ABSTRACT
The ontogeny of shivering thermogenesis was investigated in the altricial red-winged blackbird (Agelaius phoeniceus). Two indices of heat production – the rate of oxygen consumption of the bird and the electromyographic (EMG) activity of the pectoralis (PECT) and gastrocnemius (GAST) muscles – were measured simultaneously in adult and nestling red-winged blackbirds as they were subjected first to thermoneutral temperatures and subsequently to progressively colder ambient temperatures (Ta). The ontogenetic changes in both indices indicated that the capability for thermogenesis in nestling red-winged blackbirds improved markedly with age. The metabolic rates of 3-day-old nestlings decreased during exposure to gradually falling ambient temperatures; at best, these nestlings were only able to maintain mass-specific at levels similar to or slightly above the resting metabolic rate at thermoneutral temperatures (RMR) for a short time before metabolic rates decreased with further cooling. Shivering was detected only in the PECT muscles and was of a relatively low intensity (maximum of sevenfold increase in intensity over basal levels). The 5-day-old nestlings increased mass-specific modestly (approximately 1.4-fold) above RMR and attained slightly higher maximal factorial increases in the EMG activity of the PECT (maximum of 18-fold basal levels) when exposed to the same experimental conditions. Shivering was also detected in the GAST muscles of these birds. The most striking improvements in both measures observed during the nestling period occurred between day 5 and day 8. Eight-day-old nestlings increased metabolic rates by approximately 2-to 2.5-fold over basal levels and sustained these elevated rates for longer before becoming hypothermic. Both the PECT and GAST muscles contributed significantly to shivering thermogenesis, and these older nestlings attained much higher factorial increases in the intensity of shivering (up to 72-fold) during exposure to cold temperatures. In addition, both the range and magnitude of the dominant frequencies of muscle activity in the PECT increased during postnatal development.
The PECT muscles were a principal site of shivering thermogenesis in all nestling and adult red-winged blackbirds studied here. Shivering in these muscles was a ‘first line defense’ against cold; the threshold temperature for shivering in the PECT muscles coincided with the lower critical temperature for oxygen consumption (TLC), and the subsequent increases in EMG activity in this muscle with further cooling correlated well with the corresponding increases in mass-specific .
INTRODUCTION
Processes of metabolic thermoregulation in endotherms range from control of the rates of heat exchange with the environment (e.g. ptilomotor or vasomotor responses) to the active generation or dissipation of heat. Shivering thermogenesis entails the predominantly involuntary contraction of the skeletal muscles and is an important and phylogenetically primitive form of thermogenesis in mammals and birds (Calder and King, 1974; Hulbert, 1980; Hohtola and Stevens, 1986). Shivering utilizes the same effectors and mechanochemical processes as those involved in voluntary movements and, therefore, unlike non-shivering thermogenesis (NST) in the brown adipose tissue of mammals, does not require any specialized tissues or processes (Jansky, 1973; Cannon et al. 1981).
