Coordination of wingbeat and respiration in birds. II. “Fictive” flight GREGORY Department
D. FUNK,
of Zoology,
JOHN D. STEEVES, AND University of British Columbia,
WILLIAM Vancouver,
K. MILSOM British Columbia
V6T 2A9, Canada
FUNK, GREGORY D., JOHN ID. STEEVES, AND WILLIAM K. of wingbeat and respiration in birds. II. MILSOM. Coordination “Fictiue” flight. J. Appl. Physiol. 73(3): 1025-1033, 1992.-To determine whether an interaction between central respiratory and locomotor networks may be involved in the observed coordination of wingbeat and respiratory rhythms during free flight in birds, we examined the relationship between wingbeat and respiratory activity in decerebrate Canada geese and Pekin ducks before and after paralysis. Locomotor activity was induced through electrical stimulation of brain stem locomotor regions. Respiratory frequency (f,) was monitored via pneumotachography and intercostal electromyogram recordings before paralysis and via intercostal and cranial nerve IX electroneurogram recordings after paralysis. Wingbeat frequency (fw) was monitored using pectoralis major electromyogram recordings before, and electroneurogram recordings after, paralysis. Respiratory and cardiovascular responses of decerebrate birds during active (nonparalyzed) and “fictive” (paralyzed) wing activity were qualitatively similar to those of a variety of vertebrate species to exercise. As seen during free flight, wingbeat and respiratory rhythms were always coordinated during electrically induced wing activity. Before paralysis during active wing flapping, coupling ratios (f,/f,) of l:l, 21, 3:1, and 4:l (wingbeats per breath) were observed. After paralysis, f, and f, remained coupled; however, 111 coordination predominated. All animals tested (n = 9) showed I:1 coordination. Two animals also showed brief periods of 21 coupling. It is clear that locomotor and respiratory networks interact on a central level to produce a synchronized output. The observation that the coordination between f, and fV differs in paralyzed and nonparalyzed birds suggests that peripheral feedback is involved in the modulation of a centrally derived coordination.
has also been implicated in the synchronization of ventilatory pattern with locomotor pattern in rabbits (24,34) and cats (ZO), although the data to support this are equivocal, Locomotor and respiratory patterns were only occasionally coordinated in decorticate and decerebrate rabbits (24, 34), with tight coordination between these two motor systems only occurring in paralyzed spinalized animals (33-35). Furthermore, afferent feedback was not completely removed in these studies. Although vagotomy removed the influence of pulmonary stretch receptor feedback on respiratory pattern, intercostal (26) and diaphragmatic reflex pathways (30) remained intact and could have altered or contributed to the entrainment (26) of respiratory rhythm. The purpose of this study, therefore, was to determine whether the entrainment between wingbeat and respiration observed during free flight in birds could be produced by a central coupling of locomotor and respiratory rhythm generators. Birds offer unique advantages over mammalian preparations for these studies. Because of the unique structure of the avian respiratory system, birds can be unidirectionally ventilated during paralysis, eliminating the phasic afferent feedback associated with the regular mechanical deformation of the respiratory system and chest wall. In addition, problems in defining gait and/or gait transitions (24,33-35) are not a problem with birds, because once the wings are recruited into the locomotor pattern, they always beat synchronously (27-29).
locomotor-respiratory coupling; forward; flight; goose
MATERIALS
entrainment;
exercise;
feed-
Animal Preparation
relaying information regarding wingbeat rhythm appears to play an important role in the coordination of wingbeat and respiration during flight in Canada geese(16). It remains unclear, however, whether such afferent feedback is essential for the coordination of wing and respiratory movements. The two motor systems may also interact centrally to produce a coordinated output in the absence of feedback activity. This type of feedforward neurogenic drive has been shown to regulate ventilatory drive during exercise in cats (II). Neural impulses arising from the suprapontine brain that command muscle to exercise appear to “irradiate” to the respiratory centers in the medulla and cause a neurally driven increase in ventilation that is proportionate to the increase in metabolic rate. Feedforward AFFERENT
AND METHODS
FEEDBACK
0161-7567/92 $2.00 Copyright
Experiments were carried out on seven Pekin ducks (Anus platyrhynchos; 2.63 * 0.08 kg body wt) and nine Canada geese (Branta canadensis; 3.6 k 0.2 kg body wt). The surgical procedures used in these experiments for ducks and geese were similar to those described previously (14). Briefly, the animals were tracheostomized to deliver gaseous anesthesia, monitor ventilation via pneumotachography, and administer unidirectional ventilation (UDV) after paralysis. A clavicular air sac was also cannulated to provide an outflow path for the air during UDV. It was plugged during spontaneous breathing. A carotid artery and jugular vein were cannulated to monitor blood pressure and administer the paralytic agent. The birds were then supported in a sling overlying a variable-speed motor-driven treadmill. They were positioned with their legs on the treadmill and wings folded at their
0 1992 the American
Physiological
Society
Downloaded from www.physiology.org/journal/jappl at Midwestern Univ Lib (132.174.254.157) on February 12, 2019.
1025
1026
CENTRAL
COUPLING
OF
WINGBEAT
sides. The treadmill was required in these studies because walking behavior was often elicited by electrical stimulation of brain stem locomotor regions before a region that would induce wing flapping could be located. A thermometer was inserted -25 cm down the esophagus to monitor body temperature (Tb). T, was maintained constant at 41OC by use of a heat lamp and a 15-cm Ushaped copper tube inserted via the cloaca into the intestine, through which hot or cold water was circulated. The head of each bird was placed in a stereotaxic head holder, and a craniotomy was performed, followed by decerebration. The transection level extended from the caudal border of the habenular nucleus dorsally to the rostra1 border of the anterior preoptic nucleus ventrally (31). Transection at more caudal levels reduced the occurrence of electrically stimulated locomotor activity after paralysis. After completion of the decerebration, the nerve innervating the external (inspiratory) or internal (expiratory) intercostal muscles of the sixth costal space and the nerve innervating the pectoralis major muscle were isolated for recording of “fictive” respiration and wing flapping, respectively, after paralysis. Cranial nerve IX was also isolated in three animals, and its activity was used as another indicator of fictive breathing. Electromyographic (EMG) electrodes were implanted into internal and/or external intercostal muscles of the fifth and sixth costal spaces and the right and left pectoralis major muscles to monitor intercostal muscle activity and wingbeat frequency during electrically induced wing flapping. The equipment used to record respiratory, cardiovascular, and EMG data has been described elsewhere (14). After these surgical procedures, anesthesia was discontinued and the birds were allowed 21 h to recover. Brain stem stimulation (monopolar electrodes, Kopf SNE 300 X 75; pulse duration 2 ms; pulse frequency 60 Hz, stimulation intensity range 30450 PA) was then initiated using procedures described in detail elsewhere (31). High-intensity stimulation (100 PA) was used to localize a stimulation site within the brain stem (Fig. 1) that would elicit wing flapping. Stimulation intensity was then decreased to zero and slowly increased to establish the current threshold necessary to invoke wing flapping. Trials were then carried out using stimulation intensities -20% greater than threshold. The suprathreshold stimulation avoided the delay between onset of stimulation and start of locomotion often associated with stimulation at threshold levels. Artificial
Ventilation
After completion of the protocol (see below) on the nonparalyzed animals and before the administration of the paralytic agent, the birds were unidirectionally ventilated with 4% CO, in air at flow rates of 1.5-2.0 l/min. Preliminary experiments (n = 3) indicated that this flow rate and concentration of inspired gas maintained arterial blood PO, (108.5 t 4.0 Torr), PCO, (27.1 * 0.36 Torr), and pH (7.41 -t- 0.01) near the normal levels seen in ducks (6). Recent values reported for bar-headed geese (lZ), however, indicate that the geese in this study may have been slightly hvperventilated. Nonetheless, these flow
AND
RESPIRATION
FIG. 1. Composite diagram of cross sections through avian brain stem showing stimulation sites used to initiate either active (closed symbols) or fictive (open symbols) wing flapping in ducks (triangles) and geese (squares). Stimulation sites effective in evoking both active and fictive locomotion are designated with closed diamonds (ducks, n = 3) and closed circles (geese, KJ = 1). A-F run caudorostral (A, medulla; B and C, pons; D-F, midbrain). Stimulation sites are indicated at Leftof each cross section. Anatomic structures are labeled on right. AL, ansa lenticularis; AQ, cerebral aqueduct; BC, brachium conjuctivum; Cnd, nucleus reticularis medullaris pars dorsalis; Cnv, nucleus reticularis medullaris pars ventralis; DBC, decussation of brachium conjunctivum; EM, ectomammillary nucleus; EW, Edinger-Westphal nucleus; ICo, nucleus intercollicularis; III, oculomotor nucleus; IO, inferior olivary nucleus; IP, interpeduncular nucleus; IV, trochlear nucleus; LC, locus ceruleus; MLd, lateral mesencephalic nucleus; MLF, medial longitudinal fasciculus; MRF, mesencephalic reticular formation; MV, motor trigeminal nucleus; NIII, oculomotor nerve; NIV, trochlear nerve; NV, trigeminal nerve; NX, vagus nerve; OT, optic tectum; R, raphe nucleus; RP, pontine reticular formation; Rpc, pontine nucleus, parvocellular part; RPO, pontine reticular nucleus, oral part; Ru, red nucleus; ST, subtrigeminal nucleus; SV, trigeminal sensory nucleus; TPc, substantia nigra, pars compacta; TS, tractus solitarius; TTD, nucleus and tract of descending trigeminal nerve; VII, facial nucleus; X, motor nucleus of vagus; XII, hypoglossal nucleus.
rates were sufficient to maintain adequate blood gases (>1.2 l/min) yet low enough to prevent hyperventilationinduced apnea and reductions in T, (~2.5 l/min). Once UDV was established, the birds were paralyzed with an initial injection of gallamine triethiodide (Flaxedil; LO mg/kg iv). Supplemental doseswere administered as required. Nerue Recordings After paralysis, the left intercostal (either internal or external) and the left pectoralis nerves were reexposed. Tissue wells were formed around each exposed nerve and filled with mineral oil. The nerves, submerged in mineral oil, were placed on bipolar platinum hook electrodes. Under these recording conditions, the position of the birds was the same as that before paralysis. In the three animals from which cranial nerve IX activity was recorded, the bodv position of the birds was not altered:
Downloaded from www.physiology.org/journal/jappl at Midwestern Univ Lib (132.174.254.157) on February 12, 2019.
