You are on page 1of 12

Journal of Theoretical Biology 394 (2016) 137–148

Contents lists available at ScienceDirect

Journal of Theoretical Biology
journal homepage: www.elsevier.com/locate/yjtbi

How residency duration affects the outcome of a territorial contest:
Complementary game-theoretic models
Mike Mesterton-Gibbons a,n, Tom N. Sherratt b
a
Department of Mathematics Florida State University, 1017 Academic Way, Tallahassee, FL 32306-4510, USA
b
Department of Biology, Carleton University, 1125 Colonel By Drive, Ottawa, Ontario, Canada K1S 5B6

H I G H L I G H T S 

First discovers often become owners, but sometimes it is unclear who got there first. 
When rival claims to property arise, contest duration increases with residency time. 
Game-theoretic models based on increasing belief or motivation explain these effects.

art ic l e i nf o a b s t r a c t

Article history: While the first individuals to discover and maintain territories are generally respected as owners, under
Received 20 November 2015 some conditions there may be ambiguity as to who got there first. Here we attempt to understand the
Received in revised form evolutionary consequences of this ambiguity by developing a pair of game-theoretic models in which we
10 January 2016
explicitly consider rival residency-based claims to ownership. Following earlier qualitative explanations
Accepted 12 January 2016
Available online 22 January 2016
for residency effects, we assume that either the value of the territory (Model A) or an interloper's self-
belief that it is the owner (Model B) increases with duration of residency. Model A clearly demonstrates
Keywords: that if the value of a territory increases to a resident over time, so should its motivation to fight in terms
Ownership of the effort it invests in fighting. Indeed, only a small increase in territory value with residency duration
Territoriality
can be sufficient for longer established residents to win disputes, even without any arbitrary convention
Animal contests
or other form of priority effect. Likewise, Model B shows that the observed increase in fighting persis-
Game theory
Evolutionarily stable strategy tence with residency duration can be readily explained as a consequence of increasing confidence on
Asymmetric war of attrition behalf of the interloper that it is the rightful owner. Collectively, the models help to explain some general
findings long observed by empiricists, and shed light on the nature of conflicts that can arise when
individuals do not have complete information about rival claims to ownership.
& 2016 Elsevier Ltd. All rights reserved.

1. Introduction both contestants consider themselves the resident (Waage, 1988).
If a territory is large or complex, for instance, then a territory
It has long been appreciated that respect for ownership can holder may be present when another individual arrives, but the
play a role in settling disputes over contested resources, although two individuals do not detect one another until both have been
the extent to which this deference arises as a convention or as a present for some time. Waage (1983, 1988) observed precisely this
consequence of asymmetry in resource holding potential and/or set of events in territorial Calopteryx maculata damselflies and
value remains unclear (see Kokko, 2013; Kemp, 2013; Sherratt and showed experimentally that the protracted contests that arose
Mesterton-Gibbons, 2015 for reviews). Ironically, however, it under these conditions were best understood as a consequence of
appears that this respect can also result in escalated contests if “habitat-mediated confusion” over ownership. Likewise, a resident
may move away temporarily from its territory to forage or repel
intruders, only to be replaced by an interloper. Particularly long
n
Corresponding author. Tel.: +1 850 644 2580; fax: +1 850 644 4053. disputes between interlopers and the returning resident have been
E-mail addresses: mesterto@math.fsu.edu (M. Mesterton-Gibbons),
widely observed under natural conditions (e.g., Wickman and
Tom.Sherratt@Carleton.ca (T.N. Sherratt).
URLS: http://www.math.fsu.edu/  mesterto (M. Mesterton-Gibbons), Wiklund, 1983; Waage, 1988; Gribbin and Thompson, 1991;
https://carleton.ca/biology/people/tom-sherratt/ (T.N. Sherratt). Rutowski, 1992; Kemp, 2003) and they have also been frequently

http://dx.doi.org/10.1016/j.jtbi.2016.01.016
0022-5193/& 2016 Elsevier Ltd. All rights reserved.

Each of the animals is initially unaware of the other's attach- dent as a consequence of its increasing familiarity with its sur. when the based on current occupancy. The first model contests that arise on their return (Wickman and Wiklund. hawk wasps Hemipepsis usulata and found that interlopers that Accordingly. and the other individual—Player proposed that the residency-duration effect they observed in hawk 2—plays the population strategy. not all territories will increase in value to the resi. and so the respective Possible reasons for this gradual change in value of the territory site values. Indeed. Krebs. p. Alcock and Bailey. the longer a resident is absent from its territory. 1997. Kemp and affects the effort expended in a contest over ownership. Alcock and Bailey (1997) plays a potential mutant strategy. Krebs (1982) removed resident pairs of model as Model A and to the second as Model B. Walton and Nolan (1986) made a similar case for the tinuously distributed over ð0. 2012). so that the net even decline with increasing residency duration (so that V 0 ðsÞ o0). also differ. roundings and its neighbors. Model A: no role asymmetry perceived by the contestants former resident great tits and their interlopers was an increasing function of replacement time.g. V ð1Þ ¼ 1. Peixoto and Benson. Let t be the time that has then elapsed since Player suggested that it might arise indirectly through the interloper 1 discovered the territory. felt that such effects were unlikely in butterfly species. then in general sO a sI . 1Þ with probability density function importance of uncertainty over roles when describing the terri- g. when no Wiklund. 1982. Kemp. as was the probability that the First we explore how an increase of resource value with time interlopers would retain the territory. and instead suggests that the sense of roles of owner and intruder are perceived by two contestants. That is. these two animals wasps could also be mediated by a change in perceived value. ment to the site—there is habitat-mediated confusion (Waage. it rules out the possibility that animals with rival time-based The second model addresses how increasing belief in ownership claims to the same territory use entirely arbitrary conventions on the part of an interloper affects contest duration. tion and outcome of disputes involving “co-owners” with joint 1978. For example. 2000. there is clear evidence that the period of time an Parker's (1982) asymmetric war of attrition. 2004. behaviors from an adaptive perspective. the longer they were allowed to where a prime denotes differentiation with respect to argument.138 M. Specifically. the original residents often win the escalated but complementary game-theoretic models. T is a random variable. territorial conflict has been presented to help understand these Intriguingly. Separate mathema- interloper is in residence affects both the duration of the sub. from Player 1's perspective. Takeuchi. Mesterton-Gibbons. Indeed. Alcock and Bailey (1997) on territory affects effort expended in a contest over ownership. VðsO Þ and VðsI Þ. For instance. and distribution function G defined by torial behavior of prairie warblers (Dendroica discolor) after their Z s spring migration. 2000. increasing familiarity with one's sur- GðsÞ ¼ gðξÞ dξ: ð2Þ roundings and confusion over ownership are not mutually exclu. point to quantitatively evaluate earlier verbal hypotheses as to sible if. this individual specifically recognize their neighbors. 0 . before ultimately losing. so that its mean becomes smaragdinus) butterflies that replaced an original owner fought 1 and its variance is longer against original occupants as their residence duration in the Z 1 territory increased. Without loss of generality. The duration of the contests between the 2. we develop two separate resident intruders. great tits (Parus major) from their territories and released them back after replacement pairs had occupied the areas for a con- trolled period of time. ð1Þ rubecula and found that newcomers were more likely to resist eviction by the original owners. and invoked both the above arguments to σ2 ¼ ðs  1Þ2 gðsÞ ds: ð3Þ explain his results.N. the more why interlopers should fight longer to retain a territory. 2006. consider two animals that have discovered an had been on the territory for longer tended to fight longer and otherwise unoccupied territory. as far as possible. We assume that site value is include the residents' increasing familiarity with their physical zero initially and increases with time on territory towards a surroundings. Indeed a However. because role asymmetry is perceived by the contestants. the other for time sI. with two co-discoverers both continually present. Kemp. and their ultimate value may costs when negotiating boundaries with neighbors. 0 sive explanations for the time-in-residence effects. they have been in residence. V(s) is the site value to an animal that has Krebs (1982) interpreted his results to have arisen as a con. Kemp claims for ownership. 2004. 2006). assuming that it retains the resource sequence of an increase in value of the territory over time to an after contesting it. addresses how an increase of resource value with time on territory Krebs. we assume conducted removal experiments in European robins Erithacus Vð0Þ ¼ 0. value of a territory to a resident increased over time. Naturally. but also with their neighbors. but will meet. If one animal has been on site for time sO and interloper. Takeuchi. Tobias (1997) likewise maximum value of 1. Just as Krebs (1982) had proposed. from territories are small and individuals appear not to be able to whose perspective we analyze their interaction. and a decrease in value over time to the former owner. 1997. while the temporary interlopers in the above fundamental phenomena. 1983. Kemp and Wiklund (2001) 1983. it is somewhat surprising that no model of and Wiklund. tical models allow us to focus. the longer likely it will have to renegotiate its boundaries when it returns. we assume that time is measured in Takeuchi (2006) found that green hairstreak (Chrysozephyrus units of the mean of this distribution. where One of these individuals is the focal individual or Player 1. on one of these sequent contest with the former owner and the likelihood that it two effects in isolation from the other. This is a key observation. Alcock and Bailey. the inequalities assumed in (1) are the simplest starting significant asymmetry in net value favoring the interloper is pos. hold the territory. and let T be the corresponding time for being increasingly convinced that the former resident would not Player 2. 1982. 25). Tobias (1997) We note that some territories may have immediate value to resi- argued that newcomers had to pay non-transferable settlement dents on arrival (so that Vð0Þ 4 0). Eventually. conducted an analogous experiment with territorial tarantula when no role asymmetry is perceived by the contestants. been in residence for time s. To elucidate the significance of these experiments are frequently successful at repelling any non. which we assume to be con- return. Sherratt / Journal of Theoretical Biology 394 (2016) 137–148 reported in manipulative experiments following the removal and Given the role of time in residency in affecting both the dura- subsequent re-introduction of the former resident (e. Of course. T. covery is V(s). whose value at time s since dis- harder against the original owner. This model is an adaptation of Hammerstein and time. V 0 ðsÞ 4 0. but ownership and/or motivation to retain the property develops over imperfectly.. This model is new. We will refer to the first will win. Davies. Although t is then known to Player 1.

