RESEARCH STATEMENT

The aim of our current research is to explore the effects of social organization on avian learning and to find developmental principles to guide research on vocal communication in other taxa, including human infants. Of special interest is how animals, when living in an information-rich environment, selectively attend, interact, and learn from other members of their species. Given that they do not or cannot learn equally from all individuals, what rules guide access and attention to social information? Interest in the origins and nature of social organization stems from an interest in expanding the study of learning and development toward ecological concerns. We propose that much species- typical knowledge is delivered via social networks, social groups in which older organisms or peers facilitate or constrain exposure to information. The idea of networks has been bypassed in most laboratory research on learning because generally the animals are simply exposed to stimulation delivered with minimal species-typical context. To study socially guided learning, higher-level patterns of social organization need to be investigated. These are patterns that result from flock dynamics, yielding different phenotypic effects for mature and immature males and females. To explore such effects, birds are studied throughout the year in large aviaries in which the composition of groups is periodically altered and the effects measured. Social and vocal behavior is recorded, mate choice tests employed, and eggs collected and incubated to measure production, fertility, and viability. In addition, blood is taken for DNA analyses. These procedures afford the chance to explore genotype-phenotype effects on communicative competence as a function of opportunities for social learning. The work is focused on questions regarding causation and function. First, are there simple rules to guide social assortment, and how does social organization affect learning? Second, how do specific phenotypic traits, beyond age or sex, affect formation of groups and how do such groups affect developmental trajectories of individuals? The proposed work has far-reaching implications for understanding complex communication systems, including the humans. In recently completed work with human parents and their infants, hypotheses were tested about prelinguistic vocal plasticity based on avian models of social guidance. It was found that infant babbling could be socially shaped by parental responses. Parents did not vocally imitate infants' sounds, thus ruling out vocal matching as a mechanism. These are the first data to show such a relationship and they challenge long standing but untested ideas about the innate nature of early speech development. Any work that loosens the holds of biological determinism on behavior is a benefit to society because of implications for understanding plasticity in relation to education and other societal issues. Thus, showing animals, like cowbirds, presumed to develop by closed genetic programs, use open developmental programs is beneficial to educators on matters of theories of learning and plasticity. The work done with songbirds has influenced issues in human development in the past. The work revealing the use of visual displays by female songbirds is now cited as an animal exemplar for gesture and visual learning by children who are deaf or incapable of early vocalizations. The new findings on plasticity in human babbling topple a long-standing belief that it is a non-communicative by-product of motor development. Such comparative work required guidance and technology from the songbird research. Thus, efforts to study and understand how young infants learn language, the critical human skill for learning and education, will benefit from our comparative work.

