1. On the Auburn IBWO website, we see this (the bold font is mine):
...Likewise, double knocks could be atypical drum patterns of Pileated Woodpeckers, but such pileated double knocks must be rare because there is no mention of such pileated double knocks in the literature, and no members of our search team have heard pileateds make such double knocks along the Choctawhatchee or anywhere else in the southern forests...Wrong.
Check this out. Again, as noted here many times, double-knocks can also be traced to other woodpeckers, courting ducks, gunshots, wind in the trees, vehicles, etc etc.
2. An excerpt from Hill's paper:
Tanner (1942) describes the double knock as a “hard, double rap, BAM-bam, the second note sounding like an immediate echo of the first.” The timing of the putative double knocks that we recorded is consistent with this description of two raps in direct succession: the average delay from the start of the first knock to the start of the second knock was 0.115 ± 0.003 seconds (n = 99). In 45% of our putative double knocks, the first knock was louder than the second, matching Tanner’s description (Jackson 2002).Excerpt from "The Grail Bird", page 40 (the bold font is mine):
This BAM-bam is the characteristic drum of a Campephilus woodpecker, a genus found through much of South and Central America, with the ivory-bill being the northernmost representative of the group. "The second part of the double rap is so quick," said Nancy, "it sounds like an echo of the first and is nowhere near as hard." The space between the two parts of the double rap is only about seventy-five milliseconds, which is so close that some people hear them as a single rap. But the separate parts are clear if you look at a sonogram (a visual representation of a sound showing its pitch and duration).("Nancy" in the paragraph above is Nancy Tanner, who was married to the Ivory-bill expert Jim Tanner. Nancy Tanner heard real Ivory-bill double-knocks in the 1940s.)
From 11,000+ hours of ARU data, Hill et al cherry-picked the "best" 99 double-knockies that they could find. Even for those 99, the knocks are generally spaced much too far apart, and the relative amplitude of the two knocks is also generally wrong.
3. Regarding the recorded kent-like sounds, Hill et al admit there are lots of other sources:
Sounds that resemble Ivory-billed Woodpecker kent calls are produced by Red-breasted Nuthatches (Sitta canadensis), White-breasted Nuthatches (Sitta carolinensis), gray squirrels (Sciurus carolinensis), and Blue Jays (Cyanocitta cristata) (Jackson 2002, Tanner 1942), and may also be produced by Great Blue Herons (Ardea herodias) (R. Charif, pers. comm.).The admission above is all well and good. But at this point, they try to eliminate these other sources by using ridiculous arguments like this:
Great Blue Herons are common along the Choctawhatchee River, but their occasionally kent-like calls could be distinguished because they were followed in sequence by repeats of their more common squawk-like calls.Ok, so an isolated kent-like call is unlikely to be a Great Blue Heron, because of no surrounding context of convincing Great Blue Heron calls? But couldn't a nasty skeptic suggest that an isolated kent-like call is unlikely to be an Ivory-bill, because of no surrounding context of convincing Ivory-bill sounds (multiple Tanner-like double-raps, additional kent calls that actually match Ivory-bill recordings, loud wooden wing noise)?
Please also note that the duration of Hill et al's kent-like calls is far too long:
Putative kent calls that we recorded in Florida in 2006 (0.21±0.08 sec, n = 210) were slightly longer than calls recorded at the nest by Allen and Kellogg in Louisiana in 1935 (0.11±0.01 sec, n = 31).Why is the word "slightly" in the above sentence?
4. Now let's consider the "clustering argument" beloved by some. I don't see any meaningful correlation whatsoever between the recorded kent-like calls and double-knockies.
See for yourself by checking out Figure S1 on page 4 of Appendix 1 here.
In the comment section here, Andrew R wrote:
After a quick look at the "appendix1.pdf" file that has the pretty blue and red double-knockies and kents I noticed a couple of things. First, the scales of the diagrams are not the same. The increments vary from 10 minutes to 1 hour. Even if they aren't close together temporally, they sure look close together when a larger scale is used. Nothing nasty about that, but it's easily misinterpreted at first glance.If Ivory-bills were actually present in the search area, it's a near certainty that Hill et al would have captured a definitive video within a week or two of their initial "sighting". No ARUs would have ever been deployed.
