Comment 1: (Thomas Brooks should be credited for submitting these comments and devising the following system.) MacPhee and Flemming’s (1999) mammal category II was too broad for birds because it included three very different classes of unresolved extinctions:
--taxonomically invalid species;
--species that may have become extinct earlier than 1500;
--species that have not been recorded in the last 50 years but may survive.

Harrison and Stiassny’s (1999) fish category split these up but did not account for species that may have become extinct prior to 1500.

MacPhee and Flemming’s mammal category IV also grouped two different classes: those definitely lost within 50 years; and those that may survive. Brooks’ categorization system for birds split these categories, following the protocol used for fishes. Brooks agrees with the mammal survey protocol of keeping taxonomically invalid and taxonomically undescribed species separate (this was not done in the fish categories). (Similar comments have been made in Mammal Panel Feasibility Surveys). Brooks found it very difficult to separate the fish extinction categories B2/B3 and D2/D3 for some birds, ie., the two possibilities (extinction not supported by sampling or not supported by evidence of threat) can be difficult to differentiate. (Another Bird Panelist has indicated that some rigorous threat categories may overlap with the extinction categories).

Brooks' system recognizes that species can first be broken down into three main groups:

-- I. Resolved AD 1500-50 years ago
-- II.Definite but possibly lost pre-AD 1500
-- III.Definite but possibly lost in last 50 years

Brooks states that these categories can be modified by one or more additional pieces of information, as follows:

a.Taxonomy unclear, otherwise resolved
b.Undescribed, otherwise resolved.
c.May survive (on the basis of unrepresentative sampling and/or apparent environmental threat and population decline).

Hence a species is categorized not according to a dichotomous key but according the three main categories and any combination of the modifiers; eg. Gerygone hypoxantha is categorized IIIac.

(This seems a sensible and easy system to apply, probably easier than the fish key. It is also more comparable to the revised IUCN system. But for fishes I would still want to make the system more descriptive in the "may survive" category, so that species that are thought to exist because sampling has been unrepresentative are kept separate from other species where there is simply no evidence of their decline (e.g., in the case of enigmatic species that are known only from one or two specimens and were never collected again). The reason Harrison and Stiassny (1999) had steered away from a system of principal categories with modifiers is that we thought we could possibly end up with a very complex set of modifiers and submodifiers (eg. modifier c broken down into a series of submodifiers which indicate why the extinction evidence is inconclusive) - but perhaps this would be no more complex than an elaborate key?)

Comment 1: demonstrating insect extinction is notoriously difficult except for certain well-known groups (e.g. some lepidoptera, coleoptera) and some regions (e.g. parts of Europe and North America). Exploring insects on an ordinal basis will not be feasible, a family or tribal level is more realistic. Heritage lists would be a very valuable starting point for CREO's assimilation of insect data, and care should be taken not to duplicate effort of Heritage programs.

Comment 2: in general CREO is a worthwhile endeavour to undertake now. The Panelist is a little skeptical that the same criteria for determining EEDs can be applied to invertebrates and ephemeral plants as are used for vertebrates and perrenial plants, but states that's no reason not to start.

Comment 1: lepidoptera are a good group to test. Also, we should think about just considering particular regions and documenting regional extinctions (i.e. extirpations) as a study of what processes are driving extinctions at a local level. (This will provide very useful information for understanding the processes of extinctions, and the CREO methodology could be easily applied to local extinctions. However, this would mean that the project would have to cover a much larger range of species, and could duplicate some of the work of the more detailed work of SSC groups which are particularly focused on threatened rather than extinct species).

Comment 2: The values of particular Lepidoptera as 'umbrellas' may help us come to grips with declines, but I suspect that the work of the CREO group will be confined largely to a small subset of the order, butterflies, as the only members for which even reasonably comprehensive historical data will be available. Because this will be largely confined to the temperate northern hemisphere, most progress will be made by concentrating on that region as an initial thrust.

(Another entomologist has stated that many suggested insect extinctions are too premature).