Calculating the SE for N in a POPAN Model in RMark

[fishyspotz]

Hey everyone,
I created a bunch of POPAN models to calculate the superpopulation, but my superpopulation does not have an UCL or LCL. Here is my code for my best fit model:

Code: Select all

Phi.time <- list(formula=~time, link = "logit")
p.encounters <- list(formula=~encounters, link = "logit")
pent.dot <- list(formula=~1, link = "mlogit")
N.dot = list(formula =  ~ 1, link = "log")


gsb.phi.time.p.encounters.pent.dot = mark(gsb.js.proc,
                              gsb.js.ddl,model.parameters=list(Phi= Phi.time, p= p.encounters, pent= pent.dot, N = N.dot))

gsb.phi.time.p.encounters.pent.dot


# The popan.derived function of RMark estimates N
# (plus estimates SE and 95% CI using the delta method)
gsb.phi.time.p.encounters.pent.dot_popan <- popan.derived(gsb.js.proc,
                                      gsb.phi.time.p.encounters.pent.dot)$N

# Look at results
gsb.phi.time.p.encounters.pent.dot_popan


Output summary for POPAN model
Name : Phi(~time)p(~encounters)pent(~1)N(~1)

Npar : 11 (unadjusted=10)
-2lnL: 494.9219
AICc : 517.6203 (unadjusted=515.50236)

Beta

Real Parameter Phi
2015 2016 2017 2018 2019 2020 2021
0.6360242 0.9074715 0.9080846 0.7479037 0.6459913 1 0.6710513


Real Parameter p
2015 2016 2017 2018 2019 2020 2021 2022
0.0633321 0.059522 0.0744398 0.0866578 0.1629771 0.1893176 0.0934253 0.0893099


Real Parameter pent
2016 2017 2018 2019 2020 2021 2022
0.1073255 0.1073255 0.1073255 0.1073255 0.1073255 0.1073255 0.1073255


Real Parameter N
2015
1219.199

estimate se LCL UCL
Phi:(Intercept) 0.5581494 0.3537396 -0.1351801 1.2514789
Phi:time2016 1.7249966 2.7063601 -3.5794693 7.0294625
Phi:time2017 1.7323196 2.4924903 -3.1529616 6.6176007
Phi:time2018 0.5293134 0.5439685 -0.5368648 1.5954916
Phi:time2019 0.0433152 0.6121947 -1.1565865 1.2432169
Phi:time2020 24.2063360 0.0000000 24.2063360 24.2063360
Phi:time2021 0.1547943 0.7682975 -1.3510688 1.6606573
p:(Intercept) -3.0244657 0.2554491 -3.5251459 -2.5237856
p:encounters 0.0082632 0.0016942 0.0049426 0.0115839
pent:(Intercept)-0.8404664 0.3474001 -1.5213707 -0.1595622
N:(Intercept) 6.7936888 0.1632315 6.4737550 7.1136226


Occasion N se LCL UCL
1 303.2407 92.59751 121.7495 484.7318
2 323.7195 65.96147 194.4351 453.0040
3 424.6174 94.80030 238.8088 610.4260
4 516.4397 104.84027 310.9528 721.9266
5 517.0983 108.86410 303.7246 730.4719
6 464.8921 103.18220 262.6550 667.1292
7 595.7433 107.41030 385.2191 806.2675
8 530.6254 129.50149 276.8025 784.4484


So it looks like my superpopulation is 1219 if I am not mistaken. Is there a way I can calculate the UCL and LCL like what I have for my abundance estimates for each capture period?

Thanks in advance!

[jlaake]

Use summary with argument "se" set to TRUE. Below is an example

Code: Select all

library(RMark)
data(dipper)
model=mark(dipper,model="POPAN")
summary(model,se=TRUE)