organization of gdistsamp data for migration

54 views
Skip to first unread message

Auriel Fournier

unread,
Nov 4, 2015, 12:30:24 PM11/4/15
to unmarked
Hello Everyone,

I have data broken down into four primary periods and two secondary periods. 

My data is collected over fall migration and so I most certainly cannot assume closure between primary periods.The secondary periods are done on the same night by two different observers so I think closure is safer there. 

What I have done is arranged my data such that each primary period is unique site 

for example - 

      site      pp      X.0.1.1 X.1.2.1 X.2.3.1 X.3.4.1 X.4.5.1 X.0.1.2 X.1.2.2 X.2.3.2 X.3.4.2 X.4.5.2 X.0.1.3 X.1.2.3 X.2.3.3 X.3.4.3 X.4.5.3
     os21     3       0       0       0       0       0       9       2       3       1       0       0       0       0       0       0
     os25     3       0       0       0       0       0      12       8       1       1       1      11       0       2       3       2
     pool2    3       0       0       0       0       0       0       0       0       0       0      20       3       5       2       2
     os21     4      51      22      11       7     3       1       0       0       0       0       2       1       1       1       1
     os25     4       0       0       0       0       0       9       1       3       1       0      17       6       4       0       1
     pool2    4       9       5       2       1       0      12       7       8       2       5       0       0       0       0       0

pp=primary period

I've got two questions

1) Is if this is a valid way to utilize this data to obtain abundance estimates based on covariates collected during each primary period?

2) Should I be using primary period as a categorical variable in lambda? 

Thanks,

Auriel

Dan Linden

unread,
Nov 4, 2015, 10:57:37 PM11/4/15
to unmarked
Hi Auriel, how many sites do you have?  There are other examples/discussions here of folks "stacking" replicates of sites into new sites for various unmarked models.  There are potential issues with the approach (psuedoreplication) but it can sometimes also be reasonable.  It's important to understand how your inferences are affected. 

What kind of covariates are associated with the primary periods?  If they are common to all sites and vary only across periods, you are pretty limited with what you can specify and would not be able to include primary period as a factor in addition to a period-specific covariate.  But if they vary across both sites and primary periods, you have more flexibility.

Auriel Fournier

unread,
Nov 5, 2015, 8:40:05 AM11/5/15
to unmarked
I have 31 sites, each are surveyed in each primary period. I wondered if there would be pseudoreplication issues, but I wasn't sure. 

The covariates are region of the state (four regions, between 7 and 8 sites per region), percent cover of three different vegetation types and average water depth. Average water depth varies with primary period, but percent cover and region do not. 

Auriel

Dan Linden

unread,
Nov 5, 2015, 1:08:33 PM11/5/15
to unmarked
Psuedoreplication is controversial but the more you understand it, the better.  Unfortunately too many people see it as a red flag that completely invalidates statistical inferences which is not necessarily true.  But it can complicate things.

I assume average water depth also varies by site?  There's no harm in fitting a few models to see what happens.
Reply all
Reply to author
Forward
0 new messages