Publications des agents du Cirad

Cirad

Empirical knowledge to improve terrestrial ecotoxicity characterisation of trace elements with USEtox

Avadi Tapia A.D., Bravin M., Lemal L., Chevassus-Rosset C., Montes M., Lachowicz L., Tella M., Monteiro A.N.T.R.. 2020. In : Proceedings of the 12th International Conference on Life Cycle Assessment of Food 2020 (LCA Food 2020) ¿Towards Sustainable Agri-Food Systems¿. Berlin : German Institute of Food Technologies, 4 p.. International Conference on Life Cycle Assessment of Food 2020 (LCA Food 2020). 12, 2020-10-13/2020-10-16, Berlin (Allemagne).

USEtox is the consensus model for toxicity in LCA, yet it does not provide characterisation for terrestrial ecotoxicity. Owsianiak et al. (2013) suggested a new framework able to account for soil physical-chemical properties to characterise terrestrial ecotoxicity for trace elements. The performance of this USEtox-based Owsianiak framework was herein tested in the context of soil contamination by copper (Cu) and zinc (Zn) following the application of livestock faeces. To do so, one French soil sample exhibiting properties representative of the main average properties of European soils was incubated under laboratory controlled conditions alone or with the application of 31 swine or poultry faeces collected from experimental pens where livestocks was fed with different rate of Cu and Zn complements. The main endpoints of Cu and Zn availability in soils calculated in the Owsianiak framework were determined analytically on each of the 32 experimental treatments, namely: i) reactive Cu and Zn extracted with 0.43 M HNO3, ii) total Cu and Zn in extracted soil solutions, and iii) free ionic Cu and Zn in extracted soil solutions. The single value for the intermediate endpoints (i.e. fate, accessibility, bioavailability, and effect factors) and the comparative toxicity characterisation factor calculated with the Owsianiak framework for Cu using the initial soil properties were within the range of values computed by Owsianiak et al. for a variety of global soils, thereby validating the hypothesis about the representativeness of the chosen soil. From an experimental point of view, the application of swine and broiler faeces to the soil induced substantial changes in soil pH and dissolved organic carbon concentration compared to the non-amended control soil. These chemical alterations of soil solution chemistry in faeces-amended soils induced a significant decrease of Cu2+ activity for 19 out of 31 faeces compared to the non-amended soil. These observed variations suggests consequently that intermediated endpoints determined for each experimental treatment will likely show some discrepancies with those determined theoretically with the Owsianiak framework, thereby opening the discussion to suggest some potential improvements.

Documents associés

Communication de congrès

Agents Cirad, auteurs de cette publication :