Blood flow in exercising muscles by xenon clearance and by microsphere trapping

The accuracy of muscle blood flow measurement by the 133Xe clearance method (QXe) was assessed against direct venous outflow (Qv) and microsphere trapping flow (Q mu) determinations in isolated perfused dog gastrocnemius both at rest and during graded stimulation [O2 consumption (VO2) up to 12 ml X 100 g-1 X min-1] and in the gastrocnemius, vastus lateralis, and triceps of intact dogs at rest and while running on a treadmill at varied speeds up to maximum VO2. In 29 measurements performed in 11 isolated muscles, Q mu was in good agreement with Qv at rest and at all stimulation levels (Q mu/Qv = 1.0; r = 0.98). 133Xe clearance yielded much lower blood flows than the venous outflow and the microsphere trapping methods. In 43 measurements in 11 muscles, the mean QXe/Qv ratio was 0.57 +/- 0.03 (SE), independent of blood flow. Similarly, in 65 measurements in 2 intact dogs, the mean QXe/Q mu ratio in all tested muscles was 0.49 +/- 0.02 (SE), independent of blood flow. These results show that the 133Xe clearance method considerably underestimates blood flow in dog muscles.     

VO(2) on-kinetics in isolated canine muscle in situ during slowed convective O(2) delivery

The purpose of this study was to examine O(2) uptake (Vo(2)) on-kinetics when the spontaneous blood flow (and therefore O(2) delivery) on-response was slowed by 25 and 50 s. The isolated gastrocnemius muscle complex (GS) in situ was studied in six anesthetized dogs during transitions from rest to a submaximal metabolic rate (≈50-70% of peak Vo(2)). Four trials were performed: 1) a pretrial in which resting and steady-state blood flows were established, 2) a control trial in which the blood flow on-kinetics mean response time (MRT) was set at 20 s (CT20), 3) an experimental trial in which the blood flow on-kinetics MRT was set at 45 s (EX45), and 4) an experimental trial in which the blood flow on-kinetics MRT was set at 70 s (EX70). Slowing O(2) delivery via slowing blood flow on-kinetics resulted in a linear slowing of the Vo(2) on-kinetics response (R = 0.96). Average MRT values for CT20, EX45, and EX70 Vo(2) on-kinetics were (means ± SD) 17 ± 2, 23 ± 4, and 26 ± 3 s, respectively (P < 0.05 among all). During these transitions, slowing blood flow resulted in greater muscle deoxygenation (as indicated by near-infrared spectroscopy), suggesting that lower intracellular Po(2) values were reached. In this oxidative muscle, Vo(2) and O(2) delivery were closely matched during the transition period from rest to steady-state contractions. In conjunction with our previous work showing that speeding O(2) delivery did not alter Vo(2) on-kinetics under similar conditions, it appears that spontaneously perfused skeletal muscle operates at the nexus of sufficient and insufficient O(2) delivery in the transition from rest to contractions.     

A 31P-NMR study of tissue respiration in working dog muscle during reduced O2 delivery conditions.

To investigate the role of tissue oxygenation as one of the control factors regulating tissue respiration, 31P-nuclear magnetic resonance spectroscopy (31P-NMR) was used to estimate muscle metabolites in isolated working muscle during varied tissue oxygenation conditions. O2 delivery (muscle blood flow x arterial O2 content) was varied to isolated in situ working dog gastrocnemius (n = 6) by decreases in arterial PO2 (hypoxemia; H) and by decreases in muscle blood flow (ischemia; I). O2 uptake (VO2) was measured at rest and during work at two or three stimulation intensities (isometric twitch contractions at 3, 5, and occasionally 7 Hz) during three separate conditions: normal O2 delivery (C) and reduced O2 delivery during H and I, with blood flow controlled by pump perfusion. Biochemical metabolites were measured during the last 2 min of each 3-min work period by use of 31P-NMR, and arterial and venous blood samples were drawn and muscle blood flow measured during the last 30 s of each work period. Muscle [ATP] did not fall below resting values at any work intensity, even during O2-limited highly fatiguing work, and was never different among the three conditions. Muscle O2 delivery and VO2 were significantly less (P < 0.05) at the highest work intensities for both I and H than for C but were not different between H and I. As VO2 increased with stimulation intensity, a larger change in any of the proposed regulators of tissue respiration (ADP, P(i), ATP/ADP.P(i), and phosphocreatine) was required during H and I than during C to elicit a given VO2, but requirements were similar for H and I.(ABSTRACT TRUNCATED AT 250 WORDS)     

Rapid force recovery in contracting skeletal muscle after brief ischemia is dependent on O(2) availability.

