|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
1 EXPLORATORY BEHAVIOR OF FOREST BIRDS AND THE INFLUENCE OF TUFTED TITMOUSE ( BAEOLOPHUS BICOLOR ) ANTI PREDATOR V O CAL IZATIONS By PING HUANG A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FUL FILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2010
2 2010 Ping Huang
3 To my family
4 ACKNOWLEDGMENTS I would like to thank my advisor, Dr. Kathryn Sieving, for her guidance, patience, and all the sup port she provided for me. I thank my committee members, Dr. Robert Fletcher and Dr. Colette St. Mary, for their advice throughout the duration of my study. I thank Dr. Michael Avery for allowing me to use the USDA/ APHIS/ WS/ NWRC Florida Field Station to conduct my research and Kandy Keacher for helping me find my way working there. I would like to acknowledge the Ordway Swisher Biological Station for letting me use the property during my study. My sincere thanks go to my dearest fellow colleagues in Dr. S ieving s lab, Willandia Chaves Didier, Eduardo Silva, Jackson Frechette, Fangyuan Hua, Rosalyn Johnson and Chelsea Heatherington, for assisting me in set ting up the experimental cage, helping me catch my study species, generously allowing me to use the ir backyards as studying place s, and providing great insights to my study. I thank my brother Chun Huang, my soon to be sister in la w Machel Malay and a lot of friends, for helping me get used to the life in a foreign country quickly and giving me a p lace to relax Most of all, my tremendous gratitude goes to my parents, Wun Taw Huang and Hui Chen J uan, for giving me their constant and complete supports.
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 7 LIST OF FIGURES ................................ ................................ ................................ .......... 8 ABSTRACT ................................ ................................ ................................ ..................... 9 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 12 2 INTERSPECIFIC DIFFER ENCES IN EXPLORATORY BEHAVIORS AMONG FOREST BIRDS ................................ ................................ ................................ ..... 15 In troduction ................................ ................................ ................................ ............. 15 Intra specific Variation in Exploratory Behavior ................................ ................ 15 Neophobia and Interspecific Assessments of Exploration and Per sonality ...... 18 Interspecific Characterization of Exploratory Behavior ................................ ..... 20 Study System ................................ ................................ ................................ .......... 21 Research Design ................................ ................................ ................................ .... 23 Objectives and Experimental Design ................................ ................................ 23 Predictions ................................ ................................ ................................ ........ 24 Methods ................................ ................................ ................................ .................. 25 Capture and Handling ................................ ................................ ...................... 25 Testing protocol ................................ ................................ ................................ 26 Measuring Exploratory Behaviors ................................ ................................ ..... 27 Statistical Analysis ................................ ................................ ............................ 28 Results ................................ ................................ ................................ .................... 29 Interspecific Comparison ................................ ................................ .................. 29 Foraging Guild Comparison ................................ ................................ .............. 30 Social Group (Flocking Behavior) Comparison ................................ ................. 30 The Performance of Exploratory Score versus Component Behaviors ............ 31 Discussion ................................ ................................ ................................ .............. 31 Measuring Exploratory Behavior ................................ ................................ ...... 31 Social Roles and Exploratory Behavior ................................ ............................ 34 Behavior Syndromes, Neophobia and Exploration ................................ ........... 37 Flexibility, Plasticity, or Behavioral Diversity in a Changing World ................... 39 3 TUFTED TITMOUSE ( BAEOLOPHUS BICOLOR ) A NTI PREDATOR VOCALIZATIONS ALTER HETEROSPECIFIC EXPLO RATORY BEHAVIOR ....... 57 Introduction ................................ ................................ ................................ ............. 57 Animal Ecology and Information driven De cision making ................................ 57
6 Social Information and Its Influences ................................ ................................ 57 Using Social Information to Navigate Landscapes of Fear ............................... 59 Study System ................................ ................................ ................................ .......... 61 Research Design ................................ ................................ ................................ .... 63 Objectives and Experimental Design ................................ ................................ 63 Predictions ................................ ................................ ................................ ........ 64 Methods ................................ ................................ ................................ .................. 65 Capture and Handling ................................ ................................ ...................... 65 Playback Treatments ................................ ................................ ........................ 65 Testing Protocol and Exploratory Behavior Measurements .............................. 66 Statistical Analysis ................................ ................................ ............................ 67 Results ................................ ................................ ................................ .................... 68 Temporal Patterns in Movement Types ................................ ............................ 68 Temporal Patterns in Thoroughness of Exploration ................................ ......... 69 Temporal Patterns in Scanning Behavior ................................ ......................... 69 Discussion ................................ ................................ ................................ .............. 70 Interspecific Social Cues with Known Information Content ............................... 70 Exploratory Behavior under Socially Conveyed Threat of Predation ................ 71 Information, Spatial Behavior and Future Implications ................................ ..... 73 4 CONCLUSION ................................ ................................ ................................ ........ 81 LIST OF REFERENCES ................................ ................................ ............................... 87 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 101
7 LIST OF TABLES Table page 2 1 List of components of exploratory behavior their definitions / measures and how data were transformed for analysis. ................................ ............................ 48 2 2 Mean and coefficients of variation (CV) of 8 component exploratory behaviors for 6 species. ................................ ................................ ...................... 49 2 3 Comparison among species and between groups based on each component of exploratory behavior ................................ ................................ ....................... 50 2 4 Results b Correlation Analysis between component behaviors and body size. ................................ ................................ .................... 51 2 5 Tukey Test results for between species comparison WEVI = white eyed vireo, CACH = Carolina chickadee TUTI = tufted titmouse, CAWR = Carolina wren, NOCA = Northern cardinal, and EATO = Eastern towhee. MD = Mean Difference = Spp 1 Spp 2. ................................ ................................ ............... 52 2 6 Factor loadings of the 8 component behaviors o n the 2 principal component s. Eigenvalues and amount of variance explained by the respective components are given at the bottom. MAN (C) OVA results are shown in the bottom two panels. ................................ ................................ ............................. 54 2 7 St ructure coefficients/ discriminant loadings of the 8 component behaviors on the function(s) from canonical discriminant function analysis of the different group comparisons Eigenvalues and amount of variance explained by the respective components are at t he bottom. ................................ .......................... 5 5 2 8 Foraging behavior of 6 species based on different references. .......................... 56 3 1 Prediction of influence of 4 different vo calization trials through time. Sign means no changes at all, means decreasing, and means increasing. The number of signs indicates the strength of the influence. .............................. 79 3 2 Factor loadings of the 8 component behaviors on the 3 principal components after varimax rotation. Eigenvalues and amount of variance explained by the respective components are given at the bottom. MANOVA results from these three principal components for different compari sons are shown in the bottom panel. ................................ ................................ ................................ ...... 80
8 LIST OF FIGURES Figure page 2 1 T he design of the testing cage (novel environment). T he releasing cage, cove red by camouflage fabric is in the lower left corner The tripod ( at right ) held the video camera set to record all behavior s inside the cage. One perch (of 5: see Figure 2 2) is visible inside the cage. ................................ .................. 42 2 2 The assignment of volume within the cage into 27 cubes (9 ground calls 3 st r ata) for exploration by test subject s Perching substrates (sticks) intercepted a total of 21 of the 27 total volume cells. ................................ .......... 43 2 3 R esults of the d iscriminant f unction a nalysis for (A) species (1 st two discriminant functions used; WEVI = white eyed vireo, CACH = Carolina chickadee, TUTI = tufted titmouse, CAWR = Carolina wren, NOCA = Northern cardi nal, and EATO = Eastern towhee), (B) foraging guilds, and (C) flocking groups showing, for each, the two component behaviors with the highest weights on the single discriminant function generated. The factor loadings for DF analyses are listed in Table 2 4 ................................ ................ 44 2 4 Scatter plots of mean coefficients of variation (CV) across all 8 component behaviors vs. (A) max. latitudinal range (degrees) in geographical distribution, (B) max. longitudinal rang e (degrees) in geographical distribution, and (C) ratio of insect to plant food types in the diet of each species. In Panels A and B, the assumption is that generalists are on the right side, and in Panel C, they are on the left side. WEVI = white eyed vir eo, CACH = Carolina chickadee, TUTI = tufted titmouse, CAWR = Carolina wren, NOCA = Northern cardinal, and EATO = Eastern towhee. ................................ ............... 46 3 1 Bar graphs showing changes in 8 components of explorato ry behaviors over the three trial periods (before, during and after broadcasting) for each treatment (control, low risk mobbing, high risk mobbing, and seet call): (A) flight number, (B) hop number, (C) prop. of perches explored, (D) prop. of ground explor ed, (E) active scan number, (F) average active scanning time, (G) still scan number, and (H) average still scanning time. Bars show mean values with 95% confidence intervals. ................................ ................................ 76 3 2 The relati onship among social information, animal behavioral ecology and functional connectivity. ................................ ................................ ....................... 78 4 1 Linking knowledge from this study to future researches and application goals. 86
9 Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for th e Degree o f Master of Science EXPLORATORY BEHAVIOR OF FOREST BIRDS AND THE INFLUENCE OF TU FTED TITMOUSE ( BAEOLOPHUS BICOLOR ) ANTI PREDATOR V O CALIZATIONS By Ping Huang August 2010 Chair: Kathryn E. Sieving Major: Wildlife Ecology and Conservation Exploratory behavior is the gathering of information about objects or other aspects of the envir onment not reflected in the satisfaction of immediate needs Exploratory behavior is heritable and repeatable, and has a strong connection with animal personality, or beha vior syndrome s Th e first goal of my study wa s to increase understanding of how bird species differ in their exploratory behavior based on their natural history c haracteristics (e.g., foraging guilds and social roles ; Chapter 2). Additionally, I applied a novel multivariate approach to characterizing interspecific exploratory behavior not previously used, that increases the level o f detail with which exploratory behavior can be analyzed Second, I tested whether and how he ter o specific social information concerning predation risk may influence exploratory behavior (Chapter 3) In the first experiment, I recorded the explo ratory behavior exhibited by 6 diff erent forest bird species in 10 minute trials inside a novel testing cage. Different species showed different patterns across eight component behaviors of exploration that I identified a nd measured on each individual. Interspecific differences were generally related to natural histor y characteristic s (foraging guild and social roles in winter
10 foraging flocks) In discriminant function analysis foraging guild ( species grouped into either ins ectivore lower canopy / shrub gleaner vs. omnivore ground gleaner ) allowed a 72.8% correct classification of the species ; insectivores exhibited longer active scanning time s while omnivores tended to explore a higher proportion of the ground inside the cag e. When species were classified as either winter foraging flock participants or non flocking species discriminant function analysis achieved 96.3% correct class ification. Species attend ing mixed species foraging flocks in winter used flight in exploration whereas non flocking species stay ed on the ground to explore. A multi variate approach based on the 8 component behaviors I identified wa s more useful in describing interspecific differences in exploratory behavior than a single exploratory score that has been used most commonly to characterize intraspecific variation in exploratory behavior In a second experiment, I tested whether and how social information cues (predation threat encoded in anti predator vocalizations of the t ufted t itmouse Baeolophus bicolor ) influence the exploratory behavior of Northern c ardinal s ( Cardinalis cardinalis ) Both species are common and widely distributed in Eastern North America, and titmouse anti predator calls encode highly specific types of information concerning pred ation risk, that cause anti predator behaviors in a variety of taxa. I determined socially derived information (titmouse anti predator calls) encoding known predation information influenced exploration and movement in the Northern cardinal. Cardinals inter preted the anti predator signals encoded in the calls I broadcast to them correctly and responded accordingly via adjusting their activities in the novel environmental cage. When the predation threat level encoded in titmouse vocalizations increase d, cardi nals exhibited less active movements but increased their scanning frequency and time And,
11 even though I found that c ardinal s rarely exhibit still s canning during normal exploration, broadcasts of social cues from titmice indicating high predation risk sti mulate d cardinals to exhibit significant levels of this behavior Thus, titmouse calls that conveyed higher risks of predation caused cardinals to decrease their overall activity levels (exploration and movement). Since exploratory behavior is genetically related to natal dispe rsal ability of birds the results of this study could have implications for the study of functional connectivity of landscapes at large scales Given that the degree of exploratory behavior exhibited in a small cage, like the one I u sed, is directly linked to dispersal ability (more active explorers disperse further), my results suggest that social information use, in addition to species traits, may also influence large scale movements of birds. Physical landscapes with known habitat connectivity of bird species will contain bird communities that produce labile landscape of information. Through invisible to the eye, I propose, based on this work, that social information, and the environmental conditions it reflects, may h ave importan t influences on the functional connec tivity of landscape s for forest birds not reflected in physical habitat configurations.
12 CHAPTER 1 INTRODUCTION Exploratory behavior i s the gathering of information about objects or other aspects of the environment that doe s not satisfy immediate needs ; and it varies with sex, age, social status ecolo gical traits, and experience of the individua l Exploratory behavior is linked to dispers al ability, animal personality ( or behavior syndrome ) and various social roles ( e.g ., dominance; Greenberg 2003) I nformation gathering is of crucial importance in managing uncertainty in th e everyday liv e s of organisms. I nformation is gained from personal experience through explorin g the environment directly ; or via the use of social i nformation provided by other individuals ( advertently or inadvertently ) in decision making (Danchin et al. 2008) Since a nimals rel y on information to make decisions, learn ing more about the landscape of information availab le to animals at the landscape s cale, and how animals filter and then use the information in decision making regarding landscape scale behaviors is therefore necessary (Fletcher and Sieving 2010) I sought to understand, first, how exploratory behavior at a small spatial scale (in a no vel cage environment) correlated with certain natural history traits of forest birds. Since exploratory behavior is highly correlated with movement ability at larger scales th is information could prove useful in understanding species differences in suscep t ibility to forest fragmentation (Dingemanse et al. 2003). Second, I also sought to understand the influence of heterospecific social information (in the form of anti predator alarm calls) on the exploratory behavior of one species. Social information us e may strongly influence spatial behavior at lar ger scales ( Fletcher, 2009; Fletcher and Sieving 2010 ; Sieving et al. 2004 ). Therefore, I sought to test how social information influences
13 exploratory behavior at a small scale (novel cage environment) in order to guide investigations of information use and behavioral connectivity at larger scales. Specifically, in Chapter 2 I addressed the following: (1) how exploratory behavior in a novel environment generally varies among 6 different bird species using a detailed breakdown of exploratory behavior into 8 component behaviors I identified; (2) major species traits correlated with exploratory behavior, namely foraging guild and behavior syndrome (as expressed in social roles exhibited in winter foraging flock s), and (3) how my, more detailed multivariate approach to describing exploratory behavior compared to the standard exploratory score currently in use by others. I established that the 6 species tested exhibited different exploratory behavior in a novel ca ge environment Foraging guild was loosely correlated with exploratory behavior, but whether a species commonly participated in winter foraging flocks or not was tightly correlated with exploratory behavior traits (discriminant function analysis). Using a multi variate approach provided a more complete picture for describing exploratory behavior and comparing it across multiple species In C hapter 3, I explored how social information use may influence exploratory behavior. Northern c ardinals experienced voc alization playback trial s in the same novel cage environment as in Chapter 2. I presented 4 experimental playback treatments to cardinals, including t ufted t itmo use alarm calls (given to different live predators under controlled conditions; Hetrick 2006; Sieving et al. 2009) t hat encoded known but distinct p redation threats. I found that c ardinals interpreted the encoded information appropriately and adjusted their exploratory behavior in ways that would decrease their susceptibility to predation threats represented in the playback treatments A s perceived
14 risk of predation increased with the degree of danger encoded in titmouse calls, c ardinals reduced overall activity levels, expressed in active component s of exploratory behavior but increased overall v igilance related components. In particular, treatments conveying the highest risk of predation e licited significant amounts of still scanning ; a behavior not commonly observed in cardinals (Chapter 2). My results reveal ed in greater detail than previou sly examined, how exploratory behavior varies among different species, and how social information about predation risk can bring about changes in exploratory behavior at a small scale Given that exploratory behavior is highly correlated with d ispersal abi lity ( Dingemanse et al., 2003; Johnson 1988 ) my findings suggest expectations for how species traits and information use can together affect spati al behavior at larger scales For example, the functional connectivity of landscapes, or portions of them, w here social information conveys a high degree of predation risk, movement by species that can use the information to mediate their own risk of predation may be slower in such areas.
