Female and male zebra finches (Taeniopygia guttata) preferences for normal and modified songs
Description
The zebra finch is an established model species for song preference research in songbirds. In this study, a four-way choice operant paradigm (“carrousel” cage) was used to measure the spontaneous preference of adult birds of both sexes for song stimuli of four different categories: a conspecific song, a duration-stretched version and a vocoded version of the conspecific song, and a heterospecific great tit song. In an earlier operant song discrimination experiment both male and female zebra finches treated duration stretched and vocoded song versions as being different from a normal song. This study examines how the same modified songs are treated in a preference test. The results show that females exhibited a gradient of preferences: they preferred the normal and the duration-stretched song over the vocoded version of the conspecific song, but the vocoded conspecific songs were still preferred over the heterospecific great tit songs. Males preferred duration-stretched conspecific songs over heterospecific songs, but otherwise did not differentiate among the stimuli. In contrast to the discrimination study, the current study also shows a sex difference, with female song preference for normal and duration-stretched conspecific songs being stronger than male preferences. Our results demonstrate that the importance of particular song features as obtained in the context of an operant discrimination task and of a preference test can differ. This illustrates that combining different approaches and testing paradigms may help to better understand the significance of various acoustical parameters of birdsong in relation to its social and ecological functions and their relevance to both sexes.
Files
Steps to reproduce
The count of sound perch visits was chosen as a variable indicating a bird’s preference toward a particular acoustic stimulus. Four types of stimuli were used to investigate the song preferences of the birds: normal zebra finch song (referred to as 'Normal_ZF' in the Excel file), duration-stretched zebra finch song (referred to as 'DurationStretched_ZF' in the Excel file), vocoded zebra finch song (referred to as 'Vocoded_ZF' in the Excel file), and heterospecific song (referred to as 'Heterospecific' in the Excel file). The Excel file contains perch visit data from 35 individual birds (16 males and 19 females). All statistical analyses were conducted in Rstudio (R Core Team, 2016). A Generalized Linear Mixed-effects Model (GLMM) was used to assess the relative preferences that birds show among the four stimuli. We used the number of visits (refer to as 'Perch_visits' in the Excel file) for the various stimuli as response variables in GLMM. We used “Test_Stimulus” (Four versions of test stimuli), “Sex” (Male/Female), and the interaction between these two as covariates in the model with “Bird_ID” and “Stimulus_ID” as the random factors and a Poisson error structure of the “Number of visits”. To assess whether any explanatory variables in a model were significant, we used Wald Chi-Squared tests to evaluate the fixed effects of the model (using the function car::Anova, R package car). A post-hoc analysis is performed on the final model with FDR correction (using the emmeans function, R package lsmeans).