Current Research Projects

The term “social behavior” subsumes a wide range of behavioral interactions. Because of this diversity it seems difficult at times to identify tractable experimental model systems that capture the full range of processes mediating the complex behavioral interplay among individuals. In most model systems, control is achieved by removing the subject from its social environment, but doing so often also removes stimuli critical to the function of behavior. Traits and behaviors of social creatures evolved and thus were designed to function in social environments. Observing individuals in social isolation therefore may not provide an accurate view of the stimulus control or function of subjects’ behavioral repertoire.Work in the lab has been investigating ways in which the social environment can influence the development of animals' mating behaviour. Historically, in investigations of development or learning, little is known about how animals actually acquire information from others in complex social settings. Thus, a primary goal of following studies was to determine if experiments, as opposed to descriptive studies, could be carried out on animals living in groups, continuously exposed to complex auditory and visual stimulation. A major concern was whether this multivariate approach to studying learning and development could be undertaken without sacrificing scientific rigor.

The Development of Social Skills

This work started many years ago, using the outdoor aviary facilities at Indiana University and working in collaboration with Meredith West and Andrew King. To test the influence of cowbirds' social ecology on development, we housed cowbirds in large social groups and we used naturally occurring variation in social experience (the age class of individuals) to manipulate the social composition of these groups. In the first study, females were housed in one of four conditions: (1) with juvenile males, (2) with adult males, (3) with both juvenile and adult males, or (4) with no males.

During this experiment, we developed computerized system using voice recognition to record all data by speech and transcribe it to a database automatically. This allowed us to take extensive and detailed measures of birds' interactions daily throughout the year.

Developmental trajectories of all classes of cowbirds changed with social context.

Numerous differences in the birds' behaviours led to the emergence of dramatically different cultures in the four conditions as measured using auditory analysis, spatial assortment, and social interactions. The most pronounced effect was of adult male stimulation on juvenile male development. Juvenile males raised without adult males developed atypical social behaviour.

During the breeding season these males engaged in little intrasexual competition; they did not act aggressively toward one another nor did they guard females and they copulated with several different partners. These behaviour patterns differed dramatically from juvenile males that had been raised with adult males. Adult-raised juveniles behaved in a manner more typical of cowbirds in the Midwest; they were aggressive and socially monogamous. Furthermore, song analysis revealed that the songs of juveniles housed without adults were structurally different than those of juveniles housed with adult males, and were more effective in eliciting copulatory responses from females. Adult males appear to facilitate the development of competitive skills, but inhibit the learning or expression of vocal skills. These effects have now been replicated, found to be present very early in development, and persistent into adulthood. Furthermore, these patterns of behaviour (both typical and atypical) can be culturally transmitted to a new generation of juvenile males.

Social factors were of fundamental importance to the development of major aspects of cowbird behavioural repertoires. We have now documented social influences on: copulation success, egg production, mating system, aggression, mate choice, mate preferences, song development, song use, and reproductive success. This paradigm, outlined in the above studies, provides a new means to study social influences on development, cognition, perception, attention, and learning as well as on cultural evolution.

Recent work has focused on adult birds.  We were surprised to find that there was pronounced plasticity in adults social competence when they were exposed to differing types of social environments. This work has shown that social skills are highly malleable throughout life. 

Female preferences and reproductive output

Perhaps the most dramatic social effects have been on females’ reproductive physiology and behavior. Modifying social environments produced pronounced plasticity in females’ song preferences, mating patterns, egg production, and has even affected the health and viability of their offspring.

Despite the critical importance of females’ song preferences in influencing social structure, communication, assortative mating, and sexual selection, almost nothing is known about the physiological control or development of females song preferences. We undertook a series of experiments on female cowbird song preferences that was successful in documenting plasticity in their preferences, both in adult and juvenile females. When housed socially, females display marked levels of malleability in their preferences for songs. The methods we used in the experiments were effective in modifying preferences for songs of individual males, for types of songs, and even in removing preferences entirely for large-scale features of songs. We were even able to detect candidate mechanisms by which females might be communicating their preferences to one another (‘wingstroking’ in response to preferred song variants). 

