Tuesday, November 20, 2007

The American Ornithologists-Union Check-List of North American-Birds

Going through a lot of bird lists lately leaves me pondering the eternal question:

Will the AOU ever get its hyphens right?

Sure, they have consistent rules for usage now, but that doesn't change the fact that the rules are wrong. By this I mean they are at odds with standard English style and usage in the rest of the world. A "bean-goose" is some strange creature that is a chimera of bean and goose. A "night-heron" is an entity that is a mix of night and heron, which is a mythological beast at best (seen Hagrid lately?). A heron that is active at night would be "night heron" or a "nightheron." A goose that likes beans is a "bean goose," or a "beangoose" if you must compound the noun. Last time I checked, there are no leaves growing out of the "foliage-gleaners." And there is very little rain or snow coming out of a "storm-petrel." The rule is to use a hyphen to make a new noun for a thing that combines characters of the two source nouns. You do not use the hyphen to link two nouns when the first noun is just something the second noun is associated with. A plover that likes sand is a "sand plover." A plover that is made of small hard granules is a "sand-plover."

These compound hyphenated groups names in the AOU checklist are correct; all the others are wrong:

Tiger-Heron (a stripy heron, not a heron that likes to sit on the backs of tigers)
Hawk-Eagle (an eagle-ish hawk, not an eagle that eats hawks)
Quail-Dove (by now the pattern should be clear...)
Nightingale-Thrush (though this one is iffy since the original nightingale is also a thrush...)
Shrike-Tanager (but not palm-tanager, since they rarely have fronds or produce coconuts)

This only applies to the "last names" of the birds, the actual noun part of the name. The modifiers before the nouns are generally hyphenated correctly in the AOU checklist: A "red-throated loon" is a loon with a red throat. A "red throated loon" is a red loon with a throat.

And, just to prove they can get it wrong both ways, they call it a "hawk owl" which to me is an owl that eats hawks; this bird is a "hawk-owl" since it is a diurnal, long-tailed, hawk-ish owl.

And the odds that this will be corrected before the end of this century are...?

Wednesday, November 07, 2007

Lines on maps

It seems to me that many birders, especially the more hard core listers, tend to view biogeography as a collection of hot spots with a list of goodies that can be gotten at each one; sort of an avian geocaching game. My own inclinations are more towards the ecosystem and the broad pattern; birds within the context of their ecosystem, ecosystems within the context of the continent and the globe. I am especially one of those people who is always looking for the pattern within the variation, squinting to see the signal within the noise.

As such, I have always been fascinated by biogeographic classification schemes, especially those maps that delimit areas of similarity and difference. Those wiggly lines and oddly shaped polygons can hold my attention at great length. There are many of these systems out there, of course. But I never found one that exactly suited my own curiosities. Some of them seem to be both oversplit and undersplit, segregating some areas based on just a few differences while lumping together others within a vast idealized uniformity that ignores an enormous amount of difference within the borders. Many of them draw lines along physiographic/geological boundaries that to my eyes make little difference to the biota. Over the decades, I have developed my own way of thinking about the large-scale patterns of biogeography. I've wanted to be consistent about what level of similarity and difference is used to draw the border lines, making distinctions that are clearly noticeable and significant to us old fashioned naturalists.

Amusingly, one of the things that gave me a push along these lines was the (unusual) experience of seeing a movie in which the landscape scenery and the background bird sounds actually matched each other, and were correct for the locale in which the film was set! This got me thinking about seeing if I could put together a collection of what are essentially "biological landscapes," within which the common species of plants and animals that you encounter on a day-to-day basis are similar and distinctive. I decided I'd focus most on trees and birds, as these are the groups I generally know best and look at the most. Of course, the "tree" thing didn't work out so well in semiarid and desert areas, so I adjusted that.

I borrowed an idea from Mineralogy, another science that has a long history of defining categories within what are really multidimensional spectra of infinite gradation. They use a quite Aristotilian notion of the "central concept." They define idealized central concepts for the composition and structure of minerals, and then classify real minerals by comparing them to the central concepts and seeing which one they are closest to. The idea here is that the central concept marks the theoretical center of the region in multidimensional space occupied by all minerals that will be given that name. So, I decided to construct "central concepts" for biological regions.

The core of my scheme is defining areas that share half or more of the common species across the area. So, my "central concept" lists of common species for adjacent areas share about half of their species. I also wanted to avoid defining areas that were just intermediate between other regions, with no unique characeristics; I also didn't want to define a new region that was little more than an impoverished subset of another region. So in addition to the "50% difference" guidelines I also added "Rule #3" requiring each region to have some distinctive species that are more abundant in that region than in any of the adjoining regions. These species don't have to be more common there than anywhere else on earth, just more common than in the other regions that border it. Nor do they have to be restricted to the region, just significantly more common there.

