A Physiologial Comparison of Two Rowing Ergometry Protocols on Performance in Male Oarsmen

The purpose of this studM wos to exomine the effect of two strotegica11y different protocols performed on the Concept II rowinQ ergometer, on the physiological response and distance/time relationship in men's lightweight race rowing. Ten members of the University of Rhode Island men's lightweight crew team, eight of whom comprised the boot which won the 1984 Dad Vails Sma11 College Chomptonship, and two alternates volunteered for this study. Subjects performed a 3.5 mile ·an-out9 (AO) rowing protocol designed to simulate the length, duration, and strategy of traditional 2000 meter rowing race. Fourty-efght hours later the subjects performed a second 3.5 mne protocol (P), designed to simulate the pacing strategy recommended for most endurance type roces. Rowing performance was measured in elapsed time (min:sec) to complete the 3.5 mtle protocols. Stroke rate (SR) was evaluated every thirty seconds (s) with the use of o stroke watch, while metabolic efficiency was determined by thirty second calculations of heart rate (HR), absolute oxygen consumption (V02), relettve oxygen consumption (MV~t venttlatory equivalent (Ve), and respiratory quotient (RQ). Paired T-tests were applied to the forementtoned to determine possible significant differences between the two testing protocols. Total time taken to complete the two testing protocols was not significantly different, and almost identical between the tests. The mean ttme taken for completion being 379.6 +8.13, and 380.3 +-8.31 seconds in the AO and P protocols, respecttYely. SR for the AO protocol was significantly higher during the first 30s and significantly lower from the 90s point to completion of the test. HR fn the AO test was higher throughout the entire test, reachfng significance ot the 30s, 150s, ond 270s marks. V02 was also higher throughout the entire AO test, reaching signiftcance through the first 120s, and again at the 240s and 360s marks. The average across all 30s group mean values for SR, HR, absolute and relative V02, VC02, and minute volume were signf flcantly higher throughout the AO protocol. A significant difference was seen between the total energy costs of the two testing protocols, wtth the AO test cost being signiffcontly higher. These dffference, accompatnied by almost identical times taken to complete the two tests, suggest thot employment of the pacing strategy seen within the P protocol may result in more efficient mechanisms and effective uttltzatton of energy sources in the working muscles, and may result in a greater use of ·energy stores· over the last several hundred meters of 2000 meter race rowing.

Given the dem8nds of 8 rowing reg8tt8, which is 2000 meters in length for men, ond lasts between six and seven minutes, a physto1ogtco11y sound and efficient race strategy should presumably be employed. This hcts not been the case in rowing. Traditionally, o8rsmen hove violated recommended methods of pacing for endurance events which typically include a steady aerobic workload over the majority of the race with an anaerobically based sprint over the final 45 to 60 seconds of the rece (Hagerman et al., 1979;Mohler et al., 1964). Rowers, however, begin races at extremely high energy expenditures and marked ancterobic response for the first 30 to 45 seconds followed by the ·body· of the race, which ts at a slower steady state performance level for the middle 4 to 5 minutes of the race. This ts followed by another anaerobic sprint over the lost 45 to 60 seconds (Hagerman et 81., 1972;Hagerman et al., 1979). Rctce rowing sp11t times show the first 250 meters to be the fctstest, with the following 250 meter segments covered in 8 slower split time 8nd the last 250 meters being slightly faster (Hagerman et al, 1979). This strategy puts e at stress on the oxygen trcnsport systems end tncrecses cncerobtsts (Hegermcn et cl, 1972), which produces mcximal lactcte levels ct the start of the rece. This increase in cnaerobisis occurs when the individuol reoches his or her anaerobic threshold which ts defined as being the point at which there ts o non-ltneor increose in ventnatory equivalent/oxygen consumption (VE/V02), whne venti16tory equfv61ent/c6rbon dioxide (VE/VC02) rem6ins unchanged (Wasserman et el, 1973). These moxfmol loctote 1eve1s must be endured for the durctton of the rcce (Hagerman et ol, 1979).
Maximal stress testing to study oarsmen's physiologic capobt1ities hos been done primoriJy on treadmill tests or on bicycle ergometers (Clark et al, 1983;Secher et al, 1982) until the rowing ergometer was developed.
Studies have found that the results obtained on the rowing ergometer closely simulate those during actual rowing (Hagerman et al., 1979;Mahler et al, 1984). Due to its specific1ty, however, the rowing ergometer hos become the preferred method of exercise testing for ocrsmen (Cunningham et al., 1975;Hagerman et al., 1978, Hagerman et al., 1979Mf ckleson and Hagerman, 1982).
Physiological parameters have primarily been evaluated on two different protocols on rowing ergometers; 6 six-minute ·011-ou\9 (6M-AO) test, and a progressive, incremental (P) exercise test to exhausUon (Mahler et al. 1984). The 6M-AO test ts used for its close opproxtmatton of intensity, duration, end racing strategy to competitive rowing (Hagerman et al., 1978, Hagerman et cl., 1979Mohler et cl., 1984). From a ·stonding start·, each oorsmen rows the first ten strokes w1th moxtmal effort at a cadence of 40 strokes· minute -1. The stroke cadence ts then reduced to 2 34-36 strokes· minute -l untn the lost thirty seconds of the test when the stroke rote ts tncreosed to 40-42 strokes · minute -1 for the ftnol sprint (Mohler et ol.. 1964).
The P protocol corresponds to the troditiono1 form of exercise testing seen during treadmill ond bicycle ergometry tests (Nickleson ond Hogenncm. 1982). Recently, Mohler et 01., (1984), hove compared ond found no stgntftcont difference tn peok vo 2 or other peok physiologic porameters, with eoch test losUng between six ond seven minutes before exhousUon of the subject. Although each protocol hos its advontoges, there hes been no comporison of distance troveled through the different stages of en ·e11-out• test versus e ·pocing· (P) test fo11owing the recommended pocing strategy of most enduronce roces. Since success in ony roce is determined by the comporison of time needed to cover o set distonce, e study of this distonce/Ume relationship seems in order.
Understanding of this dtstonce/ttme re1otionship between the P test, which would seem more physio1ogtco11y efficient, ond the 6M-AO test, which simulates the treditiono1 epproech to recing, would seem to be tmperetive in developing e deeper understonding of the physiologico1 porometers thet e11cit opttme1 rowing performonce.

