XXV International B Ioacoustics C Ongress Murnau | B Avaria, Germany

XXV International Bioacoustics Congress Murnau | Bavaria, Germany XXV International Bioacoustics Congress Murnau | Bavaria, Germany

7 – 12 September 2015

Support 5

Important Information 7

Program Schedule 9

Overview 15

Plenary Presentations 17

Symposium Presentations 21

Poster List 57

Poster Presentations 74

Registered Participants 153 host scientific committee

International Bioacoustics Council Manfred Gahr (IBAC) Max Planck Institute for Ornithology, Germany

Henrik Brumm organising committee Max Planck Institute for Ornithology, Germany

Ole N. Larsen Manfred Gahr University of Southern Denmark Max Planck Institute for Ornithology, Germany Albertine Leitao Nicole Hoiss Max Planck Institute for Ornithology, Germany Max Planck Institute for Ornithology, Germany Fanny Rybak Antje Bakker Université de Paris Sud Max Planck Institute for Ornithology, Germany Andries ter Maat Carolina Frankl-Vilches Max Planck Institute for Ornithology, Germany Max Planck Institute for Ornithology, Germany Michiel Vellema Albertine Leitao University of Southern Denmark Max Planck Institute for Ornithology, Germany

Andries ter Maat support Max Planck Institute for Ornithology, Germany

Lisa Trost Richard Ranft Max Planck Institute for Ornithology, Germany IASA, United Kingdom

Diana Werner Michelle Rasch Max Planck Institute for Ornithology, Germany michelle und josef design, Germany

5 support

support additional support

We gratefully acknowledge the financial support Special thanks go to Richard Ranft – IASA, United of Dodotronic, Avisoft Bioacoustics, Noldus and Kingdom for his unfailing and uncomplaining Metris B.V. support regarding buildup and maintenance of the IBAC Website, to Michelle Rasch regarding the conference layout; to Swarovski Optik, who generously sponsored the Best Poster Award, the binocular SWAROVSKI OPTIK SLC 42 (SLC 10x42).

„These binoculars can go everywhere with you. You can take the binoculars with you on your extended observation trips in the surrounding area, both during the day and at twilight. The SLC 10x42 binoculars are sturdy in any weather, day in day out, and also stand out thanks to their particularly high optical quality. You will enjoy these rugged, ergonomic binoculars for many years.“

Last, but not least to Taylor & Francis Group who are enriching the conference with a special price, a 1-YEAR-SUBSCRIPTION FOR THE JOURNAL BIOACOUSTICS, for the Best Oral Presenation of Junior Participants.

Thank you very much! ■

6 Welcome to the XXV International Congress of the Bioacoustics Council. Our conference venue is oral presentations the Culture and Convention Centre / Kultur- und Tagungszentrum Murnau, KTM. Oral presentations (talks including discussion) are divided into three groups: registration Plenary talks: 60 min Symposium talks: 20 minutes The registration desk at the Room “Münter” is open Open Oral Session talks: 15 minutes the following days: To upload and test your talk, please come to the Sunday, September 06th 17.00 – 19.00 conference office the day before your talk is sche- Monday, September 07th 08.00 – 09.30 duled. PLEASE MAKE SURE YOUR TALK RUNS ON A WINDOWS 7 ENVIRONMENT. Your presentation Please follow the signs. will be deleted after your talk.

Upon registration please confirm your registration Location: LECTURE HALL 1 AND 2 for one of the excursions, for the Social Dinner and for the visit of Max Planck Institute for Ornithology in Seewiesen. poster presentation

best poster award Poster presentations are divided into two parts á two days – Monday/Tuesday and Thursday/Friday. Posters will be changed on Wednesday. The Best Poster Award is open for every poster presentation and will be awarded by the Jury. The Posters must be A0 size maximum, in portrait Poster Award Ceremony, to which everybody is invi- orientation (vertical), so 84.1 cm wide x 118.9 cm ted to join, will take place on Thursday, September high. Mounting material will be provided. 10th, after the Social Dinner. During the “Open Bar”, poster session continues.

best oral presentation for Please see poster list for scheduling (also available online) junior participants Location: ROOM ‘STAFFELSEE’ AND ROOM 3

This award is open for every oral presentation of junior participants. The winner will be selected by excursion all participants with the help of a voting sheet. The Award Ceremony will take place on Friday, September 11th, during the Closing Remarks of Ole N. Larsen. Attention: due to the large number of regis- trations for the ´Murnauer Moos hike’ (more than 120), we offer a third exciting possibility to spend exhibitors the free afternoon. We booked about 50 places on a round-trip boat on the ´Staffelsee´, the marvelous lake close to Noldus | The Netherlands Murnau. Eighty minutes of sightseeing with a wa- Metris B.V. | The Netherlands terfall, wonderful nature and a nice place to sit are Location: Foyer of the conference venue offered to all who join in the boat trip.

7 xxv international bioacoimportantustics information congress

Please indicate upon registration, whether you are willing to change. useful local information

Upon registration you´ll get more information about the excursion of your choice, the “Beer Seminar”, national emergency numbers  the “Murnauer Moos” or the “Staffelsee”. Tel. 110 / 112 Fire department, police, emergency care

visit of max planck institute hospital BGU murnau Prof.-Küntscher-Straße 8 for ornithology (MPIO) 82418 Murnau a. Staffelsee Tel. 08841 48 0

Two busses will bring you the Max Planck Institute Hospital Garmisch-Partenkirchen for Ornithology, Seewiesen, on Saturday. Departure Auenstraße 6 from the congress center (entrance hall) at 09.00, 82567 Garmisch-Partenkirchen arrival at MPIO around 10.00. Tel. 08821 77 0 We will organize guided tours for a duration of approx. 2.5 hours. After a lunchbreak at the mo- police department murnau nastery of Andechs the busses will bring you back Barbarastraße 9 to Murnau around 15.00. 82418 Murnau a. Staffelsee Tel. 08841 61 76 0 Upon registration you´ll get more detailed information. roadside assistance (ADAC) Tel. 0180 22 22 22 internet access tourist information murnau Kohlgruber Straße 1 Free WLAN Access is available at the conference 82418 Murnau a. Staffelsee center. There´s no code required. Te. 08841 61 41 0 www.murnau.de

food rent-a-car St.-Martin-Straße 17 82418 Murnau a. Staffelsee During the coffee breaks coffee, tea and water will Tel. 08821 93 42 42 be provided. You´ll have the opportunity to get more snacks and drinks at the Restaurant “Auszeit”, next taxi service murnau to the conference venue. Taxi Cavallino A range of restaurants is available within walking Tel. 08841 12 17 distance of the conference center. Stockbauer Taxi Tel. 08841 91 30

Train Station Murnau Bahnhofsplatz 1 82418 Murnau a. Staffelsee All conference staff can be recognized by their blue t-shirts and badges. If you Bus Station have any questions, anyone of us will be Kohlgruber Straße 1 happy to help. ■ 82418 Murnau a. Staffelsee

8 in the monkey brain sunday | Sept 6th, 2015 13.00 – 14.30 Lunch

17.00 – 19.00 Registration 14.30 – 16.30 Symposium 2 Communication over various spatial scales Monday | Sept 7th, 2015 Fanny Rybak Thierry Aubin

08.00 – 09.30 Registration 14.30 Rohini Balakrishnan Determinants and consequences of 09.30 – 10.00 Welcome/Opening acoustic baffling behaviour in tree Manfred Gahr crickets

10.00 – 11.00 Plenary Talk 14.50 Paulo Fonseca Mike Ryan Acoustic active space in the What you hear ain’t what you get Lusitanian toadfish and the effect of anthropogenic noise 11.00 – 13.00 Symposium 1 Control of vocal production 15.10 Max Ringler Coen Elemans Acoustic ranging in Poison Frogs. Sue Anne Zollinger Allobates femoralis uses signal cues beyond intensity to assess 11.00 Friedrich Ladich caller distance Sound-producing mechanisms and neuronal control in fishes: A 15.30 Ole N. Larsen unique diversity in vertebrates Call transmission advantage for parrots at dawn in tropical dry 11.20 Marcos Gridi-Papp forest Laryngeal muscles and vocal control in amphibians and 15.50 Clément Cornec reptiles Booming far: the long-range vocal communication of a lekking bird 11.40 Coen Elemans Sue Anne Zollinger 16.10 Solveig C. Mouterde Universal mechanisms of sound Active space of individual production and control in birds signatures in propagated calls of and mammals zebra finches: From acoustics to perception 12.00 Constance Scharff Molecular toolkits for control of 16.30 – 17.00 Coffee Break vocal production 17.00 – 19.00 Poster Session 1 12.20 Christian T. Herbst (Poster 1 – 80) Monitoring the mammalian and avian sound source with 19.00 – 20.00 Open Bar electroglottography

12.40 Steffen Hage Vocal motor control mechanisms

9 Program schedule

13.30 – 15.30 Symposium 4 tuesday | Sept 8th, 2015 Ecoacoustics Jerôme Sueur

08.30 – 09.30 Plenary Talk 13.30 Pieretti Nadia Daniel Robert Application of a recently intro- Sound frequency analysis through duced acoustic index in marine cochlea-like mechanism in an soundscapes of a shallow water insect coast (Lampedusa Island, Italy)

09.30 – 10.00 Coffee Break 13.50 Susan Parks Ecological insights form bioacou- 10.00 – 12.00 Symposium 3 stics activity levels measured from Acoustic communication long-term acoustic monitoring at of arthropods NEON observatory sites Tony Robillard 14.10 Gianni Pavan 10.00 Stefan Schoeneich SABIOD Italy: A pilot study of wild Neurobiology of acoustic soundscapes. The RNI Sassofratino communication: Neural networks pilot study for singing and song pattern recognition in crickets 14.30 Diego Llusia Heterospecific songs influence 10.20 Camille Desjonquères dawn chorusing in forest bird Acoustic diversity in freshwater communities habitats: the effect of lateral connectivity in a riverine floodplain 14.50 Stowell Dan Classifying multiple bird species 10.40 Ranjana Jaiswara in crowdsourced smartphone audio Species boundary delineation using multiple approaches: A case Frédéric Bertucci study on a field cricket genus 15.10 Large scale acoustic survey as Teleogryllus in India a new tool for the evaluation of coral reef biodiversity (Moorea, 11.00 Fabio Sarria French Polynesia) Non-invasive measurement of travelling waves in the inner ear 15.30 – 16.00 Coffee Break of neotropical bushcrickets 11.20 16.00 – 18.00 Open Oral Session 1 Rafael Rodriguez Henrik Brumm Environmentally-induced runaways: vibrational communi- 16.00 Antonieta Labra Lillo cation in Enchenopa treehoppers Interaction between calls and (Hemiptera membracidae) scents in lizard antipredator 11.40 behaviour Arne K. D. Schmidt No evidence of acoustic signal 16.15 Karla Rivera Caceres divergence in a tropical cricket Duet code development in plain assemblage wrens (Cantorchillus modestus zeledoni) 12.00 – 13.30 Lunch

10 16.30 Marie Roch 10.20 W. Tecumseh Fitch Management of acoustic metadata Songbirds without song? Corvid for bioacoustics communication from a comparative perspective 16.45 Thierry Aubin Should I stay or should I go? A 10.40 Katie E. Slocombe slight acoustic variation gives an Flexibility in chimpanzee vocal opposite meaning to a songbird communication alarm call 11.00 Thorsten S. J. Balsby 17.00 Daisuke Mizuguchi Communication and social Underwater vocal exchange bet- complexity in orange fronted ween male and female in captive conures (Aratinga canicularis) bearded seals 11.20 Julie Oswald 17.15 Sandra Goutte Geographic variation in delphinid Acoustic communication of the whistles – implications for earless toad (Brachycephalus classifier development pitanga) 11.40 All speakers 17.30 D‘Amelio Pietro General discussion: Cognition and Motor control and recognition of communication – can a comparati- unlearned vocalisations ve approach across taxa work?

Andrèa Thiebault 12.00 – 12.30 General Assembly Meeting Honking in traffic: use of calls in a 17.45 foraging seabird 12.30 – 13.30 Lunch Change of poster 18.00 Poster Session 2 (Poster 1 – 80) 13.30 Excursion

wednesday | Sept 9th, 2015 thursday | Sept 10th, 2015

08.30 – 09.30 Plenary Talk 08.30 – 09.30 Plenary Talk Julia Fischer Vincent Janik Referential communication How does vocal learning affect revisited: the case of vervet marine mammal communication? monkey alarm calls 09.30 – 10.00 Coffee Break 09.30 – 10.00 Coffee Break 10.00 – 12.00 Symposium 6 10.00 – 12.00 Symposium 5 Mechanisms of vocal learning Vocal Communication and Manfred Gahr Cognition Auguste von Bayern 10.00 Robert Lachlan Evolution and function of vocal 10.00 Elodie Briefer learning in the chaffinch Emotional contagion through (Fringilla coelebs) whinnies in horses

11 Program schedule

10.20 Veronika Beeck 14.50 Ian Agranat Call Convergence in socially affi- Unravelling zero crossing and full liated African savannah elephants spectrum, what does it all mean? (Loxodonta africana) 15.10 Ladislav Ptacek 10.40 Nicole Geberzahn Application of speaker recognition Towards an understanding of methods for chiff-chaff individual developmental and perceptual identification mechanisms of vocal individuality in the zebra finch 15.30 – 16.00 Coffee Break

11.00 Jinhong Luo 16.00 – 17.30 Open Oral Session 2 The Lombard effect emerges in Albertine Leitão bat infants: implications for the development of auditory feedback 16.00 Dylan Gomes Foraging decisions in clutter and 11.20 Alejandro Salinas-Melgoza noise: the effect of environmental The military macaw maintains complexity on multimodal signal homogeneity of a contact call use in a frog-eating bat across populations over a modified landscape in western Mexico 16.15 Nao Ota Tap dancing birds: the multimodal 11.40 Andries ter Maat mutual courtship display of Vocal learning in social groups of males and females in a socially zebra finches monogamous songbird

12.00 – 13.00 Lunch 16.30 Ivano Pelicella WaveFalcon, a new smart recorder 13.30 – 15.30 Symposium 7 Techniques in bioacoustics 16.45 Dinesh Bhatt Andries ter Maat Onset of dawn chorus in a tropical songbird, the Pied Bush Chat 13.30 Sarah Hallerberg (Saxicola caprata): relationships Bag-of-calls analysis reveals with weather and Sunrise group-specific vocal repertoire in long-finned pilot whales 17.00 Tony Robillard Evolution of mechanisms of sound 13.50 Philip Wadewitz production in Agnotecous crickets: Advances in the quantitative ana- multiple strategies to produce lysis of graded vocal repertoires highfrequencies

14.10 Marie SA Fernandez 17.15 Juliette Linossier Impact of visual contact on vocal Song and genetic divergence rela- interaction dynamics of pair ted to different migratory patterns bonded birds in a songbird, the blackcap (Sylvia atricapilla) 14.30 Karen Rowe Automated recognition software 17.30 – 19.30 Poster Session 3 improves detectability for a range (Poster 81 – 160) of bird species’ vocalizations 19.30 Social Dinner/ Poster Award 

12 13.30 – 15.30 Symposium 9 friday | Sept 11th, 2015 Music, speech and language Carel ten Cate

08.30 – 09.30 Plenary Talk 13.30 Bruno Gingras, Emily Doolittle Peter Slater A music theoretical approach to Studies of the solos, duets and birdsong choruses of “Thryothorus” wrens 13.50 Andrea Ravignani 09.30 – 10.00 Coffee Break The evolution of rhythm: a comparative approach to speech 10.00 – 12.00 Symposium 8 and music Vocalizing females Michelle Hall 14.10 Marisa Hoeschele Using bioacoustics to compare 10.00 Karan J. Odom avian and human acoustic Similarities and differences in perception female and male song in troupials (Icterus icterus): structural, diel 14.30 Buddhamas Kriengwatana and seasonal comparisons A general auditory bias for disregarding inter-speaker speech 10.20 Jacqueline R. Miller variability? Evidence in humans Sexual dimorphism in vocal beha- and songbirds viour: exploring the pheromyscine mice of North and Central America 14.50 Kurt Hammerschmidt Structural complexity can be a 10.40 Lauryn Benedict misleading indicator for advanced Female song in new world communication: the case of mouse wood-warblers courtship songs

11.00 Irina Beme 15.10 Carel ten Cate Is female song ancestral in Linguistic rule learning in birds songbirds? 15.30 – 16.00 Coffee Break 11.20 Johana Goyes Vallejos Female calling behaviour in 16.00 – 17.30 Open Oral Session 3 the Smooth Guardian Frog Michiel Vellema (Limnonectes palavanensis) of Borneo. Evidence for sexrole 16.00 Lucyna Wojas reversal? Social eavesdropping on acoustic cues in non-learning model bird 11.40 Clementine Vignal species Vocal negotiation over parental care? Partners adjust their time 16.15 Clara P. Amorim spent incubating on their acoustic Multimodal and mismatching exchanges at the nest communication in vocal fish

12.00 – 13.30 Lunch 16.30 James Campbell Sound field characteristics of small tanks in the context of noise exposure experiments

13 Program schedule

16.45 Sonja Vernes 17.00 – 19.00 Poster Session 4 A novel approach reveals first (Poster 81 – 160) molecular networks in the bat 19.00 brain: implications for vocal Closing Remarks communication Ole N. Larsen

17.00 Pauline Salvin Open Bar Functions of female signals: the copulation solicitation and the mate sampling aid hypotheses in saturday | Sept 12th, 2015 the domestic canary

17.15 Vera Klimšová 09.00 Visit to the Max Planck Institute Impalas recognize predator for Ornithology, Seewiesen vocalizations and heterospecific alarms of sympatric and allopatric species

BUY NOW Official IBAC GERMANY mugs

08.00 – 08.30

Sound frequency analysis Referential communication 08.30 – 09.00 Registration How does vocal learning affect Studies of the solos, duets an through cochlea-like mechanism revisited: the case of vervet marine mammal communication? choruses ob „Thryothorus“ wrens in an insect monkey alarm calls Vincent Janik Peter Slater 09.00 – 09.30 Daniel Robert Julia Fischer

09.30 – 10.00 Manfred Gahr Welcome | Opening Coffee Break Coffee Break Coffee Break Coffee Break

10.00 – 10.30 What you hear ain‘t what you get Mike Ryan Symposium Symposium Symposium Smyposium 10.30 – 11.00 3 5 6 8 Acoustic communication Vocal communications Mechanisms of vocal learning Vocalizing females of arthropods and cognition 11.00 – 11.30 Manfred Gahr Michelle Hall Tony Robillard Auguste von Bayern Symposium 11.30 – 12.00 1 Control of vocal production 12.00 – 12.30 Coen Elemans Gen. Ass. Meeting Sue Anne Zollinger Visit MPIO 12.30 – 13.00 Lunch Lunch Lunch Lunch 13.00 – 13.30

13.30 – 14.00 Lunch

14.00 – 14.30 Symposium Symposium Symposium 4 7 9 Ecoacoustics Techniques in bioacoustics Music, speech and language 14.30 – 15.00 Jerôme Sueur Andries ter Maat Carel ten Cate Symposium 15.00 – 15.30 2 Communication over various spatial scales 15.30 – 16.00 Fanny Rybak Coffee Break Coffee Break Coffee Break Thierry Aubin 16.00 – 16.30 Open Oral Session Open Oral Session 16.30 – 17.00 Coffee Break Open Oral Session Excursion 2 3 1 Albertine Leitão Michiel Vellema 17.00 – 17.30 Henrik Brumm

17.30 – 18.00 Poster Session Registration 1 – 80 Poster Session 18.00 – 18.30 Poster Session 81 – 160 81 – 160 18.30 – 19.00 Poster Session 1 – 80 19.00 – 19.30 Ole N. Larsen Closing remarks Bar 19.30 – 20.00 Social Dinner | Award Ceremony Bar

15 xxv international bioacoustics congrovervieessw

08.00 – 08.30

09.30 – 10.00 Manfred Gahr Welcome | Opening Coffee Break Coffee Break Coffee Break Coffee Break

13.30 – 14.00 Lunch

16 ABSTRACTS of PLENARY SPEAKERS (A – Z) Location: Lecture Hall 1/2

Wednesday | Sept 9th 08.30 – 09.30 Thursday | Sept 10th 08.30 – 09.30

Fischer, Julia Janik, Vincent M. Cognitive Ethology Laboratory, German Primate University of St Andrews, United Kingdom Center, Göttingen How does vocal learning affect marine mammal Referential communication revisited: the case communication? of vervet monkey alarm calls Abstract The most obvious cases of vocal pro- Abstract Elucidating the origins of speech has duction learning involve the copying of novel calls been a major driver in studies of animal vocal or songs that are very different from the learner’s communication. I will here discuss three aspects usual repertoire. Species such as bottlenose dol- that have been deemed central in this regard, phins and harbour seals can copy such signals, but namely semanticity, vocal learning, and intentio- little is known about the constraints in their learning nality. The classic example of semantic communi- skills and several other species use experience to cation are the alarm calls of vervet monkeys, who modify their calls in a comparatively subtle way. We give three distinct alarm calls in response to their have developed a new training paradigm that allows three main predator classes, and the calls alone us to quantify animal vocal learning capacity and are sufficient to elicit different escape strategies. compare them across species. By training animals The calls were thus deemed as being “functionally to copy sounds from their own repertoire that we referential”, and assumed to provide important then alter in a step-wise procedure we can test the insights into the evolution of speech. We used degree of modification the animals can implement, comprehensive quantitative analysis to re-assess up to sounds that are not part of their usual re- the structural variation within their vocalizations. pertoire. Grey seals trained in this way showed an Although the three different alarm calls indeed interesting tendency to copy formant modulations differ significantly, we found substantial overlap but not modulations of the fundamental frequency. in acoustic structure with calls produced in First results from bottlenose dolphins indicate that non-alarm contexts, indicating that these alarm they learn to copy novel sounds very quickly using calls are indexical rather than symbolic. Playback this paradigm. We hope that characterising vocal experiments on Green monkeys, the West African learning skills in this way will ultimately help to congeners of vervets further show that responses reconstruct the context in which vocal learning are guided by both the acoustic information evolved. available and contextual cues. Analyses of the Many seals and baleen whales produce song and variation of alarm calls in different Chlorocebus may use learning when acquiring new song types. populations as well as several studies that Grey seals, however, are not known to sing, which explored the degree of vocal production learning makes their learning skills harder to interpret. In underscore the view that the structure of non- one of our playback studies, wild grey seal pups human primate vocalizations is largely innate, copied sound combinations in their environment while there is some flexibility in call usage. I will when producing contact calls. Similarly, bottle- conclude with a brief discussion of differences nose dolphins used vocal learning for individual in the intentional structure of human and non- recognition. In laboratory and field studies, we human primate communication and provide an found that individually specific signature whistles outlook future questions that I deem promising. were used for individual recognition, negotiating

17 abstracts of plenary speakers

reunions and, when copied by other dolphins, tes that two remote organisms at the phylogenetic for addressing conspecifics. In developmental level, katydids and mammals, have evolved a series studies we have shown that signature whistles of convergent solutions to common biophysical are formed early in life and that their modulation problems, despite their reliance on very different pattern does not relate to sex, age or size of the morphological substrates. animal. Thus, new signature whistles types appear to be truly arbitrary and introduced de novo into the communication repertoire of the group. While Monday | Sept 7th 10.00 – 11.00 we still know little about the mechanism used to develop novel modulation patterns, first studies suggest that the infants use model whistles in their Ryan, Michael environment that they then alter to create a novel University of Texas, USA & Smithsonian Tropical signal. Comparing pinnipeds and cetaceans overall, Research Institute, Panama the most likely contexts for the evolution of vocal learning in marine mammals are sexual selection What you hear ain’t what you get and individual or group recognition. The degree of flexibility in learning skills seems to differ between Abstract An underlying assumption of neuroe- species and may be linked to ranging patterns and thology is that our sensory, perceptual, and cogniti- social structure. ve systems have evolved under selection to give us an accurate representation of the world around us. This need not be the case, as the brain evolves to Tuesday | Sept 8th 08.30 – 09.30 perform many tasks and its function cannot always be optimized for all domains. These issue are especially relevant to sexual selection by mate choice Robert, Daniel in which it is often assumed that all mate choice University of Bristol, United Kindom evolved under selection to choose males with “good Montealegre-Z, F. genes” (heritable variation for survivorship). We Sarria, F. know, however, that various sensory, perceptual, University of Lincoln, United Kingdom and cognitive biases also influence what “sounds Jonsson, B. T. good”. I will review how peripheral auditory tuning, University of Bristol and Lincoln, United Kingdom Weber’s Law, cocktail part effects, nonlinear interactions among auditory and visual modalities, Sound frequency analysis through cochlea-like and irrational choice all bias what choosers find mechanism in an insect attractive in the acoustic signals of their mates.

Abstract In mammals, hearing involves three canonical stages for the processing of impinging Friday | Sept 11th 08.30 – 09.30 sound waves: 1. an eardrum collecting sound, 2. a middle ear impedance converter, and 2. a cochlear frequency analyzer. Some insects are endowed with Slater, Peter J. B. an equivalent functional biophysical arrangement School of Biology, University of St Andrews, for auditory processing. In rainforest katydids, United Kingdom amongst some of the smallest ears know from animals, we could show that ears perform the crucial Studies of the solos, duets and choruses of stage of air-to-liquid impedance conversion and „Thryothorus“ wrens signal amplification. This particular process that of energy coupling is enabled through a distinct Abstract In this talk I shall review the work tympanal lever system. Further along the chain of that my collaborators and I have carried out on hearing, the important process of spectral analysis a group of Central American wrens, all formerly is achieved through dispersive wave propagation placed in the genus Thryothorus. Interest in the across a fluid substrate, similarly to what happens phenomenon of duetting led us to study the plain in the mammalian cochlea. This evidence illustra- wren (Thryothorus , now Cantorchilus, modestus) in 18 Costa Rica, and we then broadened our perspective intruders by appearing to be more numerous than to study some 23 other species demonstrating a they are. In the stripe-breasted wren (Cantorchilus variety of singing styles. We combined the study thoracicus), which is unusual in having two quite of their songs with phylogenetic analysis, and this different song forms, we have examined the diffe- led us to the conclusion that these species should rence in function between these. One, the “hoot” be more properly placed in four separate genera solo song, is especially common in unpaired males, (Cantorchilus, Pheugopedius, Thryophilus and and seems to be an advertising signal. The other, Thryothorus). These new groupings differ in their the “complex” song, with which a male’s mate most usual singing style, while within a group most sometimes joins to form duets, is mainly used in species sang in a similar fashion. However, certain interactions with other birds, and thus the most species sing differently from their closest relatives, frequent in response to playback. In the happy wren leading to questions as to why evolution has favou- (Pheugopedius felix) our results suggest that duets red such divergence. are used in cooperative territory defence rather Following on from this broad comparative work, we than mate guarding, and we have gone on to look have concentrated in particular on a few of these at how the two members of a pair integrate their species. Perhaps the most striking is the plain- contributions. The concentration of much recent tailed wren (Pheugopedius euophrys) from Ecuador, research into duetting has been on function, yet which lives in groups that sing highly stereotyped the remarkable speed and synchrony of the duets in and coordinated four-part choruses. Playback of species such as these wrens raises many questions simulated songs of groups of four and of two shows about the mechanisms involved. We have begun to that these are responded to differently in line look at these in this species, but much remains to with the numerical odds hypothesis. Furthermore, be done. The issue of development is also almost the initial response of small groups to larger ones untouched on when it comes to duetting and raises is to sing a much longer song more typical of a some particularly interesting, and challenging, larger group, suggesting that they attempt to repel issues. ■

             

BEST ORAL PRESENTATION

FOR JUNIOR PARTICIPANTS

CEREMONY ON FRIDAY, Sept 11th, 2015

during the “Closing Remarks” of Ole.N. Larsen

AWARD: 1-YEAR-SUBSCRIPTION OF JOURNAL “BIOACOUSTICS”

             

ABSTRACTS of SYMPOSIUM SPEAKERS (A – Z of first author) Presenting author is underlined

Location: Lecture Hall 1/2

Thursday | Sept 10th 14.50 – 15.10 Tuesday | Sept 8th 16.45 – 17.00

Agranat, Ian Aubin, Thierry Wildlife Acoustics, Inc., U.S.A. Université Paris Sud, France Mathevon, N. Unravelling zero crossing and full spectrum, Université Jean Monnet, St. Etienne, France what does it all mean? Magrath, R.D. Australian National University, Canberra, Australia Abstract Full spectrum and zero-crossing recording technology has been used to record and Should I stay or should I go? A slight acoustic analyze the echolocation calls of bats for decades, variation gives an opposite meaning to a and more recent advances in analysis software songbird alarm call combine these technologies by extracting zero crossing information from full spectrum recordings Abstract Predation is a powerful force shaping using different combinations of signal processing alarm acoustic signals. In many birds, the flee and techniques. We explain the physics behind full the mobbing alarm calls are both used to announce spectrum and zero crossing technologies and a predator but elicit opposite responses from recei- modern hybrid algorithms for bat biologists and vers, respectively a quick fly to cover or an approach ecologists to better understand and appreciate the and an harassment of the predator. Although pre- advantages, disadvantages, and modern capabili- vious studies have reported structural differences ties in available technology. We first look at zero between both signals in bird species, there is a lack crossing and full spectrum recording technologies, of investigations aiming to understand the acoustic how they work, and their relative advantages and basis of the switch between both functional types disadvantages in a number of areas. We then of alarm signals. In the superb fairy-wren Malurus explore a simple technique for extracting zero cyaneus, flee and mobbing calls present strong crossing data from full spectrum recordings using similitudes. They are both characterized by the band-pass filtering and adaptive thresholds. Later same high-pitched carrier frequency on which is we look at more advanced signal processing tech- applied a broad and fast frequency modulation, niques including Gaussian noise reduction, echo with the mobbing call presenting a low-pitched cancellation, call tracing and adaptive filtering, and short introductory element. By playback expe- and how these techniques can be used to enhance riments with synthetic calls, we examined which a full spectrum signal in order to extract richer structural details drive the functional difference zero crossing data. This approach also shows pro- between mobbing and flee calls. We showed that mise for analysis of birdsong as well. We conclude the functional switch between both calls is due to a that while zero crossing recordings do have some simple variation of the slope of the carrier frequen- advantages in limited circumstances, it is far cy, and does not depend on the other parameters better to record in full spectrum to record bats composing both signals. Our result highlights the and use modern signal processing techniques to importance of the carrier slope in determining the enhance the signal before either analyzing using meaning of alarm calls and supports the idea of full spectrum tools or extracting zero crossing a broadly shared coding rule among birds. Due information and analyzing using zero crossing or to their importance for survival, alarm calls have hybrid tools. evolved under strong constraints. We hypothesize

21 abstracts of symposium speakers

that, in the evolutionary history of alarm calls, Communication and social complexity in both signals have a common ancestor but that the orange fronted conures mobbing call has diverged from a pre-existent flee (Aratinga canicularis) call possibly in the context of social cooperation. Abstract During the non-breeding season the orange fronted conure lives in small flock of 2-20 Monday | Sept 7th 14.30 – 14.50 individuals, which are characterized by daily fission- fusion flock dynamics. The local population that we work with in Santa Rosa National park, Costa Rica Balakrishnan, Rohini has at least 1500 individuals based on capture Deb, R. recapture estimates. Such social complexity may Indian Institute of Science, India pose substantial challenges to the communication system. Orange-fronted conures exchange mul- Determinants and consequences of acoustic tiple contacts calls prior to decisions about flock baffling behaviour in tree crickets fusion. During these exchanges the contact calls changes and the contact calls of the interactants Abstract Male tree crickets make acoustic baff- may increase or decrease in similarity. Playback les in the form of a hole chewed into a leaf from experiments have shown that these changes have which they sing. These leaf baffles act as amplifiers signal function and that increases in similarity bet- that increase the intensity of their calling songs, ween interactants can be used to address specific which they use to attract conspecific females. The individuals within the communication network. ecological context, fitness benefits and selection The fission-fusion flock dynamics enables selective pressures that could have led to the evolution of recruitment of individuals. If individuals have the baffling behaviour are unknown and have not been cognitive abilities the decisions on who to recruit investigated so far. In the tree cricket Oecanthus may involve prior interactions with specific indivi- henryi, only a small proportion of males were duals. One experiment suggests that orange-fronted found to sing from baffles in the field although conures, in addition to discriminate between known most males were capable of building baffles. We and unknown individuals, among the known indivi- examined the factors that influence baffling beha- duals adjust their response according to how well viour and found that the decision to baffle involves the playback imitated their contact call in previous a complex interplay between available leaf size, interactions. male body size and a male’s inherent call intensity. We then examined the advantages of baffling for long-distance acoustic mate attraction in an eco- Thursday | Sept 10th 10.20 – 10.40 logically and spatially realistic context by building three-dimensional acoustic simulation models that took into account male spacing, female positions Beeck, Veronika and hearing threshold, sound transmission losses Stoeger, A. S. and sound fields of males in choruses. Baffling did University of Vienna, Austria increase numbers of females attracted but also influenced the subsequent mating behaviour of Call convergence in socially affiliated african females in ways that should increase the fitness of savanna elephants (Loxodonta africana) bafflers. The implications of these findings for the evolution of baffling behaviour will be discussed. Abstract Among just a few mammalian species, including humans, elephants have evolved advanced vocal learning skills. Since these reports Wednesday | Sept 9th 11.00 – 11.20 stem from captive individuals in rather unnatural social environments, the function of elephant vocal learning within their natural communication Balsby, Thorsten J.S. system is unknown. Vocal learning may be involved Aarhus University, Denmark in the convergence of vocalizations among socially

22 affiliated individuals when animals are housed 1998; Slater & Mann 2004; Morton, 1996). In together in captivity or naturally form social groups. 2014 Odom et al (Odom et al., 2014) have publis- In the African elephant (Loxodonta africana) the hed an article, in which the authors based on the family unit is centered around predominately, but analysis of the 1141 species claiming that female not exclusively, closely related females and their song is widespread and ancestral in passerine. offspring. It builds the stable core of a complex ma- Hear we analyzed the cases of female singing in trilineal fission-fusion social system which is based 1309 species of different passerine families (18 on mutual cooperation and maintained by acoustic complete families, two subfamilies in Thurdidae communication. Captive groups, in contrast to and three subfamilies in Sylviidae in Oscinae and 4 groups in the natural system, are often composed complete families and two subfamilies in Tiraniidae of unrelated females of different heritages. During in Suboscinae), in order to understand the female my master thesis I studied call convergence in song evolution and its connection with species bio- two groups of zoo-kept African elephants (Vienna logy and ways of settlement of the main groups of Zoo, Austria, and Basel Zoo, Switzerland), which passerines. In total 30% of analyzing species have consisted each of four unrelated, but socially affili- female song. In Australia the percent of female song ated adult females that have been housed together consist of 28, in Europe - 10%, and has maximum for several years. Using multivariate analysis, I in Africa (33%) and America (37%). We have not found that their most commonly used social call, revealed a direct relation between female song and the “rumble”, varied significantly in source and duet singing (in suboscinae duet singing consists filter-related parameters between the groups. Given of 64% of species with singing female, in oscinae the females different genetic and geographical -51%) and carotenoid-based color. According to heritage, I conclude that that the study subjects our data female song occurs in different passerine matched some of the acoustic features of their calls families and probably is not ancestral to all birds to each other along with establishing social bonds. of this order. Females of subocinae practically lose However, significant acoustic differences persisted their ability to sing after penetration into the Old between individuals of each zoo group, indicating World, while in America the proportion of female that potentially both, group and individual identity song increases. The same tendency is for oscinae. are encoded in rumbles by a different subset of Hence female song is more widespread in young acoustical parameters. Call convergence and vocal group of oscinae and suboscinae (Ericson et al., learning might function to strengthen the bonds 2002). We hypothesis that female song appear between individuals during group formation and independently in different families of passerine. may further provide means of long-distance group identification within the elephant´s natural communicative system. Friday | Sept 11th 10.40 – 11.00

Friday | Sept 11th 11.00 – 11.20 Benedict, Lauryn Najar, N. University of Northern Colorado, U.S.A. Beme, Irina Goretskaia, M.I. Female song in New World wood-warblers Lomonosov Moscow State University, Russian Federation Abstract Recent advances have revealed that female birdsong is widespread and multifunctional. Is female song ancestral in songbirds? Female song was likely ancestral among songbirds and persists in many lineages today. Nevertheless, Abstract Bird song is generally associated with many species lack female song, and researchers are males, however in recent years the number of artic- interested in understanding the selective factors les devoted to female song increase. In the tropics that promote and counter the persistence of this female singing is appear to be more widespread trait. Female song is associated with life-history (Kroodsma et al. 1996; Morton, 1996; Langmore, traits including year-round territoriality, non-migra-

23 abstracts of symposium speakers

tory behavior, sexual monochromatism, and mono- studies addressing sound production of fish com- gamy. Most studies examining these relationships munities within an entire biotope, on a long term have looked at clades with a non-migratory ancestor basis and using this information in a monitoring and have found that gains of migratory behavior perspective. Coral reefs are hot spot of biodiversity are strongly correlated with losses of female song but are strongly impacted by climate change. Their (and duetting). Here we ask if the reverse pattern monitoring and conservation are therefore crucial. exists: in a large clade of songbirds with a migratory Our study, funded by Total Foundation, aims to ancestor, do losses of migratory behavior correlate develop a methodology allowing environmental with gains of female song and visual signaling monitoring by means of a description of the marine traits? We investigated correlations between female soundscapes along the Northern coast of Moorea song, migration and dichromatism in 107 species Island. Long terms acoustic recorders were positi- of New World Warblers (Family Parulidae). All of oned at 10m depth on the outer slope of 4 Marine these species are predominantly monogamous and Protected Areas and 4 non-Marine Protected Areas territorial when breeding, 50 (47%) are migratory, and collected data daily at a rate of 5min/hour du- 49 (46%) are monochromatic, and 25 (23%) show ring 48h hours. Four replicates were made for each female song. On a robust genetic phylogeny both area i.e. three replicates during the hot season and parsimony and maximum likelihood methods reco- one replicate during the cold season. Results show ver migration and monomorphism as the ancestral that ambient sound intensity and acoustic diversity state in warblers. Female song is generally not differ between areas and that these differences reconstructed as present in any deep nodes of might be linked to the fish communities and to the the phylogeny, indicating that most extant species type of habitats observed in situ. Overall, our study with female song evolved this trait independently may pave the way for a new, large scale, acoustics and relatively recently. Gains of female song do based monitoring tool in the effort to preserve coral not correlate with losses of migration. Losses of reefs. Soundscape analysis in air has proven very dichromatism do correlate with losses of migration. useful for determining how habitat structure may Thus, in this clade, visual signals are associated predict soundscapes and for assessing species with sedentary versus migratory lifestyles, but present in habitats where other survey methods female acoustic signals are not. Our results show a may be logistically difficult. Soundscapes are now different pattern from that seen in similar studies considered an important part of landscape ecology. and support the hypothesis that losses, but not Yet, our knowledge of underwater soundscapes is gains, of female song are driven by life-history. still in its infancy, especially in coral reefs.

Tuesday | Sept 8th 15.10 – 15.30 Thursday | Sept 10th 16.45 – 17.00

Bertucci, Frédéric Bhatt, Dinesh Lecchini, D. Dadwal, N. Centre de Recherches Insulaires et Observatoire de Gurukula Kangri University, Haridwar, l‘Environnement (CRIOBE), French Polynesia Uttarakhand, India Parmentier, E. University of Liège, France Onset of dawn chorus in a tropical songbird, the Pied Bush Chat (Saxicola caprata): relati- Large scale acoustic survey as a new tool onships with weather and sunrise for the evaluation of coral reef biodiversity (Moorea, French Polynesia) Abstract The avian dawn song chorus is a daily period of high song activity performed during the Abstract Passive acoustics took advantage breeding season. The timing of dawn chorus is like- of sound production in fishes as it offers natural ly the result of a number of environmental, social acoustic tags to identify and follow populations in and intrinsic factors. In this study, an attempt has the field over long periods. However, there are no been made to understand the relationships between

24 dawn chorus start times of male Pied Bush Chat tested emotional contagion through vocalisations in (Saxicola caprata) and environmental variables domestic horses. To this aim, we played back, to such as ambient temperature, sunrise time, cloud 18 horses in 5 different farms, whinnies of familiar cover, rain rate, wind speed, wind direction, lunar (same farm) and unfamiliar horses (different farm), phase, humidity etc. We recorded dawn song chorus recorded during emotionally positive (social reuni- of the male Pied Bush Chat at 12 different sites on) and negative (social separation) situations. Our in the rural landscape of district Haridwar (29°N), results show that horses are more alert (higher head India. Sunrise time and other morning weather pa- position, more head movements, ears forward for a rameters were acquired from weather station of the longer duration, less time spent eating and faster Department of Zoology and Environmental Science, response) when hearing unfamiliar compared to fa- G.K.U., Haridwar. Results indicated that extrinsic miliar whinnies. They are also more stressed (more factors play a significant role in influencing dawn movements and vocalisations, faster respiration chorus start times in the Pied Bush Chat. Males and heart rate) when hearing negative than positive commenced their dawn chorus earlier when the whinnies produced by familiar individuals. Horses moon was above the horizon at dawn. When it was are thus able to differentiate between whinnies in- raining or there were clouds, males started singing dicating different levels of familiarity and different later. We noticed that dawn chorus started at early emotional valence. Our study suggests that hearing times as the breeding season progressed relative to negative whinnies produced by familiar horses twilight/ sunrise. Males also showed differences in within a farm could trigger stress in all horses. daily song onset during initial breeding stages such as female attraction, territorial establishment and nest sites inspections. However, once the males Friday | Sept 11th 16.30 – 16.45 have settled in their respective territories, the onset of dawn chorus occurred in a more synchronized manner. Thus, it seems that a number of environ- Campbell, James mental and social factors influenced the onset of Slabbekoorn, H. dawn chorus in Pied Bush Chat singly or jointly. Sabet, S.S. Leiden University, The Netherlands

Wednesday | Sept 9th 10.00 – 10.25 Sound field characteristics of small tanks in the context of noise exposure experiments

Briefer, Elodie Abstract Anthropogenic noise in marine Mandel, R. ecosystems can affect the fitness of fish species Maigrot, A.-L. by masking important acoustic signals as well as Briefer Freymond, S. causing immediate physiological and behavioral Bachmann, I. changes. Although all fish are capable of detecting Hillmann, E. the particle motion component of a sound field and ETH Zürich, Switzerland only species possessing a swim bladder can detect the pressure component, a majority of experiments Emotional contagion through whinnies examining the effect of anthropogenic noise on fish in horses have focused exclusively on reporting pressure. In small aquarium acoustic experiments, due to the Abstract Emotional contagion through acoustic, unique acoustic characteristics of these tanks, the visual or olfactory signals is at the basis of empathy relationship between particle motion and pressure and has been studied in only a limited number of may not be representative of open water conditions. species. This phenomenon could have a positive To compare the effects of particle motion and or negative impact on conspecifics. Indeed, con- pressure on fish behavior, we conducted a startle tagion of positive emotions (e.g. joy) could improve response experiment exposing zebrafish (Danio welfare, whereas contagion of negative emotions rerio) to a series of white noise pulses (50-2000Hz) (e.g. fear) could have the opposite effect. We in a confined region of a small glass aquarium in

25 abstracts of symposium speakers

which the measured particle motion and pressure back of modified signals we showed that the value components showed no spatial correlation. We of the 1st harmonic and the frequency modulation compared the levels of sound pressure and particle are key parameters for species identification, and motion at the fish‘s location during sound exposure that a sequence of booms elicited stronger respon- to the occurrence, intensity, and directionality of ses than a single boom. Thus, the coding-decoding the resulting startle responses. While the post- of information relies on sequentially redundant and exposure swimming speed of individuals showed no propagation-resistant features, making the booms relationship with either component, the probability particularly well adapted to exchange information of a post-exposure freezing response was positively between males separated by long distances on their correlated with particle motion, thus highlighting displays sites. Moreover we showed that during the the importance of reporting particle motion measu- booming phase of courtship, two sensory modali- rements in small tank experiments. No relationship ties - acoustic and visual - act in synergy. While was found between the direction of startle response the visual component is not necessary to induce and the direction of particle motion, but particle agonistic response, it could act as an amplifier ellipticity measurements suggest the acoustic con- and reduces the time of detection of the signaller. ditions may have inhibited the individual’s ability The utilisation of these adaptive strategies allows to localize the sound source. While current models houbara males to maximize the active space of of sound localization in fish imply that particle el- vocalizations emitted in exploded leks. lipticity is a crucial component of localizing sound sources, to our knowledge this is the first study which quantifies particle ellipticity in the context TUESDAY | Sept 8th 17.30 – 17.50 of sound source localization in fish.

D’Amelio, Pietro Monday | Sept 7th 15.50 – 16.10 Adreani, M. Klumb, M. Trost, L. Cornec, Clément Gahr, M. Hingrat, Y. ter Maat, A. Rybak, F. Max Planck Institute for Ornithology, Seewiesen, Université Paris Sud, France Germany

Booming far: the long-range vocal communica- Motor control and recognition of unlearned tion of a lekking bird vocalizations

Abstract The pressures of selection acting on Abstract In the clade of birds, according to the message transmission by acoustic signals are current phylogeny, vocal learning evolved at least particularly high in long distance communication twice. How and why this spectacular behavior networks. In the north-african houbara bustard evolved is not yet completely clear. A pathway ena- (Chlamydotis undulata undulata) males produce bling the modification of spectral features of the strikingly low-frequency vocalisations called vocalizations must have come before and, possibly, “booms” as a component of their courtship per- independently from vocal learning. One reason for formed on sites separated at least 550 m apart. the evolution of this pathway might be that it favors We investigated the acoustic features in booms vocal recognition. To be recognized, vocalizations involved in long-range coding-decoding processes need to be individually distinct, i.e. vocalizations of species-specific identity. We first assessed the show a bigger variability between individuals than degradations of acoustic parameters during booms within individuals. Vocal recognition and motor transmission in the natural habitat of the bustard control of unlearned vocalizations are not frequent- and showed that the frequency content of the booms ly investigated in avian vocal learners. However was reliably transmitted up to 600 m. Second, by their study might aid the understanding of vocal testing males’ behavioural responses to the play- evolution and potentiate comparative analysis.

26 Here I aim to present the results of 2 studies to term acoustic monitoring program was conducted show that: 1) some unlearned vocalizations types at six floodplain channels of the French Rhône of a vocal learner species are individually recog- river differing in their degree of connectivity with nized, 2) lesions of the telencephalic nuclei of the main river. The number of sound types from the vocal pathway influence the spectral features a sample of 1440 recordings of one minute was of these unlearned vocalizations whereas they do identified by visual inspection of spectrograms and not influence their temporal pattern (i.e. duetting compared with the number of arthropod species with the partner). In summary I demonstrate that of each community. We detected a high acoustic even unlearned vocalizations can be partly modified diversity of 124 different sound types with 6 +/- 2 by the vocal control system. Also some unlearned sound types per recording. Unlike what was shown vocalizations are individually recognized. Taken for the species richness, we found that the lowest together these studies shed light on to the evolution acoustic richness was reached at intermediate of the song system in the songbird brain and help connectivity values. Most of the recorded sounds formulating of hypotheses on the evolution of the could not be identified, stressing out the need to auditory dependent vocal ontogeny. learn more about the identity and behaviour of the aquatic species producing acoustic signals underwater. This study highlights that sound diversity is Tuesday | Sept 8th 10.20 – 10.40 acomplementary facet of biodiversity, linked with taxonomic diversity, but with distinctive features. Exploring the aquatic acoustic diversity could Desjonquères, Camille provide insights into conservation priorities of the Llusia, D. endangered floodplain habitats. Sueur, J. Muséum national d‘Histoire naturelle, Paris, Université Paris-Sud, France Friday | Sept 11th 13.30 – 13.50 Rybak, F. Université Paris-Sud, France Castella, E. Bruno Gingras University of Geneva, Switzerland University of Innsbruck, Austria Doolittle, Emily Acoustic diversity in freshwater Cornish College of the Arts, U.S.A. habitats: the effect of lateral connectivity in a riverine floodplain A music theoretical approach to birdsong

Abstract Little is known about bioacoustics Abstract In this talk I discuss how scientific in freshwater habitats (i.e. ponds, rivers, lakes) and musical analysis can be used together to gain despite the use of sounds by several taxa such deeper understanding of birdsong. Technological as amphibians, fish, crustaceans and insects to tools measure frequencies and durations precisely communicate underwater. In particular, arthropods and statistical analysis enables determination of show high species diversity in such ecosystems whether apparent patterns occur above chance and their acoustic productions still need to be level, while music theoretical analysis allows for investigated. In riverine floodplains, habitat dy- more nuanced understanding of pitch, rhythmic, namics and macro-invertebrate communities are and structural relationships. The song of the hermit prominently driven by the extent of connection of thrush, admired by musicians and ornithologists waterbodies to the main river channel. According alike, has long been ascribed with following “major to the intermediate disturbance hypothesis, macro- and minor keys,” using “pentatonic scales,” and invertebrate richness is expected to be maximal at “meeting perfectly the requirements of human intermediate values of connectivity. In this study, music.” These human-music-specific attributions we investigated the effect of habitat connectivity are very unlikely in birdsong. Together with music on acoustic richness in order to explore whether theorist and cognitive biologist Bruno Gingras taxonomic and acoustic diversities co-vary. A long (University of Innsbruck), evolutionary and cogniti-

27 abstracts of symposium speakers

ve biologist Tecumseh Fitch (University of Vienna) a wide range of taxa. Remarkably, we find that all and computational psychologist Dominik Endres species tested employ the myoelastic-aerodynamic (University of Tübingen), we show that the song (MEAD) mechanism, the same mechanism used to instead follows the harmonic series. This is a phy- produce human speech. Furthermore, we show sical rather than a culturally determined pattern, substantial redundancy in the control of key vocal and thus is plausible in hermit thrush song. The parameters demonstrating that vocal signals are harmonic series shows up in much human music as not specified by unique motor commands. We well, so there is indeed significant overlap between suggest that such motor redundancy can accele- pitch relationships chosen by hermit thrushes and rate vocal learning and is common to MEAD sound those which appear in some human music. The production across birds and mammals, including musician wren is also famed for the “musical” qua- humans. lity of its song. Together with ornithologist Henrik Brumm (Max Planck Institute for Ornithology), we show that it sings consonant intervals – those Thursday | Sept 10th 14.10 – 14.30 based on small integer ratios between adjacent pitches – more often than dissonant intervals, with a particular preference for the simplest integer Fernandez, Marie S. A. ratios, perfect consonances. Though “consonance” Perez, E. C. and “dissonance” are human concepts, preference Vignal, C. for small integer ratios between adjacent pitches Université de Lyon, France may occur in multiple species. Not all birdsongs Griffith, S. C. are as amenable to pitch analysis as these two, Macquarie University, Australia but in future collaborations, rhythmic, timbral, and Soula, H. A. structural analysis could equally be combined with EPI BEAGLE INRIA, Lyon, France scientific methods to gain greater understanding of a variety of birdsongs. Impact of visual contact on vocal interaction dynamics of pair bonded birds

Monday | Sept 7th 11.40 – 12.00 Abstract Animal social interactions usually revolve around several sensory modalities. For birds, these are primarily visual and acoustic. Elemans, Coen However, some habitat specificities or large University of Southern Denmark, Denmark distances may temporarily hinder or limit visual Zollinger, S. A. information transmission making acoustic trans- Max Planck Institute for Ornithology, mission a central channel of communication. Seewiesen, Germany Here we show the impact of visual limitation on the vocal dynamics between zebra finches part- Universal mechanisms of sound production and ners. Pairs were acoustically recorded during a control in birds and mammals separation/reunion protocol with gradually decreasing distance without visual contact. The vocal Abstract As animals vocalise, their vocal organ activity of each bird was recorded throughout the transforms motor commands into acoustic signals protocol. We developed an in-house software sui- for social communication. In birds, the physical te that automatically detects vocalizations from mechanisms by which vocalisations are produced the recording and allows for individual tagging and controlled remain unresolved because of the of vocalizations, as well as automatic removal extreme difficulty in obtaining in vivo measure- of non-calls (wings or cage noise). Using these ments. Here, we introduce an ex vivo preparation automatic detection/extraction algorithms, we ob- of the avian vocal organ that allows simultaneous tained the call sequences and the temporal vocal high-speed imaging, muscle stimulation, and dynamics for each individual in each condition. kinematic and acoustic analyses to reveal the Without visual contact, pairs display more corre- mechanisms of vocal production in birds across lated vocal exchanges than with visual contact.

