Using microphone arrays to explore communication in

10/14/2008

1

Using microphone arrays to explore communication strategies in songbirdsTalk Outline:1.Communication networks in

1

networks in temperate chickadees

2.Vocal duetting in neotropical wrens

Daniel MennillUniversity of Windsor Windsor, Ontario, Canada

Signaller Receiver

Animal Communication

• “Communication involves two individuals, a signaller and a receiver”

2

Bradbury& Vehrencamp’sAnimal Communication

Decision and response

Information

Environment

Communication Networks

• Communication Network: More than two individuals signalling and/or receiving simultaneously

3

McGregor’s Animal Communication Networks


• Eavesdropping: Gaining information by listening to a signalling interaction between others without being directly involved in that interaction

4


• Audience Effects: Changes in signaller behaviourthat arise from the presence of receivers

5

Acoustic Location System

6

10/14/2008

2

7

Mic #1

Mic #2

Acoustic Location System8

Mic #3

Mic #4

x=43.2, y=17.6

x=19.1, y=20.4

Acoustic Location System9

x=43.2, y=17.6

Black‐capped chickadee (Poecile atricapillus)10

Black‐capped chickadee (Poecile atricapillus)11

6

5

cy (kHz)

Variation in song timing

Territorial male chickadees engage in aggressive countersinging interactions with rivals

male 1 male 1 male 1male 2 male 2 male 2

12

0 2 4 6 8 10 12

4

3

Time (s)

Freq

uenc

Variation in song frequency

g g

Mennill & Otter (2007) Ecology and Behavior of Chickadees and Titmice

10/14/2008

3

Previous playback experiments show…13

Mennill & Ratcliffe (2004) Behaviour

Females eavesdrop on male‐male contests

Males eavesdrop on male‐male contests

…but has not been examined in a natural context.

Mennill et al. (2002) Science

Goals

Part 1: Use an Acoustic Location System to examine countersinging and movement behaviour at a network scale

l k

14

Part 2: Use multi‐speaker playback to examine neighbourhood‐level responses to aggressive interactions

15

09

06 07

1515

16

11

0909

10 14

13

BASE

12

15

010101

02

0403

08

05

1

2

3

4

5

67

98

We analyzed 100 countersinging contests in

10 neighbourhoods

10 20 30 400Time (s)

1112

13

14

1516

910

1

3

2

4

8

7

65

17

50 m

9

12

10

11

1413

16 15

Males moved apart in 46% of contests

Males approached in

18

50 m

No movement in 3% of contests

pp53% of contests

10/14/2008

4

Results of multi‐channel recordings:

• Contests in which males approached each other contained more matching exchanges

matched

son

gs

3

4

6*

5

19

Malesmove apart

Malesapproacheach other

Num

bero

f m

0

1

2

Fitzsimmons, Foote, Ratcliffe, Mennill (2008) Animal Behaviour

Multi‐speaker Playback20

Playback simulates a song contest between

two rival males

3Multi‐speaker Playback21

9

Multi‐speaker Playback22

50 m

Results of multi‐speaker playback:

• Song output from all males in neighbourhoodwas significantly higher after aggressive playback

• Strong evidence for 15

20

25

odsong

outpu

tngs/min)

*

23

gsocial eavesdropping

0

5

10

Aggressive Submissive

Playback treatment

Neighbo

urho (son

Fitzsimmons, Foote, Ratcliffe, Mennill (2008) Animal Behaviour

Conclusions from chickadee array recordings

• Song matching appears to function as a conventional signal of aggression; birds approach matching rivals

• Territorial males eavesdrop on

24

Territorial males eavesdrop on interactions that take place beyond territory boundaries

• Dyadic interactions have a ripple effect throughout neighbourhoods

10/14/2008

5

Further reading on chickadee networks• Fitzsimmons, Foote, Ratcliffe, Mennill (2008) Frequency matching,overlapping and movement behaviour in diurnal countersinginginteractions of black‐capped chickadees. Animal Behaviour 75:1913‐1920.

• Fitzsimmons, Foote, Ratcliffe, Mennill (2008) Eavesdropping andcommunication networks revealed through playback and an acousticlocation system. Behavioral Ecology 19:824‐829.

25

• Foote, Fitzsimmons, Mennill,Ratcliffe (2008) Male chickadees

t h i hb i t ti l tmatch neighbours interactively atdawn: support for the socialdynamics hypothesis. BehavioralEcology (online first).

• Foote, Fitzsimmons, Mennill,Ratcliffe (2008) Tied to the nest:Male black‐capped chickadeesdecrease dawn chorus movementbehaviour when their mate isfertile. Animal Behaviour 76:1227‐1233.

Rufous‐and‐white Wren (Thryothorus rulfabus)

26

Bird songs are complex vocalizations produced by males during the breeding season. They play an important role in territory defense and mate attraction.

