronológica de Punta Patache y la ocupación de la costa arreica del desierto de Atacama (21ºS). Estudios atacameños 68: e5060.

70.

Morán

(2007) Adaptación. In: Barfield

(ed.) Diccionario de Antropología. México D.F: Siglo XXI Editores, pp.1–2.

71.

Moreno

Martínez

Aschero

, et al. (2021) Ancient hunting strategies of wild camelids through the study of Multiple lines of archaeological evidences at southern argentine Puna. In: Belardi

Bozzuto

Fernández

(eds) Ancient Hunting Strategies in Southern South America. The Latin American Studies Book Series. Cham: Springer International Publishing, pp.1–32.

72.

Moy

Seltzer

Rodbell

, et al. (2002) Variability of El Niño/Southern Oscillation activity at millennial timescales during the Holocene epoch. Nature 420(6912): 162–165.

73.

Murra

(1975) El control vertical de un máximo de pisos ecológicos. In: Murra

(ed.) Formaciones Económicas y Políticas Del Mundo Andino. Lima: Instituto de Estudios Peruanos, pp.59–116.

74.

Murra

(1978) La Organización Económica Del Estado Inca. México D.F.: Siglo XXI.

75.

Nakatsuka

Lazaridis

Barbieri

, et al. (2020) A paleogenomic reconstruction of the deep population history of the Andes. Cell 181(5): 1131–1145.e21.

76.

Núñez

A L

Grosjean

Cartajena

F I

(1999) Un ecorefugio oportunístico en la puna de Atacama durante eventos áridos del Holoceno Medio. Estudios Atacameños 17: 125–174.

77.

Núñez

(1983) Paleoindian and Archaic cultural periods in the arid and semiarid regions of northern Chile. Advances in World Archaeology 2: 161–203.

78.

Núñez

Grosjean

(2003) Biodiversity and human impact during the last 11,000 years in North-central Chile. In: Bradshaw

Marquet

(eds) How Landscapes Change: Human Disturbance and Ecosystem Fragmentation in the Americas. Berlin; Heidelberg: Springer, pp.7–17.

79.

Olivera

(1997) La importancia del recurso Camelidae en la Puna de Atacama entre los 10.000 y 500 años A.P. Estudios Atacameños 14: 29–41.

80.

Olson

Dinerstein

Wikramanayake

, et al. (2001) Terrestrial ecoregions of the world: A new map of life on earth. Bioscience 51(11): 933–938.

81.

Pascual

Escudero

Pino

, et al. (2019) Cazadores-Recolectores del Holoceno tardío en el curso inferior de la cuenca hidrográfica del río Limarí (30° S): El sitio Tamaya-1. Revista Chilena de Antropología 40: 127–152.

82.

Pearsall

(2008) Plant domestication and the shift to agriculture in the Andes. In: Silverman

Isbell

(eds) The Handbook of South American Archaeology. New York, NY: Springer, pp.105–120.

83.

Price

Bar-Yosef

(2011) The Origins of Agriculture: New Data, new Ideas: An Introduction to Supplement 4. Current Anthropology 52(S4). Chicago, IL: University of Chicago Press, pp.S163–S174.

84.

Prugh

Deguines

Grinath

, et al. (2018) Ecological winners and losers of extreme drought in California. Nature Climate Change 8(9): 819–824.

85.

Pyke

(1984) Optimal foraging theory: A critical review. Annual Review of Ecology and Systematics 15(1): 523–575.

86.

Richerson

Boyd

Bettinger

(2009) Cultural innovations and demographic Change. Human Biology 81(2-3): 211–235.

87.

Riedinger

Steinitz-Kannan

Last

, et al. (2002) A∼ 6100 14 C yr record of El Niño activity from the Galápagos Islands. Journal of Paleolimnology 27: 1–7.

88.

Roberts

Stewart

(2018) Defining the ‘generalist specialist’ niche for Pleistocene Homo sapiens. Nature Human Behaviour 2(8): 542–550.

89.

Root-Bernstein

Ladle

(2019) Ecology of a widespread large omnivore, Homo sapiens, and its impacts on ecosystem processes. Ecology and Evolution 9(19): 10874–10894.

90.

Rothhammer

Fehren-Schmitz

Puddu

, et al. (2017) Mitochondrial DNA haplogroup variation of contemporary mixed South Americans reveals prehistoric displacements linked to archaeologically-derived culture history. American Journal of Human Biology 29(6): e23029.

91.

Salazar

Flores

Borie

, et al. (2020) Economic organization and social dynamics of Middle-Holocene hunter-gatherer-fisher communities on the coast of the Atacama Desert). In: Prieto

Sandweiss

(eds) Maritime Communities of the Ancient Andes. Gainesville, FL: University Press of Florida, pp.74–100.

