This study investigated whether physiological and performance parameters in elite rowers could be assessed with a single "2-in-1" test combining incremental exercise and a 2000m time trial, rather than with separate tests. Ten elite rowers completed an incremental exercise test, a 2000m time trial, and a 2-in-1 test on different days. Most parameters were not significantly different between the 2-in-1 test and the separate tests, indicating the 2-in-1 test can validly assess these parameters with only one test session.

1. Journal of Science and Medicine in Sport (2009) 12, 205—211
ORIGINAL PAPER
A single exercise test for assessing physiological
and performance parameters in elite rowers:
The 2-in-1 test
Pitre C. Bourdona,b,∗
, Adrian Z. Davida
, Jonathan D. Buckleyb
a South Australian Sports Institute, Australia
b Nutritional Physiology Research Centre, School of Health Sciences,
University of South Australia, Australia
Received 22 December 2006; received in revised form 31 August 2007; accepted 6 September 2007
KEYWORDS
Peak oxygen uptake;
Blood lactate
thresholds;
Rowers;
Exercise testing
Summary Testing to determine blood lactate thresholds for prescription of rowing
training is usually conducted separately from performance testing (i.e. 2000 m time
trial). The purpose of this study was to investigate whether the testing required
to determine blood lactate thresholds and performance in elite rowers could be
reduced by undertaking a single test combining incremental exercise with a 2000 m
time trial. Ten elite rowers (age 20.9 ± 2.1 years, mean ± S.D.) performed, on sep-
arate occasions and in random order, an incremental seven-step rowing test (INCR),
a 2000 m time trial (2k), or a combined test involving the performance of six incre-
mental submaximal workloads followed by 15 min of recovery and then a 2000 m
time trial (2-in-1). Physiological and performance parameters (blood lactate thresh-
olds, accumulated oxygen deficit, heart rate, work parameters) determined during
2-in-1 were not significantly different from those determined during INCR or 2k,
except for peak oxygen uptake which was higher during 2-in-1 compared with INCR
(4.23 ± 0.22 versus 4.14 ± 0.20 l min−1
, p = 0.02), and peak rating of perceived exer-
tion which was lower during 2-in-1 compared with INCR (19.4 ± 0.2 versus 19.9 ± 0.1,
p = 0.02). We conclude that physiological and performance parameters that have
traditionally been assessed during separate incremental exercise and 2000 m time
trial testing in elite rowers can be validly determined during a single combined
exercise test.
© 2007 Sports Medicine Australia. Published by Elsevier Ltd. All rights reserved.
∗ Corresponding author.
E-mail address: bourdon.pitre@saugov.sa.gov.au
(P.C. Bourdon).
Introduction
The regular physiological assessment of rowers in
the laboratory is undertaken principally to ascer-
tain whether training programs are effective, and
to determine the physiological response to exercise
1440-2440/$ — see front matter © 2007 Sports Medicine Australia. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.jsams.2007.09.007

2. 206 P.C. Bourdon et al.
in order to prescribe appropriate exercise inten-
sities. Aerobic metabolism has been shown to
contribute 67—88% of the energy requirement for
rowing performance1,2 and, since rowing perfor-
mance is highly correlated with absolute maximal
oxygen uptake (˙VO2 max),3,4 ˙VO2 max is an important
physiological parameter to assess in rowers. While
aerobic metabolism contributes most of the energy,
the contribution of anaerobic metabolism to rowing
performance is also significant and has been esti-
mated at 12—33%,1,2 making anaerobic capacity an
important parameter to assess in rowers. Anaero-
bic capacity can be estimated by calculating the
accumulated oxygen deficit (AOD).5,6
The Olympic rowing distance is 2000 m so lab-
oratory based performance tests are generally
conducted to simulate rowing this distance.4,7
While the maximal effort required, combined with
the duration of the 2000 m time trial, allow for
the determination of peak ˙VO2 and AOD, a test
of this nature cannot provide any data on blood
lactate thresholds8,9 or heart rate responses to
exercise10,11 which are important for prescribing
training and at present require a separate incre-
mental exercise test for their assessment.5
Given that contemporary elite rowing training
programs may comprise up to 14 training sessions
per week, the necessity for regular testing to deter-
mine the effectiveness of the training program, and
to set appropriate training intensities must be bal-
anced against any potential interruptions to the
training program itself. Usually, testing to deter-
mine blood lactate thresholds for prescription of
training has been conducted separately from per-
formance testing (i.e. 2000 m time trial) because
it was assumed that the performance of one test
would affect the results of the other5; however,
this assumption has not been tested. The purpose of
the present study was to determine whether incre-
mental exercise and a 2000 m time trial could be
combined into a single test without affecting the
validity of the blood lactate threshold and/or per-
formance data collected.
