If you may have further questions please explain,What is the exact nature of your research؟

I'm sorry I do not have any source available. If it is possible to send research results

Hi Maria,

I've got a couple of papers on spatial associations. One in O. hemionus sitkensis, and I cite several other species-specific studies within it. In my J. Mamm paper on parturition site selection in moose I do a deeper dive into the rest of the cervid lit; it may be beneficial to consider works outside Odocoileus spp. because there seem to be several cervid-wide commonalities. Space use and home range formation in deer species is a bit of an interest of mine, so feel free to drop me a line if you'd like to talk more about it.

Article Parturition site selection in moose (Alces alces): Evidence ...

Article Fine-scale social and spatial genetic structure in Sitka bla...

Thanks for the input Kevin!! 

I am excited about your results.

I've been intrigued about mines: genetic relatedness seems to NOT be a useful predictor of spatial association strength among ADULT mule deer. This was a bit of a surprise because many other papers have found a correlation (spatially proximate individuals are more genetically related).

However, when I look closer, the strength of the correlation between kinship and spatial separation or home range overlap varies depending on the set of deer considered in the analysis, and the scale used for measuring spatial distance. These studies have limited the analyses to pairs of deer in close proximity and high genetic relatedness to start with (i.e. they use pairs of deer captured within 1.5 km of each other, with known home range overlap, use only adult females and fawns but exclude juvenile males, etc). 

In my study, there were very few highly related adult individuals (6/982 pairs r > 0.4), therefore, we are unable to determine whether there would be an increased in association strength among close relatives. If I am interpreting this correctly... what do you think?

Then, I did another study using group membership to define associations (so, if seen in the same group, they are considered associated). Here, I considered all sex and age classes, not only adults (although the great majority are adults). The results are that genetic relatedness and sex are useful predictors of association strength.  I also run a test using physical contact rates (so, the amount of times they touch each other per minute), also using all sex and age classes. The results indicate that genetic relatedness is a useful predictor of contact rate.

So... it seems that genetic relatedness is important to determine contact rate, strength of association based on group membership, but not of strength of association based on spatial proximity. A lot to think about...  

I would love to hear your thoughts. 

Cheers,  

By spatial association, do you mean social association (existing in the same group at the same time) or general spatial proximity (existing in a similar place)? What exactly did you do for your analysis for for "association strength?" I can think of a number of different metrics that I could call "association strength" and some would be weaker than others. And there are some statistical approaches that make more sense than others, as well.

So, if I've done my math right, you have n = 32 adult deer (assuming the 982 is a typo for 992, because 992 would give a half matrix of 32 individuals)? What's this in context of your overall population density? If we sampled Toronto for 30 people, we'd get very few individuals that are related to each other by chance. However, in a village of 300, for the same average family size, obviously there's going to be more per-capita relatives. My first suspicion is that given a mere 6 individuals have high relatedness, your sampling density was low (unless mortality is high and you have a stationary or declining population density). However, this assumes you sampled your deer at random. If your sampling protocol involved exhaustively sampling all animals that had contact with a group of focal individuals, and you were comparing them to random non-contacted controls from the same population, then suddenly you're looking at a decent number. It all depends on your sampling design.

A word of caution on the contact rate stuff: 0 < r < 0.125 puts you in a foxy place. Since you're thinking about mechanistic action (the animals makes decision x based on characteristic y), you need to think how that factor is going to be assayed in nature. You can make some argument based on odour cues and imprinting for close relatives, but once you get past 2nd degree kindred you need to consider whether the animal might reliably distinguish anything further than a cousin. You might consider using "cousin level or greater" as a binary predictive factor. Also, I'm assuming that you've stripped known parent/offspring-at-heel pairs out, since it wouldn't be very surprising to see that mom pays a lot of attention to her still-young fawns. If you don't strip that, you run the risk that your effect is entirely does with fawns or yearlings at heel; while technically, that's relatedness predicting contact rate, it's a very specific type of relatedness.

The group membership thing is most interesting. There was a recent paper (Biosa et al 2015 in J. Zool.) that found that their spatial segregation had a temporal component to it, with opposite sex relatives negatively associated during breeding periods. Without knowing much more about your study and study system, I'd wonder if you considered what the deer are 'trying to do' life-history-wise the time of year you observed/sampled them.