Fig. 5: Vision and vocal communication play a role in collision avoidance during flocking flight.
a,b, In contrast to the take-off phase (LMM, estimates ± s.e.m.: 0.03 ± 0.05, P = 0.517, t = 0.65; a), call emission rates during sustained flocking flight are affected by the ambient illuminance level (LMM, estimates ± s.e.m.: −0.02 ± 0.01, P = 0.02, t = −2.33; b). Coloured circles and lines indicate individual data points and means, respectively. Colours represent bird ID. Grey diamonds and thick lines mark population means ± s.d., respectively. n = 60 for each light condition. c, The rate of collisions between birds during flocking flight is not affected by the ambient illuminance level (LMM, estimates ± s.e.m.: −0.01 ± 0.01, P = 0.5, t = −0.68). Black asterisks mark individual data points. Meaning of remaining markers as in a. n = 10 per illuminance level. d,e, In contrast to the take-off phase (LMM, estimates ± s.e.m.: 0.05 ± 0.04, P = 0.172, t = 1.37; d), call emission rates during sustained flocking flight are affected by the presence of masking noise (LMM, estimates ± s.e.m.: 0.02 ± 0.004, P < 0.001, t = 5.22; e). Meaning of colours and markers as in a. n = 60 for each noise condition. f, The rate of collisions between birds during flocking flight is significantly affected by the presence of masking noise (LMM, estimates ± s.e.m.: −0.014 ± 0.006, P = 0.036, t = −2.26). Meaning of markers as in c. n = 10 per noise condition. At the top of each panel, the P value of a linear mixed effects model is provided.
