On the evolution of acquiescence

1
On the evolution of acquiescence

• or why consent to social oppressionT.
  Wenseleers, University of Sheffield

2
Conflict cooperation

• conflicts occur in most societies
• how are conflicts resolved?
• Kin selection
• Recriprocity (tit-for-tat)
• Reputation (humans)
• Policing
• Punishment

NON-COERCIVE
COERCIVE
3
Policing in social insects

• occurs in three contexts
• conflict over male parentage(eating of worker
  laid eggs by queen or workers)
• conflict over caste fate in societies with morph.
  castes (caste fate policing by food control)
• conflict over queen replacement(prevention of
  queen overthrow)

4
Hypothesis

• policing reduces benefit of attempted
  selfishness
• could this favour individuals not even to attempt
  to behave selfishly ?
• idea of self policing or acquiescence to
  agree without protest to a system of policing

5
Aims

• formally investigate the likelihood for the
  evolution of acquiescence
• models specific for social insects, where
  several types of policing occur

6
Policing in social insects

• occurs in three contexts
• conflict over male parentage(eating of worker
  laid eggs by queen or workers)
• conflict over caste fate in societies with morph.
  castes (caste fate policing by food control)
• conflict over queen replacement(prevention of
  queen overthrow)

7
Policing in the honey bee
Workers lay male eggsrown sons gt rqueens sons
but these are eaten byother workers rother
workers sons lt rqueens sons worker policing
8
Caste fate policing in the honey bee
Females would generally benefit from becoming a
queen rather than a worker caste fate conflict
But workers only provision a few royal cells
with sufficient food to allow queen
developmentFemales in other cells are forced to
develop as workers
Bourke Ratnieks BES 1999, Wenseleers, Ratnieks
Billen J. Evol. Biol. in press
9
Conflict over queen replacement in Dinoponera ants
Policed by fellow workerspretender
punishment
Selfish antattempts to overthrow current breeder
Monnin Ratnieks BES 2001, Monnin et al. Nature
2002
10
Conflict over queen replacement
Normal queen lifespan (100)
NonproductiveLag period length L
Early queen replacement

• Benefit of early replacementno risk of queen
  being replaced by sister at later time
• Cost of early replacementdrop in average colony
  productivity, since it takes time for new queen
  to reestablish
• Selects for workers to police early replacers

11
Conflict over queen replacement
COLLECTIVE OPTIMUM ? PREVENT EARLY OVERTHROW
100
Singlemating
Doublemating
80
60
ESS age of replacement
50
INDIVIDUALOPTIMUM
L(1Rm)/(1-Rf) 0.1x1.25/0.25
40
All males queens sons
20
All males workers sons
0
10
20
30
40
50
Productivity cost of early replacement L
12
Imprisonment of virgin queens to prevent early
overthrow
In Trigonini stingless bees, e.g. Plebeia
remotaprevents virgin queens from selfishly
overthrowing mother queen - preemptive policing
Workers force virgin queens to build their own
prison and queens that escape from their prison
are executed
13
Policing and acquiescence
?
Conflict over It is selfish Policing via Acquiescence
Male parentage For a worker to lay eggs Cannibalization of w/laid eggs Worker sterility
Caste fate To become a queen Food control Accept worker caste fate
Queen replacement To replace current queen Aggression or imprisonment Not attempt to replace queen
14
Conflict over male parentage

• If workers reproduce, but their eggs are policed,
  could it favour workers not to lay eggs in the
  first place
• i.e. could policing secondarily select for worker
  sterility?
• NEW THEORY YES

15
Worker sterility as acquiescence

• single worker selected to lay when
• shows that benefit of worker laying reduces as
  the efficiency of policing P goes up
• policing can potentially make it unprofitable for
  any worker to lay

16
Conflict over male parentage

• ESS fraction of laying workers
• Function of colony size (n), relative queen
  fecundity (q), sister-sister relatedness (Rf) and
  probability that worker-laid eggs are policed (P)

