USF Libraries
USF Digital Collections

Altered features of female pigeons (Columba livia) elicit preference behavior in male pigeons

MISSING IMAGE

Material Information

Title:
Altered features of female pigeons (Columba livia) elicit preference behavior in male pigeons
Physical Description:
Book
Language:
English
Creator:
Patton, Tadd B
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla
Publication Date:

Subjects

Subjects / Keywords:
Bird -- Face Recognition -- Conspecific Recognition -- Vision -- Mate Selection
Face recognition
Conspecific recognition
Vision
Mate selection
Dissertations, Academic -- Psychology -- Masters -- USF
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Abstract:
ABSTRACT: Previous research has shown that male pigeons (Columba livia) respond with courtship displays to video playbacks of a female pigeon indicating that they 'recognize' the female as a potential mate. Courtship displays significantly decline when the head region of the female is occluded (Shimizu, 1998) suggesting that features located within the head are important for species recognition and mate selection. However, little is known about the exact visual features necessary to elicit displays. The current study examined the preference behavior of male pigeons when given a choice between photographic images of normal looking or altered female pigeon faces. The altered-face categories included: 1) enlarged or removed facial features such as the eyes or beak; 2) the eyes and beak reconfigured within the head; and 3) removed contour (outline). The results showed that subjects responded preferentially toward females with enlarged features (eyes or beak). However, subjects responded preferentially toward normal females when the alternative stimuli were faces that were missing the eyes and/or beak. Preference for normal females was also observed when females with "incorrect" configuration were shown. Finally, subjects responded significantly less to females lacking contour, even when the eyes and beak were visible. The overall findings suggest local facial components are important, although this effect diminishes if the contour of the female is not visible. These findings also suggest that pigeons attend to both local components and global configuration when they detect conspecifics and identify potential mates.
Thesis:
Thesis (M.A.)--University of South Florida, 2006.
Bibliography:
Includes bibliographical references.
System Details:
System requirements: World Wide Web browser and PDF reader.
System Details:
Mode of access: World Wide Web.
Statement of Responsibility:
by Tadd B. Patton.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 40 pages.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001911072
oclc - 173642317
usfldc doi - E14-SFE0001656
usfldc handle - e14.1656
System ID:
SFS0025974:00001


This item is only available as the following downloads:


Full Text

PAGE 1

Altered Features of Female Pigeons ( Columba livia ) Elicit Preference Behavior in Male Pigeons by Tadd B. Patton A thesis submitted in partial fulfillment of the requirements for the degree of Master of Arts Department of Psychology College of Arts and Sciences University of South Florida Major Professor: Toru Shimizu, Ph.D. Michael Brannick, Ph.D. Sarah Partan, Ph.D. Thomas Sanocki, Ph.D. Date of Approval: May 19, 2006 Keywords: Bird, Face Recognition, Conspeci fic Recognition, Vision, Mate Selection Copyright 2006, Tadd B. Patton

PAGE 2

Acknowledgements The author would like to thank the follo wing individuals for their invaluable assistance with the completion of this study a nd thoughtful review of previous drafts of this manuscript: Toru Shimizu, Gabrielle Szafranski, Justine VandenBosche, Sophia Delgado, and Joshua Nadeau.

PAGE 3

Dedication This thesis is dedicated to my friends and family that have been behind me every step of the way. A special dedicati on goes to Melinda Patton who encouraged me to finish what I started.

PAGE 4

i Table of Contents List of Tables................................................................................................................. ....iii List of Figures................................................................................................................ ....iii Abstract....................................................................................................................... .........v Introduction................................................................................................................... .......1 Importance of Head and Face for Recognition and Selection.................................2 Importance of Local Components vs. Global Configuration for Recognition.........4 Methods Subjects..................................................................................................................10 Apparatus...............................................................................................................10 Stimuli....................................................................................................................11 Procedure...............................................................................................................15 Data Analysis.........................................................................................................16 Results ....................................................................................................................... .........17 Qualitative Movement Observation.......................................................................17 Position Preference Analysis.................................................................................19 Time Spent Near Altere d vs. Original Stimuli.......................................................20 Vocalizations Near Altered vs. Original Stimuli...................................................23 Circling Displays Near A ltered vs. Original Stimuli.............................................25 Discussion..................................................................................................................... .....27 Critical and Non-Critical Features.........................................................................27

PAGE 5

ii Contour..................................................................................................................29 Global Configuration.............................................................................................29 Methodological Considerations.............................................................................31 Concluding Remarks..............................................................................................31 References..................................................................................................................... .....33 Appendices..................................................................................................................... ....37 Appendix A Summary of Cumulative Results.......................................................38

PAGE 6

iii List of Tables Table 1. Preference for one side over the other side...................................................19 Table 2. Preference for altered stimuli over original stimuli......................................21

PAGE 7

iv List of Figures Figure 1. Schematic illustration of testing apparatus...................................................11 Figure 2. Sample photographic images of each stimulus group...................................14 Figure 3. Subject movement patterns...........................................................................18 Figure 4. Boxplot of position preference scores...........................................................20 Figure 5. Boxplot of time spent near altered stimuli....................................................21 Figure 6. Boxplot of vocalizati ons toward altered stimuli...........................................22 Figure 7. Boxplot of bowing disp lays toward altered stimuli......................................23

PAGE 8

v Altered Features of Female Pigeons ( Columba livia ) Elicit Preference Behavior in Males Tadd B. Patton ABSTRACT Previous research has sh own that male pigeons ( Columba livia ) respond with courtship displays to video playbacks of a female pigeon indicating that they ‘recognize’ the female as a potential mate. Courtship di splays significantly decline when the head region of the female is occluded (Shimizu, 1998) suggesting that feat ures located within the head are important for species recognition and mate selection. However, little is known about the exact visual features necessa ry to elicit displays. The current study examined the preference behavior of ma le pigeons when given a choice between photographic images of normal looking or alte red female pigeon faces. The altered-face categories included: 1) enlarged or removed facial features su ch as the eyes or beak; 2) the eyes and beak reconfigured within the head; and 3) removed contour (outline). The results showed that subjects responded pref erentially toward females with enlarged features (eyes or beak). However, s ubjects responded preferen tially toward normal females when the alternative stimuli were faces that were missing the eyes and/or beak. Preference for normal females was also obs erved when females with “incorrect” configuration were shown. Finally, subjects responded si gnificantly less to females lacking contour, even when the eyes and beak were visible. The ove rall findings suggest local facial components are im portant, although this effect di minishes if the contour of the female is not visible. These findings al so suggest that pigeons attend to both local

PAGE 9

vi components and global configuration when they detect conspecifics and identify potential mates.

