Variation in traits is natural and expected. Individuals differ in their behavior. However, in behavioral ecology, there is a fundamental assumption that individuals behave optimality, with the expectation that individual variation in behavior would be slight. Individuals who were behaving ‘uniquely’ or far out-side of the typical range of variation were regarded as special and complicated our abilities to interpret patterns of behavior.
Animal personality research is rooted in psychology. And when ethologists became enraptured with this idea of animal personality, they often co-opted the terminlogy, testing apparatus, and interpretations provided by these pioneering animal behavior researchers.However, many soon realized that less subjective terminology and interpretations were in order. Around the year 2002, Andy Sih coined the term Behavioral Syndromes and a new Behavioral framework was created. And this Framework took the field by storm. Emphasize how this was a major paradigmExplain plasticity!!How do we measure and study this individual variation?
Initial research efforts focused on findingindividualdifferencesthatmaybelimitingbehavior of largerscales. Domain-specificresearchwasoverlookedearly onBroad behavioral syndromes much more interesting and tehfieldwereperhapsdistracted.Majority of earlyworkwasbroad, hardlyanything on domain. That wasbecausewhenthisparadigmwascreated, the call was to researchwhatmightbelimitingbehavioralplasticity on largerscales. Therefore, broadbehavioral syndrome researchwas en vogue.
Has been widely examined in many different taxa – most notably in great tits, rats, mice, gold fish.
Natural history notes: short life span, monogamous, highly social/colonial, bimodal dispersal
Very high reproductive capacity
Subjects were wild-derived because I wanted to study natural ranges of vole behavior, so no captive-evolved lines.Male and female first through third generation, lab-reared prairie voles, Microtusochrogaster, from Urbana-Champaign, Illinois, were the subjects in these behavioral tests.
I observed a cross-section of behaviors of individual prairie voles in a single context – exploration – but under different conditions (Sihet al. 2004b). Three different exploratory tests were administered. Specific behavioral activities described below were recorded. Each subject completed one or more of the tests and the order in which these three tests were presented were randomized to control for order effects.
I analyzed measures of exploratory behavior for each test individually using Principal Components Analysis (PCA). PCA separates individuals in a sample in terms of a few independent components that represent the underlying dimensions of the data, and determine which dependent variables best characterize each component. Dependent variables can be interpreted according to their loadings on the most important components, which explain how much of the variability is due to those variables. Variables having a high loading on a component are highly correlated to this component. Only components with an eigenvalue larger than 1 and with dependent variables of a loading of 0.80 or higher were retained for interpretation and cross-test comparisons. Those components accounting for only a small part of the total variability and with dependent variables of low loading values were not further analyzed.
Here are my predictions:Behavioral responses in each test would be predictive of behavioral responses in subseqent test apparatus.
Behavioral responses were scored.I expect to find a relationship among the behavioral responses across the tests.
PC1 accounted for _% of the variance of individualbehavior. Therefore the highlighteddependent variables, shownhere best explained the variance in behavioramongindividuals.EM test explanationsThe EM test is test reaction to a mysterious novel environment. When you don’t know what’s out there (like the real world), will you make relatively safe or risky choices. How far are you willing to go?Terminals don’t matter. Arm visits matter most.Upon exiting the start box the subject can see how far each arm is before visiting. Arm 1 is the shortest arm and is closed off and nearest the start box – visiting this arm means the subject is not very exploratory. There are no new areas to explore.Arm 2 is open and is a moderate distance in length (before the subject encounters or sees a wall. Get dimensions. Visiting this arm means the subject has some exploratory tendencies and this arm leads to more areas to exploreArm3 is the longest arm. Get dimensions. Visiting this arm the subject chooses to travel a long distance which leads to more areas to explore. Ranks: higher mean rank indicates a more exploratory and active individualPC 2: LatencyMost behaviors are based latency to depart start box and enter novel environment
I found norelationship among the behavioral responses across tests.
Other studies, all presented since 2008, also failed to find correlations of behaviors in a single domain and some have failed to find broad behavioral syndromes.Examples include:Fox et al. (2009). Mountain chickadees, Poecilegambeli,exploration in a novel room and with novel objectsDingemanse (2008). Sticklebacks, Gasterosteusaculeatusactivity and predator responseAdriaenssens & Johnsson (2008). Brown trout, Salmotruttaindividual behavioral responses and social dominanceMilderman (2008) Starlings, SturnusvulgarisExploration, home range size, movement patterns in the wildSnekseret al. (2008). Damselfish, StegastesleucostictusMale aggression and courtship behavior in natural & artifical settings
Newer research strongly suggests that earlier works that reported behavioral syndromes across multiple context must be re-evaluated.In fact, had I presentedmyresultsbefore 2008 myinterpretationswouldhave been verydifferentbecausetheydid not fit within the existingframework/
The previous framework doesn’t seem appropriate. It leaves to many things left unexplained.