STATEMENT OF THE PROBLEM
The purpose of this study wos to exemine the effect of two strotegico11y different protocols performed on the Concept II rowtng ergometer, on the physiologicol response, end time/distonce relationship in men·s lightweight roce rowing. Ten members of the University of Rhode Island men's lightweight crew team volunteered for this study. Eight of the subjects compnsed the boat which won the 1984 Dad Vails Sma11 College Championship. The remaining two were alternates to that boat.
The ten-subjects performed a 3.5 mile ·a11-out9 r~wing protocol, destgned to simulate the length, duration, and stategy of traditional 2000 meter race rowing. Forty-eight hours later the subjects performed a second 3.5 mtle protocol, designed to simulate the pacing (P) strategy of most enctunmce type races.
Split times were taken at each .5 mile. Also, heart rate {HR), absolute oxygen consumption {V02t relative oxygen consumption (MV02), ventnatory equivalent (Ve), fraction of expired oxygen (Fe02), fraction of expired carbon dioxide (FeC02), carbon dioxide production (VC02), and respiratory quotient {RQ) were calculated every thirty seconds for each subject. In addtUon, body compositions were also obtained for each subject.
Each oarsman was evaluated on each protocol in a laboratory setting.
Rowing performance was measured In elapsed time (min:sec) to complete the 3.5 mne protocols. Compansons of each thirty second recordings of HR, Y02, Ve, Fe02, FeC02, VC02, and RQ, were used to determine metabolic efficiency.
The following hypothesis were tested; HYPOTHESIS I Subjects w111 achieve a faster time while rowing 3.5 miles using the pacing protocol as opposed to the traditional ·a11-out9 protocol.

HYPOTHESIS II
Mean heart rotes will be higher in subjects performing the tred1t1onal ·an-out· protocol when compared to the pacing protocol.

tt'f POTHES IS II I
The mean energy cost of the P protocol wi11 be Jess than the ·an-out9 protocol as measured by mean absolute and relative oxygen consumption.

