whit! Sorting out the call notes of western Empidonax flycatchers

After a flurry of migrant Empidonax flycatchers this April in California, a number of us engaged in a discussion about their whit! call notes. Specifically, could we tell Willow from Dusky from Gray just by the whit? Personally, I’m not there yet. I usually only hear them a few times a year, which is insufficient experience, and I suspect there’s individual variation as well as the tricks that wind, humidity, and distance play on sound. But what about analyzing their sonograms with a good recording?

I looked at the very best on-line recordings from the Macaulay Library (collected via eBird, and there’s a collection of the best ones at Peterson Bird Sounds) and from xeno-canto. (Warning: the horizontal axis is different between their sonograms. Xeno-canto sonograms fit about two seconds in the space where Macaulay fits one second of recording, thus compressing the sonogram. Here, I stretched the xeno-canto sonograms horizontally to adjust for that.)

Willow whits

Dusky whits

Gray whits

It appears that Willow, with a good recording, is distinctive, with up to three harmonic tones and that downward slope after the upward slow. To the ear, the Willow whit is softer and sweeter than the others. Dusky and Gray, on the other hand, are both dry sharp whits, and virtually indistinguishable on the sonograms. Gray is more likely to go up to 10 kHz or higher, and that echo line on the first Gray sonogram was consistent on all the calls, at least for that individual recording. For the record, Least was similar to Dusky and Gray, although the triangular dark shadow to the right of the main call tended to be thicker and larger.

iPhone whit

iPhone recording

To the right is a typical iPhone recording, probably a Dusky. Based only on this sonogram, just a shadow of the good ones above, it could be any of the species. The take-home lesson is that one could probably diagnostically identify a Willow whit from a sonogram with a very good recording. The Dusky and Gray whits are too close to call.

Our discussion also focused on spring migration timing and incorporated Hammond’s (as many of these birds are silent). Again, I turned to eBird for some trends.


whit map

Spring migration for western Empids. These maps are based on 2017-18 data. I excluded a few outliers in each case. Nevertheless, these dates represent the earliest arrivals. Most birds were a week later than the dates shown.

The results are that Hammond’s is the earliest, followed by Gray and then Dusky. For all three, however, there was a pulse of records during the last week of April. Willow is a full month behind the other two. In fact, even in southern Arizona there were very few Willow records before May 10.

Kaufmann empidsMuch has been written about Empid identification. Here’s a link to the Rowland 2009 article in Birding. My personal favorite is the chapter in Kenn Kaufmann’s Advanced Birding. The diagram at left is from the old version; the new one is even better, showing variations within species.

For the Hammond’s/Dusky challenge in spring, I put together the diagram below. These identifications were confirmed by calling birds, which are easily separated (Hammond’s says peep rather than whit).


Finally, if you’re still confused, there’s this infamous meme:

empid meme

Two of the nation’s top birding spots threatened by the wall

Of the top 20 birding sites in the entire United States, based on the number of species reported on eBird, six of them are in south Texas. Two of them, Bentsen-Rio Grande Valley State Park and Santa Ana National Wildlife Refuge, are threatened by Trump’s proposed wall.


RG border map

The map above, taken from an excellent article illustrating all of the natural resources at risk from California to Texas, includes the bird totals for the eBird hotspots associated with the at-risk parks and wildlife refuges. The wall is often constructed hundreds of yards north of the actual border (the Rio Grande River). It typically includes a swath of cleared land on each side of it.  At Bentsen and Santa Ana, the wall threatens to destroy critical remaining habitat and strand the parks in “no-man’s land”, preventing public access. Dozens of articles have been written regarding the impacts to everything from butterflies to ocelots.

Sabal Palm is unique, in that the natural area is south of the wall. Visitors pass thru the wall in order to visit the park. However, there is no guarantee this arrangement will be made at other sites. Should public access be denied at Bentsen, the park could revert back to the Bentsen family per a historical agreement. The national wildlife refuges are especially at risk. As they are already federal properties, the Administration doesn’t have to deal with acquiring private property. Thus, they are the easiest places to build.



The decline of Mountain Plovers in two graphs

I live in Yolo County, California, near Sacramento, where Mountain Plovers used to be an annual winter specialty. Searching for “dirt clods with legs”, we used to be able to find dozens of these unique shorebirds, sometimes over a hundred.

Those days are over. They are now “irregular”, meaning we don’t find them every year. We’ve struck out five of the last eleven years. Before that, we averaged a high count of 72 individuals. The first graph, built from records in the Yolo Audubon Society newsletter, emails to the Central Valley Birds listserv, and eBird, shows the high count each winter in Yolo County.



This second graph backs out a bit in space and time, looking at all Mountain Plovers worldwide, starting in 1980. They are a bird of the steppe, breeding mostly on the Great Plains between the Rockies and the flat lands, between the Canadian and Mexican borders. They winter in open country in a vast arc west and south of there, mostly in California, Arizona, New Mexico, and Texas.

MOPL cbc

All the Christmas Bird Counts nationwide, plus Mexico, averaged 728 birds per count thru 1994, but have never hit that mark since. Fewer than 200 individuals have been enumerated six of the last ten years. Adjusted for party hours, the graph basically looks the same.