Part 8: Twitching and taxonomy

Conservation

I suggested at the beginning of this section that a bird’s status, species or not, is all to do with a human need to classify things, but there is a frighteningly important side to it all: conservation. Whether a bird is one species or another can influence everything from the building of an airport to the fate of a forest. Since vocalisations can form a diagnostic feature for a species, one may argue that undervaluing bioacoustics can dramatically effect biodiversity. It seems as if we are all waiting for a visionary taxonomist with a holistic view who can pull together all aspects of a bird’s biology, and hammer in place a concept that will guide conservationists and birders through the storms and tempests of climate change. 

Consider the fate of these European Storm Petrels Hydrobates pelagicus that breed in a sea cave on a small island, bathed in the night lights of the busy Spanish resort of Benidorm (CD2-47). The petrels belong to the Mediterranean subspecies H p melitensis, a warm water population thought to breed in fairly low numbers throughout the Mediterranean, differing diagnostically from North Atlantic nominate birds in measurements. The island is visited by many tourists daily, and is also home to a colony of Yellow-legged Gulls. Consequently, the petrels are vulnerable not only to climate change but to disturbance and predation, and this is a story which is repeated at other colonies, many of which are close to popular tourist areas and the gulls they attract. 

There are differences between the calls of melitensis and its North Atlantic nominate pelagicus counterpart, which breeds most numerously in Britain and Ireland (CD2-48), and that is an important reason why these two taxa are now often treated as separate species (cf Robb et al 2008). To make these recordings, Magnus had already visited several North Atlantic pelagicus colonies and was familiar with their sounds before seabird expert Eduardo Mínguez took him to the Islote de Benidorm islet. There, one particular call of melitensis immediately struck him as different. Eduardo mimics this agonistic call, and the petrels respond well to his imitations. The call is the same in both sexes (James 1984), and it is probably equivalent to the ‘flight calls’ of Grant’s Storm Petrel (cf CD2- 24), but in the two, smaller, more vulnerable European Storm Petrels they are never given in flight. Whether the calls recorded on Islote de Benidorm are representative for the whole of the Mediterranean remains to be seen but, judging from recordings (Moreno 2000), the European Storm Petrels breeding off Lanzarote, Canary Islands, may also belong to melitensis.

Pelagic islands are good places to search for endemics, and the way birds learn their sounds is part of the reason why. Song learning is a major contributing factor in the genesis of new species. The peculiar songs of Madeiran Kinglet R madeirensis (CD2-49) are thought to have evolved away from an ancestor it has in common with Firecrest (CD2-50). It has often been suggested that island endemics are relicts of birds remaining stationary in appearance and sound while their relatives on the continents evolved into ‘other’ species, under pressure of more competition or divides caused by ice ages, for instance. It is no coincidence that song learners make up about half of all bird species. The ability to learn songs vastly increases the possibilities for variation, and is thought to have led to an accelleration in the rate at which new species are evolving.

It also introduces additional complications when it comes to looking for diagnostic differences because taxonomists need to be sure that they are comparing birds at a similar stage in the learning process, preferably adult birds with fully crystallised songs.

Kroodsma, Vielliard & Stiles (1996) have pointed out that “Many recent studies have revealed that speciation may not be accompanied by much differentiation in plumage…  often, the first clue to the true status of these species was provided by a difference in vocalisations, which in turn prompted the discovery of morphological, behavioral, or biochemical differences…”.

Recently, much has been written about the use of DNA in working out what is and is not a species, and there has been a move away from using plumage and biometrics as the main indicators. Very often though, it is sounds that give the first hint that a part of a population of birds we always classified under one label has actually evolved into something else. The relative ease with which sound recordings can be obtained, makes sounds the best early warning signals about undervalued or undiscovered biodiversity: birds whose classification needs to be re-evaluated.

