Monday, April 18, 2005

ENV: A note about Parulas

Posted to the TexBirds listserv

i wrote several folks offline about this situation, preferring to keep it there, but since it has boiled up on TexBirds i think it best to try some explanation here.

this is part of an ongoing research project so some of what's going on i’m keeping out of the public eye for proprietary reasons. however, i can basically explain what i’ve found and what isn't known. hopefully i will be publishing some of this in the next couple of years.

there are a number of records of Tropical Parulas from the Edwards Plateau dating back 30 years, as others have pointed out. in recent decades a number have been found at the Devil's River SNA as well.

as long ago as the 1970s, i began questioning reports of Northern Parulas on the Edwards Plateau based on heard-only records. the short explanation for this is that i thought i had detected a preference of nesting substrate of Tillandsia usneoides (Spanish Moss) for Northern Parulas and Tillandsia recurvata and Tillandsia baileyi (Ballmoss) for Tropical Parulas. Since Spanish Moss is scarce on the western plateau i surmised that perhaps Tropical Parulas were being passed over and identified as Northerns based on voice only.

i moved to the Edwards Plateau in the mid-1980s and began seeking some of the numerous reported breeding pops of Northern Parulas. i found them only in the eastern part near New Braunfels. i did find populations of Parulas to the west which i could not identify.

then, in 1999 i was scouting a portion of the Devil's River above the SNA and encountered singing parulas. not only did i find Tropicals, but also a number of intermediates. Erik Breden and i returned in 2000 and photographed and recorded several of these birds. we revisited this site several times over the next few years for the same purpose. based on my previous premise, i began examining records, and scouting locations, and looking at birds in more depth. i travelled the length of the state looking at populations. a couple of Erik’s photos of both Northern and Tropicals from these travels can be seen on his page at:

Otterside Photography

from this work i created an index of "Northerness" and "Tropicalness" and began identifying birds based on this scale.

now, let me digress briefly. these are "phenotypic" characters. that means they are artifical characters that humans have sorted out as defning our idea of what Northern and Tropical Parulas "are". These characters may or may not be genetically accurate for determining which is which. that’s to say, perhaps we have picked characters that don’t really fit what the birds themselves are genetically made up of. (for similar confusions you might try checking out the archives of TexBirds, AZ/NM Birds and ID Frontiers for discussions on Pacific-Slope Flycatchers and Cassin’s Vireos).

what that means in a practical birding sense is that: A) rogue eye crescents (to pick an oft-cited example) may be genetically linked to "Northernness" and thus be an absolute indicator that an otherwise good-looking Tropical has Northern genes and is the result of hybridization (or introgression of subspecific isolates) somewhere along the line; or B) it may simply be within the range of variability of pure Tropicals, and thus a bird which otherwise looks Tropical probably is; or C) there are not two species involved but one, and they gradually change from one geographic area to another (clinal variation), and we're looking at quite a few Texas birds as being in this changing regime. the same thing could be said of any other of the characters that we ascribe to being either "Northern" or "Tropical."

the index i constructed used (originally) six characters that could be judged on a phenotypic basis as being either "Northern" or "Tropical". i eventually eliminated a character because it was almost impossible to determine on specimens. so, using five characters i "scored” hundreds of photos, field observations, and specimens of Parulas from the southeast US around to northern South America.

i'll show you some of the results in a minute.

this will not be factually determined until DNA work is complete. i am cooperating with LSU in doing that work, which has not yet gotten off the ground, but i hope does soon.

here's some interesting twists for those who are unaware:

Dr. Ralph Moldenhauer, formerly of Sam Houston State U. had a special interest in Parulas and published that Texas "Tropical" Parulas had songs more similar to "Western" Northern Parulas than "Western" Northern Parulas did to "Eastern" Northern Parulas. [this portion edited after posting to reflect a misconception].

those who have birded southern Mexico can attest that the voice of Tropical Parulas from Southern Mexico have a warbled, melodic song, significantly unlike our "Tropical" Parulas. i have been collecting songs of Tropical Parulas that are showing up as vagrants in Arizona and California, and at least one of those has this melodic-type song.

