The morphological data (including yield) were analyzed
using one common linear fi xed-eff ects model for each trait.
For nearly all traits there was insuffi cient statistical evidence
to justify fi tting a model with a separate slope for each area
and/or density, but including a separate intercept was necessary.
Therefore, a model used was a simple fi xed-eff ects regression of
trait value against year of introduction with a separate vertical
intercept for each combination of environment and density.
这段话让我很晕，不明白什么意思，尤其是 a separate slope和a separate intercept