British Birds: Insect, amphibian or bird?

Some years ago, I. J. Ferguson-Lees and EDHJ conceived the idea of publishing a short paper comparing the similar—and possibly confusing—sounds produced by Scops Owl Otus scops, Nightjar Caprimulgus europaeus and Savi's Warbler Locustella luscinioides on the one hand, and Midwife Toad Alytes obstetricans and two species of crickets on the other. In 1964, EDHJ drafted a short paper on the subject; in 1972, this was given to JFB, whose help was sought because of his entomological field experience and ready access to the wildlife sound collection in the BBC Natural History Sound Library and to commercially published recordings. The present paper is the outcome of our joint opinions.

The possibility of confusion between the remarkably similar sounds produced by Scops Owl and the Midwife Toad is well known. There are, however, a number of other European birds, amphibians and insects whose typical sounds could trap the unwary ornithologist into a mistaken aural identification; the danger is all the greater because, in most cases, their ranges and habitats overlap.

A feature of the sounds produced by many members of this assemblage of unrelated animals is that they are to a greater or lesser extent ventriloquial. They are also subject to modification by the acoustic properties of the environment. This means, for instance, that, even if one hears a ventriloquial sound with which one is familiar, it is often difficult to determine precisely its direction and distance. Thus, a bush-cricket stridulating on or near the ground may sound as if it is quite high up in a bush; it may, therefore, be mistaken for a bird. One of us (JFB), when he first encountered Wood-crickets Nemobius sylvestris in a pinewood in southern Europe, thought that the stridulations were emanating from high in the trees and that the insects were cicadas; only later, when he became familiar with these diminutive crickets, did he realise that they were actually present in the leaf-litter on the woodland floor. Similarly, when he first heard the Tree Cricket Oecanthus pellucens in the south of France at dusk, its penetrating trill was so loud that he thought it was coming from a frog or a toad; because of its ventriloquism, he found it extremely difficult to locate, especially in the gathering darkness; eventually, he tracked it down to some lank vegetation, where it seemed to come from near his feet, but he still could not find it; the following afternoon, in bright sunshine, he was able to pinpoint and discover the tiny singer in the field layer.

Such experiences illustrate the difficulty one may have in dense fenland of deciding, for instance, whether the continuous, reeling 'song' of an unseen singer is from a Locustella warbler or a bush-cricket. Heard close-up and in identical conditions, the song of a Savi's Warbler may sound quite distinct from the stridulation of Roesel's Bush-cricket Metrioptera roeselii, especially when both species are present for comparison; but, when heard distantly among a chorus of other wildlife sounds, it may not be so simple to distinguish it from the insect. Moreover, a chorus of bush-crickets, cicadas or amphibians may mask minor distinguishing features and make them even easier to confuse with the bird songs they resemble.

The fact that such unrelated animals have evolved very similar songs in a particular habitat may simply be an example of convergent evolution. Alternatively, it is quite possible that Locustella warblers have evolved acoustic mimicry of the more primitive singing insects because of the additional biological advantage it bestows: in which case, the statement by Armstrong (1973) that the 'Grasshopper Warbler L. naevia and Savi's Warbler are not mimetic' would be untenable. Whichever is the case, there is little doubt that the continuous reeling songs of the type produced by some Locustella warblers and bush-crickets (Tettigoniidae) in fenland confer certain selective advantages.

In the first place, as Armstrong (1973) has pointed out, the use by marsh birds of continuous song, which is more effective than discontinuous song as an advertisement, may be made possible by their comparative invulnerability in fen or swamp habitats: a ground predator finds it difficult to approach closely enough to seize a singing bird (or insect, for that matter) in a reed-bed without betraying its own presence. There is, therefore, little necessity for the singer to change its song-post as frequently as do many other species (presumably an additional advantage, since it makes it easier for a potential mate to locate a singing male). The same is true of birds that have protracted aerial song-flights, such as larks Alaudidae and pipits Anthus; they also have continuous songs. Discontinuous songs (bursts of song interrupted by frequent pauses) enable the singer, which in many cases may be partially deafened by the power of its own voice, to listen and watch for approaching predators.

Grasshopper Warbler, Brandon Marsh NR, Warwickshire (Photo: Steve Seal)

There is presumably an additional important protective advantage to a Locustella warbler in mimicking bush-crickets and cicadas: it is likely that at least some predators overlook the fewer and more widely spaced birds among the more abundant and densely concentrated insect singers. These insects are usually very numerous where they occur, thus fulfilling the requirement of effective mimicry that the model must be more common than the mimic. This may also explain why, for instance, some River Warblers L. fluviatilis sing like cicadas and others sing more like bush-crickets: which is mimicked depends upon the local abundance or otherwise of these insects. Indeed, the varied mimicry of insects by this warbler deserves special study. It could even be that an individual River Warbler may mimic a diurnal bush-cricket or cicada by day and a different nocturnal species at dusk or after dark. River Warblers sing most frequently at dawn, at dusk, and through most of the night (Bergmann & Help 1982). It would be interesting to play tape-recordings of tropical cicada species with highly distinctive songs to River Warblers which are newly arrived in their territories, to see if they eventually incorporated these exotic songs in their repertoires.