OVERVIEW OF RECENT FINDINGS

Finding 1. Cowbirds show rapid social organization with assortment patterns predictive of courtship competence. To learn about flock dynamics, from fall to late spring we studied a captive group of 74 cowbirds, adult and juvenile males and females, that were allowed to associate in a large complex of connected aviaries, the ADLIB aviary. We found that highly organized social assortment emerged rapidly: individuals formed groups with others based on age and sex. Subgroups also emerged with mixing of classes. For example, juvenile males spent 70% of their time with other juvenile males, 30% with adult females or males. The mixing had long-term consequences. Juvenile males who associated most with adult males had higher courtship success. Juvenile males who associated most with females sang less as the year progressed, suggesting that their song might have become stereotyped sooner, as we had found in previous work. Anne Smith with R.Fleischer, carried out genetic analyses to look at patterns of kinship within the ADLIB aviary flock. Kinship did not account for social affiliation. Finding 2. Emergence of social organization in flocks begins early in life and continues into adulthood. 2.1. Young cowbirds respond to adults with different patterns of social assortment. Here we asked how early in development we could see flock organization. Using four identical replicates, we housed very young juvenile birds, caught in the first few days of independence, in aviaries that either contained or did not contain adult conspecifics. Juveniles housed with adults interacted with other juveniles more often and sang significantly less than did juveniles housed without adults. Effects of adult presence or absence on social organization and singing patterns emerged rapidly and could be reversed quickly, using a within-subject ABA design. 2.2. In a flock, adult males influence juvenile males' development and courtship competence. We followed up on Finding 1 by looking more closely at mixed groups, specifically, adult influence on juvenile males. We configured two aviaries with groups of juvenile males, both housed with juvenile and adult females but one group of juveniles had adult males present (+ADM) and the other did not (-ADM). Juvenile males in the two conditions developed reliably different courtship and communication skills. Prior to the breeding season, the +ADM juveniles generally chose other juvenile males as near neighbors and sang to adult males and to females. In the breeding season, these males were aggressive with other males, and courted and copulated with females in patterns typical for cowbirds in the Midwest. In contrast, -ADM juvenile males displayed anomalous social patterns. They rarely associated with females, and sang in long bouts of undirected soliloquies. In the breeding season, they showed no aggression toward other males, courted with minimal pursuit, and even after copulation, they did not guard females, allowing multiple males to copulate with “their” females. The +ADM and -ADM juveniles also developed structurally different songs at different rates. The -ADM juveniles advanced sooner to stereotyped song and developed songs more effective at eliciting females' copulatory responses than the songs of the +ADM males. We followed the -ADM condition for another year and found the - ADM phenotypes persisted even when we introduced new females. Finding 3. Social effects in a flock are bi-directional: juvenile females and males influence learning, assortment, and courtship success. 3.1. Addition of juvenile females changes group organization and behavior. The surprising level of incompetence of the -ADM males led us to look at four replicates of that condition in 2000-2001. We also capitalized on observations from the previous year suggesting juvenile females were more interested in approaching singing males than were adult females. We found that variation in juvenile female behavior differentiated the four replicates. The one aviary with juvenile females who approached males did not show the anomalous patterns of singing and assorting documented earlier and in the other three aviaries. To understand the role of approach by females in male development, after establishing the above patterns, we exchanged small numbers of juvenile males (n=4) or juvenile females showing high levels of approach (n=4) into the four -ADM aviaries. Adding new males had no effect on overall behavior patterns: they matched the behavior of male residents. Changing juvenile females, however, did alter the behavior patterns. When new females were added, the juvenile males immediately increased countersinging (CS) and when the females were removed, CS decreased. Female stimulation in the form of proximity and approach to juvenile males affected song potency, courtship competence, and copulation success. 3.2 Presence of juvenile males affects adult male courtship success. We also looked at developmental plasticity in adult males. We compared the winter and breeding behavior of adult males housed with females but with or without juvenile males. Adults in the all-adult male condition sang more to females over the year, who, in turn, spent more time close to adult males. In the breeding season, we allowed the two groups to compete against each other with new females. The adult males housed with juvenile males out competed adult males housed without juveniles. Thus, the sub-groups of adult and juvenile males seen in the ADLIB aviary had developmental consequences for both parties. The data also show that the development of competitive ability is malleable into adulthood. Finding 4. Female assortment is affected by presence or absence of males and female egg production is affected by male-male countersinging. The ADLIB aviary indicated that females also vary in their near neighbor subgroups. To learn more about female social behavior, we housed groups of females in four conditions: 1) only adult males; 2) only juvenile males; 3) juvenile and adult males; and 4) no males. (The second and third conditions are the same individuals reported in finding 3 as -ADM and +ADM males). Females’ assortment differed across the four conditions. Females assorted by age class in the conditions where males were present, but not in the condition when males were not present. Thus, assortment appears to be a measure that reflects female perception of males. Egg production turned out to be another sensitive measure of female responsivity to male characteristics. Egg production correlated positively with CS across the male conditions. The interest in CS and egg measures led us to look at six new aviaries the following year in which we compared rates of CS and measured egg production and egg viability (# of fertile eggs through prenatal day 10). We found that the amount of CS correlated significantly with both egg measures. Undirected singing showed the opposite relationship: it was negatively correlated with both egg measures. Finding 5. Females use multiple visual displays to shape neural growth and behavior in males. An analysis of 103,320 behaviors in social interactions was completed for 12 juvenile males, each individually housed with two adult female cowbirds. The triads were videotaped in early spring. We found that non-vocal responses by females to male song facilitated vocal development. The non-singing females displayed two specific responses to song, gaping and wing stroking, both so rapid they could only be measured in frame by frame video analysis. Gaping, i.e., extension of the neck and opening of the beak, was documented for the first time. To understand the function of these responses, we correlated their relative frequency with measures of vocal ontogeny and neural growth. We found that these forms of feedback predicted song development and attentive song use. Visual responses from females also predicted the volume of a visual thalamic nucleus, Rt. Thus, song ontogeny can be affected by visual communication, a non-imitative means for males to acquire song knowledge. Finding 6. Female preferences for song are malleable. Freeberg's work established that post-fledging experience biased female mating preferences in cowbirds. To connect mating preferences to song preferences, we housed new wild caught adult and juvenile Indiana females in groups either with or without males throughout the winter and spring. We carried out playbacks in May, a week after bringing them into playback chambers from the outdoor aviaries. Unlike all previous work on female song preference where females were housed in pairs or triads, Indiana females housed in larger aviary groups without males showed no reliable preference for local over geographically distant and distinct song (South Dakota and Texas). Furthermore the juvenile females showed no concordance with other juvenile or adult females in responding to song in contrast to adult groups housed with males, which show high intra-group agreement. Finding modifiability of female preferences in aviary settings led us to investigate what information about song quality could be communicated among groups of non-singing females. We configured two outdoor aviaries both containing groups of juvenile and adult females from September to December 2001 and exposed them to singing and non-singing juvenile and adult males and to playbacks of juvenile and adult song. We made two discoveries. First, females assorted by age class in real time in the presence of song. Assortment occurred only for adult song, whether sung by males or played over a speaker. Second, because of the immediacy of the female response to song, we videotaped female behavior during playbacks and discovered that females wing stroked to song whether males were present or not, which elicited approaches and looks by other females (see figure below). These results are concordant with the hypothesis that female reactivity to sexually selected traits is malleable and can be detected early in development. These findings also suggest females may have a visual means to communicate among themselves about male traits.

