Production System and Regime · Tilapia & Catfish

Production Systems · Pillar 04

The way you grow the fish depends on your capital and your site, and you must decide before the first fingerling goes in. Cages on Lake Volta give the highest volume but carry the highest disease and water-quality exposure and depend on lake access most entrants do not have. Ponds and tanks need less water and a different kind of investment, and they suit different entrants. Choosing the wrong system is a common and expensive mistake.

A cluster of square floating tilapia cages on the open blue water of Lake Volta, Ghana, a wooden walkway between the cage frames, a farmer in a small boat, bright morning light, distant green shoreline

Site and capital first

Pick the system for your site and your money.

5 to 6 months

to a table fish of about 350 g³

80 to 90%

of Ghana’s farmed fish comes from Lake Volta, almost all from cages¹²

10 to 80%

of stocked fish killed in outbreaks on lake farms²

🐟 Three production systems 🐟

There are three main ways to farm fish in Ghana. Cage culture keeps fish in net enclosures in open water, almost all of it on Lake Volta, and it produces most of the fish. Lake Volta supplies about 80 to 90 percent of Ghana’s farmed fish, almost all of it from cages.¹² Pond culture digs or builds an earthen basin and holds its own water; it needs land and a water source but no lake access, and it is an easier start for a smallholder inland. Tanks and recirculating systems (RAS) grow fish in built tanks that clean and reuse their water; they need the least water and the most capital and skill, and they sit near cities where the fish sells fresh. A common mistake is to copy the cage model because it is the biggest, without the lake site, the capital or the tolerance for disease losses it needs.

Figure 10 Cage, pond and tank systems compared on water access, capital, disease exposure and volume

Cage culture on Lake Volta dominates by volume, but ponds and tanks demand less water access and suit a different entrant. The choice depends on capital and site.

Source: Banini et al.; Kassah; FAO.¹³⁴

Pond

An earthen pond on a Ghanaian smallholding. It needs less water access and capital than a lake cage.

Round concrete and blue plastic catfish tanks in a covered Ghanaian yard, dense dark catfish churning the water surface, a worker scattering feed pellets by hand, natural daylight, documentary style

Tank / RAS

Tanks and RAS need the least water and the most capital and skill, near cities where the fish sells fresh.

Table 5: The three production systems as a capital-and-site decision
System	Water need	Indicative capital	Disease exposure	Volume	Who it suits
Cage (Lake Volta)	open lake, flowing water	high (cages, feed, boat)	highest, as shared open water spreads disease fast	highest in total	sited, financed operators near the lake
Pond (earthen)	land plus a modest water source	low to medium	lower, as water can be managed and isolated	low to medium	inland smallholders and first-timers
Tank / RAS	least water, recirculated	highest (pumps, filters, power)	controllable but fails hard if the system stops	medium, intensive	skilled, capitalised, near-city entrants

Source: Kassah; Banini et al.; FAO for tank and RAS systems.³¹⁴

🔍 What this means

Choose your system based on whether you have secure access to lake water and how much capital you can afford to lose. Cages give the most volume but need a lake site and carry the most disease risk. Ponds are a low-capital, lower-risk start for someone with land and water but no lake. Tank and RAS are for the well-funded, skilled entrant near a city who wants control and can survive a pump failure. Match the system to your site and budget first, and plan everything else around that choice.

🐟 Why cages dominate, and what the concentration costs 🐟

Cage culture dominates the sector. The 2024 to 2028 national plan counts about 10,102 cages and 4,925 ponds, and says growth relies heavily on the large-scale cage subsector while small-scale ponds lag far behind.⁵ This concentration is the sector’s biggest structural weakness. Almost every farmed fish in Ghana grows in the same open water, so a disease, a pollution event or a dam spillage can hit the whole industry at once. The 2018 virus crash, outbreaks that still kill 10 to 80 percent of stocked fish on lake farms, the cages abandoned since, and the damage the 2023 Akosombo dam spillage did to fish farms all show the cost of that concentration.²⁶¹ Because nearly every farm uses cages on the same water, cages are the most exposed to shared risks.

