The origins of aquaculture

Measurement comparisons of ancient and modern carp push back the initial stages of aquaculture to 6000 bc, raising the possibility that rice paddy and fish co-culture systems are much older than previously thought. This research suggests carp were later independently domesticated twice, once in Europe and once in Asia.

Common carp Cyprinus carpio are the earliest truly domesticated fish1 and the third most productive farmed fish today2. Despite their modern value, questions surround their wild origins, the process by which they became used for aquaculture, and the date and location of subsequent domestication. The wild ancestor of the common carp was present in two regions, an eastern population in China, Laos and Vietnam, and a western population in the Black, Caspian and Aral sea drainages3,4. Writing in Nature Ecology & Evolution, Nakajima and colleagues show that the earliest evidence for carp aquaculture comes from China, comparing archaeological fish remains and modern carp raised in a traditional rice paddy field5.

In the Asian rice and fish co-culture system, common carp are raised alongside rice plants in an ecologically sustainable and mutually beneficial relationship. Modern trials of traditional methods show greater stability of rice yields and reduced need for fertilizers and pesticides compared with rice monoculture. Carp naturally eat and uproot weed species, and eat insect pests, all while providing natural fertilizer6. In turn, insects attracted to the rice provide food for the carp, the shade provided by the rice reduces temperature extremes, and the rice absorbs fish waste and reduces ammonia concentration7. Long-term sustainability and stable production of both rice and fish are therefore possible without intensive fertilizer use. Modern aquaculture will soon produce half of the world’s fish supply, with resulting environmental degradation, loss of biodiversity and non-native invasions8, but rice–fish co-culture systems continue a long tradition of sustainable farming without these problems. Historical records and archaeology show carp were raised in rice paddy fields during the first millennium bc. However, such fields predate those records by another four millennia, which led Nakajima et al. to ask when and where carp started to be raised in managed ponds. Archaeological fish bones provide one way of trying to answer this question.

Zooarchaeology is the study of past human interaction with animals, using identification and analysis of excavated bones and other surviving remains. Although biomolecular methods tend to make headlines, traditional analysis — involving species, skeletal element and size identification — remains crucial to research. Nakajima and colleagues have used elegant, easily replicable methods to establish body lengths of archaeological specimens at the Chinese Neolithic site of Jiahu dating to 6000 bc, comparing these to measurements of 301 modern carp raised in a traditional Japanese rice and fish co-culture system. Using the modern fish remains, they were able to establish that rice–fish co-culture produces a single-species concentration of smaller fish, in contrast to the variety of sizes and species found fishing wild individuals. Comparing the modern pattern to archaeological fish remains from other Chinese and Japanese sites, they were able to identify locations where some degree of management was being applied to carp.

This evidence for management raises questions of fish domestication. A domesticated animal has its breeding, food and habitat intentionally controlled by humans, often for economic benefit9. While it is relatively straightforward to control domestic mammal breeding, this process is more complicated in fish. Of those that could be considered domesticated (as opposed to animals bred in captivity with wild input8 such as bluefin tuna or European eel), most have only undergone this process very recently, for example, rainbow trout and Atlantic salmon8. Only two species have long been considered fully domesticated: common carp, and its close relative, the ornamental goldfish Carassius auratus1.

Nakajima et al. propose a novel three-stage process of carp aquaculture development (Fig. 1): wild populations in marshy environments fished during spawning (stage 1, recognized archaeologically by the presence of multiple species without evidence for water management); anthropogenic management of ditches and water levels to encourage natural spawning and to allow control of mass ‘harvesting’ of fish at similar life stages (stage 2, recognized by a single-taxon focus on similarly sized individuals, along with evidence of water management); and anthropogenic management of breeding, using spawning beds, and a more developed water management system, including use of ponds and rice paddy fields (stage 3, recognized by a single-taxon focus on similarly sized individuals, alongside evidence of sophisticated water management and breeding pond use). Stage 2 is equated with Zeder’s ‘commensal pathway’ where wild mammals are drawn into a mutually beneficial relationship with humans before true domestication9. Stage 2 is what Nakajima and colleagues identify at Jiahu around 6000 bc, where people were evidently managing ditches and ponds to control fish stocks for human use, even if animals were being replenished from wild stock.

Fig. 1

The three stages of common carp aquaculture in East Asia, showing progression from fishing for wild carp through to full domestication in managed ponds and rice paddies.

Although not overtly stated by Nakajima et al., stage 3 can be equated with actual domesticity, because breeding, food and habitat are all controlled by people. Previous studies have shown carp were kept in managed ponds in the Danube region during the Roman Empire — the start of carp aquaculture in Europe — and were subsequently domesticated, probably in the 12th to 14th centuries ad, driven in part by Christian fasting practices requiring regular avoidance of red meat10. Studies have questioned whether domestic carp were taken from Europe to China, or vice versa1,10. But given the evidence for the early origins of aquaculture reported here, it seems more likely that common carp were independently domesticated twice, once in Asia, and separately in medieval Europe. If we are discussing ‘stage 3’ domestication, we do not yet know the date of earliest Asian carp domestication. Historical and archaeological sources suggest this was by at least the first millennium bc, but because rice paddy fields date to the fifth millennium bc and Jiahu pushes aquaculture back to the sixth millennium bc, its origins may be much earlier.


  1. 1.

    Balon, E. K. J. Fish Biol. 65(Suppl. 1), 1–27 (2004).

    Article  Google Scholar 

  2. 2.

    The State of World Fisheries and Aquaculture 2018 — Meeting the Sustainable Development Goals (FAO, 2018).

  3. 3.

    Balon, E. K. Aquaculture 129, 3–48 (1995).

    Article  Google Scholar 

  4. 4.

    Chistiakov, D. A. & Voronova, N. V. Cent. Eur. J. Biol. 4, 304–312 (2009).

    Google Scholar 

  5. 5.

    Nakajima, Nat. Ecol. Evol. https://doi.org/10.1038/s41559-019-0966-3 (2019).

  6. 6.

    Xie, J. et al. Proc. Natl Acad. Sci. USA 108, E1381–E1387 (2011).

    CAS  Article  Google Scholar 

  7. 7.

    Lansing, J. S. & Kremer, J. N. Proc. Natl Acad. Sci. USA 108, 19841–19842 (2011).

    CAS  Article  Google Scholar 

  8. 8.

    Teletchea, F. & Fontaine, P. Fish Fish. 15, 181–195 (2014).

    Article  Google Scholar 

  9. 9.

    Zeder, M. J. Anthropol. Res. 68, 161–190 (2012).

    Article  Google Scholar 

  10. 10.

    Hoffmann, R. C. Guelph Ichthyol. Rev. 3, 57–85 (1995).

    Google Scholar 

Download references

Author information



Corresponding author

Correspondence to Jennifer Harland.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Harland, J. The origins of aquaculture. Nat Ecol Evol 3, 1378–1379 (2019). https://doi.org/10.1038/s41559-019-0966-3

Download citation

Further reading


Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing