Going Viral on the Social Network: Disease spread in killer whale social networks
- homepage
- abstract
- presentation, powerpoint, 10MB
- see also the speaker notes in the presentation
- other material
- Michael N.Weiss, Daniel W.Franks, Kenneth C.Balcomb, David K.Ellifrit, Matthew J.Silk, Michael A.Cant, Darren P.Croft. Modelling cetacean morbillivirus outbreaks in an endangered killer whale population. Biological Conservation 242:108398, 2020. doi:10.1016/j.biocon.2019.108398
Hi Dan -- thanks for what feels like a very timely presentation!
ReplyDeleteThe takehome message appears to be: protect the cetaceans through better access to food. How do we do that? Reduce fishing; supplement their food sources; or what?
Trying to reply for thre 5th time (this website is very buggy).
DeleteHi Susan.
Yeah unfortunately we won't be able to save the population from a virus like this, so it all comes down to keeping the whales in good condition in the hope they would survive it.
There are a number of things that would need ot be done. Reducing fishing would be useful, and a key action would be to remove some of the dams that are blocking the salmon runs.
Hi Dan -- looking at slide 6, you've got the plot of population since ~1970 to present day. Do you know what caused the dips at ~1982 and ~1998? Is this just natural fluctuation in population size or are they disease epidemics?
ReplyDeleteHi, "unknown"-- Please remember to comment under your (academic) name -- if for some reason you can't, do please add your name into your comment.
Delete(FYI this is James Stovold, not sure why it's anonymised me!)
DeleteHi James,
DeleteThere wasn't a known virus outbreak and I think this period is part of stochastic fluctuations. However, we know that whale deaths are not independent. For example, our previous research has shown that if an older female matriarch dies, then her offspring/grandoffspring are at risk in the years following her death (she plays an important role in her family).
Hi Dan, really interesting work. I have little knowledge on this type of modelling however I do have a few questions! I appreciate there is a lot there, but perhaps you could point me in the direction of some reading material!
ReplyDeleteOn slide 20 you present 3 different networks, even though two of the networks are more "theoretical", what I got from it is that it's effectively showing different ways that whales *could* organise themselves. The data then shows that the mean outbreak size is minimum when the network is more clustered with the most clustered network being the actual way whale groups are networked.
1. Would it be correct in saying that whales organise themselves in a way that minimises the extent that disease can spread in their overall population?
2. Is there anything in the literature that suggests that whales may instinctively "know" that clustering the way they do benefits them in this way, say due to evolution?
3. If not, what are the main processes that determined how whales group the way they do, e.g. are they optimised for mating or resources?
4. Have different patterns of organisation been observed within their social structures when unexpected events like disease or killing occur? E.g. do they stick in their groups/disperse, or migrate? If so - how significant are the differences and is this considered within the models?
Best Wishes,
Marcin
Hi Marcin,
DeleteThanks for your questions.
On slide 20, one of the networks is the actual network, and the other two networks are null models based on different randomisations of the actual network. One of the randomisation preserves the weight structure (randomised) and the other assumes completely random mixing (mean field). You're right that the network is a little bit better than random when it comes to the spread of the virus (but all are bad). We detail in the referenced published paper that theoretical models often find that networks with community structure, like this, bottleneck the virus. We don't find that here, mainly because there are just so many weak connections between the communities that it still spreads well.
1. The social organisation is better than random, but it does not appear to have evolved/adapted to minimise disease spread, given that our results show that the disease spreads through the population most of the time with this network.
2 and 3. These networks are much more driven by family structure than by anything else. Whales can recognise their family groups and members. This appears to be optimise for social activities (parenting and social foraging). I have been involved in a number of published papers on these topics, if you're interested.
4. We don't have disease event data, but we know that the networks as a whole change based on the food availability in each year. In years where food is plentiful they form larger groups (lower cost of sharing food, etc) and smaller groups of close family members when food is scarce. We're interested in knowing what happens to the network structure when important individuals (e.g. the grandmothers) die, and we're currently analysing this.
cheers, Dan