Aerobic and Anaerobic Effort: Your 2K Is More Aerobic Than You Think

The energy-system argument comes up in every club room and it is almost always settled before anyone reaches the water. The rower is told they are aerobic or anaerobic, the label is applied to a session, and the piece gets done. The problem is that the labels were inherited from a physiology that has since moved, and the numbers most rowers have been shown for a 2,000-metre effort were never measured in the first place.

Where the contribution tables came from

The classic tables — phosphocreatine supplying roughly seventy percent of ATP over six seconds, oxidative supply reaching roughly eighty percent by six minutes — came out of work in the early 1980s and have been reprinted ever since. They are useful. They are also, for rowing, mostly inferred rather than observed.

Secher's 1993 review is the reference most rowers cite for rowing physiology, and it is admirably careful about its own limits: the relative anaerobic contribution to six and a half minutes of all-out rowing, it says plainly, "has not been determined" and is estimated to fall between 21 and 30 percent ([7] Secher). That estimate became the received number. Nobody measured it, and for three decades nobody checked.

The mechanism underneath is also often taught more rigidly than the evidence supports. The stroke is an ordered sequence — catch, drive, finish, recovery — and every repetition has to run it in that order to spend energy at all ([9] Soper). But a system does not switch on when the previous one finishes. The systems run concurrently and their proportions shift; a three-hour piece still has a phosphocreatine cost at the catch of every stroke, and a four-minute race still has an oxidative system carrying most of the load from the second minute onward. The manufacturer's technique material describes the sequence as a sequence ([1] Concept2), and the sequencing is real. The handoff is not.

What happens when someone actually measures it

Four research groups have now instrumented rowing directly, using accumulated oxygen deficit rather than post-race lactate. Their results agree with each other closely, and they agree least with the inherited estimate.

In 2,000-metre race simulations, de Campos Mello and colleagues recorded the aerobic, phosphocreatine and glycolytic split on the ergometer as 84, 7 and 9 percent ([2] de Campos Mello). In the same study the water came out at 87, 7 and 6 percent. Working with thirty-one French competitive rowers aged twelve to seventeen, Diry's group put the under-eighteens racing 2,000 metres at 87.0 percent aerobic and 13.0 percent anaerobic ([5] Diry). Astridge's eighteen junior rowers gave 17.3 percent anaerobic contribution over 2,000 metres, which is 82.7 percent aerobic ([3] Astridge). Higgins and Hofmijster, with nine highly trained men, measured 17.5 percent anaerobic over a 6:30 piece — the closest equivalent to a 2K — leaving 82.5 percent aerobic ([4] Higgins).

Four labs, four methods, four different athletes, and the answer lands between 82 and 87 percent aerobic. That consistency is the finding. The classic tables implied roughly seventy-five percent for the same effort, and Secher's much-quoted estimate implied as little as seventy. Both are higher than what direct measurement returns, and both came from a method that reads anaerobic work off the blood lactate at the end of a race. That reading cannot support the claim made on it: blood lactate is a net pool shaped by appearance and removal together, so a single end-point concentration says nothing on its own about how much energy the anaerobic systems actually supplied over the course of the piece.

No gold-standard method for splitting energy contribution exists, and the researchers say so. Different laboratories measure it differently and none can claim the definitive figure. The 82-to-87 percent range should be read as a well-supported estimate with real error bars, not a physical constant.

The ergometer is not the water, but not for the reason you think

The usual explanation is that the ergometer is easier. It is not. In de Campos Mello's comparison, the water race simulation took 515 seconds against roughly 400 seconds on the ergometer — significantly longer, by a wide margin. But the energy expenditure per minute was identical across all three conditions: 99, 99.6 and 100.2 kilojoules per minute in the water, the ergometer without the slide and the ergometer with it. Heart rate, oxygen uptake and lactate did not differ either ([2] de Campos Mello).

So the ergometer is not a cheaper version of rowing. It is the same work per unit time covering less distance, because the flywheel's braking resistance behaves like a larger, faster boat than a single scull. The authors draw the obvious coaching conclusion: if you want an erg test that genuinely simulates a single scull, race 2,500 metres, not 2,000.

That is the real reason an erg 2K time is not a boat 2K time, and it is worth sitting with. The physiology transfers almost perfectly. The distance does not.

Shortening the piece raises the anaerobic share

If the anaerobic share of a 2K is around 17 percent, what raises it? Making the piece shorter. Three separate datasets point the same way.

Higgins and Hofmijster ran nine highly trained men over 6:30, 5:10 and 4:00. Anaerobic contribution climbed from 17.5 percent at the long duration to 27.5 percent at four minutes, and mean power rose 11.7 percent from 322 to 360 watts ([4] Higgins). Astridge's group got a smaller step in the same direction: 2,000 metres to 1,500 metres moved anaerobic contribution from 17.3 to 22.2 percent, with mean power up 5.2 percent from 324 to 341 watts ([3] Astridge). Diry's age categories show the same gradient across distances — 13.0 percent at 2,000 metres, 15.3 at 1,500, 19.4 at 1,000 metres ([5] Diry). Diry's comparison is between age categories rather than within the same rowers, so duration and age are tangled together there; the other two are within-athlete and are the stronger evidence.

