Some Thoughts on The Ivorybill’s Decline And A Change of Plans

Jamie Hill, who has worked with the Cornell and Auburn teams, recently posted a Facebook link to a very interesting article from the September 2014 issue of Smithsonian. Ivory-billed Woodpecker aside, the piece is well worth reading, but for the purposes of this blog, the article got me thinking about reasons for the ivorybill’s decline and the possible role of the longleaf pine. These ideas are not entirely new or original with me; Lester Short went even further, suggesting that pine might have been the ivorybill’s primary habitat; Jerome Jackson devoted several pages of In Search of the Ivory-billed Woodpecker to pines, and Fangsheath of the ivorybill researchers forum has hinted at this too.

I was struck by just how congruent the historic range of the ivorybill is with the range of the longleaf pine (Pinus pilastrus). The overlap is not exact, and the pre-Columbian range of the ivorybill extended as far north as Ohio. Nonetheless, conditions in the Singer Tract were objectively quite different from what they were in many other parts of the historic range.

A recent blog post on the Tallahassee Democrat site reiterates the conventional wisdom about the species and the reasons for its decline. Author Budd Titlow writes: “Before the Civil War, when much of the southeastern U.S. was covered with vast tracts of primeval hardwood swampland, ivory-billed woodpeckers ranged from North Carolina south to Florida, west to Arkansas and Texas, and north into Oklahoma and Missouri. Then, after the Civil War, extensive logging of these old-growth swamps wiped out most of the ivory-billed’s habitat in one fell swoop.”

While there’s some truth to this history, it’s also a stereotype that’s based in large part on an imperfect reading of Tanner’s monograph and even more on Tanner’s dedication to protecting the Tract as the last remaining extensive old-growth stand in the southeast (although the Tract contained considerably less old growth than Tanner believed). Tanner’s efforts were admirable; the loss of countless acres of magnificent old-growth swamp forest was devastating environmentally and is unquestionably something to be mourned, but it seems unlikely that the destruction of these forests was the primary cause for the ivorybill’s decline.

The species was known to be disappearing by 1890 or even earlier, and Chester Reed’s 1906 Bird Guide to Land Birds East of the Rockies stated that the birds were restricted to isolated parts of Florida and possibly to “Indian Country” (Oklahoma). In The Travails of Two Woodpeckers, Noel Snyder, who attributes the decline primarily to hunting, points out that intensive logging of bottomland hardwoods began between 1890 and 1900. Logging of pine forests began considerably earlier, and these forests were severely fragmented, even before the Civil War. Snyder reads the early record (I think selectively) as indicating that ivorybills strongly preferred bottomland hardwoods and seldom used pines, in contrast to the Cuban ivorybill and the Imperial.

Jackson takes a different view, citing multiple references to the use of pines for feeding and nesting. Where Snyder reads Alexander Wilson’s early account as reflecting a preference for “swamps and bottomlands”, Jackson reads him as describing the preferred Carolina habitat as “a mosaic of baldcypress swamp and pine uplands, similar to the habitat in Florida”. Jackson goes on to suggest that, “It appears . . . that ivory-billed woodpeckers will inhabit both hardwood forests of river bottoms and pine forests of higher elevations, particularly old growth forests supporting healthy populations of beetles. They seemed to do best at the interface of these forest types, taking advantage of the resources of each.” (Emphasis added).

This meshes well with what Allen and Kellogg observed in Florida in 1924; the birds nested and roosted in cypress and were observed and photographed foraging in open pine forest. The Lambs’ limited observations in Cuba suggest something similar, a preference for roosting in pines but an equal division between pines and hardwoods for foraging.

Thus, it seems possible that the Singer Tract was actually suboptimal habitat for the ivorybill, since it contained no pine and little cypress. I’m also led to suspect that habitat fragmentation, rather than habitat loss may have been central to the decline of the ivorybill, with hunting as one of several other contributing factors. This fragmentation actually began well before the Civil War, but it accelerated with the post-war destruction of the longleaf pine forests, followed by the logging of the bottomlands. I’m personally convinced that the species beat the odds and survived, using one or both of the strategies discussed in this post. I wonder whether some of the modern search efforts have focused excessively on the bottomland hardwood model and not enough on areas where there’s an interface between forest types.

On a different note, I had planned to make my final trip to our search area for the season during this week and next. Water levels are very high right now, so I’ve decided to postpone until late July. Better to endure the heat and humidity than to be unable to move around in the woods.