Birders can play a part in ‘uncovering new biodiversity’, not least because their increased mobility is now taking them abroad more often. The observation that ‘this species sounds totally different back home’ might just be the first of many steps towards realising that it is not the same species, even if its plumage appears totally indistinguishable. The separation of Nearctic from Palearctic taxa has also been and probably will continue to be the most productive areas of opportunity for budding taxonomists. So, it is hardly surprising that another side effect of allowing Old World participants to take part in the World Series of Birding in New Jersey, was the realisation that in the spring the North American Common Moorhen Gallinula chloropus cachinnans sounded nothing like its European counterpart G c chloropus. In New Jersey, Common Moorhens are migratory opportunists, waiting for beavers to create new ponds which they quickly populate. An uncommon species, its habitat preference seems to encourage a need to attract a mate or defend territory in a different way. In crepuscular, nocturnal and marsh dwelling birds like petrels, owls and rails, including moorhen, sound is of more importance for pair formation than appearance. Consequently, differences in sounds can indicate that speciation has taken place sooner than any other feature.

Whoever named the North American subspecies of moorhen cachinnans must have suspected that it was something quite different. As explained with the gulls (Caspian Gull), cachinnans means ‘laughing’, and this American has a laugh, louder and somehow more brash, never heard from moorhens in the Western Palearctic. Listen to these birds recorded in Florida (CD2-51) and compare them with Common Moorhens recorded in Morocco (CD2-52). The ‘Laughing Moorhen’ sounds far more rail-like. Its appearance is different too, especially the shape and size of its frontal shield, and it also has a longer bill and a more reddish brown back and rump. But these visible differences pale into insignificance when compared to the huge difference in sound.

Sometimes things are less easy to determine. In recent years, a number of ‘Maghreb’ chaffinches, either Atlas F c africana or Tunisian F c spodiogenys, have been found in western Europe with European Common Chaffinches returning from North Africa in the spring. Both ‘Maghreb’ chaffinch males differ from European chaffinches in, for instance, having a green mantle, ear coverts the same bluish colour as the crown, and paler pinkish underparts (van den Berg & The Sound Approach 2005). In CD2-53 and 54 you can hear examples of their songs. While both sound a little exotic compared to Common Chaffinch, we shouldn’t be comparing these two recordings, because the song in the Atlas recording is plastic and we will need more recordings to work out whether africana’s songs are different enough from European’s to suggest species status. However the chep calls you can hear in both recordings (right from the start) are quite unlike anything familiar to Arnoud and Magnus from European chaffinches. They also believe these sounds are diagnostic for separating them from European birds.

In this case sounds don’t make a case for Atlas and Tunisian Chaffinch to be separated from each other, despite minor differences in plumage mentioned in our paper, as they share the same chep call. Molecular data suggest however that africana and spodiogenys are less related to each other than africana is to European. So, if it were only according to DNA studies, Atlas Chaffinch would be split from European and then Tunisian would have to be split from Atlas, despite the similarity of their sounds. 

During the course of the past 15 years we have looked at a number of taxa in the same way, wondering if they differed enough vocally to be considered species, Yellow wagtails, for instance, or the two subspecies of Black-tailed Godwits. Thankfully neither these nor the Atlas or Tunisian Chaffinch appear to be greatly threatened at the present time.

Redefining relationships by sound

Do you sometimes think that we make things too complicated? In 1827 von Baldenstein came up with the idea of describing Melodious Warbler as a Hippolais. Then others took three birds from the Sylvia and two other genera and put them in von Baldenstein’s genus. History suggests it was wrong but this is the problem if you use only skins and skeletons to reach these conclusions. It lasted over 180 years until DNA studies challenged it. Had sound been the defining criteria the Hippolais would have been part of the Acrocephalus warblers and would never have been considered a separate genus. And unlike DNA, unheard of only a few decades ago, songs and calls have always been available for study. All the Hippolais songs and calls fit very comfortably within Acrocephalus. Now as the various taxonomic bodies disassemble Hippolais some are creating a new genus ‘Iduna’. Let’s look at it from The Sound Approach point of view before another 180 years pass.

