Details of the taxa included in this study.

PSMC = Pairwise Sequential Markovian Coalescent, ENM= Ecological Niche Modelling, y = yes, n = no.

Effective population size for passerines.

(a) Pairwise Sequential Markovian Coalescent (PSMC) plots using the settings –p “2 + 2 + 30 * 2 + 4 + 6 + 10” displaying reconstructed effective population size values with time for passerines based on whether habitat availability increased (left) or decreased (right) during the Last Glacial Period (LGP). Colours indicate the archipelago the bird belongs to, and the line style indicates the dietary habit of the bird species. Bold lines indicate large (> 50 g body mass) bird species. The grey bands indicate the approximate durations of the Last Interglacial (LIG) and the Last Glacial Maximum (LGM). Black arrows indicate if habitat availability increased or decreased during the LGP. A mutation rate of 1.4 x 10e–9 years/site and a generation time of 2 years for passerines, and a mutation rate of 1.91 x 10e–9 years/site and a generation time of 1 year for non-passerines were used to generate plots. Only species for which both Ecological Niche Modelling and PSMC analyses were possible are shown. Zosterops hypoxanthus is not displayed because its Ne values far exceed those of the other species. (b) Comparisons of Effective Population Size (Ne) at the Last Interglacial (LIG) and Last Glacial Maximum (LGM) incorporating bootstrapped Ne values generated using the same PSMC settings. Boxplots display bootstrapped Ne values, and blue points display the non-bootstrapped Ne value. Outlying bootstrapped and non-bootstrapped Ne values are not displayed. “***” indicates p < 0.001.

Effective population size for passerines.

(a) Pairwise Sequential Markovian Coalescent (PSMC) plots using the settings –p “2 + 2 + 30 * 2 + 4 + 6 + 10” displaying reconstructed effective population size values with time for non passerines based on whether habitat availability increased (left) or decreased (right) during the Last Glacial Period (LGP). Colours indicate the archipelago the bird belongs to, and the line style indicates the dietary habit of the bird species. Bold lines indicate large (> 50 g body mass) bird species. The grey bands indicate the approximate durations of the Last Interglacial (LIG) and the Last Glacial Maximum (LGM). Black arrows indicate if habitat availability increased or decreased during the LGP. A mutation rate of 1.4 x 10e–9 years/site and a generation time of 2 years for passerines, and a mutation rate of 1.91 x 10e–9 years/site and a generation time of 1 year for non-passerines were used to generate plots. Only species for which both Ecological Niche Modelling and PSMC analyses were possible are shown. (b) Comparisons of Effective Population Size (Ne) at the Last Interglacial (LIG) and Last Glacial Maximum (LGM) incorporating bootstrapped Ne values generated using the same PSMC settings. Boxplots display bootstrapped Ne values, and blue points display the non-bootstrapped Ne value. Outlying bootstrapped and non-bootstrapped Ne values are not displayed. “***” indicates p < 0.001, and “*” indicates p < 0.05.

Example Ecological Niche Modelling plots for species from Papua (Cnemophilus loriae), Australia (Alectura lathami), the Philippines (Irena cyanogastra), and the Caribbean (Puerto Rico, Todus mexicanus).

LIG = Last Interglacial. LGM = Last Glacial Maximum. MDH = Mid Holocene. CUR = Current. The continuous heatmap represents the probability of occurrence of the species and red points are known occurrences from GBIF. For all the plots see Figure S2.

Results of the best Bayesian multivariate regression model performed.

The response variable represents the changes in effective population size (increased (=1) or decreased (=0)) during the Last Glacial Period. Shapes represent mean values, thick lines represent the 80% confidence intervals, and thin lines represent 95% confidence intervals. Model parameters and coefficients are provided in table S6. ΔHabitat = the change in suitable habitat from the Last Interglacial to the Last Glacial Maximum.