CENTRAL
COUPLING
OF
WINGBEAT
however, the head was rotated sideways (lateral surface facing up) to facilitate formation of a tissue well. Preliminary experiments (n = 5) confirmed the previous finding (18) that intercostal nerve discharge provides a good index of respiration in birds and that unilateral recordings reliably indicate respiratory pattern. Discharge of right and left external intercostal nerves from the fifth and sixth costal spaces was always synchronous and in phase with inspiratory airflow during spontaneous ventilation. Right and left internal intercostal nerves discharged synchronously in phase with expiration. In addition, the discharge of external and internal intercostal nerves, whether ipsi- or contralateral, was always out of phase. Because the intercostal muscles may be recruited to serve a locomotor function during wing flapping, the discharge of cranial nerve IX was also examined as an index of respiration to ensure that the intercostal electroneurogram (ENG) truly represented central respiratory activity. In mammals, cranial nerve IX innervates the stylopharyngeus (an airway dilator active during inspiration) and glossopharyngeus (an airway constrictor sometimes active during expiration) muscles of the upper airway (17). Inspiratory activity in cranial nerve IX was determined by comparison with intercostal ENG activity at rest. Although bilateral recordings were never made from cranial nerve IX, its activity was always found to be in phase with external intercostal nerve discharge (i.e., active during inspiration) and/or out of phase with internal intercostal nerve discharge (n = 3). Bilateral recordings of pectoralis nerves were not taken during these experiments, because previous recordings made during spontaneous and electrically and chemically induced fictive wing flapping (27) indicated that discharge of left and right pectoralis nerves was always synchronous. All neurogram signals were amplified (X10,000), Cltered, and recorded in the same manner as EMG signals. One channel of EMG was recorded to ensure that paralysis was complete during all trials. Stimulation trials (100-400 PA) were then resumed, this time recording ENG activity rather than EMG activity. At the conclusion of each experiment, the position of the stimulating electrode was marked with an electrolytic lesion (3 mA/5 s). Histological identification of stimulation sites was performed as described previously (31). Protocol Nonparalyzed birds. Five ducks and nine geese were examined to describe the relationship between airflow and intercostal and pectoralis muscle activity during electrically induced wing flapping. Minute ventilation (VE), ventilatory frequency (f,>, tidal volume (VT), blood pressure (BP), heart rate (HR), internal or external EMG activity, and right and left pectoralis EMG activity were recorded for 30 s prestimulation, throughout stimulation, and for up to 3 min poststimulation. Most stimulation periods lasted IO-30 s. However, in two animals the duration was increased to 2 min to further describe the kinetics of the ventilatory response to electrically induced wing flapping (see Fig. 4). Paralyzed birds. Activities of the pectoralis nerve (fic-
AND
1027
RESPIRATION
TABLE 1. Different ratios of wingbeat frequency to ventilation frequency obtained with stimulation at various sites Active Stimulation Site
TTD/Cnd RP MRF ICO
Flapping
I:1
2:l
311
4:l
1:l
4 1
1 1
1
2 1
2
1
1 1 2 5
Fictive
Flapping
2:l
3:l
411
2
Values represent no. of animals (geese or ducks) demonstrating each wingbeat frequency-to-ventilation frequency ratio. Some birds showed >I ratio. TTD, nucleus and tract of descending trigeminal nerve; Cnd, nucleus reticularis medullaris, pars dorsalis; RP, nucleus reticularis pontis caudalis, pars gigantocellularis; MRF, mesencephalic reticular formation; ICo, nucleus intercollicularis.
tive flight) and internal or external intercostal nerve or cranial nerve IX (fictive respiration) were recorded for 30-60 s prestimulation, throughout stimulation (range IO-30 s), and for up to 2 min poststimulation in seven ducks and two geese. Pectoralis EMG activity, HR, and BP were also recorded. Five of seven ducks and two of nine geesewere examined before and after paralysis; however, there was no indication that those birds studied only before or only after paralysis differed from those studied under both conditions. Data Analysis Unless otherwise indicated, values are means -t SE. Analysis of variance was used to test the difference between means. P < 0.05 was assumed to be significant. RESULTS
Nonparalyzed Responses Brain stem stimulation successfully initiated active wing flapping in all five ducks and six of nine geese. All effective stimulation sites fell within areas described previously (27-29, 31) and are summarized in Table 1 and Fig. 1. No effort was made to initiate locomotion from each site listed in Table 1 for each animal. Thus the data in Table 1 do not indicate the relative effectiveness of each site at evoking locomotion. Stimulation sites were similar for ducks and geese, The stimulation intensity necessary to evoke wing flapping averaged 137 t 23 PA. During the course of the experiments, a variety of wingbeat flight patterns were observed. Figure 2 shows an entire sequence from one duck. In this particular trace, threshold stimulation was used, rather than the usual suprathreshold stimulus, which activates locomotion with little or no delay. This trace demonstrates that the dramatic alteration of respiratory pattern and increase in BP and HR that occur immediately with the onset of stimulation were not dependent on the development of wing activity. With the onset of stimulation, VT decreased from 33.3 to 16.3 ml/kg, but the dramatic increase in f, from 18.2 to 135 breaths/min observed at the onset of stimulation resulted in a large increase in VE from 606 to 2,194 ml min-l. kg? Similarly, HR increased at the onset of stimulation from 200 beats/min at rest to 513 beats/min 7 s later when wing flapping coml
Downloaded from www.physiology.org/journal/jappl at Midwestern Univ Lib (132.174.254.157) on February 12, 2019.
1028
CENTRAL
Blood
COUPLING
OF
WINGBEAT
AND
RESPIRATION
Pressure
stim. on
/
c
/
c
c
‘\ stim. off
0
2
S8C
\
FIG. 2. Sequence of events associated with wing flapping init.iated using threshold current (90 PA) in a decerebrate duck. Prestimulation and recovery records are included. Heart rate, mean blood pressure, left and right pectoralis EMG, and integrated respiratory airflow (labeled airflow, inspiration up and expiration down) are shown. Inset: expanded view of outlined section illustrating precise 21 relationship between pectoralis EMG activity and respiratory rhythm. (Unlike Fig. 5, airflow integrator was operating in reset mode. Thus, trace returned to baseline whenever flow direction changed.)
menced. Mean BP also increased over the first 7 s from 122 mmHg at rest to a maximum of 175 mmHg. By the onset of wing activity 7 s later, iiE, BP, and HR had already plateaued at these elevated levels and did not increase further. With the termination of brain stem stimulation and wing flapping, all variables gradually returned toward prestimulation levels. Similar changes in respiratory pattern during the transition from rest to wing flapping can be seen in Fig. 5. In this case, however, suprathreshold stimulation was used, and the changes in respiratory pattern preceded the onset of wing flapping by
10
WINGBEAT
rest
m stlm ESSI pora, red sttm
EZJpafa,
-+ 800
-
(mm-‘1 ‘;
400~
HA
4
AND
1029
RESPIRATION
only 60% of the level achieved preparalysis. As a result, the relationship between f, and f, also changed during fictive flight. After paralysis, all nine birds (7 ducks, 2 geese) demonstrated I:1 synchronization. Two birds (1 duck, I goose) also showed brief periods of coordination at 21. No other coordination schemes were observed after paralysis (Table 1). An example of I:1 coordination during fictive flight is shown in Fig. 6. With the initiation of fictive flight, there was a dramatic increase in fV from 6.9 bursts/min at rest (top) to 92.5 bursts/min as f, increased to match f, 1:l. During periods of 1:l coordination, the phase relationship between f, and f, was constant, Inspiratory nerve activity [external intercostal nerve (rz = 3) and cranial nerve IX activity (n = 3)] occurred out of phase with pectoralis activity (Fig. 6), while expiratory nerve activity [internal intercostal nerve (n = 4)] occurred in phase with pectoralis ENG discharge (Fig. 7). DISCUSSION
Critique
100
of Methods
Validity of decerebrate preparations: free flight vs. electrically induced wing flapping. Locomotion in decerebrate animals, whether spontaneous or produced by electrical/ chemical stimulation, has been shown to be similar to
50
-
-0
3. Minute ventilation (iiE), tidal volume (VT), breathing frequency (f,), heart rate (HR), blood pressure (BP), and wingbeat frequency (fw) for all decerebrate birds (ducks and geese; sample size indicated by no. over each column) before paralysis at rest, (rest), during electrically induced wing flapping (stim), after paralysis at rest (para, rest), and during fictive wing flapping (para, stim). Values are means -t- SE* Significant differences between groups: “preparalysis, rest vs. seim; ‘para, rest vs. para, stim; +stim vs. para, stim. FIG.
actual wing flapping, the results obtained for ducks and geese during fictive flight were virtually identical. Again, all effective stimulation sites fell within brain stem regions described previously (27-29, 31) and are summarized in Table 1 and Fig. 1. The average stimulation intensity used to evoke fictive wing flapping was 340 t 46 PA. Although fictive flight could be initiated from the more caudal brain stem stimulation sites (nucleus and tract of descending trigeminal nerve; nucleus reticularis medullaris, pars dorsalis), the more rostra1 sites (nucleus reticularis pontis caudalis, pars gigantocellularis; mesencephalic reticular formation; nucleus intercollicularis) were most effective in initiating fictive activity. Stimulation sites used to initiate flapping pre- and postparalysis were not always the same, Of the six birds (5 ducks, I goose) examined both before and after paralysis, the same stimulation sites were used pre- and postparalysis in four birds (3 ducks, 1 goose). Paralysis itself did not significantly alter HR, BP, or f, (Fig. 3). The magnitude of the increases from resting levels in HR, BP, and f, during fictive flapping were also similar to those seen during actual wing flapping (Fig. 3). The increase in wingbeat frequency (f,), however, was sienificantlv less than that before Daralvsis. increasing to
W
l>
300
-40
0
40
80
Time
120
160
200
240
(set)
FIG. 4. Kinetics of ventilatory response of 2 decerebrate geese to 2 min of electrically induced wing flapping showing VE, VT, and f,. Stimulation was turned on at t = 0 and off t = 120 s.
Downloaded from www.physiology.org/journal/jappl at Midwestern Univ Lib (132.174.254.157) on February 12, 2019.
1030
CENTRAL
R. Pect.
EMG
L. Pmt.
EMG
L. External
Air
ht.
COUPLING
OF
WINGBEAT
AND
RESPIRATION
EMG
Insp.
T
Exp.
J
Flow
R. Pect.
EMG
L. Pect.
EMG
L. External
Air Flow
ht.
FIG. 5. Relationship between right and left pectoralis EMG activity, left external intercostal EMG activity, and integrated respiratory airflow (airflow) at rest (top) and during electrically induced wing flapping in a decerebrate duck (bottom). Note different time scales. (Respiratory integrator was operating in sum mode. Thus, trace did not reset with changes in flow direction as in Fig. 2.)