Methods
Ten rowers (2 males, 8 females, age 20.9 ± 2.1
years, height 178.3 ± 7.2 cm, mass 75.0 ± 8.5 kg,
mean ± S.D.), comprising three World Champions,
five World Championship representatives and two
Australian National Championship representatives,
who were all scholarship holders with the South
Australian Sports Institute (SASI), volunteered to
participate in the study after providing written
informed consent. No effort was made to control
for stage of the reproductive cycle in the females.
Testing was conducted in the SASI Sports Physiol-
ogy Laboratory and was approved by the Human
Research Ethics Committee of the University of
South Australia.
On three separate days, and in random order,
the athletes performed either (1) an incremental
seven-step rowing test (INCR), (2) a 2000 m time
trial (2k), or (3) a submaximal six-step incremen-
tal rowing test followed by 15 min of recovery and
then a 2000 m time trial (2-in-1). Each of the three
testing sessions was separated by no less than 2 days
during which time the athletes resumed a standard-
ised training protocol. Body mass and height were
measured at the start of each laboratory visit, and
gas exchange and work parameters, heart rate and
blood lactate concentrations were monitored dur-
ing each test. Ratings of perceived exertion (RPE)
were determined at the end of each workload dur-
ing the incremental exercise tests, and at the end
of each 2000 m time trial.12
All tests were performed on the same rowing
ergometer (Concept IIC) with the drag factor set
at 110 for lightweight women, 120 for heavyweight
women and lightweight men and 130 for heavy-
weight men.
The INCR test required the athletes to perform
6 × 4 min submaximal workloads, with each work-
load separated by 1 min rest intervals. Following
the sixth workload the athletes undertook a final
maximal 4 min effort during which they performed
as much work as possible. The resistances for each
of the six initial workloads were based on each
athlete’s best time for a race simulation test per-
formed over 2000 m on the rowing ergometer during
the 2 months preceding study commencement. The
average 500 m pace of the 2000 m race simulation
plus 4 s was calculated, and then an additional 36,
30, 24, 18, 12 and 6 s was added to give target
times per 500 m for the first through sixth workloads
respectively.5 This test was the current Australian
national screening protocol and was performed rou-
tinely by all of the athletes in this study.
During the 2k test, after a self-selected warm-
up, the athletes were required to row a distance
of 2000 m in the least time possible.5 This test is
a standard criterion used for selection purposes in
many countries,2,5 and was performed routinely by
all the athletes in this study. The typical error for
performance time during a 2000 m rowing time trial
in our laboratory (expressed as a % coefficient of
variation) is 0.5%.
The 2-in-1 test was identical to INCR, with the
exception that after completing the sixth work-
load athletes had a 15 min recovery interval. This
recovery interval incorporated at least 10 min of

3. Assessing of rowers: The 2-in-1 test 207
work on the ergometer at an intensity below the
lactate threshold so as to help clear any lactate
accumulated during the submaximal stages. After
the 15 min rest interval the athletes performed a
2000 m time trial as per the aforementioned 2k test.