WHY STUDY ANIMALS IN GROUPS?

WHY STUDY ANIMALS IN GROUPS? 

We believe that studying ecological relevant groups of individuals is fundamental to understanding social development. The interaction of behavior of different classes of individuals produces non-additive and non-intuitive outcomes. Furthermore, studying the acquisition of social information requires that we do not impose the social interactions upon subjects, but instead allow the subjects to choose the stimulation most important to them- something that cannot occur in confined lab settings. We have found the following effects that differ or do not exist in the confines of traditional lab settings:

1. Modifiability of female song preferences can occur in groups in aviaries, but not in the pairs or triads in sound attenuating chambers. Females show obvious visual signals in aviaries to one another suggesting that their preferences are transmitted socially.
2. Female social proximity to juvenile males in chambers serves to increase potency of male song. Female social proximity to juvenile males in aviaries serves to stimulate male-male competition and in turn reduce male potency.
3. Group processes influence information acquisition. Groups in aviaries adjacent to one another show little cross-group interaction, no song sharing, and can even differ in reproductive stimulation, egg production, song potency and mating success.
4. Countersinging by males stimulates female egg production. Countersinging by males requires groups of males to sing together and groups of females to be stimulated by them (and to stimulate them).
5. During fall, winter and spring, individual birds self-select different social networks within a larger social group.  These social networks appear to correlate with reproductive success.

 <------click to start  Quicktime 5 or 6 2.2 MB file 

  Male cowbird: A “bird’s-eye” view of a male cowbird on a cold morning in January.

Like the sequence above, this recording was made in early March while the male’s song was variable from rendition to rendition. Prior to this sequence, this juvenile male had sung several hundred songs during the previous two hours. None of these songs had elicited any obvious reaction from the female.  In the first frame, the male sings an undirected song and the female wing strokes with her right wing while the song is being sung. He immediately jumps toward the female and tries to observe the wing stroke in frames 2 & 3. After the male has approached, the female then starts to stretch her right wing as can be seen in frames 3 and 4. The male continues to move closer to the female in frames 4, 5, attempting to observe the female wing action. By this time, the female has had enough of the male and in frame 6 she reminds him to keep his distance.