We tested the hypothesis that contracting skeletal muscle can rapidly restore force development during reperfusion after brief total ischemia and that this rapid recovery depends on O(2) availability and not an alternate factor related to blood flow. Isolated canine gastrocnemius muscle (n = 5) was stimulated to contract tetanically (isometric contraction elicited by 8 V, 0.2-ms duration, 200-ms trains, at 50-Hz stimulation) every 2 s until steady-state conditions of muscle blood flow (controlled by pump perfusion) and developed force were attained (3 min). While maintaining the same stimulation pattern, muscle blood flow was then reduced to zero (complete ischemia) for 2 min. Normal blood flow was then restored to the contracting muscle; however, two distinct conditions of oxygenation (at the same blood flow) were sequentially imposed: deoxygenated blood (30 s), blood with normal arterial O(2) content (30 s), a return to deoxygenated blood (30 s), and finally a return to normal arterial O(2) content (90 s). During the ischemic period, force development fell to 39 +/- 6 (SE)% of normal (from 460 +/- 40 to 170 +/- 20 N/100 g). When muscle blood flow was restored to normal by perfusion with deoxygenated blood, developed force continued to decline to 140 +/- 20 N/100 g. Muscle force rapidly recovered to 310 +/- 30 N/100 g (P < 0.05) during the 30 s in which the contracting muscle was perfused with oxygenated blood and then fell again to 180 +/- 30 N/100 g when perfused with blood with low PO(2). These findings demonstrate that contracting skeletal muscle has the capacity for rapid recovery of force development during reperfusion after a short period of complete ischemia and that this recovery depends on O(2) availability and not an alternate factor related to blood flow restoration.     

Effect of distribution of the stimulation frequency in the volley on post-contraction hyperemia of the gastrocnemius muscle of cats]

The influence of stimulation frequency distribution in a volley on postexercise (postcontraction) hyperemia was studied in acute experiments on cat gastrocnemius muscle. The initial high frequency of the rhythmic volley augmented the postcontraction hyperemia indices (the peak blood flow and an additional blood volume).     

Blood flow and tissue -pO2 in the trained and untrained gastrocnemius muscle of the anesthetized guinea pig.

In 10 guinea pigs the gastrocnemius muscles on one side were tenotomised. By the tenotomy the daily work load of the gastorcnemius muscle was lowered in the operated leg ("untrained muscles") and increased in the control leg ("trained muscles"). Before and several weeks after the operation blood flow was measured in the lower legs (by segmental plethysmography) and oxygen pressure was measured in the gastrocnemius muscles (by micro-Pt-electrodes) of the anesthetized animals. 4 to 6 weeks after the operation statistically significant differences between the two extermities were noted: In the operated leg the mean pO2-value was 33%lower (P is less than 0.001) and the mean blood flow value 46% higher (P is less than 0.001). These differences could be explained by a reduced number of perfused capillaries in the untrained muscles (= non uniform blood flow distribution).     

Phosphate uptake in rat skeletal muscle is reduced during isometric contractions.

During contractions, there is a net efflux of phosphate from skeletal muscle, likely because of an elevated intracellular inorganic phosphate (P(i)) concentration. Over time, contracting muscle could incur a substantial phosphate deficit unless P(i) uptake rates were increased during contractions. We used the perfused rat hindquarter preparation to assess [(32)P]P(i) uptake rates in muscles at rest or over a range of energy expenditures during contractions at 0.5, 3, or 5 Hz for 30 min. P(i) uptake rates were reduced during contractions in a pattern that was dependent on contraction frequency and fiber type. In soleus and red gastrocnemius, [(32)P]P(i) uptake rates declined by approximately 25% at 0.5 Hz and 50-60% at 3 and 5 Hz. Uptake rates in white gastrocnemius decreased by 65-75% at all three stimulation frequencies. These reductions in P(i) uptake are not likely confounded by changes in precursor [(32)P]P(i) specific activity in the interstitium. In soleus and red gastrocnemius, declines in P(i) uptake rates were related to energy expenditure over the contraction duration. These data imply that P(i) uptake in skeletal muscle is acutely modulated during contractions and that decreases in P(i) uptake rates, in combination with expected increases in P(i) efflux, exacerbate the net loss of phosphate from the cell. Enhanced uptake of P(i) must subsequently occur because skeletal muscle typically maintains a relatively constant total phosphate pool.     