Recently we have developed a new procedure to modify preferences. This work has revealed that females will copy the song preferences of other females, a mechanism akin to mate-choice copying . Graduate student, Grace Freed-Brown, used a tutoring procedure that paired male song variants with a female chatter vocalization – a vocalization that females make in response to a song of a male who is courting her. We broadcast these pairings to females and then tested their preferences for the songs. Females showed enhanced preferences for the songs that were paired with chatter. 


 This figure shows a sound spectrograph of a female's chatter vocalization experimentally added to the end of a male cowbird's song   

The majority of our work on reproduction comes from experiments studying females living and breeding in groups in outdoor aviaries. Since they are brood-parasites, female cowbirds do not build nests or raise young, but they will parasitize false nests located throughout the aviaries. We collect and incubate the eggs, and subject them to DNA parentage tests. This provides measures of reproductive success for all subjects in the groups.

One surprising discovery has been the substantial variation found across groups in egg production. Some aviaries produce hundreds of eggs in a breeding season, whereas others produce only a handful. While there is variation among females within groups, DNA parentage tests have revealed that the across-group variation is not due to a few extremely fecund females. Instead in some social conditions females as a group increase their egg production. We conducted a meta-analysis on numerous groups to investigate whether there were any social behaviors related to reproductive output. It revealed that two behaviors of males were critical to reproductive success – amounts of female directed singing produced (related to copulation success) and male-male countersinging bouts produced (related to egg production). These two behaviors are also ones that we have found to be highly influenced by social learning during development. The countersinging results were unexpected and were found to account for much of the variance in egg production across groups. Results suggest that females are eavesdropping on male-male competition and in groups where more of this information exists, females invest more in egg production.

Female control over their social network

In collaboration with Marc Schmidt, we have developed a new paradigm to study the underlying neural control of social behavior. Working with Graduate student Sarah Maguire, we conducted a lesion experiment this summer where a subset of females in aviaries were given targeted chemical lesions to a nucleus known as HVC. In males, HVC is part of the neural song system that has been implicate to control singing. In female songbirds, who do not sing, the area relates to their song preference (lesions to HVC in female songbirds results in females that are not selective in the songs to which they give copulation solicitation displays). We placed lesioned females into aviaries during the breeding season (and sham lesion controls placed in other aviaries). Results were dramatic. The lesions changed the way the individual females interacted with males, they allowed many different males to sing to them (a pattern we have never seen in any female cowbirds before). The most surprising effect in the experiment however was on the rest of the group. The changes to the lesioned females caused cascades of effects throughout the entire social network. The group-mates of the manipulated females showed pronounced changes in their social behavior, including changes in male singing patterns, pair-bonding patterns, disruption of the male dominance hierarchy, and even the way the other female interacted with males. The findings, which are the first to integrate neural manipulations with social network analyses, have two broad implications for the study of social behaviour, first they provide a means to study experimentally the role of individuals in structuring social environments, second, they provide a new way to examine the link between brain, behavior, and sociality.

Female cognition and nest selection

While collecting eggs from the aviaries, we devised experiments to determine whether selection has favored cognitive skills in female cowbirds allowing them to select nests that maximize the likelihood that their young hatch and develop successfully. We found that females can use the type and number of host eggs in the nests to determine whether the owner of the nest is a suitable host, and also to time the readiness of the nest for parasitism. This preparation has been a wildly successful tool to investigate memory, counting, and timing skills. In this past breeding season we have teased apart counting and timing skills (they do both very well), (2) determined when the nest selection decision is made (the day before laying), and (3) started to investigate the stimulus control of the decisions by breaking expectations (taking eggs away instead of adding for example). This procedure is one of the best means to evaluate animal cognition using a task that is ecologically relevant and requires no training. 