EMG
Insp.
T
Exp.
& stim. on
t
stim. off t 1
I 2 sec.
locomotion in intact animals for a large number of species (ZZ), including geese (31) and chicks (329, during wing flapping. Similarly, the respiratory pattern produced by decerebrate animals is very similar to that seen in intact animals (I3), including geese (14). Furthermore the respiratory responses of decerebrate geese to low-level exercise (treadmill walking) closely resemble the responses of intact geese (14). Because the physiological responses to hindlimb vs. forelimb exercise can differ (I), however, it was essential in this study to compare the
responses of decerebrate geese during electrically induced wing flapping with the responses of intact animals during free flight. The physiological responses of decerebrate birds to electrically induced wing flapping were considerably smaller than the responses of intact birds to free flight: f, was approximately one-half the level seen during free flight. However, assuming that metabolic rate is a function of both f, and the force of muscular contraction, a reduction in metabolic rate between free flight and elec-
Cranial Nerve IX ENG
Cranial Nerve IX ENG
L. Pect. ENG stim.
on t
stim. off T I
1 5 sec.
FIG.
6. Recording
(top) and during synchronization
of left cranial nerve IX (inspiratory activity; fictive respiration) and left pectoralis ENG at rest electrically induced fictive wing flapping (bottom) in a decerebrate paralyzed duck showing precise 1:l between the 2 patterns during fictive flight. Note different time scales.
Downloaded from www.physiology.org/journal/jappl at Midwestern Univ Lib (132.174.254.157) on February 12, 2019.
CENTRAL
COUPLING
OF
WINGBEAT
AND
RESPIRATION
1031
metabolic rate and response magnitude between intact and decerebrate animals by increasing the stimulus intensity used to evoke locomotion and thus the force of wing flapping. However, increased stimulation intensity L. Pect ENG caused a dramatic increase in metabolic heat production. 2 Lc. Highly efficient nonevaporative convective heat loss, which is responsible for 65100% of the cooling in birds during flight (5), is difficult to achieve in the laboratory. Thus, to avoid the problem of T, regulation, low-level stimulation was used to evoke locomotion and most stimulation periods were kept below 30 s. L. Pect. ENG In summary, the responses of decerebrate birds during electrically induced wing flapping appeared qualitatively similar to those anticipated to occur in intact geese during free flight. The cardiovascular adjustments observed during electrically induced wing flapping approximated those seen during exercise in a variety of vertebrate species. The kinetics of the ventilatory response and the breathing pattern responses of geese to electrically intrically induced wing flapping was expected. In turn, if duced flapping were also similar to those of free flight. the relationship between VE and 0, consumption seen Paralyzed preparations: active wing flapping vs. fictive during walking in decerebrate geese (14) also holds for flight. The physiological responses of birds during fictive flight were remarkably similar to those during active wing flapping, a reduction in TjE would also be expected. The 2,5-fold increase in VE in decerebrate birds was ap- wing flapping. As seen previously in cats (23), the cardioproximately one-fourth of the average 123-fold increase vascular responses of paralyzed and unparalyzed birds to in V, recorded from several species during free flight. fictive vs. active locomotion were very similar. In addiThus the difference in the magnitude of the ventilatory tion, although increases in f, were substantial (4%fold) response between free flight and electrically induced during fictive flight, they were less than increases during wing flapping does not appear to be due to a decrease in active flapping. Estimates of VT and, hence, overall levthe responsiveness of the decerebrate birds but more a els of ventilatory effort (7jE) could not be made for the result of the decerebrate animals performing at lower paralyzed birds during fictive flight. However, estimates metabolic rates than free-flying animals. of TjE in mammals from integrated phrenic nerve activity This suggestion is supported by three additional ob- indicate that VE increases similarly during fictive and servations. First, the kinetics of the ventilatory response active locomotion (11). to wing-flapping exercise in decerebrate birds (Fig. 4) Similar to f,, f, was reduced during fictive flight (86 t were similar to those of intact birds [black duck (3), pi- 21 beats/min) relative to active wing flapping (144 t 19 geon (9)]. Both showed a rapid increase in VE with the beats/min). A reduction in locomotor frequency after paonset of flapping, followed by a plateau and a rapid re- ralysis has been consistently observed in a variety of speturn toward resting levels with termination of flight. Sec- cies [lamprey (36), stingray (37), duck and goose (27), cat ond, like the ventilatory response of decerebrate birds, (19)] and apparently results from the removal of an excitthe increase in HR was diminished in decerebrate birds atory drive to the locomotor pattern generators that is normally provided by phasic afferent feedback. This derelative to free-flying intact animals. HR, which typically increases between 2 and 5%fold during free flight (3,9), crease in central excitability of the locomotor system is increased only L5-fold in the decerebrate birds, again also evident in that the stimulation intensity required to approximately one-fourth of the increase during free activate fictive locomotion in paralyzed birds and other flight. Third, the 1.46-fold increase in HR in the decere- species was much greater than that required to activate brate birds during wing flapping was associated with an locomotion before paralysis (19, 27, 36, 37). Once locoalmost identical 1.5-fold increase in BP from 110 to 164 motion is activated in the fictive preparation, however, mmHg. Similar increases in BP have been recorded in the locomotor pattern of the fictive preparation is very similar to that of the unparalyzed animal (19, 22, 27). intact ducks running on a treadmill at work rates that produced similar levels of ventilation [2.6 times resting 0, consumption (Z)]. The only measurements of BP durMechanisms of Entrainment ing flight in intact birds indicate only a small increase in BP from 142 to 147 mmHg in pigeons (9). However, deCoordination during fictive flight. The mechanisms inspite efforts to reduce stress in these birds, the absence of volved in controlling the gross cardiovascular and respiratory responses to exercise can be divided into two catea pressor response in these birds appeared to be due to elevated preflight BP relative to previously recorded gories: 1) feedback from humoral as well as neurogenic “resting” values (127 t 6 mmHg) (8). In fact, most ani- drives occurring after the start of exercise and 2) feedformals [humans (38), dogs (Zl), ducks (a)], including de- ward due to central coactivation of cardiovascular and cerebrate cats (II, 23), demonstrate an increase in BP respiratory regions in the brain stem by pathways involved in the initiation of locomotion. There is clear evithat is proportional to work rate during exercise. It would have been possible to reduce the differences in dence that feedforward mechanisms from locomotor L. Internal
ht.
ENG
#
Downloaded from www.physiology.org/journal/jappl at Midwestern Univ Lib (132.174.254.157) on February 12, 2019.
1032
CENTRAL
COUPLING
OF
WINGBEAT
centers are involved in the production of a proportionate drive to ventilation during exercise (Fig. 3) (11). In addition, our results provide clear evidence in birds that feedcenters not only ac tivity from the locomotor forward drive to ven tilation but is inprovides a proportional locomotor and respiratory in coordinating volved rhythms. Previous work with cats (20) and rabbits (24, 51) is suggestive of this finding; however, several factors confounded interpretation of these data. Kawahara et al. (20) observed synchronization of central locomotor and respiratory outputs in paralyzed decerebrate cats at extremely low arterial PCO, levels, when the frequency of the respiratory pump was very high (i.e., similar to the elicited locomotor frequency). However, entrainment decreased dramatically at higher arterial PCO, levels when the pump frequency was decreased. Although the cats were vagotomized to remove any vagally mediated respiratory entrainment (25), phrenic and intercostal reflexes remained partially intact and have been shown to significantly alter (26, 3O), and even entrain, respiratory rhythm (26). The strength of the central coupling between locomotor and respiratory rhythm-generating networks remains unclear in this case, because the coordination of locomotor and respiratory rhythms may have been enhanced by feedback associated with the pump ventilation. Similarly, the relationship between central locomotor and respiratory neural outputs from paralyzed decorticate and decerebrate rabbits ind cates that these two networks are only weakly coordinated (24,34). It is possible, however, that this low degree of central coordination was due to the fact that rabbits show gai t-dependent entrainment of re spiration and locomotion (7); i.e., entrainment is very low in these animals during walking but virtually 100% with the transition to trot or gallop when they breathe once per leap. Unilateral recordings of leg ENG activity were typically used to describe locomotor pattern. Because such records do not distinguish between and walking (i.e., alternating hindlimb nerve activity) hopping (i.e., synchronous hindlimb nerve activity), gait transitions, which are common during pharmacological (28, 29), may have occurred. activati .on of locomotion Failure to detect such gait changes would have produced a drastic underestimation of the degree to which locomotion and respiration are coordinated at, the central level. Thus it remai .ns unclear whether th .e inte rmittent ten .tral coord in .ation between 1ocomotion and respiration observed in rabbits is due to changes in fictive gait or weak central coupling. Evidence for tight central coordination of locomotor and respiratory networks has been found only in rabbits spinalized at C, (33-35). That locomotor output can be generated in the isolated spinal cord is well established. However, whether phrenic discharge in spinal animals represents respiration remains unclear (13), because the neural networks responsible for respiratory rhythmogenesis are predominantly located in the medulla and caudal pons (13). The present study was designed to eliminate several of these complicating factors. Unidirectional ventilation of paralyzed birds eliminated all phasic afferent feedback
AND
RESPIRATION
associated with pump ventilation, permitting more direct examination of the central relationship between locomotor and respiratory rhythm-generating networks than possible with mammalian preparations. Furthermore, because birds do not show gait changes during flight/ wing flapping, the confounding influences of gait-dependent entrainment were avoided. Our results provide the strongest evidence to date that central mechanisms are involved in coordinating locomotion and respiration. All nine animals examined during fictive flight showed a 1: 1 synchronization (2 birds also showed brief periods of 21 coupling) between the central neural outputs of the locomotor/wing and respiratory systems. The external intercostal nerve (inspiratory) discharged out of phase with the pectoralis nerve, and the internal intercostal nerve (expiratory) discharged in phase with the pectoralis nerve (Fig. 7). In addition, inspiratory activity in cranial nerve IX coordinated with pectoralis ENG during fictive flight in exactly the same manner as the external intercostal ENG (Fig+ 6), ruling out the possibility that the intercostal nerve discharge observed during fictive wing beating was locomotor in nature (Fig. 6). Thus it is clear that the neural networks responsible for the production of locomotor (wingbeat) and respiratory rhythms in birds interact centrally to produce a synchronized output in the absence of phasic afferent feedback. Coordination during free flight. The 1:l relationship between wingbeat and respiration observed in paralyzed birds differed from the range of coupling ratios seen during active wing flapping (I:1 to 4:l) when phasic mechanoreceptive feedback was present (i.e., before paralysis). Although the stimulation sites used to activate locomotor activity were not always the same before and after paralysis, the differences in coupling ratio did not appear to be a function of the stimulation sites used to activate wing flapping (Table 1). Rather, it appears that phasic afferent feedback somehow modulates the centrally derived coordination to produce a range of coupling ratios. These data agree with observations on free-flying birds, where Canada geese have shown 2 and 3 wingbeats/ breath (15) and barnacle geese have shown 2, 3, and 4 wingbeats/breath (IO). In fact, observations on a variety of bird species indicate that the majority of species demonstrate a range of coupling ratios during free flight (4). Conclusion It is clear that the networks responsible for the generation of locomotor and respiratory rhythms interact on a central level (brain stem or spinal cord), in the absence of phasic afferent feedback, to produce synchronized outputs in birds. Addition of afferent feedback does not disrupt this coordination but changes the number of wingbeats per breath. Thus, both feedforward mechanisms and mechanical feedback (16) appear to be involved in the production and control of entrainment. Under the labile conditions of free flight, when the wingbeat rhythm is constantly modified to compensate for environmental perturbations, it is most probable that the coordination of wingbeat and respiration is produced and maintained via a collective interaction of feedforward lo-
Downloaded from www.physiology.org/journal/jappl at Midwestern Univ Lib (132.174.254.157) on February 12, 2019.
