Monday, September 29, 2014

Migratory Speciation (?)

Since arriving in Laramie a few weeks ago, I've been spending a fair amount of time birding a few local spots, trying to soak in as much migration as possible before the gauntlet of grant proposals, research obligations, coursework, and -40F temperatures precludes me from getting out much.  And while I seriously enjoy all groups of birds, migration for me is really about one group: warblers.

I'm not alone in this regard.  It seems that virtually all birders I know, across all skill sets and interest levels, have an inordinate fondness for warblers.  I know several individuals who can tell me the single bird that got them hooked and the vast majority of the time that bird is a species of eastern wood warbler.  Indeed, a trip to Magee Marsh in the spring of 2009 was the single experience that pushed me from a casual bystander of ornithology to the point where I've seriously had to consider the non-zero probability that I will end up dying a hermit in an isolated shack (with good windows, of course) in the north woods because of birds.  For the record, it was a particularly striking male Cape May Warbler (Setophaga tigrina) on that trip which I suspect put me on this precarious path.

Warbler mania seems to be a sort of cliche in the birding world and, like most cliches, has its roots in some universal truth.  That truth being, of course, that eastern wood warblers are the single greatest group of birds ever (p<0.05).  I mean, look (all warbler pictures courtesy of fellow warbler fanatic, Mr. Skye Haas):

Mourning Warbler (Geothlypis philadelphia)

Cape May Warbler (Setophaga tigrina)

Black-and-white Warbler (Mniotilta varia)
American Redstart (Setophaga ruticilla)
Kirtland's Warbler (Setophaga kirtlandii)
Even in the very small sample of species here, all from Michigan, there is an astonishing amount of behavioral, morphological, and ecological diversity.  Mourning Warblers are relatively big, long-legged warblers that spend most of their time on the ground in dense vegetation.  Black-and-white Warblers have converged on a body plan and ecological niche with distantly-related nuthatches, spending most of their time climbing tree trunks in search of bark insects.  American Redstarts flash their bold, flaming tails at insects and tumble madly through vegetation in pursuit of their next meal.  The association between Kirtland's Warblers (also known as American Warblemasters) and Jack Pine (Pinus banksiana) approaches a degree of specialization that exceeds that in any host specialist insect that I can think of.

Growing up in the upper midwest, one is truly spoiled by the ease with which this diversity can be accessed.  In Michigan alone, ~34 species of warblers can be found during the breeding season, and this number increases slightly during spring and fall migration.  Unfortunately, the situation is not so idyllic out west, where warbler diversity is much lower.  Thus far, I've seen a mere 10 warbler species in and around Laramie.  To help put things into perspective, it is not terribly difficult to see more than 15 species of warblers in a single morning this time of year in Michigan.  While I do dearly miss these warblers, I'm increasingly appreciative of -- and sometimes even prefer -- being in areas with fewer species.  As great as it is to see 100+ species in a day, I find that it's very easy to overlook subtle details of single species -- details that, paradoxically, can lend insight into the dynamics governing the very diversity that so often obscures them.

Where the Laramie Valley is lacking in warbler diversity, it makes up for, almost entirely, in numbers of Wilson's Warblers (Cardellina pusilla) throughout much of August and SeptemberWhether I'm walking to campus, from campus, drinking coffee at a cafe, reading in my office, birding anywhere, struggling to wake up in the morning or going to sleep at night, my daily life is currently immersed in Wilson's Warblers.  Back home, I'm lucky if I see more than a handful of these throughout the fall, making this a great chance to look a bit closer at a species I don't know much about.   

One of the first things I noticed about Wilson's Warblers here is their voice -- specifically, the chip notes that warblers give during social contexts, flight, and even when foraging alone.  Despite my comparatively limited time with western Wilson's, it was immediately apparent to me that these birds sounded different than eastern individuals.  Have a listen!
Spectogram of typical Western Wilson's Warbler chip note

Spectogram of Eastern Wilson's Warbler chip note

The differences in amplitude (loudness, or in the case of spectograms, darkness) between the two recordings makes it a touch difficult to compare them visually, but the Eastern Wilson's call is noticeably shorter in duration and has a higher lower frequency bound than Western Wilson's.  This seems to be reflected in the Eastern call sounding sharper than the Western, though I think the amplitude differences could confound this, too.  At any rate, I feel pretty confident that there are real differences in the calls here, even if they are a touch subtle.  I'd like to examine this in more recordings, but there are surprisingly few recordings of eastern Wilson's Warblers calling (not singing).