on the other hand. TÞ ¼ VðtÞpO ðu1 þ u2 t.N. ΔsÞ ¼ ¼ : 1  β ðΔsÞ þ erðx  yÞ 1 þ efλΔs þ rðx  yÞg ð6Þ 3.s. effort) allows us to consider the possibility that individuals do not Because T is continuously distributed. y. u1 þ u2 t. A population strategy v is a strong evolutionarily stable strat- It is assumed that an animal's motivation to contest a territory egy or ESS sensu Maynard Smith (1982) when it is uniquely the can vary with its duration of residency. which requires   that even a new arrival is willing to expend on contesting the ∂f  ∂f  ¼ 0 ¼ ð13Þ resource and a rate of increase of latent effort with respect to ∂u1 u ¼ v ∂u2 u ¼ v residency duration. and so Player 1's expected payoff is We can interpret λ as the rate at which greater experience of the territory translates into a tactical advantage in a contest. v. on which v2 sO if Player 1 has been the intruder and Player 2 the owner. then the efforts expended by Player 1 and Player 2 are Thus we rule out the possibility of a Bourgeois-like convention— x ¼ u1 þ u2 sO and y ¼ v1 þ v2 sI . i. a minimum or base effort mutant strategy u a v. ΔsÞ ¼ 1  pO ðx. Because the evenly matched opponents by virtue of its greater familiarity with focal u-strategist is the owner. Sherratt / Journal of Theoretical Biology 394 (2016) 137–148 139 This is also the distribution from which t has been independently Adopting an intercept (base effort) and gradient (incremental drawn—the distribution of times to first encounter. Specifically. let pO ðx.tÞ ability of victory for the owner would increase sigmoidally with respect to effort difference x  y according to 1=ð1 þ e  rfx  yg Þ. then Player 2 is the owner and Player 1 is the intruder. the longer they are in residence must have arrived before the other. T. Their efforts may differ or be equal. v1 . v. effect of priority advantage combines with that of effort difference drawn from the same distribution as for T. sÞg 1 ðsÞ ds þ 1  GðtÞ In the case where their efforts differ. then the probability of victory is assumed to be determined by “priority advantage” β. If their E expended by either animal is assumed to cost c(E).s. v. ∂u1 u ¼ v . the possibility that they do not refer to the first arrival as the owner. they engage in a contest. which happens with probability βðΔsÞ ¼ 1 e  λΔs : ð4Þ 1  GðtÞ. which together determine the effort that the animal will actually expend in a contest over the territory.v. where we assume for simplicity that β depends on the difference Δs between the owner's and the F 1 ðu. and in no way signifies Player 1 (focal individual) or ∂f  ¼0 ð15Þ Player 2 (population strategy). t  TÞ  cðu1 þ u2 tÞ: ð8Þ intruder's time on territory according to If. y.  respectively. ð5Þ Z 1 Z 1Z 1 1 þ f1  β ðΔsÞge  rðx  yÞ 1 þ e  fλΔs þ rðx  yÞg VðtÞgðsÞgðtÞ f ðu.v. if times sO and sI have elapsed since the of which role it is in: owner and intruder are merely convenient owner and the intruder. let x 0 and y denote effort expended by owner and intruder. y. t. y. v. Thus. y. We denote the corresponding probability  cðu1 þ u2 tÞgðtÞ dt: ð12Þ of victory for the intruder by 0 1  βðΔsÞ 1 pI ðx. s. then Player 1 is the β ¼ 0) toward 1 (where β ¼ 1) in a contest between two otherwise owner and Player 2 is the intruder. but because own- population strategy v is ership is perceived only by an external observer. We will with   refer to the first quantity as base effort and to the second as ∂2 f  ∂2 f  incremental effort. v2 4 0. that is. with expected payoff contest is 1=ð2  βÞ. respectively. c0 ðEÞ 4 0. if Player 1 has been the our interest is rather how time on territory affects contest effort. and denote this pair of quantities by u ¼ ðu1 . after simplification with (5) and (6). tÞgðtÞ dt ¼ ds dt which is readily seen to satisfy pO ðx. ΔsÞ denote the probability of victory for the owner when it has been on ter. Z 1 Z 1 VðtÞgðsÞ We assume that. T 4 t. The evolutionarily stable strategy for Model A Note that our notation suppresses a dependence of pO and pI on λ and r. the unconditional reward to Player 1 from 1 1 playing strategy u against population strategy v is pO ðx. that is when f ðv. t. So the reward at time t to Player 1 from playing strategy u against then we would refer to β as owner advantage. and to the other individual as invest in fighting until they have occupied the territory for a while the intruder. TÞ ¼ VðtÞpI ðv1 þ v2 T. v1 þ v2 T. Before the focal animal dis- ritory for Δs longer than the intruder. Accordingly. u2 . ΔsÞ. where efforts are equal. respectively. as will happen often. integrating over according to that distribution. however. owner and Player 2 the intruder. ProbðT ¼ tÞ ¼ 0. Mesterton-Gibbons. M. which happens with probability G(t). ð11Þ dictor of contest outcome. have been on territory (so labels by which an external observer can distinguish the animals. Hence. a strategy for best reply to itself.  o0 ð14Þ ∂u21  ∂u22  for the focal individual and by v ¼ ðv1 . More generally. v. vÞ for any potential each player consists of two quantities. where where we interpret r as the reliability of greater effort as a pre- Q ðu. however. respectively. in the absence of any priority advantage. its probability of victory is the territory. where u1 . v1 þ v2 T. but instead scripted 1 or 2 is used for either player's base or incremental effort. sÞg 2 ðsÞ ds ¼ ds cðu1 þ u2 tÞ ð10Þ t 0 1 þ e  Q ðu.   Fðu. tÞ ¼ GðtÞ F 1 ðu. the probability that the owner wins this pO ðu1 þ u2 t. vÞ 4 f ðu. the time t is itself a random variable. v2 Þ for the corresponding u¼v u¼v population strategy. one of the two animals invest more effort in fighting. t. u2 Þ  o 0. we use priority Z t advantage instead. F 2 ðu. v. We will (gradient ¼0). and conversely. but by y ¼ u1 þ u2 sI and x ¼ v1 þ When the animals meet.. Note that F 2 ðu.e. T  tÞ  cðu1 þ u2 tÞ: ð9Þ if the animals themselves perceived an asymmetry of ownership. for an “interior” ESS such that v1 . with Δs ¼ t  T. with Δs ¼ T  t. 0Þ ¼ 1=ð2  β Þ Z 12. We assume. that neither individual is aware (intercept¼0). Then we assume that the covers the territory. y. vÞ ¼ Fðu.tÞ Z 1 pO ðx. c″ ðEÞ 4 0: ð7Þ an owner's probability of winning is increased beyond 12 (where If T ot. v2 Z 0. v. Note that a sub. prob. the degree to which cð0Þ ¼ 0. 1Þ ¼ 1 and 0 0 0 1 þ e  Q ðu. ΔsÞ ¼ ¼ . tÞ ¼ λft  sg þ rfu1 þ u2 t  v1  v2 sg. that sO 4 sI ). Effort each expends effort. t.