15 CHAPTER 2 INTERSPECIFIC DIFFER ENCES IN EXPLORATORY BEHAVIORS AMONG FOREST BIRDS Introduction Uncertainty poses significant challenges in animal decision making and carries important fitness consequences for wildlife in rapidly changing human dominated landscapes. For an animal to best exploit potential opportunities and avoid da ngers, it must anticipate changeable environments, including variation in distribution and kinds of resources and risks. Information gathering is crucial for managing uncertainty and changes affecting critical activities including: mate choice (Candolin 2003 ; Drullion and Dubois 2008 ) ; habitat selection ( Arlt and Part, 2008; Forsman et al., 2009; Parejo et al. 2007 ) ; assessing food availability ( Fletcher and Miller, 2008; Valone and Templeton 2002 ) ; whether to disperse ( Enfjall and Leimar, 2009; Ward, 2 005 ) and how far to go ( Long et al. 2005 ) ; and choosing safe paths when moving ( Roche et al. 1999 ). I n movement decisions, information about habitat and resource distributions and about predator location and predation risk is actively sought and used by vertebrates often via active exploration or investigation to obtain information (Blumstein 1999 ; Schmidt et al. 2008 ). Intra specific Variation in Exploratory Behavior Explorat ion, a nd related exploratory behavior is an important way for animal s to obt ain informatio n about their environments. Exploratory behavior is the gathering of information about objects or other aspects of the environment that doe s not satisfy immediate needs (Dall et al. 2005). Exploratory behavior has been characterized for a nu mber of organisms and is particularly well studied
16 in certain vertebrate animals including E astern c hipmunk ( Tamias striatus ; Wolfe 1969), g oldfish ( Carassius Auratus ; Kleerekoper et al. 1974), g uppy ( Poecilia reticulate ; Budaev 1997) rat ( Rattus norve gicus ; Smith et al., 2009; Raud et al. 2007 ), and a number of bird species. Major correlates of intra specific differences in e xploratory behavior include sex, age, social status e nvironmental conditions experience of the individual personality, and di spersal capability ( Arakawa 2005 ; Lodewijckx 1984 ; Mettke Hoffmann et al. 2002 ; Mettke Hoffman et al. 2006). Exploratory behavior can be highly variable among individuals of the same species depending on the combined effects of these various correlated conditions ( e.g., Dingemanse and de Goede, 2004; Dingemanse et al. 2002) Dispersal success is one potential consequence of exploratory movement because e xploration allows an animal to compare its possible success elsewhere with that of its present loca tion (Johnson 1988). I n g reat t its exploratory behavio r is heritable and has a strong positive relationship with natal disp ersal ability Daughters of males with higher exploratory behavior scores (exhibit more flights and hops in novel environments) ten d to disperse further (Dingenman s e et al. 2003). But exploratory behavior in adult birds can also be influence d by environmental exposures during the exploratory juvenile phase of development (Hopf 1985), and this has also been linked to differential dis persal success in forest birds ( C. Cornelius, personal communication ). Thus species with heritable but developmentally flexible exploratory and dispersal capabilities may adapt better to disturbed landscapes with complex patterns of habitat fragmentation t hat naturally increase the isolation of habitat patches for forest birds (Desrocher
17 and Hannon 1997; Greenberg and Mettke Hofmann, 2001; Grubb and Bronson, 2001; Grubb and Doherty 1999 ). Exploratory behavior is also strong ly connect ed with animal person ality, or behavior syndrome s A behavioral syndrome is a suite of correlated behaviors expressed either within a given behavioral context or across different contexts (Sih et al. 2004 ). Established examples of behavior syndromes include activity (Sih et al. 2003) and shyness or boldness ( Bell and Sih, 2007; Both et al., 2005 ). Individuals with different personalities can perform differently under different types of stress, and the type of personality is usually highly correlated with an ind i v i exploratory skills. In b luegill s unfish ( Lepomis macrochirus ), bolder individuals exhibit higher scoring exploratory behavior (Wilson and Godin 2009). In g reat t its, individuals with higher exploratory scores recover from being startled sooner and thus t heir latenc ies to return to a feeding table are shorter and slow explo ring tits exposed to intruders spend more time in agonistic displays and take longer to attack (Carere et al. 200 5 ). This suggests exploratory behavior may be an indicator of personali risk taking behavior ( van Oers et al. 2005 ) Finally breeding pairs with opposite exploratory scores (e.g., fast exploring male and slow exploring female) exhibit the highest reproductive success for g reat t its (Both et al. 2005) Authors speculate the contrasting parental personalities may combine the best of both worlds for reproduction. The fast explorer more effectively defends a high quality territory (Dingemanse and de Goede 2004) and the slow exploring parent can respond
18 mor e quickly to food distribution change due to more continuous exploration of alternative feeding options (Verbeek et al. 1994) In sum, exploratory behaviors documented in novel caged environments (such as flight and hop numbers ) have repeatedly been high l y correlated with personality ( Dingemanse and Rale 2005; Dingemanse et al. 200 4 ) and, in turn, with consequential vital responses of interest. Links between exploratory behavior and large scale spatial behaviors, such as natal dispersal and gap crossin g in fragmented landscapes, are of increasing interest in conservation biology. Recent definitions of landscape connectivity for animals are primarily behavior effects of landscape c onfiguration and species behavioral reactions to landscape elements (e.g., matrix and corridors between habitat patches) and assumes that different species will move differently in the same landscapes (Belisle 2005). Thus, linking interspecific variation in exploratory score to landscape level variation in distributions is a future goal of this line of work. However, recent exploratory behavior analyses only address intraspecific variation. Our best understanding of personality, spat ial movements, and links to potential conservation implications derives from work on neophobia in wild species. Neophobia and Interspecific Assessments of Exploration and Personality Neophobia is the fear of novel stimuli in the immediate environment ( Gr eenberg 1983 ), and is highly variable in its intensity among bird species. For example, when 6 warbler species a re presented with a positive, but novel foraging stimu lus (a cup full of mealworms), b ay breasted w arblers ( Dendroica
19 castanea ) a re the least h esita nt and c hestnut sided w arbler s ( Dendroica pensylvanica ) show significantly greater hesitation ( latency to approach; Greenberg 1983). As with exploratory behavior variation within species, t he difference in neophobic responses among species has been related to bold vs. shy personality and fast vs. slow exploration that, in turn, have been linked to different life histories and habitat associations (Greenberg 2003) F or example, resident Sardinian w arblers ( Sylvia melanocephala ) are less neophobic and more explorat ory than migrant g arden w arbler s ( Sylvia borin ; Mettke Hofmann et al. 2005). Parrot species that utilize complex habitats or that feed on buds show the shortest latencies to approach a novel subject in explorat ory test s (Mettke Hofmann et al 2002). Those findings suggest that exploratory behaviors (and the ir correlated personalities, such as neophobias) reflect a complex of adaptation s to different environmental conditions that vary between species (Greenberg 2003) More importantly, both exploratory scores in novel environments (within species) and neophobic latencies (between species) vary similarly with personality and risk taking behaviors. That is bold species and individuals take greater risks in aggression and food finding ( van Oers et al. 2005 ); suggesting that bolder species (like individuals) should be more active explorers, and potentially better dispersers. Interspecific comparisons of exploratory behavior are needed to make these links. One potential limiting factor in assessi ng exploratory behavior across species and ecosystems or landscapes is the very simplicity of current measures of exploration. D ifferent species move and explore in unique ways and the
20 current measurement standard for exploratory behavior (the combined n umber of flights and hops within 2 10 minutes in a novel cage environment) may not be universally applicable to interspecific comparisons. Animals can explore (gather information) using diverse cues (sounds, scents, visual and tactile stimuli) that can be gathered via active behaviors (hopping, flying, looking), but also via still scanning or listening for cues and signals that may come from further away than short hops and flights would indicate ( Fletcher et al., 2009; Grim 2008; Martin and Lopez 2009). A major goal of this work is to clarify measurement of interspecific exploratory behavior in order to facilitate comparative work. Interspecific Characterization of Exploratory Behavior I n this study, I t ook a systematic approach to characterizing inters pecific differences in exploratory behavior (in a novel cage environment) with the goal of identifying species tr aits that are most highly correlated with, both, individual components an d synthetic (multivariate) measure s of exploratory behavior. I selecte d a variety of bird species from different taxonomic groups that vary markedly in at least two ways relevant to spatial exploration or personality ; (1) foraging guild s (diet breadth, and prey search, pursuit, and capture techniques) and (2) social roles wi th respect to winter flocking. Since foraging entails intense spatial exploration in birds ( Mettke Hofmann et al. 2002 ), comparison of species exhibiting different foraging modes should ensure some variation in style and propensity to explore Similarly, I selected species with different social roles related to winter foraging flocks as a surrogate for variation in personality syndromes under the following reasoning Flock leaders are often characterized as vocal, intraspec ifically gregarious, bold ( social ly dominant ) and more vigilant
21 toward predators than flock satellites (species that follow the leader) and non flocking species ( Farley et al., 2008; Hutto 1994 ; Munn and Terborgh 1979 ). Using these two groupings (foraging guilds and flocking behavior) I could address whether interspecific variation in exploratory behavior could be used to predict larger ecological patterns in bird communities. My specific goals with this work are to help standardize ways to characterize the exploratory behavior of diver se species, and to increase u nderstanding of the complex of factors that influence exploratory behavior of different species I am particularly interested in exploratory behavior, and its causes and correlates, as an indicator of spatial behavior more bro adly for informing the conservation of behaviorally complex species in a rapidly changing world. Study System I selected 6 understory forest bird s pecies common in the South eastern United States including the tufted titmouse, w hite eyed v ireo ( V ireo gri seus ) Carolina chickadee ( Poecile carolinensis ), Carolina wren ( Thryothorus ludovicianus ), Northern cardinal and an endemic yellow eyed sub species of the Eastern towhee ( Pipilo erythrophthalmus ). All six species are permanent residents, occupying year r ound territories. A large proportion of the forest bird community participates at some level in winter foraging flocks including the titmouse, vireo, and chickadee (above, the other three study species do not; Farley et al. 2008) In North central Flori da, titmouse is a year round resident species, and in winter is reliably associated with mixed species flocks, serving as the flock leader (nuclear species; Farley et al. 2008 ; Waite and Grubb 1988 ). Titmice have bold
22 personalities and dominant social st atus. Th ey facilitate flock formation, initiate and guide flock movement ( Farley et al. 2008 ; Greenberg 2000 ), and reduce predation risk and food stres s for satellite species (Dolby and Grubb 1998 ; Ragusa Netto 2002 ; Sridhat et al. 2009). A central me chanism underlying the benefit to satellites of following titmice is the amount and precision of threat appropriate signals encoded in their alarm calls (Sieving et al. 20 10 ). Species that participate in flocks, even if they do not lead them, may exhibit different personality types than s pecies that do not join flocks. For example, flock ( Jullien and Thiollay 1998 ; Munn and Terborgh 1979 ) suggesting that flocking species m ay use space differently than non flocking species. Further, flock participants gain benefits through coordination including reduced predation risk and potentially increase d foraging efficiency ( King and Rappole 2000 ; Powell 1985 ; Pulliam 1973 ; Sridhar et al. 2009). I n order to keep up with flock movements, deal with agonistic encounters that occur at higher frequencies in mixed flocks ( Griffin et al. 2005 ; Morse 1977 ; Terborgh 1990 ), and maintain awareness of visual and vocal cues from a variety of specie s, flocking species personality syndromes are likely to be distinct from solitary foragers My 6 bird species represent ed variation in foraging guild s and in both level of participation in winter foraging flocks. These six species fall into two foraging guilds. Titmouse, chickadee, vireo and wren classify as insectivore s (& lower canopy/ shrub gleaner s), whereas, cardinal and towhee are omnivore s (& ground gleaners; de Graaf et al. 1985). Thre e species are flock participants: tufted
23 t itmice (nuclear), wh ite eyed vireo and Carolina chickadee (satellites) A nd 3 species do not follow foraging flocks : Carolina wren, Northern cardinal and Eastern towhee. Chickadees (in the same family as titmice; Paridae), have the same complex ant i predator communication sy stem as t itmice (Freeberg et al. 2003, Sieving et al. 20 10 ) but they are subordinate to t itmice when the two species co occur in flocks (Cimprich and Grubb 1994) Research Design Objectives and Experimental Design Th e objectives of this study are thre e fold : (1) to understand how exploratory behavior in a novel environment generally varies among different bird species, (2) to test (via correlation) for major species traits linked to exploratory behavior, namely foraging guild and personality (as expres sed in social roles in winter foraging flocks), and (3) to compare my measures of exploratory behavior to the standard exploratory score used by others (the accumulative number of flights and hops inside the novel environment). To address species and group differences in exploratory behavior, individual birds of each species were captured from the wild and released into a novel cage environment Birds were video taped for 1 0 minutes and then released. Using previous work on exploratory behavior of birds in similar novel environmental settings, and some measures that I developed during preliminary examination of videos of my test subjects, I quantified how the different species moved around in, and visually inspected, the space within the cage. I then examine d how each of 8 different component behaviors, associated with moving (hops, flights) and visual scanning, varied within and between the 6 species, 2
24 foraging guilds (insectivore/foliage gleaner and omnivore/ground gleaners) and 2 social groups (flocking and non flocking) using univariate and multivariate only hop and flight numbers, so that my assessments of interspecific exploratory behavior based on 8 components could be compared directly with previous intraspecific measures of expl oratory beh avior Predictions Under objective (1) I had no specific predictions, just that each species would exhibit unique exploratory behavior with respect to variation within each of the 9 component measures of exploratory behavior, and across synthetic (multivariate) measures. But in general, based on general findings from similar studies ( Dongemanse et al. 2003 ), I assumed the following: that high levels of movement (hopping and flying) a nd of visual scanning within the cage reflect high exploratory behavior, and also that if a bird actually visited a higher proportion of the volume of space within the cage then that also reflected high levels of exploration. For objective (2), foraging a nd social group comparisons, I predicted that the ground omnivores might not scan as much as the insectivores (that need to locate active prey), and that the activity of the two groups would be different in kind (above ground insectivores would fly more an d the ground foragers hop more). I also predicted that the socially dominant t ufted t itmouse would be the most vigilant (high scanning rates) and (along with) Carolina c hickadees and the other satellite species ( w hite eyed v ireo) would rank high in both e xploration thoroughness and activity levels N on flocking species (Northern c ardinal Eastern
25 t owhee, and Carolina w ren ) should be less exploratory S atellite species might explore less vigorousl y than non flocking species; if they rely on the social group to watch for predators and rely fully on the vigilant t ufted t itmouse for scanning But I expected that the ability to follow actively moving flocks would require satellites to exhibit similarly high levels of exploratory behavior as the titmice ( Griffin e t al. 2005 ). Finally, for object ive (3), while I predicted that the classic exploratory scores (hop number plus flight number ) would generally be correlated with the more comprehensive measures I use d above, I also expected that different component beha viors would be more or less useful for characterizing each species or group. I expected that the classic exploratory score may not be very useful, by itself, for describing complex interspecific differences in exploratory behavior. Methods Capture and Han dling I conducted field tests between May 2009 and January 2010 in Alachua and Putnam Counties in F lorida. Eighty on e individuals of the target species were captured either by feeder trap (wire boxes with drop doors placed on platform feeders) or mist net s Captured individual were then put into a small wire cage with a trap door (releasing cage), and that was placed inside the larger 3 3 2 m test cage. The releasing cage was covered with camouflage screen to provide semi darkness for calming the test bird while blocking the view of the novel exploratory environment. Using a string attached to the trap door, birds could be released into the testing cage remotely. The testing process and bird captures
26 took place at several different locations in Florida including the Ordway Swisher Biological Station in Melrose (Putman County) the USDA/APHIS/WS/NWRC Florida Field Station (United States Department of Agriculture, Animal & Plant Health Inspection Service, Wildlife Services, National Wildlife Research Cen te r; USDA lab) in Gainesville and 4 private yards in the city of Gainesville ( all trap sites were at least 2 Km apart). The design of the test cage (novel environment) was modified from Dingemanse et al. ( 2002) and constructed of 3 cm diameter poly vinyl chloride (PVC) pipes, plastic bird netting (mesh size = 12.7 mm), covered by white sheets. The sheets were used to block the test birds vi ew of the outside environment ; this was important because it prevented escape behaviors that would confound observati ons of exploration I placed 5 perches inside the test cage for birds to utilize during explor ation that allowed birds to perch throughout the volume of space available for exploration ( Figure 2 1 ) Testing protocol Captured birds were placed in a cloth b ag for transport to the testing environment (usually on the same site or within a five minute drive), then moved into the releasing cage. After a 10 min acclimation period in the dark ened releasing cage (draped in a camouflaged fabric ) the bird was releas ed into the larger test cage (using the pulley string from a concealed location outside the cage) O nce released, birds were allowed 10 min to explore the cag e without disturbance. A video camera (Figure 2 1 ) recorded all movements within the cage
27 Measuri ng Exploratory Behaviors I used 8 component behaviors representing the exploratory behavior for analysis in objective (1) and (2) : flight numbe r, hop number, the proportion of perche s explored, the proportion of ground explored active scan number, average active scanning time, still scan number, and average still scanning time of the tested individual in the novel cage during the tr ia l ( Table 2 1 ) I used two measures of how thoroughly an individual moved throughout the space of the cage. I identified 9 gr ound cells that could be accessed via moving on the ground or between perches, and 3 vertical strata that could be accessed by moving up or down on the 5 perches inside the cage (different perches reached to different heights; Figure 2 2). Two of the perch es reached as high as the top vertical stratum. Overall, the branching perches gave access to 21 of the 27 total cells of volume within the cage. Thus the other measure of thoroughness of exploration was the proportion of those 21 cells visited by each bir d tested in the 10 min sample period. Finally, d uring my preliminary experiments, I found tha t test individual s frequently stop ped at perches to look around to examine the novel environment T herefore I included the number and average time of scanning beha vior as traits of exploratory behavior Active scanning wa s defined as when the individual moved its head and/ or body while looking around toward different directions Still s canning was used to identify times the individual was clearly looking around bu t only the eyes moved without any body movement. Both types of scanning were co mmonly observed Dingemanse et al. (2002) use d the sum of flight number and hop number alone to score the exploratory behavior of great tits during a 2 min period
28 following rel ease into the cage. In analyses, I treat ed flight and hop number independently but I also compared the performance of the exploratory score (flights plus hops) to the other measures Statistical Analysis For objectives ( 1 ) and ( 2 ) I used a similar statist ical approach. Individual component behaviors (Table 2 1 ) were compared among 6 species using analysis of variance ( ANOVA ) and Tukey multiple comparisons t est s (t tests for the foraging guild and social group comparisons). To determine how exploratory beha vior varied across all 9 component behaviors simultaneously (i.e., in multivariate measures) I used the following sequence of analyses. To eliminate collinearity among the 9 component measures I used principal components analysis ( PCA; varimax rotation on the correlation matrix ) and then submitted the component scores with eigenvalues greater than 1 to multiple analysis of variance ( MANOVA ) to see whether the species, foraging guilds or flocking/ non flocking groups were distinct Because body size can infl uence behavior via metabolic rate and other species traits (Brown et al. 2004; Gillooly & Ophir 2009), I used it as a covariate in multiple analysis of covariance (MANCOVA) models for foraging guilds and flocking groups. Finally d iscriminant function a nalysis (DFA) was used to characterize the interspecific and intergroup distinctions Group or species misclassification rates were compared to determine the strongest discriminant model (species, foraging guild, versus social groupings). All statistical a nalyses were performed using SPSS 11. 0 for windows ; significance level was established at alpha = 0.05.
