Habitat selection decisions can impact individual fitness and ultimately scale up to mediate population dynamics. Understanding how birds select habitat is thus critical for discerning the biological processes structuring populations and developing conservation strategies, particularly for species in decline. Marbled Murrelet (Brachyramphus marmoratus; hereafter murrelet) populations have declined in recent decades due to loss of late-successional forest nesting habitat and changing ocean conditions that impact foraging success. Most other seabirds in the family Alcidae nest colonially and evidence suggests nesting murrelets may aggregate in stands, yet no studies have examined murrelet use of social information in nest site selection. In 2016 we experimentally simulated presence of murrelets at 14 randomly chosen potential breeding sites by broadcasting murrelet calls throughout the breeding period. Between broadcasting bouts, we recorded calls of wild murrelets and compared call rates with those recorded at 14 control sites (no broadcast). One year after playbacks ceased (2017) we conducted breeding season surveys to document behaviors indicative of murrelet breeding activity. Broadcasting murrelet calls in 2016 increased daily odds of wild murrelets vocalizing during the treatment period by up to 15.4× (95% CI = 2.3, 125.4) relative to control sites. During the 2017 breeding season, the odds of occupancy were 10.0× (CI = 1.2, 81.4) greater at treatment sites than control sites. These results indicate that social information influences murrelet breeding site selection because simulated conspecific presence in potential nesting habitat appeared to attract prospectors in 2016 that continued occupying treatment sites the following year. This conspecific attraction implies murrelet nesting sites are likely to remain occupied over time and that large tracts of nesting habitat may be important for supporting murrelet populations. Murrelets may also be susceptible to information-mediated Allee effects whereby a lack of conspecific information about nesting habitat could exacerbate long-term population declines.

This dataset has been formatted using Tidy data principles (Wickham 2014).  That is, each observational unit is stored in its own table, each observation is a row, and each variable is a column.  There is a meta-data tab included in the Excel workbook that explains all variables in each table.  We have also included the R code required to import these data and perform all analyses in the manuscript.

Wickham, H. (2014). Tidy data. Journal of Statistical Software 59:10.