Adult birds apparently rely either exclusively or primarily upon shivering for regulatory heat production during a cold stress (West, 1965; Calder and King, 1974; Dawson et al. 1983; Marsh and Dawson, 1989). Shivering in adult birds is first detected with electromyography (EMG) at an ambient temperature (Ta) that corresponds with the lower critical temperature for oxygen consumption (TLC), the lowest temperature at which body temperature is maintained without an increase in metabolic rate above basal levels. The intensity of shivering increases in an approximately linear manner with the severity of the cold challenge in several species (pigeons, Columba livia, Steen and Enger, 1957; Rautenberg, 1969; Hohtola, 1982a; Japanese quail, Coturnix japonica, Hohtola and Stevens, 1986; evening grosbeak, Coccothraustes vespertinus; common redpoll, Carduelis flammea; common grackle, Quiscalus quiscula; and common crow, Corvus brachyrhynchos, West, 1965). These characteristics indicate that shivering by the skeletal muscles, particularly the large pectoralis, constitutes a ‘first line defense’ in adult birds during exposure to cold (see Hemingway, 1963; Jansky, 1973; Marsh and Dawson, 1989). In contrast, very little is known about the ontogeny of shivering thermogenesis in birds, particularly in the altricial passerine species. Only a few studies have reported shivering in nestling passerines, and no study, to the best of my knowledge, uses EMG to investigate the ontogenetic improvements in shivering ability in this group. The studies that do exist rely exclusively either on visual observations (e.g. Morton and Carey, 1971) or on measurements of tremor amplitude and frequency (e.g. Odum, 1942) and establish that the capability of nestlings to shiver improves markedly during the nestling period. However, the insensitivity of the techniques employed relative to EMG measurements (Hemingway, 1963; Hohtola, 1982b) precludes accurate determination of either the threshold temperature for shivering (Tth) or the relationship between shivering intensity and metabolic rate. The use of tremors alone provides little information about the amount of heat produced, for tremor amplitude depends more on the firing asynchrony of motor units than on their total activity. Because shivering often involves antagonistic sets of muscles (Hemingway, 1963), and therefore is essentially an isometric contraction of the skeletal muscles, tremor is not a reliable manifestation of this form of thermogenesis. In fact, overt tremors may be maladaptive, given the concomitant increase in convective heat loss (Hohtola, 1982b; Hohtola and Stevens, 1986).
In addition, the role (if any) that regulatory NST (as opposed to obligatory NST; Jansky, 1973) plays in the young bird remains ambiguous. Regulatory NST in the muscles themselves has recently been implicated as a first line mode of heat production in young ducklings, based on the observation that the Tth in the leg muscles was lower than the TLC for oxygen consumption (Barré et al. 1985).
This study investigates the ontogeny of shivering ability in the altricial red-winged blackbird (Agelaius phoeniceus). Two indices of heat production – the rate of oxygen consumption of the bird and the EMG activity of selected muscles – were measured simultaneously in adult and nestling red-winged blackbirds as they were subjected to progressively colder Ta values. Previous measurements of the of individual blackbird nestlings have shown that substantial improvements in thermoregulatory capabilities occur between days 3 and 8 after hatching (Hill and Beaver, 1982; Olson, 1992). Measurements of alone, however, cannot resolve whether shivering or regulatory NST is the primary mode of heat production in red-winged blackbird nestlings in response to a cold stress. It is also unclear from these earlier studies whether the capability of some nestlings as young as 3–4 days old to sustain their metabolic rate at basal levels during cooling is due to passive (thermal inertia) or active thermoregulation (Olson, 1992). Continuously monitoring and EMG during gradual cooling at controlled rates allows both the detection of transient increases in either of these thermogenic indices and the determination of the relationship between metabolic output and shivering. If shivering in a given muscle is a dominant and first line mode of thermogenesis during a cold challenge, then (1) the Tth in the muscle will correspond with the TLC for the animal (Hemingway, 1963; Rautenberg, 1969; Jansky, 1973; Barré et al. 1985) and (2) increases in the electrical output of the muscle should be positively correlated with an increase in .
MATERIALS AND METHODS
Study animals
Adult and 3-, 5-and 8-day-old nestlings were used to investigate the ontogeny of shivering thermogenesis in the red-winged blackbird. Two nestlings at each age were collected from their nests in three marshes in Washtenaw and Livingston Counties, southeastern Michigan (see Olson, 1991, 1992, for details about the collecting sites). Nests were located before the eggs hatched and were subsequently monitored each morning or early afternoon. Therefore, the ages of the nestlings are accurate to within 1 day. Neonates were assigned an age of 0 days on the day of hatching. The physical characteristics (e.g. size, plumage, etc.) of all of the red-winged blackbirds used in this study were typical of their age (Olson, 1992).