PAGE 10

1 Introduction Bird vision is equal to, or better than, human visi on, in terms of acuity, color discrimination, and motion detection ability (Hodos, 1993; Frost & Sun, 1997; Shimizu, Patton, & Szafranski, In press). Birds use these superb abilities to navigate their environment, forage for food, and watch for pred ators from a safe distance. For instance, pigeons are able to identify members of th eir own species (consp ecific recognition) and choose potential mates (mate selection) based solely on visual cues (Shimizu, 1998). The goal of the proposed study was to extend our understanding of how birds accomplish complex visual discrimination tasks, such as the recognition of conspecifics and the selection of mates. The main question wa s: When a bird recognizes a stimulus as a conspecific or a potential mate, does it rely on specific local components of the stimulus (e.g., eyes, beak) or a global configuration of these components (e.g., the position of eyes in relation to beak), or both? In additi on, the importance of stimulus contour (outline) was investigated. The hypothesis was that pigeons depend, to some degree, on both local and global features – in the head and/or face and on the contour of the stimulus – to accomplish conspecific recognition and mate selection. The results of this study extended our understanding of the perceptual and cognitive processes birds use to accomplish such complex tasks. Furthermore, these findings can be used to draw comparisons between avian and mammalian visual processing.

PAGE 11

2 Importance of the Head and Face fo r Recognition and Selection. For many animals, characteristics of the head or face convey biologically relevant and salient information to individuals that allow them to detect members of the same species, identify potential mates and rivals, and recognize specific in dividuals within a group. This is perhaps most evident for humans. Studies have shown that we are capable of recalling the names of individuals when given photographs of only the head and face (Bahrick, Bahrick, & Whittlinger, 1975). This is somewhat remarkable considering that faces within a given species are relatively homogenous (i.e., all faces have eyes, nose, and mouth with the same spatial configuratio n). Therefore, face recognition requires a sophisticated level of visual discrimination. Studies have shown that information in the head and face is important to non human prim ates for recognition of individuals within a group. A study by Parr, Winslow, Hopkins, and de Waal (2000) demonstrated that the chimpanzee ( Pan troglodytes ) and rhesus monkey ( Macaca mulatta) could identify individuals of their own speci es based solely on information in the head and/or face region. In this study, subjects were requi red to discriminate between pairs of photographs that were either of the same individual conspecific or of different individuals. After reaching a level proficiency, features in some of the stimulus photographs were masked so that only the head and/or face region were visible. The results suggest that both species were capab le of discriminating individuals without information about the rest of the body (de Waal et al., 2000). Species recognition has been studied in many bird species as well. Previous studies have shown that bi rds, such as budgerigars ( Melopsittacus undulatus ), also use the head and/or face for conspecific and indi vidual recognition. The head and face region

PAGE 12

3 of the budgerigar is character ized by several noticeable markings on the cheeks and around the eyes that presumably convey inform ation such as age, sex, and willingness to mate (Brown & Dooling, 1993). A study by Tril lmich (1976) demonstrated the ability of budgerigars to discriminate cons pecific individuals ba sed solely on visual cues. In this study, budgerigars were trained to discriminate between two li ve conspecific individuals and between color slides (photographs) of them. When portions of the stimuli were masked so that the head was not visible, budge rigars performed less accurately than when the whole head was visible. This indicates that information in the head and/or face is important for individual discrimination for this species. In the studies discussed above, animals we re successfully trained to discriminate images of conspecific heads/faces using a ppetitive operant conditioning procedures. The assumption is that animals in natural setti ngs recognize conspecific individuals using the same discrimination ability. However, because discrimination can be accomplished without ‘recognition,’ it is not clear if th ese animals were actually ‘recognizing’ the objects in the photographs as conspecifics. An alternative to operant training procedures is to study naturally occurring responses to biol ogically relevant stim uli. For instance, courtship displays by male birds in the presen ce of females can be used as a measure of conspecific recognition and mate selection (Ryan & Lea, 1994). This ethological approach has been extens ively used in studies of Japanese quail ( Coturnix coturnix japonica ). In an experiment by Domjan and Nash (1988), male subjects placed in a test cag e could see taxidermically prepared females or males through a window. The experimenters measured the proximity of subjects to the window during the presentation of different stimuli. When the stimulus was the whole female model or

PAGE 13

4 only the head portion, male subjects showed an increased proximity to the window, suggesting that the subjects we re ‘interested’ in these stim uli. However, the proximity was significantly lower when the stimulus was a male model, inverted female model, or the body of the female model without the head. Shimizu (1998) studied the visual rec ognition capability of male pigeons by presenting video playbacks of various female pigeon stimuli. The natural response of a male pigeon toward a female conspecific is to display characteristic courtship behaviors such as bowing, tail dragging, and vocal izations (Levi, 1974; Stokes, 1979). In the study, male pigeons were sometimes presented with video playbacks of an obstructed view of a female pigeon to study the significance of di fferent physical components (i.e., head or body) for conspecific recognition. Subjects’ responses were significantly lower when only the body of a female pigeon was presen ted. However, when only the head region was presented, males reacted to the stimuli as rigorously as in the unobstructed view of the whole female pigeon. Both of these stud ies, using ethological approaches, provide additional evidence that visual components in the head and/or face are important for conspecific recognition and mate selection. Importance of Local Components vs. Global Configuration for Recognition What remains unknown is which specific information in the face birds use for recognition and mate selection. Do they us e local components, global configuration of the components, or both? There are exam ples, in which relatively simple local components can elicit distinct behaviors. In a classic ethological study investigating herring gull behavior, Tinbergen and Perdeck (1950) showed that herring gull chicks

