Zooarchaeology of Managed, Captive, Tame, and Domestic Birds: Shifts in Human–Avian Relationships

Abstract

In this paper, I review archaeological evidence for shifting human–avian interactions. Many species of birds, altering their behavior in response to anthropogenic niche construction, experienced an increased encounter rate with humans. Intensification of this relationship led to management and domestication of some taxa. An examination of the methods zooarchaeologists employ to study this changing interaction illustrates the limitations of evidence. Art history, architecture, historical sources, evidence based on modern distributions, and DNA analysis fill in some gaps in our knowledge. It is necessary to develop novel methods to investigate the earlier history of bird–human interactions. We also need to consider other taxa behaviorally amenable to domestication, as there was probably a diverse array of past human–bird relationships that remain unknown. Archaeologically, the relationship between people and birds is fundamental to understanding many symbolic and economic practices central to human societies. This review highlights the varied relationships between humans and birds globally allowing cross-regional examination.

Introduction

Throughout history, the relationship between humans and birds has been rich and multifaceted, shaped by necessity, respect, and ingenuity. From early hunters-for-agers relying on wildfowl for subsistence to the sacred role of birds in rituals and mythology, these creatures have held a central place in human culture. Over time, the domestication of birds, particularly chickens, ducks, and turkeys, has resulted in their status as industrial staples. A recent estimate of wild bird and poultry biomass (Bar-On et al. 2018) showed that the farmed birds comprise two and a half times the biomass of their wild relatives, demonstrating humanity’s profound influence on ecosystems. Furthermore, modern crop and livestock systems significantly affect biodiversity, and this compounds effects on wild fauna (e.g. see Smith et al. 2020). This transformation raises profound questions such how and when did the relationship between humans and birds change so dramatically?

The rise of industrial bird farming is rooted in the agricultural and technological revolutions of the 19th and 20th centuries, which transformed food production on a global scale. This is a culmination of the long history of human–bird relations, yet information about how our species began managing bird populations is inadequate. Niche construction activities with people modifying landscapes and waterscapes have influenced bird behavior for millennia (Yeomans et al. 2024). Following these initial responses to anthropogenic factors, humans often intentionally managed many bird species; selective breeding and transportation globally represent further inflection points in the scale of human–bird interaction. The geographical and temporal ranges covered by this altered relationship between humans and birds are wide, with discussions of archaeological evidence dispersed in regional publications and syntheses devoted to specific taxa. Larson and Fuller (2014) included a selection of bird species in their review article “Evolution of animal domestication” that highlighted pathways to domestication but did not evaluate the nature of the archaeological evidence. The edited volume The Evolution of Domesticated Animals (Mason 1984) remains an informative textbook discussing the evidence for many species including birds. However, publication of new evidence in the intervening years requires an updated synthesis. Here, I provide an update while examining limitations of zooarchaeological evidence in more detail and providing recommendations to improve our knowledge. Furthermore, a cross-comparative approach allows interrogation of data to attempt to understand the range of factors that resulted in variable trajectories of human–bird interaction. What commonalities are there between different regions? How important are behavioral and/or physical characteristics of the birds? Are there chronological patterns that might relate to environmental or cultural stimuli?

Human–bird relationships cover a spectrum of interactions. When species inhabit the same environment, they influence all other inhabitants of the ecosystem to a greater or lesser degree. At one end of the scale they may have had limited contact, although people likely have always noticed birds and almost certainly been in awe of their ability to soar in the skies above. Likewise, birds would have noticed humans and perhaps avoided being too close to them while others might have started to forage in novel anthropogenic niches. The other end of the spectrum includes birds intensively bred in industrial-scale farms, but there is an array of levels in between. To help define these interactions, I have included a table that shows categories of human–bird relationships (Table 1). This paper focuses on the categories of managed, captive, tamed, and domesticated birds. Birds operating in a synanthropic or commensal relationship with humans, such as ravens (Corvus corax) gaining access to prey hunted by humans during the Upper Palaeolithic Gravettian (Baumann et al. 2023), are not included. It is somewhat arbitrary to discuss only the more extensive species interactions since these relationships did not emerge in isolation. However, it is necessary to limit the range to allow coverage in a single paper. My focus is on archaeological evidence for managed, captive, tamed, and domestic birds. Birds have influenced art and culture throughout history, but detailed discussion is beyond the scope of this paper and brief mention is limited to iconographic sources when these are the first evidence for a change in interaction.

Table 1. Proposed definitions for the relationship between humans and birds (categories not mutually exclusive).

Status	Definition
Wild	Birds without any need for interaction with humans during their lifecycle but may have been hunted by people
Synanthropic	Birds that benefit from human changes to environments but can equally live in natural habitats
Commensal	Birds that have become accustomed to living in human altered environments and their survival in natural environments is reduced
Managed	Birds that are actively encouraged into an area by humans modifying ecological niches or birds bred in captivity and released usually as game for hunting
Captured	Birds captured from wild and kept confined by humans, but they would otherwise return to the wild
Tamed	Birds are genetically indistinguishable from wild animals and are often captured while young and therefore become accustomed to human presence
Domestic	Birds raised by humans who control their breeding; they are genetically different from wild conspecifics
Feral	Escaped or released domestic birds that retain traits of the domesticated species but survive without human intervention (often as commensals)

Firstly, I briefly discuss pathways to bird domestication before presenting evidence for geographical regions chronologically. This allows detailed discussion of characteristics of the species involved in, and cultural background to, shifts in human–bird interactions. Understanding the variety of past human–bird interactions provides insight into the deep connection between our species and the avian world. Furthermore, this yields knowledge that can aid bird ecology and conservation studies by illustrating long-term impacts of associations (Thomsen et al. 2024). Incorporating traditional ecological knowledge (TEK) and local ecological knowledge (LEK), ethno-ornithology is placed at the intersection of culture, birds, and ecology (Delfino 2024). Ethno-ornithological research in conjunction with an understanding of human–bird interactions over the longue durée provides awareness of the entangled relations between birds and humans. This study attempts to condense current information on their increased interaction through history allowing us to examine trends on a global scale. At the same time, I highlight the many gaps in our knowledge and methods to identify the nature of human–bird interaction. There were possibly many other close relationships between other species of birds and humans of which we are not aware. Unless we attempt to broaden our research and take an ecosystem approach, modeling relationships between species and actively investigating potential taxa that may have been more connected to humans, we may never be aware of the full range of past interspecies relations. TEK and LEK are fundamental to progressing this research by providing insight into the variety of human–bird interactions beyond the anticipated.

Pathways to Bird Domestication

The three trajectories toward domestication, defined by Zeder (2012), are the commensal, the prey, and the directed pathway. Commensal pathways result from nonhuman animals adapting to new ecological niches created by humans, thus encouraging closer contact between these species. Humans were not intentionally domesticating animals, but domestication was an evolutionary process as the animal population responded to selective pressures while adapting to a novel niche. The prey pathway probably also was incidental as hunters sought to influence the breeding of an animal population to increase availability, especially after a period of overhunting. Many later domestications follow the directed pathways as people intentionally sought to domestic species.

A global review of the evidence for changing human–avian relationships could be organized in many ways, and taxonomic ordering of the data allows discussion of behavioral traits that are partially related to bird evolutionary history. While past human societies would have created many groupings for birds that relate to physical appearance, behavior, or preferred habitats, these characteristics are influenced by evolutionary history. However, this does not facilitate discussion of regional and chronological trends. Alternatively, the review could be organized according to the pathways to domestication, and this would have merit in addressing how human–bird interactions intensified. However, the nature of the pathway remains unclear for some species; for example, according to Larson and Fuller (2014), Muscovy duck followed the prey pathway, but Zeder (2012) suggests a commensal pathway for this species. Instead of using taxonomy or grouping species by domestication pathway, I organize this review chronologically within in three broad zoogeographical regions of the globe (Wallace 1876): (1) the Palearctic, Oriental, and Ethiopian realms, (2) the Australasian realm, and (3) the Neotropical and Nearctic realms. These biogeographical realms apply less rigidly to birds as they can cross many of the natural barriers that resulted in these zoogeographic zones. Therefore, these are merged into the three areas to which most birds are endemic without human intervention.

The first broad region covers much of the Eastern Hemisphere. The Palearctic realm includes Europe, Central and Southwest Asia, Africa north of the Sahara, and Sinai; the Oriental realm includes the Indian subcontinent, Southeast Asia to lowland southern China, and Indonesia as far as Sumatra, Java, Bali, and Borneo; and the Ethiopian realm covers Africa south of the Sahara and southwestern Arabia. The second region is the Australasian realm, which includes Australia, New Guinea, and the eastern part of the Indonesian Archipelago. The third major region covers the Western Hemisphere. The Neotropical realm includes South America, Central America, the Caribbean Islands, and southern North America; the Nearctic realm covers most of North America, Greenland, and the highlands of Mexico. Presentation of the data regionally allows a comparison globally of trends in the human–avian relationships that are developed in the discussion below.

The following sections discuss the early evidence for managed, tamed, captive, or domestic relationships between humans and birds. How human–bird relationships changed subsequently and the nature of human interactions prior to the period of intensification provide some context to these shifts. Supplemental Table S1 lists species taxonomically, summarizing current knowledge on date and location of changes in the human–bird relationship to facilitate discussion of birds amenable to increased interaction with humans. Considerable research has been devoted to chicken and turkey domestication, so discussion of these species is limited but references are provided. My aim in this paper is to examine the diversity of avian species in its entirety rather than focusing on one or two species thoroughly covered in recent syntheses.