CENTRAL
COUPLING
OF
WINGBEAT
comotor inputs and mechanoreceptive feedback on some portion of the respiratory rhythm-generating network. This work was supported by grants from the Natural Sciences and Engineering Research Council (NSERC) of Canada (W. K. Milsom and J. D. Steeves). G. D. Funk was supported by Fellowships from NSERC of Canada and the Killam Foundation. Address for reprint requests: G. D. Funk, Systems Neurobiology Laboratory, Dept. of Physiological Science, UCLA, 405 Hilgard Ave., Los Angeles, CA 90024-1527. Received
22 January
1991; accepted
in final
form
30 March
1992.
REFERENCES 1. ASMUSSEN, E., AND M. NIELSEN. Studies on the regulation of respiration in heavy work. Actu Physiol. Stand. 10: 270-285, 1946. 2. BECH, C., AND S. NOMOTO. Cardiovascular changes associated with treadmill running in the Pekin duck. J. Exp. Biol. 97: 345-358, 1982. 3. BERGER, M., J. S. HART, AND 0.2. ROY. Respiration, oxygen consumption and heart rate during rest and flight. 2. Vergl. Physiol. 66: 201-214, 1970. 4. BERGER, M,, 0. Z. ROY, AND J. S. HART. The coordination between respiration and wingbeats in birds. 2. Vergl. Physiol. 66: 190-200, 1970. 5. BERNSTEIN, M. H. Ventilation and respiratory evaporation in the flying crow, Corvus ossifragus. Respir. Physiol. 26: 371-382, 1976. P., N. HILL, AND Y. JAMMES. Ventilatory responses to 6. BOWEROT, CO2 in intact and chronically chemodenervated Peking ducks. Respir. Physiol. 22: 137-156, 1974. D. M., AND D. R. CARRIER. Running and breathing in 7. BRAMBLE, mammals. Science Wash. DC 219: 251-256, 1983. P. J., AND E. W. TAYLOR. Responses of the respiratory 8. BUTLER, and cardiovascular systems of chickens and pigeons to changes in Pao, and PacoZ. Respir. Physiol. 21: 351-363, 1974. P. J., N. H. WEST, AND D. R. JONES. Respiratory and 9. BUTLER, cardiovascular responses of the pigeon to sustained level flight in a wind tunnel. J. Exp. Biol. 71: 7-26, 1977. 10. BUTLER, P. J., AND A. J. WOAKES. Heart rate, respiratory frequency and wingbeat frequency of free flying barnacle geese Brantu leucopsis. J. Exp. Biol. 85: 213-226, 1980. 11. ELDRIDGE, F. L., D. E. MILLHORN, J. P. KILEY, AND T. G. WALDROP. Stimulation by central command of locomotion, respiration and circulation during exercise. Respir. Physiol. 59: 313-337, 1985. 12. FEDDE, M. R., J. A. ORR, H. SHAM, AND P. SCHEID. Cardiopulmonary function in exercising bar-headed geese during normoxia and hypoxia. Respir. Physiol. 77: 239-262, 1989. 13. FELDMAN, J. L., J. C. SMITH, D. R. MCCRIMMON, H. H. ELLENBERGER, AND D. F. SPECK. Generation of respiratory pattern in mammals. In: Neural Control of Rhythmic Movements in Vertebrates, edited by A. H. Cohen, S. Rossignol, and S. Grillner. New York: Wiley, 1988, p. 73-100. 14. FUNK, G. D., W. K. MILSOM, G. N. SHOLOMENKO, AND J. D. STEEVES. Role of the telencephalon in the synchronization of locomotor and respiratory frequencies during walking in Canada geese. J. Exp. Biol. 145: 283-301, 1989. 15. FUNK, G. D., W. K. MILSOM, AND J. D. STEEVES. Entrainment of wingbeat and respiration during free-flight, “fictive” flight and passive wing flapping in geese. Sot. Neurosci. Abstr. 15: 391, 1989. 16. FUNK, G. D., J. D. STEEVES, AND W. K. MILSOM. Coordination of wingbeat and respiration in the Canada goose. I. Passive wing flapping. J. Appl. Physiol. 73: 1014-1024, 1992. 17. ISCOE, S. D. Central control of the upper airway. In: Respiratory Function of the Upper Airway, edited by 0. P. Mathew and G. Sant’Ambrogio. New York: Dekker, 1988, p. 138-148. (Lung Biol. Health Dis. Ser.) 18. JONES, D. R., AND 0. S. BAMFORD. Open-loop respiratory chemo-
AND
RESPIRATION
1033
sensitivity in chickens and ducks. Am. J. Physiol. 230: 861-867, 1976. 19. JORDAN, L. M., C. A. PRATT, AND J. E. MENZIES. Locomotion evoked by brainstem stimulation: occurrence without phasic segmental afferent input. Bruin Res. 177: 204-207, 1979. 20 KAWAHARA, K., Y. NAKAZONO, Y. YAMAUCHI, AND Y. MIYAMOTO. Coupling between respiratory and locomotor rhythms during fittive locomotion in decerebrate cats. Neurosci. Lett. 103: 326-332, 1989. 21 . KIRLIN, P. C., M. D. KI?TLESON, AND L. E. JOHNSON. Neurohumoral and cardiopulmonary response to sustained submaximal exercise in the dog. J. Appl. Physiol. 62: 1040-1045, 1987. 22. MCCLELLAN, A. D. Command systems for initiating locomotion in fish and amphibians: parallels to initiation systems in mammals. In: Neural Control of Vertebrate Locomotion, edited by S. Grillner, R. M. Herman, P. S. G. Stein, H. Forsberg, and D. G. Stuart. London: Macmillan, 1986, p. 3-20. 23. MILLHORN, D. E., F. L. ELDRIDGE, T. G. WALDROP, AND J. P. KILEY. Diencephalon regulation of respiration and arterial pressure during actual and fictive locomotion in cat. Circ. Res. 61, Suppl. I: 153-159, 1987. 24. PERSEGOL, L., M. JORDAN, D. VIALA, AND C. FERNANDEZ. Evidence for central entrainment of the medullary respiratory pattern by the locomotor pattern in the rabbit. Exp. Bruin Res. 71: 153-162, 1988. 25. PETRILLO, G. A., L. GLASS, AND T. TRIPPENBACH. Phase locking of respiratory rhythm to a mechanical ventilator. Can. J. Physiol. Pharmacol. 61: 599-607, 1983. 26. REMMERS, J. E., AND I. MARTTILA. Action of intercostal muscle afferents on the respiratory rhythm of anaesthetized cats. Respir. Physiol. 24: 31-41, 1975. 27. SHOLOMENKO, G. N., G. D. FUNK, AND J. D. STEEVES. Locomotor activities in the decerebrate bird without phasic afferent feedback. Neuroscience 40: 257-266, 1990. 28. SHOLOMENKO, G. N., G. D. FUNK, AND J. D. STEEVES. Avian locomotion activated by brainstem infusion of neurotransmitter agonists and antagonists. I. Acetylcholine, excitatory amino acids and substance P. Exp. Bruin Res. 85: 659-673, 1991. 29. SHOLOMENKO, G. N., G. D. FUNK, AND J. D. STEEVES. Avian locomotion activated by brainstem infusion of neurotransmitter agonists and antagonists. II. Gamma-aminobutyric acid. Exp. Bruin Res. 85: 674-681, 1991. 30. SPECK, D. F., AND W. R. REVELETTE. Attenuation of phrenic motor discharge by phrenic motor nerve afferents. J. Appl. Physiol. 62: 941-945, 1987. 31. STEEVES, J. D., G. N. SHOLOMENKO, AND D. M. S. WEBSTER. Stimulation of the pontomedullary reticular formation initiates locomotion in decerebrate birds. Brain Res. 401: 205-212, 1987. 32. VALENZUELA, J. I., S. J. HASAN, AND J. D. STEEVES. Stimulation of the brainstem reticular formation evokes locomotor activity in embryonic chicken (in ovo). Dev. Bruin Res. 56: 13-18, 1990. 33. VIALA, D. Evidence for direct reciprocal interactions between the central rhythm generators for spinal “respiratory” and locomotor activities in the rabbit. Exp. Brain Res. 63: 225-232, 1986. 34. VIALA, D., L. PERSEGOL, AND R. PALISSES. Relationship between phrenic and extensor activities during fictive locomotion. Neurosci. Lett. 78: 318-322, 1987. 35. VIALA, D., C. VIDAL, AND E. FRETON. Coordinated rhythmic bursting in respiratory and locomotor muscle nerves in the spinal rabbit. Neurosci. Lett. 11: 155-159, 1979. 36. WALLEN, P., AND T. L. WILLIAMS. Fictive locomotion in the lamprey spinal cord in vitro compared with swimming in the intact and spinal animal. J. Physiol. Lord 347: 225-239, 1984. 37. WILLIAMS, B. J., C. A. LIVINGSTON, AND R. B. LEONARD. Spinal cord pathways involved in initiation of swimming in the stingray, Dasyatis sabina: spinal cord stimulation and lesions. J. Neurophysiol. 51: 578-591, 1984. 38. WOLTHIUS, R. A., V. F. FROELICHER, JR., J. FISCHER, AND J. H. TRIEBWASSER. The response of healthy men to treadmill exercise. Circulation 55: 153-157, 1977. l
Downloaded from www.physiology.org/journal/jappl at Midwestern Univ Lib (132.174.254.157) on February 12, 2019.