Locustella species, such as the Grasshopper Warbler, may of course give their characteristic insect-like songs in areas where insects with similar songs are scarce or absent, especially on the northern limits of their range. This should not, however, necessarily invalidate the hypothesis that they are mimicking insect songs, since butterflies which mimic distasteful or poisonous butterflies are also sometimes found in areas where their models are absent.

It should also be mentioned that bush-crickets start to sing later than Locustella warblers, as they do not normally mature until high summer (perhaps earlier in the south). There is, therefore, a period in the spring when these warblers are likely to be confused only with the earlier-maturing crickets and cicadas. In this connection, it is worth remembering that recent studies have shown that some songbirds (e.g. the Chaffinch Fringilla coelebs) learn elements of adult song from the adults that they hear singing around them when they are in the nestling and juvenile stages. Thus, it seems possible that young Locustella warblers may similarly learn and memorise insect stridulations in their first summer, when these are at their maximum.

If those Locustella warblers which sing like insects are indeed imitating them, and thereby achieving an additional biological advantage from it—a sort of acoustic camouflage which deceives predators—then this might begin to explain why some other warblers of this genus, such as Pallas's Grasshopper Warbler L. certhiola, do not possess insect-like songs. Presumably, they would not obtain any selective advantage by doing so; perhaps because, in their adaptations to their particular environment, they have evolved other (as yet unknown) methods of avoiding or reducing the chances of attack from predators. Frequent changes of song-post may be one such alternative device.

It is interesting to note that the three Locustella species which do not sing like insects are all confined to the Eastern Palearctic, where the Lanceolated Warbler L. lanceolata is the only species with an insect-like song.

Birds which can be confused with unrelated animals

Genus Locustella

As the scientific name suggests, this group of birds has the most insect-like songs. Although some are popularly known as 'grasshopper warblers', their songs most closely resemble the stridulations of bush-crickets (Tettigoniidae), formerly known as long-horned grasshoppers, rather than true grasshoppers (Acrididae).

Bush-crickets produce their 'songs' in quite a different way from that of true grasshoppers. The latter rub rows of tiny pegs on the femurs of their hind legs rhythmically against a thickened vein on the folded forewings; the varying arrangements of the pegs and differences in the speed and rhythm of the leg movements are responsible for the song differences between species. The bush-crickets, like the true crickets, stridulate by rubbing together their forewings, the basal areas of which are modified for sound production; although some produce series of rather irregular chirps, the majority sing continuous songs like the Locustella warblers, usually of a pulsating type.

Bush-crickets, like Locustella warblers, generally shun thick woodland, frequenting open habitats, but with dense ground vegetation and scrub; they have skulking habits and are extremely well camouflaged. It is not surprising, therefore, that the songs of these warblers and of the bush-crickets have converged. Some species of bush-cricket, incidentally, such as the Great Green Bush-cricket Tettigonia viridissima, often sing quite high up in scrub vegetation and even in trees.

Of those Locustella warblers which we have studied in the field or listened to on sound recordings, Gray's L. fasciolata, Middendorf's L. ochotensis and Pallas's Grasshopper Warblers sound quite bird-like and are therefore unlikely to be confused with any amphibian or insect species known to us. The others, treated below, are distinctly insect-like. It need hardly be added that there is also a risk of confusing one Locustella with another, particularly in the case of Grasshopper and Savi's Warblers, for which see Grant (1983).

Savi's Warbler Locustella luscinioides The vibrant and buzzing, reeling trill of this species often starts with some quiet ticking notes, which gradually gather speed until they merge into the characteristic song; this then rises quickly in volume and continues level in pitch and intensity for periods of up to 30 seconds or more; the average duration is about 20 seconds, but may be as short as three to five seconds with similar pauses. It is remarkably like the stridulation of Roesel's Bush-cricket, which typically inhabits lush vegetation in marshy localities, especially in Britain where it is decidedly coastal, although it may sometimes be found well inland in quite dry situations, such as chalk downland. In Britain, Roesel's Bush-cricket is confined largely to coastal and estuarine marshes from Hampshire in the south to South Yorkshire, where it is locally common, but it has recently been discovered on the estuary of the River Dovey in North Wales; it also occurs inland in the Cambridgeshire fens. Its range, therefore, includes those areas most likely to be colonised by Savi's Warbler; moreover, it overlaps in habitat and is found around reed-beds, although not usually within them. On the Continent, it is widespread, except in the south.

