Abstract: Although the environmental impacts of cattle and poultry have been well documented in the Amazon, the resultant greenhouse gas (GHG) emissions and land use intensities of aquaculture production are poorly understood. Moreover, we know very little about whether emissions vary in aquaculture ponds dominated by different fish species, including tambaqui (Colossoma macropomum), the most common cultured species in the Amazon, pirarucu (Arapaima Gigas), and hybrids. Differences among fish species in feeding modes, morphology, and habitat use could strongly influence GHG emissions. Additionally, pre-installation land cover (e.g., unproductive pastureland versus primary forest) may have far-reaching consequences for the net GHG footprint of aquaculture systems in the Amazon. However, land-use change emissions have been largely overlooked in aquaculture sustainability assessments in the Amazon and elsewhere. Here we investigate the direct (from ponds) and indirect (land use change) GHG emissions associated with aquaculture in the Southern Amazon region, where large fish farming operations have been implemented to meet the local and regional increasing demand for fish. After, we investigate how the GHG emissions of Amazon aquaculture compare to the cattle production system. To estimate the direct GHG emissions from fish ponds, we surveyed 45 fish ponds with commonly produced species in the Rondônia State in Brazil and collected diffusive and ebullitive GHG emissions. The indirect GHG emissions from land use change were estimated by first mapping the fish farm location, after evaluating the historical trajectory of the land cover of each mapped pond, and finally, using the standard techniques proposed by the Brazilian National Inventory to estimate the GHG emission due to land use change. Our results show the GHG emission from fish ponds in Southern Amazon varied from 2.3 to 8.0 gCO2eq m-2 d-1, which means that the fluxes are lower than from fish ponds in other regions of South America with intense aquaculture. We found that the total GHG from fish ponds did not significantly differ among produced species and concluded that management strategies are the primary driver of the variation of GHG emitted from fish ponds in the Southern Amazon. Additionally, the indirect GHG emission by land use change surpasses the GHG directly emitted. Although we observed substantial carbon loss to the atmosphere, both direct and indirect GHG emissions from aquaculture are 2 to 3-fold lower than that emitted from cattle per unit of live weight produced.