All nestlings were collected in the early morning to allow enough time to complete all surgical procedures and measurements on the same day. All measurements were completed within 13 h after collection. Immediately following removal from the nest, the body mass of the nestling was measured to the nearest 0.1 g on a calibrated 50 g Pesola spring balance and the birds were immediately transported to the laboratory at the University of Michigan (less than a 30 min drive). Nestlings were fed meal worms (Tenebrio sp. larvae) and/or a diluted strained meat mixture (Gerber baby food) ad libitum throughout the day until the experiments began.
Three adults were collected in mist nets from a large common reed (Phragmites sp.) marsh that bordered Lake Erie in Huron, Erie County, Ohio. The surgical and experimental procedures on the adults were often performed on consecutive days within two-and-a-half weeks of collection. These animals were housed as needed in an outdoor aviary (approximately 2.4 m long × 1.2 m wide × 2.3 m high) and were provided food (millet, cracked corn and sunflower seeds) and water ad libitum.
Experimental procedures
Implantation of electrodes
EMG electrodes were surgically implanted in two muscles considered to be potentially important sites of shivering thermogenesis. Birds were anesthetized with an intramuscular (in the left thigh) injection of a mixture of 25 mg ml−1 ketamine and 2 mg ml−1 xylazine (Rompun) in physiological saline at an initial dose of approximately 1 ml kg−1 body mass. Additional administrations of anesthetic were given as needed at one-third to half the original dose. When the birds no longer responded to tactile stimulation, up to eight bipolar electrodes were implanted in the midbelly of the M. pectoralis pars thoracicus (PECT) and the M. gastrocnemius (GAST) (nomenclature as in Vanden Berge, 1979). Care was taken to standardize the location of electrodes in all birds. Electrodes were inserted in two sites in the PECT muscle in order to examine the possibility that a spatial heterogeneity exists within this muscle during shivering, as has been demonstrated in the pigeon during flight (cf. the sternobrachialis and thoracobrachialis heads of the pectoralis; Dial et al. 1988; see Vanden Berge, 1979, for nomenclature). One site (medial PECT), was situated in the anterio-posterior mid-belly of the pectoralis just lateral to the carina. The second site (lateral PECT), was situated lateral to the first and just posterior to the head of the humerus.
Electrodes were constructed in the ‘simple double hook’ design described by Loeb and Gans (1986, p. 116) using 0.076 mm diameter, Teflon-coated stranded stainless-steel wire. The two leads were stripped for approximately 0.5 mm at the ends and were twisted several times to hold the tips 1.5–2 mm apart. The electrodes were inserted percutaneously to a depth near the middle of the muscles (approximately 2–6 mm, dependent on the size of the bird) using a no. 23 or 24 hypodermic needle, so that the bared ends of the leads were aligned parallel to the muscle fibers. The needle was removed, and the other ends of the leads were passed subcutaneously to exit points along the midline of the back. The wires were glued to the skin at these exit points with formulated cyanoacrylate (Nexaband, Bionexus, Inc.) and soldered to gold connector pins. Birds were allowed 6–8 h (nestlings) to overnight (some adults) to recover fully from the surgery and anesthesia before they were used in the experiments. All birds tolerated surgery well and recovered fully with no obvious ill side effects.
Experimental protocol
After full recovery from surgery, the birds were fed Tenebrio larvae ad libitum, reweighed to the nearest 0.01 g on a Mettler top-loading balance (Mb,in), and their cloacal body temperature (Tb,in) was taken with a calibrated copper–constantan thermocouple. Individual birds were subsequently placed, unrestrained, into metabolic chambers constructed from new 0.95 l (nestlings) or 3.8 l (adults) paint cans. The inside of the chambers was painted black to ensure high emissivities, and the lids were equipped with three ports, one each for the incurrent and excurrent air lines and the third to accommodate the EMG leads. All three ports were sealed tightly to prevent leakage. The bird rested on a 1.9 mm mesh platform fashioned to fit snugly at a height of approximately 1.5 cm from the bottom of the chamber. Metabolic chambers were subsequently placed in a controlled-temperature cabinet (Forma Scientific) maintained at thermoneutral temperatures (range 33–40°C for nestlings and 30–35°C for adults), and the bird was allowed to equilibrate for 5–15 min before measurements were taken. Nestlings were generally inactive for the remainder of the experiment, a situation typical for recently fed passerine birds placed in a dark chamber.
