Territorial Signaling Games Dr. Tim Wright Molecular Genetics Lab, Smithsonian National Zoo All figures from Bradbury and Vehrencamp 1998 unless otherwise noted

Outline of lecture Terms and definitions The Ownership Convention Game theoretical approaches to territory signals Assessment of neighbors versus floaters Competition and deceit

What is a territory? Territory: a fixed area from which intruders are excluded by an owner using advertisement, threat, and attack Types of territories –Breeding territory- small, contains only nesting or mating sites –Feeding territory-larger, support territory holder –All-purpose territory- largest, nesting site plus food for owner –Neighborhoods- networks of contiguous territories Territorial signal: a type of long-distance threat signal

Avian territoriality: temperate vs tropical Highly seasonal vs year-round Male only vs joint defense Male song vs duets Stutchbury & Morton 2001

Territory defense signal design rules

Territory ownership and auditory signals Great tit song Experimental removal of owners Broadcast of songs from speakers After 12 hours, new owners only areas with songs Slower establishment with multiple songs

Types of intruders Neighbors: other territory owners in neighborhoods –Encounters initially aggressive, –Settle to stable associations with repeated low-level encounters –Escalated response only necessary when neighbor trespasses –Owner should recognize neighbors and judge location Floaters: Non-owners searching for a territory –Owners respond with immediate escalation –Potential asymmetry in fighting ability compared to owner –Asymmetry in value of territory with the owner Owners should distinguish between neighbors and floaters

Quantifying territory value Units of measurement = number of offspring produced in a territory (fitness) And in a fight over a territory- Territory Value = (Number offspring if fight won)- (number of offspring if fight lost)

Value of a territory differs between individuals Owners--high value to territory because high cost of losing a fight Floaters--high value to territory because low cost of losing and high benefit to winning Neighbors--low value to gaining a second territory relative to cost of defending two territories

Why do owners usually win? Bias towards owners winning often obserevd –Owners usually win contests –neighborhoods often have stable membership and boundaries Why is this the case? Asymmetry in fighting ability Asymmetry in valuation of territory Convention that owner always wins

Hawk-Dove-Bourgeois game V= territory value to winner D= cost to loser Bourgeois is a conventional strategy, uncorrelated with fighting ability and territory value Strategies: Hawk: Escalate Dove: Display Bourgeois: Hawk if owner Dove if intruder Assume owner 1/2 time Result: -Bourgeois only invades if cost of fighting greater than benefit of winning -Value of territory must be low

Do ownership conventions really exist? Speckled wood butterfly in England –Males defend sunspots to attract females –Sunspots provide warmth for high activity levels –Owners contest with intruders using spiral flight –New owners always beat former owners after experimental removal –Territory value low because of ephemeral nature of territories? But--no bourgeois convention found in Sweden –Returning owners engaged in long battles and usually won –differences in territory value or changes in fighting ability during removal? Similar studies in birds often find previous owner regains territory upon release

Alternatives to simple ownership convention Owners differ from intruders in either relative fighting ability or valuation of territory Three territory games allow assessment of fighting ability and territory value –Asymmetric hawk-dove –Asymmetric war of attrition –Sequential assessment All three games better at predicting patterns seen in nature

Asymmetric war of attrition V = benefit of winning k = rate of cost accrual V/k ratio determines how long an individual is willing to fight Contestants assess relative V/k ratio Select high or low persistence time separated by cutoff value ‘S’ Contests longer (S bigger) when V/k ratio similar between two contestants

Peak in contest duration Contests in great tits Prediction: value of territory increases with tenure, decreases with absence Assumes costs (k) similar Results: -Returning owners won after short absences, lost after longer absences -Longest contests for intermediate absences -No evidence of ownership convention -No evidence of asymmetry in fighting ability

General conclusions from territorial games Simple conventions like ‘owner always wins’ are unlikely Territory owners value territories more than intruders –Investment in reproduction –Increased probability of additional mates –Increased knowledge of food and hiding places –Stable boundaries with neighbors Territory valuation more important than fighting ability