From what I can gather from the graphs (too tired to read the data in the tables, and by "tired" I mean "buzzed") the closest kent and double-knocky pair is about 8 minutes apart - chart "D". And they were from different listening stations. (Is there a map somewhere of how the listening station were arranged? Saying they were 500 meters apart doesn't help much unless we know they were in a straight line.)
So, a kent and a knock were heard 8 minutes apart, maybe 1000 meters from each other, maybe more, maybe less. Where is the evidence that these were made by the same animal? If they were 10 seconds apart and from the same recording unit, you could make that argument. Oh, and I'm guessing the kent didn't really sound like the recording we have of an IBWO...
But leave that aside for a moment. Hill et al is suggesting that Ivory-bills were near their ARUs, day after day, yet were detected almost exclusively by isolated kent-like calls or isolated double-knockies. There's no mention of any ARU ever picking up any Ivory-bill wing noise. Did the Ivory-bills usually walk into ARU range, kent or double-knock a time or two, then silently walk out of ARU range?
More on expected sound combinations is here.
5. We're told that each of the seven ARUs was visited daily to change memory cards and batteries.
The ARUs keep picking up "Ivory-bill" sounds, day after day after day, but the person visiting all the ARUs never ever gets a decent look at a perched bird?!
6. As a wise person once said about Cornell's evidence: "There's no control for this experiment".
As a control, large amounts of ARU data could be collected at night, and also far away from the historic Ivory-bill range. It should be analyzed by people unaware of the time or place of recording. Under those conditions, I'm confident that a whole lot of "Ivory-bill" evidence would be found.
Bottom line: Hill et al's data is quite convincing proof that there were no Ivory-bills in their search area.
Posts
19 comments:
There's no mention of any ARU ever picking up any Ivory-bill wing noise.
If they had a recording of some IBWO wing noise (or some facsimile of it) they could undoubtedly use their same technique to find some "matches" in that 11,500 hours.
But that would prove nothing about the existence of a living IBWO in the area.
Tom,
I appreciate the care and thoroughness you’ve put into this. I’ll reply with equal care and thoroughness later. In the meantime my sense is that on a number of levels your approach to the audio is colored by circumstances surrounding the CLO audio, which simply do not apply here. For example, CLO made a claim of definitive proof, so merely casting credible doubt on their analysis was sufficient to undermine their claim. I don’t think that casting doubt alone is appropriate in this circumstance. Team panhandle has offered a hypothesis to explain a recorded phenomenon. Casting doubt on the hypothesis is certainly easy, but in this case only a part of the challenge. Replacing that hypothesis with a better one is where the challenge will be. I don’t think a great hypothesis is needed, just a vaguely plausible one. And I don’t think we’re close to that yet. That being said, I don’t think it’s helpful for me to respond to your post for what it’s not. I should respond to what it is. And I will.
pd
"Replacing that hypothesis with a better one is where the challenge will be."
It's no challenge at all. The Florida evidence is just another pile of the "unsubstantiated sightings, blurry photos/videos, double-knocks, kent-like calls, bark scaling, etc" stuff that we've seen countless times before.
Next season's Florida search will yield yet another pile of this evidence, but no definitive photo.
There are many non-Ivory-bill sources of all this "false positive" data, and we've discussed them ad nauseum here.
Team panhandle has offered a hypothesis to explain a recorded phenomenon.
No, pd, that's wrong. You've got it backwards.
Team panhandle offered a "recorded phenomenon" as evidence to support their stated belief that the IBWO is still alive.
But it's manufactured evidence and tells us more about the human beings who manufactured the evidence than it does about nature.
Sure, in a metaphysical sense the noises "could" be a race of IBWOs that has evolved to run away from cameras.
But in the same sense, the noises "could" be a race of reclusive aliens from outer space who recently landed in the swamp, or Bigfoot. After all, pd, you can't "prove" that the noises weren't made by Bigfoot or aliens from outer space, can you?
To be crystal clear, pd: the burden is not on the skeptics to "explain" whatever "unexplainable" garbage the IBWO peddlers cook up.
The burden is on the IBWO peddlers to explain why they can't take a frigging picture of this large noisy bird when other far more credible scientists take excellent pictures and movies of animals living at the bottom of the ocean or in the canopies of tropical rainforests.
Here's my explanation: Hill et al. are deluded amateurs. Their paper is a quintessential example of chicanery and, like the 2005 Science paper, merits every ounce of ridicule and scorn heaped on it.