3. e. Indeed the dashed curve in Fig. Explicit expressions for the quantities appearing ( in (13)–(16) are given by Appendix A.. r and especially α. Most importantly. low. λ ¼ 4. 0Þ with v1 4 0 (see.140 M. the resulting analytical expressions cðEÞ ¼ γ E2 ð18Þ are far too cumbersome to be of any use. the probability that one exceeds the have plotted them only for λ ¼ 0. ð17Þ V and c enable the left-hand sides of (A. and hence probability density Broom and Rychtář. Sherratt / Journal of Theoretical Biology 394 (2016) 137–148 owner has been in residence for time T. ESS for a uniform distribution of times to first encounter  o 0. where S interesting feature of these graphs is how little α influences the and T are random draws from the distribution of times to first intruder's probability of victory after any given time in residence.. We denote the expected How ρI (i. invariably satisfied. and Q is defined by (11). An value of 1=f1 þe  Q ðvn1 þ vn2 S. Calculation of the ESS 1 if 0 o s o 2 gðsÞ ¼ 2 ð23Þ requires specifying the value function V. r and λ and for three n n n n 1 þ e  Q ðv1 þ v2 T. by integrating over despite the increase of contest effort with α. We see that base effort time to first encounter are plotted in Fig.g. which is why we S and T (and dividing by 12.1.TÞ g conditional on T 4 S or as the expected different values of α. for increases much more rapidly with respect to α than vn1 at lower which (A. 2(c) corresponds to the limit We denote this ESS by vn ¼ ðvn1 . other. vn2 Þ. Accordingly.v1 þ v2 S. p. Mesterton-Gibbons. the expected probability ρO of vic- numerical means to yield a unique pair of values of v1 and v2. then the intruder's probability of victory is 1 pI ðvn1 þvn2 T. encounter. α ¼ 0:5. Thus. (1996. Moreover. Our chosen distributions for values of λ. T  tÞ ¼ ð20Þ 1 þ eðλ þ rv2 ÞðT  tÞ n by (6).273. θ ¼ 2. we obtain 3. Let g max denote the maximum value of g over ð0. λ and r.2. first encounter. vn1 þ vn2 t. by (3). Probability density functions of a uniform (dashed) and a Weibull (solid) ρI ð0Þ ¼ dsr g ¼ max n ð22Þ distribution with mean 1 for four different values of its shape parameter θ. Our focus is on how vn of zero priority advantage (λ ¼ 0). its graph is the dashed curve in Fig. we must condition on T 4 t. rÞ ¼ 2  ðλ þ rvn2 Þðt  sÞ ds dt From. uniform distribution with mean 1. o0 ð16Þ ∂u21  ∂u2 u ¼ v u¼v Here we assume that times to first encounter are drawn from a for a “boundary” ESS of the form v ¼ ðv1 . 1Þ. To calculate ρI .S. 2013). 1. 1.SÞ g conditional on S 4 T. 1. γ . 2 for γ ¼ 0:01. λ. depends on γ. 1Þ 0 0 Z 1Z s gðsÞgðtÞ ¼2  ðλ þ rvn2 Þðs  tÞ dt ds: ð19Þ has probability density function 0 0 1 þe  θ θ gðsÞ ¼ θ Γ ð1 þ 1=θÞ sθ  1 e  fΓ ð1 þ 1=θÞsg . which suggests that ρI ð0Þ. whereas the uniform distribution has variance 13. distribution function Let us denote its expected probability of victory at time t by ρI ðtÞ. λ. γ. namely. ð24aÞ Let us now consider the prospects at this ESS of an intruder that has been on territory for time t. respectively. the ESS equations—(13) or (15). Hence Z 1 1 gðsÞ ρI ðtÞ ¼ ds: ð21Þ 1  GðtÞ t 1 þ eðλ þ rvn2 Þðs  tÞ The probability of victory for an intruder discovered soon after arrival is well approximated by taking the limit of (21) as t-0.g. γ. as appropriate. We therefore proceed and choose either a uniform or a Weibull distribution for time to numerically instead. 0. to obtain a conditional probability density function). r ¼2 and three different effort c0 ðEÞ ¼ 2γ E increases with effort. γ . θ using notation that suppresses a dependence on α. Its effort at the ESS is E ¼ vn1 þvn2 t. The 0 1 þ eðλ þ rv2 Þs n max 0 1 þ eðλ þ rv2 Þs n λ þrv2 variances for θ ¼ 1 (exponential distribution). and that vn2 duration distribution density). No boundary ESS of the form v ¼ ð0. Although in principle this choice of g together with our choices for VðtÞ ¼ 1  e  αt . the cost function c and 0 if 2 o s o 1 the distribution of times to first encounter.e. 68).N. Thus α ¼ V 0 ð0Þ measures the rate at which the How vn and ρO depend on α (the rate at which resource value resource rises in value towards its asymptotic level. Mesterton-Gibbons et al. whereas γ rises towards its asymptotic level—see (17)) is illustrated by the measures (twice) the rate at which the marginal cost of fighting left-hand column of Fig. 0.014. We has been an occupant for a shorter time) increases with t is illu- can calculate ρO either as the expected value of 1=f strated in Fig.   ∂2 f  ∂f  3. r and any choice of g (residency effort vn2 is highest for intermediate values of α. e. T. v2 Þ function with v2 Z 0 arises. 3(a) for the same values of γ. λ. the Weibull 1 þ e distribution with mean 1 and shape parameter θ ( Z1) on ð0. λ ¼ 1 and λ ¼ 0. the influ- either the lower or the upper triangle of the joint sample space of ence of λ on these curves is likewise very weak. λ. is typically small—as with illustrated in Fig. θ ¼ 3 and θ ¼ 10 are 1.vn1 þ vn2 T. ESS for a Weibull distribution of times to first encounter Z 1Z t gðsÞgðtÞ ρO ðα.T. vn1 is strictly increasing with respect to α whereas incremental For any values of α. we satisfy (1) and (7) with and variance σ 2 ¼ 13. If the GðsÞ ¼ 1 fΓ ð1 þ 1=θÞsg ð24bÞ . namely. rÞ. α ¼ 1:5 and α ¼ 2:5. the probability of victory by a newly arrived and instantly discovered intruder. Then by standard analysis we obtain the upper bound Z 1 Z 1 gðsÞ ds g lnð2Þ Fig.132 and 0. with tory for the owner remains high even when priority advantage is (14) or (16). the probability of winning for the individual that probability of victory for the owner at the ESS by ρO ðα.3) yields explicit expressions—are readily solved by values of α..3) to be evaluated in terms of hypergeometric functions.