The continuous buzzing song of Roesel's Bush-cricket is quite penetrating and may last for up to a minute or more, especially in warm, sunny weather, when it may continue well into the night as well as during the day. Its similarity to the song of Savi's Warbler is striking; Roesel's may well be the species intended in the statement by Peterson et al. (1974) that the bird's song 'can be confused with noise made by Marsh Cricket'. Simms (1979) recounted how he and Roger Tory Peterson sat at the edge of a reed-bed in the Camargue in 1954 trying to decide 'whether a buzzing sound was that of a Savi's Warbler a hundred metres away or a bush-cricket only a few feet from us': illustrating well the problems of identification that can face even experienced observers in the field. Confusion, however, is possible in Britain only in high summer, as Roesel's Bush-crickets do not sing until they mature in July, though they mature up to a month earlier on the Continent. Because of the high frequency of its stridulation, Roesel's Bush-cricket (and Savi's Warbler!) is often inaudible to older persons.

There is an even closer similarity between the songs of Savi's Warbler and of the recently discovered Vineyard Mole-cricket Gryllotalpa vineae. Fortunately, the danger of confusing the two is probably not great, as this mole-cricket is reported to frequent much drier habitats (e.g. hill-pastures, cornfields and vineyards) than the Common Mole-cricket G. gryllotalpa (Bennet-Clark 1970 a & b). An unwary ornithologist, however, might be deluded into believing that he or she had discovered Savi's Warbler singing at night in an unusual habitat! Owing to its preference for waterside habitats, there is perhaps a greater risk of confusing the Common Mole-cricket's song with that of Savi's.

Finally, there is some resemblance between the song of Savi's Warbler and that of the Greater Sword Bush-cricket Homorocoryphus nitidulus, which is common in the Mediterranean region, but does not normally stridulate until after dark. This species often occurs in marshy, reedy places, but has a more pulsating song which, if anything, is more like that of River Warbler.

River Warbler L. fluviatilis The typical song, most often heard at dusk but often through the night and at dawn as well, is perhaps the most bush-cricket-like of all the Locustella warblers, owing to its pulsating rhythm or, as described by Peterson et al. (1974), rhythmic 'chuffing' quality, like a distant steam engine. It could be mistaken in summer, especially at a distance, for the songs of the Wart-biter Bush-cricket Decticus verrucivorus, the White-fronted Wart-biter Bush-cricket D. albifrons, or the Great Green Bush-cricket and its close relative, the Singing Bush-cricket Tettigonia cantons. There is also some resemblance to Greater Sword Bush-cricket, Roesel's Bush-cricket, Summer Cricket Tartarogryllus burdigalensis and Common Mole-cricket.

There seems, however, to be some variation between individuals: some have a harder, more metallic quality and therefore sound much more cicada-like. One River Warbler recording we have heard sounded very like a Common Cicada Lyristes plebejus; another very like the cicada Cicada orni and two others like a cross between the Great Green Bush-cricket and some other kind of cicada. As already suggested, a detailed study of this mimicry of different species of insects by individual River Warblers is needed.

As Thorpe (1957) and others have pointed out, the River Warbler's song is distinctly slower than that of either Savi's or Grasshopper Warbler: only 15 pulses per second, compared with 53 in Savi's and 31 in Grasshopper. Thorpe, timing seven songs, found an average duration of each song burst of 10 seconds, whereas our analysis of four songs on Palmér & Boswall's discs (1969–80) gave an average of 13 seconds; but three songs on Unger's tape (1971) varied from 1' 05" to 2' 49", average 1' 44".

Grasshopper Warbler L. naevia When heard at a distance, the high-pitched trill or reeling of this species is, in our opinion, most like the stridulation of the Wood-cricket and its near relation Heyden's Cricket Pteronemobius heydenii, but could also be confused with Roesel's Bush-cricket, Common Mole-cricket, or even more so with Vineyard Mole-cricket. A chorus of Summer Crickets might also suggest a Grasshopper Warbler to some hearers. As the Grasshopper Warbler may be found in drier habitats than Savi's, there is perhaps a greater risk of confusing its song with those of singing insects, which are found in greater diversity and abundance in such habitats than in marshland.

Although the stridulation of a single male Wood-cricket is quiet, a large colony can produce a considerable volume of sound and might well lead an unwary ornithologist to conclude that he is listening to a Grasshopper Warbler. Like the bird, these insects will sing continuously for hours, with only brief pauses. Grasshopper Warblers may sing for up to four hours with only brief intervals of one to two seconds between bursts, each burst lasting anything from 30 seconds to four minutes (North & Simms 1958); the average is, however, about 30 seconds (Thorpe 1957).

Lanceolated Warbler L. lanceolata The song of this species is described by Palmér & Boswall (1969´80) as 'very similar to the Grasshopper Warbler, but sharper and a fraction slower'. We concur on the whole with this statement, but consider that, like the River Warbler's, it is more pulsating and, therefore, more like that of a bush-cricket. In fact, a song recorded in Amurland, USSR, by Dr Irene A. Neufeldt (in Palmér & Boswall 1969–80) has rhythmic fluctuations which make it remarkably reminiscent of the much fainter stridulation of the Short-winged Conehead Conocephalus dorsalis. It could be confused with several other bush-crickets with which it overlaps in habitat, including Great Green Bush-cricket, Singing Bush-cricket, Greater Sword Bush-cricket and their allies, or with a chorus of Field-crickets Gryllus campestris, or even with a Common Mole-cricket.