DEFINITION OF TERMS
For the purpose of this study , the following definitions were used: Begth1ng Freayency -(BF) -The number of exhalations recorded per minute.
Heort Rate -(HR) -The number of times the heort beats fn one minute.
power Ten Strokes -The point or points in o rowing regotto to which the stroke rnte increases ond oll out effort is given on ten consecutive strokes. subjective judgement was that subjects in this study had not been exercised to mcudmal 11mits of their 8erobic energy capacity (Henderson and Haggard, 1925). These studies, though primatfve when compared to today's methods and technology, opened the door to rowing reseorch. Of equol importance, the innovation of these scientist os opp11ed to testing apporatus, identified the import~mce of specificity of testing during actual performance, a procedure current 1 y in use when measurements of optima 1 pert ormance are desired.
PhySica1 Charecteristics of Rowers lntematiomi1 coHber oarsmen and oarswomen tend to be ton, muscular, and Jean, with an age renge that shows wide variability, renging from 16 to 36 years (de Gary et ol.;. Oarswomen are rether tall athletes w1th heavy skeletomuscular structure (Hebblenick et al., 1960). Measurements collected on more than 600 oarsmen (Hagerman et al., 1979) hove shown the average height of heavyweight oarsmen to be 192 centimeters with an averege body weight of 66 kilogrems. Percentage of fat in heavyweight oarsmen has shown a decHne in recent years with reports of avereges between 91 and 10 I. ( Hagerman, 1964). An earlier study (Hagerman et ol., 1979) reported an averege of 11 I body fat for elite heavyweight oarsmen.
This testtng procedure using the bicycle ergometer as the testing apparatus was the method used on most highly trained athletes., regardless of the specific sport tn which the athlete competed.
The measure of physical working capacity in oarsmen using a rowing ergometer was first introduced in 1971 ( Hagerman and Lee., 1971). A mechanically braked rowing ergometer was used tn most of the studies done in the United States because of its mechanical operation and task specificity (Harrison, 1967;. A six minute test protocol was the designed exercise test in many of the early studies (Hagerman and Lee, 1971;Hagerman, 1975;Hagerman et al., 1972Hagerman et al., , 1975aHagerman et al., ., 1976 6. R. Hagerman, 1976). This design was used as it closely simulates rowing on B oored bmtt over the standard 2000 meter distance. A three minute protocol was later designed and used to simulate a 1000 meter race, usually rowed by women In more recent studies., (Hagerman and Nickleson., 1961;Mahler, 1983;Mahler et a1., 1983;Nickleson emd Hagerman, 1982) a variable wind resistance rowing ergometer (Concept II, Morrisvill,VT) reploced the fixed resistance rowing ergometer explained by Hagerman et al. ( 1976) as the testing apparatus. The majority of experienced oarsmen and oarswomen Pf ck the Concept 11 as the ergometer that best sf mu I ates actual rowing.
Studies have been done recently on the Concept II, to calculate power output relative to velocity and stroke rotes performed during various intensities and durations of work (Hagerman and Mtmsfield, 1984).
LocoHzed muscular fat1gue prior to the attainment of maximal working copoc1ty has been the primary reason g1ven for the depressed vo 2 max yelues tn cycle ergometry. Due to the nature of the she minute test most often used, peek V02 max values were often recorded between the second end fourth minutes of exercise, rarely in the fifth , and never 1n the sixth and final m1nute of exercise (Hagerman et el., 1979;Mohler et el 1984).
Some studies hove utilized a graded treadmill exercise test to exhaustion end found vo 2 max values equaled the h1ghest V~ measured during simulated rowing (Carey et al., 1974;Hagerman et al., 1975b). The differences in oxygen utllizetion tn these types of tests hove been attributed to differing total muscle messes 1nvolved (Hegermen, 1985).
Though both rowing end cycling ere weight support1ng exercises involving extensive quadricep muscle group action, treadmHl running end rowing use the hamstring group to a greoter extent (Hogermon, 1984).
· This attainment of physiological peak closely porrallels those found in recent s1x m1nute row1ng ergometry studies end brings to question the unique pattern of peeing that rowers use during competition.

Eva1uottng Energy Costs
The measurement of human energy expenditure at rest and during Physical octtvfty hos been of greot interest over the years to scientists and athletes alfke. Energy expenditure, or heat production hos been measured in two ways; direct calorimetry and Indirect colorimetry.
Direct colorimetry Involves ploctng subject to be tested Inside an air tight, thermally lnsu1oted chamber where heat production during the subjects activity can be evaluated. This method, though highly accurate, is tmprocttcal for evaluating energy costs during varluos sports, recreational, and occupational activity. Indirect colorimetry ts almost always the method used in these c8ses. Aerobic Importance Studies during acute and chronic hypoxic exposure have magnified the Importance of high levels of aerobic capocity to an oarsmen's performance (Hagerman, 1969;Hagerman et al., 1975b). More dysponoea-hypoia related Physical collapses were reported during the rowing competition at the 1966 Olympic games in Mexico C1ty than for ony other aerobic type event or sport.
More than 60 Incidents of physical collapse by oarsmen In the first 2 days, 13 and severel more during later races were reported by rowing officials tn Mexico City (Hagerman, 1969). Complete cessat1on of a crew during competition ts extremely rare in International Regattas and 1f so, ts usually due to mechanical difficulties. Several crews in Mexico City stopped frequently and some fettled to finish races at the 1966 Olympic Regatta. Hagerman et al., ( 1975) also reported significant effects on venttlatory adaptation In oarsmen following acute and chronic exposure to moderate altitude.