Palearctic, Oriental, and Ethiopian Realms

Swan Goose (Anser cygnoides) 5000 BC

Goose (Anser cygnoides) is currently thought to be the first species to have developed a closer than synanthropic relationship to humans in the Palearctic, Oriental, and Ethiopian realms. Our understanding of goose domestication has progressed extensively in recent years aided by genetic and new zooarchaeological evidence. DNA analysis published by Shi et al. (2006) confirms European domestic geese derive from graylag geese (see below) and Chinese geese from swan geese (Anser cygnoides). A date of 5000 BC for early management of geese at Tianluoshan in China was recently published (Eda et al. 2022). Immature goose bones and oxygen isotope analysis indicate the presence of both migrant and local populations of geese, and that breeding took place outside the current area where geese naturally breed today (Eda et al. 2022). Nitrogen isotope analysis revealed that the geese foraged in rice paddies close to the site and that the creation of a new anthropogenic niche clearly resulted in a behavioral change in these birds.

Pigeon (Columba livia) c. 4500 BC

The antecedent of the pigeons are rock doves (Columba livia), which have a long history of relationships with people. At Gorham’s Cave in Gibraltar 67,000 years ago, Neanderthals sustainably harvested rock doves over a period of 40,000 years (Blasco et al. 2014). The birds lived in large colonies, and their fast reproduction enabled exploitation for millennia. Hawes (1984) considers the pigeon to be self-domesticating as the birds were drawn to feed on human fields and settlements. Evidence in the form of dove figurines found at Tell Arpachiyah in Iraq suggests that domestication occurred by around 4500 BC. The birds featured frequently in later Assyrian reliefs (Zeuner 1963), but direct zooarchaeological evidence is absent. Ancient Egyptians kept large populations as offerings for ceremonies, and Greek statues and vases dated to the 5th century BC depict the birds held by people (Hawes 1984).

Dovecotes became common in the first millennium BC as humans began to encourage birds to roost. Pigeon towers facilitated the collection of young birds or squabs for consumption, with numerous archaeological examples including the sub-terranean dovecote in Amman hewn into the chalky limestone bedrock (Khairy and Kakish 2013). The structure was approximately 3 m in diameter and 2.8 m deep and had some 300 small niches for nesting doves; it dates between the late eight to sixth century BC (Kakish 2012). Remains of dovecotes are almost certainly under-represented as it would have been common to build them in the upper stories of structures that are unlikely to survive. Husselman (1953) provides a description of pigeon keeping in Karanis based on tax records. The site was an agricultural town in the Egyptian Faiyum that was occupied from the third century BC until the sixth century AD. A dovecote at Shivta in the Negev Desert that has been excavated dates to the Byzantine period and was probably abandoned in the middle of the sixth century AD (Ramsay et al. 2016). The excavation produced bones from 200 pigeons, many of which were almost complete skeletons. Of these bones, 24% were juvenile and 1% were neonatal (Ramsay et al. 2016). The metric data from the pigeon bones show that the birds were similar in morphology to rock pigeons although slightly smaller, perhaps reflecting Bergmann’s rule that in warmer environments animals tend to have a smaller body size to aid thermoregulation (Marom et al. 2018). Two dovecotes at Sa‘adon in the Negev Desert include one with a layer of pigeon manure and articulated pigeon skeletons with evidence of sudden destruction in the middle of the sixth century AD (Tepper et al. 2018). The tradition of dovecote building, which started in Southwest Asia, is still common in the region. It is especially prevalent at Mit Ghamr in the Egyptian Delta where hundreds of mudbrick pigeon towers line the streets.

In Europe, the medieval period witnessed widespread use of dovecotes, which were a symbol of status. Although Romans may have introduced the pigeon to Great Britain, it is more likely that it was the Normans, as there are references to pigeons and dovecotes in the Domesday Book of 1086. There was a general acceptance that only lords of manors and parish priests were entitled to build dovecotes, but this was not law until 1587 (McCann 2000). After the dissolution of the monasteries, new landowners exercised this privilege and built a large number of dovecotes. One of the sources of tension between tenants and landowners related to pigeons feeding on newly sown fields. McCann (2000) demonstrated, however, that the law allowed farmers to protect their crops and that the situation in England was not as severe as in France. Ferguson (1888, p. 413) wrote of how “swarms of hungry birds which issued from the colombiers of the great French nobles and precipitated themselves on the crops of the helpless peasants were one of the causes that promoted the French revolution.”

In the 1720s, the brown rat (Rattus norvegicus) was a stowaway on ships and found its way into mainland Britain (Armitage 1994). The ecology of the brown rat differs from that of the already present smaller black rat (Rattus rattus) with its diet of seeds and fruit presenting little danger to pigeons. From the middle of the 18th century, as the brown rat population increased, the lower levels of nest holes in dovecotes were blocked and many plastered over to create a smooth surface that brown rats were unable to climb (McCann and McCann 2006). The so-called rat ledges in earlier dovecotes were actually places for pigeons to sun themselves and rest. The only vermin to hunt from ground level prior to the brown rat were martens and polecats that could climb the square corners of dovecotes, and the corner ledges acted as deterrents (McCann and McCann 2006). It could be questioned whether the timing of the relaxation of laws allowing tenants to build their own pigeon towers and the presence of the brown rat was a coincidence. Brown rat infestation of pigeon towers may have lowered the status of pigeon keeping, meaning gentry were less inclined to ensure it was a preserve of the wealthy. Construction of later dovecotes often took place in the upper stories of other structures. This example shows how it is essential to understand the ecological relationship between multiple species in anthropogenic niches (Hulme-Beaman et al. 2016) with brown rats altering the nature of human–pigeon relationships.

The homing ability of the pigeon was used in Roman Egypt when messages started to be tied to a bird’s legs or neck. Pigeon racing is known from Palestine by about AD 200–220 (Hawes 1984), but it was after the pigeon lost its status that intensive breeding produced birds of a variety of types with many focusing on the sport of pigeon racing. There were around 150 breeds of domestic pigeon by the 19th century (Hawes 1984), and more than 800 today are bred for sport, meat, or as decorative birds. Feral pigeons are now almost everywhere on the globe and often are thought of as vermin.

Graylag Goose (Anser anser) Second or Third Millennium BC?

When European geese (Anser anser) were domesticated remains unclear. Reviews (Albarella 2005; Crawford 1984) note how in the Odyssey, Homer mentions that 20 geese were about the house eating mash out of a trough, suggesting a date going back to at least the first half of the eighth century BC. Dating to the Fifth Dynasty in Egypt (2723–2300 BC), a relief shows Anatidae, probably geese, being force-fed. This could also be Egyptian goose (see below), and other reliefs show wild animals being force-fed. Much earlier, at the site of Çatalhöyük (7500–6000 BC) in central Anatolia, Sidell and Scudder (2005, p. 118) identified the presence of graylag goose eggshell and noted “[t]he eggs could have been collected from nests of wild birds in the nearby wetland environments and/or they could represent eggs produced from birds kept on site.” A more recent study using proteomics to identify the eggshell (Demarchi et al. 2020) indicates that the eggs were dominated by Anseriformes and included numerous geese as well as ducks. Sidell and Scudder (2005) found that some of eggs were possibly hatched or from eggs with well-developed embryos, but it is not unusual for people to consume eggs at this stage of incubation (Alejandria et al. 2019). Incubation for graylag geese takes 27–28 days, and females can lay up to 12 eggs (and domestic geese even more) usually laid over successive days before incubation starts. If people were only consuming eggs with minimally developed embryos, availability would be very limited. It is interesting that, although based on a small sample size, the study by Heikkinen et al. (2015) noted that modern domestic geese in Turkey are unusually diverse, perhaps providing an indication of the longer timeframe that managed geese were present in Anatolia. DNA from 18 modern breeds had low genetic diversity, providing an indication that modern European domesticated geese descended from a narrow genetic base (Heikkinen et al. 2015). Analysis of DNA from Chinese domestic breeds demonstrated that 25 originated from swan geese but that Yili geese, with a distribution in Xinjiang province in Northwest China, originated from graylag geese (Li et al. 2011; Wen et al. 2023). DNA extracted from 122 goose bones from fourth–18th century AD sites in Russia found no evidence of European breeds crossing with Chinese domestic geese, suggesting that population mixing is a modern phenomenon (Honka et al. 2018).

Mallard (Anas platyrhynchos) 2500 BC?

Throughout prehistory, hunters have often targeted ducks, but there is minimal knowledge of how the human–bird relationship shifted to result in the situation today with domesticated birds present across the globe. Domestic ducks (Anas platyrhynchos) differ from wild ducks in that they have lost their ability to fly considerable distances because of increased size. Widespread distribution of mallards (Anas platyrhynchos), the wild ancestor of domestic ducks, makes identification of place and timing of domestication problematic. China may have been the original location with pottery figurines depicting birds at 2500 BC (Clayton 1984a). Albarella (2005) summarizes the evidence for the Roman period, arguing that the ducks may have been raised from collected wild eggs, but these tamed birds were indistinguishable from wild ones. Champagnon et al. (2023) make an interesting observation on how human alteration of environments has blurred the distinction between wild and non-wild. In southern France, for example, managed marshes are favored by released mallards instead of wild conspecifics, and gene mixing amplifies the loss of wild genotypes (Champagnon et al. 2023). Domestic mallards are now included on the Global Invasive Species Database. Lack of progress in understanding domestication of the mallard is notable, and as the genetic history is likely to be very complex, researchers have not attempted to resolve the issue with the help of DNA.

