Response to Kruse-Plass et al. (2017) regarding the risk to non-target lepidopteran larvae exposed to pollen from one or more of three Bt maize events (MON810, Bt11 and 1507)

We respond to the paper of Kruse-Plass et al. (Environ Sci Eur 29:12, 2017), published in this journal, regarding the risk to non-target lepidopteran larvae exposed to pollen from one or more of three Bt maize events (MON810, Bt11 and 1507). We emphasise that what is important for environmental risk assessment is not the number of pollen grains per se, but the degree of exposure of a NT lepidopteran larva to Bt protein contained in maize pollen. The main text of this response deals with general issues which Kruse-Plass et al. have failed to understand; more detailed refutations of each of their claims are given in Additional file 1. Valid environmental risk assessment requires direct measurement of pollen on leaves at varying distances outside a source field(s); such measurements reflect the potential exposure experienced by an individual larva on a host plant. There are no new data in the Kruse-Plass et al. paper, or indeed any data directly quantifying pollen on actual host-plant leaves outside a maize field; only data gathered within or at the edge of maize crops were reported. Values quoted by Kruse-Plass et al. for deposition on host plants outside the field were estimates only. We reiterate the severe methodological criticisms made by EFSA [Relevance of a new scientific publication (Hofmann et al. 2016) for previous environmental risk assessment conclusions and risk management recommendations on the cultivation of Bt-maize events MON810, Bt11 and 1507. EFSA Supp Publ; EN-1070, 2016], which render this estimation procedure unreliable. Furthermore, criticisms of EFSA (EFSA J 2015(13):4127, 2015) and of EFSA [Relevance of a new scientific publication (Hofmann et al. 2016) for previous environmental risk assessment conclusions and risk management recommendations on the cultivation of Bt-maize events MON810, Bt11 and 1507. EFSA Supp Publ; EN-1070, 2016] made by Kruse-Plass et al. are shown in Additional file 1 to be without foundation. We therefore consider that there is no valid evidence presented by Kruse-Plass et al. to justify their conclusions.     

Extension of sensitivity and uncertainty analysis for long term dose assessment of high level nuclear waste disposal sites to uncertainties in the human behaviour

Biosphere dose conversion factors are computed for the French high-level geological waste disposal concept and to illustrate the combined probabilistic and deterministic approach. Both ¹³⁵Cs and ⁷⁹Se are used as examples. Probabilistic analyses of the system considering all parameters, as well as physical and societal parameters independently, allow quantification of their mutual impact on overall uncertainty. As physical parameter uncertainties decreased, for example with the availability of further experimental and field data, the societal uncertainties, which are less easily constrained, particularly for the long term, become more and more significant. One also has to distinguish uncertainties impacting the low dose portion of a distribution from those impacting the high dose range, the latter having logically a greater impact in an assessment situation. The use of cumulative probability curves allows us to quantify probability variations as a function of the dose estimate, with the ratio of the probability variation (slope of the curve) indicative of uncertainties of different radionuclides. In the case of ¹³⁵Cs with better constrained physical parameters, the uncertainty in human behaviour is more significant, even in the high dose range, where they increase the probability of higher doses. For both radionuclides, uncertainties impact more strongly in the intermediate than in the high dose range. In an assessment context, the focus will be on probabilities of higher dose values. The probabilistic approach can furthermore be used to construct critical groups based on a predefined probability level and to ensure that critical groups cover the expected range of uncertainty.