Biofuels and greenhouse gas emissions from land use change: A quantitative review of the literature[Biocarburants et émissions de gaz à effet de serre dues aux changements d’usage des sols : Une revue quantitative de la littérature]

In recent years, public support to biofuels on the grounds that they contribute to the mitigation of greenhouse gas (GHG) emissions has been questioned, mainly because of the potential impacts of both direct and indirect land use changes (LUC) on carbon stocks. We provide a quantitative review of the related literature using meta-analysis techniques. 63 references that provide 706 quantitative, non-redundant and transparent evaluations of the impacts of the development of biofuels on LUC emissions were analyzed. The collected data indicate that emissions due to the development of biofuels may be of considerable importance (average LUC factor: 51 gCO2/MJ), but also characterized by a large variability. In order to disentangle what comes from differences in scenarios, assumptions and/or study-specific characteristics and what may be attributed to true uncertainty, study- and scenario-specific variables were coded and used as explanatory variables for the LUC factor in nested random-effect, unbalanced panel models. The estimated models capture an important share of the initial variability. The findings indicate that (i) the type of land-use considered (e.g. accounting for peatland), (ii) the approach (economic models vs. consequential LCA), (iii) the type of biofuel (biodiesel vs. bioethanol, 1st vs. 2nd generation), and (iv) the inclusion of market-mediated impacts (e.g. yield response to prices) matter in the overall assessment of LUC-related emissions.