Egyptian Goose (Alopochen aegyptiaca) c. 2040 to 1782 BC

Egyptian geese (Alopochen aegyptiaca) were domesticated during the Middle Kingdom (approximately 2040 to 1782 BC), although the goose was clearly important in Egypt earlier and domestication maybe as early as the third millennium BC (Albarella 2005). From the Eighteenth Dynasty (1550/1549 to 1292 BC), there is abundant evidence of goose breeding and consumption (Boessneck 1986). At the Late Bronze Age Levantine site of Lachish, people adopted the Egyptian tradition of goose keeping and consuming birds during communal feasts (Koch 2014). The goose remains from Lachish were mostly Egyptian geese rather than graylag geese (Croft 2004). Egyptian geese also occur in Late Bronze Age levels at Tell Nebi Mend in Syria (Grigson et al. 2015), and although they can occur as a winter vagrant, humans likely introduced the species. There are no domestic birds of this species today, and its disappearance might coincide with the sixth century BC Persian conquest of Egypt (Albarella 2005).

Chicken (Gallus gallus) 1650 to 1250 BC

Archaeological evidence for domestication of red jungle fowl (Gallus gallus) and sub-sequent introduction of chickens to regions outside Southeast Asia has been the subject of considerable research. Through redating and morphologically reexamining chicken bones at the Neolithic site of Ban Non Wat in central Thailand, it was found that the earliest domestic fowl dated to 1650 to 1250 BC (Peters et al. 2022). The domestic status of the birds is based on frequency in the assemblage and presence of juvenile birds (Eda et al. 2019). It was not until the late second millennium BC that domestic fowl reached central China, and Peters et al. (2022) refute previous claims for domestic chicken in the Yellow River basin dated to 9000–6000 BC (Xiang et al. 2014). These bones were misidentified (Eda et al. 2016), and the estimated distribution of wild jungle fowl does not extend into the region today, making it unlikely that the area was a center of domestication (Peters et al. 2022).

Recent work has demonstrated that humans introduced red jungle fowl to islands east of the Wallace Line in Southeast Asia around 200 BC (Meijer et al. 2023). A program of redating bones found that chickens did not arrive in Europe until the first millennium BC (Best et al. 2022). Data from the Horn of Africa suggest that chickens arrived into the continent in the ninth–sixth centuries BC (Mwacharo et al. 2013; Woldekiros and D’Andrea 2017) but did not reach West Africa until considerably later than initially suggested. Direct dating of bones from Mogador in Morocco provided a date of ninth–12th centuries AD as opposed to the seventh century BC date originally claimed (Best et al. 2022). Chickens do not appear to have been introduced prior to the Roman period, but an indirectly dated chicken bone from the first century BC at the site of Rirha (Oueslati et al. 2020) might be earlier evidence if corroborated by direct dating. In the Americas, there is a claim for the introduction of chicken via Polynesia with a bone dated to AD 1321–1407 from the site of El Arenal-1, on the Arauco Peninsula in Chile (Storey et al. 2007).

Knowledge about the spread of domestic fowl has undergone significant revision in the last couple of years (Best et al. 2022; Peters et al. 2022), illustrating the value in detailed investigation over claims based on small numbers of finds. These works also highlighted that initially chickens were exotic species valued for their symbolic importance and later became a common food animal. An interesting question remains: when did the domestic birds became more important for their egg production? The claim that eggshell fragments, identified as domestic fowl using ZooMS methods, in Central Asia that date to the fourth century BC indicate birds were laying eggs throughout the year based on their ubiquity in samples is an optimistic interpretation (Peters et al. 2024).

Pheasant (Phasinus colchicus) c. 1300 BC

The wild distribution of pheasant (Phasinus colchicus) covers the Black Sea through the Pacific, with managed birds originating from the Phasis Valley in Colchis, Georgia, around 1300 BC according to Pliny the Elder (Blank 1984). Romans introduced pheasants into western Europe, and they were reared extensively for food in the Middle Ages. Before 1800, birds were introduced into the eastern United States, reaching the western states by the later part of the 19th century (Blank 1984). In Europe, pheasant shooting became more popular after the mid-19th century as wealthy industrialists were prepared to pay the high costs associated with producing birds in large numbers (Blank 1984). Additional game farms were established, especially in England, in the later part of the 19th century and into the 20th century (Blank 1984). Managed birds were not selectively bred as the game industry tried to keep them close to wild birds in physical and behavioral traits for the sport of hunting (Blank 1984).

Peafowl (Pavo cristatus) Before 1000 BC

The peafowl was tamed in the Indian subcontinent before 1000 BC (Nair 1974). Historical sources note introduced peafowl (Pavo cristatus) in Assyria and Babylonia (Grahame 1984), and the Phoenicians also transported peafowl to Egypt having traded with the Musziri on the eastern coast of southern India. Although it features extensively in the culture of the Indian subcontinent, it remains unclear when human management of the birds started (Nair 1974). Poole (2010) reviews the evidence for the introduction of peafowl into Britain, finding reports of three bones from Roman sites and two from Saxon sites before there was an increase at Norman sites suggesting reintroduction of the species (Sykes 2007).

Guinea Fowl (Numida meleagris) at Least 500 BC

Helmeted guinea fowl (Numida meleagris) is the only bird domesticated in sub-Saharan Africa with genomic analysis suggesting a date and location of 3500 BC to AD 700 in West Africa (Shen et al. 2021). Gifford-Gonzalez and Hanotte (2011) note how these birds hunt snakes, a favorable trait that led to humans tolerating their presence around settlements as commensals. Rock art in Wadi al-Ajal in southwestern Libya is argued to show helmeted guinea fowl (Guagnin 2015), although the purported expansion of the species into the central Sahara ideally needs zooarchaeological evidence to be irrefutable and dating is problematic. Import of guinea fowl into Europe occurred by the fifth century BC (Mongin and Plouzeau 1984), but a lack of zooarchaeological data, compounded by issues with identification (MacDonald 1992), makes understanding earlier history problematic. In De Arte Venandi cum Avibus, dated to the 1240s, Frederick II mentioned the importation of guinea fowl from the Levant to Sicily. Reanalysis of the birds remains from Ukraine found no evidence for guinea fowl (Gorobets and Rudenko 2022), which had been thought to be depicted in early medieval mosaics. In the 16th century, the birds were exported to the Caribbean (Mongin and Plouzeau 1984). A presumably feral population of domesticated birds can be still found in the foothills of the middle Atlas region of Morocco (Blench 2006), and genetic evidence from this population would be of interest.

Black Francolin (Francolinus francolinus) Fourth Century BC

The black francolin (Francolinus francolinus) has a present-day distribution from Southwest and Central Asia to Bangladesh. It is a nonmigratory species found in lowland open habitats, especially cultivated and wetland edges. A study by Forcina et al. (2015) showed these birds were introduced into the western Mediterranean before being locally hunted to extinction by the late 1800s. Absence of black francolin at palaeontological sites in Cyprus suggests an introduction to the island after initial human colonization. Historically, the species is mentioned in Greek sources dating to the fourth century BC, and a resurgence in textual mentions relates to its importance as a game bird in the medieval and Renaissance periods (Forcina et al. 2015). A 14th century bird from Sicily displayed high genetic diversity suggesting birds were introduced to the island and allowed to breed before being further traded by the Portuguese to other European countries (Forcina et al. 2015).

Cormorant (Phalacrocorax carbo) at Least 221 BC

Great cormorants (Phalacrocorax carbo) were not raised in captivity but were trained to catch fish since at least 221 BC, although other historical references date the practice to AD 25–30 and AD 759–768 (Zhong-ge 1984). Gudger (1926) reviewed historical evidence and showed the earliest illustration of cormorant fishing dating to 1665 in a drawing by Johan Nieuhof. A synthetic study of cormorant bones from archaeological sites might help reveal the origin of the practice, although similar research to identify the origin of fishing using remora showed these activities are difficult to identify archaeologically (Yeomans 2021). Remora, in a manner similar to fishing with cormorants, were captured and kept alive in the boat with a line attached to their tail. Upon spotting a turtle, the fishermen would throw the remora into the sea and the fish would quickly attach itself to the turtle, allowing both to be hauled back into the boat.

Japanese Quail (Coturnix japonica) 11th Century AD

Japanese quail (Coturnix japonica), previously believed to be a subspecies of the common quail (Coturnix coturnix), was designated as a species in 1983 since no breeding took place between the two sympatric species in the same location (Cheng et al. 2010). In China, quail fighting became popular during the Sui and Tang Dynasties (AD 581–907), and by the Song Dynasty (AD 960–1279) quails were kept in captivity and bred for fighting and entertainment (Chang et al. 2005). Historical sources show that quail were widespread and that large numbers, including fried quail, were sold in markets. These quails were domesticated probably near the Song capital of KaiFeng (Chang et al. 2005). As summarized by Chang et al. (2005, p. 110) “[o]n account of … historical background and widespread existence of quail in China, Chinese quail may well be the earlier ancestor of domestic quail than Japanese wild quail.” Domesticated quail were present in Japan by the 11th century. In modern history, the quail was (re?)imported to China from Japan in 1937, and the first quail breeding farm started in 1978, with China now the focus of the quail breeding industry (Chang et al. 2005). A review of the morphological differences between wild and domestic quail notes how the latter are between 45% and 250% larger than the wild form, which weighs only 85-110 grams, and that there are changes in behavior between wild and domestic quail (Lukanov and Pavlova 2020). Additionally, wild Japanese quail lay 5–14 eggs per clutch two or three times a year, whereas domestic quail can lay more than 250 eggs per year with egg weight increasing by 20%.