I agree that Hill and company are on shakey ground here, but as pd points out, they are not CLO, and they freely acknowledge they don't know for sure what they have. Three points you made that I don't think are particularly strong:
1)"There's no mention of any ARU ever picking up any Ivory-bill wing noise. Did the Ivory-bills usually walk into ARU range, kent or double-knock a time or two, then silently walk out of ARU range?"
Um...if the double knockies are being made by other woodpeckers, especially PIWO's, doesn't this criticism apply equally to them?
2) "Please also note that the duration of Hill et al's kent-like calls is far too long:"
Yes and no. They are longer than the '30's recording. However Tom, as a "birder with many years of experience" obviously you have had the experience that a bird's call almost always changes in both pitch and length when confronted by intruders near a nest as was the case with the 30's IBWO recording. Whether the Hill recordings are too long is genuinely unknowable at this point.
3) "Even for those 99, the knocks are generally spaced much too far apart, and the relative amplitude of the two knocks is also generally wrong."
Too far apart--I think this is a legitimate point. Wrong relative amplitude--not so much. We know other Campephilus woodpeckers double knock both ways at least occasionally. I don't think it is unreasonable to conclude that IBWO's might behave similarly.
I point these out because there is no need to overextend your argument into areas where it isn't strong. In fact there is really no need to make the argument at all. Either it really is an IBWO making those noises and they will prove it with more field work, or it will remain an unknown and the IBWO hypothesis will fade into history like Arkansas.
1. "Um...if the double knockies are being made by other woodpeckers, especially PIWO's, doesn't this criticism apply equally to them?"
Um...no. Tanner wrote this about Ivory-billed Woodpeckers only:
---
The wing-feathers of Ivory-bills are stiff and hard, thus making their flight noisy. In the initial flight, when the wings are beaten particularly hard, they make quite a loud, wooden, fluttering sound, so much so that I often nicknamed the birds 'wooden-wings'; it is the loudest wing sound I have ever heard from any bird of that size except the grouse. At times when the birds happened to swoop past me, I heard a pronounced swishing whistle.
---
2. "Whether the Hill recordings are too long is genuinely unknowable at this point."
Fair enough.
3. "Either it really is an IBWO making those noises and they will prove it with more field work, or it will remain an unknown and the IBWO hypothesis will fade into history like Arkansas."
They've already had over 16 months to prove "with more field work" that Ivory-bills exist in a 2-square-mile area. That's more than enough time.
"But in the same sense, the noises "could" be a race of reclusive aliens from outer space who recently landed in the swamp.."
Alien, are you messing with the Hill o'Beans crew?
Has anyone listened through the ARU recordings for distinctive "Kneep.. kneeps..." clustered near the double knockies and kent-like utterances?
Enquiring minds want to know!
P.S. I would not want my idle speculating to taint the fine reputation of this blog. This speculation is strictly anonymous and does not necessarily reflect the opinion of this blog or its owner.
a bird's call almost always changes in both pitch and length when confronted by intruders near a nest
Hilarious.
What else do bird's almost always do when confronted by intruders near a nest?
In fact there is really no need to make the argument at all.
That's what we're trying to Hill et al.
Tom
They've already had over 16 months to prove "with more field work" that Ivory-bills exist in a 2-square-mile area. That's more than enough time.
Nothing to add except that this obvious and damning point can not be emphasized strongly enough.
I don’t think a great hypothesis is needed, just a vaguely plausible one. And I don’t think we’re close to that yet.
pd - Isn't the hypothesis that the knocks were made by other woodpeckers, the holes created by other animals/branches breaking off, the bark peeling by Pileateds, the sightings mis-identifications and the kents by other animals/things a plausible hypothesis? I realize that I present no evidence here to support this hypothesis, other than common knowledge. But in what way is the IBWO hypothesis stronger than that one? Hill et. al. agree that each bit of physical evidence could have non-IBWO sources.
Is the only compelling evidence the fact that all of these things were found in proximity to each other? Even if it is, the common knowledge hypothesis is still at least as reasonable as Hill's (IMO).
So, does this common knowledge hypothesis meet your threshold of plausibility? If not, why? Please explain your answer.
More from the Veracruz talk. The sound analysis seems to have been done visually by undergraduate students. By searching for the sound signatures of double knocks and kent calls, they were able to process a day's worth of recordings in about an hour.