28 This high correlation renders the response’s delay Monday | Sept 7th 14.50 – 15.10 more predictable than without visual contact. We also analysed the sequences of caller’s identity (turn-taking) during an exchange with or without Fonseca, Paulo visual contact. We show that in the absence of Alves, D. visual contact, the identity of a caller is well Universidade de Lisboa, Portugal predicted by the knowledge of the identity of the Amorim, M. C. P. previous one (Markov process). Thus, both the MARE, ISPA-Instituto Universitário, Portugal temporal correlation and Markov properties of acoustic interactions indicate that in the absence Acoustic active space in the of visual clues the decision to emit a call is taken Lusitanian toadfish and the effect on a very short-term basis and solely on acoustic of anthropogenic noise information (both temporal and identity of caller). Strikingly, when individuals are in visual contact Abstract Mate attraction is mediated by acoustic both these features disappear indicating that birds signals in the Lusitanian toadfish, Halobatrachus adapt their calling dynamics to cope with limited didactylus. Males aggregate in suitable shallow visual cues. Our next goal is to apply this tool to habitats and advertise their presence to females larger groups of individuals to be able to study with a loud sound, the boatwhistle (BW, c. 140 acoustic and social network in groups of birds. dB SPL, i.e. re. 1 µPa, at 0.1 m). The acoustic active space of male BW is dependent not only on propagation constraints of the habitat, but Wednesday | Sept 9th 10.20 – 10.40 also on the background noise that is affected by anthropogenic noise such as generated by boating activities. We measured attenuation of BWs and of Fitch, W. Tecumseh its representative pure tone components (120 Hz Bugnyar, T. and 240 Hz) in an estuary breeding site (the Tagus University of Vienna, Austria estuary). BWs produced by breeding males were recorded with 2 m spaced equalized hydrophones Songbirds without song? Corvid communication in a linear array. Vocalization active space was from a comparative perspective estimated by comparing the envelopes of auditory evoked potentials (AEPs) using these recordings Abstract Corvids are oscine passerines („song- as stimuli, i.e. with BW recorded at increasing birds“) from a phylogenetic viewpoint, and yet distances from the source. Another estimation was they lack the male advertisement song typical of obtained by comparing attenuation of frequency to- most other songbirds. Curiously, one of the main nes with auditory sensitivity. These measurements song areas (HVC) in corvid brains is nonetheless point to an active space for the BW at c. 2 m deep well developed. Are corvids songbirds without water ranging between 6 and 13 m, depending on song, and if so, what is the function of primitively its spectral characteristics. In the presence of boat vocal areas? We will suggest that song behavior, noise this range was strongly affected. Noise from supported by vocal learning, can occasionally be small outboard motorboats and a larger ferryboat, observed in corvids, but that its social importance presented at 130 dB SPL, reduced the active is reduced. In its place, a suite of other learned space to c. 2.5-4 m and 7-8.5 m, respectively. behaviours such as the „self-aggrandizing dis- We further measured the vocal interactions (call plays“ of ravens have evolved which utilize corvid alternation) between males at increasing distances. vocal social learning capacities, in both sexes. Vocalizations of close neighbors were entrained In general corvids have large brains and highly while the interaction pattern disappeared in distan- developed learning abilities and social cognition. ces above a few meters, pointing to an active space It is possible that this neural circuitry plays a role of BW of the same magnitude or a little longer. in other non-display behaviours as well, and that this re-purposing represents an exaptation of the „standard“ neural circuitry for song learning.

29 abstracts of symposium speakers

Thursday | Sept 10th 10.40 – 11.00 an individual signature. First results suggest, that zebra finches use the´motif and not the phrase structure of the song to recognize individual cons- Geberzahn, Nicole pecifics. Thus, while transgenerational changes in Le Maguer, L. the song seem to emphasize individual differences Nagle L. in the phrase structure, it seems to be nevertheless Université Paris Ouest Nanterre La Defense, the motif which carries the individual signature. France: Université Paris Lumières, France Zsebök, S. Aubin T. Thursday | Sept 10th 16.00 – 16.15 Université Paris Sud, France Derégnaucourt S. Université Paris Ouest Nanterre La Defense, Gomes, Dylan France; Université Paris Lumières, France; Smithsonian Tropical Research Institute (STRI), Institute Universitaire de France Panama Halfwerk, W. Towards an understanding of developmental VU Amsterdam, The Netherlands; STRI, Panama and perceptual mechanisms of vocal individua- Page, R. lity in the zebra finch STRI, U.S.A.

Abstract Understanding the origins of language Foraging decisions in clutter and noise: the ef- is a challenge in human sciences and evolutionary fect of environmental complexity on multimodal biology. If language is a human attribute, one of its signal use in a frog-eating bat components – vocal learning – is shared with other species including oscine songbirds. Experiments to Abstract Conspicuous signals are used throug- address central questions about vocal culture that hout the animal kingdom to attract mates and to cannot be conducted on humans for ethical reasons keep competitors at bay. However, eavesdroppers can be done with captive populations of birds. frequently intercept these signals. Signalling in We used a songbird, the zebra finch (Taeniopygia multiple sensory modalities can increase signal guttata) to study transgenerational transmission effectiveness to the intended receiver, especially in of song under experimental conditions. The main noisy environments, but may also give similar ad- component of the zebra finch song is the motif, a vantages to eavesdropping predators. Furthermore, stereotyped sequence of syllables, which is repea- conspecifics and predators can perceive the same ted several times in a song often with connecting signals/cues in different sensory modalities. Thus, call-like syllables in between different motif ren- variation in the sensory environment can differenti- ditions. The way motif renditions and connecting ally impact conspecific communication and preda- syllables are organized characterizes the phrase tor eavesdropping. To cope with sensory variation, structure of the song. Typically, each male finch many bats use multiple sensory modalities while produces a different song which thus carries an in- foraging. The frog-eating bat, Trachops cirrhosus, dividual signature. How do individual signals evolve uses two types of sound while hunting. It eaves- in an environment with uniform song? Which song drops on low frequency frog mating calls, and uses features are used for individual recognition in such ultrasonic echolocation calls to search for frogs. a case? To study these questions we used 10 males While the acoustic component of the frog mating producing the same song motif as founders of a display is prominent, the dynamically moving vocal colony and raised them with females in an aviary. sac provides a salient signal to female frogs (in the We found that most of the males that hatched in visual domain) and to frog-eating bats (via echo- the colony (n= 50) produced a close copy of the location). To understand how bats use multimodal colony motif. However, males largely varied in their prey signals under different sensory conditions, phrase structures. Next, we used an instrumental I used playbacks of túngara frog (Physalaemus discrimination task (Go/Nogo) to study whether pustulosus) mating calls, and a newly developed phrase structure rather than the motif serves as robotic frog model to mimic foraging opportunities

30 of a main prey species. Playbacks were broadcast by investigating the sound transmission pathway, as two-choice test, coupling one with a dynamically from the environment to the inner ear. B. pitanga inflating robofrog (multimodal signal) and the lacking a tympanum, we searched for another other with a static model, and quantified bats’ interface between the air and the body for sound approach behavior under varying levels of noise transmission. We used laser Doppler vibrometry and and echoacoustic clutter. Our results suggest that compared the resonance frequencies and oscillati- the multimodal signal enhances localization of the ons amplitude of potential entry points on the body mating display when confronted with background of the toad. To have a better understanding of the noise. Furthermore, bats show longer search times internal sound transmission pathway, we also built with the addition of echo-acoustic clutter and a 3D reconstruction of the internal ear structures. background noise. Understanding how bats rely We thereby describe in details the characteristics on passive listening (prey-generated sound) and of B. pitanga internal ear structures and propose a active foraging (echolocation) in different sensory hearing pathway for airborne sound above one kHz. environments is important for determining the These results shed light on the possible alternative impact of sensory disturbances on foraging bats as hearing pathways in other earless anurans. well as understanding selection pressures shaping the sexual advertisement calls of its prey. Friday | Sept 11th 11.20 – 11.40

Tuesday | Sept 8th 17.15 – 17.30 Goyes Vallejos, Johana University of Connecticut, U.S.A. Goutte, Sandra Toledo, L.F. Female calling behavior in the smooth guardi- Universidade Estadual de Campinas (UNICAMP), an frog Limnonectes palavanensis of Borneo. Brazil Evidence for sex-role reversal?

Acoustic communication of the earless toad Abstract Female frogs typically remain silent, Brachycephalus pitanga while males produce advertisement calls to attract mates. In some species, females also vocalize, Abstract As in most anurans, males usually as a response to a male advertisement Brachycephalus pitanga produce advertisement call. In the smooth guardian frog (Limnonectes calls to attract potential mates. However, B. palavanensis), found on Borneo, males are smaller pitanga lacks the typical anuran tympanic ear: than females and provide parental care, attending tympanic membrane, Ostia pharyngea, tympanic eggs and transporting tadpoles from the leaf annulus and stapes. This particularity makes the litter to water. I found that males call to attract tiny toad anatomically « deaf » to sounds above one females, but do so infrequently, and females call kHz – frequencies below one kHz being transmitted in response. Females also vocalize spontaneously, to the inner ear via the operculum system. The forming small choruses to which the males respond advertisement call of B. pitanga, with a dominant with a short courtship call. Behavioral observations frequency around four kHz, falls in the range showed that multiple females approach a calling normally detected by anurans via their tympanic male, and females alternate their calls when more ear. As vocalizations have a high energetic cost for than one calling female is present. In addition, I the emitting males and may attract predators, it is examined the function of female calling and its not likely that these vocalizations are an ancestral role in possible sex-role reversal. When presented behavior lacking function. We therefore postulated with a playback stimulus of a calling female, fe- that this species must have evolved a different males increased calling rate and sometimes moved hearing pathway for sounds with a frequency towards the source of the stimulus, indicating a above one kHz, such as their own advertisement predisposition for agonistic behavior. Males did not calls. In this study, we aimed at uncovering how B. exhibit aggressive behavior. These observations are pitanga perceives airborne sounds above one kHz consistent with sex-role reversal in this species.

31 abstracts of symposium speakers

Monday | Sept 7th, 11.20-11.40 behavioral pattern that can be found in all mammalian species. It can be subdivided into learned vocal patterns such as human speech, and geneti- Gridi-Papp, Marcos cally pre-programmed vocalizations, which include University of the Pacific, U.S.A. those of non-human primates and most other mammals. In my talk, I will give an overview on Laryngeal muscles and vocal control in our recent studies on vocalization-related cognitive amphibians and reptiles control processes in the cerebral cortex, and how these processes might be influenced by the auditory Abstract Most ectotherm tetrapods have vocal system. We performed these studies with a broad cords but lack intrinsic laryngeal muscles attached range of different methodological approaches to them. The arytenoid cartilages are pivoted open including neurophysiological, neuroethological by the laryngeal dilator muscle and pivoted closed and psychophysical techniques in rhesus monkeys. by two laryngeal constrictor muscles that together In particular, we studied evolutionary aspects of form a sphincter. In amphibians, a fourth muscle human speech control with our newly established called the posterior laryngeal constrictor has been primate model and recorded single neurons in the suggested to stretch the vocal cords, but experi- vocalizing rhesus monkey. We found a neuronal cor- ments of electrical stimulation rejected the idea relate of volitional call initiation in the prefrontal and revealed the the role of tis muscle is to adduct cortex that suggests a cardinal role of this structure the vocal cords into the airflow for phonation. We in cognitive processes that are involved in vocal are currently evaluating a potential fifth laryngeal motor control mechanisms. Furthermore, our data muscle in amphibians. Researchers have noted indicate that the ventrolateral prefrontal cortex that the laryngeal dilator muscle has a secondary plays an important role in audio-vocal integration attachment to the cricoid cartilage. We examined processes, which might be a pre-requisite for com- this attachment and found that in túngara frogs plex audio-vocal communication processes, which (Engystomops = “Physalaemus” postulosus) can be found in non-human primates. the laryngeal dilator is actually divided into two bundles: superficial and deep. The superficial portion of the dilator muscle corresponds to the Thursday | Sept 10th 13.30 – 13.50 traditional description of the dilator muscle. The deep bundle of the dilator muscle attaches to the arytenoid cartilage and to the cricoid cartilage. It is Hallerberg, Sarah only innervated by the short branch of the laryngeal Timme, M. nerve and it is the only laryngeal muscle innerva- Max Planck Institute for Dynamics and Self- ted by this nerve. All other laryngeal muscles are Organization, Göttingen, Germany innervated by the recurrent laryngeal nerve. We are Vester, H. currently evaluating its potential role in active con- OceanSounds e.V., Germany trol of the fundamental frequency or of the acoustic Hammerschmidt, K. complexity in male advertisement calls. German Primate Center, Göttingen, Germany

Bag-of-calls analysis reveals Monday | Sept 7th 12.40 – 13.00 group-specific vocal repertoire in long-finned pilot whales

Hage, Steffen Abstract Besides humans, several marine University of Tübingen, Germany mammal species exhibit prerequisites to evolve language: high cognitive abilities, flexibility in Vocal motor control mechanisms vocal production and advanced social interactions. in the monkey brain Here, we describe and analyse the vocal repertoire of long-finned pilot whales (Globicephalus melas) Abstract Vocal communication is a complex recorded in northern Norway. Observer based ana-

32 lysis reveals a complex vocal repertoire with 140 study the genetic foundation of vocal learning. different call types, call sequences, call repetitions Furthermore, the findings indicate that a complex and group-specific differences in the usage of call acoustic structure is not necessarily linked to gre- types. Developing and applying a new automated ater flexibility or voluntary control. This highlights analysis method, the bag-of-calls approach, we find the limited explanatory power of structural analyses that groups of pilot whales can be distinguished pu- of sounds alone to identify production mechanisms. rely by statistical properties of their vocalisations. Comparing inter- and intra-group differences of ensembles of calls allows to identify and quantify Monday | Sept 7th 12.20 – 12.40 group-specificity. Consequently, the bag-of-calls approach is a valid method to specify difference and concordance in acoustic communication in Herbst, Christan T. the absence of exact knowledge about signalers, University of Vienna, Austria which is common observing marine mammals under natural conditions. Monitoring the mammalian and avian sound source with electroglottography

Friday | Sept 11th 14.50 – 15.10 Abstract In both mammals and birds, the laryngeal/syringeal sound generator is the causal link in the sound production chain from neural motor Hammerschmidt, Kurt control to vocal output, largely governing features Fischer, J. such as the regularity of the vibratory patterns German Primate Center, Göttingen, Germany (periodic vs. irregular), the fundamental frequency, or the physical efficiency when converting aerody- Structural complexity can be a misleading namic to acoustic energy. Monitoring the dynamic indicator for advanced communication: The properties of the sound source during voice produc- case of mouse courtship songs tion is thus a central goal in bioacoustic research, yet experimental difficulties when performing Abstract In animal communication, the iden- invasive measurements such as (high-speed) video tification of structurally complex vocal utterances endoscopy in vivo prevail. In humans, an alternative often is frequently taken as evidence for flexibility low cost, non-invasive impedance measurement in vocal production, including the capacity for method for monitoring vocal fold vibration has been vocal learning. This was the case in male mouse successfully utilized in the past decades: electro- courtship songs. Their complex and variant ultra- glottography (EGG). A low intensity, high-frequency sonic vocalizations were described as “song”, and current is passed between two electrodes placed on it was suggested that they might be similar to bird each side of the larynx. The admittance variations song, and learning might be involved in their deve- resulting from vocal fold (de)contacting during lopment. Studies on cross-fostered, deaf born as laryngeal sound production are largely proportional well as deafened mice, however, failed to identify to the time-varying relative vocal fold contact area. substantial differences between the vocalizations The EGG signal has found to be a reliable physiolo- of these mice and control subjects. These findings gical correlate of vocal fold vibration, reflecting the clearly indicated that mice do not need auditory oscillatory regime (periodic or irregular, including feedback to develop and produce their strain- bifurcations), and the fundamental frequency. specific vocalizations. Moreover, the USVs of mice Research in humans suggests that landmarks in that lack Esco2 in the growth zone of the cerebral the EGG waveform may also give certain insights cortex, which results in the lack of hippocampal into pre-phonatory settings governed by intrinsic and neocortical structures, did not differ from those laryngeal musculature, and their adjustment during of control mice. Apparently, cortical structures are vocalization. In this presentation it will be demons- not necessary to develop these vocalizations or to trated that electroglottography is well suited for produce these sounds in the appropriate context. analyzing mammalian voice production, based on These results question the potential of mice to recent studies with (hemi)larynx preparations of

33 abstracts of symposium speakers

various mammalian species, utilizing simultaneous appear to have trouble with relative pitch, looking high-speed video recordings. Ongoing research by at the bioacoustics of black-capped chickadee song this author will extend the experimental paradigm and previous perceptual research, we knew that to analysis of in vivo sound production, e.g., in relative pitch was readily produced and perceived calves, dogs, and Japanese macaques. Finally, by this species in the performance and evaluation collaborative studies conducted in Dr. Elemans’ lab of their species-specific song. We found that the suggest that the EGG approach is well applicable to relative pitch ratios used in natural songs are birds, giving insights into dynamics of the oscilla- processed more readily than other arbitrary relative ting syringeal membranes. pitch relationships. In conclusion, by evaluating the natural vocalizations of species we gain insight in possible perceptual abilities that then can Friday | Sept 11th 14.10 – 14.30 be tested in laboratory studies, leading to value insights on comparative perception.

Hoeschele, Marisa Fitch, W. T. Tuesday | Sept 8th 10.40 – 11.00 University of Vienna, Austria Sturdy, C. B. University of Alberta, Canada Jaiswara, Ranjana Guillette, L. M. Dhiman, S. University of St. Andrews, U.K. Jain, M. Indian Institute of Science Education and Using bioacoustics to compare avian and Research, Mohali, India human acoustic perception Species boundary delineation using multiple Abstract The bioacoustics of animal vocaliza- approaches: a case study on a field cricket tions can give us surprising insights into acoustic genus Teleogryllus from India pattern learning abilities across species. This can be especially useful when looking for human-like Abstract Unambiguous identification of species acoustic abilities in other species. Human speech is a fundamental need for all biological studies. signals can give us insight into what kinds of fea- However, this becomes non-trivial in cases where tures are important in human language perception. individuals are inseparable morphologically. In this In one study, the perception of metrical patterns study, field cricket belonging to genus Teleogryllus in human speech were evaluated in humans and were collected and their calling songs were recor- budgerigars. First, we looked at previous research ded from different localities in India. We found studying the structure of metrical patterns in hu- external morphological differences to be very subtle man speech. We used this information to create two and characters highly overlapping. Comparison of syllable nonsense words where either the first or the Chopard’s (1969) description, based on the genita- second syllable was emphasized, and we trained lia structure, suggest that individuals belong either both species to discriminate the two patterns of to Teleogryllus mitratus (Burmeister, 1838) or to emphasis. We found that both humans and bud- Teleogryllus testaceus Walker, 1869. Townsend gerigars were able to detect metrical phonology in (1980), however, redefined T. mitratus Burmeister these signals, suggesting that detecting these kinds and concluded that testaceus Walker is a synonym of patterns may be a general acoustic perceptual of mitratus Burmeister. Undertaking both Chopard ability, at least in vocal learning animals. In ano- (1969) and Townsend (1980) documentation, there ther study, we evaluated relative pitch perception arises a problem of ambiguity in species identity in the black-capped chickadee. Relative pitch is of the Teleogryllus species. In this study we aim the ability to identify the ratio difference between to resolve this ambiguity using multiple data sets: two notes, a common perceptual phenomenon in morphology, acoustics and behavior. Preliminary humans that is used frequently in music to create results based on comparison of genitalic structures, tension and release. While many avian species acoustic analysis of calling songs and behavioural

34 isolation experiments lend support to Chopard the lion’s roaring and initiated the scan quicker af- (1969) in which he advocates the presence of two ter exposed to alarms of kudu and guineafowl rather different species. We discuss our results in light than to the jackal’s howling. According to our best of Chopard’s hypothesis of possible hybridization knowledge, this is the first study reporting ability of between the two species and the importance of ungulate to differentiate between calls of apex- and calling song of these species to serves as a means meso-predators that compares the recognition of of premating behavioural isolation. heterospecific alarms and predator vocalizations. Our findings may have an applied value in conservation. Especially large predators, important ecosystem components, frequently suffer local extinctions Friday | Sept 11th 17.15 – 17.30 and knowing prey-predator acoustic interactions may increase success of repatriation initiatives of both prey and predator species. Klimšová, Vêra Policht, R. Acknowledgement: CIGA 51120/1313/3104 and Czech University of Life Sciences Prague, IGA 51120/1312/3113 Czech Republic

Impalas recognize predator vocalizations Friday | Sept 11th 14.30 – 14.50 and heterospecific alarms of sympatric and allopatric species Kriengwatana, Buddhamas Abstract Heterospecific recognition of alarm Leiden University, The Netherlands calls and predator vocalizations has been reported in wide range of taxa. Some animal species even A general auditory bias for disregarding recognize and respond by the antipredator beha- inter-speaker speech variability? Evidence in viour to alarm calls and predator vocalizations of humans and songbirds allopatric species. Despite the fact that prey-predator interactions may be theoretically important in Abstract Human speech perception requires conservation efforts in context of acoustics, little is deciphering a highly variable signal. Vowel normali- known about this ability in ungulates. Especially, zation is a proposed method by which humans can in areas with rich assemblage of predators such as categorize vowels despite the extensive variation the Kruger National Park in South Africa a greater that stems from speaker/gender differences in vo- variability in recognition of different acoustic cues wel production. Studies in humans and non-human can be expected. By using playback experiments animals indicate that experience with very few we simulated alarm calls of kudu (Tragelaphus speakers is sufficient for vowel categorization, strepsiceros) and helmeted guineafowl (Numida and normalization of speaker differences in vowel meleagris) and vocalizations of black-backed jackal production may arise from sensitivity to formant (Canis mesomelas), spotted hyena (Crocuta crocu- ratios. This study tests the hypothesis that nor- ta) and lion (Panthera leo) and observed responses malization of speaker and gender differences in of focal impala (Aepyceros melampus) as ungulate isolated vowels does not require prior experience model species. To test whether impalas adequately with speaker-related variability because there is respond also to alarm of species not known to them an inherent tendency for the auditory system to we played calls of allopatric Eurasian jay (Garrulus transform vowel formants into formant ratios. Using glandarius). Non-alarm calls of several bird species a behavioural Go/No-go task and identical stimuli, were used as controls. The results showed that im- we compared Australian English adults’ (naïve to palas responded by quicker look towards the source Dutch) and zebra finches’ (naïve to human speech) of sound, longer scanning of environment or esca- ability to categorize /ɪ/ and /ɛ/ vowels of an novel ping to both, alarm (jay including) and predator, Dutch speaker after learning to discriminate those calls rather than to controls. They were scanning vowels from only one other speaker. Experiment 1 the environment significantly longer after hearing and 2 presented vowels of two speakers interspersed

35 abstracts of symposium speakers

or blocked, respectively. Results demonstrate that suggest that adults tend to live in pairs, we tested if categorization of vowels is possible without prior the chemical environment, in this case, substrates exposure to speaker-related variability in speech for with scents of the tested individual or from a cons- zebra finches, and in non-native vowel categories pecific of a different sex, may modulate the receiver for humans. Therefore, this study is the first to pro- response. There was an interplay between the type vide evidence for what might be a species-shared of calls and the chemical environment; individuals auditory bias that may supersede speaker-related reacted with longer immobility after complex than information during vowel categorization. It additi- simple calls, but when they were with their own onally provides behavioural evidence contradicting scents (i.e. the own territory). When exposed to the hypothesis that vowel normalization is achie- scents of a conspecific of a different sex, complex ved via the use formant ratios F1/F3 and F2/F3. calls determined a shorter immobility than simple Therefore, investigations of alternative accounts of calls. These results suggest that the weeping lizard vowel normalization that incorporate the possibility discriminates between simple and complex distress of an auditory bias for disregarding inter-speaker calls, but the triggered antipredator response is variability are warranted. modulated by “where” lizards perceived the risk; a fearful call in the own territory is riskier. Funds: Fondecyt 1120181 (AL); Conicet scholarship (MR). Tuesday | Sept 8th 16.00 – 16.15

Thursday | Sept 10th 10.00 – 10.20 Labra Lillo, Antonieta Universidad de Chile, Chile and University of Oslo, Norway Lachlan, Robert F. Monachesi, M.R. Queen Mary University of London, U.K. Instituto de Herpetología, Fundación Miguel Lillo, Argentina Evolution and function of vocal learning in the chaffinch (Fringilla coelebs) Interaction between calls and scents in lizard antipredator behavior Abstract Vocal learning in songbirds is remarkably diverse, and one facet of this is shown by variation Abstract Distress calls are vocalizations produ- between species in the precision with which they ced by individuals when they are trapped by a pre- imitate tutors. We still lack a clear understanding dator. They may enhance the survival probabilities of why this variation exists and even of why some of the caught individual and/or of its conspecifics. species learn precisely. By applying spatially explicit The weeping lizard, Liolaemus chiliensis, emits dis- cultural evolutionary simulations, it is possible to de- tress calls that seem to benefit conspecifics as they monstrate that European chaffinches learn extremely react to these calls, specifically to simple down-mo- precisely, with a “cultural mutation rate” of only 1%. dulated, with prolonged immobility, a behavior that But they do not show typical “dialectal” patterns of would reduce their probabilities of being detected song variation, and do not preferentially share songs by the predator. This behavior has been recorded with neighbours, ruling out the majority of functional when lizards are confronted with distress calls. explanations for precise learning. A recent hypothe- This lizard, however, emits a wide variety of distress sis proposes that precise learning is an assessment calls, including complex calls with different types signal, in which males attempt to learn ‘prototypical’ of non-linear phenomena. It has been proposed that versions of local song-types, and receivers learn these nonlinear sounds can honestly communicate prototypes in order to distinguish precise from im- fear, and if so, it is expected that alarmed lizards precise learners. I used a computational bioacoustic exhibit a stronger antipredator response when they analysis to identify prototypical exemplars of local are exposed to complex distress calls. We tested song types in a Catalonian chaffinch population. In this hypothesis exposing individuals to simple and a playback study in the same population, territorial complex distress calls; the latter contained chaotic males responded more aggressively to prototypical noise. In addition, and considering that evidence than to less typical versions of song types, sup-

36 porting the assessment hypothesis. When the same by motor neurons forming vocal/sonic motor nuclei stimuli were played back to a different population, (VMN). Three major organization patterns of VMNs unfamiliar with the song types, they were not dis- are known: VMNs positioned in the midline (toad- tinguished between, suggesting that prototypes were fishes) or along the ventral border (pectoral mecha- in fact learned. Finally, I will discuss evidence that nisms) of the posterior hindbrain and anterior spinal this system of precise imitation and assessment has cord. The third pattern shows swimbladder muscles rapidly broken down during the colonization of the innervated solely by spinal motor neurons (piranhas). Atlantic Island archipelagos. In these genetically Those catfishes that possess two mechanisms have a isolated populations, syllable recombination and midline VMN associated with swimbladder muscles imprecise syllable imitation occur at far higher rates and a multifunctional ventral nucleus for sonic and than on the mainland, demonstrating how learning non-sonic (locomotion) pectoral functions (pectoral behaviour can rapidly diverge between evenly closely motor nuclei, PMN). VMNs are controlled by pace- related taxa. maker neurons that receive inputs from other vocal nuclei. It remains to be investigated if vocal fishes have a common vocal pathway rostral to the VMNs. Monday | Sept 7th 11.00 – 11.20

Thursday | Sept 10th 17.00 – 17.15 Ladich, Friedrich University of Vienna, Austria Robillard, T. Sound-producing mechanisms and Legendre, F. neuronal control in fishes: A unique diversity Muséum national d´Histoire naturelle, Paris, in vertebrates France Lafond, Augustin Abstract Fishes have evolved the largest diversity Universidad Federal de Viçosa, Brazil; Muséum of sound-generating (sonic) mechanisms among national d´Histoire naturelle, Paris, France vertebrates. The main group of sonic mechanisms is based on the swimbladder. These can be vibrated Evolution of mechanisms of sound production by intrinsic drumming muscles located in the swim- in agnotecous crickets: multiple strategies to bladder wall (toadfishes) or by extrinsic drumming produce high-frequencies muscles originating on structures such as the skull, vertebral processes or body wall musculature. Abstract Crickets are acknowledged for the Extrinsic drumming muscles insert either directly on diversity of their calling songs, which play a major the swimbladder (e.g. pimelodid catfish) or vibrate role in reproduction and in preventing hybridization the swimbladder indirectly either via broad tendons in closely related species. The general mechanism of (piranhas) or bony plates (doradid catfishes). Pectoral sound production in crickets is well studied and is of- sound-producing mechanisms include the vibration ten considered as simple and stereotyped. However, of the pectoral girdle (sculpins), the rubbing of the its morphological and behavioral parameters can enhanced pectoral spine in a groove of the shoulder vary significantly, which may explain a wide array of girdle (catfishes) and the plucking of enhanced fin the acoustic diversity observed in the calling songs tendons (croaking gouramis). In addition, sounds can of these insects. We studied the evolution of the be produced by other structures such as head-neck sound generating system in the genus Agnotecous, articulation (seahorses) or pharyngeal teeth (clown- which is characterized by the production of high- fish). Two different sonic mechanisms (producing frequency calling songs. The species of this genus low- and high-frequency sounds) may be present are very similar in terms of morphology, but they simultaneously (catfishes, seahorses). Sonic organs produce songs with a wide range of frequencies (10- can be similarly developed in males and females 20 kHz). The study consisted in fine scale analysis (croaking gourami), or they are sexually dimorphic of the morphological and behavioral variables of the (toadfishes) and may hypertrophy seasonally in ma- system. These multiple variables have then been les (cods). Sound-generating muscles are controlled studied in reference to the phylogeny of the genus,

37 abstracts of symposium speakers

using methods taking into account the phylogenetic gradients creating ‘sound shadows’. To investi- relationships among the species: ancestral state gate such effects we performed a transmission reconstruction, phylogenetic PCA, comparisons of experiment in a typical conure habitat in Costa phylogenetic signal, phylogenetic generalized least Rica. We broadcast contact calls from a speaker square analyses. These analyses investigated evolu- placed at three different heights and recorded tionary links between the stridulatory behavior and with two microphones at three different distances structures in order to understand how the mechanism while simultaneously recording temperature and participated in the diversification of the acoustic si- humidity at three different heights. In the morning gnals. Unexpectedly, we have found that the studied neither temperature nor humidity gradients were species developed two different strategies to produ- found up to six meters above ground but at noon ce high-frequencies: (1) some species increased the clear gradients had developed. The experiment speed of the stridulatory movement until reaching demonstrated that temperature and humidity directly high dominant frequencies, (2) while others gradients at noon affected sound transmission exploited the vibration capacities of their forewings significantly. Temperature and humidity induced to amplify harmonic frequencies coming from slow changes in sound speed suggesting that at midday stridulatory movements. These results support the ‘sound shadow zones’ develop roughly between hypothesis that such high-frequency songs evolved 60 and 70 m from a calling bird adding to the quickly, repeatedly and stepwise, by a form of punc- excess attenuation created by the ground effect tuated evolution rather than by only the progressive and atmospheric attenuation. (Supported by the increase of the ancestral fundamental frequency. Danish Council for Independent Research - Natural Sciences #10-084844)

Monday | Sept 7th 15.30 – 15.50 Thursday | Sept 10th 17.15 – 17.30

Larsen, Ole.N. University of Southern Denmark, Denmark Linossier, Juliette Balsby, T. J. S. Zsebök, S. Aarhus University, Denmark Baudry, E. Dabelsteen, T. Aubin, T. University of Copenhagen, Denmark Courvoisier, H. Université Paris-Sud, France Call transmission advantage for parrots at dawn in tropical dry forest Song and genetic divergence related to different migratory patterns in a songbird, Abstract Central American orange-fronted the blackcap Sylvia atricapilla conures (Eupsittula canicularis) live in fission- fusion flocks outside the breeding season, where Abstract In songbirds, territorial songs are key fission-fusion events may occur on a daily basis. regulators of sexual selection and are learned When a flock flies over the forest, its members from conspecifics. The cultural transmission of usually emit loud contact calls, to which perched songs leads to divergence in song characteristics individuals may respond and attract the flock for within populations and ultimately can lead to longer exchange of contact calls. These activities speciation. Many songbirds also migrate and occur mainly in the morning and to a smaller extent individual differences in migratory behaviors can late in the afternoon. There probably are several influence population genetic structure and local reasons for this diurnal variation in calling activity. song differentiation. The complex interactions of Here we investigated whether diurnal variations in song structures, migratory routes and genetic diver- temperature and humidity and consequently in the sity can underlie ongoing selection processes, and speed of sound coincided with changes in calling remain to be better understood. Blackcaps (Sylvia activity. We hypothesized that one limiting factor atricapilla) exhibit versatile territorial songs with could be development of midday sound speed geographical variations and show a large diversity of

38 populations from sedentary to migratory. It appears examining acoustic interactions in bird dawn as a good model to study the relationships between chorus, a widespread but poorly understood phe- migratory patterns, song variability and genetic nomenon. Community acoustic monitoring was diversity. We studied a population (2 groups) of conducted in twenty forest sites in France during migratory blackcaps around Paris and a population three consecutive days. By remote-controlled (3 groups) of sedentary blackcaps in Corsica. All playback tests, forest bird communities were the birds were ringed and blood samples were exposed, only the second day, to acoustic stimuli taken in order to study genetic relatedness using of Parus major, a common species occurring in 17 microsatellite loci. A detailed song analysis all the recorded sites. The stimuli started around showed that this species has a complex repertoire 30 min before the time of the first song of P. (>100 syllables) which required the development major recorded the previous day. General linear of a semi-supervised methods to classify different mixed models indicated that this experimentally categories of syllables and compare sequences of induced temporal shift in P. major elicited, as syllables (custom-made program in Matlab). Our expected, earlier singing activity in P. major, analysis showed no genetic structuring among but more surprisingly also in other species of populations. Individuals belonging to a same po- the community, in comparison with pre- and pulation are not genetically closer that individual post-stimulus days. Our results suggest that ea- from different populations. However we found a vesdropping on other species plays an important strong morphologic differentiation between seden- role in the dawn chorusing behaviour, specifically tary and migratory populations. We also show that influencing the time of species’ singing activity geographical variation in songs relies at least on within the dawn chorus. This constitutes a novel both syllable and sequence content and, contrary example of how heterospecific acoustic signals to what we expect, that sedentary groups do not may act as cues providing information within an share more than migratory groups. Song dialects do animal community. not seem to act as barriers to gene flows and are strongly cultural rather than genetic. Thursday | Sept 10th 11.00 – 11.20

Tuesday | Sept 8th 14.30 – 14.50 Luo, Jinhong Max Planck Institute for Ornithology, Germany Llusia Diego Wiegrebe, L. Desjonquères, C. Ludwig Maximilians University, Germany Ulloa, J. S. Sueur, J. The Lombard effect emerges early in bat infants: Muséum national d’Histoire naturelle, Université implications for the development of auditory Paris-Sud, France feedback Aubin, T. Université Paris-Sud, France Abstract Auditory feedback, i.e. monitoring one’s own vocalizations, is critical in learning and Heterospecific songs influence dawn chorusing maintenance of both birdsong and human speech. in forest bird communities However, only few studies have investigated the development of auditory feedback and results Abstract Animal acoustic communication is indicate a rather late developmental origin. The primarily involved in social interactions among Lombard effect, a well-studied form of audio-vocal conspecifics. The study of species interactions interaction, refers to the phenomenon that humans/ has been mostly restricted to masking interfe- animals increase their vocal loudness in response to rence or acoustic cues within predator-prey in- increasing background noise. Here we investigated teractions. Further evidence of sound information the postnatal development of the Lombard effect sharing among animal species is still scarce. Here in echolocating bats which are the only known we report for the first time an experimental study non-human terrestrial mammals capable of vocal

39 abstracts of symposium speakers

learning. Specifically, we tracked the ability of in- interesting given the absence of size dimorphism fant bats to adjust their vocalizations to playbacks in all species examined. A phylogeny constructed of white noise, from the age of only a few days to from sequence data representing both mitochon- three months. We found that infants increased their drial and nuclear genomes (Miller and Engstrom vocal loudness in noise even in the first week of tes- 2008) provides a backdrop to explore evolutionary ting. Moreover, we found that the magnitude of the patterns in this assemblage. The results suggest Lombard effect did not change over the 3-month that dimorphism in vocal behaviour is, in part, period. These findings demonstrate that bats have phylogenetically constrained. Species whose vocali- already fully developed the capacity for audio-vocal zations are sexually dimorphic have sibling species interaction at the age of one week, challenging with sexually dimorphic vocal signals also. Only in the current view of late developmental origin of one species sampled were stereotypic and complex auditory feedback. vocalizations made solely by males. Dimorphism in acoustic signals occurs among peromyscine mice, and the existence of female stereotypic vocaliza- Friday | Sept 11th 10.20 – 10.40 tions merits further study. It is proposed that the investigation of mammalian vocal behaviour can provide a comparative context, through which Miller, Jacqueline R. deeper evolutionary and genomic influences can be Royal Ontario Museum, University of Toronto, explored. Canada

Sexual dimorphism in vocal behaviour: Tuesday | Sept 8th 17.00 – 17.15 exploring the peromyscine mice of North and Central America Mizuguchi, Daisuke Abstract Several mammalian taxa produce Tsunokawa, M. stereotypic vocalizations, representing simple Kohshima, S. sequences to complex songs. Among the best in- Wildlife Research Center of Kyto University, Japan vestigated are cetaceans, primates, carnivores and bats, particularly with reference to song. Relatively Underwater vocal exchange between male and little, however, is known regarding stereotypic acou- female in captive bearded seals stic behaviour in rodent species, with the exception of the laboratory rat and mouse. Here, as among Abstract Bearded seals produce complex sounds avian species, research is focused on the male. underwater mainly in breeding season. These sounds Relatively few studies are available regarding vocal are suggested to have territorial and/or reproductive dimorphism, and even fewer investigate the poten- function. However, little is known about the beha- tial influence of evolutionary constraint. I present vioral context of the vocalization mainly because a phylogenetic study of stereotypic vocal behavior behavioral observation is difficult in the wild. In this in the peromyscine mice, an assemblage of deer study, we recorded behaviors and sounds of an adult mice and their relatives (Cricetidae: Neotominae). male and two adult females in captivity to estimate The behaviour of both males and females of 15 spe- the function of vocalization. Underwater sounds cies was sampled, with each of the major species were recorded in Otaru aquarium, Japan, in the groups represented. Stereotypic vocalizations vary daytime between March 2012 and May 2013, using among species, and include single note pure-toned a hydrophone and a linear PCM recorder (frequency constant frequency (CF) vocalizations, harmonically range: 20 Hz to 20 kHz). Additional 24-hour recor- rich CF vocalizations with non-linear elements, dings were conducted in March 2015 to find diurnal simple frequency modulated (FM) and temporally change in vocal activity. Behavioral observation was modulated (TM) vocalizations, and complex songs conducted in the daytime to identify the caller of that are both strongly FM and TM. Both the pre- sounds and to record social behaviors. The adult sence and degree of sexual dimorphism in acoustic male vocalized from December until April with a features varies among species. This is particularly peak in March. Contrary to previous belief, two fema-

40 les also vocalized from mid March to April, breeding recovered even at distances greater than a hundred season reported in the wild. In order to examine meters. Male calls show stronger discriminability whether they conduct “vocal exchange” following at long distances than female calls, and this dif- some temporal rules, we analyzed the intervals of ference can be explained by the more pronounced vocalization between two seals that vocalized se- frequency modulation found in their calls. Operant quentially. The frequency distribution of the intervals conditioning experiments showed that female zebra were significantly different from that expected from finches were proficient in discriminating between the assumption that seals independently vocalized calls of individual males propagated at long at their own pace (P < 0.01, Kolmogorov-Smirnov distance, and even more so when they can train test) and had a sharp peak at 4.0 sec., suggesting themselves with increasingly degraded signals over “vocal exchange” between two seals. The 24-hour time. While this result emphasizes the importance recordings revealed that seals vocalized all day and of perceptual learning, electrophysiological experi- night in the breeding season, peaking in the early ments also highlighted single neurons that showed morning. No social behaviors were observed except high performances in discriminating unfamiliar muzzling by females against the throat of vocalizing individuals at long distance. male in March, which might be a signal to advertise their estrus. Our findings provide the first evidence that bearded seals use sounds for vocal exchange Friday | Sept 11th 10.00 – 10.20 that might be useful for long-range communication to advertise their reproductive status. Odom, Karan J. Omland, K. E. Monday | Sept 7th 16.10 – 16.30 University of Maryland, U.S.A. Logue, D. M. University of Lethbridge, U.S.A. Mouterde, Solveig C. Elie, J. E. Similarities and differences in female and Theunissen, F. E. male song in troupials (Icterus icterus): University of California Berkeley, U.S.A structural, diel, and seasonal comparisons Mathevon, N. Université Jean Monnet, St. Etienne, France Abstract Recent research emphasizes that female song in birds is more common than previ- Active space of individual signatures in ously thought. Yet we still know very little about propagated calls of zebra finches: From structural, temporal, or functional similarities acoustics to perception and differences between male and female song, let alone the evolutionary processes operating on Abstract Reliable transmission of acoustic song in each sex. Song in temperate-breeding male information about individual identity is of critical passerines appears to be sexually selected, but the importance for pair bond maintenance in numerous selection pressures acting on song in females may monogamous songbirds. However, information be quite different. Quantifying structural, diel, and transfer can be impaired by environmental seasonal variation between male and female song is constraints such as external noise or propagation- an important first step in forming hypotheses about induced degradation. We focused on the impact of the function and evolution of female and male song. these constraints on the active space of bird calls, We compared song structure and singing rates for and their ability to extract identity information from female and male Venezuelan troupials (Icterus highly degraded signals. We analyzed propagated icterus) for dawn versus day and for breeding versus signals using discriminant function analyses on a non-breeding seasons. Female and male songs were set of analytical parameters as well as on a com- similar in duration and frequency. Singing rates, plete spectrographic representation of the signals. however, varied substantially between males and We found that individual signature is remarkably females, particularly during dawn versus day time resistant to propagation as caller identity can be singing. Males sang the majority of dawn songs

41 abstracts of symposium speakers

during the breeding season, whereas females sang within the Mediterranean. Using Discriminant the majority of day time solos. Furthermore, fema- Function Analysis (DFA), 69% of Mediterranean les sang at similar rates during both the breeding whistles were classified to the correct species. This and non-breeding season. It is important to note, correct classification score was significantly greater during the day, most songs were answered by the than chance (χ2 test, α = 0.05). In contrast, the mate to create duets. Our finding that females sing correct classification score for the tropical Pacific more solos during the day year-round while males (55%) was not significantly greater than chance (χ2 sing more solos at dawn during the breeding season test, p = 0.37). Thus, for these species, differences indicates that female song may be functionally within the tropical Pacific are smaller than differen- distinct from male song in troupials. Such findings ces within the Mediterranean. These results suggest reinforce the need to explore alternative selection that the measured variables do not carry species- pressures contributing to the evolution of female specific information for the tropical Pacific dolphins song in birds. and that these dolphins may be using other whistle characteristics for species recognition. Delphinid repertoires are strongly influenced by vocal learning Wednesday | Sept 9th 11.20 – 11.40 such as in the development of individually distinctive signature whistle types, and playback experiments have shown that animals sometimes react to whistles Oswald, Julie N. of other species as if they came from conspecifics. Bio-Waves, Inc., Encinitas, U.S.A. We suggest that the development of many indivi- Rankin, S. dually distinctive signature whistle types may blur NOAA Fisheries, Southwest Fisheries Science species differences in areas where dolphins maintain Center, La Jolla, U.S.A. individualized social networks. Gannier, A. Fuchs, S. Group de Recherche sur le Cetaces, Antibes, Thursday | Sept 10th 16.15 – 16.30 France Janik, V. University of St. Andrews, U.K. Ota, Nao Soma, M. Acoustic species recognition in delphinids Hokkaido University, Japan Gahr, M. Abstract Interspecific variation in vocalizations is Max Planck Institute for Ornithology, Seewiesen, commonly used for species recognition, especially in Germany animals that have limited access to information from other sensory modalities. Delphinids are such a case Tap dancing birds: the multimodal mutual where acoustic information is often the only modality courtship display of males and females in a available. Yet, species differences in vocalizations socially monogamous songbird are often subtle. We studied species differences in vocalizations of two delphinid species (Delphinus Abstract Many birds perform courtship displays delphis and Stenella coeruleoalba) in the tropical that combine movements and sounds. Past song- Pacific Ocean and Mediterranean Sea to evaluate birds studies shed light on vocalization, while less the differences that exist between and within attention has been paid to non-vocal sounds. We species. Fourteen variables describing frequency, report here the first example of non-vocal sound shape and duration of whistles were measured and production of the both sexes of a songbird, the blue- compared. Between-species comparisons showed capped cordon-bleu (Uraeginthus cyanocephalus). that the degree of difference varied with location. This species is known for female song, but in addi- Few significant differences were found between tion to singing, both sexes perform courtship dance whistles produced by D. delphis and S. coeruleoalba by holding a piece of nesting material and bobbing in the tropical Pacific. In contrast, many significant up and down. Interestingly, the bobbing produces differences were found between these same species rhythmical sounds. By recording these displays with

42 a high-speed video camera, we discovered that their behaviour in a multimodal communication context. visual courtship display includes incredibly rapid We visually exposed territorial M. zebra males to a step-dancing during bobbing. Although we cannot size-matched male placed in a contiguous aquarium. deny the possibility that such stepping behavior Trials consisted in three periods of five min: PRE (vi- functions as visual and/or tactile (i.e. vibration) sual only), PBK (acoustic + visual) and POST (visual signals, it is also possible that it produces acoustic only). Acoustic treatment consisted in the playback signals due to higher sound pressure level. As a first of agonistic sounds of smaller (high frequency) and step toward understanding the function of the dance larger (low frequency) males. As a control we used display in blue-capped cordon-bleus, we investiga- white noise or silence. We observed that low fre- ted sex differences of dance performance (no. steps quency sounds but not the other treatments resulted / bob, and bobbing rate). We also examined the in a reduction of aggression during PBK and POST effects of singing and partner’s position on dance periods, suggesting that sound frequency provides performance. Sex differences in dance performance effective information of fish size. Theory predicts that were marginal as previously reported in song. Dance reduced aggression and shorter contests are expected performance changed depending on whether the when asymmetries between opponents are large. Our subjects were singing, which might contribute to results also suggest that fish evaluate the relative avoiding interference between vocal and non-vocal information provided by visual and acoustic signals sounds. However, stepping dance can also serve as during contests. When sound frequency indicates a a vibration signal because they intensified dance larger opponent (asymmetric contest) the acoustic performance when partners were on the same perch. channel overrides the visual one, but when it indica- The fact that dance display of both sexes of socially tes a smaller opponent the visual channel seems to monogamous songbirds produces non-vocal sounds prevail as fish behaved as in a matched contest. This in addition to song suggests that the evolution of study indicates that the auditory and visual channels multimodal and multicomponent courtship display as interplay in a complex way during the evaluation of an intersexual communication should be considered. an opponent to avoid ‘dangerous’ conflicts.

Friday | Sept 11th 16.15 – 16.30 Tuesday | Sept 8th 13.50 – 14.10

P. Amorim M. Clara Parks, Susan Fonseca, P.J. Denes, S. L. MARE, ISPA-Instituto Universitário, Lisboa, Varshney, P. Portugal Syracuse University, U.S.A. Beauchaud, M. Fristrup, K. Université Jean Monnet, St. Etienne, France US National Parks, U.S.A.

Multimodal and mismatching communication Ecological insights from bioacoustics activity in vocal fish levels measured from long-term acoustic monitoring at NEON observatory sites Abstract Agonistic contests are often mediated by multimodal communication but the way in which Abstract The National Ecological Observatory visual and acoustic signals interplay is not yet un- Network (NEON) has established a framework derstood in fish. Previous experiments with the Lake through which a variety of environmental metrics Malawi cichlid fish Metriaclima zebra have shown will be continuously measured for multiple decades that visual but not acoustic unimodal signals trigger at stations located across the United States. Here male aggression. However, when acoustic signals we describe a multiyear project to demonstrate the were combined with visual information they reduced benefits of incorporating continuous acoustic moni- aggressiveness in this species. Here we tested toring at NEON sites. Long-term recordings provide how the spectral information content of agonistic a wealth of data relating to species presence, animal sounds, which is size-related, modulates aggressive behavior, and anthropogenic noise disturbance with

43 abstracts of symposium speakers

minimal environmental intrusion. The phenology research in Italy is aimed at describing the acoustic of biological events such as avian, amphibian or environment (soundscape) of natural habitats with insect chorusing activity can be determined through different level of protection and human presence. automated band limited energy detectors. Data from The program started in May 2014 in 4 locations by the NEON site at Harvard Forest, Massachusetts, using autonomous recorders to measure the sounds collected between November 2013 and July 2015 produced by the animals (biophony, expressed were analyzed to assess the phenology of chorusing as biodiversity, complexity and richness) and the behaviors of bird species using customized auto- anthropogenic noise (antropophony, expressed as mated detectors. Bird vocalizations are detectable quietness). The project is supported by the Italian throughout the year, however during the spring, a Forestry Police (Corpo Forestale dello Stato, Ufficio change in vocalization behavior was clearly detec- per la tutela della Biodiversità - CFS UTB) that pro- table in the acoustic record. Assuming a qualitative vides logistic support and the access to the Integral definition of chorusing as the persistent presence of Nature Reserve “Sassofratino”, in the National bird vocalizations, the onset of bird chorusing can be Park “Foreste Casentinesi”, where the core of the identified from long-term spectral plots of the acou- research is run. This RNI, created in 1959, holds stic data and onset can be compared between years. a pristine forest, received the Europen Diploma Assessment of acoustic activity at multiple time and is taken into consideration for the UNESCO scales can provide multiple levels of data. Adding “Millenarium Forests” program. In year 2015 the acoustics to the measurements already collected un- research is mainly focused on this area with 3 re- der the NEON protocol can provide high-resolution corders coupled by the measure of temperature and information on the presence and behavior of acou- humidity, the two main parameters, with the daily stically active species in birds, frogs, and insects, solar cycle, that drive the acoustic behaviour of ani- as well as the acoustic impact of human activity at mals. Results given by (bio)acoustic indexes taken these locations and allow for long-term monitoring from literature and new indexes developed for the combined with ground truthing assessments of purpose will be compared with compact long-term biodiversity and ecosystem health from standardized spectrograms and with the analytic results given by traditional data collection methods at each site. experts. Instrumentation, software and technical details will be also presented and discussed in order to define common protocols and metrics for Tuesday | Sept 8th 14.10 – 14.30 soundscape analysis.