Recordings of Lanceolated Warbler songs we have listened to vary in duration from just under one minute to 1½ minutes.

Other birds

Nightjar Caprimulgus europaeus The Nightjar's continuous, churring song can be quite easily confused with the stridulation of the Common Mole-cricket, but much less so with that of the Vineyard Mole-cricket. The danger in Britain is, however, of rather more significance to the entomologist than to the ornithologist, as the Common Mole-cricket is nowadays very rare and localised: formerly widespread, it is now confined to the southern counties of England, where it has been reported in recent years only from the Avon and Test valleys in Hampshire and from the southeast corner of Wiltshire. Few people in Britain, even entomologists, are familiar with the Common Mole-cricket's song, and it is therefore probably much overlooked. One of us (JFB) has looked for it in its Wiltshire locality, where Nightjars happen to be common, and, although familiar with the songs of both species, nevertheless had to check each Nightjar carefully to make sure that he was not overlooking the insect.

Nightjar, Lincolnshire (Photo: Graham Catley)

On the Continent, the Common Mole-cricket is widespread and locally common, with a range which extends, like that of the Nightjar, throughout Europe as far north as southern Scandinavia, and deep into western Asia. Although this mole-cricket prefers moister places than the Nightjar (such as water-meadows, wet heathland, and the margins of ponds, canals, rivers and streams), there is often an overlap of habitat. The Vineyard Mole-cricket inhabits much drier places than the common species and probably does not occur so far north (D. R. Ragge in litt.).

Both mole-crickets and the Nightjar sing at much the same time: at dusk and after dark during the spring and summer (mid April to August), although D. R. Ragge (in litt.) concluded that Common Mole-crickets stop singing before the height of summer, perhaps by the end of June; this has been confirmed by other observers. Gilbert White (1789) knew both species well at Selborne. Writing of the Common Mole-cricket, he observed that: 'In fine weather, about the middle of April, and just at the close of day, they begin to solace themselves with a low, dull, jarring note, continued for a long time without interruption, and not unlike the chattering of the fern-owl, or goat-sucker [Nightjar], but more inward.' This inward quality in the Common Mole-cricket's song is certainly a distinctive feature.

In both cases, the song consists of a rapidly repeated, single, harsh 'click' note, of level pitch, rising and falling in intensity. The Nightjar's song is of lower pitch and less rapidly uttered than that of the Common Mole-cricket. If the Nightjar's song is described as 'churring', then that of the Common Mole-cricket is 'chirring' (a Scots voice would best express the difference, by throwing in a few extra 'rrrrr's!). Fluctuation in the Nightjar's sound is produced by the bird turning its head from side to side as it sings: giving not only a ventriloquial effect, but a rise and fall of volume with no change in the character of the even, rattling sound. The Common Mole-cricket, on the other hand, will often stridulate for periods of several minutes with no perceptible variation in intensity; when it does vary its song, however, the rise and fall takes the form of a certain 'hesitancy', whereby the component 'clicks' are produced in groups of varying speed and of different duration. This could be compared to the sound produced by holding a metallic object against the spokes of a bicycle wheel which is then rotated at a different speed every few seconds.

The effects of the two sounds may be blurred by wind and distance; and, when an observer approaches the source of the call, the Nightjar may move away and the observer's footfalls will cause the mole-cricket to become silent. Only perseverance can settle the point, and, even then, as the authors know to their cost, one species may be masking the presence of the other.

Palmér & Boswall (1969–80) included in their series of discs a recording of a Common Mole-cricket along with the Nightjar for comparison, as did Roché (1961) and Unger (1971) in their recordings.

Apart from the Common Mole-cricket, a chorus of Wood-crickets may also sound rather like a distant Nightjar. Ragge (1965) remarked that 'When one hears this song in the Wood-cricket's natural haunts it is generally a chorus of several singing males and thus sounds continuous, suggesting perhaps a distant Nightjar or Grasshopper Warbler. Although the Wood-cricket is primarily a day-loving creature, the song continues into the night in warm weather.' He has since told us (in litt.) that he himself has sometimes confused a chorus of Wood-crickets with a distant Nightjar. Wood-crickets also favour the same habitats as the Nightjar, living among the leaf-litter in woodland rides and glades, and also in areas of bracken near wood borders. The two species inhabit the same localities in the New Forest, where the Wood-cricket is common. Elsewhere in Britain, the Wood-cricket is very local, being confined to scattered localities in the Isle of Wight, Dorset, south Devon, south Wiltshire and west Surrey. On the Continent, it is widespread and locally common, especially in the south.

There is one amphibian whose incredibly loud voice could be confused with the Nightjar: the Natterjack Toad Bufo calamita. A crepuscular or nocturnal chorus of these small toads, which have the loudest voice of all European toads, can be heard over a range of 2 km or more (Arnold & Burton 1978) and might well be mistaken for the Nightjar by an inexperienced birdwatcher. Hardy (1939), writing of them in west Lancashire, observed that: 'the rattling chorus of the toads gathered in the pools is almost deafening—like thousands of Nightjars at a distance'.

