Archaeological and Anthropological Sciences (2021) 13: 152 https://doi.org/10.1007/s12520-021-01387-0 ORIGINAL PAPER Fishing during the early human occupations of the Atacama Desert coast: what if we standardize the data? Sandra Rebolledo1,2,3 · Philippe Béarez2 · Débora Zurro3 Received: 14 April 2021 / Accepted: 24 June 2021 © The Author(s) 2021 Abstract The Atacama Desert coast (18–30° S) presents one of the earliest chronologies in the South America region, whose first occupations date from ~ 13,000 cal BP. Since that time, coastal and marine resources have been a common component at sites along the littoral zone. Fish species have been particularly important, as have the fishing technologies developed and used by the coastal communities. However, even though several archaeological sites have been studied, there is no system- atic macro-regional analysis of early fisheries along the Atacama Desert coast. Furthermore, differences in theoretical and methodological approaches, as well as research objectives, hinder comparisons between ichthyoarchaeological assemblages. Here, we present a comparative analysis of the Atacama Desert fish data obtained from publications and gray literature from ten archaeological sites dating from the Terminal Pleistocene to the Early Holocene. Through the standardization of contextual and ichthyoarchaeological information, we compared data using NISP, MNI, and weight to calculate fish density, richness, and ubiquity, in order to identify similarities and differences between assemblages. This exploratory approach aims to contribute to studies of fish consumption in the area, as well as proposing new methodological questions and solutions regarding data heterogeneity in archaeozoology. Keywords Data standardization · Atacama Desert coast · Terminal Pleistocene · Early Holocene · Ichthyoarchaeology Introduction or collected, the type of project for which the data were obtained, or even theoretical-methodological constraints. In order to understand the historical trajectories of a society These factors can be challenging when attempting accurately or region, a meso-scale approach that allows comparison to integrate the information; therefore, it is essential to find between locations, time intervals, and their variability must ways to standardize the data. be considered. This necessitates the use of different sources In the modern digital age, large data sets have provided and, in many cases, dealing with disparities in information new opportunities for standardization, normalization, and as a result of several factors: when the data were produced the open access of archaeological evidence. Despite such achievements, methodological innovations can be incompat- ible with the heterogeneity of data sets (Kansa et al. 2011). This article is part of the Topical Collection on Fishing Over the Millennia This problem is also linked to the dual nature of archaeologi- cal evidence itself, as researchers usually distinguish both its * Sandra Rebolledo cultural and natural aspects. In this respect, faunal remains sandra.rebolledo@e-campus.uab.cat are a paradigmatic example where zoological classification 1 Departament de Prehistòria, Universitat Autònoma de is employed as a tool for understanding cultural practices. Barcelona, Bellaterra, Spain In the 1970s, the International Council of Archaeozoology 2 Archéozoologie, Archéobotanique: Sociétés, Pratiques et (ICAZ) started to discuss appropriate measurements to stand- Environnements (AASPE), CNRS/MNHN, Muséum National ardize and publish data, producing a series of protocols that d’Histoire Naturelle, Paris, France were uploaded in 2009 (Reitz 2009). Despite these efforts, 3 HUMANE – Archaeology and Human Ecology Research some authors still maintain that archaeozoological raw data, Group, Department of Archaeology and Anthropology, analysis, and presentation are unstandardized (Driver 1992, Institució Milà i Fontanals for Research in the Humanities, 2011; Wolverton 2013; Steele 2015). Driver (2011) states that IMF-CSIC, Barcelona, Spain 13 Vol.:(0123456789) 152 Page 2 of 18 Archaeological and Anthropological Sciences (2021) 13: 152 this is due to the lack of rules in archaeozoology and because during the Middle Holocene (Llagostera 1989, 1992; Olguín the classificatory system is based on a fixed zoological tax- et al. 2014, 2015; Salazar et al. 2015; Béarez et al. 2016; onomy. Together with identification and analysis issues, some Rebolledo et al. 2016; among others). authors have suggested that more effective recovery methods In order to evaluate this phenomenon, it is necessary to are needed to achieve better management of the faunal data compare fish remains from different archaeological contexts (Reitz and Wing 1999; Orton 2000; Butler and Campbell during this time. Unfortunately, the heterogeneity of data 2004; Lambrides and Weisler 2016). hinders a regional analysis of fish distribution and exploita- During recent years, archaeozoology has been concerned tion in the area; therefore, it is mandatory to standardize the with using data analysis to respond to large-scale questions. Sev- available data to improve our understanding of fisher-hunter- eral efforts have been undertaken in the study of ancient fisher- gatherer communities and their fishing practices. Here, we ies, using a variety of measures and estimations to compare the considered the contextual information and ichthyoarchaeo- fish assemblages of a large area and/or a long historical sequence logical material published from ten archaeological sites (e.g., Butler and Campbell 2004; Zangrando 2009; Orton et al. from southern Peru and northern Chile. We created a single 2014; McKechnie and Moss 2016; Scartascini 2016). database that allowed us to standardize the data and select The present study is a methodological exercise that useful information for a comparative analysis, applying dif- explores the limitations and potentialities of standardiza- ferent ecological indices commonly used in archaeozool- tion in order to perform accurate comparisons between ogy. Additionally, we summarized both environmental and contexts and faunal assemblages. To do so, we used fish ethological traits in order to identify historical dynamics in data available from the earliest archaeological sites regis- a large-scale data set. In this way, we aim to provide new tered along the Atacama Desert coast (18–30° S), located insights on standardization procedures and comparative between southern Peru and northern Chile, on the Pacific analysis in archaeozoology discussing the quality of fish data coast of South America. The archaeological contexts have produced, as well as contributing to the understanding of fish been dated between ca. 13,000 and 8200 years cal BP, exploitation dynamics during the early human occupation in in what are known as the late Terminal Pleistocene and this macro-region. Early Holocene periods. In southern Peru, the presence of lithic artifacts and faunal remains associated with the exploitation of littoral resources shows an early coastal Study area and environmental settings adaptation (Sandweiss et al. 1989, 1998; Lavallée et al. 1999, 2011; deFrance and Umire 2004; Reitz et al. 2016). The Atacama Desert (18–30° S) is situated along the south- Sandweiss and Rademaker (2011) have defined the phases eastern coast of the Pacific Ocean, in what is now south- of Jaguay (~ 13,000–11,400 years cal BP) and Machas ern Peru and northern Chile (Fig. 1A). Its continental (~ 10,600–8000 cal BP) in Quebrada Jaguay. Further hyperaridity is mostly explained by orographic barriers, south, the communities that inhabited northern Chile were oceanographic characteristics, and climatic conditions. The part of the Chinchorro culture, known for their exploi- coastline is characterized by a narrow littoral, with a small tation of coastal resources, as well as the development freshwater supply strongly influenced by the Pacific Ocean of distinctive mortuary practices (Standen and Santoro (Schulz et al. 2011). Belonging to the Peruvian Province 2004; Arriaza et al. 2008; Santoro et al. 2012, 2020; Carter (Camus 2001), this area is influenced by the homogenizing 2016). Finally, in the southernmost area of the Atacama effect of the Humboldt Current System (HCS) and upwell- Desert coast, several authors have identified the Huen- ings which generate high primary marine productivity. telauquén cultural complex composed of human groups In terms of fish distribution, Pequeño (2000: Fig. 7) distributed along the arid and semi-arid regions of Chile, defines this area as the Distrito Atacameño, i.e., the Ata- which developed a coastal adaptation evidenced by the cama District. The author distinguished four categories of presence of specific ecofacts and artifacts (Jackson et al. coastal fish: the first, which corresponds to coastal families 2011; Salazar et al. 2015, 2018). (including intertidal and subtidal taxa), includes Blenniidae Despite these cultural differences, the exploitation of and Labrisomidae. On the continental shelf, Centrolophidae, marine and coastal resources has been interpreted as a com- Haemulidae, Sciaenidae, Myliobatidae, Merlucciidae, and mon feature of communities living in the area. In this regard, Ophidiidae are found, among others. Pelagic-neritic fish are the capture and consumption of