top of page

Magical fishtery tour: salted-fish production around the Strait of Gibraltar

If you have a week or so free, how better to spend it than in pursuit of the Roman Empire’s favourite condiment: garum? Studying in the south of Spain for a year has had its advantages, and one of these is proximity to a frankly unbelievable number of fish-salting factories, or factorías de salazones, as they are locally known. This article aims to explore what these enigmatic ruins can teach us, whether it be about earlier Phoenician presence in the Iberian Peninsula, export and trade in the Roman world, or post-antique adaptation and abandonment of such factories.

First of all, an introduction to the theme. Roman fish sauce came in several different varieties: liquamen, muria, allex, and garum. These were produced along with salted filleted fish at fish-salting factories. According to Corcoran’s interpretations of Isidore and Pliny (1963: 204-5), liquamen was formed when smaller fish liquefied in the fish-pickling solution (salsamentum), while muria was a by-product of the liquamen-production process mixed with salt water. Meanwhile, allex was a by-product of garum production using the residue left at the bottom of the jar and was considered a cheap foodstuff. Garum itself was a luxury good made from fish intestines of multiple species of fish, with the most coveted type of garum being the garum sociorum produced in the Iberian Peninsula. This was priced at a thousand sesterces per twelve pints (Corcoran, 1963: 205). Despite the popularity of different salted-fish products, there were wildly varying opinions on it. For example, Seneca was famously derogative of garum in Ep. 95.25, writing

Quid? Illud sociorum garum, pretiosam malorum piscium saniem, non credis urere salsa tabe praecordia?

What? Don’t you believe that that garum from the provinces, the expensive guts of bad fish, inflames your stomach with its salted putrefaction?

Writers such as Martial (6.93, 7.94, 11.27.1-2) and Horace (Sat. 2.4.66) mention garum in their works with a similar ‘half-joking air of revulsion’ (Corcoran, 1963: 206). In fact, salsamentarii (workers in the fish-salting trade) were considered to be mean and of low social status, as demonstrated by the nickname Cybiosactes or ‘salt-fish dealer’ given to Emperor Vespasian by the Alexandrians for his stinginess (Lowe, 2018: 468). This was probably due to the undesirable nature of the job and the stench of the fish, which left the factories exiled to the city limits. Despite all this, salted-fish goods were popular throughout Roman society. As well as luxury garum, the cheapest fish products were available amongst the lower classes and used as food for agricultural labourers, ship crews, and the Roman army (Marzano, 2018: 441-442). Fish sauces even served a medicinal function in treating wounds (perhaps due to its high salt content) or as a laxative (Corcoran, 1963: 207).

There were multiple ways of producing fish sauce; the first involved alternately layering fish chunks, salt, herbs, and spice in a vat to ferment for 27 days, while a faster method saw a mixture of salt and fish cooked in a bronze pot. The results of both methods would then be drained and bottled for consumption. Another technique included a longer fermentation period of three months with the fish placed in a salt solution (Lowe, 2018: 469). Whichever production method was employed, the primary ingredients were salt and fish, meaning it was ideal for production to take place by the sea. It is for this reason that most of the fish-salting factories discussed in this article are located along either coast of the Strait of Gibraltar between Europe and Africa.


We start in Seville, or Roman Hispalis. Unlike the other factories discussed here, this factory is not on the south coast of Spain but a good 80 kilometres inland, in Seville’s Plaza de la Encarnación. At first glance, it seems odd for there to be a fish-salting factory here. As well as the city being very far inland, the factory, dating from the 1st century AD, is just over a kilometre away from the Guadalquivir River. However, this can be explained by the dramatic shift in the course of the river which took place over time; during the Roman period, it was a mere 350 metres away from the fish factory (Borja Barrera, 2014: 284). Fish production inland on the banks of the Guadalquivir was not such a bad idea considering that the river, known then as the Baetis, was navigable from the sea as far north as the provincial capital Córdoba (Corduba). In fact, analysis of the amphorae found in Rome’s ancient rubbish tip, Monte Testaccio, suggests that over the course of 250 years, 25 million amphorae from Hispania Baetica reached Rome. This means that to supply the city of Rome alone, at least 1400 trips had to take place down the Guadalquivir (Hidalgo, 2012). However, fish-salting factories stank (not surprising when fish were being fermented for up to three months) and this shows in the positioning of the factory in the northern industrial outskirts of the Roman city. While fulfilling important preservative purposes, salting was also used to improve flavour, as in the case of the grey mullet salted in Hispalis’ fish-salting factory, whose flavour was allegedly greatly improved after processing (Marzano, 2018: 442).


