The advent of rowing ergometry has facilitated not only environmentally controlled training and monitoring of rowers but more critically the establishment of important physiological determinants of 2000-m on-water performance (13). Logically, ergometry testing cannot account for differences in boat speed associated with crew number, oar blade surface area (sculling or sweep-oar rowing), and synchronicity or effectiveness of work performed with the oar blade (feathering, entry, work-slip arcs, stroke rate, and consistency) (1,10,16,24). However, with all such factors remaining equal, rowers who sustain greater average net propulsive power achieve faster boat speeds (9,34). Therefore, in the quest to maximize average boat velocity over a 2000-m race, various rowing ergometer protocols are deployed to evaluate individual physiological needs and to detect changes in performance capacities (13,15,20,22,25,27).
Acknowledging that there are limitations to single-factor performance models, the strongest “determinants” of an elite rower's 2000-m ability, identified from multiple regression models (R2, ∼96%, SE, ∼1.5%) are the power at V[Combining Dot Above]O2max (r = 0.84–0.93, p < 0.001) and the peak stroke power (r = 0.82–0.88, p < 0.001) (13,25). Given these determinants are, physiologically, relatively polarized, and considering a strong correlation between muscle strength and rowing peak stroke power (e.g., r = 0.8–0.9; p < 0.05) (2,13,25), it is not surprising that weight training and benchmarks for strength testing are commonly prescribed as part of preparation plans in rowing (5,8,21).
Absolute maximal strength (in kilogram, Newton, or watts), but not relative maximal strength (i.e., in kilograms, Newton, or watts per kilogram of body mass), is a strong discriminator of rowing ability (17–19,28,30,33). Peak isokinetic quadriceps strength (r = −0.41 to −0.68, p < 0.05), along with V[Combining Dot Above]O2max (r = −0.43 to −0.61, p < 0.05) and lean body mass (r = −0.40 to −0.73, p < 0.05), has been reported as a key (multiple regression) “determinant” of 2000-m performance (28,33). However, given a large aerobic energy contribution in 2000-m racing (around 70–85%) (7,20,32,35), some debate exists as to whether testing and training using weight room exercises should be directed at local muscle endurance, rather than maximum strength adaptations (3,4).
Muscular endurance adaptations have been assessed using the maximum repetitions attained with a 50% 1RM load (14) or the absolute quantity of work (i.e., mass × vertical distance × repetitions = total joules per exercise) performed with a 70% 1RM load (18) during weight room exercise. Researchers using absolute leg press muscular endurance exercise results report strong to modest correlations with ergometer time (r = −0.68 and −0.51, respectively, p < 0.05) (14,18). By way of contrast, poor correlations were associated with rowing performance and absolute upper-body muscular endurance testing (e.g., prone bench pull, r = −0.25; p > 0.05) (18). This is somewhat surprising, given that rowing involves all the major muscle groups and the proven usefulness of upper-body muscular endurance benchmarks in differentiating rowers of varying competitive ability (17).
Given the relatively small muscle mass of the arms and limited contribution to rowing power (16,36), it may be that absolute upper-body muscular endurance testing is relevant only when considered as part of multiple regression models of performance; however, such analysis has not been explored. Alternatively, absolute upper-body muscular endurance tests may be “determinants” of rowing performance measures other than 2000-m (e.g., 5000-m time), but to the knowledge of these authors, such data have not been reported. In terms of weight training, some sense of the relationships between upper- and lower-body exercises and various ergometer tests used to assess rowers would prove useful for the development and assessment of strength and conditioning programs. However, information addressing such contentions is limited. Subsequently, the purpose of this study was to establish strength, power, and muscular endurance exercises for weight room training, which are strong determinants of success in performance measures used to assess elite rowers.
Experimental Approach to the Problem
We developed a cross-sectional research design in collaboration with the National Centre for Rowing Excellence to determine strong common variances between weight room exercises used as part of preparations (predictor variables) and measures of rowing performance (dependent outcomes). The national squad of heavyweight male rowers was invited to participate in this study conducted early off-season (i.e., October) in place of normally scheduled endurance rowing and weight training. Over 12 days of investigation, we assessed each rower's repetition maximum loads for a range of dynamometer and “free weights” exercises (predictor variables) and performance in a range of rowing ergometer tests (dependent outcomes). Rowers were familiar with all exercises and testing protocols because they were used regularly as part of national team preparations.
