Advances in Studies of Avian Sound Communication
Total Page:16
File Type:pdf, Size:1020Kb
The Condor96:817-830 0 The CooperOrnithological society 1994 HISTORICAL PERSPECTIVES ADVANCES IN STUDIES OF AVIAN SOUND COMMUNICATION Lurs F. BAFTISTA Department of Ornithology and Mammalogy, California Academy of Sciences, San Francisco,CA 94118 SANDRAL. L. GAUNT Borror Laboratory of Bioacoustics,Department of Zoology, The Ohio State University, Columbus,OH 43221 Abstract. Ground-breaking advances in the study of animal sound communication per se, and avian vocal behavior in particular, awaited instrumentation that enabled capture of sound signals for objective description and analyses. Once these technological advances becamegenerally available, a groundswellof activity in vocal studies,bioacoustics, provided the “raw material” for hypothesistesting pertaining to various biological and evolutionary principles. All aspectsof this field cannot be covered in one review. Thus, themes developed herein deal primarily with questions of vocal acquisition such as factors affecting song learning, the sensitive phase and stimulus filtering hypothesespertaining to the learning process,the effect of vocal behavior on population dynamics, the effect of hormones on the songlearning processand production, photoperiod and song as they are related to gonadal recrudescence,advances in the neurophysiologicalcontrol of song acquisition and produc- tion, and the incidence of female song and function acrosstaxa. Key words: Song; songlearning: vocal acquisitions;vocalizations. All disciplines in biology begin with observation and field biologist, who had to wait until the instrument description, and when the field matures, develop into was declassifiedafter the war for its introduction to the an experimental-hypothesistesting stage(Stresemann mass market. Prior to that time, important develop- 1947). Although the vocalizations of most birds are ments in microphone and parabolic reflector use had yet undescribed, sufficient specieshave been studied been introduced for field biologists by Kellogg, Allen that experimentation and hypothesis testing have ad- and Tanner (Kellogg 1962). Associated with tape-re- vanced. Concomitant with this maturation have been corderswere parallel improvements in magnetic tapes continuing technological innovations which have en- and microphones. Today digital technologyis proving abled scientiststo not only capture,reproduce and thus as revolutionary in bioacousticsstudies as the advent describethese otherwise ephemeral behaviors, but also of the magnetic tape recorder both in digital sound to objectively analyze the sounds that birds produce, recording and computer technologyfor sound synthe- to study the development of these sounds,to hypoth- sis, manipulation and analysis. esize and then designexperiments to test the functions The special requirements of the field biologist for of those sounds, and to interpret these data in terms recording sound under less than ideal conditions are of evolutionary principles. Thus studiesof sound com- seldom the concern of commercial instrument devel- munication behavior, or bioacoustics,can and is con- opers. Thus, advances in conceptual issues in bio- tributing to our growing understanding of many un- acousticshave played leap frog with instrumentation derlying principles of biology. developments for the medical, military and entertain- In this essay, a brief survey of recent advances in ment industries. Insights into animal behavior have instrumentation which have facilitated the analysesof been coupled with advances in technologies not de- bird soundsis presentedfollowed by a review of some veloped for their specificapplication. A prime example of the contributions that avian bioacousticshas made of this mismatch was the introduction and develop- to various fields of biology. ment of sound spectrographic(also termed sonograph- ic) technology that has been the work-horse in bio- INSTRUMENTATION acousticanalysis for over 40 years. Early descriptionsof bird soundswere onomatopoeic Spectrographictechnology was originally designed (Saunders1935), or used musical notations (one of the by the Bell Laboratory prior to World War II to pro- last being Miller 1952). Though methods for capturing duce pictures of human speechthat could be used by sounds existed in the 1800s and early 1900s (phono deaf personsto “read” spoken language(Potter 1945). discs, wire recorders,motion picture film), none were Only one test subject ever developed a rudimentary particularly well suited for field conditions. The in- ability to decipher spoken words from spectrograms novation ofthe magnetictape recorderby the Germans (also termed sonograms).However, the U.S. Navy rec- during World War II was the true breakthroughfor the ognized the utility of the