29 To assess objective (3), the value of the classic exploratory score for interspecific characterization of exploratory behavior I used t test and MANOVA to test fo r inter group differences. Results Interspecific Comparison All 8 component behaviors of exploratory behavior varied significantly with species (ANOVA p < 0.05; Table 2 3 ) and only 3 of them were not correlated with body size (proportion of pe rches explored, active scan number and exploratory score; Table 2 4 ) Each component behavior of exploration differentiated between species (Tukey pairwise comparisons: q 0.05, 80, 5 = 4.16, p<0.05; Table 2 5 ). In PCA, the first principal component (PC1) wa s weighted most heavily on flight number, proportion of perches explored and active scan number (active above ground exploration) PC2 however, was dominated by hop number and average still scanning time (active ground exploration; Table 2 6 ) T he first t wo principal components explained 63 8 % of the variance Both components varied significantly with species (MANOVA p<0.05; Table 2 6 ). In the DFA, two functions were generated for species identification. The first factor (F1) was mostly determined by propo rtion of ground explored, still scan number, and flight number, and F2 was determined mainly by average active scanning time (Table 2 7 ). The DFA produced 77.8 % correct classification of the 6 speci es The resulting graph (Fig. 2 3 A) shows 3 distinct clus ters; the white eyed vireo, Carolina chickadee and tufted titmouse (centroids 1 3) were in 1 group, the Northern cardinal was by itself (centroid 5) and the Eastern towhee and Carolina wren formed the third cluster (centroid 4 and 6).
30 Foraging Guild Compar ison Six out of 8 component behaviors differed significantly with foraging guilds (t test p<0.05; Table 2 3 ), but active scan number (t 79 = 0.98, p=0.330), average still scanning time (t 79 = 0.37, p=0.710), and exploratory score (t 79 = 0.37 p=0.715) did not In MANOVA, component 1 (PC1) vary significantly with foraging guilds (F 1, 80 =2.26, p=0.136), but the second component (PC2) did (MANOVA p<0.05; Table 2 6 ). However, in MANCOVA, after adding body size as a covariate, neither PC1 nor PC2 varied betwe en guilds (MANCOVA p=0.44 and 0.66 respectively; Table 2 6 ). In the DFA only one discriminant function was significant and was comprised mainly by the proportion of ground explored and average active scanning time (Table 2 7 ). The DFA produce d a 72.8% corr ect classification of the groups. The scatter plot (Fig. 2 3 B) showed that the insectivore (& lower canopy/ shrub gleaner) group differed most from the omnivore (& ground forager) guild based on proportion of the ground explored. Social Group (Flocking Be havior) Comparison Six out of 8 component behaviors varied significantly between the two groups (Flock vs. Non flock; t test p<0.05; Table 2 3 ). The two groups did not differ in the proportion of perches explored ( t 79 = 1.64 p=0. 106), active scan number ( t 7 9 = 0.36 p=0. 718), or exploratory score ( t 79 = 0.12 p=0. 907). In MANOVA, PC1 and PC2 varied significantly with flocking group; and this was the same results for MANCOVA (with body size as covariate; all p<0.01; Table 2 6 ). In DFA, only one discriminant func tion was used, defined primarily by proportion of the ground explored and flight number (Table 2 7 ). This DFA provided 96.3%
31 correct classification of flocking groups. Fig. 2 3 C shows how flocking/ non flocking groups separated on flight number and propor tion of ground explored The Performance of Exploratory Score versus Component Behaviors In the 6 species comparison (ANOVA) exploratory score varied significantly among species (p=0.004), but the degree of resolution between individual species was low. Multiple comparisons tests with exploratory score identified 2 subsets while other component behaviors grouped the 6 species in to 2, 3 or 4 subsets (Table 2 5 ). The exploratory score did not vary between either the 2 foraging guilds or social groups (p v alue=0.715 and 0.907 respectively). Discussion Measuring Exploratory Behavior F our types of measures are commonly used to quantify behavior: latency (time to onset or cessation) frequency ( number of occurrences) duration (time), and intensity (Martin an d Bateson 2007). While latency is the primary measure used to describe neophobia (Greenberg 2001), both l atency and frequency measures are commonly used to describe intraspecific exploratory behavior ( Dingemanse et al. 2002 ; Verbeek et al. 1994 ). Drent et al. (2003) use d latency to visit a given proportion of the perches whereas Dingemanse et al (2002 ; 2003 ; 2004 ) set the total number of flights and hops within the first 2 minutes as an index of exploratory score. In this study I explore d the utility and relevance of a larger diversity of measures of exploration in order to expose the richness of variation in how different species gather information immediately available in space Greater complexity of measurement will better expose the potential comp lexity of behaviors that serve critical functions (Sieving et al. 20 10 ). In this
32 case, exploratory behavior may best be characterized as exploratory syndrome or a complex of individual measures of exploration that can vary differently across species, guilds, and ecological conditions. In taking this view, my work provided several key insights into behavioral variations not evident when only flight and hop number were used (Table 2 3 ; 2 6 ). All 8 component behaviors provid ed ways to distinctively chara cterize individual species. For example, Carolina c hickadees fly significantly more than any other species, and never stopped moving to scan while still (Table 2 2 2 5 ) In contrast the con fam ilial and socially dominant tufted t itmouse spent more time scanning actively than any other species. Yet no single measure varied so much between species that more than 4 homogeneous sub sets were detected in multi ple comparisons tests (Table 2 5 ) suggesting that ecologically similar birds explore their environme nts in similar ways My analysis also identified behavioral measures that would be useful (or not) for intra specific examination of exploratory behavior. For example, the proportion of ground explored or still scanning behavior would be useless for int raspecific studies with c hickadees, because they did not stay stil l or touch the ground (Table 2 2 ). Additionally, there were large differences in inter individual variation across species and traits that could help guide species and trait selecti on for fu ture studies. Eastern t owhees and Carolina w rens both exhibited high means for proportion of the ground explored (80%) with low variation (CV = 25%), whereas, Northern c ardinals had both the highest mean active scan number (12.6) and highest CV for that sa me trait (47.6%). Thus, we identified
33 species and traits with demonstrably low and high intraspecific variation knowledge useful for designing studies involving response to sele ction on spatial behaviors ( Greenberg 2003 ). With respect to comparisons of behaviors varied significantly between groups (foraging and flocking). This suggests that for inter guild comparisons of exploratory behavior, the choice of components to measure may be critical And I found t exploratory score based on the combined number of hops and flights would largely be useless for making inter group comparisons, and somewhat less powerful for inter species comparisons than the other 8 compo nent behaviors I used (Table 2 3 ) Additionally over half of the traits were significantly correlated with body size suggesting that this needs to be considered in selection of behaviors for interspecific comparison s (Table 2 4 ; Gillooly and Ophir 2010 ). Overall we required multi ple analyses with component behaviors to provide clearer understanding of how species and ecological groupings of species differed with respect to exploratory behavior ( or syndrome; Figure 2 3; Table 2 6 ). As is the case in other aspects of behavioral ecology (e.g., acoustic communication) multi variate analytical approaches to tracking behavioral variation will limit understanding less than univariate approaches ( Arnold et al., 2008; Dawson et al. 2006; S ieving et al. 20 10 ) and, therefore, shou ld be incorpora ted into the d esign of future comparative studies of exploratory (and other complex) behavior s (intra and inter specific, and inter group).
34 Social Roles and Exploratory Behavior I identified that a social group ing (flocking versus non flocking) of the 6 s pecies generated the best discriminent model of the three I compared ( 96.3% correct classification vs. 77.8% for the species model and 72.8% for foraging guilds ) This likely relates to the necessity of flocking species to move in similar ways; species tha t flock together have a similar set of behaviors that relate to the ones we measured. I n subtropical and temperate forests, species attending flocks utilize the same forest strata ( Ghizoni and Guimaraes Azevedo 2006; Herzog et al. 2002; Jullien and Thiol lay 1998; Munn and Terborgh 1979; Thiollay and Jullien 1998; Tubelis 2007 ). This may derive from a potential cost for species attending flocks ; that is whether they have to compromise their optimal foraging speed in order to follow the flock (Hutto 19 88 ; Jones, 1977 ). For example, a slow exploring (hopping) ground gleaner cannot participate in a sub canopy foliage gleaning flock that flies between microsites. Actively foraging above ground flocks move very quickly, relative to solitary foragers (Chen a nd Hsieh 2002) .I found that flocking species have many very similar behaviors related to rate of movement An important component that separated flocking and non flocking species in this study was flight number (Figure 2 3 C ; Table 2 3 ) ; f locking birds ex hibit ed significantly higher number s of flights than non flocking species Another shared trait for flocking species appeared to be their vigilance rates ; flocking species actively scan ned the cage ( for predators or other information; Figure 2 3 A ; Table 2 7 ). Overall, my analysis identified the 3 flocking species as active (flying, active scanning) above gro und explorer s
35 (Figure 2 3A ; Table 2 7 ), whereas, non flocking species were mostly ground explorers. My findings underscore the linkage between the t w o central hypotheses regarding the formation of mixed species foraging flocks ; namely, the anti predator (e. g. Gaddis 1980; Thiollay and Jullien 1998) and foraging efficiency hypothesis (e. g. Macdonald and Henderson, 1977; Powell, 1985; Pulliam 1973 ). The anti predator hypothesis is based on the high degree of alertness observed in nuclear species around which mixed flocks form (Gaddis 1980; Greig Smith 1981; Munn and Terborgh, 1979; Ragusa Netto 2002). In concert with alertness, nuclear species ten d to provide abundant and specific vocal inf ormation ( S ieving et al. 20 10 ) that satellite species use to locate both the flocks and nearby predators ( Goodale and Kotagama, 2005; King and Rappole 2000 ). Most flocking species exhibit high levels of vigila nce toward potential predators ( Chen and Hsieh, 2002; King and Rappole, 2000; Latta and Winderle 1996 ) and vigilance levels can be very high, especially in the flock leaders (commonly they are parids in the Holarctic; Farley et al. 2008 ; Langham et al. 2006; Tubelis 2007 ) Interestingly, vigilance often decreases with group size ( Beauchamp, 2001; Lima et al. 1999 ), suggesting the existence of shared vigilance (more eyes together decreases the need for individual vigilance; Treves 1998 ). High vigilanc e among all flock members will also enhance c ollective detection of predator attack; when all members of feeding group can be alerted quickly to an attack if at least one group member detects and alters its behavior in a way that signals the detection has been made ( via fleeing or freezing; Lima 1995). Having
36 similar capabilities for high scanning frequency would insure that all flock members could share vigilance and predator detection, thus providing the benefits underlying the facilitative social inter actions in flocks (biological markets theory; Hoeksema and Schwartz 2001) I found that the three flocking species I tested exhibit ed significantly higher average active scanning ( Function 2 was weighted heavily on this component; Figure 2 3 A ), supporti ng the functional links that have been proposed between how species move and scan, and their propensity to participate in close, actively moving, interspecific foraging associations (flocks). The performance of the foraging guild model was not as good as the flocking model (Fig ure 2 3 B, C ). Rather than concluding that foraging modes have little relationship to exploratory behavior, this result may relate more to the inconsistent and incomplete guild classifications that are available in the literature. I elected to use foraging guilds to test for similarities in exploratory behavior because s pecies placed in the same foraging guild s tend to exhibit similar responses to spatial variation in their environments (e.g., toward landscape fragmentation ; Kling beil and Willig 2009; Varasteh Moradi and Zakaria 2009) But guild construction is a difficult task and not very precise (Jones and Sieving 2006; Somasundaram and Vijayan 2008 ) With respect to foraging guilds, bird species exhibit a great deal of int ra annual variation in food types (e.g., insectivorous during breeding, frugivorous during migration) not accounte d for in terminology that captures the mean or modal food types over the annual cycle ( e.g., Ehrlich et al., 1988; de Graaf et al. 1985; Rule 1993 ). In my study, the
37 Carolina Wren readily classified as an insectivore, but it is likely to share omnivorous habits with other ground foragers (Table 2 8 ; E h rlich 1988; Rule 1993). In co ntrast, my determination of flock participation was very prec ise and locally determined for my study species (Farley et al. 2008). Hence the used to classify species exploratory tendencies. Finally, p hylogen et ically related species may share similar behaviors (Dobsen 1984; Skinner 1984) ; but these relationships were not considered here Applying the comparative method (Harvey and Pagel 1993) in future interspecific studies of exploratory behavior should prove fruitful. Behavior S yndromes, Neophobia and Exploration Individuals usually show consistent differences in their behavioral tendency (Gosling 2001) and this can influence individual fitness of the individuals through adult survivorship reproduction parental care behavio r ( Both et al. 2005 ; Rale and Festa Bianchet 2003 ). Similar suites of correlated behaviors with fitness effects are referred to as behavioral syndromes (Sih et al. 2004). Both mechanistic and functional studies are applied to behavior syndromes in orde r to understand how the variation persists, how it results from the combin ed effects of genetic and environmental factors, and how selection operates on the personality (Dingemanse et al. 2002 ; Stamps 2003). Neophobia is the avoidance of novel objects or environment s, defined by their dissimilarity from what the individual h a s previously explored (Greenberg 1983). Exploration is defined as an individual s search pattern for and active investigation of novel situations in the absence of pressing physiol ogical need (Immelmann and Beer
38 1989 ; similar to neophilia or curiosity ) Neophobia and neophilia together provide important conceptual implications for how an individual organism deals with the changing world. Neophobia and neophilia are often studied by novel object testing. Researchers examine neophobia by measuring latency and number of unsuccessful approaches to feed when novel objects are placed near a familiar feeding cup or plate filled with preferred food ( Greenberg 1983 ; Mitchell 1976 ). While s tudying neophilia, objects are presented without food so that the ir intrinsic attractiveness to the individual can be measured ( Mayeaux and Mason 1998 ; Mettke Hoffman et al. 2002 ; Negro et al. 1996 ). In reality, a ttraction to explore novelty and the opp osite ( avoid ance or neophobia) are distinct responses that when they interact result in ambivalence (Greenberg and Mettke Hoffman 2001). I n the wild animals do not face simplified situations of either/or; dangers and attract ive resources are mixed in time and space. Yet animals must constantly collect all types of information through explor ation (Danchin et al. 2008), and if we are to understand how exploratory syndrome s are related to ultimate responses, such as fitness and personality in a heterogeneous world, we need a framework for characterizing the variety of related proximate responses, such as curiosity, avoidance, and neutral exploration. The approach developed here, based on identifying finer component behaviors th at characterize more complex responses (exploratory or personality syndrome) could contribute to advancing understanding of animal ecology in changing landscapes ( Greenberg 2003 )
39 Flexibility Plasticity or Behavioral Diversity in a Changing World Behavi oral flexibility is viewed as unlimited and allows individuals to maximize their fitness in the many different environments they encounter during life (Sih et al. 2004). Ecological stereotype s (Klopfer 1967), on the other hand, describe those behaviors t hat remain specialized even in the face of environmental change. However, as Greenberg and Mettke Hodmann (2001) point out, ecological plasticity and stereotypy could be traits with identifiable physiological underpinnings determining the capacity for beha vioral change in the face of resource changes driven by anthropogenic forces acting on landscapes ( Greenberg and Droege 1999 ; Morton 1998 ). Therefore, an understanding of species tendencies toward behavioral stereotyp y or plasticity could be important f or conservation Access to this understanding may be had through studies of neophobia and exploratory behavior. N eophobia and exploration have been related to ecological plasticity (Klopfer 1967) innov ative behavior (Lefebvre et al. 1997) whether a sp ecies classifies as a specialist or generalist ( Greenberg 1992 ; Webster and Lefebvre 2000 ) and even to food preference and geological origins (Mettke Hoffman et al. 2002). Identifying ecological specialist versus generalist species is critical in ecol ogy, evolution and conservation (Devictor et al. 2010). Efforts to quantify species niches in order to identify their status as speciali sts or generali sts ha ve a long history in ecology ( Bazzaz, 1991; Hutchinson 1957; Levins, 1968; Whittaker 1956 ). Att empts to classify species according to their degree of specialization are often problematic because choices of environmental gradients that define a species niche are somewhat subjective and also because of the high variability
40 of existing definitions and methods ( Devictor et al., 2010; Fridley et al., 2007; Witkowski and Lamont 1997). Even so, e cological specialization is classically viewed as a constraint on a species to adapt to environmental changes and, therefore as a strong contributor to extinction risk (McKinney 1997). Specialists have even been labeled the great losers of past and current global change (Colles et al. 2009). As Colles et al. (2009) point out however, useful understanding of how a species degree of specialization affe cts its success in this changing world, integrati on of different approaches (ecology, paleobiology and phylogeny) will be necessary. In the behavioral arena, g eneralist birds are characterized as less neophobic than specialists (Greenberg 1984; 1990; 199 2) Webster and Lefe bvre (2000) show that differences in ecological specialization are linked to variation in neophobia those, in turn, allow predict ecological plasticity. My results may reflect similar relationships if I assume that the ecological plasticity (Table 2 2; Greenberg 2003 ). Two ecological axes of specialization commonly used for birds include diet (diversity of food types consumed) and geograp hic range ( Mettke Hoffmann et al. 2002 ). Plotting each exploratory behavior) on the Y axis, versus (respectively; Fig. 2 4 A C) latitudinal, longitudinal range, and diet specializat ion, Northern cardinal and tufted titmouse stand out as the most behaviorally variable/flexible (highest CVs) and with the broadest geographic ranges and diets. These two species are identifi ed as
41 extreme generalists in majo r references ( e.g., Birds of Nor th America Online; http://bna.birds.cornell.edu/bna/ ), and they exhibited extreme intraspecific variation in exploratory behavior. On the other end of the spectrum is the Eastern towhee ; the distribution range displayed in Fig. 2 4 is for white eyed (pale straw yellow iris ) subspecies that is largely endemic to Florida and all towhees used in this study were of this particular morphological form ( Batten, 2008; Dickinson 1952 ). My data suggest that th is subspecies may exist in a limited geographic al area, in part, because of the lack of variation (plasticity) in exploratory behavior (Fig. 2 4A, B); though we note its diet is not particularly specialized (Fig. 2 4 C). Keeping in mind the caveat that this study was designed to detect interspecific (n ot intraspecific) variation, my findings do seem to support the idea that behavioral diversity along eco logical axes of specialization. In conclusion, my findings parallel other avenues of study linking behavioral plasticity with key traits relevant to survival and reproduction, as exploratory behavior surely is, and continued work should of environmental change.