D. FUNK,
of Zoology,
JOHN D. STEEVES, AND University of British Columbia,
WILLIAM Vancouver,
K. MILSOM British Columbia
V6T 2A9, Canada
FUNK, GREGORY D., JOHN ID. STEEVES, AND WILLIAM K. of wingbeat and respiration in birds. II. MILSOM. Coordination “Fictiue” flight. J. Appl. Physiol. 73(3): 1025-1033, 1992.-To determine whether an interaction between central respiratory and locomotor networks may be involved in the observed coordination of wingbeat and respiratory rhythms during free flight in birds, we examined the relationship between wingbeat and respiratory activity in decerebrate Canada geese and Pekin ducks before and after paralysis. Locomotor activity was induced through electrical stimulation of brain stem locomotor regions. Respiratory frequency (f,) was monitored via pneumotachography and intercostal electromyogram recordings before paralysis and via intercostal and cranial nerve IX electroneurogram recordings after paralysis. Wingbeat frequency (fw) was monitored using pectoralis major electromyogram recordings before, and electroneurogram recordings after, paralysis. Respiratory and cardiovascular responses of decerebrate birds during active (nonparalyzed) and “fictive” (paralyzed) wing activity were qualitatively similar to those of a variety of vertebrate species to exercise. As seen during free flight, wingbeat and respiratory rhythms were always coordinated during electrically induced wing activity. Before paralysis during active wing flapping, coupling ratios (f,/f,) of l:l, 21, 3:1, and 4:l (wingbeats per breath) were observed. After paralysis, f, and f, remained coupled; however, 111 coordination predominated. All animals tested (n = 9) showed I:1 coordination. Two animals also showed brief periods of 21 coupling. It is clear that locomotor and respiratory networks interact on a central level to produce a synchronized output. The observation that the coordination between f, and fV differs in paralyzed and nonparalyzed birds suggests that peripheral feedback is involved in the modulation of a centrally derived coordination.
has also been implicated in the synchronization of ventilatory pattern with locomotor pattern in rabbits (24,34) and cats (ZO), although the data to support this are equivocal, Locomotor and respiratory patterns were only occasionally coordinated in decorticate and decerebrate rabbits (24, 34), with tight coordination between these two motor systems only occurring in paralyzed spinalized animals (33-35). Furthermore, afferent feedback was not completely removed in these studies. Although vagotomy removed the influence of pulmonary stretch receptor feedback on respiratory pattern, intercostal (26) and diaphragmatic reflex pathways (30) remained intact and could have altered or contributed to the entrainment (26) of respiratory rhythm. The purpose of this study, therefore, was to determine whether the entrainment between wingbeat and respiration observed during free flight in birds could be produced by a central coupling of locomotor and respiratory rhythm generators. Birds offer unique advantages over mammalian preparations for these studies. Because of the unique structure of the avian respiratory system, birds can be unidirectionally ventilated during paralysis, eliminating the phasic afferent feedback associated with the regular mechanical deformation of the respiratory system and chest wall. In addition, problems in defining gait and/or gait transitions (24,33-35) are not a problem with birds, because once the wings are recruited into the locomotor pattern, they always beat synchronously (27-29).
locomotor-respiratory coupling; forward; flight; goose
MATERIALS
entrainment;
exercise;
feed-
Animal Preparation
relaying information regarding wingbeat rhythm appears to play an important role in the coordination of wingbeat and respiration during flight in Canada geese(16). It remains unclear, however, whether such afferent feedback is essential for the coordination of wing and respiratory movements. The two motor systems may also interact centrally to produce a coordinated output in the absence of feedback activity. This type of feedforward neurogenic drive has been shown to regulate ventilatory drive during exercise in cats (II). Neural impulses arising from the suprapontine brain that command muscle to exercise appear to “irradiate” to the respiratory centers in the medulla and cause a neurally driven increase in ventilation that is proportionate to the increase in metabolic rate. Feedforward AFFERENT
AND METHODS
FEEDBACK
0161-7567/92 $2.00 Copyright
Experiments were carried out on seven Pekin ducks (Anus platyrhynchos; 2.63 * 0.08 kg body wt) and nine Canada geese (Branta canadensis; 3.6 k 0.2 kg body wt). The surgical procedures used in these experiments for ducks and geese were similar to those described previously (14). Briefly, the animals were tracheostomized to deliver gaseous anesthesia, monitor ventilation via pneumotachography, and administer unidirectional ventilation (UDV) after paralysis. A clavicular air sac was also cannulated to provide an outflow path for the air during UDV. It was plugged during spontaneous breathing. A carotid artery and jugular vein were cannulated to monitor blood pressure and administer the paralytic agent. The birds were then supported in a sling overlying a variable-speed motor-driven treadmill. They were positioned with their legs on the treadmill and wings folded at their
0 1992 the American
Physiological
Society
Downloaded from www.physiology.org/journal/jappl at Midwestern Univ Lib (132.174.254.157) on February 12, 2019.
1025
1026
CENTRAL
COUPLING
OF
WINGBEAT
sides. The treadmill was required in these studies because walking behavior was often elicited by electrical stimulation of brain stem locomotor regions before a region that would induce wing flapping could be located. A thermometer was inserted -25 cm down the esophagus to monitor body temperature (Tb). T, was maintained constant at 41OC by use of a heat lamp and a 15-cm Ushaped copper tube inserted via the cloaca into the intestine, through which hot or cold water was circulated. The head of each bird was placed in a stereotaxic head holder, and a craniotomy was performed, followed by decerebration. The transection level extended from the caudal border of the habenular nucleus dorsally to the rostra1 border of the anterior preoptic nucleus ventrally (31). Transection at more caudal levels reduced the occurrence of electrically stimulated locomotor activity after paralysis. After completion of the decerebration, the nerve innervating the external (inspiratory) or internal (expiratory) intercostal muscles of the sixth costal space and the nerve innervating the pectoralis major muscle were isolated for recording of “fictive” respiration and wing flapping, respectively, after paralysis. Cranial nerve IX was also isolated in three animals, and its activity was used as another indicator of fictive breathing. Electromyographic (EMG) electrodes were implanted into internal and/or external intercostal muscles of the fifth and sixth costal spaces and the right and left pectoralis major muscles to monitor intercostal muscle activity and wingbeat frequency during electrically induced wing flapping. The equipment used to record respiratory, cardiovascular, and EMG data has been described elsewhere (14). After these surgical procedures, anesthesia was discontinued and the birds were allowed 21 h to recover. Brain stem stimulation (monopolar electrodes, Kopf SNE 300 X 75; pulse duration 2 ms; pulse frequency 60 Hz, stimulation intensity range 30450 PA) was then initiated using procedures described in detail elsewhere (31). High-intensity stimulation (100 PA) was used to localize a stimulation site within the brain stem (Fig. 1) that would elicit wing flapping. Stimulation intensity was then decreased to zero and slowly increased to establish the current threshold necessary to invoke wing flapping. Trials were then carried out using stimulation intensities -20% greater than threshold. The suprathreshold stimulation avoided the delay between onset of stimulation and start of locomotion often associated with stimulation at threshold levels. Artificial
Ventilation
After completion of the protocol (see below) on the nonparalyzed animals and before the administration of the paralytic agent, the birds were unidirectionally ventilated with 4% CO, in air at flow rates of 1.5-2.0 l/min. Preliminary experiments (n = 3) indicated that this flow rate and concentration of inspired gas maintained arterial blood PO, (108.5 t 4.0 Torr), PCO, (27.1 * 0.36 Torr), and pH (7.41 -t- 0.01) near the normal levels seen in ducks (6). Recent values reported for bar-headed geese (lZ), however, indicate that the geese in this study may have been slightly hvperventilated. Nonetheless, these flow
AND
RESPIRATION
FIG. 1. Composite diagram of cross sections through avian brain stem showing stimulation sites used to initiate either active (closed symbols) or fictive (open symbols) wing flapping in ducks (triangles) and geese (squares). Stimulation sites effective in evoking both active and fictive locomotion are designated with closed diamonds (ducks, n = 3) and closed circles (geese, KJ = 1). A-F run caudorostral (A, medulla; B and C, pons; D-F, midbrain). Stimulation sites are indicated at Leftof each cross section. Anatomic structures are labeled on right. AL, ansa lenticularis; AQ, cerebral aqueduct; BC, brachium conjuctivum; Cnd, nucleus reticularis medullaris pars dorsalis; Cnv, nucleus reticularis medullaris pars ventralis; DBC, decussation of brachium conjunctivum; EM, ectomammillary nucleus; EW, Edinger-Westphal nucleus; ICo, nucleus intercollicularis; III, oculomotor nucleus; IO, inferior olivary nucleus; IP, interpeduncular nucleus; IV, trochlear nucleus; LC, locus ceruleus; MLd, lateral mesencephalic nucleus; MLF, medial longitudinal fasciculus; MRF, mesencephalic reticular formation; MV, motor trigeminal nucleus; NIII, oculomotor nerve; NIV, trochlear nerve; NV, trigeminal nerve; NX, vagus nerve; OT, optic tectum; R, raphe nucleus; RP, pontine reticular formation; Rpc, pontine nucleus, parvocellular part; RPO, pontine reticular nucleus, oral part; Ru, red nucleus; ST, subtrigeminal nucleus; SV, trigeminal sensory nucleus; TPc, substantia nigra, pars compacta; TS, tractus solitarius; TTD, nucleus and tract of descending trigeminal nerve; VII, facial nucleus; X, motor nucleus of vagus; XII, hypoglossal nucleus.
rates were sufficient to maintain adequate blood gases (>1.2 l/min) yet low enough to prevent hyperventilationinduced apnea and reductions in T, (~2.5 l/min). Once UDV was established, the birds were paralyzed with an initial injection of gallamine triethiodide (Flaxedil; LO mg/kg iv). Supplemental doseswere administered as required. Nerue Recordings After paralysis, the left intercostal (either internal or external) and the left pectoralis nerves were reexposed. Tissue wells were formed around each exposed nerve and filled with mineral oil. The nerves, submerged in mineral oil, were placed on bipolar platinum hook electrodes. Under these recording conditions, the position of the birds was the same as that before paralysis. In the three animals from which cranial nerve IX activity was recorded, the bodv position of the birds was not altered:
Downloaded from www.physiology.org/journal/jappl at Midwestern Univ Lib (132.174.254.157) on February 12, 2019.