Our next stop is the city of Cádiz, known by the Romans as Gades, and even earlier as Gadir by the Phoenicians. Cádiz has a rather complicated geography and in antiquity was an archipelago, composed of at least three separate islands. Two of these have now become one due to the silting up of the canal separating them. This means that yet again, the location of the city’s surviving fish-salting factories does not make immediate sense to the modern visitor. And yes, that is multiple factories. Upon visiting one Roman fish-salting factory built over the top of a much older Carthaginian settlement, I was surprised to find another one a mere hundred metres or so away. In fact, both these factories were built close to the edge of the ancient canal separating the two islands, meaning that they were in a prime location for the delivery of fish by boat. Each can teach us different things. The first one, situated in the Yacimiento Arqueológico Gadir, dates from the first century BC and comes complete with a cistern for cleaning out the vats. This factory was subsequently abandoned around the mid-2nd century AD. The second factory, located just down the road from the first, was in use from the 1st century BC to the 5th century AD (Enclave Arqueológico: Factoría de salazones), attesting to the thriving fish trade and its longevity in the ancient world. According to Ferrer, the garum from Gades was ‘el más valorado en las mesas romanas’ (the most valued at Roman tables), and the variety even had its own name: ‘Flor de Garum’ (lit. flower/best of garum). This goes some way towards explaining the proliferation of fish-salting factories in Gades; if two have survived to the modern era, it’s reasonable to expect that there may have been many more in antiquity. The fish-processing trade was especially long-lived in Gades, being practised there by the Carthaginians long before the Romans arrived. Garum from Gades was ‘even in the fourth century B.C…. rated a delicacy in Greece’ (Corcoran, 1963: 208).


We now arrive in Bolonia, the site of the ancient Roman city of Baelo Claudia. The city was founded in the 2nd century BC, and according to Strabo (Geog. III 8), was the port from which people would sail to Tingis (Tangier), the provincial capital of Mauretania Tingitana. This makes sense upon visiting the site, as Morocco is clearly visible across the Strait. Strabo also tells us that at Baelo Claudia there were ‘trading-places and establishments for salting fish’ (Jones 1917: 15-16). Like many Roman coastal towns on the Strait of Gibraltar, the fishing industry was Baelo Claudia's main source of income. The fish-salting factories excavated there constitute one of the largest industrial sites excavated in the Iberian Peninsula and were active between the 2nd century BC and the 3rd century AD. Not only do the fish-salting vats survive, but two domus-style houses dating from the 2nd century AD have also been conserved, probably belonging to the factory owners or traders. The size of the fishing industry in Baelo Claudia can be partly attributed to the site’s natural advantages. The city is located in a cove, making for a natural harbour, while its surroundings hold countless smaller sheltered coves perfect for mooring fishing vessels. Like other sites nearby, Baelo Claudia was ideally located for catching tuna, which pass through the natural bottle neck of the Strait of Gibraltar twice a year to spawn in the Mediterranean before returning to the Atlantic. This frequent migration combined with the effects of the natural bottleneck formed by the Strait of Gibraltar made (and still make) the surrounding area lucrative for fishing.