42 Figur e 2 1. T he design of the testing cage (novel environment) T he releasing cage, covered by camouflage fabric is in the lower left corner The tripod ( at right ) held the video camera set to record all behavior s inside the cage. One perch (of 5: see Figure 2 2) is visible inside the cage.
43 Figure 2 2. The assignment of volume within the cage into 27 cubes (9 ground calls 3 st r a ta) for exploration by test subject s Perching substrates (sticks) intercepted a total of 21 of the 27 total volume c ells.
44 (A) Figure 2 3. R esults of the d iscriminant f unction a nalysis for (A) species (1 st two discriminant functions used; WEVI = white eyed vireo, CACH = Carolina chickadee, TUTI = tufted titmouse, CAWR = Carolina wren, NOCA = Northern ca rdinal, and EATO = Eastern towhee), (B) foraging guilds, and (C) flocking groups showing, for each, the two component behaviors with the highest weights on the single discriminant function generated. The factor loadings for DF analyses are listed in Table 2 4.
45 (B) (C) Figure 2 3. Contin ued.
46 (A) Figure 2 4. Scatter plots of mean coefficients of variation (CV) across all 8 component behaviors vs. (A) max. latitudinal range (degrees) in geographical distribution, (B) max. longitudinal range (degrees) in geographical distribution, and (C) ratio of insect to plant food types in the diet of each species. In Panels A and B, the assumption is that generalists are on the right side, and in Panel C, they are on the left side. WEVI = white eyed vireo, CACH = Caro lina chickadee, TUTI = tufted titmouse, CAWR = Carolina wren, NOCA = Northern cardinal, and EATO = Eastern towhee.
47 (B) (C) Figure 2 4. Continued
48 Table 2 1. List of component s of exploratory behavior their definitions / measures and how data were tran sformed for analysis. Component Name (ID) Definition Transformation Flight number (FN) Number of flights in ten minutes Hop number (HN) Number of hops in ten minutes Prop o f perches explored (PP) The number of perch spaces one individual explored in ten minutes divided by the total number of perch spaces (3 strata 9 cells where a segment of perches is present = 21 cells with perch segment out of 27 total ) A rcsinsqrt Prop o f ground explored (PG) The number of ground cells one individual explored i n ten minutes divided by the total number of cells A rcsinsqrt Active scan number (ASN) The number of active scans (l ooking around with head or body movements) in ten minutes Average active scanning time (AAST) Accumulated active scanning time divided by the active scan n umber Still scan number (SSN) The number of still scans (looking around with no head or body movements, merely the eye activities) in ten minutes Average still scanning time (ASST) Accumulated still scanning time divided by the still scan number
49 Table 2 2. Mean and coefficients of variation (CV) of 8 component exploratory behaviors for 6 species. Species Name (Species ID) FN HN PP PG ASN AAST SSN ASST White eyed v ireo (WEVI) Mean 16.9 5.9 0.3 0.0 7.3 54.5 1.5 9.1 CV 58. 0% 54.2% 33.3% N/A 23.3% 15.8% 106.7% 95.6% Carolina c hickadee (CACH) Mean 37.7 17.9 0.5 0.1 12.0 39.2 0.0 0.0 CV 15.6% 33.0% 40.0% 200.0% 15.0% 15.1% N/A N/A Tufted t itmouse (TUTI) Mean 27.7 12.4 0.5 0.1 8.8 58.0 0.5 12.5 CV 39.7% 52.4% 40.0% 200. 0% 26.1% 20.3% 160.0% 307.2% Carolina w ren (CAWR) Mean 10.5 26.0 0.5 0.8 7.5 48.3 2.8 33.4 CV 64.8% 76.5% 20.0% 25.0% 25.3% 18.6% 32.1% 118.9% Northern c ardinal (NOCA) Mean 16.9 20.1 0.4 0.4 12.6 37.3 1.1 6.3 CV 72.8% 74.1% 50.0% 75.0% 47.6% 23.9% 127.3% 169.8% Eastern t owhee (EATO) Mean 8.4 32.9 0.2 0.8 6.1 44.8 3.4 29.6 CV 96.4% 59.6% 100.0% 25.0% 29.5% 14.3% 50.0% 68.2% Total Mean 19.2 18.7 0.2 0.2 8.9 48.0 1.6 15.8 CV 68.8% 84.5% 50.0% 150.0% 41.6% 24.0% 106.3% 177.2%
50 Table 2 3 Compari son among species and between groups based on each component of exploratory behavior Statistical Analysis: ANOVA Comparison Trait Results F df p value Species Flight number 16.4 5,80 0.0 1 Hop number 7.0 5,80 0.0 1 Proportion of perches explored 6.8 5,80 0.0 1 Proportion of ground explored 41.9 5,80 0.01 Active scan number 10.0 5,80 0.01 Average active scanning time 11. 7 5,80 0.01 Still scan number 1 5.0 5,80 0.01 Average st ill scanning time 3. 4 5,80 0.0 1 Exploratory Score 3.8 5,80 0.01 Statistical Analysis: T Test Comparison Trait Results t df p value Foraging Guild Flight number 3.0 79 0.01 Hop number 3.0 79 0.01 Proportion of perches explored 3.0 79 0.01 Propo rtion of ground explored 4.5 79 0.0 1 Active scan number 1.0 79 0.33 Average active scanning time 4.0 79 0.0 1 Still scan number 2.3 79 0.02 Average still scanning time 0. 4 79 0.71 Exploratory Score 0. 4 79 0.7 2 Flocking Group Flight number 5. 8 79 0.01 Hop number 4. 8 79 0.01 Proportion of perches explored 1.6 79 0.1 1 Proportion of ground explored 11.6 79 0.01 Active scan number 0. 4 79 0.7 2 Average active scanning time 3.5 79 0.01 Still scan number 5.2 79 0.01 Average still scanning t ime 2.5 79 0.01 Exploratory Score 0.1 79 0.9 1
51 Table 2 4. Results of Kendall s Tau b Correlation Analysis between component behaviors and body size Flight number Hop number Prop. of perches Prop. of ground Active scan number Average active scannin g time Still scan number Average still scanning time Exploratory score Body size Correlation C oefficient 0.30 0.32 0.07 0.55 0.06 0.21 0.26 0.23 0.74 Significanc e (2 tailed) 0.0 1 0.0 1 0.40 0.0 1 0.49 0.02 0.01 0.01 0.40
52 Table 2 5 Tukey Test results for between species comparison WEVI = white eyed vireo, CACH = Carolina chickadee, TUTI = tufted titmouse, CAWR = Carolina wren, NOCA = Northern cardinal, a nd EATO = Eastern towhee. MD = Mean Difference = Spp 1 Spp 2. Flight Number Hop Number P. of Perches Explored P. of Ground Explored Active Scan Number Spp 1 Spp 2 MD p value MD P value MD p value MD p value MD p value W E VI CACH 20. 8 0.0 1 12. 0 0.26 0.2 0.09 0.1 0.99 4.7 0.0 1 TUTI 10.8 0.03 6.4 0.76 0.2 0.04 0.2 0.36 1.4 0.77 CAWR 6.5 0.42 20.1 0.0 1 0.1 0.21 0.8 0.0 1 1.3 1.00 NOCA 0.0 1.00 14.1 0.07 0.1 0.64 0.5 0.0 1 5.3 0.88 EATO 8.5 0.19 27.0 0 .01 0.1 0.35 0.8 0.0 1 1.3 0.0 1 CACH TUTI 10.0 0.10 5.5 0.91 0.0 1.00 0.1 0.85 3.3 0.08 CAWR 27.2 0.0 1 8.1 0.68 0.1 0.99 0.7 0.0 1 4.5 0.01 NOCA 20.8 0.0 1 2.2 1.00 0.1 0.82 0.4 0.0 1 0.6 1.00 EATO 29.3 0.0 1 15.0 0.11 0.3 0.0 1 0.8 0.0 1 5.9 0.0 1 TUTI CA WR 17.2 0.0 1 13.6 0.06 0.0 0.98 0.6 0.0 1 1.3 0.83 NOCA 10.8 0.04 7.7 0.64 0.1 0.76 0.3 0.0 1 3.9 0.01 EATO 19.3 0.0 1 20.5 0.0 1 0.3 0.0 1 0.7 0.0 1 2.7 0.19 CAWR NOCA 6.5 0.47 5.9 0.85 0.0 0.99 0.3 0.0 1 5.2 0.0 1 EATO 2.1 0.99 6.9 0.77 0.3 0.0 1 0.1 0.96 1.4 0.83 NOCA EATO 8.5 0.23 12.8 0.17 0.2 0.01 0.4 0.0 1 6.5 0.0 1 Subsets 2 3 2 3 3
53 Table 2 5 Continued Average Active Scanning Time Still Scan Number Average Still Scanning Time Exploratory Score Spp 1 Spp 2 MD p v alue MD p value MD p value MD p value WEVI CACH 15.3 0.0 1 1.5 0.03 9.1 1.00 32.7 0.0 1 TUTI 3.5 0.88 1.0 0.18 3.4 1.00 17.2 0.13 CAWR 6.2 0.42 1.3 0.06 24.3 0.12 13.6 0.37 NOCA 17.2 0.0 1 0.5 0.92 2.8 1.00 14.1 0.37 EATO 9.7 0.07 1.9 0. 0 1 20.5 0.34 18.5 0.13 CACH TUTI 18.8 0.0 1 0.5 0.91 12.5 0.84 15.5 0.33 CAWR 9.1 0.13 2.8 0.0 1 33.4 0.03 19.1 0.14 NOCA 1.9 1.00 1.1 0.29 6.3 0.99 18.6 0.19 EATO 5.6 0.68 3.4 0.0 1 29.6 0.10 14.3 0.49 TUTI CAWR 9.7 0.04 2.3 0.0 1 2 0.9 0.24 3.6 1.00 NOCA 20.7 0.0 1 0.6 0.80 6.2 0.99 3.1 1.00 EATO 13.2 0.0 1 2.9 0.0 1 17.1 0.53 1.3 1.00 CAWR NOCA 11.0 0.02 1.7 0.0 1 27.1 0.08 0.5 1.00 EATO 3.5 0.91 0.6 0.78 3.8 1.00 4.9 0.99 NOCA EATO 7.5 0.29 2.3 0.0 1 10.5 0.24 4.3 0.99 Subsets 4 4 2 2
54 Ta ble 2 6 Factor loadings of the 8 component behaviors on the 2 principal component s. Eigenvalues and amount of variance explained by the respective components are given at the bottom. MAN (C) OVA results are shown in the bottom two panels Factor loadings Parameter PC1 PC2 1 Flight number 0.83 0.1 6 2 Hop number 0.05 0.86 3 Prop of perches explored 0.7 8 0.14 4 Prop of ground explored 0.4 4 0.59 5 Active scan number 0.74 0.28 6 Average active scanning time 0.71 0.08 7 Still scan number 0.6 7 0.1 1 8 Average still scanning time 0.33 0.8 8 Eigenvalue 3.11 1.99 % variance explained 38.9 24.9 MANOVA results Comparison PC1 PC2 F df p value F df p value Species 15.5 5 ,80 0.0 1 23.0 5 ,80 0.0 1 Foraging guild 2.3 1 ,80 0.14 26.6 1 ,80 0.0 1 Flocking group 13.5 1 ,80 0.0 1 92.2 1 ,80 0.0 1 MANCOVA results Comparison PC1 PC2 F df p value F df p value Foraging guild 0.60 1 ,80 0.44 0.20 1 ,80 0. 66 Body size 3.44 1 ,80 0.07 7.83 1 ,80 0.01 Flocking group 8.40 1 ,80 0.0 1 37.38 1 ,80 0.0 1 Body size 0.64 1 ,80 0.43 0.07 1 ,80 0.79
55 Table 2 7 Structure coefficients/ discriminant loadings of the 8 component behaviors on the function(s) from canonical discriminant function analysis of the different group comparisons Eigenvalues and amount of variance explained by the respective components are at the bottom. Parameter Species Foraging Guild Flocking Group Function1 Function2 Function1 Function1 1 Flight number 0.38 0.20 0.23 0.3 4 2 Hop number 0.2 1 0.0 5 0.3 3 0.27 3 Prop of perches E xplored 0.1 5 0.07 0.4 1 0.1 4 4 Prop of ground E xplored 0.68 0.05 0.6 2 0.68 5 Active scan number 0.12 0.29 0.11 0.02 6 Average active scanning time 0.12 0.9 2 0.55 0.2 1 7 Still scan number 0.38 0.2 5 0.2 8 0.3 1 8 Average still scann ing time 0.04 0.2 9 0.0 8 0.02 Eigenvalue 5.72 0.77 0.6 8 3.68 % variance explained 79.6 10.8 100 .0 100 .0
56 Table 2 8 Foraging behavior of 6 species based on different r eferences Species Name D e Graaf et al. 1985 Ehrlich et al. 1988 Rule 19 93 White eyed v ireo (WEVI) Insectivore: lower canopy/ shrub gleaner s Insect: foliage gleaner 85% insects 15% plants Carolina c hickadee (CACH) Insectivore: lower canopy/ shrub gleaner s Insect, conifer seed, and fruit: foliage and bark gleaner 70% insects 30% plants Tufted t itmouse (TUTI) Insectivore: lower canopy/ shrub gleaner s Insect, seeds, and fruit: foliage and bark gleaner 65% insects 35% plants Carolina w ren (CAWR) Insectivore: lower canopy/ shrub gleaner s Insect: ground and foliage gleaner 95% insects 5% plants Northern c ardinal (NOCA) Omnivore: ground forager Insect, seeds, and fruit: ground gleaner 65% plants 35% insects Eastern t owhee (EATO) Omnivore: ground forager Insect, seeds, and fruit: ground and foliage gleaner 65% plants 35% insects
57 CHAPTER 3 TUFTED TITMOUSE ( BAEOLOPHUS BICOLOR ) ANTI PREDATOR VOCALIZATIONS ALTER HETEROSPECIFIC EXPLORATORY BEHAVIOR Introduction Animal Ecology and Information driven Decision making Behavior is often defined as the way by which organisms adjust to en vironmental variation. Adjustment requires reactions to cues in the environment signifying that organisms make decision s based on information they gather and process. Therefore, the study of behavior can thus be viewed as the study of decision making emp hasizing information gathering, memory storage and processing (Danchin et al. 2004; Danchin et al. 2008). The value of information resides in its power (accuracy and precision) to aid in identify ing current conditions or to predict future conditions. Re search e fforts to understand information use are widespread across molecular, developmental and neurobiolog ical contexts ( Adami, 2004 ; Maynard Smith 1999; Wagner 2007). However as Dall et al. point out (2005), information use in behavioral and evolution ary ecology has been narrowly restricted to particular contexts ; e.g., foraging and mate choice and navigation. Given that i nformation use is critical to animals adaptive behavior both Dall et al. (2005) and Danchin et al. (2008) emphasize an urgency to develop information driven approaches across animal ecology in order to speed fundamental integration and enhance the predictive power of ecological theory Social Information and Its Influences Biological information is defined as gene tic and environmenta l cues that reduce uncertainty, potentially allowing a more adaptive response (Danchin et al.