so that the Weibull distribution has a lower no increase in fighting effort with residency duration. which it falls to zero. is that whereas vn2 is always positive for a uniform dis- Γ ðzÞ ¼ 01 wz  1 e  w dw).193 for λ ¼ 4. λ ¼ 0 corresponds to no priority advantage. but is off the drawn from a Weibull distribution is illustrated by the right-hand graph to the right for λ ¼ 0 and λ ¼ 1. λ ¼ 2 (thin solid curves) and λ ¼ 0 (thick dashed curves). The graph of g is plotted in Fig. In effect. M. r ¼2 and three different values of λ in (and a Weibull distribution of residency duration) there should be the case where θ ¼ 2. where RΓ denotes Euler's Γ function (defined for z 40 by however. Mesterton-Gibbons.N. 1 for four tribution. Left-hand column: How (a) base effort vn1 . 2 for γ ¼ 0:01. 2(e). How ρI . A difference. for a Weibull distribution there is a critical value of α at different values of θ. with γ ¼ 0:01 and r ¼2 for λ ¼ 4 (thick solid curves). it is approximately 2. (e) vn2 and (f) ρO vary with α at the ESS when time to first encounter has a Weibull distribution with shape parameter θ ¼ 2 (and hence variance 4π  1  0:273). for the same values of γ. incremental effort 2θΓ ð2=θÞ σ2 ¼ 1 ð25Þ vn2 is highest for intermediate values of α and the expected prob- fΓ ð1=θÞg 2 ability ρO of victory for the owner remains high. Sherratt / Journal of Theoretical Biology 394 (2016) 137–148 141 Fig. T. α. by (2) and (3). Right-hand column: How (d) vn1 . Again we see that base effort vn1 is strictly increasing with respect to α. and hence. This critical value decreases with λ as illu- How vn and ρO depend on α when residency durations are strated in Fig. r and λ. variance variance than the uniform distribution. for high values of α column of Fig. (b) incremental effort vn2 and (c) expected probability ρO of victory for the owner vary with maximum rate of resource value increase. at the ESS when time to first encounter has a uniform distribution. By (4). 2.

672) have shown that at the considers contest effort only. α and λ have only by A. with one contestant in the favored role. 13). when the roles of owner  0 if x r s and intruder are imperfectly perceived by two contestants. but effort is costly. A difference this time is that whereas ρI is favored are defined with respect to payoffs.142 M. 671–672). Interim conclusions disfavored) if VA VB Our model assumes that the territory has value to both animals 4 : ð26Þ CA CB at the time of their encounter. p. Under these conditions. Then. denoted by B a weak influence on the intruder's probability of victory after any (Hammerstein and Parker. pp.3. ð29Þ the animals to observe a role asymmetry. let VI an accelerating function of t for a uniform distribution. if the distribution of residency times is identifies with role I. and this value is non-transferable. Moreover. p. and 1 qV ¼ η.N. provided there is at least a small no effect of residency duration on fighting effort—individuals are positive probability that both opponents identify with the same highly motivated to fight from the get-go. contestants rerðs  xÞ if s o x o 1 Z x GA ðxÞ ¼ g A ðξ Þ dξ Here we explore how increasing belief in ownership on the part 0 of an intruder affects contest duration. it is denote the value of the resource in role I and let CI denote the cost decelerating for a Weibull distribution. until the animal with the shorter persistence time withdraws. but anteed. increases with t is illustrated in Fig. qV ¼ : ð27Þ ner will retain its value in the site while the loser will have to seek CB VB it afresh elsewhere. the contestant with the longer residency—does not require 0 o P A o 1. we have assumed that the win. a contestant identifying with role I should be prepared to contest duration. CA VA qC ¼ . denoted 0 if s o x o 1 . α ¼ 1:5 (thin solid curves) and α ¼ 2:5 (thick solid curves). but we need only the special case of their ( Ke  rx if 0 ox os model in which there exists an “objective asymmetry” between g B ðxÞ ¼ . Mesterton-Gibbons. hence 1  P I the probability that it identifies clumped (Weibull distribution) and α is high. even though it has been in residence for a shorter period of time than its rival. T. Let PI be the probability that a contestant in role I low. Favored and dis- given time in residence. Model B: roles of owner and intruder perceived by the 0 if 0 o x o s g A ðxÞ ¼ . which is more likely to win because longer residency induces greater most readily satisfied by setting incremental effort. qV by to the individual only. For I ¼ A. and the other contestant in the disfavored role. In other cases. However. which reinforces its priority advantage. then there may be with the alternative role. Hammerstein and Parker's asymmetric and war-of-attrition model. In both cases. Again. especially when α is imperfectly. the model Hammerstein and Parker (1982. 0 o P B o1. B. the longer the individual has been on the territory. sensu Kokko (2013. the owner is much Then a sufficient condition for (26) to hold is qC o 1 o qV . role in a contest. How an intruder's expected probability of victory increases with time t in residence when time to first encounter has (a) a uniform distribution and (b) a Weibull distribution with θ ¼ 2 when γ ¼ 0:01. and because winning is not guar- with η 4 1. The roles A and B are perceived by the contestants. Sherratt / Journal of Theoretical Biology 394 (2016) 137–148 Fig. qC ¼ ð28Þ neither contestant knows either which animal has priority η advantage or the extent of it. in that the benefits of residency accrue Let us define cost and value quotients qC. base effort is expected to be smaller. and persist for a time drawn from the continuous distribution with so we turn to Model B. For that purpose we need a different model. namely. probability density function gI and cumulative distribution func- tion GI. Base effort also reinforces priority advantage. 3. 3(b). two contestants. requiring Model A shows that a high probability of victory by the owner— that is.5). 1982. per unit time of contesting it. where  4. the higher its chances of winning (up to a maximum of 0. For this ¼ ð30aÞ 1  erðs  xÞ if x 4 s purpose we adapt an existing game-theoretic model of a contest over a resource. Then role A is favored (and role B 3. and so does not directly address ESS. r¼ 2 and λ ¼ 0 for α ¼ 0:5 (dashed curves).