Trip Report, Part One: March 31-April 5, 2015

As always, my time in our search area was very productive – inspiring new insights and ideas and producing suggestive but inconclusive evidence that Ivory-billed Woodpeckers are present in this location and have been for years. The weather was considerably more cooperative this trip than on the two or three preceding ones, although temperatures edged toward the uncomfortable – mid 80s and humid from Tuesday-Friday – and rain limited field time on Saturday and Sunday. I was alone from Tuesday-Thursday, and Frank Wiley joined me from Friday-Sunday. Later this week, I’ll post a day-to-day log that includes more about possible encounters and some additional images,

For reasons that should become clear, we are starting to think there may be a home range in an area of over four square miles (and possibly considerably more than that), much of which we have not yet explored, and some of which is very difficult to reach – a two mile walk from the nearest road and bisected by deep sloughs and streams. We have some reason to suspect that this range has been used for a number of years. This is in very mature bottomland forest, logged between 1905 and 1915, and it includes the stand of sweet gums where we found a cavity cluster last year.

Also on this trip, we did more experimenting with playbacks; I actually began the experiment shortly before I left for Louisiana, with a Pileated Woodpecker in my yard outside New York City. She responded with considerable agitation to my playback of Pileated calls and drums – calling and flying over at very close range while looking directly at me. She did not react at all to playback of ivorybill calls and pounding from the Singer Tract (the iBird Pro selections). Several species in our search area seem to react to ivorybill playbacks. Pileated, Red-bellied, and Red-headed Woodpeckers frequently react with drumming and scolding. In one instance, a calling Pileated Woodpecker went silent and flew away immediately after a playback. Barred Owls will often call immediately after, as will American Crows. In one case, a pair of crows came in to within 80 feet, apparently to investigate; in another, a Red-shouldered Hawk did the same.

There were three instances of possible ivorybill interaction with or response to playback. Two of them were very weak possibles, meriting only this passing mention. The third was a little more interesting and will be discussed in the day-by-day account. We will continue the experiment, both in Louisiana and New York (to see if and how various species react). We’ve recently been informed, by “Motiheal” from, that a Red-headed Woodpecker in Virginia approached in response to the playback of five kents.

One of the reasons we’re optimistic about having pinpointed a home range is the abundance of feeding sign in the area. In addition to the work sign from this area discussed in previous posts, there’s an abundance of older work, like this scaling on a hickory snag. OldHickory3 oldhickory2

Extensive scaling on an older hickory snag. Work goes from near the base to near the top. Atypically, some of the work at the base appears to be more recent than the rest.

Extensive scaling on an older hickory snag. Work goes from near the base to near the top. Atypically, some of the work at the base appears to be more recent than the rest. While the work is not fresh, and the bark has started to loosen, it is still within the decay stage described by Tanner as being preferred by ivorybills, as shown by the presence of twigs.

According to Tanner (p. 47), “Trees and limbs almost two years dead have lost almost all twigs, some small branches, and bark is loosened on some small branches.” Of course, the decay process is not as linear as Tanner’s description implies, and scaling of bark itself hastens the loosening of whatever remains. Thus, on scaled branches and boles, bark is likely to have loosened considerably unless the work is very fresh. Still, the presence of leaves and/or twigs is a strong indicator of recent death, perhaps even more so on blowdown, for which the decay process is likely hastened by proximity to the ground. In terms of more recent work, I found two sweet gums with sign on large high limbs, perhaps the most dramatic scaling that closely matches Tanner’s description we’ve found to date. Not only is it very extensive; the scaled limbs are quite recently dead. While it’s not possible to test the tightness of the bark, the presence of leaves in the case of the more recent scaling and twigs with buds in the case of the somewhat older work suggest that the limbs died within a six months to, at most, two years. SweetGum1 SweetGum2 It has been suggested that ivorybills are largely birds of the canopy that seldom if ever feed near the ground and that this behavior might account for the difficulty in obtaining clear photographs. Despite the fact that Allen and Kellogg observed a female bird feeding on the ground like a Flicker, and Tanner himself reported observations of foraging close to the ground, the idea that the species is limited to the canopy has become a kind of conventional wisdom. As I’ve discussed in previous posts, I don’t accept this notion and much of the feeding sign we’ve found has been low on standing trees and snags and on blowdown or slash. In the last trip report, I discussed feeding sign found on recently downed sweet gums (just outside of what we believe to be the hot zone, although possibly within it if it is larger than we currently suspect). On this trip, I found over two dozen examples of extensive bark scaling on downed sweet gum tops and limbs. IMG_0888 DSC00076 DSC00070 DSC00074 This work was so commonplace that photographing additional examples seemed redundant. In all cases, the blowdowns were recent and involved very freshly dead wood. At least some leaves were still attached, making it likely that these limbs and tops had fallen in the last six months to one year. In the hot zone, I found only two sweet gum tops or large limbs that had not been scaled. Most of the scaling was recent to very fresh, probably one or two days old in one instance (unfortunately, it had rained the night before, so any scat had been washed away.) I do not believe that all of this is the work of ivorybills. Nonetheless, I suspect that much of it is, due to its abundance and extensiveness and in light of Tanner’s study and the preference he found in the Singer Tract ivorybills for recently dead and dying sweet gums (this even though I believe Tanner overstated this preference and did not sufficiently account for specific conditions in the Singer Tract).