Pearson correlation moments and stepwise multiple linear regression calculations were used to establish strong common variances shared between predictors and outcomes. We anticipated marginal differences in testing results given training histories and competitive rankings and given the small sample size we selected measures with robust intraclass correlation coefficients (ICC ≥ 0.95) or coefficients of variation (CV, ∼5%). For the purposes of this study, predictors that shared strong common variances with rowing ergometer performance were recommended as exercise combinations for weight room training.
Nineteen heavyweight male rowers volunteered for this study conducted “off-season” (October). These elite open-age and elite under-23 age (U23) rowers had recently competed in “A finals” at the Rowing World Championships, and included 14 gold and 2 bronze medallists. The mean ± SD characteristics of these rowers were age, 24.3 ± 4.0 years; rowing experience, 9.4 ± 4.1 years; height, 189.7 ± 5.3 cm; arm-span, 193.4 ± 6.2 cm; sitting height, 97.8 ± 2.5 cm; and body mass, 87.3 ± 10.1 kg.
Institutional Review Board approval for the study protocol was gained from the AUT University Ethics committee. Written (signed) informed voluntary consent was attained from rowers after project information and familiarization seminars were attended. Rowers who were cleared by medical staff for participation in regular training were allowed to test.
Rowing Ergometer Performance Measures
All rowing performance measures used an air-braked rowing ergometer (Model D; Concept 2, Inc., Morrisville, VT, USA), with the drag factor setting to 130. Rowing performance measures included a 500-m, 2000-m, and 5000-m time trials (in seconds); a 1-hour trial (in meters); and greatest peak stroke power (in watts) attained over 15 maximal strokes. The “aerobic condition” was assessed using a 7-stage incremental test. At the completion of each 6-minute stage, the workload was increased by 15 W by increasing stroke rate, until blood lactate equaled or exceeded 6 mmol·L−1. The blood lactate response was determined using a Yellow Spring Instruments 2300 blood lactate analyser (Yellow Springs, OH, USA) from a 25-μl pipette sample drawn from the earlobe within 30 seconds of the commencement of the 1-minute rest interval allowed between stages. From these data, average power at 4 mmol·L−1 was determined using scatter plot graphing (MS Excel). Previous test-retest reliability reported for these tests were CV = 2.0–3.1% and ICC ≥ 0.96 (12,29).
Leg Pressing and Seated Arm Pulling Dynamometry
A “Dynamic Strength Training” dynamometer (DYNO; Concept 2, Inc.) was used to measure average concentric work produced during leg pressing and seated arm-pulling exercise. The dynamometer provides accommodating resistance by flywheel fan inertia combined with air braking to create drag, comparable in design with a Concept 2 rowing ergometer. A processor chip in the dynamometer console uses changes in flywheel fan motion (acceleration and velocity), measured by a magnetic position transducer, to calculate work (in Joules) each repetition. Because the moment of inertia of the flywheel calibrated at the factory is a known constant, the console calculates the component of work to overcome air drag as a function of the speed of the flywheel and the air drag coefficient (or “load”). The air drag coefficient is determined at the completion of each repetition from the deceleration of the flywheel. For all rowers, the load was set at about 120 with 2 air dampeners opened for leg pressing and at about 100 with all dampeners closed for the seated arm pulling.
Participants adjusted the set-up for leg pressing and seated arm pulling (or bench pull) exercise on the dynamometer before testing (http://concept2.co.uk/dyno/exercises). During horizontal leg pressing, participants strapped their feet into an adjustable foot stretcher, kept their back against a seat rest, and held onto seat handles. Leg pressing repetitions were commenced from a flexed knee position (i.e., the shin was vertical to floor) and completed at full extension. For seated arm pulling, participants sat astride the bench pull seat and adjusted the height of the handle grips (attached to the vertical carriage rail post of the machine) to align with their elbows during flexion. Repetitions were commenced at full elbow extension and by keeping their chest against the support pad and feet positioned on the floor in a forward split stance at all times, completed when the handles carriage rail reached the stops at the end of the support beam.