(29) is not the only condition that PA and PB must Here r. P B Þ must lie for the ESS to exist (light and dark shading) and be unique (dark shading). We show in Appendix B that existence also requires C A þC B ð1  P A ÞC A þ P B C B 1  r r¼ . for various values of η. K and s are defined by satisfy for this ESS to exist and be unique. Thus whenever the two contestants identify with different roles. T. s ¼  ln 1  ð31Þ V A þV B P B ð1  P A ÞðV A þ V B Þ r K PA 4 1  PB . Note that we have recast the expressions given by 8 <K Hammerstein and Parker (1982. : 1 if x 4 s: Nevertheless. Region of the unit square in which the point ðP A . 672) to present the results we ð1  e  rx Þ if x r s ¼ r ð30bÞ need in a form more suited to our purpose. the contestant in the favored role identifies with it more . ð32Þ with (29) implying K 4r.N. 4. Sherratt / Journal of Theoretical Biology 394 (2016) 137–148 143 Fig. The significance of the L-shape inside the top right-hand quadrant is discussed in Appendix B. Z x invariably prepared to persist for longer than the animal in the GB ðxÞ ¼ g B ðξ Þ dξ 0 disfavored role. p. the animal in the favored role is that is. K¼ . M. Mesterton-Gibbons.

Mesterton-Gibbons. By (32)–(34) and (36). 5. P B Þ to lie in the dark shaded curvilinear quadrilateral within the upper right-hand quadrant. Sherratt / Journal of Theoretical Biology 394 (2016) 137–148 Fig. than its opponent does. the ESS is unique only for 0 o ω o 45 (where the curve is shown solid) but continues to exist for 45 r ω o P A (where the curve is shown dashed). V B ¼ 1 ¼ C B and C A ¼ 12 (implying r ¼ 12) to satisfy (28) with η ¼ 2. P B Þ to lie in this region is η þ1 η PA 4 . Then the temporary resident will win with probability 12 if either player is mistaken about its role (so that both draw either from the lower or the upper part of the distribution) and with probability 1 if both players are mistaken. 5.95. We further show in Appendix B that uniqueness additionally requires ðP A . 4. P B Þ lies in the (light or dark) shaded triangle in Fig. By (32)–(34). Let us make this correspondence explicit  PA  ϵ 1 δ ln P A  η o t o 1 where the curve is shown dashed. Let us now identify ownership with role A and temporary residency with role B. r ¼ 12) to satisfy (28) with η ¼ 2. V B ¼ 1 ¼ C B and C A ¼ 12 (implying (33) is satisfied for ω o 25. Thus the probability that the temporary resident wins is 1 2  P B ð1  P A Þ þ 12  ð1  P B ÞP A þ 1  ð1  P B Þð1  P A Þ þ 0  P B P A ¼ 12 f1  P A þ ωg ð35Þ where ω ( ¼ 1  P B ) is now the probability that the intruder believes itself to be the owner. but will lose if both players are correct. The above expression is strictly increasing with respect to ω. and that a sufficient con- dition for ðP A . where we also calculate its mean μ and variance σ2.144 M. μ. the ESS is unique only for  Increasing 1  P B corresponds to increasing time on territory.95 and η ¼ 2. 6. 1Þ of contest duration is determined in Appendix C. μ and σ2 versus intruder's probability ω ¼ 1 P B of belief in ownership for 0 o ω o P A with PA ¼ 0. the point ðP A . so that in belief in ownership δ with PA ¼ 0. The distribution over ð0. but continues to exist for 2η þ 1 which we denote by t. VA ¼ 2. 2η þ 1 . ωo ð33Þ 2η þ 1 2η þ 1 where ω ¼ 1  PB : ð34Þ It is convenient for our purposes to assume that (33) holds. In Fig. σ2 and pI versus time on territory t for ϵ ¼ 0:025 and three rates of increase ω ¼ 1  P B of belief in ownership for PA ¼0.N. T.95. VA ¼ 2. When (32) holds. 0o t o 1δ ln P APA  ϵη where the curve is shown solid. μ and σ2 are plotted against the intruder's probability Fig.

e. model (B) considers a contest over a resource that remains fixed in although the two are clearly interlinked. we feel it is appropriate to spread the net three different values of δ. or behaviorally mediated confusion nant and reversing the contest outcome in staged competitions. The duration of these fights in such instances was umn. 6(b) and (c) shows the corre- wide and consider examples of disputes over all resources. and both models consider contests a recent study. μ. We have developed two complementary models to understand 1991). In Fig. We have presented our model in terms of 2f1  P A þ ϵg as t-0 to 2 as t-1. resources become more valuable to residents over time. but spend most of their time hunting. 2). seemingly temporarily unaware of when the distribution of residency times is clumped and the rate each other's presence. there is considerable evidence that the value of a resource influences an individual's motivation to obtain or retain 5. 1983). It is probability). with the resource going to the individual effect. Fig. winning against a co-owner might vary with the duration of (2010) found that male Pararge aegeria butterflies that had inter- residency. ϵ (4). 1982). beha- pI ¼ 12 1  ðP A  ϵÞe  δt : ð37Þ vioral ecologists have long tended to consider contests over non- breeding resources under the same general umbrella as conflicts This expression increases with t at a rate determined by δ. because rival time-based claims the lower limit is small (  12ϵ). Sherratt / Journal of Theoretical Biology 394 (2016) 137–148 145 with experience of it—but is also a quite distinct effect. but when they do a fight gen- may not increase with time in residency (Fig. Silk is a costly material to value over time. The ESS identified in this second model is a this non-refundable investment that may explain the “endowment persistence-time bid. since the katydid resources are entirely to lose than a more temporary resident if they fail to keep it. just as territory n o occupancy is frequently disputed. because ϵ is small with P A  1. erally ensues. (2014) found that the signalling rates of which are decided on the basis of the costs individuals are pre. the mean contest are much more likely to arise under these conditions. and pI. σ2. from over breeding resources. VB. Discussion it. VA. and if the caterpillars have invested limited time and dependent degree of confusion over which role an interloper is in resources in creating a shelter. effect was a marginal one). In over ownership (Waage. Mesterton-Gibbons. An exception to this general rule arises only are unusual in that two wasps. pillars and shells for hermit crabs are not breeding territories—but Then the probability that the temporary resident wins becomes they are objects of value that are fought over. This transferable and therefore equally valuable to both parties. since it arises  δt even in the absence of priority advantage. i. Moreover.. the model predicts that if found to increase with the total number of katydids in the nest. Collectively. 1991) simply metric war of attrition. 1984). reflecting a less than those occupying an intact shelter. which implies β ¼ 0 in where δ is the rate at which belief in ownership increases. Hammerstein and Parker. 6(a). make. (small) is initial belief in ownership and the intruder's belief in Resources such as cases for caddis flies. when discussing how investment in fighting might change values of PA. CA and CB as in Fig. By contrast. Never- duration. any tactical advantage in contesting the territory from greater the wasps count only their own investment when considering .. One effect of longer residency reinforces any priority advantage—that possibility. Intriguingly. The first model (A) assumes no with increased duration on the leaf prior to introducing an asymmetry in roles (neither individual knows the arrival time of intruder. long-term owners tend to invest more in fights the wasp who had brought most prey to the nest (although the and thereby win.. Both models assume conditions under which there is a acted with a female had a higher probability of becoming domi- degree of habitat-mediated. Bergman et al. provision the same burrow with insect prey of increase of resource value with time in residency is very high. a low α). while spiders increase their signalling rates when they why both the contest length and the probability of an individual occupy webs that are larger in area (Riechert. 5 with ϵ ¼ 0:025 in (36) and with territory value. suggested by Dawkins and Brockmann (1980). Intrigu. residents should invest more to retain the loser had contributed. rather sponding curves for the variance.e. T. resident masked birch caterpillars (Drepana arcuata) increased pared to incur in order to win. thick solid curves). because the consequences of loss become higher for an investing Model A shows that if a territory becomes more valuable to a resident than a random intruder. In short. squatters placed on a leaf containing a full silk the other) and simply investigates how a change in value of the shelter (made by another individual) signalled more to intruders resource during occupancy affects the ESS contest effort on than did caterpillars placed on a fresh leaf a similar amount of encounter.e. M. an asym.” with individuals valuing the objects they own more than prepared to contest the resource for the longest (i. since incremental effort tends to be digger wasp Sphex ichneumoneus. In this regard. then they will have gradually (the real owner continues to consider itself the owner with high accumulated costs that are at least in part non-transferrable. This ESS contest effort takes the form of an initial time. 1 1 disputes over territory occupancy. even for a relatively modest increase in the value support for the intuitive proposal that individuals should be more of the resource to an occupant over time (i. One might wonder why the greater transferable costs into the territory and will therefore have more investor tends to win. it provides quantitative ingly. Although our first model was based on their resources the longer they have been in residence. 2. however. because they have already invested more non. These instances of co-occupation positive at the ESS. contest effort rather than fight duration. then the evolutionarily stable effort expended In a classic paper. Dawkins and Brockmann (1980) reported in excess of base effort is in general positive and increases with field observations of 23 incidences of co-occupation of nests by the time spent in residency. their potential increase in the motivation to win as the resource experiments indicate that the motivation of caterpillars to signal is becomes more valuable to the resident. silk shelters for cater- ownership (1  P B ) approaches that of the true owner (PA) as t-1. is that is. the ESS effort is high (because base effort is high). in (paralyzed katydids).N. when λ ¼ 0 (corre- P B ¼ 1  P A þ ðP A  ϵÞe ð36Þ sponding to the dashed curves in Fig. then but was most strongly correlated with the number of prey items whenever a dispute arises. identical objects owned by others (Kahneman et al. right-hand col. but assumes an asymmetry in roles and a time. Given the close parallels. than only territorial ones. Put simply. the second a function of both prior residency duration and territory quality. The co-occupants rarely meet because they this case.. while residents whose shelters had been removed signalled investment and a time-dependent additional effort. resident over time. the winner of the above contests tended to be ensues. is plotted against time on territory t for the same theless. Yack et al. Caddis fly larvae residents alter their aggressive response according to the quality of the case they occupy (Englund and Otto. the longer motivated to fight to retain a property when it is more valuable to residing co-owner is much more likely to win any contest that them.