CENTRAL
COUPLING
OF
WINGBEAT
however, the head was rotated sideways (lateral surface facing up) to facilitate formation of a tissue well. Preliminary experiments (n = 5) confirmed the previous finding (18) that intercostal nerve discharge provides a good index of respiration in birds and that unilateral recordings reliably indicate respiratory pattern. Discharge of right and left external intercostal nerves from the fifth and sixth costal spaces was always synchronous and in phase with inspiratory airflow during spontaneous ventilation. Right and left internal intercostal nerves discharged synchronously in phase with expiration. In addition, the discharge of external and internal intercostal nerves, whether ipsi- or contralateral, was always out of phase. Because the intercostal muscles may be recruited to serve a locomotor function during wing flapping, the discharge of cranial nerve IX was also examined as an index of respiration to ensure that the intercostal electroneurogram (ENG) truly represented central respiratory activity. In mammals, cranial nerve IX innervates the stylopharyngeus (an airway dilator active during inspiration) and glossopharyngeus (an airway constrictor sometimes active during expiration) muscles of the upper airway (17). Inspiratory activity in cranial nerve IX was determined by comparison with intercostal ENG activity at rest. Although bilateral recordings were never made from cranial nerve IX, its activity was always found to be in phase with external intercostal nerve discharge (i.e., active during inspiration) and/or out of phase with internal intercostal nerve discharge (n = 3). Bilateral recordings of pectoralis nerves were not taken during these experiments, because previous recordings made during spontaneous and electrically and chemically induced fictive wing flapping (27) indicated that discharge of left and right pectoralis nerves was always synchronous. All neurogram signals were amplified (X10,000), Cltered, and recorded in the same manner as EMG signals. One channel of EMG was recorded to ensure that paralysis was complete during all trials. Stimulation trials (100-400 PA) were then resumed, this time recording ENG activity rather than EMG activity. At the conclusion of each experiment, the position of the stimulating electrode was marked with an electrolytic lesion (3 mA/5 s). Histological identification of stimulation sites was performed as described previously (31). Protocol Nonparalyzed birds. Five ducks and nine geese were examined to describe the relationship between airflow and intercostal and pectoralis muscle activity during electrically induced wing flapping. Minute ventilation (VE), ventilatory frequency (f,>, tidal volume (VT), blood pressure (BP), heart rate (HR), internal or external EMG activity, and right and left pectoralis EMG activity were recorded for 30 s prestimulation, throughout stimulation, and for up to 3 min poststimulation. Most stimulation periods lasted IO-30 s. However, in two animals the duration was increased to 2 min to further describe the kinetics of the ventilatory response to electrically induced wing flapping (see Fig. 4). Paralyzed birds. Activities of the pectoralis nerve (fic-
AND
1027
RESPIRATION
TABLE 1. Different ratios of wingbeat frequency to ventilation frequency obtained with stimulation at various sites Active Stimulation Site
TTD/Cnd RP MRF ICO
Flapping
I:1
2:l
311
4:l
1:l
4 1
1 1
1
2 1
2
1
1 1 2 5
Fictive
Flapping
2:l
3:l
411
2
Values represent no. of animals (geese or ducks) demonstrating each wingbeat frequency-to-ventilation frequency ratio. Some birds showed >I ratio. TTD, nucleus and tract of descending trigeminal nerve; Cnd, nucleus reticularis medullaris, pars dorsalis; RP, nucleus reticularis pontis caudalis, pars gigantocellularis; MRF, mesencephalic reticular formation; ICo, nucleus intercollicularis.
tive flight) and internal or external intercostal nerve or cranial nerve IX (fictive respiration) were recorded for 30-60 s prestimulation, throughout stimulation (range IO-30 s), and for up to 2 min poststimulation in seven ducks and two geese. Pectoralis EMG activity, HR, and BP were also recorded. Five of seven ducks and two of nine geesewere examined before and after paralysis; however, there was no indication that those birds studied only before or only after paralysis differed from those studied under both conditions. Data Analysis Unless otherwise indicated, values are means -t SE. Analysis of variance was used to test the difference between means. P < 0.05 was assumed to be significant. RESULTS
Nonparalyzed Responses Brain stem stimulation successfully initiated active wing flapping in all five ducks and six of nine geese. All effective stimulation sites fell within areas described previously (27-29, 31) and are summarized in Table 1 and Fig. 1. No effort was made to initiate locomotion from each site listed in Table 1 for each animal. Thus the data in Table 1 do not indicate the relative effectiveness of each site at evoking locomotion. Stimulation sites were similar for ducks and geese, The stimulation intensity necessary to evoke wing flapping averaged 137 t 23 PA. During the course of the experiments, a variety of wingbeat flight patterns were observed. Figure 2 shows an entire sequence from one duck. In this particular trace, threshold stimulation was used, rather than the usual suprathreshold stimulus, which activates locomotion with little or no delay. This trace demonstrates that the dramatic alteration of respiratory pattern and increase in BP and HR that occur immediately with the onset of stimulation were not dependent on the development of wing activity. With the onset of stimulation, VT decreased from 33.3 to 16.3 ml/kg, but the dramatic increase in f, from 18.2 to 135 breaths/min observed at the onset of stimulation resulted in a large increase in VE from 606 to 2,194 ml min-l. kg? Similarly, HR increased at the onset of stimulation from 200 beats/min at rest to 513 beats/min 7 s later when wing flapping coml
Downloaded from www.physiology.org/journal/jappl at Midwestern Univ Lib (132.174.254.157) on February 12, 2019.
1028
CENTRAL
Blood
COUPLING
OF
WINGBEAT
AND
RESPIRATION
Pressure
stim. on
/
c
/
c
c
‘\ stim. off
0
2
S8C
\
FIG. 2. Sequence of events associated with wing flapping init.iated using threshold current (90 PA) in a decerebrate duck. Prestimulation and recovery records are included. Heart rate, mean blood pressure, left and right pectoralis EMG, and integrated respiratory airflow (labeled airflow, inspiration up and expiration down) are shown. Inset: expanded view of outlined section illustrating precise 21 relationship between pectoralis EMG activity and respiratory rhythm. (Unlike Fig. 5, airflow integrator was operating in reset mode. Thus, trace returned to baseline whenever flow direction changed.)
menced. Mean BP also increased over the first 7 s from 122 mmHg at rest to a maximum of 175 mmHg. By the onset of wing activity 7 s later, iiE, BP, and HR had already plateaued at these elevated levels and did not increase further. With the termination of brain stem stimulation and wing flapping, all variables gradually returned toward prestimulation levels. Similar changes in respiratory pattern during the transition from rest to wing flapping can be seen in Fig. 5. In this case, however, suprathreshold stimulation was used, and the changes in respiratory pattern preceded the onset of wing flapping by
10
WINGBEAT
rest
m stlm ESSI pora, red sttm
EZJpafa,
-+ 800
-
(mm-‘1 ‘;
400~
HA
4
AND
1029
RESPIRATION
only 60% of the level achieved preparalysis. As a result, the relationship between f, and f, also changed during fictive flight. After paralysis, all nine birds (7 ducks, 2 geese) demonstrated I:1 synchronization. Two birds (1 duck, I goose) also showed brief periods of coordination at 21. No other coordination schemes were observed after paralysis (Table 1). An example of I:1 coordination during fictive flight is shown in Fig. 6. With the initiation of fictive flight, there was a dramatic increase in fV from 6.9 bursts/min at rest (top) to 92.5 bursts/min as f, increased to match f, 1:l. During periods of 1:l coordination, the phase relationship between f, and f, was constant, Inspiratory nerve activity [external intercostal nerve (rz = 3) and cranial nerve IX activity (n = 3)] occurred out of phase with pectoralis activity (Fig. 6), while expiratory nerve activity [internal intercostal nerve (n = 4)] occurred in phase with pectoralis ENG discharge (Fig. 7). DISCUSSION
Critique
100
of Methods
Validity of decerebrate preparations: free flight vs. electrically induced wing flapping. Locomotion in decerebrate animals, whether spontaneous or produced by electrical/ chemical stimulation, has been shown to be similar to
50
-
-0
3. Minute ventilation (iiE), tidal volume (VT), breathing frequency (f,), heart rate (HR), blood pressure (BP), and wingbeat frequency (fw) for all decerebrate birds (ducks and geese; sample size indicated by no. over each column) before paralysis at rest, (rest), during electrically induced wing flapping (stim), after paralysis at rest (para, rest), and during fictive wing flapping (para, stim). Values are means -t- SE* Significant differences between groups: “preparalysis, rest vs. seim; ‘para, rest vs. para, stim; +stim vs. para, stim. FIG.
actual wing flapping, the results obtained for ducks and geese during fictive flight were virtually identical. Again, all effective stimulation sites fell within brain stem regions described previously (27-29, 31) and are summarized in Table 1 and Fig. 1. The average stimulation intensity used to evoke fictive wing flapping was 340 t 46 PA. Although fictive flight could be initiated from the more caudal brain stem stimulation sites (nucleus and tract of descending trigeminal nerve; nucleus reticularis medullaris, pars dorsalis), the more rostra1 sites (nucleus reticularis pontis caudalis, pars gigantocellularis; mesencephalic reticular formation; nucleus intercollicularis) were most effective in initiating fictive activity. Stimulation sites used to initiate flapping pre- and postparalysis were not always the same, Of the six birds (5 ducks, I goose) examined both before and after paralysis, the same stimulation sites were used pre- and postparalysis in four birds (3 ducks, 1 goose). Paralysis itself did not significantly alter HR, BP, or f, (Fig. 3). The magnitude of the increases from resting levels in HR, BP, and f, during fictive flapping were also similar to those seen during actual wing flapping (Fig. 3). The increase in wingbeat frequency (f,), however, was sienificantlv less than that before Daralvsis. increasing to
W
l>
300
-40
0
40
80
Time
120
160
200
240
(set)
FIG. 4. Kinetics of ventilatory response of 2 decerebrate geese to 2 min of electrically induced wing flapping showing VE, VT, and f,. Stimulation was turned on at t = 0 and off t = 120 s.
Downloaded from www.physiology.org/journal/jappl at Midwestern Univ Lib (132.174.254.157) on February 12, 2019.
1030
CENTRAL
R. Pect.
EMG
L. Pmt.
EMG
L. External
Air
ht.
COUPLING
OF
WINGBEAT
AND
RESPIRATION
EMG
Insp.
T
Exp.
J
Flow
R. Pect.
EMG
L. Pect.
EMG
L. External
Air Flow
ht.
FIG. 5. Relationship between right and left pectoralis EMG activity, left external intercostal EMG activity, and integrated respiratory airflow (airflow) at rest (top) and during electrically induced wing flapping in a decerebrate duck (bottom). Note different time scales. (Respiratory integrator was operating in sum mode. Thus, trace did not reset with changes in flow direction as in Fig. 2.)
