Wednesday, March 14, 2007

A bit more about the Avian Phantoms

One final thing I want to do here on this topic of overlooked and unseen birds...

Though it isn't possible to rule out the persistence of the phantoms, I think it is possible to make some more ballpark estimates, this time of the maximum plausible population size that could be consistent with the data we have. Swallowing the big frog first, I'll start with the Ivory-billed Woodpecker. Please accept that I am going to use the most generous and optimistic interpretations of the existing data that I think are plausible, and don't fill up my comments box with all the (well-known) alternative interpretations that yield much lower (i.e. zero) population estimates. We can take that as a given.

Some very generous, hopeful, and optimistic things have been said about potential population sizes for this species, such as "a population near carrying capacity in the White River NWR" and "as many as 10 pairs on the Choctawhatchee." Most of these were said early on, and I think it is safe to say that they have proven untrue. For the Big Woods, given something on the order of 20,000 hrs of field time in recent years (I'm guessing at that number, it's 10 parties in the field 10 hrs a day 5 days a week for 40 weeks), that gives 2000 square km covered (obviously a lot of duplicate coverage there) and no nesting pair found or even hinted at. I think the most generous plausible interpretation of their 10 or so "robust" sightings, ARU results, and one lousy video is the presence of only 1 or 2 non-breeding birds that might not even be in the area all year. I also think it is clearly implausible to expect there to be 10 pairs in the Choctawhatchee area. With a team of full-time observers, brief sightings are happening every few weeks; the more robust sightings with more detail have come only a couple of times a year. One thing I am not buying in to is the notion that the Ivorybill is preternaturally wary and elusive. To me, this low encounter rate indicates a very low abundance, no more than one bird per roughly 10 square km, not magical evasive powers. That doesn't leave anywhere near enough room for 20 adults plus young of the year in that river basin. There is another point also. If there were that many Ivorybills in the Choctawhatchie, one might expect similar colonies to show up in other rivers in the Florida panhandle. Dispersal between basins should not be a big problem. But if we had 10 pairs in the Escambia, Perdido, Coldwater, Blackwater, Appalachicola, etc. then we'd be talking about 100s of Ivorybills in an area with many highway crossings and swamp-front houses. Such a density in this area would not be overlooked for a year, much less 50 years. Birders traversing I-10 would see Ivorybills cross the highway in front of them on a regular basis. In fact, this is only reported to happen every few decades (and then in Louisiana, not Florida). I think the most generous interpretation that can be put on the Choctawhatchee at this point is a few individuals, maybe a pair plus a couple of non-breeding offspring, and a bunch of Blue Jays that have gotten very good at imitating Ivorybill calls to confuse the field workers and the ARUs. And I think we also must conclude thay this density is not "typical" for the panhandle rivers; that is, there is not a pair or two in every swamp down there, just a very fortunate few.

What about the Pearl, Atchafalaya, Congaree, all the other places? Well, for the Pearl, the same arguments apply as above. If there were more than an isolated pair (maybe two?) in the area, photos and multiple sightings by multiple observers would have almost surely happened by now. Mike Collins has not been the only camera-equipped birder there in recent years. The Atchafalaya is certainly big, difficult of access, and mostly private. But even our resident Ivorybill-searching duck hunter down there (the enigmatic "Choupique") only claims a couple of personal sightings in his lifetime. Again, this suggests to me no more than one bird per every 100 square km in the basin. This is only a handful of pairs at best. As for all the other areas, we have isolated sightings, and those only very occasionally. I personally have probably spent over 1000 hrs in "potential Ivorybill habitat" in Georgia and several other states, without a glimpse, kent, or rap, not even a suggestion of such a thing. Once again, this implies less than a bird per 100 square km.

My "optimistic" conclusion from all this is not in fact very optimistic. If we have an average density of <0.01 bird per square km in our approximately 10,000 square km of this habitat, then we have a global population of less than 100 birds. Personally I feel that is too generous, and a number of no more than a few dozen is all that is likely supported by the evidence. This might not sound too dreadful at first glance; after all we've had other species drop to those levels and then rebound. The difference though, is that those Cranes, Condors, and Warblers were all in one place, where they could be monitored and managed and where they could find each other. These couple of dozen woodpeckers, if they exist, are heavily dispersed, with no more than single digits in any individual State. And this is an upper estimate; the lower estimate would be in the single digits worldwide. There is no way to describe this other than with words like "bleak," "dire," and "desparate."