fish resources had socio- most represented by Engraulidae, Scombridae, Carangidae, economic and ecological implications for human groups, as and Clupeidae. Finally, epipelagic fish families are primarily well as land/seascapes. Particularly, the Early Holocene has represented by Triakidae, Carcharhinidae, and Gempylidae. been understood as the beginning of a progressive speciali- Although this area has been defined as a single ecore- zation in the use of coastal and marine resources in some gion (Houston and Hartley 2003; Spalding et al. 2007; Dürr areas of the Atacama Desert coast, which was consolidated et al. 2011), the geomorphological configuration, fog zone 13 Archaeological and Anthropological Sciences (2021) 13: 152 Page 3 of 18 152 Fig. 1  A—Terminal Pleistocene-Early Holocene archaeological sites along the Atacama Desert coast; B—Exorheic coast (Caleta Vítor); C— Arheic coast (San Ramón) distribution, and loma vegetation produce certain latitudinal due to the surge effect. Also, the presence of surficial water heterogeneity (Thiel et al. 2007; Tapia et al. 2018; Bartz flows allows the presence of oasis areas, characterized by et al. 2020), which has implications for access to the sea and diverse fauna and flora (Fig. 1B). The Arheic coast (AC) its resources. The northern portion (18–21° S), or Exorheic in the southern part (21–30° S) has an extremely narrow coast (EC), has an active cliff which is constantly decreasing littoral, with a marine terrace between the coastline and 13 152 Page 4 of 18 Archaeological and Anthropological Sciences (2021) 13: 152 the Cordillera de la Costa (Fig. 1C). Despite the absence which is characterized by a sandy beach (Sandweiss et al. of a freshwater supply, an abundance of coastal resources is 1989; Reitz et al. 2016). Excavations consisted of three observed as a consequence of marine fogs, locally known as columns (A, B, and D) and five pits (C, L, M, J, and camanchacas (Paskoff 2010; Walk et al. 2020). TP-1), with a cultural stratigraphy from the Terminal The Pleistocene–Holocene transition enormously affected Pleistocene until the Middle Holocene. The published the morphology and climate of the Atacama Desert. The faunal data are from Unit C, which is considered repre- increase of continental and oceanic temperatures pro- sentative of Early Holocene occupation (Group 1). duced marine transgressions with different results along • Quebrada Tacahuay (QT): open-air site located in an the coast. Throughout the course of the Early Holocene alluvial fan at 47–56 masl. The site is located 0.3 to (ca. 11,700–8200 year cal BP), the aridization process, an 0.4 km from the present shoreline, consisting of a sandy increase of SST (Sea Surface Temperature) and ENSO (El beach with a rocky promontory (Keefer et al. 1998). Cul- Niño/La Niña Southern Oscillation) cyclical events have tural deposits come from strata units 4, 4c3, 5B, 8, 8A, produced interannual and interdecadal variability, influenc- and 8B, which represents continuous human occupation ing marine primary production (Latorre et al. 2003; Var- from the Terminal Pleistocene until the Middle Holocene gas et al. 2006). It is during the final phase of the Terminal period (deFrance and Umire 2004). Pleistocene that the first archaeological evidence has been • Quebrada Los Burros (QLB): located in the middle part recorded. of the eponymous ravine Quebrada Los Burros, at 170 masl, QLB is an open-air site 2 km from the actual shore- line, characterized by a rocky environment with some Archaeological context(s) sandy beaches (Lavallée et al. 2011). Extensive open-area excavations (150 m ­ 2) revealed six occupational levels The archaeological sites studied here were selected accord- from the Early Holocene to the Middle Holocene epoch ing to the following criteria: data availability (in terms of (Lavallée and Julien 2012). Early Holocene sequences contextual information, fish data, and radiocarbon dates), were found in levels N4, N5, N6, and N7 (Fontugne time span, and geographical location. Four of the ten sites 2012). are located in southern Peru (Quebrada Jaguay, Ring Site, • Caleta Vítor 3 (CV3): an open-air site located at the Quebrada Tacahuay, and Quebrada Los Burros), while six mouth of Quebrada Vitor at ~ 25 masl, in front of a sandy are from northern Chile (Caleta Vítor 3, La Chimba 13, beach. Its cultural sequence starts in the Early Holocene Paposo Norte 9, Alero 224-A, Alero 226–5, and Morro period with occupations continuing until recent times. Colorado) (Table 1). Even though we recognize the Atacama The site has been divided into seven areas (CV1 to CV7), Desert coast as a whole macro-region, we also recognize with the Early Holocene occupation starting in the deep- latitudinal differences associated with cultural differences est deposit of CV3 (Carter 2016), a shell midden 0.3– (Llagostera 2005; Santoro et al. 2020). For this reason, we 0.4 km from the present shoreline. Ichthyoarchaeologi- have classified the archaeological sites according to the two cal data came from the excavations of two 0.5 × 0.25 cm types of coast mentioned above: the Exorheic coast to the profiles (CV3-Perfil 1 and CV3-Perfil 2) (Rebolledo et al. north and the Arheic coast in the south. 2021). Exorheic coast archaeological sites Arheic coast archaeological sites • Quebrada Jaguay (QJ-280): an open-air site with shell • La Chimba 13 (LC13): located at the mouth of Quebrada deposits located on an alluvial terrace 40 m above sea Las Conchas, La Chimba 13 is an open-air site at ~ 260 level (masl). It is located 2 km from the modern shore- masl. It is located 4 km from the current shoreline, which line, which, in the vicinity, consists of sandy beaches is characterized by a sandy beach. The site is composed with a rocky promontory (Sandweiss et al. 1998; Reitz of two depositional events with Early Holocene occupa- et al. 2016). Excavations were classified into four sectors tions (Llagostera et al. 1997). The ichthyoarchaeological (I, II, III, IV) evidencing a human sequence that starts in data presented here comes from two 2 × 2 m units: N23/ the Terminal Pleistocene and ends in the Early Holocene W07 and N54/W05 (Llagostera 1979; Llagostera et al. period. Additionally, QJ-280 Early Holocene occupations 1997). were subdivided in two subperiods: EHI and EHII (Sand- • Paposo Norte 9 (PN-9): a rock shelter located in a weiss et al. 1998). marine terrace at ~ 25 masl and 0.2 km from the cur- • Ring Site (RS): open-air site consisting of an extensive rent shoreline, which is characterized by a sandy-rocky shell midden deposit situated on a marine terrace 50 beach (Casteletti 2007). Human occupations at this masl. It is located 0.75 km from the present shoreline, site are from the Early and Late Holocene (Casteletti 13 Archaeological and Anthropological Sciences (2021) 13: 152 Page 5 of 18 152 Table 1  Archaeological sites from: TP Terminal Pleistocene, EH Early Holocene, MH Middle Holocene, LT Late Holocene Area Site Location Archaeological information Geogra- masl Distance Type Type of site Units/levels used Chronology phical actual of here location shore- beach (Screen size) TP EH MH LH line(km) QJ- Alluvial 40 2 Sandy/ Open air Sector I, II, III and 280 terrace Rocky Shell-matrix IV (1.7 mm) RS Marine 50 0.75 Sandy Open air 3 columns; 5 pits terrace Shell-matrix (6.35 mm) QT Alluvial fan 47-56 0.3-0.4 Sandy/ Open air Profile 3C: Unit 5 EC Rocky Soil-matrix Bloque 3: Units 5B; 8 (3.175 mm) QLB Middle part 170 2 Sandy/ Open air N4; N5; N6 of a ravine Rocky Shell-matrix (3.175 mm) CV3 Mouth of a 25 0.3-0.4 Sandy Open air CV3-P1; CV3-P2 ravine Shell-matrix (2, 1, <1 mm) LC13 Mouth of a 260 4 Sandy Open air N23/W07; ravine Shell-matrix N54/W05 (2 occupational levels) (n/i) PN-9 Marine 25 2 Sandy/ Rock shelter 3 units; 2 pits terrace Rocky Soil-matrix (2-3 mm) AC 224-A Marine 20 0.6 Rocky Rock shelter 3 units; 2 pits terrace Soil-matrix (2-3 mm) 226-5 Slope next 100 0.3 Rocky Rock shelter Units (A; B; C; D) to mouth of Soil-matrix (2-3 mm) a ravine MC Marine 20 0 Sandy Open air Units (2; 3; 4A; terrace Shell-matrix 4B; 5) (2-3 mm) 2007; Salazar et al. 2015, 2018). Fish remains came • Alero 224-A: this site consists of a rock shelter located on from three 1 × 1 m units (Unidad 1, 2, and 3) and two a littoral terrace at ~ 20 masl and 0.6 km from the present 0.5 × 0.5 m pits (Pozo 1 and 2). coastline, characterized by a rocky beach (Salazar et al. 2018). Human occupations date from the Early and Late 13 152 Page 6 of 18 Archaeological and Anthropological Sciences (2021) 13: 152 Holocene. Ichthyoarchaeological data came from three associated with archaeozoological indices and the etho- units (Unidad 5, 6, and 7) and two pits (Pozo 1 and 2). logical information related to each species. We have also • Alero 226–5: this rock shelter is situated on a slope next considered NISP (Number of Identified Specimens) and to the mouth of Quebrada Cascabeles, at ~ 100 masl. This MNI (Minimum Number of Individuals) (Casteel 1974; site is less than 0.3 km from the actual shoreline, char- Grayson 1984). Almost all sites presented both indices, acterized by a rocky beach (Salazar et al. 2018). Human with the exception of La Chimba 13, where fish infor- occupations are from the Early and Late Holocene mation is restricted to otoliths and vertebrae. However, (Casteletti 