After this initial equilibration period, and EMG were monitored continuously for the duration of the experiment, first at thermoneutral temperatures and then during gradual cooling. Ambient temperatures were lowered at a rate of approximately 0.3°C min−1 (see Olson, 1992). Metabolic rates were measured using open-circuit respirometry (Depocas and Hart, 1957). Flow rates (; cm3 min−1) and the fractional concentration of oxygen in the excurrent air were collected throughout the experiment at a sampling rate of 50 min−1via an Interactive Systems A/D converter and Apple IIe microcomputer, averaged over the minute, and the resulting data were recorded on diskette for calculation of metabolic rates (see below). was accurately regulated in the incurrent stream on a Brooks model 5841 mass flow meter and was based on dry volumes converted to standard temperature and pressure (STPD). Flow rates were kept stable during an experimental run, but varied among runs according to the size of the bird (range 296 cm3 air min−1 for 3-day-old birds to 656 cm3 min−1 for adults) to ensure optimal differences in oxygen concentration between the incurrent and excurrent circuits. was measured in dried, CO2-free excurrent air using a calibrated Applied Electrochemistry S-3AII O2 analyzer. Drierite (8 mesh; W. A. Hammond Co.) and ascarite (8–20 mesh; Thomas Scientific) were used to absorb water vapor and carbon dioxide, respectively.
For the measurements of muscle activity, the electrical output from each electrode was amplified 1000-fold with a calibrated Tektronix type FM-122 preamplifier (bandpass 8 Hz and 10 kHz). The amplified EMG signals were monitored continuously on an oscilloscope and concurrently stored on a Sony four-channel FM reel-to-reel tape recorder (model TC 788-4) at approximately every 1°C interval (approximately every 3 min) throughout the experiment and at the threshold temperatures for each muscle (determined from output on oscilloscope). In addition, EMG signals were recorded during voluntary movements for each muscle. Electrical activity of the heart did not interfere with the EMG measurements in the PECT in any bird. The integrity of all electrodes from which data were collected was verified by inducing movements at the beginning and end of each experimental run.
Ambient temperature was measured with calibrated copper–constantan thermocouples and stored on a Kaye multichannel data logger. At the termination of the experiment, the birds were immediately removed from the chamber, their body temperature (Tb,out) measured within 20 s, and then weighed to the nearest 0.01 g on a top-loading Mettler balance (Mb,out). Animals were not allowed to eat or drink during the experiment.
Data analysis
Metabolic rates
Muscle activity
Mean rectified values appear to be a reliable index of the intensity of shivering thermogenesis. First, Umrv correlates well with the rate of heat production in adult birds, in which shivering is the primary mode of regulatory thermogenesis (Hohtola, 1982a). Second, the calculation of Umrv is robust and is less sensitive to sampling rate; it does not rely on the determination of the number of spikes, but is derived from a direct integration of the signal (Jayne et al. 1990).
