RELEASE Test 3.SR male/female

questions concerning analysis/theory using program MARK

RELEASE Test 3.SR male/female

Postby constant survivor » Wed Mar 25, 2020 1:21 pm

Hi,
I have a dataset of bird mist netting. 31 Occasions (years). I grouped individuals according to sex.
Birds get colour marked and there are lot of transients in the dataset. Also the best models are exclusively TSM models among all species (it's a intra-specific analysis). Small Passerine species are the subject.

When I run RELEASE and look at the cumulative results of Test 3.SR for both groups it seems that females often have (highly) insignificant results while test results for males are (highly) significant (alpha 0,05). Or the other way around...)

I would have suggested that the test results for both sexes should be at least somewhat close to each other but for some species they are much different (in terms of p-value).

Is there an obvious explanation for this situation? I thought it may be due to differences in encounter probability or 'catchability' for both sexes and therefore smaller datasets or something like this...

I appreciate any suggestions on that if this is possible with the scarce information I gave on the dataset.
Thanks a lot and kind regards
Hannes
constant survivor
 
Posts: 36
Joined: Wed Dec 11, 2019 12:20 pm

Re: RELEASE Test 3.SR male/female

Postby cooch » Wed Mar 25, 2020 1:39 pm

constant survivor wrote: I would have suggested that the test results for both sexes should be at least somewhat close to each other but for some species they are much different (in terms of p-value).


Why would you have that expectation? Unless a species is monogamous at some level, there is no a priori reason why I'd expect males and females to 'be the same', in any number of ways.

More to the point, unless you've maded a mechanical error (unlikely -- RELEASE just...runs), the results are the results. Assuming they're not spurious artifacts of spare data (often the case for small bird studies), then your results are: males and females differ for some species.
cooch
 
Posts: 1628
Joined: Thu May 15, 2003 4:11 pm
Location: Cornell University

Re: RELEASE Test 3.SR male/female

Postby constant survivor » Fri Mar 27, 2020 9:46 am

Hi cooch,
thank you. I of course support the 'result' that males and females differ. I am glad they do.

But as to the specific question of Test 3.SR I could not think of any reason why females should be "less transient" than males but maybe this is my mistake of interpreting the test and it's implications wrong (I am pretty sure I do).
I mean I thought that the procedure of separating phi for TSM1 and TSM2+ should also be required for females when assuming that the proportion of transient and resident birds is similar between sexes. But I also see in the model results that sex is no significant factor for 1YSM estimates (excuse my insufficient description).

Well, at least the model results clearly indicate the better fit of TSM models so maybe I stop caring (because it exceeds the scope of my study).

I thought that some of the experienced biologists and MARK users here could give some common thesis on this subject.

Best
Hannes
constant survivor
 
Posts: 36
Joined: Wed Dec 11, 2019 12:20 pm

Re: RELEASE Test 3.SR male/female

Postby Jochen » Fri Sep 11, 2020 3:08 am

Hannes,
passerine females are usually less site-faithful than males. Within the same species, you may find more transient females than males. But you don't mention age at first capture, or if the study was done with local breeders or on migration, so it is hard to guess what is happening. Maybe using U-CARE instead of RELEASE makes interpretation easier.
Jochen
 
Posts: 26
Joined: Fri Nov 17, 2006 12:04 pm
Location: Germany

Re: RELEASE Test 3.SR male/female

Postby constant survivor » Fri Sep 11, 2020 6:38 am

Hi Jochen,
what a coincidence to meet you here :)

Thanks for reply. You may also recognized that I posted the calculation path once again in this forum.

Age at marking: second calendar year or above

The study is supposed to be a breeding bird survey but surely there are a lot of migrants caught in beginning of May - End of May. There are a lot of birds with only one '1' in capture history. Re-encounter rates vary between species within a range of 7-20% .

Best
Hannes
constant survivor
 
Posts: 36
Joined: Wed Dec 11, 2019 12:20 pm


Return to analysis help

Who is online

Users browsing this forum: No registered users and 14 guests