The first recordings of animal sounds date from the end of the 19th century. The earliest known example is Ludwig Koch's recording of a captive bird, a shama, made on wax cylinder in Germany in 1889. The first wild bird sound recordings were probably those of a song thrush and a nightingale made by Cherry Kearton, also on wax cylinder, in England in 1900. The pioneers,  such as Ludwig Koch in Germany, Belgium and Britain,  Carl Weismann in Denmark, and A.R. Brand, Paul Kellogg and A.A. Allen in the United States, made many bird and other animal recordings by direct cutting of discs. The cumbersome equipment required was a great hindrance and it was the development of portable magnetic tape recorders in the 1950s which made possible the collection of wildlife sound recordings all over the world on a larger scale and on a systematic basis. This was also greatly facilitated by other technical developments like the invention of the parabolic reflector and the gun microphone.

As long ago as the early 1930s Ludwig Koch cherished the idea of founding an institute to preserve sound recordings of all kinds and he tried successively, but failed, to get one set up in Germany, Belgium and Britain. At about the same time in the United States Peter Paul Kellogg and his associates began to build up, from their own recording activities, a collection of bird and some other animal sound recordings at Cornell University which was to become the Library of Natural Sounds at the Laboratory of Ornithology there. This was the first, and is still the largest, wildlife sound archive in the world. By 1980 it held about 40,000 tape recordings of some 4,000 species - mainly birds - from all over the world. Two other important collections were established in North America around 1950: the Borror Laboratory of Bioacoustics at Ohio University, specialising in insect and bird sounds of North America, primarily for internal research; the Gunn Library of Wildlife and Environmental Sounds in Toronto, which is a commercial organisation, supplying many recordings for broadcasting and films - but also for research - and comprises animal sounds from North America and some areas of South America and Africa.

It was not until the 1960s that the movement spread to other parts of the world. It saw the formation of sound libraries in Denmark (at the Bioacoustics Laboratory of the Natural History Museum, Aarhus), Hungary (Animal Sound Archives of the Academy of Sciences in Budapest), Britain (British Library of Wildlife Sounds at the British Institute of Recorded Sound in London), USSR (Phonotek of Animal Sounds at the Institute of Biophysics, Pushchino, the Moscow University Library of Animal Voice, as well as a specialised collection at Leningrad University), Australia (at the CSIRO Division of Wildlife Research, originally in Canberra but now in Western Australia), New Zealand (Wildlife Service Sound Library, Wellington), South Africa (Fitzpatrick Bird Communication Library formerly in Natal but now at the Transvaal Museum, Pretoria) and in Brazil (Laboratorio de Bioacoustica at the University of Campinas).

These ‘libraries' developed for different reasons, and few had as their primary objective the establishment of a national archive of wildlife sound. Some of them sprang from the accumulation of an individual's recordings and some from the need for a research resource, but all do provide a centralised collection which can be drawn upon by any person or organisation with a serious use for such recordings. All of them are associated with zoological institutions of one kind or another except the British Library of Wildlife Sounds (BLOWS).

Some broadcasting organisations have in their sound archives sizeable collections of natural history recordings, notably those in Sweden, Finland, Japan, Czechoslovakia, Britain and South Africa. Smaller, specialised collections have been built up in some zoological institutions, such as the amphibian sound recordings in the Natural History Museum in Geneva and the grasshopper and cricket recordings in the Natural History Museum in London. Some private collections have almost the status of a sound library, e.g. those of Claude Chappuis and Jean-Claude Roche in France, Claus König in Germany, Ken Scriven in Malaysia, Leslie McPherson in New Zealand, John Kirby in Britain and Tsuruhiko Kabaya in Japan.

