Skip to main content
. 2015 Nov 16;6:8877. doi: 10.1038/ncomms9877

Table 1. Summary of the multi-model inference conducted to explain avian responses to human disturbance.

Predictor Levels Estimate s.e. Importance
All birds (180 species, 457 effect sizes)
 Intercept   0.736 0.599  
 Habitat contrast        
  Natural versus urban −0.941 0.632 1
  Rural versus suburban −1.438 0.762
  Rural versus urban −1.689 0.589
  Suburban versus urban −1.690 0.740
  Low versus high in urbanized areas −1.274 0.635
  Low versus high in recreational nature −0.684 0.593
  Low versus high in islands −0.894 0.717
  Low versus high in reserve −1.622 0.723
 Body mass   −0.315 0.075 0.98
 Clutch size   0.052 0.021 0.93
 Diet        
  Herbivorous −0.232 0.131 0.91
  Omnivorous −0.260 0.105
 Habitat openness   0.182 0.091 0.83
 Migration   0.127 0.080 0.70
 Group size        
  5–50 Individuals −0.060 0.088 0.51
  >100 Individuals 0.097 0.138
 Foraging habit   −0.007 0.109 0.41
         
Rural versus urban birds (103 species, 324 effect sizes)
 Intercept   −0.662 0.309  
 Body mass   −0.403 0.088 0.99
 Diet        
  Herbivorous −0.376 0.164 0.88
  Omnivorous −0.244 0.130
 Habitat openness   0.215 0.105 0.85
 Clutch size 0.035 0.025 0.64
 Migration   0.123 0.095 0.62
 Group size      
  5–50 Individuals −0.100 0.108 0.56
  >100 Individuals 0.091 0.176
 Foraging habit   0.079 0.130 0.46

Results are shown from both a meta-analysis using the full data set (all birds) and a meta-analysis focusing on the contrast between rural and urban populations. Values are average coefficients of models (estimate), their associated standard error (s.e.), and the importance of each factor in explaining species responses to human disturbance (the closer than 1, the most important the factor). Habitat contrasts presented as ‘low versus high in' refer to contrast between populations experiencing low and high human disturbance within a given habitat type.