PAGE 14

5 would peck at a red spot on their mother’s beak to elicit regurgita tion of food. It was subsequently determined that the red spot located on the mother’s beak was the key stimulus that acted as a releaser for this be havior. In another example, some birds (e.g., jays) are acutely sensitive to the brightly colored designs on the forewings of various moths that appear to be large eyes. The eyespot patterns on the moth serve as effective deterrents to small insectivorous passerines presumably because they mimic the eyes of large avian predators, such as owls (M cFarland, 1985). Eyespot mimicry has been extensively and systematically investigated in laboratory studies. In a study conducted by Blest (1957), birds were first trained to eat mealworms off of a specially designed box. Then, different stimulus shapes would be il luminated near the mealworm in the box as birds approached. The stimuli ranged from simp le lines to more complex eyelike shapes. The findings showed that the more eyelike th e stimuli, the more effective they were in eliciting avoidance behavior from the birds. The releasers used in these examples above are not directly involved in conspecific recognition or mate selecti on. In social commu nications between conspecifics, the relative importance of local components and global c onfiguration is still in debate. Research in human face perception has shown that humans use different strategies for recognition depending on conditi ons, such as familiarity and orientation (Schwaninger, Lobmaier, & Collishaw, 2002). Specifically, when the face is familiar or in its normal, upright position, we rely on a global configuration of diverse components for recognition. That is, successful face re cognition does not necessarily depend on any particular local components within the face (e.g., shape, size, or color of eyes), but on the Gestalt of all the parts (Farah, Tanaka, & Drain, 1995; Sergent, 1984). However, when

PAGE 15

6 the face is unfamiliar or inverted (upside-dow n), subjects attend to local components for recognition (Thompson, 1980). Similar results have been found in nonhuman primates suggesting they use similar processes for face recognition. For instance, one study showed that chimpanzees exhibit the face inversion effect for humans and chimpanzees, but not for capuchins ( Cebus apella ) or automobiles (Parr, Dove, & Hopki ns, 1998). However, there is some evidence that when a face is unfamiliar or partially masked, specific components are important for recognition. For instance, when chimpanzees and rhesus monkeys were forced to rely on local components of the face, they appear to pay attention to eyes rather than other components. In a previous ly discussed study by Parr et al., (2000), investigators masked portions of chimpan zee and rhesus monkey face photographs (e.g., eyes and/or mouth), and found that both spec ies had more difficulty when the eyes of these photographs were masked than when mout hs were masked. The results suggest that eyes are particularly salient local com ponents used for individual and conspecific recognition, and are consistent with other primate studies of individual recognition (McKelvie, 1976; Keating & Kea ting, 1993). As in primates, previous studies of avian visual behaviors suggest th at birds can also use both local components and global configuration in order to discriminate conspecific indi viduals. Using appetitive operant conditioning procedures, Brown and Dooling (1992; 1993) conducted a series of studies using budgerigars to study this issue. In one study (Brown & Dooling, 1992), they trained budgerigars to discriminate whether two color photographs of conspecific faces belonged to the same bird or not. It was determined that subjects successfully discriminated them based solely on local com ponents within the face. By analyzing the

PAGE 16

7 reaction time to these stimuli using multi-dime nsional scaling, the re searchers identified which specific local components acted as sali ent cues for accomplishing this task. The results showed that birds could discrimi nate individuals, using local components including face color and pattern of markings in the head region. The results also showed that specific features of their eyes, such as da rkness of the iris and si ze of pupil, provide important information for id entifying indi viduals. Global configuration is also important according to a subsequent series of experiments by Brown and Dooling (1993). In the study, budgerigars were presented with synthetic (computerized) conspecific faces that had been modeled after real budgerigars. When some local components of th e stimulus faces were scrambled so that all components remained in the stimuli, but were not in their co rrect location, reaction times for scrambled faces were significantly longer compared to non-scrambled (normal) stimuli. The results demonstrated the im portance of global configuration and correct orientation of specific components. Unlike budgerigars, pigeons do not have many noticeable facial components. Nevertheless, there is evidence that both local components and gl obal configuration are important for conspecific recognition and mate selection in pigeons. Watanabe and Ito (1991) investigated the ability of pigeons to discriminate photographs of the head regions of other pigeons, in which they arbitrarily assigned some as positive stimuli (S+) and others as punishment (S-). Using appe titive operant conditio ning procedures, once subjects were trained to discri minate between the S+ and Sbirds, some face components were altered and subjects were tested for th eir responses to the new stimuli. Pigeons showed a significant reduction in pecking at S+ birds when eyes or beak were completely

PAGE 17

8 removed. The results suggest that eyes a nd beak are important for discrimination of individuals. The importance of global configuration wa s also tested in this study. When subjects were shown a silhouette (no visibl e local component) of the S+ pigeon, they responded to this stimulus almost as strongl y as when all compone nts of the S+ were visible. In addition, when the stimuli were scrambled so that the local components were visible, but not in the corre ct locations (e.g., eyes were positioned below the beak; the beak and eyes were positioned away from the head), responses to these stimuli were greatly reduced. These results suggest that global configuration, such as the contour, or outline of the face and the correct orientations of these components, were critical as well. These findings indicate th at pigeons use both indivi dual components (especially eyes) and global configuration of these components for conspecific recognition. However, these studies used ope rant conditioning procedures to force birds to choose one or the other stimulus with food as reward. The relative importance of these components may change in natural settings. In order to understand the function and evolution of the visual discrimination ability in birds, it is essential to st udy how birds use diverse visual components in natural settings for consp ecific recognition and mate selection. The goal of the proposed study was to identify: 1) the specific local facial features important for species recognition and mate sele ction; 2) the importance of configuration of these local facial features; and 3) the importance of the contour (outline) of the stimulus. It was hypothesized that pigeons differentially rely on each these features attributes in order to accomplish cons pecific recognition and mate selection.