A Ghanaian fish farmer kneeling on a wooden cage walkway on Lake Volta, lifting the edge of a net cage to inspect tilapia, calm water reflecting the sky, documentary style, natural light

One lake, shared risk

Almost every farmed fish grows in the same open water, so one event hits the whole industry at once.

~10,102

cages counted in the 2024 to 2028 national plan⁵

4,925

ponds, which lag far behind the cage subsector⁵

📡 Market signal: the crowded lake is also the opening

The heavy reliance on one lake also creates an opening. The sector depends heavily on Lake Volta cages, so the national plan wants more pond and tank capacity away from the lake and better-sited cages on it. An entrant who can farm profitably away from the crowded lake avoids the shared-water risk and adds capacity the sector lacks.

🐟 Daily farm management decides survival and cost 🐟

Figure 11 The tilapia grow-out cycle, stage by stage from stocking to harvest

Survival and cost are set at stage two and defended every day at stage four.

Source: Kassah; Agyakwah et al.³⁷

50 to 80

fish a cubic metre, the recommended cage stocking³

1.5 to 2.0

cage tilapia feed conversion on 28 to 45 percent-protein pellets³⁷

Whichever system you pick, the day-to-day regime is what turns a stocked cage or pond into a profit or a loss. Three practices matter most. First, stocking density. The recommended stocking is 50 to 80 fingerlings a cubic metre, each above about 10 grams, and fish reach the standard table size of about 350 grams in five to six months.³ Actual densities on lake farms range from 7 to more than 130 fish a cubic metre.² Crowding lowers oxygen and raises stress, although a 2024 survey of lake farms found no clear link between stocking density and disease deaths, because the pathogens circulate lake-wide.² Second, feeding to a target feed conversion ratio (FCR). Cage tilapia has an FCR of about 1.5 to 2.0 on floating pellets of about 28 to 45 percent protein by fish size.³⁷ Feed is about three quarters of the cost of a small-scale cage,⁸ so any feed wasted adds straight to cost. Feeding little and often to keep the FCR low is the most important habit. Third, water management. Most farmers name poor water quality as a cause of outbreaks, yet about 73 percent of farms do not monitor it.² An ordinary site can do well if these three are right. If they are wrong, even the best fingerlings will not save the batch.

🔍 What this means

Survival and cost depend on stocking and on daily care after that. Sensible density, careful feeding to a low FCR and active water management cost less than any input you can buy, and they matter more than the strain. A small, well-run unit can earn more than a bigger, badly run one, because the operator's skill affects results more than the equipment does.

🐟 Where to farm: use the regional map as a shortlist 🐟

Choosing the site is a large part of the production decision. Ghana’s farmed fish is heavily clustered: the Eastern, Greater Accra and Volta Regions together produced about 82 percent of Ghana’s farmed fish in 2021.⁹ Those three regions lead because they have Lake Volta water for cages and are close to the Accra market for ponds and tanks that sell fresh fish. For a newcomer, this has pros and cons. Farming where the cluster already is means tested water, feed dealers, hatcheries and buyers on hand, but also the most crowding and the highest shared-disease risk. Farming outside it means thinner support and logistics, but cleaner water, less competition and a chance to serve a regional market that local fresh fish currently cannot reach.

Figure 12 Where the fish is farmed: the Eastern, Greater Accra and Volta Regions made about four-fifths of national output in 2021

Farmed fish clusters on and around Lake Volta, in the Eastern, Greater Accra and Volta Regions, which together made about four-fifths of national output in 2021.

Sources: Mapedza et al.; Ghana National Aquaculture Development Plan.⁹⁵

🎯 Opportunity

01

Pond or tank grow-out near a secondary city, away from the crowded lake: cleaner water, lower disease risk, and a fresh-fish market that imports and distant cages struggle to serve.

02

A better-sited cage within the cluster, chosen for flow, depth and isolation, which helps the fish survive a hot, still week.

03

Site-selection and set-up advice. Many entrants take whatever water is available, and a service that matches system to site prevents one of the most expensive early mistakes.