Pacing Strategies During Rowtng
Men's 2000 meter compettttve race rowing has been previously establtshed to be a predominately aeroblca11y based exercise with aerobic metaboltsm yielding 751 to 651 of the total energy cost (Connors, 1974).
Gtv~n this energy demand, a physiologica11y sound and efficient racing strategy would presumably be employed. Ironically, this has not been the case. Traditionally, oarsmen have violated the recommended methods of pacing for endurance events: a steady aerobic workload over the majority of the race with an anaerobically based sprint over the final 45 to 60 seconds of the race (Hagerman et al. 1979;Mahler et al. 1964). During traditional race rowing, oarsmen begin races at extremely high energy expenditures and marked anaerobic response for the first 30 to 45 seconds before they settle at D slower steady rate of performtmce for the middle four to five minutes of the race. This ts followed by tmother anoeroblcolly based sprint over the last 45 to 60 seconds (Hagerman et al., 1972;Hagerman et al., 1979). Race rowing spHt times show the first 250 meters to be the fastest, with the 14 followtng 250 meter segments covered in equal slower times with the last 250 meters betng slightly faster (Hagerman et al., 1979). This strategy put a great stress on the oxygen transport system and increases 8naerobisis (Hagerman et aJ., 1972), which produces maximal lactate levels at the start of the nsce. These lactate levels must be endured for the duration of the race whtle inhibiting aerobic performance (Hagerman et al., 1972).
Aerobic Metabo1fsm of Oarsmen limited research was conducted with oarsmen during the 1920's (Henderson and Haggard, 1925;Ltljestrand and Undhard, 1920), and not until the late 1960's did physiological data of oarsmen and oarswomen begin to appeer fn scfentiffc Htensture (Astrand, 1967;Astrand and Rodahl, 1977;Hagerman, 1969;Hamby and Thomas, 1969;Hay, 1968;lshiko, 1967;Mellerowicz tmd Htmsen, 1965;Nowacki et al., 1969;Saltin and Astrand, 1967). In 1967, high levels of aerobic capacity were reported in Swedish oarsmen performing on bicycle ergometers (Astrand, 1967). The average of these high levels of aerobic capacities ranked these oarsmen behind only biathletes, cross country skiers, and orienteering athletes. Saltln and Astrand ( 1967) studied the effects of maximal treadmill running and bicycle ergometry work on VD:2 and heart rate 1n several h1ghly cond1t1oned 8thletes from the sports of canoeing and rowing. These subjects performed both arm end leg exercises on a specially designed bicycle ergometer. The oarsmen's maximal vo 2 in ltters per mtn compared favorably wtth the results achieved by the c~moeists, cyclist, middle distance runners, and biathlon competitors, ell ech1ev1ng an average VCJi max of 5.1 to 5.4 11ters per minute. When oarsmen's vo 2 max was expressed 1n relaUve terms, m11111tters per ktlogram per min (ml/kg/min), an average of 62 ml/kg/min was obtained.
This value was well below the mean values acheived for other endurance type othletes, reportedly due to consistently greater body weights in the oorsmen.

Absolute and Relative Values
Numerous studies have shown that maximal aerobic capacities of elite oarsmen and oarswomen are among the highest recorded (Di Prampero et al., 1970;Hegermtm and Lee, 1971;Hagerman et el., 1972Hagerman et el., , 1975aHagerman et el., , 1978Hagerman et el., , 1979Jackon andSecher, 1973, 1976;Larson end Forsberg, 1980;Mahler et al., 1983Mahler et al., , 1984Nowacki et el, 1969Nowacki et el, , 1971aSaltin and Astrand, 1967;Secher, 1963;. Absolute vo 2 max values have been measured in excess of 7 Uters per minute (1/m1n) 1n 2 e11te oarsmen, end over 5 1/m1n 1n 3 females (Hagerman at al., 1979). Translated to relative terms these values were over 60 ml/kg/min for the men and over 70 ml/kg/min for the women Ughtweight oarsmen have attetned the highest relative vo 2 max WUh values exceeding 80 ml/kg/min (Hagerman et al., 1979). Due to their body size and total muscle mess, Hghtweight ottrsmen would be expected to achieve this greater v~ max in relative terms whlle et the same time not reechlng the absolute values achieved by heavyweight oarsmen.
Meteboltc data collected on more than 2000 oarsmen and oarswomen 16 with tntemot1onol experience oppeors to 1nd1cote th8t 1f rowing oth etes expect to become successful ot this level, moles should be oble to ochieve o w 2 mex of 6 1/min wh11e females should be oble to ech1eve 8 VOi mex of 4 1/min (Hagerman, 1984).
The fact thot oarsmen's and oarswoman's weight is supported in the boat, seems to point toward absolute V02 max as the more relevant of the two measurements of oxygen uptake. The Germen Democretic Republic uses such on objective criterion to assist in identifying athletes with outstanding physiological cepacitfes ( FISA Coeches Conference Report, Rome, Italy, October 1980). Athletes from East Germany must be able to attain a VOi max of 6 1/min and 4 1/min for men and women respectively to be considered for netionel team selection. This selection process ettempts to tdenttfy those athletes heving one of the mejor components ettributed to intemeUonal rowing success; e highly developed oxygen transport end deUvery system.

