Did Upper Palaeolithic people provision food for their raven neighbours?

Did Upper Palaeolithic people provision food for their raven neighbours?

Human-wildlife interaction has emerged as a pressing theme in the Anthropocene, as wild animals are increasingly drawn into human spaces, unsettling our treasured Western divisions between nature and culture, and necessitating new forms of engagement and governance. Interactions with wild animals are now often in the news, from urban foxes to tree-felling beavers and rampaging wild boars. Some of these species have become flagship examples of ‘green’ cities, providing attractive, and sometimes preferred, habitats for a wide range of nonhuman species, especially evocative during the global Covid-19 pandemic. Yet many of these species are also regarded as pests by modern city dwellers as they interfere with human projects, often with negative effects. Their return to human spaces is viewed cautiously. Misunderstanding and conflict abound, we can learn much from taking a deep dive into the evolutionary history of our species, as humans have interacted with wild animals for thousands, even millions of years. These relationships were always subject to change and not all animals were equally prominent interlocutors, and not in all contexts. Although wild animals continuously lived in close spatial proximity to humans, some animals, at some point in the more than two million years of human history, started to adapt to human neighbourhoods and to actively take advantage of what these had to offer. These animals are called ‘early adopters’ as they are capable to quickly exploit the opportunities created by human co-habitation, notably the availability of easy-to-access food stuff at human settlements. But what were these animal species? And when did such processes begin? These are questions that can only be addressed by combined archaeological, zoological, and biomolecular research. We here report on such research suggesting that common ravens were among the first animals to exhibit such behaviour around 30,000 years ago in what is today Moravia.

As many science stories, asking these questions and finding a suitable context to properly address them has an unlikely story itself. In our case, it all began with one of us (me/STH) being interested in owls and how these fascinating birds are represented in Palaeolithic art. Owls are extraordinarily rare in the art left behind by European Ice Age hunter-gatherers. Some depictions are found in caves and a very peculiar set of owl-shaped pendants and clay figures is known from the so-called Pavlovian culture (ca. 27-31,000 years ago) in today’s Czech Republic. I am interested in the deep history of human-animal relationships and Palaeolithic art depicting animals is a valuable source of knowledge in this regard. While investigating the archaeological and zooarchaeological context of the owl representations to show that this art can tell us something important about how humans and owls met in the past, I realised how unusually abundant corvid bones are in the zooarchaeological assemblages associated with the same sites harbouring the owl-related art. What is more, the vast majority of the corvids were identified as common ravens (Corvus corax), and as such this constitutes a notable anomaly in the earlier part of the European Upper Palaeolithic. I immediately began to ask myself what the reasons for this raven anomaly could be. Already at the time, I suspected that corvids were in some way facilitated by Pavlovian foragers and must have lived close to them, and in larger numbers. I noted that the relative abundance, visibility and audibility of ravens as creatures of the daylight contrasts with the invisibility and elusiveness of many owls as creatures of the dusk and night. While there is evidence that Pavlovian people would extract the shining feathers of ravens, they depicted owls in their art and referenced them in personal adornment. I therefore proposed that the duality between owls and ravens was culturally important to Pavlovian people and tells us how they lived together with these animals and related to them.

Left: view from the Dyje river to Dolní Věstonice 1 with the Pavlovian Hills in the background. Photography by A. Bohmers, approximately 1941/42; archive of the Moravian Museum, Brno. Right: excavations at Pavlov 1 led by B. Klíma in 1957. Photograph by K. O. Hrubý; archive of the Moravian Museum, Brno.

When discussing these ideas with Chris, we decided to test the hypothesis that these ravens somehow benefited from Pavlovian settlements, in turn explaining their relative abundance at the respective archaeological sites. One possibility to directly and independently test such a scenario is to examine the diets of the birds, as early adopter animals often target novel food resources made available by human occupations. Chris’ research focuses on the reconstruction of trophic interactions and food-webs in Late Pleistocene ecosystems using biomolecular methods and especially stable isotope analysis, which is a common way of examining the diets of past animals. To analyse Pavlovian common raven bones in this manner requires well-preserved bone collagen and it wasn’t clear whether bird bones that are potentially around 30,000 years old would qualify. This is also because birds are not normally in the focus of bioarchaeological research. Isotope analysis has so far mainly been used to assess the ecology and diet of large mammals, especially herbivores, as these animals are assumed to have been of key importance to human hunter-gatherers, and to best track environmental and climatic changes. There was therefore no guarantee for success, and it was generally unclear what to expect. We teamed up with Martina, the curator of the Quaternary palaeontological collection of the Moravian Museum in Brno, where the raven remains from the Pavlovian key sites of Pavlov I, Předmostí I and Dolní Věstonice I are housed. We were especially keen to sample raven specimens from the site of Předmostí since previous stable isotope research had created a good dataset comprising varied animal species including Artic fox, reindeer, horse, wolf and mammoth, with which the raven data could be compared. We initially sampled 15 raven individuals, a snowy owl, and a vulture.

