model edits

[zenrabbit]

Also a suggestion for your emmeans, I wouldn't test every single pair of possible interactions. Testing High Phacelia vs. Low Salvia is not particularly informative since both factors are being changed and you're just over-inflating your multiple comparisons (probably why you have so many Inf values). Instead do two sets, one that is pairwise~Temp|species and one pairwise~species|Temp. One will compare all temperatures within a species and the other compares all species within a temperature. I would advise doing this unless you are really interested in comparing all temperature levels of each species with all temperature levels of another species.

[MarioZuliani]

I have tried this and is now added to the rmd.