A console displayed the elapsed time and information about the cadence and work performed on completion of each repetition. Because work is the product of the displacement and force, participants were advised to maximize the “length” and “strength” of each repetition effort they performed. Ten seconds after the completion of each test, the dynamometer console produced a summary of the average work per repetition. These data were recorded as the test result.
Test 1 involved 120 repetitions (120RM) of leg pressing within a target completion time of 240 seconds. All subsequent tests involved both leg pressing and seated arm pulling. Test 2 involved 5 repetitions (5RM, target time 10 seconds), with the best of 3 trials (3 minutes rest between attempts) recorded as the result. Test 3 involved 1 trial of 30 repetitions (30RM, target time 60 seconds) and test 4, 60 repetitions (60RM, target time 120 seconds). A time sheet and a timer were used to ensure that each participant had a rest interval of at least 10 minutes after any 5RM, 30RM, or 60RM test, and that at least 20 minutes rest followed a 120RM test. Previous test-retest reliability trials associated with these methods established using a repeat-measures design (same assessors and participants) in our laboratory were CV = 2.5–6.2% and ICCs ≥ 0.96.
Free Weight Exercises
A 1 repetition maximum (1RM) load (in kilograms) for the “power clean” was determined using free weights (Fitness Works Pty. Ltd., Auckland, New Zealand) within 3 sets (3 sets × 1RM) after warm up (3 sets, 75–90% of anticipated maximum), allowing 3 minutes rest and increments of 2.5 kg per attempt. A six repetition maximum (6RM) load (in kilograms) for the “prone bench pull” (Fitness Works Pty. Ltd.) was determined within 3 sets after a warm up (3 sets, 70–85% anticipated 6RM), allowing 3 minutes rest and increments of 2.5 kg per attempt. “Bench pull power” was assessed by connecting and aligning a linear position transducer (IDM instruments Pty. Ltd, Hallam, Victoria, Australia) to the path of the bar. Data were sampled at 200 Hz with the force time series processed by the Ballistic Measurement System software (version 2009 1.4; www.innervations.com, Australia) to compute peak concentric power (in watts). Previous test-retest reliability trials associated with these methods established using a repeat-measures design (same assessors and participants) in our laboratory were CV = 2–5% and ICC ≥ 0.96.
A common on- and off-water endurance taper (∼50% reduced training volume) was also performed by rowers, who maintained a similar diet and hydration status monitored throughout the duration of the 12 days of the investigation. On all test occasions, rowers were physically prepared by “easy” rowing on an ergometer for at least 20 minutes before any initial test (apart from the aerobic condition test) to ensure maximal efforts were attained during assessment and to reduce measurement errors associated with procedures.
The 5 rowing ergometer tests were distributed over three consecutive days in place of regular endurance rowing (day 1: 5000-m time trial; day 2: 60-minute distance trial, and day 3: 2000-m time-trial in morning, with 500-m time trial and peak stroke power test scheduled in the afternoon). Five days later, the 10 weight room exercises were distributed over 2 training days, in place of normally scheduled strength and conditioning (day 8: power clean [1RM], prone bench pull [6RM] and 120RM leg pressing; day 9: prone bench pull power [in watts], 5RM, 30RM, and 60RM leg pressing, and seated arm pulling Concept 2 DYNO tests). Finally, after 2½ days of rest, the 7-stage incremental power aerobic condition test with blood lactate response was performed (day 12).
All weight room exercises (predictors) and rowing ergometer performance measures (dependent variables or outcomes) were determined and data presented as mean ± SD.
