Lessons from vervets and macaques

1
Lessons from vervets and macaques

• MSc ACSB module 2006/07
• AC Session 3

2
Vervet monkey alarm calls

• Cheney Seyfarth "How monkeys see the world
  Vervet monkeys (tree ground living monkey,
  Africa)
• Predators Leopard, Monkey-eating Eagle, Python,
  baboons, etc, and 3 calls
• Eagle gets a Cough,
• Snake gets a Chutter
• Leopard gets a Bark

3
Vervet alarm-call responses

• Behave appropriately when they hear one of these
  calls (run down from treetops / walk carefully /
  run up into trees)
• Do they know what messages the calls carry? (e.g
  There is an Eagle, / a Snake, / a Leopard)
• Film response to plausible taped call no real
  caller whose behaviour might give hearers a clue
  to the right response
• 3 responses are given in appropriate contexts
  using just the information in the call itself,
  showing the monkeys are responding to the
  acoustic signals, not just to callers concurrent
  behaviour

4
Vervet calls on www

• http//www.wjh.harvard.edu/mnkylab/media/vervetca
  lls.html

5
Vervet alarm calls provide information

• about environment, not about signallers
  motivational state (e.g. different levels of
  fear) equivalent of words for the 3 predators?
• Nearer to language than the displays considered
  last time is call-use learned c.f. word-use?
• Are monkey vocalisations acquired/learned?
• Not in squirrel monkey - innate.
• Perhaps in X-fostered Japanese rhesus macaque
  food-coos
• Learning probably contributes in vervets

6
Ontogeny of vervet predator calls

• Vervet infants give alarms to appropriate class
  of stimuli, but too wide a spread of targets
  within each class
• Leopard to many large ground animals
• Eagle to birds of all sorts
• Snake to sticks and other long thin objects
• As grow up, they focus alarms down on the real
  predators, the class-members that spell danger

7
Vervet call development (2)
8
Vervet call development (3)
Probability of adult alarm call after infant has
given eagle alarm call high only to Martial
eagle 1 other sp.
Infants give fewer wrong and more adult-like
responses to alarm call playback as they grow
older
9
Vervet call development (4)

• Narrowing down of call-triggers may depend on
  response from adults take up and repeat alarm to
  real hazards, ignore it to a harmless stimulus
• Responses to alarm calls not fully adult
• Initially respond after looking at an adult which
  has started to respond
• More often show adult-like response when near
  mother than when mother has wandered away
• Vervets use wrrr-call to indicate threat from
  another group experience shapes its correct use
  over 1st two years of life (earlier if more
  frequently in contact with other groups)

10
Involuntary or voluntary?

• High ranking vervets call more often, and are
  more often the first to call but dont scan for
  predators more often. So subordinates must also
  scan and detect predators, but omit call
• Females call more readily if kin present
• Captive males call more when female companion(s)
  than when companion is male
• Never call eagle when should say leopard

11
Is this alarm call system unique?

• Calls provide info about dangers, not level of
  fear
• Vervet monkey grunts (Cheney Seyfarth)
• Can't be distinguished by ear by humans
• 4 types DomgtSub, SubgtDom, Move Into Open, see
  Another Group
• Difference in response to taped grunts indicates
  monkeys can separate them, appropriate
  information conveyed, e.g.
• MIO listener looks towards loudspeaker
• AG looks away towards where loudspeaker points

12
Vervet grunts
16 acoustic parameters from one female 82
correct classification of her calls and others'
calls
Spectrographically distinct but cannot be
distinguished by ear by humans
13
Rhesus macaque screams

• Rhesus pigtail macaque screams studied by
  Gouzoules
• Rhesus has 5 types of scream code for
• Rank of the opponent
• Whether a relative (safer) or non-kin (risky)
• Whether or not any physical contact
• Pigtail has 4 types of scream

14
Rhesus screams (2)

• High rank, contact
• Low rank, no contact
• Relative, or
• High rank, no contact
• Relative
• High rank, no contact

15
Interim conclusions

• In Vervet alarm call system, information is
  encoded in specific calls coding is partly
  pre-wired but is refined by experience
• Several other call systems which communicate
  environmental information
• Kitui used the leopard call (sans leopard) to
  halt a fight that his troop were losing but
  then walked across ground repeating the call,
  which made it plain to humans that there was no
  real danger

16
What information is in a call?

• Do primates lump-together different calls that
  refer to the same thing?
• Habituation
• Do primates learn to ignore specific calls, or to
  distrust a mental state (eg fear) in the caller?
• Are changes in risk tied to a particular threat?
• Do callers aim to inform, or to trigger a
  specific response?

17
Rhesus food calls

• 4 food calls
• Warble, Harmonic Arch (Good food)
• Coos, Grunts (low-quality food)
• S1 and S2 initially elicit orientation
• Habituate S1, then test S2, where S2 may be a
  different signal for same quality of food, or a
  different signal for different-quality food

18
Hausers results

• Hauser, 1998, Anim Behav 55, 1647-1658
• Habituate response to one HQ food call
• Eliminates response to other HQ call
• Leaves intact response to LQ calls
• Habituate response to one LQ food call
• Leaves intact responses to HQ food calls

19
Cheney Seyfarth - Vervet

• Inter-group calls
• Wrr (low arousal just spotted) Chutter (high
  arousal scrap going on or likely)
• Habituation paradigm
• Test Chutter habituate Wrr (same ) re-test
  Chutter
• Decreased response if all 3 stimuli for same
  hazard, from same , not if different monkeys
  calls used
• Implications
• know that A and B represent the same threat,
  conclude that this has become unreliable about
  other groups
• No decrement if calls represent different threats

20
Superb staring alarms

• Aerial and ground predator alarms
• Test starling alarms habituate vervet eagle
  alarm test starling alarms again
• Decreased response to starling eagle alarm
• No decrement for starling ground predator call
• Have learned to be sceptical about (any) warnings
  about aerial predators, not just habituated to
  vervet coughs specifically

21
What does caller aim to achieve?

• In Cameroon, vervets attacked by feral dogs
• Dogs trigger leopard alarm, troop runs into
  trees
• Elsewhere, hunted by men with dogs guns
• Leopard alarm would attract attention and a shot
• So dogs trigger a quiet call that allows troop to
  flee silently
• Monkey link signals to the action that the signal
  needs to achieve

22
Limits on understanding

• Kitui used a leopard-call to stop a fight
  (deception?), but then walked across ground
  showing that there was probably no leopard none
  of the hearers noticed the incongruity
• Vervets also cant recognise other indirect cues
  to danger e.g., snake track on ground, or
  antelope carcass stored in tree (which signals
  that a leopard is nearby)

23
References session 6

• Cheney Seyfarth (1992) Behavioral and Brain
  Sciences, 15, 135-147 (commentary 147-182)
• Cheney Seyfarth (1990) How monkeys see the
  world, Ch. 3-6.
• Seyfarth Cheney (2003) Meaning and emotion in
  animal vocalizations. Annals of the New York
  Academy of Sciences, 1000, 32-55.
• Hauser (1997) The evolution of communication.
  Ch. 5, 7