One thing these studies cannot tell you is which of the two changes is doing the work. In every comparison the piece gets shorter and faster at the same time — Higgins's 4:00 average was 360 watts against 322 at 6:30 — so duration and intensity are confounded throughout. Read them as showing that a shorter, harder piece carries more anaerobic contribution, not as isolating duration as the cause.

What the evidence does separate out is stroke rate. It did not differ significantly across Higgins's three durations. Faced with a much more anaerobic piece, these athletes did not simply start rowing faster — they produced more power. Whatever drives the higher contribution in a shorter piece, it is not a rower sprinting up the rate to meet it.

Where feeling and measurement come apart

Here is the finding most likely to change how you pick a session.

Astridge's 1,500-metre trial had the higher anaerobic contribution and the higher mean power. It also had a lower maximum heart rate, a lower rating of perceived exertion and a lower blood lactate concentration than the 2,000-metre trial ([3] Astridge).

The shorter, faster, more anaerobic piece was less brutal by every measure of suffering.

The reason is duration. A 2K has more time at a high fraction of maximum oxygen uptake, produces more lactate in total, and creates a larger clearance demand — even though it hands a smaller percentage of the bill to the anaerobic systems. The share went up and the total load went down at the same time.

So the two questions a coach asks are genuinely separate. How hard is this? and How anaerobic is this? do not have the same answer, and choosing a workout because it feels anaerobically brutal will systematically push you toward long pieces at high intensity rather than the short pieces that actually develop the anaerobic contribution you are after.

Astridge's group also found that pacing behaved the same way at both distances, and that the rowers who gained the most power going from 2,000 to 1,500 metres were the ones who paced the shorter race most evenly. Smooth is fast, and it is worth practising deliberately rather than hoping for it.

Lactate is a traffic report, not a fuel gauge

Lactate has been treated as a fatigue meter for as long as rowing has been measured, and the assumption that a high number means you are close to empty has never been true. Maciejewski's group gave seventeen trained rowers a three-minute all-out erg test and used the accumulated oxygen deficit to estimate their non-oxidative energy supply, then tracked lactate through recovery.

The non-oxidative supply correlated with how much lactate the working muscles actually released (r = 0.58), with the muscle content of the lactate transporters MCT1 and MCT4 (r = 0.63), and with the removal rate constant (r = 0.55) ([6] Maciejewski). The same physiological number is entangled with how well a rower moves lactate, not only how fast they make it. Two rowers with identical 2,000-metre splits can have genuinely different underlying physiology because they clear lactate differently.

There is a practical consequence. Chasing a post-session lactate reading is chasing a number that partly reflects how your recovery went, not purely how hard you trained. The reading is real. It is just not measuring what people think it measures.

What the 2,000 m actually measures

The 2,000 metres is usually described as a synthesis of aerobic and anaerobic energy provision. The measured regression says something narrower and more specific. Ingham's group tested forty-one elite championship finalists and built a model of 2,000-metre speed from four variables: power at VO2max, oxygen uptake at the blood lactate threshold, power at the 4 mmol/L blood lactate concentration, and maximal power. Together these explained 98 percent of the variance in speed, and the model was then validated in eighteen further elite rowers with agreement limits of roughly minus 1.5 to plus 6.9 seconds ([8] Ingham).

Three of those four predictors are power or oxygen. The 2K is a power test. It rewards how much force you can put into the handle and for how long, which is a different question from which energy system is paying.

And the shape of that power matters, because peak handle force does not arrive at the start. In race simulations, handle force peaked at 46 percent of the stroke cycle in male rowers and 43 percent in women ([10] Ledergerber). The stroke builds, and the finish is where the power is taken back out. A fast start followed by a fade is not a smooth power curve with a bad first phase; it is the shape the test is specifically built to punish.

What to do with this article

Measure your own anaerobic contribution before you plan around it. The rates on either side of your 2,000-metre pace define the range where the balance shifts fastest, and the distances where that balance actually flips are shorter than most club sessions assume.

When a piece feels disproportionately hard for its duration, check what you actually did before concluding you have found your threshold. A 2,000-metre effort can produce a higher lactate and a higher heart rate than a faster, shorter one while containing less anaerobic work overall.

Compare like with like. The federation's indoor distances — 500 m, 1,000 m, 2,000 m and 30 minutes ([11] World Rowing) — differ by enough that the energy balance between them is not a rounding error. And when you compare an erg result to a boat result, compare the work, not the distance: the per-minute cost is the same and the boat simply takes longer to cover it.