Of course, this isn't too surprising.  There are loads of species that have distinct western and eastern populations that differ in coloration, morphology, ecology, vocalizations, etc.  However, experiencing these differences in Wilson's Warblers in the field prompted an evening of reading about this species and the discovery of some rather unique features about these populations.  But first, a bit more detail on the biogeography and natural history of Wilson's Warblers is in order.

Wilson's Warblers breed throughout most of Canada, in a large portion of the inter-mountain west, and along most of the pacific coast.  Based on very (and I do mean very) subtle differences in coloration between these populations, 3 subspecies have been described that correspond to these biogeographic and morphological differences.

Range map of Wilson's Warbler subspecies (from Irwin et al., 2011)

These populations are genetically distinguishable, but little was (and kind of still is) known about if and how breeding range and winter range differences were connected -- are different populations overwintering in different regions?  Darren Irwin's group from the University of British Columbia looked at this by sampling birds on the breeding and wintering grounds, trying to find the connections.  And while they didn't find any wintering eastern birds in the area they sampled (points on the map above), they did find something rather remarkable: eastern and western (lumping interior west and pacific populations) birds are as genetically divergent as many well-recognized species.  So divergent, that they are thought to have diverged some 2.3 million years ago, which is more than enough time for speciation in birds.  They didn't even find any birds that were genetically intermediate between the two populations, strongly suggesting that there is substantial reproductive isolation between these populations.

What is most impressive about this to me is that there is so little morphological differentiation between these two populations.  The biggest difference that I know of concerns plumage coloration, with pacific birds being a brighter, more saturated yellow than eastern birds.  Of course, there are also differences in song, which could help maintain reproductive isolation in an area where the two populations might meet and thus potentially hybridize (such an area of contact has not yet been found to my knowledge, though Irwin has looked).  But the fact that they are so similar in appearance after all this time is so fascinating.  Take a look for yourself!

From Sibley

Furthermore, the distinct breeding ranges suggest that differences in migratory behavior have played and are currently playing an important role in maintaining the distinctiveness of these two populations.  Additional study of winter ranges will be especially interesting to this end: are the two populations migrating to the same wintering areas, taking different routes in spring, or are they even overwintering in different areas?  What is the genetic basis of migratory differences -- do hybrid birds have an intermediate migratory route?  If so, are there any costs associated with this intermediate route, as has been implicated in Blackcaps?  If they are overwintering in the same region, why didn't Irwin's group find any eastern birds?  Are they using a different microhabitat? 

This is a system that I will be keeping a close eye on in the coming years, in hopes that Irwin manages to find an area of contact.  It will be particularly interesting to know to what extent and how these two populations are reproductively isolated.  Who knows -- we could have another species of warbler on our checklists in a few years!


Irwin, D.E., J.H. Irwin, and T.B. Smith. 2011. Genetic variation and seasonal migratory connectivity in Wilson’s Warblers (Wilsonia pusilla): species-level differences in nuclear DNA between western and eastern populations. Molecular Ecology 20: 3102-3115.

Tuesday, September 9, 2014

Making the Case for Crossbills

When I tell fellow birders that I study Red Crossbill (Loxia curvirostra) call types, the vast majority of responses range from "what's a call type?" to "wait -- people actually think call types are real?".  I'd increasingly like to devote more time to crossbills on this blog, so I think an overview of the natural history of crossbills is probably in order right about now.  In particular, I'd like to target people on each end of the response spectrum above and hopefully convince anyone unfortunate enough to be reading this that crossbill call types are real and worthy of your attention in the field and the attention they get from evolutionary biologists.  I won't be delving too much into how to identify each call type, as Matt Young from Cornell  has already put together a fantastic piece on this here that you absolutely should read.  Instead, I'd like to focus on the evidence that North American Red Crossbills are composed of distinct evolutionary lineages and some of the evolutionary mechanisms that may account for this pattern. 