Pavan, Gianni Thursday | Sept 10th 16.30 – 16.45 Favaretto, A. Bovelacci, B. Scaravelli, D. Pelicella, Ivano Macchio, S. Balsamo, F. Università di Pavia, Italy Dodotronic, Castel Gandolfo, Italy Glotin, H. Pavan, G. Université de Toulon, France Università di Pavia, Italy

SABIOD Italy: A pilot study of wild areas‘ WaveFalcon, a new soundscape smart recorder soundscapes. The RNI Sassofratino pilot study Abstract Recording soundscapes is mostly based Abstract The research was developed within the on autonomous remote recorders to be programmed SABIOD (Scaled Acoustic Biodiversity) project gran- for a given recording schedule, e.g. 10 minutes ted by the French CNRS MASTODONS Project on Big every half-hour, deployed in the location to be mo- Data acquisition, analysis and processing. SABIOD nitored, and then recovered to download data and started in 2014 in cooperation with the University replace batteries. Their use implies several techni- of Toulon (DYNI laboratories) and the JASON (Joint cal constraints, mainly related to storage and power. Acoustic Survey for Online Biodiversity) Project. The Nowadays the storage space is not the main issue

44 as SDXC cards reach 512GB of capacity; batteries round and can lead to a biased interpretation of are the most limiting factor and their replacement the biological sounds. Indeed, most of geophonies must be programmed prudentially to guarantee the and anthropophonies have a low frequency range, continuity of the data acquisition. The WaveFalcon i.e.less than 1kHz, which overlaps with the fre- is a smart autonomous low-power recorder able to quencies used by fishes for communication and work either in stand-alone mode or web connected. other biologically important activities. In the pre- Several web connection modes are available to ba- sent study we used the Acoustic Complexity Index lance battery usage and available bandwidth on the (ACI) to elaborate one year of recordings (from 22 wireless channel that can be a Wi-Fi or a cellular June 2013 to 6 June 2014) carried out in shallow network, possibly 3G or faster. To limit bandwidth coastal waters inside the Marine protected Area and battery usage, the wireless connection can be of Lampedusa, Italy. Data were collected using an limited to monitoring status (storage space availa- autonomous recorder (SM2, Wildlife Acoustic, US; ble, battery level, scheduling of recording and of recording bandwidth: 2Hz – 192kHz; sensitivity: web access) to allow timely maintenance operations -165dB re: 1 V/uPa). The ACI was chosen since it (data recovery and battery replacement), or it can is an algorithm designed to measure the spectral be expanded to perform scheduled fpt of summary complexity of soundscapes, which tends to enhance data or of recorded data, including real-time or the intrinsic variability of biotic sounds, and filter/ near-real-time display on a public page. Recording reduce the more constant intensity values that cha- of audio and ultrasonic data from different sen- racterize many types of geophonies or anthropopho- sors can be continuous, scheduled, or on trigger. nies. The ACI was previously widely used to track WaveFalcon provides temperature, humidity and the dynamics of terrestrial communities, but it was light level recording as well as photo and video cap- only recently applied to marine environments. Pros ture possibilities. An onboard GPS allows mobile and cons of its extensive temporal application on applications and the synchronization of a cluster of one entire year of recordings of a marine sounds- units for sound source localization and wide area cape are here described and evaluated. monitoring. Examples of use in the SABIOD-Italy project will be presented. Thursday | Sept 10th 15.10 – 15.30

Tuesday | Sept 8th 13.30 – 13.50 Ptacek, Ladislav Machlica, L. Pieretti, Nadia Linhart,P. Ceraulo, M. Jaska, P. Mazzola, S. Muller, L. Filiciotto, F. University of South Bohemia, Czech Republique Maccarrone, V. Farina, A. Application of speaker recognition methods for Buscaino, G. chiff-chaff individual identification University of Urbino, Italy Abstract This paper deals with the identifica- Application of a recently introduced acoustic tion of chiff-chaff individuals using the Gaussian index in marine soundscapes of a shallow Mixture Model and Universal Background Model water coast (Lampedusa Island, Italy) (GMM-UBM) supplemented by a Voice Activity Detection (VAD) algorithm. Contrary to usual Bird Abstract Quantifying and characterizing the Individual Identification systems the one introduced acoustic production of animals in marine sound- here processes the live recordings without editing scapes can be sometimes a challenging task to (cutting, song or syllable extracting, de-noising). address. Geophonies (sounds from wind, rain The testing recordings were made outdoors under and waves) or anthropophonies (fishing activities, natural conditions and thus contain unwanted cruise ships and other vessels) are present all year sounds and noises: crackling, aircraft engines,

45 abstracts of symposium speakers

blowing, simultaneous bird singing of the same First, audio-motor synchronization experiments or different species, etc. The method uses MFCC are being performed in an increasing number parametrization with the added VAD algorithm for of species. Mapping synchronization abilities automatic recognition of bird song activity. The to species either closely related by descent or identification runs on the open set, which means occupying similar niches will help reconstruct the that any number of unknown birds may appear evolution of rhythm not only in those species, but anytime during the process as is common in the also in humans. Second, interacting individuals can nature. The achieved success rate of the identifi- influence each other’s timing, both across several cation varies between 69.6% and 93.5%, and the modalities and in different animal taxa. A large overall achieved accuracy is 78.6%. The results body of observational and experimental research suggest that the task of automatic identification of on animal chorusing is already available. Once its individual birds, even within one species, is feasi- focus is shifted from the spectral to the temporal ble in natural conditions. Main goal of the research domain, published research on animal spontaneous was to prove the method of Automatic Individual chorusing can inform hypotheses of rhythm evolu- Identification using GMM-UBM with VAD could tion in humans. Finally, empirical findings should supplement and even supersede the bird ringing be synthesized and combined via mathematical using real recordings without post-editing. models of the evolutionary dynamics of rhythm. These models could resemble classical models of behavioural traits in evolutionary biology, must be Friday | Sept 11th 13.50 – 14.10 informed by empirical bioacoustics, and should aim at generating additional testable hypotheses.

Ravignani, Andrea We acknowledge MASTODONS CNRS mission Vrije Universiteit Brussel, Belgium for supporting SABIOD http://sabiod.org & Univ Toulon Axe Information for supporting JASON. The evolution of rhythm: a comparative approach to speech and music Monday | Sept 7th 15.10 – 15.30 Abstract The human sense of rhythm has traditionally been studied along two dimensions, stemming from language sciences and experimen- Ringler, Max tal psychology. Linguists have performed acoustic Hödl, W. analyses of phonological patterns in languages, University of Vienna, Austria comparing syllable stress and timing within and Ringler, E. between human languages. Experimental psycholo- University of Vet. Medicine, Vienna, Austria gists interested in music have performed a variety of rhythm perception and synchronization experi- Acoustic ranging in poison frogs – Allobates ments, where human participants were required femoralis uses signal cues beyond intensity to to listen to series of marked events over time, and assess caller distance possibly synchronize their movements by e.g. finger tapping. The little comparative research relating Abstract Acoustic ranging – to properly asses the non-human animal communication and perception distance of a sound source, like a competing male to speech and music has mostly focused on spectral – is especially important in the context where indi- rather than temporal features. Here, I: (i) review viduals live in highly territorial systems. There it is and compare a number of contrasting hypotheses often used as a mechanism to maintain individual on the evolution of rhythm; (ii) propose crucial spacing to avoid costly physical conflicts between experiments and model species to support or reject neighbouring individuals. At the same time accu- these evolutionary hypotheses; (iii) suggest how rate ranging allows territory holder to make reliable three complementary approaches (observational, decisions on when to engage other proximate callers experimental, and mathematical) can elucidate the that might be competing for space and mates. As evolutionary history of the human sense of rhythm. signal intensity is not a reliable cue for acoustic

46 ranging – a caller might not only be near or far, but then plain wrens must pair with individuals from also could vocalize more or less loud, be blocked the opposite sex that already share their specific by obstacles, or vocalize directionally – accurate codes. Such a pairing scheme seems too restrictive, ranging requires the integration of further signal so it seems more likely that the wrens can learn new cues. For birds and mammals, including humans, duet codes as adults. We performed a removal ex- it has been shown previously, that reverberation periment in adult plain wrens to determine whether characteristics and signal degradation are used to individuals are able to acquire new answering rules for acoustic distance assessment. In frogs so far as adults. We found that adult plain wrens are able clear evidence is missing, that signal characteris- answer to unfamiliar phrase types and when they tics other than signal intensity are used to asses do, their coordination is poorer than for familiar signaller distance. To asses intensity independent phrases. This suggests that individuals can potenti- ranging in frogs, we performed playback with the ally modify their duet codes during adulthood. If so, Neotropical poison frog Allobates femoralis. Males then what is important for juveniles to learn may be of this species are known for their stable, long- the general rules governing duet codes, rather than term territories, which are announce by prolonged the specifics of the codes used by their models. advertisement calling. We used naturally degraded and reverberated signals which were recorded simultaneously across 0.75, 1.5, 3, 6, 12, and 24 m Tuesday | Sept 8th 16.30 – 16.45 in Neotropical rainforest, the natural habitat of the study species. These signals were normalized and played back at equal intensity from equal distances Roch, Marie to territorial males. The significantly differential re- Herbert, S. sponses of the focal males to these playback clearly San Diego State University, U.S.A. shows, that also frogs can perceive and evaluate Baumann-Pickering, S. signal characteristics beyond intensity to evaluate Hildebrand, J.A. the distance of a sound source. Širović, A. Scripps Institution of Oceanography, La Jolla, U.S.A. Tuesday | Sept 8th 16.15 – 16.30 Berchok, C.L. Cholewiak, D. Oleson, E.M. Rivera Caceres, Karla van Parijs, S. Searcy, W.A. Soldevilla, M.S. University of Miami, U.S.A. NOAA National Marine Fisheries Service, Fisheries Guerrero, E.Q. Science Centers, U.S.A. University of St. Andrews, U.K. Risch, D. Scottish Association for Marine Science, Argyll, Duet code development in plain wrens U.K. (Cantorchilus modestus zeledoni) Management of acoustic metadata Abstract One of the most striking properties of for bioacoustics avian duets is the occurrence of answering rules (“duet codes”) that individuals use to link their own Abstract Recent expansion in the capabilities song types to those of their partners. The mecha- of passive acoustic monitoring of sound-producing nisms by which such rules are acquired have been animals is providing expansive data sets in many little studied. In particular it is not clear at which locations. These long-term data sets will allow the stage individuals learn their duet codes. In a pre- investigation of questions related to the ecology of vious study with juvenile plain wrens (Cantorchilus sound-producing animals on time scales ranging modestus zeledoni) we found that individuals learn from diel and seasonal to inter-annual and decadal. a code during early development. However, if codes Analyses of these data often span multiple analysts acquired during early development do not change, from various research groups over several years of

47 abstracts of symposium speakers

effort, and as a consequence have begun to gene- “black box” approach captures features that vary rate large amounts of scattered acoustic metadata. genetically among social groupings or host plants, It has therefore become imperative to standardize providing great analytical power. We find that social the types of metadata being generated. A critical groupings and host plants generate varied feedback aspect of being able to learn from such large loops between signals and preferences as causes varied acoustic data sets is providing consistent and targets of selection. For example, experience and transparent access. This is juxtaposed with of signaling modifies female mate preferences, the need to include new information for specific which is expected to influence variation in signals research questions that evolve over time. Hence, a and in turn alter females’ experience of signaling, method is proposed for organizing acoustic metada- and so on. Some feedback loops are stabilizing and ta that addresses many of the problems associated maintain variation. Others may lead to runaways. with the retention of metadata from large passive For example, we find strong signal-preference phe- acoustic data sets. A structure was developed for notypic covariance among treehoppers that develop organizing acoustic metadata in a standardized on different host plants genotypes (and in different manner along with a mechanism that implements social groupings on different plants). Thus, con- the data representation into a client-server oriented ditions that are common in nature (variation in database. The implementation was extended to social and other biotic aspects of the environment) access Internet available data products, permitting may often generate divergent sexual selection and access to a variety of environmental information promote speciation. (e.g. sea surface temperature, ephemeris, etc.) from a wide range of sources as if they were part of the data service. This metadata service is in use at Tuesday | Sept 8th 11.00 – 11.20 several institutions and is used track and analyze millions of acoustic detections. Sarria-S, Fabio A. Jonsson, T. Tuesday | Sept 8th 11.20 – 11.40 Montealegre-Z., F. University of Lincoln, U.K.

Rodriguez, Rafael Non-invasive measurement of travelling waves University of Wisconsin-Milwaukee, U.S.A. in the inner ear of Neotropical bushcrickets

Environmentally-induced runaways: vibrational Abstract In mammals, frequency analysis of communication in Enchenopa treehoppers acoustic signals is possible due to the mechanical (Hemiptera Membracidae) anisotropy of the basilar membrane (BM) in the cochlea. Travelling waves (TW), propagating inside Abstract Insects that communicate with subst- the cochlea fluid-filled cavity, generate amplitude rate-borne vibrational signals lead rich and complex maxima responses at frequency-specific locations lives. Juveniles and adults cooperate and compete along the BM, providing a spatial map of sound using varied signal repertoires, and their interac- frequencies. Physiological and biophysical pro- tions unfold on complex settings, such as social perties of TWs in the mammalian cochlea remain groupings on their host plants. Here I summarize elusive, yet are vital to understand tonotopy and work by my research group on variation in mating active amplification. In fact, a valid question is signals and mate preferences arising from insect whether TW actually exist in the mammalian inner social and host plant environments. Our study ear. Frequency analysis has also been observed in animals, Enchenopa treehoppers, are plant-feeding the hearing system of acoustically communicating vibrational insects wherein species evolve as novel insects through non-contact mechanical measure- host plants are colonized and signals and preferen- ments using Laser Doppler Vibrometry (LDV). For ces diverge. We use treehopper and plant genotypes instance, in locusts, frequency analysis is directly as treatments to assess the variation that they achieved in the tympanal membrane, while in bus- induce in treehopper behavioral phenotypes. This hcrickets, it is achieved through a more complex

48 process, using a cochlea-like sensory transduction workshop organ constituted by the crista acustica (CA) and the associated auditory vesicle (AV). TWs have been observed in the CA-AV of two bushcricket species Thursday | Sept 10th 15.30 – 17.30 using different in vivo experimental methods and (parallel session) LDV. One is invasive as it requires the rupturing of the dorsal cuticle protecting the hearing organ for CA’s surface scans. The other one takes advantage of the transparency of the cuticle of some species, WORKSHOP: ANALYSIS OF ANIMAL allowing access to the CA-AV structure in a non- ACOUSTIC SIGNALS IN R invasive manner. This study focuses on the diversity of TW in the ears of different tropical species accessing the CA-AV in a non-invasive approach. This workshop is intended for students and re- Using LDV and a range of pure tones stimulation, searchers interested in analyzing the structure we measured and compared the biophysical of animal acoustic signals. The workshop will principles of TWs in living and intact specimens. focus on functions from the R package warb- Demonstrating the variability of travelling waves in leR. This package is designed to streamline the ears of different species as a response to sound bioacoustic analysis, providing users with a frequency and distance. Finally, Bushcrickets’ workflow that facilitates automated spectro- TWs and tonotopy response, suggest an alternative graphic visualization and acoustic parameter experimental model to study processes otherwise measurements. impossible to visualise in the mammalian cochlea.

The functions in warbleR allow users to download recordings from the Xeno-Canto Thursday | Sept 10th 14.30 – 14.50 online repository, create maps of recording locations, automatically (or manually) detect signals, print spectrograms of full recordings Rowe, Karen or individual signals, and perform acoustic Adams, A. L. measurements. In addition we will review Museum Victoria, Melbourne, Australia workflow procedures for quantitative (acoustic parameters) and qualitative analyses (visual Automated recognition software improves classification). detectability for a range of bird species’ vocalizations Basic familiarity with the R environment is recommended. Participants are expected Abstract Effective conservation and manage- to bring their computers and install R and ment planning requires accurate assessment of Rstudio in advance. R codes and sample sound species’ presences across their distribution, but files for running the analyses will be provided. traditional survey methods can be prohibitively The workshop will focus on bird vocalizations, costly. Automated recording devices (ARDs) are but the methods and procedures are also increasingly used in place of observational tech- applicable to signals from other taxa. niques to survey avian communities within a region. Emerging automated pattern recognition technology has seen the development of software where users can generate species-specific recognizers theoretically capable of autonomously scanning the large volumes of collected audio recordings to detect vocalizations of a target species. We investigated the use of the Song Scope (Wildlife Acoustics, Maynard, MA, USA) bioacoustics monitoring software as a survey tool for generating

49 abstracts of symposium speakers

species-specific recognizers to autonomously scan threatened Military Macaw (Ara militaris), among audio files collected by ARDs, and evaluated their three populations in a modified landscape along ability to correctly detect a range of species with the coast of Jalisco, Mexico. Under the vicariance varying vocalization types when compared to liste- and drift model we predicted a discrete pattern of ning and visually scanning spectrograms. Although divergence in calls between populations suggesting recognizers varied among species in their ability to that habitat modification restricts movements of correctly discern vocalizations of our target species macaws, whereas in the isolation by distance model and were generally associated with large false po- we predicted a gradual clinal pattern of similarity sitive rates, we found that recognizers, particularly in calls, indicating no influence of habitat modifi- second pass recognizers which incorporated known cation. We evaluated call features among sites and vocalizations identified in the field recordings by with distance by comparison of specific parameter the first pass recognizer, detected more species measures using Principal Components Analysis in more surveys and sites with consistently higher with ANOVA, and similarity values obtained with detection probabilities than listening and visually spectrogram cross-correlations and Mantel tests. scanning spectrograms. Our results highlight the Acoustic parameter analysis found a significant potential of using ARDs coupled with automated difference in one principal component between species recognizers as an effective survey method the two furthest sites, while the spectrogram for a range of species’ vocalization types. The abi- similarity analysis showed an association of calls lity to autonomously detect species has significant with site, and decreased similarity in calls with implications for informing conservation decisions distance. This demonstrates a clinal differentiation and management protocols, and we suggest ways of acoustic signals with distance, supporting the to improve the efficacy of self-generated automated Isolation by Distance model. Our results suggest recognizers for a range of users. that large-bodied non-oscines, such as the Military Macaw, may be able to maintain homogeneity of culturally transmitted traits in modified landscapes Thursday | Sept 10th 11.20 – 11.40 through population processes of dispersal and cultural diffusion. Therefore species-specific dispersal capabilities may drive the likelihood that habitat Salinas-Melgoza, Alejandro modification leads to call divergence in species Centro Tlaxcala de Biologia de La Conducta, with vocal learning capabilities. México Renton, K. Estacion de Biologia Chamela IB-UNAM, México Friday | Sept 11th 17.00 – 17.15

The military macaw maintains homogeneity of a contact call across populations over a Salvin, Pauline modified landscape in western Mexico Leboucher, G. Amy, M. Abstract Anthropogenic land-use change may Université Paris Ouest Nanterre La Defense, create barriers that restrict the movement of France individuals among populations, thereby disrupting population processes that would geographically Functions of female signals: the copulation homogenize culturally transmitted acoustic signals. solicitation and the mate sampling aid hypo- This may lead to divergence in vocalizations, po- theses in the domestic canary tentially creating behavioral barriers, and may be more severe for threatened species with reduced Abstract Since Darwin, studies on the sexual populations. However, few studies have evaluated selection theory mainly focused on male secondary the impact of habitat modification on geographic sexual characters. For instance, studies on animal variation in acoustic signals of large-bodied, communication in songbirds have mainly focused wide-ranging species. We evaluated two models on male song. Yet, female vocalizations in song- for geographic variation in a contact call of the birds are growing in interest among behavioural and

50 evolutionary biologists because these vocalizations conserved molecular toolkits that act as genetic can reveal the female’s preferences for male traits modules. These may specify the neural circuits and may affect male display. Two studies were subserving these particular behaviors, and organize designed to test whether female domestic canaries their function. Elucidating these genetic modules in (Serinus canaria) can signal their preferences and different animal models may inform the evolution of whether these different signals can incite males to language and other complex traits. display. In a first experiment, female canaries were daily exposed to three types of song performance, differing in the repetition rate of sexy syllables. Tuesday | Sept 8th 11.40 – 12.00 This experiment revealed that females are engaged in multimodal communication during sexual interaction. Female used Female-Specific Trills as well Schmidt, Arne K. D. as Copulation Solicitation Display to indicate their Römer, H. readiness to copulate. These results support the University Graz, Austria copulation solicitation hypothesis for the Female- Specific Trills. In a second experiment, still in No evidence of acoustic signal divergence in a progress, male canaries were daily exposed to visual tropical cricket assemblage and acoustic female canary signals: two different female calls (a simple call and a Female-Specific Abstract Efficient acoustic communication in Trill) and a female in Copulation Solicitation Display multispecies assemblages will strongly depend on (or not). The results will be discussed in light of the the degree of signal overlap. Masking interference sexual selection framework. and signal confusion of similarly structured signals can affect detection, recognition and localization, and thus might impose fitness costs. Therefore, it Monday | Sept 7th 12.00 – 12.20 is widely believed that masking interference should drive signal divergence to reduce overlap in signal space and to promote intraspecific communication. Scharff, Constanze In this study we examined the role of acoustic com- Free University Berlin, Germany petition in a tropical cricket community and asked whether call structures of species that co-signal at Molecular toolkits for control of vocal the same time and place are more dissimilar than production species that are spatially and/or temporally separated. Moreover, we examined whether species with Abstract Spoken language and birdsong share higher spectral overlap of their calling songs show a number of striking parallels. Comparing the higher dissimilarity in their temporal call structures biologically tractable cognitive abilities necessary than those species that have low levels of spectral for language and for birdsong is a fruitful endeavor overlap. We used non-metric multidimensional to identify, which properties are shared and which scaling (NMDS) and pairwise comparison of acoustic are unique to each. I will review evidence for the similarity/dissimilarity and found no significant relevance of the FoxP2 gene and its associated difference in acoustic signals for species pairs molecular network for speech and its role in mo- co-signaling or showing higher spectral overlap dulating variability in the songbird basal ganglia compared to species that were spatially/temporally circuit relevant for the acquisition and production segregated or had low levels of spectral overlap, res- of birdsong. However, I will argue that the simila- pectively. Therefore, for the acoustic cricket commu- rities between human language and songbirds are nity investigated here we found no evidence for the not limited to sensorimotor processes – but may hypothesis that masking interference is a significant extend to other structural and functional properties. driver of signal divergence. Alternatively, we suggest Many questions regarding the similarities between that a suite of sensory/neuronal mechanisms, such spoken language and birdsong remain unanswered, as narrow filtering, spatial release from masking, or but increasing evidence suggests that human and gain control processes play a more important role to non-human communication systems may rely on ensure reliable communication under noise.

51 abstracts of symposium speakers

Tuesday | Sept 8th 10.00 – 10.20 changes of time constants in the networks for signal generation and recognition during speciation.

Schöneich, Stefan Hedwig, B. Wednesday | Sept 9th 10.40 – 11.00 University of Cambridge, U.K.

Neurobiology of acoustic communication: Slocombe, Katie E. Neural networks for singing and song pattern University of York, U.K. recognition in crickets Flexibility in chimpanzee vocal communication Abstract From human language to birdsong and the chirps of insects, acoustic communication is Abstract Non-human primate receivers have the information transfer by sound signals. Male demonstrated considerable flexibility to learn crickets use species-specific calling songs to associations between vocalisations and events attract conspecific females. They produce the and to integrate contextual information and exis- genetically determined sound pulse pattern by ting knowledge with information extracted from rhythmic movements of their forewings. By intra- vocalisations to respond to conspecific calls in cellular recordings in the CNS of singing males I an adaptive manner. In contrast, it has often been identified key neurons of the central pattern gene- assumed there is little complexity or flexibility rator (CPG) that drives the motor pattern for sound in the production of non-human primate voca- production. The CPG is located in the anterior lisations. Recent findings in chimpanzees are, abdominal neuromers and comprises interneurons however, beginning to challenge this assumption. showing mutual inhibition and postinhibitory Firstly, field experiments presenting wild chim- rebound (PIR). Recordings in females revealed a panzees with model snakes have shown that some small network of auditory brain neurons that re- of their alarm calls are selectively produced, cognizes the species-specific sound pulse pattern directed at others and goal directed, rather than and exhibits properties fundamental to a feature simply emotional responses to the stimulus. detection based on delay-line and coincidence- These alarm calls meet the same criteria previ- detection mechanism. An internal delay that ously used to identify first order intentionality matches the pulse period of the male calling song in child and ape gesture production. Secondly, is provided by a non-spiking neuron. Upon acou- whilst chimpanzees, like other non-human prima- stic stimulation it receives transient inhibition tes, cannot generate new vocalisations, a recent that triggers PIR. Direct and delayed excitations study has shown that there is some flexibility converge in a coincidence detector neuron, which for chimpanzees to alter the structure of their responds best to the pulse pattern of the male referential food calls to match those of close song. The pattern selective output of the feature social partners, independently of their preference detector circuit closely matches the tuning of the for the food. Together, these studies indicate that female’s phonotactic behavior. For effective com- chimpanzees have more control and flexibility munication the sender and the receiver need to over both the structure and usage of their calls be attuned to the same signal and the generation than previously assumed. and recognition of the song pattern are genetically coupled in crickets. Modifications of the membrane conductance that set the time constants of the Tuesday | Sept 8th 14.50 – 15.10 PIR are likely to adjust the phonotactic selectivity in species that use different pulse rates. Delayed excitation by PIR is not only crucial for song pat- Stowell, Dan tern recognition but also essential in the singing Queen Mary University, London, U.K. pattern generating network. Therefore, coupling at the level of membrane proteins that are controlled Classifying multiple bird species in by the same genes seems to ensure concomitant crowdsourced smartphone audio

52 Abstract Given an amateur sound recording elements not heard before, or new combinations from a smartphone app, it is challenging to auto- of familiar elements. The responses of the zebra matically recognise the set of bird species that are finches showed that they seem to discriminate active in the recording. I describe our method for the strings by attending to the ordinal positions multilabel classification of bird sounds, which was of elements within the strings. However, the bud- validated through the BirdCLEF challenge on 500 gerigars generalized their discrimination to strings and 1000 tropical bird species, and then deployed consisting of new sounds, thus revealing evidence as a popular smartphone app for UK birds. I dis- of having formed an abstract representation of cuss unsupervised feature learning and classifier the grammar underlying the training strings. This configuration, and also illustrate some unexpected implicates that although the ability to abstract and aspects of the „soundscapes“ captured on mobile generalize structural rules is not readily found in phones. the animal kingdom, there are species that share this mechanism with humans.

Friday | Sept 11th 15.10 – 15.30 Thursday | Sept 10th 11.40 – 12.00 ten Cate, Carel Spierings, M. ter Maat, Andries Leiden University, The Netherlands Poot, H. Gahr, M. Linguistic rule learning in birds Max Planck Institute for Ornithology, Seewiesen, Germany Abstract The ability to abstract a grammar rule that underlies strings of seemingly meaningless Vocal learning in social groups of zebra sounds is an important aspect of language proces- finches sing. Rule abstraction allows infants to learn about regularities in their native language and to genera- Abstract Male zebra finches sing repetitions of lize these regularities to novel input. This ability is unique song motifs that consist of 3 to 14 syllab- a core mechanism underlying the language faculty. les, which are learned by imitation of adult tutors But is it specific to humans and language lear- during a sensitive phase in early life. Experiments ning? Birds provide interesting models to address with single tutors or tape relays result in almost this question. Like spoken language, songs of perfect copies of the tutor song. Therefore, it was songbirds are characterized by rapidly produced, assumed that juveniles learn the song of their structured sequences of more or less stereotyped fathers rather than that of others. However, wild elements (‘notes’). Birds might therefore also show zebra finches live and breed in large flocks with some ability to abstract structural rules underlying many potential song tutors. In our experiment we strings of sounds. We present an artificial gram- explored possible alternative scenarios where the mar learning (AGL) experiment examining the rule social environment consisted of more than one learning abilities of a songbird, the zebra finch, possible tutor. We kept 10 zebra finch pairs in and a parrot species, the budgerigar. In particular each of 4 aviaries and recorded the song of all we focus on these species’ abilities to distinguish adult males and of the subsequently generated XYX from XXY strings (X and Y denoting different male offspring. The juvenile males imitated song vocal elements), an ability demonstrated in human syllables rather than complete song motifs and infants (with X and Y being speech syllables). the syllables were a mixture of exact copies, imi- Subjects were trained to discriminate between tations, variations, elements and new developed strings of sounds with either an XYX or an XXY syllables. Preferred tutors were not necessarily grammar. After this discrimination was acquired, social or genetic fathers or dominant males. Tutor each subject received a number of test strings choice cannot be ascribed to one favoured tutor (intermixed with the training sounds) that followed type, but seems to be a composition of convenient the same grammar rules, but consisted of either song parts of different tutors.

53 abstracts of symposium speakers

Tuesday | Sept 8th 17.45 – 18.00 Firzlaff, U. TU Munich, Germany

Thiebault, Andréa *These authors contributed equally to this work Tremblay, Y. Mullers, R. A novel approach reveals first molecular Pistorius, P. networks in the bat brain: implications for Nelson Mandela Metropolitan University, vocal communication South Africa Abstract Bats are able to employ an astonishin- Honking in traffic: use of calls in a foraging gly complex vocal repertoire for navigating their seabird environment and conveying social information. A handful of species also show evidence for vocal Abstract Seabirds are known to be very noisy at learning, an extremely rare ability shared only their breeding colonies, when aggregated in high with humans and few other animals. However, densities on land. Calls are used for individual despite their obvious potential for the study recognition between partners and between parents of vocal communication, bats remain severely and offspring. But when at sea, do seabirds com- understudied at a molecular level. To address municate and for what reasons? We deployed 35 this fundamental gap we performed the first Cape gannets (Morus capensis) with GPS devices transcriptome profiling and genetic interrogation and video-cameras with microphones during the of molecular networks in the brain of a highly vo- breeding season of 2010-2011 at Bird Island cal bat species, P. discolor. To identify functional, (Algoa Bay, South Africa) to study their foraging biologically relevant gene networks, we utilized behaviour at sea. Group formation and size of ag- two contrasting co-expression network analysis gregations were derived from the camera footage. methods with distinct underlying algorithms; During ~42h of video recordings, 89 gannet calls WGCNA and MCLUST. These methods typically were recorded at sea. These were only made when need large sample sizes for correct clustering, conspecifics were observed in the surroundings which can be prohibitive where samples are of the equipped birds, and most of them when limited, such as in this study. To overcome this, foraging in large groups (65 calls during 11 group we built on the WGCNA and MCLUST methods to events). Some of the calls in flight (9 out of 29) develop a novel approach for identifying robust were concomitant with jerky camera movements, co-expression gene networks using few samples suggesting their use to avoid collisions with sur- (≤6). Using this approach, we were able to ge- rounding birds. In addition, 25 dives were directly nerate tissue-specific functional gene networks preceded with one or two calls. We will discuss the from the bat PAG, a brain region fundamental use of calls in foraging Cape gannets. for mammalian vocalization. The most highly connected of the networks identified in our study represented a cluster of genes involved in glu- Friday | Sept 11th 16.45 – 17.00 tamatergic synaptic transmission. Glutamatergic signaling plays an essential role in vocalizations elicited from the PAG, suggesting that the gene Vernes, Sonja C. network uncovered here is mechanistically impor- Max Planck Institute for Psycholinguistics, tant for vocal-motor control in mammals. These Nijmegen, The Netherlands; Donders Centre for findings show that our innovative gene clustering Cognitive Neuroimaging, The Netherlands approach can reveal robust biologically relevant Rodenas-Cuadrado, P.* gene co-expression networks with limited sample Chen, X.S.* sizes. Moreover, this work reports the first gene Max Planck Institute for Psycholinguistics, network analysis performed in a bat brain and Nijmegen, The Netherlands establishes P. discolor as a novel, tractable model Wiegrebe, L. system for understanding the genetics of vocal Ludwig Maximilians University, Munich, Germany communication.

54 Friday | Sept 11th 11.40 – 12.00 Abstract To identify the selective factors that shape the structure of a species’ acoustic communication system, comparative analyzes between Vignal, Clementine species are of crucial importance. A major chal- Boucaud, I.C.A. lenge in this endeavor is to standardize analytical Mariette, M. M. procedures to make different vocal repertoires Université Jean Monnet, Saint-Etienne, France comparable. Here we compare the vocal repertoire of chacma baboons (Papio ursinus) with the vocal Vocal negotiation over parental care? repertoire of Barbary macaques (Macaca sylvanus). Partners adjust their time spent incubating The two subsamples were selected from recordings based on their acoustic exchanges at the nest that were taken from wild chacma baboons in the Okawango Delta, Botswana und from semi-wild Abstract Bi-parental care in birds occurs Barbary macaques that live in an outdoor enclosure throughout an individual’s adjustment of its at Rocamadour, France. Both datasets were adjusted workload to that of its partner. Communication to similar sample size, age and sex composition. 38 within a pair might be crucial to achieve this acoustic features that describe the calls in terms of adjustment. Zebra finches, Taeniopygia guttata, frequency and time dynamics were extracted from form life-long monogamous pair bonds, in spectrograms. To quantify the graded structure which partners are highly coordinated and both of the datasets, fuzzy c-means clustering was incubate eggs. When relieving each other during applied. The results reveal a significant higher level incubation, partners perform a structured call of gradation in the Barbary macaque repertoire. duet at the nest. If this duet functions to coor- However, even the chacma baboon repertoire which dinate incubation workload, disrupting the pair’s was formerly described as being rather discrete is usual nest-relief pattern by delaying the male’s showing an array of intermediate call structures return to the nest should affect the structure between typical call types. Another finding is that of the duet. Using domesticated birds breeding the level of gradation differs not only between but in a large aviary, we found that delaying the also within the repertoires and some call types male’s return induced shorter duets with higher show more distinct boundaries than others. A big call rate. In addition, we tracked individuals’ challenge that still remains unsolved is the deter- location with a transponder at the nest and the mination on how many call types can be found for feeder, and showed that these accelerated duets a given species. For graded systems, there does were associated with increased haste of the not necessarily have to be an optimal solution and partners to take turns incubating and foraging. subsequent analyzes should be understood as one Females also spent less time incubating during possible solution amongst others. To describe and their subsequent shift, and females´ time off- understand a repertoire in detail, different cluster nest was best predicted by their mates’ calling solutions should be evaluated and compared. behaviour in the previous duet. Taken together, these results suggest that duets may function as ‘vocal negotiation’ over parental care. Friday | Sept 11th 16.00 – 16.15

Thursday | Sept 10th 13.50 – 14.10 Wojas, Lucyna Osiejuk, T.S. Adam Mickiewicz University, Warszawa, Poland Wadewitz, Philip Hammerschmidt, K. Social eavesdropping on acoustic cues Fischer, J. in non-learning model bird species German Primate Center, Göttingen, Germany Abstract The social eavesdroppers are individu- Advances in the quantitative analysis of als that gather the information about the observed graded vocal repertoires opponents based on interactions between them.

55 abstracts of symposium speakers

This behavior helps to adjust the defensive strategy played back with a set of four speakers, imitating to rival’s capabilities and minimizes the costs of the call exchange, approach, and a physical fight; territory defense. In this study we’ve tested if corn- (2) a simulated intrusion of either winner or crake males eavesdrop on vocal interactions taking loser of earlier interaction into the territory of the place close to the boundaries of their territories. focal individual. In addition, we perform a control Corncrake (Crex crex) is a nocturnal species inha- experiment where in the first stage there were not biting wet meadows with dense vegetation. Males clear winner and loser. In each experiment (n=58), are territorial and the effective communication male’s response was recorded by a microphone between conspecifics is limited to the acoustic array. Males response differ both in behaviour and channel. Corncrake’s territorial call is loud (ca. vocal reaction. Males respond stronger to the intru- 95 dB SPL at 1m) and reaches many conspecifics sion of aggressive individual than to the submissive simultaneously. The rhythm of calling is a conven- and to the control. Corncrakes treated with control tional signal of males’ aggressive motivation and experiment increase their rhythm. Our results help call characteristics are used by males for individual to understand the relationships between territorial discrimination. All the mentioned features make males of the non-learning rail species relying solely corncrake an ideal model for studying the commu- on vocal communication, and suggest that corn- nication networks. To test our hypothesis we use crake males eavesdrop on interaction taking place two-stage experiment: (1) a simulated vocal fight close to the boundaries of their territories. ■

BUY NOW Official IBAC GERMANY mugs

56 POSTER LIST In alphabetical order

Presenting author | Poster number | Date

A Adreani, N. N˚138 Thursday/Friday

Telemetric audio radio-transmitters: a powerful device for studies in vocal communication between group-interacting small animals

Amaya, J.P. N˚101 Thursday/Friday

Development of vocalizations in a fossorial rodent (Ctenomys sp.)

Anichini, M. N˚32 Monday/Tuesday

Singing plasticity and signalling strategies in the bushcricket Poecilimon ampliatus (Orthoptera: Tettigoniidae)

Araya-Salas, M. N˚93 Thursday/Friday

Does vocal learning accelerate acoustic diversification? Evolution of contact calls in Neotropical parrots

Attia, J. N˚159 Thursday/Friday

Acoustic playbacks elicit more responses in dim light than in daylight conditions in the cichlid fish Metriaclima zebra

Aumann, H. N˚126 Thursday/Friday

The RF microphone: a new way of monitoring honey bee sounds

B Baier, L. N˚15 Monday/Tuesday

Breaking the waves: spatial frequency in bat biosonar

Bakhturina, D. N˚143 Thursday/Friday

The structure, context and functions of duet displays in three auk species (Charadriiformes, Alcidae)

57 poster list

Blum, C. N˚109 Thursday/Friday

Vocal ontogeny in Common ravens (Corvus corax)

Bolaños Sittler, P.R. N˚155 Thursday/Friday

Quetzal echoes in mayan temples

Boucaud, I. N˚73 Monday/Tuesday

Does vocal communication at the nest during incubation signal female’s hunger in great tits?

Brazao Protazio, J.M. N˚136 Thursday/Friday

Mathematical-statistical methods applied to bioacoustics – a case study applied to zebra finches (Taeniopygia guttata)

Brumm, H. N˚58 Monday/Tuesday

Traffic noise drowns out great tit alarm calls

Budka, M. N˚145 Thursday/Friday

Duet singing behavior in Common crane

Cardoso, G.S. N˚97 Monday/Tuesday C

Vocal performance of Thamnophilidae (Aves, Passeriformes) species in different biomes of South America

Chabrolles, L. N˚77 Monday/Tuesday

Do personality and social status influence sounds production in Metriaclima zebra

Chivers, B. N˚39 Monday/Tuesday

Bilateral asymmetry in the distribution of sound pressure levels around a singing cricket

Codarin, A. N˚128 Thursday/Friday

An inexpensive Arduino-based autonomous recording unit: open source tool for remote acoustic studies

58 Collier, C. N˚69 Monday/Tuesday

Dwarf mongoose alarm calls: a complex system

Coquereau, L. N˚55 Monday/Tuesday

Acoustics as a potential tool to monitor variabilities under anthropogenic pressure in a marine biogenic habitat

Crisostomo Valdoria, C.M. N˚26 Monday/Tuesday

Early multimodal communication in the territorial Lusitanian toadfish

Curé, C. N˚123 Thursday/Friday

An automatic acoustic recognition system for monitoring an endangered amphibian species, the common spadefoot Pelobates fuscus

D Dadwal, D. N˚30 Monday/Tuesday

Examination of parental investments in the nest defense and anti-predator behavior of a tropical songbird, the Pied Bush Chat

Davis, J. N˚110 Thursday/Friday

Experimental assessment of distance-from-source effects on spectral classification of bird vocalizations recorded in natural environments

Diniz, P. N˚61 Monday/Tuesday

Vocal response by social pairs to playbacks of solos and duets varies with the presence of juveniles in a Neotropical ovenbird

Dorado Correa , A. N˚105 Thursday/Friday

Vocal plasticity in mallards: multiple signal changes in noise and the evolution of the Lombard effect in birds

Duarte, M.H.L. N˚45 Monday/Tuesday

Mining noise affects acoustic communication by the rofous-collared sparrow (Zonozhichia capensis) in a tropical forest

E Eckenweber, M. N˚160 Thursday/Friday

Individual recognition in wild greater sac-winged bats (Saccopteryx bilineata)

59 poster list

Emam, T. N˚53 Monday/Tuesday

The Yorkshire Soundscape Project: a case study in soundscape comparison and composition

Español Jimenez, S. N˚23 Monday/Tuesday

First study of presence of humpback whales vocalizations in a feeding ground of the Southern Chile

Ey, E. N˚148 Thursday/Friday

New insights into spontaneous ultrasonic vocalisations in mice

Faragó, T. N˚71 Monday/Tuesday F

Non-linear phenomena in dog whines as potential indicators of separation anxiety

Favaro, L. N˚3 Monday/Tuesday

Acoustic cues to individuality and species in Spheniscus penguins: a source-filter theory approach

Feng, N. N˚85 Thursday/Friday

The circadian vocal rhythm of a singing fish

Ferhat, A.T. N˚76 Monday/Tuesday

Contexts of emission of ultrasonic vocalisations during social interactions in mice

Florentin, J. N˚57 Monday/Tuesday

Wildlife presence in a rural soundscape near a wind turbine

Florentin, J. N˚111 Thursday/Friday

Acoustic characterization of woodpecker drumming in view of European species identification

Frommolt, K.H. N˚50 Monday/Tuesday

Information obtained from bioacoustics monitoring – lessons from systematic acoustic monitoring during the breeding season

60 g Garcia, M. N˚4 Monday/Tuesday

Acoustic allometry in mammals: a comparative study of over 60 carnivore and primate species

Gervaise, C. N˚122 Thursday/Friday

3D imagery of the marine biophony with compact sensor array

Gessinger, G. N˚9 Monday/Tuesday

Nasal and oral echolocation call emission in phyllostomid bats

Gifford, T. N˚139 Thursday/Friday

Vocalisation characteristics of the soniferous catfish species Neoarius graeffei

Gill, L. N˚79 Monday/Tuesday

Who ya gonna call? Individual recognition in jackdaw (Corvus monedula) contact calls

Glotin, H. N˚118 Thursday/Friday

Scaled soundscape automatic annotation – evidences from the Sabiod EADM challenge

Goerlitz, H.R. N˚41 Monday/Tuesday

Linking lab physiology to behaviour in the wild: the multi-species predator-prey interaction of bats and moths

Gogala, M. N˚34 Monday/Tuesday

Differences in song parameters and taxonomy of singing cicadas (Hemiptera, Cicadidae)

Goncharova, M. N˚70 Monday/Tuesday

Acoustic call variables indicate hunger level in Siberian crane chicks (Grus leucogeranus, Gruidae)

Gonzalez-Terrazas, T.P. N˚156 Thursday/Friday

The sound of cactus: a flower-visiting bat’s perspective

61 poster list

Grison, M. N˚102 Thursday/Friday

Spatial and temporal dynamic in a lek of a tropical bird: the Lipaugus vociferans

Hallerberg, S. N˚132 Thursday/Friday h

A bag-of-coefficients approach towards call categorization

Hoebel, G. N˚16 Monday/Tuesday

Does audiovisual integration change leading call preferences in treefrogs?

Hoier, S. N˚144 Thursday/Friday

Context specific ultrasonic vocalizations in wild female mice (Mus musculus)

Honaiser Rostirolla, T. N˚133 Thursday/Friday

Model reliability for automated bioacoustic recognition through Arbimon

Hrouzková, E. N˚27 Monday/Tuesday

Vibrational communication of blind mole rat (Spalax ehrenbergi) superspecies

Hügel, T. N˚40 Monday/Tuesday

Behavioural audiograms to study anti-predator adaptation in eared moths

Hügel, T. N˚157 Thursday/Friday

Acoustic and ecological similarity: Eavesdropping on heterospecific echolocation calls in Myotis bats

Ilina, I. N˚103 Thursday/Friday i

Comparative analysis of the song variation across the ranges of Chiffchaff (Phylloscopus collybita) and Willow Warbler (Ph.trochilus)

Jakubowska, A. N˚42 Monday/Tuesday j

Song duration mediates response of territory owner during simulated intrusion in a songbird species with small song repertoire

62 Jansen, M. N˚83 Thursday/Friday

Individuality in face of plasticity: signal variation during sustained calling in frogs

Jayasena, N. N˚62 Monday/Tuesday

Interspecific communication for predator awareness between yellow-billed babblers and common palm squirrels k Kerchev, I. N˚35 Monday/Tuesday

Investigation of acoustical communication and behavior of invasive fir bark beetle Polygraphus proximus Blandford

King, S.E. N˚6 Monday/Tuesday

Sound-induced hearing loss and recovery in the axolotl (Ambystoma mexicanum)

Kipper, S. N˚18 Monday/Tuesday

Song features are predictive for male paternal effort in Common nightingales (Luscinia megarhynchos)

Klenova, A. N˚72 Monday/Tuesday

Hunger advertising by begging calls in three auk species (Alcidae, Charadriiformes)

Klenschi, E. N˚38 Monday/Tuesday

Technological advances in 3D micro-scanning laser Doppler vibrometry: investigation of the locust ear

Kriesell, H.J. N˚87 Thursday/Friday

Communication in colonial long-lived seabirds: relative contribution of acoustic and visual signalling in sexual selection in penguins

Kruger, D.J.D. N˚21 Monday/Tuesday

The extraordinary vocal characteristics of the South African Common river frog (Amietia quecketti)

Kugler, K. N˚17 Monday/Tuesday

Echo-acoustic flow guides flight in bats

63 poster list

Lampe, H.M. N˚92 Thursday/Friday l

Exploring the relationship between song characteristics of fathers and sons in pied flycatchers

Lankau, H.E. N˚129 Thursday/Friday

Developing standards and methods for bioacoustic data collection and processing in Alberta, Canada

Larsen, O.N. N˚7 Monday/Tuesday

In-air and underwater hearing of the cormorant studied with ABR

Lattenkamp, E.Z. N˚104 Thursday/Friday

Do horseshoe bats complement their echolocation with listening for prey-generated sounds?

Laverde-R., O. N˚112 Thursday/Friday

Bird songs in the shelf: temporal patterns of songs and vocal output are lost in sound archives

Le Maguer, L. N˚95 Thursday/Friday

Song learning and cognitive abilities in zebra finches (Taeniopygia guttata)

Leitner, S. N˚90 Thursday/Friday

Hatching late in the season requires flexibility in the timing of song learning

Lewanzik, D. N˚158 Monday/Tuesday

Low-intensity echolocation plus intensity compensation – a means for barbastelle bats to remain undetected by eared moth

Linke, S. N˚52 Monday/Tuesday

Real-time acoustic monitoring of freshwater ecosystems

Lossent , J. N˚117 Thursday/Friday

Mapping marine biophony

64 Luís, A.R. N˚121 Thursday/Friday

Analyzing a graded acoustic repertoire: the use of discriminant analysis to classify burst-pulsed signals of bottlenose dolphins m Ma, S. N˚10 Monday/Tuesday

Sensorimotor feedback maintains auditory objects formation

Maciel, I. N˚146 Thursday/Friday

High incidence of non-linear phenomena in Antarctic Phocidae

Maeder, M. N˚44 Monday/Tuesday

„Phyto-acoustics“: acoustic emissions of plants

Maigrot, A.L. N˚68 Monday/Tuesday

Vocal expression of emotions in Przewalski horses

Manfrini, V. N˚80 Monday/Tuesday

Acoustic monitoring of a pregnant bottlenose dolphin (Tursiops truncatus) under human care

Marchenko, A. N˚108 Thursday/Friday

Ontogenesis of vocalization in diurnal bird of prey (Falconifornes)

Marfull, R. N˚91 Thursday/Friday

The effect of sibling competition on the acoustic component of begging behaviour in blue tits (Cyanistes caeruleus)

Martin, J.G. N˚78 Monday/Tuesday

Differences in the vocal emissions of bottlenose dolphins during surface feeding and bottom foraging events

Martinez-Alvarado, D. N˚56 Monday/Tuesday

Does noise affect song activity of oscine and suboscine species between urban and rural areas?

65 poster list

Martínez-Serena, A. N˚127 Thursday/Friday

Mexican bat acoustic monitoring system (SIMMA)

Martins, J. N˚150 Thursday/Friday

Rhythms from the animal world in the practice of composers

Matrosova, V. N˚147 Thursday/Friday

What affects alarm call structure in speckled ground squirrels: multifactor analysis

Matsinos, Y. N˚54 Monday/Tuesday

Vocalization attributes of specific song notes and distress calls (churring) between urban and rural great tits (Parus Major)

Matsinos, Y. N˚59 Monday/Tuesday

Defining quietness in urban green areas

60Matsinos, Y. N˚60 Monday/Tuesday

Soundwalk as a landscape remodeling driving tool

McLeod, L. N˚114 Thursday/Friday

Using emerging bioacoustic technologies to model for Yellow Rail occupancy

Mélotte, G. N˚1 Monday/Tuesday

Sound production in piranhas and relatives: comparisons between species

Meyer, J. N˚149 Thursday/Friday

Speaking and singing with whistles around the world: speech and music in a simple narrow frequency band

Mishima, Y. N˚75 Monday/Tuesday

PS1 call is determined as an affiliative contact call in captive belugas (Delphinapterus leucas)

66 Montealegre-Z, F. N˚33 Monday/Tuesday

Song quality and sound-generator assymetry in bushcrickets

Montealegre-Z, F. N˚37 Monday/Tuesday

Biophysics of a pronotal Helmholtz resonator enhancing sound radiation in a neotropical bushcricket

Montes-Medina, A.C. N˚63 Monday/Tuesday

Functional diversity in the vocal repertoire of the lilac-crowned parrot

Moskalenko, V. N˚96 Thursday/Friday

Song types are genetically determined in birds with the single song type

Motes Rodrigo, A. N˚84 Thursday/Friday

Testing the nutritional stress hypothesis and learning capacity in pied flycatchers n Njoroge, A.W. N˚31 Monday/Tuesday

Acoustic characteristics of Prostephanus truncatus (Horn) (Coleoptera: Bostrichidae) and Sitophilus zeamais (Motschulsky) (Coleoptera: Curculionidae)

Nolan, H. N˚24 Monday/Tuesday

Dingo communication

Noriega Romero Vargas, M.F. N˚151 Thursday/Friday

Bigrams in sequences of pilot whale calls o Osiejuk, T. N˚65 Monday/Tuesday

Soft song use in the presence of predator and during rival‘s intrusions in the Ortolan Bunting (Emberiza hortulana)

Oswald, J. N˚124 Thursday/Friday

Geographic variation in delphinid whistles – implications for classifier development

67 poster list

Padilla de la Torre, M. N˚67 Monday/Tuesday p

Vocal expression of emotion in cattle

Palhares, C.O.M. N˚154 Thursday/Friday

Description of the vocalization of Melanopareia torquata bitorquata (D‘orbigny & Lafresnaye, 1837) in Mato Grosso, Brazil

Parks, S. N˚88 Thursday/Friday

Description of the ontogeny of acoustic communication behavior of North Atlantic right whales

Penna, M. N˚47 Monday/Tuesday

Communication amidst biotic and abiotic interference in frogs of the austral temperate forest

Pereira Chaves, P.A. N˚2 Monday/Tuesday

Development of the inner ear morphology and auditory sensitivity for social acoustic communication in the Lusitanian toadfish

Petruskova, T. N˚28 Monday/Tuesday

Song convergence and aggressive interactions of two closely related songbird species in their contact zone: the case of Nightingales (Luscinia spp.)

Popova, D. N˚99 Thursday/Friday

Ontogenesis of vocalization in several owl species (Strigiformes)

Porteš, M. N˚13 Monday/Tuesday

Who is who? Acoustic monitoring of migratory passerines with complex and simple song as a tool for individual recognition

Prior, N. N˚140 Thursday/Friday

Acoustic behavior in paired, breeding and non-breeding condition zebra finches

Ptacek, L. N˚137 Thursday/Friday

Improving the automatic bird individual identification method by parametrization data merging

68 q Quispe, R. N˚22 Monday/Tuesday

Seasonal and daily timing of dawn chorus in an equatorial songbird r Riebel, K. N˚100 Thursday/Friday

How ‚crystallised‘ is crystallised birdsong? Changes in singing over time in adult male zebra finches

Rivera-Gutierrez, H.F. N˚82 Thursday/Friday

Does song-learning play a role in acoustic adaptation to urban environments?

Romero-Vivas, E. N˚135 Thursday/Friday

Spectrogram differences approach for extracting fundamental frequency variation of dolphin whistles

115Rosa, F. N˚115 Thursday/Friday

Looking for number of degrees of freedom at Orcinus orca calls for the design of a classifier

Ruiz, T. N˚120 Thursday/Friday

Software-aided identification of Neotropical bat calls: an active learning approach s Sarria-S., F.A. N˚131 Thursday/Friday

Loud extreme ultrasonic frequencies produced by a small generator

Schneiderová, I. N˚130 Thursday/Friday

Species determination of captive lesser galagos (Galago spp.) integrating molecular-genetic and bioacoustic data

Schneiderová, I. N˚64 Monday/Tuesday

Differences in alarm calls of juvenile and adult European ground squirrels: findings from a semi-natural enclosure

Schoeppler, D. N˚12 Monday/Tuesday

Non-paradoxical Doppler shift compensation in Rhinolophus paradoxolophus

69 poster list

Schulz-Mirbach T. N˚8 Monday/Tuesday

Diversity in swim bladder morphology affects hearing in cichlid fishes

Seltmann, S. N˚5 Monday/Tuesday

Influence of sleep on song-related neuronal activity in RA – what role does Melatonin play?