The fish-salting vats from Ceuta, dating from the 4th to 6th centuries AD, are the only ones in this article located outside of the province of Hispania Baetica, and instead are from the North African province of Mauretania Tingitana. While these fish salting vats are no longer in situ, they are from Ceuta, or the Roman city Septum, demonstrating the extent of the fishing industry on both sides of the Strait of Gibraltar. The original location of these fish-salting vats was the Paseo de las Palmeras, a little closer to the coast than the museum where they are now kept. According to Pliny the Elder (Nat. Hist. III 1), the narrowest part of the Strait can be found around the area between ‘Abyla Africae, Europae Calpe’ (Abyla in Africa and Gibraltar in Europe), i.e., the Pillars of Hercules, which the legendary hero Hercules was said to have pulled apart in mythology. The contenders for the ancient Mons Abyla are either Monte Hacho in Ceuta, or Jebel Musa in Morocco. This locates Ceuta in one of the narrowest parts of the Strait of Gibraltar, making for very lucrative fishing with more fish confined into a smaller space. The presence of these fish-salting vats in Septum serves as a reminder of the interconnectivity of the Roman world, with similar industrial fishing practices employed on either side of the Strait of Gibraltar.


Coming back from North Africa to the other side of the Strait of Gibraltar, we find ourselves at the site of Carteia, in modern-day San Roque. This has been considered the first Roman settlement outside of Italy to be given Latin rights; this was to accommodate for the mixed children of Roman soldiers and Hispanic women. Located almost directly across the Strait of Gibraltar from Roman Septum, Carteia similarly benefitted from the fish migrations through the Strait and became a prosperous fishing settlement. Unlike the other fish-salting factories mentioned here, the vats in Carteia have remained filled in to protect them from wear. The vats’ extreme proximity to the sea means that wind-borne grains of sand pose an erosive threat to the integrity of the ruins. The fish-salting factory pictured here is only one of the fourteen which have been identified onsite, dating between the 1st and 6th centuries AD (La historia de los talleres salazoneros romanos de Carteia, al descubierto). This quantity of fish-salting factory ruins attests to the enormity of the trade in antiquity.


Malaca, or modern-day Málaga, was another previously Punic settlement taken over by the Romans and home to multiple fish-salting factories. I had actually visited Málaga to see the city’s Roman theatre, and just stumbled across this fish-salting factory in the middle of the street. Like Carteia, not just one fish-salting factory has survived, but there are three more in the basement of the Museo Picasso, in the Rector’s office of the Universidad de Málaga, and in the Museo Carmen Thyssen. The factory pictured here has been dated to the second half of the 4th century AD, which again bears testament to the strength of the fish trade despite the political instability and disruption which arose across the Roman Empire during the Crisis of the 3rd Century. According to Corcoran, the city’s name may have even been derived from a Carthaginian word meaning the ‘fish-processing place’ (Corcoran, 1963: 208).


Torrox was a bit of a strange one; the modern town is located directly on top of the ancient town of Caviclum, and its ruins are hidden around corners behind buildings or, in the case of the fish salting factory, underneath the dusty glass viewing platform pictured above. This is not the only reason for describing the town as strange; Caviclum’s fish-salting factory was recycled in the 4th or 5th centuries AD as a cemetery, with some of the fish-salting vats seemingly subdivided into individual tombs. In the picture displayed above, the hollow on the right looks to have been a fish-salting vat owing to the characteristic curvature of its corners, which made the vats easier to empty. Meanwhile, the difference between the impermeable opus signinum coating used in the right-most boundary of the second grave (which also has curved corners) compared to the material separating the two graves suggests that this may have been a vat divided into two graves. Another interesting feature of the ruins at Caviclum was the survival of two pottery kilns which, among other things, produced amphorae to hold and transport the salted-fish goods produced there.


Almuñécar, or Roman Sexi definitely had the best fish factory in my book. As can be seen from the photo (which does not even show the full scale of the site), the ruins of this fish factory are extensive with the largest surviving number of vats of any of the fish factories covered in this article. Not only does much of the production area survive, but the administrative offices and warehouse area adjacent to this zone have also been preserved. The administrative function of this section is suggested by the large quantity of coins found there, as well as an absence of the opus signinum which was traditionally used in the vats. As already seen in Gades, the factory required a water supply, and in Sexi this was provided through a vaulted pipeline which drew water from one of the city’s aqueducts. Like in Gades and Malaca, fish processing here had its origins in Phoenician fish-salting practices, with the factory itself dating from the 4th century BC, and remaining in use until the 4th century AD, a whopping eight centuries. Excavations at the Puente de Noy necropolis in Almuñécar have revealed fish plates buried as Phoenician burial goods; these were plates with a depression in the centre where it is posited that sauces such as garum would have been placed for easy dipping (de Sousa, 2019: 301). As if that were not enough to attest to the importance of the fish trade in Sexi, the city had its own mint which, during the Phoenician period, minted coins prominently featuring depictions of fish. This attests to the economic importance of the fish-processing trade for the town.