B. as appropriate. temporary replace. between 1 and 2 days before consistently dominating intruders. for any values of α. that is. As ser. Likewise. a COFRS award because ProbðAÞ ¼ 12 ¼ ProbðBÞ.1) reduces (13) to however. r and any choice of g. s. we have considered a continuous game. Naturally. s. J ¼ A. ABÞ ¼ P A P B =fP A P B þ ð1  P B Þð1  P A Þg and ProbðBj ABÞ ¼ ð1  P B Þð1  P A Þ=fP A P B þ ð1  P B Þð1  P A Þg. tÞÞgðsÞgðtÞ ¼  r2 ds dt have held the territory in the past. ProbðIJj KÞProbðKÞ ProbðK j IJÞ ¼ ProbðIJj AÞProbðAÞ þ ProbðIJj BÞProbðBÞ Acknowledgements ProbðIJj KÞ ¼ ðB:3Þ ProbðIJj AÞ þ ProbðIJj BÞ This work was partially supported by a grant from the Simons Foundation (#274041 to Mike Mesterton-Gibbons). v. between territorial convict cichlids. yet tend to be short with ran. Mesterton-Gibbons et al. Setting u1 ¼ v1 and u2 ¼ v2 in (A. Player 1 identifies Nevertheless. ∂u22 2 0 0 f1 þ coshðQ ðu. T. this work Appendix B. 2002). Continuing in this fashion. (A. the model predicts that  c″ ðu1 þu2 tÞgðtÞ dt ðA:2aÞ in the absence of any RHP asymmetry. The ESS equations for Model A V AA ¼ 12 ðV A þ V B Þ P P B V A þ ð1  P B Þð1  P A ÞV B From (12) we obtain V AB ¼ A P A P B þ ð1  P B Þð1  P A Þ Z 1Z 1 Z 1 ð1  P A Þð1  P B ÞV A þ P A P B V B ∂f 1 V ðtÞgðsÞgðtÞ V BA ¼ ¼ r ds dt  c0 ðu1 þ u2 tÞgðtÞ dt. allowing us to Z 1Z 1 Z 1 consider contest duration directly. then the duration of the contest with the ∂u12 2 0 0 f1 þ coshðQ ðu. v. but not after 2 h removal. Mesterton-Gibbons. Here. J ¼ B. v. Sherratt / Journal of Theoretical Biology 394 (2016) 137–148 Z 1 Z 1 Z 1 value. Moreover. and for contestants to have role wBB ¼ P B ð1  P A Þ ðB:1Þ asymmetry with imperfect identification of roles. as has a time-dependent increase in the probability that ðA:3aÞ the interloper will win (see Section 1). tÞÞ 0 ðA:1aÞ V BB ¼ 12 ðV A þV B Þ . and isolation but also ensure that the models are analytically tractable. from Florida State University and a grant from NSERC (to Tom Then ProbðABj AÞ ¼ P A P B and ProbðABj BÞ ¼ ð1  P B Þð1  P A Þ. v. 1 VðtÞgðsÞgðtÞ r ds dt ¼ c0 ðv1 þ v2 tÞgðtÞ dt An increase in contest duration with residency has been widely 2 0 0 1 þ coshðfλ þrv2 gft  sgÞ 0 observed. 1 tVðtÞgðsÞgðtÞ r ds dt ¼ tc0 ðv1 þ v2 tÞgðtÞ dt ments that were allowed to be resident for 4 days fought for 2 0 0 1 þ coshðfλ þrv2 gft  sgÞ 0 longer and won more contests against the original owners than ðA:3bÞ interlopers who were resident for only 2 days (Johnsson and For. with (14) or (16). 1 conditional upon event (or situation) IJ. Sherratt).3a) for a boundary ESS. a belief Z 1Z 1 that we have assumed to increase with time in residency (see also ∂2 f 1 VðtÞsinhðQ ðu. First suppose that I ¼A. v. invariably satisfied. perhaps (as the authors suggest) as while Player 2 identifies with role J for I. tÞÞgðsÞgðtÞ ¼  r2 ds dt Alcock and Bailey. v. In addition. it is quite possible for a territory to increase in net wBA ¼ 12 P B P A þ 12 ð1  P A Þð1 P B Þ ¼ wAB value with increasing familiarity. tÞÞg 2 Z 1 dom intruders. wAA þwAB þ wBA þ wBB ¼ 1: ðB:2Þ Indeed. by treating these phenomena separately we can not with role I while Player 2 identifies with role J. ∂u2 2 0 0 1 þ coshðQ ðu. Figler and Einhorn (1983) staged contests noted in Section 3. tÞÞ 0 Our complementary model B confirms that as an interloper ðA:1bÞ grows increasingly confident that it is the rightful owner. in which case our model readily explains why the individual ∂f 1 tVðtÞgðsÞgðtÞ ¼ r ds dt  tc0 ðu1 þ u2 tÞgðtÞ dt.146 M.N. γ. s. 1997). s. with v2 ¼ 0 in (A. Such a model with would involve a modification of Model B to allow territory value to vary with time in residency. as currently assumed in Model A. we readily obtain Appendix A. P A P B þ ð1  P A Þð1 P B Þ ∂u1 2 0 0 1 þ coshðQ ðu. to Player interloper is eventually in the favored role with respect to payoffs. in aquarium Z Z Z 1 1 1 experiments with brown trout (Salmo trutta). Then. (2014) analysed a pair of territorial games with discrete strategies and also found that  t 2 c″ ðu1 þ u2 tÞgðtÞ dt: ðA:2bÞ 0 confusion over roles increases the frequency of fighting. the probability that the 0 interloper will win in a contest against the original resident rises and as the interloper's belief in ownership increases. J and K all be either A or B. let ProbðDj EÞ only elucidate the effects of motivation and confusion over roles in denote the probability of event D. they found that residents temporarily removed from their territories after 3 days dominated their interloper after 1 h removal. who places greater value on the resource tends to win it. Constraints on Hammerstein & Parker's ESS demonstrates that the prior-residency effect among individuals with competing claims to ownership is not instantaneous and Let wIJ be the probability that Player 1 identifies with role I develops over a period of time. s. tÞÞg 2 Z 1 original resident will increase. Thus the model Z 1Z 1 2 helps explain directly why fights are longer when both contestants ∂2 f 1 t VðtÞsinhðQ ðu. for an interior ESS. Collectively. For example. Cichlasoma nigrofasciatum and Eqs. respectively. conditional upon event E. Then a consequence of the decline in a fear response (another form of wAA ¼ P A ð1  P B Þ non-transferable cost analogous to negotiating boundaries with wAB ¼ 12 P A P B þ 12 ð1  P B Þð1  P A Þ neighbors). λ. by Bayes' Theorem.3) are readily solved by numerical means to yield a unique found that individuals must be the sole resident for somewhere pair of values of v1 and v2. s. We thank our editor and an anonymous reviewer for from which V AB ¼ V A  ProbðAj ABÞ þ V B  ProbðBj ABÞ with ProbðAj helpful comments. let events I. when this arises there may be conditions under which the Let VIJ and CIJ denote value and cost.