Sibiryakova, O. N˚86 Thursday/Friday

Acoustic variation in Siberian wapiti Cervus elaphus sibiricus: effects of sex and age

Simon, R. N˚119 Thursday/Friday

Driverless vehicle guidance by using bio-inspired sonar reflectors

Singh, Y. N˚89 Thursday/Friday

Vocalization a passive indicator of developmental stages of dairy animals

Singh, Y. N˚113 Thursday/Friday

Bioacoustic sensing an authentic descriptor of individual‘s identity

Souriau, A. N˚107 Thursday/Friday

Asymmetrical song convergence in two hybridizing nightingales: The mystery of “doublespeaking” males

Sprau, P. N˚43 Monday/Tuesday

Plasticity or consistency? Individual differences in territorial defence along urban gradients in Great tits (Parus major)

Stepien, A.E. N˚152 Thursday/Friday

Syllable-specific rules in the microstructure of birdsong

Stowell, D. N˚116 Thursday/Friday

From raw audio to songbird communication networks: fully automating the analysis

70 Sueur, J. N˚36 Monday/Tuesday

Ecoacoustics: echoes from a new discipline

Szymanski, P. N˚51 Monday/Tuesday

Vocal communication behaviour as an indicator of changes in bird populations: a case study of farmland species t Tanttu, J.T. N˚25 Monday/Tuesday

Time-frequency analysis of the Marvellous spatuletail (Loddigesia mirabilis) vocalizations

Thomsen, H. N˚106 Thursday/Friday

Individual variation and vocal matching by imitation in peach-fronted conures (Aratinga aurea)

Tipp, C. N˚134 Thursday/Friday

Archiving the sounds on life on earth u Ueno, A.S.O. N˚74 Monday/Tuesday

Barking in capybaras: more than just an alarm call

Ulloa, J.S. N˚20 Monday/Tuesday

Temporal and spatial dynamics of a tropical soundmark: the song of the Screaming piha Lipaugus vociferans v v. Merten, S. N˚153 Thursday/Friday

The song of wild mice – acoustic divergence in natural populations of mice (Mus sp.)

van Boheemen, J.S. 81 Thursday/Friday

Simple song doesn‘t mean low variability – individual variation in the Yellowhammer vocalization

Vellema, M. 98 Thursday/Friday

Prior singing experiences accelerate subsequent development of vocal skills

71 poster list

Villain, A. 141 Thursday/Friday

Behavioural changes in response to background noise sign the importance of intra-pair communication at the nest in a songbird.

Vokurkova, J. 29 Monday/Tuesday

Seasonality in singing activity of bird communities along elevational gradient of Mount Cameroon

Zagkle, E. 14 Monday/Tuesday z

To buzz or not to buzz? Situation-dependent behavioural flexibility in bat biosonar during object approach

Zaytseva, A. 66 Monday/Tuesday

Estimating discomfort in a fat-tailed gerbil by ultrasonic joint-calls

Zollinger, S.A. 11 Monday/Tuesday

Great tits stay on pitch even as vocal amplitudes and background noise levels increase

Zsebök, S. 19 Monday/Tuesday

What does the hierarchical analysis of quantitative and qualitative song traits tell us about signal design and reliability?

72 Vocalizations and behavior

• Describe behavior accurately and quantitatively • Collect and integrate rich data • Use a wide variety of parameters and statistics • Discover our mobile and stationary solutions • Join a worldwide community of users

UltraVox™ XT – Software to capture the full The Observer® XT – Software to collect rich acoustic spectrum of animal vocalizations, behavioral data live or from video and synchro- including ultrasonic calls. These are visualized nize it with other data such as vocalization data in a waveform and a spectrogram from which from UltraVox XT, audio files, and physiological you can select calls. Then categorize calls signals. You can set-up your ethogram and score based on frequency, amplitude, duration, and at any level of detail, continuously or using time gaps and perform analysis. You can easily time-sampling, and perform in-depth analysis. combine your UltraVox XT data with other data Pocket Observer offers a mobile solution for in The Observer XT. smart phones.

Innovative solutions for animal behavior research 73 www.noldus.com abstracts poster

ABSTRACTS POSTER (A – Z of first author) Presenting author is underlined Poster number in left upper corner of abstract

Location: Room “STAFFELSEE” and Room 3

138 the same time. Finally, we show some examples on the application and advantages of this device in the Adreani, M. Nicolás study of vocal communication in small animals. The D´Amelio performance of our device and its characteristics P., Gill, L. make it a great tool for studies in vocal commu- Gahr, M., ter Maat, A. nication between group-interacting small animals. Max Planck Institute for Ornithology, Seewiesen, Germany 101 Telemetric audio radio-transmitters: a powerful device for studies in vocal communication Amaya, Juan P. between group-interacting small animal CRILAR (Centro Regional de Investigaciones Científicas y Técnicas de La Rioja), Argentina Abstract One of the major challenges in the Areta, J.I. study of vocal communication is to record every (IBIGEO-CONICET) acoustic signal that is performed by freely-behaving Basso, G. animals involved in a given interaction. So far, the Farina, A. most widely-used devices have been the classic Universidad Nacional de La Plata, Argentina dynamic or condenser microphones. However, except in few cases, these devices are generally Development of vocalizations in a fossorial insufficient to study individuals in a group because rodent (Ctenomys sp.) of the impossibility of identifying the signals of every sender. In recent years, data loggers and Abstract Ctenomys are highly vocal subterrane- telemetric devices have increased their popularity an rodents endemic to South America. Acoustic and their development has improved surprisingly communication is an integral part of their behavior fast. Work on cetaceans (Deecke et. al., 2013) and they exhibit a remarkable vocal repertoire of and big mammals (Lynch et. al., 2013) have shown airborne signals. The S-type vocalizations are terri- the relevance of using such approaches. However, torial and/or warning long-range signals conformed when it comes to studies in small animals such as by two parts: Part 1 consists on a variable number songbirds, amphibians or small mammals, the main of series of two, three or four notes and Part 2 limiting factors of these devices are: portability, is a train of individuals notes. We analyzed the battery life, storage and synchronicity. In this work development of S-type vocalizations of one female we present a wireless microphone transmitter of and one male litter of Ctenomys sp. from La Rioja, 0.6 grams (battery included), with a battery life of Argentina. Litters were captured in the field and 14 days that allows for avoiding frequent handling sound recorded in our laboratory during the next 5 of the animals and a recording distance of up to months beginning one week after capture. We re- 15 meters. We show that it takes only 4 days for corded 65 vocalizations from the male (initial/final the birds to regain their normal movement and weight 70/135g) and 15 vocalizations from the fe- vocalizing behavior, after the transmitter is placed male (initial/final weight 30/120g). Peak frequency on the animal, and that they are able to perform values decreased in both individuals through time normal behaviors while carrying the transmitters. approaching the adult values: in the male from Furthermore, we present data demonstrating the 222±28.8 Hz [range 183-269] to 175±31.5 Hz synchronicity between 8 transmitters recorded at [129-247] (n=3 in each case), and in the female

74 from 519±209 Hz [323-904] to 318±5.6 Hz was more severe the closer the distance between [312-323] (n=3 in each case). Peak frequency in the competing singers was. Song competition adult males (weight 240 g) is 181±8.4 Hz [161- seems to be strong in this species and should have 193] (n=3), while we lack certain data on female consequences for the signalling strategy in the vocalizations (weight 160 g). Litters gave only Part natural chorus, where more than two competitors 1 which also had fewer series than in adults (ma- sing simultaneously. ximum number of registered series were 6 in litter female, 17 in litter male and 98 adult males) and with longer silences between successive series than 93 in adults (in litter male silences between series durations was 7.4±6.2 sec [1.9-27.5], in litter female Araya-Salas, Marcelo 4.4±0.2 sec [1.6-11.4] in while in adults male was Medina-Garcia, A. 0.7±0.1 sec [0.4-1.1]. This study is the beginning Wright, T. to understand the S-type vocalizations developed New Mexico State University, U.S.A in this specie and allowed to observe how the peak frequency values of these two litters change during Does vocal learning accelerate acoustic part of the period of growth. diversification? Evolution of contact calls in Neotropical parrots

32 Abstract Learning has been traditionally though to accelerate the evolutionary change of Anichini, Mariana behavioral traits. We evaluated the evolutionary Lehmann, G. rate of learned vocalizations and the interplay of Humboldt-Universität zu Berlin, Germany morphology and ecology in the evolution of these signals. We examined contact calls of 51 species of Singing plasticity and signaling strategies Neotropical parrots from the tribe Arini. Parrots are in the bushcricket Poecilimon ampliatus ideal subjects due to their wide range of body sizes (Orthoptera: Tettigoniidae) and habitats, and their open-ended vocal learning that allows them to modify their calls throughout Abstract Males of several acoustically commu- life. We estimated the evolutionary rate of acoustic nicating insect species use these signals to attract parameters of parrot contact calls and directly a conspecific partner for mating. Song in most compared them to those of morphological traits cases seems to be honest signals and song traits and habitat. We also evaluated the effect of body seems to be under sexual selection. In previous mass, bill length, vegetation density, and species studies on the unidirectional communicating bush- interactions on acoustic parameters of contact calls cricket species Poecilimon zimmeri (Tettigoniidae: while controlling for phylogeny. Evolutionary rates Phaneropterinae), females prefer the heavier of two of acoustic parameters did not differ from those of competing males (Lehmann & Lehmann 2008), our predictor variables except for spectral entropy, which transfer the bigger nuptial gift during mating which had a significantly slower rate of evolution. (Lehmann & Lehmann 2009). Using the congeneric We found support for correlated evolution of call bushcricket species Poecilimon ampliatus as a duration, and fundamental and peak frequencies model, we investigated condition dependent song with body mass; and of fundamental frequency with characters. Heavier males produce songs with hig- bill length. The degree of sympatry between species her duty cycles (number of syllables produced per did not have a significant effect on acoustic para- time unit) and obviously signal their condition with meters. We demonstrate that parrot contact calls, the song. When placed in male-male competition which are learned acoustic signals, show similar trials, males speed up song production. However, evolutionary rates to morphological traits. This is matching males with a significant weight difference the first study to our knowledge to provide evidence (heavy and light individuals) showed that heavier that change through cultural evolution does not males outcompete their rivals in song characters. necessarily accelerate the evolutionary rate of traits This song contest with heavier males as the winner acquired through life-long vocal learning.

75 abstracts poster

159 Abstract A radio frequency (RF) microphone has been developed as a new tool for studying incidental Attia, Joel and deliberate insect sounds. It was specifically Université Jean Monnet, St. Etienne, France designed to record the sounds produced by honey bees inside and outside a beehive without disturbing Acoustic playbacks elicit more responses in the bee colony. Acoustic methods of recording bee dim light than in daylight conditions in the sounds are well known. However, acoustic micropho- cichlid fish Metriaclima zebra nes must be inserted into the bee hive. They have very little directionality and the measurements are Abstract The Malawian cichlid fish Metriaclima corrupted by environmental noise.Optical sensors zebra emits sounds. These sounds are produced have also been used to study bee vibrations, but they in the same time as body quivers, behaviours that require very precise pointing and visual access to the occur during aggressive interactions. Aggressive normally dark beehive interior. The RF microphone interactions notably occur during the territory is based on an adaptation of a Doppler radar to defence. Playback experiments at Laboratory have measure vibrations of a dielectric object. It is not put in evidence the attenuator role of the sounds sensitive to airborne pressure waves. If the vibration the aggressiveness. These results were obtained onal amplitude of the object is much less that the under daylight conditions, when visual perception radar wavelength, the frequency of the mechanical is optimal. Nevertheless, aggressive interactions vibration and the frequency of the phase modulated can occur during dawn, dusk, or night. Do low light radar reflection are the same. The RF microphone conditions, disturbing the visual perception, modify derives its narrow beamwidth from the high di- the attenuator role of sounds on aggressiveness rectivity of the radar antenna. By operating in the and/or the sound production? To answer to these unlicensed 5.8 GHz Industrial, Scientific and Medial questions, we have investigated the impact of acou- (ISM) radio frequency band, the instrument can be stic playbacks on the aggressive behaviour of adult implemented safely and inexpensively with readily males defending their territory in very low light available wireless RF components. The output of conditions (“dim light”, 3.2 lx) and under daylight the RF microphone is an audio signal no different conditions (“daylight”, 165.0 lx) for comparison. from the audio signal that might be recorded with a We also recorded the sounds emitted by the fish microphone on a laptop computer. As such the same during the aggressive interactions under these two data collection and processing techniques can be lighting conditions. We found that sound produc- applied. Airborne sounds and vibrations produced by tion is significantly higher in „dim light“ than in honey bees inside and outside an observation bee- „daylight“ conditions. The acoustic playbacks hive are compared. The radar based RF microphone make decreasing the aggressiveness of the receptor is shown to penetrate the opaque beehive wall and fish both in „dim light“ and „daylight“ conditions. measure vibrations which are otherwise inaudible. But the decrease of aggressiveness is four times higher in „dim light“ than in „daylight“ conditions, for the same amount of sounds sent. Metriaclima 15 zebra seems to react more to sounds when visual perception is disturbed. These results suggest Baier, Leonie that the environmental constraints can impact the Ludwig Maximilians University Munich, Max multimodal communication of Metriaclima zebra. Planck Institute for Ornithology Seewiesen, Germany Goerlitz, H. 126 Max Planck Institute for Ornithology Seewiesen, Germany Aumann, Herbert Wiegrebe, L. University of Maine, U.S.A. Ludwig Maximilians University Munich, Germany

The RF microphone: A new way of monitoring Breaking the waves: spatial frequency in bat honey bee sounds biosonar

76 Abstract The concept of spatial frequency was Abstract A number of hypotheses explain the instrumental in understanding how visual periphe- presence of duets in behavioral repertoires of some ral representations of objects are processed by the passerine birds. But suitability of such hypotheses brain. In echolocation there is no direct correlation for non-passerine birds is still unclear. Here we between an echo’s acoustic features and the spa- compared the structure and context of duet dis- tial layout of the environment. Bats must use the plays (so-called billings) in three auk species. The spectro-temporal and binaural echo cues to create objects, namely crested auklet (Aethia cristatella), an internal representation of their surroundings. parakeet auklet (Cyclorrhynchus psittacula) and We hypothesised that despite the fundamentally horned puffin (Fratercula corniculata), are closely different peripheral representations of spatial related, monogamous, colonial breeding and highly frequency in vision and echolocation, both systems socially active seabirds. They spend on duetting utilise the concept of spatial frequency to segregate from 2 to 10% of their time during social activity on objects from their surroundings. Here we quantified the colony. In 2010-2011 and 2013 years we ob- perceptual sensitivity to spatial frequency in a for- served duet behaviour of the species on the natural mal psychophysical detection experiment. Six bats breeding colony of Talan Island, Sea of Okhotsk. We of the species Phyllostomus discolor were trained described the situational context of duet displays to discriminate 3D-objects exhibiting wave patterns and found that it differ significantly among the with spatial frequencies in the range of 4.4/m to species (Kruskal-Wallis ANOVA, p<0.001). Most 17.8/m from a reference object with a smooth sur- frequently individuals of all species performed face. Our results show that wave amplitudes requi- duets due to conspecific approach (47.2, 25.7 and red for detection decrease with increasing spatial 19.7% of cases in crested auklet, parakeet auklet frequency, i.e. animals were much more sensitive and horned puffin, respectively) or in course of in- to higher spatial frequencies than to lower frequen- teractions between the pair mates (31.1, 50.5 and cies. We argue that this perceptual high-pass filter 74.2% of cases, respectively). The structural vari- is directly comparable to a visual or optical spatial ables of duets, namely total duration and partners’ high-pass filter that allows extracting the contours postures, had significant interspecific differences; of objects in a complex scene. Ecologically, this is the duet intensity didn’t differ among species directly relevant to prey detection on water, as prey- (Kruskal-Wallis ANOVA). In all species the structure induced disruptions in spatial frequency patterns of of duets was affected significantly by the number of natural water surfaces may become more detecta- conspecifics in the nearest environment (Spearman ble. Consequently, the formal detection experiment rank), but not affected by the situation that caused is currently complemented with a field experiment duet (Kruskal-Wallis ANOVA). Crested auklet’s du- using water specialists, Daubenton’s bats. If our ets were the most attractive for conspecifics (they expectations hold to be true, this would identify caused the largest number of approaches), while spatial frequency as a cue for prey detection on duets of parakeet auklet and especially horned agitated water surfaces and provide first insights puffin were less attractive (0.6±0.9, 0.2±0.5 and into how a biosonar system solves this challenging 0.1±0.3 conspecifics’ approaches during duetting, task despite its lack of an explicit spatially arranged respectively). We discussed the possible functions sensor array. of duet displays in auks like mate guarding, signaling about quality and commitment to the partner. The study was funded by Russian Scientific 143 Foundation (grant 14-14-00237).

Bakhturina, Daria Klenova, A. 109 Lomonosov Moscow State University, Russian Federation Blum, Christian Pika, S. The structure, context and functions of duet Max Planck Institute for Ornithology, Seewiesen, displays in three auk species Germany (Charadriiformes, Alcidae) Böckle, M.

77 abstracts poster

Danube University Krems, Austria Quetzal Echoes in Mayan Temples Bugynar, T. University of Vienna, Austria Abstract The species Pharomachrus mocinno (Resplendent Quetzal) is a very important bird Vocal ontogeny in common ravens for Guatemala, since ancient times. The Mayas (Corvus corax) represented it in their art and temples, and was considered a god. In modern culture the species is Abstract Language clearly distinguishes us also very important and it is represented in textiles from the rest of the animal kingdom. To investigate and Mayan art. The quetzal is the Guatemalan the evolution of this ability researchers have been national bird, it is part of the flag shield, it is a focusing on complex communication systems of freedom symbol and it is also the name of the other animal species. Bird song represents one national currency. At the Mayan cities it is possible of the most complex systems. It is characterized to listen a phenomenon caused when a person claps by an early critical period in communicative deve- in front of some temples, which cause a chirp echo lopment, vocal production learning and individual very similar to the call of the bird. It has been dis- recognition. However, little is known about the cussed if this phenomenon was caused deliberately complexity underlying calls, which unlike song by the Mayas, or if this was only a causality. There are short, flexible vocalisations and are produced are a few studies about this effect at Chichen Itzá by both sexes. They occur independent of season, (México), to elucidate how this happens, from an are strongly related to specific contexts and can acoustic point of view, and a very simple compa- function as referential signal. Concerning usage, risons of the sounds of the echo and the bird. It calls may therefore be more cognitively demanding has been proposed that the temples could be the than song, equalling the complexity of referential first sound preservation system made by humans; it vocalizations of some nonhuman primate species has even been suggested that this echoes could be (e.g. vervet monkeys). The aim of the present study used to attract quetzales to the temples, as a kind was to investigate i) emergence and extinction of of “play back”. The major concentration of Mayan specific calls, ii) order of emerging call-contexts, cities are located at Guatemala, as well as the such as affiliative and agonistic interactions and tallest temples. One of the more important cities functional reference, iii) changes in physical is Tikal, which means “City of Voices”, and is con- parameters like chaotic parts of vocalisations, sidered as a capital for the Mayas at their heyday. frequency and duration of calls in the course of For this research I did a comparison between the vocal ontogeny. As study subjects we chose ravens, echoes of the temples, and the calls of the quetzal a species of songbird, whose vocal repertoire at their habitat. The obtained results show a great consists entirely of calls with varying degrees of resemblance between the two sound types, and will acoustic complexity, featuring context specificity be useful to support future studies to answer other and functional reference. Our data was collected important questions regarding this effect. from 13 individuals during their first year of life, covering the time from fledging to vocal maturity. We recorded 45 minutes per bird per week (ob- 73 servational focal sampling) and analysed resulting data via a clustering method. Preliminary results, Boucaud, Ingrid such as change in number of clusters over time Aguirre Smith, M. and order of emergence of new call-groups are Valere, P.A. discussed. Vignal, C. Université Jean Monnet, St. Etienne, France

155 Does vocal communication at the nest during incubation signal female’s hunger in great tits? Bolaños Sittler Pablo R. Abstract Although most bird species show University del Valle de Guatemala, Guatemala monogamous pair bonds and bi-parental care, little

78 is known on how mated birds coordinate their ac- stand to benefit from the application of more mo- tivities. Acoustic communication between partners dern and precise analytical techniques. The aim of at the nest is a good candidate but as seldom been this study is to empirically examine the relationship considered. The great tit (Parus major) represents between acoustic properties of vocalization, body an interesting study system to investigate intra-pair size, and larynx size, across a diverse sample of communication at the nest, as males are known to carnivore and primate species (n=60+) using address songs to their female while she is in the rigorous methods. For each species in our study nest cavity, and females have been shown to answer we obtained vocalization recordings, average body their male with calls. In a first study, we recorded size measurements, a laryngeal specimen, and an the vocalizations and observed the behavior of great associated femoral length measurement. Acoustic tit pairs around the nest at different breeding stages analyses are made and the results are compared (laying, incubation and young nestlings). We obser- to the anatomical data. In addition to traditional ved vocal exchanges between the female inside the approaches that correlate acoustic properties with nest and her male outside in three situations with body size, we additionally examine how the same different outcomes: (i) the female left the nest, (ii) acoustic properties correlate with laryngeal anato- the male entered the box to feed the female, (iii) my (as assessed by measurements of 3D models mates stopped calling but did not move out/in the derived from CT-Scan images). Given that the larynx nest. The comparison of the acoustic structure of is the organ of vocal production, we predict that its these three situations of vocal exchanges suggests properties are likely to be more directly related to that one possible function of this communication the acoustics of vocalization than body size. is for the female to signal her need for food to her mate. To test for this hypothesis, we conducted a second study in which each pair was recorded on 136 two days during incubation: one day without food supplementation and one day with a feeder of me- Brazao Protazio, J. Marcelo alworms into the nest box. We found that females Universidade Federal do Pará, Brazil; Max Planck called less with food supplementation. This result Institute for Ornithology, Seewiesen, Germany confirms our hypothesis that females indicate their Leitão, A. need for food in vocal exchanges with their mates. ter Maat, A. Males’ response to this information remains to be Gahr, M. determined. Max Planck Institute for Ornithology, Seewiesen, Germany

4 Mathematical-statistical methods applied to bioacoustics – a case study applied to zebra Bowling, Daniel L. finches (Taeniopygia guttata) Garcia, M. Fitch, T.W. Abstract The main objective of this work is University of Vienna, Austria create a methodology able to model the dynamic of the song learning process of the Zebra finches Acoustic allometry in mammals: a comparative (Taeniopygia guttata) from his first vocalizations study of over 60 carnivore and primate species until the song crystallization. For that, we use mathematical-statistical methods to generate in- Abstract Understanding the information commu- dices based on spectral properties of the sound to nicated by animal vocalizations is one of the main describe individual characteristics of the song. With focuses of bioacoustics research. Vocal allometry these parameters, we developed a function based on seeks to determine what the acoustic characteris- similarity indices, to model the development of the tics of an animal’s vocalizations can tell us about individual‘s song in relation to its crystallized song the size of their body. Despite considerable interest, and in relation to its tutor song. In the study, we there are only several studies in the literature that apply the methodology to the Zebra finches, but the compare data across species, and their methods procedure can be used for any other object of study.

79 abstracts poster

58 145

Brumm, Henrik Budka, Michal Zollinger, S. A. Kokocinski, P. Max Planck Institute for Ornithology, Seewiesen, Adam Mickiewicz University, Poznan, Poland Germany Templeton, C.N. Duet singing behavior in common crane University of St. Andrews, U.K. Abstract Singing in duet is one of the most ama- Traffic noise drowns out great tit alarm calls zing communication strategy, both from behavioral and evolutional point of view. Researchers still Abstract In the recent years, anthropogenic have tried to explain wherefore such coordinated noise pollution has been receiving increased acoustic display, uttered by two individuals, have interest by bioacousticians and conservationists evolved. Despite these efforts, the function, causes alike. Perceptual studies with birds in the la- and ontogeny of such joint acoustic display remains boratory have shown that high noise levels can unclear, since too little species have been exami- constrain their call detection but the fitness ned to make any generalization. Here we focused consequences of these impairments are only on duet singing behavior in Common Crane (Grus poorly understood. Here, we investigated whether grus) - a large-size bird species, which inhabits traffic noise impairs the communication of free- northern parts of Europe and Asia. We analysed re- ranging birds in a task with important fitness cordings of 39 wild-living pairs, collected in Poland consequences, alarm call detection. Birds use in 2014 and 2015. We measured various temporal alarm calls to communicate about the presence and spectral call characteristics to examine, (1) the threat level of predators. Therefore, failure which of them could be responsible for identity to immediately detect an alarm call may be fatal coding and (2) body size signaling. Additionally, for a bird. We studied alarm call production we performed experiment with song transmission and perception with a combination of lab and in which we tried to show how signal is degraded field experiments in great tits (Parus major), a during sound propagation for a long distance. We common songbird that inhabits noise-polluted found that variation observed within-individual environments. First, we measured noise-related and within-pair in song characteristics is lesser call changes in the lab and field and found that than those observed between different individuals birds increased the amplitude of their calls and different pairs. Thus, in this species, theo- in response to increased background noise. retically, song characteristics could be used both Based on these measurements we were able to for individuals and pair discrimination. Moreover, construct realistic call playback stimuli that we we found significant differences in formant fre- tested in combination with realistic traffic noise quencies between individuals which sing in duet. playbacks in the territories of great tits close This may suggest that formant frequencies could to country roads in Bavaria, Germany. Our data be acoustical body size signal, since in Common demonstrate that current levels of road traffic Crane sexual dimorphism in body size is observed – noise mask the alarm calls of this species at a males are larger than females. We also showed how distance of 20m, leading to markedly reduced signal is degraded during transmission. We found responses to alarm calls in the presence of that formant frequencies, especially those lower, traffic noise compared to control trials with no are distinguishable and weakly distorted even in noise. These results show that despite the vocal the distance 500 m from the speaker. This finding adjustments used to compensate for anthropoge- suggest that formants could be important acoustic nic noise, great tits could not restore the active information carrier. We hope that our study will space of their calls. As a consequence, birds are be the first step to the better understanding the likely to suffer from increased predation risk in function, causes and ontogeny of singing in duet noise, which could affect both their behaviour in Common Crane. This study was supported by and population dynamics in noise-polluted the Polish National Science Centre (grant number areas. 2013/09/N/NZ8/03214).

80 97 77

Cardoso, Guilherme S. Chabrolles. Laura Donatelli, R.J. Thibault, T. São Paulo State University, Brazil Attia, J. Vocal performance of Thamnophilidae (Aves, Beauchaud, M. Passeriformes) species in different biomes of Université Jean-Monnet, St. Etienne, France South America Amorim, C. Fonseca, P. Abstract The vocal performance refers to the MARE, ISPA-Instituto Universitário, Lisboa, ability to execute challenging vocalizations closer Portugal to the physical constraints of the birds. This feature provides information about male quality and may Do personality and social status influence interfere with female choice for mates. The trilled sounds production in Metriaclima zebra vocalizations are limited by a performance tradeoff between the syllable repetition (trill rate) and the Abstract Many gregarious species have a social range of frequencies that the syllables can scope hierarchy (1). It is the case for Metriaclima zebra, (frequency bandwidth). Birds of Thamnophilidae a sounds producer cichlid’s fish. In this species, family are known to have species that produce dominant fish have favored access to resources trilled vocalizations. This study sought to determine because of their social status. Social status is whether different species of Thamnophilidae within therefore a very important criteria to qualify an the same geographic distribution show similar individual. Otherwise, individuals from a same vocal performances. We surveyed 645 songs from group can show different levels of locomotor 31 species. We clustered the vocalizations based activity, attractiveness for novelty, willingness or on the geographic distribution of the species. We avoidance for fights with fellow creatures (2). All obtained three groups of species: the Amazonian, of these behavioral expressions form what we call the Atlantic and the Central species. The trill «temperament» or «personality» in animals (3). rate (TR) and the frequency bandwidth (FB) were But the link between social status and tempera- calculated for each song. We applied the Spearman ment is not always so obvious. Moreover, even if correlation on each group separately. Then, a vocal we begin to understand the role of acoustic sig- performance limit was established based on an nals during interactions between two individuals, upper boundary linear regression between those we do not know yet the impact of the social status two variables. We calculated the performance of and the temperament on the sounds production vocalizations using the distance from the point to for Metriaclima zebra. Firstly in this study we the upper boundary line. We conducted a Kruskall- used neophobic device to test fish personality. Wallis test with Dunn-Bonferroni post-hoc test on Then we performed dyadic interactions between the groups using the vocal performance, the trill fish showing different status and personality. All rate and the frequency bandwidth of each song to behaviors and sounds emitted during the fights determine the relationships between the biomes. have been recorded and analyzed. First results The trill rate had significant correlation with the showed that the dominant fishes were bolder and frequency bandwidth on the Amazonian and Central more aggressive than the dominated fishes. But species. The three groups have vocalizations with now remains the question about a correlation bet- low performance, but the Amazonian and Atlantic ween the sounds production and the protagonist’s species present a considerable number of samples status. Is there any difference in sounds produc- in which performance is closer to the limit. The tion between a fight among two dominant fishes Kruskall-Wallis test showed that the groups differ and another fight among two dominated fishes? the trill rate and in the frequency bandwidth. (1) Jameson et al., 1999 (2) Gosling, 2001 (3) According to the post-hoc test, the Amazonian and Réale et al., 2007 the Atlantic groups are more similar in frequency bandwidth than the Central one.

81 Avisoft SASLab Pro Powerful Sound Analysis and Synthesis Software for Investigating Animal Acoustic Communication

Automated sound parameter measurements Call classification based on template comparison using spectrogram cross- correlation Versatile sound editing and filtering Graphic Synthesizer for generating modified songs and calls Pulse Train Analysis tool for quantifying temporal signal pattens Octave and Third-Octave Analysis ...

High-quality sampling at up to 1 MHz / 16 bit Single or multichannel operation Compact and rugged design (USB interface) Various ultrasound microphones (both electret Schönfließer Str. 83 and sensitive solid-dielectric condenser types) 16548 Glienicke, Germany Ultrasound speakers and power amplifiers for [email protected] playback experiments Versatile recording and playback software www.avisoft.com UltraSoundGate 82 Integrated Ultrasound Recording Systems for Field and Lab Use 39 128

Chivers, Benedict Codarin, Antonio Jonsson, T. ARPA FVG - Regional Agency for Environmental Jackson, J. Protection of Friuli Venezia Giulia, Palmanova, Windmill, J. Italy Montealegre-Z., F. University of Lincoln, U.K. An inexpensive Arduino-based autonomous recording unit: open source tool for remote Bilateral asymmetry in the distribution of acoustic studies sound pressure levels around a singing cricket Abstract Autonomous acoustic recorders are Abstract Male field crickets generate mating very important and increasingly popular tools for calls to distant females through tegminal stridu- acoustically and bioacoustically characterization lation (the rubbing together of the overlying right of specific environments over a long period, both wing which bears a file of teeth against a ple- terrestrial and aquatic. Often, however, their exces- ctrum on the underlying left wing). Specialised sive cost inhibits their use. During the course of areas of membrane on each wing oscillate my work I have designed and built a water-proof together to produce acoustic radiation. Maximum recorder, driven by Arduino UNO board programmed transmission effectiveness will be facilitated by to record, continuously or at specific intervals, the omnidirectional propagation at high intensities as noise present in a certain area, for the cost of the location of the females in the environment is about 300 Euros. The recorder is built with readily unknown to the calling males. However, produ- available materials and can easily be assembled, by cing an omnidirectional sound field of maximum anyone, with maintenance costs of only the internal sound pressure levels may be problematic due a batteries. It provides high-quality audio data (up mechanical asymmetry found in the cricket sound to 96kHz, 24-bit) in .wav format saved on a SD generation system. Mechanical asymmetry occurs card and can be used both in terrestrial and aquatic by the right wing coming to partially cover the environments, connected to an external microphone left wing (and the left wing membranes) during or a hydrophone, and fixed to a suitable support. the closing stroke, the effective stage of sound The recorder has been tested to assess the acoustic production. As such it is hypothesised that the environment, as required by the Marine Strategy sound field on the left-wing side of the animal European Directive, in the Gulf of Trieste, an area will contain lower sound pressure components with intense vessel traffic (north Adriatic Sea, 25 than on the right-wing side as a result of this meters depth). Results of one minute recordings coverage. This hypothesis was tested using a no- made every 15 minutes, 24 hours a day, will be vel method to accurately record a high resolution, reported. The low cost, flexibility and increasing three dimensional mapping of sound pressure popularity of systems based on Arduino should levels around restrained field crickets in which provide biologists, teachers and students with a stridulation had been elicited by pharmacological tool which is easily customized for studies in a wide brain stimulation. Acoustic recordings were made range of environments. For this purpose possible through the use of a robotic arm, which is capa- changes to the recorder will be discussed in order ble of moving a condenser microphone across a to extend the recording time and the maximum mesh of points around a central position whilst amount of recordable data, and make it completely maintaining a consistent distance and aspect to submersible to a medium depth. the source. The results indicate that across individuals, a bilateral asymmetry is present, with greater amplitude components being present in 69 the right-wing side of the animal, the uncovered wing. Individual variation in the focus of sound Collier, Katie pressure to either the right or left-wing side is Townsend, S.W. also observed. Manser, M.B.

83 abstracts poster

University of Zurich, Switzerland Di Iorio, L. Radford, A.N. Fondation of Grenoble Institure of Technology, University of Bristol, U.K. Grenoble, France

Dwarf mongoose alarm calls: a complex system Acoustics as a potential tool to monitor variabilities under anthropogenic pressure in a Abstract Dwarf mongooses (Helogale parvula) marine biogenic habitat produce a large number of alarm calls to warn conspecifics of danger. Whilst previous work has Abstract Maerl beds are a marine biogenic suggested the function of these calls is to refer to substrate, which has been the subject of increa- different predator types (Beynon and Rasa, 1989) sing attention in recent decades because of its these observational data have not been verified high importance as reservoirs of diversity, often with playback experiments demonstrating receiver called “temperate waters reefs”. However, human comprehension. We revisited the alarm call system activities, such as maerl extraction, fishing and eu- of dwarf mongooses living on Sorabi Rock Lodge, trophication, continue to degrade beds worldwide. Limpopo, South Africa. Through recording calls Acoustic monitoring of invertebrate sounds may be given during natural predator encounters as well indicative of ecosystem health since changes in as experimental predator presentations, we concur biodiversity as a result of degradation likely change that dwarf mongooses produce structurally distinct the acoustic features of underwater soundscapes. alarm calls to different types of predators: aerial However, limited records of sound production and terrestrial. Playback experiments further de- by invertebrates have restrained the application monstrated distinct predator-specific responses to of acoustic monitoring at the individual and the the playbacks of these two alarm calls, suggesting community level in benthic habitats. The first step these alarm calls are functionally referential. Dwarf of our work was to identify and characterize sound- mongooses also produce a third alarm call that is producing invertebrates in maerl beds. Aquarium the combination of the two other calls, though its recording sessions of twenty invertebrates from production does not seem to be tied to a specific Atlantic maerl beds revealed fifteen different sound predator type. Instead, this call combination seems types. The behaviors corresponding to these sounds to be given when a predator of either type is present were feeding, moving, snapping, and probably but not an immediate threat to the mongooses, con- communication. These soniferous species included forming with the hypothesis that call combinations three species of sea urchins, one species of snap- are more likely to happen in less dangerous situ- ping shrimp, one of spider crab, and the invasive ations. Furthermore, behavioural observations and slipper limpet. These species play key ecological playback experiments indicate receiver response and economic roles but their sound production has to this third alarm call varies in both quality and not been described previously despite its potential intensity depending on the context it is produced for ecological monitoring. Ongoing work conducted in; namely, whether it is produced alone or is in poorly and heavily dredged maerl beds suggests preceded by predator-specific alarm call. Our data variabilities between both habitats in seasonal indicate dwarf mongooses have a complex alarm sound production. Acoustics might help in moni- call system with both functionally referential calls toring maerl beds health state, and more generally and call combinations whose importance seems to benthic habitats. be context specific.

123 55 Curé, Charlotte Coquereau, Laura Dutilleux, G. Grall, J. Cerema, Dter Est, Laboratoire Régional de Chauvaud, L. Strasbourg, France Institut Universitaire Européen de la Mer, Brest, France An automatic acoustic recognition system for

84 monitoring an endangered amphibian species, University of Saint Joseph, Macau, SAR China the common spadefoot Pelobates fuscus Chaves, P. University of Saint Joseph, Macau, SAR China and Abstract Acoustic recordings of animal sounds MARE, ISPA, Lisboa, Portugal provide an effective tool for monitoring acoustically Amorim, C.P. active species and species which are difficult to MARE, ISPA-Instituto Universitário, Lisboa, observe in the wild. Automated acoustic recordings Portugal allow to collect sounds for extended periods, increasing the probability of detecting a given Early multimodal communication in the species with little disturbance to the vocalizing territorial Lusitanian toadfish animals. This method implies the processing of a huge acoustic dataset. Although it can be done Abstract Ontogenetic development of acoustic manually by listening to the collected acoustic communication has been widely reported in birds recordings, it represents a very long and tedious and mammals but scarcely investigated in other work. The aim of our study was to get rid of this vertebrates such as fish. Certain species as the operator-dependent processing of the acoustic data Lusitanian toadfish (Halobatrachus didactilus) and to develop an automatic acoustic recognition start producing sounds (grunts) and visual displays system for monitoring the common spadefoot toad remarkably early in development, suggesting that (Pelobates fuscus), an amphibian species threate- these signals may convey important information for ned in France. Automated acoustic recordings were social communication. The biological function of carried out in pools of different locations throug- such multimodal communication in early fish deve- hout four successive annual breeding seasons. A lopment has never been properly investigated. We routine was developed in Scilab to automatically investigated the social role of acoustic and visual detect the spadefoot calls in the collected acoustic signaling among early juveniles H. didactilus (3-6 recordings. The recognition system was based upon months old; 2-3 cm total length). Our goals were: 1) i) inspection of the frequency band corresponding determine how social context affects acoustic and to spadefoot call frequencies carrying the highest visual signaling; and 2) verify whether signaling is amplitudes and ii) difference of sound levels related to territory ownership (social status). Six between this frequency band and the background groups of four similar-sized juveniles were isolated noise. Best configuration of the system retained a in small tanks provided with two shelters to promote signal as being a common spadefoot toad call if the territorial competition. Two fish were initially placed maximum sound level, calculated in the octave for in each tank to become “residents”; 24h later two which the center frequency is 800 Hz, exceeded other individuals were introduced. Audio-video re- the average sound level of the background noise of cordings were carried for 1h, twice a day, over 4-12 +14dB. This system showed 52% of true positives days. Data was analyzed per individual regarding and 5.4% of false positives, providing a reliable number of sounds and visual displays, and social method compared to existing anuran automatic context of signal production. Most juveniles firstly detectors. Using this system, we explored temporal introduced in the tanks readily occupied the availa- variation in the spadefoot calling activity throug- ble shelters becoming residents and defending the hout the breeding season and found that beside the territory throughout the whole observation period. typical and expected nocturnal vocal peak activity Juveniles relied on both visual and acoustic signals of the species, a diurnal period of calling was also that were often exhibited as a multimodal display. present. The most common visual signal was attack. Visual signalling was equally observed during territorial defense and other social interactions, but frontal 26 and lateral displays were also exhibited. The social context had a significant effect on sound Crisostomo Valdoria, Ciara M. C production, with the feeding context evoking the Ramos, A. highest number of sounds. Residents, or dominant Lam, M.K. individuals, revealed higher visual and acoustic Vasconcelos, R.O. signaling rates compared to non-residents. We

85 abstracts poster

provide evidence that multimodal communication Freeberg, T. plays important role in the social communication of University of Tennessee, Knoxville, U.S.A. a vocal teleost species. Experimental assessment of distance-from- source effects on spectral classification 30 of bird vocalizations recorded in natural environments Dadwal, Navjeevan Bhatt, D.C. Abstract Propagating sounds degrade with Sethi, V.K. distance from their source. Animal signal degra- Gurukula Kangri University, Haridwar, dation can influence information content, value Uttarakhand, India and, in turn, organismal responses; therefore it is of conceptual importance in ecological acoustics. Examination of parental investments in the However, acoustic degradation also influences nest defense and anti-predator behavior of a accuracy in detecting and classifying animal tropical songbird, the Pied Bush Chat signals, and may be especially problematic in the interpretation and effectiveness of spectral analy- Abstract Most of the avian species expend con- ses using automated recordings. We take a robust siderable efforts in nest defense through different experimental approach to determining the effects strategies including use of acoustic signals in the of distance-from-source on acoustic signal detec- form of threat/alarm calls. In the present study, tion using spectral tags and automated searches an attempt was made to examine the parental (Avisoft SAS-Lab Pro). We broadcast replicated and investments in nest defense behaviour by the Pied calibrated high fidelity recordings containing > 20 Bush Chat (Saxicola caprata). We quantified the vocal elements produced by 2 species of birds in alarm and territorial calls in the presence of the family Paridae in natural habitats within the spe- decoy avian predator in the territories of 12 males. cies’ ranges. Broadcast sounds were then recorded Results revealed that the nest protection intensity at 1m and at all 10m increments up to 100m away increased with the progression of the nesting cycle from the sound source. Experiments with identical i.e. the parental aggression was more intense to procedures were conducted in FL, IN, and TN, USA protect the nestling than that of the eggs. Also, in the same months. Spectral tags were created significant difference was noted for the usage of for each vocal element from both the high fidelity the threat calls over the territorial calls during and field recordings across the distance gradient this experiment. The assessment of the individual and are being used in tag sets containing different parental aggression between male and female put titrations of classifiers made from degraded and the male on the winning side i.e. for the most un-degraded vocal elements. Statistical models will cases, males were found more aggressive than that evaluate, for each vocal element, the relationships of the females (on the basis of the calls rates they between detection accuracy, distance-from-source, uttered). It was interesting to note that consistent tag set composition, spectral model parameters, use of the same predator throughout the study did background noise, habitat type, and region. Results affect the parental aggression behaviour. Thus, this will be discussed in light of the most common study does not support the prey- predator familia- objectives and strategies for automated biophonic rity hypothesis. recording and signal detection (e.g., community surveys vs. species-centered behavioral outputs).

110 61 Davis, Jordon Sieving, K. Diniz, Pedro University of Florida, Gainesville, U.S.A Rech, G. Lucas, J. Ribeiro, P. Purdue University, West Lafayette, U.S.A. Macedo, R.

86 Universidade de Brasília, Brazil Brumm, H. Max Planck Institute for Ornithology, Seewiesen, Vocal response by social pairs to playbacks Germany of solos and duets varies with the presence of juveniles in a Neotropical ovenbird Vocal plasticity in mallards: multiple signal changes in noise and the evolution of the Abstract Duets may occur between breeding Lombard effect in birds pairs due to conflict (e.g., when an individual answers its partner’s song to acoustically mate- Abstract Acoustic communication is crucially guard partner) or as a cooperative behavior (e.g., constrained by environmental noise. As a response, if the individuals share territory defense benefits many animals adjust the properties of their voca- by coordinating their songs into duets). The pre- lizations in noise to maintain signal transmission sence of juveniles, however, may affect the relative despite interference. A particular form of this noise- cost for territorial birds, of either losing a partner dependent vocal plasticity is the Lombard effect, in versus losing a territory, thus contributing to how which a signaler increases its vocal amplitude in duetting functions. We studied the Rufous Hornero, response to an increase in the background noise. a year-round territory defending species with The Lombard effect is a well-studied feature of delayed juvenile dispersal. We hypothesized that human speech. It has also been reported in other the cost of losing a territory is higher for pairs with mammals as well as several bird species. However, juveniles than for those without, because territory the evolution of the Lombard effect and other rela- loss may affect both juvenile and adult survival. ted vocal adjustments in birds is still unclear. Thus To test this hypothesis, we played back (1) duets, far only three of the five major avian clades have (2) male solos, (3) female solos, and (4) a control been studied. We report the first evidence for the heterospecific duet to social pairs, with and without Lombard effect in an Anseriform bird, the mallard juveniles, in the non-breeding season. Both male (Anas platyrhynchos). Besides the increase in call and female adults sang promptly after conspecific amplitude, the birds also raised the peak frequency playbacks (57 times faster compared to control), in- of their calls in noise. However, in contrasts to dependently of juvenile presence. Most of the song the increase in amplitude the observed increased response (94%) to conspecific playbacks consisted in call frequency did not yield any release from of duets, of which 69% were male-initiated. Pairs masking because we used white noise in our experi- with juveniles responded faster to the playback of ment. Therefore the frequency change is unlikely to duets and female solos than pairs without juveniles. be adaptive in the context of noise mitigation. Our Pairs without juveniles responded faster to male so- findings support the notion that all extant birds use los than to any other playback treatment. Playback the Lombard effect to solve the common problem of treatments did not influence male and female song communication in noise, i.e. it is an ancestral trait duration, duet duration, and probability of males shared among all living avian taxa, which means initiating a duet. Our results suggest that duets are that it has probably evolved more than 100 million cooperative displays in the Rufous Hornero. Territory years ago within that group. seems to be more valuable for pairs with juveniles than pairs without juveniles, and social pairs with juveniles seem to coordinate their songs to defend 45 their territory. In contrast, male intrusion seems to be more threatening to territorial males without juve- Duarte, Marina H.L. niles, though females seems to cooperate to prevent Pinto Viana Pontificial Catholic, Y., Federal their partners from being overthrown. University of Minas Gerais, Brazil

Mining noise affects acoustic communication 105 by the rofous-collared sparrow (Zonozhichia capensis) in a tropical forest Dorado Correa, Adriana Zollinger, S.A Abstract Several species depend on acoustical

87 abstracts poster

signals for communication, and many studies have individually plays a crucial role in shaping animal shown that noise can affect communication by societies. Under natural conditions, vocal indivi- masking vocalizations, preventing animals from dual recognition has predominantly been studied communicating effectively. An important part of in primates, despite its importance for other the Brazilian economy is based on mining, which mammalian taxa. Here, we present experimental produces noise from heavy machinery, explosions, evidence that wild bats, Saccopteryx bilineata, can sirens, and road traffic. Mining is often conducted recognize individual group members using voice close to biodiversity hotspots, but studies on the cues in distress calls. Bats produce distress calls effects of mining noise on animals are lacking. when being caught by a predator, and distress The rufous-collared sparrow is a passerine found in calls often attract conspecifics. We conducted two Atlantic forest areas in Brazil. The aim of this study distress calls playback experiments to simulate pre- was to analyze the effect of noise from mining trucks dation events in the vicinity of the bats’ day-roost. on vocalizations of Zonothichia capensis. Data were In the first experiment, roosting bats approached collected in an Atlantic forest fragment located the speaker broadcasting distress calls of group close to a mine at Peti Environmental Station in members only when the simulated scenario was Southeast Brazil. Two digital field recorders (SM2, plausible (when the stimulus donor had just left the Wildlife Acoustics) were installed 150m from each roost), but never when the simulated scenario was other and 25m from a mining road. The SM2 were physically impossible (when the stimulus donor was set to record at 44.1kHz from 05:00 to 09:00 am still perched among the focal bats). The lack of re- during 7 continuous days in October 2012. The sponsiveness when conflicting information was pre- data analysis was done using Raven Pro 1.4 soft- sented (i.e. the physical presence of an individual ware according to the following steps: 1) Manual in one place and its voice in another place) strongly search for truck noise; 2) Check for presence of suggests that S. bilineata is capable of recognizing Z. capensis vocalizations in intervals of 1 minute individual conspecifics based on voice cues in before the truck transit; 3) Analyze the spectral distress calls. Since a simulated predation event at characteristics of the truck noise (duration, maxi- the day-roost likely represents a highly relevant but mum and minimum frequencies, and bandwidth); nevertheless rare scenario, we conducted a second and 4) Analyze the spectral characteristics of the experiment to test the bats’ responsiveness to vocalizations 1 minute before and during and 1 distress calls of unfamiliar conspecifics at different minute after the truck transit (duration, maximum distances to the day-roost. Focal bats reacted to and minimum frequencies, number of notes, the broadcasted distress calls when the speaker interval between the calls, and latency period after was placed directly at the day-roost or in close pro- the truck transit). Results showed that the species ximity to it, but not when it was placed at a nearby stops calling or decreases the duration and number foraging site, suggesting that foraging S. bilineata of notes during truck transit. These results indi- are not particularly attracted to conspecific distress cate that mining noise can affect a species’ vocal calls. Thus, the bats’ degree of responsiveness to behavior, which is important for the development distress calls apparently depends on the perceived of conservation strategies in natural areas close to social relevance of a given situation, and distressed mining operations. group members calling at the day-roost elicit the strongest response.

160 53 Eckenweber, Maria Knörnschild, M. Emam, Tariq University of Ulm, Germany Mackay, R. Barnard, M. Individual recognition in wild greater sac- University of Hull, U.K. winged bats (Saccopteryx bilineata) The Yorkshire Soundscape Project: a case study Abstract The ability to recognize conspecifics in soundscape comparison and composition

88 Abstract Recent developments in the field of Abstract Humpback whales, Megaptera novae- ecoacoustics have yielded different approaches angliae inhabit all major ocean basins and under- to environmental monitoring through the use of take long-distance seasonal migration between sound recording, including the capturing and ana- productive high-latitude areas where whales feed lysis of entire soundscapes, rather than individual in the summer and fall, and low-latitude tropical species (Pijanowski 2011). May 2015 has marked waters where mating and calving occur during exactly forty years since the World Soundscape winter and spring. Southern Chile, especially Project (from Simon Fraser University, Vancouver) the Chiloense Ecoregion, have been recognized visited places within Europe in order to produce as an important summer feeding ground for the Five Village Soundscapes and the accompanying southeastern Pacific humpback whale. The song European Sound Diary (both Schafer 1977). This of humpback whales is a loud, complex series of project focuses on the capturing and analysis of a sounds repeated over and over. It occurs primarily range of soundscapes using rural Yorkshire as its (although not exclusively) during the breeding study area, developing effective sound recording season, is sung only by males, and its composition and analysis techniques in order to enhance envi- changes as it is being sung with all singers in a ronmental monitoring; taking an interdisciplinary population singing the same version at any one approach, combining practices in sonic arts and time. In this investigation, we used the songs of ecology. It begins with a comparative analysis of humpbacks whales to study the seasonal patterns archived sound recordings from 1975 as part of in the presence of humpback whales in this area. the World Soundscape Project database at Simon We deployed one marine autonomous recordings Fraser University. This paper will present the first unit (MARUs) during six months (January to June year of study and results that have put into effect 2012). We used SONS-DCL software to automati- relevant field and compositional techniques to cally extract acoustic events from this unit. Our re- fuse methodologies on theoretical and practical sults indicated that there was a constant presence levels. These include field recording, ecoacou- of humpback whale in the south of Chiloense stics composition, spectral analysis, working Ecoregion, but there were an increases in the with aural archives and their philosophical presence in the last months studied. So, in June challenges. Composed sonic material will be we found the maximum presence of humpback previewed accompany the writing. The aims of whale vocalizations. In addition, we studied the this paper are in line with the growing focus on diary or circadian pattern of these vocalizations for ecology and contemporary composition; together, understanding more of the activities they were do- arts and sciences have seen, and continue to see, ing in these months over this area. In this context, a convergence of human feeling and ecological our results showed that were more vocalizations concern (e.g. from Payne 1970 to Krause 2014). in periods with low light or in dark (at night). Soundscape recording and ecoacoustic composi- These are the first recorders of humpback whale tion ‘could be a powerful means of raising awa- in this area, so for complete this investigation, we reness of acoustic biodiversity and its heritage.’ continue studying these vocalizations, as well as (Monnachi 2013). other sounds and biology variables that maybe will be influenced in humpback whale behavior.