device to detect subtle dif- B171 818 HISTORICAL PERSPECTIVES ferences between individual submarines, and thus ( Wilsonia citrina) and Wood Thrush (Hylocichla mus- spectrographictechnology was classifiedfor the dura- telina) songs often consisted of notes opposite in di- tion of the war. Several technical reports published rection of frequencymodulation and thus must be pro- after the war suggestedthe power of the instrument- duced by the two separatesound sources.Greenewalt especiallyto describesuch sounds as producedby birds (1968) described this two-voice phenomenon in other that were often discreteand poor in overtone structure oscinesand first developed the thesis that each side of (Potter 1945). the syrinx contributed to each “voice.” EuropeanStar- Spectrographicanalysis provided scientistswith an lings (Sturnus vulgaris) may actually imitate the vo- image of the soundsanalyzed (spectrograms)and per- calizationsof two allospecificssimultaneously (Jenkins mitted quantification of sound parameters from that in Baptista 1990a), a most unusualexample of the two- display. These data could then be subjectedto statis- voice phenomenon. There has been much advance- tical treatment (Collias 1991). The first studiesof bird ment of our knowledge of syringealfunction and song songusing the sound spectrographwere published in- production since Greenewalt’s pioneering text, but re- dependently by Borror-and Reese (1953), Kellogg and view of this area of avian communication is beyond Stein (1953). and Collias and Joos (19531, and avian the scopeofthe presentwork. We refer thoseinterested bioacdustics’passed from anecdotal to objective de- to the reviews of Gaunt and Gaunt (1985) and Gaunt scription. (1987). Use of the sound spectrographmachine was limited to two or four seconds of analyses at a time. This REPERTOIRE SIZE sufficedwhen usedto analyze soundsof short duration. Hartshome (1956; seealso Kroodsma and Vemer 1978) Analyses of long continuous songssuch as those pro- describedthe manner of songdelivery in birds as “con- duced by Mockingbirds (Mimus polyglottos)and Ca- tinuous” versus “discontinuous.” Continuous singers tharusthrushes (Howard 1974, Raitt and Hardy 1970) are thosebirds, suchas mockingbirdsor skylarks(Alau- were formidable tasks indeed, until the development da arvensis),that sing long songswith short intervals ofthe continuousspectral analyzer (Hopkins et al. 1974). of silencebetween bursts of sound, whereasdiscontin- This technology has evolved to digital systemswhich uous singers are those with short songs (two to four allow for real-time spectrographicdisplay and digital seconds long) divided by longer intervals of silence. analysis such as the dedicated DSP Sona-graph ma- The intervals of silence labeled as “cadence” by Rey- chine, Kay Elemetrics, or computer programsfor per- nard (1963) may be speciesspecific. Within a species, sonal computers such as SIGNAL from Engineering cadence may be used to advertise individual fitness Design and CANARY from the Cornell Laboratory of (Payne and Payne 1977). Ornithology. These advancesin computer technology Thus repertoire size in bird song may be described have enabled investigatorsto analyze sound and sub- in terms of syllabletypes or number of themes or both. ject massesof data to statistical treatment in a timely Speciesmay differ considerablyin repertoire size even manner. Cross-correlation programs permit investi- within a family. For example, among emberizid finches gators to compare syllables uttered by different indi- we find a continuum between the Chipping Sparrow viduals, populations and widely separatedgeographic (Spizellapasserina) at one extreme who singsonly one areasin an objective manner (Gaunt et al. 1994). Com- songtype to the Five-striped Sparrow (Aimophila quin- puters also enable bioacousticiansto manipulate cap- quefasciata)who may singover 200 songtypes (Marler tured natural sounds or synthesize sounds for use in and Isaac 1960, Groschupfand Mills 1982). In between tutoring (Marler and Peters 1977, Konishi 1985) and are suchspecies as the Black-chinnedSparrow (SpizeNa playback. atrogularis) who sings two song types (Baptista, un- published),or Dark-eyed Juncos(Junco hyemalis) who DESCRIPTIVE STUDIES sing three to seven or Song Sparrows (Melospiza me- lo&a) who sing 10 or 20 (Williams and MacRoberts THE INDIVIDUAL SONG 1977, Mulligan 1966). Irwin (1988) has presented a Speciesspecificity in bird sounds may be encoded in thesis that in some groups,including emberizids, song duration or rhythm (seconds),in frequency (kHz), or complexity has been reducedduring the courseof evo- in tonal quality (Becker 1982). Tonal quality may be lution. coded in the harmonic structure of the sound (Davis 1964, Mailer 1969) or by singing two unrelated notes SONG SHARING or series of notes simultaneously