Hemodynamic responses to heat stress in the resting and exercising human leg: insight into the effect of temperature on skeletal muscle blood flow

Heat stress increases limb blood flow and cardiac output (Q) in humans, presumably in sole response to an augmented thermoregulatory demand of the skin circulation. Here we tested the hypothesis that local hyperthermia also increases skeletal muscle blood flow at rest and during exercise. Hemodynamics, blood and tissue oxygenation, and muscle, skin, and core temperatures were measured at rest and during exercise in 11 males across four conditions of progressive whole body heat stress and at rest during isolated leg heat stress. During whole body heat stress, leg blood flow (LBF), Q, and leg (LVC) and systemic vascular conductance increased gradually with elevations in muscle temperature both at rest and during exercise (r(2) = 0.86-0.99; P < 0.05). Enhanced LBF and LVC were accompanied by reductions in leg arteriovenous oxygen (a-vO(2)) difference and increases in deep femoral venous O(2) content and quadriceps tissue oxygenation, reflecting elevations in muscle and skin perfusion. The increase in LVC occurred despite an augmented plasma norepinephrine (P < 0.05) and was associated with elevations in muscle temperature (r(2) = 0.85; P = 0.001) and arterial plasma ATP (r(2) = 0.87; P < 0.001). Isolated leg heat stress accounted for one-half of the increase in LBF with severe whole body heat stress. Our findings suggest that local hyperthermia also induces vasodilatation of the skeletal muscle microvasculature, thereby contributing to heat stress and exercise hyperemia. The increased limb muscle vasodilatation in these conditions of elevated muscle sympathetic vasoconstrictor activity is closely related to the rise in arterial plasma ATP and local tissue temperature.     

Effect of local pulmonary blood flow control on gas exchange: theory

The effect of local pulmonary blood flow control by local alveolar O2 tension on steady-state pulmonary gas exchange is analyzed with techniques derived from control theory. In a single homogeneous lung unit with normal inspired and mixed venous blood gas composition, the homeostatic effect on local ventilation-perfusion ratios (VA/Q) regulation occurs over a restricted range of VA/Q. The homeostatic effect is maximal at a moderately low VA/Q (about 0.4) due to the slope of the O2 dissociation curve. In a multicompartment lung with a lognormal distribution of VA/Q, regulation of arterial O2 tension varies with the extent of inhomogeneity. At mild degrees of inhomogeneity where local pulmonary blood flow (Q) control acts predominantly on the lower VA/Q of the Q distribution, the regulatory effect is best. At severe degrees of inhomogeneity where local Q control acts mainly on the higher VA/Q of the Q distribution, the regulatory effect is worse, and positive-feedback behavior may occur. Local Q control has the potential of reducing the deleterious effects of lung disease on pulmonary gas exchange particularly when it operates in association with other regulatory mechanisms.     

Blood flow in skeletal muscles of chicken in the embryonic and early postembryonic periods

In chicken Leghorn, blood flow volume speed in pectoralis and gastrocnemius muscles was measured on 15 and 19 day-old embryos and at the 1st and the 10th days alter hatching. It was revealed that in the last quarter of embryogenesis BF in muscles did not vary remaining in both muscles in identical limits. Similar BF parameters in pectoralis and gastrocnemius muscles and their age-dependent dynamics were observed at embryos with the detained development (with the body weight 2-fold less than the norm). After hatching, the blood flow in both muscles was grown, on the average, 2.4-fold and remained high by the 10th day, a little decreasing in the pectoralis muscle. It was shown, that increase of a muscular blood flow after hatching was accompanied by different changes of anatomic lumen of the arteries addressed in pectoralis and gastrocnemius muscles: in the former it decreased, in the latter--increased.     

Adaptation of respiratory muscle perfusion during exercise to chronically elevated ventilatory work.