To conclude, the prominence of fish-salting factory ruins on the south coast of Spain and the north coast of Africa demonstrates the importance of the fish trade for the Roman Empire, and its impact on local livelihoods around the Strait of Gibraltar. Although there is no doubt that fish-salting was unglamorous work (hence snooty attitudes about the trade and the factories themselves being pushed as far from the city centre as possible), the archaeology makes it clear that fish-salting could make or break a Roman town, supplying jobs to fishermen, those who supplied the salt, the salsamentarii slicing and fermenting the fish, those who transported it, and even the potters and kiln masters who produced the containers. Salted fish products were widespread around the Strait even before the Romans arrived, and the fish-salting factories discussed here inherited a long-lasting legacy from the Phoenicians and Carthaginians. The salted fish story does not end there; fermented fish products are not merely a thing of the past but can still be found all over the world in products such as Worcestershire sauce, Colatura di alici, and oyster sauce. It seems that fermented fish sauces are here to stay.


Ayuntamiento de Sevilla (2020) Sala Antiquarium. Available at: (Accessed: 3 May 2022).

Borja Barrera, F. (2014) ‘Geoarqueología urbana en Sevilla’, in J. Beltrán Fortes and O. Rodríguez Gutiérrez (eds.), Sevilla arqueológica: la ciudad en época protohistórica, antigua y andalusí. Seville: Universidad de Sevilla, pp. 276-303.

Junta de Andalucía (2018) Enclave Arqueológico: Factoría de salazones. Available at: (Accessed: 4 May 2022).

Ferrer, J. F. (2014) El garum y los sabores de Gades. Available at: (Accessed: 6 May 2022).

Jones, H. L. (1933) Geography, Volume II: Books 3-5. DOI: 10.4159/DLCL.strabo-geography.1917. (Accessed: 22 May 2022).

EuropaSur (2022) La historia de los talleres salazoneros romanos de Carteia, al descubierto. Available at: (Accessed: 29 May 2022).

Gummere, R. M. (1925) Epistles, Volume III: Epistles 93-124. DOI: 10.4159/DLCL.seneca_younger-epistles.1917. (Accessed: 26 May 2022).

Lowe, B. (2018) ‘Manilius and the Logistics of Salting in the Roman World’, Journal of Maritime Archaeology, 13, pp.467-480.

Marzano, A. (2018) ‘Fish and Fishing in the Roman World’, Journal of Maritime Archaeology, 13, pp.437-447.

Hidalgo, R. (2012) The Encyclopedia of Ancient History. Available at: (Accessed: 31 May 2022).

Tasting History with Max Miller (2020) I finally made GARUM | Ancient Rome's favorite condiment. 9 June. Available at: (Accessed: 3 May 2022).

With thanks to:

Sala Antiquarium, Sevilla

Yacimiento Arqueológico Gadir, Cádiz

Enclave Arqueológico Factoría de Salazones, Cádiz

Conjunto Arqueológico de Baelo Claudia, Bolonia

Museo de la Basílica Tardorromana, Ceuta

Enclave Arqueológico de Carteia, San Roque

Yacimientos de Torrox-Costa, Torrox

Factoría de Salazones El Majuelo, Almuñécar

All images are my own.

Rene Russell is an MML with Classics undergraduate at St John’s College, Cambridge. She has a particular interest in ancient religion and in the intersection between the two halves of her degree, namely the Roman history of the Iberian Peninsula, and the ways in which the archaeological heritage of Spain has been received and preserved. She is also The Antiquarian’s Archaeology Editor.

102 views0 comments


bottom of page