Starting slowly in the early 1900s, but at an accelerating rate in the 1960s and 1970s, a large number of wildlife sounds on commercial gramophone records and cassettes have been published all over the world (see the discographies by Boswall and others¹) and over a thousand may now be listed. A collection of these is therefore an important resource for the would-be listener to the sounds of particular species which may not be available amongst the unpublished tape recordings in an archive. Partly as a result of the availability of these commercially published discs and cassettes, but even more as a result of radio and television programmes, an increasing interest in the sounds and language of animals is being taken by the general public. At the same time zoologists have been giving more attention to the study of bioacoustics and the expansion of this research into animal sounds has been stimulated by the greatly increased possibilities which now exist  of acquiring recordings.

Thus there is a growing demand for wildlife sound recordings at the popular, educational and scientific levels. A centralised archive of such recordings has an important role to play in satisfying this demand.  It may be drawn upon for general interest, for the publication of discs or cassettes, for film and broadcasting use, for various educational applications and for the provision of sound •specimens' for scientific research.

1. Boswall, J. 'A bibliography of wildlife discographies' in Recorded Sound No.54; 1974
   Boswall, J. 'A supplementary bibliography of wildlife discographies' in Recorded Sound No. 74-75; 1979

A wildlife sound archive may be defined as a centralised collection of animal sound recordings properly preserved and documented for their cultural and scientific value. Such an archive may house recordings of all classes of animals from all parts of the world, as does BLOWS. Alternatively it may have a much more limited s~0pe and collect recordings only from the local area or only of one class of animals. The Australian CSIRO collection, for example, is virtually restricted to birds of the Australasian zoogeographical region.

The primary concern of this chapter will be the setting up of archives devoted to collecting wildlife sound recordings from their own local geographical areas. The resources required to develop an archive of worldwide scope are probably such that their establishment is justified in only a few centres in the world. It may be that the ideal course is the setting up of many national collections of recordings of the fauna of individual countries which can both supply and draw upon comprehensive archives of world-wide scope in a few international centres.

Whatever the scope, any wildlife sound.archive should aim to include the whole vocabulary of all the species occurring within the geographical and faunistic limitations set. In addition the more examples of, for instance, each 'song' the better, especially from a wide range of localities, dates and timings. Research is often concerned with the study of these variations and the analysis of as many different examples as possible.

Any policy for building up a wildlife sound archive should give priority to preserving the following kinds of material:

the sounds of endangered or vanishing species and those from disappearing habitats;

the original material of pioneers in wildlife sound recording and large important collections of other recordists;

recordings made for major research projects;

published recordings on disc or tape while they are available.

Clearly, an essential starting point for setting up an archive is the identification of existing sources of recordings. For wildlife sounds, the following are likely to be the most fruitful.

The collections of major pioneer wildlife sound recordists should be eagerly sought. Such people as Ludwig Koch in Britain, Myles North in East Africa and Paul Schwartz in South America fall into this category. All their recordings have now been archived. The acquisition of such collections may involve the archive in a great deal  of sorting, copying and documentation.

Individual zoologists who make recordings as part of their research may contribute valuably to archive collections. Of similar importance are zoology departments of universities and other research institutions which have collected recordings for scientific projects. The archiving of such recordings will enable future workers to have access to material on which the research was based and, in many cases, from which published sonograrns were prepared. A good example is the collection of recordings at Cambridge made for Professor W.B. Thorpe's classic study of song learning in birds, with special reference to the chaffinch, of which copies are preserved in BLOWS.

Natural history museums and learned societies may also be rich sources. Thus, in the former category, comes the Orthoptera (i.e. grasshoppers, crickets, etc.) recordings collected by Dr David Ragge and Jim Reynolds at the British Museum (Natural History) in London, copies of which are now at BLOWS. Expeditions to zoologically rich but unexplored areas often obtain valuable recordings. Thus Julian Dring made many unique recordings of frogs in Sarawak during an expedition run by the Royal Geographical Society and contributed copies of them to BLOWS. Similarly, there may be private expeditions organised by individuals or groups primarily for the purpose of recording wildlife sounds. Some fine batches of bird sound recordings, for example, have been contributed to BLOWS by Terry White from expeditions to the Gambia, Trinidad, Tobago, Australia and Brunei.

