Spatially resolved inventory and emissions modelling for pea and lentil life cycle assessment

Abstract

Pulses are an increasingly popular source of plant-based protein among food manufacturers, food service providers, and retailers globally. Canada is currently a global leader in pulse production and exports. Given the large geographical scope of their production within Canada, pulses are produced under a variety of conditions reflecting differences in soil, climate, and characteristic management practices. In order to understand the influence of regional factors on the impacts of pulse production, high-resolution, regionalized life cycle assessments (LCAs) were carried out using both region-specific input/output data and emissions modelling. Six hundred Canadian pea and lentil farmers were surveyed to collect detailed data regarding characteristic farm inputs, yields and management practices at the reconciliation unit level of spatial resolution, which reconciles Canadian provincial borders with terrestrial ecozones based on soil and climate factors. The process-based model DeNitrification DeComposition (DNDC) was used to estimate regionally specific N emissions. This was compared to emissions estimates generated using the IPCC Tier II empirical models, which are typically used to estimate greenhouse gas emissions for national inventory reports. The main contributors to the life cycle environmental impacts of pea and lentil production were fertilizer and fuel use. This was consistent across all levels of regional aggregation including the total Prairie province average, as well as ecozone and provincial levels. There was variation in the magnitude of the impacts in each region assessed, which was mainly attributable to differences in yield, as well as reported fertilizer application rates and related emissions and fuel use for field operations. Significant differences were found in N emissions estimates between the DNDC and IPCC models, the magnitude of which varied by region and by the N emissions model employed. DNDC was found to provide better-resolved emissions estimates at the ecozone scale. This demonstrates the relevance of regionally specific emissions modelling since local soil and climate conditions had a large impact on the emissions estimates. In addition, the levels of uncertainty in the models were generally higher at the provincial and prairie province scale than at the ecozone scale. This may indicate that farming practices and associated resource/environmental impacts are more strongly influenced by soil and climate conditions than by provincial standards and guidelines. These results underscore the necessity of spatially resolved data collection and modelling to provide accurate estimates of the environmental impacts of crop production and to support more sustainable management of arable crop production.