Throughout each rowing test, the athletes
breathed through a face mask (Hans Rudolph, Series
7500) attached to a two-way non-rebreathing valve
(Hans Rudolph 2700), with a pre-calibrated large
flow turbine transducer (Morgan Mark 2 ventilation
meter) connected to the inspiratory port to mea-
sure ventilatory volumes. Expired air was collected
into a 2.6-l mixing chamber (Sportech, Australian
Capital Territory) from which dried gas was sam-
pled continuously (∼500 ml min−1) and passed to
an oxygen analyser (Ametek S-3A/I) and a carbon
dioxide analyser (Ametek CD-3A), both of which
had been calibrated prior to each test with three
commercially produced gas mixtures of known oxy-
gen and carbon dioxide composition (BOC Gases,
Australia). The electrical outputs from the gas anal-
ysers and ventilation meter were integrated using a
personal computer, which calculated the necessary
ventilatory variables as 30 s averages using cus-
tom software. The values for submaximal ˙VO2 were
measured by averaging the final 2 min of each sub-
maximal workload of the incremental rowing tests.
Peak ˙VO2 values were obtained by averaging the
two successive highest 30 s data points during the
2000 m time trials and the final maximal 4 min effort
in the INCR test.5
Heart rate (HR) was recorded throughout each
rowing test as 5 s averages using a Heart Rate Mon-
itor (Polar Accurex Plus, Polar Electro). The HR
values averaged over the final 30 s of each work-
load for the incremental rowing tests, and the peak
value attained in the 2000 m time trials and the
final maximal 4 min effort in the INCR tests, were
recorded as the measured values.
Finger prick blood samples for determination of
blood lactate concentrations were collected imme-
diately prior to the commencement of all rowing
tests. Samples were also collected at the end of
each workload during the incremental tests, and
immediately after the completion of each 2000 m
time trial. Aliquots (25 ␮l) were analysed using an
automated lactate analyser (Yellow Springs Inter-
national, 1500 Sport).
The lactate threshold (LT) was identified using
the ADAPT LT method5 which is represented by the
last completed workload preceding the point on
the lactate curve where the concentration reached
a level 0.4 mmol l−1 above the minimum recorded
lactate reading. The anaerobic threshold (AT) was
identified using a modified Dmax method5,13 which
is represented by the point on the third order
polynomial regression curve of blood lactate con-
centration that yielded the maximal perpendicular
distance to the straight line formed between the LT
and end data point.
Accumulated oxygen deficit was estimated
according to the method described by Hahn et al.5
Briefly, the ˙VO2 requirement to support the work
done during each 30 s split of the 2k was extrap-
olated from the submaximal ˙VO2 and power data
collected during INCR, while the submaximal data
collected in the 2-in-1 were used in the determi-
nation of the ˙VO2 requirement of the 2000 m time
trial incorporated in the 2-in-1. In both cases mea-
sured ˙VO2 was then subtracted from the estimated
˙VO2 requirement for each 30 s split to give the oxy-
gen deficit for each 30 s. The oxygen deficit over
the entire time trial was then summed to estimate
AOD.
Athletes were required not to train in the 12 h
preceding each test. On the day before each test
session training was limited to no more than 12 km
of low intensity aerobic rowing. In the 24 h pre-
ceding each test athletes were required to avoid
heavy resistance training or exercise to which they
were unaccustomed. Athletes were requested to
eat a high carbohydrate meal on the evening pre-
ceding each test and if scheduling allowed, also
on the day of the test. Instruction was also given
to maintain good hydration in the lead up to
each test.5 All assessments were conducted dur-
ing the general preparation phase of each athlete’s
training program with the athletes undertaking
∼17 h of training per week in the month preceding
testing.
Normality of the distribution of the data was
confirmed using the Shapiro—Wilk’s W-test. Peak
˙VO2, peak HR, peak lactate at the end of exer-
cise and peak RPE were determined during all
three exercise tests and the values achieved during
each test were compared using one-way analysis of
variance (ANOVA) with repeated measures. Where
ANOVA showed a significant main effect, differ-
ences between means were determined by post hoc
analysis using a test of least significant differences.
Average power output, work done and AOD were
determined during each of the two 2000 m time tri-
als, and these values were compared using paired
t-tests. Paired t-tests were also used to compare
values for power output, ˙VO2 and HR at LT and AT
determined during INCR and 2-in-1. Absolute relia-
bility of outcome measures was determined using
limits of agreement analysis.14 Linear regression
and Pearson product—moment correlations were
used to determine relationships between values.
