This accepted yet to be published #paper cited the work that I help do on agricultural commodities in England and Wales.

I really like how thoroughly the authors have modelled what is a global supply chain into European cereal products. The Irish are very big breakfast cereal eaters (>8kg/ head/ year) and Italian’s the least (<1 kg/head/year). It does give good systematic insights and indicates leverage points for improvements.

The bits that made me think:

• Rice paddy fields consume a lot of water, but most of it flows into the next paddy field -which is not a net consumption, unlike drainage and evapo-transpiration losses. I am never sure that is is properly considered in many estimates of water use in agriculture.
• When is a waste a by product? When you can find someone who will buy it! In Life Cycle Assessment (LCA) that transition means that you go from  waste disposal burdens charged to the primary product to an allocation of the primary product’s burdens onto the by-product. A green circular economy means we should be doing more of this and I don’t think that should make a difference to the burdens of the primary product and certainly not mask the independent potential for improvement in the primary product. The assumptions behind allocation decisions such as this are known to dramatically alter LCA results.

The biggest environmental impact of agriculture is the decision to farm.

• That is a mantra that I picked up from Seale-Hayne Agricultural college in the 1980. So to is it here if the impact of Cocoa production and possible deforestation re included in the analysis. One of the challenges with Land Use Change and soil carbon  is to justify the time horizon over-which your work applies. All farm land was once something else and moving to tillage crops does shed soil carbon over hundreds of years. If it reasonable to have a 20 year cut-off, as commonly adopted, and ignore Land Use Change before that? I’d argue that impact of Land Use Change should be averaged out and accounted for against all future cropping. I’d go further and suggest that unless the previous land use was a carbon accumulating peat bog then the long-term cycle of carbon is at equilibrium with no net loss or gain for all land uses. For similar reasons I’ve always been sceptical of land use change as carbon sequestration option. Yes you can sequester carbon to move to a new high soil carbon state, but unless you can hold that there for geological time scales then you are not countering the anthropogenic carbon cycle
• Another interesting area was the conclusion that a big burden hotspot of cereal manufacture is the agricultural phase. It does create big burdens. However, agriculture’s case is not helped by food waste down stream in storage, transport, procession, retail, storage, consumption. Accidents do happen and all those little percentages lost soon back multiply to expand the size of the agricultural industry to deliver a set amount of nutrition to a consumer. I’ve never been happy with the way that that shifts, in conventional LCA, the hotspots in the direction of the primary industries whilst partly masking those that wasted it.
• A final remark is that the authours compare a kg of cereal a dispatched from the manufacturer with a kg of cereal consumed with milk in a bowl that has to be washed up. To be fair they are honest that that does not give the consumer two identical nutritional experiences and thus is not a like for like comparison. However, due to the heavy burdens of milk production a superficially comparative evaluation lures the reader into the impression that processing is quite well run, but it is a shame about the farmers (point above) and the consumers!

  

• One little additional thought with their improvement scenarios would have been to have tried Monte-Carlo simulation across the ranges of feasible improvement. It would have given an idea of what combinations of improvements lead to significant change and how significant that would be on average. This would help justify and prioritize investments in improvements.
• My little quibble is that ‘corn’ is ambiguous each side of the Atlantic and explicitly stating maize or #wheat is better.