PAGE 18

9 In the present study, we used face pictures of female pigeons as stimuli to test which facial components of females were critical for triggering male courtship behavior. Male subjects were forced to choose betw een photographic images of normal female faces and those that had been digitally alte red along a specific dimension (e.g., faces with enlarged eyes or removed eyes). In pa rticular, we manipulated the size of local components (eyes and beak), as well as sp atial configuration of these components and contour (outline) of a female stimulus. We analyzed their preference for an altered stimulus over the original stimulus by meas uring stereotyped courts hip behaviors toward each stimulus, as described elsewhere (Shimi zu, 1998; Partan, Yelda, Price, & Shimizu, 2005). In the present study, it was necessary to use static images instead of video playback because the stimulus manipulations would require frame-byframe alteration, a procedure which was deemed inefficient ea rly in the stimulus production process. Although a previous study showed the importa nce of biological motion (Shimizu, 1998), we reduced the potential for “boredom” by su ccessively presenting static images from different view angles every five seconds dur ing a trial. The procedure was successful enough that birds reacted vigorously to these stationary stimuli.

PAGE 19

10 Methods Subjects Twelve (12) male white Carneaux pigeons ( Columba livia ) obtained from Palmetto Pigeon Plant, South Carolina, were used in this experiment. Bird weights ranged from 500 to 600 g, and they were sexually matured. The sex of each bird was confirmed before the experiment, by observing its response to a live female pigeon. All birds were housed individually in cages, sepa rated by opaque partitions to limit visual and physical interaction with ot her birds for the entire durat ion of the experiment. The vivarium was maintained on a constant 12 h li ght/12 h dark cycle. All the procedures were reviewed and approved by the Instituti onal Animal Care and Use Committee at the University of South Florida. Apparatus The present study used a testing appara tus large enough for a pigeon to move freely (90 x 90 x 60 cm) (see Figu re 1). The walls of the a pparatus were made of clear Plexiglas covered with water-resistant cont act paper (Tuff-Ware, Inc. almond-90.9983). A wire-mesh was placed over the top to keep su bjects from flying away during the trials. Two opposing sides of the testing apparatus had openings (30 x 35 cm) to accommodate a liquid crystal display (LCD) monitor (17" color flat pane l, Dell UltraSharp 1702FP) or a clear Plexiglas chamber (41 x 41 x 30 cm) us ed for presenting live-bird stimuli. An observation video camera (Panasonic wv-BP134) was positioned above the testing

PAGE 20

11 apparatus to record all m ovements of the subjects. A microphone (Audio Technica Omnidirectional ATR35s) was attached to th e wire-mesh covering the testing apparatus to record all vocalizations. MonitorScreens computer 1 computer 2 Figure 1 Schematic representation of the te sting apparatus (view from above). Dashed line represents the midline. Stimul us presentations were controlled via two personal computers located in an adjacent room. Stimuli Two female white Carneaux pigeons we re individually ph otographed using a digital camera (Nikon Coolpix 4800) while a male pigeon in a clear cage was presented to each of them. The presence of the male in creased the female’s activity, which, in turn, increased the likelihood of obt aining a variety of female stimulus viewpoints.

PAGE 21

12 Twelve color photographs of each female stimulus bird were selected. These comprised of four images from three different camera angles (0, 45, and 90 degrees). All photographs were cropped just below the fema le’s breastbone and applied to a black background. These images were then copied and horizontally flipped to make a complete set of 24 images from each stimulus bird. This procedure was followed for each of the eight stimulus conditions tested in this experiment. Thus, there were a total of 192 original stimulus photos (24 original photos x 8 stimulus conditions) from each stimulus bird. To make altered stimulus images, the 24 original stimuli were digitally manipulated with photo editing software (A dobe Photoshop v. 8.0). The following eight stimulus conditions were prepared for each original stimulus: 1. Enlarged Eyes (LE) original female with eye size enlarged by 180%. The increment scale was selected to contain the enlarged eyes inside the head; 2. Enlarged Beak (LB) original female with beak size enlarged by 180%. The increment scale was selected to be consistent with the eye enlargement; 3. Removed Eyes (NE) original female with eyes removed; 4. Removed Beak (NB) original female with beak removed; 5. Removed Eyes and Beak (NEB) original female with eyes and beak removed; 6. Shifted Feature Configuration (SEB) or iginal female with shifted eyes and beak position. Using photo editing software the eyes and beak were selected as one object and rotated approximately 180 de grees so that the beak was at eye level and eyes were at beak position.

PAGE 22

13 7. No Contour, with visible faci al features (NC+EB) original female with Gaussian blur applied to the entire im age except for the eyes and beak; and 8. No Contour (NC) original female with Gaussian blur applied to the entire image including the eyes and beak.

PAGE 23

14 OriginalLELB NENBNEB SEBNC+EBNC Figure 2. Sample photographs from each stimul us type. See above for abbreviations. Figure 2 shows examples of an original stimulus female face and its eight altered stimulus faces. During each trial, one of the eight altered stimulus conditions was presented with their matching original female face images. In tota l, each subject was exposed to 384 (24 originals x 8 stimulus condi tions x 2 female exemplars) stimulus pairs in the experiment. Each trial lasted tw o minutes, during which 24 stimulus pairs were consecutively presented for five seconds e ach. The slide show presentations were rendered using visual-effects software (Adobe After Effects 5.5), and played back on Windows Media Player (v. 9).

PAGE 24

15 Procedure Birds used in the experiment were acclimat ed to the testing apparatus for at least 20 minutes per day for five consecutive days or until they walked freely within the apparatus. Following acclimation, testing tr ials began. During the testing trials, to eliminate the possibility of position/monito r preference, each stimulus condition was presented twice by alternating presentation positions (i.e., left and right monitor). Therefore, there were 32 trials in total (8 stimulus conditions x 2 exemplars x 2 positions). The presentation order of the 32 trials was counterbalanced between subjects. Only one trial per day was conducted every other day. In each trial, a subject was placed in the testing apparatus and allowed to acclimate for at least five minutes. After the acclimation was accomplished, and before the testing trial began, the subj ect was exposed to alternatin g “preview” on both monitors for 30 seconds each. This was a preview of the upcoming stimulus (i.e., original or altered) to be presented on the monitor screen. Once a trial started, stimuli were presented only on the monitor that was on th e side where a subject was located. The other monitor was dark and the corresponding stimulus category was not visible unless a subject crossed the midline of th e testing apparatus. The cente r of the head of the subject was used as a reference point for the bird’s location. A two-minute trial was followed by two minutes of a post-trial phase, during wh ich no stimulus was presented on either monitor. Subjects were returned to thei r home cage following a post-trial phase. Subjects were exposed to a live female wh ite Carneaux pigeon in a Plexiglas stimulus chamber periodically (once a week) throughout th e experiment. This procedure was used