The level of statistical significance was set at an ˛
level of p ≤ 0.05. All data values cited in the text,

4. 208 P.C. Bourdon et al.
and shown in the tables represent mean ± standard
error unless otherwise stated.
Results
Performance and physiological parameters for the
2000 m time trials determined during 2k and 2-in-
1 are shown in Table 1. There were no significant
differences in any parameters between tests apart
from peak HR which was lower during 2k than dur-
ing 2-in-1 (p = 0.05) and INCR (p = 0.002). There
were relatively narrow limits of agreement for most
parameters for 2-in-1 and 2k, and no systematic
errors between values (p > 0.26).
Parameters determined during the incremen-
tal exercise tests (INCR and 2-in-1) are shown in
Table 2. The peak ˙VO2 achieved was significantly
lower during INCR compared with 2-in-1 (p = 0.02)
and 2k (p = 0.02). Limits of agreement were nar-
row for most parameters but there was a systematic
error in peak ˙VO2 measurements between INCR and
2-in-1, with the 2-in-1 test providing increasingly
higher values compared with INCR as peak ˙VO2 val-
ues increased (r2 = 0.56, p = 0.02). Time to peak
˙VO2 was significantly shorter in the final stage of
INCR compared to both the 2-in-1 (p < 0.001) and
2k (p < 0.001). Peak RPE was significantly higher in
INCR compared with 2-in-1 (p = 0.02). There were
no other significant differences between parame-
ters determined during INCR and 2-in-1.
Discussion
This study demonstrated that physiological and per-
formance parameters which are routinely assessed
during separate 2000 m time trials and incremen-
tal exercise tests in elite rowers can be validly
determined during a single test session combining
incremental exercise with a 2000 m time trial.
There was no difference in the time taken to per-
form a 2000 m time trial whether performed as a
separate stand-alone test (i.e. 2k) or following a
series of incremental workloads. This indicates that
combining incremental exercise with a 2000 m time
trial in the 2-in-1 test did not significantly influ-
ence 2000 m time, which is considered to be the
most important measure of rowing performance in
the laboratory.1,15 Indeed, the mean difference of
1.5 s in performance time between the two 2000 m
time trials (with 2-in-1 taking on average 1.5 s
longer) represented only 0.3% of the performance
time for the 2k. This was within the 0.5% typi-
cal error (expressed as % coefficient of variation)
for this test in our laboratory, and also fell within
Table1Physiologicalandperformancecharacteristicsfor2000mrowingtimetrials
Parameter2-in-12kp95%CIrBiasLimitsofagreement
Time(s)430.0±7.3428.5±7.20.22−1.1to4.30.99−1.5−9.0to5.8
Peak˙VO2(lmin−1
)a
4.23±0.224.22±0.210.86−0.12to0.140.75−0.09−0.26to0.09
Timetopeak˙VO2(min)a
5.7±0.35.9±0.20.86−0.3to0.70.570.2−1.2to1.7
Peakheartrate(beatsmin−1
)193.1±2.3191.5±2.40.050to3.20.24−1.6−4.7to2.7
Peakbloodlactateconcentration(mmoll−1
)9.8±0.510.8±0.30.13−0.4to2.40.150.98−2.3to4.3
Averagepoweroutput(W)286.7±16.8288.6±17.20.54−4.6to8.20.991.9−16.0to19.6
Workdone(kJ)122.2±4.6122.6±4.70.61−1.3to2.10.990.4−4.7to5.5
PeakRPE(arbitraryunits)19.4±0.219.2±0.20.32−0.6to0.20.41−0.2−1.4to1.0
AOD(l)3.30±0.343.36±0.370.84−0.49to0.590.750.05−1.53to1.64
AOD:accumulatedoxygendeficit;Biasrepresents2kvalueminus2-in-1value;RPE:ratingofperceivedexertion;2-in-1:incremental6×4minworkloadsfollowedby15minrecovery
andthena2000mtimetrial;2k:2000mtimetrial.
aDataforonlynineathletesshownduetomissingvaluesasresultoftechnicalproblems.

