1) The document lay out requires some attention. Some examples follow: a) The document is unclear as to whether it is a proposed requirement to protect individuals, or populations of animals and plants. For example, para 3 indicates it is aimed at species/ecosystem level, yet by para 8 the document is discussing reference levels (DCL) that refer to individuals in nominated species. However, para 352 onwards seems to introduce concepts that were lacking earlier in the document. b) It is appreciated that there is no data about the effects of radiation on populations, whereas there is limited data about radiation effects on individuals in representative species, but this key issue is buried in par 17. c) The dosimetric quantity of interest is only identified in par. 76 in Chapter 4, it should be identified earlier and clearly. As it stands, the document is confusing and muddled; it should state more strongly and earlier what ICRP recommends should be protected (presumably the ecosystem, not at the individual level), and how this is to be achieved. If restructured along these lines, with key concepts and issues drawn together much earlier in the document, then some of these concepts are of merit. 2) It is unclear as to whether the report is a discussion document, identifying gaps in knowledge and hence research needs, or is guidance to users on the development of environmental protection standards. It serves the former purpose much better than the latter. 3) The concept of the dose consideration level has some value as a practical tool for a proactive assessment modelling. However, for retrospective assessment, there is much more value to be gained by looking at the ecosystem level. 4) The set of 12 reference animals and plants seems to have been chosen on the basis that data exists for them, or they are visible. However, there are other species which are better indicators of the health of an ecosystem. The species chosen should include species that contribute to carbon/nitrogen cycling, and/or others that are already used as indicator organisms in toxicity studies, such as micro-organisms, or insects. These are ubiquitous in all ecosystems and are in general easier to study than mammals. 5) Vegetation is also poorly represented; mosses and lichens (which are used as early indicators of air pollution) are not mentioned, neither are large leaf plants. Similarly, filter feeders in the aquatic environment are omitted. 6) It would be useful to ensure continuity between different ecosystems by including reference organism(s) common to more than one ecosystem (ideally all). 7) There is no mention of bio-availability or chemical characteristics of the radionuclide for intake scenarios. However, speciation would not be an issue if an ecological approach was used. 8) The whole document is too granular and subject to misinterpretation. The impact on the nuclear industry/MOD could be negative, with much time spent on justifying the interpretation of it, with challenges, attendant costs and adverse publicity.