2007). The fish remains studied here came we have still decided to integrate this site and interpret from units A, B, C, and D (Casteletti 2007; Salazar et al. it as a special case. As excavation areas differ between 2018). sites, we have combined the NISP and MNI per excavated • Morro Colorado (MC): an open-air site located on a 20 m volume in order to calculate the density of fish remains rocky promontory, next to a sandy beach. This site pre- per ­m3 (Jerardino 1995; Zangrando 2009), and we have sents a continuous cultural sequence from the end of the also included the fish weight in those cases where infor- Early Holocene until the Late Holocene (Andrade and mation was available. Moreover, we have incorporated Salazar 2011; Salazar et al. 2018). The ichthyoarchaeo- the richness and ubiquity index (Ui) in order to evaluate logical data came from five units (Unidad 2, 3, 4A, 4B, the presence of fish species from a regional perspective and 5) and a column (Rebolledo 2017). through time. The richness corresponds to the number of taxa counted at an archaeological site (Grayson 1984). As this index is directly affected by the size of excavation Materials and methods units (Grayson 1984; Cruz-Uribe 1988), it is mandatory to be cautious when analyzing the tendencies of fish assem- We generated a new database to include fish information blages between archaeological sites. For this reason, we from publications and gray literature available from 1977 to also have compared richness with NISP and MNI. 2021 (Llagostera 1977, 1979; Llagostera et al. 1997; Sand- The ubiquity index (Popper 1988), widely used in weiss et al. 1989, 1998; deFrance and Umire 2004; Caste- archaeobotanical studies, measures the presence or letti 2007; Béarez 2012; Reitz et al. 2016; Rebolledo 2017; absence of a taxon in the total number of analyzed sam- Salazar et al. 2018; Rebolledo et al. 2021). All data were ples. This index is calculated as the percentage of samples integrated into a single database based on three types of in which a taxon is found, with values ranging from 0 information: regional, archaeological, and ichthyoarchaeo- (taxon not observed) to 1.00 (taxon observed in every sam- logical. Contextual regional information includes the spatial ple or feature) (McKechnie and Moss 2016; Diehl 2017). position with regards to the two types of coast: Exorheic Following Diehl (2017), the ubiquity index formula is: and Arheic. The archaeological information comprises the Utaxon = Ntaxon ∕Ntotal type of site and chronology, integrating radiocarbon dates available from published papers. Radiocarbon dates were In this case, Utaxon corresponds to the measured ubiquity calibrated using ShCal20 and Marine20 curves in OxCal of the fish taxon, Ntaxon is the number of sites where the (Heaton et al. 2020; Hogg et al. 2013). It is important to taxon appeared, and Ntotal corresponds to the total number note that the radiocarbon dates were selected according to of archaeological sites. a margin of error ≤ 100 years (Table 2). Radiocarbon dates We have also included the habitat and ecology of the from La Chimba 13 were not considered since the margin of identified fish species. Four types of littoral allow the clas- error exceeded 100 years; therefore, we have considered this sification of fish taxa: (1) coastal-sandy; (2) coastal-rocky; archaeological site to be part of the Early Holocene. (3) neritic-pelagic; and (4) pelagic-oceanic. These catego- Additionally, archaeological information included strati- ries have been linked to the fishing techniques commonly graphical data from the sites. In this regard, layers and/or used for each species. levels have been considered, as well as the excavated area Results have been divided into two sections: archaeo- and cubic meter of the sample matrix ­(m3). The latter has logical and fish standardization. The former is essential for been considered when evaluating the compatibility between the analytical procedures required for the second, since the fieldwork methods and representativity regarding site size. “degree” of standardization of archaeological data permits Concerning the ichthyoarchaeological information, comparative analysis between ichthyoarchaeological assem- all scientific names have been standardized and updated blages. Data integration has been summarized following spa- following Fricke et al. (2020). In those cases where taxa tial and chronological scales in order to identify differences could not be identified, we have labeled them as “indeter- and similarities between the types of coasts, as well as vari- minate” and have not considered them during the analysis. ations and continuities throughout the sequence. Additionally, we have used the available quantitative data 13 Archaeological and Anthropological Sciences (2021) 13: 152 Page 7 of 18 152 Table 2  Radiocarbon dates (100 years) per archaeological site. Radiocarbon dates from La Chimba 13 were not considered for calibration Site Lab. no Material 14C yr. BP Reference QJ-280 1959 Charcoal 7690 ± 100 Sandweiss et al. 1998 1958 Charcoal 7620 ± 100 Sandweiss et al. 1998 RS SI-6784 Shell 7675 ± 60 Sandweiss et al. 1989 PITT-0142 Charcoal 7415 ± 65 Sandweiss et al. 1989 QT 108692AMS–97 Charcoal 10,750 ± 80 deFrance and Umire 2004 172645AMS–01 Charcoal 10,690 ± 60 deFrance and Umire 2004 160706AMS–01 Charcoal 10,660 ± 80 deFrance and Umire 2004 122821AMS–98 Charcoal 10,510 ± 50 deFrance and Umire 2004 160707AMS–01 Charcoal 10,050 ± 90 deFrance and Umire 2004 172615AMS–01 Charcoal 9010 ± 40 deFrance and Umire 2004 109354Conv–97 Charcoal 7990 ± 80 deFrance and Umire 2004 QLB SacA 7591 Charcoal 7390 ± 35 Fontugne 2012 GifA 100,342 Charcoal 7880 ± 90 Fontugne 2012 GifA 100,345 Charcoal 7980 ± 90 Fontugne 2012 SacA 7595 Charcoal 8235 ± 35 Fontugne 2012 GifA 102,444 Charcoal 8260 ± 80 Fontugne 2012 Gif-11738 Charcoal 7280 ± 60 Fontugne 2012 GifA 102,445 Charcoal 8120 ± 90 Fontugne 2012 CV3 ANU31016 Charcoal 8420 ± 40 Santoro et al. 2017 (from Carter 2016) UGAMS 10,516 Wood 8260 ± 30 Santoro et al. 2017 (from Roberts et al. 2013) ANU31014 Charcoal 8075 ± 40 Santoro et al. 2017 (from Carter 2016) LC13 P-2702 Charcoal 9400 ± 160* Llagostera 1977 P-2702 Charcoal 9680 ± 160* Llagostera 1977 PN-9 D-AMS 008,355 Charcoal 9813 ± 37 Salazar et al. 2018 (from Salazar et al. 2015) D-AMS 011,300 Charcoal 9489 ± 37 Salazar et al. 2018 224-A UGAMS 13,089 Shell 10,530 ± 30 Salazar et al. 2018 (from Salazar et al. 2013) D-AMS 018,343 Charcoal 10,441 ± 44 Salazar et al. 2018 D-AMS 018,344 Charcoal 9955 ± 33 Salazar et al. 2018 226–5 Beta-190672 Shell 10,290 ± 60 Salazar et al. 2018 (from Castelleti 2007) Beta-190671 Shell 10,040 ± 60 Salazar et al. 2018 (from Castelleti 2007) MC UGAMS 13,094 Charcoal 7470 ± 30 Salazar et al. 2018 (from Salazar et al. 2015) Results presents radiocarbon dates from Quebrada Tacahuay and Paposo Norte 9, there is no ichthyoarchaeological informa- Archaeological standardization: calibration tion for this time period. In fact, available fish data for the of the radiocarbon dates and excavated volume Exorheic coast come from the Quebrada Jaguay and La Chimba 13 sites, although their radiocarbon dates could To standardize the information and identify similarities not be calibrated (Llagostera 1977; Reitz et al. 2016). and differences with a diachronic perspective, Fig. 2 shows During the 10,000–9000 years cal BP interval, we found the radiocarbon dates according to ~ 1000 year time slots, only three sites from the Exorheic coast: Quebrada Taca- starting at the Final Pleistocene–Early Holocene transi- huay, Quebrada Los Burros, and Caleta Vítor 3. Lastly, tion until the Early–Middle Holocene transition. Accord- the interval 9000–8000 years cal BP is represented by the ing to the radiocarbon dates, two sites were occupied dur- Quebrada Jaguay, Ring Site, Quebrada Tacahuay (with- ing 13,000 and 12,000 years cal BP (Quebrada Tacahuay out fish data), Quebrada Los Burros, Caleta Vítor 3, and and Alero 224-A). Following Reitz et al. (2016), we also Morro Colorado sites. include Quebrada Jaguay in this interval. Later, the period Regarding the distribution of radiocarbon dates, some between 12,000 and 11,000 years cal BP is only repre- tendencies were observed: a major concentration of remains sented by Arheic coast sites: Paposo Norte 9, Alero 224- from 9000 to 8000 years cal BP, while the period between A, and Alero 226–5. Even though the following interval 11,000 and 10,000 presented the fewest. 