Adult male type 9 Red Crossbill
Crossbills are cardueline finches, a group of predominately herbivorous finches that are mostly specialized for foraging on seeds.  There is considerable variation in the degree of specialization on particular types of seeds in this group, with crossbills representing the extreme end of high specialization.  The vast majority of crossbill diet is composed of nothing more than seeds from inside the cones of coniferous trees.  Crossbills are able to extract seeds from closed cones of any stage (fresh green cones or old weathered cones) due to their unique bill morphology.  Specifically, the vertical curvature of both the upper and lower mandible allow them to exert strong biting forces between the scales of cones to open them, which would not be easy if both mandibles were straight.  Additionally, the lower mandible is curved to either the right or the left of the upper mandible, which then allows the bird to laterally abduct the lower mandible to pry open the scales on a cone, giving it access to the seeds inside.  Once a seed is extracted, it is pushed into a groove on one side of the bill.  The lower mandible then puts pressure on the seed and removes the husk with the aid of the tongue.

From Newton (1973)
A South Hills Crossbill (type 9) foraging on Lodgepole Pine cones.  Note the cones are already opened a bit, but the foraging method is illustrative of that on closed cones.

Like all specialists, crossbills face unique challenges that revolve around their food source.  The most obvious challenge being the need to consume A LOT of seeds to survive, especially during colder times of year and when feeding young (unlike many seed-eating birds, crossbills feed their young primarily seeds, not insects, by grinding seeds up into a paste-like consistency).  For example, White-winged Crossbills (Loxia leucoptera), a closely related species, can eat up to 3,000 seeds in a day.  Assuming ~12 hours of foraging time in a day, that's 250 seeds an hour, a little over 4 seeds per minute.  Obviously, this number will be higher during winter when day lengths are shorter, further increasing the need for efficient foraging.  I don't have numbers readily accessible, but I would be willing to bet this number is higher for Red Crossbills, given that they are larger than White-wingeds and thus likely have higher energetic demands.

Red Crossbills have my mind captivated, but White-wingeds will forever hold my heart (Photo credit: Skye Haas)
Another non-independent problem associated with specialization on conifers is the year-to-year variability in cone production in most conifer species.  It is not uncommon for conifers across a large geographic area to produce almost no cones in any given year.  Obviously, this is a huge problem for crossbills and there appears to be two partial solutions to this problem.  One solution is to specialize on conifers that produce more regular cone crops, especially those conifers that also retain seeds throughout winter and early spring instead of dispersing them all (so-called 'key conifers').  However, even still, cone crops will inevitably fail at some point.  When faced with such a situation, crossbills will leave an area en masse and travel widely across the country in search of a sufficient food supply.  In the fall of 2012, the midwest saw a massive influx of type 3 crossbills from the pacific northwest, likely in response to such a cone crop failure.  Another possible stimulus for such irruptions is that crossbills could produce an excess of young in a given year due to favorable cone crops, resulting in strong competition for food and an effective food shortage per individual.  If sufficient food sources are found during these irruptions, crossbills will sometimes even stick around in an area and breed, no matter the time of year (this is responsible for odd breeding records from states such as Kansas).

eBird map of type 3 sightings from 2010-2011 (left) and 2012-2013 (right)
Of course, different conifer species -- even different key conifers -- produce cones that are very different in size, shape, and, importantly for crossbills, the thickness of the scales that protect the seeds.  Key conifer cones for Red Crossbills range in size from the diminutive cones of Western Hemlock (Tsuga heterophylla) to large, hard-scaled pines such as Ponderosa Pine (Pinus ponderosa).