5. Assessing of rowers: The 2-in-1 test 209
Table2Physiologicalandbloodlactatethresholdparametersforincrementalexercisetests
Parameter2-in-1INCRp95%CIrBiasLimitsofagreement
Peak˙VO2(lmin−1
)a
4.23±0.224.14±0.200.02−0.16to−0.020.75−0.09−0.26to0.09
Timetopeak˙VO2(min)a
5.7±0.33.6±0.2<0.001−3.1to−1.30.15−2.2−4.4to0.1
Peakheartrate(beatsmin−1
)193.1±2.3194.3±2.30.14−0.5to2.90.951.2−3.4to5.8
Peakbloodlactateconcentration(mmoll−1
)9.8±0.510.0±0.40.13−0.5to0.10.560.2−2.4to2.8
PeakRPE(arbitraryunits)19.4±0.219.9±0.10.020.1to0.90.270.5−0.5to1.5
PoweroutputatLT(W)169.3±8.6169.5±8.60.59−0.6to1.00.990.2−2.0to2.4
˙VO2atLT(lmin−1
)2.89±0.132.86±0.120.15−0.07to0.010.99−0.03−0.15to0.08
HeartrateatLT(beatsmin−1
)154.9±2.2155.1±2.20.79−1.5to1.90.950.2−4.3to4.7
PoweroutputatAT(W)234.7±13.2233.8±13.40.58−4.4to2.60.99−0.9−10.6to8.8
˙VO2atAT(lmin−1
)3.69±0.173.67±0.160.54−0.09to0.050.98−0.02−0.20to0.16
HeartrateatAT(beatsmin−1
)179.4±1.9180.4±2.00.36−1.4to3.40.861.0−5.4to7.4
Averagepoweroutputforfinalworkload(W)286.7±16.8290.1±18.20.52−8.1to14.90.963.4−28.0to34.8
aDataforonlynineathletesshownduetomissingvaluesasresultoftechnicalproblems.BiasrepresentsINCRvalueminus2-in-1value.RPE:ratingofperceivedexertion.LT:
lactatethreshold;AT:anaerobicthreshold;2-in-1:incremental6×4minworkloadsfollowedby15minrecoveryandthena2000mtimetrial;INCR:incremental6×4minworkloads
followedby4minmaximaleffort.
the 0.6% coefficient of variation for 2000 m row-
ing time trials reported by others.7 Therefore, the
average 1.5 s longer taken to perform the 2000 m
time trial as part of 2-in-1, compared with per-
forming a stand-alone test, was within the normal
test—retest variation in performance time for a
2000 m time trial, indicating that the difference
in the performance times for these two tests was
trivial. The 2000 m time determined during the 2-
in-1 test therefore represents a valid assessment of
simulated rowing performance.
Given that blood lactate thresholds occur at
submaximal exercise intensities, and the first six
workloads of both incremental exercise tests (i.e.
INCR and 2-in-1) were the same, it is perhaps not
surprising that there were no differences in blood
lactate thresholds between tests. Since the INCR
was the Australian nationally endorsed rowing test
protocol at the time of data collection,5 and the
2-in-1 values for blood lactate threshold measures
were not significantly different from INCR, it would
appear that the 2-in-1 is a valid test for assessing
blood lactate thresholds and related measures in
elite rowers.