Two of the Hippolais that have created a lot of interest are Booted Warbler A caligatus and Sykes’s Warbler A rama (lets follow our beliefs and the CSNA and call them ‘Acrocephalus’). These two species’ songs, while fitting into Acrocephalus very comfortably, also differ from each other in striking ways. Adult Sykes’s (CD2-55) follows most Acrocephalus in singing a stream of song which contains the odd impersonation. The song jumps abruptly and rapidly between high and low sections, and doesn’t accelerate or rise in pitch. With its steady unchanging and fast tempo it can sound like the ultimate Acrocephalus, European Reed Warbler, but turbocharged. Booted’s adult song rarely repeats adjacent notes. As in all Acrocephalus it alternates between low and high notes but it has a unique spiralling quality (CD2-56). Starting quiet and low, it rises, gains volume and accelerates a little, twisting and turning. The most obvious difference between the two is that Sykes’s regularly repeats notes, often three times or more repetitively, which eliminates any spiralling effects.

These recordings were made when Magnus visited Kazakhstan in 2000, where he joined Kazakh ornithologist Andrei Gavrilov and a group of other interested birders on a visit to a site where Booted Warbler and Sykes’s Warbler had been discovered breeding side by side. This confirmed what had already been suspected for a long time: that these are best treated as two separate species (Sangster et al 1998).

Booted and Sykes’s Warbler’s calls

Acrocephalus warblers in the WP use both tak or chak type sounds and rattling, chattering or churring type calls in their repertoire. As a recent paper in Birding World has shown (Lindholm & Aalto 2005), Booted and Sykes’s Warblers have short zak-type calls differing from each other, and this is the sound most likely to be heard from an autumn vagrant in western Europe. The calls of Booted are of a longer duration, and sound like somebody striking a match (CD2-57). Sykes’s show shorter calls, while sounding more like a tongue-click (CD2-58). Both would fit snugly beside any Acrocephalus.

In fact all Hippolais possess both tak or chak, or a slight variation thak in Icterine, and rattling, chattering or churring type calls. Icterine also has a somewhat melodic call sounding vaguely reminiscent of the word hip-po-lais (rising with the last syllable), which is often incorporated into the song.

In a recent paper Svensson (2001) compares Western Olivaceous Warbler A opacus (CD2-59) to European Reed Warbler (CD1-09 and CD2-40) or Moustached Warbler A melanopogon. We agree and as a general rule an adult Acrocephalus warbler’s tone when singing is unmusical, harsh or buzzy. You couldn’t relax to their songs; they aren’t uplifting, more irritating if anything. When I first listened to an Eastern Oliveaceous Warbler A pallidus singing, I vividly remember being very disappointed when I heard it’s acro-like song, not at all sexy. Listen to an Eastern Olivaceous (continuing our theme we call it an Acrocephalus) of the Saharan subspecies A p reiseri, which Arnoud recorded in Tunisia. Typically for Eastern Olivaceous, it has a habit of repeating sections of song several seconds long before moving on, like a vinyl record with a scratch in it (CD2-60). This is a habit it shares with some other Hippolais, eg, Upcher’s Warbler H languida (CD2-61), but also with several Acrocephalus warblers, especially Blyth’s Reed Warbler (cf CD2-41). In fact, it seems difficult to find vocal traits not shared between Hippolais and Acrocephalus warblers, which makes it all the more surprising that sounds were not given more serious consideration in a recent review of the taxonomy of this group (Parkin et al 2004).

Here is another Hippolais song for you to listen to while making your mind up (CD2-62). 

One Acrocephalus warbler we have kept out of the discussion so far is Cape Verde Warbler A brevipennis, an endemic Cape Verde Islands species occurring in the south-westernmost Western Palearctic (CD2-63). It belongs to a group of African species called ‘swamp warblers’ that have different, bulbul-like songs. Indeed, on vocal characters, it seems less qualified to be a member of Acrocephalus than any of the ‘tree warblers’ (including Hippolais). Let’s consider the main acro-points. 