23 148 Español Jimenez, Sonia Hucke-Gaete, R. Ey, Elodie Austral University, Argentina de Chaumont, F. van der Schaar, M. Olivo-Marin, J.-C. Universitat Politecnica de Cataluña, Spain Bourgeron, T. Institut Pasteur, Paris, France First study of presence of humpback whales vocalizations in a feeding ground of the New insights into spontaneous ultrasonic Southern Chile vocalisations in mice

89 abstracts poster

Abstract Rodent ultrasonic vocalisations are 2002 suggested several adaptive functions of these currently broadly used as proxies for communica- irregularities. Later the ‘unpredictability hypothe- tive abilities in the behavioural characterisation of sis’ was tested in several species (meerkats, prairie models for neuropsychiatric diseases. These com- dogs) providing support for the attention evoking munication signals are recorded in experimental function, but we still have little knowledge about contexts in which the motivation of the animals is whether the NLP provide information about the manipulated. This artificiality in the contexts does caller’s identity, quality or its inner state. Blumstein not allow mice to express their complete natural and Récapet suggested that nonlinearities can act as behavioural repertoire and might lead to biases in honest signals of the level of arousal and several in- interpreting data and difficulties in reproducing direct evidence suggest a link between the arousal results. We now aim at defining more ethologi- of the individual and the harshness of their sounds. cally relevant conditions to test mouse models of However, to our knowledge no studies attempted to neuropsychiatric disorders, and more specifically test this link directly so far. Dog whines can provide to evaluate social communication. As a first step, an excellent model for this, as NLP is often present we concentrated on the contexts of emission of in them, and we can collect and compare whines spontaneous ultrasonic vocalisations. We followed from dogs experiencing different level of stress continuously several groups of 3 adult female mice during separation from their owner. Thus in this during 3 days and 4 nights. Ultrasonic vocalisations study with using questionnaires about separation were recorded automatically, and simultaneously and vocal behaviour, we recruited family dogs that the corresponding video recordings were triggered. whine during separation, and according their Spontaneous isolated ultrasonic vocalisations were owners have or lack separation-related stress. We emitted when mice were foraging or resting without recorded their whines and measured the occurrence any contact. In contrast, sequences of complex of NLP. We compared the number of nonlinear ultrasonic vocalisations were emitted mostly during events and biphonations between dogs reported to social interactions of at least two mice. It appeared have separation problem and not. Our preliminary that the most complex sequences were emitted when results from the data of 50 subjects showed that all three mice were sniffing each other when arousal there were difference only between the occurrences appeared to be high. Such approaches will deepen of nonlinearities in the first fundamental frequency. our understanding of rodents as such, and therefore Significantly more nonlinear phenomena were found lead to more reliable characterisation of behavioural in the whines of the stressful subjects. Additionally phenotypes in rodent models of neuro-psychiatric we are currently running quantitative acoustical diseases. This information on mouse vocal communi- analysis and behaviour coding to find additional cation needs to be complement but it will be of high associations between the inner state of the subjects interest in mouse models of social communication and the acoustical parameters of their whines. Our deficits such as autism spectrum disorders. preliminary results are promising, suggesting that nonlinear phenomena in whines are possibly act as indicators of stress level in dogs. 71

Faragó, Tamás 3 MTA-ELTE Comparative Ethology Research Group, Budapest, Hungary Favaro, Livio Marx, A. Gamba, M. Eötvös Loránd University, Budapest, Hungary Fissore, C. Gnone, G. Non-linear phenomena in dog whines as Pessani, D. potential indicators of separation anxiety University of Turin, Italy

Abstract Although nonlinear phenomena are Acoustic cues to individuality and species in often present in healthy vocalizations, their com- Spheniscus penguins: a source-filter theory municative functions are still unclear. Fitch at al. in approach

90 Abstract The diversity of colonial lifestyles Yip, F. displayed by penguins presents different chal- Liao, C. lenges for vocal production and recognition. Bass, A.H. Banded penguins (Spheniscus spp.) are unusual Cornell University, Ithaca, U.S.A. as they are highly territorial, and build nests in underground burrows. Banded penguins also have The circadian vocal rhythm of a singing fish complex vocalisations (display songs), which are relevant for mate and parent-offspring recognition. Abstract Biological rhythms enhance animal Moreover, they produce contact calls to maintain survival and reproduction by synchronizing social relationships and group cohesion. We coll- behavior and physiology to cyclical changes in ected acoustic recordings from two large ex-situ the external environment. The hormone mela- colonies of African penguins (Zoom Torino, Italy) tonin is the chemical expression of darkness in and Magellanic penguins (Acquario di Genova, vertebrates, and plays a central role in entraining Italy) between 2012 and 2014. To determine daily activity to the day-night cycle. Although which acoustic proprieties of vocalisations have regulation of cyclical locomotion and feeding the potential to discriminate between different behaviors has been well characterized, circadian individuals, sexes, and species, we measured tem- and melatonin regulation of vocal behavior is not poral (e.g. duration), source-related (fundamental well understood. Knowledge of nocturnally vocal frequency, F0), and filter-related (formants) species is especially lacking, as diurnal songbirds acoustic features of each vocal signal. The acou- are the predominant models for studying vocal stic parameters were used to carry out a series of behavior. We investigated the nocturnal courtship stepwise, permuted discriminant function analysis vocalization (“hums”) of a highly vocal fish, the (with cross-validation). Our results showed that plainfin midshipman (Porichthys notatus). To test vocalisation could be classified according to the hypothesis that hums are under circadian con- the emitter, in a manner shown to be far greater trol, we recorded them from individual males held than that attributable to chance. Moreover, we under normal light-dark (LD, 15:9) and constant were able to determine the acoustic features dark (DD) cycles in artificial habitats mimicking that are potentially relevant for individual vocal natural conditions. The results strongly suggest recognition in both species. We also showed that that the daily cycle of midshipman vocalization species discrimination by calls is possible, despite is controlled by a circadian clock, although the the limited genetic distance between the African strength of the clock varies across individuals. penguin and the Magellanic penguin. However, we To test the hypothesis that melatonin stimulates failed to find sex-related vocal distinctiveness. midshipman vocalization, males who had been Overall, we demonstrated that penguin calls can humming under LD were transitioned to constant be studied by considering independent contribu- light (LL) to suppress endogenous melatonin tions from three different parts of the respiratory levels. On the day of transition to LL, one group apparatus, namely lungs (temporal patterns), was implanted with 2-iodomelatonin (2-Imel; vocal production organ (source, determining the potent melatonin receptor agonist) and another F0) and vocal tract (filter, responsible for formant with vehicle control. Preliminary data suggest that peaks). Our results provide the first evidence that humming activity decreases under LL in control- the source-filter theory can lead to a far better un- implanted and non-implanted males, but persists derstanding of biologically meaningful information in 2-Imel-implanted males. Together, our results contained in nesting penguin calls. We suggest demonstrate for the first time a circadian rhythm that this approach should be adopted to study in fish vocal behavior, and that melatonin action vocal communication in other penguin species. stimulates nocturnal vocal behavior. Knowledge of circadian and melatonin control of nocturnal vocalization in midshipman fish will complement 85 investigations in diurnal birds and inform comparative investigations in other nocturnally vocal Feng, Ni vertebrates. Support from NSF IOS1406515 and Ahn, G. IOS1120925.

91 abstracts poster

76 Kouroussis, G. Verlinden, O. Ferhat, Allain-T. Université de Mons, Belgium Ey, E. de Chaumont, F. Wildlife presence in a rural soundscape near a Institut Pasteur, Paris, France wind turbine

Contexts of emission of ultrasonic vocalisations Abstract Studies of wind turbine noise yield during social interactions in mice generous amounts of low quality sound data in various environments. In the present context, audio Abstract Social communication is heavily af- recordings near a turbine in the rural setting of fected in patients with neuropsychiatric disorders. Chevetogne, Belgium, are exploited to highlight Accordingly, mouse models designed to study the wildlife presence. Two weeks of 12000 Hz conti- mechanisms leading to these disorders are tested nuous sound data were measured at three positions for this phenotypic trait, especially for ultrasonic vo- around the turbine, at distances of 100 and 200 calisations during social interactions. Nevertheless, meters. Upon inspection, the audio records were little is currently known on the precise contexts of found to contain elements of anthropophony (the emission of these ultrasonic vocalisations during turbine, various vehicles, voices, church bells), social interactions. Therefore, in the present study, geophony (wind) but most interestingly a rich we aimed at synchronising the analyses of ultrasonic biophony (birds, insects and mammals). The de- vocalisations and specific events in social interac- tected bird species are common in the area. Color- tions during male-male interactions in freely moving composite spectrograms, assembled from acoustic animals (occupant / new-comer). We tested 8 pairs indicators that target wildlife detection, allow a of male mice in free dyadic social interactions, in direct visualization of 24 hours of sound data at three cage formats (rectangle, round, square). Under once. The acoustic complexity index (ACI) excels at these conditions, we examined the emission of highlighting the contributions of passerines, while ultrasonic vocalisations during the different types of a modified spectral entropy (Hs) captures the voices social contacts, approach-escape sequences, follow of birds with more monotonous songs. A third behaviour, and the time each animal spent in the dimension, derived from the sound pressure level, vision field of the other one. We did not highlight any gives a sense of the weight of human activities on significant effect of cage shape / size on the number the soundscape, namely through road traffic and of ultrasonic vocalisations recorded as well as on agricultural work. The final spectrograms display the behavioural events examined. In contrast, we the musical score of the day, showing birds singing showed that vocalisations were emitted during spe- on and off in succession or simultaneously but in cific behavioural sequences, especially during close different frequency ranges. Observations related to contacts or approach behaviours. A large proportion the coexistence of birds and the wind turbine are in of these vocalisations (30- 50%) occurred when the line with existing knowledge; common passerines are occupant was behind the new-comer. The types of heard at short distances from the turbine and limited ultrasonic vocalisations emitted also varied accor- raptor presence on site is captured (Buteo buteo). ding to the behavioural events in which they were emitted. The present study provides new information about the context of emission of ultrasonic vocalisa- 111 tions in a protocol reliably eliciting social contacts in adult male mice. This knowledge will be valuable in Florentin, Juliette standardizing investigation of social interactions in Kouroussis, G. mouse models of neuropsychiatric disorders. Verlinden, O. Université de Mons, Belgium

57 Acoustic characterization of woodpecker drumming in view of European species Florentin, Juliette identification

92 Abstract Recent studies on bird song identifica- restoration in Mecklenburg-Western Pomerania in tion have primarily used mean spectral content to northeastern Germany. Here we present the results describe vocalizations. This strategy is certain to of the first year of systematic evaluation of species fail for some species, for example when the tempo- composition during the breading time. Recordings ral structure of the song is critical. The drumming were made in 2012 from March 29th until May of woodpeckers falls into this last category. For 30th with a focus on nocturnal birds. Six sound woodpeckers, drumming achieves territory defense snippets of 15 min duration were recorded daily, and partner attraction to such an extent that starting at 3 a.m., 5 a.m., 7 a.m., 8 p.m., 10 p.m. some species have no song (Dendrocopos major, and 12 p.m. In total 372 sound snippets (93.5 Dendrocopos leucotos). Hence drumming sounds hours of four-channel recordings) were obtained. bear species markers and their acoustic description All recordings were evaluated by listening to the is a necessity. For that purpose, a database of 361 recordings supported by spectrographic visuali- drumming recordings is assembled from the online zation. Vocalizations of 62 bird species could be archives Xeno-Canto (the Xeno-Canto Foundation) detected. There was no recording with absolute and Tierstimmen (Museum für Naturkunde Berlin) silence. The number of bird species detected on with samples from the nine European woodpecker the sound snippets varied from 3 to 22 (median species that drum. Our objective is the recognition 12). The temporal distribution patterns allowed an of drumming in a continuous sound flow and the assessment of the status of the species. A regular identification of woodpecker species through a presence can be used as an indicator of breeding. generic supervised classification algorithm such as Applying pattern recognition techniques, the tem- KNN. 2666 drumming rolls are thus extracted from poral patterns of vocal activity could be described the Xeno-Canto and Tierstimmen sound files using on the example of the Eurasian bittern (Botaurus a characterization of drumming as the repetition of stellaris). Also the number of calling birds of this a brief signal with a beat between 30 ms and 120 species could be estimated based on the evaluation ms and with the bulk of the spectral components of four-channel recordings. The bioacoustic moni- lying below 1500 Hz. The drumming rolls are also toring revealed the occurrence of rare nocturnal sorted in series of the same bird hitting on the same birds such as Little Crake (Porzana parva), Spotted tree; this allows taking into account the possible Crake (Porzana porzana) and Short-eared Owl duets between different birds or the potential tree (Asio flammeus). However, some species which are changes. On that basis, the characteristic time known as common breeding birds in the study area interval between successive drumming rolls of the could not be found on the recordings. The potential same bird is evaluated. Other acoustic features of bioacoustic methods for long term monitoring of used in the classification are the duration of bird populations will be discussed. drumming rolls, their time structure (acceleration or deceleration while hitting) and a rough spectral description. The final feature ranges are compared 122 to existing literature data. Gervaise, Cédric chair CHORUS, Fundation of Grenoble Institute of 50 Technology & GIPSA LAB, France Lossent, J. Frommolt, Karl-Heinz France Energies Marines, Brest, France Museum für Naturkunde - Leibniz Institute for Di Iorio, L. Evolution and Biodiversity Science, Germany Fundation of Grenoble Institute of Technology, France Information obtained from bioacoustics monitoring – lessons from systematic acoustic 3D imagery of the marine biophony with monitoring during the breeding season compact sensor array

Abstract Using a four-channel recording setup Abstract In this communication, we address we provided a bioacoustics monitoring in a wetland the development of protocol (instrumentation,

93 abstracts poster

algorithms to process the data) to localize and them are narrow space gleaners and utilize on first characterize the marine biophony in 3D for coastal glance very similar echolocation calls. Foraging environment in a 100 m x 100 m x 15 m volume of style demands for special adaptations of call para- water. For the biophony, we focuss on the impulses meters such as intensity, frequency, call duration, ([1 ms, 20 ms] duration range, [2 kHz, 50 kHz] beam shape and bandwidth. Phyllostomid bats use bandwidth) produced by a large range of marine typically rather short multi harmonic and steeply species (snapping shrimps; sea urchin, bivalves, frequency modulated calls. Preliminary observa- noise from s. To localize these impulses, we use tions showed some phyllostomid species flying a small compact array of 4 hydrophones located mostly with opened mouth, whereas others have within a 2 m x 2 m x 2m cube with a ad hoc geome- their mouth closed during flight. We hypothesized try, for each impulse several features are extracted that open versus closed mouth might be used to (peak frequency, SPL, etc) and then the distribution modulate call parameters. We therefore recorded of the features in space may assessed and may help flying individuals of different species with a for the identification of species, behaviors or acti- multi microphone array (four microphones, three vities. We will present in detail the processing of CM16, Avisoft Bioacoustics, and one calibrated the data (detection of impulses, estimation of tdoa, ¼” G.R.A.S. type 40BF) and a high speed camera estimation of positions) and study the accuracy of (frame rate 500 fps). All experiments were per- localization both with theoretical argument and formed in Panama under standardized conditions in situ calibration using synthetic emitters. Then in a flight cage using animals captured on Barro we present the result of our 3D imagery of marine Colorado Island and the surrounding islands. biophony for two 7 days at sea experiments ran The multi microphone array allowed calculating in 2014 and 2015 to study via passive acoustics distance and direction to the bat and hence call monitoring the marine fauna and its dynamic at intensity, whereas frequency, call duration and artificial reefs (hosting rich marine biodiversity) on bandwidth were determined from single micropho- a soft substrat (hosting poor biodiversity). The first ne recordings. We plan to test for correlations of site is the artificial reef of Prado (Marseille, France) call parameters with type of emission and the life developed 6 years ago to mitigate the negative history of the species animals. Bats having their impact of Marseille of marine life. The second site mouth closed during flight emit calls exclusively is an unassigned petroleum terminal at Lanveoc through the nostrils, while bats with an open (Rade of Brest, France) where several pylons (10 mouth may emit calls through the mouth and/ m diameter) are settled by marine life. In each or the nostrils. Preliminary results show that the case, the ground truth for the distribution of marine type of emission varies distinctly between the sub- fauna is performed with divers. families: Frugivorous Carolliinae and sanguivorous Desmodontinae fly with an open mouth. So do the animalivorous Phyllostominae with the exception 9 of the omnivorous genus Phyllostomus, which flies with mouth closed. Frugivorous Stenodermatinae Gessinger, Gloria have their mouths closed during flight and emit the University of Ulm, Germany echolocation calls exclusively through the nostrils. Tschapka, M. University of Ulm, Germany and Smithsonian Tropical Research Institute, Balboa, Panama 139 Surlykke, A. Syddansk Universitet, Denmark Gifford, Toby Linke, S. Nasal and oral echolocation call emission in Karakonstasis, C. phyllostomid bats Barclay, L. Griffith University, South Brisbane, Australia Abstract The phyllostomids form the ecologically most diverse neotropical bat family. Although Vocalisation characteristics of the soniferous they show a huge range of feeding habits, most of catfish species Neoarius graeffei

94 95 96 Abstract The Blue Catfish is an introduced sion, and seem to differ between individuals, even soniferous fish abundant in the Brisbane River to the (trained) human ear. Whether the birds use which emits a distinctive grunt, suggesting the this information to communicate specifically is, to possibility of automated passive acoustic popu- date, unknown. We performed a two-level playback lation monitoring for this species. We have made study with a captive population in which focal indi- in-situ field recordings and analysed the acoustic viduals were held in short-term acoustic and visual characteristics of Blue Catfish daytime vocalisa- isolation from their conspecifics. We played back tions. During periods of high activity vocalisations stimulus calls of the partner and of other familiar are common - on the order of 1000 per hour for a group members, and, due to the pair’s crucial role in school. We found typical grunts to be 50-100ms the colony, expected different reactions if partners in duration, with 3-8 spectral bands with a falling could be vocally distinguished (level 1: recognition). contour of between 40-120Hz drop over the length In addition, these stimulus calls had been recorded of the grunt. There is a clear harmonic structure, previously, while stimulus birds were listening to with absent fundamental implied between 165- calls of our focal or of a different individual (level 190Hz, commonly 3-4 strong harmonic overtones, 2: addressing). Preliminary results suggest that jack- weak second and fifth harmonics, and strong third daws respond more strongly to their partner’s calls and fourth harmonics. We report on algorithmic and even more so when “addressed” by the partner. detection of grunts from a stereo hydrophone array We conclude that jackdaws are able to use vocal based on this heuristic characterisation. cues to distinguish between individuals, and that their contact calls mean more than merely eliciting group cohesion, but have the potential to be used in 79 communication between specific individuals.

Gill, Lisa Goymann, W. 118 ter Maat. A. von Bayern, A. Glotin, Hervé Gahr. M. Université Toulon, France Max Planck Institute for Ornithology, Seewiesen, Pavan, G. Germany University of Pavia, Italy

Who ya gonna call? Individual recognition in Scaled soundscape automatic annotation - jackdaw (Corvus monedula) contact calls evidences from the Sabiod EADM challenge

Abstract The ability to recognise familiar indivi- Abstract This talk presents the SABIOD EADM duals underlies most animal societies, to maintain (Environmental Acoustic Data Mining) challenge on social networks and ensure investment in closely the categorization of terrestrial soundscapes applied related individuals (e.g. parental care towards own to 40 days of recordings made in SABIOD-Italy. The offspring). To communicate, many species use vo- task consists into the classification in 11 categories calisations that may show inter-individual variation and the maximisation of the mean average precision reflecting differences in the anatomy of vocal organs. metrics computed from the similarity scores. Six of Some animal species have been shown to use the classes are biophony: insects, low song (at least such individual signatures at the receiver‘s end to 10 continuous seconds with structured patterns differentiate between individuals. But only very few of different notes below 1.5 kHz), high song (as species have been reported to use this information at before, but higher than 1.5 kHz), low call simple the sender‘s end – to “address” specific individuals. note (eventually in series) below 1.5 kHz (may be by Jackdaws (Corvus monedula) are highly social, long- birds or frogs), high call (as before, but higher than lived birds that nest in colonies and share nest de- 1.5 kHz), mammals (low frequency sounds, tonal or fence and parental care with their life-long partner, harsh). Other categories are anthropophony (planes), and utter various, not yet well-studied vocalisations. geophony (rain, wind, thunder), and an “other” event Contact calls are often used to maintain group cohe- class. We summarize and discuss the performances

97 abstracts poster

of the challengers, the used signal representation, nal adaptation for detecting bat species calling at classification method, and the requested CPU time. different frequencies at distances where they pose a Metrics to describe soundscapes will be discussed. similar threat. Combining neuronal measurements, EADM workshop will take place at the IEEE bat call recordings and bioacoustic modelling, we ICDM 2015 conference. http://sabiod.univ-tln.fr/ show that bat call frequencies predict bat threat eadm/#challenge predation threat. that moth auditory cells detect bats long before bats can detect the moth, and that this neuronal activity translates into similar temporal 41 safety margins for initiating escape flight across all sympatric bat species. Goerlitz, Holger R. Max Planck Institute for Ornithology, Seewiesen, Germany 34 ter Hofstede, H.M. Dartmouth College, Hanover, U.S.A Gogala, Matija Holderied, M.W. Trilar, T. University of Bristol, U.K. Slovenian Museum of Natural History, Ljubljana, Slovenia Linking lab physiology to behaviour in the wild: the multi-species predator-prey Differences in song parameters and taxonomy interaction of bats and moths of singing cicadas (Hemiptera, Cicadidae)

Abstract The ears of many insects are a classic Abstract During the last few decades it became example for predator-mediated selection. Eared obvious that by differences in song parameters most moths possess ears to detect and evade bats. species of close related cicadas can be easily dis- However, while the neurobiology of moth hearing tinguished in contrast to morphological comparison. is well studied, the escape behaviour is much less In many cases acoustic differences are very clear understood, despite being the predator-mediated and easily definable (e.g. Cicadetta montana, C. phenotype. Here, we aim to close this gap by, brevipennis, C. cantilatrix - sympatric and syntopic first, establishing lab-derived neural audiograms in Slovenia). However, recent detailed studies of the as a tool to estimate predator detection distances complex of mountain cicadas and some other groups in the wild and, second, to test the function of the of close related species of Tibicininae show in many seemingly exaggerated neuronal detection distances cases much subtler differences between the songs for predator escape behaviour. For our first aim, we of certain populations like C. olympica from Mt. recorded neuronal audiograms of twelve species of Olympus and related population from Mt. Taygetos noctuid moths in the lab, and additionally measured in Greece. Another example are close related species their detection thresholds for bat calls in the lab of Cicadivetta from Southern Greece. The question and field, and furthermore their detection distances is, how big must be differences in acoustic signals to for free-flying bats in the field. Next, we predicted prevent cross mating in the field. Behavioural tests neuronal call detection thresholds and -distances in such animals are almost impossible. Therefore based on these audiograms and compared them to differences must be statistically validated and com- the measured lab and field data, showing that the pared with possible zoogeographical, morphological predicted data were similar or slightly larger than the and molecular differences. But even DNA analysis of measured one. Thus, neuronal audiograms, which COI and COII does not always show clear results in can be readily measured in the lab, allow to predict distinction of putative species. predator detection distances in the wild, which are the ecologically and evolutionary relevant behavioural parameter that determines the distance and time 70 available for moths to escape from an attacking bat. In our second approach, we hypothesized that the Goncharova, Maria frequency-dependant hearing of moths is a functio- Klenova, A.V.

98 Lomonosov Moscow State University, RUS Simon, R. Friedrich-Alexander University Erlangen- Acoustic call variables indicate hunger level Nuremberg, Germany in Siberian crane chicks (Grus leucogeranus, Schnitzler, H.-U. Gruidae) University of Tübingen, Germany Tschapka, M. Abstract It is well known that chick calls can University of Ulm, Germany; Smithsonian Tropical convey information about the physiological state. Research Institute, Balboa, Panama However the vast majority of studies have focused on chick’ vocal activity in connection with hunger The sound of cactus: a flower-visiting bat’s level, and they agreed that it was a good indicator perspective of chick’s needs or condition. Fewer researches were devoted to the temporal-frequency variables of the Abstract Bat-pollinated plants present specia- begging calls and different studies showed opposite lized characteristics that attract and guide their results. Here we experimentally tested whether vocal pollinators towards the flowers. Recent studies activity and temporal-frequency variables of the have shown that some of these flowers evolved begging calls of human-raised Siberian crane chicks characteristic acoustic features that facilitate their are related with their hunger level. The study was detection by echolocating bats. Many species of conducted in 2012 in Oka Crane Breeding Centre cacti depend on bats for pollination, however, the (Russia). We recorded calls of 9 individual chicks in echo-acoustic properties of their flowers are largely two conditions: after experimental food deprivation unknown. For assessing what acoustic cues are (hungry chicks) and than during additional feeding available to bats for finding open cactus flowers (replete chicks). Each recorded session lasted 5 we used a biomimetic sonar head to measure echo min. We measured 9 variables in first 20 calls in impulse responses of an empty cactus branch of each recording session (360 calls in total), as well as Pachycereus pringlei, as well as branches with a total time spent on vocalization in each session. We bud and an open flower, respectively.Additionally, found that all measured frequency variables (except in order to identify the floral structures relevant minimum fundamental frequency and second energy for acoustic localization of the flower opening we quartile) in hungry chicks were significantly higher measured echo impulse responses of an intact than in replete ones (e.g. maximum fundamental cactus flower and four cactus flowers that had been frequency in hunger chicks: 3.41±0.53, in replete: modified in different ways. We found thatechoes 3.34±0.53; F1, 350=5.80, p<0.05). Calls of reflected by a plain cactus branch differed subs- hunger chicks also were longer, noisier and followed tantially from those reflected by one with an open each other through smaller gaps. The total time of flower. Flowers reflected echoes of higher intensity vocal activity in hungry chicks was also greater than and over a wider angle than a plain cactus branch, in replete chicks, though this difference was insig- which makes them acoustically highly conspicuous nificant. Thus the level of vocal activity, according to the bats. Flowers without petals showed the to previous studies of bird begging behaviour, is weakest echo of all targets and the most different quite universal indicator of hunger level in chicks. reflective pattern. Our results show that the petals At the other hand, we found that temporal-frequency are the acoustically most perceivable structures variables of Siberian crane chick calls also related that may help bats to locate the opening of the with hunger level. We discuss how the way of vocal cactus flowers. Additionally, our result suggest that hunger level expression is related with ecological the distinct reflections of cactus flowers may allow traits of different species. The study was funded by the bats to distinguish between cactus branches Russian Scientific Foundation (grant 14-14-00237). with or without flowers by just using echolocation.

156 102

Gonzalez-Terrazas, Tania P. Grison, Matthieu University of Ulm, Germany Sèbe, F.

99 abstracts poster

Grison, J. studied species, leks constitute permanent social Mazagol, P.-O. networks with stable and long-term hierarchies, Mathevon, N. and that this social structure primarily results from Université Jean Monnet, St. Etienne, France signaling interactions between males. Aubin, T. Huetz, C. Université Paris Sud, France 132 Quispe, R. Max Planck Institute for Ornithology, Seewiesen, Hallerberg, Sarah Germany Meigel, F. da Silva, M.L. Dahlkemper, M. Universidade Federal do Pará, Belém, Brazil Timme, M. Max Planck Institute for Dynamics and Self- Spatial and temporal dynamic in a lek of a Organization, Göttingen, Germany tropical bird: the Lipaugus vociferans Vester, H. Ocean Sounds e.V., Germany Abstract In some bird species, males use com- Hammerschmidt, K. munal area, known as “leks”, to perform courtship German Primate Center, Göttingen displays. Leks are generally restricted to a mating season and are of primary importance for females’ A bag-of-coefficients approach towards mate choice. In some species however, leks remain call categorization active all year-round and the dynamics of their social organization remain poorly explored. One functional Abstract Technical advancements in bioacou- hypothesis is that these leks could constitute social stics: From sound analysis to telemetry. Besides networks where males compete for both spatial humans, several marine mammal species exhibit and temporal resources to establish long-term and prerequisites to evolve language: high cognitive possibly stable hierarchies. Here we focus on the abilities, flexibility in vocal production and ad- Screaming Piha Lipaugus vociferans, where males vanced social interactions. Here, we describe and use to form permanent leks of 10 or more individu- analyse the vocal repertoire of long-finned pilot als. They produce one of the loudest songs, which whales (Globicephalus melas) recorded in northern consists in two parts: one introductive bass roo-roo Norway. Observer based analysis reveals a complex followed by loud syllables pee-peeyo. We studied vocal repertoire with 140 different call types, call the acoustic and social networks of Screaming sequences, call repetitions and group-specific Piha’s leks during a whole year, providing the first differences in the usage of call types. Recently, description of the spatial and temporal dynamics the bag-of-calls approach has been proposed to of a permanent lek in a non songbird species. 44 distinguish ensembles calls recorded from different individuals from 4 leks have been recorded and groups of pilot whales. In this contribution we test localized during three recording sessions between in how far a similar concept, the bag-of-coefficients April 2011 and April 2012. By analyzing the approach is sufficient for distinguishing different acoustic structure of males’ song, we found an call types within the recordings of one group. individual signature and characterized its coding features. We showed that this signature remains stable enough during one year to survey each male 16 by its vocalizations only. Using microphone arrays, we thus investigated the year-round dynamics of the Hoebel, Gerlinde lek’s acoustic network. We found that individuals University of Wisconsin- Milwaukee, U.S.A. are faithful to the same lek all year, and the lekking place in the center of the lek is more stable than Does audiovisual integration change leading the fringes. Our study gives interesting insights of call preferences in treefrogs? the functional dynamics of lek in tropical forest, and supports the hypothesis that, at least in the Abstract Humans perceive two sounds in

100 close temporal succession as a single event ori- Abstract Mice (Mus musculus) produce complex ginating from the location of the leading sound. vocalizations in the ultrasonic range (ultrasonic Surprisingly, a visual cue associated with the vocalizations = USVs) that remind of high pitched leading sound enhances sound localization, while birdsong when shifted into the human hearing ran- a visual cue associated with the lagging sound ge. Male USVs show signs of individuality as well inhibits it, indicating that auditory spatial percep- as family and population membership and could tion in humans is a fundamentally multisensory be a signal that is evaluated by females during process. Many animals also attend to the sequence mate choice. USVs are however not restricted to at which mating displays are perceived. For ex- the mating context, suggesting that they serve a ample, females of many frog species will approach more general function in social communication of the leading of two calls in close succession, resul- mice. We investigated the vocalization behavior ting in higher mating success for the leading male. of partially interacting female mice under semi- Combining techniques used to study communica- natural conditions over an extended time course. tion in frogs with those used to study perception We analyzed whether certain features of their USVs in humans, I investigated whether audiovisual were specific to environmental stimuli (food, water, integration affects leader preferences in treefrogs. nesting site, male cues) or to two types of social in- If so, addition of a visual cue to the leading call teraction (direct and indirect). Our results support should induce or strengthen a leader preference. the idea that female USVs are mainly used as social Conversely, addition of a visual cue to the lagging signals and that they play a role during hierarchy call should reduce a leader preference. I examined establishment and familiarization. We found that this phenomenon in two species of treefrogs that female mice use their USV differently depending differ in their leader preferences: Gray treefrogs on the type of social interaction. Females emitted have calls comprised of a series of pulses – they do less complex vocalizations at lower frequencies not have a preference for leading calls in general, when they encountered another female, as opposed but they do prefer leading pulses. Green treefrogs to high pitched complex vocalizations when they have tonal calls, and they have a preference for encountered another female’s urine and deposited leading calls. I found that addition of a visual cue their own. The two vocalization types seem to affects leader preferences under some circum- resemble what is known from birdsong as directed stances, but that audiovisual integration is not and undirected song and are likely to serve distinct a general feature of treefrog sensory perception: functions in mouse communication. addition of a visual cue does not change existing leader preferences (i.e., leading call preference in green treefrogs and leading pulse preferences in 133 gray treefrogs remain unchanged), but it can induce a slight leader preference where there initially Honaiser Rostirolla, Tomás was none (i.e., induce a leading call preference in de Araújo, G.F. gray treefrogs). I discuss the implications of these Barboza Capella, D. findings in light of sensory perception and mate Teixeira Moreira, A.P. choice. Universidade Federal de Santa Catarina, Brazil

Model reliability for automated bioacoustic 144 recognition through Arbimon

Hoier, Svenja Abstract With the recent development of better Von Merten, S. software recognition technology, perspectives for Pfeifle, C. bioacoustic monitoring arise as a reliable source of Max Planck Institute for Evolutionary Biology, data regarding studies in biodiversity, environmen- Plön, Germany tal policies, restoration initiatives and so on. This work investigate the efficiency of Arbimon-acoustic Context specific ultrasonic vocalizations in Web Aplication for automatic recognition models on wild female mice (Mus musculus) three bird species found at Santa Catarina, Brazil.

101 abstracts poster

The recordings belong to Arquivo Bioacústico perspecies have been widely studied also regarding Catarinense, where we selected 178 sound sam- vibrational communication. The four chromosomal ples of Attila rufus, Basileuterus culicivorus and species differ in their courtship calls as well as in Setophaga pitiayumi. Analysing song’s characters seismic signals.We studied the whole repertoire of presence stabilit, we created two range models for all four species of S. ehrenbergi superspecies. The each species: whole song (WS) and stable song ele- repertoire consists of six true vocalizations (court- ments (SSE), calculated through a χ²-distribution ship, cheep, chirp, snort, squeak and squeal) and (p > 0,95). Samples were uploaded at the Arbimon two mechanical sounds (teeth grinding and seismic II platform and used to create the range models, signal). The four species differ in use of the calls each with 5 random samples, tested with fitting of probably in accordance with their diverse degree of 20 present and 20 absent sound samples. The suc- aggressiveness. cessful detection rate of species presence on the whole list, for each range model was: A. rufus: WS = 62% and SSE (GL=4) = 57%; B. culicivorus: WS = 157 73% and SSE (GL=3) =63%; and S. pitiayumi: WS = 66% and SSE (GL=13) = 89%. In most cases, Hügel, Theresa whole song range models made the most reliable Max Planck Institute for Ornithology, Seewiesen, models for automatic recognition, a good perspecti- Germany and Julius Maximilians University ve for monitoring works started from scratch. Stable Würzburg, Germany song elements range model got promising results, van Meir, V. although it requires a prior library of song samples Clarin, B.-M. for modeling, available at Arquivo Bioacústico Siemers, B.M. Catarinense and other sound libraries. From further Goerlitz, H.R. work with other species and false-positive detection Max Planck Institute for Ornithology, Seewiesen, more accurate models should come. At last, our be- Germany lief is that Arbimon represent a great breakthrough in biodiversity monitoring and automated analysis Acoustic and ecological similarity: eavesdrop- tool, and because of its global range for biodiversity ping on heterospecific echolocation calls in analysis at long term, and we strongly suggest that Myotis bats monitoring initiatives consider the possibility of joining this network and also have a sound library Abstract The heterospecific recognition hypothe- supporting them. sis predicts that animals recognize heterospecifics and respond with species-specific adaptive behaviours. This ability should be adaptive, provided that 27 species-specific reactions to interspecific encounters are advantageous, for example by following Hrouzková, Ema foraging individuals that occupy a similar ecolo- Matêj, Šklíba J. gical niche. Increasing evidence shows that bat Šumbera, R. echolocation calls, primarily used for orientation University of South Bohemia, Ceské Budêjovice, and foraging, also enable communication between Czech Republic individuals within a species. However, data about the role of echolocation calls for interspecific infor- Vibrational communication of blind mole rat mation transfer is scarce. In contrast, the acoustic (Spalax ehrenbergi) superspecies similarity hypothesis predicts that animals simply react to heterospecific calls with similar spectro- Abstract The blind mole rat (Spalax ehrenbergi) temporal structure, without explicitly distinguishing superspecies comprise four chromosomal species between con- and heterospecific calls. Here, we in Isreal. Their habitat vary in degree of aridity in tested if bats are attracted to the echolocation calls temperature, with S. judaei inhabiting dry and hot of species that occupy the same foraging niche, southern deserts and S. golanii living in humid and independent of the calls’ spectro-temporal struc- relatively cold Golani heights. The S. ehrenbergi su- ture. We conducted three different playback expe-

102 riments with one focal species, Myotis capaccinii, effectively measure the flight of tethered moths in a European insectivorous trawling bat, testing its a standardized way and in response to different reactions to its own calls and those of four different acoustic stimuli. Our second aim is to use this sys- playback species with similar or dissimilar foraging tem to obtain behavioural audiograms of different ecology and call structure. In a flight room (EXP. moth species differing in neurobiology, ecology 1), which was separated into two compartments, we and additional anti-predator strategy. We therefore measured individual flight duration in the playback stimulated moths with pure tones of different compartments as a measure of attraction. In the frequencies and increasing intensities to obtain the field, we evaluated the effect of playback type on threshold intensities at which their flight behaviour individual flight behaviour using acoustic tracking changes. These behavioural audiograms will first be of individual flight trajectories (EXP. 2), and on ove- compared to existing neuronal audiogram to estab- rall group activity by counting the number of flight lish a link from neuronal activity to elicited flight passes (EXP. 3). Results differed between experi- behaviour. Second, audiograms will be compared ments: M. capaccinii was attracted to the calls of across species to test the hypothesis that due to a M. capaccinii and M. daubentonii (EXP. 1), only to higher detection risk, larger moths initiate evasive M. capaccinii (EXP. 2) or none at all (EXP. 3). These flight at larger distances than smaller moths. Third, results support neither hypothesis, but suggest that audiograms will be compared across species to test that both conditions, ecological and acoustical si- the hypothesis that moths that are protected by milarity, are required for information transfer across additional passive or active defence mechanisms species boundaries, yet also highlight the difficulty possess higher thresholds than moths without such of assigning degrees of “similarity” to different additional mechanisms. sympatric competing species.

103 40 Ilina, Irina Hügel, Theresa Marova, I.I. Goerlitz, H.R. Moscow State University, Russian Federation Max Planck Institute for Ornithology, Seewiesen, Germany Comparative analysis of the song variation across the ranges of Chiffchaff (Phylloscopus Behavioural audiograms to study anti-predator collybita) and Willow Warbler (Ph.trochilus) adaptation in eared moths Abstract We studied song variations across the Abstract Animal prey exhibit many different ranges of two closely related species: Common adaptations to predation threat. One common Chiffchaff and Willow Warbler. In Willow Warbler adaptation is unpredictable movement known as 3 poorly differentiated subspecies are normally “protean behaviour”. The defence behaviour of ea- recognized. By contrast, Chiffchaff is treated red moths´ consists of two stages, directional flight recently as superspecies including 12 races 4 of in reaction to a distant bat, and erratic flight, which which have status separate species. In order to moths exhibit to escape attacking close-by bats, a estimate the extent of intraspecific differentiation classic example for protean behaviour. While the in vocalization of both species we studied 10 geo- neurobiology of moth hearing is well studied, the graphical populations of 8 subspecies of Chiffchaff: escape behaviour is much less understood, despite Ph.c.collybita (W Europe), Ph.c.abietinus (Moscow, being the predator-mediated phenotype. Since the Kiev, E Ukraine), Ph.c.tristis (Central Siberia), predator-prey interaction between echolocating Ph.c.menzberii (W Kopetdag), Ph.s.sindianus bats and eared moths is solely mediated by sound, (Pamir-Alai Mts), Ph.s.lorenzii и Ph.c.caucasicus it is a powerful model system to study the function, (Caucas), Ph.brehmii (Iberian p-la) as well as 8 ecology and evolution of auditory-guided behaviour. population belonging to 3 subspecies of Willow Our first aim was to establish a sensitive computer- Warbler: Ph.tr.trochilus (W Europe), Ph.tr.acredula controlled force-transducer system, enabling us to (Moscow, Kostroma, North and South European

103 abstracts poster

Russia, South Ural, Mirnoe), Ph.tr.yakutensis of rivals. We considered two scenarios: (1) song (Eastern Siberia). We use canonical discriminate duration conveys information about male identity analysis of a number of characteristics of songs: and males should respond stronger both to elon- maximum and minimum frequency, bandwidth, gated and shortened song phrases in comparison number of notes and note types per one song, etc. to natural duration neighbour songs; (2) song Acoustic differences between different populations duration changes are associated with aggressive of Chiffchaff (Wilks Lambda =0.00357; p<0,001) motivation, thus elongation and shortening should were found to be better expressed then between affect response of rivals in opposite directions. We populations of Willow Warbler (LW = 0,042; p<0, tested 18 and 17 different males in two paired 001). The Chiffchaff subspecies fall into two experiments (response to natural vs. elongated or major groups: Europe green races and Asian brown natural vs. shortened songs). We found no support races by complex genetic and bioacoustic parame- for the first hypothesis. The way of response to dif- tries. Difference between brown races are greater ferent treatments reflected motivation hypothesis. (LW=0,00669, p<0.001) then between green races Shortened songs did not evoke such immediate and (LW=0,08843, p<0.001). The taxonomic structure strong response as normal duration song did, and of Willow Warbler is not clear so far, different re- response to shortened songs was weaker than to searcher separate from 3 up to 5 subspecies at the elongated songs. We also found that the order of different areas. According to our data only western- treatments affected response significantly. If birds most and easternmost geographical populations of were first responding to normal duration songs, they that species show reliable bioacoustic differences approached loudspeaker faster in treatment with from other population (LW = 0,144; р<0,001) .It’s shortened song. Our results suggest that song phra- possible that different taxonomic and bioacoustic se duration is used for mediating escalation level structure of these two closely related species of territorial conflicts. (Financial Support: National of leaf-warblers could be explained by different Science Center, Poland, grant no. NN303807340). history of dispersion which was connected with the Glacial epoch. 83

42 Jansen, Martin Masurowa, A. Jakubowska, Aleksandra O´Hara, R.B. Osiejuk T.S. Senckenberg Research Institute and Natural Adam Mickiewicz University, Poznán, Poland History Museum, Frankfurt, Germany

Song duration mediates response of territory Individuality in face of plasticity: signal owner during simulated intrusion in a songbird variation during sustained calling in frogs species with small song repertoire Abstract Recent developments in passive acou- Abstract Song duration is an important dimen- stic monitoring and acoustic pattern recognition sion of bird song variability affecting response of now enable us to comprehensively record and ana- signal receivers. In discontinuous singers males lyze acoustic events over longer periods of time. In increase singing effort rather by increasing song this project we studied the call variation of several rate than by elongating phrase duration, however, individual white-lipped frogs (Leptodactylus) over often they also change song phrase duration and several nights. We tested the potential of individual it could be meaningful too. In Emberiza buntings coding (PIC), i.e. if calls were statistically discrimi- dominates relatively short songs birds often pro- nated between each individual. In contrast to other duce song phrases with reduced duration but the studies where the test of PIC is usually based on meaning of such behaviour is not well understood. a rather small sample size, our analysis is based We tested experimentally whether elongation on the whole nightly calling activity of individuals and shortening of neighbour song in the ortolan (usually more than 10.000 calls of one individual bunting (Emberiza hortulana) affects response per night), and thus, it takes into account an

104 unexpected high intra-individual variation in spec- with alarm calls to predators in all of the trials tral call traits which are usually suggested to be even when foraging alone (n=18), while squirrels responsible for individual recognition in other frog foraging alone responded with only behavioural species. We report on variation in several other call changes (n=19). Neither species reacted with traits including calling effort during and between alarm calls to the control (spotted dove) trials. Both nights, and we ask if the markedness of the po- babblers and squirrels responded to the alarm calls tential individual signature varies during the night. of the other species with behavioural changes in all trials (n=4 each) and sometimes with alarm calls of their own. The results so far indicate that babblers 62 communicate predator threats to both heterospecifics and conspecifics, while both squirrels and Jayasena, Nilmini babblers obtain information from alarm calls of the Galappaththi, H.G.S.S. other species. Alahakoon, A.A.C.B. University of Peradeniya, Sri Lanka Sieving, K. 37 University of Florida, Gainesville, U.S.A. Jonsson, Thorin Interspecific communication for predator Montealegre-Z., F. awareness between yellow-billed babblers and University of Lincoln, U.K. common palm squirrels Clarke, D. Robson-Brown, K.-A. Abstract In multi-species associations, animals Robert, D. may gather information from heterospecifics as well University of Bristol, U.K. as from conspecifics. We hypothesise that the common palm squirrel (Funambulus palmarum) forms Biophysics of a pronotal Helmholtz resonator foraging associations with flocks of yellow-billed enhancing sound radiation in a neotropical babblers (Turdoides affinis) to acquire information bushcricket on predators. We predict that babblers would respond with alarm calls to predator presence more Abstract In the animal kingdom, male calling frequently than squirrels and that squirrels would songs are produced for the purpose of attracting change their behaviour in accordance to babbler females over long distances. In many cases, calls are alarm calls. We also hypothesise that each species loud and repeated at a high rate over a long time span. would respond to the other’s alarm calls indicating These advertisement calls involve comparatively the ability to gather information from heterospecific high sound pressure, as these propagate over greater vocalisations. We are testing these hypotheses by distances and therefore have greater attraction. But conducting playback trials of predator calls with producing loud calls is metabolically costly, and, presentation of predator models using predators due to small body size, animals like insects can be common to both species (shikra and domestic poor sound radiators. Therefore, in some acoustically cat). We are conducting similar trials on squirrels active animals, there is significant benefit from pro- and babblers foraging alone to investigate whether ducing sound amplification using specialised body presence of either species influences vocalisation. structures, or exploiting signal-enhancing structures We are testing for interspecific communication from the environment, like baffles or burrows. In this between the two species by playing back heteros- work we report on the biophysics of an unconventio- pecific alarm calls to each species when foraging nal body cavity, the pronotal chamber, developed in alone. In squirrel-babbler associations, babblers males of a small, neotropical bush-cricket species responded with alarm calls to predators in all of (Acanthacara acuta; Tettigoniidae: Agraeciini) to en- the trials (n=19). In 95% of the trials the initial hance signal transmission. Apart from the holotype, squirrel response was to babbler alarm calls and only a few specimens have been collected from A. not to the playback. None of the focal squirrels acuta. Generally, male bush-crickets produce mating responded with alarm calls. Babblers responded calls by tegminal stridulation, where one wing pos-

105 abstracts poster

sesses a scraper that is rubbed against a file on the by males with evidence of antiphon (frequency contralateral wing. Males of A. acuta are one of the range: 1400–22000 Hz), that are produced during loudest animals heard at night in the Ecuadorean contact of two or more individuals. In addition, for cloud forest, with the main spectral energy of their males was registered production of more complex calling songs peaking at ca. 10 kHz. They are flight- chirps before each act of copulation. In courtship less and the wings have been reduced to stridulatory chirps production also used ridges on costal margin structures that are completely hidden under a large of elytra and hind legs. The same structures on inflation of the pronotum, the pronotal chamber. costal margin of elytra previously discussed as a Using micro-scanning laser Doppler vibrometry, and possible stridulatory apparatus on other species of broadband stimulation, both passive non-invasive Curculionidae, but their involvement in sound pro- methods, we show that this chamber resonates at duction hasn´t been observed before. This study was the frequency of the calling song and at the resonant supported by the Russian Federation Presidential frequency of the wings. 3D computer models of a Grant: MK-4422.2015.4 male A. acuta obtained from micro-CT scans, and finite element modelling suggest that the physical properties of the pronotal chamber approach those 6 of a Helmholtz resonator. King, S.E. Fehrenbach, A.K. 35 Johnson, J.R. Smith, M.E. Kerchev, Ivan Western Kentucky University, Bowling Green, Institute of Monitoring of Climatic and Ecological, U.S.A. Systems of the Siberian Branch of the Russian Academy of Sciences, Tomsk, RUS Sound-induced hearing loss and recovery in the axolotl (Ambystoma mexicanum) Investigation of acoustical communication and behavior of invasive fir bark beetle Abstract The axolotl (Ambystoma mexicanum) is Polygraphus proximus Blandford used as a model organism for studying development, genetics, and regeneration. Although the sensory Abstract Four eyed fir bark beetle is the natural hair cells of the axolotl lateral line are able to re- destructor of weakened fir trees in its origin area in generate, it is unknown whether this also occurs in the South of Russian Far East, China, Japan and the inner ear.We recorded auditory evoked potentials Korea. Presumably P. proximus was transported in (AEPs) of six axolotls at eleven frequencies in order Southern Siberia by the Trans-Siberian Railway on to produce baseline audiograms of underwater pres- unpeeled wood packaging material. In the region sure sensitivity. Individuals were then subjected to of invasion it became the most destructive pest of a 48-hour, 150 Hz sound exposure at approximately Siberian fir Abies sibirica Ledeb. due to absence on 170 dB (re 1 µPa). AEPs were then performed to its new host tree effective defense against invasive measure hearing thresholds at 0, 2, 4, and 8 days pest and its ofiostomatoid associates. The aim of post-sound exposure (DPSE). In the baseline audio- this study was to eliminate a whitespace in know- gram, axolotls were most sensitive at 600 Hz, with ledge of P. proximus behavior with emphasis on its an additional peak of sensitivity at 3 kHz. Following acoustic component. The presence of elytro-tergal sound exposure, axolotls experienced a 6 to 12 dB stridulatory organ has been described for this spe- temporary threshold shift (TTS). Hearing sensitivity cies by Sasakawa and Yoshiyasu (1983). According returned to control levels within 8 DPSE, indicating to critical comments on paper mentioned above that axolotl hearing sensitivity recovers following from C.H.C Lyal &T. King (1996), the morphology acoustical trauma.This study is the first to document of P. proximus males stridulatory ogan in our study hearing loss and recovery in the axolotl. Future stu- was revised. The normal plectrum instead plots of dies are needed to correlate these shifts in hearing spines as it described previously is found. Behavioral sensitivity to sensory hair cell loss and regeneration studies have shown the producing of rivalry chirps in the axolotl inner ear.

106 18 Hunger advertising by begging calls in three auk species (Alcidae, Charadriiformes) Kipper, Silke Technische Universität München, Germany Abstract Begging behaviour is important element Bartsch, C. in the parent-offspring conflict. However the majority Freie Universität Berlin, Germany of the researches in this area have fully been based Weiß, M. on call counts, and they agreed that vocal activity Federal Institute for Risk Assessment, Berlin, was a good indicator of chick’s nutritional need Germany and/or condition. Fewer researches were devoted to the temporal-frequency variables of the begging Song features are predictive for male paternal calls themselves and different studies showed effort in Common nightingales (Luscinia opposite results. I studied begging behaviour in megarhynchos) three burrow-nested, uniparous species of auks: crested auklet (Aethia cristatella), parakeet auklet Abstract The ‘good-parent’ model predicts that (Cyclorrhynchus psittacula), and horned puffin male sexual ‘ornamentation’ honestly advertises (Fratercula corniculata). These objects provides the the degree of paternal investment. In passerines, opportunity to analyze the signaling value of begging male song is indicative of male attributes and for calls in the absence of confounding factors such few species it has been shown that song features as nestling competition and predation pressure. also signal paternal investment to females. Males of The study was conducted in 2013 on the seabird nightingales Luscinia megarhynchos are famous for colony of Talan Island (Sea of Okhotsk). I recorded their elaborate singing but so far there is only little calls of individual chicks (7-9 chicks per species) knowledge on the role of male song characteristics in two conditions: during natural feeding and after in intersexual communication, and it is unknown experimental four-hour food deprivation. I found that whether male song predicts male parenting abilities. almost all measured acoustic variables contain in- We studied the potential relation between courtship formation about the chick’s state in all species. The song and later parental investment on 20 males hungry chicks produced calls higher in fundamental of a color-ringed and RFID-tagged population of frequency and power variables and at higher calling nightingales in Golm (Potsdam, Germany). Using rate compared to naturally feeding chicks (e.g. in RFID technology to record male feeding visits to the hungry crested auklet chicks the mean peak funda- nest we found that nightingale males substantially mental frequency was 2,65±1,16 kHz; in naturally contribute to chick feeding. In addition, we analyzed feeding chicks – 1,76±0,66 kHz; ANOVA: F=98,7 male nocturnal song (prior to breeding) with focus on p<0,001). The effect of food deprivation on acoustic song features that have been shown to signal male variables exceeded both the effects of individuality quality before. We found that several song features, and species. In all species, the frequency variables namely measures of song complexity and song se- were stronger affected by hunger than the calling quencing, were correlated with male feeding rates. rate and call duration. I suppose that so strong Moreover, the combination of these song features changing of acoustic variables after food deprivation had strong predictive power for male contribution to may be explained by the absence of vocal individual nestling feeding. These results suggest that females identification in these birds. As parents do not need may assess future paternal care on the basis of song to check their chick’s individuality in the burrow, features identified in our study. Additionally we un- which they find visually, the chicks could use all of derline the importance of multiple acoustic cues for the acoustic variables to communicate about their female mating decisions especially in species with nutritional needs Supported by Russian Scientific complex song such as the nightingale. Foundation (grant 14-14-00237).

