The Sound Approach guide to nocturnal flight calls

Magnus Robb,
Lukas Pelikan &
The Sound Approach
NFC, Night flight calls
22nd March 2021

Mass nocturnal bird migration witnessed on board a ship in the Baltic, just northeast of Rügen, Germany, 22:37, 18 October 2015 (Martin Grimm – BIA).

Species

 

Contents

Introduction

At some point you may have heard that a vast multitude of birds migrate at night. While it is now fairly popular to watch birds migrating at daytime, you may have wondered, how on earth can we experience bird migration at night?

In the distant past, ornithologists actually took quite a while to be sure that birds migrate at all. Nowadays we can observe migration directly through ‘vismig’ (from ‘visible migration’): sitting in a fixed spot and watching as masses of birds fly in the same general direction, either singly or in groups, to or from their wintering grounds. When flying past at speed, many such as smaller passerines are easier to identify by ear than by eye; others, especially seabirds such as shearwaters can only be identified visually.

For some groups such as waders or most herons, migration by day or by night is much the same thing, and the one often flows seamlessly into the other. This means that at night you can hear the same diagnostic calls crossing the sky, just as you could during the day, only now you rely entirely on your ears to know what is going on.

Listening to nocturnal migration is a niche activity but growing in popularity to the point that it has its own nickname: ‘nocmig’ (from ‘nocturnal migration’). A big difference from ‘vismig’ is that many birds pass undetected: we can only detect the ones that call. Many individuals and even flocks belong to species with flight calls but choose not to use them. In addition, there are large taxonomic groups for which we have not yet detected any flight calls, such as most warblers Sylviidae. We think that this disadvantage is more than compensated by everything we can learn, and would otherwise never know, from the birds that do call. If you are interested in a more exhaustive introduction to nocturnal activity in birds, see Martin (1990).

In the past, a number of studies tried to quantify nocturnal migration (Martin 1990 and Newton 2008). Radar studies, for example, have the advantage of being able to detect and quantify birds flying at high altitudes, but a disadvantage is that most birds cannot be identified to species level. When listening to nocturnal migration it is basically the other way around. We can identify most callers to species level, but it is more difficult to estimate numbers, and impossible to detect birds that do not call, except for those passing close enough for their wingbeats to be audible to us.

There are two main approaches to nocmig: you can either sit outside, perhaps even at your favourite vismig spot, and listen to the birds calling overhead, or you can leave recording gear outside, giving you the opportunity to check identifications later, or simply because you need your nights for sleeping. In fact, the strongest nocturnal flight call activity often occurs towards the end of the night, when very few people are awake.

Magnus Robb listening to nocturnal migration at the fortress of Sagres, Vila do Bispo, Portugal, 11 September 2010 (Pim Wolf). Sagres is the town closest to the southwest corner of Europe, the last land many migrants pass before crossing the sea to Africa. It is an excellent place to listen to nocturnal migration, and many of the recordings in this guide were made in the area.

Aims of this guide

Many of us have struggled to make sense of the birds we hear at night. We do our best to compare calls with what we remember from daytime, recordings from our own collections and archives such as xeno-canto. We consult collections of daytime flight calls such as Out of the blue or Birds in flight, and articles on nocturnal flight calls on websites such as this one or nocmig.com. At the time of writing, only Eastern North America has a comprehensive nocmig guide – Flight calls of Migratory Birds (Evans & O’Brien 2002).

This guide will try to fulfil the need for a thorough guide to nocturnal flight calls in Europe, while serving both approaches to nocmig. For the sheer pleasure of sitting outside and listening live, it can help with carefully chosen examples, descriptions of calls, and information on where and when you are most likely to hear a certain species. For the analysis of your sound-recording, it explains how to identify calls via sonagrams (graphs plotting frequency along time, also known as spectrograms), and the measurements that can be obtained from them. If you are unsure where, when and how to record and what to log, there is A Protocol For Standardised Nocturnal Flight Call Monitoring, which gives further advice for obtaining nocmig data.

The first phase will be to cover the 50 nocturnal migrants we judge most likely to be recorded giving flight calls during nocmig sessions in western and central Europe. We will not include wildfowl among the first 50, as their calls are the same day and night, and there are plenty recordings available elsewhere. For Common Scoter see this earlier article. Perhaps you may notice that a species regular in your area is not among the first 50. We will include it later. Over time we hope to give better coverage for eastern Europe, and to be able to include many times more than 50 species. Let’s see how far we can go.

