Animal resources
A settled population needs a predictable supply of energy and nutrients throughout the year. Beans provide plant protein, but fishing, trapping and raising small animals may have supplemented a changing diet. The archaeological question is how to demonstrate this for each place and period.
Permanent settlement in the Aburrá Valley required crops, stored supplies and access to different environments to work together. Maize and beans provided plant protein; guinea pigs and, in Proyecto AD’s interpretation, some of the dogs mentioned by Sardella broadened the range of foods available. The 1541 chronicle describes abundant provisions, but does not allow each food’s contribution to the diet to be calculated.20, 166
Modern records identify other animals living in, or documented within, the territory: chachalacas, guans, tinamous, armadillos, agoutis and two different paca species, as well as fish from rivers and streams. Camera-trap surveys by the Área Metropolitana and Instituto Humboldt, studies in Envigado and aquatic inventories provide a verifiable basis for investigating these possibilities. Presence during those surveys does not, by itself, demonstrate pre-Hispanic capture or consumption.167, 168, 169
Fishing connects this diversity with more recent memories. A family account of sabaleta brought home around 1988 from the stretch now occupied by Parque de las Aguas places a household practice on the Aburrá River. Scientific sources on the Porce provide biological context, whose regional scope must be specified in each case.170 Understanding how much these resources contributed to ancient populations requires dated animal remains, abundance and seasonality data, and dietary analysis. Read the account and its references.
A historical hypothesis · Aburrá, 1541
Proyecto AD’s interpretation is that at least some of the dogs described by Sardella were intended for consumption. The main argument comes from the account itself: they appear alongside maize, beans and guinea pigs in a list of abundant food resources.
After describing the fighting and Indigenous weapons, Sardella explains why Robledo moved his camp to the settlement: to attend to the wounded and obtain food. He then lists the provisions found in the houses, ending with the dogs. This continuity of subject favours reading them as food resources, beyond their mere presence in a domestic setting.20
‘When the messengers sent by Gerónimo Tejelo reached the Captain and told him what had happened, he immediately set out for that place with the entire camp, for the sake of the wounded and because there was enough maize food there for more than two months. He lodged there, where in the houses, apart from what was in the fields, an immense abundance of food was found: maize and beans, which are almost like peas, and many guinea pigs, which are like rabbits except smaller and very good to eat, and many medium-sized dogs like those of Castile, except that they are mute.’
«Como los mensajeros que Gerónimo Tejelo envió llegaron donde el Capitán estaba y le dieron mandado de lo que había sucedido, se partió luego con todo el real para allá, por amor de los heridos y porque allí había comida de maíz para más de dos meses, y se aposentó en él, donde en los bohíos, sin lo que en el campo estaba, se halló mucha infinidad de comida, así de maíz como de frisoles, que casi son como alverjas, y muchos curíes, que son como conejos, salvo que son más chiquitos, que tienen muy lindo comer, y muchos perros medianos como los de Castilla, salvo que son mudos.»
The phrase ‘very good to eat’ refers directly to the guinea pigs; it is not an explicit statement that dogs were eaten. Nevertheless, this does not remove the significance of their inclusion in the same list. The food-resource hypothesis has a textual basis, although the chronicle does not describe the slaughter, preparation or consumption of these dogs. If that use is confirmed, the presence of many animals would be consistent with raising them as a meat reserve; deliberate fattening is a further possibility, not a practice described by Sardella.
The hypothesis proposes that a tendency to bark little may have been desirable when keeping dogs in domestic spaces and raising animals for food. Barking is useful for giving warning and in some forms of hunting, but it is not essential to every hunting activity: Oviedo describes Indigenous dogs used for hunting that he also characterises as mute.176 Silence therefore does not establish an exclusive function or demonstrate deliberate selection. ‘Mute’ records a European judgement; it does not settle whether those dogs were unable to bark or simply vocalised little.