I did not find this type of work in brief visits to areas outside the hot zone, where it was ubiquitous; nor have I seen anything quite like it elsewhere. I did not see anything like it on other species of downed trees; the only partial exception was some scaling on longer dead parts of a live downed hickory. As an aside, it’s worth mentioning that the species of hickory in our area were not present in the Singer Tract, although their congeners, pecans and water hickory were. Unlike Tanner, we’re finding scaling on hickories that likely exceeds their relative abundance. We’re also finding considerably less scaling on various oak species.

In addition to the work on freshly downed sweet gums, I found two standing, recently dead young sweet gums that had been worked on in unusual ways. Both showed signs of infestation by insects that bored into the heartwood. Both had been very heavily scaled, one with minimal excavation only around the insect tunnels. The other had been hacked up in a way that, in the words of several people, looked as if someone had taken a hatchet to it; the wood was hard and not at all punky. Whatever did this work chopped through a small branch to the point where it broke off and almost severed the top of the tree as well.

SmashedSaplingLongShot SmashedSGSapling IMGP0564

Close-up of 2" diameter limb that apparently broke off while being fed upon

Close-up of 2″ diameter limb that apparently broke off while being fed upon

In his report on Cuban ivorybills, George Lamb described something similar:

Soon after we observed a female ivory-bill . . . feeding on the dead branch of a Hilacho tree (Torrubia obtusata) in a small stand of hardwoods. Suddenly the branch broke off while she was still perched on it . . . The Hilacho limb previously mentioned as breaking while being fed on, represents a type of feeding which was neither scaling nor digging. The limb was vertical and had probably originally been about three inches in diameter. Possibly it had once been scaled, but when recovered showed evidence of feeding to the extent that hardly anything was left. The wood was very punky and hand been chipped away from the perimeter to of the limb all along it’s 2 1/2 foot length. The chips, some of which we gathered, were long and splintery appearing, and were riddled with beetle larvae “tunnels”. 

Our broken branch is approximately 2″ in diameter, while the top appears to be more than 3″. Unlike the Hilacho tree, the wood on this sweet gum was hard, not punky.

While I suspect that some of the work on these trees, the very targeted work on the limbs (small rectangular scaling/digging), may have been done by Hairy Woodpeckers, the bulk of it is extremely unusual, inconsistent with any Pileated Woodpecker work I’ve ever seen and with Tanner’s description of that species’ foraging preference for longer dead wood; the type of prey is consistent with what would be expected for ivorybills. While the work on ‘hatcheted’ sapling doesn’t meet the diagnostic criteria we’ve developed over the years, we think it highly likely that this is Ivory-billed Woodpecker work. The scaling on the other small sapling is generally consistent with our criteria, although it has some very limited excavation, clearly aimed at expanding existing tunnels, rather than digging into the wood in the manner typical of Pileated Woodpeckers. Top of Snag LittleSapling ExitTunnelExpanded IMG_0876_2 Again, from the Lamb report: At one point she was only about 25 feet away while she was feeding around the base of a small pine. She began working “barking” this tree around 30 inches from the ground and slowly worked her way up to the top.

Stay tuned for the second installment, which will also include details of a sighting Frank Wiley had on Friday, April 3.

George and Nancy Lamb’s 1957 Cuban Ivorybill Study

I have been re-reading George Lamb’s 1957 report on the Cuban Ivory-billed Woodpecker. A number of items struck me as potentially significant for North American searchers, some for how they diverge from Tanner and others for their level of detail. Since this report is likely unfamiliar to many, I thought I’d do a quick post listing some of the more interesting observations

Lamb references a number of local sightings of “groups” of ivorybills, with one report to John Dennis that involved six birds. Notwithstanding, Lamb estimated the population density in Cuba to be much thinner than in the Singer Tract, at one pair per 12-25 square miles. He also pointed out that “ . . . the Cuban Ivory-bills are living for the most part in a cut-over pine forest where only small and deformed trees remain.”

The Cuban ivorybills fed on pines and hardwoods more or less equally, although most of the feeding sign was found on pines, due to the difficulty of searching for sign in the denser hardwood habitat. Roosts and nests were found exclusively in pines (one unused cavity was found in a hardwood), which is interesting in light of the fact that hardwoods were also available. Cavities were found at heights ranging from under 20 feet to nearly 60 feet. Cavities were higher in mature forest; Lamb suggested but did not conclude that the preference was for higher cavities and that the lower ones reflected an adaptation to cut-over conditions.