🐟 Lighter stocking can earn more 🐟

Figure 13 Profit margin by fingerling size and stocking density in a Ghana pond trial

Small 2-gram fingerlings out-earned 10-gram fingerlings at both densities, and the heavily stocked 10-gram ponds earned least.¹⁰

Source: Addo et al.¹⁰

A pond trial at the CSIR-Water Research Institute station in Akosombo tested two fingerling sizes and two densities over 154 days. When bigger fish earned a premium, the most profitable option stocked small 2-gram fingerlings lightly, at about 3 fish a square metre, for a profit margin of 38 to 76 percent on variable costs, depending on the premium paid for bigger fish. Small fingerlings stocked at 6 a square metre still returned about 50 percent, and at a flat price they made the most profit; only 10-gram fingerlings at 6 a square metre fell to about 9 percent.¹⁰ Doubling density cut survival from 90 to about 83 percent and slowed growth, but feed conversion did not differ between treatments, and fingerling size mattered as much as density.¹⁰ For cages the evidence is weaker: a 2024 survey of lake farms found no clear link between stocking density and disease deaths.² The right density balances how many fish the water can carry healthily against how much each fish is worth at harvest. It is usually well below the most a cage or pond can hold.

38 to 76%

margin on variable costs, 2-gram fingerlings at about 3 fish a square metre¹⁰

~9%

10-gram fingerlings stocked at 6 fish a square metre¹⁰

🔍 What this means

Stocking density and fingerling size together affect profit. In the Akosombo trial the best lightly stocked pond returned up to 76 percent and a dense pond of small fingerlings about 50 percent, while dense stocking of larger fingerlings returned about 9 percent. This means an entrant does not need to fill every cubic metre to make money, and overstocking can cut the margin sharply. Stock for healthy growth and harvest weight, then set the number from the margins.

🐟 Managing the water from a phone 🐟

Keeping the water right is a daily job that an affordable digital tool can now help with. AquaMet Technologies, a Ghanaian start-up, sells a solar-powered water-quality probe that sends SMS alerts. The company reported reaching more than 145 farmers by June 2025 and says its tools cut feed waste and fish mortality by up to 25 percent;¹¹ users report yields 10 to 15 percent higher.¹² Poor water and low oxygen are ranked as the top cause of loss, so a probe that texts an alert before an oxygen crash protects against a major risk the farmer can control. Solar sensors with text alerts, feeding-record apps and mobile-money input finance are within reach of smallholders now. Automatic feeders, camera disease detection and drone siting remain tools for commercial-scale farms.

A technician on a Lake Volta cage walkway lowering a handheld dissolved-oxygen meter on a cable into the water beside a net cage, reading the display, early morning light on calm water, documentary style

Sensing beats guessing

A probe that texts an alert before an oxygen crash protects against a major risk the farmer can control.

145+

farmers reached by AquaMet’s solar probe and SMS alerts by June 2025¹¹

10 to 15%

higher yields reported by early users¹²

📡 Market signal: sensing beats guessing

The production edge over the next few years is likely to come from cheaper sensing and faster alerts on water and feeding. An entrant who builds or resells this monitoring, or uses it well on their own cages, protects survival, which decides the result of each cycle. Pillar 8 covers which tools are practical now and which suit only commercial farms.

🐟 The risks that sit inside the production-system choice 🐟

Overstocking the cage

HIGH

What it is

Packing more fingerlings into a cage or pond to chase volume, on the belief that more fish means more money. Crowding drops dissolved oxygen, raises stress and disease, and slows growth, giving a weaker, later harvest.

Evidence

In a Ghana pond trial, 10-gram fingerlings stocked at 6 fish a square metre returned about 9 percent against 38 to 76 percent for 2-gram fingerlings at 3 a square metre, and doubling density cut survival from 90 to about 83 percent.¹⁰ Lake cage farms stock anywhere from 7 to more than 130 fish a cubic metre.²

Who it hits

Volume-chasing entrants and anyone stocking to the cage's physical limit rather than its healthy carrying capacity.

