Tuesday, July 17, 2012

Baited Shark Dives and Shark Attacks?

Outright Shark feeding - even worse than chumming?

Oh well.
So much for being too busy to blog.

There has been a Shark strike in Western Australia.
From what I can discern, it very much appears to have been a predatory attack (Christopher???) and it comes right on the heels of Mr. Moore's decision to preemptively ban cage diving there. The apparent reason for that politically motivated stupidity was that South Australia had restricted cage diving there because Barry Bruce et al had discerned that berleying (= chumming) had caused a minimal change in GW distribution.
Nothing whatsoever to do with Shark attacks - but who gives a shit provided that one can engage in populist grandstanding, right? And this? Indeed: Neanderthal reactions!

Which brings me straight to the following.
I rather despise those couch farting Shark lists - but people do keep me in the loop about the rare noteworthy discussions; and so I got sent this remarkable, and provided that you got a sense of humor, incredibly witty post by the Grand Mufti of Shark behavior, non other than the forever unequaled Doc.
I hope he doesn't mind me re-posting it verbatim in all of its glory.

A key criticism of the ruling in Western Australia is that researchers were not able to show a causal link between shark attack and chumming at shark dives. Lets see how they might prove such a link using the scientific method. How about setting up two dive sites within say 5 km of each other. At one you use chum and feed the sharks from a feeding tube or box. At the other you attract sharks by low frequency sound attraction and again feed shark from some device that precludes odorant from getting to the test-beach 5 km distant.

Now if there are sharks around I wager they will very quickly be attracted to the baiting station due to currents which rapidly brings the olfactants to their nostrils---so you might get them trained up in just a few days. For the acoustic attractions the sharks have to be within about 500 m of the hydrophone and be fairly motivated at the time. This might take several months but eventually the sharks at both sites will be feeding regularly, and reliably but probably seasonally.

To be further certain about your experimental study population you would have to tag the sharks and possibly take genetic samples to be sure you are working with different animals at the two sites...ok it could be done. Thus you have established two relatively equal feeding sites with a control and a treatment trial....good!

So after about two years of preparations you can now set up an hypothesis: H0, the null hypothesis is that there is no statistical difference between the control and treatment groups with respect to shark attacks; H1 is the proof of the pudding---There is a statistical difference between the two.

Ahhh, but here's the rub: You would of course have to continue your trials until you collect and adequate sample size for your appropriate statistical tests that you would used to either accept or reject H1 or H0. Based only on the annual rate of shark attacks world wide, this will take on the order of 1,000 years depending on the seasonality of the sharks. Considering the paucity of attacks and the time required for a human generation, you should put in your will that for the next 50 generations your offspring must continue the experiment on a daily basis until each season ends, then start again next year when the season begins. Get my drift? So it goes for all such shark-attack research.

OK...I put it up there now you shoot it down. I personally would love to understand the motivation of why sharks bite naked apes. (see Gruber, S.H. 1988. Why do sharks attack humans? Naval Research Review 40(1):2-19.)

Bingo! :)
As I've said all along, those musings about Shark attacks amount to nothing more than conjecture, i.e. sometimes plausible but ultimately untestable hypotheses.

But is the whole story?
Is there really no evidence whatsoever to point us one way or the other?

I believe that in fact, there is.
Yes George the great doyen of Sesselfurzing should indeed finally creep out from behind his desk and have a reality check by doing a couple of Shark dives; but in this specific specific case, he may even be useful by staying right where he is, smack in the middle of his Sammelsurium of Shark attack data - and ditto for Collier!

What I'm suggesting is this.
Get a Masters student to mine those data and to analyze whether globally, there is any statistically relevant correlation between the location of Shark strikes and that of baited Shark dives!
Of course correlation, even if it were found, is not causation. And, this would not be testing the hypothesis that baiting leads to more Shark attacks. For that, one would have to proceed like Doc suggests and wait for several centuries. But it would at least give us an indication about whether the hypothesis is at least plausible and thus, worth testing!

Betcha the result it is a big fat zero!
And if so, wouldn't that be all the more surprising as common wisdom would suggest the exact contrary! Here's the long version of why this should be the case - but if you can't be bothered, here's the synopsis.