13 152 Page 8 of 18 Archaeological and Anthropological Sciences (2021) 13: 152 Fig. 2  Calibrated radiocarbon dates from Terminal Pleistocene and Early Holocene occupations: A—terrestrial samples; B—marine samples 13 Archaeological and Anthropological Sciences (2021) 13: 152 Page 9 of 18 152 The available data from the selected archaeological sites included in the taxonomic classification, such as billfishes revealed important disparities. The Quebrada Jaguay, Ring and Chondrichthyans. Site, and Quebrada Tacahuay sites provided no information The 63 fish taxa recognized are differentially distributed on excavated volumes, but the fieldwork data from the Que- in each archaeological context, with a higher number of brada Los Burros site enabled an estimation to be made. species at the Morro Colorado site (N = 34), and fewer This was also the case for the Caleta Vítor 3, La Chimba (N = 5) at Quebrada Tacahuay and Alero 226–5. Except for 13, Paposo Norte 9, Alero 224-A, Alero 226–5, and Morro Alero 226–5, the Arheic Coast presents the highest num- Colorado sites (Table 3). However, given the problems ber of taxa at its sites. The richness of Arheic coast sites associated with the ichthyoarchaeological study of the La is similar to the Quebrada Los Burros site, the only one Chimba 13 site, i.e., that only otoliths were counted (see the with more than 20 species in the northern area (Fig. 4). “Materials and methods” section), we decided not to include When comparing richness with NISP and MNI, there are this site in the fish density estimate. some similarities between them (Fig. 5): the higher the NISP, the higher the richness. Quebrada Los Burros and Fish standardization: fish density, richness, Morro Colorado present the highest values in these three and ubiquitous index indices, with a richness of over 25 species. However, although Alero 224-A does not have a high NISP or MNI, In total, NISP amounted to 14,669 and MNI resulted in 3752 it shows a high richness (N = 31). The higher richness in individuals (Appendix 1). Most of the NISP came from the 224-A is mainly due to the presence of a greater number Exorheic Coast (53.22%) while MNI was greater along the of coastal-rocky taxa (Appendix 1), possibly reflecting a Arheic coast (57.19%). Although the excavated volume is more intensive exploitation of this habitat and a reduced not available for all sites, we decided to perform an estima- exploitation of the pelagic domain. On the other hand, tion of fish density based on the analysis. Figure 3A shows those archaeological sites with a lower richness present the Alero 226–5 site with the greatest excavated volume, lower NISP and MNI values, such as Quebrada Tacahuay followed by Alero 224-A, and Quebrada Los Burros, while and Alero 226–5, both with 5 species. The exception is Caleta Vítor 3 presents the lowest value. However, these ten- Caleta Vítor 3 (N = 10), with a richness close to that of dencies change when considering fish density: Morro Colo- Quebrada Jaguay (N = 12). rado presents the highest value, followed by Quebrada Los Since the fish density estimation could not be carried out Burros, and Alero 224-A (Fig. 3B). Additionally, the total for all archaeological sites, the ubiquity index (Ui) was used fish weight at Morro Colorado and Quebrada Los Burros is for comparison. The most ubiquitous taxa are sciaenids with considerably higher than the other sites (Fig. 3C). Callaus deliciosa and Cilus gilberti, then the Carangidae The taxonomic compositions of all archaeological sites family, which is almost exclusively represented by Trachurus from the Early Holocene in the Atacama Desert coast are murphyi. The Latridae (Chirodactylus variegatus), Haemuli- represented by 63 fish taxa. Depending on each archaeo- dae (Anisotremus scapularis and Isacia conceptionis), and logical site, taxa details are described according to family, Serranidae (Paralabrax humeralis) families are also well genus, and/or species (Appendix 1). It is important to men- represented (Fig. 6). As shown in Table 4, there is an inverse tion that even though a standardization of scientific names relationship between the number of taxa and the ubiquity was made, there are still some specimens that could not be index. Table 3  Quantitative information available on contextual and ichthyoarchaeological data Area Site m3 NISP MNI NISP/m3 MNI/m3 EC QJ-280 - 1091 417 - - RS - 576 179 - - QT - 1538* 39 - - QLB 3.08 4511 843 1465 274 CV3 0.939 110 57 117 61 AC LC13 5.154 - 545** - - PN-9 1.5135 212 126 140 83 224-A 3.535 688 257 195 73 226–5 4.2 149 53 35 13 MC 2.1284 4023 1064 1890 500 * Includes total number of fragments(not necessary with taxonomic identification). **Considers only otoliths and vertebrae 13 152 Page 10 of 18 Archaeological and Anthropological Sciences (2021) 13: 152 Fig. 3  Excavated volume and fish density considering B— NISP/m3 and MNI/m3, and C—g/m3 per archaeological site With only the exception of billfishes and K. pelamis, chronological sequence: on the coast, rocky littoral areas which thrive in pelagic-oceanic waters, the fish taxa present a decrease from 13,000 to 9000 years cal BP, and identified in the total assemblage inhabit coastal and an increase around 8000 years cal BP. Regarding sandy neritic-pelagic zones. Between regions, taxa richness areas, the highest values are found during 11,000–10,000, does not show important differences: the Exorheic coast and 9000–80,000 years cal BP. It is interesting to note the is mostly represented by neritic-pelagic and coastal- progressive augmentation of a variety of species from the sandy taxa, while the Arheic coast is mostly represented neritic-pelagic area, while the oceanic pelagic species are by neritic-pelagic fish (Appendix 1). Major differences found during 12,000–11,000 and 9000–8000 years cal BP appear when considering habitat richness through the (Fig. 7). 13 Archaeological and Anthropological Sciences (2021) 13: 152 Page 11 of 18 152 coastal-sandy zone, with the largest NISP corresponding to C. deliciosa. In the case of La Chimba 13, the most repre- sented fishing areas are the rocky coast and the sandy coast, with a predominance of A. scapularis and Cynoscion analis, respectively. In terms of fishing techniques, fish taxa suggest the use of hooks and lines, nets and spears. During the 10,000–9000 years cal BP occupations, even though there is a higher richness of neritic-pelagic taxa in Quebrada Tacahuay, Quebrada Los Burros, and Caleta Vítor 3, there is an NISP predominance of coastal-sandy species, especially with C. deliciosa. As in the previous period, pos- sible fishing techniques used correspond to hooks and lines, nets, and spears. Fig. 4  Richness according to the region and archaeological site Finally, during 9000–8000 years cal BP, Quebrada Jag- uay presents a major proportion of coastal-sandy fish spe- The highest richness of coastal-rocky fish is observed at cies, and Quebrada Los Burros has an equal amount of the Alero 224-A site between 13,000 and 12,000 years cal coastal-sandy and neritic-pelagic taxa, with an NISP pre- BP, with a NISP predominance of Scartichthys spp. At the dominance of the former due to the presence of C. deliciosa. Exorheic sites, Quebrada Jaguay presents a major richness of The neritic-pelagic zone is also most represented at Caleta coastal-sandy species, while neritic-pelagic fish are mostly Vítor 3 (with E. ringens), and particularly in Morro Colo- represented in Quebrada Tacahuay, with a major NISP rado, where an absolute NISP predominance of T. murphyi abundance of Engraulidae. The most commonly used fish- is observed. Finally, the coastal-rocky area is well repre- ing techniques for this type of species are hook and line and sented at the Ring Site with C. variegatus. In terms of cap- fishing nets; to a lesser extent, the presence of some species ture techniques, there was probably a preference for fishing that inhabit the intertidal zone suggests the possible use of nets, but hooks and lines were also employed. The presence scoop nets and spears. of coastal-rocky species suggests the use of scoop nets, and Between 12,000 and 11,000 years cal BP, a similar rep- the presence of large pelagic-oceanic taxa suggests the addi- resentation of coastal-rocky fish is observed at the Paposo tional use of harpoons. Norte 9, Alero 224-A, and Alero 226–5 sites. In the first two sites, the highest NISP values corresponds to Scartichthys spp., while Semicossyphus darwini is the most represented Discussion species in the latter. In addition, it is important to note that Paposo Norte 9 and Alero 226–5 also have a good represen- Data standardization as an archaeological problem tation of coastal-sandy taxa (sciaenids). During this period, species caught mainly by hook and line are still predomi- Data standardization arises as one of the main problems nant, while those evidencing scoop net and spear fishing are when developing a large-scale study, whether it is a regional in the minority. approach or a historical sequence. In the case of the Final Subsequently, from 11,000 to 10,000 years cal BP, Pleistocene–Early Holocene archaeological sites along Quebrada Jaguay only presents three species from the the Atacama Desert coast, ichthyoarchaeological informa- tion needs to be evaluated using contextual data. This is especially important not only to compare assemblages, but also to evaluate the possibility of being able to make such comparisons. The calibration of radiocarbon dates allows for the posi- tioning of each archaeological site and its occupations in a defined range of time. In our case study, we positioned archaeological sites within 1000-year time slots, considering this division as a tool to order the data but not as fixed cat- egories of human occupations. The time range begins around 13,000 cal BP, and continues to approximately 8200 cal BP. This wide time interval might hinder a detailed approach, Fig. 5  Distribution of richness, NISP, and MNI according to archaeo- but it does give us a general picture of the human communi- logical sites ties’ dynamics and their fishing practices. Additionally, it 13 152 Page 12 of 18 Archaeological and Anthropological Sciences (2021) 13: 152 Fig. 