Skip to main content

Table 2 Best-fit PGLS models. Overall, wing patches were best predicted by climate, while we found no relationship between face color and biogeographic variables. Models were selected for each patch subset and each color principal component. Coefficients are presented in the order that they are listed under the “predictors” column, with the intercept value as the last coefficient

From: Macroevolutionary bursts and constraints generate a rainbow in a clade of tropical birds

PredictorsAICAdj-R-SqauredP-ValueCoefficientsPatch SubsetColor PC
Ppc1 + Alt219.40.03670.07192.42167x + 0.00029x + −0.00835Wing1
Ppc1 + Ppc3279.7−0.00320.42732.97335x + 0.00028x + 0.00094Wing2
Tpc1 + Tpc3 + Ppc1 + Ppc3 + Alt236.20.12030.00681.42404x + −0.00068x + 0.03842x + 0.0016x + 0.00191x + 0.01336Wing3
Ppc3 + Alt257.20.01520.18943.60072x + −0.00101x + 0.006Face1
Ppc1 + Alt259.40.01050.23353.16067x + −0.00024x + 0.00789Face2
Tpc3 + Alt269.0−0.00530.46872.74926x + 0.02832x + 0.00556Face3
Ppc3 + Alt247.20.05610.02934.54306x + −0.00171x + − 0.00519Abdomen1
Ppc1 + Alt303.80.01330.20643.93479x + −0.00054x + 0.00722Abdomen2
Tpc1 + Ppc2280.30.02570.11852.18623x + −0.00026x + 0.00086Abdomen3
Ppc2 + Ppc3469.30.04990.039217.80537x + −0.00178x + − 0.00575All1
Ppc1 + Ppc2477.6−0.00120.391519.72418x + −0.001x + − 0.00154All2
Tpc1 + Ppc1 + Ppc3505.60.03310.116113.56358x + −0.00497x + 0.0089x + 0.00577All3