It can sing fairly continuously, but its streams of song are more broken up than in typical Acrocephalus songs, tending to consist of short, loud and varied trills, separated by gaps of roughly three seconds containing just a few quiet calls. The timbre of its song is not really harsh, nor is it pure and lyrical-sounding. Rather, it has a thrush-like, or better still bulbul-like quality, combining angular lines with a somewhat mellow timbre, sounding like the song of a larger bird. It is very much a daytime singer, although it has been heard to sing in the dark after rains (BWP). Cape Verde Warbler seems to be the only Western Palearctic Acrocephalus without mimicry of other species in its song, but it is possible that we have simply not detected it. Although it does repeat longer chunks of song, it does so with the repetitions separated by long gaps. It also has quite different calls and with its bulbul-like song, this species seems to fit rather uncomfortably in Acrocephalus, unlike the Hippolais warblers. 

We have been unable to find any vocal characters that separate the ‘hippos’ cleanly from the ‘acros’. Hippolais, on vocal characters at least, is a poorly defined genus. The Dutch committee for avian systematics (CSNA), whose taxonomy we follow in this book, was able to move only the four olivaceous and booted warblers, the so-called Iduna group, from Hippolais to Acrocephalus, because the DNA work their decision was based on did not look at all members of the traditional Hippolais group. Vocally, there seems to be nothing clearly separating the remaining Hippolais members from olivaceous and booted warblers (or Acrocephalus, for that matter).

Jeff Groth first published his work on ‘vocal types’ of North American crossbills in 1988. It was and still is a shock. He described two populations of crossbills that differed strongly in their flight calls, excitement calls and alarm calls. In plumage they were identical, showing only slight but nevertheless significant differences in their measurements, especially the dimensions of the bill. These could later be explained as adaptations to foraging principally on spruce Picea in one, and pine Pinus in the other (eg, Groth 1993). Because they were breeding in the same Appalachian forests but maintained their distinctness, Groth argued for them being separate, ‘cryptic’ species. Later, his research across the whole continent, published in 1993, resulted in as many as eight of these new crossbills, and subsequently Craig Benkman (eg, 2003) discovered a locally endemic ninth one in the South Hills of Idaho, USA. In his North American bird guide, Sibley (2000) lists and describes these nine types as “possibly nine separate species”. All are specialists at extracting seeds from a particular type of cone, and all differ slightly in measurements, though not in plumage. At one time it was suggested that vocal types involved dialects but, by definition, dialects are linked to a specific area; they might meet but they don’t mix. In contrast, these nomadic vocal types overlap in their breeding distributions, not incidentally here and there, but as a matter of course over large areas of North America’s coniferous forests.

When I was introduced to Magnus by Arnoud at the Dutch Birding annual meeting in February 1999, he had been up all night, making copies of a CD of crossbills recorded in Europe over the previous three years. Inspired by Groth’s work, which had been summarised in Dutch Birding (Sangster 1996), he had recorded and analysed flight calls, excitement calls and alarm calls in the Netherlands and beyond, and was able to show that at least six additional vocal types of ‘common crossbill’ could be found in Europe. The sleeve notes explained that crossbills paired according to type, and that differences in calls were not a matter of age or sex. Wow! What were these birds, and where did they come from? How did they relate to Two-barred L leucoptera, Scottish L scotica and Parrot Crossbills L pytyopsittacus? And when and where would it be published? At that time, I seemed to be constantly explaining to various enthusiasts that they couldn’t claim something was a separate species unless they had eliminated all other explanations. Yet Magnus was making no such claims; he just showed how it worked. I bought all the CDs, not wanting anyone to run away with his results before they had been published properly. The editors of Dutch Birding soon became aware of the work, and in spring 2000 the paper was published with a comprehensive audio CD comparing the six types of ‘common’ plus Two-barred, Scottish and Parrot Crossbills (Robb 2000). 