For there to be a Shark strike, a person and a Shark need to be in the same place at the same time (dooh...) and consequently, anything increasing the rate of such encounters increases the probability for such a strike.
When it comes to baited Shark dives, it thus follows that, equally trivially
  • the more Sharks are present, the higher the risk of a Shark bite
  • the more divers are present, the higher the risk of a Shark bite
  • the more dives are being conducted, the higher the risk of a Shark bite
  • the closer the divers and the Sharks interact, and the more frequently those interactions occur, the higher the risk of a Shark bite - meaning that protocols are incredibly important!
And then there's also the element of location, i.e. where those baited Shark dives are being conducted.
They are obviously being conducted where there are Sharks (dooh...) - but often, they are even located where Sharks aggregate naturally to mate and feed, e.g. around Dyer and Seal Islands in South Africa, the Neptunes in Southern Australia, Guadalupe, etc., meaning that those places are more dangerous to start with, something one may want to consider when analyzing the data.

But then again, maybe the above is a fallacy.
Maybe Shark dives habituate the Sharks to the presence of people, meaning that they become more relaxed or "tame" and thus refrain from biting. Let's not forget that agonistic (= essentially, defensive) behavior may well be one of the principal causes of Shark bites on divers! And if we put too many divers in the water, the Sharks may well decide to simply go away.
Yes as always it's complicated!

So how about it.
Will those guys finally let somebody analyze those data, and will those stupid assertions be finally outed for what they really are, nothing but rubbish and completely unsupported and probably even falsified by the very data collected by the people proffering them!

PS Patric here (thanks!) and Jillian here!
PS great comment on the attacks here!


OfficetoOcean said...

Five shark attacks in ten months, all fatal, three victims consumed and sharks protecting their kill from rescuers, returning at least once for another go in all of the attacks and the response is to pre-emptively ban something which isn't even being done and which hasn't even been suggested as far as I'm aware, that and try to do what fails about 95% of the time and "catch the killer shark."

Would it be mischievous to once again suggest that not having shark tourism there certainly doesn't seem to be making the beaches any safer and to perhaps give it a go to keep the sharks away from the beaches? Of course, that's stupid but then so is 99% of the stuff said by people in regards to this subject.

Thankfully you can see through that and have once again written one of the few pieces worth reading.

The study of shark attacks can yield positive results and has done so in the past but what is often overlooked is the causal factors emanating from the victim and human impact in the geographical location of the attack including on land. The simple fact is that people don't want to hear the truth, particularly those in industries which affect the risk of shark attack (The abbatoir in Recife, canning factories and fish processing plants in Australia, the whaling industry in South Africa...) and so turn on the one industry it's easy to scapegoat, shark tourism.

As is often the case in Australia in particular, shark attacks are often preceded by warning signs, dogs go missing, fish are snapped from lines and most importantly, sharks are sighted. This shark in the most recent attack had been hanging around for days, long enough to get a nickname and STILL surfers took the risk, as is their want. It's a tragedy but the shark was just behaving perfectly naturally.

It's like the age old religious debate, the burden of proof is not on the athiest but the believer, there is not a single jot of evidence connecting shark dives with shark attacks, just shouting it loudly doesn't make it so.

OfficetoOcean said...

Incidentally, in the interests of placating the pedants out there, my referebce to the whaling industry in SA is retrospective, based on historical events there surrounding the spate of attacks around KZN, particularly Amanzimtoti...

DaShark said...

Ya know David, these events are so incredibly rare that making any generalist statements about their causes is always going to be conjecture pure and simple because we just simply lack the required sample size.

The fact is, there are Sharks in the ocean and very rarely, one of them will attack a person.
That's just how it is and anybody entering the water should know that he incurs that minuscule risk - and that's really all we can say.

One is still orders of magnitude more likely to drown, or to be involved in a fatal car crash whilst driving to the beach.
This does not detract from the individual tragedy of these isolated events - but it is what it is and short of staying out of the water, there is really nothing we can do to prevent the same happening in the future.

Cristina said...

DaShark, if we take in consideration your points:

the more Sharks are present, the higher the risk of a Shark bite
the more divers are present, the higher the risk of a Shark bite
the more dives are being conducted, the higher the risk of a Shark bite
the closer the divers and the Sharks interact, and the more frequently those interactions occur, the higher the risk of a Shark bite - meaning that protocols are incredibly important!

and compare them with locations around the world where baited shark dives are conducted in all shapes and forms,(good protocols and sometime bad protocols included) we could gather and provide enough data to prove that in general feeding and attacks are NOT related. I do understand that the death of a young man is indeed very sad but the reaction of some as always proves also the death of common sense, something we failed to protect long time ago. Thank you for your blog and for posting Doc comments.

DaShark said...

but the reaction of some as always proves also the death of common sense, something we failed to protect long time ago

Amen to that Cristina!