Tom tells us, “there are many non-ivory-bill sources to this “false positive” data and we’ve discussed them ad nauseum here.”
Unfortunately, none fit the context audio recordings. As we’ve seen, it’s difficult to support or refute a theory about a single out of context sound as we were forced to do with CLO, but when we have hundreds of sounds in a context of time a space, and when we can establish enough similarity between many of them to determine that the majority are likely to have a similar source, then many theories become very easy to refute. Interestingly it’s the non-IBWO theories that are refuted by the context here. Consider a distant stop light on a truck route where semis backfire from time to time, bang bang. There are at least five contextual reasons why those backfires can’t explain the majority of the double knocks. Backfire would occur round the clock, they are remote to the site, in this example the source is stationary, backfires are extremely loud, and they are occasional (not occurring in “clusters”). Given the movement of the sound among various ALS locations, any one of these five conditions alone is sufficient to preclude a potential “false positive” from explaining the chronology of the evidence. When you’ve ruled those out, so far as my vivid imagination can reasonably determine, we are left with exactly one possible “false positive”: that would be one or more diurnal animals moving freely on the site.
That leaves us a very limited number of choices: geese, moorhens, cranes, jays, none to my ear typically vocalize in a way that in even nearly fits the very consistent range of the many recorded toots. (I don’t know about local mammals, but that would be a good thing to check out.) For Tom to be ruling out IBWO as a source for toots based on a one tenth of a second measurements from a single historic sample fails to recognize the well established variability of bird sounds within a species. When refuting CLO it was fine to say, “yes, but remember blue jay vocalizations vary greatly”. But now the variability that was promoted to refute CLO, is denied to refute Choctawhatchee. In terms of sound similarity alone, and from what we know of IBWO, IBWO is simply the best fit, maybe the only fit. I remain open to other proposals.
The same is true of the double knocks. Tom denies the possibility of a 40 millisecond variability in knock measurements, while advancing a non-typical pileated behavior to account for regularly occurring clusters of 10 or 15, and at one time 45 Campephilus-like double knocks. OK, but of course the point is that the knocks are Campephilus-like and IBWO again is simply the better fit. We’re stuck reaching for non-typical explanations, something that was never necessary with the CLO evidence.
By repeatedly saying “there are all kinds of possibilities”, when in fact there are very very few, the mind is encouraged to wander… imagining it is yet to find a fit for the audio that isn’t such a reach. No, there are only a few possibilities. And they are a reach.
Because there is considerable in overlap in Tom’s points one through four from the original post, I’ve somewhat grouped them together in my comments above, but I would like to address Tom’s points five and six separately. In five, Tom is surprised that they checked the ARU’s daily but never got a photo. True, but that’s meaningless in analyzing the audio. They didn’t get a photo of IBWO vocalizing, just like they didn’t get a photo of a crane vocalizing, or a gate hinge squeaking. It adds nothing to discussion of the audio. In six Tom states there are no controls for this experiment. Absolutely untrue. The Pearl and the Big Wood provide tens of thousands of hours of control. Nothing like the Choctawhatchee audio could be found there, even by imaginative CLO types.
pd
pd,
you're falling into all of the traps that we've been trying to point out all along.
First, I don't think you can assume that these sounds were only recorded in the daytime, Hill et al don't tell us anywhere that they searched ALL of their recordings, and even if they did the students searching for these sounds would be inclined to speed through the dark hours or to ignore possible sounds at night since we all know that Ivory-billeds wouldn't be active then.
The so-called "groups" of sounds look like little more than random distribution of these recordings. Imagine that you're birding and you hear two sounds separated by two hours - would you call that evidence of a single bird? None of these "groups" consist of a nice clear series of dozens of kent calls and double knocks in a five minute span. (Now that would be convincing)
As for imagining other sources, it's not our job, but the recording labelled "4 double knocks" from Jan 20 sounds like a small woodpecker tapping idly as it searches for food. I suspect that many of the other knocks are also simply foraging birds, but we only hear them in isolation, not in context.
Contrary to your suggestion, there is wide variation in these recordings, and it's upside-down to say that this just shows the expected variation in kent calls of Ivory-billed Woodpeckers! This is like Cornell's claim that Ivory-billeds flap more quickly than pileateds, and the now-popular myth that you can identify an Ivory-billed by it's "loon-like" flight! It is all speculation and proves nothing!!!!! ARRGGH!!!!!!!