Pearson correlation moments and respective 95% confidence intervals (95% CI) were calculated to determine the common variances shared between predictors and dependent variables. The determination and interpretation of correlation coefficients was dependent on fulfillment of statistical criteria, namely the assumption of normality, linearity, and homoscedasticity in the distribution of data. These criteria were examined using SPSS histograms, normal probability plots, a Shapiro-Wilks test, and bipolar plots and least squares regression analysis. The interpretation of correlation moments was r = 0.0–0.09 (trivial); 0.1–0.29 (small); 0.3–0.49 (moderate); 0.5–0.69 (strong); 0.7–0.89 (very strong); 0.9–0.99 (nearly perfect); and 1.0 (perfect) (11). A paired t-test was used to establish significant correlations between predictors and outcome variables (where p ≤ 0.05). The magnitude of differences in correlations was taken to be significant where the 95% CIs did not overlap or where p ≤ 0.05 of the z-score transformation.
Simple linear and 2-factor multiple regression models of dependent variables were computed to determine predictors or combination of 2 predictors provided greater explained variances (i.e., R2 > 50%) for each rowing performance measure of interest. Colinearity between predictors was assessed initially by cross-examination and exclusion from entry to regression models of highly correlated co-predictors (i.e., r ≥ 0.8). Using SPSS (version 17, SPSS, Chicago, IL, USA), stepwise and then staged backward removal of the strongest predictors were computed against each dependent variable of interest. The probability of the change in explained variance (R2 change) F value for entry was ≤0.05 and for removal was ≥0.10.
The strength of each model fit to the data (R of Model) was compared with use of the means by analysis of variance and considered satisfactory where p ≤ 0.05. Selected models had lower standard error of the estimates (SEE) along with narrowly dispersed 90% CI for each of the b value coefficients (i.e., did not include positive and negative values). To evaluate 2-factor predictor models, partial and part correlations were examined to assess the common variance between each predictor and the outcome variable, and between each predictor and outcome variable while controlling for effect of other variables (i.e., the unique relationship of assessed variable). Selected models had Durbin-Watson values of around 2.0 to confirm the assumption that errors in regression were independent, variance inflation factors (VIF) <10, with average VIF values substantially <1.0 and tolerance estimates above 0.2 to ensure that multicolinearity between predictors was eliminated. Models were excluded if the plot of predicted and the residual of the estimate and observed data displayed skewness, heteroscedasticity, and nonlinearity to plots.
Predictors identified in each selected regression models were tabulated against dependent outcomes with respective R2 and SEE reported. For purposes of this study, these predictors were recommended by equipment type (i.e., dynamometry or free weights) as exercises or combinations of exercises for weight room training.
The mean ± SD values for weight room exercises and rowing ergometer performance data are reported in Table 1. The 5RM Concept 2 DYNO tests resulted in greater average work than 30RM (leg pressing: +23% and seated arm pulling: +30%), 60RM (leg pressing: +39% and seated arm pulling: +51%), and, in the case of leg pressing only, the 120RM test (+59%).
Highly correlated co-predictors were identified as 5RM, 30RM, and 60RM Concept 2 DYNO tests (leg pressing range r = 0.9–0.96, and seated arm pulling r = 0.79–0.84), and prone bench pulling 6RM and power (r = 0.82; 95% CI = 0.57–0.94, p < 0.01). A very strong correlation was found between 5RM leg pressing and 5RM seated arm pulling (r = 0.78; 95% CI = 0.46–0.92, p < 0.01), and between the 60RM seated arm pulling and 120RM leg pressing (r = 0.77; 95% CI = 0.4–0.92, p < 0.01). Given the observed magnitude of correlations, these co-predictors were not entered together in 2-factor regression models of rowing performance.
The 1RM power clean was very strongly correlated to 6RM bench pull (r = 0.77, p < 0.05) and strongly with bench pull power (r = 0.66, p < 0.05). Power clean and Concept 2 DYNO correlations (leg pressing and seated arm pulling) were strong and ranged from r = 0.59 to 0.62 (p < 0.05). Seated arm pulling assessments were strongly associated with 6RM prone bench pulling (range r = 0.60–0.66, p < 0.05), but lower correlations were noted to prone bench pull power (range r = 0.36–0.5, p > 0.05).
Significant correlations were observed between weight room exercises and measures of rowing ergometer performance—Table 2. The 5RM leg pressing (r = 0.51–0.69, p < 0.05), 6RM prone bench pulling tests (r = 0.57–0.75, p < 0.05), and the higher repetition 60RM seated arm pulling (r = 0.51–0.66, p < 0.05) were frequently observed predictors of rowing ergometer performance. Observed differences in correlation coefficients between predictors and performance factors of interest lacked statistical significance after z-score transformation.