58 2004). Non genetic information can be acquired in three ways: (1) parental effects on offspring phenotype; (2) personal information gathered by one individual s interaction with its environment; and (3) social information that is gathered by observing other individuals. Social information can be derived from cues produced inadvertently by individuals engaged in performance of activities intentionally between individuals ( Danchin 2008). O btaining and utilizing social information is increasingly recognized as a common and important mechanism underlying animal decision making, including foraging, habitat choice, and mate choice (Bonnie an d Earley 2007) Social information use can strongly influence both survival and fitness ( Marzluff et al. 1996 ). For example, in an English r aven ( Corvus corax ) roost, individuals follow successful foragers to food which enhances their survival ( Mennill et al. 2002 ). And e ven though most examples derive from intraspecific studies social information is often exchanged between heterospecifics. For example, n ine spined sticklebacks ( Pungitius pungitius ) choose foraging patches based on the foraging prefere nce of three spined sticklebacks ( Gasterosteus aculeatus ; Coolen et al. 2003). Using information gathered from heterospecifics through their unique sensory adaptations or greater experience, animals may be able achieve more accurate information than the y can obtain alone (Sepp nen et al. 2007). V igorous interspecific competitors, class ically thought to have globally negative influence on one another, apparently can provide exceptionally relevant information for each other about food resources or predatio n risk ( Forsmann et al., 2007; Monkk nnen and Forsmann 2002 ; Templeton and Greene 2007).
59 Therefore interspecific social information use is likely critical for understanding interaction s among species; yet is a topic largely absent from community ecologi cal literature ( Dall et al. 2005; Sepp nen et al. 2007 ; but see Fletcher 2008 ). One area where interspecific information use has been explored most fully, however, is regarding interspecific use of anti predator alarm calls. Using Social Information to Navigate Landscapes of Fear Perhaps the most important of situations animals face is the risk of death or mortal injury by predator attack. Real probabilities of attack on prey organisms vary across landscapes in predictable ways based on changes in predat or activity and abundance with changes in land cover, habitat types, and distances among habitats (e.g., Kuehl and Clark 2003 ; Phillips et al. 2004 ; Sieving and Willson 1998 ). While direct predation events obviously reduce individual fitness, the nonlet hal (largely behaviorally mediated) effects of predation risk also impact fitn ess in a variety of ways (Lima 1998 ; Lima and Dill 1990 ). Non lethal predation effects, often termed threat fear risk or non consumptive effects, are now thought to play a major role in population and community dynamics, and potentially a much greater role even than direct predation ( Creel and Christianson 2008 ; Cresswell 2008 ; Preisser et al. 2007 ). It follows that information about predators (location, type, hung er state) and the relative risk of predation would be a valuable commodity that individuals could share with others using various types of cues and signals ( Caro 2005 ; Hoeksema and Schwartz 2001 ) with important consequences, including the determination o f animal distributions at large scales (Goodale et al. 2010).
60 Social information use by animals has potentially profound influence s on important spatial decisions of animals including habitat selection For example, the real or perceived presence and den sity of resident birds are used by migrant s in breeding habitat selection; in part as indicators of breeding patch quality ( Fletcher 2008 ; Thomson et al. 2003) The perception of shared predation risk can also be transmitted between species ( Ito and Mor i, 2010; Sieving et al. 2010 ), and this can influence spatial behaviors at small (Schmidt et al. 2008) and larger scales ( Manning et al., 2009; Plotnick et al., 2010; Sieving et al. 2004; Walker et al. 2007 ). The study of social information use in hete rogeneous landscapes of fear is an active area of conservation, and is of great value to the study of animal spatial ecology more generally (Willems and Hill 2009). In this study, I examined one of the mechanisms whereby a species use of socially derived information about predation risk can influence spatial behavior in birds. Exploratory behavior is the gathering of information about objects or other aspects of the environment that does not satisfy immediate needs (Dall et al. 2005), and can be charact erized using a simple measures of movement and scanning behaviors in novel environments (Chapter 2). Exploratory behavior is both heritable and highly correlated with dispersal movements at large scales (Dingemanse er al., 2002; 2003). Therefore, explorato ry behavior as described in experimental (novel) environments can be used as a surrogate for large scale movement behaviors (e.g., natal dispersal; Dingemanse et al. 2003). Therefore, I experimentally manipulate d exploratory behavior with interspecific so cially derived information about predation risk as a means to generate hypotheses
61 about how interspecific information use may be influencing landscape level spatial behaviors such as exp loration and habitat selection Here, I used a selection of well chara cterized anti predator calls of the common, socially dominant tufted titmouse ( Baeolophus bicolor ) recorded under known predation threat contexts, as heterospecific social i nformation and examined the effects of these calls on the exploratory behavior of sympatric Northern cardinals ( Cardinalis cardinalis ) Study System I s e lected 3 different anti predator vocalization of tufted titmouse as the social information. As other species in the Paridae family, titmouse is a very vocal species ( Harrap and Quinn 1 995) Its most common form of risk related vocalization is the chick a dee call a predator mobbing call that is typical of most species in the same family and given when facing a non immediate threat like a perched avian predator (e.g., the Eastern s cre ech o wl Megascops asio ; Otter 2007 ) Individuals of c losely related b lack cappe d c hicka d ee ( Poecile articapillus ) vary the composition of their notes in response to the presence of different predators and th ese mobbing calls are displayed in a progressi ve manner consistent with the threat level of different predators smaller raptors being perceived as more dangerous (Baker and Becker 2002 ; Ficken 1989 ; Gaddis 1985 ; Templeton et al. 2005) T itmouse also exhibit s the same mobbing call system, where t he numbers of chick and dee notes per chick a dee mobbin g call vary inversely with the perceived level of risk (Hailman 1989 ; Sieving et al. 20 10 ) They also give a different anti predator call for actively attacking raptors i.e., peak levels of mortality risk. These calls are commonly
62 high frequency, low amplitude calls ; Hetrick 2006; Sieving et al. 20 10 ). A coustic information content in t itmouse calls given to live predators clearly and appropriately distinguish th e different threats they represent, and Carolina c hickadee another Paridae species, responses clearly and appropriately distinguish predation threats encoded in t itmouse calls ( Hetrick 2006 ) Potentially many other species (birds and mammals) use the in formation from parid alarm calls across the Holarctic region in decision making concerning anti predator behaviors (Langham et al 2006 ; Schmidt et al. 2008) and, potentially, in such spatial behaviors as boundary crossing (Sieving et al. 2004) Previo us work s on the communication networks surrounding titmice and other parids provid e strong indications that the vocalizations of parids represent important, community level information sources that inform risk appropriate decision making Here, I tested th is hypothesis directly by seeking to generate adjustment s of behavior in one species toward the information provided by an other Auditory cues are among the most important forms of social information transmission ( e. g. Clara et al. 2008 ; Grim 2008 ; Welb ergen and Davies 2008) widely used within and between species ( e. g. Griffin et al. 2005 ; Schmidt et al. 2008) This is especially the case in bird communities, where acoustic cues and signals carry the majority of social information available to inte rested con and heterospecifics (Caro 2005; Kroodsma and Miler, 1996; Langham et al. 2006; Sieving et al. 2 008 ). In this study I tested whether and how acoustic information about predation threats, produced in known anti predator contexts by one
63 specie s, influenced the exploratory behavior of another species known to participate in the same alarm calling network. I selected the Northern cardinal as a focal species for receiving the alarm calls of titmice for several reasons. My previous study suggests that they exhibit large intraspecific variation in certain components of exploratory behavior, providing the potential for detectable behavioral differences among treatments; the species is common, widely sympatric with titmice, and participates in the com munication network centered on titmice (i.e., demonstrates appropriate responses to alarm calls of known context; Langham et al. 2006 ; Sieving et al. 2004) Research Design Objective s and Experimental Design I hypothesized that t he e xploratory behavior of Northern cardinal will be influenc ed by t ufted t itmouse vocalizations, and such influence will be determined by the type of predation risk encoded in the t itmouse vocalizations. This will be tested by comparing exploratory behavior of individual cardina ls before, during, and after informational cues (titmouse alarm calls) are broadcast ed to them while held in a novel environmental cage (Chapter 2) I exposed cardinals to 4 acoustic treatments: 1) control (no acoustic information); 2) low risk mobbing cal ls (the proportion of chick/D notes was 3/2; 3) high risk mobbing calls (the proportion of D/chick notes was 3/2 (using Hetrick 2006; Sieving et al. 20 10 ); 4) seet call (comprised of several seet notes per call). Cardinal exploratory behaviors were video taped for 9 minutes; 3 min before b roadcasting, 3 min during broadcasting, and 3 min after broadcasting. I then quantified 8 component activities that reflect
64 key exploratory behaviors (Chapter 2) for each period (before, during, after) and examined whether and how exploratory behavior chan ged among treatments and among periods of each playback trial Predictions Because no predators were visible to birds during trials, I assumed that reduced as the level of thre at of predation increased from control to low risk to high risk mobbing to seet calls, respectively. Mobbing aggregations typically generate high amounts of movements by birds inspecting perched or ground predators that have been spotted, but when mobbin g calls are given, responding individuals do not exhibit vigorous activity until the predator is spotted ( Sieving et al. 2004 ). Thus, in this experiment, the mobbing call treatments represent the possibility that a perched predator is nearby, but invisibl e ; therefore movement would be increasingly risky as the threat level encoded in titmouse calls goes up. And this response (less movement) should be extreme in the seet call treatment because parids use this call to identify a flying, potentially attackin g, hawk in close proximity ( Sieving et al. 20 10 ). The most common response to this call by same and different species is to freeze in place until danger passes ( Gaddis 1985 ; Haftorn 2000; Hetrick 2006). With respect to the 4 treatments, then, I predict ed generally less activity (hopping, flying, active scanning) from control to seet call treatment, and more still scanning. With respect to temporal changes in behavior across the treatment periods, I expected that exploration would decrease during the th ree titmouse treatments, and then rebound in the post manipulation period, but the rebound should be less with increasing threat level.