CRC Press.. Entomol. Then V AB V BA fð1  P A Þð1 P B Þ þ P A P B gðP A þ P B  1ÞðV A C B  V B C A Þ  ¼ ProbðDAA 4 τ Þ ¼ ProbðX A 4 τ. P B Þ to lie in the σ2 ¼ ðτ  μÞ2 hðτÞ dτ ¼ 0 4r 2 still smaller square with solid lower and left-hand boundaries. we require ¼ ProbðDAA r τÞ  wAA þ ProbðDAB r τÞ  wAB þ ProbðDBA r τÞ wBA þ ProbðDBB r τÞ  wBB ð2P A 1ÞP B qV þ ð1  P A Þð1  2P B Þ 4 0 ðB:7aÞ ¼ P A ð1  P B ÞGA ðτÞf2  GA ðτÞg þ f1  P A þ P B gGB ðτÞ and  P B ð1  P A ÞfGB ðτÞg2 ðC:3Þ ð1  2P A Þð1  P B ÞqC þ ð2P B  1ÞP A 4 0 ðB:7bÞ after simplification with (B. So existence requires only (29) and (32): Bergman. Wiklund.N. KðKð1 P A Þ  rÞðKP B  rÞð1  P A þ P B Þs2 þ The weak asymmetry condition is a sufficient condition for r3 uniqueness. Z 1 4  ðP A  P B Þð2 þ P A  P B þ 4ð1  P A þ P B ÞKsÞ A sufficient condition for (B. Favored is first defined by inequality ProbðDBB 4 τÞ ¼ ProbðX B 4 τ.e. P B Þ lies within r3 this square when (33) holds. we find that D has density where qC and qV are defined by (27). Within this smaller square. the ESS exists throughout the shaded upper triangle of Fig. and let Player 2's draw when identifying with role J be denoted by YJ..8) to hold is for ðP A . let P A P B þ ð1  P B Þð1  P A Þ random variable D denote (unconditional) contest duration. It Kð1  P A ÞP B ðð3ðP A  P B Þ  1ÞKs P A þ P B  1Þ þ is now straightforward to show from (B. Contest outcome in a territorial butterfly: the role of motivation. which lies wholly within the region that is shaded dark in Fig. Ecol. not a necessary condition for existence. Differentiating and using (30). M.5) to be positive. both wAB V AB 4 wBB H AB ðτÞ ¼ ProbðDAB r τÞ ¼ GB ðτÞ V BB and wBA C BA 4 wAA C AA (Hammerstein and Parker. 653). 4. let Player 1's draw when identifying with and hence role I be denoted by XI. M. W. M. Sherratt / Journal of Theoretical Biology 394 (2016) 137–148 147 C AA ¼ 12 ðC A þ C B Þ Appendix C.7b) excludes P A Z 12. 377–383. P B Þ lies in the top right-hand quadrant of the square ( defined by (29). p. C. which translates to (26). Proc. it fills the ðC:4Þ top right-hand quadrant of the unit square defined by (29) in the limit and variance as η-1. if (32) holds.1). the ESS defined by (30) and (31) will still exist as long as r4 w2AB C BA V AB 4 wAA wBB C AA V BB (Hammerstein and Parker. 653) as requiring the left- ðC:1Þ hand side of (B. i.1). Even if it fails K 2 ð1  P A ÞP B fð1  P A ÞP B ð2Ks  1Þ  ðKð1  P A Þ  rÞðKP B  rÞs2 g þ : to hold. 672) only if also implying P A þ P B 4 1. Player 1 identifies with role I while Player 2 identifies with role J. 1997. 4 for various values of η. (B.8) that ðP A . (B. Success in territorial defence by male tarantula hawk wasps Hemipepsis ustulata: the role of residency. Raton. 1982. Olofsson. (B. B 277. The uniqueness of the ESS also H AA ðτÞ ¼ ProbðDAA r τÞ ¼ 1  f1  GA ðτÞg2 requires the “weak asymmetry” condition.7a) excludes hðτÞ ¼ H 0 ðτÞ P A r 12. (B. that is. ðqV þ 2ÞP B  1 ðη þ 2ÞP B  1 PA 4 ¼ 2ðqV þ 1ÞP B  1 2ðη þ 1ÞP B  1 mean P þ ð2P A  1ÞqC ðη þ 2ÞP A 1 Z 1 PB 4 A ¼ ðB:8Þ 2  P A þ P B þ2ðr  KP B Þs Kð1  P A ÞfðKP B  rÞs  P B g 2P A þð2P A  1ÞqC 2ðη þ 1ÞP A  1 μ¼ τhðτÞ dτ ¼ þ 0 2r r2 This region is shaded dark in Fig. p. p. ðC:5Þ 665) in addition to (29) and (32). Because (B. It is straightforward to show that (B. 1982.. . Lond.4). Y A 4 τÞ ¼ ProbðX A 4 τÞ C AB C BA fP A P B C A þð1 P A Þð1  P B ÞC B gfð1  P A Þð1  P B ÞC A þP A P B C B g ðB:5Þ ProbðY A 4 τÞ ¼ f1  GA ðτÞg2 ProbðDAB 4 τÞ ¼ ProbðY B 4 τÞ ¼ 1  GB ðτÞ in which the denominator and the term in curly brackets in the ProbðDBA 4 τÞ ¼ ProbðX B 4 τÞ ¼ 1  GB ðτÞ numerator are both positive.. 4. that is.. Boca although it is only in the darker region that the ESS is unique.J. which is always positive. P B Z 12. P B r 12. J. that is. Y B 4 τÞ ¼ ProbðX B 4 τÞProbðY B 4 τÞ ¼ f1  GB ðτÞg2 (5) of Hammerstein and Parker (1982. 22. Bailey. Game-Theoretical Models in Biology. 1982.. R.7) can hold only if PA and PB both exceed 12. Mesterton-Gibbons. P B r 12 and (32) excludes ¼ 2P A ð1 P B Þg A ðτÞf1  GA ðτÞg þ f1  P A þ P B gg B ðτÞ P A r 12. J. 2010.7) will actually hold Kf1 P A þP B  2KP B ð1  P A Þð1  e  rτ Þ=rge  rτ if 0 o τ o s ¼ within the region defined by the inequalities 2P A ð1  P B Þre2rðs  τÞ if s o τ o1. Rychtář. FL. Broom. The contest duration distribution for Model B P A P B C A þ ð1  P B Þð1  P A ÞC B C AB ¼ To calculate the distribution function for contest lengths. H BA ðτÞ ¼ ProbðDBA r τÞ ¼ GB ðτÞ We have   H BB ðτ Þ ¼ ProbðDBB r τÞ ¼ 1  f1  GB ðτÞg2 ðC:2Þ wAB V AB  wBB V BB ¼ 12 ð2P A 1ÞP B V A þð1 P A Þð1  2P B ÞV B   and hence wBA C BA  wAA C AA ¼ 12 ð1  2P A Þð1  P B ÞC A þ ð2P B  1ÞP A C B ðB:6Þ HðτÞ ¼ ProbðD r τ Þ For both these quantities to be positive.. p. Let D or DIJ have distribution function H or C BB ¼ 12 ðC A þC B Þ ðB:4Þ HIJ and density h or hIJ. 2013. M. and ð1  P A Þð1  P B ÞC A þ P A P B C B let random variable DIJ denote contest duration conditional upon C BA ¼ P A P B þ ð1  P A Þð1 P B Þ situation IJ. V A =C A 4 V B =C B (Hammerstein and Parker.. Soc. ðB:9Þ Alcock. that is. 3027–3033. T. if  2P B ð1  P A ÞGB ðτÞg B ðτÞ the point ðP A . (27) and (28) imply References w2AB C BA V AB  wAA wBB C AA V BB ¼ 14 ðP A þ P B  1ÞfP A P B V B C B  ð1 P A Þð1  P B ÞV A C A ¼ 14 ðP A þP B  1Þ2 V B C B .