Table 3 summarizes weight room exercises that shared greater variances (R2 = 46–68%), with elite rowing peak stroke power, and 500-m and 2000-m ergometer performance with a comparison shown to rowing ergometer tests alone (R2 = 75–83%). Single-factor regression models using weight room exercises were not useful to predict aerobic condition, 5000-m, or 60-minute ergometer tests (R2 < 33%).
The recommended combinations of weight room exercises for program prescription or assessment (monitoring) of elite rowers that explained greater variances in rowing peak stroke power, 500-m, or 2000-m time performances are summarized in Table 4. Two-factor multiple regression models using weight room exercises were not useful (R2 < 33%) to predict aerobic condition, 5000-m, or 60-minute outcomes. Combinations including power cleans (in kilograms) and bench pull power (in watts), 5RM leg pressing (in Joules), and 6RM bench pull load (in kilograms) were identified as the main weight room exercises recommended to predict specific measures of rowing performance. On the basis of coefficient estimates (b values), the greater explained variance of two-factor predictor models (R2 = 59–73%) lacked significance (p > 0.05) to population groups other than the specific sample of elite rowers examined in this investigation.
Exercises used during weight room training that can provide robust diagnostics about an individual's strengths or weaknesses as a rower are of interest to coaches and rowers alike. This study identified that strength, power, and muscular endurance weight room exercises were strong predictors of rowing peak stroke power, 500-m, and 2000-m performance. Therefore, practitioners who wish to select exercises for training, which periodically could be used as assessments indicative of potential success in these rowing performance measures, should consider including (1RM) power cleans, (6RM) bench pulls, (5RM) leg presses, or (60RM) seated arm pulling as part of strength and conditioning plans.
In this study, no single weight room exercise emerged as a “universal determinant” of the various rowing performance measures. That is, predictors varied in accordance with the specific rowing ergometer performance measure of interest, highlighting the diverse range of muscular requirements to excel at rowing (Table 4). Including a range of strength (5RM leg press and 6RM bench pull), power (1RM power clean and bench pull power), or muscular endurance (60RM seated arm pulling) exercises year-round would therefore seem to be a good strength and conditioning practice. Alternatively, these recommended exercises might periodically be used as tests to evaluate training efficacy, which may facilitate a better understanding of effect mechanisms between weight room training and rowing performance. However, the balance between time required, information gained, energy demands, and risks in administering a range of weight room exercises as tests should not be overlooked. For example, very strong correlations indicated that a high degree of common variance was shared between 5RM, 30RM, and 60RM seated arm pulling Concept 2 DYNO tests (range r = 0.79–0.84); thus, any small advantages in terms of explained variance gained on, including a 60RM, could be outweighed by the extra duration and effort required, when compared with the briefer 30RM (r = 0.84, p < 0.05).
Unsurprisingly, linear regression analysis confirmed the greater specificity of ergometer data (R2 = 75–83%) over weight room dynamometer data (R2 = 35–47%) or free weights exercise (R2 = 46%–68%) in the determination of rowing peak stroke power, 500-m, and 2000-m performance. Coaches might therefore consider ergometer tests as better measures for rowers, negating any need for weight room testing other than for the assignment of loads for strength exercise. However, from this study, it was observed that ergometer and weight room exercises in combination were better determinants of rowing performance together, than either measure in isolation. For example, peak stroke power with either 30RM or 60RM seated arm pulling, accounted for 87% of 500-m time and 79% of 2000-m time, respectively (Table 4). It is also worth considering that, unlike ergometer testing, weight room exercise data may prove useful determinants for rowers unable to participate fully in rowing testing, because of injury (e.g., low back).