The Natterjack's vocal period, which may extend from early spring to August in Britain, and up to a month later in southern France, overlaps the song period (Ma—August) of the Nightjar. It frequents sand-dunes and moist sandy heaths, which are also often inhabited by Nightjars. Abroad, Natterjacks are found over much of western and north–central Europe, apart from Italy and most of Scandinavia. In Britain and Ireland, the Natterjack has disappeared from many of its inland haunts, but is still locally common in southwest Ireland, on the Cumbrian coast and the Solway, in west Lancashire and in East Anglia.

Red-necked Nightjar Caprimulgus ruficollis Although we cannot liken the song of this Iberian nightjar to that of any particular insect, it is nonetheless suggestive of a cicada, especially the female Red-neck's rasping call. We can only advise ornithologists to be cautious in assigning a distantly heard nocturnal sound to this bird species.

Eurasian Scops Owl, Turkey (Photo: Rob Smallwood)

Scops Owl Otus scops The resemblance of the typical call of this owl to that of the Midwife Toad is well known (e.g. König 1968; Thönen 1968). The extent to which even experienced ornithologists can be misled by the calls of amphibians or insects with which they are not familiar is illustrated by an error on Dr C. König and Dr G. Thielcke's record Vögelstimmen aus Sudeuropa, released in 1965 by Kosmos of Stuttgart: the recording of a Scops Owl is, in fact, of a Midwife Toad! In their series of discs, also available on cassettes, Palmér & Boswall (1969–80) have included a recording of the Midwife Toad for comparison.

The call of each species is, superficially at any rate, a single note of short duration, 'musical' in nature and variously described as resembling a whistle or a bell. Each individual repeats this note monotonously at regular intervals of a few seconds, but the effect of this regularity may be entirely lost when the listener is among a scattered group of owls or toads (or both) calling at different intervals. Especially in hilly country, great difficulty is sometimes experienced in exactly locating a calling individual. It is, however, often quite easy to call up the Scops Owl by imitating its call and then, by using a torch, to obtain a sighting.

Attentive listening and a close approach to a Scops Owl reveal that its call is, in fact, disyllabic: a short high note, immediately followed by a lower, longer note of similar intensity—this could be written 'KeOOOO'—the whole call lasting for approximately half a second. The Midwife Toad, on the other hand, has a shorter note of apparently absolute purity and of even pitch and intensity. Perhaps it can best be described as 'electronic', since it closely resembles the sound produced by an oscillator, with no component which could be represented by a consonant and no perceptible variation in pitch. It could be expressed simply as 'ooo!' and lasts about one-eighth of a second. Johnny Morris (verbally) has aptly likened the call of a single Midwife Toad to the Greenwich time signal.

The diagnostic characters of the two calls are apparent only at close range, for a variety of reasons. Distance, especially among trees, tends to absorb the second, quieter, part of the Scops Owl's call, with the result that it sounds shorter and monosyllabic; this is well illustrated if one listens to a bird close at hand being answered by another at a distance of 100 m or so. The call of the Midwife Toad, among rocks and against buildings, can be lengthened by resonance; an echo at close quarters can even appear to give it a second component. Where both animals exist together, great patience is necessary to assess the relative status of each.

In Europe, the Scops Owl is found from central France and northwest Germany southwards and eastwards throughout the Continent. The Midwife Toad occurs throughout the Iberian Peninsula and France, its range extending eastwards from Belgium through western Germany to the Hartz Mountains and the western and southern slopes of the Black Forest. The Midwife Toad inhabits hilly, stony country which often coincides with the Scops Owl's habitat of scattered trees, plantations, gardens and old buildings.

The clear, musical, but mournful call of the Fire-bellied Toad Bombina bombina might also mislead birdwatchers in east-central Europe and the southern USSR, but not as easily as the call of the Midwife Toad. The same is true of its close relative, the Yellow-bellied Toad B. variegata, in south-central Europe. Both these toads, however, usually sing in chorus.

In hill and mountain regions, the call of Scops Owl might also be mistaken for that of the Mountain Cricket Eurgryllodes pipiens. We must also draw attention to the similarity between the typical calls of Pygmy Owl Glaucidium passerinum and Scops Owl. The call of female Scops is especially like that of male Pygmy.

Pygmy Owl Glaucidium passerinum König (1968) and Thönen (1968) have pointed out independently that there is a risk of confusion not only between the calls of Scops Owl and Midwife Toad, but also with the very vocal Pygmy Owl, especially the unmated male. Having listened to sound recordings of the latter and compared them with recordings of the other two species, we can confirm that the danger is a real one, although the similarity between Pygmy Owl and Midwife Toad is less striking than that between the toad and Scops Owl. In our opinion, the call of a single Midwife Toad most closely resembles that of a Pygmy Owl when a sound recording of the latter is played at low level; perhaps the greatest danger is that a distantly calling Pygmy Owl might be passed over as a Midwife Toad!

