“…Group size (mean ± SE = 4.9 ± 0.1, range = 3-7) was included in all models as it not only reflects potential benefits of passive group augmentation (Figure 1), but is also necessary to control for potential load-lightening effects in larger groups. 34,70 Although group size is negatively correlated with the probability of breeding position inheritance (LM: R = À0.55, t = À7.84, p < 0.01), both variables were included in models since for models without inheritance (Table S2, models 2, 5, 6 and 7), this yielded a better fit to the data compared to the same models excluding group size (DAICc = 4.4, 4.8, 4.3 and 4.3, respectively), while excluding group size yielded a non-significantly better fit for models including inheritance (Table S2, models 1, 3, 4, and 8; DAICc = 0.9, 0.9, 0.4 and 0.4, respectively). Additional variables that are expected to affect the probability to engage in nest defense were also included in all models: sex (male, female), age (first-year, older), brood size, nest stage (eggs, nestlings), predator type (goanna, goshawk, cuckoo), trial no.…”