Having now gotten the big frog out of the way, the remaining two (Eskimo Curlew and Bachman's Warbler) are relatively easy. During the latter decades of the 20th Century, they were reported at roughly the same frequency as the Ivory-billed Woodpecker: a few reports a decade, and "good" reports less than once a decade. So, skipping all the intermediate steps, I come to the same conclusions, and the same bleak estimate for an upper limit for the population that is consistent with "phantom" status: a few dozen at best. Zero at worst. But the difference between zero and slightly more than zero is the difference between night and day.


At 1:46 PM, Blogger cyberthrush said...

one small point Bill: just in fairness to Dr, Hill, I believe the "10 pair estimate" pertains to the entire length of the Choctawh. river basin from AL. thru FL. Panhandle and not just the small patch (2 sq. mi. I believe) that the Auburn group is focussed on -- if you assume even just 1 pair where they are, the 10 pair figure for the whole basin isn't as totally far-fetched as it may first sound.

At 1:59 PM, Blogger Bill Pulliam said...

I was basing my estimate on the whole basin as well. If they really had a pair of birds living within that 5 sq. km, that's a density of 0.4/km2 and they should be having real, in-the-flesh Ivorybill encounters at a rate of something like 1 per every 20 hrs of field work. Their actual encounter rate seems to be much less than this, so if they have that pair of birds, they are ranging over a much larger area than just those 2 sq. mi. As I said, I don't buy the hyper-elusive hypothesis. There is no a priori reason to expect them to be more difficult to detect than any other large forest bird. I think the difficulty of finding these birds is purely a function of there being almost none of them out there.

At 4:37 AM, Anonymous Anonymous said...

Very interesting series of posts. I'm inclined to agree with your general point that we're dealing with very small populations (I'm undecided about the Bachman's Warbler and Eskimo Curlew, leaning toward extinct for the former and extant for the latter. . .just my purely speculative take). In your previous posts you drew a distinction between migratory and non-migratory birds, and I think that raises some questions about the final one, questions that make the Ivory-bill's survival (assuming it has survived) all the more mysterious.

It's easy to understand how a small population (say 2 dozen pairs) of migratory birds could survive, at least for a while, assuming the birds share summer territory. Ivory-bills, conversely, are non-migratory, so it strains credulity to think that the Choctawhatchee, Congaree, Atchafalaya, etc. each support only one, two or, at most a handful, of breeding pairs. If that were so, I would think these scattered tiny populations would have blinked out due to predation, inbreeding, disease, etc.

Also, if the Arkansas sightings are of a transient or transients or a dispersing male (I think I may have been the first to suggest this, at least online) it raises several questions: why, why now and where did it/they come from?

Although I'm persuaded the Ivory-bill has survived, this strikes me as the strongest argument the skeptics have; a couple of pairs, or a handful of scattered, isolated ones, does not a viable population make.

I don't buy into the hyper-wariness theory either, but I suspect there must be more than a handful of birds in one or two of these areas, making possible both dispersal and a slow but steady increase in population.

At 8:20 AM, Anonymous Anonymous said...

"If there were more than an isolated pair (maybe two?) in the area, photos and multiple sightings by multiple observers would have almost surely happened by now."

Campephilus are still detectable when reduced to a single pair. Here is a quote about Crimson-bellied Woodpecker from "The Birds of Ecuador Field Guide" (p. 436 of Status, Distribution, and Taxonomy): "[recorded] south to Tinalandia, where a pair persisted into at least the early 1990s;v.o." Thus, various observers were able to locate ca. 2 Campephilus, presumably while wearing normal clothes, using of ARUs, and without staking out intriguing cavities or bark-scaling.

Closer to home, Pileated Woodpeckers are routinely detected even in areas at the edge of their range where their population size is minimal, even in densely forested areas such as Mt. Tamalpais, Marin County, California.

At 8:47 AM, Blogger Bill Pulliam said...

As I mentioned before, I don't believe the "extreme wariness" theory, and the numbers I scribble on the backs of my envelopes are detectabilities comarable to Pileateds in the same habitats. The only explanation I find acceptable for the difficulty in locating these birds is extremely low population density. Looking at videos of other Campephilus sp., these birds seem to be double-rapping all the time. I can accept that a single, unpaired bird might be less "rappy" and less vocal, but it should still be only a matter of degree and not kind. And if we have nothing but unmated birds, there's no viable species anyway. I can also accept a larger home range for non-breeding birds in habitat with a suboptimal food density. These are the reasons I maintain that a small dispersed population could easily remain almost undetected (by birders) for decades on end. But somewhere out there, if there is any meaningful survival, there has to be at least one mated pair rapping and vocalizing. If this pair doesn't exist, the species might as well be extinct even if there is a bird here or there.

Sort of a side note -- seeing videos of how communicative pairs of tropical Campephilus sp. are actually encourages me, in an odd way. It means that if a regular birder did happen within 100 meters of a nesting pair of Ivorybills, s/he has a good chance of noticing them and being able to positively observe them. If, if, if...