The only differences in parameters obtained dur-
ing the INCR and 2-in-1 tests were that 2-in-1
peak RPE was significantly lower, and peak ˙VO2 was
significantly higher. The reason for these differ-
ences could not be determined from the available
data. One possible explanation is that the 15 min
active recovery interval incorporated into the 2-
in-1 may have allowed athletes to recover better
prior to the supramaximal efforts and their subse-
quent perceived exertion was reduced. Recovery
intervals greater than 6 min have recently been
recommended for rowers performing high-intensity
efforts.16
The fact that the athletes achieved a signifi-
cantly higher peak ˙VO2 during both 2-in-1 and 2k
suggests that peak ˙VO2 was underestimated dur-
ing INCR. Other studies have also reported that
peak ˙VO2 was higher in elite rowers when mea-
sured using a sport specific 2000 m race simulation
compared with an incremental stage test.3 Pierce
et al.17 however reported that in well-trained row-
ers ˙VO2 max could be obtained from a discontinuous
incremental test involving seven stages of 3—4 min
duration. Their study however, did not compare the
discontinuous incremental test with a 2000 m time
trial but rather compared it with a 1 min continuous
incremental test as the criterion measure, which is
an unfamiliar test to most rowers. In the present
study however, since the six incremental workloads
during both 2-in-1 and INCR were identical, and
the average power achieved during the final work-
load of INCR and during the two 2000 m time trials

6. 210 P.C. Bourdon et al.
were similar, the higher peak ˙VO2 achieved dur-
ing 2k and 2-in-1 is most likely explained by the
longer durations for which the average power out-
puts were maintained. The final workload during
INCR was 4 min in duration, while the final work-
load during 2-in-1 and 2k lasted almost twice as
long (i.e. ∼7 min 10 s). Studies of oxygen uptake
kinetics suggest that the attainment of a steady-
state is delayed during high-intensity exercise (i.e.
there is a slowing of ˙VO2 kinetics).18 In the present
study, the final workload of INCR and the work-
loads during the final stage of 2-in-1 and during
2k were performed at a supramaximal intensity,
which should therefore have resulted in a slowing
of the ˙VO2 kinetics. This may not have allowed suf-
ficient time for ˙VO2 to reach its true peak during
the shorter maximal stage performed during INCR.
The times taken to reach peak ˙VO2 during the 2-
in-1 and 2k tests support this concept with peak
˙VO2 being achieved in 5.7 and 5.9 min in the 2-
in-1 and 2k tests, respectively (Tables 1 and 2).
Therefore, if during maximal rowing exercise it
requires 5.7—5.9 min to reach peak ˙VO2, then the
4 min available during INCR would not be sufficient
for peak ˙VO2 to be achieved. Irrespective of the
mechanism, it appears that INCR tended to provide
an underestimation of actual peak ˙VO2 compared
with 2-in-1 and 2k, suggesting that 2-in-1 provided
a more valid estimate of actual peak ˙VO2.
Another advantage of the 2-in-1 test is that it
also allows for the assessment of anaerobic capac-
ity. In this study there were no differences between
2-in-1 and 2k for any of the anaerobic capacity
related measures. The mean AOD for 2-in-1 and 2k
were 3.30 and 3.36 l, respectively, which compares
favorably with the value of 3.4 l reported by Prip-
stein et al. for 16 female rowers assessed during a
2000 m time trial.2
A limitation of the present study was that there
was no effort made to control for phase of the
female reproductive cycle. While this should not
affect blood lactate responses,19 there is some
evidence of a trend for improved time-trial perfor-
mance during the late follicular phase.20 A further
limitation of this study was the small sample size
but accessing large numbers of rowers of the stan-
dard used in the current study is difficult, and the
use of this small homogenous sample may there-
fore preclude generalising the findings of this study
to non-elite rowers.
The main finding of this study was that valid
physiological and performance data could be
obtained when an incremental exercise test was
combined with a 2000 m time trial during a single
test session. The adoption of a single combined test
(i.e. 2-in-1 test), rather than having athletes under-
take separate incremental and time trial tests,
would reduce the number of routine testing ses-
sions required, resulting in cost savings and less
disruption to athlete’s training programs. While the
current study has validated the use of the 2-in-1
test in elite rowers it may be possible to extend
this test to other endurance sports, but further val-
idation studies would be required. The 2-in-1 test
provides an alternative to traditionally used tests
for the physiological assessment of rowers in the
laboratory.
Practical implications
• A test which combines incremental exercise
with a 2000 m time trial allows physiological
function and performance data to be collected
in a single assessment, saving time and money
and reducing disruption to training programs.
