Mixed-species flocking is associated with low arthropod detectability and increased foraging efficiency by Yungas forest birds in Argentina

Mixed-species flocks presumably provide birds with antipredator and foraging benefits. The foraging benefits hypothesis predicts that a reduction in arthropod abundance will trigger flocking activity; however, flocking activity may also be influenced by the difficulty of detecting arthropods, a seldom explored possibility. We found that environmental traits (temperature and foliage density) combined with arthropod abundance explained arthropod detection by birds in the Yungas foothill forest of NW Argentina. Prey detection was inversely related to ambient temperature and foliage density while positively associated with arthropod abundance. Based on this result, we built a structural equation model using a latent proxy variable for arthropod detectability, arthropod crypsis, integrating ambient temperature, foliage density, and proportion of immature arthropods. This model allowed us to compare the relative importance of arthropod abundance and the difficulty in detecting prey items as predictors of flocking propensity. After two years of studying 129 mixed-species flocks, 1,344 bird foraging sequences, and 25,591 arthropod captures, we found that the flocking propensity of birds was only significantly correlated with arthropod detectability and not with arthropod abundance. Flocking propensity peaked when the arthropod community was comprised of proportionately more immature and non-flying arthropods, the temperature was low, and the foliage cover was more dense; all factors are contributing to a low arthropod detectability. Finally, we evaluated whether joining mixed-species flocks provided foraging benefits such as increased foraging efficiency. Individuals benefited from joining flocks by an average increase of their prey-capture attempt rate of 40%, while the search rate increased by 16%. Our results add a new perspective on the drivers of mixed-species flocking by showing that the capacity to find prey items may have a more significant effect than prey abundance per se.Data were collected from 20 independent transects 100 m long each. Surveys were performed on a seasonal basis. Measured variables: 1. Environmental (temperature and foliage density) 2. Arthropod surveys (two techniques, beating and passive pitfalls) 3. Birds foraging behavior (behavioral observations) Three tables are presented: Table 1. For SEM model and arthropod seasonality Table 2. For Model selection in relation to environmental and arthropod abundance variables Table 3. Birds behavior regarding their social conditio