Corvids (Crows, Rooks, and Ravens) Middle Ages

Evidence of tamed corvids mainly derives from historical sources, with Serjeantson (2023) suggesting finds of old ravens remains many have been families of druids or companion birds in the Iron Age. In reviewing corvid remains from Britain and Ireland, she argued that the remains of immature birds from the early medieval period near Winchester likely were birds taken from the wild, but the most convincing evidence comes from historical sources that note how various people kept the birds as pets. A find from medieval Brussels (De Cupere and Vanhuysse 2024) of a raven whose scapulae had pathological lesions are possibly from a bird that damaged its wings by flapping in its cage.

(Serinus canaria) 15th Century AD

The canary (Serinus canaria) is native to the Canary Islands and Madeira and was domesticated at the start of the 15th century when European sailors found the originally grayish green birds attractive (Parsons 1987). Quality of song was the focus of breeding in Harz Mountains in Germany, while the Low Countries and Britain concentrated on producing show birds. The bird sellers (oiselier) engaged in increasing trade over the 16th century, and the distinctive yellow color became characteristic of the birds in the later part of the 17th century (Parsons 1987). Flemish textile workers had them as pets and brought the canaries with them as they migrated to London. In the textile towns in Lancashire and Yorkshire, the working class benefited from the profits of breeding the birds as well as gaining pleasure from their presence (Parsons 1987). The craze for canaries peaked at the end of the 19th century, and the Huguenot immigrants in Norwich made the city the most important center of canary breeding (Parsons 1987). Canaries also were taken into mines as an early warning of gas. Germany became the center for export to the United States. In the 1880s, three quarters of the occupants of the former mining village of St. Andreasberg were raising canaries for the export market (Parsons 1987).

Lovebirds (Agapornis spp.) 1860

There are nine species of lovebirds (Agapornis spp.), all of which are allopatric. They were first domesticated in 1860, but in the 1970s many color mutations were bred into the birds (Lever 1984). Originally, they were found in sub-Saharan Africa, but feral populations of Agapornis roseicollis have become established in the southwest United States and in Europe, and Agapornis personatus and hybrids populations are present in southern Europe (Mori et al. 2020).

Ostrich (Struthio camelus) 1860s

Ostrich (Struthio camelus) eggs have been exploited extensively throughout human history as containers as well as raw materials for items of ornamentation (Collins and Steele 2017; Gorzalczany and Rosen 2022). Evidence for hunting ostrich is rare with a notable exception at the site of Medjez II in northern Algeria, where ostrich bones date to the earliest phase at about 8000 BC (Merzoug 2011). A review of the evidence from the Arabian Peninsula (Potts 2001) suggested that ostriches were rarely hunted for meat, but their feathers were desired. Rock art depictions showing ostriches hunted from horseback could date as early as the last centuries BC (Potts 2001). Introduction of firearms and motorized vehicles rapidly led to the extinction of the Arabian ostrich (Struthio camelus syriacus) in 1966. Siegfreid (1984) suggests that ostriches were perhaps semidomesticated by Egyptian, Greek, and Roman civilizations with fattened birds incorporated into feasts. Egyptian and Roman women allegedly rode the birds at ceremonies, but no archaeological evidence has been found despite the historical sources. In the 19th century, a high demand for feathers was fulfilled by hunting. Successful artificial hatching first took place in 1857 in Algeria, and in 1869 an artificial incubator for ostrich eggs was invented (Deeming 1999), which lead to commercial farming in South Africa (Shanawany 1995). By 1913, the number of commercially raised ostriches increased to over 1,000,000 and spread across the globe, with Egypt, Australia, New Zealand, the United States, and Argentina adopting the industry at scale (Shanawany 1995). The market for feathers all but disappeared with the First and Second World Wars, so farms expanded into the meat and hide industries, which fueled global growth (Shanawany 1995).

Other Birds

Falconry is an example of interspecies cooperation without domestication. Birds were, until recently, collected from the wild, individually tamed, and then often released to breed after a period of time of hunting with humans (McGough 2019). It is so embedded in many human cultures that in 2010 it was added to the Representative List of the Intangible Cultural Heritage of Humanity by UNESCO (Negro 2018). But identifying early evidence of falconry is problematic archaeologically. The practice may go back deep into the history of our species; feathers and talons of birds of prey were used by Neanderthals (Finlayson et al. 2012; Morin and Laroulandie 2012). Recently, an aerophone found at the southern Levantine site of Eynan-Mallaha was shown to be capable of producing sounds similar to raptor calls (Davin et al. 2023), perhaps also reflecting the symbolic importance of the birds. Large numbers of raptors are present at many Late Pleistocene and Early Holocene sites across Southwest Asia. Dobney (2002) argued that such accumulations may relate to early falconry but pointed out that it was impossible to demonstrate the practice. Hawking is generally assumed to have been developed by nomadic pastoralists in the steppe area of Inner Asia, but archaeologically it is difficult to demonstrate (Müller 1993).

Falconry was popular in Europe by the Middle Ages, and many lines of evidence can be used to demonstrate the practice by this later period. These include falcons interred with humans, paraphernalia associated with the sport, imagery, pathology as well as species of falcon outside their natural range. To give a few examples, white gyrfalcons from Iceland (Falco rusticolus) were highly regarded and brought to central Europe. A grave of a noble, dated to the late fifth or early sixth century AD, at Quedlinburg-Bockshornschanze, contained skeletons of a female goshawk and two dogs (Müller 1993). Female birds of prey, being larger in size than the males, were ideal for falconry. There are many other examples including ship burials, and remains from cremation pyres are too numerous to mention here. The burials are of wealthy people who could afford to be interred with their prized falcons, making falconry obvious in the archaeological record. Earlier, before the practice became a noble pursuit, it was probably practiced by less wealthy individuals and therefore may be less obvious in the archaeological record.

The oldest known representations of falconry scenes are from Argos in the Peloponnesus, where two scenes dating to about AD 500 are depicted (Müller 1993). Most of the representations date from later (11th century) and include the Bayeux tapestry showing King Harold with a falcon on his left wrist (Müller 1993). Water-logged leather straps around a left tarsometatarsus of a male gyrfalcon (Falco rusticolus) or female peregrine (Falco peregrinus) were found in deposits dating to the second half of the 13th century AD at Tver in Russia; they were clearly jesses used to keep hold of the bird when on the arm (Zinoviev 2017). It also is possible to find evidence of healed fractures; since the birds were valuable and people sought knowledge of why deaths occurred, dissected bones provide evidence of falconry. Overall, there is a substantial body of evidence to inform on falconry from the Middle Ages (Prummel and Gersmann 2018). More intensive investigation of material from the Asian steppe region, applying methods that have helped identify falconry in the Middle Age, as well as investigating isotopic variability between wild and tamed hawks, may prove useful to clarify the earlier history of falconry and hawking.

The mute swan (Cygnus olor) is the only swan that was bred in captivity. Occasionally cygnets are separated and raised in pens. Records of swan ownership date back to the 12th century in England, and by the 15th century most swans in England were owned (Ogilvie 1984). The birds were kept as a symbol of status and were given as presents as well as consumed at feasts (Ogilvie 1984).

Blank (1984) notes how partridges (Perdix perdix, Alectoris rufa, and A. chukar) have been kept for a long time but only recently have they been reared in captivity. Introduced Indian chukar are bred in several states in the US, including Nevada, California, Washington, and Idaho, before being released into the wild. In Michigan, 200 gray partridges were released in 1910 (Dale 1943), and in Ohio between 1909 and 1916 around 2000 were released (Westerskov 1956). In Europe, some gray partridges were bred for release following the Second World War, but this was only a temporary activity with red-legged partridges increasingly reared in the late 20th century. More recent changes to farmland have reduced the number of red-legged partridges even in areas where they were previously common (Cabodevilla et al. 2021).

The idea that Egyptians raised sacred African ibises (Threskiornis aethiopicus) in captivity has been replaced by the interpretation that priests would have raised tamed birds in natural wetlands (Ikram 2005). A study of the mitochondrial DNA from 14 sacred African ibis from catacombs dated to about 600 BC found no evidence of inbreeding that might indicate a farmed population (Wasef et al. 2019). Ikram (2020) suggests that many priests were involved in breeding and feeding ibis flocks, and additional birds also could have been raised by villagers living in surrounding areas. More work on the nature of human–bird relations is needed to understand how the vast numbers of birds used as mummified offerings were obtained, as many species including falcons (Rowland et al. 2013) were part of these rituals.