65 If still scanning increases during playback (as expected), it should reach its highest level in the seet call treatment, and decrease again after playback stops but to a lesser degree with increasing threat level. A summary of predicted responses is in Table 3 1. Methods Capture and Handling I conducted field tests from May, 2009 to March 2010 in North central Florida. No rthern cardinals were captured either by feeder trap (wire boxes with drop doors placed on platform feeders) or mist net s Field work took place at the Ordway Swisher Biological Station in Melrose FL (Putman County) ; the USDA/APHIS/WS/NWRC Florida Field S tation (United States Department of Agriculture, Animal & Plant Health Inspection Service, Wildlife Services, National Wildlife Research Cente r; USDA lab) and at 5 private yards in the city of Gainesville FL ( Alachua County; all sites where birds were tra pped were at least 2 Km apart). Within 10 min of capture on average (up to 1 hour) Northern cardinals were placed in a small wire box with a swinging door, covered with camouflage cloth (to darken the cage, and hide its outline). This box was placed insid e the test cage (Chapter 2) which serve d as a novel environment for recording exploratory behavior The bird was released from the holding cage at the beginning of the trial (within 10 minutes of being placed there; see below). Playback Treatments All reco rdings were constructed to play titmouse calls in realistic sequences for 3 minutes total, during each trial. Playback of each recording followed a silent period of 3 minutes, followed by an silent period
66 of observation lasting 3 minutes. To minimize pseudoreplication ( Kroodsma 1989 ) each recording played during a trial had calls from 3 different groups of 3 titmice presented with the same live predator stimulus under controlled aviary conditions (low and high risk mobbing and seet calls recorded by Hetrick 2006). The 3 different sets of calls were played in random order on each playback recording, for a total of 6 possible recordings per treatment type to select for a given trial. For each cardinal captured and released with in the test cage, treatment type was randomly selected from the 4 trials (high and low risk mobbing, seet calls, and controls) to complete one full replication of treatments before completing the next replicate set (to maintain equal sample sizes among tre atments as testing progressed). I applied a priori power analyses several times as data were collected to identify how many total samples were needed to detect the least difference of interest in analysis (using G*power 3.1; http://www.psycho.uni duesseld orf.de/aap/projects/gpower/ ). Testing P rotocol and Exploratory B ehavior M easurements Following capture and handling, each bird was given 10 min to acclimate to the small releas ing cage (draped in a camouflaged fabric ) ; then each bird was released into th e larger test cage with perches ( Figure 2 1, 2 2 ). The test individual experienced a n initial 3 min period of silence, then 3 m in of broadcast of one treatment type (control, low risk mobbing call, high risk mobbing call, and seet call), and then another 3 min of silence. B irds were kept in the cage for a total of 19 minutes. From the moment of placement into the releas e cage (opened via a string and pulley apparatus from 20 m away), test birds w ere not
67 exposed to sight or sound of human observers A video camera recorded all movements of subjects within the cage (details described in Chapter 2) I used 8 components of exploratory behavior (as in Chapter 2) : flight numbe r, hop number, the proportion of perche s explore d, the proportion of ground explo red ac tive scan number, average active scanning time, still scan number, and average still scanning time Summaries of all 8 behavioral components were produced for each period (before, during and after playback), and submitted to analyses. Statistical Analysis I conducted a princip al component analysis (PCA) based on all 8 component behaviors of exploratory behavior and pooling across the three periods of testing to obtain the princip al components (PC) having an eigenva lue greater than 1. Based on this given p rincip al component model I extracted the PC values of each treatment at every testing period as currency for later comparison To test for the combined effects of treatment and time period (before, during, after), I first calculated the differences in the PCs for the two transitions of interest; During Before (D B), and After Before (A B). These differences would capture the magnitude of initial responses (during playback; D B) and the degree to which those responses lasted during the post playback period (A B). I then performed two MANOVA s on each of these PC differences, respectively, with treatment as the predictor. In the D B analysis, I expect ed to see a general /significant decreas e in movement (hops, flights, thoroughness, active scanning) with incre asing predation threat (i.e., from control to low high risk mobbing, to seet calls) In the A B analyses, PC differences capturing movement
68 should not be so large, and treatment effects not so distinct (assuming rapid return to exploration following cess ation of playbacks). But behaviors reflecting vigilance (i.e., scanning) should remain higher (differences between before and after periods should be greater) for the highest risk treatments (see Predictions, above, and Table 3 1) Results I tested a tota l of 32 cardinals ( 8 individuals per treatment type). The results of the PCA on the 8 component behaviors are summarized in Table 3 2. Only components with an eigenvalue greater than 1 were subsequently analyzed The first principal component (PC1) was we ighted most heavily on flight number and active scan number (active movement ) The second component (PC2) was dominated by still scan number and average still scanning time ( still scanning behavior ) The third component (PC3) was dominated by hop number an d proportion of ground explored (ground exploration). T he first three principal components explained 78.4 % of the variance In MANOVA, all three principal components varied significantly with treatments in the Before During analysis, and two of the thre e were significant in the After Before analysis; PC3 did not vary among treatments anymore (p value = 0.19; Table 3 2). Temporal Patterns in Movement Types The s eet call treatment caused a significant decrease in flight number during broadcast, and foll owing broadcast, flight number did not return to levels preceding broadcast High risk mobbing calls caused the same pattern but not as apparent as in the seet call treatment (Figure 3 1 A). In contrast, hop number dropped during, and then fully rebounded after playback, for all titmouse call
69 treatments, to levels observed before playback; the drop in hop number during playback was most significant in the seet call group (Figure 3 1 B). Temporal Patterns in Thoroughness of Exploration For the proportion of perches explored, both low risk and high risk mobbing caused a slight increase during the broadcast. Following broadcast individuals kept exploring perches in low risk mobbing treatment but not so much in the high risk mobbing treatment. The seet call cau sed a decreas e in the proportion at first but then it largely rebounded post broadcast (Figure 3 1 C). Individuals explored a slightly smaller proportion of the ground for all experimental treatments during broadcasting. After broadcasting, cardinals expl ored either similar or greater proportion s of the ground inside the cage (Figure 3 1 D). Temporal Patterns in Scanning Behavior A ctive scan number increased during broadcast then decreased slightly after the low risk mobbing treatment was played In contra st, the high risk mobbing and seet call treatments caused a decrease and then an increase in active scan number ; and the decrease during broadcasting was more significant in the seet call group (Figure 3 1 E). As for active scanning time, all three e xperimental treatments showed a quadratic pattern of chang e with an increase during and a decrease after broadcast (Figure 3 1 F). Still scanning, on the other hand, in both high risk mobbing and seet call treatments increased during broadcast (both stil l scan number and time) and decreased after broadcast (both still scan number and time) T he decreas e after broadcast was less significant in the seet call treatment (Figure 3 1 G, H) ; still scanning remained pretty high
70 Discussion Interspecific Social Cues with Known Information Content The basic function and structure of chick a dee calls is highly conserved across the Paridae family and has been well de s cribed for several parid species (Langham et al. 2006 ; Templeton and Greene 2007; Templeton et a l. 2005; Sieving et al. 2009; Soard and Ritchison 2009 ). The encoded information indicates different degrees of perceived predation threats from perched predators (avian, mammalian and/or reptilian; Sieving et al. 20 10) and is expressed in the pattern of variation in chick and D notes per chick a dee call ; more D notes per call signifies a more dangerous predator is being addressed by the signaler, and the greater the level of threat encoded in the call ( Hetrick 2006; Templeton et al. 2005 ; Sieving et al. 20 10 ). In contrast, seet calls are given when attacking (nearby flying) predators are detected and these calls elicit escap e or fearful responses ( Hetrick, 2006; Lima and Dill 1990; Templeton and Greene 2007). Appropriate responses to these calls, and to the different types of predator situations and threat levels encoded in the calls, can be generally classified into, first, escape or predator avoidance and, second, predator inspection or mobbing behaviors. Each, escape or inspection/mobbing, has associated predictable patterns in scanning (vigilance) or movement (hopping/flying) behaviors; all of which are components of exploratory behavior ( Dugatkin and Godin, 1992 ; Lima 199 8 ; Chapter 2). Escape / avoidance behaviors, as from diving or flying h awks, would include an initial flight or hop into cover, followed by stillness in hiding and increased vigilance (Lima 199 8; Templeton and Greene 2007 ). When cover is not readily available, freezing, and using still scanning in small birds has been
71 obse rved (Ficken and Witkin 1977 ; Gaddis 1985 ; Haftorn 2000 ). Mobbing calls can generate increased activity levels (Langham et al. 2006; Templeton et al. 200 5 ; Soard and Ritchison 2009) associated with predator inspection and attack, but this may not al ways be the case when a perched predat or has not been located In this study (as in Hetrick 2006), no predator is locatable; therefore depending on the perceived threat level represented by the predator, responses should vary from indifference to escape or avoidance behavior. Given the known level and kinds of threat known to be encoded in the calls that I broadcast, and given assumptions relating the different components of exploratory behavior I measured to appropriate types of responses to the calls, m y results generally fit expected pattern s Northern cardinals showed significant changes on exploratory component behaviors across time at different treatments (Table 3 2). Exploratory Behavior under Socially Conveyed Threat of Predation As the encoded thr eat level in titmouse calls increased, levels of cardinal activity generally decreased, consis tent with avoidance behaviors. Seet call s reduced the flight number and hop number of individuals during treatment broadcast s ; high risk mobbing generated the sam e but less distinct pattern of chang e while low risk mobbing caused no specific changes (Figure 3 1 A, B). Similar patterns were observed for the measures of exploratory thoroughness (proportion of perches and ground explored; Figure 3 1 C, D). On the othe r hand, calls with higher threat level s encoded in them induced marked expansion of time spent scanning the environment, either active ly (with head/body turning) or via still scanning (head and body still, eyes moving) during broadcast of vocalization s (Fi gure 3 1 F, H). I therefore concluded that Northern cardinals not only
72 accurately interpret ed the social information from tufted titmouse but responded appropriately with finely tuned variations in diverse components of exploratory behavior My analysis reveal ed the highest level of detail to date in understanding the various components of appropriate anti predator behaviors Compared to the other 5 species tested in Chapter 2, Northern cardinal s (as non flocking omnivores) had average hop and flight num bers, thoroughness, and active scanning behaviors, but very low s till scanning propensity in the unmanipulated novel exploratory cage (Table 2 1). In general, increasing risk of potential attack encoded in calls broadcast in this study decreased all compon ent behaviors that involved active movement (hops, flights, thoroughness of ground and perch exploration and active scan number) while still scanning and active scanning time increased (Figure 3 1). Notable exceptions included an increasing proportion of perches explored (and slight increase in flight numbers) across time periods in the low risk mobbing treatment, suggesting that slightly elevated predation risk caused cardinals to seek higher vantage points and potentially greater overall exploration of t he cage, presumably to seek information on the threat before it becomes critical. This supports the general hypothesis that one function of mobbing aggregations is to locate a predator for future referen ce (Goodale and Kotagama 2005 ; King and Rappole, 20 00 ), while also suggesting the more specific hypothesis that cardinals used titmouse call structure to decide between two very different anti predation strategies. When the encoded threat level is low (a large clumsy owl is near; Hetrick 2006 ; Templeton e t al. 2005 ), or perhaps a greater danger is still far away, it is a good time to search vigorously
73 for the threat (eliciting greater inter perch movement), but once a threat becomes critical (e.g., titmice are signaling a dangerous predator is pe rched nea rby; Sieving et al. 20 10 ) the best bet is to make oneself less active (hide). A large amount of variation in pre treatment periods across treatments (Figure 3 1) within most component behaviors reflects the high intraspecific variation in cardinal explora tory behavior, relative to other species, that I detected in Chapter 2 (Table 2 1). The exception to this high variation lies in the active scanning measures number and time (Figure 31 E, F). The overall measures during the pre broadcast period were ver y consistent, suggesting that changes in these behaviors are very precise measures of perceived predation risk. T his relates to other findings for changes in vigilance with perceived risk (e.g. Beauchamp 2003; Jones and Whittingham 2008; Newey 2007 ), a nd environments; a measure of exploratory behavior that is largely overlooked in recent studies characterizing it in novel environments. Information, Spatial Behavior and Future Implication s Every species, as Danchin et al. (2008) point out, at some point in their lives, have to answer the following question: Do I have all that I need here and now? Then the animal has to choose when, where, and whether to move. Animals use info rmation to decide where to go, where to forage, where to breed, and where to live T herefore the study of behavior can actually be viewed as the study of decision making (Danchin et al. 2004). In turn, spatial behavior is, in many ways, the defining featu re of non sessile animals and its analysis addresses fundamental aspect s of animal decision making process es
74 Information plays a critical role in decision making. Success fully adaptive respon ses to the environment depend on the accurate estimation of rel evant ecological parameters in order to adjust behaviors (Dall et al. 2005). Even though one cannot see it a landscape of information exists and produces strong influence s on animal behavior (Plotnick et al. 2010). Social information about predation t hreats (location, type, etc.) is an important and widely produced and used class of signals and cues that, collectively, define what has been called (e.g. Apfelbach et al. 2005; Dall et al. 2005; Valeix et al. 2009; Willems and H ill 2009) Use of anti predator information can profoundly influence a species use of space, interactions, population, and community dynamics ( Cresswell 2008) ; responses collectively identified as non letha l effects of predation (Lima and Zollner 1996) Evidence is accruing that the alarm calls of titmice and other Paridae may define a landscape of fear for a large number of species that share predators with them (La ngham et al. 2006 ; Templeton and Greene 2007 ; Schmidt et al. 2008). One important co nsequence with potentially far reaching conservation implications is how parid information may influence small and large scale movements in complex landscapes, or functional connectivity for north temperate forest birds (e.g., Sieving et al. 2004). Lands cape connectivity refers to the functional relationship among habitats, owing to the spatial contagion of habitat and the movement responses of organisms to landscape structure (Taylor et al. 1993; With et al. 1997). Functional connectivity has been defi ned to express an organism centered view of movement responses that are not solely determined by landscape structure
75 ( Crooks and Sanjayan 2006 ; Wiens, 1989 ). By including explicitly behavioral motivations and constraints in models of animal movement for c omplex landscapes (Zollner and Lima 1998), the predictive power of spatial models used in conservation planning can be enhanced (Belisle 2005 ; Castell n and Sieving 2007). My work can inform future efforts to understand how behavioral (functional) conne ctivity varies in time and space. For example, the finely tuned predation et al. 20 10 ). Titmice and other Paridae are widespread and abundant in the Holarctic, and most specie s are highly vocal ( Olsen and Grubb 2007 ). A dynamic map of Parid anti predators calls would represent a landscape of fear that likely Mnkknen and Forsman 2002 ; Sieving et al. 2004 ) and foragin g (Dolby and Grubb 2000) and breeding success ( Hogstad 1995 ). Directed studies of dispersal behavior in complex landscapes, by species that participate in the parid communication network, could be an effective way to test for the influence of information on connectivity, and its relevance to conservation. Since exploratory behavior is highly correlated with the dispersal ability of birds (Dingemanse et al. 2003), my work strongly suggests that dispersal (timing, distance, path) could be influenced by tit mouse anti predator information In turn, the connectivity of landscapes for animals privy to titmouse information would be influenced in ways not visible on an aerial photograph (Figure 3 2).
76 (A) (B) (C) (D) Figure 3 1 Bar graphs showing changes in 8 components of exploratory behaviors over the three trial periods (before, during and after broadcasting) for each treatment (control, low risk mobbing, high risk mobbing, and seet call): (A) flight number, (B) hop number, (C) prop. of perches explored, (D) prop. of ground explored, (E) active scan number, (F) average active scanning time, (G) still scan number, and (H) average still scannin g time. Bars show mean values with 95% confidence intervals.
77 (E) (F) (G) (H) Figure 3 1. Continued.
78 Figure 3 2. The relationshi p among soc ial information, animal behavioral ecology and functional connectivity.
79 Table 3 1. Prediction of influence of 4 different vocalization trials through time. Sign means no changes at all, means decreasing, and means increasing. The number of signs indicates the strength of the influence. Treatment Time Flight Number Hop Number Prop. of Perches Explored Prop. of Ground Explored Active Scan Number Average Active Scanning Time Still Scan Number Average Still Scanning Time C ontrol Before vs. During During vs. After Before vs. After Low risk Mobbing Before vs. During During vs. After Before vs. After High risk Mobbing Before vs. During During vs. After Before vs. After Seet Call Before vs. During During vs. After Before vs. After
80 Table 3 2. Factor loadings of the 8 component behaviors on the 3 principal components after varimax rotation. Eigenvalues and amount of variance explained by the respective components are given at the bottom. MANOVA results from these three principal components for different compari sons are shown in the bottom panel. Factor loadings Parameter PC1 PC2 PC3 1 Flight Number 0.81 0.23 0.22 2 Hop Number 0.38 0.24 0.79 3 Prop. of Perches Explored 0.57 0.21 0.62 4 Prop. of Ground Explored 0.16 0.07 0.80 5 Active Scan Number 0.75 0.41 0.09 6 Average Active Scanning Time 0.69 0.40 0.37 7 Still Scan Number 0.17 0.92 0.07 8 Average Still Scanning Time 0.14 0.92 0.09 Eigenvalues 3.19 1.80 1.29 % of Variance 39.88% 22.45% 16.10% MANOVA results PC1 PC2 PC3 F df p valu e F df p value F df p value Treatments (During Before Diff ) 8.46 3,31 0.01 3.83 3,31 0.02 6.00 3,31 0.01 Treatments (After Before D iff ) 4.34 3,31 0.01 4.01 3,31 0.02 1.72 3,31 0.19
81 CHAPTER 4 CONCLUSION Exploratory behavior exhibits great intraspecific variation; sex, age, social status e nvironmental conditions and experience of the individual can all cause variance in exploratory behavior within species ( e. g. Arakawa 2005 ; Lodewijckx 1984). Exploratory behavior is highly correlated with other behaviors (e.g., dispersal ) and is a reliable indicator of animal pers onality, or behavior syndrome s (Dingemanse et al. 2003; Sih et al. 2004). Neophobia, one trait that has been linked to different life histories and habitat associations (Greenberg 2003), together with exploration represent two distinct spatial response s The former is a spatially negative or avoidance behavior, and the latter is a spatially positive, or attraction behavior, with respect to environmental cues. When the two interact in an individual, spatially (or decisively) ambivalent responses can re sult (Greenberg and Mettke Hoffman 2001). The environment is full of dangers and critical resources for every species at all scales and times, and spatial decision making (navigation of landscapes) must be continuously responsive to relevant information ( navigation to information scapes) to maximize gain and minimize risks in survival and reproduction ( Danchin et al. 2008 ). To enhance understanding of animal spatial ecology, measures of spatial behavior in response to stimuli must encompass the range of spatial stimulus response dynamics, from avoidance and escape to attraction. And measures of utility in landscape ecology and conservation should be useful in comparative studies among taxa of animals. Exploratory behavior as I have described it in Chapt er 2 is a step in this direction. Neophobia has been to show significant interspecific variation in avoidance behaviors that are correlated widely with species traits linked to extinction sensitivity
82 (Greenberg 1983; 2003; Mettke Hofmann et al. 2002; 2 005; 2006). W ork on animal personalities and behavioral syndromes are opening up our understanding of the mechanisms underlying ecological variation of importance to animal conservation (e.g., dispersal, reproduction). However, investigation of behavior co mplexes linked to spatial exploration has been focused primarily within species. Therefore, the strong links that have been established between exploratory behavior and its diverse correlates in personality ( Sih et al. 2004 ), reproduction ( Dingemanse and R ale 2005 ), and movement ( Dingemanse et al. 2003 ), do not allow interspecific comparisons. To provide tools and insights for enhancing interspecific analyses of behavior syndromes, my first experiment focused on interspecific comparison of exploratory behavior. I used multivariate approach es to represent exploratory behavior instead of the simplified and traditional exploratory score The 8 component activities comprising exploratory behavior include d multiple measures each of movement, exploratory tho roughness, and scanning exhibited during exploration inside a novel cage environment (Table 2 1) Variation across all 8 components was significant across explorer s (Table 2 2, 2 5 Figure 2 3 A); some component behaviors were more sensitive than the others (Table 2 4). The most appropriate grouping derived from a model that placed species together that either did or did not participate in winter foraging flocks ( 96 .3% correct classification of species into the two groups); higher than a model that used foraging guild (7 7.8% ), or no groupings (species all separate; 72.8% ) F locking species exhibited a higher mean number of flights, and non flocking species exhibited a larger proportion of ground explored ( Figure 2 3 C ) F indings confirm that
83 interspecific groupings of exploratory behavior should reflect interspecific groupings of correlated behavior syndromes ( Greenberg, 2003; Mettke Hofmann et al. 2002 ). The multiv ariate approach used here achieves a broader, more complete view of exploration, beyond what previous measures of exploratory or neophobic behaviors could capture. And i nterspecific comparison s provide d insights on how behavior (decision making) may be sha ped by the interaction of an animal and its environment (Figure 4 1, left side) To explore another aspect of h ow interspecific interaction s can influence exploratory behavior, and the behavioral syndromes related to it, I conducte d a second experiment involving provi sion of social information from the tufted titmouse to Northern cardinals. The anti predator vocalization s are highly conserved in the Paridae, provide reliable information about predation threats occurring in different contexts in the environment, and those of the tufted titmouse are representative of the family. chick a dee and seet calls (for attack) have been well characterized and encode tremendous specificity with respect to predation ri sk ( Sieving et al. 2009 ; Templeton et al., 2005 ) Many forest bird species in North America respond appropriately to titmouse anti predator calls ( Hetrick 2006 ; Hurd 1996; Langham et al. 2006; Sieving et al. 2004); including Northern cardinals Using calls with known different levels of predation threat encoded in the m, I broadcasted them to cardinals and was able to set clear expectations about how exploratory behavioral components should change given appropriate ant predator responses (Table 3 1). R esults matched prediction s quite well: with increasing threat level encoded in titmouse vocalizations, active movement s declined to a greater degree (Figure 3 1 A, B) Measures of
84 thoroughness of space explored generally followed the same pattern as flight s and hops (above; Figure 3 1 D). Scanning behavior, generally held to reflect vigilance levels, were influenced by vocalization treatments as well. Cardinals significantly increas ed their scanning time with increasing predation risk (Figure 3 1 F) and the rarely seen still scanning behavior was induced by the highest risk treatments (Figure 3 1 G, H). Thus, making was very finely tuned to, and influenced by, socially derived information about predation risk (Figure 4 1, right si de). Though I only used one species in this experiment, results from the first experiment strongly suggest that the changes in exploratory behavior with perceived risk (as from social information) will manifest differently across species. Ground, canopy, mid story, or social versus solitary species, because of the unique combinations of behavioral components in the make up of their behavioral syndromes, will act out avoidance or attraction or vigilance in different ways. T he results of this experiment hav e generated many ideas about how exploratory behavior and its correlates, can be affected by heterospecific interaction through social information exchange. In summary, I established first, how personal information gathering expressed as exploratory be havior varied across species with different and similar natural history traits and how ecologically relevant social information can affect exploratory behavior at a small scale. Given the profoundly strong correlations established between many aspects of behavioral ecology and exploratory behavior, including large scale spatial behaviors such as dispersal, future studies of exploratory behavior may provide understand ing of functional connectivity of landscapes (Figure 4 1) For so many species, movement among critical habitat patches and regions will be the key to
85 persistence and continued evolution under the dramatic land cover and climate change scenarios that are being developed for global environments (Virkkala et al. 2008 ). This analysis provides an other potentially useful tool for helping to understand and predict species persistence in a rapidly changing world.