Carcass of a young bison attended by ravens, Yellowstone National Park, January 2022. Photograph courtesy of Mike Dunn (https://roadsendnaturalist.com/)
Carcass of a young bison attended by ravens, Yellowstone National Park, January 2022. Photograph courtesy of Mike Dunn (https://roadsendnaturalist.com/)

When the measurements returned after pre-treatment and laboratory work, we were extremely happy that 12 of the 15 sampled raven bones yielded enough collagen to allow further analysis. We then also decided to radiocarbon date a subsample of these bones to make sure that all examined ravens are associated with the Pavlovian and to assess their internal chronology, if any. The raven bone collagen was analysed in terms of carbon, nitrogen, and sulfur stable isotopic ratios. Carbon and nitrogen were used to establish general dietary profiles, while sulfur was measured in order to investigate mobility and land-use differences. When we plotted the raw measurements for diet and compared them with the available stable isotope data for other animals as well as human individuals from the Pavlovian, it quickly became clear that the raven data was very interesting. Already on first visual inspection, two ravens from Pavlov I showed very similar carbon and nitrogen values as the Pavlovian humans, for which both stable isotope and zooarchaeological research has suggested mammoth-rich diets. When using statistical methods to estimate the relative composition of raven diets, it became clear that ravens must have mainly eaten large herbivores, especially horse and bison, and that mammoth was a substantial dietary component for most ravens. The radiocarbon dates further suggested that there are no notable differences in when these ravens died. The obtained dates may actually suggest that the raven phenomenon is mainly linked to the later stages of Pavlovian settlement in Moravia. We were now increasingly convinced that this pattern had something to do with the presence of people in the same environments and did not simply reflect the natural food consumption of ravens in such environments. To build up their isotopic signature, ravens have to consistently feed on the respective animal resources over many years. Moreover, it seems unlikely that the strong alignment in feeding preferences between Pavlovian humans and ravens, in terms of specific species, is merely a coincidence. We indeed suggest that it is rather impossible coincidence, and that other possible food resources cannot sufficiently explain the data we obtained.

Ravens foraging at a human dumpster, Gardiner, Montana, close to Yellowstone National Park, USA. Photograph courtesy of Dr. Matthias-Claudio Loretto.

In our paper, we therefore propose that common ravens regularly scavenged larger herbivore carcasses, notably dead mammoths, which Pavlovian hunter-gatherers accumulated and deposited in the landscape, especially in the vicinity of their settlement sites. Ravens were thus commensal to humans, they co-fed on the latter’s hunting spoils. We interpret this as the until now earliest known evidence for what we call bird synanthropes. Synanthropes are animals that benefit from and thrive in human-shaped ecologies. We argue that the isotope data reported in the paper provides evidence that some ravens exploited a dietary niche fundamentally sustained by human predatory behaviour and the elevated but nonetheless selective availability of large herbivore carrion that resulted from it. This interpretation gains further credence, we believe, by the larger archaeological context of Pavlovian settlement and ecosystem impact. The Palvovian sites in Moravia belong to a larger cluster of archaeological sites with often extensive occupational horizons, dense material culture, bone deposits, and some built structures, including pits and possibly meat caches. Other contextual evidence similarly suggest that the landscape was probably occupied by Pavlovian people nearly year-round, and this kind of human ecological fingerprint is certainly not the norm in the larger context of the timeframe. Another interesting observation is that evidence for direct carnivore interference with the accumulated faunal remains at the main Pavlovian sites is surprisingly low, in fact virtually absent, and this may suggest that larger carnivores such as wolves and lions were suppressed wherever people settled. Some of them were hunted by Pavlovian people as testified by their cut-marked bones, perhaps as they tried to claim carcasses. We suggest that this would have reduced the competition for carcasses within the larger scavenger guild of the Palvovian, and such human-dominated landscapes would naturally promote smaller fast-moving, difficult-to-expel avian scavengers such as ravens. This again emphasises the likely role of past human behaviour in contributing to the ecological dynamics that may have led to the archaeological accumulation of the relatively high numbers of raven remains with similar dietary preferences as contemporaneous humans.

Schematic outline of the early emergence of human-sustained feeding niches for early adopter animals during the Upper Palaeolithic.

In total, our research draws attention to the involvement of some Upper Palaeolithic hunter-gatherers in the manipulation of carrion supply dynamics already 30,000 years ago. Such manipulation, in turn, provided unique conditions for some early adopter animals such as ravens to ‘colonize’ these human-shaped landscapes and to prosper in them. By constructing specific micro-ecologies, past human foragers drew ravens into their activity sphere and thereby developed new forms of relationships with them. Humans and ravens were strongly exposed to each other, and ravens became a regular and integral part of people’s everyday experiences. As a result, ravens also became culturally important for Pavlovian people, and this is reflected in the extraction and use of their feathers. The consistent association of ravens with dead animals at the periphery of human domestic spaces thereby helped to shape particular understandings of the birds, some of which may still resonate in some circumarctic cosmologies today, where ravens are often regarded as important transformative agents, ‘tricksters’ or ‘world-makers’. The form of coexistence between humans and ravens observed during the Pavlovian at least seems to have resulted in both species orienting their behaviours towards each other, and we suggest that the Pavlovian is therefore a good model system for better understanding the relationship between human ecosystem impact and animal behavioural responses that ultimately underpins much Anthropocene unease in the face of returning wildlife. Archaeologists hold some of the knowledge that is relevant here and archaeological research can thus critically contribute to the development of viable and sustainable forms of human-wildlife coexistence in the present and future.


Contacts for the media

Shumon T. Hussain, Department of Archaeology and Heritage Studies, Aarhus University


Chris Baumann, Department of Geosciences and Geography, University of Helsinki & Department of Geosciences, University of Tübingen 


Martina Roblíčková, Anthropos Institute, Moravian Museum, Brno


Please sign in or register for FREE

If you are a registered user on Ecology & Evolution Community , please sign in