Well it has been a busy week on the Trans Mountain Expansion (TMX) Project file with the NEB initiating their review of tanker traffic. In all the news surrounding the plan to update the marine assessment one topic that caught my eye  was a series of news reports about the Southern Resident Killer Whales (SRKWs) featuring some really scary “facts” about acoustic threats to the SRKWs. I have written previously about the risks posed by the TMX to the SRKWs but only gave a minor consideration to the noise generated by the project. Upon hearing the news stories I feared that I had missed something. Essentially every media report had the same paragraph, which said:

Misty MacDuffee, a biologist with Rainforest Conservation Foundation on Vancouver Island, said additional noise from just six extra oil tankers each week would raise the possibility of extinction for the Southern resident killer whales to between 15 and 24 per cent. Right now, that risk is less than 10 per cent, she said.

Needless to say I found this level of precision to be very surprising and went looking to find where these numbers came from. The rest of this blog post will go into detail on what I found and will show how activists are selling a bill of goods to suit their faulty narrative. In the process I hope to explain, using the concept of “survivorship bias”, another reason why the TMX may be a positive for the SRKWs.

Let’s start with the basics, where did those numbers (the 15-25% chance of extinction) come from? It appears they came from an Raincoast Conservation Society Report titled: Report on Population Viability Analysis model investigations of threats to the Southern Resident Killer Whale population from Trans Mountain Expansion Project. Specifically Table 2 (page 34 of the .pdf file) from that report.

The two values represent the likelihood that the population numbers for the SRKWs will drop below 30 individuals (called “quasi-extinction” in the report) under the “Noise92.5” and “Noise100” scenarios. Not to be a pedant, but the report provides a likelihood of true extinction of the population as 0.3% and 1%, respectively under the two scenarios. That is a lot less than 15% and 24% cited in the media reports. It is funny how Ms. MacDuffee chose to use the term “extinction” even when the report clearly doesn’t use that term for these values.

So let’s look at the Noise92.5 and Noise100 scenarios. They are defined on page 33 as:

• Noise92.5 = Feeding time reduced because of boat presence 92.5% of the time
• Noise100 = Feeding time reduced because of boat presence 100% of the time

I hear you asking: what do they mean by feeding time reduced? Well in the report the authors identify a baseline scenario. In it they argue that due to the existing conditions in the Sea, the SRKW spend up to 85% of their foraging time in situations where noise from nearby boats interferes with their ability to hunt/communicate. Now here is the interesting part, the authors then decided to look at what it would be like if the noise was increased so that the SRKW encountered this level of noise for both 92.5% and 100% of the time. This is where the science simply disappears.

You see the authors don’t conduct any quantitative analyses to create these two numbers. The 100% is used as a worst-case scenario and the 92.% is used because it is half-way between 85% and 100%. So when Ms. MacDuffee uses that 92.5% value as the lower possible outcome in her media interviews she does so with zero basis in real-world conditions. As for the 100% value, that is even better. In the paper the authors make this statement:

With increased shipping traffic associated with the Project, Southern Residents could be around boats up to 100% of the time – according to the proponent tankers would be a “near continuous” presence. Thus, feeding could potentially be reduced by up to 19.5%.

That’s right they argue that six additional tankers a week (in a Strait that sees approximately 23,000 ship movements a year) will change the SRKWs’ exposure to noise from 85% to 100%. Yes you read that right, that is the entire basis of their argument. They don’t count the number of ships or look at population ranges, they just throw out that number and voila…a new “fact” has been created.

I asked the Raincoast Conservation Society about the source of their numbers and they referred me to a Nature Scientific Reports article (Evaluating anthropogenic threats to endangered killer whales to inform effective recovery plans) which is simply a  summary of their research output which was peer reviewed. The Nature Scientific Reports article, however, doesn’t even pretend to explain how they came across the 92.5% and 100% numbers, it just applies them. I’m guessing the peer reviewers didn’t buy the arguments they made in their initial paper so they just excluded the origin of the numbers for their article.

In a Twitter thread I pointed out that 100% is impossible since it involves tankers chasing whales throughout their range. The Raincoast Conservation Society responded that technically the tankers didn’t have to follow the SRKWs all the time, just in the designated critical habitat as defined by NOAA and the government of Canada as their summer foraging grounds. So when they say 100% of the time what they really mean is 100% of the time, during a narrow stretch in the summer. Except that is an entirely different story and not what was input into the model used in the report.

Why is my argument important? The first thing to notice is the overall habitat for the SRKW extends from California to Haida Gwaii. This information is available in the NOAA presentation: Distribution and Diet of Southern Resident Killer Whales.

