wild salmon dissonance and patchwork quilts for fisheries managers

Dissonance, the Free Online Dictionary suggests means:

Wild Salmon Policy?

1. A harsh, disagreeable combination of sounds; discord.

2. Lack of agreement, consistency, or harmony; conflict.

Under the first definition I was privy to some serious dissonance this past week while sitting in salmon-related  meetings listening to a senior Department of Fisheries and Oceans manager run circles around questions. At one point, he was even asked for a “yes” or “no” answer; yet a long-winded response would start winding up…

“No, no… I said YES or NO, please” said the questioner again.

“Well… yeah… but… you need to understand…” said the bureaucrat.

One of the most offensive ways to answer a question or begin any explanation is: “you have to understand…”

_ _ _ _

Part of the meeting involved discussions around Fraser sockeye and predicted returns. Of course, this type of discussion involves the computer simulation model Fraser River Spawning Sockeye Initiative (FRSSI) which apparently pumps out fishing plans, or “TAM rules” (total allowable mortality). This model is great for theoretical understandings of biological organisms — like Fraser sockeye — but an absolute joke for “managing” biological organisms in the wild — like Fraser sockeye.

Worse yet (and you can read more on previous posts on this site) the Fraser River is suggested to have approximately 200 distinct sockeye populations or stocks; and approximately 150 different spawning areas, and countless nursery lakes. The FRSSI is based on information on 19 stocks or populations, much of that information spotty at best, and only about 50 years of data which means approximately 12-15 life cycles of sockeye. Those 19 stocks are further simplified into four groups based on the timing of their upstream migration and spawning.

The model has various productivity scenarios fed in (again spotty estimates); with the added benefit that DFO only studies two sockeye rearing lakes (yeah that’s 2).

Worse yet, the FRSSI model does not incorporate any data prior to about 1948 when DFO started keeping records.

Worse yet, the model has no ecological values factored in — things like seals, bears, eagles, and so on. It simply sets fishing rates and guesstimates how many sockeye need to reach spawning grounds.

It’s supposed to be “Pilot Study” and is listed as such in DFO promotional material… yet, it’s being used to “manage” sockeye on the Fraser, even though there hasn’t been a commercial fishery in three years (that will probably change this year…).

However, I suppose one positive out of this is that the maximum fishing rates (maximum sustained yield – MSY) is now 60% of total run size, not 80% as it has been in the past….

Salmon think tank... salmon fisheries in the tank...

_ _ _ _

The second definition of dissonance occurred in the same meetings last week when DFO reps started presenting information on Fraser River Chinook: pre-season forecasts, proposed exploitation (a.k.a. fishing) rates, and various forecasted population scenarios with proposed fishing rates:

The 5 represents the age of these fish and the small subscript 2 represents how many winters they spent in fresh water as babies (i.e. fry) and the Spring and Summer referring to timing of their migration.

Estimates for this year suggest a run of these Chinook somewhere between 58,000 and 62,000; however, not a lot is known about these fish and they represent a huge geographic area in the various Fraser tributaries that they spawn in – from the far upper Fraser to tributaries downstream through Williams Lake (Chilcotin) and Thompson River and tributaries.

I asked what seemed like the obvious question to me: “what is the rate for maximum sustained yield (MSY)?… is it 80% of the run — like in the past; 60% like it is with sockeye now; 50% like many Alaskan salmon runs are managed to?”

“Well, you have to understand…” began the DFO rep…

Eghad, here we go again.

“…these Chinook are managed differently… it’s based on habitat capacity and output…so it’s not a set rate”

“Gee… that sounds accurate”, was the response that slipped out of my mouth.

I was given a name of a DFO scientist working in Nanaimo that is apparently the expert on this: “Parken”.

_ _ _ _

I looked up his work (apologies, I’m assuming it’s a he… i think I heard the pronoun “he”). I came across a report on the Canadian Science Advisory Secretariat: Habitat-based methods to estimate escapement goals for data limited Chinook salmon stocks in British Columbia, 2004

Seems like some pretty interesting work… although I was struck immediately by a key phrase in the title: “data limited“.

There’s another way to phrase that, that might be more accurate: “limited data“.

In other words… we don’t really know because we have limited data.

