Thursday, December 07, 2023

— Just An Experiment That Ought Be Done (…but won’t be) —

 -------------------------------------------------------

Place ARUs (autonomous recording units) in some northern forests (where there are NO IBWOs), Maine, Michigan, Washington state, or the like. Let them run for a month or two, then collect the recordings and use a computer program or AI to analyze the results looking for “kent”-like and double-knock-like sounds. Once isolated, these can be further spectrographically-analyzed for ones perhaps in the range of the IBWO (understanding that the true range isn’t really known precisely). Kents and double-knocks are waaay more generic than many acknowledge. Often just a small handful of examples of things that may sound like them are mentioned, when there may be dozens of similar sounds in deep woods (not to even mention, in the instance of southern forests, there may be IBWO searchers moving around making such sounds, which are then picked up by others — in the last dozen or so years, even in my NON-IBWO area, I have never gone into deep woods without at some point attempting a double-knock imitation on a tree out of sheer curiosity whether anything would respond to it).


All of this to say, we don’t have a base or control value for kents and DKs in deep woodland… we are clueless how many may appear regardless of any IBWO presence. And it gets worse, we have but a tiny (almost statistically meaningless) sample size of genuine “kents” from known IBWOs (in one locale and circumstance, at a point in time almost 80 years ago), recorded on large antiquated equipment which may enter variables or artifacts into the recordings that aren’t fully known -- the tiny sample size may not necessarily yield a true spectrographic analysis of what today's birds might sound like.  And we have no past recorded IBWO DKs, though I’ll grant that those of other Campephilus species are probably good for comparison-sake. We are, to some scientific degree, operating blindly.


Long time readers here may recall that I essentially feel that auditory evidence sucks (okay, maybe I'm biased by all the junky recordings sent to me over the years)! -- not ALL of it, but enough of it that I don’t place much weight on it. It’s nice to have WHEN directly in conjunction with good sightings (or film clips), but otherwise it doesn’t mean much…. indeed if all the nice-sounding auditory recordings were real one might expect there to be many more good (or even fleeting) sightings by now, when in fact sightings are rare.


 It’s bad enough that most film/photographic evidence is ambiguous at best; acoustic data is perhaps even more ambiguous (and not diagnostic) because of all we don’t know. Similarly, for decades searchers have tried to come up with diagnostic features of IBWO cavities and foraging work, only to have automatic cameras trained on such activity fail to produce a single clear IBWO, despite literally millions of photo frames. There are of course many cavities and tree-work that can be ruled OUT as coming from IBWOs, but nonetheless, ambiguity rules; we seem incompetent to actually pinpoint such necessary avian work.


Thus, we still need to find CURRENT, ACTIVE (daily used) roostholes, nestholes, foraging sites… all the other evidence turns to mush without those. I've lost track now of how many years we've been told the definitive evidence was just around the corner... when in fact the only thing around the corner, was.... another corner... on and on.


With all that said, my basic view remains that Ivorybills likely exist in at least 3 states (Louisiana, Florida, Arkansas, Mississippi, being the top contenders, but at least a half-dozen other states are possible!), perhaps spending most of their lives in the upper canopies (far from people and cameras), yet so too remains my pessimism that human intervention can do much long-term good -- still, stranger things, and remarkable successes, have at times happened….


-------------------------------------------------------

5 comments:

john said...

Perfect timing! You must be psychic. I just gave a Mission Ivorybill slide presentation on this very thing (kents, not DKs for this one). It answers most of your questions and shows that the study of kents is valid, that the sample size is actually rather large for the conclusions reached, that most of the recorded kents were from expeditions with associated sightings, that it is not necessary to sample areas with no IBWOs as a control, that there are not many other species that make sounds that are comparable in spectrograms-- in fact, there are none. Thank you. Here is the slideshow for interested readers. Cyberthrush and I write quite differently as a reader will see-- https://docs.google.com/presentation/d/19RkgzdT4yMPu7cepPEaGqyxf_VrhZNdX4MRFGrpOafQ/edit?usp=sharing

john said...

Addition-- I forgot to mention that both Hill's and Cornell's expeditions, and probably others (Principalis), had protocol to not use playback, or to record the exact times when they did. This can discount the idea that there were people in the woods playing kents.

cyberthrush said...


John, I’m quite aware of the Cornell/Auburn protocols — which in NO WAY bear on random, other searchers invading an area and playing tapes or simulating DKs (lots of folks, unknown on the internet) occasionally search on their own. That’s why I mentioned that even I simulate DKs when I’m in the woods, and have no idea what some other birder a half-mile away may think upon hearing such.

And I saw your Zoom talk but it involves a lot of simple, circular, speculative reasoning… the ONLY solid auditory data KNOWN to come from IBWOs (and it too may be flawed in various ways), are the old Cornell recordings showing a Hz significantly higher than the 580-or-so you focus on — I think your data is inconlusive, but if forced to make a conclusion, the most logical, scientific one would be that those lower Hz sounds derive from some other extraneous critter or activity common to southern woods (of which you test only a few). Yes, those sounds could be from IBWOs and you can invent explanations of how/why they might differ from Cornell, but it is sheer speculation. That’s part of the point of “control” data, to see if the same 580 Hz level kents show up in various northern woodland which would completely compromise your conclusions. In general, people are allowing their own biases to let them ignore the problems, variables, and variety of interpretations of acoustic data.
On a side note, it’s not clear to me if you used ALL of Mennill’s recordings in constructing your chart (many of which almost certainly were not IBWOs), or you used only a subset which is then subject to your own biases of selection…

What we have in the IBWO arena is lots of acoustic encounters, leading to relatively few visual encounters, leading to no clearcut photographs or videos. It’s not a good look for us (…even though there are possible explanations).

john said...

I disagree and interested readers can see my logic in the Google Slides presentation attached in the first comment. It strains credulity to entertain the thought that there were hidden persons, playing kent calls, in the Dennis recordings-- and the Mennill recordings-- and the Cornell recordings-- and the Project Principalis recordings-- and two others that I was given to examine in confidence, where the searcher was not even announcing intent. This is simply a "no".

It strains credulity to believe that, of the hundreds of thousands of hours of recordings in other areas of the US (as found on xeno-canto or Macaulay Library for example), as Cyberthrush wishes as a "control" (which is really for experiments not discoveries-- I taught this for years), not a single person has demonstrated a sound that matches spectrographically with an IB sound. Not a kent, not a DK. So Cyberthrush already has his wish. AI is not needed; competent earbirders would have noticed an unusual unidentified sound and inquired. My presentation is valid; the hypothesis is there for future research to compare to. Here is the link again; notice there is a Slide 38 for logic-- https://docs.google.com/presentation/d/19RkgzdT4yMPu7cepPEaGqyxf_VrhZNdX4MRFGrpOafQ/edit?usp=sharing

cyberthrush said...


For those interested in this, it may be very difficult to actually follow all John’s points and reasoning by just viewing his slides without explanatory narration, so by all means if Matt uploads the Zoom talk to YouTube look for it there to see John's arguments (and assumptions) fleshed out more.
(I’ll try to remember to add a link here if it becomes available)