PAGE 25

16 to reduce stimulus habituation and served to rekindle the courtship displays for the remaining conditions. Data Analysis Video and audio-recordings of each tria l were analyzed for two well-documented courtship behaviors (Shimizu, 1998; Partan et al., 2005). One measurement was circling display frequency, which was the number of time s a subject turned in full circles with its neck feathers ruffled and its head lowered. The other measurement was vocalization frequency, which was the number of times a subject made a coo sound. These courtship displays were chosen as dependent measur es because they are robust stereotypical behaviors easily observed dur ing courtship (Levi, 1974; St okes, 1979). The amount of time subjects spent on each half of the te sting apparatus was also measured. Preference scores, (‘altered’) / (‘altered ’ + ‘original’), were obtained for each stimulus condition. In preference calculations ‘altered’ and ‘origina l’ were the averaged measurements (i.e., time, frequency) expressed near the altered or original stimuli. The preference scores became 1.0 when a subjec t showed behavior exclusively near the altered stimuli, whereas the scores became zero when a subject preferred the original completely. The overall effect of the stimulus conditions on preference score was evaluated by one way repeated measure anal ysis of variance (ANOVA). For each of the stimulus conditions, the preference score was an alyzed using single sample t-tests with a hypothetical mean of 0.5. The binomial exact probability was also calculated to determine whether significantly more subjects prefer the altered stimuli over the original stimuli.

PAGE 26

17 Results Eleven out of the 12 subjects showed so me courtship behaviors (i.e., ‘coo’s and circling) when presented with the described stimuli on the monitor screens. Data from the subject that exhibited no courtship behavi or was excluded from further analysis. Qualitative Movement Observation Figure 3 shows the path of a subject as observed during three separate trials. Figure 3A shows a pattern in which the subj ect did not exhibit much movement. The pattern in Figure 3B shows a subject that paced back and forth with no indication of stimulus preference. The pattern in Figure 3C shows that the subject spent time on both sides of the testing apparatus, yet showed a preference for the stimulus displayed on the left side by circling more often on the left si de. During tests, most subjects exhibited clear courtship displays as exemplified in Fi gure 3C rather than st aying still (Fig. 3A) or pacing back in forth between the two sides (Fig. 3B).

PAGE 27

18 A C B Figure 3. Three distinct patte rns of subject’s movement obs erved during testing: A) low activity level and no courtship; B) high activity level with no stimulus preference; and C) high activity level with clear stimulus preference.

PAGE 28

19 Position Preference Analysis The mean preference score for time spent on one half of the apparatus versus the other half was 0.48, with sc ores ranging between 0.44 and 0.50. A one way repeated measure ANOVA showed that there was no si gnificant overall difference among stimulus conditions [ F (7,70) = 0.56, p = 0.79]. Table 1 shows the number of animals that preferred the left side to the right side of the testing appa ratus (i.e., preference score > 0.5). There were no significant differences in the number of subject s preferring one side to the other for any stimulus conditions (i.e., binomial p > 0.05). These results confirmed that the subjects did not show any position/ monitor preference during the experiment. Table 1 Number of male subjects that preferred the one side over the other side (N=11) Stimulus Type Measure LE LB NE NB NEB SEB NC+EBNC Position 5 5 6 3 5 5 4 5

PAGE 29

20 1.0 0.5 0 LELBNENBNEBSEB NC+ EB NC Left Right Figure 4. Boxplot showing the position preferen ce scores between the original and altered stimuli based on time dura tion spent on each side of th e testing apparatus. Boxes represent the interquartile range between the 25th and 75th percentiles. Horizontal lines represent medians and circles represent means. Vertical lines represent the entire range of scores. See the Method section for abbreviations. Time Spent Near Altered vs. Original Stimuli Figure 5A is a boxplot showi ng the preference scores base d on time spent near altered stimuli relative to time spent near original stimuli. The mean preference score was 0.47, with scores ranging between 0.40 and 0.54. A one-way ANOVA showed that different stimulus conditions have differential effects [ F (7,70) = 2.58, p < 0.05]. Birds spent more time near altered faces with large beaks (LB) than near original unaltered faces [t(11) = 1.95, p < 0.05]. In contrast, birds spent less ti me near the faces without beaks (NB) [t(11) = 1.95, p < 0.05] and those without contour, but with eyes and beak intact (NC+EB) [t(11) = 2.15, p < 0.05]. No other comparisons of tim e spent near altered faces compared to normal faces were significant. However, the number of subjects that spent more time

PAGE 30

21 near altered faces compared to normal faces was also evaluated (see Table 2) Binomial tests showed that significantly more birds pref erred the original intact faces than altered faces without eyes and beak (NEB, p< 0.05), or faces without contour (NC, p< 0.01). Table 2 Number of male subjects that preferred the a ltered stimuli to the original stimuli (N=11) Stimulus Type Measures LE LB NE NB NEB SEB NC+EBNC Time Spent 5 7 5 3 2* 7 3 1** Vocals 10** 10** 5 3 0** 3 0** 1** Bowing 4 9* 3 1** 1** 3 0** 0** 1.0 0.5 0 LELBNENBNEBSEB NC+ EB NC Altered Original Figure 5A. Boxplot showing the preference scores of time spent near th e altered stimulus for each stimulus pair. Data points represent the median durations. Boxes represent the interquartile range between the 25th and 75th percentiles. Horizontal lines represent

PAGE 31

22 medians and circles represent means. Vertical lines represent the enti re range of scores. See the Method secti on for abbreviations.

