strata as fixed effects for global density estimation

21 views
Skip to first unread message

Silka Finkbeiner

unread,
Mar 26, 2006, 7:48:47 PM3/26/06
to distance...@googlegroups.com
List,

I am graduate student investigating the effects of three grazing systems =
on the density of various grassland bird species. I have an equal =
amount of line transect on each of the four replicates of my three =
treatments. Each pasture was visited nine times a year during each of =
the three years of the study. The strata and sample levels are defined =
as the grazing system and pasture, respectively. I am calculating =
density/year/species. =20

I am having what I hope is a simple problem with operating the program. =
For each year of our study, I would like to estimate the global density =
of species across grazing systems examined. To do this, I have 'global =
density to be estimated as the mean of stratum estimates', which are my =
grazing treatments, and 'weighted by the total effort in each stratum' =
in the 'Estimate' tab of the appropriate model definition properties =
box. I would like to treat the grazing systems as a fixed effect. To =
do this I need to have the 'Strata are replicates' box unchecked. For =
some reason, I can not remove the check mark. It seems as if for the =
settings I have tried, treating the strata as replicates (random effect) =
is the default. =20

I have created three model definitions; one at each level of resolution =
(global, strata, and sample) to estimate detection function. In each of =
the three models I always estimate global, stratum, and sample level =
densities; encounter rate and cluster size are always at the sample =
level. =20

I have tried to change old models from 'strata are replicates' to no =
replicates and by creating a new model with random effects for global =
density estimates, but I cannot get the combination I need. =20

I have referred to the program reference in the appendix (p177) of the =
online manual and I seem to understand that calculating global densities =
with strata as a fixed effect is possible, but I can't figure out why I =
can't do get it done. Also, for the models described above, I am using =
the conventional distance sampling engine. I am working with Distance =
4.1 Release 2. Thanks for any guidance!

Silka Kempema
sfink...@pie.midco.net
University of Nebraska-Lincoln

--
distance-sampling list home page:
http://www.jiscmail.ac.uk/lists/distance-sampling.html
To unsubscribe, go to the home page and click "Join or leave the list",
or email jisc...@jiscmail.ac.uk with the following message body:
leave distance-sampling
Before posting to the list, read the list guidelines at
http://www.jiscmail.ac.uk/files/distance-sampling/
distance-sampling.welcome
Any questions, email the list owner at
distance-sam...@jiscmail.ac.uk
--

Distance Development

unread,
Mar 31, 2006, 4:22:54 AM3/31/06
to distance...@googlegroups.com
Hi Silka, list,

I was able to reproduce the problem you report in Distance 4.1 release 2, but
not in Distance 5.0 Beta 5 (the latest beta). So, I guess it's an example of
a problem I've fixed without even knowing it was there! It only occurs in
D4.1 when you estimate density by sample, stratum and globally, and I wasn't
clear why you need density by sample, so one option is to turn that off. The
other is to upgrade to D5.0.

Cheers,

- Len


--
Len Thomas, Co-ordinator, Distance Development Project
Research Unit for Wildlife Population Assessment
Centre for Research into Environmental and Ecological Modelling
The Observatory, University of St Andrews, Scotland KY16 9LZ

email: dist...@mcs.st-and.ac.uk
Distance home page: http://www.ruwpa.st-and.ac.uk/distance/
Reply all
Reply to author
Forward
0 new messages