Inconsistencies between regional- and field-scale biodiversity indicators within life cycle assessment: the case of rice production systems in Japan

Abstract

Characterization factors for biodiversity impact assessment derived from ecological zoning and land use classification have been proposed within life cycle assessment (LCA). However, their applicability to LCA of agricultural production systems has not yet been elucidated. This study clarifies correlations between regional-scale biodiversity indicators (characterization factors) and field-scale biodiversity indicators and estimates the degree of macro-micro inconsistencies in biodiversity indicators. Correlation coefficients were calculated between two types of variables. One is biodiversity (potential species loss) at the ecoregion level provided in UNEP/SETAC (2017) and Chaudhary and Brooks (2018), and the other is biodiversity (species richness) at the field level surveyed during a research project on biodiversity in Japan. The data on two taxa (amphibians and plants) in paddy fields were used for the analyses. Two types of correlation coefficients (the Pearson’s product-moment correlation coefficient and the Spearman’s rank correlation coefficient) were calculated. Uncertainties of the correlation coefficients were estimated by statistical resampling because the number of surveyed years and regions were limited. Although in most cases the signs of the coefficients were consistent with theoretical expectations that the correlation between potential species loss and species richness must be negative, the absolute values were low for all cases (especially for the case of amphibians and for the case of using the UNEP/SETAC characterization factors). It was difficult to estimate field-scale biodiversity from ecoregion-scale biodiversity. The introduction of land use intensity into the calculation of biodiversity at the ecoregion-scale increased the correlation coefficients for plants. Uncertainties due to limitations in the number of surveyed locations were larger than those arising from limitations in the number of surveyed years. These results highlighted the existence of macro-micro inconsistencies and the necessity of developing constructive approaches for biodiversity assessment in agriculture. It is concluded that employing characterization factors based on ecoregions and land use categories was not useful when assessing the biodiversity impacts of rice production systems at the field-scale because of the existence of macro-micro inconsistencies. Use of field monitoring methods, in addition to approaches to construct biodiversity indicators based on management practices, will be necessary for establishing sustainable agricultural production systems.