Table 2

Likelihood ratio tests comparing generalized linear mixed models
Effect tested Fixed effects Random effects AIC X2 d.f. p
a)
P, L, P×L C, P 2149
Capsule (C) P, L, P×L P 2350 203.1 1 < 0.001
Position (P) P, L, P×L C 2169 21.8 1 < 0.001
Population × P, L C, P 2131 14.2 16 0.58
Location (P×L)
Population (P) L C, P 2135 12.1 4 0.017
Location (L) P C, P 2400 277.4 4 < 0.001
b)
P, L, P×L C, P 936
Capsule (C) P, L, P×L P 1061 126.9 1 < 0.001
Position (P) P, L, P×L C 934 0.1 1 0.78
Population × P, L C 935 9.2 4 0.055
Location (P×L)
Population (P) L C 943 15.8 4 0.003
Location (L) P C 942 8.3 1 0.004

(GLMMs) on the effects of population origin, rearing location, position in capsule, and capsule number on a) embryo survival and b) hatching rate among survivors within the locations Giesse and Worble only (because hatching rates or survival were too low in the other populations). Akaikes information criterion (AIC) describes the quality of fit of each model. To evaluate the significance of fixed and random effects, alternative models without the variable of interest were compared to the full model (bold) using likelihood ratio tests. If an alternative model had a significantly better fit (bold), this model was subsequently compared against further reduced models.

Stelkens et al.

Stelkens et al. BMC Evolutionary Biology 2012 12:247   doi:10.1186/1471-2148-12-247

Open Data