86 Figure 4 1. Linking knowledge from this study to future researches and application goals.
87 LIST OF REFERENCE S Adami C 2004 Information theory in molecular biology. PhLRv 1 : 3 22. Apfelbach R, Blanchard CD Clanchard RJ Hayes RA, McGregor IS 2005 The effect of predator odors in mammalian prey species: A review of field and laboratory studies. Neurosci Biobeha v Rev 29:1123 1144. Arakawa H 2005 Interaction between isolation rearing and social development on exploratory behavior in male rats. Behav Processes 70:223 234. Arlt D Part T 2008 Post breeding information gathering and breeding territory shifts in n orthern wheatears. J Anim Ecol 77:211 219. Arno ld K, Pohlner Y, Zuberbugler K, 2007. A forest monkey s alarm call series to predator models. Behav Ecol Sociobiol 62: 549 559. Baker MC, Becker AM 2002 Mobbing calls of black capped chickadees: effects of ur gency on call production. Wilson Bull 114:510 516. Batten JS, 2008. The white eyed eastern towhee: a molecular assessment of the validity of Pipilo erythrophthalmus alleni Master of Science thesis. Las Vegas: University of Nevada. Bazzaz FA 1991 Habit at selection in plants. Am N at 137:116 130. Beauchamp G, 2003. Group size effects on vigilance: a search for mechanisms. Behav Process 63: 111. Belisle M, 2005. Measuring landscape connectivity: the challenge of behavioral landscape ecology. Ecology 86: 1988 1995. Bell A M Sih A 2007 Exposure to predation generates personality in three spined sticklebacks ( Gasterosteus aculeatus ). Ecol Lett 10:828 834. Blumstein DT 1999 Alarm calling in three species of marmots. Behaviour 136:731 757. Bonnie KE Earley RL 2007 Expanding the scope for social information use. Anim Behav 74:171 181. Both C, Dingem anse NJ, Drent PJ, Tinbergen JM 2005 Pairs of extreme avian personalities have highest reproductive success. J Anim Ecol 74:667 674. Brown JH, Gillooly JF, Allen AP, Savage VM West GB 2004 Towards a metabolic theory of ecology. Ecology 85 : 1771 1789.
88 Budaev SV 1997 "Personality" in the guppy ( Poecilia reticulata ): a correlational study of exploratory behavior and social tendency. J Comp Psychol 111:399 411. C andolin U 2003 The use of multiple cues in mate choice. Biol Rev 78:575 595. Carere C, Drent PJ Privitera L Koolhaas JM Groothuis TGG 2005 Personalities in great tits, Parus major : stability and consistency. Anim Behav 70:795 805. Caro T 2005 Anti predator defenses in birds and mammals. Chicage : Uni versity of Chicago Press Casellas E, Gautrais J Fournier R Blanco S Combe M Fourcassie V Theraulas G, Jost C 2008 From individual to colle c tive displacements in heterogeneous environments J Theor Biol 250:424 434. Chen C Hsieh F 2002 Composition and foraging behaviour of mixed species flocks led by the Grey cheeked Fulvetta in Fushan Experimental Forest, Taiwan. Ibis 144:317 330. Cimprich DA, Grubb TC 1994 Consequences for Carolina Chickadees of foraging with Tufted Titmice in winter. Ethology 75:1615 1625. Clara E, Tomassi L Rogers LJ 2008 Social mobbing calls in common marmosets ( Callithrix jacchus ): effects of experience and associated cortisol levels. Anim Cogn 11:349 358. Colles A, Lio w LH, Prinzing A 2009 Are specialists at risk under environmental change? neoecological, paleoecological and phylogenetic approaches. Ecol Lett 12:849 863. Coolen I, V anBergen Y Day RL, Laland KN 2003 Species differences in adaptive use of public info rmation in sticklebacks. Proc R Soc B 270:2413 2419. Cornelius C, 2010, Feb 22. Email contact. Creel S Christianson D 2008 Relationships between direct predation and risk effects. Trends Ecol Evol 23:194 201. Cresswell W 2008 Non lethal effects of pre dation in birds. Ibis 150:3 17. Dall SRX, Giraldeau L Olsson O McNamara JM, Stephens DW 2005 Information and its use by animals in evolutionary ecology. T rends Ecol Evol 20:187 193. Danchin E Giraldeau LA, Valone TJ, Wagner RH 2004 Public informatio n: from noisy neighbors to cultural evolution. Science 305 : 487 491 Danchin E, Giraldeau LA, Cezilly F 2008 Behavioural Ecology New York: Oxford University Press Inc
89 Dawson MRW, Bloomfield LL, Charrier I, Sturdy CB, 2006. Statistical classification of b lack capped ( Poecile attricapillus ) and mountain chickadee ( Poecile gambeli ) call notes. J Comp Psychol 120: 147 153. De Graaf RM, Tilghman NG Anderson SH 1985 Foraging guilds of North American birds. Environ M anage 9:493 536. Desrochers A Hannon SJ 199 7 Gap crossing decisions by forest songbirds during the post fledging period. Conserv Biol 11:1204 1210. Devictor V, Clavel J Julliard E Lavergne S Mouillot D Thuiller W Venail P Vill'eger S, Mouquet N 2010 Definin g and measuring ecological specia lization. J Appl Ecol 47:15 25. Dickinson JC, Jr., 1952 Geographic variation in the red eyed towhee of the eastern United States. Bull Mus Comp Zool 107:271 352. Dingemanse NJ, Both C Drent PJ V anOers K, V anNoordwijk AJ 2002 Repeatability and heritabil ity of exploratory behavior in great tits from the wild. Anim Behav 64:929 938. Dingemense NJ, Both C, Van Noordwijk AJ Drent PJ 2003 Natal Dispersal and personalities in great tits ( Parus major ). Philos T rans R Soc London Ser B 270:741 747. Dingemance N J De Goede P 2004 The relation between dominance and exploratory behavior is context dependent in wild great tits. Behav Ecol 15:1023 1030. Dingemanse NJ, Both C, Drent PJ Tinbergen JM 2004 Fitness consequences of avian personalities in a fluctuating environment. Philos T rans R Soc London Ser B 271:847 852. Dingemanse NJ Reale D 2005 Natural selection and animal personality. Behaviour 142:1165 1190. Dobson FS, 1984. The use of phylogeny in behavior and ecology. Evolution 39: 1384 1388. Dolby AS Gru bb TC Jr ., 1999 Functional roles in mixed species foraging flocks: a field manipulation Auk 116:557 559. Dolby AS Grubb TC Jr ., 2000 Social context affects risk taking by satellite species in a mixed species foraging group. Behav Ecol 11:110 114. Dre nt PJ, van Oers K, van Noordwijk AJ, 2003. Realized heritability of personalities in the great tit (Parus major). Proc R Soc Lond B 270:45 51.
90 Drullion D Dubois F 2008 Mate choice copying by female zebra finches, Taeniopygia guttata : what happens when m odel female provide inconsistent information? Behav Ecol Sociobiol 63:269 276. Dugatkin LA, Godin J GJ 1992. Prey approaching predators: a cost benefit perspective. Ann Zool Fennici 29: 233 252. Enfjall K and Leimar O 2009 The evolution of dispersal the importance of information about population density and habitat characteristics. Oikos 118:291 299. Enrlich PR, Dobkin DS, Wheye N 1988 The birder's handbook: a field guide to the natural history of N orth A merican birds, pp 422 562 New York: Simon & Schuster Inc. Farley EA, Sieving KE Contreras TA 2008 Characterizing complex mixed species bird flocks using an objective method for determining species participation. J Ornitho l 149:451 468. Fernandez Juricic E, Delgado JA Remacha C Jimenez MD Garci a V, Hori K 2009 Can a solitary avian species use collective detection? an assay in semi natural condition. Behav Processes 82:67 74. Ficken MS 1989 Acoustic characteristics of alarm calls associated with predation risk in Chickadees. Anim Behav 39:400 401. Fletcher RJ, Jr, 2008. Social information and community dynamics: nontarget effects from simulating social cues for management. Ecol Appl 18: 1764 1773. Fletcher RJ, Jr, 2009. Does attraction to conspecifics explain the patch size effect? An experimen tal test. Oikos 118: 1139 1147. Fletcher RJ Jr., Miller CW 2008 The type and timing of social information alters offspring production. Biol Lett 4:482 485. Fle t cher RJ, Jr., Sieving KE, 2010. Social information use in heterogeneous landscapes: a prospect us. Condor In press. Fletcher N, Storey EJ Johnson M Reish DJ, Hardege JD 2009 Experience matters: female use smell to select experienced males for parental care. PloS One 4:e7672. Fors man JT, Thomson RL, Seppanen J, 2007. Mechanisms and fitness effect of interspecific information use between migrant and resident birds. Behav Ecol 18: 888 894. Forsman JT, Hjernquist MB, Gustafsson L 2009 Experimental evidence for the use of density based interspecific social information in forest birds. Ecography 32:5 39 545.
91 F reeberg TM, Lucas JR, Clucas B 2003 Variation in chick a dee calls of a population of Carolina chickadees, Poecile carolinensis : identity and redundancy within note types. J Acoust Soc Am 113:2127 2136. Fridley JD, Vandermast DB Kuppinger DM M anthey M, Peet RK 2007 Co occurrence based assessment of habitat generalists and specialists: a new approach for the measurement of niche width. J Ecol 95:707 722. Gaddis PK 1985 Structure and variability in the vocal repertoire of the mountain chickad ee. Wilson Bull 97:30 46. Ghizoni IR, Jr Guimaraes Azevedo MA 2006 Composition of mixed flocks of understory forest birds in areas of the Atlantic coast and lowlands of Santa Catarina State, southern Brazil. Biotemas 19:47 53. Gillooly FG Ophir AG 200 9 The energetic basis of acoustic communication. Proc R Soc B 277:1325 1331. Goodale E Kotagama SW 2005 Testing the roles of species in mixed species bird flocks of a Sri Lankan rain forest. J Trop Ecol 21:669 676. Goodale E, Beauchamp G, Magrath RD, N ieh JC, Ruxton GD, 2010. Interspecific information trandfer influences animal community structure. In press. Gosling SD 2001 From mice to men: what can we learn about personality from animal research. Psychol Bull 127:45 86. G ram WK 1998 Winter partici pation by Neotropical migrant and resident birds in mixed species flocks in northeastern Mexico. Condor 100:44 53. Greenberg R 1983 The role of neophobia in determining the degree of foraging specialization in some migrant warblers. Am Nat 122:444 453. G reenberg R 1984 Differences in feeding neophobia in the tropical migrant wood warblers Dendroica castanea and D. pensylvanica J Comp Psychol 98:131 136. G reenberg R 1990 Feeding neophobia and ecological plasticity: a test of the hypothesis with captiv e sparrows. Anim Behav 39:375 379. Greenberg R 1992 Differences in neophobia between naive Song and Swamp sparrows. Ethology 91:17 24. Greenberg R 2000 Birds of many feathers: the formation and structure of mixed species flocks of forest birds. In On t he move: how and why animals travel in groups ed Boinski S, Garber PA pp 521 558. Chicage: University of Chicago Press.
92 Greenberg R 2003 The role of neophobia and neophilia in the development of innovative behaviour of birds. In Animal Innovation e d R eader SM, Laland KN pp 175 196. Oxford: Oxford University Press. Greenberg R Droege S 1999 On the decline of the rusty blackbird and the use of ornithological literature to document long term population trends. Conserv Biol 13:533 559. Greenberg R Met tke Hofmann C 2001 Ecological aspects of neophobia and neophilia in birds. Curr Ornithol 16 : 119 178. Griesser M Nystrand M 2009 Vigilance and predation of a forest living bird species depend on large scale habitat structure. Behav Ecol 20:709 715. Gri ffin AS, Savani RS Hausmanis K, Lefebvre L 2005 Mixed species aggregations in birds: zenaida doves, Zenaida aurita respond to the alarm calls of carib grackles, Quiscalus lugubris Anim Behav 70:507 515. Grim T 2008 Are blackcaps ( Sylvia atricapilla ) defending their nests also calling for help from their neighbours? J Ornithol 149:169 180. Grubb TC, Jr Bronson CL 2001 On cognitive conservation biology: why chickadees leave a patch of woodland. J Avian Bio l 32:372 376. Grubb TC, Jr Doherty PF Jr 1999 On homerange gap crossing. Auk 116:618 628. Haftorn S 2000 Contexts and possible functions of alarm calling in the willow tit, Parus montanus ; The principle of 'better safe than sorry'. Behav 137:437 449. Hailman JP 1989 The organization of majo r vocalizations in the Paridae. Wilson Bull 101:305 343. Harrap S, Quinn D, 1995. Chickadees, tits, nuthatches and treecreepers New Jersey : Princeton University Press. Harvey PH Pagel MD 1993 The comparative method in evolutionary biolo gy. Oxford : Oxfo rd University Press Herzog SK, Soria AR Troncoso A Matthysen E 2002 Composition and structure of avian mixed species flocks in a High Andean Polylepis forest in Bolivia. Ecotropica 8: 133 143 Hetrick SA 2006 Investigation of tufted titmouse ( Baeolo phus bicolor ) anti predator vocalization. M aster of Sc ience thesis Gainesville: University of Florida. Hoeksema JD, Schwartz MW 2001 Modeling interspecific mutualisms as biological markets. I n Economics in nature ed Ne R van Hooff J Hammerstein P, p p 173 186. Cambridge: Cambridge University Press
93 Hogstad O, 1995. Alarm calling by willow tits, Parus montanus as mate investment. Anim Behav 49: 221 225. Hopf S 1985 Development of attachment and exploratory behavior in infant squirrel monkeys under c ontrolled rearing conditions. Int J Behav Dev 8:55 74. Hurd CR, 1996. Interspecific attraction to the mobbing calls of black capped chickadees ( Parus atricapillus ). Behav Ecol Sociobiol 38: 287 292. Hutchinson G.E. 1957 Concluding remarks Cold Spring Har b Symp Quant Biol 22 : 415 457. H utto RL 1987 A description of mixed species insectivorous bird flocks in western Mexico. Condor 89:282 292. H utto RL 1988 Foraging behavior patterns suggest a possible cost associated with participation in mixed species bird flocks. Oikos 51:79 83. Hutto RL 1994 The composition and social organization of mixed species flocks in a tropical deciduous forest in western Mexico. Condor 96:105 117. Immelmann K Beer C 1989 A dictionary of ethology Cambridge: Harvard Univer sity Press Ito R, Mori A, 2010. Vigilance against predators induced by eavesdropping on heterospecific alarm calls in a non vocal lizard Oplurus cuvieri cuvieri (Reptilia: Iguania). Proc R Soc Long B Bio 277:1275 1280. Johnson ML 1988 Exploratory behavi or and dispersal: a graphical model. Can J Zoo l 67:2325 2328. J ones SE 1977 Coexistence in mixed species antwren flocks. Oikos 29:366 375. Jones GA, Sieving KE 2006. Intercropping sunflowers in organic vegetables to augment avian predators of arthropod pests. Agric Ecosyst Environ 117: 171 177. Jones KL, Whittingham MJ, 2008. Anti predator signals in the Chaffinch Fringilla coelebs in response to habitat structure and different predator types. Ethology 114: 1033 1043. Jullien M, Thiollay JM 1998 Multi species territoriality and dynamic of Neotropical forest understorey bird flocks. J Anim Ecol 67:227 252 King DI Rappole JH 2000 Winter flocking of insectivorous bird in montane pine oak forests in middle America. Condor 102:664 672. Kingbeil BT Willig MR 2009 Guild specific responses of bats to landscape composition and configuration in fragmented Amazonian forest. J Appl Ecol 46:203 213.