Source of the recordings

The recordings in this guide will be mostly from the collections of The Sound Approach, Lukas Pelikan and some long-term ‘nocmig colleagues’ including Nick Hopper, Steve Klasan and Tom Wulf. When they are from The Sound Approach, the recordist’s name is indicated by his initials in the serial number, given at the end of the caption. Most are by Magnus Robb (MR); others are by Arnoud B van den Berg (AB), Mark Constantine (MC), Dick Forsman (DF), Killian Mullarney (KM) and Paul Morton (PM). To diversify our collection, we have sent equipment to collaborators in interesting locations who have recorded for us. Starting with the top contributors, they are: Geoffrey Morrison (GM) at Sagres in Portugal, the Besh Barmag (BB) team in Azerbaijan, John Bowler (JBr) on Tiree in Scotland, Nuno Vieira (NV) of the STRIX team in eastern Tunisia, Jem Babbington (JBn) near Dhahran in Saudi Arabia, Jon Strandberg (JS) at Paralimni on Cyprus, Thor Veen (TV) in British Columbia, and finally Pierre-André Crochet (PAC) on Corvo and Alexandre Leitão (AL) on Flores, both islands in the Azores. All such recordings were later analysed by MR, whose initials are included in the file names. The initials of the actual recordist, as indicated above, will be shown in brackets after the location. Many other recordists have kindly made their own recordings available to use and their names will be given in full in the captions. See the acknowledgments for a full list.

Besh Barmag migration bottleneck, Azerbaijan, 28 May 2018 (Michael Heiß). The coastal plain narrows to a thin strip here, squeezed between the Caspian Sea on one side and the start of the Caucasian mountain range on the other. The town on the right is Zarat, where the Besh Barmag team recorded for us nightly from early September to early December in 2018. For examples of nocmig counts from this location, see here.

The species accounts

Each species account starts with a recording that aims to reflect a favourable listening experience with the species, often including calls at a range of distances, either the same individual or several. This recording is meant to give a first impression of how the species may sound in the environment of a listening post. If you only listen to one example per species, listen to this one. It has been chosen with care. Accompanying the recording is a representative sonagram at a scale that we would recommend using while searching through sonagrams: typically 4 seconds to a line and frequencies up to 10 kHz. Sometimes we will show a longer stretch when appropriate. There is a scale bar at the bottom left to make it easier to compare this sonagram’s scale to that of the more detailed sonagrams in the identification section below.

Immediately below the first recording, you can read about some of the peculiarities of this species’ migratory behaviour and NFCs (nocturnal flight calls): for example, whether they differ from daytime calls, how many types, and how we came to know them. If the species uses a different flight call at night than during the daytime, or has no diurnal flight call at all, then we will present a convincing example to show how, during the day, it does use calls similar to the NFC but possibly in other contexts.

 

Identification

The identification section provides relevant features for the identification of the species’ NFC. First, it presents mean values for measurements of duration and frequency. The standard deviation and sample size are given in parentheses. Sample size gives number of both individuals and calls, which are the same: from any given recording, we have chosen one call, the clearest one available for measurement.

What follows is a block of typically 8-12 sonagrams of typical NFCs of the relevant species, covering the variation of the call. The number of NFCs shown will depend on the degree of variation in the call. Note that sonagrams in this section will be at a different scale from the one at the top of the page (see the scale bar at the bottom left). Occasionally, for particularly low-pitched calls, we may choose to show a different frequency scale from the usual 0 – 10 kHz. If so, we will draw attention to this. Each single call in these sonagrams has a letter above it. The letter refers to the longer recording and its caption below, which gives information about location, date, and source. The letters also appear in the descriptions at the start of the identification section, where they indicate typical examples of a feature being described. The call shown is usually the first and loudest in the recording. When this is not so, we tell you at what point in the recording the call occurs.

You should be aware that calls of birds at either very close or distant range may appear slightly or strongly different to those presented here. In addition, the acoustics of your location (e.g., presence of walls or cliffs causing echoes) and the quality of your recording equipment will have consequences for how the sound changes between leaving the bird and arriving at your ear. In a subsection with the title Effects of recording quality, we draw attention to those aspects of the call that may be particularly subject to variation.

 

Similar NFCs

You should certainly have a look at this! Many times, NFCs of one species are easily confused with those of another (eg Common Moorhen Gallinula chloropus and Bar-tailed Godwit Limosa lapponica, Redwing Turdus iliacus and Common Blackbird T merula, or Spotted Flycatcher Muscicapa striata and European Robin Erithacus rubecula). It may take some time until you finally work out the correct species. We also made plenty of errors in the past and this section is intended so you don’t have to make the same ones. If you want to see sonagrams of the comparison species, click on its name and you will be redirected to its species page. If there is no hyperlink, we have still to construct a page for that species.