In 2000, when we started The Sound Approach, Arnoud concentrated on recording regular species in mainland Europe, and I worked with Dick and Killian in Finland and Norway. Magnus specialised in visiting areas where endemics and ‘new species’ could be found. His adventures could fill another book, but one of the easier trips involved a visit to Canada especially to record the sounds of White-winged Crossbills L l leucoptera. We knew that this smaller North American version of the Two-barred Crossbill L l bifasciata of Eurasia had a finer bill, and a more contrasting plumage that included more black, especially on the lores and surrounding the ear coverts (Cramp & Perrins 1994). Published recordings show that White-winged of both sexes, young and old, have a vocabulary of sounds very different from Two-barred. In many respects, White-winged seem closer to redpolls, the nearest relatives of crossbills (Arnaiz-Villena et al 2001). It is not difficult to distinguish between the various sounds of White-winged and Two-barred Crossbills. Listen for yourself (CD2-64 to 69). The wing-barred crossbills of North America and Eurasia seem to have diverged not only in song and calls, but also in ecology, measurements and plumage. Hispaniolan Crossbill L megaplaga of the Caribbean, another crossbill with wing bars, was recently split from leucoptera and bifasciata for similar reasons (Smith 1997, American Ornithologists’ Union 1998, Boon et al 2006) and for consistency, White-winged and Two-barred are probably also best treated as separate phylogenetic species. 

Although Dutch Birding had published the paper and CD, interest in the vocal types didn’t really take off. Birders didn’t realise how important it was to know the various sounds of the different vocal types if you were to accurately identify Two-barred Crossbill and Parrot Crossbill by call. In autumn 2002, migration watchers in the Netherlands started reporting a few Parrot Crossbills moving through the country. The apparent invasion gained support from observers at Falsterbo, the famous migration bottleneck at the southern tip of Sweden, who were reporting numbers of them passing through there too. At the same time, several ringers at different spots in the Netherlands, who were catching hundreds of birds, were unable to lend any credence to an invasion of Parrots: they didn’t catch a single one, and an invasion of this species never did materialise that year. When Magnus went out in the hope of recording them at various local hotspots, he found ‘common’-sized crossbills, giving a new flight call very similar to Parrot, but with excitement and alarm calls that were quite different. They became known as Parakeet Crossbill L c type X, so called because they sound like Parrot Crossbill but are smaller. They have been one of the commonest in the Netherlands since that time, and we have also recorded them in Poland and the UK. 

Parakeet Crossbill isn’t on the Dutch Birding CD, but in retrospect this is hardly surprising, because some of the six other crossbills which were around up to 1998 have been very scarce since. The only one which has remained equally widespread since 1998 is Glip Crossbill L c type C, named after the sound of its flight call. This highly mobile crossbill has been recorded in a variety of conifer species, from Finland to Spain, and in the UK from Shetland to Dorset, but seems to be particularly at home in Norway Spruce Picea abies forests in Scandinavia and the Alps. Biometrics gathered from hundreds of sound recorded crossbills at four ringing sites in the Netherlands were unanimous on the average bill depth of these two crossbills. Contrary to any suggestions evoked by Parakeet’s more powerful-sounding calls, Glip Crossbill actually has the deeper bill measurement (Edelaar et al 2004), albeit only by a few 10ths of a millimetre. 

When you hear two clearly different calls from a flock of crossbills, you are not necessarily hearing two types. There are two commonly heard calls that are useful in identifying Crossbills: those made in or just before flight, and excitement calls. 

Flight calls are the classic crossbill sound and are the first thing to learn. Some describe them as metallic, but no matter how much I bash two hammers together they don’t sound that way to me. In fact the attempt to describe them accurately nearly had me hitting Magnus’s head with the hammers. In true Mission Impossible style The Sound Approach called in Nick Hopper whose pragmatic approach to learning them was an inspiration. Between us all we found that the flight sounds varied between those that would be expected from a lost chick through to others that could be given by a passing Meadow Pipit A pratensis. 

The different excitement calls are more subtle and sound superficially like a blackbird alarming at a cat but at various distances. They are more powerful, nasal and insistent than flight calls and typically sound lower-pitched, prompting fantasies of a bigger crossbill. Don’t be fooled: all crossbills including Scottish and Parrot use them. In Two-barred the equivalent is the trumpet call. Essentially, there are calls for reinforcements, whether to mob an owl or just to liven up the party. They are highly infectious, and a translation would be something close to “here, here…”. A good way to elicit excitement calls is to ‘spish’ quietly. 

