New Page

New Page Sim-Songbird Background:
The ability of males to sing "local" song types can be an important factor determining a male's reproductive success. Our lab has further shown that during aggressive interactions, song sparrows use song types shared with neighbors in complex ways to communicate levels of aggression and intent to attack. Thus, song sharing is an important (and often overlooked) behavior in many species. The purpose of this model is to test out specific hypotheses we have about cultural transmission of a particular behavior: the sharing of distinctive song types among neighboring birds. We also are interested in testing hypotheses we have about how complex patterns of song sharing have arisen in the bird neighborhoods within our study populations. Below are a few of the many variables that may affect sharing and the transmission of song types to succeeding generations of songbirds.

Environment:

Habitat patchiness (often caused by human disturbance) can decrease the number of song tutors a young bird has to learn from.
Habitat structure can also determine who birds learn from as well. Various barriers to movement, such as open fields for species who prefer wooded areas, can create complex patterns of song sharing at the local level.
Mortality rates vary with geographic location, capacity of habitat to sustain individuals, preditor densities, bird age, etc. Does mortality rate affect 1) what songs a bird learns, and 2) how many neighbors.

Behavioral strategies:

Timing of dispersal is a crucial event for young birds, and can have a direct effect on how many songs they share with their territorial neighbors. If birds leave their natal area after they have memorized their songs, they will likely not share song types with any neighboring birds. Likewise, if birds have the ability to learn later in life, they can drop songs that are different from those in the area they have settled and add "local" song types.
Song learning strategies have a direct impact on sharing: if birds prefer to learn shared songs, then song sharing should generally increase in the population. However, birds may use other learning strategies, such as learning from the most successful male, or learning from as many neighbors as possible, or even randomly selecting songs. Without a model, it is hard to make predictions about what effects these strategies have on sharing.

Pre-existing cultural conditions:

The songs that already exist in a neighborhood are the only material, besides improvisation, that young birds have to use for their own song learning. This recursive connection to previous generations can create complex patterns of sharing that, in turn, affect successive generations and can only be seen in a generational model such as the one I am developing.

The Program:
Sim-Songbird simulates the song dynamics of a neighborhood of songbirds over time. Simulated birds hatch, disperse, memorize songs from tutor birds, select their repertoire, establish territories, and tutor new birds. Important variables such as when birds learn, who they learn from, environmental factors, etc are adjustable as parameters. Over years, a pattern of song sharing develops that can be compared with real bird populations. The screen picture below is from the model in its current state of development. Although not complete, this model still generates patterns of song sharing that look like natural conditions and which can be manipulated by changing parameters.

(Click to see an expanded picture)