 

Where and when?

The three points refer to:

Where? Diurnal as well as nocturnal migration is thought to happen mostly over a broad front. However, at certain points both diurnal and nocturnal migration can be canalised along rivers, coastlines and peninsulas or deviated by mountain ranges, large lakes or oceans. At night, these effects can be different from daytime, and species-dependent. For example, waterbirds are more likely to be recorded over or near water bodies. We will note here if we have recorded a species over all kinds of habitats or mainly over certain habitats.

How nocturnal? Some species use more or less the same flight-call during nocturnal migration as during their daytime flights, while others have a different flight-call at night.

What time of night? Some species rarely appear except during the last couple of hours before dawn (eg, Dunnock Prunella modularis). Others like Common Scoters Melanita nigra may peak at a particular hour at your location, depending on how long it takes to arrive from the coast, or some other migration stopover.

 

Phenology graphs

The phenology graphs are based purely on nocmig data and are intended to give an idea of the seasonal occurrence of nocturnal flight calls in two different areas within the Western Palearctic. As there are birds that travel mainly via an eastern route to their wintering grounds in Africa, and others that use a more westerly route, the phenology can differ markedly from place to place. For this guide we decided to use central Europe and south-western Europe. We hope to include south-eastern Europe later when there is enough data. Phenology for countries like France or the UK is likely to be intermediate between central Europe and south-western Europe. The data behind the graphs was mainly obtained from Germany (especially Potsdam, Brandenburg) for central Europe and from Portugal (especially Cabriz, Sintra and Sagres, Vila do Bispo) for south-western Europe. We also consulted additional data from other locations.

The graphs show maximum estimates of individuals per night for ten-day periods in three steps, which can be read as a low, medium or high level of nocturnal flight call activity.

As nocmig is still in its infancy, the phenology data is only provisional and likely to be adjusted in the future.

 

Note of caution

For nocturnal flight calls, it is important always to be realistic about how much we still don’t know, and this section draws attention to challenges yet to be tackled. These often concern less well-known species we may need to exclude from our identifications, geographic variation still to be studied in more detail, or matters concerning the flight behaviour of certain species at night.

 

Further reading

Here we list publications referred to in the species page, or which are a useful reference for the species. It is by no means intended to be a complete list.

Definitions

Night and nocturnal flight calls

For this guide we define night as the period between civil dusk and civil dawn, when the sun is 6° or more below the horizon. This is a fairly generous definition of night, and for a short period at the start and end of it some species will be using their daytime call repertoire. However, we have chosen this definition because virtually all observation of migration during this period will have to be done by ear: it is too dark to recognize most bird species visually. Any examples recorded outside this period will be marked clearly as such, even if they miss it only by a minute or two.

The term nocturnal flight call or NFC is, for the purposes of this guide, mainly used for calls given during migration flights. We also include calls in flights between roosting and feeding areas that take place at night (eg, Jack Snipe Lymnocryptes minimus), mainly because these are indistinguishable from migration calls. We also include vocalisations of some species that may form part of nocturnal song flights, ie, prospecting for breeding partners and/or unoccupied territories, where there is evidence that these may also be used during migration flight. Examples include Common Quail Coturnix coturnix, Little Grebe Tachybaptus ruficollis, Common Moorhen Gallinula chloropus and Great Spotted Cuckoo Clamator glandarius). Occasionally even sandpipers Scolopacidae may give snatches of song while migrating, and where we have examples that were clearly nowhere near breeding habitat we may include them. We will not include nocturnal songs that are normally only given on breeding territories, such as those of Eurasian Woodcock Scolopax rusticola or larks Alaudidae. Note that for some species, wingbeats can be more useful than calls for identification (eg, Mute Swan Cygnus olor or Common Goldeneye Bucephala clangula) and we will also mention them, although they are not NFCs.

Bioacoustic and descriptive terms

Pitch and frequency are not synonyms, although they are related. Frequency is an objective measurement of sound waves per second, whereas pitch is a subjective sensation of how ‘high’ or ‘low’ a sound may be. Pitch is limited by our human abilities, and we are often unable to resolve details of frequency change at fine temporal scales. In general, our impression of the pitch of a sound is based on the features with the longest duration or the loudest component. So, for example, we may hear the wit of a European Pied Flycatcher Ficedula hypoleuca as simply rising in pitch, whereas at the end of the call the frequency actually often descends to a point nearly as low as the starting point, doing so in such a short time that we cannot hear this.