72 38

Klenova, Anna Klenschi, Elizabeth Lomonosov Moscow State University, RUS Guarato, F.

107 abstracts poster

Jackson, J.C. Centre Scientifique de Monaco (CSM), France; Windmill, J.F.C. Université de Strasbourg, France; Centre National University of Strathclyde, Glasgow, U.K. de la Recherche Scientifique, Strasbourg, France Aubin, T. Technological advances in 3D micro-scanning Université Paris XI, France laser Doppler vibrometry: investigation of the Bonadonna, F. locust ear Université de Montpellier, France

Abstract In order to better understand insect Communication in colonial long-lived seabirds: hearing, different approaches must be combined to relative contribution of acoustic and visual investigate how insect ears detect the different pro- signalling in sexual selection in penguins perties of a sound wave, how this signal is transmitted to the central nervous system, and how this in- Abstract In order to understand fundamental formation translates into an appropriate behavioural evolutionary processes, we need to study beha- response. The tympanal hearing of the desert locust vioural and physiological mechanisms involved Schistocerca gregaria has already been the subject in sexual selection. Sexual selection is based of numerous studies investigating membrane on intraspecific signalling modalities including mechanics, neurophysiology, and ethology over the visual and acoustic signals. In this study, we aim past few years. In this context, we are studying the to investigate the role of calls and ornaments in impact of tympanal structure on frequency analysis sexual selection in a non-nesting versus a nesting in the locust. Membrane mechanics have already penguin species: the king penguin (Aptenodytes been investigated in locusts using laser Doppler patagonicus) and Adélie penguin (Pygoscelis vibrometry to record tympanal displacement in adeliae).Data will be collected at three different previous studies. Owing to this approach it is known field sites: Crozet and Kerguelen sub-Antarctic that, in locusts, sounds of given frequencies will Archipelagos and Terre Adélie in Antarctica. generate travelling waves across the tympanum that We will record a) display calls of courtshipping propagate to different locations depending on the individuals, b) physiological (blood samples, body frequency of the initial stimulus. These travelling mass) and morphological parameters (beak and waves thus allow for the incoming stimulus to be flipper length), c) reflectance measurements and concentrated in the region of the tympanum best size of king penguin’s ornaments, and d) nest or tuned to its frequency, a phenomenon which can be territory features. The majority of those birds are regarded as a first, mechanical step in the process being long-term monitored using automatic identi- of frequency analysis. The research addressed in fication systems, and we have precise life-history this work applies recent technological advances information about them: 10.000 king penguins in 3D micro-scanning laser Doppler vibrometry and 2000 Adélie penguins have been fitted with to measure membrane displacement. Thus, in passive integrated transponder tags since 1998 addition to the travelling waves identified through and 2009, respectively.Although both penguin vibration measurements in only one dimension by species emit calls that are similarly structured, previous studies, this work has for the first time consisting of a series of syllables separated by detected and identified membrane oscillations in strong amplitude declines, calls differ in their three dimensions simultaneously. The results allow complexity and characteristics. Also, unlike the us to improve our understanding of the impact of mutually ornamented king penguin, Adélie pen- membrane mechanics on frequency analysis in the guins do not have ornaments (unless cryptic).We locust ear. predict that the information signalled by display calls of Adélie and king penguins are complementary to the information conveyed by the nest 87 and the ornaments, respectively. We also aim to identify what kind of information is encoded in Kriesell, Hannah J. acoustic (individual display calls) and visual cues Le Maho, Y. (nest-associated features and ornaments) in the Le Bohec, C. mate selection process.

108 21 Ludwig Maximilians University, Munich, Germany

Kruger, David J.D Echo-acoustic flow guides flight in bats Heyns du Preez, L. North-West University, Potchefstroom, Abstract Common to all airborne animals is the South Africa need to react fast and correct to rapid changes in their environment during flight. Visually guided The extraordinary vocal characteristics of the animals tackle this challenge by evaluating South African Common River Frog (Amietia visual flow, generated by their movement through quecketti) structured environments. Echolocating bats flying in complete darkness cannot make use of visual Abstract Call repertoire variation for the majority flow; they can navigate solely by echolocation, i.e, of anuran species usually comprise only of slight the auditory analysis of self-generated sounds. In modifications of the signature advertisement call, contrast to vision, echolocation provides explicit e.g. consisting of a single tonal or pulsatile note, distance information through the analysis of echo a series of identical repeated notes, or a long delay. Here we show that bats exploit echo-acoustic trill. The extensive variation in the call repertoire flow to navigate rapidly through narrow passages. of the South African Common River Frog (Amietia Specifically, we find that bats’ navigation between quecketti) has previously been overlooked. This lateral structures is significantly affected by the study demonstrates the uniqueness of the call pro- echo-acoustic salience of those structures, inde- perties of A. quecketti by evaluating the extensive pendent of their physical distance. This is true variation in call note types and the occurrence of despite the stroboscopic nature of echolocation four nonlinear phenomena. Three call-types for A. which interferes with a motion percept and alt- quecketti are described, namely biphasic adver- hough echolocation, unlike vision, provides explicit tisement call, pulsatile aggression/territorial call, distance cues.The results demonstrate that sensory and release call. The aggression note emitted by flow elicited by self motion is a ubiquitous principle this species is the first reported instance for the for guidance of flight in the animal kingdom, inde- Amietia genus. Advertisement calls consisted of an pendent of the fundamentally different peripheral extraordinary variation with four types that are de- representation of flow information across the senses scribed, namely tonal-notes, creak-notes, iambic- of vision and echolocation. notes, and whine-notes). The calls range widely in spectral characteristics from a narrow-band tonal to broadband chaotic phases, to even broader band, 92 pulsatile advertisement click-notes and pulsatile iambic aggression notes consisting of up to 13 Lampe, Helene M. pulses. Temporal variation within pulsatile note- Labra, A. types could be produced at different rates, from University of Oslo, Norway 16 pulses/second in advertisement click-notes, to 90 pulses/second in rip-notes, to extremes of 170 Exploring the relationship between song pulses/second in creak-notes. Nonlinear events characteristics of fathers and sons in pied include biphonation, subharmonics, deterministic flycatchers chaos and frequency jumps. Abstract Song may play a key role in sexual selection in birds. Males use song to attract females 17 and defend territories. In pied flycatchers (Ficedula hypoleuca), the complexity (e.g. repertoire size, Kugler, Kathrin versatility) of male song is a good proxy of their Greiner, W. quality (e.g. condition, experience), and females Luksch, H. use this honest information to select their mate. Firzlaff, U. Males sing until they have attracted a female, but Wiegrebe, L. do not use song to defend the territory and more

109 or less stop singing after pairing, unless singing the surveys. While getting more data is always good, in their secondary territory to attract a second it brings challenges in extracting the information female. If songs are relevant for male fitness, it from vast numbers of acoustic files. Within the may be expected that sons inherit song characte- Bioacoustics Unit, formally established in 2014 as ristics. In addition, fathers may be important song a collaboration between the University of Alberta tutors for their sons, as has been found in other and the Alberta Biodiversity Monitoring Institute, species. Therefore, fathers and sons could share our research team is working on solutions to the song syllables since males sing occasionally when challenges presented by ARU data. Our efforts fall feeding nestlings. We tested the two hypotheses into three main categories: 1) Developing standard using song recordings from a long-term study of protocols for ARU deployment in the field that can pied flycatchers collected during 31 years. We had easily be used by other researchers, industry, and recordings of both fathers and sons in 28 cases. For communities for their own acoustic monitoring each male, we analyzed 25 songs, measuring seven programs so that they can collect the data nee- variables: 1) repertoire size (different syllable types ded for their specific projects and collaborate to in 25 songs), 2) total syllables sung, 3) N° different generate large data sets over large spatial scales. syllables per song, 4) N° syllables per song, 5) song 2) Research into efficient and effective ways of ext- sample versatility (var1/var2), 6) song versatility racting species data from the recordings using both (var3/var4), and 7) duration. We found that there human observers and auto-recognition software. 3) is no relationship between any of these variables Developing a training program for acoustic techni- between fathers and sons even when considering cians where new technicians receive mentorship the age of the individuals. The percentage of sylla- and training from experienced staff in both detailed bles shared by fathers and sons ranged from 0 to species identification of complex recordings and in 60%, and did not correlate with any of the sons’ the effective use of auto-recognition software. We song characteristics. In summary, results indicate are successfully collaborating with other research that in pied flycatchers the capability to sing more groups and with industry. We have gather more data attractive songs for females (e.g. complex) is not on rare species that any other single project to date inherited, and fathers play a variable role as song partly because we have developed recognizers than tutor. can process thousands of files in a short time period

129 7

Lankau, Hedwig E. Larsen, Ole N. Bayne, E. Wahlberg, M. University of Alberta, Canada Christensen-Dalsgaard, J. University of Southern Denmark, Denmark Developing standards and methods for bioacoustic data collection and processing In-air and underwater hearing of the cormorant in Alberta, Canada studied with ABR

Abstract Acoustic monitoring using autonomous Abstract Numerous studies have mapped recording units (ARUs) is fast becoming the most the hearing abilities of birds in air but currently common way of surveying avian communities. By there is little or no data on how diving birds hear placing an ARU at a station, it is possible to record or react to sound under water. Therefore, it is multiple times per day across multiple days. This unknown whether the ears and auditory system of is particularly beneficial for surveying species that diving birds are adapted to underwater hearing. In are rare, nocturnal and/or vocalize infrequently the present study, we measured audiograms from because the longer an ARU records at a station, the wild-caught cormorant fledglings (Phalacrocorax more likely the species is to be detected. Because carbo sinensis) using auditory brainstem response there is no human observer present, there is little (ABR). Each bird was anesthetized and the ABR risk that individuals will behave differently during was measured, first, in a sound attenuated and

110 anechoic box in air and, secondly, with its head and frequencies. We predicted that bats will react to neck submerged approximately 10 cm under water prey-generated rustling sounds by steering their in a large water filled-tank while being artificially sonar beam towards the position of the sound for ventilated. ABR-responses to calibrated tone bursts further biosonar-based evaluation. To present prey produced by a woofer and an underwater speaker, rustling sounds and monitor bat echolocation, we respectively, were measured at different intensities developed a spherical three loudspeaker - eight and frequencies to obtain hearing threshold values microphone array. Each loudspeaker was placed in air and under water. The shape of the audiogram next to one microphone and was symmetrically was similar to that reported for birds of the same surrounded by three other microphones. Bats were size in air. The bandwidth and slopes of the au- trained to perch in the centre of the spherical diogram were similar in air and water. In air the array. Per trial, we played back a rustling sound highest sensitivity was found at 1-2 kHz, while of a moth fluttering in vegetation from one of the the most sensitive response was displaced towards loudspeakers, or alternatively phase-randomized lower frequencies (lower than 1 kHz) under water. or amplitude-inverted versions of the recordings to Generally, the cormorant ear was not very sensitive control for temporal and spectral cues. We recorded to sound, neither in air nor under water. The hearing the bat’s echolocation behaviour simultaneously abilities in water, however, were better than what with all eight microphones for offline analysis of would have been expected for a purely in-air ad- relative call intensity and thus call direction. apted ear. (Supported by the Carlsberg Foundation 2009_01_0292 and the Danish Council for Independent Research 4002-00536) 112

Laverde-R., Oscar 104 Caycedo-Rosales, P. Pulgarín-R., P.C. Lattenkamp, Ella Z. Cadena, C.D. Kaučič, R. Universidad de los Andes, Colombia Koselje, K. Goerlitz, H.R. Bird songs in the shelf: temporal patterns Max Planck Institute for Ornithology, Seewiesen, of songs and vocal output are lost in sound Germany archives

Do horseshoe bats complement their echolocation Abstract Understanding the frequency with with listening for prey-generated sounds? which animals sing (vocal output) is of critical importance to address a variety of research questions Abstract Echolocation is a highly sophisticated in behavioral ecology and sexual selection. However, sensory system for actively probing light-deficient information on vocal output, a central component environments. However, due to the stroboscopic of the investment in signaling, is lacking for most and directional emission of the calls and the species and is difficult to collect. We asked whether strong attenuation of ultrasonic frequencies, the the number of recordings of avian vocalizations in space that can be probed by biosonar is limited three different sound archives and the times when both temporally and spatially. We hypothesised such recordings were obtained reflect estimates of that this limitation will favour the opportunistic vocal output and temporal patterns of vocal activity use of additional information for prey detection, obtained through systematic monitoring of wild bird such as prey-generated rustling sounds, and that populations in three tropical forest sites. Based bats thus exploit a much wider range of environ- on a sample of 43 montane species, we found mental information than previously believed. We significant relationships between the number of tested this hypothesis in greater horseshoe bats recordings of species detected through continuous (Rhinolophus ferrumequinum), whose echolocation monitoring over several months and the number of is specialised for the detection of fluttering insects, recordings archived in sound collections, especially but spatially strongly limited due to high call when accounting for the area of distribution of each

111 abstracts poster

species. In addition, daily activity patterns based similarity between the song of the offspring and the on data collected through continuous monitoring founders; 2) song complexity. We observed an inter- over several days did not differ from those based individual variability in the song: some individuals on recordings archived in sound collections in 12 have learned to produce a very close copy of the of 15 species of lowland birds. Annual patterns founders’ song while others produced a poor copy. in vocal activity of two species estimated based We observed a positive correlation between the on recordings in collections closely resembled similarity score and the song complexity index, sug- previously published patterns. We conclude that gesting that the most similar songs to the model are recordings in sound collections contain valuable also the most difficult to produce. We investigated information about the vocal output and temporal whether these proxies of song learning (similarity patterns in vocal activity of birds. This opens the to the song model, complexity index) were linked possibility of using sound collections to assess to the cognitive abilities of the males by evaluating vocal output and to consider it as a variable of their performance during a reversal learning (colour interest in studies on the ecology and evolution of association) task. We did not observe a significant birds and other animals that use acoustic signals correlation between proxies of song learning and for communication. We discuss potential caveats of male performance to solve the cognitive task. These this approach and make recommendations for field results do not support the idea that the acoustic workers contributing to sound archives to maximize structure of song could signal cognitive skills in the the potential usefulness of their material. Zebra Finch. However, additional behavioural tests and song analysis are required in order to confirm that song learning is not linked to other cognitive 95 abilities in this species.

Le Maguer, Lucille Geberzahn, N. 90 Université Paris Ouest Nanterre La Defense, France, Université Paris Lumières, France Leitner, Stefan Derégnaucourt, S. Teichel, J. Université Paris Ouest Nanterre La Defense, ter Maat, A. France, Université Paris Lumières, France and Voigt, C. Institut Universitaire de France Max Planck Institute for Ornithology, Seewiesen, Gemany Song learning and cognitive abilities in zebra finches (Taeniopygia guttata) Hatching late in the season requires flexibility in the timing of song learning Abstract Despite a recent interest in studying cognitive abilities in oscine songbirds, few re- Abstract Most songbirds learn their songs from search focuses on connecting these capacities to adult tutors, who can be their father or other male singing behaviour in general and to song learning conspecifics. However, the variables that control in particular. Song learning itself is a challenging song learning in a natural social context are largely cognitive task and there is evidence in many oscine unknown. We investigated whether the time of hat- songbirds that sexual selection favours precisely- ching of male domesticated canaries has an impact learned song. Recently, it has been proposed that on their song development. Average age difference the accuracy with which songs are learned may between early and late hatched males was 50 days serve as an indicator of other cognitive abilities for with a maximum of 90 days. Song activity of adult females in mate choice decisions. Here we mea- tutor males decreased significantly during the sured the performance of adult male zebra finches breeding season. While early hatched males were (Taeniopygia guttata) in a cognitive task. We used exposed to tutor songs for on average the first 99 male offspring of a colony where founder males were days, late hatched peers heard adult song only du- originally selected to produce the same song. Using ring the first 48 days of life. Remarkably, although a semi-automated procedure, we measured: 1) song hatching late in the season negatively affected body

112 ELECTRONIC DEVICES

USB MICROPHONES USB HYDROPHONES

PARABOLIC MICROPHONES

ULTRASONIC ANALOGIC MICROPHONES

SMART REMOTE RECORDER

MULTICHANNEL HI SPEED RECORDER

www.dodotronic.com [email protected] 113 abstracts poster

condition, no differences between both groups of have combined a mechanism for intensity control males were found in song characteristics either with low-intensity echolocation to counter moth in autumn or in the following spring. These data hearing until the final capture of the moth. This no- suggest that late hatched males must have under- vel strategy might have given them exclusive access gone accelerated song development. Furthermore, to a formerly unavailable food source, thereby pro- the limited adult tutor song exposure did not affect bably altering competition between sympatric bat song quality. species tremendously. In turn, selection pressure for eared moths might have increased, potentially fostering new anti-bat traits to evolve. 158

Lewanzik, Daniel 52 Goerlitz, H.R. Max Planck Institute for Ornithology, Seewiesen, Linke, Simon Gemany Gifford, L. Barclay, L. Low-intensity echolocation plus intensity Karaconstatis, C. compensation – a means for barbastelle bats Griffith University, South Brisbane, Australia to remain undetected by eared moth Tonolla D. Zuerich HAW, Switzerland Abstract The arms race between insectivorous bats and their insect prey lead to the convergent Real-time acoustic monitoring of freshwater evolution of ultrasonic sensitive ears in several ecosystems insect taxa. Eared moths perform evasive flight manoeuvres when they hear bat echolocation Abstract Traditional methods of freshwater calls, thereby significantly decreasing their risk ecosystem monitoring have three key disadvantages of predation. Yet, the barbastelle bat, Barbastella due to their invasive nature. The survey techniques barbastellus, is able to almost exclusively prey bear a) risks to fish health and habitat integrity, on eared moths. Recently, it has been shown that b) introduce bias, because it might cause fright barbastelle bats emit low-intensity search calls that responses in key aquatic species and c) standard are not audible to moths at distances at which the surveying only produces a snapshot from the time bats are already able to detect a moth. However, of surveying – which in many cases does not happen while bats move towards their prey, the intensity of more than once a year. Non-invasive passive bioa- their calls at the moth’s ears increases, eventually coustic monitoring can improve all three problems, becoming audible and potentially eliciting evasive however it can also offer a holistic real-time glimp- flight. We hypothesized that barbastelle bats lower se into aquatic ecosystems. We argue that a holistic call intensity while closing in on their prey, such ecosystem monitoring program is feasible in fresh- that the intensity at the moth’s ear remain below the water ecosystems as a comprehensive literature moth’s hearing threshold and the moth fails to elicit review has identified considerable knowledge about the life-saving evasive flight manoeuvre. We tested sonically identifying three key elements in a river this hypothesis in the wild by offering tethered ecosystem. 1. Freshwater fish: at least 20% of taxa moths to free-flying barbastelle bats. We computed are estimated to be soniferous. 2. Invertebrates: the bats’ three-dimensional flight paths based on key indicator families among benthic invertebrates time-of-arrival differences of the echolocation calls are – at least in part – soniferous. This includes at a four-microphone array. We then calculated call indicators of good water quality (Trichopetera, source levels and received levels at the moth’s po- Odonata) as well as indicators of degraded systems sition based on an additional miniature microphone (Coleoptera, Hemiptera) 3. Geophysical habitat: in at the tethered moth. We show that barbastelle bats the last decade, half a dozen studies have measu- continuously reduced call intensity upon detecting red habitat quality and diversity by sonic proper- a moth and that this intensity reduction is suited to ties. With these three main classes of a freshwater remain undetected by their prey. Barbastelle bats ecosystem covered, a real-time holistic ecosystem

114 assessment seems feasible in the not too far future. to the rocks and sand habitats “listen” to the rocks We will demonstrate the feasibility of this approach and the coralligenous. with identified sounds of fish, invertebrate and riverine habitats from Queensland, Australia. 121

117 Luís, Ana R. Couchinho, M. Lossent, Julie dos Santos, M.E. France Energies Marines, Brest, France MARE - Marine and Environmental Sciences Gervaise, C. Centre, ISPA - Instituto Universitário, Lisboa, Di Iorio Portugal L.,CHORUS, Fundation of Grenoble Institute of Technology, France Analyzing a graded acoustic repertoire: the use of discriminant analysis to classify burst- Mapping marine biophony pulsed signals of bottlenose dolphins

Abstract 1922 snapshots (10s-long) of sounds Abstract Common bottlenose dolphins (Tursiops have been measured in a small 4km2 coastal area truncatus) present a complex acoustic repertoire that (depth from 0 to 40m) near the pointe of Revelatta comprises three major types of signals: frequency- (Calvi, Corse). Our aim is to build the soundscape modulated tonal whistles, echolocation click trains of this complex area hosting four interleaved ha- and other pulsed-sounds with high repetition rate. bitats (rock, seagrass, sand, coralligenous). Wide While most acoustic studies of dolphins focus on Band [2kHz, 40kHz] SPL is maximum at the rocks whistles and/or echolocation clicks, research on and decreases offshore with the regression law: SPL burst-pulsed sounds (BP sounds) is still scarce. A (dBre.1µPa) = 147 – 14log10(r) with r the distance set of distinct packets of pulses, usually classified to the coastline. The range of detection of the bio- according to their aural appearance fall under the phony of the rocks extends from 1800 to 8300m. designation of BP sounds. However, BP sounds are This regression law based on “energetic criteria mainly ultrasonic signals, with inter-click-intervals only” allows to think that the rocks’ biophony is (ICI) that cannot be perceived by the human loud and propagates to the other habitats. We assess hearing; thus, the use of subjective classifications the diversity of the spectral shape of the biophony is doubtful. Multivariate classification techniques by computing the PCA of 2 million of measured based on time-frequency parameters may help to spectra. The two first eigenvalues explain 50% of overcome the methodological difficulties faced the spectral variability. The spectra and their scores in BP sounds’ categorization. Field recordings of 1 and 2 create a continuous cloud in {score 1, score bottlenose dolphins’ emissions were collected in 2} plan. Then, we propose to segment the data in the Sado estuary, Portugal, from 2011 to 2015. 4 quadrants depending of the sign of scores 1 and We analyzed the acoustic characteristics (minimum 2 of the PCA. Each quadrant produces a family of frequency, maximum frequency, peak frequency, spectral signatures (spectra with high peak at 3 duration, repetition rate and inter-click-interval) and 4KHz, spectra with wide bump between 5 and of 526 BP sounds classified in three categories 12kHz, spectra with maximum at high frequency (“Buzzes”,”Creaks” and “Squawks”), using aural more than 20kHz). The 4 proportions of impulses features and graphical representation aspects. from a given habitat in each of the 4 quadrants Preliminary results have shown significant differen- are proposed to describe the diversity of spectrum ces in peak frequency (FWelch (2, 165.99) = 5.99, shape of this habitat. For our data base, each habi- P=0.003), duration (FWelch (2, 357.07) = 239.34, tat hosts the four families of spectral signature but P<0.001), repetition rate (FWelch (2, 154.22) = in different proportions. The coralligenous hosts its 721.27, P<0.001) and inter-click-interval (FWelch own sound production with high frequency impul- (2, 214.50) = 447.47, P<0.001). Additionally, we ses. The rocks hosts its own soundbproduction with applied a discriminant function analysis (DFA) to low frequency impulses. Seagrass habitats “listen” previously classified BP sounds. Using a stepwise

115 abstracts poster

DFA, three variables were identified as statistically 146 significant for the discrimination among different burst-pulsed signals: ICI (F(2,412)=735.97, Maciel, Israel p<0.001), duration (F(4,822)=369.35, p<0.001) Marino Simão, S. and repetition rate (F(6,820)=285.75, p<0.001). Federal Rural University of Rio de Janeiro Discriminant and classification functions were (UFRRJ), Brazil also obtained, and our model correctly classified Figueiredo, L.D. 87.9% of the analyzed BP sounds in their a priori Instituto Federal do Rio de Janeiro (IFRJ), Brazil categories. Considering human auditory limitations, the use of more objective procedures may help to High incidence of non-linear phenomena in improve the categorization of animal signals, espe- Antarctic Phocidae cially in graded repertoires. Abstract Nonlinear phenomena (NLP) in vocalizations are still poorly studied, possibly because of 10 the difficulty in identifying that phenomenon, their causes and biological meaning. The non-linearity Ma, Shouwen in vocalizations can be generated by anatomical ter Maat, A. differences in the vocal tract, whether pathological or Gahr, M. because it is young individuals with vocal tract still Max Planck Institute for Ornithology, Seewiesen, under formation or else be intentionally produced by Germany animal. This study aims to identify these phenomena in Antarctic Phocidae, as well as a discussion of Sensorimotor feedback maintains auditory their biological meaning. Data were obtained by objects formation hydrophone of Brazil’s Navy, fixed to three meters deep, located on Half Moon Island (South Shetland Abstract In an international cocktail party, peo- Archipelago, Antarctica). The data was collected by a ple are able to identify those who speak the same hydrophone ITC, model 8073, pre-amplified (20dB) as languages from background noises easier than sampling rate of 22 kHz. Since the sample was passive those who speak other languages. It suggests that and Phocidae vocalizations have many similarities, we there must be neural mechanisms for listener to identify until the lowest taxonomic level as possible. maintain the language sensitivity of his own langu- The sampling was recorded from 12 to 30 of November ages. Reports in human studies that patients with /2012, totalling 372 hours and 24 minutes. The re- damages to the Broca’s area suffered from aphasia, cording was divided into 7,448 files with 3 minutes and also had difficulty in speech comprehension, long. They were detected the occurrence of Phocidae imply that sensorimotor feedback may influence from 12 to 16 November /2012, totalling 834 files speech perception. In zebra finch, the sensorimotor (41 hours and 42 minutes). Data were analysed nucleus HVC (higher vocal center) corresponds func- using Raven Pro 1.4 software. Were found 1,786 tionally to Broca’s area. Both HVC and Broca’s area vocalizations, composed of 930 (52.22%) containing were mostly studied in motor domain, whereas the NLP‘s (e.g. sideband, harmonic, sub-harmonic and feedback from motor to auditory domain in speech bifonation) and 851 (47.78%) without NLP’s. Most perception is less known. The aim of this study is to of the vocals without NLP’s consist of pulsed sounds. investigate the influence of HVC on auditory object Bifonic calls regularly showed the same pattern and formation with regard to the international cocktail are unlikely to the incidence of duets. Witch one of party problem. We employed electrophysiological the different vocalizations types showed consistently methods to examine responses to auditory objects the same type of nonlinearity. In addition, NLP‘s in both male and female auditory cortices. We occurred every day in which there was presence of found that the habituation rate to auditory objects Phocidae, indicating with production intend a biologi- was higher in male than in female. Inactivation of cal meaning in communication. Despite the contribu- HVC in male resulted in decrease of habituation tion of this study to understanding NLP‘s in Phocidae, rate. Our results provide evidence that sensorimotor are necessary more studies seeking understood the integration affect auditory object formation. meaning of these NLP‘s in communication.

116 44 ETH Zürich, Switzerland

Maeder, Marcus Vocal expression of emotions in Zurich University of the Arts ZHdK, Switzerland Przewalski horses Zweifel, R. Swiss Federal Institute for Forest, Snow and Abstract Expression of emotions plays an im- Landscape Research WSL, Birmensdorf, portant role in social species, including humans, Switzerland because they regulate social interactions. However, studies testing a direct link between emotions and „Phyto-acoustics“: acoustic emissions vocal structure in non-human animals, and especially of plants wild animals, are rare. In particular, little is known about how animals encode in their vocalizations, Abstract Plant physiologists have known that information about the valence (positive/negative) of plants emit sounds for several decades now. Many the emotion they are experiencing. In this study, we of these sounds are of transpiratory/hydraulic origin combined new methods developed to study animal and are therefore related to the circulation of water emotions and vocalizations, in order to investigate and air within the plant as part of the transpiration how Przewalski’s horses encode information about process. Some of the acoustic emissions are indi- emotional arousal and valence in their vocalizations. cations of embolism in the water transport system, To this aim, we recorded vocalizations produced by which occurs when a plant is subjected to drought the animals in naturally occurring situations charac- stress and desiccation. The excessive water tension terized by different emotional arousal and valence. in the water-conducting system leads to the rupture Negative situations consisted in agonistic interac- of the water columns in the plant vessels. Each tions and positive situations in affiliative interactions plant species—in fact each plant individual—has and food anticipation. Behavioural measures taken its own acoustic signature, related to its structure when the animals are vocalizing confirmed their and to the local climatic conditions. Investigating emotional states. The results of these experiments the acoustic emissions of a tree in response to provide better knowledge of the way in which emo- dynamically changing climatic conditions might tions are encoded in wild mammal vocalizations. The reveal biological or physical properties that place vocal indicators of emotions that we found would be them in a broader ecophysiological context and particularly useful for welfare assessment of wild enable us to explain processes that are not yet species, which cannot be easily approached and fully understood. In our research, we found an manipulated, such as the Przewalski’s horse that is impressive number of acoustic emissions that show endangered. Finally, comparisons of our results with different characteristics and occurance than the those of similar studies in domestic species may „classical“ Cavitation pulses caused by embolisms reveal interesting findings about the effect of dome- in the plant water transport system. Some of these stication on the structure of animal vocalizations. sounds lie in the audible domain and occur also at night. Such acoustic emissions – which are, as far as we know, for the first time examined in more 80 detail – seem to arise from different processes than light-driven transpiration and relate to other or new Manfrini, V. models of water transport, distribution of nutrients Zoomarine Italia, Rome, Italy and plant growth. Pelicella, I. Dodotronic, Rome, Italy Basile, E. 68 Zoomarine Italia, Rome, Italy Pavan, G. Maigrot, Anne-Laure Università di Pavia (CIPRA), Italy Anne, C. Hillmann E. Acoustic monitoring of a pregnant bottlenose Briefer, E.F. dolphin (Tursiops truncatus) under human care

117 abstracts poster

Abstract The observations under human care Abstract Most of knowledge about bird voca- allow to study the behavioral patterns of the lization ontogenesis comes from studies of song pregnancy and mother-calf relationship thanks development in passerine birds (Passeriformes), to non-invasive methods. The same observations but development of vocalization in other taxa is in the wild are not easy. Veterinary examinations poorly known. Some Anseriformes and Galliformes (e.g. ultrasound) together with the collection of demonstrate gradual changes of call-frequency physiological data (e.g. feces) and behavioral (e.g. variables values during development, while some flexions) allow to monitor the pregnancy over time Gruiformes, Coraciiformes and Columbiformes show and to determine the date of birth. This protocol jump-like changes of them. We analyzed early vocal was used to determine the date of calf birth of a development in two diurnal bird of prey species Tursiops truncatus female hosted at Zoomarine (Falconifornes): Saker Falcon (Falco cherrug) and Italia (Rome). In addition to the observation, we Northern Goshawk (Accipiter gentilis). All diurnal also collected acoustic data of the pregnant female birds of prey have semi-altricial chicks that hatch (20 hours) and the couple mother-calf (20 h.) over sighted and stay in nests until fledging. We recorded several days and at different times. In the pool there calls from 7 Saker Falcon and 4 Northern Goshawk was the female and then the mother-calf pair. The chicks between hatching and fledging age stages other bottlenose dolphins were in other connected in Vitasfera Rare Bird Species Breeding center. We pools, but the gates were closed by watertight doors. collected 15 signals per chick every 3 days (1160 The recordings were made by an Aquarian Audio signals in total). We found two main call types - hydrophone model AS-1 with piezoelectric sensor begging for food and discomfort call - in chick vocal (not pre-amplified) connected to a preamplifier repertoire for both species. Chicks used calls of and A / D converter with USB interface Dodotronic both types since hatching. Call-frequency and call Hydromic200K and this to a laptop with Windows duration variables were observed only for begging XP. The analyses were performed using the software for food call. Fundamental frequency gradually de- SeaPro (developed by CIBRA) and Audacity. The creased between calls of hatchlings and fledglings signals in the audio and ultrasonic band were in both species, but it was higher than that in calls sampled at 200kHz / 16-bit with a bandwidth of 95 of adults. Call duration decreased in saker falcon, kHz. The acoustic recordings were coupled with the but merely fluctuated in northern goshawk with observations of the animals in the pool. This study age. According to fundamental frequency changes is a contribution to collecting acoustic recordings in we divided period between hatching and fledging in captivity and especially during the pregnancy (very two: early and late ones. Thus, we confirm gradual sensitive moment of the life). Although the pool vocal development in diurnal birds of prey. was a reflective environment that introduces many reverberations, it has been interesting to study the female vocalizations (signature whistles, stress 91 signals, echolocation), the mother-calf communication, and behaviors associated to some sounds. Marfull, Reinaldo Komdeur, J. Rijksuniversiteit Groningen, The Netherlands 108 Korsten, P. University of Bielefeld, Germany Marchenko, Alexandra Beme, I. The effect of sibling competition on the Lomonosov Moscow State University, Moscow, acoustic component of begging behaviour Russian Federation in blue tits (Cyanistes caeruleus) Sarychev, E. Vitasfera Rare Bird Species Breeding Centre, Abstract Begging behaviour of nestling birds Moscow region, Russian Federation mediates parent-offspring interactions through continuous adjustments of postures and vocalizations Ontogenesis of vocalization in diurnal bird of for food. In this context, competitive state and past prey (Falconifornes) experience may be key factors to assess costs and

118 benefits of begging by the offspring. It has been dolphins (Tursiops truncatus) in Europe. These shown that begging postures of hand-raised nest- dolphins’ complex social behaviour entails a par- lings can be modified as a result of training regimes, ticularly diverse vocal repertoire. They emit mainly and that these different levels of postural begging three different types of sounds: 1) tonal, frequency- may persist later on. Here, we looked at whether modulated whistles, 2) echolocation clicks and 3) competition during early-life influences the acou- rapid repetition rate “burst-pulse” packets, which stic component of begging behaviour between- and include so-called “buzzes”, “creaks”, “squawks”, within-nests. Using brood size manipulations, we and “bray series”, composed of “gulps”, “squeaks” created different sibling competition levels for 36h and “grunts”. Underwater vocal behaviour during in natural nests (i.e. reduced and enlarged nests), feeding and foraging events is of particular interest, from which the second lightest and the second hea- since these are critical activities carried out by dol- viest nestlings were chosen to be exposed to 30, 90 phin groups. Behaviourally, predation episodes on and 150 min of food deprivation. We show that iso- fish schools near the surface are considerably diffe- lated nestlings that were satiated to equal levels at rent from the dolphins’ search for prey hidden near the start responded with increased call output (e.g. the substrate. Nevertheless, little is known about bout duration, frequency and sound level) with an how echolocation clicks and burst-pulse emissions increase in food deprivation. Remarkably, nestlings differ during these events. Recordings of dolphin reared in broods with low levels of sibling compe- vocal behaviour during surface feeding events and tition (i.e. reduced) showed short and narrow calls bottom foraging events were collected in the Sado with lower dominant frequency and sound level. No Estuary between 2011 and 2015 (213 events). within-nest differences were found. These results A Mann-Whitney U test revealed that the overall indicate that the acoustic component of begging emission rates differed significantly between sur- is dynamically adjusted according to the present face feeding and bottom foraging events (U(1,212) hunger level and the intensity of past sibling com- = 3195.5, p < 0.001). Further analysis showed petition, and hence expand the evidence on the role that emission rates of “creaks”, “squawks” and of the past experience in shaping relevant early-life echolocation clicks were significantly higher during behaviours. Potentially, individual offspring may be surface feeding events (U(1,212) = 2363.5, p < able to match begging behaviour to optimal levels 0.001, U(1,212) = 3689, p = 0.001 and U(1,212) in relation to family interactions. = and 3281.5, p < 0.001, respectively). These results suggest that, while foraging, these dolphins might depend mainly on passive hearing for prey 78 detection, whereas surface feeding may require more active acoustic emissions. While “creaks” Martin, Juliette G. may possess echolocation value, “squawks” may University of Edinburgh, U.K. and MARE - Marine represent social communication signals used in and Environmental Sciences Centre, ISPA - close proximity. However, the specific functions Instituto Universitário, Lisboa, Portugal of individual burst-pulses requires further investi- Luís, A.R. gation. Also, the question of whether the observed MARE - Marine and Environmental Sciences acoustic behaviour is associated with cooperative Centre, ISPA - Instituto Universitário, Lisboa, (e.g. sharing of information) or competitive interac- Portugal and Projecto Delfim - CPEMM dos Santos, tions remains to be answered. M.E., MARE - Marine and Environmental Sciences Centre, ISPA - Instituto Universitário, Lisboa, Portugal and Projecto Delfim - CPEMM 56

Differences in the vocal emissions of bottleno- Martinez-Alvarado, Dariel se dolphins during surface feeding and bottom Jaramillo-Calle, V. foraging events Rivera-Gutierrez, H.R. Universidad de Antioquia UdeA, Medellin, COL Abstract The Sado Estuary in Portugal harbours one of the few resident communities of bottlenose Does noise affect song activity of oscine and

119 abstracts poster

suboscine species between urban and rural Gonzáles-Terrazas, T.P. areas? Ruiz., T. Jung, K. Abstract Many bird species use song for commu- University of Ulm, Germany nicating, and it is well known that birdsong plays Tschapka, M. an important role for finding mates, defending University of Ulm, Germany and Smithsonian territories and for social integration. Recent studies Tropical Research Institute, Balboa, Panama on birdsong in urban environments have found that increased noise may limit acoustic communication Mexican bat acoustic monitoring system in birds by masking the song. Bird passerine (SIMMA) species are classified in two main groups, birds that learn their song (Passeriformes: Oscines), Abstract Despite the current threats faced by and birds that do not learn to sing (Passeriformes: many bat species (e.g., habitat destruction, wind Suboscines). While the effect of urban noise on farming, diseases), large-scale, long-term monito- acoustic communication of oscine species has been ring programs for determining bat population trends thoroughly studied, little is known on the responses are non-existent in tropical countries. Existing mo- or strategies of suboscine species. Given that spec- nitoring programs are limited in time and space due tral characteristics of song in suboscine species to high costs and the amount of human resources may be consistent through different environments, needed. We designed and implemented in a local non-learning species may use different strategies scale a standardized acoustic bat monitoring proto- for avoiding the masking effect of anthropogenic col for Mexico (SIMMA given by Sistema Mexicano noise. We hypothesized that suboscine species may de Monitoreo Acústico). The first implementation of shift the timing and places in which they sing (song the monitoring program was conducted in a tropical activity) for noise avoidance. Here we present a rainforest at Montes Azules Biosphere Reserve in study on the song activity of eight passerine species Southeast Mexico. We selected four permanent (four oscine, four suboscine) in eight different loca- acoustic recording stations, each station consisting lities along an urban-rural gradient, and that differ of a forest and a gap or edge site. We used one in noise levels. Five different points were randomly bat detector at each site (SM3, Wildlife Acoustics) selected at each locality, and song activity of the and programmed to record consistently 4 hours chosen species was measured by registering the per night (18:00 to 22:00) during 30 consecutive number of strophes sung by species. Song activity nights. To assess the performance of our approach and noise levels at each point were measured during we analyzed the recordings, focusing on the total five minutes, every hour, during four hours, starting of collected call sequences, effects of weather and at sunrise. Order of points was randomized. Each moonlight, and the influence of open and closed locality was visited during four consecutive or semi spaces on species diversity. In addition, we compa- consecutive days, in case of rain, or bad weather red two different recording schedules: one recording conditions. Our results suggest that song activity the first 4 hours, as established in our protocol, and differs between oscine and suboscine species, and one recording 12 hours per night. This comparison between urban and rural areas. Our study shows for will help us to evaluate the effectiveness of our pro- the first time variation in song activity of oscine tocol, improve the next implementations and allow and suboscine species in urban and rural localities. us to establish a standardized long-term system at national scale.

127 150 Martínez-Serena, Abigail Ibarra-Macías, A. Martins, Joao Torres, L. Fiat Sonitus, Cologne, Germany Víquez-R., L. Medellín, R. Rhythms from the animal world in the practice Universidad Nacional Autónoma de México,MEX of composers

120 Abstract The pervasiveness of animal sounds in affecting evolution of alarm communication in musical activity gives plenty of space for investiga- speckled ground squirrels Spermophilus suslicus. tion of the relevance of animal sounds in musical In individually chip-marked animals from natural practice. If music has any parallel to other repre- colonies, we investigated effects on the alarm sentational arts, namely painting and sculpture, call structure of 1) caller identity, 2) sex, 3) age, then animal sounds must have been a constant 4) time-span between recordings, 5) population presence in early stages of human development. It membership and 6) genetics. In total, we examined is not possible to listen to these songs today, but by 171 individuals, 10 calls per animal, 8 acoustic looking into the archaeology of musical instruments variables per call. Caller identity had strong effects and to the recorded practice of musicians, it is on all acoustic variables. Age effects were signifi- possible to get some insights into the music of the cant on call duration and fundamental frequency time. As soon as a fully fledged music notation sys- bandwidth. Age-class differences in the maximum tem was developed in 13th century, with melodic and dominant fundamental frequencies of alarm and rhythmic components, music was represented calls lacked despite prominent age differences in with a relatively precise method and from the body weight (222±50g vs. 88±34g). No sex effect beginning we can observe human interpretations on call variables was found. Twenty individuals of animal sound. A direct reference can be heard were recorded 4 times with time-spans of 1 day, in the original work of the 13th century „Sumer Is 2 weeks, and 1 year. Calls from individual callers Icumen In“, where the cuckoo song is developed were highly similar within recordings, but vocal as a canon. Examples of this practice can be found individuality decreased drastically with increase in the works of numerous composers to this day, of time span between recordings. However, calls with a special mention to Olivier Messiaen who remained individualistic, allowing distinguishing exhaustively notated the sounds of birdsong. This individual callers higher than was expected by chan- work explores and analyses the representation and ce. We compared alarm calls and polymorphism of fixed media examples of music that have drawn mtDNA control region for 90 individuals from 6 influence from animal vocalization. Examples of scattered populations, 15 animals per population. musical works that have explicit connections to Population membership and individual identities animal vocalisations, whether imitative or symbolic, affected strongly the acoustic variables, especially will be shown in connection to their original animal the mean fundamental frequency and duration correlates. The rhythmic features in both vocalisa- of the calls. Comparison among populations by tions and musical representations are analysed. acoustic, genetic and geographic distances with Mantel test revealed a highly significant positive correlation between the genetic and geographical 147 distances (r=0.97, p<0.001) as well as significant correlations between acoustic and genetic (r=0.66, Matrosova, Vera p<0.01) and between acoustic and geographical Engelhardt Institute of Molecular Biology RAS, distances (r=0.74, p<0.01). Genetic differences Russian Federation among populations gradually increased with the Volodin, I. A. increase of the geographical distance. Acoustic Lomonosov Moscow State University differences followed the same pattern, in support Moscow, Russian Federation of gene drift hypothesis. Financial Support: RFBR Volodina, E.V. grant 15-34-20589. Scientific Research Department, Moscow, Russian Federation 59 What affects alarm call structure in speckled ground squirrels: multifactor analysis Matsinos, Yannis Economou, C. Abstract Processes, responsible for vocal University of the Aegean, Mytilene, Greece variation in mammals are poorly understood so far. This 10-years-long study focuses on factors Defining quietness in urban green areas

121 abstracts poster

Abstract The existence of green areas in a city urban and rural landscape, in order to listen to has proven to be beneficial due to their ability to the environment. The purpose of this procedure is increase biodiversity and provide city dwellers with soundscape evaluation, primarily using the sense several direct and indirect ecosystem services. The of hearing and secondary the other senses-feelings provision of quietness, as a direct ecosystem ser- (mainly optical stimuli). Soundwalks are considered vice, is a major indicator of environmental quality. tools in soundscape research that could be used Nevertheless, the way that city inhabitants perceive in order to collect the necessary qualitative and their acoustic surroundings could determine the quantitative data, concerning a specific location character of the landscape along with the quality of or a larger area. The main driving force of sound- the soundscape and define the meaning of quietness walking, is the acoustic perception of individuals that still remains vague. Furthermore, the assess- while its flexible methodology, allows application ment of acoustic perception in an urban green area in various interdisciplinary areas such as ecology could enhance knowledge regarding quiet areas and urban design. This research focuses on the in agglomerations. The way that soundscapes are way that urban park users perceive their acoustic being perceived by urban park users, is an issue that surroundings, incorporating the ecology of bird could be investigated using both, quantitative and habitat and vegetation. The participant’s acoustic qualitative data. A number of noise measurements perception was assessed during a soundwalk speci- and sound recordings were conducted, in order to fically designed for two urban parks, located in the portray the acoustic identity of two parks located in center of the coastal city of Mytilene, which is a the center of the coastal city of Mytilene, which is a small urban setting in Lesvos Island (North Aegean, small urban setting in Lesvos Island (North Aegean, Greece). Using a novel soundwalk methodology that Greece). The noise maps that were created, accom- differs from the common practices, both quantita- panied with the analysis of several acoustic indices tive and qualitative data were collected. The scope (Acoustic Complexity Index, Normalized Difference of this research was to utilize the findings of this Soundscape Index), provided the necessary quan- soundwalk and visually present them, in order set titative data. Additionally, a soundwalk including the ground for a future landscape remodeling. As the two urban green areas under consideration was a result, amongst others, the areas vulnerable to conducted in order to assess the perceived noise intrusive noise derived. Furthermore, the soundwal- level (PNdB) of individuals with the results with kers highlighted the “soundmarks” of their prefe- the accompanied actual measurements. From the rence and proposed new ones in order to compose a statistical analysis that followed resulted that the more desirable soundscape and therefore, reshape quality of each soundscape is directly associated the landscape. by both audible and visual attributes. Furthermore, the soundwalkers highlighted the “soundmarks” of their preference in order to compose a more 54 desirable soundscape. Quietness in an urban environment, as difficult as it is to be achieved, Matsinos, Yiannis requires soundscape planning and environmental Economou, C. awareness by its users. University of the Aegean, Mytilene, Greece

Vocalization attributes of specific song notes 60 and distress calls (churring) between urban and rural great tits (Parus Major) Matsinos, Yiannis Economou, C. Abstract 40 years after its proposal, the acoustic University of the Aegean, Mytilene, Greece adaptation hypothesis remains a provisional idea whose merit still requires evaluation. The idea of Soundwalk as a landscape remodeling animals adaptively responding to the alteration of driving tool the acoustic background, by respectively altering their vocalizations, as appealing as it sounds to Abstract A soundwalk is any excursion in the classic biologists, undergoes investigation from

122 the scientific community. Skepticism is solely the software Song Scope, to automatically detect based upon the mechanism utilized towards such and classify vocalizations of the Yellow Rail. This an alteration –whether a shift in pitch (AAH) or species is a monitoring priority in the province gi- a shift in amplitude (Lombard Hypothesis) takes ven our lack of knowledge regarding its population place. Up to now, studies investigating possible in the region, as well as, its proximity to energy differences between bird vocalizations have been development. It is rare and poorly detected by mainly focused on song features. Latter studies att- traditional surveys because it is nocturnal and its empted to distinguish between these hypotheses by wetland habitats are remote and difficult to access. investigating how calls other than songs (e.g. alarm Occupancy modelling provides us with a useful calls) are affected. Our study tried to combine the framework for understanding differential habitat two approaches and compare specific vocalization use and changes in sites occupied by the rail. Here, attributes of both songs and calls. We compared we make use of detection histories derived from urban and rural great tits, Parus Major, specific our Yellow Rail classifier to produce occurrence song notes and distress calls (churring), recorded estimates. By comparing these estimates with those in situ across Lesvos Island in Greece, and selected generated from human processing in a manner akin recordings containing both types of vocalizations to traditional surveys we gain important insight into from the same individual, thus to eliminate the the robustness of occupancy models to a when the amplitude difference parameter. We found -as input is ‘scaled-up’. We aim to determine how the expected- that rural song notes had lower mean fre- results of the approaches compare. quencies than urban, but no significant difference was found for the distress calls. This latter runs contrary to expectations as suggested by both the 1 hypotheses. Furthermore, the frequency differences observed in low pitched notes of the songs were Melotte, G. smaller than the high pitched ones, both findings Michel, C. being totally opposite from the expectation of the Parmentier, E. response to the low frequency positioned masking University of Liège, Belgium Having in mind the words of Hainman (1989) “It may therefore seem ironic that I remain uncertain Sound production in piranhas and relatives: about vocal organization in this species-but for a comparisons between species Mélotte, G. simple reason: it mimics other species, including other tits”, we conclude that there exists a growing Abstract Acoustic communication plays an necessity of incorporating ecological aspects in our important role in the life of many teleost species studies. where it is mainly involved in agonistic and/or courtship behavior. Despite the large number of Serrasalmidae species (92), sound production has 114 been described only in the genera Serrasalmus and Pygocentrus. The aim of this study was first McLeod, Logan to investigate the sound producing abilities of dif- Bayne, E. ferent Serrasalmidae species and then to describe University of Alberta, Canada and understand the corresponding mechanisms. One herbivorous species, Piaractus brachypomus, Using emerging bioacoustic technologies to produces sounds composed of a single pulse. The model for Yellow Rail occupancy mechanism would involve the vibration of the bladder due to the hypaxial musculature contrac- Abstract New technology for acoustic recordings tion. In contrast, the calls emitted by Serrasalmus and signal processing have opened up new ave- rhombeus, Serrasalmus compressus, Serrasalmus nues for monitoring animal populations. Passive, marginatus, Serrasalmus elongatus, Pygocentrus automated recording units (ARUs), are being used nattereri and Pristobrycon eigenmanni are all widely in Alberta, Canada, for biodiversity moni- harmonic sounds composed of several pulses wit- toring efforts. We have used a classifier, built in hout inter-pulse interval. They all show the same

123 abstracts poster

kind of mechanism: the sound results from the into whistled melodies. After 12 years of inquiry forced vibration of the swimbladder following the we found that the speaking mode of whistling is contraction of sonic muscles that are attached to distributed worldwide. It is not limited to a given a perpendicular tendon surrounding ventrally the continent, language family or language structure. bladder. A last species, Pygopristis denticulata, is One of the most striking aspects of this whistled able to produce another type of sounds. It consists transformation of words is that the whistled senten- of several pulses with irregular pulse period and is ces remain highly intelligible to trained speakers likely produced by a sonic muscle inserting on the in several languages, despite a reduced acoustic skull and on the rostral part of the swimbladder. channel to convey meaning. By comparing whistled The relatively high diversity of sound types and speech and whistled singing in several languages of mechanisms in Serrasalmidae will be used in the the world we also managed to explain their typology future to understand the evolutionary development and to explore the intricate relationship between of this particular behavior. Do the mechanisms music, speech and language in whistling. evolve separately or is there a continuity between them? 75

149 Mishima, Yuka Morisaka,T. Meyer, Julien Mori, Y. Laboratoire sur le Langage le Cerveau et la Miyamoto, Y. Cognition (L2C2), CNRS, Lyon, France Tokyo University of Marine Science and Technology, Japan Speaking and singing with whistles around the world: speech and music in a simple narrow PS1 call is determined as an affiliative contact frequency band call in captive belugas (Delphinapterus leucas)

Abstract The main focus of this paper deals Abstract Previous studies on one captive beluga with the results of a worldwide inquiry on whistled population in Japan indicated that they frequently speech and whistled singing. Around the world, it exchanged one of the pulsed sounds (“PS1 call”). is rather common to hear people whistling the tune PS1 was the most common call type in isolation of a song by following the prosody of its sung lyrics. context. Some acoustic parameters of PS1 were This kind of performance can be defined as the individually different, especially inter-pulse interval ‘whistling mode of singing’ or the ‘singing mode of patterns (IPI contours), which were stereotyped whistling’. It is characterized by a musical valence intra-individually. PS1 was thus thought to func- that puts rather little pressure on the whistler for tion as an individualized contact call, but larger the transmission of the meaning of the lyrics. By samples are required. To verify the PS1 characters contrast, several remote rural populations around and functions which previous studies suggested, we the world also use a special speech register based collected PS1 calls from other captive belugas kept on whistling that fulfils purely communicative at Shimane Aquarium in Japan. There were two purposes: commonly called ‘whistled speech’ or beluga pools, the pool A containing three belugas the ‘speaking mode of whistling’. In these local and the pool B containing four belugas. Recordings communities, the speakers learn since childhood to were made in both pools during free swimming. The copy any sentence of their language into a simpler pool A belugas produced PS1 calls most frequently whistled signal with the clear objective to transmit with 40% (971 calls) of total calls during twelve spoken messages to the listener(s). Whistled 30-min sessions. Analysis of Inter-PS1 intervals speech is always based on a spoken language. represented that PS1 calls were produced in bouts Whistled Greek is still Greek but adapted to another with a criterion interval of 5.1 s. Only 2 out of 69 speech register. Basically, people articulate words aggressive behavior occurred in PS1 bouts and while whistling and thereby transform spoken ut- the number of aggressive behavior per unit time terances by simplifying them, syllable by syllable, was significantly smaller in PS1 bouts than other

124 times (Wilcoxon Signed-ranks Test, P=0.0224). strong tendency to have all these areas reduced to a PS1 sequences by different belugas occurred most single major cell (the mirror) in the PBW for sound commonly within an interval of 0.5 s (28.0%), and radiation (some species however, show PBWs with a 75.1% of the sequences occurred within 2.0 s. complex arrangements of cells.) In this research we Individual comparison on temporal parameters of test the hypothesis that the observed morphological PS1 calls revealed that duration and the number wing asymmetry in katydids has an effect on the of pulses were significantly different (one-way resonances of both wings. Using micro-scanning ANOVA, F(2,151)=52.72, P<0.0001 and Kruskal- laser Doppler vibrometer and broadband sound Wallis test, χ2=64.98, P<0.0001, respectively). stimulation we measured the resonance and wing Furthermore, the belugas had individually specific, deflection of both the PBW and FBW across diffe- stereotyped IPI contours in PS1 calls with some rent species of katydids. Results show that across variables at extremely lower rate. These results sug- species the stridulatory area of the PBW dominates gested that PS1 was commonly produced, affiliative in the amplitude of deflection, while the FBW tends call type and used for vocal exchange not only in to be damped. These measurements also allow us one captive population but as a species character to quantify the levels of mechanical asymmetry of belugas. The results of pool B belugas will also between both wings. Interestingly, the degree of be presented. asymmetry is highly associated with the purity of the call at ultrasonic frequencies.