Which is what the 3 remaining chapters of my dissertation was all about.I examined the roots of these 3 different state responses. Each test is an examination of a different state response. Different independent variables explained the reasons for these different behavioral responses in subjects.
Open-field: INTERACTVITY – small litters, older, and males more interactiveNovel Choice: PROACTVITY – all generally proactive, fast moversMaze: ACTIVITY – opposite-sex individuals entered sooner and females traveled fartherMost individual in most families behaved the same. 1/3 to ¼ were the differently behaving ones. Animals in my study might over represent more exploratory types in the wild that are highly active, long distance dispersers
An investigation of behavioral syndromes and individual differences in exploratory behavior in prairie voles, Microtus ochrogaster
An Investigation of Behavioral Syndromes and Individual Differences in Exploratory Behavior of Prairie Voles, Microtusochrogaster<br />Danielle N. Lee<br />Department of Biology<br />University of Missouri<br />
Animal personality research<br />Behavioral Syndromes: Suites of related behaviors across situations or contexts<br />Correlation of behaviors<br />Behavioral plasticity limited<br />Stability<br />Context <br />3<br />
Behavioral Stability<br />State response (Unstable)<br />Behavioral responses that occur in a specific situation (not consistent from one situation to another)<br />Trait response (Stable)<br />Behavioral responses that are consistent from one situation to another<br />Kopp, Voge & Misslin 1999<br />4<br />
Behavioral Stability<br />Behavioral Context<br />Domain specific<br />Correlations of behavior within a single context or functional behavioral category<br />Broad<br />Correlations of behavior across different contexts or functional categories<br />Trait response (Stable)<br />Behavioral responses that are consistent from one situation to another<br />State response (Unstable)<br />Behavioral responses that occur in a specific situation (not consistent from one situation to another)<br />Kopp, Voge & Misslin 1999<br />5<br />
Domain-specific Behavioral Syndrome<br />Correlations of behavior within a single context or functional behavioral category<br />Verbeek et al. 1994<br />Great tits, Parus major<br />Single Context (Exploration)<br />Fast novel room explorers are also fast novel object explorers<br />Image credit: http://www.thebirdsofsussex.co.uk/<br />6<br />
Broad Behavioral Syndrome<br />Correlations of behavior across different contexts or functional categories<br />Wilson et al. 1993<br />Pumpkin-seed sunfish, Lepomis gibbosus<br />Multiple Contexts (Exploration & Social)<br />Bold fish were more likely to approach novel objects and swim in closer proximity to other fish compared to shy fish<br />Image credit: Encyclopaedia Brittanica<br />7<br />
Behavioral Syndromes Hypotheses Hierarchy<br />H0: no correlations of behavior responses across situations or contexts = no syndrome = STATE RESPONSE<br />H1: correlations of behavior response across situations or contexts = syndrome = TRAIT RESPONSE<br />H1a: correlations of behavior responses within one context = DOMAIN SPECIFIC<br />H1b: correlations of behavioral responses across two or more contexts = BROAD <br />9<br />
Experimental Design<br />OBJECTIVE<br /><ul><li>To examine individual variation in a single context – EXPLORATION</li></ul>QUESTIONS<br /><ul><li>Is variation in individual behavior consistent across different tests?