The adaptive landscape for Red Crossbills

Taking into consideration the intense pressure on crossbills to forage efficiently, the potential benefits of specializing on a key conifer species, the diversity of key conifer cones, and the method that crossbills use to extract seeds from these cones, it is perhaps not surprising that crossbills have diversified into several forms (call types), many of which appear to be more or less specialized for foraging on a single or a few key conifers (Benkman, 1993; 2003; Irwin, 2010).  Much like the more famous case of Darwin's Finches to which they are often compared, the primary axis on which this divergence has occurred is bill morphology -- birds specializing on small cones have small bills whereas birds specializing on large cones have large bills.  While the association between bill depth (how 'tall' a bill is) and cone size is pretty striking, there's more evidence inside each bill that some degree of specialization is occurring.  Specifically, there is a near perfect relationship between the width of the husking grooves in a bill and the size of the seeds that many of the different call types specialize on (Benkman, 1993; 2003; Irwin, 2010).  It's worth noting that, based on the presence of stable cone crops and the retention of seeds throughout winter, Craig Benkman actually predicted the existence of a call type that should specialize on Sitka Spruce (Picea sitchensis) in the coastal pacific northwest, even down to the bill depth and width of the husking groove.  In 2010, Ken Irwin described such a bird and there is excellent agreement between Ken's measurements and Craig's predictions. 

7 of 10 call types in North America showing some of the diversity in body size, bill morphology, and flight calls 

In North America, Red Crossbills are currently divided into 10 populations or call types, a few of which have only been described very recently.  In most birds, such populations are categorized as subspecies, which can be thought of as geographical replacements of each other that may differ in any number of traits (e.g. vocalizations, habitat use, coloration, etc.) but are still considered the same species.  Although there is a rough pattern of geographical separation among crossbill call types, there are several places where more than one call type regularly occurs.  The general term used to describe such organisms that overlap extensively yet show differences in ecology and various traits is "ecotypes".  The call type system in crossbills is nothing but a spin on this -- calls are emphasized because different populations are most reliably distinguished in the field by the flight calls they give (see Matt Young's article).

And where more than one call type occurs, all available evidence suggests that hybridization is surprisingly rare (Groth, 1988; Smith and Benkman, 2007).  In other words, crossbill call types behave like different species.  The fact that crossbills have maintained significant genetic (Groth 1993; Parchman et al., 2006, unpublished data), morphological (Benkman, 1993), and vocal differences despite co-occurring together is made more impressive by their frequent nomadic movements across the country.  This should increase the opportunity for gene flow between populations, eroding any differences that had previously accumulated.  Furthermore, it is currently thought that most of the diversity we currently see in crossbills has arisen within the last ~12,000 years, coincident with the northward expansion of different conifers following the retreat of the last glaciers.  This represents one of our biggest gaps in understanding the story of crossbill evolution and is in need of further study.  However, if this scenario is accurate, the diversification of crossbills becomes even more remarkable, as 12,000 years is a mere blink of an eye in evolutionary terms.

This is not a crossbill.  I don't have any applicable pictures to place here, so enjoy this Ring-necked Snake (Diadophis punctatus) that I found in the Upper Peninsula.
The big question, then, is how are these differences maintained -- what barriers to hybridization exist in crossbills?  So far, only one study (Smith and Benkman, 2007) has looked at this in any detail and only in a single group of call types, but it seems likely that the results are somewhat generalizable to other crossbill pairs.  I'm also going to go out on a limb and suggest some other barriers to hybridization based on my own experiences with crossbills, fully acknowledging that they are in need of further study.

1) Habitat isolation: Given that different call types forage most efficiently on different conifers, we might expect to find them in different habitats, which should reduce the probability that they encounter (and thus breed with) each other.  Many conifer species are found in large homogeneous tracts of forest, making it less likely that different call types will encounter each other.  That being said, crossbills regularly fly long distances, so this alone will not be enough to prevent hybridization.  Additionally, there are quite a few conifer species that seemingly all crossbills can utilize well, so the importance of habitat isolation can potentially vary.

2) Immigrant inviability: Say a type 3 crossbill (a western hemlock bird) were to wander into a patch of lodgepole pine where a bunch of type 5 were hanging out.  Given the very small bill of this type 3, it is going to have a very hard time foraging on the big cones of lodgepole pine.  As such, if it were to stay there, the odds that it would survive are not good, given the insane amount of seeds that a crossbill needs to consume.  Even if it were to survive, it may not be able to acquire enough resources to get itself in breeding condition or feed young ('immigrant infecundity').  There is somewhat suggestive evidence that something like this may be happening with type 5 crossbills in the South Hills of Idaho (Smith and Benkman, 2007).

Fast forward to 15:04 for an illustration of how immigrant inviability could work.  Otherwise, do check this video out if you have time!