33 63 Montealegre-Z., Fernando Sarria, F. Montes-Medina, Adolfo C. Chivers, B. D. Renton, K. Jonsson, T. Universidad Nacional Autónoma de México University of Lincoln, U.K. Salinas-Melgoza, A. Robert, D. Centro Tlaxcala de Biologia de La Conducta, University of Bristol, U.K. México

Song quality and sound-generator asymmetry Functional diversity in the vocal repertoire of in bushcrickets the lilac-crowned parrot

Abstract Male bushcrickets (or katydids) Abstract Understanding the role of avian vocal generate mating songs to call distant females by communication in reproduction and social orga- wing stridulation. The calls are usually ultrasonic nization requires greater knowledge of the vocal although many species communicate in the human repertoire used to convey information. Parrots have audio range, and the call varies from sharp pure complex social systems, and are vocal learners able tones (tonal calls of high purity) to broadband to transmit acoustic signals through social interac- (non-resonant calls with noisy spectra). The right tion. We describe the vocal repertoire of the Lilac- wing possesses a plectrum (the plectrum bearing crowned Parrot (Amazona finschi) in nine behavioral wing PBW), while the left wing bears a file (referred contexts (alarm, threat, flight, take-off, landing, to here as the file-bearing wing FBW) and sits on foraging, perched, food soliciting, and nesting). top of the PBW. Different to crickets, in which We spectrographically classified 101 note-types, both PBW and FBW are more or less mirror images though for analysis we considered 38 note-types of each other to the human eye, katydids exhibit that were emitted at least 5 times. We evaluated remarkable bilaterally asymmetrical wings, with 15 spectral and temporal variables for each note, the FBW usually showing atrophied sound radiating and identified four Principal Components that areas. The PBW on the other hand has well deve- explained 75.15% of the variation among notes. loped membranous areas for sound radiation. But Statistical validation of these components using different from crickets, which have at least three Discriminant Function Analysis determined a final major cells for sound radiation, katydids show a 60.6% correct classification. Vocalizations emitted

125 abstracts poster

in nine behavioral contexts could be separated into et al. 2003). In our work we test the hypothesis two main groups based on frequency and temporal that certain song characteristics may be genetically characteristics of short-range, low-intensity calls determined. The Wood Warbler (Phylloscopus and long-range, high-intensity calls. The Lilac- sibilatrix) singing a very simple song has been crowned Parrot had a diverse vocal repertoire with chosen as the model species. Wood Warbler songs the highest variety of note-types of any Neotropical consist of two phrases. The frequency modulation parrot, where only Australian psittacines have a of the first phrase is similar in all males, the second richer vocal repertoire. Threat interactions had the phrase varies among males and has three types greatest variety of notes and proportion of exclusive which differ significantly in frequency band and trill notes, and alarm calls had the highest frequency of rate (p<0.001, 60 males, 10-15 songs per male). notes/sec. The most common notes used by parrots Every male performs only one type of the second were notes C (26.4% of notes) and B (23.7%) phrase. We expect that if the second phrase type presented in all behavioral contexts but emitted is genetically determined then the close related mainly in long-range, high intensity calls to attract males should have the same phrase type. For the attention of conspecifics. These features indicate preliminary analysis we have studied 95 songs of that the Lilac-crowned Parrot has a complex system 19 males, the genetic distances between them were of communication, where differing functions were estimated in GenAlEx 6.5 on the basis of 5 micro- achieved by modifying the composition of notes, satellite loci analysis. Pairwise individual genotype emission rate, and acoustic parameters of notes in and individual second phrase structure were tested each context. Finally, we propose a standardized for correlation using Mantel tests. Mantel test nomenclature and statistical analysis to permit showed significant (p<0.006) correlation between comparative studies of cross-functional use of matrices of male genetic distances and the second vocalizations by parrots to elucidate signal design phrase structures. This result supports the hypothe- rules for differing social functions. sis that the Wood Warbler second phrase structure has some genetic determination.

96 84 Moskalenko, Viktoria Goretskaia M. Motes Rodrigo, Alba Belokon, M. Labra-Lillo, A. Belokon Y. S. Lampe, H.M. Lomonosov Moscow State University, Russian University of Oslo, Norway Federation Politov, D. Testing the nutritional stress hypothesis Vavilov Institute of General Genetics RAS, Russian and learning capacity in pied flycatchers Federation Abstract In songbirds song is used in sexual Song types are genetically determined in birds selection; males sing to compete with other males with the single song type over territories and/or to attract females. More complex songs are related to high-quality males Abstract Bird song is a sexually selected male that have more possibilities of being chosen by a fe- trait where females select males on the basis of male. It could be possible for a male to exaggerate song quality. It is well known that in Passerine birds his quality when singing in order to cheat females. song learning plays an essential role in song deve- However, if song features are costly to produce, lopment (Hultsch, Todt, 2004; Catchpole, Slater, they can be reliable signals of male quality. The 2008). Genetics is also important in song develop- nutritional stress hypothesis is an explanation of ment, but little is known about relations between how songs can be costly. It suggests that if a male individual genotype and song structure. However nestling is well fed, he has more energy to develop females selecting males by preferring certain song the brain structures needed for singing more com- characteristics may choose a good genes (Foerster plex songs later, while an underfed male needs all

126 the energy he gets for developing other parts of the in an anechoic chamber were analyzed. Sitophilus body and brain. Thus, males that receive less food zeamais larvae displayed continuous low amplitude should have less complex songs. The pied flycat- impulses whereas P. truncatus larvae displayed cher is a good species for testing this hypothesis continuous higher amplitude impulses. In addition, since the males use songs almost exclusively to spectrograms associated with S. zeamais larvae had attract females, and females are found to prefer fewer high energy regions as compared to those of large syllable repertoires. We analysed if there P. truncatus indicating that movement and feeding was a relationship between the versatility of pied by the S. zeamais larvae is generally associated flycatcher songs (number of different syllables per with low energy sound. Contrastingly, oscillograms song or song sample) and their tarsus length and of P. truncatus adults were characterized by lower body mass measures when nestlings, but found no amplitude impulses compared to those of the S. such relationship. They are open-ended learners, so zeamais adults. Analysis of impulse and burst rates we also analysed males returning to breed a second revealed that rates of noise impulses in P. trun- time to see if they could improve their songs from catus and S. zeamais sounds varied significantly. their first breeding season to their second. We For adults, the rates of burst and the impulses found that they had on average doubled their sylla- per bursts were significantly (P < 0.05) different. ble repertoires. Thus, female pied flycatchers could However, for the larvae, there was no significant (P use song to separate between males with breeding > 0.05) difference in the rate of bursts or rate of experienced and inexperienced male. impulses although the impulses per burst varied significantly (P < 0.05). These findings would be useful in developing maize acoustic sensors for 31 real-time detection of P. truncatus and S. zeamais in bulk storage warehouses. Njoroge, Anastasia W. Affognon, H. Mutungi, C. 24 International Center for Insect Physiology and Ecology, Nairobi, Kenya Nolan, Huw Richter, U. Brown, W. Hensel, O. Ballard, G. Mankin, R. Zewe, F. University of Kassel, Germany University of New England, Armidale, New South, Australia Acoustic characteristics of Prostephanus truncatus (Horn) (Coleoptera: Bostrichidae) and Dingo communication Sitophilus zeamais (Motschulsky) (Coleoptera: Curculionidae) Abstract Dingoes (Canis familiaris) have a controversial place in the Australian landscape. Abstract Prostephanus truncatus (Horn) They are apex predators that many people believe (Coleoptera: Bostrichidae), and Sitophilus zeamais are vital for conservation while to others they are (Motschulsky) (Coleoptera: Curculionidae) are invasive pests to be controlled. Despite this cont- highly damaging storage insect pests for maize. roversy, little is known about their behavior in the Timely detection of infestations using acoustic wild. As with many canid species, dingoes have means can enable effective and timely control complex vocal repertoires that, if better understood, measures that help to avoid destructive losses. might be used for conservation and management. There is little research on acoustic detection of However, we remain largely ignorant of the role these two pests. Laboratory experiments were of vocal communication in dingo social systems. conducted to characterize the acoustic profiles of Initially, to understand the vocal repertoire of wild the last instar larvae and adults of P. truncatus dingoes we quantitatively described the recorded and S. zeamais on stored maize. Spectral and vocalizations of 45 trapped animals using acoustic temporal features of recorded sound signals taken parameters including spectral (such as frequency

127 abstracts poster

and amplitude shifts) and temporal (such as call explore the sequences of pilot whale calls recorded length) components. Based on this analysis, voca- from six different groups of animals. We observe lizations were categorized into seven distinct call that the distribution of time intervals between con- types. Dingoes used these call types singly or in secutive calls is similar for all six groups of whales. combination, with gradation present between some Additionally, we analyse pairs of consecutive calls calls. Additional analysis of vocalizations of a ran- (call bigrams). We are interested in the question dom subsample of dingoes from four ecologically whether such combinations occur more, or less distinct habitats (five dingoes per habitat) demons- often than expected when the calls are assumed to trated there were habitat-specific differences (p be independent. < 05) in their acoustic structure, consistent with predictions of the acoustic adaptation hypothesis. Currently, we are investigating the variation in vo- 65 calizations of dingoes recorded in a single habitat. Data collection is on-going but recordings analyzed Osiejuk, Tomasz to date indicate there are acoustic differences in Jakubowska, A. vocalizations associated with factors such as size, Adam Mickiewicz University, Poznan, Poland status, sex and degree of hybridization. In addition, we are finding differences in the acoustic struc- Soft song use in the presence of predator and ture of calls in different contexts, e.g. presence/ during rival‘s intrusions in the Ortolan Bunting absence of humans. Understanding the differences (Emberiza hortulana) in vocalizations of dingoes in varying environments and situational contexts is vital to the success of Abstract The soft (low amplitude) song is a remote monitoring and can help improve wildlife poorly understood acoustic communication pheno- management and increase our knowledge of the mena despite the fact that it seems to be common ecology of the dingo. and widespread in birds. Several hypotheses were raised in order to explain soft song functionality. We present results from a project design to test all 151 working hypotheses on soft song functionality in the ortolan bunting (Emberiza hortulana). The first Noriega Romero Vargas, Maria F. experiment (27 males tested twice) was designed to Hallerberg, S. test hypothesis that males use soft song during ter- Timme, M. ritorial intrusions to avoid being predated. We used Max Planck Institute for Dynamics and Self- microphone array allowing for amplitude measure- Organization, Göttingen, Germany ments in the field and simulated territorial intrusion Vester, H. into ortolan bunting territories by a stranger male OceanSounds e.V., Germany (playback + model presented) in a paired design Hammerschmidt, K. experiment. In random order, we presented (treat- German Primate Center, Göttingen, Germany ment) or did not (control) a predator model and its calls before simulated intrusion of a non-neighbour Bigrams in sequences of pilot whale calls male. We found no significant relationship between presence of a predator and response of males to Abstract Vocal communication is highly evolved playback with soft song. However, the presence of in social marine mammals such as killer whales, pi- other conspecific males in vicinity affected respon- lot whales or other dolphins. Their vocal repertoires se. In the second experiment with similar design consist of multiple sounds like clicks, buzzes, we initially lured male to loudspeaker with regular grunts, and a variety of pulsed calls including (loud) songs and after approaching to a standar- whistles. There is evidence that killer whales have dized distance we played back regular or soft song specific call combination and that they commu- (22 males tested twice). We found that loud songs nicate by exchanging vocal signals. However, the evoked stronger response than soft songs. However, way these vocal signals are combined and used by tested males used soft songs and it seems that this the animals is largely unknown. In this study we is an acoustic tool for verifying if the potential rival

128 is close. These results suggest different function Within species, 72% of S. coeruleoalba whistles of soft song than in any hypotheses presented ear- and 73% of D. delphis whistles were classified to lier. (Financial Support: National Science Center, the correct ocean basin. Within the tropical Pacific, Poland, grant no. 2013/09/B/NZ8/03275). many significant differences were found between populations of D. delphis recorded in the North and South Pacific and between subspecies of S. 124 attenuata (coastal – S.a. graffmani, and offshore – S.a. attenuata). Using DFA, 95% of D. delphis Oswald, Julie whistles and 80% of S. attenuata whistles were Bio-Waves, Inc., Encinitas, U.S.A. classified to the correct population or subspecies. Rankin, S. These results suggest that classifiers should be NOAA Fisheries, Southwest Fisheries Science trained using data collected in the location where Center, La Jolla, U.S.A. they will be used. Gannier, A. Fuchs, S. Group de Recherche sur le Cetaces, Antibes, 67 France Padilla de la Torre, Monica Geographic variation in delphinid whistles – Briefer, E. implications for classifier development Hillmann, E. ETH Zurich, Switzerland Abstract The Real-time Odontocete Call Classification Algorithm (ROCCA) is a whistle classi- Vocal expression of emotion in cattle fication module in the acoustic processing software platform, PAMGuard. ROCCA contains classifiers for Abstract Emotions have a crucial function for the tropical Pacific and northwest Atlantic oceans. an animal’s life, because they facilitate responses Can these classifiers be used to analyze data coll- to external or internal events of significance for ected in other locations? Characteristics of whistles the organism; positive emotions trigger approach produced by three delphinid species (Delphinus behaviour towards stimuli that enhance fitness delphis, Stenella coeruleoalba, S. attenuata) in the (“rewards”), whereas negative emotions trigger tropical Pacific Ocean and Mediterranean Sea were avoidance behaviour when encountering stimuli examined to evaluate the differences that exist bet- that threaten fitness (“punishers”). Expression ween and within species. Fourteen variables descri- and perception of emotional states play an bing frequency, shape and duration were measured important role in social species, because they and compared (t-tests and Mann-Whitney U tests, inform individuals about the probable intention of α=0.05). Between-species comparisons showed behaviours of others and therefore, regulate social that the degree of difference varied with location. interactions. Although some studies have suggested Few significant differences were found between that emotions might be encoded in vocalisations, whistles produced by D. delphis and S. coeruleoalba it has never been tested in cattle. To achieve this in the tropical Pacific. In contrast, many significant goal, we triggered emotions in beef cattle using differences were found between these same spe- mother-offspring separation as a negative context, cies within the Mediterranean. Using Discriminant including low arousal (partial isolation – physical Function Analysis (DFA), 69% of Mediterranean but not visual) and high arousal (full isolation), whistles were classified to the correct species. This and mother-offspring partial and full re-union as a correct classification score was significantly greater positive context Behavioural and physiological data than chance (χ2 test, α = 0.05). In contrast, the were collected for each situation, in order to con- correct classification score for the tropical Pacific firm underlying emotional arousal and valence of (55%) was not significantly greater than chance the situations. Preliminary results showed that cow (χ2 test, p = 0.37). This implies that, for these and calf heart rates were very high during visual re- species, differences within the tropical Pacific are union and rapidly decreased after full-reunion. In smaller than differences within the Mediterranean. accordance with recent findings about ear positions

129 abstracts poster

in cattle, mothers and calves spent longer times lowed by fifth harmonic. It was found in some notes displaying forward ear positions during negative descending frequency modulation until 516,8 Hz contexts (partial isolation and full isolation) than from the beginning to the end of the fundamental during positive contexts (partial re-union and full frequency. There was a significant variation in va- re-union). Preliminary acoustic analyses of cow lues of interseries interval (867,41 ms ±1222,49; calls suggested that source-related parameters, n=78; CV=104,9%) and number of notes per series such as fundamental frequencies (F0), were lower (2,81 notes ±1,76; n=68; CV=62,5%). Our results during partial isolation than during full isolation. As suggests that duration and frequency of the notes expected, F0 decreased after full re-union with their as well as intensity and frequency of the harmonic calves. Nevertheless, calves produced vocalisations components, can be used as useful measures for exclusively during full isolation. Two different types taxonomic comparisons between subspecies of M. of vocalisations (high and low frequency calls) torquata. were produced in this context. These non-invasive indicators of emotions could allow us to develop better and more accurate welfare measurements for 88 domestic cattle worldwide. Parks, Susan Root-Gutteridge, H. 154 Cusano, D.A. Syracuse University, U.S.A. Palhares, Cíntia O.M. Conger, L. Lopes, L.E. van Parijs, S.M. Federal University of Viçosa, Brazil NOAA Northeast Fisheries Science Center, U.S.A.

Description of the vocalization of Description of the ontogeny of acoustic commu- Melanopareia torquata bitorquata (D‘orbigny & nication behavior of North Atlantic right whales Lafresnaye, 1837) in Mato Grosso, Brazil Abstract The vocal ontogeny of species can Abstract Biocoustics has proven to be a very provide insight into their physical and social de- useful technique for accessing the taxonomy of velopment and capacity for social learning. North Suboscines passerines. Preliminary taxonomic Atlantic right whales (Eubalaena glacialis) are a studies conducted by LEL using external mor- highly endangered species of baleen whale with a phological data suggested that M. t. bitorquata complex range of vocalizations used in a wide range should be given full species status. This taxon is of social contexts, and show evidence of behavioral endemic to the open grasslands of eastern Bolivia plasticity of signal production in the presence of and south-west Brazil and its conservation status noise. In this study, we systematically characterized is unknown. This study aims to describe the far- developmental changes in the sound production reaching vocalization of M. t. bitorquata, providing behavior from shortly after birth through adulthood. the bases for future taxonomic studies. We analyzed Calls were recorded from 47 different individuals 11 tape recordings obtained on December 2009 between 2001 through 2015, and include data and November 2014 in the municipality of Vila from calves as young as 1 month of age to adults Bela da Santíssima Trindade, Mato Grosso. We > 30 years of age. Data included recordings from used the Raven Pro 1.5. software for evaluating the hydrophone arrays and non-invasive suction cup following parameters: duration (ms), energy (dB), digital acoustic recording tags. Only calls that maximum frequency (Hz) of the notes; internote could be confidently attributed to a specific whale and interseries intervals (ms); dominant frequency of known age were used in the study. Analyses (Hz) and number of notes per series. The notes have indicate a gradual maturation of sound production a harmonic spectrum structure with until seven with increasing age of individuals, with typical harmonic components visible, on the best quality stabilized ‘adult’ vocalizations identified around 4 records. The fundamental frequency presents the years of age. Early sound production by calves in highest intensity (2,587 kHz ±0,128; n=304), fol- their first months of life are generally unstructu-

130 red, amplitude modulated, and short in duration, antartandica and B. leptopus did not alter their often produced during social interactions with their calling rates. When presented with heterospecific mother. Specific landmarks for vocal development calls, B. taeniata and B. antartandica decreased included first occurrence of particular calls types, their vocal output as compared to their responses along with documented evidence of increasing to conspecific signals, while the responses of B. vocal control with maturation including increasing leptopus to conspecific and heterospecific calls stability of tonal frequencies, a reduction in the did not differ over most playback conditions (GLM presence of chaotic segments within calls, and ANOVAs and Tukey post-hoc tests, P < 0.05). Vocal increasing frequency bandwidth and modulations activation overcomes limitations to active space, rates. The developmental changes observed in right while decrease in vocal output restricts energy whale sound production track those described for expenditure amidst interference. The different other typical mammalian systems and suggest that responses in the presence of prolonged noises and vocal repertoire development in right whales may temporally-structured signals by these anurans rely upon a combination of physical maturation and contribute further considerations on the adaptive social learning. significance and evolutionary determinants of such vocal strategies by sound communicating animals. Supported by FONDECYT grant 1140014. 47

Penna, Mario 2 Quispe, M. Muñoz, M. Pereira Chaves, Patricia A. Ramos, N. University of Saint Joseph, Macau, SAR China and University of Chile, Chile MARE, ISPA, Lisboa, Portugal Vasconcelos, R. O. Communication amidst biotic and abiotic Valdoria, C.M.C. interference in frogs of the austral temperate University of Saint Joseph, Macau, SAR China forest Development of the inner ear morphology Abstract Interfering sounds from biotic and and auditory sensitivity for social acoustic abiotic origins are likely to shape the responsive- communication in the Lusitanian toadfish ness of sound communicating animals. Anurans adopt different strategies to communicate in noisy Abstract Fish represent the largest group of environments: some species increase, while others sound producing vertebrates and exhibit the high- decrease their vocal activity in the presence of in- est diversity in peripheral auditory mechanisms. terfering sounds. The acoustic environments where The inner ear endorgans, including the otoliths and frogs from austral temperate latitudes communica- sensory epithelia, grow in most fish species throug- te are relatively simple; few anuran species vocalize hout their lifespan. However, studies addressing in synchrony, in contrast with jammed tropical the functional consequences of such development sound landscapes. To gain an understanding of the of the peripheral auditory system in this taxon are abilities of anurans from temperate environments rather sparse compared to other vertebrate groups. to overcome interferences of various origins, we The highly vocal Lusitanian toadfish (Halobatrachus explored the vocal activity of male frogs Batrachyla didactylus) has become a particularly interesting in the southern temperate forest in Chile during model to investigate questions concerning the exposures to prolonged natural abiotic noises of development of auditory-vocal systems as social wind, creek, rain, and to a band-pass noise cente- acoustic communication is initiated early in deve- red at about the dominant frequency of their calls lopment. A previous study has shown an improve- (2000 Hz). Males were also exposed to synthetic ment in auditory sensitivity during development, conspecific and heterospecific calls. In response namely among juveniles (> 5 cm SL), which was to prolonged noises, males of Batrachyla taeniata coincident with a significant increase in the vocal increased their vocal output significantly, while B. repertoire. The goal of this study was to investigate

131 abstracts poster

the morphological development of the inner ear came recently into secondary contact in Northern saccule (the main auditory endorgan in teleosts) and Eastern Europe, where they occasionally hybri- in H. didactylus throughout ontogeny, comparing it dize. They are morphologically and ecologically very with the development of auditory saccular sensiti- similar; however, Thrush Nightingales are slightly vity in this species. Sensory epithelia were removed larger and several lines of evidence indicate com- from the inner ear saccules of previously fixed petitive dominance of this species. Nightingales in specimens, ranging from fry (<1,7 cm SL) to the the contact zone show asymmetric convergence in adult stage (>25 cm). Hair cells were stained with song: most Thrush Nightingale males include in phalloidin and their number, density and orienta- their repertoires song types copied from Common tion patterns determined across the different size Nightingales but not vice versa. This convergence groups. Even though the number of hair cells and apparently does not increase hybridization, as epithelium size increased with fish growth, the hair hybrids are relatively rare and most of them result cell density remained similar within the 1,7-25 cm from mating of Thrush Nightingale females with range. However, fry <1,7 cm revealed higher hair Common Nightingale males. As vocalization is an cell density than older developmental stages. We important component of territorial defence, we propose that other morphological features or steroid evaluated in playback experiments the strength of levels might be involved in the saccular sensitivity aggressive responses of both species to conspecific improvement in this species. Furthermore, the and heterospecific song stimuli. The species did increased hair cell density early in ontogeny may not differ in reactions to heterospecific playbacks, result from a rapid increase in the hair cell number suggesting that competitive dominance is not relative to the growth of the sensory epithelium, associated with higher interspecific aggressiveness which slows and reverses later during development. in nightingales. Interestingly, while the Common Nightingale reacted more aggressively to the conspecific stimulus, the Thrush Nightingale show- 28 ed similarly strong responses to both stimuli. We suggest that this pattern results from mixed singing Petrusková, Tereza of the latter species in sympatry, as males may Reif, J. not distinguish conspecifics from heterospecifics Vokurková, J. by song alone. These results are consistent with Jiran, M. the concept of convergent agonistic character dis- Petrusek, A. placement, which suggests that vocal convergence Reifová, R. might be adaptive in species that overlap broadly in Charles University in Prague, Czech Republic resource use. However, the potential role of female Kipper, S. preferences in interactions between the two spe- Free University of Berlin, Germany cies, and specifically in song convergence, should also be evaluated. Song convergence and aggressive interactions of two closely related songbird species in their contact zone: the case of Nightingales 99 (Luscinia spp.) Popova, Daria Abstract Understanding mechanisms causing Beme, I. reproductive isolation and interspecific compe- Lomonosov Moscow State University, RUS tition between incipient bird species can provide Sarychev, E. important insights into the speciation process. We Vitasfera Rare Bird Species Breeding Centre study these mechanisms in closely related species pair of songbirds, Common Nightingale (Luscinia Ontogenesis of vocalization in several owl megarhynchos) and Thrush Nightingale (L. lusci- species (Strigiformes) nia), combining multiple approaches ranging from ecology and genetics to bioacoustics and behaviou- Abstract In comparison with Passeriformes, ral experiments. Populations of these two species ontogenesis of birds of prey vocalization is poorly

132 studied at the present time. The main goal of our spp.). From a three year study on Tree Pipits (A. research was to study owl vocalization ontogenesis. trivialis), which often sing in flight or from perches, We were first to trace acoustic signals structure we conclude that acoustic monitoring based on and their time-frequency parameter values change the syllable repertoire can be substantially more with age (from hatching to fledging and further) in efficient for individual recognition than colour rin- chicks of two owl species: tawny owl (Strix aluco) ging. Repertoire of each banded male was distinct and great gray owl (Strix nebulosa). In particular, and stable within as well as between seasons; and we have studied mechanisms of transformation of males with similar syllable repertoires differed in juvenile calls into calls of adult birds. Also, we syntax. Thus, we were able to unambiguously iden- made an attempt to uncover possible reasons of tify singing males in the studied population from owl calls transformation with age. We choose chick analysing a recording containing 20–30 songs, calls that we defined as ones of begging for food and use the data to track territory changes and as the object of our study; in adult birds we used detect returning individuals. We will contrast these all recorded acoustic signals as study objects. In results with those from terrestrial Tawny Pipit (A. addition, comparison between signals of chicks campestris) with much simpler song. For this spe- and adults of the same species was made. In our cies, we will present first data about song stability study we checked hypothesis about abrupt changes of individually banded males recorded repeatedly in signal structure and time-frequency parameters during one breeding season. values in chick calls after their eyes opening. We carried out a comparison of acoustic signal parameters values and ways of their transformation with 140 age between owl species: we checked hypothesis about connection between signal structure and Prior, Nora time-frequency parameters with habitat features Vignal, C. and life habits of the birds. The study was conduc- Université Jean Monnet, St. Etienne, France ted at Vitasfera Rare Bird Species Breeding Centre Fernandez, M. and supported by RFBR grants № 11-04-00062а Université Jean Monnet, St. Etienne and INSA and № 14-04-00108/14. Lyon, France Soula, H. INSA Lyon, France 13 Acoustic behavior in paired, breeding and Porteš, Michal non-breeding condition zebra finches Petrusková, T. Pišvejcová , I. Abstract Zebra finches are native to the deserts Petrusek, A. of Australia, where environmental conditions are Charles University in Prague, Czech Republic unpredictable. To facilitate survival in this harsh climate, zebra finches have evolved to be oppor- Who is who? Acoustic monitoring of migratory tunistic breeders and form life-long monogamous passerines with complex and simple song as a pair bonds. Zebra finches integrate several environ- tool for individual recognition mental cues in order to make breeding decisions, and in the laboratory, manipulating water availabi- Abstract In field studies, it is often necessary to lity is sufficient to affect reproductive condition. identify specific individuals. Colour rings frequently Importantly, both the initiation and cessation of used to mark birds are often difficult to observe, breeding must also be coordinated at the level of especially in small passerines and dense habitats. the pair. However, it is unclear how pairs coordi- Acoustic-based monitoring detecting individuals by nate these decisions. It is possible that acoustic their characteristic vocalization is a potentially sui- communication between partners facilitates this table alternative, but this approach is challenging decision-making. There are at least two ways that a in species with complex songs. Here we present zebra finch could use vocalisations to inform their its application on two species of pipits (Anthus partner about their state of breeding readiness: (1)

133 abstracts poster

the use of vocalization subtypes (i.e., tet call, dis- Each set consists of approximately 2,000 songs; tance call, male song) may change throughout the about 1.3 million trials were processed in total. breeding cycle; and/or (2) the acoustic structure of The number of merged data was incrementally vocalizations may change, functioning as an ‘honest increased from two to ten for each experiment. We signal’ for the hormonal or reproductive state of an provided 90 experiments ie. combination of all sets individual. Here, we recorded acoustic behavior x nine merging levels. An improvement in accuracy within zebra finch pairs that were visually isolated. of recognition was achieved for 71.4% of these When acoustic contact is the only available commu- experiments. The achieved improvement was 1.7 nication channel, pairs have more synchronized and % on average, and the highest improvement was structured calling exchanges. Pairs were recorded in 13.4%. The best results consist of the merging of 4 non-breeding condition (water restriction), breeding songs, where an average of 2.1% improvement was condition (water ad lib), and while actively incuba- achieved. There was no direct correlation between ting eggs. Firstly, we quantified overall vocal activity any particular merging level and improvement. The and temporal aspects of the calling exchange: call results lead to the assumption that the method rate, time spent, the cross-correlation of female- cannot be apprehended as a universal means for male calling, and the frequency and duration of improvement. However it should be useful when it male song in visually-separated pairs. Secondly, we is generally tuned for acquisition of particular data quantified structural parameters of short calls and and specific requirements of the experiment. The male song. Determining whether the structure and/ results demonstrate how the method can partially or use of vocalizations shifts across breeding cycles eliminate disadvantages for cases when only a will help elucidate the mechanisms zebra finches limited number of records is available. The method use to coordinate reproductive attempts, and more should be beneficial in the cases for which there broadly contribute to our understanding of partner are a small number of songs, short vocal duration, dynamics in pair-bonded individuals. or insufficient recording quality.

137 22

Ptacek, Ladislav Quispe, Rene Vanek, J. Gahr, M. Linhart, P. Max Planck Institute for Ornithology, Seewiesen, Muller, L. Germany University of South Bohemia, Ceske Budejovice, Sebe, F. Czech Republic Université Jean Monnet, St. Etienne, France

Improving the automatic bird individual Seasonal and Daily Timing of Dawn Chorus in identification method by parametrization an Equatorial Songbird data merging Abstract In songbirds, the most prominent Abstract This research deals with a method for period of singing occurs prior to sunrise during improving an automatic identification of individual the reproductive stage, creating the so-called birds for cases when only a limited amount of ‘‘dawn chorus’’. Several factors are involved in the recorded data is available. The method applies regulation of singing at dawn. One important aspect pre-merging of the parametrization data before is the close relationship between dawn chorus processing the model of an individual. Although the and territorial establishment, which defines its model is based on the same amount of data, it can behavioral context within the annual cycle. Dawn be made more robust and discriminative because chorus can be stimulated by increased levels of of the data bundling. The testing was based on the song-controlling hormones, such as testosterone. Gaussian Mixture Model and Universal Background Besides, environmental factors including food Model (GMM-UBM) identification method. Ten sets availability, photoperiod, sunrise time, and climatic of chiff-chaff recordings were randomly compiled. variables can influence dawn chorus performance.

134 Thus, dawn singing is intimately linked to the the study of vocal learning and sexual selection. phenological characteristics of the species and Males’ undirected singing was recorded twice, ap- their niches. However, much less is known about proximately two and a half years apart. Longitudinal this behavior in tropical songbirds and there is even comparisons showed changes in performance fewer information available regarding equatorial parameters and song length. We discuss potential populations. We studied the dawn chorus timing in implications for intraspecific communication and silver-beaked tanagers (Ramphocelus carbo) from its relation to senescence. an equatorial population at the lowland Amazon of Brazil. During the course of a year, we recorded the seasonal and daily onset of dawn chorus. We 82 correlated this with circulating cortisol and testosterone levels, molting rate and the diet composition Rivera-Gutierrez, H.-R. estimated by stable isotopes measurements. In Jaramillo-Calle, V. addition, we analyzed these data in the context Martinez-Alvarado, D. of local abiotic changes, including sunrise and Universidad de Antioquia, Medellin, Colombia twilight timing, photoperiod and precipitation rate. The dawn chorus was displayed by silver-beaked Does song-learning play a role in acoustic tanagers for a long period during the year. We found adaptation to urban environments? well-defined seasonal/daily timing of dawn singing that differentially correlates with the expression of Abstract Bird species living in the cities may distinct life history traits. This study contributes to face limitations for communicating in noisy urban the scarce knowledge of song behavior in equatorial environments. Urban birds can avoid the masking songbirds, in the frame of the annual cycle and effect of urban noise by rapidly adjusting temporal the environmental characteristics that define the or spectral parameters of the song during noise lowland Amazon regions. exposure. For instance, they may increase the song amplitude, sing at higher pitch, or change the timing of singing. Whereas the strategies for 100 noise avoidance are well known for bird species that learn their song (Passeriformes: oscine), little Riebel, Katharina is known on the adaptations of Passerine species Varkevisser, J. that do not learn to sing (Passeriformes: subosci- Kriengwatana, B. nes). Given that many suboscine species inhabit Verhulst, S., Leiden University, The Netherlands and are common and successful in the cities, we tested whether song learning plays a role in the How ‚crystallised‘ is crystallised birdsong? acoustic adaptation of passerine species in urban Changes in singing over time in adult male environments. We hypothesized that suboscine zebra finches species may face limitations for rapid changes in acoustic or temporal patterns of their song (song Abstract Morphological sexual ornaments plasticity), and therefore, they may have developed and mate-advertising displays often show age- different strategies, compared to oscine species. dependent variation. Many songbird species are Using a correlational study in a paired design closed-ended learners that learn their songs early in (urban-rural) we tested whether song activity differs life during a limited sensory-motor learning phase between suboscines and oscines. In addition, using and do not change their song repertoire thereafter. a comparative phylogenetic analysis, we tested if However, even if no new songs are learned, singing song structure in suboscines that inhabit urban performance might change over time: song features environments is a preadaptation for effectively such as vocal consistency or tempo have been communicating in noisy places. Our results suggest suggested to change with age after song crystalli- that suboscine species lack plasticity and show zation. We therefore investigated whether long-term larger song activity, even when noise may mask changes occur in the crystallized song of the zebra acoustic communication. On the other hand, song finch, Taeniopygia guttata, an important model for characteristics of suboscines seem to confer an

135 abstracts poster

advantage for communicating in noisy places. Our Laguna, Spain study tests for first time the adaptations in both Almunia, J. oscine and suboscine species for communicating in Loro Parque Foundation, Tenerife, Spain urban environments. Looking for number of degrees of freedom at Orcinus orca calls for the design of a classifier 135 Abstract Orcinus orca acoustic behavior has Romero-Vivas, Eduardo revealed as a source of non-linear calls that inclu- Centro de Investigaciones Biologicas del Noroeste, des harmonic limit cycles, biphonations and even CIBNOR. La Paz, B.C.S, Mexico chaotic events. The sound production is based on Leon-Lopez, B. a complex nasal system, and its precise source is University of St Andrews, U.K. under discussion since various decades. Here we try to establish a methodology from differential Spectrogram differences approach for equation modeling that allows to calculate the extracting fundamental frequency variation number of degrees of freedom of the sound proof dolphin whistles duction device in these odontocetes. These number is related with the ability of the animal to choose Abstract Studying acoustic communication in a call and to control its production, something the wild usually presents the challenge of detecting that must be in a slower temporal scale than the signals of interest in noisy environments, which is fundamental frequency. We have developed a smart a common situation for underwater species. In ad- net of hydrophones that have been tested at Loro dition, the use of stationary underwater devices has Parque facilities in Tenerife Island. We start with increased the capacity of obtaining large amount the acquisition of data and the detection of call of acoustic data, leading to an increasing need for events, but our main objective is to produce a good accurate automatic detection methods of animal algorithm to classify these calls and even the caller. signals. Dolphin species are well known for having a The determination of the number of degrees of free- complex acoustic repertoire which includes pulsed dom, is directly related with the space parameter (echolocation clicks) and tonal sounds (whistles). to classify. We present a method, based on the differences of contiguous windows of the spectrogram, to detect and identify modulation of the fundamental fre- 120 quency of tonal sounds. We tested the effectiveness of the method using recordings from two dolphin Ruiz, A.Tonatiuh species, long-beaked common dolphins (Delphinus Jung, K. capensis) and bottlenose dolphins (Tursiops trunca- Schwenker, F. tus) in the Gulf of California, Mexico. Results show University of Ulm, Germany that the structure of whistles can be clearly iden- Tschapka, M. tified even in noisy environments. This approach University of Ulm, Germany and Smithsonian is simpler than other current methods, but with a Tropical Research Institute, Balboa, Panama high effectiveness in signal detection and it could be used to complement acoustic classification Software-aided identification of Neotropical methods. bat calls: an active learning approach.

Abstract Recent years saw the development of 115 computational methods for the acoustic identification of bat species, using supervised-learning Rosa, Fernando techniques. In order to train an accurate model it is Sanluis, J.C. necessary to dispose of a large call library previous- Lüke, J.P. ly verified by human experts. However, usually the Universidad de La Laguna, San Cristobal de La manually verified data is scarce since the manual

136 species identification of calls is time-consuming of the right wing, which bears a scraper. Sound is and requires high skills. To address this problem produced when the scraper is passed across the we propose a novel method for software-aided file. Bushcrickets song frequencies usually range identification of bat species implementing active from audible signals (5 kHz) to low ultrasonic (30 learning algorithms. This method aims to minimize kHz). We report on the discovery of Supersonus, a the human supervision for training accurate classi- new genus with three new species of predaceous fication models. As database we used a library of katydid (Insecta: Orthoptera) from Colombia and echolocation calls of Neotropical bats from Mexico, Ecuador in which males produce the highest fre- Panama and Costa Rica with more than 30 species. quency ultrasonic calling songs ever recorded from We designed a user interface (BatAL) to detect and an arthropod. Males of Supersonus spp. call fema- classify sequences of bat calls in field recordings. les at 115 kHz, 125 kHz, and 150 kHz. Exceeding In addition to an automated classification process the human hearing range (20 Hz - 20 kHz) by we implemented an active learning algorithm (que- one order of magnitude, these insects emit their ry strategy) that allows to retrain the classification ultrasound at unusually elevated sound pressure model with new call data. The active learning levels (SPL). In all three species these calls exceed algorithm processes the new training data and 110 dB SPL rms re 20 µPa (at 15 cm). Males of selects the most informative samples. Then this Supersonus spp. have unusually reduced forewings selection is presented to the user (human expert) (<0.5 mm2). Using high-speed video synchronized to reconfirm the calls´ species suggested by the with ultrasound-sensitive equipment, and Laser current classification model. After each query the Doppler Vibrometry, we demonstrate that the right classifier model is retrained, including the newly wing radiates appreciable sound, and that the file- verified calls. Additionally, to asses which is the bearing wings (left wing) does not show a particular optimal configuration of classification models and resonance. In contrast to most katydids, males of query strategies we conducted a performance test. Supersonus spp. position and move their wings We evaluated the accuracy of query strategies such during sound production so that the concave aspect as Uncertainty Sampling and Query-By-Committee of the right wing, underlain by the insect dorsum, combined with classifiers such as K Nearest forms a contained cavity with sharp resonance. The Neighbors and Random Forest configured singly or observed high SPL at extreme carrier frequencies is as an ensemble. Our results show that our active better explained by wing anatomy, a resonant cavity learning approach is more efficient than supervised with membrane, and cuticle deformation. learning techniques, as we achieve higher identification accuracy with less training call samples. 130

131 Schneiderová, Irena Šmíd, J. Sarria-S., Fabio A. Smetanová, M. Montealegre-Z., F. Lhota, S. University of Lincoln, U.K. Brandl, P. Morris, G.K. Bolfíková, B.C. University of Toronto at Mississauga, Canada Czech University of Life Sciences, Prague, Czech Windmill, J.F.C. Republique Jackson, J. University of Strathclyde, Glasgow, U.K. Species determination of captive lesser galagos (Galago spp.) integrating molecular- Loud extreme ultrasonic frequencies produced genetic and bioacoustic data by a small generator Abstract Lesser galagos (Galago spp.) typically Abstract Male bushcrickets sing by rubbing their show a lack of overt morphological diversity, which wings together to attract distant females. The left complicate their correct determination, thus wing bears a vein modified as a file and sits on top breeding management at zoos. On the other hand,

137 abstracts poster

they produce highly species-specific vocalizations significantly different from those of adults and that that can be recommended for species determina- the alarm calls underwent nearly the same changes tion. We used molecular-genetic and bioacoustic in all studied individuals. All juveniles emitted approach to determine the lesser galagos with alarm calls consisting of one element with almost unclear taxonomic status kept at European zoos. constant frequency, but their alarm calls included For this purpose, all major lines were sampled a second frequency-modulated element after their for DNA and their species-specific advertisement first hibernation as adults. Our data show that calls were recorded. Both approaches were applied the duration of the first element is significantly successfully, and showed high potential for species shorter in adults than in juveniles. Additionally, the determination of the lesser bush babies. Together frequency of the first element is significantly higher with data extracted from studbooks, our results in adults than in juveniles. Similar to previous confirmed that G. moholi is kept at five zoos, and findings in other Palearctic ground squirrel species, G. senegalensis at 18 zoos. Surprisingly, molecular- our data are inconsistent with the assumption that genetic data showed that there are two clearly juvenile mammals emit vocalizations with higher separated genetic lineages of G. senegalensis at fundamental frequencies than adults. However, European zoos. The first lineage is represented by our results do not support the previously suggested animals originating from Ghana and Togo and the hypothesis that juvenile ground squirrels conceal second lineage by animals originating from Guinea. information regarding their age in their alarm calls Bioacoustic data indicate that the observed variabi- because we found significant differences in alarm lity is at an intra-species level. Study was supported calls of juveniles and adults. by Internal Grant Agency of the Czech University of Life Sciences in Prague (CIGA CULS Prague) project no.: 20134311. 12

Schoeppler, Diana 64 Schnitzler, H.-U. Denzinger, A. Schneiderová, Irena University of Tuebingen, Germany Schnitzerová, P. Uhlíková, J. Non-paradoxical Doppler shift compensation in Brandl, P. Rhinolophus paradoxolophus Mateju, J. Czech University of Life Sciences, Prague, Czech Abstract Horseshoe bats emit long CF-FM sig- Republique nals to detect prey in narrow space. By compensa- ting Doppler shifts created by their own flight speed Differences in alarm calls of juvenile and adult the echo frequency is kept in the auditory fovea European ground squirrels: findings from a which allows the evaluation of the prey generated semi-natural enclosure flutter information in the echoes. In horseshoe bats, the species specific CF-frequencies of the echolo- Abstract The European ground squirrel cation calls correlate negatively with body size. (Spermophilus citellus) emits alarm calls that However, members of the poorly investigated R. warn conspecifics of potential danger. Although philippinensis-group are outliers as they emit unit has been observed that inexperienced juveniles expected low call frequencies despite of their small of this species emit alarm calls that sound similar body size. Rhinolophus paradoxolophus, a member to those of adults, studies focusing on juvenile of this group, was named according to the deviating alarm calls are lacking. We analyzed the acoustic and paradoxical morphology of the nose leaf and structure of alarm calls emitted by six permanently the ears. In this study we investigated whether this marked European ground squirrels living in a species also use Doppler shift compensation, and semi-natural enclosure when they were juveniles if yes, whether the precision of the Doppler shift and after one year as adults. We found that the compensation system is similar to that of other hor- acoustic structure of the juvenile alarm calls was seshoe bats. We trained three R. paradoxolophus to

138 fly from a starting point to a landing grid and ana- Etroplus maculatus possessed anterior extensions lyzed the echolocation and flight behavior. During connecting the swim bladder to the inner ears. approach bats shortened call duration and pulse Hemichromis guttatus and E. maculatus were able interval and increased duty cycle and terminal FM to detect frequencies up to 3 kHz (SPL) whereas bandwidth. They also compensated Doppler shifts S. tinanti only responded to frequencies up to 0.7 with high precision and kept the calculated echo kHz. In addition, E. maculatus showed significantly frequency constant at a reference frequency inde- higher auditory sensitivities at 0.5 and 1 kHz than pendently from their flight speed. Even the offset the two species lacking anterior swim bladder between the reference and resting frequency was extensions. Our results indicate that anterior swim lower than in other horseshoe bats. Despite of the bladder extensions seem to improve significantly au- paradoxical morphology, the echolocation behavior ditory sensitivities between 0.5 and 1 kHz. Besides in approach flight and Doppler shift compensation anterior extensions, the size of the swim bladder of R. paradoxolophus are non-paradoxical and very appears to be an important factor for extending the similar to those of other horseshoe bats. detectable frequency range (up to 3 kHz).

8 5

Schulz-Mirbach, Tanja Seltmann, Susanne Ludwig Maximilians University Munich, Germany Max Planck Institute for Ornithology, Seewiesen, Ladich, F. Germany University of Vienna, Austria The influence of sleep on song-related Diversity in swim bladder morphology affects neuronal activity in RA – what role does hearing in cichlid fishes Melatonin play?

Abstract For animals, the sense of hearing plays Abstract Sleep is essential for song learning and an important role to gain information about biotic song production in birds. The song control nuclei (e.g. predators, prey, conspecifics) and abiotic HVC and RA, which both show changes in their sound sources in their environment. Several bony neural activity patterns during sleep, play a crucial fish species have evolved anterior extensions of the role in these processes. The presence of Mel-1B swim bladder which come close to or directly contact receptors in both nuclei suggests that melatonin the inner ears which may result in improved hearing. may be directly involved. However, it is still largely So far only few comparative studies on two groups unknown how melatonin affects song structure and (drums and squirrelfishes) have shown that these its neuronal correlates. To address this question we morphological specializations may enhance hearing used wireless transmitters (~1g including batteries) abilities; however, outcomes of these investigations to record the neuronal activity of individual premotor were in part contradictory. Here we thus asked neurons in the RA of free-ranging zebra finches up whether the diversity of swim bladder morphology in to several days while simultaneously manipulating the speciose family Cichlidae affects their hearing melatonin levels. By using melatonin-containing abilities. We studied swim bladder morphology cream we were able to simulate and maintain natu- and inner ears by dissections and by making 3D ral melatonin levels for several hours with a single, reconstructions based on high-resolution microCT non-invasive treatment. Our results suggest a direct imaging. Auditory sensitivities were determined in influence of melatonin on the activity pattern in terms of sound pressure levels (SPL) and particle RA. Melatonin was found to increase the interspike acceleration levels (PAL) using the auditory evoked interval during the steady firing that RA projection potential (AEP) recording technique. The swim blad- neurons show during non-vocal periods, an effect ders in the jewel cichlid Hemichromis guttatus and normally observed in the transition to sleep. the rheophilic Steatocranus tinanti lacked anterior Furthermore, we found that melatonin is crucial for extensions and the swim bladder was distinctly small the occurrence of an auditory response in RA that in the latter species. In contrast, orange chromide occurs spontaneously, or can be triggered by play-

139 abstracts poster

back of bird’s own song, during sleep. In absence call fundamental frequency. Siberian wapiti voca- of melatonin these spiking patterns did not occur. lizations were more similar to respective call types We therefore speculate that the effects of melatonin of American subspecies than those of European red on RA firing properties are important for the sleep deer, with intermediate values of fundamental fre- related maintenance of song structure and could quency between them, substantially higher than in also be involved in song learning. European subspecies, but substantially lower than in American subspecies. The research was funded by grants from the Russian Scientific Foundation, 86 grant No 14-14-00237.

Sibiryakova, Olga Volodin, I.A. 119 Lomonosov Moscow State University, Russian Federation Simon, Ralph Volodina, E.V. Rupitsch, S. Moscow Zoo, Russian Federation Friedrich-Alexander University Erlangen- Nuremberg, Germany Acoustic variation in Siberian wapiti Cervus Steckel, J. elaphus sibiricus: effects of sex and age Peremans, H. University of Antwerp, Belgium Abstract Red deer Cervus elaphus is the widespread species that forms many subspecies Driverless vehicle guidance by using displaying a strong divergence of vocal characteris- bio-inspired sonar reflectors tics. Studying vocal divergence across subspecies and sex and age-classes of Cervus elaphus might Abstract In tropical South and Central America, help to trace the evolution of vocal communication some 1000 plant species chose a quite rare in this species. Unlike well-studied European pollination system: they are pollinated by bats. As subspecies, acoustic variation of Asian subspecies bats are active at night, so too the flowers open is poorly understood. We analyzed frequency, and produce nectar at night. These plants must temporal and power variables of 251 contact and surmount a major challenge; in the absence of rutting calls, collected from 20 individuals (3 light they cannot attract their pollinators with calves, 9 hinds and 8 stags) of the most abundant conspicuous colors, as other plants do. Instead Asian subspecies, Siberian wapiti Cervus elaphus they have found a sophisticated solution: they sibiricus. We registered open-mouth (oral) and the have developed floral reflectors, which reflect the closed-mouth (nasal) contact calls in all sex and ultrasound signals of bats in special ways, making age-classes, whereas the open-mouth rutting calls the flowers acoustically conspicuous to the bats. were only produced by stags. The maximum fun- The trick of these concave shaped reflectors is that damental frequency was similar between contact the echoes are not only broadcast over a wide range calls of calves and hinds (in oral contact calls 1.50 angles of sound incidence but are also marked with vs. 1.39 kHz, and in nasal contact calls 1.06 vs. a spectral signature, which is acquired through an 0.81 kHz, in calves and in hinds respectively). The interference mechanism. Experiments show that maximum fundamental frequency of stag oral con- such passive acoustic signals can reduce bats’ tact calls (0.87 kHz) was significantly lower than search time for flowers by 50%. Moreover, these those of contact calls of hinds or calves, and lower signals not only help guide bats to small targets than in stag rutting calls (1.20 kHz). This study is (i.e., flowers) in acoustically complex environments, the first reporting emission of contact calls (oral they also provide information about the flowers’ and nasal) by stags of Cervus elaphus; previously, orientation, helping coordinate the animals’ ap- contact calls were found only in hinds and calves. proach. We think that such floral reflectors, shaped Siberian wapiti contact calls were close in values of by evolution to specifically „respond“ to the bats fundamental frequency between hinds and calves, sonar signals, could inspire the design of artificial what means only a minor ontogenetic decrease of sonar reflectors that may facilitate orientation as

140 well as navigation of robots and perhaps of cars, as Murrah buffaloes from their herd. The mean call they too are typically equipped with sonar systems. duration, 1st formant, periodicity and degree of Such systems are often guided by artificial land- voice breaks of adult lactating Murrah buffaloes marks, including visual signals, laser beacons, and were observed to be 2.15±0.05 sec, 900.11 ± omni-directional radar reflectors, which are already 4.21 Hz, 95.31% (183.95 periodic pulses out of in use to streamline the navigation efficiency of 192.99 pulses) and 20.78 ± 0.89% respectively. robots. However, there is a lack of suitable sonar By using MFCCs the machine learning algorithm landmarks for airborne sonar systems. We think along with J48 classifier revealed maximum ef- that sonar reflectors inspired by the form of floral ficiency of accurate recognition of Murrah calves, reflectors can easily be used as landmarks. They heifers and adults with a recognition rate of could help guide towards targets and they could 95.67% (k= 0.9515), 94.55% (k = 0.9378), and function as beacons indicating obstacles. Moreover 94.50% (k = 0.937) respectively; while it could different forms and types of reflectors may transmit actually differentiate these with a recognition rate certain kinds of information. of 79.72% (k = 0.7727), 73.89 % (k = 0.702), by 74.21% (k=0.7037) respectively.