PAGE 32

23 Vocalizations Near Altered vs. Original Stimuli Figure 5B shows the preference scores based on the number of coo vocalizations subjects produced near altered faces, rela tive to the number made near original faces. The scores (mean = 0.42, with scores ranging between 0.26 a nd 0.57) showed rather a large variance among conditions. A one-way ANOVA confirmed that subjects reacted differentially depending on stimulus conditions [ F (7,70) = 5.67, p < 0.01]. Single sample t-tests showed that birds vo calized more near LB faces than near faces [t(11) = 4.28, p < 0.01]. In contrast, the birds vo calized more near original faces than near NB faces [t(11) = 2.05, p < 0.05], NEB face [t(11) = 5.11, p < 0.01], SEB faces [t(11) = 1.91, p < 0.05], NC+EB faces [t(11) = 3.75, p < 0.01] and NC faces [t(11) = 5.51, p < 0.01]. As seen in Table 2, significantly more birds vocalized near LE faces (binomial p <0.01) and LB faces (binomial p <0.01) than near orig inal faces. In contrast, subjects produced fewer vocalizations near NEB faces (bi nomial p <0.01), or NC+EB and NC faces (binomial p <0.01). 0.5 0 LELBNENBNEBSEB NC+ EB NC A lte r e d Original 1.0

PAGE 33

24 Figure 5B. Boxplot showing the preference scores of vocalizations near the altered stimulus for each stimulus pair. Data points represent the median durations. Boxes represent the interquartile range between the 25th and 75th percentiles. Horizontal lines represent medians and circles represent means. Vertical lines represent the entire range of scores. See the Method s ection for abbreviations.

PAGE 34

25 Circling Display Near Altered vs. Original Stimuli Figure 5C shows the preference scores ba sed on the number of circling displays near the altered stimuli, re lative to the original stimu li. The mean score was 0.40, ranging between 0.17 and 0.64, suggesting even a wider variance than the vocalization measure. A one-way ANOVA showed that diffe rent stimulus conditions had differential effects [ F (7,70) = 4.36, p < 0.01]. Birds tended to circ le more often near original faces than near NB faces [t(11) = 1.82, p < 0.05], NEB faces [t(11) = 3.93, p < 0.01], or NC+EB faces [t(11) = 4.07, p < 0.01] and NC faces [t(11) = 5.86, p < 0.01]. As seen in Table 2, significantly more birds circled near LB faces (binomial p <0.05) than original facef. In contrast, fewer circled near NB and NEB faces (binomial p <0.01), or NC+EB faces and NC faces (binomial p <0.01). 0.5 0 LELBNENBNEBSEB NC+ EB NC Altered Original 1.0 Figure 5C Boxplot showing the preference scores of circling displays near the altered stimulus for each stimulus pair. Data points represent the median durations. Boxes represent the interquartile range between the 25th and 75th percentiles. Horizontal lines

PAGE 35

26 represent medians and circles represent means. Vertical lines represent the entire range of scores. See the Method section for abbreviations.

PAGE 36

27 Discussion The results showed that male subjects responded differentially to the various stimulus types. Overall, the present study clarified the importance of the presence of local facial features in visual discrimina tion and species recognition. Furthermore, the present study revealed that the global configuration of facial features was not as crucial as it is in other species, such as humans, but th e contour was. The following discussion will focus on several issues related to the significance of 1) local facial features, 2) global configuration, and 3) contour. Critical and Non-Critical Features In general, more subject birds preferre d female pictures with unnaturally large local components to the origin al pictures with normal size features. The results suggest that these local components provide important information that facilitates conspecific recognition and mate selection in the natura l setting. The finding th at unnaturally large features can initiate strong preference behavior is remini scent of the well-established notion of supernormal stimuli introduced by ear ly ethologists. Their studies showed that, when certain dimensions of a releasing stimul us (e.g., size or color) are exaggerated, such supernormal stimuli initiate more vigorous be havior than the normal stimuli (Baerends & Drent, 1982; Tinbergen & Perdeck, 1951; Ba solo, 1990). In the present study, the enlarged local components within the face might have functioned as supernormal stimuli.

PAGE 37

28 The importance of local components was c onfirmed by the effect of removal of them. More subjects preferred the original stimuli to those stimuli in which facial features were removed (NEB and NB, but not NE stimuli). This finding is consistent with other research on the subject. In a study on visual discrimination of individual pigeons, Watanabe and Ito (1991) rained pigeons to peck at photographs of certain pigeons (S+) and not at other pigeons (S-). Then, the eyes and beak of the S+ pigeons were removed, which resulted in significan tly decreased pecking accuracy, suggesting that these local facial components served as critical disc riminative stimuli in their task. Similarly, the present study s uggests that birds use these local facial components as critical signals for elicit ing courtship display. While local features are clearly important, it is difficult to determine which of the two local face components (eyes or beak) is more crucial for species recognition based on this study. There are two reasons why this co mparison is impossible. First, although the eyes and beak differ naturally along many physi cal dimensions such as shape, size, and color, only their physical size was altered in the present study. Litt le is known about the salience of these various physical attributes for eliciting courtship. It is most likely that there were differential effects of such alterati on on the salience of eyes and beak, and thus the results of the size alteration cannot be compared directly. Second, as discussed below, the present study showed that the cont our itself played an important role in eliciting courtship display and the alterati on of eyes and beak affected the contour differently. When the eyes were enlarged or removed, the contour as a whole was not affected. Alterations of beak size changed the overall cont our of the stimuli, making it difficult to isolate the importanc e of the beak as a specific lo cal component. That is, were

PAGE 38

29 birds influenced caused by changes to the beak itself or to the alte red contour? Thus, the relative influence manipulations (enlargeme nt and/or removal) have on the eyes compared to the beak is not po ssible based on the current data. Contour Most birds showed a strong preference for the original stimuli to those stimuli in which the contour was removed (NC condition) This is, perhaps, unsurprising, since there were no ‘bird-like’ global, or local, facial components visi ble in this altered stimulus condition. However, the preference fo r original faces persisted even when the eyes and beak were visible (NC + EB condition) This finding is in agreement with the results presented by Watanabe and Ito (1991) in which pigeons showed a significant decrease in pecks toward S+ photographs when only the eyes and beak were visible (i.e., contour had been removed), suggesting that th e eyes and beak alone were not sufficient for visual discrimination of individual pigeons Together, these results suggest that the ‘bird-like’ contour of the stimulus is esse ntial for species recogn ition, and that, without the contour, specific facial features such as the eyes and beak are ineffective (or not sufficient to elicit courtship). Global Configuration Birds produced more coo vocalizations near the original face than near the SEB face. Although this was not statistically signi ficant, this pattern of preference for the original face was supported by the fact that eight of the eleven s ubjects produced more coo vocalizations and circles near the original fa ce than near the SEB face (Table 2). It is