How to manage it, and the opening

Stock within the recommended 50 to 80 fish a cubic metre, feed to a tight FCR, and monitor oxygen and mortality daily. The opening is careful, lightly stocked production that earns more than crowded neighbours, and advisory or monitoring services that help farmers find the right density.

Siting the farm in the wrong water

HIGH

What it is

Choosing a site because it is close or available, without checking it is suitable. Examples are stagnant, shallow, polluted or poorly flowing water, or a crowded stretch of lake where disease already circulates. A good system can still fail if the water is poor.

Evidence

About 80 to 90 percent of farmed fish grows in the same Lake Volta water, so shared-water disease and pollution hit the whole cluster at once, and outbreaks on lake farms kill 10 to 80 percent of stocked fish.¹²

Who it hits

First-time entrants who take whatever water is nearest, and cage operators who crowd into the busiest, highest-risk part of the lake.

How to manage it, and the opening

Assess flow, depth, oxygen and isolation before committing, keep the 1 kilometre separation from other farms and at least 2 metres between cage bottom and lake bed that the Water Resources Commission guidelines set,¹³ and consider pond or tank sites off the crowded lake entirely. The opening is well-sited capacity away from the crowded parts of the lake, which spreads the sector’s risk.

Buying the wrong system for your capital

MEDIUM

What it is

Copying the cage or RAS model because it looks the most productive or the most modern, without the lake access, the working capital or the technical skill each demands. A tank system with no reliable power, or a cage with no lake site, is money spent on the wrong kind of business.

Evidence

Cage, pond and tank differ sharply on water need, capital and skill, and RAS in particular fails hard if the pumps or power stop; cage depends entirely on secure lake access.⁴

Who it hits

Ambitious entrants who size the system to their ambitions and not to their site and cash, especially those drawn to high-tech tanks without the power and skill to run them.

How to manage it, and the opening

Start with the system your site and capital actually support, prove the operating skill on a small unit, and move to a bigger or more intensive system only once the cash flow and skills are in place. The opening is advice that helps entrants choose the right system from the start.

Dead tilapia floating in a Ghanaian cage after a mass mortality event

🐟 Key takeaways 🐟

01

Production method is a capital-and-site decision. Cage, pond and tank differ on water need, capital, skill and disease exposure; pick the one your site and budget support.

02

Lake Volta cages supply about 80 to 90 percent of output, and that concentration on one lake is a major structural risk for the sector.

03

Daily management (sensible density, careful feeding to an FCR of about 1.5 to 2.0 and active water management) decides survival and cost more than the strain or the equipment does.

04

Siting matters a great deal. Three regions make about 82 percent of output. Farming inside the cluster gives support but more crowding, while farming outside gives cleaner water but weaker logistics.

05

Less can earn more: in a pond trial, 2-gram fingerlings at 3 fish a square metre returned 38 to 76 percent, against about 9 percent for 10-gram fingerlings at 6 a square metre. Stock for healthy growth.

Where this connects

Stocking density and daily management affect survival, which links to disease and biosecurity in Pillar 3, especially on the crowded lake. The feeding and FCR discipline flows straight into the cost model in Pillar 5, where feed is the dominant expense, and the screened, monosex fingerlings these farms depend on are the seed supply problem covered in Pillar 1.

Written for each reader

🐟 Practitioner intelligence 🐟

Hover any card to pause and lift it.

For students

If you want to learn production, start on a small earthen pond or a couple of tanks inland, where the capital at risk is low and the water is yours to manage. Learn the three practices that decide the result. These are stocking density, feeding to a low feed conversion ratio, and daily water management. Catfish in a pond is the most forgiving place to build these habits. Avoid copying the biggest system before you can run a small one profitably, because the operator's skill affects the numbers most.

For entrepreneurs

Decide your system from your site first and then your capital, before you buy anything. If you have secure lake access and the money to absorb disease risk, a cage gives the most volume; if you have land and water but no lake, a well-run pond is the lower-risk way in; if you are near a city with capital and skill, tanks can serve fresh at a premium. A costly mistake is buying the wrong kind of system, such as a tank with no reliable power or a cage on crowded, poorly flowing water. Stock within 50 to 80 fish a cubic metre, resist the urge to crowd, and prove the regime on one unit before you scale.