When European expeditions traveled through the Antilles, Mesoamerica, Central America, Tierra Firme, and the Andes between the late 15th and 17th centuries, chroniclers repeatedly recorded with amazement the presence of Indigenous domestic canids (Canis lupus familiaris) that they called “mute dogs” (perros mudos) or gozques because they did not bark like European dogs, emitting instead only low whimpers, snorts, or muffled growls [20][176][241].
In the Aburrá Valley, Juan Bautista Sardella's account (1541) of the entry of Jorge Robledo's expedition explicitly documents the presence of “many medium-sized dogs like those of Castile, except that they are mute” inside the Indigenous bohíos, listed alongside maize, beans, and guinea pigs [20]. Sixteenth-century chronicles from other regions of the Americas—especially those by Gonzalo Fernández de Oviedo [176], Friar Bartolomé de las Casas [241], Friar Diego de Landa [242], Friar Bernardino de Sahagún [243], Inca Garcilaso de la Vega [244], and Father Bernabé Cobo [245]—demonstrate that native dogs fulfilled simultaneous and complementary functions in pre-Hispanic societies: hunting auxiliaries through scent tracking, domestic companions, and food or ceremonial reserves [174][175].
“All these dogs, here on this and the other islands, were mute, and even if beaten or killed, they did not know how to bark; some yelp or whimper softly when harmed.”
«Eran todos estos perros, aquí en esta e las otras islas, mudos, e aunque los apaleasen ni los matasen, no sabían ladrar; algunos gañen o gimen baxo cuando les hacen mal.»
“They had some little dogs like those we call lapdogs, mute, which did not bark but growled, and these served only to be eaten, and with them they also hunted hutias...”
«Tenían unos perrillos chequitos como los que decimos de falda, mudos, que no ladraban sino gruñían, y éstos no servían sino para los comer, y también con ellos cazaban las hutías...»
“...where in the bohíos, besides what was in the fields, an immense amount of food was found, both maize and beans, which are almost like peas, and many guinea pigs, which are like rabbits, except that they are smaller and very good to eat, and many medium-sized dogs like those of Castile, except that they are mute.”
«...donde en los bohíos, sin lo que en el campo estaba, se halló mucha infinidad de comida, así de maíz como de frisoles, que casi son como alverjas, y muchos curíes, que son como conejos, salvo que son más chiquitos, que tienen muy lindo comer, y muchos perros medianos como los de Castilla, salvo que son mudos.»
Present-day hairless dogs from Mexico and Peru provide a visual reference for imagining one possibility. The hypothesis suggests that hairlessness may have made culinary preparation easier and could therefore have been favoured in breeding. Such a preference has not yet been documented for Aburrá. Consumption did not require hairless animals either: Oviedo records American dogs with different coats and, when describing their preparation as food in Nicaragua, mentions removing the hair.176 The Peruvian museum also notes that the hairless dog was only one of the pre-Hispanic canine forms.177
The image brings both possibilities together without identifying a local breed. The genetic history of dogs in the Americas reveals profound changes after colonisation, so present-day populations are not a direct portrait of those described in 1541.178 Use as food could coexist with companionship, hunting or ceremonial roles; Andean contexts also show diverse relationships between dogs and people.179
The expressions ‘enough maize food for more than two months’ and ‘an immense abundance of food’, together with the beans, guinea pigs and resources in the fields, constitute indications of food availability and storage capacity. They support the case for a material basis conducive to food-related well-being at the moment described: a community able to produce, gather and store provisions. Raising dogs for consumption is a coherent hypothesis within that economy.
The abundance encountered by an invading expedition does not, by itself, reveal how food was distributed among households, how long stores lasted in other seasons, or the health of the population as a whole. Food-related well-being is used here as a limited interpretation of availability, not a general diagnosis. The arrival of troops and their appropriation of supplies may have disrupted precisely the conditions glimpsed in the account.
Read as a whole, the chronicle offers three demographic clues that complement the food-resource hypothesis:20
The narrator estimates two successive gatherings during the fighting. They suggest a capacity to mobilise people and resist. The figures must not be added together: the same people may have participated, and the numbers do not come from a census.