PAGE 39

30 possible that the preference fo r the original faces over SEB faces might have been even stronger if we used more extreme violati ons of global confi guration (i.e., local components positioned outside of the head region) in the SEB condition. Thus, although there were numerous possibili ties for configural alterati on, we used only one stimulus condition (SEB). With this caveat in mind, the present findi ng of the effect of configural alteration is in contrast to other obs ervations about visual discri mination behavior of pigeons (Cavoto & Cook, 2001; Cerella, 1980; Delius & Hollard, 1995; Hollard & Delius, 1982). For example, experiments conducted by Ho llard and Delius (1982, 1995) showed that, unlike humans, pigeons could recognize visual stimuli presented at various orientations without significant response decrement. Cerella (1980) showed similar results, suggesting that configuration is not n ecessarily an important cue for visual discrimination, as pigeons responded to local el ements regardless of sp atial configuration. These previous studies suggest that, in visual discrimination tasks which used artificial geometrical stimuli, pigeons tend to attend to local features, rather than global configuration or orientation of local features. However, the present results are consistent with previous studies using more naturalistic, biologically relevant s timuli, such as color photographs of the head of pigeons used by Watanabe and Ito (1991), who showed that pigeons responded to global configuration, as we ll as local features, in order to perform the visual discrimination task. Similarly, Johnson and Horn (1988) reported that newly hatched chicks, which are known to show a natural predisposition to approach a conspecific adult (hen) rather than less natura listic objects preferre d a stuffed hen, when features of their head and neck we re in their correct arrangement.

PAGE 40

31 Methodological Considerations The present study used simultaneous pres entation of two visual stimuli and analyzed the preference behavior, instead of se quential presentation of different stimuli. The simultaneous presentation procedure succes sfully revealed diffe rential responses to various stimulus types. This procedure of presenting two stimuli at the same time was in contrast to previous studies conducted in th e same laboratory (Shi mizu, 1998; Partan, et al., 2005), which used sequential presentation procedures. Subjects in these previous experiments were exposed to one stimulus type during a given trial. Although this is an effective procedure for systematically eval uating the importance of specific features, there were always concerns about potential ce iling effects. When only one stimulus is presented during a trial, the subject may res pond constantly and vigorously at asymptote. In such a case, it is difficult to determine pr eference for particular stimuli. By using simultaneous procedure, this potential problem was avoided in the present study. In this sense, the present procedure was effective a nd useful for future studies with similar interests. Concluding Remarks In the present study, pigeons showed a pr edisposition to attend local components, but also global configurati on. In contrast, in many visu al discrimination tasks using operant conditioning techniques, pigeons tended to pay attention to local features rather than global features of complex stimuli. Pigeons appear to have the predisposition to attend to local elements and not to group these features in order to perceive the global

PAGE 41

32 aspect of visual stimuli in these tasks. Th ese findings could be related to the fact that these tasks often required pigeons to discrimina te artificial, unnaturalistic stimuli, which may not be biologically salient enough to f acilitate their ability to perceive global features. Pigeons and chicks appear to atte nd to global configurati on when biologicallyrelevant visual stimuli, such as faces of conspecific animals, were used in a discrimination task (Watanabe & Ito, 1991) and also in a preference test (Johnson & Horn, 1988; the present study). Although pigeons appear to show a natural predisposition toward local elements over gl obal elements in many discrimination tasks, this does not mean that pigeons have a genera l deficiency in perceiving global features. Rather, such a “deficiency” may be contingent on the type of visu al stimuli presented (i.e., artificial, biologically le ss relevant, geometrical figures vs. naturalistic). Future research with more systematic analysis on this issue is clearly warranted.

PAGE 42

33 References Baerends, G.P. & Drent, R.H. (1982). The herr ing gull and its egg: II. The responsiveness to egg-features. Beha viour, 82(1-4), 416. Bahrick, H.P., Bahrick, P.O., & Wittlinger, R.P. (1975). Fifty years of memory for names and faces: A cross-sectional approach. Journal of Experimental Psychology: General 104, 54-75. Basolo, A.L. (1990). The female preference for male sword length in the green swordtail, Xiphophorus helleri (Pis ces: Poeciliidae). Animal Behaviour 40, 332-338. Blest, A.D. (1957). The function of ey espot patterns in the Lepidoptera. Behaviour 11, 209-256. Brown, S.D. & Dooling, R.J. (1992). Per ception of conspecific faces by budgerigars ( Melopsittacus undulatus ): I. natural faces. Journal of Comparative Psychology 106, 203-216. Brown, S.D. & Dooling, R.J. (1993). Per ception of conspecific faces by budgerigars ( Melopsittacus undulatus ): II. Synthetic models. Journal of Comparative Psychology, 107, 48-60. Cavato, K. & Cook, R.G. (2001). Cognitive precedence for local information in hierarchical stimulus processing by pigeons. Journal of Experimental Psychology: Animal Behavior Processes 27, 3-16. Cerella, J. (1980). The pigeons’ analysis of pictures. Pattern Recognition 9, 1-6.

PAGE 43

34 Cook, R.G. (2000). The comparative psyc hology of avian visual cognition. Current Directions in Psychological Science 1, 83-89. Delius, J.D. & Hollard, V.D. (1995). Orienta tion invariant pattern recognition by pigeons (Columba livia) and humans (Homo sapiens). Journal of Comparative Psychology 109(3), 278-290. Domjan, M., & Nash, S. (1988). Stimulus contro l of social behavior in male Japanese quail, Coturnix coturnix japonica, Animal Behaviour, 36, 1006-1015. Farah, M.J., Tanaka, J.W. & Drain, M.H. ( 1995). What causes the face inversion effect? Journal of Experimental Psychology: Human Perception and Performance 21, 628-634. Frost, B.J. & Sun, H. (1997). Visual moti on processing for figure/ground segregation, collision avoidance, and optic flow analys is in the pigeon. In M.V. Srinivasan, & S. Venkatesh (Eds.), From living eyes to seeing machines (pp. 80-103). New York: Oxford University Press. Hodos, W. (1993). The visual ca pabilities of birds. In H. P. Zeigler & H.-J. Bischof (Eds.), Vision, brain, and behavior in birds (pp. 77-98). Cambridge, MA: MIT Press. Hollard, V.D. & Delius, J.D. (1982). Rotation invariance in visual pattern recognition by pigeons and humans. Science 218, 4574. Johnson, M.H. & Horn, G. (1988). Developmen t of filial preferen ces in dark-reared chicks. Animal Behaviour, 36, 675-683.