From figure S1 I, J, and K we can see that all of the recorded kent calls on those days came from a single listening station. This is interesting, and suggests that the sounds perhaps don't move around as much as you imagine. But Hill et al don't tell us anything about the spatial distribution of the sounds. Maybe something in the metal stand holding the microphone was squeaking when the wind blew?
The fundamental point here is that nothing is KNOWN. You say it's a reach to find any other possibility, and speculate along with Hill et al that these sounds are Ivory-billeds, but THAT is really a reach.
Consider for a moment, along with all of the other evidence, that 11,000+ hours of sound recordings has yielded an assortment of only 300 unexplained sounds, all similar to Ivory-billed sounds but NONE of which actually match known recordings of Ivory-billed.
pd
The Pearl and the Big Wood provide tens of thousands of hours of control. Nothing like the Choctawhatchee audio could be found there, even by imaginative CLO types.
Pure self-serving rhetoric, pd, without any basis in fact.
Until the experiment is done, you have no idea what "cannot be found" by several students searching all summer through ten thousand hours of audio from Big Wood for "matches" to IBWO "audio imprints."
"The so-called "groups" of sounds look like little more than random distribution of these recordings."
No look again, they do not look at all random. Not even close.
Because they have a number of sounds an hour before sunrise, but all sounds, without exception, end an hour or more before sunset and none of the three hundred sounds occur in the last hour of the day, it seems to me they did as they suggest, and analyzed 24 hour of audio. Its hard to believe they analyzesd before sunrise but quit before sunet.
pd
pd,
I take it your point is that in the Florida case, the IBWO hypothesis is actually the most reasonable/likely explanation for the kent sounds. You base this on what you say is the fact that other possible explanations are much less likely ("a stretch"). I disagree with that. Or at least, I don't think the other explanations have yet to be proven to be less likely.
Let's say that the sounds were produced by a diurnal animal. I think you're right on that one. I have not listened to all of the kent sounds, and I know nothing about spectral analysis. To me, some of the sounds sounds like they come from a Canada Goose. Some don't, however. But you have said they couldn't be geese because there are too many recorded sounds that are similar, yet would be atypical for a goose. Others claim that nothing in the southern woods makes those sounds. Fine, but I think that needs to be proven, not just stated and accepted as fact. In fact, those are things that, I assume, could be relatively easy to prove. There are enough of those recordings and God knows there are enough Canada Geese to enable some scientist to prove whether or not those sounds could have come from geese. (Hell, we could ask Mike Collins to do the experiment since he's an unbiased expert on how sounds travel through forests.) I'd be interested to know if the recordings were searched for typical Canada Goose sounds and other diurnal animal sounds, and how those sounds correlate with the kent calls.
And while they're conducting the goose experiment, they can do the same for all of the ducks and mammals in the area, including juveniles. You know, do some science. The facts are out there just waiting. It's low hanging fruit. This process of ruling out alternative hypotheses just doesn't seem to be too popular in Florida. Too much effort is being put into getting "positive data" and very little getting "negative data". Maybe that's what they have next on their to-do list. I hope so, because that's where you're really going to get your bang-for-the-buck (or doe!).
Now, if it is _shown_ that geese/ducks/mammals could have made those sounds, then there's your reasonable and more likely alternative hypothesis. If, however, it is shown that ducks/geese etc. could not have made those sounds, well, then you're one step closer to being on to something. But until then, positing the existence of a bird generally believed to be extinct and not sighted with certainty for many years, while at the same time saying that the currently available hypotheses are a "stretch" compared to the IBWO hypothesis, is, with all due respect, the pot calling the kettle black.
As an aside, I just want to add that the double-knock recordings are nearly pointless. Without a BAM-bam, all you've got are woodpeckers. If you follow around several woodpeckers for 4 months, which is in essence what they did, you're going to hear them peck wood twice in a row, whether it's "in the literature" or not. I'm sure all of them are fully capable of knocking at the rates that were recorded. I can make those double-knockie noises with a little practice.
jesu christo, we are actually arguing over this terrible terrible data that Auburn has submitted?
Are you guys and gal out of your collective mind?
jesu christo, we are actually arguing over this terrible terrible data that Auburn has submitted?
Are you guys and gal out of your collective mind?
We need to keep pd occupied so he/she doesn't flee the country with my $1000.
Post a Comment