Knowledge for better food systems

GHG impacts of seafood and fish

Sue Dibb's picture

Can FCRN members advise how fish and seafood compare with different types of meat re their GHG impacts? 

I recently saw this article re research into the carbon footprint of (warm water) prawns/shrimps which quotes J Boone Kauffman, an ecologist from Oregan State University: “The carbon footprint of the shrimp is about 10-fold greater than the land use carbon footprint of an equivalent amount of beef produced from a pasture formed from a tropical rainforest,”

Is this new research and does it compare with existing research?  And how might this compare with cold water prawns which is what I think we mainly get in the UK?

In advising people to eat less and better meat, Eating Better doesn’t encourage a shift to increased fish consumption due to the unsustainability of many fish stocks.  But it would also be good to be able to share information re fish and seafood’s GHG impacts.

Thank you

Sue Dibb, Coordinator

Eating Better alliance

intolife's picture
Submitted by intolife on

Hi Sue, not sure about this particular study in terms of methodology (thats out of my skill range!), but it relies heavily of the carbon release from land-use change of course, rather than the production-side emissions of the shrimp. Norwegian-based research has a much lower (much much lower than this study) CFP for cold water prawns.  Somewhere around 3kg GHG per kg of produce.  But land-use impact of sea-caught prawns does not factor here in the same way as the Kauffman measurements...

In general, cod-type fish have been shown to be around 2.5, similar for farmed salmon, haddock around 3.5, mussels around 1.5, herring and mackerel both under 1.

There is research implying that seafood has a peculiar lifecycle pattern because the transportation and cold-storage impact can be significant too.

Like you, I don't use fish as a strong case for meat alternatives, simply that in general certified sustainable seafood is going to be lower than most meat choices.  Due to the limited availability of certified seafood, I take the same approach in a meal composition for seafood and meat - that they should not be the main factor in a meal, but part of the plate.  You know what I mean.

Not sure that has answered your question about the Kauffman research... my meal comparisons graphs would look a bit insane if I used his CFP and compared different meals!


Will Nicholson


Food Sustainability in the Restaurant Industry's picture

Hi Will/Sue,

Yes accounting for land use/land change is what it is all about here and the analysis has given an extreme and skewed picture. Global Aquaculture Alliance, an aquaculture advocacy organisation, responded to the original article at the time

Shrimp production has become increasingly diverse over the last ten years ranging from traditional and improved extensive systems with little to no reliance on external feed inputs through to super-intensive systems that are inherently feedlot systems. This makes for very large, and system specific differences in environmental impacts. In a recent study we found that values ranged from <1 to around 16 kg GHG per kg product in Bangladesh and between3.5-9 in Thailand.  (this is liveweight equivalent data)

Most of the wild shrimp fishery data is much higher but again varies greatly-whereas a trawl fishery for a southern pink shrimp product was 38.1 it was only 7.76 for an artisanal fishery of the same species.(Ziegler et al 2009)


Hope this helps






StVitus's picture
Submitted by StVitus on

I think a lot of that article was debunked on several points. When talking about land use and land use change (LULUC) there will always be certain assumptions. Firstly the time scale that we are talking about for LUC. The main drivers for deforestation are often logging and then the land sometimes is used for agriculture subsequently, such as soy bean or for pasture. Generally speaking, beef produced on any pasture, no matter what its previous use was, has a much larger GHG emission because of the enteric methane produced. Shrimp and beef have a massive range of values for GHGs because of the broad range of production systems from extensive polyculture - intensive monoculture and pasture to feedlot respectively. Salmon in contrast has a narrow range of values typically around 3000kg/tonne live weight because they are produced in very similar conditions globally. Another point of contention is comparing live weights because of the edible portions of the animal vary and then the utilisation of the by-products also as cattle produce leather which has no nutitional value but large economic value. In any case in most animal prodcution systems the critical input is the feed. If you can reduce that then the GHGs, land and freshwater inpacts reduce almost in parallel. Bivalves such as mussels are not fed but have large footprints associated with servicing the production and depuration. However, they have distinct advantages in freshwater and land usage over all other forms of food production.

Angus Garrett's picture

Hi Sue,

To me, three questions arise from your comment:

1. How do warm water prawn GHG emissions compare to beef?

2. How do warm water prawn GHG emissions compare with cold water prawn?

3. Is there information we can share on fish and seafood GHG emissions impacts?

Unfortunately the response in each case is ‘not easily answered’!  I would make the following observations from an industry perspective:

GHG emissions impacts can show large variation depending on the methodology used to calculate them.  This means results should be treated with caution.  In seafood, we have tried to reduce variation by standardising the methodology used.  Working with British Standards, the FAO and stakeholders around the world we were the first to draw up specific guidelines for calculating GHG emissions for aquatic foods.

Seafood supply chains are complex and diverse and can change quite rapidly - calculating impacts can therefore be time consuming and expensive for industry operators.  To date we have focussed on wild capture seafood, conducting a number of studies and producing an online tool that can help operators explore and produce impact profiles at a lower cost.  We have generated initial product comparisons across various species but for supply chains reliant on wild capture only (not in aquaculture and not in agriculture).

Given the differences in supply chains and species, and methodologies used, there is ample scope for specific results to be used/abused.

A final comment is that there is limited impact data to be found in the public domain – although this is growing.  This is partly a consequence of the above, but additionally because data is proprietary across - often fragmented and changing  - supply chains.  Data can be sensitive to commercial operations, and particularly so given the scope to use/abuse specific results.  This makes data collection and publishing a challenging endeavour.

Overall it depends on the objective behind the questions.  Acknowledging different interests, an industry interest is to know whether seafood performs well and how to improve performance.  In this regard general patterns (hotspots) can be observed and it is perhaps better to focus on these, what’s driving them and the actions needed to address them than making claims about particular products.  That is, looking ‘one level up’ at the general storyline on food types rather than looking for answers relating to specific species/products.

Hope this helps.