Table 3 Binomial generalized linear mixed models testing the effect of colony, distance, sex (female as reference level) and their interaction on the use of glacier fronts by black-legged kittiwakes at two different scales: (1) Regional scale, based on all glacier fronts occurring in a radius of 50 km from Kongsfjorden’s centroid and (2) Fjord scale, based on all glacier fronts occurring in Kongsfjorden, Svalbard. Colony was included as a fixed effect to account for their average difference in glacier front use.

From: Fine-scale spatial segregation in a pelagic seabird driven by differential use of tidewater glacier fronts

Effect

Distance

Sex

Distance × sex

Colony

cR2

Model H0

Sex + colony

Distance + colony

Distance + sex + colony

Distance + sex

–

Model H1

Distance + sex + colony

Distance + sex + colony

Distance × sex + colony

Distance + sex + colony

–

Regional scale

\({{\varvec{X}}}_{\mathbf{1}}^{\mathbf{2}}\)= 549.78; p < 0.001

\({X}_{1}^{2}\) = 0.31; p = 0.581

\({X}_{1}^{2}\) = 1.72; p = 0.190

\({X}_{4}^{2}\) = 7.96; p = 0.093

0.59

Fjord scale

\({{\varvec{X}}}_{\mathbf{1}}^{\mathbf{2}}\) = 24.58; p < 0.001

\({X}_{1}^{2}\)= 0.31; p = 0.578

\({X}_{1}^{2}\) = 3.12; p = 0.077

\({{\varvec{X}}}_{\mathbf{3}}^{\mathbf{2}}\) = 7.86; p = 0.049

0.04

  1. The individual ID was fitted as random effects. Likelihood ratio Chi-squared statistic and associated p values are given, with models specifying the null (H0) and alternative (H1) hypotheses specified for each test. cR2 is the conditional R2 (i.e., for both fixed and random effects) for the model using the distance, sex and colony as additive fixed predictors.
  2. Significance (shown in bold) based on the nominal threshold of 0.05
Back to article page