The peak frequency is the frequency that carries most power, ie the loudest part within the call. It is not necessarily the maximum frequency, nor in the middle of the call. As the loudest part within a call, it is the frequency that dominates the pitch (how high or low a call appears to sound). It can be measured automatically in many bioacoustics programs or measured by hand by visually inspecting a sonagram. In a black and white sonagram it is the part with the deepest black.

A frequency band is a descriptive term that makes most sense when looking at a sonagram. It describes a trace moving in time from left to right, inflected up or down as it goes. The term is most useful when describing calls with more than one band simultaneously, such as a European Robin’s NFC.

A harmonic partial, usually abbreviated to ‘harmonic’, is a component of a sound with a frequency that is a multiple of the basic or fundamental frequency. In sonagrams it usually appears as a copy of the ‘actual’ call, above it, and typically there are several. In calls that sound pure, the harmonics are very faint. In calls that sound nasal (eg, a toy trumpet), several harmonics will have equally strong power. It is the relative strength of the various harmonics that determines the timbre of the call. Several types of frequency bands exist, but harmonics are always multiples of a basic frequency. A bird can also produce two signals at once, one with each syrinx or vocal organ. These can be entirely unrelated in their frequency contour and are not called harmonics.

The fundamental frequency is traditionally defined as the lowest frequency component of a call. Harmonics are multiples of the fundamental frequency, and the fundamental is considered to be the first harmonic. In some calls, the fundamental can be strongly suppressed by the bird and therefore inaudible as well as being invisible in sonagrams (eg, in many calls of Tree Pipits Anthus trivialis).

A ‘spike’ is a brief, sharp frequency modulation pointing upwards from the baseline of an otherwise unmodulated frequency band. A classic example of spikes can be seen in most (but not all) Spotted Flycatcher NFCs.

A ‘foreleg’ is a near-vertical ascent at the start of a call, and the shape of this structure can be important for identifying certain species, eg wagtails Motacilla.

Similarly, a ‘hindleg’ is a near-vertical descent at the end of a call. The presence or absence of this structure can be important for distinguishing certain species. Eg, European Pied Flycatcher often has it whereas Spotted Flycatcher very rarely does.

A concave shape in a sonagram is a curve which dips downward, i.e. it dips noticeably below the most direct line from its starting point to its ending point. For example, 1) a U-shaped line, 2) a descending curve that descends more rapidly at the start, then levels out, or 3) an ascending line that starts off rather flat then rises more sharply towards the end.

A convex curve in a sonagram is a line which arches upward, i.e., rises noticeably above the most direct line from its starting point to its ending point. For example, 1) a simple arch, 2) an ascending curve that ascends more steeply at the start, then levels out, or 3) a descending line that starts off rather flat then descends more sharply towards the end.

The most generous definition of modulation in bird calls includes any change to the frequency or amplitude of a wave. However, here we use a narrower definition, referring to more or less regular oscillations in the frequency or amplitude, producing ‘zigzags’ (frequency modulation) or vertical hatching (amplitude modulation). Either way, the sound becomes harsher with rapid modulation than without it. Note that certain sonagram settings, such as window size or bandwidth have a strong influence on the appearance of modulations. At higher window size and smaller bandwidth values, fine modulations will disappear. If your Redwing Turdus iliacus calls show parallel bands instead of fine zigzag modulations, this is the reason. See also A word about measurements.

Secondary modulation occurs in some species when a second, usually slower layer of modulation is superimposed on the more regular, usually faster layer characteristic for the species. In Europe, this can be seen most often in European Pied Flycatcher and Redwing.

Taxonomy and English names

The taxonomy used for this guide follows the decisions of the Dutch Committee for Avian Systematics (CSNA), while English names are those used in the IOC list (Gill & Donsker 2019)

A word about measurements

For some of the more difficult species, measurements on a sonagram will be one of the best ways to secure the identification. We make ours in Raven Pro, currently version 1.6, although there are several other applications that can be used for this.

Measurements of duration are fairly straightforward. It is worth bearing in mind that the accuracy will be greater when the ‘window size’ of your sonagram is set to a lower value and you zoom in to make the call as big as possible on the screen. If your value seems to be on the low side, bear in mind that calls recorded at a distance may be missing details at the start or end of the call, making them appear shorter. Or the call you are measuring may simply have a duration at the shorter end of the range.

Measurements of frequency should always be taken at the vertical mid-point of the frequency band, at whatever point in time you are taking the measurement, even when you are measuring the maximum frequency. When measuring the top frequency of a rapid modulation, the bandwidth may be difficult to judge. In this case look at the width of the band in a less rapidly modulated part of the call and from this, estimate how far below the apparent peak of modulation you need to take your measurement.