94 Kleerekoper H, Matis J, Gensler P Maynard P 1974 Exploratory behavior of goldfish Carrassius auratus Anim Beh av 22 : 124 132. Klopfer PH 1967 Behavioral stereotypy in birds. Wilson Bull 79:290 300. Kroodsma DE 1989. Suggested experimental designs for song playbacks. Anim Behav 37: 600 609. Kroodsma DE Miller EH, 1996. Ecology and evolution of acoustic communic ation in birds. New York : Cornell University Press. Kuehl AK, Clark WR 2002 Predator activity related to landscape features in northern Iowa. J Wildl Manage 66:1224 1234. Langha m GM, Contreras TA, Sieving KE 2006 Why pishing works: Titmouse (Paridae) scolds elicit a generalized response in bird communities. Ecoscience 13:485 496. L atta SC, W underle JM 1996 The composition and foraging ecology of mixed species flocks in pine forests of Hispaniola. Condor 98:595 607. Lefebvre L, Whittle P Lascaris E, Finkelstein A 1997 Feeding in n ovations and forebrain in birds. Anim Behav 53:549 560. Levins R 1968 Evo lution in c hanging e nvironments Princeton: Princ eton University Press. Lima SL, Dill LM 1990 Behavioral d ecisions m ade under the r isk of p redation a r eview and p rospectus. Can J Zool 68:619 640. Lima SL 1995 Back to the basics of anti predatory vigilance: the group size effect. Anim Behav 49:11 20. Lima SL 1998 Nonlethal effects in the ecology of predator prey interactions. Biosci 48:23 34. Li ma SL Zollner PA 1996 Anti predatory vigilance and the limits to collective detection: visual and spatial seperation between foragers. Behav Ecol Sociobiol 38:355 363. Lodewi jckx E 1984 The influence of sex, sexual condition and age on the exploratory behavior of wild wood mice ( Apodemus sylvaticus L. ). Behav Processes 9:431 444. Long ES, Diefenbach DR Rosenberry CS Wallingford BD, Grund MRD 2005 Forest cover influences dispersal distance of white tailed deer. J Mammal 86:623 629. Ma nning AD, Gordo n IJ, Ripple WJ, 2009. Restoring landscaped of fear with wolves in the Scottish Highlands. Biol Cons 142:2314 2321.
95 Martin P, Bateson P, 2007. Measuring behaviour: an introductory guide. Cambridge: Cambridge University Press. Martin J, Lopez P 2009 Multi ple color signals may reveal multiple messages in male Schreiber's green lizards, Lacerta scheriberi Behav Ecol Sociobiol 63:1743 1755. Marzluff JM, Heinrich B Marzluff CS 1996 Raven roosts are mobile information centres. Anim Behav 51:89 103. Mayeaux DJ Mason WA 1998 Developement of responsiveness to novel objects in the titit monkey, Callicebus moloch Primates 39:553 560. Maynard Smith J 1999 The idea of information in biology. Q Rev Biol 74 : 395 400 McKinney ML 1997 Extinction vulnerability an d selectivity: combining ecological and paleontological views. Annu Rev Ecol Syst 28 : 495 516. Mennill DJ, Ratcliffe LM, Boag PT 2002 Female eavesdropping on male song contest in songbirds. Science 296:873. Mettke Hofmann C, Winkler H, Leisler B 2002 Th e significance of ecological factors for exploration and neophobia in parrots. Ethology 108:249 272. Mettke Hofmann C, Ebert C Schmidt T Steiger S Stieb S 2005 Personality traits in resident and migratory warbler species. Behaviour 142:1363 1381. Mett ke Hofmann C, Rowe KC Hayden TJ, Canoine V 2006 Effects of experience and object complexity on exploration in garden warblers ( Sylvia borin ). J Zool 268:405 413. Mitchell D 1976 Experiments on neophobia in wild and laboratory rats: a reevaluation. J C omp Phys Psych 90:190 197. Mnkknen M, Forsman JT 2002 Heterospecific attraction among forest birds: a review. Ornithol Sci 1:41 51. Morse DH 1977 Feeding behavior and predator avoidance in heterospecific groups. Bio s cience 27 : 332 339. Morton ES 198 8 Pairing in mallards and american black ducks: a new view on population decline in american black ducks. Anim Cons 1 :239 244. Munn CA Terborgh JW 1979 Multi species territoriality in neotropical foraging flocks. Condor 81:338 347. Negro JJ, Bustamante J Milward J, Bird DM 1996 Captive fledgling american kestrels prefer to play with objects resembling natural prey. Anim Behav 52:707 714.
96 Newey P, 2007. Foraging behaviour of the common myna ( Acridotheres tristis ) in relation to vigilance and group siz e. EMU 4: 315 320. Olsen JR Grubb TC Jr ., 2007 Winter adaptations in chickadees and titmice and the added effect of habitat fragmentation. In Ecology and Behavior of Chickadees and Titmice: An Integrated Approach ed Otter KA pp 263 273. Oxford: Oxford University Press. Otter K (ed.) 2007. The ecology and behavior of chickadees and t itmice: an integrated approach. Oxford: Oxford University Press. Parejo D White J Clobert J, Dreiss A, Danchin E 2007 Blue tits use fledgling quantity and quality as pu blic information in breeding site choice. Ecology 88:2373 2382. Plotnick RE, Dornbos SQ, Chen J 2010 Information landscapes and sensory ecology of the Cambrian Radiation. Paleobiology 36: 303 317. P owell GVN 1985 Sociobiology and adaptive significance o f interspecific foraging flocks in the Neotropics. Ornithol Monogr 36:713 372. Ragusa Netto J 2002 Vigilance toward raptors by nuclear species in bird mixed flocks in a Brazilian savannah. Stud Neotrop Fauna Environ 37:219 226. Raud S, Sutt S Plaas M L uuk H Innos J Philips MA Koks S, Vasar E 2007 Cat odor exposure induces distinct changes in the exploratory behavior and Wfs1 gene expression in C57B1/6 and 129Sv mice. Neuroscience 426:87 90. Reale D Festa Bianchet M 2003 Predator induced natural selection on temperament in bighorn ewes. Anim Behav 65:463 470. Reale D, Gallant BY, Leblanc M Festa Bianchet M 2000 Consistency of temperament in bighorn ewes and correlates with behaviour and life history. Anim Behav 60:589 597. Phillips BL Brown GP Shine R 2004 Assessing the potential for an evolutionary response to rapid environmental change: invasive toads and an Australian snake. Evol Ecol Res 6:799 811. Prei sser EL, Orrock JL, Schmitz OJ, 2007. Predator hunting mode and habitat domain alter n onconsumptive effects in predator prey interactions. Ecology 88: 2744 2751. Roberts DL, Cooper RJ, Petit LJ 2000 Flock characteristics of ant following birds in premontane moist forest and coffee agroecosystems. Ecol Appl 10:1414 1425. Roche BE, Schulte Hostedee AI, Brooks RJ 1988 Route choice by deer mouse ( Peromyscus maniculatus ): Reducing the risk of auditory detection by predators. Am Midl Nat 142:194 197.
97 Rule M 1993 Songbird diet index: a guide to the natural food habits of Ohio songbirds and su bstitute diets for use in wildlife rehabilitation facilities pp 26 141 Pompano Beach : C oconut Creek Publishing Company P ulliam HR 1973 On the advantages of flocking. J T heor Biol 38:419 422. Schmidt KM, Lee E Ostfeld RS Sieving KE 2008 Eastern chi pmunks increase their perception of predation risk in response to titmouse alarm calls. Behav Ecol 19:759 763. Seppanen J, Forsman JT Monkkonen M, Thomson RL 2007 Social information use is a process across time, space, and ecology, reaching heterospecif ics. Ecology 88:1622 1633. Sieving KE, Willson MF, 1998. Nest predation and avian species diversity in northwestern forest understory. Ecology 79: 2391 2402. Sieving KE, Contreras TA, Maute KL 2004 Heterospecific facilitation of forest boundary crossing by mobbing understory birds in North central Florida. Auk 121:738 751. Sieving KE, Hetrick SA, Avery ML 20 10 The versatility of graded acoustic measures in classification of predation threats by the tufted titmouse Baeolophus bicolor : exploring a mixed f ramework for threat communication. Oikos 119:264 276. Sih A, Kats LB, Maurer EF 2003 Behavioural correlations across situations and the evolution of antipredator behaviour in a sunfish salamander system. Anim Behav 65:29 44. Sih A, Bell AM Jo h nson JC Z iemba RE 2004 Behavioral syndromes: an integrative overview. Q Rev Biol 79:241 277. Skinner BF, 1984. The phylogeny and ontogeny of behavior. Behav Brain Sci 7: 669 677. Smith JW, Benkman CW, Coffey K 1999 The use and misuse of public information by re d crossbils. Behav Ecol 10:54 62. Smith RG, Kember RL Mill J Fernandes C Schalkwyk L Leonard C Buxbaum JD, Reichenberg A 2009 Advancing paternal age is associated with deficits in social and exploratory behaviors in the offspring: a mouse model. Plo S One 4:e8456. Soard CM, Rit chison G, 2009. Chick a dee calls of Carolina chickadees convey information about degree of threat posed by avain predators. Anim Behav 78: 1447 1453. Somasundaram S Vijayan L 2008 Foraging behavior and guild structure of b irds in the Montane wet temperate forest of the Palni hills, South India. Podoces 31:79 91.
98 Sridhar H, Beauchamp G, Shanker K 2009 Why do birds participate in mixed species foraging flocks? A large scale synthesis. Anim Behav 78:337 347. Stamps J 2003 question comes of age. Anim Behav 66:1 13. Stephens DW 1989 Variance and the value of information. Am Nat 134 : 128 140 Taylor PD, Fahrig L, Henein K, Merriam G, 1993. Connectivity is a vital element of landscape structure. Oikos 68 : 571 573. Templeton JJ Giraldeau LA 1995 Patch assessment in foraging flocks of European starlings: evidence for the use of public information. Behav Ecol 6:65 72. Templeton CN, Greene E 2007 Nuthatches eavesdr op on variations in heterospecific chickadee mobbing alarm calls. PNAS 104:5479 5482. T empleton CN, Greene E Davis K 2005 Allometry of alarm calls: Black capped chickadees encode information about predator size. Science 308:1934 1937. Terborgh J 1990 Mixed flocks and polyspecific associations: costs and benefits of mixed groups to birds and monkeys. Am J Primatol 21 : 87 100 Thiollay J Jullien M 1998 Flocking behavior of foraging birds in a neotropical rain forest and the antipredator defence hypoth esis. Ibis 140:382 394. Thoms on RL, Forsman JT, Monkkonen M 2003 Positive interactions between migrant and resident birds: testing the heterospecific attraction hypothesis. Oecologia 134:431 438. Treves A, 1998. The influence of group size and neighbors on vigilance in two species of arboreal monkeys. Behaviour 135: 453 481. Tubelis DP 2007 Mixed species flocks of birds in the Cerrado, South America: a review. Ornitologia Neotropical 18:75 97. Valeix M, Loveridge AJ Chamaille Jammes S Davidson Z Muri ndagomo F Fritz H, Macdonald DW 2009 Behavioral adjustments of African herbivores to predation risk by liond: spatiotemporal variations influence habitat use. Ecology 90:23 30. Valone TJ Templeton JJ 2002 Public information for the assessment of qual ity: a widespread social phenomenon. Philos T rans R Soc London Ser B 357:1549 1557. Van Oers K, Klunder M, Drent PJ 2005 Context dependence of personalities: risk taking behavior in a social and nonsocial situation. Behav Ecol 16:716 723.
99 Varasteh Moradi H Zakaria M 2009 Insectivorous bird community changes along an edge interior gradient in an isolated tropical rainforest in Peninsular Malaysia. Malay Nat J 61:48 66. Verbeek MEM, Drent PJ Wiepkema PR 1994 Consistent individual differences in early e xploratory behaviour of male great tits. Anim Behav 48 : 1113 1121. Virkkala R, He ikkinen RK, Leikola N, Luoto M, 2008. Projected large scale range reductions of northern boreal land bird species due to climate change. Biol Cons 141: 1343 1353. Wagner A 200 7 From bit to it: How a complex metabolic network transforms information into living matter. BMC Syst Biol 1:33. Waite TA, Grubb TC Jr ., 1988 Copying of foraging locations in mixed species flocks of temperate deciduous woodland birds: an experimental st udy. Condor 90:132 140. Walker ABD, Parker KL, Gil lingham MP, Gustine DD, Lay RJ, 2007. Habitat selection by female Stone s sheep in relation to vegetation, topography, and risk of predation. Ecoscience 14: 55 70. Ward MP 2005 Habitat selection by disper sing yellow headed blackbirds: evidence of prospecting and the use of public information. Oecologia 145:650 657. Webster S Lefebvre L 2000 Neophobia in the lesser Antillean bullfinch, a foraging generalist, and the bananaquit, a nectar specialist. Wilso n Bull 112:424 427. Welbergen JA Davies NB 2008 Reed warblers discriminate cuckoos from sparrowhawks with graded alarm signals that attract mates and neighbours. Anim Behav 76:811 822. Whittaker RH 1956 Vegetation of the Great Smoky Mountains. Ecol M onog 26 : 1 80. Wiens J A 1989 Spatial scaling in ecology. Funct Ecol 3:385 397. Willems EP Hill RA 2009 Predator specific landscapes of fear and resource distribution: effects on spatial range use. Ecology 90:546 555. Willis EO Oniki Y 1992 Birds and army ants. Bol Mus Para Emilio Goeldi ser Zool 8:123 150. Wilson ADM Godin JJ 2009 Boldness and behavioral syndromes in the bluegill sunfish, Lepomis macrochirus. Behav Ecol 20:231 237. With KA, Gardner RH, Turner MG 1997 Landscape connectivity and p opulation distributions in heterogeneous environments. Oikos 78:151 169.
100 Witkowski ETF Lamont BB 1997 Does the rare Banksia goodii have inferior vegetative, reproductive or ecological attributes compared with its widespread co occurring relative B. gard neri? J Biogeogr 24 : 469 482. Wolfe JL 1969 Observations on alertness and exploratory behavior in the eastern chipmunk. Am Midl Nat 81:249 253.
101 BIOGRAPHICAL SKETCH Ping Huang was born and raised in Taiwan. She is the youngest child and has one older br other. Ping Huang inherited her parents appreciation o f nature. Observing the metamorphosis of caterpillars, watching firebugs flying in front of her grandfather s farm visiting the wetland preserve of Black faced Spoonbill in Tainan and ot her similar ou tdoor activities we re always a part of life in her childhood upbringing. Her interests in animals and natur al science rooted on those experiences. Even t hough diagnosed as a cancer patient at age 16, she never ga ve up her dream in studying biology and ecol ogy in the future Actually, the disease tau ght her to cherish what she has and to own the faith on herself. A ll the struggles paid off eventually Ping Huang later entered Life Science Department in Tunghai University, Taiwan. It was during her years as a n undergraduate student there that she learn ed her life s passion would be behavioral ecology, ornithol ogy and conservation activities She graduated from Tunghai University with a BS degree in the s pring of 200 6. After spending times assisting several dif ferent field researches, she fortunately got the chance to enter the master s program in Wildlife Ecology and Conservation Department, University of Florida in the fall of 2008. She then chose bird exploratory behavior as h er master s study. She just got the Ph D admission from Department of Biology, University of Florida; the higher education will start in the fall semester of 2010 Her long term career goal is to find a position in a museum, government agency, non governmental organization concerned with conservation issues, or a university department where she can conduct basic research es and educate more people about the urgency of our environmental conservation issues as well