PAGE 44

35 Keating, C.F. & Keating, E.G. (1993). Monkeys and mug shots: Cues used by rhesus monkeys ( Macaca mulatta ) to recognize a human face. Journal of Comparative Psychology 107, 131-139. Levi, W.M. (1974). The pigeon Sumter, SC: Levi. McFarland, D. (1985). Animal behavior. Reading, MA: The Benjamin/Cummings. McKelvie, S.J. (1976). The role of eyes and mouth in the memory of a face. American Journal of Psychology 89, 311-323. Parr, L.A., Dove, T., & Hopkins, W.D. (1998). Why faces may be special: Evidence of the inversion effect in chimpanzees. Journal of Cognitive Neuroscience 10, 615622. Parr, L.A., Winslow, J.T., Hopkins, W.D., & de Waal, F.B.M. (2000) Recognizing facial cues: Individual discrimination by chimpanzees ( Pan troglodytes ) and rhesus monkeys ( Macaca mulatta ). Journal of Comparative Psychology 114, 47-60. Partan, S., Yelda, S., Price, V, & Shimiz u, T. (2005). Female pigeons, Columba livia, respond to multisensory audio/video pla ybacks of male courtship behavior. Animal Behaviour 70(4), 957-966. Ryan, C.M.E. & Lea, S.E.G. (1994). Images of conspecifics as categories to be discriminated by pigeons and chickens: Slid es, video tapes, stuffed birds and live birds. Behavioural Processes 33, 155-176. Schwaninger, A., Lobmaier, J.S., & Collis haw, S.M. (2002). Role of featural and configural information in familiar and unfamiliar face recognition. Lecture notes in computer science 2525, 643-650.

PAGE 45

36 Sergent, J. (1984). An investigation into co mponent and configural processes underlying face perception. British Journal of Psychology 75, 221-242. Shimizu, T. (1998). Conspecifi c recognition in pigeons ( Columba livia ) using dynamic video images. Behaviour 133, 43-53. Shimizu T, Patton T B, Szafranski G (In Press) Evolution of the visual system in birds. In M. D. Binder, N. Hirokawa, U. Windhorst & M. C. Hirsch (Eds.), Encyclopedic Reference of Neuroscience, He idelberg, Germany: Springer Stokes, D.W. (1979). A guide to bird behavior Boston, MA: Little, Brown. Thompson, P. (1980). Margaret Thatcher: A new illusion. Perception 9, 483-484. Tinbergen, N., & Perdeck, A.C. (1950). On th e response situation releasing the begging response in the newly hatched Herring Gull chick ( larus argentatus agrentatus Pont. ). Behaviour 3, 1-39. Trillmich, F. (1976). Learning experiment s on individual recognition in budgerigars ( Melopsittacus undulates ). Zeitschrift fr Tierpsychologie 41, 372-395. Watanabe, S. & Ito, Y. (1991). Discri mination of individuals in pigeons. Bird Behaviour 9, 20-29. Watanabe, S., Lea, S.E.G., & Dittrich, W.H. (1993). What can we learn from experiments on pigeon concept discrimination? In H.P. Zeigler & H.-J. Bischof (Eds.), Vision, brain, and behavior in birds (pp. 351-376). Cambridge MA: MIT Press.

PAGE 46

37 Appendices

PAGE 47

38 Appendix A Summary of Cumulative Results

PAGE 48

39 Appendix A (Continued)

PAGE 49

40 Appendix A (Continued)


xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 001911072
003 fts
005 20071003110328.0
006 m||||e|||d||||||||
007 cr mnu|||uuuuu
008 071003s2006 flu sbm 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0001656
040
FHM
c FHM
035
(OCoLC)173642317
049
FHMM
090
BF121 (ONLINE)
1 100
Patton, Tadd B.
0 245
Altered features of female pigeons (Columba livia) elicit preference behavior in male pigeons
h [electronic resource] /
by Tadd B. Patton.
260
[Tampa, Fla] :
b University of South Florida,
2006.
3 520
ABSTRACT: Previous research has shown that male pigeons (Columba livia) respond with courtship displays to video playbacks of a female pigeon indicating that they 'recognize' the female as a potential mate. Courtship displays significantly decline when the head region of the female is occluded (Shimizu, 1998) suggesting that features located within the head are important for species recognition and mate selection. However, little is known about the exact visual features necessary to elicit displays. The current study examined the preference behavior of male pigeons when given a choice between photographic images of normal looking or altered female pigeon faces. The altered-face categories included: 1) enlarged or removed facial features such as the eyes or beak; 2) the eyes and beak reconfigured within the head; and 3) removed contour (outline). The results showed that subjects responded preferentially toward females with enlarged features (eyes or beak). However, subjects responded preferentially toward normal females when the alternative stimuli were faces that were missing the eyes and/or beak. Preference for normal females was also observed when females with "incorrect" configuration were shown. Finally, subjects responded significantly less to females lacking contour, even when the eyes and beak were visible. The overall findings suggest local facial components are important, although this effect diminishes if the contour of the female is not visible. These findings also suggest that pigeons attend to both local components and global configuration when they detect conspecifics and identify potential mates.
502
Thesis (M.A.)--University of South Florida, 2006.
504
Includes bibliographical references.
516
Text (Electronic thesis) in PDF format.
538
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
500
Title from PDF of title page.
Document formatted into pages; contains 40 pages.
590
Adviser: Toru Shimizu, Ph.D.
653
Bird
Face Recognition
Conspecific Recognition
Vision
Mate Selection
Face recognition.
Conspecific recognition.
Vision.
Mate selection.
690
Dissertations, Academic
z USF
x Psychology
Masters.
773
t USF Electronic Theses and Dissertations.
4 856
u http://digital.lib.usf.edu/?e14.1656