Implications for territorial signals Territory signals should be designed to: –Primarily display value placed on territory ownership (motivation to fight) –Secondarily display fighting ability –Be individually distinctive Example: Territorial defense calls in cricket frogs –Frequency of initial call by territory owners correlated with body size –Subsequent responses to intruders lower in frequency –Drop in frequency correlated with likelihood of attack –Calls contain information on both fighting ability and of motivation to fight

Assessment of neighbors versus floaters Neighbors and floaters differ in degree of threat to owners Owners need to assess these differences Neighbors need to be recognized as such Production of individually variable signals Discrimination of these signals Memorization of proper location of a neighbor

Individual signatures in kangaroo rats Individually distinctive temporal patterns of hindfoot drumming Signatures change when individuals move to new territories Signatures shifted to make them maximally distinctive from neighbors Randall 1995 AnimBehav. 49:

Neighbor-stranger discrimination Neighbor’s song ignored when broadcast from proper territory Aggressive response when neighbor’s song broadcast from different territory Rule of thumb: recognize neighbor when in own territory, treat all other songs as if from floaters

Sound degradation and ranging Signal amplitude Frequency dependent attenuation Reverberation Owners benefit from knowing distance to singing neighbor Avoid wasting energy in investigating Avoid unnecessary fights with neighbors Termed ‘ranging’

Examples of ranging Playbacks of song in territories of Carolina wrens –Respond to undegraded song with approach of loudspeaker –Respond to degraded song by singing at a distance Playbacks of song to Kentucky warblers –Fly directly to speaker playing undegraded song –Fly past a speaker playing degraded song

Neighbor versus floater intrusions

Neighbors as ‘dear enemies’ Owners respond less to neighbors then floaters, –During playback experiments at territorial boundaries –Also during territorial intrusions Termed the ‘dear enemy’ phenomenon Consistent with War of Attrition, where relative V/k ratio more easily assessed between neighbors than between floater and owner Also consistent with sequential assessment game, with different behaviors used in different contests

Competition in territorial signals Variation in singing strategies Song matching and song dialects Song type repertoires Eavesdropping Territorial cheaters

Singing strategies in territorial songbirds Some singers have multiple songtypes in repertoires -Mixed-mode singing = ABCDABCDABCD -Bout signing = AAAABBBBCCCCCCDDDD -Infinite = really big number of songtypes

Single songtypes: song dialects Song dialects in white-crowned sparrows Song learning used to match neighbor’s song after juvenile settlement Result in song ‘dialects’-patchwork variation Matching neighbors song appears to be important for obtaining territories Does not preclude individual recognition Dialects more common in stable environments like the tropics Smaller dialects in sedentary populations of WCS Larger, looser dialects in migratory populations

Multiple songtypes: Why different repertoire sizes? Anti-habituation (receiver) Anti-exhaustion (sender) Beau Geste –Multiple songtypes deceive intruders Ranging –New songtypes make neighbor ranging more difficult Escalation by matching

Multiple songtypes: song matching Song matching in song sparrows Neighbors share some, but not all, songs in their repertoire In contests, individuals can -songtype match, -repertoire match, or - fail to match Interactive playbacks show rate of singing, rate of songtype switching, and type of matching all indicate aggressive motivation Intruders with larger repertoires who can match songtypes are more successful gaining territories

Eavesdropping in mammal olfactory marking Permanent olfactory marks are available for eavesdropping by intruders Intruders will competitively overmark signals with own signals Birds will assess intruders by observing interactions with neighbors

Territorial cheating Selection for alternative mating strategies if territories are very limited –noncalling male frogs that intercept females –streaking spawners in fish –Non-harem-holding deer Individuals using alternative strategies avoid territorial signals or mimic non-territory holding sex. –Bluegill sunfish (large territorial males, small sneakers) –Ruff plumage polymorphisms Stategies can be –Condition dependent (Bluegill sunfish) –Genetic polymorphoisms (Ruffs)

Genetic polymorphism in ruff 16% light males, 84% dark males. Dark is dominant to light. Dark males are territorial, only a few mate, while light males follow females. Average mating success is the same.

Any Questions? Contact Tim Wright at