The scientific method, it seems, isn’t just a professional ethos. It’s a way of life.
I thought about this today as I was fly fishing up a tricky stretch of Chalk Creek, an icy, milky-green stream that cascades downward through the pine forests and aspen groves of the San Isabel National Forest in the Colorado Rockies. A fresh breeze funneled down the creek bed, tempering the heat of the morning and foreshadowing the usual afternoon rainstorm. Besides the roar of the water, all I could hear was the scolding of chickadees and the aerial battles of hummingbirds overhead. The far bank, lush with moss and ferns and dappled with sunlight, overhung a murky indigo pool where, I imagined, the largest trout remained immune to my arts. The gravel bank under my feet, meanwhile, was festooned with a tangle of saplings, shrubs and overhanging branches that seemed specially designed to ensnare the angler’s line.
Small fry magnet? The Royal Coachman (out of the box) and friends
I’d been having great luck with one particular kind of fly, the Royal Coachman. But with it, I seemed only to be able to land small trout in the neighborhood of six to seven inches long: beautiful creatures, glistening olive green with scarlet spots, but hardly the stuff of campfire legend. While some consider small trout to be the tastiest, I only killed one on our first evening, poached over the fire in white wine and garlic as a sort of arrival ritual. I had decided not to kill one again until I’d bagged something over ten inches long.
As I worked my way up the bank, I began entertain two disparate hypotheses:
1. The stream was currently overpopulated with small trout, so these were the ones I was most likely to catch
or
2. Only the indiscriminate small fish were interested in the Royal Coachman, and the larger ones would prefer to strike on some other fly.
I decided to try to test these hypotheses, so I sat on a fallen stump in the dusky sunshine, cut off the Coachman and tied on something completely: an Elk Caddis Brown. But as I tightened on and waterproofed the new fly, it occurred to me that the experiment was flawed from the beginning. In the ten minutes since I’d caught something small with the Coachman, the sun had shifted in the trees and set a different part of the water in shadow, the temperature had increased – hundreds if not thousands of microscopic variables had altered. If I caught a big trout now, I’d never know if it was down to the new fly. How far this situation seemed from the regulated, air-conditioned world in my lab.
Or was it? I considered a hypothetical row of twenty-four Eppendorf tubes, each filled with the same substance and treated in batch. Could these tests really be considered to be exactly comparable? I might have pipetted more carefully into the first few tubes but grown more lax and cavalier by the end of the row, or a 37-degree heat block might not provide the exact incubation temperature in each of its wells, or there might be variability in the composition of the mass-produced plastic tubes. We might think we are being careful, but it is probably impossible to treat a control in the same way as an experimental sample. On further reflection, I decided that this wasn’t such a bad thing. If general lab conditions really are so variable, it must be a particularly robust biological phenomenon that can provide reproducible results on different days or when performed by different people. That we can learn anything in the face of our chaotic environment is something to be proud of.
Now you’ll have to excuse me. After having blown into this dusty cowboy town to refuel on supplies and file this report, it’s time to get back to the more important business of fishing, reading, swimming, napping, hiking, eating, gold panning – and trying not to think too much about science.
Get back on holiday woman.
(Glad you’re having a fun time. Very jealous)
Yes, sir!
I have an image of Jenny sitting in the lab with a fishing rod and 24 oversized tubes each containing a 12-inch trout. I’m sure NERC or BBSRC would fund it.
That’s a good point about controls and robustness though.
However, the main reason for being so careful is not necessarily that changes/effects aren’t robust, but that they might be due to something else. Which is why I’m thinking about different ways of doing the same experiment.
(And the important question of course is … did you catch the bigger fish?)
I caught my first fish recently. I think I’m hooked (pun intended), especially when garlic and wine and lemon juice are involved. It wasn’t a sophisticated operation like fly fishing though.
Enjoy the rest of your break – it sounds wonderful!
The solutions to your experimental design worries are replication and randomization. As worked out by, naturally, Fisher.
So we need replicate Jennies, all casting at random points along the bank?
I’ll write the grant.
And at random times too.
As long as I get to be the clone that catches the big one…
Size matters?
snort
Apparently.
Got several gorgeous 15-inchers today, on the Arkansas River!
Unfortunately it was catch and release. Mom’s cooking salmon tonight as a consolation prize.
Argh!
I’m sure they wouldn’t have missed just one…
I’m a law-abiding citizen, sir. The sheriffs out here (or “shurfs”, as the locals have it) are pretty bored and have nothing better to do than go after damsels in distress with illicit trout.
I’m moving to Colorado. It’s much more fun than here.
Is a ‘damsel in distress with an illicit trout’ one of those modern idioms that I have yet to understand. I was recently similarly embarassed over ‘being on the bus to Hebden Bridge’, although Queen Victoria denied such things happened.
Don’t ask me Charles: I’m snagged as a wombat’s tonker.
Lots of things in Colorado (prickly and otherwise) that a wonker’s tonker definitely doesn’t want to get snagged on.
a wonker’s tonker? What the blazes are you talking about?
Sunstroke.
A useful ailment to cover a variety of sins.