17
Worker sterility as acquiescence

• when there is no policing (P0) and for large
  n,ESS is for colony to have ratio of
  Rson-Rnephew laying workers to
  RnephewRsister sterile workers i.e. z
  (1-Rsister)/(1Rsister) 14 under single
  mating
• but fewer workers selected to lay when there is a
  high chance that their eggs are policed

18
Worker sterility as acquiescence
honey bee, polyandrous queen
0.5
0.4
0.3
ESS fraction of laying workers
0.2
single monandrous queen
0.1
0.2
0.4
0.6
0.8
1
Efficiency of policing (P)
19
Empirical data honey bees and vespine wasps
Dolichovespula saxonica
Dolichovespula sylvestris
5.00
Dolichovespula media
Apis cerana
Vespa crabro
of laying workers
0.50
Vespula vulgaris
0.05
Apis mellifera
? -0.8, p lt 0.01
84
86
88
90
92
94
96
98
100
efficiency of policing( of worker eggs eaten)
20
Empirical data honey bee

• Only 1 in 10,000 workers lay eggs
• Models predictions
• n 35,000 workers
• q 25 (counting only queens male eggs)
• Rf 0.3 (multiple mating)
• If policing were absent 54 of all workers
  should lay
• But as a result of highly effective policing
  there is less to gain from layinge.g. with P
  0.99 only 1 in 10,000 laying workers
  predicted
• Fits observed data very well

21
Theoretically predicted values
14
? 0.8, p lt 0.01
Dolichovespula saxonica
12
10
8
Dolichovespula sylvestris
observed of laying workers
6
Dolichovespula media
4
Apis cerana
2
Vespa crabro
Vespula vulgaris
0
Apis mellifera
0
2
4
6
8
10
12
14
16
theoretically predicted of laying workers
22
Caste fate policing

• Given that there are individuals benefits of
  developing as as queen, why do females accept
  their worker caste fate?
• Even when females are only given little food they
  could potentially evade their caste fate by
  becoming a dwarf queen
• Occurs in stingless bees and some ants

23
Stingless beePlebeia remota
dwarf queen
normal queen
2 mm
24
Acquiescence to caste fate

• dwarf queens may have lower fecundity than normal
  queens
• producing 100 of sisters offspring (r 0.375)
  is as good as producing 75 own offspring (r
  0.5) !
• favours acquiescence

25
Favouritism for large queens

• Schwarziana quadripunctata
• 89 of all queens produced are dwarf queens
• Yet only 31 of all colonies are headed by dwarf
  queens
• Seems to be the result of favouritismworkers
  selectively kill dwarf queens and prefer to
  swarm with normal queens
• This also reduces the benefit of becoming a dwarf
  queen

26
single mating, all males queens sons
0.2
0.15
single mating, all males workers sons
ESS prob. of becoming dwarf queen
0.1
Schwarziana quadripunctata
0.05
0.2
0.4
0.6
0.8
1
Worker preference for normal queens (P)
Parameters 1 in a 100 cells are normal sized
queen cells
27
Resolution of conflict over queen replacement

• Dinoponera antsworkers are selected to
  overthrow mother at 50 her normal lifespan, but
  attempted overthrow is punished
• Does punishment of pretenders reduce the
  conflict?
• YES, it reduces the benefit of becoming a
  dominant resisting worker

28
0.2
0.15
ESS fraction of resisting workers
0.1
0.05
All males gamergates sons
All males workers sons
0.2
0.4
0.6
0.8
1
Effectiveness of pretender punishment (P)
29
Conclusion

• Varied mechanisms of policing in social insects -
  aggression, imprisonment, punishment, etc
• Policing does not just repress individual
  selfishness, it also makes it unprofitable for
  individuals to act selfishly in the first place
• Can explain why
• workers in many species are sterile
• females usually accept worker caste fate
• breeder isnt always challenged in totipotent
  societies

30
Acknowledgements

• Collaborators
• Steve Frank, Adam Hart, Heikki Helanterä,
  Thibaud Monnin, Francis Ratnieks
• Funding
• INSECTS network
• EC Marie Curie Fellowship