Australasian Realm

Cassowary (Casuarius casuarius) 16,000 BC

Possible management of the southern cassowary (Casuarius casuarius) has been discussed by Douglass et al. (2021). Their study analyzed 1019 eggshell fragments from Late Pleistocene to mid-Holocene sites in the montane forests of New Guinea, and eggshell microstructural variation indicated that many eggs contained embryos in later stages of development. However, lack of cassowary bones, especially bones from juveniles, makes interpretation problematic. Furthermore, developing embryos of other birds are eaten nowadays in a dish known as balut (Alejandria et al. 2019). The eggshell fragments could derive from eggs collected from wild nests. Three radiocarbon dates from the earliest eggshell fragments date to the Last Glacial Maximum at around 16,000 BC and, if the management interpretation were correct, would be the earliest evidence for any bird. Offering support is an argument that cassowaries were translocated in prehistoric times. Heinsohn (2014) suggested that the southern cassowary may have been intentionally introduced from New Guinea to Seram Island and that the dwarf cassowary (Casuarius bennetti) was taken to New Britain in prehistory. Both are argued to have been taken as traded captives valued for their meat, feathers, and bones, but archaeologically there is no information on how and when translocation occurred.

Parrots and Perching Birds (Psittaciformes and Passeriformes) 1840s

The budgerigar (Melopsittacus undulatus) is native to the central and eastern regions of Australia, and in 1840 John Gould brought two bred birds back to England (Lever 1984). Budgerigar was introduced into France around 1845, Belgium and Netherlands about 1850, and Germany in 1855; later, at the turn of the 20th century, they were introduced into the United States and South Africa (Lever 1984). Belgium, Netherlands, and then France (especially in Toulouse) were the first to increase the scale of breeding (Lever 1984), with yellow birds developed in 1871–1872 in Belgium and blue varieties in the Netherlands in 1878 (Lever 1984). Some breeds of budgerigar are three times the size of their wild conspecifics and are unable to breed without human assistance (Sossinka 1982). The zebra finch (Taeniopygia castanotis) probably descended from Australian birds, and although the date of domestication is unknown, it was imported into Europe by the 1840s (Ellis 1984). Other finches are well suited to aviculture, including the society finch or Benalese (Lonchura striata domestica), and may have been domesticated in China (Ellis 1984). The cockatiel (Nymphicus hollandicus) is native to most of Australia and probably was domesticated slightly before 1850. The first color mutations only occurred in 1950 in the United States, after which the bird became very popular in both Europe and the United States, especially at the end of the 1960s (Lever 1984).

Emu (Dromaius novaehollandiae) 1930s

Emu (Dromaius novaehollandiae) is a Casuariiformes native to Australia. The importance of emu for Indigenous people is discussed by Raven et al. (2021) who note that the collection of eggs and bird hunting usually took place at certain times of the year. As male emus incubate eggs and raise chicks, they would be hunted only prior to breeding while females can be hunted only after egg laying. All parts of the carcass were used (Eastman 1969), and generally the relationship between emus and the Indigenous population was in ecological balance, although there is evidence that hunting may have diminished the population in the Northern Territory (Hermes 2018). Creation beliefs viewed humans as emus and emus as humans with distinction between animal and human blurred (Raven et al. 2021), and respect for emus created relatively sustainable hunting practices (Hermes 2018). Excavations have revealed many sites with remains of emu, and while bones are rare they often contain large amounts of eggshell (Cosgrove et al. 2010). When Europeans arrived, hunting emus provided fat for their lamps as well as food, and firearms rapidly reduced the population. By the 1920s emus were seen as pests and farmers were financially encouraged to hunt the birds. Although imported ostriches were being farmed in Australia by the late 19th century, farming of emus only began in the 1930s with minimal growth in the industry until the 1970s (Sales et al. 1999). The case of the emu also raises a number of interesting political and legal issues. Since the 1980s, medical and dermatological companies have been interested in emu oil for various products and have applied for patents, ignoring the long history of innovations made by Indigenous experimentation in the use of materials obtained from emus (Raven et al. 2021).

Neotropical and Nearctic Realms

Macaws (7400–7100 BC)

There is substantial evidence for captured and bred macaws across the American Southwest and Northwest Mexico where they must have been imported from the south. Bones of macaw (Ara sp.), dating to 7400–7100 BC, have been found outside the known range of any macaw species at Sabana de Bogotá, Columbia (Zorro-Luján et al. 2021). The site, Nemocón IV (Fig. 1), yielded 13 bones of what was probably a single bird directly radiocarbon dated (Zorro-Luján et al. 2021). Comparison of the bone morphology suggested that it was most similar to Ara ararauna, but Ara macao and Ara chloroptera could not be dismissed. Paleoenvironmental data from the site suggest similar conditions as today, and currently no species of macaw occurs locally in the wild (Zorro-Luján et al. 2021). The mummified head of a military macaw (Ara militaris) found in the Cueva de Avendaño rock shelter in Mexico dates to about 100 to 50 BC (Cruz et al. 2023; Gallaga Murrieta et al. 2021). Carbon isotope values of the macaw were comparable to maize-fed domesticated turkeys (Gallaga Murrieta et al. 2021). Species distribution models (Fig. 1) covering the last 2000 years demonstrate that the bird was not local (Cruz et al. 2023).

Fig. 1.

Examples of translocations of macaws with current range for blue and yellow macaws, scarlet macaws, and thick-billed parrot compared to the early sites where the species have been identified; modeled range for military macaw for 50 BC (Cruz et al. 2023).

Direct dating of scarlet macaw bones from Pueblo Bonito in Chaco Canyon in New Mexico (Fig. 1) show the presence of introduced birds as early as AD 900–975 (Plog et al. 2022; Watson et al. 2015). At the Mexican site of Paquimé (c. AD 1200–1450) 322 bones of scarlet macaws (Ara macao) were found. The range of this tropical parrot species today is at least 500 km away (Fig. 1) in a different ecotone (Somerville et al. 2010). Many of the bones were recovered from 56 adobe pens that also contained feces and fragments of eggshell. Unlike the turkey pens, the macaw pens had stone door plugs as the birds peck organic ones. The presence of a further 125 stone plugs probably derived from upper stories of pens suggest a large number of macaws were kept there (Somerville et al. 2010). The carbon isotopic evidence suggests that the macaws had a maize-based diet. The difference in the oxygen isotope values of the earlier and later fusing elements of the femur and humerus are opposite to what would be expected if the birds were imported from their natural range, and therefore Somerville et al. (2010) suggest that breeding of macaws took place at Paquimé, with the eggshell supporting this interpretation. Comparison of the DNA from five sites in Chaco Canyon and the Mimbres area in New Mexico that date to AD 900–1200 shows low diversity, suggesting that the birds descended from a limited genetic pool, probably part of an independent breeding stock exported from the natural range of the birds (George et al. 2018). Furthermore, comparison of the mitochondrial DNA of recent macaws showed that the archaeological specimens were from a specific haplotype now found in the wild macaws of Mexico and northern Guatemala. This evidence may help pinpoint the region from which the original breeding macaws were collected (George et al. 2018).

Bones of thick-billed parrot (Rhynchopsitta pachyrhyncha) have also been found at Chaco Canyon (AD 850–1150), and while not as common at the bones of scarlet macaw, Bishop (2022) notes that these were probably outside the native range for the species. The study of the aDNA from bones of Hispaniolan parrot (Amazona ventralis) suggests that Indigenous people introduced birds to Montserrat at least 2500 years ago and perhaps Grand Turk around 1250–1050 years ago (Oswald et al. 2023). The Hispaniolan parrot was native to the Bahamas and may have had a native range that included Grand Turk, but there is no evidence that the birds were native to Monsterrat, which is more than 700 km away (Oswald et al. 2023).

Considerably later, in the 1850s, scarlet macaws were introduced into San Francisco during the gold rush as they provided companionship to people often far from their families (Conrad 2022). Bones of either Aratinga sp. or Pionus sp. were found on the wreck of the El Buen Jesús y Nuestra Señora del Rosario, dated to 1622, indicating the transport of smaller parrots from the New World to Europe (Cooper and Armitage 2017). Bones of lilac-crowned amazon (Amazona finschi) dating to the 1870s were found in San Francisco, and blue-cheeked amazon (Amazona dufresniana) bones have been dated to the 1880s (Conrad 2022). Blue-fronted amazon (Amazona aestival) are known in archaeological material from Heyward-Washington House in Charleston, South Carolina, dated to the 18th and 19th centuries (Zierden et al. 2019).

Muscovy (Cairina moschata) Earlier than AD 1350

We have minimal information about domestication of the large forest duck, the Muscovy (Cairina moschata). Muscovy have a native distribution across South America, except the western and southern fringes, and their range includes the coastal strip of Central America. When the Spanish arrived at the end of the 15th century, they found people had a domesticated form of duck, probably the Muscovy (Clayton 1984b). By the middle of the 16th century, the species had been introduced into West Africa (Clayton 1984b). Feral and domesticated Muscovy ducks now have a global distribution and breed with other ducks and are often deliberately encouraged to do so by humans, since hybrids have higher egg production (Stahl 2005).