And this is laid out quite clearly in the report, even in the abstract:

Our habitat-based model can generate biologically-based escapement goals, rooted in fish-production relationships, for data limited stocks over a broad range of environments. This simple approach requires easily acquirable data and makes few assumptions. However, spawner escapements of known accuracy and reliability are required, which may impede implementation for some systems. The approach is well suited for most data limited stocks in BC and can be tested and refined as new stock-recruitment data become available. Since the habitat-based method was more accurate than the interim method for BC Key Streams, we recommend applying it for data limited stocks in BC to establish escapement goals until more stock-specific data are available. [my emphasis]

Again… as mentioned previously in other posts. My comments are not meant to be a jerk; just pointing out some gaping voids and massive assumptions. Sometimes these assumptions are pointed out in papers and reports; yet, many/most fisheries science folks still talk in circles around them. It’s as if the massive assumptions and “data limited” experiments are a personal affront against them and their “science”.

Our goal was to develop a habitat-based approach to generate escapement goals for data limited Chinook stocks in British Columbia (BC). We focused on developing a model with general applicability that could be applied inexpensively and quickly, while making sufficiently accurate predictions to suit fisheries management purposes.

That’s the problem in most cases… “fisheries management” is more speculative then investing in penny stocks sold out of some guys basement. And thus the “purpose” of fisheries management is to carve out as many fish as possible for human consumption, economic opportunities, and social considerations — then think about the environmental/ecological implications…thus “sufficiently accurate” for fisheries management?! Yikes…

We focused on developing simple models that lacked biological detail, yet described general biological patterns across a range of environmental conditions and Chinook salmon biology. Inasmuch as high precision and accuracy are desirable properties of models, we aimed to develop a method with reasonable accuracy and precision for most domestic and international fisheries management purposes.

But isn’t “biological detail” the whole purpose of looking after fish populations and all the critters that depend on them?

Yes, precision and accuracy are desirable properties… especially if you work for the same federal department responsible for decimating North Atlantic Cod.

There’s a big important word missing after the “properties or models” and the “,” (comma); it’s “BUT”…

As in “… , but we aimed to develop a model with reasonable accuracy”

“Reasonable” accuracy…?! what the ^*!#?

reasonable to whom…?

Is this like the legal test: “a reasonable person…” or is this like “reasonable” to fisheries management folks… or “reasonable” to a Fisheries minister with a long distinguished career with Revenue Canada.

Sorry folks, but the history of “reasonable accuracy and precision for domestic and international fisheries management purposes” is brutal. There’s a reason why the oceans have lost 90% of large predator fish and 75% of the world’s fish populations subject to fisheries pressure are in trouble.

Time for a new paradigm.

Is this a wet blanket?

And maybe time for DFO to stop managing salmon via a patchwork quilt of methods…

Or at least fully admit limitations — especially “data limited” limitations — and get a whole lot more precautionary. Oh wait, isn’t the precautionary approach part of the Wild Salmon Policy…?

One thought on “wild salmon dissonance and patchwork quilts for fisheries managers

  1. tlellami

    Salmon Guy,

    FRSSI is a model/process I’m just beginning to understand, but I believe it does attempt to set total allowable mortality of Fraser River sockeye (total mortality = natural and fishery related mortality). Presumably, fishing plans will be made pre-season to ensure total mortality is not exceeded. Once the acceptable total allowable mortality for each conservation unit (CU) is reached (a CU being DFO’s new management unit as per the Wild Salmon Policy), there will be no more fishing allowed on that specific CU. On your comment that it does not take into account ecological values (seals, bears, etc.), I would argue that this is indeed captured as part of the “natural mortality” in FRSSI – though values used here are, in all probability, highly uncertain.

    Also, to your comment that FRSSI does not incorporate any data prior to 1948, I have to ask, “Should it?” The world (and its environment) today is very different than it was in mid-century, and I would argue that perhaps we should be using only the most recent 20 years’ or so worth of data, as this would be most representative of the “current state” of our ocean/rivers/predator-prey interactions, etc. This point is similar to those raised about the poor performance of some of the sockeye forecast models (of which I believe there are 8 or so) – that they rely too heavily on trends from decades past…trends that are no longer relevant 50 years on.

    In general, I also am having a harder and harder time accepting that the money that DFO is putting into its sockeye program is doing us any good. As you make clear, we really don’t know all that much about Fraser sockeye (200 populations, 19 stocks that we monitor, 4 run timing groups that we manage), and still we do a horrible job trying to make sense of it all. If we spend millions monitoring and managing these fish and still can’t do a very good job, I’d be inclined to totally re-think what we are doing before we invest more money and time to try to monitor all 200 populations and more than 2 rearing lakes, as I think you may have been implying.


Leave a Reply

Your email address will not be published. Required fields are marked *