6  The three most ubiq- uitous species throughout the sequence Table 4  Ubiquity index (Ui) UI N Taxa and number of taxa (N) 0.9 2 Callaus deliciosa; Cilus gilberti 0.8 1 Trachurus murphyi 0.7 3 Anisotremus scapularis; Chirodactylus variegatus; Isacia conceptionis; Sciaenidae 0.6 2 Paralabrax humeralis 0.5 4 Semicossyphus darwini; Pinguipes chilensis; Cheilotrema fasciatum; Sarda chiliensis Engraulis ringens; Girella laevifrons; Labrisomidae; Labrisomus philippii; Paralichthys 0.4 6 adspersus; Sebastes oculatus Galeichthys peruvianus; Scarthichtys spp.; Clupeidae; Graus nigra; Mugil sp.; Cyonoscion 0.3 9 analis; Acanthistius pictus; Hemilutjanus macrophthalmos; Serranidae Aplodactylus punctatus; Scartichthys gigas; Scartichthys viridis; Seriolella violacea; Ethmidium maculatum; Engraulidae; Thyrsites atun; Sicyases sanguineus; Chondrichthyes; Auchenionchus microcirrhis; Auchenionchus spp.; Genypterus spp.; Paralichthys sp.; Chromis intercrusma; 0.2 16 Paralonchurus peruanus; Katsuwonus pelamis Odontesthes regia; Bovichtus chilensis; Carangidae; Seriola lalandi; Carcharhinus sp.; Sardinops sagax; Clupeiformes; Haemulidae; Billfish; Kyphosidae; Labridae; Auchenionchus variolosus; Merluccius gayi; Prolatilus jugularis; Micropogonias altipinnis; Robaloscion wieneri; Scomber 0.1 21 japonicus; Scombridae; Scorpaena histrio; Mustelus mento poses interesting questions concerning the prioritization of In addition to the information on radiocarbon dating, the dating the “earliest” occupations over others, a phenomenon excavation data allows us to evaluate other aspects relating already observed in other areas (e.g., Campbell and Qui- to the ichthyoarchaeological recovery methods. This stand- roz 2015). However, there is also a concentration of dates ardization exercise can open new debates about the possibil- between 9000 and 8000 cal BP which could elucidate other ity of making comparisons in the Atacama Desert coast, and factors affecting the presence of human occupation, as well the need for collaborative work between different research as demographic variations, or differential use of space and teams. This will undoubtedly lead to a broader discussion archaeological site functions. In any case, these proposals— about how to work with this kind of data, as well as its asso- previously discussed by several authors—must be analyzed ciation with other resources such as invertebrates, seabirds, with more information. mammals, and artifacts (e.g., fishing gear). 13 Archaeological and Anthropological Sciences (2021) 13: 152 Page 13 of 18 152 Fig. 7  Fishing strategies accord- ing to taxa presence through time The problem of data standardization is not new in archae- their absolute frequencies. To overcome those difficulties, ozoology. In this regard, O’Connor (1985: 29) indicates that we have emphasized the distribution of taxa and fish habi- “the quantification problem might be better approached by tat representation considering latitudinal and chronological considering frequency rather than abundance.” Following differences. this author, we chose to focus on the frequency of taxa, Disparities in the contextual information available have using fish density, richness, and the ubiquity index depend- hindered a regional comparison using fish density. For those ing on the variables that we were comparing. Regarding fish sites from which data were compared, it must also be noted density, this presents problems due to disparities of contex- that the excavated volume was strongly influenced by the tual information, which hinder the estimation of excavated site’s function and formation processes. In this regard, some volume. Even though this measure is normally recorded in authors have pointed out the importance of depositional pro- fieldwork diaries, the information is often missing when cesses, as well as the type of matrix for each context, which results are published. This could be a useful tool, not only affect the preservation of fish remains through time (Jerar- to standardize the archaeozoological information but also to dino 1995; Zangrando et al. 2021). In our case study, open deepen the discussion about defining shell middens, espe- air sites with a predominant shell matrix presented higher cially in the case of coastal contexts. fish density values than rock shelter sites with a soil matrix. Richness gives us a first approach to characterize the For this reason, interpretations concerning this type of index total assemblage. It allows us to identify common elements should be treated with caution. among the assemblages, and the standardization of species Even though the fish density estimation could not be names allows us to compare their presence or absence in a applied to all sites, tendencies associated with the exca- particular sample. This first stage was essential for applying vated volume show interesting differences between them. the ubiquity index and to identify the most ubiquitous spe- Considering the total fish weight, Morro Colorado presents cies in the total assemblage, but also to analyze data with the highest value, followed by Quebrada Los Burros, both spatial and temporal perspectives. chronologically located between 10,000 and 8000 years cal Despite the methodological difficulties we had to face BP. In the case of Caleta Vítor 3, according to Table 3, this during this analysis, it is always convenient to work with site presents the smallest screen size of all the assemblages, different types of indices (cf. Zohar et al. 2018). In addition which could affect the results associated with an abundance to cultural, natural, and/or taphonomic issues, other vari- of fish remains and taxonomic representation, especially for ables such as fieldwork procedures, laboratory analysis and small pelagic fish (Rebolledo et al. 2021). Beyond spatial the way the results are published, also have an impact on differentiation between coasts, it seems that temporal vari- archaeological interpretations at local and regional scales. ations will be more relevant to understand fish distribution in archaeological sites. Fish standardization in the Atacama Desert coast Along with fish density, other indices focused on the taxonomic distribution were developed. The 63 fish taxa Previous analyses have determined the way we standard- evidence a greater richness within the total assemblage for ize ichthyoarchaeological materials during the early coastal a region with temperate waters, even though the species are occupations registered along the Atacama Desert coast. differentially distributed. Here, it is important to note the NISP and MNI have been considered, but the disparities bias of this index regarding the NISP and MNI values at of fish data hinder a comparison between samples using each archaeological site, and to examine the trends relating 13 152 Page 14 of 18 Archaeological and Anthropological Sciences (2021) 13: 152 to each type of coast. In terms of richness, no major dif- where the coastline is sandy and similarly exploited between ferences were found between the two coasts, although the 12,000 and 9000 years cal BP. values appeared to be somewhat higher and maximum bio- Concerning differences between the assemblages, there diversity reached 34 species at one site (MC) on the Arheic is little variation on the type of littoral represented along coast. This shows the ecological homogeneity of the two the Atacama Desert coast. Even though in both regions, coastal areas and supports the choice to study the Atacama the coastal species are well represented, there is a high Desert coast as a whole. Despite the consistency between number of rocky shore fish on the Arheic coast. This is the three indices, there are some interesting exceptions: consistent with the geographical features mentioned above, Caleta Vítor 3 on the Exorheic coast and Alero 224-A on where the lithological composition suggests a northern the Arheic coast. Caleta Vítor 3 presents a high richness coast with sandy substrate and a southern coast with a value despite the low NISP and MNI, a phenomenon that prevalence of rocky shores (Tapia et al. 2018). As a con- could be partially explained by the recovery methods already sequence, a variation of species dominance is observed: mentioned, which allowed the identification of small fish sciaenids are more present on the Exorheic coast, with the species (Rebolledo et al. 2021). For Alero 224-A, a larger presence of C. deliciosa and C. gilberti. In the south, the excavation volume could explain the identification of a wide Arheic coast is also dominated by sciaenids, but other fam- variety of species, but this is not completely consistent with ilies such as Haemulidae, Blenniidae, Paralichthyidae, and the NISP and MNI values. It may be that the sample size Carangidae are well represented. It is important to note has allowed the necessary size to be reached and exceeded that Carangidae is the most represented neritic-pelagic in order to obtain the maximum diversity represented at this taxa, and its importance is primarily explained by Morro site. With this in mind, a minimum of 700 NISP could be Colorado. The higher number of taxa, combined with the recommended for the analysis of fish remains in the area, lowest diversity value in Morro Colorado, suggests that to avoid the risk of being affected by the sample size