Oxygen Consumption During Rowing
Oxygen consumption tn rowers parallels the demands of the athlete on the oxygen trensport system as is the case in all physical events. A cereful examination of the aerobic curve in oarsmen shows e unique response to a predominately aerobic event. With the exception of the first minute of exercise, oarsmen perform at near moximol oerobic capacities for the entire duretion of e six minute exercise test (Hegerman et al. 1979). The lergest portion of a very high oxygen deficit is incurred during the first 30 to 90 seconds of exercise after which they must call on their highly developed aerobic capactttes to meet the energy requirements of the next four minutes (Connors, 1974;Hagerman et al., 1976: G. C. Hagerman, 1976Poltnskt, 1976). Anaerobtsts ts called upon over the last 30 to 60 seconds when a boat's final all-out sprint to the ftntsh occurs. This approach appears to eltctl a rether Inefficient approach to energy production.
An estimated oxygen cost of 4611ters was reported by DIPrampero et al., (1971) for rowing a paired boat over 2000 meters. This value was obtained during tank or basin rowing where water and current conditions are often Jess than optimal (Hagerman and Lee, 1971 ). Oxygen consumption during actual rowing was reported by Jackson and Secher ( 1976)  during paired rowing as opposed to eight oarsmen sharing the work load may explain these oxygen cost differences. It ts also possible thet the rowing ergometer slightly underestimates an oarsman's maximal aerobic capacity.
Differences in boat length, weight, and design have an been mentioned as important fectors that mey alter an oarsman's energy production.

Aerobic Capacities
Studies on the aerobic capacities of rowers have consistently put them emong the e11te 8thletes 1n terms of 8ch1ev1ng extremely h1gh mox1mum v~ values. Although thts me8sure of V~ max 1s helpful tn ossesstng othlettc potential and performance, rowers' most impressive phystolog1ca1 attribute seems to be their ability to sustain extremely high percentages of their obsolute v~ max even after they have exceeded their anaerobic threshold levels (Hagerman et al., 1978).

Energy Costs of Rowing
V~ and velocity measurements of an eHte paired-oared crew were used to calculate the metaboHc cost of rowing at racing speed. An everage of 6.38 liters/minute was reported (Jackson and Secher, 1973;) An oxygen cost contlnuem every ten years for race rowing from 1919 to 1979 wes later shown using this same data (Sacher, 1963 Mangarhfs formula (Mangano et al., 1963(Mangano et al., , 1964 was applted to post-exercise venous blood lactic acid concentrations and indicated that glycolysts provided 4.1 kcal/minute or 13.61 of total energy expenditure. Secher (1963) proposed a simHcr percentage of 141 attnbuted to anaerobic metabolism.

Aneerobic Threshold Limits
A greded exercise protocol on the rowing ergometer has been used in 21 recent studies for the purposes of determining both anaerobic threshold nd vo 2 max. Graded exerc1se test1ng a11ows for a gradual 1ncrease 1n metabolic end cerdiorespiratory acUvtty. This gradual increase a11ows detectton of the point where exercise intensity begins to exceed the cepacity of the oxygen dellvery system to sustain exercise. This progressiYe incremental exercise test allows the athlete to continue exercise fn a stepwise fashion untt1 V~ fs reached, sometime after the observation of the anaerobic threshold.
Anaerobic threshold, or an increase in anaerobisis, has most often been defined as the point in a graded e>eercise test at which there is a non-linear tncreose tn Ve/vo 2 , while Ye/Yco 2 remains unchanged (Wasserman, Whipp, Keyel, end Beaver, 1973). This progressive incremental testing procedure hes also elicited vo 2 max results similar to the ma><1mum or peak values of vo 2 recorded during three and six minute rowing ergometer tests designed to s1mulete race condit1ons end not just en acceptable anaerobic measure of anaerobic threshold (Hagerman end M1ckleson, 1981;Mahler, 1983;Mahler et 11. 1963;M1ckleson end Hagerman, 1982).
A great deal of controversy sun surrounds the accuracy end mean1ng of enaerobfc threshold measurements. Recent studies (Hagerman and M1ckleson. 1981;Mahler. 1983;Mahler et al., 1983;M1ckleson and Hagerman. 1962) have reported anaerob1 c thresho 1 d measurements that ere 1 ndeed est1mates, but never the less prov1de useful Information to athletes and coaches a11ke to better evaluate relative fitness levels and to determine tre1n1ng 1ntens1t1es. Anaerob1c thresholds of 851 to 951 of V02 max (Hagerman and M1ckleson, 1981;Mahler, 1983;Mahler et el., 1983;M1ckleson and Hagerman, 1962) attest to the very high aerobic capacities of rowing athletes.
Anaerobic threshold measurements ere most commonly reported es a percentage of V~ max. In studies performed on oarsmen dur1ng the off season, measurements were stgntf1cantly lower, 701 to 751, than those measurements taken only weeks before the World Rowing Championships, when they were reported et bet ween 851 end 951 of vo 2 max (Hagerman and Mtckleson, 1961;Hagerman end Staron, 1983;Nickleson and Hagerman, 1982).
In terms of enhancing performance, the increase in oxygen utilization could delay the possible deleterious side effects of increasing lactic acid during high intensity exercise (Astrand and Rodahl, 1977;Hagerman et al., 1978). An increased ability to utilize lactic acid es a fuel for exercise may be en important attribute of the endurance trained athlete with high anaerobic thresholds (Ortelt, 1970;Spitzer, 1974). Utilization of lactic acid during a six minute rowing ergometer test reported by Hagerman et al., ( 1978) was proposed since there was either a slight reduction or no change tn this variable from its peek concentration et the second minute of exercise until cessation of exercise, despite signiflcent involvement of the eneerobtc energy system.

