marginally significantp = 0.060
The results showed significant (except for “bite” data for wild‐caught individuals) or marginally significant (“bite” data for wild‐caught individuals: p = 0.060) overdispersion, and a negative binomial GLM was built and tested the model using the glm.nb function in the R package MASS (Venables and Ripley 2002 ).