The prehistoric distribution of domesticated Muscovy duck is uncertain; depictions on clay vessels made by the Moche of northern Peru and by the Chimú empire, which succeeded the Moche, provide some indication. Two bones of Muscovy duck were found in Milagro, in the Tucson Basin of the American Southwest, dating to AD 1350 in an area modified into raised agriculture fields. Another duck bone with a similar date was recovered from Peñón del Río and also from coastal Ecuador. Of these three bones, Stahl (2005), using criteria developed to separate male and female and wild from domestic Muscovy, identified one as a large domestic male. Outside the natural range for Muscovy, these ducks were potentially traded from the initial area of domestication on the eastern side of the Andes in Peru. Sugiyama et al. (2020) present isotopic data for Muscovy ducks from the sites of Cerro Juan Díaz and Sitio Sierra in Panama (dated to 200 BC–AD 1600 and 200 BC–AD 1521, respectively). Their results show that the Muscovy (as well as crested guans and other species in the parrot family) consumed C4-based plants (maize). Both sites are characterized as villages with mixed economies and extensive agriculture, and birds may have been feeding in the fields. The isotope results of both wild and domestic Muscovy ducks suggest “that morphometric changes conventionally used as quantifiers to define domesticated species can be inadequate for resolving longterm, often household-level taming/habituation and capture practices recorded in many settlements across the New World” (Sugiyama et al. 2020, p. 11). Knowledge about domestication is summarized by Stahl (2008) who notes that possible centers of domestication may include the southern Caribbean, Paraguay, the Gran Chaco, and/or possibly the middle and lower Amazon. Domestication was aided by the Muscovy’s omnivorous diet and attraction to crop fields, and they might have helped control insects.

Turkey (Meleagris gallopavo) 1800 BC–200 BC/AD 250 in the American Southwest

Many publications summarize current knowledge of turkey (Meleagris gallopavo) domestication (see bibliography of recent literature and references in Speller [2020]). Occasional turkey remains from central Mexican highlands, Oaxaca, the Gulf Coast, and the Yucatan from outside their natural range suggest some management in the first half of the Preclassic or Formative period (1800 BC–AD 250) (Speller 2020). Thornton and Emery (2017) note how osteological morphology has been unable to distinguish between wild and precolonial domestic birds. The genetic markers for American Southwest turkeys are known, but it still not possible to separate Mesoamerican wild and domestic turkeys. As Thornton and Emery (2017, p. 331) state: “[i]n the absence of direct morphological and genetic markers for Mesoamerican turkey domestication, it is difficult to identify early examples of tamed or captive reared turkeys that appear within the natural range of the species’ wild progenitor.” Thornton and Emery (2017) discuss how other lines of evidence, such as frequency within assemblages, pathology, dietary reconstruction, and penning are not useful identifying early domesticated turkey because turkeys were used as insurance against scarcity and in feasting and ceremonies. It is therefore more complicated to recognize domestication archaeologically than it is for species that quickly became the mainstay of subsistence. Finds just south of the range of wild turkeys might be remains of traded dead animals or skins (Thornton and Emery 2017). The use of stable isotopes based on the higher carbon isotope ratios (δ13C) in managed animals is likely to be more effective, and this line of evidence has been applied in the American Southwest (Thornton and Emery 2017). Mitochondrial DNA from turkey bones and 29 coprolites demonstrates that separate domestication of turkey occurred in the precontact southwestern United States at about 200 BC (Speller et al. 2010).

Rhea (Rhea americana) 1980s

The Greater rhea (Rhea americana) has a natural range covering Argentina, Bolivia, Brazil, Paraguay, and Uruguay, and overlaps with the range of Darwin’s rhea (Rhea pennata) in Argentina. Separation of species using morphology of osteological remains is only possible on tibiotarsus and tarsometatarsus bones or based on pore density of eggshell (Abbona et al. 2019; Medina et al. 2019), although there are still problems with identification even using these methods (Navarro et al. 2020). Traditional boleadoras (weighted stones attached with leather straps) facilitated hunting of wild rhea. Darwin’s rhea (Rhea pennata) are depicted in the rock art of Cueva de las Manos dated to 7350 BC in the Patagonia region of Argentina (Aschero 2018), but no claims have been made for management of these birds prior to recent history. Since the mid-1980s, there has been an increase in the farming of rhea for feathers, meat, skin, and oil with birds exported outside their natural range (Glatz and Miao 2008). In Germany, seven birds that escaped from a farm in the early 2000s have resulted in a feral population of several hundred (Göttert and Perry 2023).

Contested Species

Cooke and Olsen (1984) report the presence of white-faced whistling duck (Dendrocygna viduata) dated to the first half of the first millennium AD in central Panama. At the site of Sitio Sierra, a partial skeleton was found as well as bones of young ducks that could not be identified to species, but Cooke and Olsen (1984) noted the possibility that these could be domesticated. The extensive agriculture and faunal remains indicative of pond/marsh conditions would have been an ideal location that attracted whistling ducks. In Mexico today, the black-bellied whistling ducks (Dendrocygna autumnalis) are kept in captivity with the eggs of wild birds incubated by domestic birds (Feekes 1991).

Wild turkey and oscillated turkey (Meleagris ocellata) have separate natural geographic ranges, but domesticated wild turkeys were introduced into the Maya region. At El Mirador, remains of at least three wild turkeys were found in sealed contexts (300 BC–AD 100) (Thornton et al. 2012). The ability of wild and ocellated turkeys to hybridize complicates identification, and although it has not been argued that ocellated turkey was domesticated, Thornton and Emery (2017) suggest it may have been reared in captivity by ancient Maya populations.

Historical accounts suggest that great-tailed grackle (Quiscalus mexicanus) was introduced from the Mexican Gulf Coast into the Valley of Mexico by Ahuitzotl (Haemig 2014). This would suggest that these birds were tamed by the Aztec around AD 1500 allowing the birds to be translocated into the new location where they flourished.

Grayson (2001) provides a discussion of extinct Antillean cave rail (Nesotrochis debooyi), a member of the Gruiformes, that was introduced at an unknown date to several of the Virgin Islands from Puerto Rico. Recent DNA evidence suggests that this genus belongs to the Sarothruridae (flufftails) (Oswald et al. 2021).

New World quails (Colinus spp.) have been traditionally kept as game birds and for meat and eggs (Kozicky 1993). Archaeologically, there is no evidence for when the practice started, but the possibility that the birds were managed in pre-Columbian contexts needs considering. Four sets of bobwhite wings dated to AD 1300 at the C. C. Witt site, an earth lodge and burial mound complex from the Smoky Hill phase Central Plains Tradition at Grandview Plaza in Geary County, Kansas, are discussed by O’Brien and Post (1988) as potentially having ritual meaning by the ancestors of the Pawnee.

Sugiyama et al. (2020) found nitrogen isotopes for crested guan (Penelope purpurascens) at Cerro Juan Díaz and Sitio Sierra in Panama and indicated that this species in the Cracidae family may also have been drawn to agricultural fields, but it is unknown if this was deliberately managed.

Discussion

The previous sections and Table S1 illustrate that the early history of human–bird intensification is obscure for many species of poultry, game, and pet birds. We often lack the zooarchaeological evidence to provide a date and location for the initial change in human–bird relations. A current timeline (Fig. 2) of the earliest evidence for the changing relationship between different species of birds and humans will need refining with additional research.

Fig. 2.

Timeline showing current evidence for initial management or domestication of avian species. Colors show the zoogeographic regions; black—Palaearctic, Oriental, and Ethiopian realms, red—Australasian realm, green—Neotropical and Nearctic realms.

There is evidence for numerous bird species that display a changing relationship with humans, but these are limited to relatively few taxonomic orders (Fig. 3), which infers behavioral traits amenable to domestication. Investigating these taxonomic groups in more detail enables discussion of the overarching characteristics of pathways to domestication. The long-term history of human interactions with ratites is often a story of domestication or decimation. Pachystruthio was a giant (about 450 kg) flightless bird found in the area between the Black Sea and the northern Caucasus in the Late Pleistocene (Zelenkov et al. 2019). Very little is known about these large birds or why they became extinct (Zelenkov et al. 2019), but overhunting is a possibility. We know the fate of the three species of Madagascar’s elephant birds (Aepyornis hildebrandti, Mullerornis modestus, Vorombe titan), which became extinct around 800–1000 BC as forests were replaced with grassland for livestock (Hansford et al. 2021). In New Zealand, nine species of moa (Dinornithiformes) were probably hunted to extinction after the arrival of the first settlers around AD 1450. If some birds survived, Europeans finished them off on their arrival (Foxon 2024). Either way, there was a short-lived timeframe when both humans and moa inhabited New Zealand. Even small ratites, such as the five species of kiwi, because of invasive animals introduced by humans, are vulnerable or near threatened. Although not a ratite and more closely related to Anseriformes, Newton’s mihirung (Genyornis newtoni) became extinct in Australia 50,000 years ago (Demarchi et al. 2022). The large flightless bird is behaviorally similar to ratites and easily overexploited by humans arriving to the continent. Recent farming of ratites is evidence for how well these species can adapt to management. With the exception of the cassowary, claims for an earlier form of management have been made only for semidomestication of ostriches in Egypt based entirely on historical evidence. Careful consideration of archaeological data for human–ratite interactions may reveal other ways that we interacted with these birds in the past.

Fig. 3.

Phylogenetic tree of bird orders (redrawn from Stiller et al. 2024) highlighting those with birds that developed a close relationship with humans. Excludes Suliformes whose placement is disputed within the Pelecaniformes or as a separate order.

