Abstract The hypothesis that Neanderthals exploited birds for the use of their feathers or claws as personal ornaments in symbolic behaviour is revolutionary as it assigns unprecedented cognitive abilities to these hominins. This inference, however, is based on modest faunal samples and thus may not represent a regular or systematic behaviour. Here we address this issue by looking for evidence of such behaviour across a large temporal and geographical framework. Our analyses try to answer four main questions: 1) does a Neanderthal to raptor-corvid connection exist at a large scale, thus avoiding associations that might be regarded as local in space or time?; 2) did Middle (associated with Neanderthals) and Upper Palaeolithic (associated with modern humans) sites contain a greater range of these species than Late Pleistocene paleontological sites?; 3) is there a taphonomic association between Neanderthals and corvids-raptors at Middle Palaeolithic sites on Gibraltar, specifically Gorham's, Vanguard and Ibex Caves? and; 4) was the extraction of wing feathers a local phenomenon exclusive to the Neanderthals at these sites or was it a geographically wider phenomenon?. We compiled a database of 1699 Pleistocene Palearctic sites based on fossil bird sites. We also compiled a taphonomical database from the Middle Palaeolithic assemblages of Gibraltar. We establish a clear, previously unknown and widespread, association between Neanderthals, raptors and corvids. We show that the association involved the direct intervention of Neanderthals on the bones of these birds, which we interpret as evidence of extraction of large flight feathers. The large number of bones, the variety of species processed and the different temporal periods when the behaviour is observed, indicate that this was a systematic, geographically and temporally broad, activity that the Neanderthals undertook. Our results, providing clear evidence that Neanderthal cognitive capacities were comparable to those of Modern Humans, constitute a major advance in the study of human evolution.

Citation: Finlayson C, Brown K, Blasco R, Rosell J, Negro JJ, Bortolotti GR, et al. (2012) Birds of a Feather: Neanderthal Exploitation of Raptors and Corvids. PLoS ONE 7(9): e45927. https://doi.org/10.1371/journal.pone.0045927 Editor: Michael D. Petraglia, University of Oxford, United Kingdom Received: April 20, 2012; Accepted: August 23, 2012; Published: September 17, 2012 Copyright: © Finlayson et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Funding: The excavations and scientific research associated with Gorham's Cave have been funded by the Government of Gibraltar and, additionally between 2002–04, by the European Community - Program Interreg IIIB: 2002-02-4, 1-U-048 -. R. Blasco and J. Rosell are part of the following projects: 2009 SGR 188 of Generalitat de Catalunya and CGL2009-12703-C03-02 of the Spanish Goverment. A. Sánchez Marco belongs to the CGL2011-28681 project of the Spanish Government. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing interests: The authors have declared that no competing interests exist.

Introduction The regular and systematic exploitation of flying birds for food is considered to be a hallmark of behavioural modernity, exclusive to anatomically modern Homo sapiens (Modern Humans) after 50 thousand years ago (kya) [1], [2]. The prevailing paradigm among Palaeolithic archaeologists today is still one which regards flying birds to have been difficult prey to capture and beyond the capabilities of all hominins prior to 50 kya and non-modern hominins (including the Neanderthals) even after the 50 kya threshold [1], [2]. The corollary, which has been applied to the Neanderthals for the period after 50 kya, is that they only targeted birds once easier prey (presumed to be energetically less costly to obtain than birds) were exhausted [3], [4]. Even when evidence that the Neanderthals took prey commonly regarded as difficult has been presented [5], the argument that these are examples of opportunistic and unsystematic captures has been used in response [6]. These interpretations have been contested from an ecological perspective which suggests that Neanderthals were equally versatile omnivorous hunter-gatherers [7], who even included marine mammals in their diet when available [5]. Recently, evidence has been accumulating that strongly suggests that Neanderthals regularly exploited birds as part of a varied diet within coastal Mediterranean regions [8], [9], [10]. However, the hypothesis that Neanderthals exploited birds for the use of their feathers or claws as personal ornaments in symbolic behaviour [11], [12] is revolutionary as it assigns unprecedented cognitive abilities to these hominins. Specifically, raptors (Orders Accipitriformes and Falconiformes) and corvids (Family Corvidae in the Order Passeriformes) were among the bird taxa found associated with Neanderthals at Riparo Fumane, Italy [11] and, Combe-Grenal and Les Fieux, France [12]. The suggestion that Neanderthals exploited birds for ornamental purposes has added a further and important dimension to the debate, that of their cognitive capacities. This hypothesis has, however, been put forward on the basis of very small samples and is thus open to the criticism that it does not represent regular or systematic behaviour. To assess the existence of universal patterns of early use of feathers for ornamental and symbolic purposes, here we examine the relationship between Modern Humans, Neanderthals, raptors and corvids across a broad temporal and geographical framework: the Palearctic Region in the Middle and Late Pleistocene. These taxa of birds are chosen for the present study because (a) they are frequently present in sites occupied by hominins; (b) they represent taxa that are not typically consumed by hominins; and (c) they are carnivores that often scavenge the corpses of medium and large mammals, so that they were likely to frequently come into close contact with humans. They may have, in all likelihood, also been regular scavengers around Palaeolithic camp sites [13], as they are today in urban areas and garbage dumps in many parts of the world [14]. Corvids are abundant species in many Eurasian landscapes while raptors - apex predators - tend to be scarcer. Both groups include rock dwelling species that would be naturally expected to accumulate close to nesting sites but there are no known taphonomic processes that would concentrate the remains from these taxa any more than other rock-dwelling birds. In any case our findings also include species that are tree nesters as well. For these reasons the palaeontological sites would seem to reflect natural accumulation rates. From a multi-scale approach, we show that strong positive relationships exist between Neanderthal-raptor and corvid. On the other hand, we confirm, using taphonomic data from three sites in Gibraltar (Gorham's, Vanguard and Ibex Caves), that the relationship involves active processing of raptors and corvids by Neanderthals for the purpose of wing feather removal. The temporal and geographical extent of the connection, along with the direct taphonomic evidence, establishes that Neanderthals systematically targeted these birds for purposes other than food.

