Wednesday, January 18, 2006

Summary of the Arkansas Ivorybill debate points

Trying to consider the major points that have been brought up concerning the credibility and strength of the Big Woods Ivorybill evidence, and if they can be resolved without additional evidence.

1. Arguments about the possibility or impossibility of the species having survived 60 years without physical evidence or unequivocal photographs. These amount to differing opinions about the probability of a single event. They will not lead to any definite conclusion about whether this event, however likely or unlikely, actually occured.

2. Arguments about the suitability or unsuitability of the Big Woods habitat. These are somewhat speculative and hypothetical, given the relatively small number of life history studies that were conducted before the species' disappearing act. Extrapolation from then until now is interesting, but not definitive, and again will not lead to any definite determination about whether any Ivorybills are in fact utilizing the habitat in question.

3. Alternate interpretations of the Luneau video. To me, this would seem to be the one place were a rigorous analysis could actually make a determination between the alternatives. There are two widely proposed hypotheses: (a) The images are of an Ivory-billed Woodpecker, and (b) The images are of a Pileated Woodpecker. For now I will set aside the abberant Pileated alternative. It seems to me that it should be straightforward to determine unequivocally which of these hypotheses matches the images. It should be possible to analyse the flight of the bird using digital 3D models based on documented flight of at least the Pileated, and determine the actual pose of the bird relative to the camera frame-by-frame. We should not have to argue about whether the upperwing or underwing is visible in each frame, this should be able to be determined without subjective guessing. It should then be possible to photograph models of each species placed in these positions, using the exact same video equipment, under lighting conditions that simulate those of the original video. From this it should be simple to determine if the apparent white on the upperwing (suggesting Ivorybill), and the apparent black trailing edge (suggesting Pileated) could both be generated by either species. It should also establish definitively whether or not underwing white could appear so brilliant in an image collected by that camera under those conditions. There will be complications with motion effects, but these should also be subject to quantification and analysis. This analysis should conclude that the images could have been made by one species and not the other. Or, they might conclude that neither is a match, in which case we arrive at the abberant pileated hypothesis as the probable best explanation. I know some simulation with models was done. But was it done as throughly as I just described? If so, why are we still arguing this point? If not, why hasn't it been done?

4. Reliability of sight records. There seems to be no progress to be made on this point.

5. Audio data. There is probably little additional analysis to be done here.

6. The abberant pileated hypothesis. Though often cited, there seem to be very little actual data that have been made available for evaluation of this. The references to the existence of these birds have been in passing, and none of the supposed photographs have been produced. This is a place where more data is sorely needed. There are fewer publicly known reports of sightings purported to be abberant Pileateds than those purported to be Ivorybills. It does seem that a clear statement and release of data on this issue would go a long way towards resolving it.

So in summary, the one piece of evidence that might be subject to further rigorous analysis is the Luneau video. If no more evidence is obtained in this current search season, this is probably the one place that it will be possible to determine if there actually was one Ivorybill there last year and it vanished, or if indeed there never was one at all.

Comments welcome; however, unsigned anonymous comments will be deleted. Thank you.


At 6:01 PM, Anonymous Patrick Coin said...

Good to see you writing about this. A very good distillation of the issues.

If you would send me your e-mail address, I will e-mail you the series of frame captures I did of the Luneau video. It is a very interesting way to look at the video, and I feel it is easy to pick out the individual wingbeats. I'd like to know what you think. I have some more work to do on it, such as add frame numbers, but perhaps you'd like to see it now.

My e-mail is
Simply replace the "SPAMNOT" with "@" of course. That's my feeble attempt to defeat SPAMbots.

At 9:10 PM, Anonymous Don Crockett said...

Using 3-d modeling software one should also be able to simulate the camcorder's view of a 3-d model as far as pixels and shutter speed.

At 3:01 AM, Blogger Bill Pulliam said...

Very true, and it's a complete mystery to me why the Cornell team never did an analysis like that


Post a Comment

<< Home

Site Meter