The typical call or song of Pygmy Owl is also, in our opinion, more like the call of female Scops. The two overlap in parts of Europe (e.g., the southern Alps and parts of the Balkans), and Scops has been extending its range northwards in recent years, for example, to central Germany (Hoyer 1967).

König (1968) described the 'song' of the unmated Pygmy Owl as a series of monotonous drawn-out whistling sounds, like 'duh' or 'duhb' These are quite like the calls of Midwife Toads, bearing in mind that there is a good deal of individual variation in the case of the toads, but they lack their 'Greenwich time signal' electronic quality. The difficulty arises from the likelihood of not being able to compare the two animals in the field, but should be resolved by carefully locating the source of the sound. If the toads are responsible, there should be little trouble in tracking them down.

There is a possibility of mistaking the calls of Pygmy Owl for those of Fire-bellied Toad, and even more for Yellow-bellied Toad, where their ranges overlap. The owl's distribution overlaps with that of the Fire-bellied Toad in Poland, Russia, Czechoslovakia, Austria and Romania; and with the Yellow-bellied in eastern France, Switzerland, extreme northern Italy, Austria, Czechoslovakia, southern Germany and Poland, Romania and northwestern Yugoslavia.

Baillon's Crake Porzana pusilla Palmér & Boswall (1969–80), in their notes on this species for disc RFLP 5003, mentioned that its jarring, trilling calls may at a distance be confused with the Common Frog Rana temporaria. Having listened to recordings, there is, in our opinion, a similarity which might well mislead the unwary ornithologist during the spring. Feindt (1968) warned from his own field experience that confusion is also possible at a distance with the Edible Frog R. esculenta. The calls of individual European Tree Frogs Hyla arborea and of the Stripeless Tree Frog H. meridionalis are crake-like, and may lead birdwatchers to think they are listening to Baillon's or Little Crake P. parva. For an English translation of Feindt (1968), see Boswall (1969).

Little Crake Porzana parva The characteristic 'quek-quek-quek' call of the Little Crake might, in some circumstances, be mistaken for the 'krak-krakkrak' calls of an isolated European Tree Frog, or even the similar, but slower and deeper calls of the Stripeless Tree Frog (JFB has followed up intermittent calls of the last-named species in the Camargue during the daytime, uncertain whether it was a crake or not). Usually, the Little Crake accelerates its calls towards the end, whereas the reverse happens with the European Tree Frog. The Stripeless Tree Frog tends to maintain a slow, steady rhythm.

Corncrake Crex crex The Corncrake's grating double-note is easily recognised, but it is not unlike the sonorous croak of the male Parsley Frog Pelodytes punctatus. The rhythm of their calls is different, of course, but one might conceivably mistake a couple of unseen, but fairly close Parsley Frogs answering each other for a distant Corncrake. Both species call by day or night and their breeding ranges overlap, notably in France.

Coot Fulica atra The weak, rather plaintive piping croak of the Common Toad Bufo bufo during the spring resembles to a certain extent the quieter 'piping' calls of the Coot, so frequently heard at this time of the year; but this is more likely to mislead the inexperienced herpetologist than the ornithologist. Nevertheless, the latter should bear the similarity in mind when relying on sound only.

Little Bustard Otis tetrax Palmér & Boswall (1969–80) pointed out that the snorting 'prett' calls of the Little Bustard may be confused with the croak of the Parsley Frog, and give a recording of this amphibian on disc RFLP 5003 for comparison. We agree that they are similar and that confusion could occur in the breeding season, when the frog is often active by day. The distributions of both species overlap to a considerable degree: the Little Bustard breeds throughout much of the Iberian peninsula and France, except the north and east, Sardinia, Sicily, southern Italy and part of the Balkans; while the Parsley Frog is found over most of France, extreme northwestern Italy and the whole of the Iberian peninsula (Arnold & Burton 1978).

Woodcock Scolopax rusticola The croaking of breeding Common Frogs at dusk or dawn in spring could, in certain circumstances, mislead an incautious ornithologist into believing that he is listening to a roding Woodcock, especially as the two may occur in the same habitats. It is also possible that some croaks of the Edible Frog or the Marsh Frog Rana ridibunda may mislead birdwatchers on the European mainland.

Little Bittern Ixobrychus minutus Kettle & Boswall (1981) have drawn attention to the fact that a recording made in Germany and published by Palmér (1958-63) purporting to be of a Little Bittern is actually of a Fire-bellied Toad. JFB has listened to this and to recordings of genuine Little Bitterns (he has also heard them in the field in the Netherlands); he considers that confusion is more likely between some of the resonant 'awk' or 'quawk' calls of Edible or Marsh Frogs, although the similarities are not great, the bird's call sounding sharper.

Conclusions

There are a surprising number of amphibians and insects whose sounds resemble the calls or songs of European birds and could be confused with them. Although in some cases the resemblances are remarkably close, there is little real risk of confusion, so long as ornithologists are aware of the danger and take the trouble to ensure that errors of identification are avoided. This may mean less hasty judgements, and a determination to track down an unseen singer until its identity is beyond doubt.