The gist of this is that if you are in a spot within a give region, and you pull out the "central concept" list of common species for that region, that will cover at least half of the common species you actually find at your locale. And there should be some distinctive species that you can find that are a sort of a signature for that particular region.

So what does "common" mean? For trees I just used standard terms thrown about in botanical writings, like "frequent" and "abundant." For birds I used detection rates on things like the BBS, CBCs, eBird, and my own field notes. I defined "common" to mean 0.1bird/hr, and abundant to mean 1 bird/hr. So "common" birds are the ones that you'd be likely to find in a weekend of birding through a variety of local habitats. "Abundant" species are the ones you would expect to find most of the time, even on fairly short outings of only an hour or so. Of course these numbers are all from different methodologies, and all observers have their own idiosyncratic biases. Because of this, and the fact that bird populations are alway shifting, any "central concept" list I come up with for a region includes a fair bit of subjectivity, and the choices as to where I draw lines are very much a function of personal preference, bias, and ignorance. I've been playing with this on and off for a decade or two; everytime I sit down and look at it again I make more changes.

Another note: Even though I focused on bird and trees, the regions seem to be appropriate for wildflowers, herps, mammals, butterflies, etc. as well. However, they are NOT relevent to fishes. Fish have entirely different rules governing their distributions; the fish map (I have tried to put one together) looks completely different. In particular you see a lot of very small regions in the southeast encompasing single river basins, and some absolutely huge regions covering large fractions of the continent in the north.

After all those preliminaries, here is my reinvented wheel, a very rough, approximate sketch of how I have come to think of the large-scale biogeography of North America north of Mexico excluding Greenland, also known as the ABA listing area (click for a larger but no more detailed view):

Please, please, please remember this is just a scribble, the lines are rough and only indicate the general idea, not precise boundaries. There are many little bits, wiggles, inholdings, and outliers that I haven't even remotely tried to represent above. The real point is the general arrangement and nature of the 33 "central concept" regions I have come up with. Actually, there are more than these, since I've decided that alpine and coastal areas need their own categories; that's a subject and a map for a later date.

I know it looks like splatter paint, and not necessarily different from a hundred other maps you have already seen. I just find this scheme works for me; some others might find it interesting too. The jumble of blobs is actually more orderly than it first seems. I think of them arranged in series from west to east and/or north to south, like so:

I. The Arctic:

Treeless tundra of the north.

1. High Arctic. The far north; dry, cold, little vascular plant life. This is probably a circumpolar region.

2. Bering Arctic. Affinities to the old world; I've wondered if the extreme eastern tips of Siberia would map into this region too.

3. Central Arctic. Vast areas of tundra, the core of the New World arctic biota.

4. Atlantic Arctic. A region I have some trouble with. Bird-wise it's mostly just a depauperate version of the Central Arctic, which breaks my rule #3 above. Botanically it is more distinctive. I expect southern Greenland would map here; northern Iceland might too. The alpine summits of New England seem to fit here, too, at least botanically.

II. Pacific coast lowlands:

Coastal lowlands, hills, and lower mountains from western Alaska to Baja California.

5. Aleutians. Treeless, but more of a maritime subarctic heathland than an arctic tundra. Strong old-world influences westwards.

6. Northern Rainforest. Actually, I haven't settled on a good name for this region. It is the cold, wet, spruce-fir-hemlock-cedar forest lands of the rugged coastal slopes from southern Alaska to central British Columbia. I think of it as "Southeast Alaska" but that understates its extent.

7. Pacific Northwest Lowland. The southward continuation; more wet conifer forests, some differences in species.

8. Northern California. Land of redwoods, doug firs, mixed evergreen forest, wet coastal prairies and brushlands. South at least to Monterey, including the western (foggy) parts of the SF Bay area.

9. Interior California. The Central Valley and the grass, chaparral and oak covered foothills.

10. Southern California. Also valleys, grasslands, chapparal, and oaks that are quite similar to the interior, but distinct enough in species composition to warrant splitting. I think. The Yellow-billed Magpies just won't let me lump the two regions, and you don't want to piss off a magpie. Continues down into northwestern Baja California.

III. Pacific Crest:

The Cascades, Sierra Nevada, and Peninsular Ranges from BC to Baja. Split in half, and extended inland (as is traditional) to parts of the northern Rockies. Generally wetter and more diverse forests than the Cordilleran regions to the east.

11. Pacific Northwest Highland. Quite a variety of intermingled forests lumped here, from wet mixed hemlock-fir-etc. forests of the west slopes to much drier Ponderosa Pine lands inland. Still, overall distinct from the coastal lowland rain forests and the cold, continental Rockies.