The occupation of a settlement and the arrival of people from neighbouring villages point to a territory inhabited by several communities, with connections between settlements. They do not allow population density per square kilometre to be calculated.
The provisions and the expedition’s stay are consistent with substantial productive capacity. The twenty foot soldiers and twelve horsemen belonged to Tejelo’s advance party, not to Robledo’s entire camp.
The defensible conclusion is that the episode describes a settled population, connected to neighbouring communities and equipped with resources to sustain itself and mobilise people. The abundance of dogs fits that household economy and strengthens the case for investigating their use as food. The text does not establish a total population or a growth rate. The number of households, the expedition’s full size, ration quantities and the proportion of inhabitants mobilised are unknown.
The end of the account records pursuit, captivity and episodes of suicide. These indicate a violent rupture in local life, but are insufficient to quantify its demographic impact. Religious explanations or fear of beards attributed to Indigenous people belong to the colonial narrative framework and should not be adopted as sufficient explanations for those deaths.
Juan Bautista Sardella’s account of the conquistadors’ entry into the Valley mentions maize, beans and ‘many guinea pigs’.20 This is direct historical evidence for the sixteenth century, but must not automatically be projected across all preceding millennia without archaeological contexts.
Forests and hillsides of the Valley
Modern records identify animals that may have been among the food resources available to communities. This selection brings together documented presence in the Aburrá Valley, accounts of food use and species worth investigating. Each entry distinguishes established information from open questions. Information from other regions is explicitly identified.
The survey by the Área Metropolitana del Valle de Aburrá, conducted with the Instituto de Investigación de Recursos Biológicos Alexander von Humboldt, documents 3,568 records of 82 species using 65 cameras installed across the ten municipalities between December 2021 and October 2023. Version 1.1 of the dataset was published in 2024. Locations were selected on the basis of earlier wildlife reports; this sampling does not establish the abundance of each species.167
Nothocercus julius
Envigado camera trapping, 2022; 2023 SILAP guide, table 3.168
Broadens the investigation of ground-dwelling birds. The source establishes presence, not local pre-Hispanic consumption.
Crypturellus soui
Regional list reproduced by Muñetón Jácome (ITM, 2023), appendix 1, drawing on the Valley’s photographic guide.183
A regional bibliographic reference without a date for an individual observation; this species belongs to the tinamou family.
Odontophorus hyperythrus
Camera traps in Envigado, 2022; table 3 of the 2023 guide.168
A wood-quail in the family Odontophoridae, distinct from tinamous. Included as a candidate for testing the food-resource hypothesis.
Cuniculus taczanowskii
Recorded by camera traps in Envigado’s SILAP in 2022.168
Corantioquia identifies hunting for meat among the pressures on this species.186
Dark, long, dense fur; a robust body, tiny tail and pale spots in rows along the flanks. Reference: Parque Nacional Cajas guide, p. 139.187
Cuniculus paca
Included in the metropolitan camera-trap survey published in 2024, with records in the dataset extending to 2023.167
A different species from Cuniculus taczanowskii, the mountain paca illustrated above. Their records must not be merged under a single common name.
Shorter, reddish-brown fur, white spots in rows and raised hindquarters. Reference: CONABIO CITES guide, pp. 363–364.188
Syntheosciurus granatensis
Recorded in the Aburrá Valley camera-trap survey published in 2024.167
The name used in that inventory is retained; other sources list it as Sciurus granatensis. Possible local use as food remains to be tested.
The river and its tributaries
Basin inventories and the Convenio BIO provide records of native fish. Research on the Porce’s ecosystem services helps examine their possible role as food, without treating every catchable fish as a species whose consumption has been demonstrated.169, 189, 190
The illustrations were reviewed against identified photographs, plates and descriptions in the cited sources. They are AI recreations: useful for comparing forms, but not substitutes for taxonomic identification. Names from the POMCA are retained as recorded; uncertainties are noted in each entry.