Weight room exercises shared little common variance with aerobic condition (5RM leg pressing: r = 0.58, r2 = 34%), 5000-m (6RM prone bench pull: r = −0.57, r2 = 33%), or 60-minute rowing performances (60RM seated arm pulling: r = 0.51, r2 = 26%). This is not unexpected, given that the ability to sustain stroke power and stroke rating is more strongly associated with the ability to aerobically metabolize fuel substrates (7,31,32). However, we were surprised that the 120RM (4-minute) leg-pressing test was not found to be a predictor of rowing endurance performance (e.g., 60-minute r = 0.20, 95% CI = −0.37 to 0.71), given previous research. Elite rowers are likely to have similar aerobic power (13,20,25); thus, we suspect that within this elite group, differences in 60-minute ergometer performance were not adequately explained by muscular endurance. It may also be the case that specific tapering was required for the 120RM test.
In terms of the “gold-standard” 2000-m time trial, 5RM leg pressing, in combination with either 60RM seated arm pulling or 6RM bench pull exercise, were found to be strong co-predictors of performance (R2 = 57%, SEE = 6.4 seconds; and R2 = 59%, SEE = 6.3 seconds, respectively) and thus recommended for weight room exercise or assessment. It may be that intensive lower-body strength exercise, in combination with either high- or low-repetition upper-body exercise, provides a training stimulus of equal relevance to 2000-m performance, for rowers of comparable aerobic power (3,4). Maximal loads (1RM power clean [in kilograms], r2 = 61%) and upper-body power (bench pull power [in watts], r2 = 52%) tests were also very strong determinants of 500-m time (R2 = 70%, SEE = 1.8 seconds; see Tables 3 and 4). In this study, approximately equal shared variances in 2000-m ergometer performance could be explained using either 500-m time (r = 0.87) or aerobic condition (r = 0.82), which together, account for up to (R2) 92% of 2000-m performance. Therefore, if rowers have no technique or injury concerns, we recommend power cleans (1RM) and prone bench pulls (using 6RM loads) to be included as part of weight training or strength testing.
Isoinertial exercises like prone bench pulls do not offer the most practical mode for the assessment of strength qualities over a wide continuum of repetition maximums (e.g., 6RM, 30RM, 60RM, or 120RM), given a trial and error approach to load selection. Rowers may also be concerned about the possibility of (temporary) rowing performance losses or the risk of injury perceived with muscle soreness associated with such assessments (6). In contrast, the dynamometer (DYNO; Concept 2, Inc.) provided an accommodating concentric-only drag inertia resistance to quantify average repetition work readily during leg pressing or seated arm pulling exercise. Although the dynamometer has been used to assess rowers in a field setting (5), we were unable to find any published research to justify the dynamometer testing protocols. We observed from our study that, apart from minimizing muscle soreness, specific RM dynamometer protocols appeared to be both reliable and valid measures (i.e., shared strong common variances with specific measures of rowing performance) for rowers (see Tables 3 and 4). Indeed, the dynamometer made the administration of tests both practical and efficient for the assessment of strength qualities using an array of repetition maximums. As the dynamometer was simple to operate, was portable, and, in the opinion of the authors, proved robust for testing or training, it may provide sport scientists and rowing coaches with the opportunity to administer tests efficiently to a broad range of individuals with diverse training histories for sport selection purposes.
Without doubt, the validity of any strength or muscular endurance weight room exercise should be evaluated in context of models incorporating on-water performance data (2000-m time or rank). That is, efficiency on an ergometer is only a rough estimate of on-water rowing performance. Research indicates that 2000-m ergometer times appear only useful for performance models that are constrained to small boat crews, such as singles, doubles, or pairs, and even then, only modest predictive certainty of on-water performance is attained (R2 = 0.60–0.55, SEE 4.3–5.4 seconds) (23). Moreover, whereas allometric scaling to normalize 2000-m ergometer times improved prediction of single sculling rowing speed (R2 = 59.2%, error = 3.1%), the significant effect of body mass on boat drag cannot be overlooked (26). However, there are limitations to the reliability and precision of such on-water testing because of the large SEE of data (23). Accordingly, sport scientists and coaches will continue to use an array of physiological data considered important in the monitoring and assessment of the rower's physical development off-water (500-m time trial, power at 4 mmol lactate, or 1-hour distance challenge) (20) while on-water trials continue to be used to determine combinations and seat positions of rowers within crews (i.e., seat racing).