3) Assortative flocking: This is one of the coolest parts of the crossbill story.  Using playback experiments to types 3 and 4 in an area where they were co-occurring, Smith et al. (2012) found that wild crossbills flying over showed a strong tendency to approach speakers playing their own call type, but not of the different, co-occurring call type.  This is likely to be very important for crossbills, given that they spend much of their lives in flocks.  Furthermore, they seem to choose their mates from within flocks -- if crossbills are, on average, staying with their own call type, this will automatically reduce the probability of hybridization.  Based on my own experiences in the South Hills (with types 2, 5, and 9) and here in Laramie (with types 2 and 5), this tendency for assortative flocking seems to be present in other crossbill groups.  Others have noticed this behavior, too!

4) Conspecifics are sexy: Snowberg and Benkman (2007) gave captive female crossbills a choice of associating with either a male of her own call type or of a different call type (type 2 and 9 were used in this study).  Males were matched for several traits such that the biggest differences between them were their calls.  Despite this, females showed a clear preference for males of their own call type.  Many people I've talked to find this and the assortative flocking results problematic, given that crossbills learn their flight calls from their parents (Groth, 1993) and there has been a documented case of a hybrid pair where the female later changed her call type to match her mate's (Keenan and Benkman, 2008).  As such, if a breeding pair of crossbills was composed of two different call types, the hybrid offspring would inevitably learn one call or the other and then be able to hybridize with members of that call type.  And if crossbills are changing their calls to match their mates, how do we know that a bird giving a type 2 call is actually a 'genetically' type 2 bird?

The first thing to say here is that this call type switching is exceedingly rare.  In the 2008 study, only 1 out of 79 birds (1.27%) changed their call types.  Since 2008, thousands of additional birds from this same area have been banded and recorded and we have not yet found another bird that has changed its call type.  Furthermore, crossbills appear to learn flight calls very early on, after which there is virtually no modification of their distinctive call type (Sewall, 2009; 2011).  So, given that crossbills overwhelmingly prefer their own call type as mates (Snowberg and Benkman, 2007; Groth 1988; Smith and Benkman, 2007) and that we can expect that at least ~99% of these pairs to actually be the 'right' call type, the opportunity for mismatches between flight calls and genetic identity is exceedingly minimal.  There also seems to be a weird notion that learned traits of organisms don't constitute a viable means of inheritance, which is a misguided view (Pfennig and Servedio, 2013).  As a side note, a huge number of birds that we generally consider different species learn their songs just like crossbills learn calls, yet we're perfectly comfortable using songs to delineate species, even when there are well-documented cases of birds giving the 'wrong' song (see link below).

A neat story on a White-throated Sparrow that learned the song of a Black-throated Green Warbler

5) are good foragers: Snowberg and Benkman (2009) also found that females prefer males who are efficient foragers.  By preferring efficient foragers, female mate choice will reinforce natural selection for foraging efficiency.  This will also accelerate divergence between call types and further prevent hybridization.  For example, if a male type 3 joined a flock of large-billed type 2 (or vice versa) foraging on Ponderosa Pine, not only would he struggle to access food, but he would also struggle to attract a mate!

Wow.  You're still reading?  Well, here's a dog (Henslow) in a silly hat.

6) Singing different songs: Surprisingly, differences in the songs that males use to attract females and deter other males have been very under-studied in crossbills.  I've heard rumors that a paper will be published soon on differences in songs among call types, but until then (or at least until I can figure out how to embed my recordings), you'll just have to take my word: there are drastic differences in song!  The differences are so that I'm cautiously suspicious that we will begin using songs more to identify call types in the field.  You'll also have to take my word that these differences in the songs that males produce seem to matter to females.  I've done a few informal playback experiments in the South Hills and have seen striking differences in how females react to songs of their own call type versus those of other call types.  There's also the intriguing possibility that bill size could influence song production, which Julie Smith and I are currently investigating.  Depending on what we find, this could be another example where mate choice reinforces divergent natural selection.