Energy Contribution of Anaerobic Metabolism
Aerobic meteboltsm es expressed by oxygen uptake is a highly reliable measurement and has been reported extensively in numerous studies.
Aneerobtc meteboltsm, on the other hand, ts extremely dtfftcult to assess.
Thts relattve contribution of anoerobtc metoboltsm to rowtng has en estimated in G number of wttys including the mettsurement of oxygen deficit, oxygen debt, ttnd the energy equivttlent of post-exercise lttctttte concentrations (Connors, 1974;Httgermttn et al., 1974Httgermttn et al., , 1978G. R. Hagerman, 1976;PoHnsk1, 1976 Each subject was coaxed through his customary start from the St8t1onary position of 3/4 slide, 1/2 slide, 3/4 slide, 4/5 slide, 5/6 slide, full slide, followed by 20 fu11 strokes of maximal effort at a stroke rate codence of 36 strokes per minute. The stroke rate was then reduced to a cadence of 30 strokes per minute until the three mile mark, where the subject was coaxed for hts final all-out sprtnt over the last .5 mtle. After a forty eight hour pertod, tn which all subjects completed the exact training sessions, subjects returned to perform the second exercise testing protocol. This Pacing (P) protocol was designed to follow the generally employed strategy in most exercise events of the aerobic type.
Each subject agatn performed a five minute warm up on the Concept 11 rowing ergometer at a flywheel speed of twenty-two mph, (35 kpm).
After this warm up, the subject was given a five minute rest as he received tnstructtons from his coxswain on the exercise test that would follow.
Each subject was again coaxed through his customary start from the stationery position of 3/4 slide, 1 /2 slide, 3/4 slide, 4/5 slide, 5/6 slide, full sltde, f on owed by 20 full strokes of maximal effort at a stroke rate cadence of 32 strokes per minute. The stroke rate was then reduced to a cadence of 31 strokes per minute until the three mile mark, where the subject was coaxed for his final an-out sprtnt over the last .5 mile.
Two sets of power ten strokes were ttlso performed by each oarsmen 1n eech exercise testing protocol. These power tens were done at a cadence of two strokes per minute faster then the settle stroke rate in the body of the given protocols, or at 32 strokes per minute in the ·an-our protocol and 33 strokes per minute In the pacing protocol. These strokes were done at the 1.5 and 2.3 mile marks in each of the two protocols which simulates closely the points in a rowing race where power tens are usually performed.

.Qgdyed Measurements
Derived measurements include percent body fot from hydrostoUc weighing ond stroke rote from stroke wotch reading.