Pneumonectomy (PNX) leads to chronic asymmetric ventilatory loading of respiratory muscles (RM). We measured RM energy requirements during exercise from RM blood flow (Q) using a fluorescent microsphere technique in dogs that had undergone right PNX as adults (adult R-PNX) or as puppies (puppy R-PNX), compared with dogs subjected to right thoracotomy without PNX as puppies (Sham) and to left PNX as adults (adult L-PNX). Ventilatory work (W) was measured during exercise. RM weight was determined post mortem. After adult and puppy R-PNX, the right hemidiaphragm becomes grossly distorted, but W and right costal muscle mass increased only after adult R-PNX. After adult L-PNX, the diaphragm was undistorted; W and left hemidiaphragm RM Q were elevated, but muscle mass did not increase. Mass of parasternal muscle did not increase after adult R-PNX, despite increased Q. Thus muscle mass increased only in response to the combination of chronic stretch and dynamic loading. There was a dorsal-to-ventral gradient of increasing Q within the diaphragm, but the distribution was unaffected by anatomic distortion, hypertrophy, or workload, suggesting a fixed pattern of neural activation. The diaphragm and parasternals were the primary muscles compensating for the asymmetric loading from PNX.     

Effects of inertial load and countermeasures on the distribution of pulmonary blood flow.

We assessed the influence of cranial-to-caudal inertial force (+G(z)) and the countermeasures of anti-G suit and positive pressure breathing during G (PBG), specifically during +G(z), on regional pulmonary blood flow distribution. Unanesthetized swine were exposed randomly to 0 G(z) (resting), +3 G(z), +6 G(z), and +9 G(z), with and without anti-G suit and PBG with the use of the Air Force Research Laboratory centrifuge at Brooks Air Force Base (the gravitational force of the Earth, that is, the dorsal-to-ventral inertial force, was present for all runs). Fluorescent microspheres were injected into the pulmonary vasculature as a marker of regional pulmonary blood flow. Lungs were excised, dried, and diced into approximately 2-cm(3) pieces, and the fluorescence of each piece was measured. As +G(z) was increased from 0 to +3 G(z), blood flow shifted from cranial and hilar regions toward caudal and peripheral regions of the lung. This redistribution shifted back toward cranial and hilar regions as anti-G suit inflation pressure increased at +6 and +9 G(z). Perfusion heterogeneity increased with +G(z) stress and decreased at the higher anti-G suit pressures. The distribution of pulmonary blood flow was not affected by PBG. ANOVA indicated anatomic structure as the major determinant of pulmonary blood flow.     

Effect of ventilation-perfusion inhomogeneity and N(2)O on oxygenation: physiological modeling of gas exchange.

Ventilation-perfusion (VA/Q) inhomogeneity was modeled to measure its effect on arterial oxygenation during maintenance-phase anesthesia involving an inspired mixture of 30% O(2) and either N(2)O or N(2). A multialveolar compartment computer model was constructed based on a log normal distribution of VA/Q inhomogeneity. Increasing the log SD of the distribution of blood flow from 0 to 1.75 produced a progressive fall in arterial PO(2) (Pa(O(2))). The fall was less steep in the presence of N(2)O than when N(2) was present instead. This was due mainly to the concentrating effect of N(2)O uptake on alveolar PO(2) in moderately low VA/Q compartments. The improvement in Pa(O(2)) when N(2)O was present instead of N(2) was greatest when the degree of VA/Q inhomogeneity was in the range typically seen in anesthetized patients. Models based on distributions of expired and inspired alveolar ventilation give quantitatively different results for Pa(O(2)). In the presence of VA/Q inhomogeneity, second-gas and concentrating effects may have clinically significant effects on arterial oxygenation even at "steady-state" levels of N(2)O uptake.     

Muscle O2 deficit during hypoxia and two levels of O2 demand

We have examined the relative deficits in tension development and O2 uptake in contracting skeletal muscle during severe hypoxic hypoxia. Anesthetized mongrel dogs were ventilated to maintain an end-tidal PCO2 between 35 and 40 Torr. Venous outflow from the gastrocnemius muscle was measured using an electromagnetic flow probe. The tendon was cut and attached to a strain gauge. The muscle was stimulated to contract isometrically at 2 or 4 Hz for 20 min. Hypoxia (9% O2 in N2) was then imposed for 30 min, followed by 30 min of normoxia. Blood flow first increased in proportion to the contraction frequency and then increased further a similar amount in both groups during hypoxia. O2 extraction and blood flow reached maximal levels during hypoxia in the 2-Hz group. The further O2 deficit that was accumulated during 4 Hz and hypoxia was, therefore, a result of the greater discrepancy between O2 supply and demand. O2 uptake decreased more in hypoxia than did developed tension. These results are best explained by ATP supplementation from nonaerobic energy sources that was promoted by the free-flow condition of hypoxic hypoxia.     

