Assessing variability in carbon footprint throughout the food supply chain: a case study of Valencian oranges

Abstract

PURPOSE: This study aims to analyse the variability in the carbon footprint (CF) of organically and conventionally produced Valencian oranges (Spain), including both farming and post-harvest (PH) stages. At the same time, two issues regarding sample representativeness are addressed: how to determine confidence intervals from small samples and how to calculate the aggregated mean CF (and its variability) when the inventory is derived from different sources. METHODS: The functional unit was 1 kg of oranges at a European distribution centre. Farming data come from a survey of two samples of organic and conventional farms; PH data come from one PH centre; and data on exportation to the main European markets were obtained from official secondary sources. To assess the variability of the farming subsystem, a bootstrap of the mean CF was performed. The variability of the PH subsystem was assessed through a Monte Carlo simulation and a subsequent subsampling bootstrap. A weighted discrete distribution of the CF of transport and end-of-life (EoL) was built, which was also bootstrapped. The empirical distribution of the overall CF was obtained by summing all iterations of the three bootstrap procedures of the subsystems. RESULTS AND DISCUSSION: The CF of the baseline scenarios for conventional and organic production were 0.82 and 0.67 kg CO2 equivalent•kg orange-1 , respectively; the difference between their values was due mainly to differences in the farming subsystem. Distribution and EoL was the subsystem contributing the most to the CF (59.3% and 75.7% of the total CF for conventional and organic oranges, respectively), followed by the farming subsystem (34.1% and 19.8% for conventional and organic oranges, respectively). The confidence intervals for the CF of oranges were 0.72-0.92 and 0.61-0.82 kg CO2 equivalent•kg orange-1 for conventional and organic oranges, respectively, and a significant difference was found between them. If organic production were to reach 50% of the total exported production, the CF would be reduced by 5.4-8.4%. CONCLUSIONS: The case study and the methods used show that bootstrap techniques can help to test for the existence of significant differences and estimate confidence intervals of the mean CF. Furthermore, these techniques allow several CF sources to be combined so as to estimate the uncertainty in the mean CF estimate. Assessing the variability in the mean CF (or in other environmental impacts) gives a more reliable measure of the mean impact.

Integrating Protein Quality and Quantity with Environmental Impacts in Life Cycle Assessment

“…While most studies report carbon footprints as point values, there can be significant uncertainty and variability surrounding such point values [35]. Variability can be a result of differences in farm features, soil type, and management practices [36]. Reviewing the parameters of LCA data sources included indicated that the most common attributes were a system boundary of cradle to gate, economic-based allocation, attributional LCA, and a European region.…”

Modeling the carbon footprint of fresh produce: effects of transportation, localness, and seasonality on US orange markets

“…Past cradle-to-farm gate life cycle assessments (LCAs) of citrus production report life-cycle carbon footprints ranging from 0.07 to 0.64 kgCO 2 e per kilogram of citrus produced with a median value of 0.29 [7][8][9][10][11][12][13][14][15][16][17][18][19][20][21][22][23]. While these studies represent a variety of geographic regions, only two are specific to the United States.…”

Sustainable Development in the Agri-Food Sector in Terms of the Carbon Footprint: A Review

“…Emissions from organic orchards were 52.5% lower, which was associated with the use of only natural fertilizer and a significant reduction in the use of pesticides, because the other stages of the orange's life cycle, such as packaging, transport, and distribution, are the same regardless of the origin of the fruit. In both cases, it was observed that the carbon footprint decreased with the increase in orange harvest per hectare, although in organic farming this trend was more pronounced [66].…”