The yooper type 10: one of many crossbill mysteries (Photo credit: Skye Haas)
Really, there's quite a lot more that could be added to this, but I think this post is already much too long.  For example, what happens to hybrid crossbills in the rare event that they're formed?  Do they have bill morphologies that are intermediate between their parents', leaving them poorly adapted to either conifer that their parents specialize on?  A lot of research on crossbills still needs to be done, which is very exciting to me as a birder, crossbill fanatic, and evolutionary biologist.

One point that I'd like to stress is that crossbills are a perfect example of the potential of citizen scientists to make serious contributions to our understanding of the natural world.  Perhaps one of the biggest unresolved areas in crossbill research concerns the distribution and irruptive tendencies of different call types.  While we have a decent idea of where different call types regularly breed, it seems to me that there are several important details that need to be worked out.  Exactly where are different call types co-occuring in the northwest?  Do different suites of call types co-occur in different forest types -- perhaps forests with conifers that all call types can easily exploit?  Are these fine-scale distributions stable over time?  During irruptions, what call types are moving where?  Are they sticking around to breed?

Answers to these questions will greatly enhance our understanding of the ecological and evolutionary dynamics that are writing the crossbill story, even as we speak.  The single best way you can contribute to this is to record crossbills that you encounter in the field so that they can be identified to call type.  Don't have a recorder?  The vast majority of recording/voice memo devices on cell phones are completely capable of capturing sufficiently high quality recordings to identify crossbills (I made the xeno canto recording in this post with my iPhone).  Once you have a recording, you have two options.  You can identify the birds to call type on your own by making a spectogram of the flight calls and visualizing them, which is, in many cases, the only way to confirm call type identification.  For this, you can download a free spectogram-producing software here and then compare the images to the ones in Matt Young's article (see link at the beginning of this post).  Alternatively, you can contact Matt Young (his email is in that article) or myself at empidonaxdvg "at" gmail "dot" com for help.  Finally, once you have your crossbills identified, please submit your observations to eBird, an online database that has bird sightings from all around the world, all contributed by volunteers.


Benkman, C. W. 1993. Adaptation to single resources and the evolution of crossbill (Loxia) diversity. Ecological Monographs 63:305-325.
Benkman, C. W. 2003. Divergent selection drives the adaptive radiation of crossbills. Evolution 57:1176-1181.

Groth, J.G. 1988. Resolution of cryptic species in Appalachian Red Crossbills. Condor 90:745-760.

Groth, J.G. 1993. Evolutionary differentiation in morphology, vocalizations and allozymes among nomadic sibling species in the North American Red Crossbill (Loxia curvirostra) complex. University of California Publications in Zoology 127: 1-143.

Irwin, K. 2010. A new and cryptic call type of the Red Crossbill. Western Birds 41:10-25.

Keenan, P. C., and C. W. Benkman. 2008. Call imitation and call modification in Red Crossbills. Condor 110:93-101.

Parchman, T. L., C. W. Benkman, and S. C. Britch. 2006. Patterns of genetic variation in the adaptive radiation of New World crossbills (Aves: Loxia). Molecular Ecology 15:1873-1887.

Pfennig, D. W. and Servedio, M. R. 2013. The role of transgenerational epigenetic inheritance in diversification and speciation. Non-genetic Inheritance 1: 17-26.

Sewall, K.B. 2009. Limited adult vocal learning maintains call dialects but permits pair distinctive calls in red crossbills. Animal Behavior 77:1303-1311.

Sewall, K.B. 2011. Early learning of discrete call variants in red crossbills: implications for reliable signaling. Behavioral Ecology and Sociobiology 65:157-166.

Smith, J. W., and C. W. Benkman 2007. A coevolutionary arms race causes ecological speciation in crossbills. American Naturalist 169:455-465.

Smith, J. W., S. M. Sjoberg, M. C. Mueller, and C. W. Benkman. 2012. Assortative flocking in crossbills and implications for ecological speciation. Proceedings of the Royal Society of London Series B 279:4223-4229.

Snowberg, L. K., and C. W. Benkman. 2007. The role of marker traits in the assortative mating within red crossbills, Loxia curvirostra complex. Journal of Evolutionary Biology 20:1924-1932.

Snowberg, L. K., and C. W. Benkman 2009. Mate choice based on a key ecological performance trait. Journal of Evolutionary Biology 22:762-769.