This figure shows eggs collected from 3 types of nests across 3 days of testing. A nests contained one egg on  day one, increased to 2 eggs on day 2 and increased to 3 eggs on day 3. B nests contained two eggs on day one, increased to 3 eggs on day 2 and remained at 3 eggs for day 3. C nests never changed, containing 3 eggs on each of the three days. Females showed preferences to lay eggs in nests that had changed from the day before (White et al. 2010) 

Social influences on nest selection results: 

All Data is based on preference for the "Change" Nest All Data is based on preference for the "Change" Nest Exp1
Female Band Experimental Control Difference between Exp. & Control Preference Score When Demonstrating Score of the Demonstrators Observer Score (4DEM) Demonstrator Score (4DEM) Score of the Demonstrators during 4DEM Observer Score (2DEM) Demonstrator Score (2DEM) Score of the Demonstrators during 2DEM Difference between 4DEM & 2DEM Observer Score (0DEM) Demonstrator Score (6DEM) 2DEM 4DEM Avg. Conformity Score of the Demonstrators - Control
Avg. Score Max. Score Min. Score Avg. Score Max. Score Min. Score Avg. Score Max. Score Min. Score DEM OBS DEM OBS DEM OBS Both Avg. Score Max. Score Min. Score
F2H 0.44 0.22 0.23 0.44 0.59 0.94 0.24 0.17 0.13 0.41 0.60 0.25 0.15 0.92 0.13 0.25 0.00 0.02 0.13 0.34 0.51 0.58 0.56 0.56 0.54 0.57 0.54 0.44 0.87 0.00
F2P 0.25 0.00 0.25 0.00 0.59 0.94 0.24 0.77 0.39 0.93 1.00 0.79 0.84 0.46 0.13 0.25 0.00 0.07 0.86 0.32 0.60 0.67 0.56 0.50 0.58 0.58 0.56 0.20 0.25 0.14
FDN 0.75 0.53 0.22 N/A 0.59 0.64 0.55 1.00 0.25 0.75 1.00 0.46 0.38 0.62 0.73 0.50 0.62 0.42 0.49 0.63 0.49 0.52 0.51 0.71 0.87 0.54
FHB 0.44 0.27 0.17 0.14 0.52 0.61 0.44 0.78 0.48 0.68 1.00 0.17 0.00 0.25 0.23 0.46 0.00 0.78 0.82 0.07 0.45 0.61 0.50 0.60 0.47 0.61 0.54 0.44 0.87 0.00
FHN 0.14 0.00 0.14 0.61 0.59 0.94 0.24 0.65 0.94 0.21 0.56 0.00 0.50 0.50 0.60 1.00 0.00 0.15 0.83 0.52 0.91 0.72 0.56 0.62 0.73 0.72 0.44 0.87 0.00
FLO 0.82 0.78 0.04 N/A 0.59 0.64 0.55 0.82 0.00 0.75 1.00 0.46 0.79 0.79 1.00 0.50 0.03 0.59 0.40 0.27 0.71 0.61 0.71 0.44 0.46 0.71 0.87 0.54
FMW 0.68 0.68 0.00 0.87 0.59 0.94 0.24 0.64 0.90 0.21 0.56 0.00 0.00 1.00 0.23 0.46 0.00 0.64 0.20 0.23 0.35 0.69 0.50 0.25 0.42 0.47 0.47 0.20 0.25 0.14
FNP 1.00 0.02 0.98 0.24 0.52 0.61 0.44 1.00 0.00 0.93 1.00 0.79 0.00 0.00 0.13 0.25 0.00 1.00 0.00 0.00 1.00 0.00 1.00 0.92 0.20 0.25 0.14
FNR 1.00 0.00 1.00 0.25 0.52 0.61 0.44 1.00 0.13 0.68 1.00 0.17 0.00 0.00 0.23 0.46 0.00 1.00 0.86 0.00 0.40 0.75 0.92 0.57 0.92 0.73 0.44 0.87 0.00
FPB 0.69 0.35 0.34 N/A 0.59 0.64 0.55 0.83 0.46 0.17 0.39 0.00 0.42 0.62 0.73 0.50 0.41 0.16 0.06 0.46 0.34 0.39 0.34 0.43 0.47 0.71 0.87 0.54
FPH 0.00 0.35 0.35 N/A 0.59 0.64 0.55 0.88 0.00 0.17 0.39 0.00 0.30 0.79 1.00 0.50 0.58 1.00 0.00 0.49 0.38 0.00 0.43 0.50 0.71 0.87 0.54
FWH 0.92 0.39 0.53 0.94 0.52 0.61 0.44 0.15 0.50 0.41 0.60 0.25 0.46 0.73 0.54 1.00 0.00 0.31 0.74 0.50 0.55 0.65 0.64 0.58 0.60 0.59 0.20 0.25 0.14