92.

Santoro

Capriles

Gayo

, et al. (2017) Continuities and discontinuities in the socio-environmental systems of the Atacama Desert during the last 13,000 years. Journal of Anthropological Archaeology 46: 28–39.

93.

Santoro

Castro

Carter

, et al. (2019) Marine Communities in the Atacama Desert. Maritime Communities of the Ancient Andes. Gainesville, FL: University Press of Florida, p.39.

94.

Schiappacasse

Niemeyer

(1984) Descripción y Análisis Interpretativo de Un Sitio Arcaico Temprano En La Quebrada de Camarones. Arica: Ministerio de Educación Pública, Dirección de Bibliotecas, Archivos y Museos.

95.

Smith

Bettinger

Bishop

, et al. (1983) Anthropological applications of optimal foraging theory: A critical review [and comments and reply]. Current Anthropology 24(5): 625–651.

96.

Spencer

(2010) Territorial expansion and primary state formation. Proceedings of the National Academy of Sciences 107(16): 7119–7126.

97.

Standen

Santoro

Arriaza

(2004) Síntesis y propuestas para el período arcaico en la costa del extremo norte de Chile. Chungará (Arica) 36: 201–212.

98.

Surovell

Waguespack

(2009) Human prey choice in the late Pleistocene and its relation to megafaunal extinctions. In: Haynes

(ed.) American Megafaunal Extinctions at the End of the Pleistocene. New York, NY: Springer, pp.77–105.

99.

Tchilinguirian

Morales

(2013) Mid-Holocene paleoenvironments in Northwestern Argentina: Main patterns and discrepancies. Quaternary International 307: 14–23.

100.

Tchilinguirian

Morales

Oxman

, et al. (2014) Early to Middle Holocene transition in the Pastos Chicos record, dry Puna of Argentina. Quaternary International 330: 171–182.

101.

Thomas

(1969) Great basin hunting patterns: A quantitative method for treating faunal remains. American Antiquity 34(4): 392–401.

102.

Tomczak

(2003) Prehistoric diet and socioeconomic relationships within the Osmore valley of southern Peru. Journal of Anthropological Archaeology 22(3): 262–278.

103.

Tomka

(2013) The adoption of the bow and arrow: A model based on experimental performance characteristics. American Antiquity 78(3): 553–569.

104.

Troncoso

Vergara

Pavlovic

, et al. (2016) Dinámica espacial y temporal de las ocupaciones prehispánicas en la cuenca hidrográfica del río Limarí (30 Lat. S.). Chungará (Arica) 48(2): 199–224.

105.

Ugan

(2005) Does size matter? Body size, mass collecting, and their implications for understanding prehistoric foraging behavior. American Antiquity 70(1): 75–89.

106.

Ugan

Bright

Rogers

(2003) When is technology worth the trouble? Journal of Archaeological Science 30(10): 1315–1329.

107.

Urban

Barbieri

(2020) North and South in the ancient Central Andes: Contextualizing the archaeological record with evidence from linguistics and molecular anthropology. Journal of Anthropological Archaeology 60: e101233.

108.

Vilá

(2012) Camélidos Sudamericanos. Buenos Aires: Eudeba.

109.

Williams

Santoro

Smith

, et al. (2008) The impact of ENSO in the Atacama Desert and Australian Arid Zone: Exploratory time-series analysis of archaeological records. Chungara: Revista de Antropología Chilena 40: 245–259.

110.

Wolverton

Dombrosky

Lyman

(2016) Practical significance: Ordinal scale data and effect size in Zooarchaeology. International Journal of Osteoarchaeology 26: 255–265.

111.

Yacobaccio

(1994) Biomasa animal y consumo en el Pleistoceno-Holoceno Surandino. Arqueología 4: 43–71.

112.

Yacobaccio

(2001) La domesticación de camélidos en el Noroeste Argentino. In: Berberian

Nielsen

(eds) Historia Argentina Prehispánica. Córdoba: Editorial Brujas, pp.7–40.

113.

Yacobaccio

(2006) Intensificación económica y complejidad social en cazadores-recolectores surandinos. Boletín de Arqueología PUCP 10: 305–320.

114.

Yacobaccio

(2017) Peopling of the high Andes of northwestern Argentina. Quaternary International 461: 34–40.

115.

Yacobaccio

Madero

Malmierca

, et al. (1997) Caza, domesticación y pastoreo de camélidos en la Puna argentina. Relaciones de la Sociedad Argentina de Antropología 22–23: 389–418.

116.

Yacobaccio

Vilá

(2013) La domesticación de los camélidos andinos como proceso de interacción humana y animal. Intersecciones en Antropología 14: 227–238.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.04 MB

0.02 MB

0.11 MB