 

Modulation rate is the rate at which a sound oscillates up and down in frequency or in amplitude, typically producing what looks like vertical hatching in a sonagram (see definitions). Often the speed of this modulation is highly characteristic for a given species. Presumably it relates to the size or proportions of certain anatomical structures such as the bill and the vocal apparatus. Rates can be measured in Hz or oscillations per second, the same units used to express frequency (when expressing modulation rates, the values are much lower). In most cases the call will be well under a second long, but we can measure a smaller part of the call and derive the rate per second from this. So, when measuring flycatcher calls, a 50 ms stretch can be multiplied by 20 or a 100 ms stretch by 10 to obtain a measurement in Hz. Even if the accuracy is not particularly high, this can be a helpful measurement.

UFOs and rusty bicycles

Some calls can be very hard to identify by ear, especially if they are rare in your area, or differ strongly from the norm for a given species. Whether you are listening live or studying sonagrams, it is perfectly normal for many calls to remain unidentified. If you make recordings, you will probably accumulate many ‘UFO’s: tantalising mysteries that may bug you for years until, eventually, you solve a few of them. Beware that mammals (especially Red Fox Vulpes vulpes: for examples see British Library 2012), as well as amphibians, reptiles and insects also vocalise at night. An unexplained, bird-like sound could even come from something as innocent as a rusty bicycle hitting a small pothole two blocks away. In short, it is always sensible to err on the side of caution, especially until you have built up some experience. We strongly recommend that you avoid the temptation to name every sound that you hear.

Here is a suggestion: during your first few nocturnal recording sessions, sit outside while the recorder is running, listen to everything you can hear and take notes when you hear something man-made. If you cannot identify it immediately, try to find out what caused it, even if it was only a chestnut falling on the roof of your neighbour’s garage. Later you may stumble upon the same sound in your recording and will not lose much time wondering what it was.

Something else to be wary of is this. In many locations, local birds may start singing several hours before dawn, both in spring and in autumn, when they will be mostly European Robins. If these birds are far enough away, only fragments of their songs may reach you or your microphones, and they may lead to false impressions of distant flight calls. Most experienced nocmig enthusiasts could probably tell you a story or two about when they were fooled in this way, and we ourselves are no exception.

We think this quotation from Louis Pasteur (1854) is an appropriate one to end with: “In the field of observations, fortune favours only the prepared mind.”

Recording ‘nocmig’ from an apartment in Freiburg, Baden-Württemberg, Germany, 5 April 2020 (Ralph Martin). The instructions for a home-made parabolic microphone like this one can be found here.

Acknowledgments

We would like to say a special thank you to those who have recorded nocturnal migration for us. Our most dedicated helpers have been Geoffrey Morrison at Sagres in Portugal in 2017-2019 and the Besh Barmag team in Azerbaijan in 2018 and 2019 (especially Michael Heiß, Zülfü Fərəcli and Elvin Məmmədsoy). We are also grateful for the help of Jem Babbington in Saudi Arabia, John Bowler in Scotland, Pierre-André Crochet on Corvo and and Alexandre Leitão on Flores, Azores, Jon Strandberg in Cyprus, Klas Strandberg in Sweden, Thor Veen in British Columbia and Nuno Vieira in Tunisia.

Many correspondents have shared recordings and ideas, and to all of them we are very grateful. For the species covered so far, contributions from Jonas Buddemeier, Frédéric Cazaban, David Darrell-Lambert, Margus Ellermaa, Nicholas Galea, Nick Hopper, Steve Klasan, Ralph Martin, Ireneusz Oleksik, Eduardo Realinho, Julien Rochefort, Seán Ronayne, Albert Cama Torrell, Stanislas Wroza and Tom Wulf have been particularly helpful. Other names will appear here as the guide expands.

Special thanks to Martin Grimm, Michael Heiß, Ralph Martin and Pim Wolf for the photos that appear in the introduction.

For inspiration and setting high standards, we would like to honour pioneers in North America: in particular Bill Evans, Andrew Farnsworth, Michael Lanzone and Michael O’Brien. And last but not least, a European pioneer from the 19th century: Johann Friedrich Naumann, who was not only a true birder and keen observer but also knew a lot about nocturnal flight calls.

 

In order to proceed to the family and individual species pages, click on the links below.

Species

 

Literature

British Library 2012. Sounds of the night: An audio guide to Britain’s nocturnal species. Audio CD. London.

Evans, W R & O’Brien, M 2002. Flight calls of migratory birds. Eastern North American landbirds.

Gill, F & Donsker, D 2019. Birds of the world: recommended English names.

Martin, G 1990. Birds by night. London

Newton, I 2008. The migration ecology of birds. London.