Not only is Glip Crossbill the commonest; it’s also the one that has been used in the field guides, where its flight calls are described as a metallic glip. Crossbill excitement calls create another slight conflict as their sonagrams show a few strange phenomena. There are bands of sound within all but Glip and Scarce Crossbill L c type F excitement calls that provide a nasal quality but are not true harmonics. Glip Crossbill has none of these bands, only true harmonics. So although it’s the most often described, its excitement calls are not typical. They sound rather like a falsetto version of the others. Compare the four sounds that you would be likely to hear from a mixed flock of Parakeet and Glip Crossbills (CD2-70 to 73). In my opinion Parakeet flight calls are like a lost chick while the corresponding sound from Glip is more like a sparrow. Parakeet’s excitement call is very like a Common Blackbird. If this doesn’t work for you, look carefully at the sonagrams and memorise them that way. 

Wandering Crossbill L c type A is one of the more irruptive and widespread crossbills. It has a distinctive flight call, a short incisive, sharp keep, a bit like the kip of Arctic Tern or the contact geg sound Redwing makes when calling from a bush. Its excitement call is a fraction longer than the equivalent in Parakeet Crossbill, but sounds the same. Bohemian Crossbill L c type B has a diagnostic flight call, sounding like a passing Meadow Pipit. Its excitement call is higher pitched than most and has a hard and nasal quality. Bohemian seems to occur mainly in Black Pine Pinus nigra forests from Bohemia and the eastern Alps through the Balkans and into Turkey. 

The next three crossbills can be identified by listening to both flight and excitement calls. This enables accurate identification every time. Phantom Crossbill L c type D, so called because it seems capable of disappearing for long periods then popping up unexpectedly, has so far only been recorded in the Benelux and southern England. It could well be missed purely because its flight calls are easily confused with Scarce Crossbill. Phantom’s flight calls are reminiscent of redpoll flight calls while its excitement calls remind me of a Black-throated Thrush T atrogularis I once heard alarming at a Northern Goshawk.

British Crossbill L c type E is apparently the commonest crossbill in northern England and Scotland, especially in new plantations of North American conifers (only one flock has ever been recorded elsewhere, in 1997/98 on the Dutch coast). Its flight calls sound similar to Parakeet Crossbill’s but its excitement calls are very different and remind me of a Western Jackdaw C monedula.

Flight calls of Scarce Crossbill are like Phantom Crossbill’s but heavier and remind me of Common Chaffinch rain calls in southern Britain. Its excitement calls sound like Glip Crossbill’s but more reminiscent of a Great Spotted Woodpecker kit note. No other crossbill combines these two calls. Scarce Crossbill has only been recorded in the Benelux, although it must also occur elsewhere as it was absent from the Benelux from 2001 until it turned up again in late 2005. 

While learning the sounds of Scottish Crossbill, Magnus had studied previously published sonagrams (eg, BWP) and in March 1999, when he visited Abernethy forest, he recorded large-billed crossbills with similar calls for his Dutch Birding paper (Robb 2000). Sadly he wasn’t in on one closely guarded secret; he didn’t know that there were also large numbers of Parrot Crossbills in the Scottish Highlands. This put him at a considerable disadvantage and the sample of recordings that he published as Scottish Crossbill probably included both large-billed crossbill species.

Ron Summers is a researcher employed by the RSPB to study rare Scottish breeding species. An expert in crossbills he along with Craig Benkman, Mark Herremans, Alan Knox and myself were all sent the paper before publication for peer reviewing. None of us commented on the mistake. 

Ron, David Jardine, Mick Marquiss and Rob Rae had been catching Parrot Crossbills between 1995 and 2000 and published their work on crossbill diversity in the Scottish Highlands in Ibis (Summers et al 2002). Through catching and ringing, Ron’s team had discovered that there were actually five crossbills occurring in Scotland including not only Scottish Crossbill but also considerable numbers of Parrot Crossbill. The secret was out, and Magnus’s mistake was pointed out in their paper.