Pbystcol Measurements
Height wos recorded to the nearest quarter of on inch, cmd weight was recorded to the nearest quarter of a pound on a phys1c1on's scale.  Mehler et al., 1964), subjects in the present study tended to be slightly shorter tmd heavier, with Mahler et al. ( 1963) reporting an average height of 163 +-3 cm and weight of 72.2 +-1.4 kg.
The percentage of body fat of the subjects in this study was higher ( 11 .261) when compared to the 7 to 6 percent reported by Hagerman et al., ( 1964) and the 8.5 percent reported by Burke (1980) for lightweight oarsmen. Subjects tested in the present study were tested in the off season and were not in peak condition. This moy explain some of the dHference between group scores in percentoge of body fat. Hagerman ond Staron (1963) found elite oarsmen to hove quite extreme seasonal vor1ot1ons across all physiological variables.
TesUng Times Table 2 reports the mean group scores of total time taken to complete the ·all-oue and pacing exercise tests. The group average for the completion of the 3.5 mHe rowing ergometry test employing the all-out strategy was 6' 19.6. +-6.13·. The group average for total Ume when subjects employed the pacing strategy was 6·20.3· +-8.31 •. As these times  Stroke Rate Figure 1 shows the group mean stroke rates across each 30 seconds of the two testing protoco1s. Stroke rates for subjects performing the AO protoco1 were constderob1y higher during the first 30 seconds when compared to the rates over the fjrst 30 seconds of the P test. Fo11owing sfmflar rates recorded at the 60 second mark, subjects achieved and maintained higher stroke rates using the P protoco1 for the remainder of the test. Tab1e 3 shows the group mean va1ues of subjects' stroke rates across each 30 seconds of the two testing protoco1s a1ong with the average stroke rates across a 11 30 second ti me pert ods. A 1 so shown in T ob 1 e 3 are probabtlity values resulting from dependent T-tests appHed to the two exercise tests.  29 .. · . · .. · . · ·· · · · · · · · · · · ·· · · · · ·· · · · · · · -·. -. · .· .· . -. ·. · . -.... · . · . · . · .·~. · . · . -. · . · . · . :· i.;;.· 2~~~·.· . · . · . · . · . · . · . · . · . ·-~~==-:=. · . · . · . · . · . · . · . · . -. · . · . · . · . · . · . · . · . · . -. · . · . · . · . · . -. · . · . · .  tn the AO protocol and from 32 to 31 SPM 1n the P protocol for the body of the test. After this cross over, subjects performing the P protocol achieved stgnlflcanUy higher stroke rotes over each 30 second mark for the duration of the p test until the 360 second mark, where significance at the .05 level wos barely missed. These differences were designed within the protocols prtor to testing cmd show that distinct protocol differences were achieved. ourtng the last .5 mile of each protocol, or approximately the last minute of the test, subjects were coaxed into their final sprint In which they were to perform at maximal effort until the 3.5 mile mark was achieved. During this time period, subjects performing the P test were able to increase their stroke rates faster and maintain this increased rate longer when compared to their AO test peformance. Over the entire test, subjects averaged o greater stroke rote during the P test when compared to the AO test.
In summary, with the exception of the 60 second stroke rote scores, a si.gnificcmt difference at the .01 level was seen between all 30 second mean scores of the AO test when compared to the P test. Subjects attained a htgher stroke during the first 30 seconds of the AO test, o simtlar stroke rote at the 60 second mork, end a signtffcont1y higher rote during the last 5 minutes of the P protocol. Subjects were oble to maintain this higher stroke rate over the entire P test. Distinct protocol differences were designed and achieved "tYithin each of the two testing protocols.

Heart Rate
Heart n~te (HR) data were only collected on nine subjects. Figure 2 shows the group mean heart rote values across each 30 seconds of the two testfng protocols. Heart rotes recorded during the AO protocol were higher   beats -minute -I +-7 beats · minute -l was achieved during the first mtnute which gradually increosed to 183 +-6 b ·min -l by the sixth minute.
Subjects in the present study responded similarly at the 1 minute mark with an average of 175.2 during the AO test which remained sHghtly higher with an average of 185.4 at the sixth minute.
In summary# the average HR recorded throughout the 3.5 mile AO protocol was significantly higher then the heart rate values recorded for the 41 p protocol. Meon group HR's were stgnificontly higher throughout the first 90 seconds fn the AO testing protocol and remoined higher throughout the enUre test. Averoges ocross 011 30 second scores for the entire test were el so sfgnf ff contly higher in the AO test despite a sf gnificontly lower stroke rote performed during the AO test.
Absolute Oxygen Consumption In summory, the total oxygen cost to complete the AO test was stgniffcantly higher when compored to the P test. Thts oxygen consumption releites directly to energy production over the fndfvtduo1 30 second seg-  Relative Oxygen Consumption Figure 4 shows mean relative oxygen consumption (MV02) across each 30 second period for both exercise protocols. The subjects relative oxygen consumption was greater at the 30 second mark in the AO protocol and remained higher throughout the duration of the test. Table 6 reports the group mean relative oxygen consumption values across e8ch 30 seconds of both the AO and P protocols. Also shown in Table 6 ore the significant differences between the subjecrs performance during the two tests.   Subjects wnl achieve a faster time whne rowing 3.5 mnes using the P protocol as opposed to the traditional AO protocol. Table 6 reports the mean group scores of total time taken to complete both the AO and P exercise tests. The group average for the completion of the 3.5 m11e rowing ergometry test employing the AO strategy was 6 · 19.6. +-e.13·. The group average for total ttme when subjects employed the P strategy was 6. 20.3· +-8.31 •. Though employing quite different testtng strategies over the duration of the two testing protocols, the difference in total times taken to complete the 3.5 m11e protocols were not significantly different. Therefore, hypothesis 1 wos rejected.

Hypothesis 11
Meon heort rates w111 be higher in subjects performing the AO protocol when compared to the P protocol.

