Use the hurdle() function in pscl to repeat the analysis of the beetle egg example from Section 4.4 with a hurdle model. Compare the results to those from the zeroinflated Poisson model. Are there any differences that lead you to prefer one model over the other? In terms of the beetles, what is the difference in assumptions between the two models regarding the source of zero counts? (Note that the coefficients in the probability model estimate P(Y > 0|z), but predict(..., type = "prob") returns P(Y = 0|z).)