27

Alcock’s Animal Behavior Catchpole & Slater’s Bird Song

Duets: Complex, coordinated vocalizations produced jointly by members of a mated pair

28

Rufous‐and‐white Wren (Thryothorus rulfabus)

29

Stu

Resident neotropical songbirds

30

10/14/2008

6

31 32

33 34

35

3

4

cy (kHz)

Rufous-and-white Wren Song36

0

1

2

0 21

Freq

uenc

Time (s)

0.5 1.5

Mennill & Vehrencamp (2005) Auk

10/14/2008

7

0

1

2

3

4

Rufous-and-white Wren Duets

malefemale

0 421 30

1

2

3

5

4

37

0 421 3

malefemale

Rufous-and-white Wren Duets

0 421 30

1

2

3

5

4

38

0 421 3

• Recorded 19 colour-banded pairs of rufous-and-white wrens with an eight-microphone acoustic location system during the early breeding

Spatial Analysis of Duetting39

the early breeding season

Mennill & Vehrencamp (2008) Current Biology

20

01

50

25

0 Mic 1

Mic 2

Mic 3

Mic 4

01

02

0403

05

40

01

00

0 100

50

50 150Meters

Mic 5

Mic 6

Mic 7

Mic 8

06

07

08

• Tremendous variation in the distance between male and female duet partners

50

60

130

140

150

160

Number of duets

Minimum: 0.39 m

Average: 19.2 m

41

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150

Distance between duet partners (n = 525)

0

10

20

30

40

50of duets recorded

Maximum: 144.3 m


• Partners were closer together during duets where female sang first, male sang second

Distance betweenduetting 10

15

20 *

42

Paired t-testt=2.1, p=0.05, n=16

Sex of duet creator

duetting partners (m)

Male Female0

5

10


10/14/2008

8

20

01

50

25

0

• Duets were sometimes given in bouts where pairs sang several duets in a row

• During a significant majority of these bouts (44 out of 64) the male and female

43

01

00

0 100

50

50 150Meters

moved closer together between subsequent duets

(Binomial test: p=0.01)


Polo!Marco?

44

20

03

00

25

0

• Duets were given throughout territories, not concentrated at territory boundaries

45

Meters0 10050 150

01

00

50

15

0

250200 300 350 400 450

A A B

• Duets were given throughout territories, not concentrated at territory boundaries

Distance

25

30

46

ANOVAF2,54=7.8p=0.001

to territory edge (m)

15

20

DuettingMales

DuettingFemales

RandomPoints


• Duets were often given near nests2

00

30

02

50

N

Meters

47

Meters0 10050 150

01

00

50

15

0

250200 300 350 400 450

N

N

• Duets were often given near nests

A A B

Distance 50

70

60

48

ANOVAF2,54=7.8p=0.001

Distance to nest (m)

30

40

50

DuettingMales

DuettingFemales

RandomPoints


10/14/2008

9

• Duetting behaviour is consistent with the acoustic contact hypothesis– Birds duet when they are physically separated,

often by great distances– Duets are not focused at territory boundaries

Conclusions from wren array recordings49

– Birds approach each other during duet bouts

Further reading on wren duets• Mennill, Vehrencamp (2008) Context‐dependent functions of avian duetsrevealed through microphone array recordings and multi‐speaker playback.Current Biology 18:1314‐1319

• Topp, Mennill (2008) Seasonal variation in the duetting of rufous‐and‐white wrens. Behavioral Ecology & Sociobiology 62:1107‐1117.

• Mennill (2006) Aggressive responses of male and female rufous‐and‐whitewrens to stereo duet playback. Animal Behaviour 17:219‐226.

• Mennill, Burt, Fristrup, Vehrencamp (2006) Accuracy of an acoustic

50

location system for monitoring the position of duetting tropical songbirds.Journal of the Acoustical Society of America 119:2832‐2839.

Using microphone arrays to explore communication strategies in songbirdsConclusions:• Microphone array recordings enhance our understanding of animal communication

• They facilitate monitoring at a

51

They facilitate monitoring at a neighbourhood scale

• They facilitate monitoring behaviour in visually‐obstructed environments

Daniel MennillUniversity of Windsor Windsor, Ontario, Canada

AcknowledgementsCollaborators: Stéphanie Doucet (University of Windsor),

Laurene Ratcliffe (Queen’s University), Sandy Vehrencamp (Cornell University)

Grad Students: Nicole Barker, David Bradley, Sandra Gallo, Julie Koloff, Van La, Karan Odom, Anneka Osmun, Lauren Fitzsimmons, Kyle Swiston, Steph Topp, Sarah Tremain, & Sandra Valderrama

52

The Mennill Sound Analysis Lab 2008

Technical Assistance: John Burt (www.syrinxpc.com), the Bioacoustic Research Program at Cornell’s Lab of Ornithology

Field Assistance: J. Baldock, R. Bull, S Doucet, R. Jamieson, S. Lippold, A. McKellar, D. Potvin, K. Winger (chickadee research), and V. Connolly, D. Moseley, S. Doucet (wren research)

Logistic Support: Queen’s University Biology Station (chickadee research), Santa Rosa National Park (wren resesarch)

Acknowledgements53

NSERC of Canada

University of Windsor

National Geographic Society

Canada’s Foundation for Innovation

Ontario’s Early Researcher Award Program

Association of Universities and Colleges of Canada

Funding:

Documents

Using microphone arrays to explore communication in