To assess the patterns of recruitment within the muscles of birds at several ages, spectral analyses of the power output in the frequency domain were performed on selected electromyograms using the Fast Fourier Transform (FFT) algorithm of Cooley and Tukey (1965). Data were digitized at 4000 Hz, maintaining a frequency sensitivity in the analysis of 2000 Hz based on Nyquist sampling criteria. The relatively wide bandpass (8 Hz to 10 kHz) provided an ample range of frequencies to explore the possibilities that the spectral characteristics of the electrical output of the muscles change during ontogeny and with thermal stress (see West et al. 1968; Hohtola, 1982a). Three analyses were performed. First, FFTs were performed on EMG traces collected during shivering in the medial PECT and GAST of birds of all ages to determine whether the spectral characteristics of shivering change ontogenetically. Next, identical analyses were performed on EMGs during movements in these muscles (using the same electrodes as those above) to examine the possibility that the dominant frequencies of involuntary and voluntary movements differ. Finally, the effect of hypothermia on recruitment in these muscles during shivering thermogenesis was assessed.
Statistical analyses
Parametric statistics were used to test for differences among groups. An analysis of covariance (ANCOVA) was used to evaluate inter-individual variation and the effect of age on the relationships between temperature and several variables related to shivering intensity. In cases where temperature did not have a significant effect, an analysis of variance (ANOVA) was performed to test for differences among age groups. Because the data were not orthogonal, all ANOVAs and ANCOVAs were calculated using least-squares analysis (SYSTAT; Wilkinson, 1989). To determine the relationship between the EMG output of the two sites of the PECT muscle, the shivering intensities in the two lobes of the PECT were regressed using a model II regression analysis (non-linear least absolute deviations algorithm) available in SYSTAT, because both the x and y variables were empirically determined and therefore were subject to an equal amount of variability. Even though care was taken to standardize the placement of electrodes across animals, quantitative comparisons of absolute EMG activities among tissues are avoided here (see Hemingway, 1963). However, quantitative analyses of the relative power output of a muscle based on the recordings from a given electrode (e.g. factorial increases) are considered valid, because neither the electrodes nor the animal were disturbed throughout the duration of the experiment. Statistical significance was accepted at the 0.05 level or, where appropriate, was adjusted to a lower level with the Bonferroni procedure.
RESULTS
Pattern of shivering
Two distinct patterns of shivering exist in birds. One pattern is characterized by a continuous electrical output by the muscle (‘continuous’), although variations in the intensity of the output may exist over the short term as well as over the long term (see West et al. 1968). The second pattern (true ‘bursting’) is characterized by an alternation of electrically silent periods with periods of more-or-less intense activity. These patterns are easily distinguished with electromyographic techniques (Fig. 1), as are both from the myoelectric events associated with movements.
Shivering in the PECT and GAST muscles was nearly always continuous in all the red-winged blackbirds monitored in this study, despite significant changes in the Umrv with age and Ta. However, the distribution of rectified values over the 5 s of data that were digitized (20 000 points) in the PECT and/or GAST of some animals varied significantly from a normal distribution (P<0.002, P<0.0001 and P<0.05 for the medial PECT, lateral PECT and GAST, respectively). In all of these cases, the distribution of values was platykurtic relative to a Gaussian distribution (i.e. g was significantly greater than 1.253; P<0.05 in a paired t-test), indicating that the intensity of shivering varied significantly over the short term (see Figs 1 and 2). These deviations, especially marked in the lateral PECT of adult-1 and the GAST of 8d-2 (Fig. 2), were not accounted for by changes in either temperature or mass-specific metabolic rate (P>0.05; ANCOVA). True bursting was only rarely observed and was confined to the GAST muscle of birds 5d-2 and 8d-2 and the medial PECT of 3d-2.
The ontogeny of heat production in red-winged blackbirds Oxygen consumption
The mass-specific rate of oxygen consumption represents the overall metabolism of an animal including the costs associated with maintenance, thermoregulation and activity. Costs of maintenance (including SDA, specific dynamic action) are estimated by measuring the RMR, or the metabolic rate at thermoneutral temperatures when the bird is inactive. The RMRs of the six nestling and three adult red-winged blackbirds were similar to those reported in a larger study (N=69) of the development of thermoregulation in this species (Olson, 1992; compare filled circles with regression line in Fig. 3). As in the earlier study, RMR increases very nearly isometrically with body mass throughout the nestling period. Consequently, the mass-specific RMR of the six nestlings was constant, averaging 3.13±0.22 cm3 O2 g−1 h−1 (S.E.M.), a value very close to that (3.1 cm3 O2 g−1 h−1) previously reported for red-winged blackbird nestlings (Hill and Beaver, 1982; Olson, 1992). As in the larger study, the corresponding basal rate for the adults (2.83±0.13 cm3 O2 g−1 h−1; N=3) was slightly below that of the nestlings.