Within the Galliformes order, three families in the Phasianoidae superfamily (Numididae, Odontophoridae, and Phasianidae) include managed or domesticated species. Guinea fowl may have been domesticated after being drawn to human settlements and tolerated because the birds hunted snakes. The same pattern of birds being enticed to the human niches is probable for chicken domestication, but further research on the response of wild red junglefowl to anthropogenic habitats would be valuable. Game birds were captured and released into new habitats for the sport of hunting, but genetic traits of wild birds were retained. These birds, sometimes raised in captivity, have been translocated to a wide range of environments. Peafowl and its role as a bird for ornamentation and food for banquets illustrate how Galliformes can be a status symbol of prestige. Indeed, the introduction of chickens outside their native range initially reflected the consumption of high-status foods before they became commonplace, being kept for everyday meat and later eggs. Across the globe, many species of Anatidae have been drawn to habitats modified by humans. Eventually, selective breeding led to the presence of fully domesticated birds, but archaeological evidence for when this process began is very difficult to discern. Issues in morphologically identifying duck bones to species compounds the challenge (Haruda et al. 2024; Poland 2018). There is potential for many duck species to have had a closer relationship to humans, and again, it is important to understand how Anatidae respond to anthropogenic habitat creation. The long-term history of humans and landscape is influenced by feedback on biodiversity as humans modified environments such as at Tianluoshan in China where there is evidence of the earliest management of geese (Eda et al. 2022). Such realignments between humans and the natural world altered the trajectory of human history, and the evidence reviewed here suggests that these events were more common than generally acknowledged. The mandarin duck (Aix galericulata) has an important role in Chinese, Korean, and Japanese cultures, and in recent history it was introduced into western Europe and a small population became established in North America. The wood duck (Aix sponsa) in North America was probably influenced by human settlements and habitat modifications. These species may reveal a close relationship to humans in the past, but we need further research to test this hypothesis.

For the Neoaves (the clade of all modern birds with the exception of ratites and Galloanserae), the history of interactions with humans can be summarized in three groups. 1) Rock doves were attracted to human occupation, and later their commensal nature was encouraged by the construction of dovecotes. In addition to the meat of birds, humans made use of eggs as food, feces as fertilizer, and for its chemical properties that facilitated leather production. Later on, the birds’ homing ability and plumage were developed by breeders and pigeon fanciers to exaggerate these characteristics. 2) Cormorants, falcons, and hawks were tamed by humans keen to use the hunting skills of birds. These birds are usually raised from chicks or juveniles collected from the wild and are genetically inseparable from their wild conspecifics. Behavioral changes in these birds relate to their familiarity with humans rather than genetic change. 3) Birds in the Psittaciformes and Passeriformes families were captured and sometimes bred mainly for their visual appearance and song. Feathers incorporated into cultural artifacts was a common motivation for capturing and breeding, but many people have found the companionship of the birds, as well as the physical and audible beauty, a source of pleasure. Other groups of birds are less well represented, and those, such as the Sacred African ibis and Antillean cave rail, seem to be captured wild animals.

Differences Across Zoogeographic Regions

Comparison of the nature of the human–bird interactions across the three zoogeographic regions provides interesting contrasts. In the Neotropical and Nearctic regions, many of the birds had a ritual role. Crown (2016) provides a detailed interpretation of the human–macaw relationship, arguing that ritual discard of both feathers and birds created on-going demand for the birds. She suggests that birds were sacrificed at a young age and had a greater value in death than in life. Crown (2016) also suggests that clans or societies probably owned the birds and that they were not an indicator of individual wealth accumulation. In a contextual analysis of bird interments from the American Southwest, Hill (2000) suggests that several species of birds (raptors, turkeys, macaws, and parrots) were often ritually disposed of as ceremonial waste. This represented a respectful disposal of bodies after the birds were sacrificed. Macaws were often killed at a young age, just after fledging, with uniform means of dispatching the birds followed a ritual methodology. The rainforest environments in the southern part of the continent are home to the impressive parrots and macaws that unsurprisingly inspired human culture. As these birds have a behavior amenable to living and breeding under captive conditions, it is easy to see why humans wanted to obtain these majestic birds. Analysis of the bones of macaws found that many had been plucked for feathers throughout their life (Fladeboe and Taylor 2022). Fothergill (2016) also found evidence that turkeys were plucked in the American Southwest; turkeys seem to have a dual purpose as they were also raised as a source of meat. The Muscovy duck was clearly symbolically important, featured on ceramic vessels, but zooarchaeological evidence for the nature of relationship between humans and these birds is so limited that it is impossible to make generalizations.

In North America, humans had a deep connection with many species of water birds (e.g., Sloan 2014), but there is currently no evidence that humans and these birds ever developed a closer relationship. Perhaps this relates to the behavior of birds that are more wary of humans and less likely adapt to human-altered environments, but again the belief systems need to be considered. With respect to the northern and southern parts of the continent, detailed cross-comparison to understand the completely different trajectories of human–bird interactions would be helpful. It is beyond the scope of this paper and could be targeted by scholars with firsthand knowledge of these regions and the complexities of human settlement patterns through time.

What is notable for the Australian region is the limited evidence for intensification of human–bird interactions until the arrival of European colonists. These included many botanists and zoologists who sought knowledge of the natural world by collecting specimens. The capture and breeding of birds in Australia in the 19th century was a continuation of these traditions following the practice of breeding canaries and pigeons. The same is true for the capture of lovebirds in Africa. Prior to the arrival of colonists, people in Australia had synanthropic relationships with birds, and there is extensive evidence for humans modifying the landscapes and ecosystems to increase productivity. However, this does not seem to have extended to deliberate management of the birds, or at least we do not have any evidence for the practice yet.

The relationship between humans and birds in the Palearctic, Oriental, and Ethiopian biogeographic realms reflects the deep cultural, ecological, and economic significance of avifauna across diverse landscapes. Many of the relationships intensified as humans modified the landscapes and birds were drawn to new habitats. The challenge for those of us working in this region is to define when these were deliberate strategies on the part of humans. Many birds followed the commensal pathway to domestication, and only later, with the capture of game birds, do we see clear intentionality in the domestication process. Symbolism in art and mythology also propelled shifts in human–bird interactions. For example, the striking appearance of peafowl made them symbols of beauty, power, and prosperity, and they were often associated with royalty, divinity, and prestige. The range of human–bird interactions in the Old World is remarkably diverse, reflecting the complexity of its cultures, societies, and environments, as well as the wide array of human adaptations over time. Within these geographical realms, wetlands in Southwest and Southeast Asia provided ideal conditions for early cultivation. These environments provide ideal conditions for attracting and sustaining wild bird populations, making them accessible for human communities to manage and domesticate avian species. The close interaction between humans and birds in these ecosystems facilitated the selective breeding of desirable traits. Urban environments also played an important role in the domestication and management of birds, offering unique opportunities for human–bird interactions. These settlements provided concentrated food sources that attracted wild birds to live in closer proximity to humans. Urban environments also offer protection from predators, and over time, the human control of breeding in these settings allowed for the development of traits favorable for utility or aesthetics.

Bird Behavior and Human–Bird Interactions

Understanding the behavior of birds helps us visualize how they adapted to human societies and allows us to target research into past human–bird interactions not yet revealed. This systematic review has illustrated how close human–bird interactions across the globe focus on several taxonomic groups with adaptable behavior. Zooarchaeologists have long considered behavioral characteristics of mammals in discussions of species domesticated by humans (Clutton-Brock 1992; Hemmer 1990); similar discussions of bird traits highlight other potential species to consider further. Pre-adaptations in birds, as noted by Sossinka (1982), show that many domesticated species are found in arid or semiarid regions and often consume a diet of grass or seeds, which allows them to forage in diverse habitats. Increased filial imprinting and a wider range of breeding stimuli are also likely to be important. Other possible characteristics include flock size, unpronounced territoriality, maturity rate, feeding adaptability, clutch size, and tolerance of human presence.

By reviewing species of birds that developed a close relationship to humans, even those that occurred since the industrial era, we gain insight for other species on an archaeological timescale. The recent domestication of ratites makes domestication of cassowary in the Last Glacial Maximum plausible, even if more evidence is required to support the claim. Given the limitations of zooarchaeological evidence (Table S1), it is conceivable that early management of more species took place prior to historical and iconographic evidence. Species such as the ocellated turkey, bobwhite quail, and whistling duck have behaviors suited to a close relationship with humans, but current evidence is inconclusive. Many cases I have covered in this review also highlight how many species of birds adapted to anthropogenic changes to the landscape, and archaeologists need to investigate changes to the ecosystem as a whole. Questioning how birds responded to niche construction activities by humans will improve insights into the initial stages of human–bird intensification.

Timing of Shifts in Human–Bird Interactions?

An argument has been made (e.g., Sossinka 1982) that there were three periods in human history that witnessed significant realignments of human–bird interactions. The second of these, visible in Figure 2, with a concentration of bird domestication in the first and second millennia BC maybe a product of population expansion increasing the quantity of archaeological evidence and trade translocating birds to new environments, which increased their visibility in the archaeological record. Portrayal of birds in cultural artifacts is also associated with the period. Whether this was a period that witnessed significant reorientation of human–bird relationships, or if timing reflects availability of evidence with smaller-scale management of many bird species in preceding periods, is open to debate. The last several hundred years of an increased rate of domestication is not the result of improved evidence but relates to the human control of nature since the Industrial Revolution. More research will help us understand the early history of human–bird relationship before we can test the hypothesis of three contracted periods of realignment of the relationship between humans and birds.

Allochthonous Aves

Birds have been translocated for millennia. Once human-mediated movement of species across the natural barriers that limit their geographic range has occurred, populations sometimes expand dramatically in the new habitats. Given the lack of other methods to identify early management of birds, species presence outside their natural range is one of the earliest forms of zooarchaeological evidence for changing human–bird relationships. Families of birds with more successful introductions as part of wildlife conservation efforts (Skikne et al. 2020) are generally those that were translocated in the past, supporting the notion that certain taxonomic groups are particularly amenable to these actions. Western colonists were responsible for many bird translocations during the 18th and 19th centuries that had significant impacts on ecosystems. Movement of bird populations can influence biodiversity, both for the areas where introductions occur and sometimes for the areas where the birds were taken, depending on the scale of the exploitation.