A further word of caution: ornithologists must appreciate and allow for individual variations in both amphibians and insects. This often depends upon the prevailing temperature, which affects the frequency and volume of the sounds produced by these animals: as one would expect, they respond more actively to high temperatures. As Arnold & Burton (1978) have pointed out, the pitch of amphibian vocalisations can vary with the size of the calling individual.

It should also be noted that the stridulations of many bush-crickets and other insects may become inaudible, or nearly so, to persons who have reached middle age, as may the higher frequencies of some bird songs.

We are not familiar with all the sound-producing insects (mostly grass-hoppers, bush-crickets, crickets and cicadas) in Europe, and it is highly likely that there are other species not listed here whose songs closely resemble certain species of birds. We would therefore be grateful for any new cases to be brought to our notice. We are, however, confident that there are not likely to be any new examples among the amphibians as we have listened to almost all known European species, either in the field, or on tape or disc, or both.

Discs and tapes consulted

ANDRIEU, A. J., & DUMORTIER, B. 1963. Chants d'Insectes. One 30-cm 33.3rpm disc, no. LDP-E 7331-Med. Pacific. Neuilly, France.
BBC SOUND ARCHIVES. Collection of Natural History Recordings.
FEINDT, P. 1968. Vier Europäische Rallenarten. One 17-cm 45rpm disc, OV1. Hildersheim, German Federal Republic.
KÖNIG, C , & THIELCKE, G. 1965. Vögelstimmen aus Sudeuropa. One 17-cm 45rpm disc, no. 75-09255. Kosmos, Stuttgart, German Federal Republic.
NAUMOV, R., & VEPRINTSEV, B. N. 1964. The Voices of Birds in Nature: 4 Birds of Siberia. One 25-cm 33.3rpm disc, 14867-8. Melodia, Moscow, USSR.
NIKOLSKY, I. 1979. Voices of Amphibians. One 25-cm 33.3rpm disc, no. 41503/04 Melodia 33M71. Moscow, USSR.
ORSZAG, M. 1980. Bird Songs of Hungary: Sounds of Woods and Reeds. One 30-cm 33.3rpm disc. Hungaroton, Budapest, Hungary.
PALMÉR, S. 1958-63. Radions Fågel Skivor. 35 17-cm 45rpm discs, nos. RFEP 201-35. Sveriges Radio, Stockholm.
— & BOSWALL, J . 1969–80. The Peterson Field Guide to the Bird Songs of Britain and Europe. 15 30-cm 33.3rpm discs, RFLP 5001-5015. Swedish Radio Company, Stockholm. (Also available on 16 boxed cassettes, SRMK 5021-36, published 1981.)
RAGGE, D. R., BURTON, J . F., & WADE, G. F. 1965. Songs of the British Grasshoppers & Crickets. One 17-cm 33.3rpm disc, no. 16108-9. Warne, London.
ROCHÉ, J. C. 1961. Les Voix de la Nuit. One 17-cm 33.3rpm disc, LVBL. La Vie des Betes, Paris.
— 1965. Guide Sonore du Naturaliste: 1 Insectes. Guide Sonore du Naturaliste: 2 Batraciens. Two separate 17-cm 45rpm discs. Roché, Haubourdin, France.
SIMMS, E. 1969. British Mammals and Amphibians.
BBC Records, Wildlife Series. One 30-cm 33.3rpm disc no. RED 42M.
TEMBROCK, G., & SCHUBERT, M. 1970. Stimmen der Vögel Mitteleuropas: II Wasservögel. One 30-cm 33.3rpm disc, Eterna 8 20 999. VEB Deutsche Schallplatten Berlin, German Democratic Republic.
UNGER, O. 1971. Zvukový Atlas Prírody I (Sound Atlas of Nature.); Ptáci v leseauvody. Czechoslovak Radio (Ceskoslovensky Rozhlas), Prague.

Acknowledgments

We are grateful to I. J. Ferguson-Lees for the initial conception of this paper and for his valuable suggestions and criticisms in its early stages of preparation. Dr D. R. Ragge and Eric Simms also read an early draft and we are indebted to them too for their comments. Moreover, we are particularly grateful to Dr Ragge for answering many entomological questions with exemplary promptness and for providing details of the distribution of some of the insects discussed. A special word of thanks is also due to Stanley Cramp, Peter Grant and Dr J. T. R. Sharrock for encouragement in seeing this paper through to its conclusion, and to Dr Hans-Heiner Bergmann for drawing our attention to relevant papers published in Germany. We have enjoyed the benefit also of discussions on particular points with Jeffery Boswall (whose admirable discographies were a splendid aid to this study) and Ron Kettle (British Library of Wildlife Sounds). Mesdames Helga Burton and Libuše Taylor kindly provided translations from German and Czech respectively, while Miss Margaret Reese made a marvellous j ob of the final typescript from a jumble of amendments and inserts to the original draft. Finally, J FB would like to express his appreciation to his colleague Michael Bright, Senior Radio Producer in the BBC Natural History Unit, for his interest and encouragement. The BBC's collection of wildlife sounds has, of course, been a tremendous source of comparative material and we are indeed grateful for access to it.