http://dx. Melanitis leda (l. contest in a green hairstreak Chrysozephyrus smaragdinus (Lepidoptera: and status quo bias. 11.) Waage. Imperfect information and the persistence of pre- Kemp.. 11. Biol. Wiklund. P. Cues underlying context-associated (Sulzer) (Zygoptera: Coenagrionidae). Behav. T. 1992. 427–432. Matter of size or matter of residency experience? Territorial Kahneman. Krebs. Dyadic contests: modelling fights between two individuals. R. Johnsson.. J. 2001.. M.A. Marden. Ethology 112. J. R. The effects of size and residency on territorial Satyrinae).. 293–299.J. Confusion over residency and the escalation of damselfly terri- Hardy. http://dx.. Sociobiol.R. 157–183. 28.. Behav.. Ethology 120. J. 181. 2000. Animal Contests.D.I. H. R. 591–596. Kemp.. Nat. 185–194.. The evolution of respect for property. game revisited: the effect of ownership uncertainty on Bourgeois as a pure Englund. Mesterton-Gibbons. Anim. Residency duration and Kokko. Behav. Econ.J. cichlids (Cichlasoma nigrofasciatum Gunther): temporal aspects of establish. 2013.. 586–595. Male mate-locating behavior in the common eggfly. 429–442. 2006. 46. 892–896. 2013. Wickman. T. M. (Eds. Evolution and the Theory of Games.. Cambridge University Press.E. The endowment effect. J. Soc. Proc. weight asymmetry. 66. In: shelter quality influence vibratory signalling displays in a territorial caterpillar.K. W. Theoret.. Do digger wasps commit the concorde fallacy? Mesterton-Gibbons. The iterated Hawk-Dove Anim. torial disputes. 30. Forser. Gill. M. C. 1185–1202. 2014. Residence duration influences the outcome of ter. Behav. disputes and short-term mating success in the damselfly Pyrrhosoma nymphula Riechert. D. 113–120. Otto. Mesterton-Gibbons. 7–13. Ecol. Dugatkin. Parker. Lycaenidae). Briffa. Anim..B.N.. contest competition in butterflies. Twilight fighting in the evening brown butterfly. Ecol. Dyn. Behav.H. R.). 1988.. V..N. Evol.. loss aversion.. M.org/10. Behav. 65–83.doi.M. Ecol. 1980. J. The territorial prior residence effect in convict without assessment. Drummond-Main.org/10. Sherratt. and case convention. Territorial defence in the speckled wood butterfly (Pararge Maynard Smith. Dawkins. 5–32. H. 28.E. J. J.J. Cambridge.J. from damselflies (Odonata). 51. Ecol. J.). Effects of ownership status.. In: Waage.K. Peixoto. M. Nolan Jr. Kemp. Sexual selection.. 1983. 4. Behav.. J. III.J. J. ritorial conflicts in brown trout (Salmo trutta). 26.H. Soc.A. Theoret. Contest behaviour in butterflies: fighting without weapons.H.148 M. Behav. Press. 1982. Contest behavior in territorial male butterflies: does size matter?.L. Residency effects in animal contests. 2002.. (Trichoptera: Phryganeidae) larvae.. 193–206. Briffa. Behav.. Entomol. Hypo- Biol. 407–431. Yack... Knetsch..W. T. Territorial defence and its seasonal decline in the Kemp.1111/jeb. Anim. D. Hardy.doi. Anim.C. S. . Lepid. Games spiders play. Fighting without weaponry: a review of male-male tenders: male prairie warblers contesting for territory. 1997. Walton.. 1996. 2014.. Anim. 31. C. 96. 1991. Influence of previous residency and body mass ment and retention and proximate mechanisms.E. Anim.. 29.12648. Einhorn. N. 128.. Territorial defence in the great tit (Parus major): do residents always win?. Fla. 5. Cambridge. M. Ecol. J. D.. Brockmann. Cambridge University Press. 354–364. 1978. Rutowski. J. 282–286. D.. Ethol. Mesterton-Gibbons. 49. 1991. 32. Behav. 9–21.. D. Cambridge University aegeria): the resident always wins. 1982. 1982. T. 1986. 138–147. L.. pp. D. D. Karabiyik. Sherratt. Sociobiol. 54. P. D. changes in agonistic behaviour.. The asymmetric war of attrition. 1984. Games Appl. 1983. Sherratt. S. 61–68.. Thompson. Tobias.L. Benson. Ecol.1007/ fit on the outcome of case contests in two populations of Agrypnia pagetana s13235-014-0111-5. 1991. 54. P. 36. On wars of attrition Figler. (Eds. 2003. C. 2004. limnas bolina (nymphalidae). Thaler. 1–15. G.. Wiklund.N.. 689–695. J. 24–38. K.. Sociobiol.. Animal Contests. Behav.N.. S. Takeuchi. Behav.. G.. J. settlement costs. speckled wood butterfly Pararge aegeria. C.. ESS theory and insect behavior: some examples (Nymphalidae): age and residency effects. Perspect. B 271. R. 134–146. Am. 2012.. in the territorial contests of the butterfly Hermeuptychia fallax (Lepidoptera: Gribbin. Sherratt / Journal of Theoretical Biology 394 (2016) 137–148 Davies. Asymmetric territorial contests in the European robin: the role of Behav. J. 41. 1206–1216. Kemp. A. pp. 647–682. 2015. Sociobiol... C.. 1983. Behaviour 85... Sociobiol. 1707–1711. Lond. M. 19–31. T.O.. Hammerstein. Cambridge. M.. Biol. Wiklund.J.