• It is possible that this test procedure could
be suitable for application to other endurance
sports.
Acknowledgements
The authors would like to thank Ms. Sarah Woolford
(Sport Physiologist) and Mr. Tom Stanef (Biomedical
Technician) for their assistance. Above all we would
like to thank the athletes for giving of their time
and effort. This work was conducted using internal
funds of the South Australian Sports Institute and
the University of South Australia.
References
1. M¨aestu J, J¨urim¨ae J, J¨urim¨ae T. Monitoring of performance
and training in rowing. Sports Med 2005;35(7):597—617.
2. Pripstein L, Rhodes E, McKenzie D, et al. Aerobic and anaer-
obic energy during a 2-km race simulation in female rowers.
Eur J Appl Physiol 1999;79:491—4.
3. Hagerman FC. Physiology and nutrition for rowing. In: Lamb
DR, Knuttgen HG, Murray R, editors. Perspectives in exer-
cise science and sports medicine: physiology and nutrition
for competitive sport. Cooper: Indianapolis; 1994.
4. Ingham S, Whyte G, Jones K, et al. Determinants of 2000 m
rowing ergometer performance in elite rowers. Eur J Appl
Physiol 2002;88:243—6.
5. Hahn A, Bourdon P, Tanner R. Protocols for the physiologi-
cal assessment of rowers. In: Gore CJ, editor. Physiological
tests for elite athletes. Human Kinetics: Champaign; 2000.
6. Medbø J, Mohn A, Tabata I, et al. Anaerobic capacity deter-
mined by maximal accumulated O2 deficit. J Appl Physiol
1988;64:50—60.
7. Schabort E, Hawley J, Hopkins W, et al. High reliability of
performance of well-trained rowers on a rowing ergometer.
J Sports Sci 1999;17:627—32.

7. Assessing of rowers: The 2-in-1 test 211
8. Bourdon P. Blood lactate transition thresholds: concepts
and controversies. In: Gore C, editor. Physiological tests
for elite athletes. Champaign: Human Kinetics; 2000.
9. Steinacker J. Physiological aspects of rowing. Int J Sports
Med 1993;1:3—10.
10. Bunc V, Hofmann P, Leitner H, et al. Verification of the heart
rate threshold. Eur J Appl Physiol 1994;70:263—9.
11. Conconi F, Ferrari M, Ziglio PG, et al. Determination of the
anaerobic threshold by a noninvasive field test in runners.
J Appl Physiol 1982;52(4):869—73.
12. Borg G. Perceived exertion as an indicator of somatic stress.
Scand J Rehab Med 1970;23:92—8.
13. Cheng B, Kuipers H, Snyder AC, et al. A new approach for
the determination of ventilatory and lactate thresholds. Int
J Sports Med 1992;13(7):518—22.
14. Bland J, Altman D. Statistical methods for assessing agree-
ment between two methods of clinical measurement.
Lancet 1986;1:307—10.
15. Soper C, Hume P. Reliability of power output during rowing
changes with ergometers type and race distance. Sports
Biomech 2004;3(2):237—47.
16. Mavrommataki E, Bogdanis G, Kaloupsis S, et al. Recovery of
power output and heart rate kinetics during repeated bouts
of rowing exercise with different rest intervals. J Sport Sci
Med 2006;5:115—22.
17. Pierce SJ, Hahn AG, Davie A, et al. Prolonged incremental
tests do not necessarily compromise VO2max in well-trained
athletes. J Sci Med Sport 1999;2(4):356—63.
18. Roston W, Whipp B, Davis J, et al. Oxygen uptake kinetics
and lactate concentration during exercise in humans. Am
Rev Respir Dis 1987;135:1080—4.
19. Janse de Jonge XA. Effects of the menstrual cycle on exer-
cise performance. Sports Med 2003;33(11):833—51.
20. Oosthuyse T, Bosch AN, Jackson S. Cycling time trial perfor-
mance during different phases of the menstrual cycle. Eur
J Appl Physiol 2005;94:268—76.
Available online at www.sciencedirect.com