Nowadays, the pet trade is responsible for reducing the biodiversity of wild bird populations and practice, with 75–90% of birds dying before reaching owners (Peng and Broom 2021). At the ecosystem level, introduction of birds may cause eutrophication of waterbodies as additional droppings create enriched nutrient levels and alteration of plant communities through novel pollination and seed dispersal mechanisms (Martin-Albarracin et al. 2015). Introduction can also affect other bird species through hybridization (Andersen and Kahlert 2012), increased competition, predation, and disease transmission (Castle and Christensen 1990; Gottdenker et al. 2005). In one study, birds of the Anatidae and Psittacidae families were found to have the greatest impact on biodiversity (Kumschick and Nentwig 2010), possibly reflecting the adaptable nature of their behavior that suited changes in their relationship to humans in the first place. Avian influenza traces back to natural reservoirs of the virus in wild birds, particularly waterfowl, which carried the disease for centuries without showing severe symptoms. The prevalence of bird flu increased significantly with the domestication of birds, particularly poultry, because domesticated birds were more densely packed, which created an ideal environment for the virus to spread. As poultry farming became more industrialized and globalized, birds were transported, inadvertently helping the virus spread rapidly across regions and continents. Migration patterns of wild birds also spread the virus and introduced new strains to poultry populations where they could rapidly spread and mutate.

Not all impacts on biodiversity are negative, and it is important to incorporate information from modern studies to understand how birds may have influenced environments in the past. Approaches that consider ecosystems as a whole have seen a resurgence (Crabtree and Dunne 2022), facilitated by methodological (Crabtree et al. 2021) and theoretical developments (Hussain 2024; Pilaar Birch 2020). Avifauna would have been integral elements of past environments, and early translocations have important implications for ecosystem realignments.

Conclusion and Future Perspectives

Interactions, both utilitarian and symbolic, between humans and birds have been pivotal in shaping human societies over millennia. The humble pigeon has a fascinating relationship to humans and tracks the history of our own species closely (Fig. 4): from the rise of urban environments, as a status symbol of landowners during periods of prevalent inequality, as a working-class pastime during the industrial era, as war hero in the turmoil of earlier 20th century conflicts to urban pest as populations expanded and people are increasingly divorced from the natural world. If the history of our species were to be embodied in a single animal, the pigeon seems an appropriate choice. This example shows that by furthering our understanding of our changing relationship with a bird allows us to visualize the history of our species and our effects on the natural world.

Fig. 4.

a King pigeon on dovecote ledge. This breed, developed in the 1890s in America, produces large numbers of squabs and matures quickly to reach a large size. b Shape variation in pigeon breeds (Königsberg color-head tumbler, king pigeon, and an English magpie pigeon) compared to rock dove (black), redrawn based on Sossinka (1982, fig. 1). c Dovecote and d nesting boxes from Wadenhoe, UK, dated to c. 1800 and probably used to provide birds for shooting competitions.

We still have much to learn about the earlier history of our interactions with birds. Many, even the turkey and chicken, were not domesticated initially for subsistence, so we often lack characteristic indicators of domestication such as frequency increases or changes in mortality profiles. We need to focus research, guided by knowledge of potential challenges and general patterns of human–bird relationships, to search for new case studies. As Tella (2011) notes, many ancient cultures had a fascination for exotic birds, and we should expect that numerous early introductions across the world are still to be discovered. The archaeology of human–bird interactions is certainly more complex than the evidence currently suggests. More species were potentially tamed, and intensification of the human–bird relationship was likely to have been earlier than currently known. Testing these hypotheses would have major implications for archaeological research, and techniques including isotope analysis, geometric morphometrics, genetics, palaeoproteomics, and species distribution modeling could add to our understanding of human–bird interactions. This review demonstrates limitations of archaeological evidence for early bird management (Table 2). Given the difficulties archaeologists face in determining the presence of managed, tamed, and domesticated birds prior to iconographic and textual sources, we need to develop new approaches and revisit these questions, keeping in mind the possibility of earlier intensifications. Interdisciplinary projects combining genetics (McHugo et al. 2019), systematic reviews of the archaeological evidence and modeling could aid our understanding of bird communities as they responded to anthropogenic alteration of the landscape and as humans deliberately managed birds. One study modeling domesticated cattle in Neolithic Britain (Cummings and Morris 2022) illustrates how quickly herds of a domesticated animal can increase. Similar modeling approaches might be employed for bird species using ethnographic evidence as baselines of household economic practices. Hindcasting species distributions could also be compared to estimated zoogeographic ranges at different periods, allowing better evaluation of species as exotic (Franklin et al. 2015). So far, the most detailed use of this technique is for the military macaw (Cruz et al. 2023) but it has potential for a range of species.

Table 2. Limitations in zooarchaeological methods for identifying early bird management and potential developments that may improve our understanding of human–bird interactions.

Zooarchaeological evidence	Limitations	Potential developments
Species outside natural range	Estimated past ranges, trade in dried skins or dead animals outside natural range, length of time before translocation outside natural range	Greater use of species distribution models to aid reconstruction of past ranges
Genetic evidence	Preservation limitations, costs of analysis, and many studies based on limited sample sizes	Organized sample collection methodologies with results integrated with other lines of evidence
Frequency of species in assemblage	Symbolic importance of early domesticated birds can lead to low numbers being kept	Regional syntheses may aid interpretation
Sex ratio	Preservation of osteometric data, short period of medullary bone formation limits utility, challenges in interpretation	Publication of large datasets of osteometric data from modern collections
Frequency of juvenile birds	Taphonomic bias against preservation and recovery, species identification	Studies on bone growth to aid mortality profile reconstruction and compare to ethnographic evidence for household economies
Eggshell frequency and microstructure	Difficult to quantify, consumption of eggs with developing chicks means microstructure evidence is not conclusive	Recovery of eggshell in archaeological deposits and marker development for species identification using proteomics
Morphological evidence	Changes to osteology not manifested early in domestication, species level identification of closely related taxa	Combined use with proteomics or geometric morphometries to improve species identification and morphological changes in early domestication of birds
Isotopic evidence	Not yet extensively employed but method has potential for our understanding of human-bird interactions	Widespread application of techniques
Architecture	Early structures likely to be built from wood may not be preserved, upper stories unlikely preserved in archaeological record	Collaboration between zooarchaeologists and field archaeologists might improve sampling around urban areas to identify structures through the presence of eggshell and/or juvenile bird bones
Artifacts associated with keeping birds	Specialized equipment might take time to develop, preservation of paraphernalia made from organic materials	Insights from historical periods could be used to help interpret earlier archaeological evidence
Burials of humans with birds	As evidence for falconry this might be limited to higher-status individuals	Regional syntheses may aid interpretation
Pathological evidence	Difficult to interpret and present only in a limited number of cases	This may be used as supportive evidence, but it is unlikely to be convincing by itself

Problems with species-level determination hinder progress in our understanding of past human–bird interactions. Improvements in proteomic identifications helping separate closely related taxa is one potential solution (Eda et al. 2020; Richter et al. 2022; Yeomans et al. 2024). Identification of bones using geometric morphometric approaches has been successful for later periods, and the separation of domestic and wild graylag geese (Oueslati and Gruwier 2023) might help reveal subtle morphological changes in earlier populations of birds. Recent analysis of osteometric data from wild red junglefowl, captive wild red junglefowl, and domesticated chicken shows that the morphological changes should be expected early in the domestication of this species (Eda 2022). Given these promising results, geometric morphometric techniques are likely avenues for development. There are very promising results from isotopic analysis of bird bones, but these methods could be expanded and applied further. Additionally, genetic evidence has an important role but there is a need to ensure integration with other lines of evidence and that sampling protocols are adequate to address research questions.

In sum, the three broad biogeographical regions experienced different trajectories in the relationships among humans and birds. In the Old World, initial management of birds followed from birds adapting to human niches, whereas in the Neotropical regions many birds were captured from the wild for ritual and symbolic practice and exported at an early date to the southern part of the Nearctic region. Fewer birds were domesticated in the Nearctic region, but a separate domestication of the turkey is notable and the tradition of humans keeping birds for feathers was adopted. In the northern part of the Nearctic region, there is need for further research to test how birds responded to habitat modifications. In the Australasian region, the relationship between humans and birds was synanthropic until the arrival of colonists. Regardless of the region, the behavioral and physical characteristics influenced the pathway to domestication, with birds in the Galloanserae superorder (fowl) adapting to niche creation and modifiable diets facilitated domestication. In the future, research revisiting these questions should improve our knowledge on the range human–avian interactions. In this review, I have also suggested that certain environmental zones were important loci for increased human–bird interaction. Wetlands are critical ecosystems that play a vital role in supporting biodiversity, and they have fostered many human innovations including avian management. Similarly, the Neotropics support more than 3000 species of birds, many that are not found elsewhere in the world, and diversity of birdlife is likely an important factor driving complex human–avian interactions. The relationship between people and birds is fundamental to many cultures and societies throughout history; understanding the symbolic and economic practices encompassed by human–bird interactions is therefore an essential part of future archaeological research.