Regional muscle blood flow capacity and exercise hyperemia in high-intensity trained rats

The purpose of this study was to determine the effects of high-intensity treadmill exercise training on 1) the regional distribution of muscle blood flow within and among muscles in rats during high-intensity treadmill exercise (phase I) and 2) on the total and regional hindlimb skeletal muscle blood flow capacities as measured in isolated perfused rat hindquarters during maximal papaverine vasodilation (phase II). Two groups of male Sprague-Dawley rats were trained 5 days/wk for 6 wk with a program consisting of 6 bouts/day of 2.5-min runs at 60 m/min up a 15% grade with 4.5-min rest periods between bouts. After training, blood flows were measured with the radiolabeled microsphere technique (phase I) in pair-weighted sedentary control and exercise-trained rats while they ran at 60 m/min (0% grade). In phase II of the study, regional vascular flow capacities were determined at three perfusion pressures (30, 40, and 50 mmHg) in isolated perfused hindquarters of control and trained rats maximally vasodilated with papaverine. The results indicate that this exercise training program produces increases in the vascular flow capacity of fast-twitch glycolytic muscle tissue of rats. However, these changes were not apparent in the magnitude or distribution of muscle blood flow in conscious rats running at 60 m/min, since blood flows within and among muscles during exercise were the same in trained and control rats.     

Lactate metabolism in resting and contracting canine skeletal muscle with elevated lactate concentration.

This study was undertaken to quantitatively account for the metabolic disposal of lactate in skeletal muscle exposed to an elevated lactate concentration during rest and mild-intensity contractions. The gastrocnemius plantaris muscle group (GP) was isolated in situ in seven anesthetized dogs. In two experiments, the muscles were perfused with an artificial perfusate with a blood lactate concentration of ~9 mM while normal blood gas/pH status was maintained with [U-(14)C]lactate included to follow lactate metabolism. Lactate uptake and metabolic disposal were measured during two consecutive 40-min periods, during which the muscles rested or contracted at 1.25 Hz. Oxygen consumption averaged 10.1 +/- 2.0 micromol. 100 g(-1). min(-1) (2.26 +/- 0.45 ml. kg(-1). min(-1)) at rest and 143.3 +/- 16.2 micromol. 100 g(-1). min(-1) (32.1 +/- 3.63 ml. kg(-1). min(-1)) during contractions. Lactate uptake was positive during both conditions, increasing from 10.5 micromol. 100 g(-1). min(-1) at rest to 25.0 micromol. 100 g(-1). min(-1) during contractions. Oxidation and glycogen synthesis represented minor pathways for lactate disposal during rest at only 6 and 15%, respectively, of the [(14)C]lactate removed by the muscle. The majority of the [(14)C]lactate removed by the muscle at rest was recovered in the muscle extracts, suggesting that quiescent muscle serves as a site of passive storage for lactate carbon during high-lactate conditions. During contractions, oxidation was the dominant means for lactate disposal at >80% of the [(14)C]lactate removed by the muscle. These results suggest that oxidation is a limited means for lactate disposal in resting canine GP exposed to elevated lactate concentrations due to the muscle's low resting metabolic rate.     

Gravity is an important but secondary determinant of regional pulmonary blood flow in upright primates

Original studies leading to the gravitationalmodel of pulmonary blood flow and contemporary studies showinggravity-independent perfusion differ in the recent use of laboratoryanimals instead of humans. We explored the distribution ofpulmonary blood flow in baboons because their anatomy, serialdistribution of vascular resistances, and hemodynamic responses tohypoxia are similar to those of humans. Four baboons wereanesthetized with ketamine, intubated, and mechanically ventilated.Different colors of fluorescent microspheres were given intravenouslywhile the animals were in the supine, prone, upright (repeated), andhead-down (repeated) postures. The animals were killed, and their lungswere excised, dried, and diced into~2-cm³ pieces with the spatialcoordinates recorded for each piece. Regional blood flow was determinedfor each posture from the fluorescent signals of each piece. Perfusionheterogeneity was greatest in the upright posture and least when prone.Using multiple-stepwise regression, we estimate that 7, 5, and 25% ofperfusion heterogeneity is due to gravity in the supine, prone, andupright postures, respectively. Although important, gravity is not thepredominant determinant of pulmonary perfusion heterogeneity in uprightprimates. Because of anatomic similarities, the same may be true for humans.     