Discussion and Conclusions The strong relationship between Neanderthals, corvids and raptors requires explanation as does the clear evidence of direct action on the bird bones. If this processing of raptors and corvids by Neanderthals had been related to consumption, then we would have expected a concentration of anthropic marks in parts of the anatomy linked to the fleshy regions of the body (e.g. the sternum which holds the large pectoral muscles). Instead, it is the wing bones, low in meat but anchors for the large flight feathers, which were processed. The overrepresentation of raptor and corvid wing bones in Neanderthal sites cannot thus be interpreted in any way other than the use of their feathers. This is supported by the statistically significantly high proportion of individual wing bones (Goodness of Fit, G 3 = 139.849, p<0.0001; Table S6) and the fact that these had a statistically significant higher frequency of anthropic marks than other bones (Goodness of Fit, G 2 = 29.2568, p<0.0001; Table S6). Within the wing bones, humeri and ulnae – bones that support the large flight feathers - appeared to have the highest frequency of anthropic marks (Table S2 and Text S3). The carpo-metacarpi - also supporting flight feathers - might not, we suspect, require as much processing because of their small size, and this may explain the relatively low proportion with anthropic marks. Support that the processing by Neanderthals involved feather removal, and not food, comes from the observation that raptors and corvids are not regularly eaten in any culture, confirmed by the lack of data of corvid or raptor consumption in the ethnographic literature. Feathers as such are not edible either, and they are rapidly disintegrated by feather-degrading bacteria in the soil [24]; thus their use for bedding on cave floors is precluded. The most parsimonious explanation for feather use by Neanderthals would be the same as for tribal Modern Humans: ornaments on their heads and bodies. Why were dark raptor and corvid feathers selected preferentially over others? These bird species are related to rocky outcrops for nesting and roosting and savannah-like habitats for foraging [7], [17]. They would have therefore been familiar to the Neanderthals and a part of their daily lives; opportunities for obtaining feathers from live birds, at nests or roosts, or from individuals that died and fell to the ground would have been plentiful. They may even have shared the same food resources, as both humans and these scavengers would have coincided around ungulate carcasses. These birds may well have acted as indicators of freshly dead animals to the Neanderthals. Carcasses would have become focal points of convergence for large numbers of vultures, other raptors and corvids, as they still do today. These would have been ideal conditions allowing the Neanderthals the possibility, which would have necessitated a degree of planning and anticipation, of capturing the large birds as they gorged themselves. The behaviour might therefore have originated in the practice of following large birds to fresh carcasses for food. The apparent selection for feathers of specific color, that our results show, adds yet another dimension, requiring sophisticated cognitive processes, to the demonstrated non-random use of feathers. Lacking previous examples of feather use by Neanderthals, except the valuable recent suggestions by Peresani et al. [11], we have reviewed use of feathers by the only surviving Homo species Modern Humans, H. sapiens. Current or historic use of feathers by Modern Humans is widespread and spans practically every culture that has been studied, including modern western civilization as well as numerous tribal peoples in every permanently inhabited continent (Table S7). This pattern of feather use for adornment appears to be part of the universal human psyche. The Neanderthals clearly shared this invariant behaviour [25] with Modern Humans, suggesting that it may have been a common characteristic of the two lineages, although we cannot determine if one learnt the behaviour from the other or if it was, instead, present in the common ancestor. Focusing on tribal examples, and assuming they may represent ancestral traditions, we observe that in a majority of cases the use of feathers is ornamental, in the form of headdresses [26], cloth decorations, as in skirts or belts, or even full feather cloaks or capes [27], as those worn by Hawaiian or Maori chiefs. A common characteristic of ornamentation, of which jewellery is the best example, tends to require valuable items that are not easy to replace. Feathers as ornaments seem to follow this