12. Sierra Nevada. This is also not the best name, since this region extends well beyond the Sierra Nevada proper. Includes the southern Cascades and the high parts of the Transverse and Peninsular ranges, plus the inland higher coast ranges and Klamath/Siskyou area. Wonderfully diverse conifer forests, many distinctive species. I tried splitting out the southern California mountains with their Coulter Pines and Bigcone Douglas Firs, but there's just not quite enough difference, especially considering the high diversity that prevails throughout the region.

IV. Arid Lands:

Deserts of the intermountain rain shadows.

13. Northern Intermountain. Sage and grasslands from south-central BC to Wyoming and the northern Great Basin. Also some of the larger mountain valleys of the Rockies, like North and South Parks in Colorado. Division between this and the next is mushy.

14. Southern Intermountain. Sage, playa lakes, and grasslands as above, division is a problem. Central BC and the Painted Desert are very different, but the line between could be drawn many places.

15. Mojave-Sonoran Desert. Think Creosote bush. In spite of the wide range from Joshua trees to Saguaros to Death Valley, there's enough widespread plants and animals across the area that I've kept it together from southern Nevada to eastern California and most of western Arizona. Continues far down into Mexico in Baja, Sonora, and northern Sinaloa.

16. Chihuahan Desert. Somewhere east of Tuscon the desert becomes grassier, shrubbier, and just plain different. A large area with quite a lot of variation. Continues well down into Mexico on the Altiplanice.

V. Cordillera:

From the Alaska Range to the Sierra Madre; even farther in fact.

17. Subarctic Cordillera. Yukon and Interior Alaska. A western variant of the Taiga, with classic spruce-tamarack muskeg mixed with rugged mountains and some species more characteristic of the Rockies. This is the region that often surprises visitors to interior Alaska when they see "eastern" Myrtle Warblers and Slate-colored Juncos swarming over the US's westernmost state.

18. Northern Rockies. Spruce, fir, lodgepole, aspen, meadow. The classic landscape of the high mountain west. I drew it as a northern blob, but the high mountain forests as far south as New Mexico and Arizona belong here, too. To define it by absence, there are virtually no Ponderosa Pines here.

19. Southern Rockies. Ponderosa country. Also includes many of the mountain ranges within the Intermountain region, in a fine-grained mosaic.

20. Northern Sierra Madre. Fabled among ABA listers. The US version of this is a bit depauperate, lacking many plants and animals that are common just south of the border. There is also a bit of influence here from the dry pacific forests of northwest Mexico; many of the rarities and strays in southeast Arizona are more characteristic of those forests than the Sierra Madre itself.

VI. Central Plains:

From the Tundra to the Rio Grande.

21. Taiga. Transitional between the tundra and the boreal forests, this area also has some distinctive species that prefer it over areas either farther north or south. I've not split it east versus west; it might bear a closer look.

22. Central Boreal. The great north woods, western version. Drier and less diverse than the eastern north woods.

23. Aspen Parklands. Another area that seems transitional but finds a variety of species that prefer it over its neighbors. Essentially a western extension of the Northeastern forest.

24. Northern Plains. As with the intermountain, the split between northern and southern plains is not clearcut. Also mingles with the intermountain in Wyoming and Montana.

25. Southern Plains. Continues the vast expanses southward.

26. Tamaulipas Brushlands. Another mecca for ABA listers. From the north it looks like a northern outpost of the tropics; but from the other side of the Rio Grande it looks like a southern outpost of the great plains. More than just an intergrade, though, it is a distinct region with many characteristic and unique species.

VII. Eastern Forests:

One of the largest and most diverse areas of temperate forests on earth, vastly underappreciated by the 200 million or so people who live here.

27. Eastern Boreal. The moister, richer end of the Boreal continuum. A warbler factory.

28. Northeast. Another blend zone between the boreal and temperate zones with many species that prefer the middle over either end. Also a major warbler factory. Includes the spruce-fir summits of the southern Appalachians; a case could be made for including the high balds and cove forests as far south as Georgia.

29. Midwest. A hard region to define, biologically as well as culturally. It is flatter and drier than the Appalachians, wetter than the plains, colder and drier than the southeast. It does have enough distinctive species to satisfy rule 3 and not be carved up as just a blend zone.

30. Appalachian. Hills and forests, extremely high diversity of trees, one of the most diverse temperate forests on earth.

31. Southeast. Hills and plains, much geological diversity hidden under the rain and trees. Often described as having a "subtropical" climate, but winter temperatures are much to low for this to be an accurate description. "Warm Temperate" fits better.

32. Subtropical. The area where you actually don't expect hard freezes every winter. Think live oaks, pines, and palmettos.

33. Bahamas. The Florida Keys are more tropical in the plants than in the birds, but I've still put them with Bahama. Though only a few of the distinctive west indian birds occur regularly here, you can find most of the trees.

On beyond these there are several alpine regions and a bunch of coastal ones that I'll talk about later. In the coming weeks I'll run through this list and post more details about each region.

Site Meter