For investors

Start due diligence by checking that the system fits the site, because many ventures fail there. Ask whether the water is secure and suitable, whether the site sits in the crowded, high-disease part of Lake Volta or safely off it, and whether the operator can run the grow-out regime rather than just build cages. A crowded cage on shared water concentrates disease risk. Back well-sited, well-run farms, especially ponds and tanks away from the lake. They carry less risk, and the national plan is trying to fund this kind of spread. Judge the site and the operator before the tonnage target.

For ecosystem actors

The aim should be well-sited production spread away from the most crowded stretches of the lake. Support pond and tank development away from Lake Volta, tie any siting approvals to flow, isolation and carrying capacity, and back extension that teaches the grow-out regime rather than just handing out fingerlings. Measure success as healthy production spread across suitable sites. Avoid subsidising more crowding on the same stretch of water, which raises the shared disease risk the sector already carries.

For students

If you want to learn production, start on a small earthen pond or a couple of tanks inland, where the capital at risk is low and the water is yours to manage. Learn the three practices that decide the result. These are stocking density, feeding to a low feed conversion ratio, and daily water management. Catfish in a pond is the most forgiving place to build these habits. Avoid copying the biggest system before you can run a small one profitably, because the operator's skill affects the numbers most.

For entrepreneurs

Decide your system from your site first and then your capital, before you buy anything. If you have secure lake access and the money to absorb disease risk, a cage gives the most volume; if you have land and water but no lake, a well-run pond is the lower-risk way in; if you are near a city with capital and skill, tanks can serve fresh at a premium. A costly mistake is buying the wrong kind of system, such as a tank with no reliable power or a cage on crowded, poorly flowing water. Stock within 50 to 80 fish a cubic metre, resist the urge to crowd, and prove the regime on one unit before you scale.

For investors

Start due diligence by checking that the system fits the site, because many ventures fail there. Ask whether the water is secure and suitable, whether the site sits in the crowded, high-disease part of Lake Volta or safely off it, and whether the operator can run the grow-out regime rather than just build cages. A crowded cage on shared water concentrates disease risk. Back well-sited, well-run farms, especially ponds and tanks away from the lake. They carry less risk, and the national plan is trying to fund this kind of spread. Judge the site and the operator before the tonnage target.

For ecosystem actors

The aim should be well-sited production spread away from the most crowded stretches of the lake. Support pond and tank development away from Lake Volta, tie any siting approvals to flow, isolation and carrying capacity, and back extension that teaches the grow-out regime rather than just handing out fingerlings. Measure success as healthy production spread across suitable sites. Avoid subsidising more crowding on the same stretch of water, which raises the shared disease risk the sector already carries.