If so, then do these consistent differences in behavior contribute to a behavioral syndrome?</li></ul>10<br />
Chapters<br />Examining individual differences and exploratory behavioral responses in prairie voles: In search of a behavioral syndrome.<br />Open-field and with novel objects<br />Novel choice apparatus<br />Complex maze apparatus<br />11<br />
Domain specific behavioral syndromes: Hypothesis<br />H0: behavior responses across different tests are not correlated<br />H1: behavior responses across different tests are correlated<br />Exploratory Behavior<br />12<br />
Exploratory behavior<br />Tendency to investigate novel environments and/or objects<br />Functional behavior category widely studied<br />Spontaneous behavioral responses<br />Indicates how animals might gather information about local environment and its resources<br />13<br />
Open-field with Novel Objects Test<br />1 m<br />1 m<br />Dependent Variables:<br /> Latency<br /> Time in the novel environment <br /> Time with novel objects<br /> Returns<br /> Instantaneous scan of location<br />a) Total squares<br />b) Visits to each section<br />18<br />
500 mm<br />Novel Choice Test<br />Dependent Variables:<br /> Latency<br /> Time to reach the first terminal<br /> Time to reach the second terminal<br /> Total time to complete the test<br />19<br />
Complex Maze Test<br />396 cm<br />610 cm<br />Dependent Variables:<br /> Latency<br /> Number of times each arm was entered<br /> Returns to the start box<br />20<br />
Statistical Analysis<br />PCA analysis for each exploratory test<br />Identified the key dependent variables<br />PC1 accounted for the highest % of variance<br />All dependent variables that scored above .80 that explained PC1 – called Key Dependent Variables<br />Ranked each Key DV<br />Computed the overall Exploratory Score for each individual in each test<br /><ul><li>Correlate Exploratory Scores across tests</li></ul>21<br />
Overall Predictions<br />less exploratory more exploratory<br /><ul><li>Lower latency
More time spent interacting with novel stimuli
Discussion<br />Behavior in one situation did not correlate or predict behavior in a different situation<br />Behavioral responses are distinct and stand alone<br />Animals interpret context differently<br />Null Hypothesis in hierarchy supported<br />29<br />
Behavioral Syndromes Hypotheses Hierarchy<br />H0: no correlations of behavior responses across situations or contexts = no syndrome = STATE RESPONSE<br />H1: correlations of behavior response across situations or contexts = syndrome = TRAIT RESPONSE<br />H1a: correlations of behavior responses within one context = DOMAIN SPECIFIC<br />H1b: correlations of behavioral responses across two or more contexts = BROAD <br />30<br />
Calling behavior of roosters<br />Nelson et al. (2008). Roosters, Gallus gallus domesticus<br /> Behavioral syndromes in stable social groups: an artifact of external constraints? Ethology, 114: 1154-1165<br />31<br />
Calling behavior of roosters<br />Nelson et al. (2008). Roosters, Gallus gallus domesticus<br /> Behavioral syndromes in stable social groups: an artifact of external constraints? Ethology, 114: 1154-1165<br />32<br />
Fox et al. (2009). Mountain chickadees, Poecilegambeli,<br />exploration in a novel room and with novel objects<br />Dingemanse (2008). Sticklebacks<br />Adriaenssens & Johnsson (2008). Brown trout<br />Milderman (2008) Starlings<br />Snekseret al. (2008). Damselfish<br />33<br />Other studies<br />
Historical contextStudies in Psychology<br />Context: Learning ability and cognition in lab rats in labyrinth mazes<br />No two trials of the exact same test correlated with any degree of reliability. Tolman (1924)<br />Learning ability is specific to the apparatus. <br /> Tryon (1940), Searle (1949)<br />34<br />
<ul><li>No single test can serve as a proxy for an entire domain
Behavioral syndromes may not be a general attribute, as previously proposed
Demonstrate domain-specific correlations first, then broad</li></ul>Cautions<br />35<br />
Heuristic Value of Behavioral Syndrome Framework<br />H0: no correlations of behavior responses across situations or contexts = no syndrome = STATE RESPONSE<br /><ul><li>State responses still reveal high individual variation</li></ul>37<br />
Chapters 2, 3, 4<br />OBJECTIVE<br />To examine roots of individual variation in exploratory behavior responses of prairie voles<br />QUESTION<br />What are the shared characteristics of individuals who show similar exploratory tendencies?<br />HYPOTHESES<br />1: subjects from similar early social environments behave similarly<br />2: behavioral similarity is best explained by age or sex<br />38<br />
Overall Results<br />Open-field<br />Litter size matters, litter sex ratio contributes<br />Sex differences are consistently different, but n.s.<br />Age matters<br />Novel Choice<br />No differences for any of the comparisons<br />Complex maze<br />Litter sex ratio matters<br />Sex differences are consistently different, but n.s.<br />39<br />
Conclusions<br />Individual variation in behavior was observed across all tests<br />Individual behavioral plasticity is high in voles<br />Plasticity might be beneficial for this species<br />40<br />
ACKNOWLEDGEMENTS<br />David Chisholm, Lyndell Bade, Tom Kent<br />Jasmine Clayton, Robert Clayton, Allison Clayton, MeitalLaks, Dominique Craven, Evyn Craven, AnnaLynn Harris, Dianne Voorhis, Angel Scarfino, Alecia Hoyt<br />Stan Braude, Karen Norberg, Lon Wilkens, George Taylor, Godfrey Bourne, Patrick Osborne, Maryann Hempenn<br />Elizabeth Congdon, GenaSbeglia, Laura Kent, Javier Hernandez, Tim Lescher<br />Darcella Craven, Piia Hanson<br />Family, Friends, Blog readers and SM supporters<br />NSF <br />AGEP Fellowship<br />Doctoral Dissertation Improvement Grant<br />G-K12 MO-STEP Fellowship<br />American Society of Mammalogists GIAR<br />TWA Scholarship for Environmental Science <br />UMSL – The Graduate School BRIDGE Fellowship<br />41<br />