If, on any type of expedition, there is a competent wildlife sound recordist with proper equipment present, so that good recordings seem likely to be obtained of animal sounds particularly required by the archive, support may be given to the work at least to the extent of supplying tape and equipment. Such support should be subject to a signed agreement that all the recordings will be deposited in the archive within a stipulated time and with proper documentation.

Recording expeditions organised by the archive itself can be a most profitable source when a new collection is being set up in a country without a tradition of animal sound recording. Obviously areas with a particularly interesting or rich fauna or, more importantly, disappearing habitats with threatened species will be chosen for such fieldwork. It is always wise to consult experts when organising such expeditions, such as Dr James Gulledge, Director of the Cornell Library of Natural Sounds, who has led collecting expeditions to Arizona and the Galapagos Islands to obtain recordings for the Library. Suitable high quality equipment will be needed (see 5 below) and at least one member of the party must be a competent wildlife recordist. There must also be an experienced naturalist capable of identifying all the species recorded and it is important to keep the fullest possible documentation (see also 5 below).

Broadcasting organisations may have extensive collections of natural history sound recordings in their archives. Since the primary purpose of these collections is to serve the needs of broadcasting they are not readily available for use by outside people. The broadcasting authority may, however, be willing to supply a copy of the recordings to a national archive -provided copyright is safeguarded -thus enabling them to be more widely used without involving the authority in any further work. The deposit of duplicates of the BBC's natural history recordings in BLOWS is an outstanding example of this kind of co-operation.

Individual recordists should by no means be ignored as acquisition sources for archives, especially since an increasing number of people are taking up wildlife sound recording as a hobby. The major wildlife sound recordists in Britain, for instance such people as John Kirby, Lawrence Shove, Ray Goodwin, Victor Lewis, Patrick Sellar, Richard Margoschis and other leading British bird and animal recordists, have extensive collections of first-class material which has been used in published records, films and broadcasting. Many others who record wildlife sounds as a hobby also make excellent recordings. In Britain such enthusiasts are easy to identify because most are members of the Wildlife Sound Recording Society. Elsewhere they may be harder to discover. Some will be members of natural history societies. Others, who are primarily interested in sound recording for its own sake, may belong to tape recording societies through which they can be discovered. The latter often make recordings of good quality, but may be weak on identification and documentation, which can detract from the value of their tapes.

There are many problems to overcome in making good wildlife sound recordings in the field and some special equipment and techniques are necessary.

Most animals are difficult to approach closely enough to record their utterances with an ordinary hand-held microphone. This can be overcome by carefully placing the microphone, which is connected to the recorder by a long lead, near the spot where the animal is expected to utter its sound, such as by the song-post of a bird. Another method is to use a parabolic reflector, which concentrates the sound waves and reflects them onto the microphone, thus effectively amplifying the sound.

Extraneous noise is usually an inherent problem when trying to record any particular sound in an outdoor environment. The amount of background sound may be considerably reduced by the use of the highly directional parabolic reflector, which excludes most of the sound emanating from other than the direction of aim of the parabola. Another directional device is the gun microphone, which similarly excludes sounds other than those coming from a narrow frontal arc. An ordinary directional microphone close to the subject is ideal where possible. The unwanted noise caused by wind is a frequent problem, but there are some fairly effective windshields that can be fitted to the microphone to reduce this.

Field recorders need to be high quality, light-weight, reel-to-reel portables and the choice is limited to the very few machines in the Nagra and Uher categories. Monitoring by headphone is essential and the recorder, preferably, should provide this facility directly offtape. Since wildlife sounds cover a very wide range of frequencies, the equipment's frequency response must be the widest possible (ideally 20 to 20,000 Hz.). Finally, as many animal sounds are of a low level relative to the ambient noise, it is often necessary to use a high gain setting in order to record them loud enough. For this reason it is important the recorder should have as high a signal-to-noise ratio as possible.

