Skip to main content

Table 2 Evaluation of the Chi-square test statistic (χ2) for fit of observed data to expectations under alternative quantitative genetic models.

From: Retention of low-fitness genotypes over six decades of admixture between native and introduced tiger salamanders

Trait Model χ 2 P Λ
Mass μ 0 28.384 0.0080 NA
  μ0 + bS* 5.979 0.9171 0.0000
  μ0 + bS + bH 5.917 0.8788 0.8038
  μ0 + bS + bH + bSS 5.533 0.8528 0.5357
  μ0 + bS + bH + bSS + bHH 4.791 0.8522 0.3888
  μ0 + bS + bH + bSS + bHH + bSH 4.157 0.8427 0.4262
SVL μ 0 8.714 0.7942 NA
  μ0 + bS* 2.163 0.9991 0.0105
  μ0 + bS + bH 2.129 0.9980 0.8530
  μ0 + bS + bH + bSS 2.042 0.9960 0.7688
  μ0 + bS + bH + bSS + bHH 1.795 0.9943 0.6193
  μ0 + bS + bH + bSS + bHH + bSH 1.527 0.9923 0.6042
Tmet μ 0 45.252 0.0000 NA
  μ0 + bS 36.341 0.0003 0.0028
  μ0 + bS + bH* 17.657 0.0899 0.0000
  μ0 + bS + bH + bSS 17.163 0.0708 0.4821
  μ0 + bS + bH + bSS + bHH 16.781 0.0523 0.5368
  μ0 + bS + bH + bSS + bHH + bSH 16.782 0.0325 1.0000
Growth μ 0 6.568 0.9230 NA
  μ0 + bS* 1.316 0.9999 0.0219
  μ0 + bS + bH 1.317 0.9998 1.0000
  μ0 + bS + bH + bSS 1.137 0.9997 0.6720
  μ0 + bS + bH + bSS + bHH 0.968 0.9995 0.6811
  μ0 + bS + bH + bSS + bHH + bSH 0.757 0.9994 0.6457
  1. The first model (including only μ0) is a null model of no variation among cross types. Significant chi-square values ( < 0.05) indicate lack of fit relative to a saturated model. Significant (PΛ < 0.05) likelihood-ratio tests (Eq. 2) indicate improvement in fit of higher-order models over simpler models. Asterisks denote the best-fit model for each phenotype.