bias. the main differences in the sample could be related to this The ubiquity index has made it possible to establish site. In this regard, the predominance of Carangidae—and comparisons through the presence/absence of the species, specifically T. murphyi—in Morro Colorado is one of the and the recurrence of certain fish in the total archaeological major variations in taxonomic distribution. The high per- record. The three most represented species (Ui > 0.8) reflect centage of this taxon at this site also affects the picture the relevance of Sciaenidae and Carangidae in the data and of the most exploited habitat during 9000–8000 years cal archaeological interpretation. It is worth asking whether this BP, when neritic-pelagic fish overtook nearshore fish trend is related to the methods of recovery and analytical from previous periods. In comparison with the Exorheic procedures of bones and otoliths, pre- and post-depositional coast, Morro Colorado is contemporaneous with the lat- taphonomic processes, fishing techniques, abundance, and est occupations of Quebrada Los Burros and Caleta Vítor variability of the species in the sea, human food prefer- 3, two sites that present a considerable percentage of T. ences, or a combination of these factors. Although we are murphyi. This species is a particularly interesting case as not considering the taphonomic processes of fish remains it is already evidenced at Morro Colorado and plays a key identified in this analysis, the question about the over and/or role during the Middle Holocene, as well as at Quebrada underrepresentation of certain species should be posed. For Los Burros on the Exorheic coast and Zapatero on the instance, the available data does not permit the comparison Arheic coast (Béarez 2012; Rebolledo et al. 2016; Salazar of this kind of information between archaeological sites, but et al. 2018). it is possible to compare the Ui tendencies with the Sur- Considering that the biogeographical background is simi- vival and Recovery Index (SRI) proposed by Falabella et al. lar in the Arheic coastal area, divergences between Morro (1994). The species identified in our case study are part of Colorado and the other sites could be explained by its posi- SRI groups III (recovered only under special excavation pro- tion in the occupational sequence. On a temporal scale, cedures) to VI (best possibilities to recover). Furthermore, Morro Colorado dates fall in the transition from the Early S. violacea, T. atun, G. nigra, and Genypterus spp. stand to the Middle Holocene. At this time, an increase in SST out as the species with lower Ui values, although they are has been registered by Flores and Broitman (2021) in Tal- well represented in terms of SRI. Regarding those species tal (Arheic coast), which may have had a local impact on with the highest Ui values, it is interesting to note that the fish harvesting at this time. How this may have affected the sciaenids dominate all the cultural sequences. This could be northern area of the Atacama Desert coast is still an open partially explained by the good bone and otolith preserva- question that needs to be answered in light of further palaeo- tion of C. deliciosa in deposits (Béarez 2012), its abundance ecological and archaeozoological studies. throughout the year, and/or cultural preferences. This could The comparative analysis has been adjusted to consider also be related to the location of human settlements in areas the characteristics of fish data and, as we have mentioned on several occasions, limitations in the data have not led us 13 Archaeological and Anthropological Sciences (2021) 13: 152 Page 15 of 18 152 to the detailed insights for which we were aiming. Even so, information about palaeoclimatic features and archaeologi- from the results obtained, it is possible to establish some cal contexts on both coasts, in order to understand fish har- guidelines regarding the fish capture strategies, as well as to vesting during different time–space intervals. In addition, propose future questions for regional ichthyoarchaeological we need to increase our ichthyoarchaeological sample and studies. In this respect, the high variety of species can be improve available methodological tools to make better com- understood as an “opportunistic catch” strategy, suggesting parisons between assemblages. Further studies must include the use of coastal and maritime zones and their resources. other lines of evidence, especially concerning marine (e.g., The wide variety of species and their ethological traits mollusks) and terrestrial resources. This could help explain suggest the use of different fishing strategies. Among these, how the trophic web has evolved from the earliest interac- the possible use of hooks and lines and nets seems to be pre- tions with human communities throughout the chronologi- dominant over other techniques such as scoop nets, spears, cal sequence, and also how different ecological conditions or harpoons. Scoop nets and spears may have been used interact with cultural differences. to catch fish in intertidal pools or along the beach, while Finally, the standardization exercise presented here con- harpoons would have been used from a boat to catch large stitutes a synthesis generated from a compilation of ichthyo- pelagic fish as they struggle and need to be tired before they archaeological data corresponding to the end of the Pleisto- can be subdued. Concerning the most ubiquitous species, cene and the beginning of the Holocene, using information sciaenids and carangids are well represented throughout the from published papers and gray literature. This exercise is sequence. Fish from both families live in schools, which also, therefore, a reflection on the way we construct data facilitates their mass capture and the appearance of other and approach archaeological knowledge. Consequently, this species as by-catch. As other authors have inferred, this information should be made available to a wider and more could reflect the use of nets, again evidencing the relevance public audience, in order to build new bridges from which of this fishing gear (Béarez, 2012; Disspain et al. 2016; Mar- we can approach the past, where knowledge is constructed tens and Cameron 2019). including all perspectives, allowing us to create new ones. The spatial distribution does not seem to be as relevant as variations through the chronological sequence. At the Supplementary Information The online version contains supplemen- tary material available at https://d​ oi.o​ rg/1​ 0.1​ 007/s​ 12520-0​ 21-0​ 1387-0. beginning of the sequence, until 11,000 years cal BP, there was a major diversification of fishing strategies employed. Acknowledgements The authors gratefully acknowledge Danièle Lav- In contrast, between 9000 and 8000 years cal BP, two strate- allée, Michèle Julien, and Jérôme Louvet for providing useful informa- gies became predominant: fishing nets and hooks and lines, tion about the Quebrada Los Burros site. The authors would also like accompanied by a slight contribution from spearfishing. Dur- to thank Francisca Santana and Yerko Fuentealba for their help, and to Jill Cucchi for copy editing. Finally, the authors gratefully acknowledge ing this period, a major richness is also observed; thus, cap- the suggestions of the two anonymous reviewers. ture techniques could be related to the exploitation of fish on a larger scale. This behavior may be relevant for subsequent Funding This work was funded by the National Agency for Research periods, especially during the Middle Holocene, where the and Development (ANID)/Beca Doctorado en el Extranjero/DOCT- presence of T. murphyi in the archaeological contexts eclipses ORADO BECAS CHILE/2018—72190047. S.R. received financial support from the Agencia Nacional de Investigación y Desarrollo the other species, associated with an intensification in fish- (ANID, Chile), Millennium Nucleus UPWELL [Millennium Science ing (Béarez 2012; Salazar et al. 2015, 2020; Rebolledo et al. Initiative Program UPWELL– NCN19_153]. 2016). Declarations Conclusion Competing interests The authors declare no competing interests. This study shows the fish composition of ten Terminal Open Access This article is licensed under a Creative Commons Attri- Pleistocene-Early Holocene archaeological sites on the bution 4.0 International License, which permits use, sharing, adapta- Atacama Desert coast, between southern Peru and north- tion, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, ern Chile. Methodological decisions to standardize data and provide a link to the Creative Commons licence, and indicate if changes develop a comparative study highlight the advantages and were made. The images or other third party material in this article are disadvantages of the archaeozoological method. This study included in the article's Creative Commons licence, unless indicated also attempts to integrate environmental and archaeological otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not information, to propose new lines of discussion about fishing permitted by statutory regulation or exceeds the permitted use, you will harvesting through the Terminal Pleistocene and the end of need to obtain permission directly from the copyright holder. To view a the Early Holocene. It remains necessary to obtain further copy of this licence, visit http://​creat​iveco​mmons.