Avadí et al. 2015 assessed the effects of different feeds on the environmental impacts of trout (Oncorhynchus mykiss), tilapia (Oreochromis spp.) and tambaqui (Colossoma macropomum) production in Peru. There was one fish oil and fish meal production Cycle, seven feed blending Cycles, and eight fish production Cycles.
Conversion steps
We largely followed the same steps as the Myanmar example, with the following additional steps:
We created six Sites, two for feed processing in Chile and Peru, one for the trout cage systems, one for the tambaqui pond systems, one for the semi-intensive tilapia system, and one for the intensive tilapia system. Because multiple Sites were surveyed and then aggregated, we specified the numberOfSites.
We added Infrastructure representing the cages or ponds, and specified the lifespan of the Infrastructure.
We created 16 Cycles, linked them to the Sites, and set the functional unit to relative. Relative means that the functional unit is defined by the quantity of the Products.
The fish processing Cycle creates two Products: fish oil and fish meal. These Products are subsequently used as Inputs into the feed blending Cycles. Cycles cannot be directly linked in HESTIA and must be linked via an Impact Assessment. Therefore, we therefore created the fish processing Cycle with multiple Products, and then created two Impact Assessments for the oil and meal.
We created the seven feed blending Cycles and added each Input, its quantity, and the country of origin of each Input. For the fish oil and fish meal, we also linked the Input to the Impact Assessment. We specified the Product of these Cycles as Concentrate feed blend. We used Properties to specify the nutrient composition of the blended feed. We also created seven Impact Assessments for these feed Cycles. The authors provided substantial data to populate these Impact Assessments (such as GHG emissions, ecotoxicity, land occupation) and we added these data.
We created the eight aquaculture Cycles and added the Inputs and Products. We linked the Concentrate feed blend Input to the Impact Assessment using cycle.inputs.0.impactAssessment.id.
The aquaculture Cycles had two Products: fish and excreta. We added them and also added the split between solid and liquid excreta.
Following excretion, microorganisms change the mass and composition of excreta, creating Emissions such as NH3 and CH4. A share of that excreta is sedimented at the bottom of the pond and it is often removed and stored in an excreta management system. Here, microorganisms convert it into a different Product again. To represent the transformations of excreta, we added Transformation blank nodes to the Cycles. Specifically, the name of the first Transformation is the first excreta management system, Excretion into water body, and the Input is Excreta, solid and liquid, fish/crustaceans (kg N). The second excreta management system was not specified in the paper, so a generic term was used which was Liquid/Slurry, and we assumed that the amount sedimented was inputted into that system.
Validation
The phosphorus excreted was negative in the study, which was raised as a hard validation error. On inspection of the data the authors were including phosphorus uptake by the fish from sediment and water as a negative amount in the excreta. This should be entered as an Input to the fish, but insufficient data were provided to split the excreta from the uptake, so to resolve this, we simply did not add the phosphorus content of the excreta.