113 89 Singh, Yajuvendra College of Veterinary Science and Animal Singh, Yajuvendra Husbandry; DUVASU, Mathura, India UP Pandit Deen Dayal Upadhyaya Pashu Chikitsa Lathwal, S. S. Vigyan Vishwavidyalaya Evam Go-Anusandhan National Dairy Research Institute, Karnal, Sansthan, Mathura, India Haryana, India Lathwal, S. S. National Dairy Research Institute, Karnal, Bioacoustic sensing an authentic descriptor Haryana, India of individual‘s identity Vocalization a passive indicator of Abstract The present investigation was perfor- developmental stages of dairy animals med to investigate the significant differences for various acoustic features of vocal signals articu- Abstract Vocal differences within and between lated from different individuals of a buffalo herd. the individual animal was used as a basis for discri- Analysis of bioacoustics features extracted from mination of calves, heifers and adult individuals of 300 voice samples of ten adult lactating Murrah KF cows and Murrah buffaloes in present investiga- buffaloes revealed that differences for amplitudes tion. The vocal signals produced from ten individu- (minimum, maximum and mean), total energy, als of each and every category were recorded and mean power, pitch (median, minimum, range and edited in order to prepare smaller stationary frames mean), mean (p<0.05), intensities (minimum and of 10 ms from a total of at least 50 vocal signals maximum), formants (F1, F2, F3, F4 and F5), from each and every individual. Various vocal bandwidths (B1 (p<0.05), B3 and B4), number of signals were analyzed in order to find out the best pulse, number of period, mean period, unvoiced suited acoustic feature for identification of calf, frames, degree of voice breaks, jitter, shimmer, heifer and adult stages of dairy animals. The vocal mean noise to harmonic ratio (%) and mean har- signals from calves, heifers and adults KF cows and monic to noise ratio (dB) were highly significant Murrah buffaloes differed significantly (p<0.001) (p<0.001). Out of these only few acoustic features for acoustic features such as amplitude (maximum), viz. formants (F1, F2, F3, F4 and F5), number of total energy, mean power, pitch (Q50%, minimum, pulse, number of period and degree of voice breaks maximum and mean), intensity (mean, minimum were observed to have significant difference for and maximum), formants (F1, F2, F3, F4 and F5) each and every individual dairy buffalo, hence only bandwidth (B1, and B4), number of pulse, number these features could be selected as the best suited of periods, mean period, unvoiced frames, degree of acoustic features for discrimination of individual voice breaks, mean noise/harmonic ratio and mean

141 abstracts poster

harmonic/noise ratio. Regression coefficient ‘b’ for if improved communication with common nigh- linear trend line of resonance frequency curve was tingales reduces the cost of competition between observed to be the best suited acoustic feature in males. Last, mixed singing might simply result from discrimination of calf, heifer and adult stages of misdirected learning of thrush nightingales, if not dairy animals. For pattern recognition MFCCs were costly enough to be counter-selected. On the poster, extracted from all the frames of vocal signals and we will present the first detailed comparison of the were utilized in preparation of training and test repertoire size between sympatric and allopatric models for decision making using machine learning thrush nightingale males, focusing on the song type algorithm. The machine learning algorithm with diversity level. Apart from a direct additive effect of J48 classifier was able to discriminate the calves, whole-song copying, analysis of song syllables show heifers and adults of KF cows and Murrah buffaloes the integration of heterospecific syllables within with an overall rate of discrimination 87.91% and typical thrush nightingale song types. Surprisingly, 87.20% respectively. Moreover calves of both the some sympatric males also display a higher rate of species could be discriminated with a higher rate song improvisation, out of usual species repertoire. of recognition (98.83 and 92.35% respectively) Furthermore, we will present spatial distribution of than those of heifers and adults. The efficiency of males in sympatric area and investigate the rela- the machine learning algorithm for actual discri- tionship between the extent of mixed singing and mination of calves, heifers and adults of KF cows density of both species. Though the previous hypo- and Murrah buffaloes was sufficient enough (k = theses would predict to find more mixed singers in 81.77% and 80.75% respectively). close range of common nightingale locations, the actual territory partitioning seems more complex. We conclude by proposing some leads on further 107 potential explanations for the occurrence of mixed singing. Souriau, Abel Charles University of Life Sciences in Prague, Czech Republic 43

Asymmetrical song convergence in two Sprau, Philipp hybridizing nightingales: The mystery of Dingemanse, N. “doublespeaking” males Ludwig Maximilians University, Munich, Germany and Max Planck Institute for Ornithology, Abstract Nightingale species are considered Seewiesen, Germany among the most skilful song performers of European songbirds, by both the diversity of their song and Plasticity or consistency? Individual their learning abilities. The common nightingale differences in territorial defence along urban (Luscinia megarhynchos) and thrush nightingale gradients in great tits (Parus major) (Luscinia luscinia) are two closely related species sharing a recent sympatric zone across central Abstract Next to climate change the United Europe. Interestingly, many thrush nightingale ma- Nations considers urbanisation the biggest envi- les in this contact zone develop a “mixed singing” ronmental challenge of our time. In the coming strategy, copying parts or sequences of songs from decades the growth of urban areas is expected to the other species’ repertoire. Several hypotheses attain unprecedented levels with detrimental ef- were proposed to explain the occurrence of such fects on ecosystems. Whether wild animals will be song convergence. First, evolution of mix-singing able to survive in an increasingly urbanized world behaviour might be driven by female preferences will depend on their capacity to adapt to these new for males with larger song repertoire, as shown environmental conditions. In this study we investi- in several other songbird species. Second, the gated the impact of a suite of environmental traits two species have similar habitat preferences, and – i.e. noise, temperature, light, human activity and interspecific territoriality occurs in sympatry. Mixed vegetation structure – that were quantified along singing may constitute an advantageous strategy urban gradients on territorial behaviour in an urban

142 great tit (Parus major) population. Using playback also track whether the sign of correlation between experiments we recorded male responses towards pitch and tempo of song syllables is a feature simulated territorial intruders. These intrusions inherited by juveniles from their tutors or can it were simulated repeatedly to reveal whether birds be individually changed. Further, we explored the respond plastic to intruders depending on the relationships between pitch/tempo fluctuations and environment or whether birds respond consistently brain temperature (above HVC, the ‚song-tempo ge- across different environmental conditions. This nerator‘) as well as air-sack pressure, both of which study will show decisively how urbanisation alters may independently exert influences on pitch and behavioural patterns within and between individu- tempo. Project supported by the Swiss National als and thereby provides important insights into Science Foundation Project 31003A_127024 and human-induced micro-evolutionary processes. ERC-2010-AdG Project No 268911 VOTECOM.

152 116

Stepien, Anna E. Stowell, Dan Canopoli, A. Clayton, D. Vyssotski, A. Queen Mary University of London, U.K. Hahnloser, R.H.R. University of Zurich and ETH Zurich, Switzerland From raw audio to songbird communication networks: fully automating the analysis Syllable-specific rules in the microstructure of birdsong Abstract There has been recent progress and much interest in fully automatic bioacoustic Abstract Male zebra finches learn their song by analysis such as multiple species recognition. Full imitating the song of a tutor. After fledging they un- automation is important because any step that dergo a practicing period during which they display requires manual intervention (such as selecting high variability in song structure. Once they reach regions of song) becomes a bottleneck in scaling adulthood, they sing a stable song that consists of up to big data scenarios such as sound archives a number of syllables produced almost always in or large acoustic monitoring programmes. Species the same sequence. How stable is the adult song? detection and recognition give us very useful mo- We find that in most birds the pitch of individual nitoring data, but beyond those „labelling“ tasks syllables during undirected song tends to follow a we would like to extract details of inter-individual circadian pattern that is characteristic for that bird. communication patterns such as contact calls and Typical pitch fluctuations consist of higher pitch in counter singing. We will describe recent work in the morning and lower pitch in the evening, with machine learning methods for analysing songbird overnight discontinuities. Given the known daily communication networks, using a case study with variation in song tempo (syllable duration), we ex- zebra finch recordings. This includes automatic call plored whether there existed correlations between detection designed to handle multiple overlapping tempo and pitch within a syllable and whether such calls, and automatic inference of inter-individual correlations agreed with a simple stretching hypo- communication tendencies from group calling thesis according to which the pitch trace within a patterns. syllable is fully determined by song tempo, i.e., by the time-warped pitch trace of a reference syllable. Correlations between tempo and pitch allow us to 36 identify different types of syllables. Some syllable types agree with the stretching hypothesis. But Sueur, Jerôme others, such as those involving down-sweeps, beha- Muséum national d’Histoire naturelle, Paris, ve in anti-stretched manners. That is, in the latter France syllable types, the pitch in shorter renditions of the syllable is higher than that in longer renditions. We Ecoacoustics: echoes from a new discipline

143 abstracts poster

Abstract In the past few years, numerous works consequence of social learning behaviour) between dealing with animal acoustics did not strictly con- territorial males as an indicator of changes in sider sound as a research topic in animal behaviour populations of Ortolan Bunting Emberiza hortulana but rather as a tool to tackle questions in ecology. - a farmland species undergoing steep declines in Sound emanating from animals, whichever their most parts of Europe. We showed that in an agri- identity or habitat, can be used as an indicator cultural landscape of Western Poland the pattern of ecological processes and thus as a material to of repertoire sharing varied between landscape investigate biodiversity monitoring, community plots and depended on habitat fragmentation. In ecology, landscape ecology, population biology, two more fragmented populations, males (N=73) biogeography, and conservation biology. This new tended to share less repertoire with neighbors way to see, or better said to listen, animal sound than with non-neighbors. On the other hand, in mainly results from a change of scale and goes two populations with higher densities and smaller hand in hand with the development of autonomous distances between males (N=66), Ortolan Buntings recording methods, batch sound analysis tools and shared higher proportion of the repertoire with peculiar theoretical concepts. This new research their closest neighbors than with non-neighboring can be formalised under a new discipline named males. Our results suggest that changes in vocal ecoacoustics (Sueur & Farina, 2015). In the int- communication behaviour can reflect the emergent roduction to the session „Ecoacoustics: a change population status of species vulnerable for rapid of scale in the study of animal sounds“, I will try decline due to habitat destruction. We emphasize to briefly define ecoacoustics and to introduce the that studying animal cultural traits in landscapes main challenges this new discipline will have to that are changing due to human activity can be face in the future. more effective and informative than traditional methods based on long-term population monitoring.

51 25 Szymanski, Pawel Osiejuk, T.S. Tanttu, J.T. Adam Mickiewicz University, Poznan, Poland Puhakka, L. Ijäs, A. Vocal communication behaviour as an indicator Tampere University of Technology, Pori, Finland of changes in bird populations: a case study of farmland species Time-frequency analysis of the Marvellous Spatuletail (Loddigesia mirabilis) vocalizations Abstract The increasing risk of extinction of wildlife populations is caused to a great extent by Abstract Marvellous Spatuletail (Loddigesia human-driven habitat fragmentation. Therefore, it mirabilis) is a hummingbird species endemic to is crucial for conservation biologists to work out Peru. The species exhibits lek behavior: during effective methods for monitoring and management breeding season, individuals gather in lekking sites of declining populations. Such methods should where the males perform so-called display flights provide valuable information on current status of to attract mates. During these display flights, the population which could serve as an early warning male raises its long spatule-like tail feathers to system when negative changes in a core population its sides and produces a high frequency clicking start to proceed. Vocal communication is essential sound. This sound was originally suspected to ori- for reproduction and any variation in functional ginate from the tail feather shafts snapping against significance can negatively affect individual fitness each other; however, high-speed camera footage and population growth rate. Thus, incorporating the recently obtained by the BBC has proved this not study of cultural traits in a conservation has been to be the case. We recorded Marvellous Spatuletail recently suggested as a novel way to monitor the vocalizations (male song, bursts of display clicks diversity and status of animal populations. Here produced by the males, and female song) in three we explored the potential of sharing repertoire (a lekking sites in Huembo, Northern Peru, in autumn

144 2011. The recordings were made using in the left in mediating selective associations during fission- channel of the recorder (Fostex FR-2) an ultrasound fusion events in peach-fronted conures. In this microphone (Laar 10-2) and in the right channel study we conducted a semi-interactive playback a ‘normal’ audio microphone (Sennheiser ME-66). experiment on peach-fronted conures in captivity. We collected the recordings opportunistically, The study was conducted in order to determine waiting at known lekking sites in the early mornings if contact calls emitted during vocal interactions and late afternoons, and recording individuals that contain information about the identity of the vo- vocalized spontaneously. In this paper we present calizing individual that could facilitate individual the results of an extensive time-frequency analysis recognition during fission-fusion events. We found of the vocalizations. Besides the traditional spec- that the contact calls emitted as responses to our trograms we also apply the Wigner-Ville distribution playbacks contained information about the identity to study the energy distribution of the sounds. This of the vocalizing individual and that the individual approach proved useful especially with the clicking distinctiveness persisted in interactions where the sounds, which have very short time duration and individual imitated the playback. The existence of contain a wide range of frequencies up to about such a robust individually distinctiveness could 30 kHz. Based on this analysis we also discuss allow not only individuals taking an active part in some potential production mechanisms for the high interaction but also listeners extracting information frequency clicking sounds. from the interaction, to integrate the identity of the caller with social knowledge, such as memories of previous social interactions, that could be used 106 in deciding who to group with during fission and fusion events. Thomsen, Heidi Dabelsteen, T. University of Copenhagen, Denmark 134 Balsby, T.J.S. Aarhus University, Denmark Tipp, Cheryl The British Library, London, U.K. Individual variation and vocal matching by imitation in peach-fronted conures Archiving the sounds of life on earth (Aratinga aurea) Abstract The British Library Sound Archive has Abstract The peach-fronted conure, Aratinga au- one of the world’s largest and most comprehensive rea, lives in a fission-fusion social system, charac- collections of bioacoustic sound recordings. terized by frequent changes in flock composition, Founded in 1969, the collection now contains more during the non-breeding season. The inherent flui- than 200,000 catalogued recordings of wildlife dity of their social dynamics presents peach-fronted and environmental recordings, many of which are conures with a unique set of challenges, such as unpublished and unique to the British Library. The frequent establishment of dominance hierarchies wildlife collection was originally formed to serve as and fluctuating levels of competition for resources. an international resource for bioacoustics research, Therefore, the ability to choose good group mates providing copies of recordings to researchers and may be essential for individuals in order to reduce archiving field recordings for future generations to competition and aggression.During flock encoun- benefit from. Today, recordings are also provided for ters, peach-fronted conures exchange loud contact educational projects, exhibitions, private listening, calls just prior to fissions or fusions. During these artistic sound installations, musical compositions, vocal interactions peach-fronted conures will mo- broadcasting and commercial publications. Visitors dify the fine-scale structure of their contact calls to to the British Library’s Reading Rooms can engage imitate that of another interactant. Vocal matching with the collection in various ways. Digitised con- by imitation has been suggested to be a flexible tent can be accessed through dedicated terminals way to address specific individuals in dynamic com- while analogue material is currently delivered munication networks and could thus be important through an appointment based service. Remote

145 abstracts poster

access is available through British Library Sounds, in the other members of the group. Furthermore, an online resource which provides streaming access the bark was not exclusively used as an alarm call to thousands of hours of recordings from across and, in fact, was never used in this situation by the the sound archive, including a range of wildlife pups. Three other stimulus evoked barks: stimulus and environmental sounds from 1936 until the associated with food delivery (elicited more barks present day. Soundcloud is another platform by adults than by juveniles), with playing behavior that is being increasingly utilised as a means of (elicited more barks by juveniles), and the “stan- providing free, downloadable access to a growing ding up and moving away” behavior performed by number of recordings. In May 2015, the British social companions that were resting close to the Library was awarded a Heritage Lottery Fund grant emitter. Whether barking in these situations means of £9.5 million to help digitise the sound archive’s the same thing or is a matter of maturing and lear- most unique and at risk recordings. This grant will ning skills, we still have to investigate. We suggest provide much needed digitisation of the library’s that the social status and age of the caller are analogue wildlife and environmental recordings, si- recognized by the conspecifics through their bark gnificantly improving both onsite and offsite access and that it influences the response of the others for the researchers of today, as well as preserving the collection for the researchers of tomorrow. 20

74 Ulloa, Juan S. Sueur, J. Ueno, Adriana S.O. Muséum national d’Histoire naturelle, Paris, France Geraldi, M. Gasc, A. Monticelli, P.F. Muséum national d’Histoire naturelle, Sorbonne University of São Paulo, Brazil Universités, Paris, France, Gaucher, P., CNRS USR 3456 Guyane, France Barking in capybaras: more than just an Aubin, T. alarm call Université Paris-Sud, Orsay, France Réjou-Méchain, M. Abstract Alarm calls can be subjected to the Université Paul Sabatier-ENFA, Toulouse, France influence of the social, ontogenetic and environmental pressures. We investigated the influence Temporal and spatial dynamics of a tropical of such factors in barking behavior, traditionally soundmark: the song of the Screaming piha described as a warning social signal, in a captive Lipaugus vociferans group of capybaras, a species with a high level of sociality structurally organized by a linear domi- Abstract Acoustic monitoring has proved to be nance hierarchy, at least at the higher positions. an efficient approach to monitor wildlife, notably We studied a group of 10 individually recognized in environments with limited visibility. Surprisingly, animals (8 adults and 2 3-month-old juveniles) in large bioacoustic monitoring programs focusing on a zoo in the municipality of Ribeirão Preto, Brazil, tropical environments are rather rare. The coexis- between February and May 2014. We sampled bar- tence of many tropical species generates a fasci- king behavior during two-hour observation sessions, nating and complex acoustic environment, called repeated twice a week. Whenever a bark was heard, soundscape, which remains poorly understood in its we registered the identity of emissor (social status structure and dynamics. Here, we adapted signal and name), the reactions of all the other members processing techniques to reveal the spatio-temporal of the group, and the stimulus that triggered the dynamics of a remarkable soundmark of the Amazon emission. Our results showed that the dominant basin: the song of the Screaming Piha (Lipaugus male barked more often than the other adults vociferans). Using recordings made with an array and almost in the same rate of juveniles; and that of 24 microphones in the French Guiana lowland social status of the emitter correlated positively forest, we implemented and tuned a detection with its success in promoting defensive behaviors system based on spectrogram cross-correlation to

146 trace the vocalizations of L. vociferans. The system structures in bird recordings and use them for in- scanned nearly 960 hours of recordings, revealing dividual identification. In species with very simple novel ecological information and rising further vocalizations, individuals may differ in frequency questions about this species. We found that spatial or temporal characteristics of songs; in those with patterns of lower activity corresponded to the zone complex songs, individual-specific characteristics with lowest trees—a known liana-infested forest can be found in repertoire composition. Our model patch—reflecting that habitat of the Screaming species with a relatively simple song, in which we Piha is the mature primary forest. The sampling want to apply repertoire-based individual acoustic sites near the creeks had more detections than the monitoring, is the Yellowhammer (Emberiza sites further away, revealing that the lek mating citrinella). In an extensive database of over 3000 arena could be distributed strategically to have a recordings available from a citizen science project nearby source of water. We also found a marked monitoring the distribution of Yellowhammer dia- temporal pattern. The lek was active during the lects throughout the Czech Republic, we observed whole day, from sunrise to sunset, with two peaks a substantial variation in initial song phrases. of activity shifted by more than two hours from the Almost every male’s recording, of a sufficient dawn and dusk chorus. This noticeable presence quality, contained a unique set of initial phrases, on the diurnal soundscape confirms its status of indicating a high variation among individuals. soundmark of the Amazon. Species associated with These results suggest that despite a relatively these iconic sounds could provide a focus for rising simple Yellowhammer song structure, the variation awareness and action; a flagship species in terms of in initial phrases can be used in field research for conservation biology. recognition of males based on their vocalisation. To test whether such approach is indeed feasible, we repeatedly (multiple times a week) record 81 colour-ringed males in a dense Czech population of Yellowhammers throughout the breeding season van Boheemen, Sharina J. of 2015. Allowing us to evaluate the repertoire Kauzal, O. stability within individuals and among-male vari- Samperi, F. ation on a detailed temporal and spatial scale. On Cinerová, M. the poster, we will present the first results of this Diblíková, L. detailed pilot study. Petrusek, A. Petrusková, T. Charles University of Life Sciences, Prague, 98 Czech Republic Vellema, Michiel Simple song doesn‘t mean low variability – University of Southern Denmark, Odense, Denmark individual variation in the Yellowhammer Zsebök, S. vocalization Université Paris-Sud, France Bascones, S. Abstract When studying wildlife ecology, it is van der Linden, A. often important to distinguish between different University of Antwerp, Belgium individuals. In some cases, acoustic monitoring, Gahr, M. which is less invasive than traditional marking Max Planck Institute for Ornithology, Seewiesen, methods, could be used for that purpose. It is well Germany known that songbirds can use song characteristics to distinguish between conspecifics; species with Prior singing experiences accelerate both simple and complicated songs typically dif- subsequent development of vocal skills ferentiate between vocalizations of an unfamiliar bird and a known neighbour. This opens the ques- Abstract Seasonally breeding songbirds such tion to what extent people (researchers in particu- as the canary (Serinus canaria) need to re-acquire lar) are also able to detect individual-specific song vocal skills each year during the early breeding

147 abstracts poster

season. The detailed temporal dynamics of INRIA,France vocal skill development are not well understood Soula, H.A. however, and it is unclear in what way the initial EPI BEAGLE INRIA, INSERM INSA LYON, France development of song differs from the seasonal re- acquisition of song. Female canaries seldom sing Behavioural changes in response to back- under natural conditions, but administrating the ground noise sign the importance of intra-pair hormone testosterone stimulates them to develop communication at the nest in a songbird singing skills that are typically reserved for males. In this study we investigated the dynamic acqui- Abstract Intra-pair communication might sition and re-acquisition of song after repeated play a key role in the coordination of behaviours testosterone treatments. The testosterone-induced between mates, which is a central aspect of song of female canaries went through the cha- the biology of monogamous pair bonds in birds. racteristic phases of song development known Surprisingly, little is known on the structure and for male canaries: subsong, plastic song, and function of this communication. If communication stable song. Although the syllable repertoire was between mates participates in pair coordination established early during song development, both during breeding, partners should show strategies temporal patterning and frequency modulation to maintain information transmission even under of song syllables continued to develop, slowly difficult environmental conditions. We examined consolidating over a period of five months. Song the effect of an increased level of background noi- production was discontinued for three months se on intra-pair acoustic communication around by withholding testosterone, and subsequent re- the nest in the zebra finch. Using a playback of application of testosterone resulted in songs with wind noise inside the nest box, we monitored striking similarity in temporal and spectral organi- how partners adapted their acoustic interactions. zation compared to the initial songs, independent During incubation, zebra finch mates perform of exposure to singing conspecifics. Interestingly, three types of structured call duets at the nest. the song structure consolidated much more rapidly When one mate returns to the nest and meets its after the second testosterone treatment than after incubating partner, a call duet is performed with the first, suggesting that canaries retain some of two possible outcomes: the returning mate either their vocal skills even in periods when no song is relieves its partner and takes its turn incubating produced. It is unclear however, what constitutes the eggs (relief duets), or leaves the nest at the this type of skill memory or where this memory is end of the duet (visit duets). Other call duets are located. We observed that the hormone dihydrotes- performed with one mate inside the nest and the tosterone (DHT), a metabolite of testosterone, did other outside (sentinel duets). Despite the experi- not stimulate naive female canaries to start sin- mental noise constraint, mates still performed call ging. However, applying DHT to birds with previous duets but with several behavioural changes. Mates singing experience did stimulate them to sing. increased the number of visit duets, increased These findings suggest that singing experiences their spatial proximity during sentinel duets, and irreversibly modify the hormonal responsiveness emitted louder, higher pitched and less broadband of brain and/or peripheral pathways that control calls inside the nest box. Relief duets were par- song production, facilitating the rapid retrieval of ticularly affected as partners increased regularity previously acquired vocal motor memories. and precision in their response to each other in noisy conditions. This may sign the importance of relief duets in coordinating partners during the 141 crucial moments of incubation shifts. Our results show that birds use several signal strategies to Villain, Avelyne cope with noise during intra-pair communication Bouchut, C. and strongly suggest a role of vocal interactions Vignal, C. at the nest in coordinating mates’ behaviours. Université Jean Monnet,Saint-Etienne, France These data reveal how environmental noise might Fernandez, M.S.A. interfere with reproductive success by affecting Université Lyon/Saint-Etienne, EPI BEAGLE intra-pair communication.

148 29 The song of wild mice – acoustic divergence in natural populations of mice (Mus sp.) Vokurkova, Jana Ferenc, M. Abstract Mice produce ultrasonic vocalisation Horak, D. (USV) in a range of social contexts. To study a pos- Sedlacek, O. sible evolutionary impact of mate choice based on Charles University, Prague, CZE USV, we analysed the vocalisation behaviour of four populations of wild mice: two populations of the Seasonality in singing activity of bird western house mouse (Mus musculus domesticus), communities along elevational gradient one population of the eastern house mouse (M. m. of Mount Cameroon musculus) and one population of the steppe mouse (M. spicilegus). We recorded and analysed the USV Abstract In relative comparison to the temperate of males and females from all populations emitted zone, environmental seasonality in tropical areas is in different social contexts. For acoustic analysis, much lower. This applies in particular to the tempe- we extracted several temporal and spectral features rature and the length of the daylight. Environmental from the recorded songs and compared them stability throughout the year is reflected in rela- between populations and sexes. Using a multiva- tively stable food availability. It has been generally riate analysis, we found quantitative and structural expected that tropical bird species can breed differences in the USV between sexes and between year-round and timing of breeding is influenced by different populations. These results suggest an less predictable environmental determinants such acoustic divergence of natural populations of wild as precipitation or peaks of availability of food mice. We further investigated whether the USV sources. Consequently, it is suggested that also show a syntactical pattern, analysing if different annual cycles of singing activity may not be clearly elements of vocalisation (syllable types) are arran- defined in the tropics. However, such predictions ged sequentially. Our results support the hypothesis are rarely supported by empirical studies so far. that mice do not emit syllable types randomly but The newly developed bioacoustic approaches using use syntactical structure in their songs. automatic recording of sounds provide probably the easiest way how to obtain the data on a year-round singing activity of bird communities, even on long- 14 term scales. In our study we focuse on seasonality in singing activity of forest birds along elavational Zagkle, Elisavet gradient of Mt. Cameroon. We compare seasonality Greif, S. and diversity patterns between lowland (350m a. s. Kaucic, R. l) and upper montane rain forest (2200m a. s. l). Goerlitz, H.R. Knowledge of the distribution of reproductive effort Max Planck Institute for Ornithology, Seewiesen, of bird species in the tropics are a key requirement Germany for any considerations about the differences in the evolution of life histories along altitudinal and To buzz or not to buzz? Situation-dependent latitudinal gradient. behavioural flexibility in bat biosonar during object approach

153 Abstract Behavioural flexibility is the ability of animals to react to and adaptively adjust their von Merten, Sophie behaviour in response to external stimuli. Bats Adam Mickiewicz University, Poznán, Poland navigate their dark environment with biosonar, a Hoier, S. highly flexible sensory system that probes its envi- Pfeifle, C. ronment by continuously emitting ultrasonic calls, Max Planck Institute for Evolutionary Biology, whose spatio-spectro-temporal call parameters are Plön, Germany constantly adapted to the task at hand. During object approach, however, insectivorous bats show

149 abstracts poster

a stereotypic change in echolocation behaviour, levels of discomfort, only call number per time unit called terminal phase, or buzz, characterized by is commonly estimated, whereas other acoustic decreasing call duration and increasing call repeti- variables are often neglected. In this study, we tion rate. Here, we tested the hypothesis that bats applied the approach based on joint-calls, for esti- flexibly adapt the spectro-temporal call structure mating vocal discomfort of pups of the fat-tailed even during the terminal phase to their task and gerbil Pachyuromys duprasi. Calls of 35 pups from environment. We quantified the echolocation beha- 9 litters were recorded a few times along ontogeny viour of three bat species (long-fingered bat, Myotis from 8 to 40 days. During a recording trial, a pup capaccinii; Schreiber’s bat, Miniopterus schreiber- was initially isolated on the table surface for 4 min sii; greater horseshoe bat, Rhinolophus ferrume- (isolation context), and immediately after this was quinum) in two different contexts, namely when handled for 4-5 min (handling context). We sup- approaching a standard reflective object (prey, wall) posed that the handling context provoked stronger and when approaching a water surface for drinking. discomfort compared to the isolation context, so we Echolocation during drinking has been scarcely expected that variables of ultrasonic calls might investigated, yet might be very different due to reflect the increase of discomfort at transition the unique specular reflection properties of water between these contexts. Only trials where pups surfaces. We recorded the echolocation behaviour produced ultrasonic calls in both contexts (n=73 during object approach in three to six individuals trials) were included into analysis. Joint-calls were from each species (871 approaches in total) and prepared for each context of each recording trial, measured temporal and frequency call parameters by cut of intercall intervals. In total, 146 joint- in ten echolocation sequences per individual and calls contained 3955 ultrasonic calls. From each context. When approaching standard reflective joint-call, we measured 4 power variables: the peak objects, the bats produced typical approach and frequency and 3 quartiles, covering respectively terminal call sequences, with species-specific dif- 25, 50 and 75% of call energy. The values of the ferences. For example, M. schreibersii continuously peak frequency and of all the three quartiles were decreased its pulse interval and then emitted the significantly higher at handling compared to isola- buzz II part of the terminal phase. M. cappaccinii tion context (repeated-measures ANOVA, p<0.001 produced grouped approach phase calls and both for all comparisons). The most prominent differen- a buzz I and II in the terminal phase. In contrast, ces were found between 20 and 32 days of life. most bats completely omitted any terminal phase These results, indicating higher values of power calls when approaching water. Our results show variables with increase of discomfort in ultrasonic that bats can control their biosonar emissions also calls of gerbil pups are similar to data obtained during the last moments of object approach and previously with the same method for audible joint- flexibly adapt them to their sensory needs. calls of fur-farmed red foxes Vulpes vulpes. Thus, power variables might encode discomfort in both ultrasonic and audible calls of mammals. Financial 66 support: The Russian Scientific Foundation, grant No 14-14-00237. Zaytseva, Alexandra Volodin, I.A. Lomonosov Moscow State University, Russian 11 Federation Ilchenko, O.G. Zollinger, Sue Anne Volodina, E.V. Brumm, H. Moscow Zoo, Russian Federation Max Planck Institute for Ornithology, Seewiesen, Germany Estimating discomfort in a fat-tailed gerbil by Nemeth, E. ultrasonic joint-calls BirdLife Österreich, Vienna, Austria

Abstract Ultrasonic vocalizations often attend Great tits stay on pitch even as vocal discomfort states in rodent pups. For estimating amplitudes and background noise levels increase

150 Abstract For communication to occur, a signal Estación Biológica de Doñana-CSIC, Seville, Spain must meet two criteria: it must be transmitted through the environment to reach a receiver, and it What does the hierarchical analysis of quanti- must encode information. A signaler may encounter tative and qualitative song traits tell us about trade-offs in the maintenance of these two criteria signal design and reliability? if the same vocal parameter is involved in both. For example, birds and mammals have been found Abstract Bird song in some passerine species to use a vocal phenomenon called the Lombard is extremely plastic sexual signal, which, beside effect to flexibly adjust the vocal amplitude of their the considerable between-individual variation, is signals in response to fluctuating background noise loaded with a remarkable within-individual variati- levels, maintaining signal transmission through the on. However the theory of reliable signalling only noise. Frequently the rise in amplitude due to the assumes that features of song must consistently Lombard effect is accompanied by a rise in pitch, differ between individuals, while the role of which can also improve signal detection in noise. the within-individual plasticity of song typically However, if signal frequency is an important part of remains neglected. To get more insight on the the signals information content, then shifts in pitch importance of different components of variation in could also hinder effective communication. In hu- the song we studied how the repeatability of song mans, the increase in pitch that often accompanies parameters changes through different hierarchical the Lombard effect is under voluntary control, and levels (syllables, songs and song sequences) at a singer can learn to stay on pitch at any vocal am- different time scales (within-day, between-day and plitude. We tested whether songbirds, like humans, between-year) in the collared flycatcher (Ficedula are able to produce their songs at increasing vocal albicollis). We recorded the songs of 51 males amplitude without also increasing song frequency. twice across different time windows, and charac- We recorded wild-caught adult great tits (Parus ma- terized several variables to describe the physical jor) in quiet, sound attenuated rooms, and during and compositional structures of syllables, songs playback of 70 dB(A) broadband noise. We then and song sequences by spectrographic analysis that compared the amplitude and frequency of each song was subsequently submitted to hierarchical mixed type produced in noise and no-noise conditions for model analyses. For all variables, we found low each individual. We found that great tits did not within-song and within-song-sequence repeatability increase the frequency of their songs even as their that was consistent across different time scales. vocal amplitude increased during Lombard-induced However, at the level of song sequences, we found songs. These results demonstrate that songbirds, high repeatability for repertoire size and entropy, like trained human singers, can decouple the link which declined over longer time scales. We also between amplitude and frequency to maintain found evidence for changes in song composition, signal transmission in noise without compromising as high repeatability of repertoire size was asso- song frequency. Moreover, our observation that the ciated with moderate levels of syllable turnover. birds stayed on pitch despite a rise in amplitude We conclude that not necessarily the features of may suggest that frequency is a critical part of the particular syllables or songs but potentially their signal’s information content. longer sequence may code useful information on individual properties in this species. ■

Zsebök, Sándor Herczeg, G. Blázi, G. Laczi, M. Nagy, G. Török, J. Eötvös Loránd University, Budapest, Hungary Garamszegi, L.Z. No liability for the information provided.

151 abstracts poster

SLC 8x42 WB

One for all

The SLC 8x42 are real multipurpose binoculars that have had their optics and coating system optimized specifically for contrast and focus. This model also stands out for having a large field of view and being extremely rugged thanks to its magnesium housing. The ergonomic design ensures optimum viewing and operating comfort even when observing over long periods.

THE SOLID, MULTIPURPOSE BINOCULARS

Real multipurpose binoculars. The SLC 8x42 and the SLC 10x42 binoculars are sturdy in any weather, day in day out, and also stand out thanks to their particularly high optical quality. You will enjoy these rugged, ergonomic binoculars for many years.

152 list of participants (A – Z)

A Adreani Nicolas Balakrishnan Rohini MPI for Ornithology, Germany Indian Institute of Science, Bangalore, India [email protected] [email protected]

Agranat Ian Balsamo Francesco Wildlife Acoustics, Inc., U.S.A. Dodotronic, Italy [email protected] [email protected]

Almunia Portoles Javier Balsby Thorsten Loro Parque S. A., Tenerife, Spain Aarhus University, Denmark [email protected] [email protected]

Amaya Juan Pablo Barcante Ferreira Luciana CRILAR, Argentina University of Salford, U.K. [email protected] [email protected]

Anichini Marianna Beeck Veronika Humboldt University Berlin, Germany University of Vienna, Austria [email protected] [email protected]

Araya-Salas Marcelo Beme Irina New Mexiko State University, U.S.A. Lomonosov Moscow State University, RUS [email protected] [email protected]

Attia Joel Benedict Lauryn University of St. Etienne, France University of Northern Colorado, U.S.A. [email protected] [email protected]

Aubin Thierry Bertucci Frédéric University Paris Sud, France CRIOBE, Moorea [email protected] [email protected]

Aumann Herbert Bhatt Dinesh Chandra University of Maine, U.S.A. Gurukula Kangri University, India [email protected] [email protected]

B Baier Leonie Blum Christian MPI for Ornithology, Germany MPI for Ornithology, Germany [email protected] [email protected]

Bakhturina Daria Boucaud Ingrid Lomonosov Moscow State University, RUS Université of St. Etienne, France [email protected] [email protected]

Bakker Antje Bowling Daniel MPI for Ornithology, Germany University of Vienna, Austria [email protected] [email protected]

153 list of participants

Braulio Leon-Lopez Cornec Clément University of St. Andrews, U.K. University Paris Sud, France [email protected] [email protected]

Brazao Protazio Joao Marcelo Courvoisier Hélène Federal University of Para, Brazil University Paris Sud, France [email protected] [email protected]

Briefer Elodie D D´Amelio Pietro ETH Zurich, Switzerland MPI for Ornithology, Germany [email protected] [email protected]

Brumm Henrik Dadwal Navjeevan MPI for Ornithology, Germany Gurukula Kangri University, India [email protected] [email protected]

Budka Michal Davis Jordon Adam Mickiewicz University, Poland University of Florida, U.S.A. [email protected] [email protected]

C Campbell James Derégnaucourt Sébastien Leiden University, The Netherlands University Paris West, France [email protected] [email protected]

Cardos Guilherme Sementili Desjonquères Camille Universidade Estadual Paulista Museum national d‘Histoire naturelle, France „Júlio de Mesquita Filho“, Brazil [email protected] [email protected] Diniz Pedro Chabrolles Laura University of Brasilia, Brazil University of St. Etienne, France [email protected] [email protected] Dittrich Falk Chivers Ben MPI for Ornithology, Germany University of Lincoln, U.K. [email protected] [email protected] Donatelli Reginaldo José Codarin Antonio Universidade Estadual Paulista ARPA FVG, Italy „Júlio de Mesquita Filho“, Brazil [email protected] [email protected]

Collier Katie Dorado Correa Adriana University of Zurich, Switzerland MPI for Ornithology, Germany [email protected] [email protected]

Coquereau Laura dos Santos Manuel Institut Universitaire Européen de la Mer, MARE/ISPA, Portugal Brest, France [email protected] [email protected] Dropková Kamila MPI for Ornithology, Germany

154 [email protected] Fitch W. Tecumseh University of Vienna, Austria Dutilleux Guilaume [email protected] Cerema / Dter Est, France [email protected] Florentin Juliette University of Mons, Belgium e Eckenweber Maria [email protected] University of Ulm, Germany [email protected] Fonseca Paulo University of Lisboa, Portugal Economou Chris [email protected] University of the Aegean, Greece [email protected] Forkosh Oren Weizmann Institute of Science, Israel Elemans Coen [email protected] University of Southern Denmark, Denmark [email protected] Frankl-Vilches Carolina MPI for Ornithology, Germany Ey Elodie [email protected] Institute Pasteur, France [email protected] Frommolt Karl-Heinz Museum fuer Naturkunde, Germany f Faragó Tamás [email protected] MTA-ELTE Comp. Ethol. Research Group, HUN [email protected] g Gahr Manfred MPI for Ornithology, Germany Favaro Livio [email protected] University of Turin, Italy [email protected] Garcia Maxime University of Vienna, Austria Feng Ni [email protected] Cornell University, U.S.A. [email protected] Geberzahn Nicole Université Paris Ouest La Defense, France Ferhat Allain-Thibeault [email protected] Institut Pasteur, France [email protected] Gessinger Gloria University of Ulm, Germany Fernandez Marie [email protected] Université of St. Etienne, France [email protected] Gingras Bruno University of Innsbruck, Austria Firzlaff Uwe [email protected] TU München, Germany [email protected] Gill Lisa MPI for Ornithology, Germany Fischer Julia [email protected] DPZ Göttingen, Germany [email protected] Goerlitz Holger MPI for Ornithology, Germany [email protected]

155 list of participants

Gogala Matija Hoebel Gerlinde Slovenian Academy of Sciences and Arts, SLO University of Wisconsin-Milwaukee, U.S.A. [email protected] [email protected]

Gomes Dylan Hoeschele Marisa Smithsonian Tropical Research Institute, Panama University of Vienna, Austria [email protected] [email protected]

Goncharova Maria Hoier Svenja Lomonosov Moscow State University, RUS MPI for Evolutionary Biology, Germany [email protected] [email protected]

Goutte Sandra Hoiss Nicole University Estadual de Campinas, Brazil MPI for Ornithology, Germany [email protected] [email protected]

Goyes Vallejos Johana Honaiser Rostirolla Tomás University of Connecticut, U.S.A. University Federal de Santa Catarina, Brazil [email protected] [email protected]

Goymann Wolfgang Hrouzková Ema MPI for Ornithology, Germany University of South Bohemia, CZE [email protected] [email protected]

Gridi-Papp Marcos Hügel Theresa University of the Pacific, U.S.A. MPI for Ornithology, Germany [email protected] [email protected]

h Hage Steffen i Ilina Irina University of Tuebingen, Germany Lomonosov Moscow State University, RUS [email protected] [email protected]

Hall Michelle L. j Jaiswara Ranjana University of Melbourne, Australia Indian Institute of Science Education & [email protected] Research, India [email protected] Hallerberg Sarah MPI for Dynamics and Self-Organization, Germany Jakubowska Aleksandra [email protected] Adam Mickiewicz University, Poland [email protected] Hammerschmidt Kurt DPZ Göttingen, Germany Janik Vincent [email protected] University of St. Andrews, U.K. [email protected] Herbst Christian Palacký University Olomouc, CZE Jansen Martin [email protected] Senckenberg Research Institute Natural History Museum, Germany Hödl Walter [email protected] University of Vienna, Austria [email protected]

156 Jayasena Nilmini [email protected] University of Peradeniya, Sri Lanka [email protected] Kruger Donnavan North-West University, South Africa Jonsson Thorin [email protected] University of Lincoln, U.K. [email protected] Krüger Frauke Echolot GbR, Germany k Kerchev Ivan [email protected] Inst. of Monitoring of Climatic & Eco. Systems, RUS l Labra Lillo Antonieta [email protected] University of Chile, Chile [email protected] King Steven Western Kentucky University, U.S.A. Lachlan Robert [email protected] Queen Mary University of London, U.K. [email protected] Kipper Silke TU München, Germany Ladich Friedrich [email protected] University of Vienna, Austria [email protected] Klenova Anna Lomonosov Moscow State University, RUS Lampe Helene M. [email protected] University of Oslo, Norway [email protected] Klenschi Elizabeth University of Strathclyde, U.K. Lankau Hedwig E. [email protected] University of Alberta, Canada [email protected] Klimsova Vera Czech University of Life Sciences, CZE Larsen Ole N. [email protected] University of Southern Denmark, Denmark [email protected] Ko Maggie MPI for Ornithology, Germany Laschober Monika [email protected] University of Vienna, Austria [email protected] Koselj Klemen MPI for Ornithology, Germany Laverde-R. Oscar [email protected] University de los Andes, Colombia [email protected] Krasheninnikova Anastasia MPI for Ornithology, Germany Le Maguer Lucille [email protected] University Paris West, France [email protected] Kriengwatana Buddhamas Leiden University, The Netherlands Leitao Albertine [email protected] MPI for Ornithology, Germany [email protected] Kriesell Hannah Joy Centre Scientifique de Monaco, France

157 list of participants

Leitner Stefan [email protected] MPI for Ornithology, Germany [email protected] Martin Juliette University of Edinburgh, U.K. Lewanzik Daniel [email protected] MPI for Ornithology, Germany [email protected] Martínez Alvarado Dariel Andrés University of Antioquia, Colombia Linke Simon [email protected] Griffith University, Australia [email protected] Martínez-Serena Abigail National Autonomous University of Mexiko, Linossier Juliette Mexiko City, Mexiko University Paris Sud, France [email protected] [email protected] Martins Joao Llusia Diego Fiat Sonitus, Germany Museum National D´Histoire Naturelle, France [email protected] [email protected] Mathevon Nicolas Luis Ana Rita University of St. Etienne, France MARE, ISPA - Instituto Universitario, Portugal [email protected] [email protected] Matrosova Vera Lossent Julie Engelhardt Institute of Molecular Biology RAS, RUS France Energies Marines & IUEM Gipsa-lab, [email protected] Brest, France [email protected] McLeod Logan University of Alberta, Canada Luo Jinhong [email protected] MPI for Ornithology, Germany [email protected] Meigel Felix MPI for Dynamics and Self-Organization, Germany m Ma Shouwen [email protected] MPI for Ornithology, Germany [email protected] Mélotte Geoffrey University of Liège, Belgium Maciel Israel [email protected] Federal Rural University of Rio de Janeiro, Brazil [email protected] Miller Jacqueline Royal Ontario Museum, Canada Maigrot Anne-Laure [email protected] ETH Zurich, Switzerland [email protected] Mishima Yuka Tokyo University of Marine Science and Marchenko Alexandra Technology, Japan Lomonosov Moscow State University, R.F. [email protected] [email protected] Miyammoto Yoshinori Marfull Reinaldo Tokyo University of Marine Science and Rijkuniversiteit Groningen, The Netherlands Technology, Japan

158 [email protected] Oswald Julie Bio-Waves, Inc., Encinitas, U.S.A. Mizuguchi Daisuke [email protected] Kyoto University, Japan [email protected] Ota Nao Hokkaido University, Japan Montealegre-Zapata Fernando [email protected] University of Lincoln, U.K. [email protected] p P. Amorim M. Clara MARE, ISPA, Portugal Montes-Medina Adolfo Christian [email protected] University Nacional Autónoma de México, México [email protected] Padilla de la Torre Monica ETH Zurich, Switzerland Moskalenko Viktoria [email protected] Lomonosov Moscow State University, RUS [email protected] Palhares Cíntia University Federal de Vicosa, Brazil Motes Rodrigo Alba [email protected] University of Oslo, Norway [email protected] Parks Susan Syracuse University, U.S.A. Mouterde Solveig C. [email protected] University of California Berkeley, U.S.A [email protected] Pelicella Ivano Dodotronic, Italy Mundry Roger [email protected] MPI for Evolutionary Anthropology, Leipzig, Germany Penna Mario [email protected] University of Chile, Chile [email protected] n Nemeth Erwin University of Vienna, Austria Pereira Chaves Patricia Alexandra [email protected] ISPA, Portugal, University of Saint Joseph, Macao [email protected] Njoroge Anastasia Wanjiru International Center of Insect Physiology and Petrusková Tereza Ecology, Kenya Charles University, Prague, CZE [email protected] [email protected]

Noriega Romero Vargas Maria Florencia Pfeifle Christine MPI for Dymanics and Self-Organization, Germany MPI for Evolutionary Biology, Germany [email protected] [email protected]

o Odom Karan Pieretti Nadia University of Maryland Baltimore County, U.S.A. University of Urbino, Italy [email protected] [email protected]

Osiejuk Tomasz Popova Daria Adam Mickiewicz University, Poland Lomonosov Moscow State University, RUS [email protected] [email protected]

159 list of participants

Porteš Michal [email protected] Charles University, Prague, CZE [email protected] Robotka Hermina MPI for Ornithology, Germany Prior Nora [email protected] University of St. Etienne, France [email protected] Roch Marie A. San Diego State University, U.S.A. Ptacek Ladislav [email protected] University of South Bohemia, CZE [email protected] Rocha Diales Mariana MPI for Ornithology, Germany q Quirós Guerrero Esmeralda [email protected] University of St. Andrews, U.K. [email protected] Rodriguez Rafael University of Wisconsin-Milwaukee, U.S.A. Quispe Rene [email protected] MPI for Ornithology, Germany [email protected] Rosa Fernando University of La Laguna, Spain r Ramenda Claudia [email protected] MPI for Ornithology, Germany [email protected] Rowe Karen Museum Victoria, Australia Ranft Richard [email protected] International Association of Sound and Audiovisual Archives (IASA), U.K. Rusfidra Rusfidra [email protected] Andalas University, Indonesia [email protected] Ravignani Andrea Vrije Universiteit Brussel, Belgium Ryan Mike [email protected] University of Texas and Smithsonian Tropical Research Institute, Panama Riebel Katharina [email protected] Leiden University, The Netherlands [email protected] Rybak Fanny University Paris Sud, France Ringler Max [email protected] University of Vienna, Austria [email protected] s Salinas-Melgoza Alejandro University Autonoma de Tlaxcala, México Rivera Caceres Karla [email protected] University of Miami, U.S.A. [email protected] Salvin Pauline University Paris West, France Robert-Nicoud Daniel [email protected] University of Bristol, U.K. [email protected] Sanluis Jose Carlos University of La Laguna, Spain Robillard Tony [email protected] Museum national d‘Histoire naturelle, France

160 Sarria Fabio [email protected] University of Lincoln, U.K. [email protected] Soma Masayo Hokkaido University, Japan Scharff Constance [email protected] Free University Berlin, Germany [email protected] Souriau Abel Charles University, Prague, CZE Schmidt Arne [email protected] University Graz, Austria [email protected] Sprau Philipp Ludwig Maximilians University, Munich, Germany Schneiderová Irena [email protected] University of Life Sciences in Prague, CZE [email protected] Stepien Anna Ewa ETH and University of Zurich, Switzerland Schoeppler Diana [email protected] University of Tübingen, Germany [email protected] Stowell Dan Queen Mary University of London, U.K. Schöneich Stefan [email protected] University of Cambridge, U.K. [email protected] Sueur Jerôme Museum National D´Histoire Naturelle, France Sellar Patrick [email protected] President of IBAC, U.K. [email protected] Szymanski Pawel Adam Mickiewicz University, Poland Seltmann Susanne [email protected] MPI for Ornithology, Germany [email protected] t Tanttu Juha T. Tampere University of Technology, Finland Sibiryakova Olga [email protected] Lomonosov Moscow State University, RUS [email protected] ten Cate Carel Leiden University, The Netherlands Sieving Kathryn [email protected] University of Florida, U.S.A. [email protected] ter Maat Andries MPI for Ornithology, Germany Simon Ralph [email protected] Friedrich-Alexander-University, Erlangen- Nuremberg, Germany Thiebault Andréa [email protected] Nelson Mandela Metropolitan University, South Africa Slater Peter [email protected] University of St Andrews, U.K. [email protected] Thomsen Heidi University of Copenhagen, Denmark Slocombe Katie [email protected] University of York, U.K.

161 list of participants

Tipp Cheryl von Merten Sophie The British Library, U.K. Adam Mickiewicz University, Poland [email protected] [email protected]

Trost Lisa w Wadewitz Philip MPI for Ornithology, Germany Cognitive Ethology Laboratory, Germany [email protected] [email protected]

u Ueno Adriana Sicuto de Oliveira Werner Diana University of Sao Paulo, Brazil MPI for Ornithology, Germany [email protected] [email protected]

Ulloa Juan Wojas Lucyna Museum National D´Histoire Naturelle, France Adam Mickiewicz University, Poland [email protected] [email protected]

v van Boheemen Sharina Jarmila Zagkle Elisavet Charles University, Prague, CZE MPI for Ornithology, Germany [email protected] [email protected]

Varkevisser Judith z Zaytseva Alexandra Leiden Unversity, The Netherlands Lomonosov Moscow State University, RUS [email protected] [email protected]

Vellema Michiel Zielinski Piotr University of Southern Denmark, Denmark University of Wroclaw, Poland [email protected] [email protected]

Vernes Sonja Zollinger Sue Anne MPI for Psycholinguistics, The Netherlands MPI for Ornithology, Germany [email protected] [email protected]

Vignal Clementine Zsebök Sándor University of St. Etienne, France Eoetvoes Lorànd University, Hungary [email protected] [email protected]

Villain Avelyne University of St. Etienne, France [email protected]

Villavicencio Camila MPI for Ornithology, Germany [email protected]

Vokurkova Jana Charles University, Prague, CZE [email protected] von Bayern Auguste M.P. MPI for Ornithology, Germany [email protected]

162