​org/​licen​ses/​by/4.​0/. 13 152 Page 16 of 18 Archaeological and Anthropological Sciences (2021) 13: 152 References Falabella F, Vargas L, Meléndez R (1994) Differential preservation and recovery of fish remains in Central Chile. Annales du Musée Royal de l’Afrique Central, Science Zoologiques 274:25–35 Andrade P, Salazar D (2011) Revisitando Morro Colorado: Compara- Flores C, Broitman BR (2021) Nearshore paleoceanogaphic conditions ciones y propuestas preliminares en torno a un conchal arcaico en through the Holocene: shell carbonate from archaeological sites of las costas de Taltal. Taltalia 4:63–83 the Atacama Desert coast. Palaeogeogr Palaeoclimatol Palaeoecol Arriaza BT, Standen VG, Cassman V, Santoro CM (2008) Chinchorro 562:110090. https://​doi.​org/​10.​1016/j.​palaeo.​2020.​110090 culture: pioneers of the coast of the Atacama Desert. In: Sil- Fontugne M (2012) Dataciones radiocarbónicas de los depósitos de la verman H, Isbell WH (eds) The Handbook of South American Quebrada de los Burros. In: Lavallée D, Julien M (eds) Prehis- Archaeology. Springer, New York, pp 45–58 toria de la costa extremo-sur del Perú, Institut Français d'Études Bartz M, Walk J, Binnie SA, Brill D, Stauch G, Lehmkuhl F, Ho Andines and Fondo Editorial de la Pontificia Universidad Católica D, Brückner H (2020) Late Pleistocene alluvial fan evolution del Perú, pp 91–96 along the coastal Atacama Desert (N Chile). Glob Planet Chang Fricke R, Eschmeyer WN, Fong JD (2020) Eschmeyer’s catalog of 190:103091. https://​doi.​org/​10.​1016/j.​glopl​acha.​2019.​103091 fishes: Genera, Species, references. IOP Publishing CAS http://​ Béarez P (2012) Los peces y la pesca. In: Lavallée D, Julien M (eds) resea​rchar​chive.​calac​ademy.​org/​resea​rch/​ichth​yology/​catal​og/​ Prehistoria de la costa extremo-sur del Perú, Institut Français fishc​atmain.​asp. d'Études Andines and Fondo Editorial de la Pontificia Universidad Grayson DK (1984) Quantitative zooarchaeology. Topics in the analy- Católica del Perú, pp 99–123 sis of archaeological faunas. Academic Press, London Béarez P, Fuentes-Mucherl F, Rebolledo S, Salazar D, Olguín L (2016) Heaton T, Köhler P, Butzin M, Bard E, Reimer R, Austin W, Bronk Billfish foraging along the northern coast of Chile during the Mid- Ramsey C, Grootes P, Hughen K, Kromer B, Reimer P, Adkins dle Holocene (7400-5900 cal BP). J Anthropol Archaeol 41:185– J, Burke A, Cook M, Olsen J, Skinner L (2020) Marine20 - the 195. https://​doi.​org/​10.​1016/j.​jaa.​2016.​01.​002 marine radiocarbon age calibration curve (0–55,000 cal BP). Pan- Butler VL, Campbell SK (2004) Resource intensification and resource gaea. https://​doi.​org/​10.​1594/​PANGA​EA.​914500 depression in the Pacific Northwest of North America: a zooar- Hogg AG, Hua Q, Blackwell PG, Niu M, Buck CE, Guilderson TP, chaeological review. J World Prehist 18:327–405. https://​doi.​org/​ Heaton TJ, Palmer JG, Reimer PJ, Reimer RW, Turney CSM, Zim- 10.​1007/​s10963-​004-​5622-3 merman SRH (2013) SHCal13 Southern Hemisphere Calibration, Campbell R, Quiroz D (2015) Chronological database for Southern 0–50,000 Years cal BP. Radiocarbon 55:1889–1903. https://​doi.​ Chile (35°30’-42° S), ~33000 BP to present: human implications org/​10.​2458/​azu_​js_​rc.​55.​16783 and archaeological biases. Quat Int 356:39–53. https://​doi.​org/​10.​ Houston J, Hartley AJ (2003) The central andean west-slope rain- 1016/j.​quaint.​2014.​07.​026 shadow and its potential contribution to the origin of hyper- Camus P (2001) Biogeografía marina de Chile continental. Rev Chil Hist aridity in the Atacama Desert. Int J Climatol 23:1453–1464. Nat 74:587–617. https://​doi.​org/​10.​4067/​S0716-​078X2​00100​03000​08 https://​doi.​org/​10.​1002/​joc.​938 Carter CP (2016) The economy of prehistoric coastal Northern Chile: Jackson D, Maldonado A, Carré M, Seguel R (2011) Huentelauquén case study: Caleta Vitor. Thesis submitted for the degree of Doctor cultural complex: the earliest peopling of the pacific coast in of Philosophy of The Australian National University, The Austral- the South-American southern cone. In: Vialou (ed) Peuple- ian National University ments et préhistoire en Amériques. Documents Préhistoriques Casteel RW (1974) On the remains of fish scales from archaeological 28:221–232 sites. Am Antiq 39:557–581 Jerardino A (1995) The problem with density values in archaeological Castelleti J (2007) Patrón de asentamiento y uso de recursos a través de analysis: a case study from Tortoise Cave, Western Cape, South la secuencia ocupacional prehispana de la costa de Taltal. Memo- Africa. The South African Archaeological Bulletin 50(161):21– ria para optar al título de master, Universidad Católica del Norte 27. https://​doi.​org/​10.​2307/​38892​71 - Universidad de Tarapacá Kansa EC, Kansa SW, Watrall E (2011) Archeology 2.0 New Cruz-Uribe K (1988) The use and meaning of species diversity and approaches to communication and collaboration. Cotsen Digital richness in archaeological faunas. J Archaeol Sci 15:179–196. Archaeology 1, UCLA, California https://​doi.​org/​10.​1016/​0305-​4403(88)​90006-4 Keefer DK, deFrance SD, Moseley ME, Richardson JB, Satterlee deFrance SD, Umire A (2004) Quebrada Tacahuay: Un sitio marítimo DR, Day-Lewis A (1998) Early Maritime Economy and El Niño del Pleistoceno Tardío en la costa sur del Perú. Chungara Events at Quebrada Tacahuay, Peru. Science 281:1833–1835. 36:257–278 https://​doi.​org/​10.​1126/​scien​ce.​281.​5384.​1833 Diehl M (2017) Paleoethnobotanical sampling adequacy and ubiquity: Lambrides ABJ, Weisler MI (2016) Pacific Islands ichthyoarchaeology: an example from the American Southwest. Adv Archaeol Pract implications for the development of prehistoric fishing studies and 5(2):196–205. https://​doi.​org/​10.​1017/​aap.​2017.5 global sustainability. J Archaeol Res 24:275–324. https://​doi.​org/​ Disspain MCF, Ulm S, Santoro CM, Carter C, Gillanders BM (2016) 10.​1007/​s10814-​016-​9090-y Pre-Columbian fishing on the coast of the Atacama Desert, north- Latorre C, Moreno P, Vargas G, Maldonado A, Villa-Martínez R, ern Chile: an investigation of fish size and species distribution Armesto J, Villagrán C, Pino M, Núñez L, Grosjean M (2003) using otoliths from Camarones Punta Norte and Caleta Vitor. Late Quaternary environments and palaeoclimate. In: Gibbons Journal of Island and Coastal Archaeology 00:1–23. https://​doi.​ W, Moreno T (eds) The Geology of Chile. London Geological org/​10.​1080/​15564​894.​2016.​12043​85 Society Press, London, pp 309–328 Driver JC (1992) Identification, classification and zooarchaeology. Lavallée D, Julien M (2012) Prehistoria de la costa extremo-sur del Circaea 9:35–47 Perú: Los pescadores arcaicos de la Quebrada de los Burros Driver JC (2011) Identification, classification and zooarchaeology. Eth- (10000-7000 a. P.). Institut Français d'Études Andines and Fondo nobiology Letters 2:19–39. https://d​ oi.o​ rg/1​ 0.1​ 4237/e​ bl.2.2​ 011.3​ 2 Editorial de la Pontificia Universidad Católica del Perú Dürr HH, Laruelle GG, Kempen CM, Slomp CP, Meybeck M, Mid- Lavallée D, Julien M, Béarez P, Bolaños A (2011) Quebrada de los delkoop (2011) Worldwide typology of nearshore coastal systems: Burros: Los primeros pescadores del litoral Pacífico en el extremo defining the estuarine filter of river inputs to the oceans. Estuar sur peruano. Chungara 43:333–352. https://​doi.​org/​10.​4067/​ Coasts 34:441–458. https://​doi.​org/​10.​1007/​s12237-​011-​9381-y S0717-​73562​01100​03000​02 13 Archaeological and Anthropological Sciences (2021) 13: 152 Page 17 of 18 152 Lavallée D, Julien M, Béarez P, Usselmann P, Fontugne M, Bolaños A Rebolledo S, Béarez P, Zurro D, Santoro C, Latorre C (2021) Big fish (1999) Pescadores-recolectores arcaicos del extremo sur peruano. or small fish? Differential ichthyoarchaeological representation Excavaciones en la Quebrada de los Burros (Tacna, Perú). Prim- revealed by different recovery methods in the Atacama Desert eros resultados 1995-1997. Bulletin de l’Institut Français d’Études Coast, northern Chile. Environ Archaeol J Hum Palaeoecol. Andines 28:13–52 https://​doi.​org/​10.​1080/​14614​103.​2021.​18866​47 Llagostera A (1977) Ocupación humana en la costa norte de Chile Reitz EJ (2009) International Council for Archaeozoology (ICAZ) asociada a peces local-extintos y a litos geométricos: 9680±160 professional protocols for archaeozoology. https://​doi.​org/​10.​ A.P. Actas del VII Congreso de Arqueología en Chile, Kultrun, 1017/​CBO97​81107​415324.​004 Santiago de Chile, pp 93–113 Reitz EJ, McInnis HE, Sandweiss DH, deFrance SD (2016) Terminal Llagostera A (1979) 9,700 years of maritime subsistence on the Pacific: Pleistocene and Early Holocene fishing strategies at Quebrada an analysis by means of bioindicators in the north of Chile. Am Jaguay and the Ring Site, southern Perú. J Archaeol Sci Rep Antiq 44:309–324 8:447–453. https://​doi.​org/​10.​1016/j.​jasrep.​2016.​05.