Lamb describes a female scaling bark: “At this point she was only about 25 feet away while she was feeding around the base of a small pine. She began ‘barking’ this tree about 30 inches from the ground and slowly worked up to the top.” Dennis too had observed birds scaling small pines. They found more scaling than excavation.

This apparent preference for pines, including small ones, may be significant, particularly since the hardwood areas were “relatively untouched”.

An estimated 17 birds were killed by humans over a ten year period, a huge number for such a small population. And it seems an open question whether the thinner population density noted by the Lambs was due to habitat quality, hunting pressure, or a combination of the two.

Regarding flight style: “. . .the flight of the Cuban Ivory-billed Woodpecker was always level and purposeful. They are strong fliers, capable of covering considerable distance in little time, as indeed they must to live successfully in cut-over woodlands. Although the Ivory-bill did not seem to undulate in its flight, the wing beats were not steady, having an almost imperceptible 2-3-2-3 rhythm.”

There’s no mention of double knocks, but calls are discussed. Lamb describes the sound as like the “note of a penny tin trumpet . . . short and usually repeated in a series of single-double-single beats, or it may begin with a double call: that is a high nasal “pent, pent-pent, pent”, or just “pent-pent”. On several occasions the female Ivory-bill most frequently observed made a few long and very loud calls, soon after leaving here roost tree in the early morning. The notes were of greater duration than normal and were repeated in a series of sixteen to twenty-two kients.”

Food for thought . . .

Additional Thoughts on Behavior and Rapid Evolution

I’d like to address an interesting post from “Sidewinder” on the Ivory-billed Woodpecker Researchers’ Forum on the rapid evolution question. His key points:

“Cyberthrush and others have suggested that natural selection has favored high levels of wariness and human avoidance in the IBWO. This position assumes that the change has a genetic rather than experiential (learned) basis. I have questioned this possibility based on the simple fact that behavior is usually one of the fastest traits to evolve. I have no problem with intense human predation on the IBWO resulting in increased vigilance among the surviving remnant, but if the IBWO persists, human predation has been absent now for dozens of generations. While many studies demonstrate that predation pressure can select for increased wariness in animals, what about the inverse? Can multiple generations of relaxed selection over a relatively short term (<100 years) result in relaxed vigilance?”

He concludes that the evidence is mixed and that the, “ . . . findings do not really support or refute Cyberthrush’s hypothesis. Clearly, we need more study–particularly of birds–to learn whether avoidance of humans might persist for many generations after selective pressure (predation) no longer exists to maintain vigilance. In the meanwhile, let’s acknowledge the highly tentative nature of this hypothesis.”

I hadn’t considered relaxed vigilance as a possibility, and it’s an interesting idea. With regard to the general evolved vigilance hypothesis, it’s certainly possible; I just don’t see it as being necessary to explain the difficulty of detection. I think normal wariness, difficult habitat, and extremely low densities suffice.

The hypothesis that the IBWO would dramatically change its foraging behavior, which is to a large degree morphologically determined, is considerably more extreme than the idea that the species became more wary. I have taken issue with the notion (or simplistic reading of Tanner), that the species is (or was) an extreme specialist, but its anatomy and historic range point to some degree of specialization – considerably more than exists in the PIWO.

I suspect that Tanner significantly overstated matters when writing about the canopy and high branch work, but even Tanner made it clear that IBWOs foraged at all levels. Some of the known prey species primarily feed and develop in the boles and in some cases quite near the ground (H. polita, for example). I suspect the high branch foraging Tanner observed was for larvae that he dismissed as being unsuitable because they feed on longer dead wood – Tenebrionidae in particular, although there’s no evidence from stomach contents to support this idea. The larvae we found under bark of this downed sweet gum have been id’d as belonging to that genus, and the tree was not very long dead.

One of the Singer Tract (in a pin oak stub, Mack’s Bayou) was in a clearing.

Mack's Bayou Ivorybill nest tree. Courtesy of the Division of Rare and Manuscript Collections, Cornell University Library

Mack’s Bayou nest tree. Courtesy of the Division of Rare and Manuscript Collections, Cornell University Library

John Dennis’s photos of Cuban IBWOs at a nest also appear to be from a very open area (and even if the Cuban IBWO is a different species, it’s a very close relative, and the hunting pressure there was almost certainly equal, if not more intense.) These seem odd nest locations for a bird that has rapidly evolved to hide in the canopy.

It’s also pretty clear to me that the John’s Bayou birds learned to tolerate human presence, while other IBWOs in the Singer Tract did not. As I’ve pointed out in several posts, Tanner and Kuhn (to a lesser extent) had a difficult time finding ivorybills in other parts of the Tract. This also suggests a behavioral rather than a genetic basis for the wariness or at the very least a substantial behavioral component.