In their work, birds were measured and sorted into Scottish, Parrot and ‘common’-sized, as well as being sound recorded. Summers and his co-workers developed their own system of classifying calls. Flight calls (four variants numbered 1 to 4) and excitement calls (five variants A to E) were considered separately, and these turned out to be used in consistent combinations: 1A was British, 2B was Wandering, 2D was Parrot, 3C was Scottish, and 4E was Glip. In addition to biometric and vocal data, Summers’ team collected DNA samples, and this led to a startling discovery (Piertney et al 2001). None of these birds, not even Parrot Crossbill, seemed to be genetically distinct, meaning one of three things. Either they had evolved very recently, their genetic differences too subtle to be detected by the method used, or they were frequently hybridising species, or the different crossbills were not species but some kind of morphs, able to interbreed freely. The five crossbills in the Scottish Highlands, occurring in the same mixture of native Scots Pine forests and non-native commercial plantations, all seem to be capable of retaining their distinctness; data collected in Scotland suggest that mixed pairs are extremely unusual (Summers pers comm). Such a situation is not unprecedented: it has proven equally difficult to find genetic support for the five redpolls Acanthis confidently identified by keen birders: Lesser cabaret, Mealy flammea, Greenland rostrata, Coues’s Arctic exilipes and Hornemann’s Redpolls hornemanni. Another parallel is the Galapagos finches discovered by Charles Darwin. When he visited the Galapagos Islands in 1835, these birds, with their amazing array of bill sizes for very specialised food sources, became one of the principle inspirations for The origin of species (1859). Galapagos finches are among the most closely studied birds in the world, and they also differ from each other in ecology, calls and songs. More than 100 years later, their evolution has actually been witnessed taking place (Weiner 1995), and yet their genes have proven to be an equally murky pool. 

Now the secret was out, Magnus returned to Scotland to record birds for The Sound Approach. Staying with Ron, he was taken to some of the best sites. Parrot Crossbills aren’t easy to record, and after several trips to Finland to capture them, Dick and I spent a desultory week in 2005 deep in the snow, trying to record them on the Finnish/Russian border with only one sighting. What we have now worked out is that flight calls of Parrots are quite variable. They differed for example between the 1983 and 1991 invasions in the Netherlands, although importantly, Parrot excitement calls were in all cases the same. 

Describing the differences between Scottish and Parrot Crossbill’s sounds (CD2-84 to 87) is like describing a malt whisky. The diagnostic Scottish excitement calls have a musical doubled note quality that, in true Scottish whisky salesman language, is peaty with more than a hint of a bee-eater flight call about it. Meanwhile, Parrot’s excitement calls remind me of the same salesman tapping out his receipt on an old fashioned typewriter.

So where does this leave Scottish and Parrot Crossbills? Really keen twitchers like Lee Evans struggle to know what can and cannot be ticked. Lee is a chronicler of twitching lore and only trusts his eye when identifying birds. This book however is primarily about sounds, not taxonomy, so thankfully we don’t need to decide here and now whether to give all those crossbills ‘species’ labels. Several teams are analysing them in various different ways that will eventually resolve their status; European work in progress is summarised in Edelaar et al (2003). Using a new technique, a team in North America has found small but significant genetic differences between the vocal types of crossbill there (Parchman et al 2006), and it seems likely that this can be done in Europe as well. The crucial point to understand in the meantime is this: that all these plain-winged crossbills are in the same boat. Scottish and Parrot Crossbills just happen to be the largest, which allowed them to be discovered early on, back in the days when only silent, dead birds in museums were studied in any detail. 

So for Lee and those wondering where this leaves the crossbills’ status as species? Galapagos finches are far too high profile to be left in this species’ no-man’s land and scientists are already debating their status (Zink 2002). In the same way crossbills’ status will be decided. Meanwhile sound as a primary means of determining species is starting to come back into its own and vocally distinct populations of crossbills exist. So, when you are birding in the UK, it is best to start sorting out British from Scottish Crossbills.