Cooperation In Free-ranging Dog Packs

[corvus]

I found this study quite interesting:

Pattern of individual participation and cheating in conflicts between groups of free-ranging dogs

Roberto Bonanni, Paola Valsecchi, Eugenia Natoli

Animal Behaviour 79 (2010) 957–968

Cooperative intergroup aggression provides an example of a costly cooperative behaviour whose benefits spill over to noncooperative animals as well. Consequently, investigating factors that promote individual participation in intergroup contests should prove useful for understanding how cooperation may persist in animal societies despite cheating. Here, we examined variables affecting individual participation in naturally occurring conflicts between groups of free-ranging dogs, Canis lupus familiaris. The overall proportion of cooperating group members decreased significantly with an increasing number of group members present. In one pack, the individual probability of active participation decreased significantly when this pack had a numerical advantage over opponents. Dogs belonging to the smallest pack tended to be more cooperative than those belonging to larger groups. Social prestige (measured as the number of submissions received during greeting) did not appear to be a consequence of cooperative behaviour. Individual participation increased with an increasing number of affiliative partners. Young and high-ranking dogs tended to cooperate more when their group was outnumbered by opponents but did not stay at the front of the pack during conflicts. These results emphasize the greater opportunity for

cheating in larger groups and the complexity of dogs’ behaviour. Cooperation appears to be conditional on both the ‘adversity of the environment’ (as measured by relative group size) and the identity/behaviour of companions.

The most interesting finding to me was that dogs that received (or gave) a lot of affiliative attention such as muzzle licking and nosing were more likely to cooperate in any situation, but this was independent of social status. Hmmmmm

ETA journal details.

Edited March 29, 2010 by corvus

[Aidan]

Glad you could understand it!

[corvus]

Did you get hold of the full paper, Aidan? That abstract isn't the clearest I've ever seen... The general gist of it all is that there is no simple reason why dogs cooperate with one another in conflicts with other packs. In bigger packs it seems like individuals are less likely to get involved in aggressive encounters when the other group is much smaller than theirs, maybe because the extra numbers aren't needed. Smaller packs where lots of social bonding behaviour goes on have individuals that are more likely to cooperate in any aggressive encounter with another pack. That may be because they are related, or maybe because all that bonding makes them more cohesive as a group. Dogs were given social status ranks, which sometimes predicted where a dog might be in an aggressive encounter (leading, participating near the front) but sometimes seemed a lot less important. I think it suggests that there's more to dog sociality than status or resource holding potential, and more to it than learning theory.

[Aidan]

The broader field of psychology has never considered behaviourism to hold all the answers, but I'm not sure how this paper accounts for anything more than the fact that "dogs tend to do what works for them"? Which in itself is a very good thing, and I'm glad that scientists are continuing to be objective about the myths that "pop dog psychology" perpetuates.

Probably if you were to analyse "patterns of individual participation and cheating in ruck and maul contests between groups of field-contained rugby teams" you would see striking parallels ;-)

[corvus]

I don't think it does account for anything at all, frankly! But that's what I find interesting. IME it's often hard to find patterns in animal behaviour. Lots of "noise", and coming from a wildlife perspective, it was usually impossible to know what the animals had learnt in their life. Even a little bit. I reckon you could model when dogs are likely to form quite cohesive groups and when they are not based on access to resources and the personalities of the dogs. Maybe that's a little ambitious, though. Perhaps I'll start with Rugby instead. :D

In seriousness, though, if you only do what works for you, why all the variation within the one environment? I'm not saying that there should be just one evolutionarily stable way for dogs to live in social groups, but I'm wondering what drives the variation and if it is predictable. I think that dogs are wonderfully adaptable when it comes to being social, and trying to come up with simple rules about how or why they group is probably going to meet with limited success because it's not simple. I think that this paper hints at the wonderful complexity of sociality in dogs. Or maybe I've got stars in my eyes at the moment.

[Aidan]

I reckon you could model when dogs are likely to form quite cohesive groups and when they are not based on access to resources and the personalities of the dogs. Maybe that's a little ambitious, though.

If you had time and money you could probably do a meta-analysis based on the exisiting data and start to form some conclusions. e.g stray dogs in cities as opposed to stray dogs in rural settings (it would appear that city dogs are less likely to "pack" than rural dogs, could this be because scavenging from bins doesn't require a team effort and running dog wallabies does?)

Perhaps I'll start with Rugby instead.

Not as silly as it sounds! Lots of readily available data and nicely contained variables (except they keep changing the blasted rules).

if you only do what works for you, why all the variation within the one environment? I'm not saying that there should be just one evolutionarily stable way for dogs to live in social groups, but I'm wondering what drives the variation and if it is predictable.

My argument (allowing me to take a position here) is that "doing what works" accounts for all the variability quite happily. Using the rugby analogy, you only throw in as many players as you need to secure your position or position yourself well for the next phase. Every player plays for their position because that is what works best; if you're fat you play in the forwards, if you're skinny and take care of yourself you play in the backs.

[corvus]

if you only do what works for you, why all the variation within the one environment? I'm not saying that there should be just one evolutionarily stable way for dogs to live in social groups, but I'm wondering what drives the variation and if it is predictable.

My argument (allowing me to take a position here) is that "doing what works" accounts for all the variability quite happily. Using the rugby analogy, you only throw in as many players as you need to secure your position or position yourself well for the next phase. Every player plays for their position because that is what works best; if you're fat you play in the forwards, if you're skinny and take care of yourself you play in the backs.

Okay, makes sense.... What about the dogs that were equally likely to participate in an agonistic interaction with another group regardless of whether their group outnumbered the other group or not?

[Aidan]

Okay, makes sense.... What about the dogs that were equally likely to participate in an agonistic interaction with another group regardless of whether their group outnumbered the other group or not?

They were the "forwards", just doing their job.

[corvus]

I get the sense this is boring to everyone else. Might continue via PM.

[Are You Serious Jo]

Occam's razor - plasticity- solved :)