I know a bit about global, national and state ranks and offer to share some of it here. First of all, it's important to understand that the ranks are outputs of something called the NatureServe Rank Calculator. It's just a macro-enabled spreadsheet that the user enters values into, and then pushes a button to get the rank.
The basic categories of stuff the calculator considers to derive its output are 1) rarity, 2) threats, and 3) trends
(both long-term and short-term).
NatureServe produces global and national ranks. When applicable, they prefer to use data provided by member programs (typically called Natural Heritage Programs) - in the USA most of them are at the state level. There are also Navajo Nation and Tennessee Valley Authority programs. When I say "state level" I don't mean those programs are operated
by state agencies - most are, but about 1/3 are not. That appears to be the case in Oregon - it seems to be a university thing there. Most non-agency state-level programs are operated by universities. Two states I have lived in, Colorado and Florida, have this situation. It is not infrequently the case that when a universirty runs a Heritage program, the state wildlife agency has little to do with the program.
So just to repeat - S ranks are not produced by NatureServe. They are produced by state-level programs, and reported to NatureServe by participating or member programs. Participating in the NatureServe network, that is.
...I came away with the conclusion that EOs are based on locality sightings which may be clumped into groups the then call EOs..
Yeah that's about right. It's a bit of an abstraction. There is some direction given for e.g. colonial animals - it would be silly to have 2 (or 200) observations of individual prairie dogs from the same town each get their own EO. The dog town is the EO. Same goes for, e.g., a pelican or tern nesting colony, a sea turtle nesting beach, etc. For solitary animals, conceptually, a spatial buffer is applied to the observation. More about that below, on "grid cells".
Well, depending on the species, there can be little correlation between the number of locality sightings and the numerical abundance of species. So the use of EOs to assess the status of species is a terribly flawed practice, especially as it pertains to secretive species such as many species of herps. The concept likely has more merit when applied to plants
If you look more closely at the rank calculator you will note that there are 2 basic ways to handle that "rarity" criterion. One is abundance, one is extent of distribution. Very few species are evaluated on the basis of abundance, for the simple reason that such data do not exist for most species.
So let's take extent of distribution. In the calculator, species occupying points or polygons are evaluated based on how many 4 sq km grid cells have an EO (all or part of them) in them. Species occupying lines (e.g. stream organisms) are evaluated based on how many 1 sq km grid cells they occupy. If a species in Oregon occurred statewide, in every conceivable habitat, it would occupy about 63,700 of the larger grid cells. A species occupying 3 springs in the same small valley might occupy 1 or 2 or 3 grid cells.
Restricted range is a widely-accepted predictor of imperilment. So is low abundance, but to a much smaller extent, particularly if distribution is large. Take large carnivores as an example. They are naturally "rare", but by no means may they be assumed to be imperiled.
In the calculator, rarity commands most of the weight driving the output - 85% in the current version if I am not mistake. Threats and trends take up the rest. I quibble with this, I think threats and trends ought to get more weight. But this is inviting a distraction...just remember that threats and trends are also included. How they are included can be utterly subjective, or it can be strongly data-driven. But just as an example, take the lower 48 grizzly population. The long term trend is terrible - they've been extirpated from >95% of their former range. However the short-term (last 20 years) trends are excellent - their distribution and
abundance (the latter well-known in this freak case) have at least tripled. This is the case because most of the serious threats to grizzlies have been sufficiently managed to allow births and survival to greatly outpace deaths.
The important thing to keep in mind, is that the ranks are simply an index of the relative degree of imperilment
a species faces. Forget trying to define exactly what a 3 is - the important thing is, it's better than a 2 and worse than a 4. You don't make a decision about a particular course of action just based on the number - you need to look at the specifics of rarity, threats, and trends.
And as is typical now of state wildlife agencies, neither NatureServe nor the ODFW have any evidence in support of that S-3 ranking for the Night Snake in Oregon.
Again, I would look to whose rank that is, and I would further request of them the rank calculator report showing what inputs they used. The easiest way to "uprank" that animal would be to add more legitimate dots to the map - ones not in already-recorded grid squares. This is an important contribution FHF members can make.
I have now twice taken time to inform myself about NatureServe and their methods. To sum up my position, it amounts to junk science at it finest.
I think this is a bit extreme and I hope I can convince you to take a broader view. Perhaps not in how ODFW is using the ranks to do whatever it is trying to do, but simply to recognize that the Rank Calculator outputs are actually a very helpful device to help overworked, understaffed managers look at the entirety of a flora and fauna (virtually all of which is data deficient
), and quickly determine what they do not need to worry about for the moment, and what needs some attention. What kind of attention
is a whole 'nother thing entirely.
In my case, when I see an S rank that implies
imperilment, but I "know" from experience that it ain't so, the first thing I do is try to round up more "locality sightings" as you call them. I want to paint the entire presumed distribution of the animal in locality sightings, if possible one every few miles in all directions. I do not rush to call something imperiled when it's obviously a simple case of insufficient data. Others may behave differently...but it can be good to ask first, before assuming any motives.
In the case of an animal that is restricted in its distribution, I look to the threats. If there are obvious threats the first order of business is to go about trying to reduce them, preferably starting with the worst ones first. Typically, those manifest in habitat loss and degradation, or excessive mortality or insufficient recruitment. Note that these are drivers of distribution and abundance - exactly the things the rank calculator gives most weight to.
Finally, operators of the calculator can always do a manual override. If for example a species has a naturally small distribution, but there has been little or no discernible change in that distribution, and there are no credible existential threats that have not been managed, if the calculator spits out a "1" the operator can always manually assign a "2" or "3".
Anyway, I hope this is informative. I will close by suggesting that to demand virtual omniscience as a prerequisite to taking some sort of action, is a pretty specious argument. You're damned if you do and damned if you don't, so you might as well do something that seems about right - that's the position conscientious managers are in. The trick is in figuring out what "right" is, and not just grabbing at the first thing you see. Those who appear guilty of doing that need to be called out, starting with an approach that is more likely to have them engage constructively, rather than retreat to the bunker.