Bryconamericus caucanus
Capture recorded in the Aburrá basin characterisation (POMCA, 2016), table 458, p. 1388.169
Elongated, silvery body, yellowish tail base and a central dark patch. The revision retains these general features; it does not validate scale or fin-ray counts. Reference: Porce guide, p. 32.170
Hyphessobrycon inconstans
Capture recorded in the Aburrá basin characterisation (POMCA, 2016), table 458, p. 1388.169
Form reviewed against figure 4 of the original 1907 description: a deep body, long anal fin and a tail spot that does not extend onto the middle fin rays. Colour is interpretative: the source is a monochrome plate, not a photograph of a local specimen.192
The name follows the POMCA. The taxonomic catalogue notes uncertainty about type locality and distribution; the inventory specimens warrant re-examination.193
Caquetaia umbrifera
Capture recorded in the Aburrá basin characterisation (POMCA, 2016), table 458, p. 1388.169
The POMCA lists it under the name Kronoheros umbrifer.
Andinoacara latifrons
Capture recorded in the Aburrá basin characterisation (POMCA, 2016), table 458, p. 1388.169
Identification unresolved
Trichomycterus medellinensis
Name published by Posada (1909), included in the 2020 update of the Colombian checklist.195 EPM’s 2023 news item mentions it among the Convenio BIO’s findings.190
However, the 2025 field guide by the Grupo de Ictiología and collaborators identifies the illustrated small catfish as Trichomycterus caliensis (figure 17). This discrepancy requires the record to be reviewed; it does not establish that the two names are synonyms.194
Species-specific illustration pending. No image with a verifiable identification of T. medellinensis was found. The generic recreation was removed to avoid assigning another species’ appearance to it.
One species, several memories of water
Brycon henni
The EPM and Universidad de Antioquia guide to fish of the Porce canyon records this species in the river, streams and reservoirs of the study area. This is regional evidence, whose scope must not automatically be extended to every stretch of the Aburrá.170
There is an explicit reference for the northern stretch of the Aburrá River. In February 2014, El Colombiano reported a statement by Víctor Vélez Bedoya, Wildlife Coordinator at the Área Metropolitana, about sabaleta upstream of the stretch known as the Porce. The article does not identify a stream or sampling point in Barbosa.196
Head proportions, the patch on the gill cover, the red above the eye and the central tail pigmentation were reviewed. View the illustrated reference, p. 38.
Family memory · around 1988
My father and grandfather fished in the Aburrá River when I was about eight years old and brought sabaleta home.
Mi padre y mi abuelo pescaban en el río Aburrá cuando yo tenía unos ocho años y llevaban sabaletas a la casa.
The participant dates this memory to around 1988, on the stretch of river beside what is now Parque de las Aguas. Family account shared with Proyecto AD on 7 October 2026; wording standardised from the participant’s messages and clarifications. Translated from Spanish.
This is a situated oral memory of fishing for household use. ‘Sabaleta’ is the remembered name: no specimens or photographs are preserved here to verify its biological identification. The date refers to the narrated experience, not to an observation of fishing in 2026.
The video Pesca de Sabaleta y su conservación, published by ImportaExpertos on 13 July 2021, presents a conversation with Juan Diego Monsalve. Between 42:51 and 43:17 the discussion addresses sabaleta fishing near Medellín.197 The review used the automatic transcript, which contains errors in place names; without checking the audio, the passage is not assigned to a particular tributary in Barbosa.