Now remember the article only used summer foraging grounds. You may wonder why Raincoast was so adamant we only look at summer feeding grounds? It would be because of this:

and this:

You see the winter feeding grounds for J-pod are mostly in the northern Strait of Georgia while K- and L-pods travel along the Oregon and California coasts in winter. That means for essentially half of the year the SRKWs aren’t anywhere near where the TMX tankers routes. That really blows a hole in the whole 100% of the time narrative.

However, in the summer it is a different story.

In the summer they do indeed choose to forage near the tanker route, but this brings us to the second topic of our post: “survivorship bias”. Survivorship bias is:

a cognitive bias that occurs when someone tries to make a decision based on past successes, while ignoring past failures

The most famous example is from World War II (image source)

As described in Wikipedia:

During World War II, the statistician Abraham Wald took survivorship bias into his calculations when considering how to minimize bomber losses to enemy fire. Researchers from the Center for Naval Analyses had conducted a study of the damage done to aircraft that had returned from missions, and had recommended that armor be added to the areas that showed the most damage. Wald noted that the study only considered the aircraft that had survived their missions—the bombers that had been shot down were not present for the damage assessment. The holes in the returning aircraft, then, represented areas where a bomber could take damage and still return home safely. Wald proposed that the Navy reinforce areas where the returning aircraft were unscathed, since those were the areas that, if hit, would cause the plane to be lost.

So why is survivorship bias important on this topic? Let’s look again at that map of the SRKW summer foraging and notice where we do not see foraging. Notice how the south half of the Strait of Juan de Fuca all the way to the Puget Sound (the most active shipping route), is empty of sightings while the north half has lots of sighting. That should tell us something. Now look at Rosario Strait (labelled below for those who don’t know the regional geography).

Notice how there are no sightings in Rosario Strait while there are lots of sightings around San Juan Island. That should be raising a whole lot of red flags. If it doesn’t then this heat map makes it really obvious.

The SRKWs are apparently avoiding Rosario Strait and Anacortes as well as the US marine shipping routes. This is a pretty significant thing to recognize because fishers will tell you that salmon can be found in those waters. It should be prime habitat for the SRKWs but there were very few sightings in that area. What could be causing this effect?

Oh right, those red markings on the drawing are tankers supplying the US refineries with their crude. Now look again at that heat map. Notice that the whales are not avoiding the main shipping route along Haro Strait. That is because the wide main  channel and deep water attenuates the sound and so the whales are willing to use those waters for foraging. The whales are, however, avoiding the rich fishing grounds of the San Juan Islands, this is a very important thing to recognize if you are trying to help the endangered community recover.

I know that was a lot of pictures but it is pretty important to see. Taking survivorship bias into consideration we can recognize that the real challenge for the SRKWs appears to be US-bound oil tankers. These tankers, running along the south of the Strait of Juan de Fuca, through Rosario Strait and in the Anacortes, have essentially taken that entire section of the Salish Sea out of the foraging habitat for the SRKWs. That is a huge area of incredibly rich fishing grounds that the SRKWs are not using. Any project that reduced the number of large, loud tankers running in Rosario Strait and the Anacortes would be a potential bonanza for the SRKWs by re-opening that habitat for use.

Now I have to ask, can anyone remember what project is intended to supply the Anacortes and Puget Sound refineries with oil via a pipeline, thus reducing the number of tankers that have to make the Rosario Strait run? You are right, that would be the TMX.

Reading back over this blog post I can’t help but get angry. We have activists dominating the media narrative making claims that are completely unsupported by the science. That 15% increase in likelihood of extinction (but not really extinction) is based on a number completely unrelated to the number of ships moving in the area (the Noise92.5 scenario). The 25% increase in likelihood of extinction (but not really extinction)  is based on an assumption that that the six additional tankers a week will result in a massive increase in ship noise over the entire year, over the entire range of the SRKWs habitat when the area already is a busy shipping route and the SRKWs only use that area for foraging for the summer months and are still willing to use that area even with all the ship traffic.

It is unconscionable that activists can make these claims and have the media credulously repeating them in print. The numbers are essentially a fiction. The supporting references show that the numbers have no basis in quantitative science and to believe that the 15% and 25% numbers are anything other than a useful modelling exercise means completely ignoring the underlying science.

This is simply another case of activists making false narratives about a complex topic safe in the knowledge that few real scientists will recognize their falsehoods and even fewer will call them out on it. This is the advantage the activists have in this debate. They are not limited by things like the truth or provable facts. They can rely on truthiness and information that feels right and still be featured in the morning paper. Meanwhile a project that has the potential to re-open huge swathes of habitat for an endangered species gets short shrift because it simply doesn’t sound right…even though it is.