the key, from a birding standpoint is this. if you want to assure yourself of having seen a Tropical Parula (at least based on the current definition) you should take pains to observe one well, assuring yourself that all the fieldmarks are present is the key. even in summer, just hearing a Parula is no guarantee that it's a Tropical. and at this time of year, Northerns are passing through.

now go to: Parula Project Maps

these maps are the inital construct from my project. i have additional data that has not yet been incorporated, but no data that changes any of the information here.

basically what you’re seeing is that each circle represents a mean, an average, of all the specimens scored from a particular location. a solid white circle would represent a “pure” Tropical Parula in a phenotypic sense. a solid black circle would represent a “pure” Northern. the “pie slices” are indicative of the percentage of the phenotype of each species in that population (note that this is not a genetic percentage – which a) we can’t determine without DNA studies, and for which we have no understanding of whether or not one character may have a stronger phenotypic expression than another – it’s just a measure of the outward characters we see in a population). you can see that far afield of the expected populations many birds may be showing characters that we have defined from the other species. so that in the Tennessee you can find birds with “Tropical” characters and in Venezuela you can find birds with “Northern” characters.

triangles represent either populations of which i know but haven’t yet visited; or populations that can be heard from a distance but cannot be approached close enough to determine. in some cases these have been determined by good observers to be either Tropical or Northern in character, but have not been scored for mapping purposes. FYI, the Travis and Bastrop County county birds are Northern types (Brush Freeman and photos by John Ingram), the Bandera and Edwards County birds are unknown (tg and Kelly Bryan), the Val Verde County birds are both (tg, Mark Lockwood, Kelly Bryan, John Karges), and all the remainder are Tropical types.

you will note that between the South Texas Plain and the East Texas Woodlands there is a basic changeover in “more” Tropical-looking birds and “more” Northern-looking birds; and a similar demarcation between western Edwards Plateau and East Texas Woodland birds.

the southern- and western-most “Northern” appearing birds occur in stands of Tillandsia usneoides (Spanish Moss) at Goliad State Park in Goliad County, and in Spanish Moss along the Guadalupe River at Hill Country State Natural Area on the Comal-Kendall County line. A population of Parulas in the westernmost known stand of Spanish Moss in western Bandera County has never been approachable close enough to observe. recordings are likewise indeterminate.

whether that invalidates the use of these characters for identification purposes or is indicative of hybridization/introgression is what we don’t know, and won’t know for some time.

some last notes, in response to previous posts:

it is clear now that the "normal" breeding ranges of Tropical and Northern Parulas are not thousands of miles apart. there is probably a continuous band of breeding Parulas along the Texas coastal plain, although there is certainly a sort of bottleneck there. Brush i think is right that the Sand Shield population is the largest. the Devil’s River population is rather dense but is limited to a narrow strip on the river, and for a short length of the river (at least as we know it now).

the ID Frontiers and Birdchat discussions are instructive. however, at that time there was little evidence of sympatric populations of birds that could be realistically described as “pure” Tropical and “pure” Northern. we have that now on the Devil’s River and associated drainages. the possibility/probability that the rogue eye-crescents are indicative of hybridization now legitimately presents itself. again, we won’t know for sure until DNA work is complete. and i might remind that such characters might be the result of multiple generations – thus the idea that a mixed nest is needed to prove such a situation is not true, although it would help.

i’m not sure the direction of an “ice age” comment but i suspect it may be in terms of incipient speciation in post-glaciation isolates. if so, it’s also possible that this speciation event is not complete, or that it is reversing due to the construction of hybrid habitats through the influence of man. that’s probably another long dissertation.

as Stevan Hawkins has pointed out, there’s probably a number of other populations that we’ll be finding/learning about as we gain more access to the coastal plain just south of San Antonio, the same as what we’ve learned in recent years about Green Jays and Audubon’s Orioles (see recent discussions on those). i suspect we’ll eventually fill in a lot of gaps for this bird. i don’t recall but possibly Martin Reid has found these at Halff Bros. R.?

