Tuesday, December 14, 2010

Bird species discovery odds and ends

Here are a couple more neat graphs about the rate of discovery of new bird species that I couldn't fit well into my previous two posts about the subject (10,000 birds? Part 1 and Part 2).

In Part 1 I briefly introduced the species accumulation curve as a means of comparison of rates of discovery between different groups. Here, I wanted to parse apart any broad-scale patterns in the bird data. I started by splitting the birds into their two broadest groupings - passerines (5927 species) and non-passerines (4100 species) - and then I split passerines again into two more broad classes - oscines (4668 species) and suboscines (1259 species). These new species accumulation curves are plotted against the overall bird curve (in black):

(click to enlarge)

The non-passerine curve falls generally to the left of the passerines and birds overall, indicating that non-passerines were generally described earlier than passerines and fewer non-passerines have remained undescribed until recently. The passerine, oscine, and suboscine curves are generally similar in slope and position, although suboscines lag in the last 100 years or so. The suboscine curve is also the only one with a noticeable uptick - a bunch of new species described - in the last few decades.

Those who know me when I get into projects like this know I like to take things to ridiculous excess. In this case, I decided to investigate family-level patterns of description. To get fairly smooth accumulation curves requires a fairly large number of taxa, so I picked the bird families recognized by Birdlife that contain more than 150 species. These twenty families are:

Family name Species
Tyrannidae 414
Psittacidae 374
Trochilidae 337
Timaliidae 326
Columbidae 318
Emberizidae 313
Sylviidae 293
Muscicapidae 286
Thraupidae 256
Furnariidae 241
Accipitridae 238
Picidae 218
Thamnophilidae 218
Strigidae 186
Phasianidae 181
Meliphagidae 177
Fringillidae 177
Turdidae 173
Anatidae 164
Rallidae 156

And their species accumulation curves, plotted against the all-bird curve in black:

(click to enlarge)

I don't expect anyone to actually tease apart that rainbow coalition of squiggles in detail, especially because they form such a tight column of similarly-sloped curves. That in itself is a pretty neat result - the family-level differences aren't as noticeable as I thought they would be, and teasing them apart in detail is fairly pointless. There are, however, two curves that stand apart from the rest.

On the far left in pale blue is Anatidae. 50% of the currently known ducks were described by 1800, decades before any other family, and 90% were described by 1870. Ducks are big, obvious birds with common interaction with humans (in the form of hunting) and often bold, distinct male plumages in each species. It is thus not surprising that ducks were described comparatively early in history and very few new species have been described in the last 100 years.

In contrast, on the right side of the curves, the pale reddish curve for Strigidae falls below the others in the last 100 years or so. The slope is pretty similar to the other curves for this time period, so I don't think the overall rate of discovery in owls is much lower. Instead, the curve is displaced lower by the flush of new species described in the last few decades - remember in my last post that owls had the highest number of new species described of any family since 1942.

That's all I have for now - have fun staring at those colorful squiggles.


  1. Nick - I've really enjoyed this series of posts, and love the "ridiculous excess" you go to. One trend that caught my eye in these similarly-sloped curves is the relatively sharp upward tick around 1815(-ish). Any speculation on what was going on culturally, scientifically, socially, or otherwise at that time? What allowed for that rapid increase in newly-described species?


  2. Mike,
    Compare it with the raw data graph from the first post (http://i199.photobucket.com/albums/aa220/slybirdsly/Papers/Fig1-BirdSpeciesByYear.png). The first fifty years or so are really ragged curves, the result of small annual numbers of new species punctuated by huge monographs. Then in the early 1800s it transitions to fewer huge monographs and a much higher annual number of new birds - this results in the increase in all of the curve slopes in this time. I don't really know anything about the social forces shaping this change, but I would like to find out.

  3. I'm currently working on a monograph that describes 400+ new species of wasps, just from a single genus, and just from Costa Rica.

    You bird people are so cute with your tiny little numbers of species.

  4. Thanks for sharing! Very informative!

    Kah Wai