​035 Llagostera A (1989) Caza y pesca marítima (9.000 a 1.000 a.C.). In: Reitz EJ, Wing ES (1999) Zooarchaeology. Cambridge University Hidalgo J, Schiappacasse V, Niemeyer H, Aldunate C, Solimano Press, Cambridge, Cambridge Manuals in Archaeology I (eds) Prehistoria: desde sus orígenes hasta los albores de la con- Roberts A, Pate F, Petruzzelli B, Carter C, Westaway MC, Santoro quista. Editorial Andrés Bello, Santiago de Chile, pp 57–79 CM, Swift J, Maddern T, Jacobsen GE, Bertuch F, Rothham- Llagostera A (1992) Early occupations and the emergence of fishermen mer F (2013) Retention of hunter-gatherer economies among on the Pacific Coast of South America. Andean Past 3:87–109 maritime foragers from Caleta Vitor, northern Chile, during the Llagostera A (2005) Culturas costeras precolombinas en el norte chil- late Holocene: Evidence from stable carbon and nitrogen iso- eno: secuencia y subsistencia de las poblaciones arcaicas. In: topic analysis of skeletal remains. J Archaeol Sci 40:2360–2372. Figueroa E (ed) Biodiversidad marina: valoración, usos, perspec- https://​doi.​org/​10.​1016/j.​jas.​2013.​01.​009 tivas ¿Hacia dónde va Chile? Editorial Universitaria, Santiago de Salazar D, Andrade P, Borie C, Escobar M, Figueroa V, Flores C, Chile, pp 107–148 Olguín L, Salinas H (2013) Nuevos sitios correspondientes al Llagostera A, Kong I, Irachet P (1997) Análisis ictioarqueológico del complejo cultural Huentelauquén en la costa de Taltal. Revista sitio La Chimba 13 (II Región, Chile). Chungara 29:163–179 Taltalia 5–6:9–19 Martens T, Cameron J (2019) Early coastal fiber technology from Salazar D, Arenas C, Andrade P, Olguín L, Torres J, Flores C, Vargas Caleta Vitor archaeological complex in Northern Chile. Lat Am G, Rebolledo S, Borie C, Sandoval C, Silva C, Delgado A, Lira N, Antiq 30:1–13. https://​doi.​org/​10.​1017/​laq.​2018.​78 Robles C (2018) From the use of space to territorialisation during McKechnie I, Moss ML (2016) Meta-analysis in zooarchaeology the Early Holocene in Taltal, coastal Atacama Desert, Chile. Quat expands perspectives on Indigenous fisheries of the Northwest Int 473:225–241. https://​doi.​org/​10.​1016/j.​quaint.​2017.​09.​035 Coast of North America. J Archaeol Sci Rep 8:470–485. https://​ Salazar D, Figueroa V, Andrade P, Salinas H, Olguín L, Power X, doi.​org/​10.​1016/j.​jasrep.​2016.​04.​006 Rebolledo S, Parra S, Orellana H, Urrea J (2015) Cronología y O’Connor TP (1985) On quantifying vertebrates - some sceptical organización económica de las poblaciones arcaicas de la costa observations. Circaea 3:27–30 de Taltal. Estudios Atacameños 50:7–46. https://​doi.​org/​10.​4067/​ Olguín L, Castro V, Castro P, Peña-Villalobos I, Ruz J, Santander S0718-​10432​01500​01000​02 B (2015) Exploitation of faunal resources by marine hunter- Salazar D, Flores C, Borie C, Olguín L, Rebolledo S, Escobar M, gatherer groups during the Middle Holocene at the Copaca 1 Cifuentes A (2020) Marine communities in the Atacama Desert. site, Atacama Desert coast. Quat Int 373:4–16. https://​doi.​org/​ Economic Organization and Social Dynamics of Middle-Holocene 10.​1016/j.​quaint.​2015.​02.​004 Hunter-Gatherer-Fisher Communities on the Coast of the Atacama Olguín L, Salazar D, Jackson D (2014) Tempranas evidencias de Desert (Taltal, Northern Chile). In: Prieto G, Sandweiss DH (eds) navegación y caza de especies oceánicas en la costa pacífica Maritime communities of the ancient Andes. University Press of de Sudamérica (Taltal, ~7,000 cal A.P.). Chungara 46:177–192 Florida, Florida, pp 74–100 Orton C (2000) Sampling in archaeology. Cambridge, Cambridge Sandweiss DH, Mcinnis H, Burger RL, Cano A, Ojeda B, Paredes Orton C, Morris J, Locker A, Barrett JH (2014) Fish for the city: R, Sandweiss C, Glascock MD (1998) Quebrada Jaguay : Early Meta-analysis of archaeological cod remains and the growth of South American Maritime Adaptations. Science 281:1830–1832 London’s northern trade. Antiquity 88:516–530. https://​doi.​org/​ Sandweiss DH, Rademaker K (2011) El poblamiento del sur peruano : 10.​1017/​S0003​598X0​01011​52 Costa y sierra. Bolet Arqueol PUCP 15:275–293 Paskoff R (2010) Geomorfología costera. In: Frutos J (ed) Díaz- Sandweiss DH, Richardson JR III, Reitz EJ, Hsu JT, Feldman RA Naveas J. Geología Marina de Chile, Comité Oceanográfico (1989) Early maritime adaptations in the Andes: Preliminary stud- Nacional, Santiago de Chile, pp 76–83 ies at the Ring Site, Peru. In: Rice DS, Stanish C, Scarr PR (eds) Pequeño G (2000) Delimitaciones y relaciones biogeográficas de Ecology, Settlement, and History in the Osmore Drainage, Peru, los peces del Pacífico Suroriental. Estudios Oceanol 19:53–76 BAR International Series, vol 545, pp 35–84 Popper VS (1988) Selecting Quantitative Measures in Paleoethnobot- Santoro CM, Castro V, Carter CP, Valenzuela D (2020) Marine com- any. In: Hastorf CA, Popper VS (eds) Current paleoethnobotany: munities in the Atacama Desert. Masters of the sub-tropical analytical methods and cultural interpretations of plant remains. Pacific Coast of South America. In: Prieto G, Sandweiss DH (eds) University of Chicago Press, Chicago, pp 53–71 Maritime communities of the ancient Andes. University Press of Rebolledo S (2017) Exploitation des ressources halieutiques sur la Florida, Florida, pp 39–73 côte Pacifique du Chili au cours de l’Holocène Ancien: les sites Santoro CM, Rivadeneira MM, Latorre C, Rothhammer F, Standen VG Paposo Norte 9, Alero 224-A et Morro Colorado de la commune (2012) Rise and decline of Chinchorro sacred landscapes along de Taltal. Mémoire de Master. Muséum National d’Histoire the hyperarid coast of the Atacama Desert. Chungara 44(4):637– Naturelle de Paris 653. https://​doi.​org/​10.​4067/​S0717-​73562​01200​04000​07 Rebolledo S, Béarez P, Salazar D, Fuentes F (2016) Maritime fish- Scartascini FL (2016) The role of ancient fishing on the desert coast ing during the Middle Holocene in the hyperarid coast of the of Patagonia, Argentina. J Island Coast Archaeol 00:1–18. https://​ Atacama Desert. Quaternary International 391:3–11. https://​doi.​ doi.​org/​10.​1080/​15564​894.​2016.​11723​80 org/​10.​1016/j.​quaint.​2015.​09.​051 13 152 Page 18 of 18 Archaeological and Anthropological Sciences (2021) 13: 152 Schulz N, Boisier JP, Aceituno P (2011) Climate change along the arid Vargas G, Rutllant J, Ortlieb L (2006) ENSO tropical-extratropical cli- coast of northern Chile. Int J Climatol 32:1803–1814. https://​doi.​ mate teleconnections and mechanisms for Holocene debris flows org/​10.​1002/​joc.​2395 along the hyperarid coast of western South America (17°-24°S). Spalding MD, Fox HE, Allen GR, Davidson N, Ferdaña ZA, Finlayson Earth Planet Sci Lett 249:467–483. https://​doi.​org/​10.​1016/j.​epsl.​ M, Halpern BS, Jorge MA, Lombana A, Lourie SA, Martin KD, 2006.​07.​022 McManus E, Molnar J, Recchia CA, Robertson J (2007) Marine Walk J, Stauch G, Reyers M, Vásquez P, Sepúlveda FA, Bartz M, ecoregions of the world: a bioregionalization of coastal and shelf Hoffmeister D, Brückner H, Lehmkuhl F (2020) Gradients in areas. BioScience 57:573–583. https://​doi.​org/​10.​1641/​b5707​07 climate , geology , and topography affecting coastal alluvial fan Standen VG, Santoro CM (2004) Patrón funerario arcaico temprano morphodynamics in hyperarid regions – The Atacama perspec- del sitio Acha-3 y su relación con Chinchorro: Cazadores, pes- tive. Glob Planet Chang 185:102994. https://​doi.​org/​10.​1016/j.​ cadores y recolectores de la costa norte de Chile. Lat Am Antiq glopl​acha.​2019.​102994 15(1):89–109. https://​doi.​org/​10.​2307/​41415​65 Wolverton S (2013) Data Quality in Zooarchaeological Faunal Identi- Steele TE (2015) The contributions of animal bones from archaeologi- fication. J Archaeol Method Theory 20:381–396. https://​doi.​org/​ cal sites: the past and future of zooarchaeology. J Archaeol Sci 10.​1007/​s10816-​012-​9161-4 56:168–176. https://​doi.​org/​10.​1016/j.​jas.​2015.​02.​036 Zangrando AF (2009) Is fishing intensification a direct route to Tapia J, González R, Townley B, Oliveros V, Álvarez F, Aguilar G, hunter-gatherer complexity? A case study from the Beagle Chan- Menzies A, Calderón M (2018) Geology and geochemistry of nel region (Tierra del Fuego, southern South America). World the Atacama Desert. Anton Leeuw Int J Gen Mol Microbiol Archaeol 41:589–608. https://​doi.​org/​10.​1080/​00438​24090​33638​ 111:1273–1291. https://​doi.​org/​10.​1007/​s10482-​018-​1024-x 48 Thiel M, Macaya EC, Acuna E, Arntz WE, Bastias H, Brokordt K, Zangrando AF, Tivoli AM, Alunni DV, Pérez SA, Martinoli MP, Pinto Camus PA, Castilla JC, Castro LR, Cortes M, Dumont CP, Escrib- Vargas G (2021) Exploring shell midden formation through tapho- ano R, Fernandez M, Gajardo JA, Gaymer CF, Gomez I, Gonzalez chronometric tools: a case study from Beagle Channel, Argentina. AE, Gonzalez HE, Haye PA, Illanes JE, Iriarte JL, Lancellotti DA, Quat Int 584:33–43. https://​doi.​org/​10.​1016/j.​quaint.​2020.​04.​050 Luna-Jorquerai G, Luxoroi C, Manriquez PH, Marin V, Munoz P, Zohar I, Dayan T, Goren M, Nadel D, Hershkovitz I (2018) Opportun- Navarrete SA, Perez E, Poulin E, Sellanes J, Sepulveda HH, Stotz ism or aquatic specialization? Evidence of freshwater fish exploi- W, Tala F, Thomas A, Vargas CA, Vasquez JA, Vega JMA (2007) tation at Ohalo II- a waterlogged upper paleolithic site. PLoS ONE The Humboldt current system of Northern and Central Chile. 13:13–17. https://​doi.​org/​10.​1371/​journ​al.​pone.​01987​47 Oceanographic processes, ecological interactions and socioeco- nomic feedback. Oceanogr Mar Biol 45:195–344. https://​doi.​org/​ Publisher’s note Springer Nature remains neutral with regard to 10.​1091/​mbc.​E04-​05-​0427 jurisdictional claims in published maps and institutional affiliations. 13