Practical Implications
The purpose of the present study was to determjne jf the pacing strategy employed in the majority of aerobjcally based endurance events could be applied to the sport of race rowtng where a unique pattern of racing strategy has evolved over the years. This strategy is often referred to as the ·an-out9 pace. Extremely high energy expend1tures and elevated oxygen costs are demanded over the jnit1al 30 to 45 seconds of the race followed by a slower steady rate performance level for the middle 4 to 5 minutes of the race, 8nd finally an anaerobjc sprint over the last 45 to 60 seconds of the race. The pacing strategj es used 1 n the ma j on ty of endurance events, which typi ca Hy include a steady aerobic workload over the majority of the race wtth an anaerobically based sprint over the ftnal 45 to 60 seconds of the race was what the P protocol in the present study simulated. Sjgniftcant differences in stroke rates and selected physiologic responses were found between the two protocols. However, when consjdenng performance over the 3.5 m11e rowing ergometer tests, Httle djfference was reported jn times to complete the two exercjse tests.
Absolute and relotive oxygen consumption values were stgnjffcontly higher over the fjrst two minutes and on average for the ent1re AO test, which clearly indicate a higher metabolic cost for ·a11-out9 performance.
Subjects expended significantly more energy during the AO test wjthout any advantage jn actual performance. Conversely, the metaboHc data suggests that the oarsmen "conserved· stgntftcantly more energy during the P test wtthout any odvontage in actual performance. At o glonce, thjs fjndjng seems contradictory. However, 1t should be noted that all subjects were coaxed in the identical manner during the Jost minute of each test when the14 were instructed ·an-out·. During this time, the subjects performing the P test were able to increase their stroke ntte more quickly as well as mointoin higher stroke rates throughout the completion of the test.
From these observations ft may be implied thot the AO test or trodftfonal approach did not elicit a favorable response from the subjects 1n that 1ts start required too great an energy demand from which subjects never appeared to recover. This was demonstrated in the difficulty to increase and maintain a greater stroke rate over the last minute of the race.
The strategy employed in the P protocol allowed the subjects to increase their stroke rates and maintain them until the 3.5 m11e mork wos reached.
These data suggests that the stroke rates employed in the P protocol may not have been demanding enough to push each subject to his maximal performance. Thfs ·settled· rate of the P test may not have optimized the conditioning of these wen trained athletes, Jeoving them with o greater energy reserve than necessary for the middle and final stages of the 3.5 mf1e row. A strategy employing the P tests mol<e-up ot a slightly higher ·settled· stroke rote through the body of the test would allow for s1fghtly faster stroke rates without the drastic changes employed in the traditional strategy.
The prnctical question facing researchers and cooches alike lies in how one con determine the proper pacing strategy to o11ow for optimal increases in the oxygen transport system without any effect of anaerobic metabolism.
In many endurance events the onoerobic threshold is used to determine this pofnt. However, due to the significant workload placed on oarsmen ot the beginning of a n~ce to get the rowing she11 in motion, on anaerobic response 51 ts Inevitable in the first 1 to 2 minutes.

Recommendations for Future Research
Race rowing, as in all other competitive racing events, poses the challenge to athletes and coaches alike to develop a racing strategy that will elicit optimal performance In the event of their choice. To this point, tradition, more than scientiHc knowledge and experiments hos dictated the unique strategy employed by most elite oarsmen and rowing teams. The pattern of beginning races at extremely high energy expenditures along with a marked anaerobic response has remained the sport's trademark. The present study presents metoboHc information which questions such an approach. Though subjects were unaccustomed to the P protocol in the present study and expended significantly Jess energy during this test, times for .the P test were nearly identical to those achieved during the AO test.
Conversely, subjects performing the AO protocol, a racing strategy that was famnar and practiced, had a metabolic cost which was significantly greater without ~my improvement In actual performance. Of interest to both the athlete and coach alike would be a similar study which employed a period of experimenting with stroke rotes in an effort to find the optfmol rote for the P test. This type of test would o11ow for the subjects to become fomnor with o steadier race pace than they tradftiono11y fo11ow ond moy result in subjects .. finishing strong· os opposed to ·holding on· during the final minute of o rowing race.
A second study should include the training of oarsmen in the fashion presented by the present studies P test. By keeping the pacing strategy relat1vely consistent and increas1ng the 1ntensity across an entire tra1n1ng session or competition as subjects become more fit and not any particular portion of the race over the other portions, an oorsman would opproach h1s training and performance in a more traditional and phsiologically sound approach.
Cone 1 usi ons 1) A significantly higher stroke rate may be maintained dunng a 3.5 mne rowing ergometry test when using a pacing strategy as opposed to an ·an-out· strategy.
2) Heart rates responses are significantly higher during an ·a11-out9 rowing test of 3.5 m11es when compared to a rowing protocol employing a pacing strategy.
3) Energy requirements as estimated from oxygen consumption are sfgniffcantly higher during et 3.5 mf1e row employing en •t111-out9 strotegy when comp6red too pocing strategy.