Microdialysis and the measurement of muscle interstitial K+ during rest and exercise in humans.

The purpose of this study was to examine whether microdialysis and the internal reference thallium-201 ((201)Tl) could accurately measure muscle interstitial K+ (Ki+) before, during, and after exercise. The relative loss of (201)Tl and simultaneous relative recovery of K+ were measured in vitro for 12 microdialysis probes that were bathed in Ringer acetate medium and perfused at various flows (3-10 microl/min). (201)Tl loss was linearly related to K+ recovery, and their level of agreement was not different from zero. Microdialysis and (201)Tl were then used to measure Ki+ in the gastrocnemius medialis muscle of four humans during rest and static plantar flexion exercise. At rest, Ki+ was 3.9-4.3 mmol/l when the perfusate flow was 2 or 5 microl/min. During exercise, Ki+ increased from 6.9 +/- 0.4 to 7.5 +/- 0.3 mmol/l at low to high intensity and declined to 5.2 +/- 0.3 mmol/l after exercise. These results suggest that large changes in Ki+ in human skeletal muscle can be accurately measured by using microdialysis and (201)Tl.     

Role of convective O(2) delivery in determining VO(2) on-kinetics in canine muscle contracting at peak VO(2).

A previous study (Grassi B, Gladden LB, Samaja M, Stary CM, and Hogan MC, J Appl Physiol 85: 1394-1403, 1998) showed that convective O(2) delivery to muscle did not limit O(2) uptake (VO(2)) on-kinetics during transitions from rest to contractions at approximately 60% of peak VO(2). The present study aimed to determine whether this finding is also true for transitions involving contractions of higher metabolic intensities. VO(2) on-kinetics were determined in isolated canine gastrocnemius muscles in situ (n = 5) during transitions from rest to 4 min of electrically stimulated isometric tetanic contractions corresponding to the muscle peak VO(2). Two conditions were compared: 1) spontaneous adjustment of muscle blood flow (Q) (Control) and 2) pump-perfused Q, adjusted approximately 15-30 s before contractions at a constant level corresponding to the steady-state value during contractions in Control (Fast O(2) Delivery). In Fast O(2) Delivery, adenosine was infused intra-arterially. Q was measured continuously in the popliteal vein; arterial and popliteal venous O(2) contents were measured at rest and at 5- to 7-s intervals during the transition. Muscle VO(2) was determined as Q times the arteriovenous blood O(2) content difference. The time to reach 63% of the VO(2) difference between resting baseline and steady-state values during contractions was 24.9 +/- 1.6 (SE) s in Control and 18.5 +/- 1.8 s in Fast O(2) Delivery (P < 0.05). Faster VO(2) on-kinetics in Fast O(2) Delivery was associated with an approximately 30% reduction in the calculated O(2) deficit and with less muscle fatigue. During transitions involving contractions at peak VO(2), convective O(2) delivery to muscle, together with an inertia of oxidative metabolism, contributes in determining the VO(2) on-kinetics.     

Selected contribution: redistribution of pulmonary perfusion during weightlessness and increased gravity.

To compare the relative contributions of gravity and vascular structure to the distribution of pulmonary blood flow, we flew with pigs on the National Aeronautics and Space Administration KC-135 aircraft. A series of parabolas created alternating weightlessness and 1.8-G conditions. Fluorescent microspheres of varying colors were injected into the pulmonary circulation to mark regional blood flow during different postural and gravitational conditions. The lungs were subsequently removed, air dried, and sectioned into approximately 2 cm(3) pieces. Flow to each piece was determined for the different conditions. Perfusion heterogeneity did not change significantly during weightlessness compared with normal and increased gravitational forces. Regional blood flow to each lung piece changed little despite alterations in posture and gravitational forces. With the use of multiple stepwise linear regression, the contributions of gravity and vascular structure to regional perfusion were separated. We conclude that both gravity and the geometry of the pulmonary vascular tree influence regional pulmonary blood flow. However, the structure of the vascular tree is the primary determinant of regional perfusion in these animals.