In summary, a range of strength, power, and muscular endurance measures from weight room exercises seemed to be strong predictors of specific ergometer tests used to assess elite rowers, notably the 2000-m or 500-m time trial or a peak stroke power test. Therefore, practitioners who wish to select exercises for weight room training or researchers seeking assessments for training interventions indicative of rowing strength qualities, should consider (1RM) power cleans, (6RM) bench pulls, (5RM) leg presses, or (60RM) seated arm pulling as part of strength and conditioning plans.
This study examined an array of weight room strength, power, and muscular endurance exercises involving knee extension (leg pressing), shoulder adduction (seated arm pulling and prone bench pulls), and whole-body activities, such as the power clean. These activities are commonly prescribed as part of the weight room training of rowers and therefore could also be used periodically as valid assessments of development. However, because of the level of skill and perceived potential risk of injury associated with traditional free weights, such as delayed on-set of muscle soreness or acute muscle damage (6), isoinertial exercises may not be as easily administered with “novice” or “untrained” individuals. Subsequently, the portable air-braked concentric-only accommodating resistance, provided by the Concept 2 DYNO, seems highly suitable for testing rowers who are not performing weight room training (e.g., as part of controlled research designs) or more particularly, individuals participating in testing as part of ‘talent identification” programs for rowing.
The authors have no conflicts of interest to report, or professional relationships with companies or manufacturers who will benefit from the results of this study. No funding from the National Institutes of Health (NIH), Welcome Trust, Howard Hughes Medical Institute (HHMI), or others was received for this work. Results of this study do not constitute endorsement by the authors or the National Strength and Conditioning Association.
1. Baudouin A, Hawkins D. Investigation of biomechanical factors affecting rowing performance. J Biomech 37: 969–976, 2004.
2. Bell GJ, Petersen SR, Arthur Quinney H, Wenger HA. The effect of velocity-specific strength training on peak torque and anaerobic rowing power. J Sports Sci 7: 205–214, 1989.
3. Ebben WP, Kindler AG, Chirdon KA, Jenkins NC, Polichnowski AJ, Ng AV. The effect of high-load vs. high-repetition training on endurance performance. J Strength Cond Res 18: 513–517, 2004.
4. Gallagher D, DiPietro L, Viser AJ, Bancheri JM, Miller TA. The effects of concurrent endurance and resistance training on 2000 meter rowing ergometer times in collegiate male rowers. J Strength Cond Res 24: 1208–1214, 2010.
5. Gee T, Olsen P, Berger N, Golby J, Thompson K. Strength and conditioning practices in rowing. J Strength Cond Res 25: 668–682, 2011.
6. Gee TI, French DN, Howatson G, Payton SJ, Berger NJ, Thompson KG. Does a bout of strength training affect 2,000 m rowing ergometer performance and rowing-specific maximal power 24 h later? Eur J Appl Physiol 111: 2653–2662, 2011.
7. Hagerman FC. Applied physiology of rowing. Sports Med 1: 303–326, 1984.
8. Hay JG. Rowing: An analysis of the New Zealand Olympic selection tests. NZ J Health Phys Ed Rec 1: 83–90, 1968.
9. Hofmijster MJ, Landman EHJ, Smith RM, Van Soest AJK. Effect of stroke rate on the distribution of net mechanical power in rowing. J Sports Sci 25: 403–411, 2007.
10. Hofmijster MJ, Van Soest AJ, De Koning JJ. Rowing skill affects power loss on a modified rowing ergometer. Med Sci Sports Exerc 40: 1101–1110, 2008.
12. Hopkins WG, Schabort EJ, Hawley JA. Reliability of power in physical performance tests. Sports Med 31: 211–234, 2001.
13. Ingham SA, Whyte GP, Jones K, Nevill AM. Determinants of 2000-m rowing ergometer performance in elite rowers. Eur J Appl Physiol 88: 243–246, 2002.
14. Jürimäe T, Perez-Turpin JA, Cortell-Tormo JM, Chinchilla-Mira IJ, Cejuela-Anta R, Mäestu J, Purge P, Jürimäe J. Relationship between rowing ergometer performance and physiological responses to upper and lower body exercises in rowers. J Sci Med Sport 13 434–437, 2010.