The ontogeny of thermoregulatory capabilities during the postnatal period similarly conformed to that described in the earlier study. Any incremental increases in mass-specific above RMR during cooling reflect the costs associated with thermogenesis, providing that the animal remains inactive. As expected, the mass-specific of the three adults increased linearly during gradual cooling below the TLC (Fig. 4A). Both the TLC values for each animal (22, 30 and 27.5°C for adult-1, adult-2 and adult-3, respectively) and the slope of the curve relating mass-specific to Ta below TLC were lower than those of the nestlings, partly because of the relatively well-developed insulation of mature birds. None of the adults reached summit metabolism in the conditions of this study.
The metabolic rates of 3-day-old nestlings decreased substantially in response to gradually cooling Ta, although nestling 3d-1 stabilized its mass-specific at a value similar to or slightly above (approximately 15% at 27°C) mass-specific RMR down to a Ta of 25°C before decreasing to the low level of the other bird (Fig. 5A). As a result of their limited thermoregulatory capabilities, these birds were severely hypothermic when removed from the metabolic chamber at the end of the experiment (Tb,out=15.6 and 20.1°C for nestlings 3d-1 and 3d-2, respectively).
The thermoregulatory capabilities of the two 5-day-old nestlings exceeded those of their younger counterparts. The TLC of each nestling was recognizable (37 and 30°C for 5d-1 and 5d-2, respectively), and these nestlings were capable of increasing their mass-specific slightly above standard levels (Fig. 6A) before succumbing to the deleterious effects of decreasing Ta. Bird 5d-1 increased its by a maximum of 30% above RMR at 29°C during cooling, whereas bird 5d-2 achieved a 42% increase in at 27°C. However, these increases were transient and both metabolic rate and body temperature decreased with further cooling (Tb,out=16.0 and 17.1°C for 5d-1 and 5d-2, respectively).
The thermoregulatory ability of nestlings improved even more markedly between days 5 and 8 after hatching. Nestlings 8d-1 and 8d-2 exhibited more mature endothermic responses below the TLC for each bird (33 and 34.5°C, respectively; Fig. 7A). Furthermore, summit metabolic rates were reached at the lower Ta values of 22 and 12°C, at which temperatures the nestlings achieved a 1.9-and 2.4-fold increase, respectively, in mass-specific over basal levels. Metabolic rates decreased with a further drop in Ta, however and, consequently, these animals were also hypothermic at the end of the experimental run (Tb,out=17.5 and 30.0°C, respectively).
All nestlings tolerated hypothermia, recovering completely after removal from the metabolic chamber. Similar abilities have been noted for other nestlings (e.g. willow ptarmigan, Aulie, 1976; ring-billed gulls, Dawson et al. 1976).
Shivering
The ontogenetic improvements in the capability of nestlings to shiver roughly paralleled those previously described for metabolic rate. Thermogenesis by adult red-winged blackbirds always balanced the rates of heat loss experienced under the conditions of this study. Therefore, shivering in both the PECT and GAST muscles probably did not reach maximal intensities (Fig. 4B–D). In all adults, Tth for shivering in the medial PECT was identical to that in the lateral PECT, and both corresponded with TLC. The Tth in the GAST was always below the TLC. The Umrv of the medial PECT of adult-2 did decrease slightly between 15 and 10°C (Fig. 4B). The reasons for this decrease were not clear. However, it certainly does not represent either a maximal expandability for this muscle or a failure of the overall thermogenic response, for this animal did not become hypothermic. It is intriguing that the decrease in the EMG activity of the medial PECT coincided with the intensification of shivering in the GAST muscle in this bird (Fig. 4). This correspondence suggests that the heat produced by the GAST muscles obviated the need for maintaining the high shivering intensities in the PECT muscles.