rule, common to any biological signal, that is, they are costly to produce or to maintain [28]. The bird species used by humans, such as the Golden Eagle Aquila chrysaetos in the case of the Amerindians, were either scarce in the environment [29] or many individuals were needed to produce the elaborate feathered ornaments, as was the case for the red and yellow birds used in Hawaiian capes, in which thousands of individuals were killed to make a single garment [30]. The use of feathers, or the application of other species trophies as adornment on the body, is an exclusively human trait. Feather adornments, however, are not the earliest cases of ornamentation in humans. For Modern Humans, ochre pigment use has been suggested as the first manifestation of symbolic behaviour, in South Africa over 160 thousand years ago [31]; it has been heralded as evidence of the transition to “modernity” in humans. Recently, similar evidence of pigment use has been found in the case of early Neanderthals at Maastricht-Belvédère, Netherland (200–250 kya) [32]. With more recent chronologies, the use of manganese and iron oxides by Late Pleistocene Neanderthals seem to be documented from at least 60 kya onward [33], [34], [35]. In spite of this, the absence of beads, portable figurines or cave art in Neanderthal sites continues to be cited as evidence of their inferior cognitive capacities [36]. That Neanderthals shared this uniquely human trait of feather ornamentation with Modern Humans, provides a further bridge that brings them closer to each other. Recent evidence seems to have resolved the question of Neanderthal-Modern Human gene interchange [37], showing that such exchange in all likelihood occurred in the course of the history of the two lineages. The biological differences between the two could therefore not have been as great as previously envisaged if they were able to interbreed. But the debate of cognitive differences remains open. Discussion of the cognitive abilities of the Neanderthals has a protracted history which came to the fore with the debate on whether ornamentation found associated with Neanderthals in France was autochthonous or was instead the product of acculturation from Modern Humans or trade with them [38], [39]. This debate continues to generate controversy [40], [41] and leaves the question of Neanderthal cognitive capacities unresolved. The results presented here show that extraction of feathers from birds by Neanderthals was a temporally and geographically widespread phenomenon. The results are reinforced by evidence of repetition of this behaviour across a substantial time period of thousands of years in Gibraltar. The earliest observation of this behaviour in Gibraltar preceded the arrival of Modern Humans in Europe by several thousand years. There is therefore no possibility that the practice was acquired from Modern Humans. Thus Neanderthals, though different in a number of ways from Modern Humans had comparable cognitive capacities that included symbolic expression. The observed behavioural differences between them therefore have to be related to distinct cultural trajectories, as would have been the case between different Modern Human populations [42], [43]. We have shown that Neanderthals were associated with raptors and corvids of particular characteristics (dark remiges, scavenging or colonial cliff nesters) across the entire geographical space of the Palearctic and they directly processed their bones for their feathers. In this respect they were distinctly human. The absence of parietal art in caves occupied by Neanderthals, and also of bone and shell ornaments, is a key argument cited in support of the superior cognitive capacities of Modern Humans. Our results put this long-standing contention in doubt, by providing strong evidence that Neanderthals simply used media, other than cave walls, to express themselves.

Acknowledgments The authors wish to dedicate this paper to the memory of our friend and colleague - the late Gary Bortolotti - who was inspirational in the development of this paper and was sadly unable to see the final version in print. We would like to thank the Editor, Michael D. Petraglia, R. Lee Lyman and one anonymous reviewer for their comments on the manuscript that have greatly improved the final version.

Author Contributions Analyzed the data: CF KB RB JR ASM. Contributed reagents/materials/analysis tools: JJN GRB GF. Wrote the paper: CF RB JR JJN GRB. Palaeoecology: CF GF JJN GRB JSC DF. Statistical analyses: CF RB JR JJN GRB GF. Palaeontology: ASM. Geology and Geomorphology: JRV JMRL. Taphonomy and Zooarchaeology: KB RB JR. Lithic technology: FGP. Ethnology: JJN GRB.