General Comments on the Draft for discussion produced by the ICRP on “Environmental Protection: the concept and Use of reference Animals and plants” (Only general comments are given herafter) The approach that is described throughout the document intends to demonstrate why and how the protection of the environment per se could be implemented for radioactive substances and for the situations of interest (i.e., planned, existing, emergency situations) described in the main ICRP recommendations (ICRP 103, 2008). The background is well reminded in the introduction highlighted the need for consistency with the human radioprotection and with the existing approach for chemical substances. The RAP concept strongly mimics what has been developed for the human radioprotection system and even though the report only presents the preliminary reasoning of the approach, the adoption of the RAPs concept seems to ensure the consistency with human radioprotection system will be met. Concerning the other major point to respect while developing the system of radioprotection of the environment i.e. to fit as far as possible the existing approach and associated methods for chemical substances, the preliminary reasoning is clearly too poor to convince. References to methods of ecological risk assessments for chemicals are missing. Discussions on the variety of the objects to protect and of the levels of protection that are targeted (e.g., the structure and functioning of a given ecosystem or the reproductive capacity inside a particular population of a given species, or each individual for endangered species, etc) are not developed enough and the use of RAPs and needed extrapolations to ensure a given object of protection and a given level of protection is poorly done. In addition, the relationships between the exposure level, the dose or dose rate and the biological effects are only narrative, restricted to the limited set of RAPs. The additional value of the section in comparison with previous reviews such as UNSCEAR and/or IAEA is not clear. Here again, references are missing. Finally, the way proposed to use this information as an help for a risk indicator is explained in a very confusing manner. Those DCLs said to be based on background dose rates in the text, only appear as a given range (e.g., 0.1 to 1 mGy/day) in the DCLs tables, those latter referring to ranges of effects. Expert judgement is herein with no indication of the accuracy associated with the bounds of the ranges that are given, and with no explanation concerning the way these ranges were selected. Moreover a very confused explanation is given about their potential interpretation (bullet 353:…it does not imply that higher dose rates would be environmentally damaging nor that lower dose rates were in some way safe or non-damaging…). It is also unclearly explained how propagation at higher level of organisations than individuals could be managed practically and how indirect effects through food webs could be taken into account. The proposed approach needs to gain credibility while reorganising deeply the reasoning and the use of RAPs throughout this reasoning. The objective is to give priority to the methods that could be implemented to demonstrate ecological objects are protected at a given level. The reasoning suffers from the lack of clarity on how RAPs should be used. Major advances in this field have been met during the ERICA project and now during PROTECT, while proposing a risk-based indicator (Hazard quotient or probabilistic approach making use of species sensitivity distributions to support in a transparent way the knowledge used, the gaps and therefore the inherent limits, the extrapolation needed and associated uncertainties). References about these aspects are clearly missing (e.g., JER ERICA special issue; ERICA D5; Garnier-Laplace et al., ES&T 2006; Alonzo et al., Aquatic Toxicology 2006,). Moreover the applicability of the approach has been demonstrated throughout a number of case studies during the framework of ERICA (e.g., ERICA D9, D-ERICA Beresford et al., 2007). This is a clear strength for this ERA-type approach that is not used within the reasoning of ICRP C5. ICRP should encourage and justify the prioritisation of research development in general and not only through the RAPs. For some of them, little radioecology data are available and acquiring more can be hardly implemented. This underlines that the use of RAP could be preferably limited to an help for dosimetric considerations while extension to all non human biota for effects analysis would allow to enlarge the use of all available knowledge. Conclusion: The report is not reflective enough of current thinking and scientific progress in the field of ecological risk assessment associated to radioactive substances. The applicability of the described approach is clearly not demonstrated. The follow-up of the C5 work plan should be revised to give priority to the revision of this preliminary basic approach to ensure (and then to demonstrate) its applicability to different exposure situations. This is a pre-requisite before delivering recommendations on more focused topics such as those listed in bullet 388 (i.e. databases for modelling exposure, RBE…). Some more specific comments on dosimetric aspects: The dosimetric chapter is too much detailed but not sufficiently accurate. The choice to name the dosimetric coefficients DCF (§.85) is contrary to the agreement reached in the IAEA EMRAS BWG where this factor was named DCC (Dose Conversion Coefficients). Some reasoning is not clear enough, especially concerning whether the DCF were weighted or not (§.100). If yes, which values were applied as weighting factors? There are wrong references in the text (§.101, eq. 4.1 & 4.2). Other points need clarification: distance to the soil (§.107, deer; §.119, bee), external exposure of pine tree (§.122 & 124), 4Õ/2Õ geometry (§.113), meaning of a blank versus a “-“ (appendix B, table B-3).