and finally, once again i encourage people to make use of the archives when they are seeking info on TexBirds. it really is a goldmine of information. you can find it here:

TexBirds Archives

a large number of people over the years have contributed information or helped in the field on this, including: Erik Breden, Tim Brush, Kelly Bryan, Mark Lockwood, Chuck Sexton, Greg Lasley, Brush Freeman, Chris Sharpe, Barbara Bickel, John Maresh, Craig Farquhar, Dick Payne, Donna Dittmann, Steve Cardiff, Robb Brumfield, John Karges, John Arvin, Eric Carpenter, Stevan Hawkins, are who i remember right off, and i’m sure there are others, but i’m writing this from a location away from my notes.

tony g
hunt, tx, usa

Here are some previous requests i made to TexBirds and various other listservs and outlets.


I would like to strongly encourage taking time with identifying Parulas, especially outside of the east Texas/coastal Spanish Moss (Tillandsia usneoides) zone. Any bird that seems to be setting up territory in woodlands composed only of Ballmoss (Tillandsia recurvata with or without Tillandsia baileyi) should be suspect. I suspect that more than a few Parulas outside the valley/lower coast over the years may have been passed off as Northerns without critical examination. Please get a good look.

I would be interested in any and all Texas reports of these things: 1) Parulas of any stripe that seem to be setting up shop in stands of pure Ballmoss (T. recurvata); 2) Parulas of any stripe that are setting up territories west of the Balcones Fault, south of I-10, or west of the coastal Spanish Moss zone, and away from the lower valley; 3) any Parulas anywhere that seem to have anomalous plumage features (photos and/or recordings of odd birds would be most appreciated); or 4) Parulas that seem to be resident in areas devoid of any species of Tillandsia, including in areas where Usnea lichens seem to be the preferred nesting strata.

Any such information and photos would be useful in a large scale population project i am doing, and you are promised acknowledgment for any contributions.

Folks wishing to be engaged in data collection for the project may also contact me for more information.

I can be reached at

tony g
hunt, texas

I'm going to assume some familiarity with Spanish Moss (Tillandsia usneoides), the long pendulous gray "moss" that hangs from trees, usually in wetter areas and usually associated with oaks. This plant is an epiphyte, a member of the Bromeliad family (sometimes referred to as the Pineapple family), gaining all its nutrients from the air and rain. It is not a parasite, in that it sinks no roots into the tree to "steal" nutrients. It can damage trees however, by becoming so thick as to shade out sunlight, and sometimes becomes so thick that when wet it becomes heavy and breaks off twigs and small limbs. Some consider it a scourge and a spray has been developed to kill it. It is an ideal nesting material for Parulas since it is already pendulous and becomes thick in portions. Parulas burrow into thick mats of it and nest in the pseudo-cavity.

It is found throughout the south and in much of east Texas, though it can be local. South of the Houston area (and a line drawn west along I-10) the occurrence of this plant roughly shrinks to a narrow band along the coast, becoming more and more uncommon and local, until it shrinks from view in the King Ranch area. It appears again in the valley at least as far inland as Bentsen State Park. It is also found in Mexico south into South America.

Ballmoss (Tillandsia recurvata) is a cousin plant with a very different growth pattern. Instead of long pendulous strands (which are leaves) it forms compact tufts of short twisted leaves, rarely exceeding 2 inches in diameter for each plant. They are also epiphytic but not parasitic, though they do firmly attach themselves to bark. Usually associated with oaks as well. They can form dense colonies on limbs and do the shading and weight thing as well. The plant itself has always seemed marginal at best as nesting strata for Parulas, but in the last couple of years we have been finding more and more birds in stands of this. I'm trying to delineate what using this odd habitat fixture may be doing to influence populations of Parulas.

Ballmoss is apparently spreading north through Texas, but is most often associated with riparian woodlands in south Texas and with the Edwards Plateau. But it also occurs in the Trans-Pecos in spots, and overlaps Spanish Moss along the coast and up into the Blackland Prairie and Live Oak belts. It can be found west into Arizona, east into Louisiana, and is also in Florida; and occurs south through Mexico.

