| |
Bivariate model
|
Univariate model
|
|---|
|
\(\varvec{\sigma}_{\varvec{P}}^{2}\)
|
\(\varvec{\sigma}_{\varvec{A}}^{2}\)
|
h2
|
r
|
βf
|
|---|
|
Grazing toxic fescue
|
0.90
|
0.38
|
0.40
|
0.45
|
0
|
|
Not grazing toxic fescue
|
0.95
|
0.30
|
0.34
|
0.34
|
− 0.59 hair shedding score units
|
- The estimated phenotypic variance (\(\sigma_{P}^{2}\)), and repeatability (r) from a bivariate model and fixed effect of grazing versus not grazing fescue from a univariate model. Additive genetic variance, heritability, and repeatability are higher for hair shedding recorded while grazing toxic fescue when treated as a different trait from hair shedding while not grazing toxic fescue. When fescue grazing status is fit as a fixed effect in a univariate model, the estimated effect of toxic fescue on hair shedding score (βf) is also higher (i.e. later shedding animals)
