Content bias in the cultural evolution of house finch song

Detecting cultural evolution in a songbird species using community-science data and computational modeling

“…Under a neutral model of song learning, a juvenile randomly chooses a tutor's song to imitate; with conformity bias, a juvenile is disproportionately likely to choose a tutor with the most common song; and directional selection operates to favor certain song properties such as rate of syllable production or higher frequency bandwidth (Podos, 1997;Podos & Nowicki, 2004), such that juveniles are more likely to learn songs that exemplify better performance. Here, directional selection is somewhat analogous to the content bias observed in house finches (Youngblood & Lahti, 2022) since juveniles are choosing to learn a syllable based on its properties and not its frequency of occurrence. Chipping sparrows only learn a single syllable, so in our model, directional selection operates on a continuous feature of a syllable-the rate of syllable production-instead of the selection of certain syllable types to compose a song.…”

“…By comparing data from real populations to predictions from evolutionary models, researchers have identified which of these models best explains the data (Akashi & Schaeffer, 1997;Gutenkunst et al, 2009;Kryazhimskiy & Plotkin, 2008;Williamson et al, 2005). Some researchers apply this theoretical approach to the cultural evolution of song by examining the distribution of song within populations (Lynch et al, 1989;Lynch & Baker, 1993, 1994Mcgregor & Krebs, 1982;Parker et al, 2012) and by developing individual-based or agent-based simulations of song learning that are compared to field-site data (Crozier, 2010;Ellers & Slabbekoorn, 2003;Lachlan et al, 2018;Lachlan & Slater, 2003;Slater, 1986;Wheelwright et al, 2008;Youngblood & Lahti, 2022). Such agent-based simulations have been used in conjunction with birdsong data to infer the learning strategies used by swamp sparrows and house finches (Lachlan et al, 2018;Youngblood & Lahti, 2022).…”

“…Some researchers apply this theoretical approach to the cultural evolution of song by examining the distribution of song within populations (Lynch et al, 1989;Lynch & Baker, 1993, 1994Mcgregor & Krebs, 1982;Parker et al, 2012) and by developing individual-based or agent-based simulations of song learning that are compared to field-site data (Crozier, 2010;Ellers & Slabbekoorn, 2003;Lachlan et al, 2018;Lachlan & Slater, 2003;Slater, 1986;Wheelwright et al, 2008;Youngblood & Lahti, 2022). Such agent-based simulations have been used in conjunction with birdsong data to infer the learning strategies used by swamp sparrows and house finches (Lachlan et al, 2018;Youngblood & Lahti, 2022). With these comparisons of field recordings and results, researchers found evidence for different cultural transmission biases in different species.…”

Cultural Evolution and Music

“…By comparing these models with real datasets, either descriptively or within a statistical framework like approximate Bayesian computation, researchers can make inferences about the processes that in uence the evolution of non-human song. In birds, such models have been used to investigate the emergence of local dialects (Strigul, 2009), the e ect of dispersal on song diversity (Ellers and Slabbekoorn, 2003), drivers of context-dependence in song syntax (Katahira et al, 2011;Morita et al, 2021), and the in uence of transmission biases on population-level cultural diversity (Lachlan et al, 2018;Youngblood and Lahti, 2022). In whales, researchers have used such models to highlight the importance of shared breeding grounds on song evolution (Mcloughlin et al, 2018;Zandberg et al, 2021), explore how di erent life history characteristics interact in driving dialects (Filatova and Miller, 2015), and investigate the e ect of conformity and homophily in stabilizing structural patterns (Cantor et al, 2015).…”

pykanto: A python library to accelerate research on wild bird song

“…To illustrate this point, as part of our research on a wild population of great tits Parus major , we record around 50,000 songs every year, which translates to well over half a million discrete acoustic units. Any analysis that required finding, labelling and characterising them, if done manually—as is still often the case in wild bird vocalisation research (Beecher et al, 2020; Demko & Mennill, 2018; McLean & Roach, 2020; Pipek et al, 2018; Youngblood & Lahti, 2022)—would take a very long time to complete. This bottleneck, in turn, severely limits researchers' ability to ask questions that require large datasets to answer—such as those about social learning, vocal development, large‐scale cultural diversity, and the syntactic organisation of animal vocalisations (Aplin, 2019; Kollmorgen et al, 2020; Lachlan et al, 2018; Sainburg et al, 2019).…”