NaNs produced for SE, z and P(>|z|) in occupancy model

[Joshua Jones]

I'm trying to create an occupancy model for the beetle family Curculionidae. My data is spread across 378 sites and I have 795 recorded presences. The mean number of observations per site is 3.02 and 357 sites have at least one detection. I've run a simple model with the following code:

 occu(formula = ~days # detection formula

                                    ~acd,  # occupancy formula

                 data = umf,

                 se = TRUE,

                 engine = "C")

When I go to get the summary I get the following warning message:

Warning message:

In sqrt(diag(vcov(obj, fixedOnly = fixedOnly))) : NaNs produced

And the summary output looks as follows:

Call:

occu(formula = ~days ~ acd, data = umf, se = TRUE, engine = "C")

Occupancy:

            Estimate  SE   z P(>|z|)

(Intercept)   0.0551 NaN NaN     NaN

acd           5.5144 NaN NaN     NaN

Detection:

            Estimate    SE      z P(>|z|)

(Intercept)    1.154 0.722  1.598   0.110

days          -0.107 0.237 -0.452   0.651

AIC: 1410.39 

I'm pretty new to occupancy modelling so would anyone be able to tell me why I am getting NaNs?