For outdoor use microphones need to be robust, have high sensitivity for comparatively low level sounds and a good frequency response. Only dynamic (or moving coil) and capacitor (or condenser) microphones are suitable for fieldwork.

There are particular problems and equipment requirements in some types of wildlife recording, such as that of the ultrasonics of bats and underwater creatures, for which specialist advice should be taken.

As a minimum for fieldwork, the archive will need to have two recorders (preferably Nagras), one directional and one omnidirectional microphone, a gun microphone, a 20 or 24-inch parabolic reflector and monitoring headphones. For archive work (copying, editing and 'processing' recordings), three studio recorders with ancillary equipment and playback machines for discs and cassettes are minimum requirements.

For field recording, long-play tape is probably the best type to use with most portable recorders, although the use of standard-play has the advantage of also being suitable for storage if the original material is being archived.

The special problems, and the requirements for overcoming them, are a considerable hindrance to the acquisition of good quality wildlife recordings. For most scientific purposes, and indeed for most other uses, it is essential that recordings are of high quality, free from distortion and obtrusive extraneous sounds. The very highest technical standards must be aimed at. For more detailed information the reader is referred to the technical chapter and bibliography in this book.

Wildlife sound recordings without adequate documentation are of little value. The fuller the documentation the better. The minimum data required are: name of recordist; recording speed and duration; locality; date; species name; type of sound; number of animals and sex if known. The following further details are needed if the recordings are to have real scientific value: recording equipment; time of day; weather (especially temperature for insects and amphibians); habitat; associated behaviour and circumstances. This information should be noted down in some form in the field at the time the recording is made. Dictating notes onto the tape immediately after the recording is a particularly good method, or the details can be entered in a field notebook.

Eventually there must be a completed standard data sheet to accompany each tape recording in the archive (a specimen of the form used by BLOWS is shown on page 44.). It is best if these are completed by the recordist. Since writing down all the information is a considerable task, a form on which the recordist can simply tick-off the appropriate pre-printed details has some advantages. Dr James Gulledge has devised such a form for recordings contributed to the Cornell Library of Natural Sounds (see figure 1). This lends itself to quick completion in the field and has the further advantage that it can be used directly for computer input.

Figure 1

The standardisation of species nomenclature is a difficult question, but it is important to be consistent. The scientific name must, of course, be the definitive one, since vernacular names vary widely. One authority should be followed rigidly for each class of animal. A local archive might do best by following the recognised standard work (book or check-list) for the various fauna of its area. Archives which are international in outlook should use a world species list widely recognised as the best authority and used by other such wildlife sound libraries e.g. Morony, Bock and Farrand Reference Lists of Birds of the World; American Museum of Natural History: New York; 1975.

The data sheets, referred to in section 6, filed in systematic order will effectively provide a catalogue of the tape recordings in the archive. For computer storage the data must be susceptible to direct translation for computer input. In practice it is easy to discover from the file of data sheets what tape recordings are held for any species or group of animals and these can be readily located from the reel references on the sheets. However, it is useful for the potential user also to have a printed catalogue. A simple one giving basic details is comparatively easily prepared by a small archive, as has been done, for instance, by the New Zealand Wildlife Service library and the two main Russian Phonoteks (see section I above). The entries in such a catalogue might show: species name, type of sound, place, date, recordist, quality, duration, reference and reel numbers. A typical entry in such a catalogue might be:

ROBIN Erithacus rubecula

1. Song. Yorkshire. 20 October 1980. C. Smith.
   B. 3'40". 3054. S2/4.
2. Calls of pair alarming near nest. Surrey.
   5 May 1960. B. Jones. A. 2'05". 2531. Cl/5

The entries would be listed in systematic order, separately for each class of animal. The need for constant up-dating is an obvious problem.

For commercial records and cassettes, the cards completed should include a list of the species to be found on each published item. In addition a species card index file should be compiled showing, for each species, what records or cassettes its sounds occur on. If these cards also show that there is a related field recording in the archive and that it occurs in a separate special collection (e.g. a broadcasting collection), then all information about recordings of the species concerned is usefully available in the one place.