More information about the laboratory preparation:The housing conditions underlying this work is a significant departure from other experimentally controlled investigations in animal behaviour. We provide our subjects with a great deal of freedom to engage, or not engage in social interactions as they see fit. While we can control the individuals' present in the conditions, the interactions, choices, decisions of the birds is left to them. The subjects effectively have the freedom to select their own independent variables in the experiments. It is thus our task to figure out what variables they choose. 

We have moved to these types of conditions for our lab work because we have been influenced by the work of birdsong field biologists. Results of field studies often reveal that when birds are given the opportunity to have more degrees of freedom in their social behaviour, new patterns of learning emerge that can be very different from patterns seen in confined, socially isolated conditions.

The results of our work have revealed the value of this approach. When the birds are given the opportunity to select the variables in the studies, they invariably select important variables. Variables that we would never have thought to have studied.

It is however important to remember that this is not field work. The birds are held in a laboratory and thus the concerns of studying animals in captivity apply. The aviaries cannot be considered analogous to the wild. The housing conditions constrain subjects' range, they force individuals into the groups we create, birds are not exposed to predation, birds have constant access to food, water and shelter at all times. All of these factors and more could influence subjects' behaviour. We attempt to deal with these concerns by following these principles:

1.Using field data to guide experimental design. The variation that we study across social conditions occurs naturally across the cowbirds' range. We strive to create groups differing in social composition to mirror social compositions that have been reported in the wild. For example, in some western populations, adult birds move off their breeding areas quickly in search of food. In such populations, juvenile cowbirds may never interact with adults in their first year. In other locations, juveniles join flocks while adults are still in the final days of breeding and remain with the adults for their lifetimes. This variation served as the impetus for the study described above. In another example, in newly settled areas, cowbirds experience more social fluctuation with more individuals entering and leaving groups, whereas in other older populations there is more social stability.

It becomes clear when investigating field reports of cowbird social structure is that there is no longer a “species-typical social composition”. Cowbirds are non-territorial and range across the North American continent. Cowbirds seem to be able to exploit any type of habitat effectively and have moved into new areas as a result of human-induced habitat changes. Flocks composed of all females, all juveniles, and all classes have been reported in the field. There are also reports of variation across populations in: group density, breeding area climate, sex ratio, timing of the breeding season, and migration.

2.Comparisons are made across conditions . Because of the artificial conditions, our studies cannot provide direct information about patterns of cowbird behaviour in the wild. However, because we  can influence social dynamics in a way that cannot be controlled in the field, we are able to have a window into behaviour that cannot be studied in the field. It is our hope that these lab findings will lead field investigations (see 3 below). We  always construct the compositions of the conditions to allow for comparisons across groups so that differences across conditions may reveal the variables that are important to cowbird behaviour. For example, we cannot and do not ever attempt to suggest that the number of eggs females lay in the aviaries is representative of the number of eggs females would lay in the wild because of the variables that differ in captivity (such as nutrition, calcium supplements, nest availability, etc). That being said, looking at individuals across years and across conditions reveals that certain characteristics of the social environment (male countersinging for example) play a role in influencing female decision processes for trading-off current and future reproductive success. This example also highlights the power of the aviary preparation, as effects like this could not be studied in the wild. In the wild it would be impossible (a) to find every egg females laid within or across years, and (b) to be able to compare patterns of egg laying across different types of social groups that do not differ in some other ecological variable.


