Accounting for biodiversity in life cycle impact assessments of forestry and agricultural systems - the BioImpact metric

Purpose: Life cycle assessment (LCA) is a useful method for assessing environmental impacts at large scales. Biodiversity and ecosystem diversity are site-specific, often complex, and difficult to generalise within an LCA framework. There is currently no globally acceptable means of assessing biodiversity within the LCA framework. We introduce, test and revise BioImpact, a method for incorporating biodiversity into an LCA framework, on four production systems (native forestry, plantation softwood timber production, cropping and rangeland grazing) in Australia.

Methods: Our proposed method, a metric we call BioImpact, incorporates biodiversity and ecological impacts through a series of semi-quantitative questions, published data and expert opinion which aim to encapsulate the main issues relating to biodiversity within a disturbance impact framework appropriate to LCA. Results are scaled to a single biodiversity measure that can be incorporated into LCA. We test and revise BioImpact scores on four production systems (native forestry, plantation softwood timber production, cropping and rangeland grazing) in comparison to species richness and net primary productivity (NPP) for these production systems. We demonstrate how the scores can be incorporated into LCA using SimaPro as a platform.

Results and discussion: For pine plantation, cropping/pastures and rangeland grazing, BioImpact demonstrated greater impact, which represents biodiversity loss for multiple species groups. Native forestry scored significantly lower impact than that of other land uses. As a comparison, all production processes scored highly for species richness of main multiple species groups (vascular plants, invertebrates, birds) and were not different in terms of NPP. Integration of BioImpact into LCA found that the softwood system, despite having a higher biodiversity impact per ha year, had a marginally lower BioImpact score per cubic metre compared to native forestry. This was possibly due to cumulative effects and consideration of the reference benchmark, e.g., low levels of pre-harvest biodiversity when not established on native forests; fewer threatened species (and lesser impact) compared to native forestry; questions not weighted sufficiently; and the difference between establishment on either agricultural cleared land or native forest area. Improvement in scaling and/or weighting within the BioImpact scores within each question is discussed.

Conclusions: BioImpact encapsulates different components of biodiversity, is transparent, easily applied (subject to literature/ecological experts) and can be incorporated into LCA. Application of BioImpact for LCA requires co-ordination to identify key regions and production systems; develop the relevant scores with the assistance of ecologists; and make the results available in public LCA databases.