Footnotes

Banini, Philip Kwasi, Kofi Ferni Anyan, Jacob Zornu, Mabel Ackah, David Narteh Batsa, Kwame Issifu, Abigail Amankwah, Shimaa E. Ali, Samuel Addo, and Kofitsyo S. Cudjoe. "Rethinking Freshwater Cage Aquaculture: A Case in Ghana." Water 16, no. 21 (2024): Article 3054. https://doi.org/10.3390/w16213054.↑
Baah, Doris Y., Regina Edziyie, Amr A. A. Gamil, Kwasi Adu Obirikorang, and Øystein Evensen. "Linking Knowledge, Attitudes, and Practices to Disease Dynamics in Tilapia (Oreochromis niloticus) Aquaculture on Volta Lake." Frontiers in Aquaculture 5 (2026): Article 1710076. https://doi.org/10.3389/faquc.2026.1710076.↑
Kassah, Jemimah Etornam. Best Management Practices Guidelines for Small-Scale Tilapia Cage Aquaculture in Ghana and Nigeria. WorldFish Manual 2023-09. Penang, Malaysia: WorldFish, 2023. https://digitalarchive.worldfishcenter.org/items/fa7eaabd-3058-4f86-93cb-93abf520e6d3.↑
Péteri, András, Thomas Moth-Poulsen, Éva Kovács, Imre Tóth, and András Woynárovich. African Catfish (Clarias gariepinus, Burchell 1822) Production with Special Reference to Temperate Zones: A Manual. Budapest: Food and Agriculture Organization of the United Nations, 2015. https://openknowledge.fao.org/server/api/core/bitstreams/f8fb572b-7d08-45cc-ac5c-4aa7ef4c5e4d/content.↑
Ministry of Fisheries and Aquaculture Development and Fisheries Commission. Ghana National Aquaculture Development Plan (GNADP) 2024-2028. Accra: MoFAD and Fisheries Commission, 2024.↑
Zornu, Jacob, Saraya Tavornpanich, Edgar Brun, Paul A. M. van Zwieten, Ingrid van de Leemput, Patrick Appenteng, Janet Anchirinah, and Kofitsyo S. Cudjoe. "Understanding Tilapia Mortalities and Fish Health Management in Lake Volta: A Systematic Approach." Frontiers in Sustainable Food Systems 7 (2023): Article 1249898. https://doi.org/10.3389/fsufs.2023.1249898.↑
Agyakwah, Seth K., Ruby Asmah, Emmanuel T. D. Mensah, Catherine Ragasa, Sena Amewu, Nhuong Tran, Mathew Oyih, and Peter Ziddah. Farmers' Manual on Small-Scale Tilapia Cage Farming in Ghana. Accra: CSIR-Water Research Institute; International Food Policy Research Institute, 2020. https://cgspace.cgiar.org/items/bff79246-b074-4a61-9ef2-793b51e83483.↑
Magna, Emmanuel Kaboja, Emmanuel Tetteh-Doku Mensah, Franklin Nantui Mabe, Mercy Johnson-Ashun, Lilly Osei Konadu, and Ebenezer Koranteng Appiah. "Profitability Analysis of Small-Scale Cage Aquaculture Farms in the Volta Lake of Ghana." Aquaculture Research 2023 (2023): Article 1314660, 1-10. https://doi.org/10.1155/2023/1314660.↑
Mapedza, Everisto, Marie-Charlotte Buisson, Giulia Zane, Sarah Appiah, Ruby Asmah, Lawrence Ahiah, and Emmanuel Mensah. Review of Fisheries and Aquaculture Policies in Ghana: Technical Report. Colombo, Sri Lanka: International Water Management Institute (IWMI); CGIAR Initiative on Aquatic Foods, 2024. https://cgspace.cgiar.org/items/a20c555b-c2b0-4429-89ab-b550e790adb6.↑
Addo, Isaac, Seth Koranteng Agyakwah, Francis K. Y. Amevenku, Ruby Asmah, Emmanuel Tetteh-Doku Mensah, Catherine Ragasa, and Trong Quoc Trinh. "Assessment of Production and Profitability of Stocking Sizes and Densities of Nile Tilapia (Oreochromis niloticus) for Small-Scale Pond Culture in Ghana." GSJ (Global Scientific Journal) 12, no. 4 (2024): 1390-1415. https://www.globalscientificjournal.com/researchpaper/Assessment_of_production_and_profitability_of_stocking_sizes_and_densities_of_Nile_tilapia_Oreochromis_niloticus_for_small_scale_pond_culture_in_Ghana.pdf.↑
Citi Newsroom. "Meet Aquamet Technologies, a Ghanaian Startup Powering Sustainable Aquaculture with Smart Innovation." Citi Newsroom, 30 June 2025. https://citinewsroom.com/2025/06/meet-aquamet-technologies-a-ghanaian-startup-powering-sustainable-aquaculture-with-smart-innovation/.↑
Royal Academy of Engineering. "Frank Owusu, Ghana, AquaMet." Africa Prize for Engineering Innovation, 2025 cohort, accessed 29 September 2026. https://africaprize.raeng.org.uk/2025-cohort/frank-owusu.↑
Water Resources Commission (Ghana). Technical Guidelines for Sustainable Cage Aquaculture. Accra: Water Resources Commission, 2013. https://wrc.gov.gh/wp-content/uploads/2025/05/Technical-Guideline-for-Sustainable-Cage-Aquaculture.pdf.↑