Without computerisation it is difficult for an archive of any size to locate recordings with special features, such as sub-song or mimicry, or those which come from particular habitats or areas. For further information on computerisation and on cataloguing generally the reader is referred to the separate chapter on this subject.

In an archive, recordings for each species are best held on separate reels, each successive recording (or series of recordings) received being spliced onto the relevant reel. At the same time the reel and 'cut' (i.e. individual recording) number will be entered on the data sheet. The reels should be housed in boxes labelled on the 'spine' with the species name and stored in alphabetic or some other systematic order. Obviously reels for the different classes of animals should be kept separate and BLOWS uses a different colour for labelling each class of animal. A sheet or card giving brief details of each 'cut' should be kept in the box (or they could be written on the box). An announcement dictated onto the tape, giving data sheet number, species name and recordist, should precede each 'cut'. This not only identifies each recording but also facilitates its location if leader tape is not used between recordings.

Normally the archive recordings will be first generation copies of selected portions of original field tapes (which usually contain much that is not worth preserving and, in any case, are often long-play). An additional copy of each batch of recordings received (or the selection made from them) should be made and stored separately. This will serve as a duplicate copy for security purposes.

Ideally recordings will be submitted to the archive as first generation copies on standard play tape at a speed of 19 cm/sec or 38 cm/sec with announcements already dictated onto the tape and with data sheets completed. It is then simply a matter ·of adding those of each separate species to the appropriate reels and making a straight security copy of the whole batch as a 'collection'. In practice, however, recordings are received in many different forms requiring different kinds of processing. The most basic form might be a collection of unedited original recordings with field notes from a recordist who has died. Since some of the material will - by the very nature of wildlife sound recording in the field - be unsuitable for archive use, a selection will have to be copied off onto species reels, with announcements added at the beginning of each 'cut' and data sheets completed from the field notes. Contributions may also be received in a variety of other forms. They may each require different treatment to achieve copies of the required 'cuts' on species reels and a duplicate copy as a collection.

Clearly an archive's staff needs are dependent on many factors, particularly the size and scope of the collection. A suitable complement would be:

(a) Curator:

Responsible for the efficient running of the natural history sound library, the curator should have a zoology degree and experience in bioacoustic research as well as in recording wildlife sounds.

(b) Field Recordist:

Responsible to the curator for carrying out a programme of collecting wildlife sound recordings for the archive, the field worker should be an experienced and competent wildlife sound recordist with good knowledge of the fauna of the areas to be covered. He should also be a reliable identifier of species in the field and have a thorough knowledge of their sound vocabularies.

(c) Technical Assistant:

An assistant would be needed, responsible to the curator, for the processing of acquired recordings and their incorporation in the archive as well as for the preparation of copy recordings for users. He should be a well qualified audio technician with an interest in wildlife sounds.

(d) General Assistant:

Responsible to the curator for documentation, cataloguing, filing, indexing, typing and other routine work, a general assistant should have a good general education, a systematic mind, an interest in wildlife and a knowledge of classification of species.

The main purpose of a natural history sound archive is to make wildlife recordings available for study and use. As well as offering listening facilities on the premises, the service of supplying copies of the tape recordings to bona fide person~ for non-commercial use should be offered. A form of application and agreement protecting copyright needs to be drawn up for all users to complete. Figure 2 shows the form used by BLOWS. All contributors should sign an agreement stipulating the terms under which their recordings are held by the archive and may be copied for users. A typical arrangement is that material can be supplied for non-commercial use, but all applications for commercial purposes (e.g. films and broadcasting) have to be referred to the recordist or copyright holder. A specimen of the contributor's agreement form used by BLOWS is shown in figure 3.

Given the wide range of potential uses that a wildlife sound archive is open to, which are described in section 1, the need to control users and to safeguard the interests of depositors is self-evident.

Figure 2

Figure 3