3.Predictions for field study. In point one above, we discussed how using information from field reports is important in generating theory and experiments in the lab. It has, however, been extremely difficult to examine whether the variables that are revealing themselves as important in the lab are actually important in the field. This is a result of the difficulty of studying some of these early developmental experiences in a non-territorial species in the field. Our work does provide a number of testable predictions for cowbirds in the wild. For example, results from White et al. (2002c) suggest that juvenile males living in flocks without adult males may experience less aggression in response to their directed song overtures. In the aviaries, one consequence of this was that juveniles without adult male exposure developed songs with more notes and more note clusters than did the juvenile males who were housed with adult males. Freeberg has studied the songs of two subspecies of brown-headed cowbirds, Molothrus ater artemisiae, and M. a. ater. Artemisiae males from his study population were more likely to be found in large all- juvenile flocks, whereas the ater juveniles were more likely to be found in mixed age-class flocks. Song analysis revealed that artemisiae males had significantly more note clusters and more notes within clusters in their songs than did the ater birds (Freeberg, 2001). This study reveals that it may be possible to study the social effects in which we are interested using aviary and field studies in an integrative manner.


4.Some benchmarks in aviary behaviour. Even though the above points argue for the use of cowbirds in our experiments, they do not speak to the issue of whether there may be problems with social housing of birds in these conditions. It is possible that confinement at such densities might produce abnormalities in behaviour such that results would not provide valid information about sexual selection or social learning. Here are some measures of behaviour in the lab that suggest we are not generating abnormal behaviour patterns.

Density . In Indiana, we studied several flocks in a very large aviary system (comprised of four smaller aviaries positioned end-to-end; spanning the length of a football field). While this is still smaller than their natural range, birds can disperse such that they would not be highly clustered. We have housed flocks in the large aviary complex four times now (including Smith, et al. 2000, for example), and each time we have done so, we have seen that birds did not disperse but instead clustered together in one quadrant of the large aviary in much higher densities than we had expected. The flock did move around in the large enclosure, and sub flocks formed, but bird always remain very closely clustered. This is at least suggestive that when they have more choice, they choose to remain in a group. 

Wild caught birds . We are able to catch birds from the wild up to one month prior to the breeding season, place them in aviaries, and have them reliably sing, compete, court, mate, and reproduce. One of the main problems with captive breeding programs in other species is that the animals do not have the natural stimuli available to elicit reproduction. The aviaries provide males and females with the stimuli needed to reproduce. Wild caught adult males come into aviaries exhibiting the same degree of variation in social behaviour as the males raised in different social conditions in the experiments. In addition, females caught up to two months prior to the breeding season will respond in playback experiments and their patterns of responding with copulation solicitation displays to male songs are similar to the responses of females who have been in aviaries for their lifetimes. Also, in playback studies, responses by females to songs of wild caught males do not differ significantly from responses to songs of aviary-raised males.

False nests and female laying behaviour. Another abnormal aspect of the aviaries is that there are no host species available. We make sure the false nests are as realistic as possible, using mock eggs in the nests (one is placed into a nest each day and new nests are added and old ones removed throughout the breeding season).  In the past we have experimented with hosts (juncos) in aviaries in the past, we never found any differences between patterns of laying when hosts were present and when absent. Cowbirds and juncos rarely ever interacted in the aviaries.

Finally, we have as little contact with the birds as possible. Birds are handled only rarely, and they are given ad lib food so that they do not associate humans with feeding schedules.  

In sum, we strive to study variables that have been shown in the wild to be important to cowbird behaviour across their range. Cowbirds are an excellent species for studying social behaviour because there is such dramatic variability in social behaviour and social structure in the wild. While the data from the field certainly are indicative that social ecology affects learning, development and reproductive success, it is difficult to make such a causal statement without the experimental control provided by the lab environment.