[Babase] Finding intergroup interactions
Susan Alberts
alberts at duke.edu
Mon Apr 19 15:16:45 EDT 2010
Thanks Niki, see below.
On Apr 19, 2010, at 2:10 PM, Niki Learn wrote:
> I gave up trying to query it for now. I just went through the list
> of 358
> mismatched agonisms and declared them all mm, mf, fm, or ff
> interactions.
> It was faster. I might be slightly off in my counts if I made a
> booboo or
> two but I checked genders of individuals I wasn't sure of against
> biograph
> so it should be accurate enough for this purpose. I got 25 female-
> female
> interactions (regardless of maturity) where the two females were in
> different groups, 16 where a female won over a male that was
> supposed to be
> in another group, 157 where a male won over a female listed as being
> in a
> different group, and 160 where it was a male-male interaction with
> males
> from two different groups (and we know 98 of these involve adult
> males only
> so 62 involve at least one juvenile or subadult male). So 25 vs 160 -
> pretty big difference.
Yes indeed. Plus the 157+16 male-female or female-male, also a puzzle
>
> What if anything do you want me to investigate further? I would not
> be
> surprised if the female-female interactions are more likely to have
> a demog
> note or other groups note about an interaction between the two
> groups than
> the male-male seem to be. I could check a handful at random to see?
Yes I think this makes sense. I am guessing that virtually all of
these are errors but it will instructive to find out.
Just as a note, I think that if the animal is marked absent on a given
day and still occurs in agonism and grooming records and there is no
demog note, it's tough to call whether the census is in error or the
ad lib data. It will depend partly on what the context is. Someone
gone for only a day, that might be a census error. Someone gone for
weeks or months, less likely.
>
> Curiously three of the female to female interactions are Viola's
> females
> picking on Edith, who was a white infant in Omo's group (actually,
> the only
> white infant we've seen in Omo's group). The three interactions are
> all on
> different dates. ??? That's the only thing I noticed resembling a
> pattern
> in the female-female data.
It seems very likely that they were confusing her with a female infant
in Viola's. Yum and Vava are likely, as they are female white infant
in VIO at about the same time and same age. But let's see what your
search turns up.
Susan
>
> -----Original Message-----
> From: babase-bounces at eeblistserv.Princeton.EDU
> [mailto:babase-bounces at eeblistserv.Princeton.EDU] On Behalf Of Niki
> Learn
> Sent: Monday, April 19, 2010 12:14 PM
> To: 'The Baboon Database Project'
> Subject: RE: [Babase] Finding intergroup interactions
>
> A clarification (sorry, did this a couple weeks ago):
>
> There are 358 total agonism mismatches for 2000 - 2008a (male-male,
> male-female, and female-female, regardless of age). That is 0.12%
> of the
> 300k agonisms for that period. 98 of the 358 are adult male to
> adult male
> (this is the 0.89% out of just shy of 11k adult male to adult male
> interactions). It sounds to me from those numbers that this problem
> is
> skewed toward males. A preliminary look at the list of 358 suggests
> that
> most of the other cases are either male-male where at least one male
> was not
> an adult and male-female interactions. I am working on a query that
> will
> list how those that are actually female-female.
>
> I know that some of the adult male to adult male cases I looked at
> included
> multiple interactions for the mismatched male, often including
> interactions
> with females. I would guess that the same males from the 98 cases
> for adult
> male to adult male account for a large portion of the adult male to
> female
> interactions. There are also a small number in which the female wins,
> including a very small number in which she wins over an adult male -
> these
> may be cases where she defeated a visiting male who was not listed
> in the
> census (but possibly there are demog or other groups notes to explain
> these).
>
> On the bright side, it sounds like there are a lot fewer of these
> mismatches
> overall in recent data (0.12%) than in older data, given the 0.76%
> Karl
> found below for all years.
>
> -----Original Message-----
> From: babase-bounces at eeblistserv.Princeton.EDU
> [mailto:babase-bounces at eeblistserv.Princeton.EDU] On Behalf Of Karl
> O. Pinc
> Sent: Monday, April 19, 2010 11:49 AM
> To: The Baboon Database Project
> Subject: Re: [Babase] Finding intergroup interactions
>
> FYI
>
> I get an overall not-in-the-same-supergroup to
> total interaction ratio of 5689 / 744082,
> or about 0.76%.
>
> On 04/19/2010 10:31:01 AM, Karl O. Pinc wrote:
>>
>> SELECT interact.iid, interact.act, interact.date
>> , actor.sname, supergroup(actorgrp.gid, interact.date) as
>> actorsupergroup
>> , actee.sname, supergroup(acteegrp.gid, interact.date) as
>> acteesupergroup
>> , 'Warning: Interactions should be between individuals in the
>> same
>> supergroup'
>> FROM interact, parts AS actor, parts AS actee
>> , members AS acteembr, members AS actormbr
>> , groups AS acteegrp, groups AS actorgrp
>> WHERE actor.iid = interact.iid
>> AND actor.role = 'R'
>> AND actormbr.date = interact.date
>> AND actormbr.sname = actor.sname
>> AND actorgrp.gid = actormbr.grp
>> AND actee.iid = interact.iid
>> AND actee.role = 'E'
>> AND acteembr.date = interact.date
>> AND acteembr.sname = actee.sname
>> AND acteegrp.gid = acteembr.grp
>> AND supergroup(actorgrp.gid, interact.date)
>> != supergroup(acteegrp.gid, interact.date)
>> ORDER BY interact.date, interact.act, interact.iid, actor.sname,
>> actee.sname;
>
>
>
>
> Karl <kop at meme.com>
> Free Software: "You don't pay back, you pay forward."
> -- Robert A. Heinlein
>
>
> _______________________________________________
> Babase mailing list
> Babase at www.eco.princeton.edu
> http://www.eco.princeton.edu/mailman/listinfo/babase
>
> _______________________________________________
> Babase mailing list
> Babase at www.eco.princeton.edu
> http://www.eco.princeton.edu/mailman/listinfo/babase
>
> _______________________________________________
> Babase mailing list
> Babase at www.eco.princeton.edu
> http://www.eco.princeton.edu/mailman/listinfo/babase
--------------------------------------------------------
Susan Alberts, Dept of Biology, Duke University, Box 90338, Durham NC
27708, 919-660-7272 (Ph), 919-660-7293 (Fax)
More information about the Babase
mailing list