Summary

The danger of confusing the songs of the Nightjar Caprimulgus europaeus and the Common Mole-cricket Gryllotalpa gryllotalpa, and the calls of Scops Owl Otus scops and the Midwife Toad Alytes obstetricans are well known; less well known is a similar risk of mistaking the sounds of a number of other European amphibians and insects for birds, especially when the listener is in unfamiliar territory. The authors have investigated, as far as possible, using published and unpublished sound recordings, and drawing upon their considerable field experience, all other likely instances.

The following bird species are further examples and are discussed in this paper, together with the Nightjar and Scops Owl: Savi's Warbler Locustella luscinioides, River Warbler L. fluviatilis, Grasshopper Warbler L. naevia, Lanceolated Warbler L. lanceolata, Pygmy Owl Glaucidium passerinum, Baillon's Crake Porzana pusilla, Little Crake P. parva, Corncrake Crex crex, Coot Fulica atra, Little Bustard Otis tetrax, Woodcock Scolopax rusticola and Little Bittern Ixobrychus minutus.

The sounds made by species of amphibian and insect with which they can be confused are described and compared, and also listed systematically in a cross-reference section, which we hope will prove useful to entomologists and herpetologists as well as to ornithologists.

The circumstances under which confusion between the sounds of these unrelated animal groups can occur are also discussed. The biological advantages to birds of producing amphibian- or insect-like sounds are also considered briefly. In particular, it is postulated that some of the Locustella warblers do in fact mimic the more abundant insects, such as certain bush-crickets, crickets and cicadas, thereby deceiving potential predators through a sort of acoustic camouflage. Moreover, it is suggested that this may explain why individual River Warblers may sing like cicadas and others like bush-crickets, and it is therefore further suggested that this species should be the subject of a special study.

References

ARMSTRONG, E. A. 1973. A Study of Bird Song. New York.
ARNOLD, E. N., & BURTON, J. A. 1978. A Field Guide to the Reptiles and Amphibians of Britain and Europe. London.
BENNET-CLARK, H. C. 1970a. The mechanism and efficiency of sound production in mole-crickets. J . Exp. Biol. 52: 619-652. 1970b.
— A new French mole-cricket, differing in song and morphology from Gryllotalpa gryllotalpa L. (Orthoptera: Gryllotalpidae). Proc. R. Ent. Soc. Lond. (B) 39 (9-10): 125-132.
BERGMANN, H-H., & HELB, H-W. 1982. Stimmender Vögel Europas. Munich.
BOSWALL, J. 1966. A discography of Palearctic insect sound recordings. Entom. Record 87: 202-206.
— 1969. New Palearctic bird sound recordings during 1968. Brit. Birds 62: 271-281.
GRANT, P.J. 1983. Identification pitfalls and assessment problems: 2. Savi's Warbler Locustella luscinioides. Brit. Birds 76: 78-80.
HARDY, E. 1939. A rare toad (Natterjack). Field 174: 415.
HARRISON, C. 1982. An Atlas of the Birds of the Western Palaearctic. London.
HOYER, G. 1967. Zwergohreule—Otus scops—und Rauhfusskauz—Aegolius funereus—im Burgwald, Landkreis Marburgs. Luscinia 40 (1): 19.
HVASS, H. 1972. Reptiles and Amphibians in Colour. London.
KETTLE, R., & BOSWALL, J. 1981. A discography of amphibian sounds. Recorded Sound 79: 51-75.
KÖNIG, C. 1968. Zur Unterscheidung ähnlicher Rufe von Zwergohreule (Otus scops), Sperlingskauz (Glaucidium passerinum) und Geburtshelferkrote (Alytes obstetricans). Orn. Mitt. 20: 35.
NORTH, M., & SIMMS, E. 1958. Witherby's Sound Guide to British Birds. Part Two. London.
PALMÉR, S., & BOSWALL, J. 1969–80. Special note to The Peterson Field Guide to the Bird Songs of Britain and Europe (set of 15 discs). Stockholm.
PETERSON, R., MOUNTFORT, G., HOLLOM, P. A. D. 1974. A Field Guide to the Birds of Britain and Europe. London. Third edition.
RAGGE, D. R. 1965. Grasshoppers, Crickets and Cockroaches of the British Isles. London.
THÖNEN, W. 1968. Die Ähnlichkeit der Rufe von Zwergohreule, Sperlingskauz und Geburtshelferkrote. Orn. Beob. 65: 17-22.
THORPE, W. H . 1957. The identification of Savi's, Grasshopper and River Warblers by means of song. Brit. Birds 50: 169-171.
UNGER, O. 1971. Zvukový Atlas Prírody I: Ptáci v leseauvody. Prague.
WHITE, G. 1789. The Natural History of Selborne. London.