Visita Concepción provides information on local fishing and mentions sabaleta in the Porce system. The reference to the road towards Barbosa appears in the caption of an image of Uvita; it does not, by itself, establish a record in an Aburrá tributary in Barbosa. It is retained as local public-information context, separate from sources used for scientific identification.198
In ‘Ecología de los caribes’ (1995), Priscilla Burcher de Uribe relates fish availability to the size and distribution of settlements. On page 92, she argues that pressure on resources encouraged settlements to separate, and links warfare to protein supply. Drawing on Daniel Gross, she uses an average requirement of 50 grams a day and states that ‘there can be no settlements of more than two thousand people’ among groups dependent on fishing. Her approach invites questions about how populations were fed; its figures and explanations require testing in specific contexts.199
The FAO/WHO/UNU consultation sets a safe protein intake level for healthy adults of 0.83 grams per kilogram of body weight per day: approximately 50 grams for a person weighing 60 kilograms. This reference is intended to meet the requirements of almost all healthy adults, not to define an identical minimum for every individual. Age, weight, pregnancy and breastfeeding affect requirements.200 WHO recognises animal and plant sources and notes that protein digestibility and quality also matter.201
Fifty grams of protein are not the same as fifty grams of meat, nor must all that protein come from fish. As a hypothetical exercise, 2,000 people at 50 grams each would require a total of 100 kilograms of protein a day. If half came from other foods, fish would need to supply 50 kilograms of protein. This calculation illustrates why knowing the diet matters: it does not represent an actual Aburrá population or allow nutritional requirements to be converted into kilometres of river without measuring sustainable catches, losses and edible portions.
Overexploitation can reduce available resources. That risk does not, however, imply a universal maximum number of inhabitants per settlement. Environmental productivity, seasonality, capture techniques, preservation, exchange and access rules alter the possible supply. An extensive floodplain and an Andean stream do not necessarily offer the same resources per kilometre. Village size, regional population and population density per unit area are also different measures.
An archaeological study published in 2024 documents fish-trapping canals and ponds at Crooked Tree, Belize. Its authors estimate an annual supply for around 3,700 to more than 25,000 people, depending on yields, assuming consumption of 40 kilograms of fish per person per year and simultaneous operation of the facilities. These are estimates, not a census; nor do they represent all the protein needed. The study does not demonstrate a single settlement of that size, but challenges the idea of a fixed fishing capacity independent of technology and landscape.202
Contemporary evidence also shows the effects of management: a study of 83 Amazonian lakes found averages of around 305 arapaima in lakes protected by communities, compared with nine in open-access lakes. These are numbers of fish, not people fed, but they show that the organisation of resource use can substantially alter availability.203 Likewise, protein needs alone are insufficient to explain warfare or cannibalism: that explanation requires independent evidence for each society.
Sardella describes ‘enough maize food for more than two months’ and ‘an immense abundance of food’, listing beans, guinea pigs and dogs. The passage supports substantial availability of stored food at the time of the conquest; it does not measure rations or each food’s nutritional contribution.20 Maize and beans both provide protein, and their amino-acid profiles complement one another, as the FAO notes. In sufficient quantities and within a diet that meets energy needs, this combination can contribute substantially to protein supply.166
Guinea pigs were another food source explicitly mentioned by the chronicler. Proyecto AD argues that at least some dogs were also intended for consumption, because they appear in the list of provisions. This hypothesis has textual support, although Sardella does not describe their preparation or consumption. Fishing may have supplemented the diet, but the passage does not mention it, and species recorded in the basin today do not allow the quantity of fish eaten in 1541 to be calculated. We do not know what proportion of protein came from each resource, or how often animals were eaten.
The strongest hypothesis is a diversified food economy, able to accumulate reserves and combine plant and animal protein. This makes it plausible that stable communities could be sustained and inhabitants mobilised from several settlements. The abundance described by Sardella is a positive indication of food-related well-being, but does not demonstrate that every person received 50 grams of protein daily throughout the year, or allow the Valley’s population to be calculated. Testing this requires local data on production, storage, plant and animal remains, diet and health. Burcher’s enduring question is how resources and inhabitants were balanced; answering it for Aburrá requires studying the whole diet and its distribution, without imposing a ceiling of 2,000 people in advance.
Vancouver referencing: numbering follows first appearance across the project. Proyecto AD bibliography.
