# I have performed three likelihood ratio tests on a factor in R. Do i need to adjust the p values?

I am a student doing a thesis for zoology. I am looking at how factors like season (winter/summer), fencing (inside fence/outside fence), and the interaction between season and fencing affects invertebrate abundance. I have included moisture as a random effect in a GLMM analysis.

I am doing likelihood tests on my GLMM to see if these two factors and the interaction have a significant affect on invertebrate abundance using anova.

e.g., to find effect of season:

x <- glmer(amphipoda ~ fence + season + (1 | moisture), family = poisson(),
data = data)

y <- glmer(amphipoda ~ fence + (1 | moisture), family = poisson(),
data = data)

anova(x, y)


I do the same thing when I want to find the effect of fence and the effect of interaction. This means that I am doing three likelihood ratio tests on one thing (eg: amphipoda). Does this mean that I have to change the P value confidence of each of these tests from 0.05 to something lower using something like the Bonferroni method? From searching online I have also found something called Harmonic P value, but I cannot figure out how to find this value because there are many things about statistics that I do not understand.