15. Kendall KL, Smith AE, Fukuda DH, Dwyer TR, Stout JR. Criticial velocity: a predictor of 2000-m rowing ergometer performance in NCAA D1 female collegiate rowers. J Sports Sci: 29: 945–950, 2011.
16. Kleshnev V, Kleshnev I. Dependence of rowing performance and efficiency on motor coordination of the main body segments. J Sports Sci 16: 418–419, 1998.
17. Koutedakis Y, Sharp NCC. A modified Wingate test for measuring anaerobic work of the upper body in junior rowers. Br J Sports Med 20: 153–156, 1986.
18. Kramer JF, Leger A, Paterson DH, Morrow A. Rowing performance and selected descriptive, field, and laboratory variables. Can J Appl Physiol 19: 174–184, 1994.
19. Lawton TW, Cronin JB, McGuigan MR. Strength testing and training of rowers: a review. Sports Med 41: 413–432, 2011.
20. Maestu J, Jurimae J, Jurimae T. Monitoring of performance and training in rowing. Sports Med 35: 597–617, 2005.
21. McNeely E, Sandler D, Bamel S. Strength and power goals for competitive rowers. Strength Cond J 27: 10–15, 2005.
22. Mikulic P. Anthropometric and metabolic determinants of 6,000-m rowing ergometer performance in internationally competitive rowers. J Strength Cond Res 23: 1851–1857, 2009.
23. Mikulic P, Smoljanovic T, Bojanic I, Hannafin JA, Matkovic BR. Relationship between 2000-m rowing ergometer performance times and World Rowing Championships rankings in elite-standard rowers. J Sports Sci 27: 907–913, 2009.
24. Millward A. A study of the forces exerted by an oarsman and the effect on boat speed. J Sports Sci 5: 93–103, 1987.
25. Nevill AM, Allen SV, Ingham SA. Modelling the determinants of 2000m rowing ergometer performance: A proportional, curvilnear allometric approach. Scand J Med Sci Sports 21: 73–78, 2011.
26. Nevill AM, Beech C, Holder RL, Wyon M. Scaling concept II rowing ergometer performance for differences in body mass to better reflect rowing in water. Scand J Med Sci Sports 20: 122–127, 2010.
27. Riechman SE, Zoeller RF, Balasekaran G, Goss FL, Robertson RJ. Prediction of 2000 m indoor rowing performance using a 30 s sprint and maximal oxygen uptake. J Sports Sci 20: 681–687, 2002.
28. Russell AP, Le Rossignol PF, Sparrow WA. Prediction of elite schoolboy 2000-m rowing ergometer performance from metabolic, anthropometric and strength variables J Sports Sci 16: 749–754, 1998.
29. Schabort EJ, Hawley JA, Hopkins WG, Blum H. High reliability of performance of well-trained rowers on a rowing ergometer. J Sports Sci 17: 627–632, 1999.
30. Secher NH. Isometric rowing strength of experienced and inexperienced oarsmen. Med Sci Sports 7: 280–283, 1975.
31. Secher NH. Physiological and biomechanical aspects of rowing. Implications for training. Sports Med 15: 24–42, 1993.
32. Shephard RJ. Science and medicine of rowing: a review. J Sports Sci 16: 603–620, 1998.
33. Shimoda M, Fukunaga T, Higuchi M, Kawakami Y. Stroke power consistency and 2000 m rowing performance in varsity rowers. Scand J Med Sci Sports 19: 83–86, 2009.
34. Smith RM, Spinks WL. Discriminant analysis of biomechanical differences between novice, good and elite rowers. J Sports Sci 13: 377–385, 1995.
35. Steinacker JM. Physiological aspects of training in rowing. Int J Sports Med 14: S3–S10, 1993.
36. Tachibana K, Yashiro K, Miyazaki J, Ikegami Y, Higuchi M. Muscle cross-sectional areas and performance power of limbs and trunk in the rowing motion. Sports Biomech 6: 44–58, 2007.