At 8:48 AM, Anonymous Anonymous said...

I meant WITHOUT using ARUs of course.

If Eskimo Curlew persisted it would be seen in fall migration in New York and New England, among other places. These are not underbirded areas. Likewise, Bachman's Warbler, if extant, would be seen in migration in Florida. Given that shorebirds with minimal population sizes (usually absent, but likely only 1 or at very most a few vagrants when present) in North America are found and photographed...Little Curlew, Terek Sandpiper, Spoonbill seems logical that a larger population of Eskimo Curlews, even if as small as 5-10 individuals, would be well documented.

At 8:52 AM, Anonymous Anonymous said...

Regarding Neotropical Campehilus:

"if a regular birder did happen within 100 meters of a nesting pair of Ivorybills, s/he has a good chance of noticing them and being able to positively observe them."

Actually, even a NON-birder has a good chance of noticing a calling Campephilus pair, as I've witnessed myself.

At 10:11 AM, Blogger Bill Pulliam said...

Actually, we have no idea how many individuals of those vagrant shorebirds occur in North American every year, but it seems ludicrous to me to think that we find most of them or anything even remotely approaching that. It is a logicl fallacy to deduce "we only see one or two a year, therefore there are only one or two a year, therefore we detect all the ones thta occur." We also don't know how much of any autumn Eskimo Curlew migration would take place through the populated parts of the US versus the very sparsely populated portions of eastern Canada, where they would be very easily overlooked. There is also no reason to assume that all (or even the majority) of the population of Bachman's Warblers would pass through the Florida Keys or Peninsula. Just because there are multiple historical sightings doesn't mean that is a mandatory stopover for the entire population. There are scattered Kirtland's Warbler records from east coast states; but the bulk of the population flies over these areas without stopping every year. Cuba to South Carolina is a quite ordinary distance for migrating warblers to cover in a single hop. These are examples of the sorts of certainties that birders like to declare, but even a cursory examination of the probabilities involved suggests otherwise.

At 11:28 AM, Anonymous Anonymous said...

Implicit in your comments is the notion that overlooked individuals of extremely rare Eurasian shorebirds such as Spoonbill Sandpiper, Little Curlew, and Terek Sandpiper collectively occur in Canada and the lower 48 more frequently, conspicously, or in larger numbers than a viable population of Eskimo Curlew, as the former have been documented recently by birders but not the latter. Do you really think there are that many overlooked Terek Sandpipers and Little Curlews? I find it impossible to believe that a viable population of a shorebird known to migrate through New England, New York, and maritime Canada could go undetected for decades. There are no recent records from these areas because the bird is extinct, not because these areas were outside its migration route.

At 11:46 AM, Blogger Bill Pulliam said...

Implicit in your comments is the notion that overlooked individuals of extremely rare Eurasian shorebirds such as Spoonbill Sandpiper, Little Curlew, and Terek Sandpiper collectively occur in Canada and the lower 48 more frequently, conspicously, or in larger numbers than a viable population of Eskimo Curlew, as the former have been documented recently by birders but not the latter.

Of course, and I spelled out the logic behind this in my earlier posts in this series. I won't repeat that all here. My whole argument is that even given the lack of observations we can't discriminate between a very small population and no population in a statistically meaningful way.

Eskimo Curlew autumn migration took place primarily along the arctic shores of Canada eastwards and then southwards over the Atlantic. The bulk of the population did not pass through the eastern US or the heavily populated regions of Canada.

At 7:28 AM, Anonymous Anonymous said...

I think the hyper-elusive hypothesis should be considered seriously. As a European birder with a little bit of North-American experience, I can say that most large birds there (Pileated Woodpeckers, Turkey Vultures and others) appeared extremely tame to me. (An exception were Wood Duck, which seemed very wary.) Perhaps hunting has been more intense during a longer period here in Europe, forcing more species to develop shyness. When I look at the proportion of good observations of Ivory-bills to fleeting observations lasting a few seconds or sound detections, this reminds me of the extremely shy Green Woodpecker here in Europe. There must be one pair using the part of the village that I live in. Yet, in spite of houses and people everywhere, and few trees to hide in, I have only one good observation of this bird in more than a year. I hear it regularly and sometimes see it flying away. One should realize that when we see a bird, this is almost always because the bird allows us to. Of course, no bird has supernatural powers and good observations of IBWO's have occurred (e.g. by Kulivan quietly sitting in camouflage) but concluding from the rareness of such encounters that there can't be 10 pairs on the Choctawatchee seems unreasonable.


Post a Comment

<< Home

Site Meter