EMG
Insp.
T
Exp.
& stim. on
t
stim. off t 1
I 2 sec.
locomotion in intact animals for a large number of species (ZZ), including geese (31) and chicks (329, during wing flapping. Similarly, the respiratory pattern produced by decerebrate animals is very similar to that seen in intact animals (I3), including geese (14). Furthermore the respiratory responses of decerebrate geese to low-level exercise (treadmill walking) closely resemble the responses of intact geese (14). Because the physiological responses to hindlimb vs. forelimb exercise can differ (I), however, it was essential in this study to compare the
responses of decerebrate geese during electrically induced wing flapping with the responses of intact animals during free flight. The physiological responses of decerebrate birds to electrically induced wing flapping were considerably smaller than the responses of intact birds to free flight: f, was approximately one-half the level seen during free flight. However, assuming that metabolic rate is a function of both f, and the force of muscular contraction, a reduction in metabolic rate between free flight and elec-
Cranial Nerve IX ENG
Cranial Nerve IX ENG
L. Pect. ENG stim.
on t
stim. off T I
1 5 sec.
FIG.
6. Recording
(top) and during synchronization
of left cranial nerve IX (inspiratory activity; fictive respiration) and left pectoralis ENG at rest electrically induced fictive wing flapping (bottom) in a decerebrate paralyzed duck showing precise 1:l between the 2 patterns during fictive flight. Note different time scales.
Downloaded from www.physiology.org/journal/jappl at Midwestern Univ Lib (132.174.254.157) on February 12, 2019.
CENTRAL
COUPLING
OF
WINGBEAT
AND
RESPIRATION
1031
metabolic rate and response magnitude between intact and decerebrate animals by increasing the stimulus intensity used to evoke locomotion and thus the force of wing flapping. However, increased stimulation intensity L. Pect ENG caused a dramatic increase in metabolic heat production. 2 Lc. Highly efficient nonevaporative convective heat loss, which is responsible for 65100% of the cooling in birds during flight (5), is difficult to achieve in the laboratory. Thus, to avoid the problem of T, regulation, low-level stimulation was used to evoke locomotion and most stimulation periods were kept below 30 s. L. Pect. ENG In summary, the responses of decerebrate birds during electrically induced wing flapping appeared qualitatively similar to those anticipated to occur in intact geese during free flight. The cardiovascular adjustments observed during electrically induced wing flapping approximated those seen during exercise in a variety of vertebrate species. The kinetics of the ventilatory response and the breathing pattern responses of geese to electrically intrically induced wing flapping was expected. In turn, if duced flapping were also similar to those of free flight. the relationship between VE and 0, consumption seen Paralyzed preparations: active wing flapping vs. fictive during walking in decerebrate geese (14) also holds for flight. The physiological responses of birds during fictive flight were remarkably similar to those during active wing flapping, a reduction in TjE would also be expected. The 2,5-fold increase in VE in decerebrate birds was ap- wing flapping. As seen previously in cats (23), the cardioproximately one-fourth of the average 123-fold increase vascular responses of paralyzed and unparalyzed birds to in V, recorded from several species during free flight. fictive vs. active locomotion were very similar. In addiThus the difference in the magnitude of the ventilatory tion, although increases in f, were substantial (4%fold) response between free flight and electrically induced during fictive flight, they were less than increases during wing flapping does not appear to be due to a decrease in active flapping. Estimates of VT and, hence, overall levthe responsiveness of the decerebrate birds but more a els of ventilatory effort (7jE) could not be made for the result of the decerebrate animals performing at lower paralyzed birds during fictive flight. However, estimates metabolic rates than free-flying animals. of TjE in mammals from integrated phrenic nerve activity This suggestion is supported by three additional ob- indicate that VE increases similarly during fictive and servations. First, the kinetics of the ventilatory response active locomotion (11). to wing-flapping exercise in decerebrate birds (Fig. 4) Similar to f,, f, was reduced during fictive flight (86 t were similar to those of intact birds [black duck (3), pi- 21 beats/min) relative to active wing flapping (144 t 19 geon (9)]. Both showed a rapid increase in VE with the beats/min). A reduction in locomotor frequency after paonset of flapping, followed by a plateau and a rapid re- ralysis has been consistently observed in a variety of speturn toward resting levels with termination of flight. Sec- cies [lamprey (36), stingray (37), duck and goose (27), cat ond, like the ventilatory response of decerebrate birds, (19)] and apparently results from the removal of an excitthe increase in HR was diminished in decerebrate birds atory drive to the locomotor pattern generators that is normally provided by phasic afferent feedback. This derelative to free-flying intact animals. HR, which typically increases between 2 and 5%fold during free flight (3,9), crease in central excitability of the locomotor system is increased only L5-fold in the decerebrate birds, again also evident in that the stimulation intensity required to approximately one-fourth of the increase during free activate fictive locomotion in paralyzed birds and other flight. Third, the 1.46-fold increase in HR in the decere- species was much greater than that required to activate brate birds during wing flapping was associated with an locomotion before paralysis (19, 27, 36, 37). Once locoalmost identical 1.5-fold increase in BP from 110 to 164 motion is activated in the fictive preparation, however, mmHg. Similar increases in BP have been recorded in the locomotor pattern of the fictive preparation is very similar to that of the unparalyzed animal (19, 22, 27). intact ducks running on a treadmill at work rates that produced similar levels of ventilation [2.6 times resting 0, consumption (Z)]. The only measurements of BP durMechanisms of Entrainment ing flight in intact birds indicate only a small increase in BP from 142 to 147 mmHg in pigeons (9). However, deCoordination during fictive flight. The mechanisms inspite efforts to reduce stress in these birds, the absence of volved in controlling the gross cardiovascular and respiratory responses to exercise can be divided into two catea pressor response in these birds appeared to be due to elevated preflight BP relative to previously recorded gories: 1) feedback from humoral as well as neurogenic “resting” values (127 t 6 mmHg) (8). In fact, most ani- drives occurring after the start of exercise and 2) feedformals [humans (38), dogs (Zl), ducks (a)], including de- ward due to central coactivation of cardiovascular and cerebrate cats (II, 23), demonstrate an increase in BP respiratory regions in the brain stem by pathways involved in the initiation of locomotion. There is clear evithat is proportional to work rate during exercise. It would have been possible to reduce the differences in dence that feedforward mechanisms from locomotor L. Internal
ht.
ENG
#
Downloaded from www.physiology.org/journal/jappl at Midwestern Univ Lib (132.174.254.157) on February 12, 2019.
1032
CENTRAL
COUPLING
OF
WINGBEAT
centers are involved in the production of a proportionate drive to ventilation during exercise (Fig. 3) (11). In addition, our results provide clear evidence in birds that feedcenters not only ac tivity from the locomotor forward drive to ven tilation but is inprovides a proportional locomotor and respiratory in coordinating volved rhythms. Previous work with cats (20) and rabbits (24, 51) is suggestive of this finding; however, several factors confounded interpretation of these data. Kawahara et al. (20) observed synchronization of central locomotor and respiratory outputs in paralyzed decerebrate cats at extremely low arterial PCO, levels, when the frequency of the respiratory pump was very high (i.e., similar to the elicited locomotor frequency). However, entrainment decreased dramatically at higher arterial PCO, levels when the pump frequency was decreased. Although the cats were vagotomized to remove any vagally mediated respiratory entrainment (25), phrenic and intercostal reflexes remained partially intact and have been shown to significantly alter (26, 3O), and even entrain, respiratory rhythm (26). The strength of the central coupling between locomotor and respiratory rhythm-generating networks remains unclear in this case, because the coordination of locomotor and respiratory rhythms may have been enhanced by feedback associated with the pump ventilation. Similarly, the relationship between central locomotor and respiratory neural outputs from paralyzed decorticate and decerebrate rabbits ind cates that these two networks are only weakly coordinated (24,34). It is possible, however, that this low degree of central coordination was due to the fact that rabbits show gai t-dependent entrainment of re spiration and locomotion (7); i.e., entrainment is very low in these animals during walking but virtually 100% with the transition to trot or gallop when they breathe once per leap. Unilateral recordings of leg ENG activity were typically used to describe locomotor pattern. Because such records do not distinguish between and walking (i.e., alternating hindlimb nerve activity) hopping (i.e., synchronous hindlimb nerve activity), gait transitions, which are common during pharmacological (28, 29), may have occurred. activati .on of locomotion Failure to detect such gait changes would have produced a drastic underestimation of the degree to which locomotion and respiration are coordinated at, the central level. Thus it remai .ns unclear whether th .e inte rmittent ten .tral coord in .ation between 1ocomotion and respiration observed in rabbits is due to changes in fictive gait or weak central coupling. Evidence for tight central coordination of locomotor and respiratory networks has been found only in rabbits spinalized at C, (33-35). That locomotor output can be generated in the isolated spinal cord is well established. However, whether phrenic discharge in spinal animals represents respiration remains unclear (13), because the neural networks responsible for respiratory rhythmogenesis are predominantly located in the medulla and caudal pons (13). The present study was designed to eliminate several of these complicating factors. Unidirectional ventilation of paralyzed birds eliminated all phasic afferent feedback
AND
RESPIRATION
associated with pump ventilation, permitting more direct examination of the central relationship between locomotor and respiratory rhythm-generating networks than possible with mammalian preparations. Furthermore, because birds do not show gait changes during flight/ wing flapping, the confounding influences of gait-dependent entrainment were avoided. Our results provide the strongest evidence to date that central mechanisms are involved in coordinating locomotion and respiration. All nine animals examined during fictive flight showed a 1: 1 synchronization (2 birds also showed brief periods of 21 coupling) between the central neural outputs of the locomotor/wing and respiratory systems. The external intercostal nerve (inspiratory) discharged out of phase with the pectoralis nerve, and the internal intercostal nerve (expiratory) discharged in phase with the pectoralis nerve (Fig. 7). In addition, inspiratory activity in cranial nerve IX coordinated with pectoralis ENG during fictive flight in exactly the same manner as the external intercostal ENG (Fig+ 6), ruling out the possibility that the intercostal nerve discharge observed during fictive wing beating was locomotor in nature (Fig. 6). Thus it is clear that the neural networks responsible for the production of locomotor (wingbeat) and respiratory rhythms in birds interact centrally to produce a synchronized output in the absence of phasic afferent feedback. Coordination during free flight. The 1:l relationship between wingbeat and respiration observed in paralyzed birds differed from the range of coupling ratios seen during active wing flapping (I:1 to 4:l) when phasic mechanoreceptive feedback was present (i.e., before paralysis). Although the stimulation sites used to activate locomotor activity were not always the same before and after paralysis, the differences in coupling ratio did not appear to be a function of the stimulation sites used to activate wing flapping (Table 1). Rather, it appears that phasic afferent feedback somehow modulates the centrally derived coordination to produce a range of coupling ratios. These data agree with observations on free-flying birds, where Canada geese have shown 2 and 3 wingbeats/ breath (15) and barnacle geese have shown 2, 3, and 4 wingbeats/breath (IO). In fact, observations on a variety of bird species indicate that the majority of species demonstrate a range of coupling ratios during free flight (4). Conclusion It is clear that the networks responsible for the generation of locomotor and respiratory rhythms interact on a central level (brain stem or spinal cord), in the absence of phasic afferent feedback, to produce synchronized outputs in birds. Addition of afferent feedback does not disrupt this coordination but changes the number of wingbeats per breath. Thus, both feedforward mechanisms and mechanical feedback (16) appear to be involved in the production and control of entrainment. Under the labile conditions of free flight, when the wingbeat rhythm is constantly modified to compensate for environmental perturbations, it is most probable that the coordination of wingbeat and respiration is produced and maintained via a collective interaction of feedforward lo-
Downloaded from www.physiology.org/journal/jappl at Midwestern Univ Lib (132.174.254.157) on February 12, 2019.