Bailey Ballmoss (Tillandsia baileyi) is a rare, rather restricted, species of larger Tillandsia that looks very much like T. recurvata but grows to five inches in diameter, when groups of the plant clump together. It gets larger south into Mexico. It is the only U.S. epiphytic bromeliad not found in Florida besides the common ballmoss and Spanish Moss. Currently in Texas it is known from only three counties -- Kenedy, Cameron and Willacy -- but also possibly occurs in Brooks, Kleberg, and Hidalgo. It's significance is that until the last three years, the only known nesting Tropical Parulas in Texas that were not in Spanish Moss, were in mixed T. baileyi/recurvata and thus we assumed there might be some correlation between the birds and Bailey Ballmoss. We now know that may not be the case.

There are, by the way, about six hundred species of Bromeliads in the genus Tillandsia, and Parulas down through South America may be utilizing other species in the group, though Spanish Moss seems uniquely suited for their purposes, and no other species of Tillandsia has that pendulous structure.

Usnea Lichens are not at all related to the above plants. In fact they are not technically plants at all but mixed growths of a fungus and an algae. There are many mixes of closely related lichens in this group, but the ones of interest are those that grow on smaller limbs of trees (often oaks) and become long and pendulous, similar in nature to Spanish Moss. In fact, that is how Spanish Moss gets its specific epithet usneoides (Usnea-like).

In the northeast, where Spanish Moss is lost (from not being as coldhardy), Usnea becomes the preferred nesting strata for Parulas.

In Texas pendulous Usnea is not very common or prevalent, but it does occur -- we have documented it in our research in at least Trinity, Washington, Newton and Jasper Counties. But in these places there is also abundant Spanish Moss. Our questions here are twofold: are there areas in Texas where Usnea occurs, but not Spanish Moss, and is it then occupied by Parulas; and in places where both "plants" occur, do Parulas sometimes utilize Usnea anyway. There is some evidence from Texas’ premier Parula biologist, Ralph Moldenhauer, that they do. We have no idea how extensive the Texas range of these pendulous lichens may be.

Besides its pendulous nature, Usnea is fairly easy to identify. It is composed of "slivers" of forking matter radiating from plaques that often seem to grow right out of the bark of a tree. They really don't look plant-like at all when closely examined. The flattened "leaves" have no veins, are often pockmarked with what look like moon craters, may be fringed, and terminate in an array of oddball shapes, including little club shapes. The Tillandsia species' leaves are all consistent in shape and roughly in length -- the leaves curl, and have a grayish cast to them. When wet they are green. The Usnea "leaves" almost always have a seafoam green color to them, although they can range from dark green to gray to brown.

Two quick additional notes on the Parula topic.

Chris Benesh wrote and questioned my description of California records of Tropical Parula. He is exactly right. It was a lapsus on my part, related to my receiving recordings of extralimital Northern Parulas from California. Sorry about that.

Note that some genetic work has already been completed and the authors (Lovette and Bermingham) suggest that Tropicala and Northern Parulas are the same species. The full article can be accessed below.

Some references:

Two Song Populations of the Northern Parula by Ralph Moldenhauer

Discrimination Between Regional Song Populations in the Northern Parula by Daniel Regelski and Ralph Moldenhauer

Mitochondrial perspective on the phylogenetic relationships of the Parula Wood-warblers by Irby J. Lovette and Eldredge Bermingham

Moldenhauer, R. R., and D. J. Regelski. 1996. Northern Parula (Parula americana). In The birds of North America, No. 215 (A. Poole and F. Gill, eds.). Philadelphia: The Academy of Natural Sciences; Washington, D. C.: The American Ornithologists' Union.

D. J. Regelski, D. J. and R. R. Moldenhauer. 1997. Tropical Parula (Parula pitiayumi). In The birds of North America, No. 293 (A. Poole and F. Gill, eds.). Philadelphia: The Academy of Natural Sciences; Washington, D. C.: The American Ornithologists' Union.


Post a Comment

<< Home