CENTRAL
COUPLING
OF
WINGBEAT
comotor inputs and mechanoreceptive feedback on some portion of the respiratory rhythm-generating network. This work was supported by grants from the Natural Sciences and Engineering Research Council (NSERC) of Canada (W. K. Milsom and J. D. Steeves). G. D. Funk was supported by Fellowships from NSERC of Canada and the Killam Foundation. Address for reprint requests: G. D. Funk, Systems Neurobiology Laboratory, Dept. of Physiological Science, UCLA, 405 Hilgard Ave., Los Angeles, CA 90024-1527. Received
22 January
1991; accepted
in final
form
30 March
1992.
REFERENCES 1. ASMUSSEN, E., AND M. NIELSEN. Studies on the regulation of respiration in heavy work. Actu Physiol. Stand. 10: 270-285, 1946. 2. BECH, C., AND S. NOMOTO. Cardiovascular changes associated with treadmill running in the Pekin duck. J. Exp. Biol. 97: 345-358, 1982. 3. BERGER, M., J. S. HART, AND 0.2. ROY. Respiration, oxygen consumption and heart rate during rest and flight. 2. Vergl. Physiol. 66: 201-214, 1970. 4. BERGER, M,, 0. Z. ROY, AND J. S. HART. The coordination between respiration and wingbeats in birds. 2. Vergl. Physiol. 66: 190-200, 1970. 5. BERNSTEIN, M. H. Ventilation and respiratory evaporation in the flying crow, Corvus ossifragus. Respir. Physiol. 26: 371-382, 1976. P., N. HILL, AND Y. JAMMES. Ventilatory responses to 6. BOWEROT, CO2 in intact and chronically chemodenervated Peking ducks. Respir. Physiol. 22: 137-156, 1974. D. M., AND D. R. CARRIER. Running and breathing in 7. BRAMBLE, mammals. Science Wash. DC 219: 251-256, 1983. P. J., AND E. W. TAYLOR. Responses of the respiratory 8. BUTLER, and cardiovascular systems of chickens and pigeons to changes in Pao, and PacoZ. Respir. Physiol. 21: 351-363, 1974. P. J., N. H. WEST, AND D. R. JONES. Respiratory and 9. BUTLER, cardiovascular responses of the pigeon to sustained level flight in a wind tunnel. J. Exp. Biol. 71: 7-26, 1977. 10. BUTLER, P. J., AND A. J. WOAKES. Heart rate, respiratory frequency and wingbeat frequency of free flying barnacle geese Brantu leucopsis. J. Exp. Biol. 85: 213-226, 1980. 11. ELDRIDGE, F. L., D. E. MILLHORN, J. P. KILEY, AND T. G. WALDROP. Stimulation by central command of locomotion, respiration and circulation during exercise. Respir. Physiol. 59: 313-337, 1985. 12. FEDDE, M. R., J. A. ORR, H. SHAM, AND P. SCHEID. Cardiopulmonary function in exercising bar-headed geese during normoxia and hypoxia. Respir. Physiol. 77: 239-262, 1989. 13. FELDMAN, J. L., J. C. SMITH, D. R. MCCRIMMON, H. H. ELLENBERGER, AND D. F. SPECK. Generation of respiratory pattern in mammals. In: Neural Control of Rhythmic Movements in Vertebrates, edited by A. H. Cohen, S. Rossignol, and S. Grillner. New York: Wiley, 1988, p. 73-100. 14. FUNK, G. D., W. K. MILSOM, G. N. SHOLOMENKO, AND J. D. STEEVES. Role of the telencephalon in the synchronization of locomotor and respiratory frequencies during walking in Canada geese. J. Exp. Biol. 145: 283-301, 1989. 15. FUNK, G. D., W. K. MILSOM, AND J. D. STEEVES. Entrainment of wingbeat and respiration during free-flight, “fictive” flight and passive wing flapping in geese. Sot. Neurosci. Abstr. 15: 391, 1989. 16. FUNK, G. D., J. D. STEEVES, AND W. K. MILSOM. Coordination of wingbeat and respiration in the Canada goose. I. Passive wing flapping. J. Appl. Physiol. 73: 1014-1024, 1992. 17. ISCOE, S. D. Central control of the upper airway. In: Respiratory Function of the Upper Airway, edited by 0. P. Mathew and G. Sant’Ambrogio. New York: Dekker, 1988, p. 138-148. (Lung Biol. Health Dis. Ser.) 18. JONES, D. R., AND 0. S. BAMFORD. Open-loop respiratory chemo-
AND
RESPIRATION
1033
sensitivity in chickens and ducks. Am. J. Physiol. 230: 861-867, 1976. 19. JORDAN, L. M., C. A. PRATT, AND J. E. MENZIES. Locomotion evoked by brainstem stimulation: occurrence without phasic segmental afferent input. Bruin Res. 177: 204-207, 1979. 20 KAWAHARA, K., Y. NAKAZONO, Y. YAMAUCHI, AND Y. MIYAMOTO. Coupling between respiratory and locomotor rhythms during fittive locomotion in decerebrate cats. Neurosci. Lett. 103: 326-332, 1989. 21 . KIRLIN, P. C., M. D. KI?TLESON, AND L. E. JOHNSON. Neurohumoral and cardiopulmonary response to sustained submaximal exercise in the dog. J. Appl. Physiol. 62: 1040-1045, 1987. 22. MCCLELLAN, A. D. Command systems for initiating locomotion in fish and amphibians: parallels to initiation systems in mammals. In: Neural Control of Vertebrate Locomotion, edited by S. Grillner, R. M. Herman, P. S. G. Stein, H. Forsberg, and D. G. Stuart. London: Macmillan, 1986, p. 3-20. 23. MILLHORN, D. E., F. L. ELDRIDGE, T. G. WALDROP, AND J. P. KILEY. Diencephalon regulation of respiration and arterial pressure during actual and fictive locomotion in cat. Circ. Res. 61, Suppl. I: 153-159, 1987. 24. PERSEGOL, L., M. JORDAN, D. VIALA, AND C. FERNANDEZ. Evidence for central entrainment of the medullary respiratory pattern by the locomotor pattern in the rabbit. Exp. Bruin Res. 71: 153-162, 1988. 25. PETRILLO, G. A., L. GLASS, AND T. TRIPPENBACH. Phase locking of respiratory rhythm to a mechanical ventilator. Can. J. Physiol. Pharmacol. 61: 599-607, 1983. 26. REMMERS, J. E., AND I. MARTTILA. Action of intercostal muscle afferents on the respiratory rhythm of anaesthetized cats. Respir. Physiol. 24: 31-41, 1975. 27. SHOLOMENKO, G. N., G. D. FUNK, AND J. D. STEEVES. Locomotor activities in the decerebrate bird without phasic afferent feedback. Neuroscience 40: 257-266, 1990. 28. SHOLOMENKO, G. N., G. D. FUNK, AND J. D. STEEVES. Avian locomotion activated by brainstem infusion of neurotransmitter agonists and antagonists. I. Acetylcholine, excitatory amino acids and substance P. Exp. Bruin Res. 85: 659-673, 1991. 29. SHOLOMENKO, G. N., G. D. FUNK, AND J. D. STEEVES. Avian locomotion activated by brainstem infusion of neurotransmitter agonists and antagonists. II. Gamma-aminobutyric acid. Exp. Bruin Res. 85: 674-681, 1991. 30. SPECK, D. F., AND W. R. REVELETTE. Attenuation of phrenic motor discharge by phrenic motor nerve afferents. J. Appl. Physiol. 62: 941-945, 1987. 31. STEEVES, J. D., G. N. SHOLOMENKO, AND D. M. S. WEBSTER. Stimulation of the pontomedullary reticular formation initiates locomotion in decerebrate birds. Brain Res. 401: 205-212, 1987. 32. VALENZUELA, J. I., S. J. HASAN, AND J. D. STEEVES. Stimulation of the brainstem reticular formation evokes locomotor activity in embryonic chicken (in ovo). Dev. Bruin Res. 56: 13-18, 1990. 33. VIALA, D. Evidence for direct reciprocal interactions between the central rhythm generators for spinal “respiratory” and locomotor activities in the rabbit. Exp. Brain Res. 63: 225-232, 1986. 34. VIALA, D., L. PERSEGOL, AND R. PALISSES. Relationship between phrenic and extensor activities during fictive locomotion. Neurosci. Lett. 78: 318-322, 1987. 35. VIALA, D., C. VIDAL, AND E. FRETON. Coordinated rhythmic bursting in respiratory and locomotor muscle nerves in the spinal rabbit. Neurosci. Lett. 11: 155-159, 1979. 36. WALLEN, P., AND T. L. WILLIAMS. Fictive locomotion in the lamprey spinal cord in vitro compared with swimming in the intact and spinal animal. J. Physiol. Lord 347: 225-239, 1984. 37. WILLIAMS, B. J., C. A. LIVINGSTON, AND R. B. LEONARD. Spinal cord pathways involved in initiation of swimming in the stingray, Dasyatis sabina: spinal cord stimulation and lesions. J. Neurophysiol. 51: 578-591, 1984. 38. WOLTHIUS, R. A., V. F. FROELICHER, JR., J. FISCHER, AND J. H. TRIEBWASSER. The response of healthy men to treadmill exercise. Circulation 55: 153-157, 1977. l
Downloaded from www.physiology.org/journal/jappl at Midwestern Univ Lib (132.174.254.157) on February 12, 2019.