In sharp contrast, 3-day-old nestlings possessed only a rudimentary ability to generate heat through shivering. Significant shivering was detected in only one animal (3d-1) and was confined to the PECT, particularly the lateral PECT; no accompanying change in the EMG activity was evident in the GAST (Fig. 5D). The increased EMG activity of the PECT in this bird was first detected at 31°C and increased to its highest, albeit still relatively low, value at 20°C in both the lateral PECT (6.8-fold increase in Umrv above baseline) and medial PECT (2.8-fold increase) before decreasing to the baseline value at 10°C (Fig. 5). This pattern of shivering coincided temporally with the extended plateau in metabolic rate observed in this bird (Fig. 5). The PECT of 3d-2 began to shiver at 37.5°C; however, the response was of an extremely low intensity. The lack of a significant shivering thermogenesis in this nestling was also consistent with the pattern of changes (immediate decreases) in mass-specific observed with cooling.
Shivering thermogenesis improved in both 5-day-old nestlings, judging from the higher maximal increases in the intensity of shivering in their PECT muscles (4.3-and 9.3-fold in the medial PECT and 17.8-and 14.8-fold in the lateral PECT of birds 5d-1 and 5d-2, respectively; Fig. 6). The Tth values of the medial and lateral PECT of each bird were identical to the corresponding TLC (37 and 31°C for 5d-1 and 5d-2, respectively) and the pattern of increases and subsequent decreases in shivering intensity in the two sites during cooling were very similar (e.g. highest factorial increases at 25–30°C; Fig. 6). The GAST muscles only contributed modestly to shivering in both 5-day-old nestlings; the Tth for shivering in the GAST was lower than the TLC in both cases (29 and 26°C, respectively), and the GAST in 5d-1 achieved only a 4.9-fold increase in Umrv above baseline at 20°C and that in 5d-2 achieved a 3.5-fold increase at 15°C.
The capability for muscular thermogenesis improved markedly between 5 and 8 days of age. The EMG activity of the medial PECT muscles of the two 8-day-old nestlings increased by a maximum of 35.8-and 24.2-fold at 10 and 15°C relative to the baseline values at thermoneutral temperatures, while the corresponding increases for the lateral PECT were 72.1-and 54.9-fold over baseline (Fig. 7). Again, the two sites of the PECT muscle were coordinated in both the Tth (33 and 34.4°C for 8d-1 and 8d-2) and overall pattern of increases in EMG activity throughout the gradual cooling experiments (Fig. 7), even though the magnitude of the increases differed. Unlike the younger nestlings, however, significant shivering was detected in the GAST of one 8-day-old nestling. The GAST of bird 8d-1 began shivering at 31°C, a Tth only slightly below those of the medial and lateral PECT muscles, and achieved a maximum 22.3-fold increase in EMG activity over basal values (Fig. 7). The Tth of the GAST of the other 8-day-old (8d-2) was 27°C, and the EMG activity increased only 3.8-fold before reaching a plateau (Fig. 7). As illustrated in Fig. 8, the maximal factorial increases (Umrv/Umrv at thermoneutral temperatures) in shivering intensity for all three muscle sites in 8-day-old nestlings were higher than predicted assuming a simple linear pattern of ontogenetic improvements. The maximal factorial increases in shivering increased further after day 8 (data not shown); however, the extent of these improvements is unknown.
Overall, changes in metabolic rate corresponded with changes in the electrical output from the PECT (but not from the GAST) muscle. The EMG output of the PECT was significantly and positively correlated with oxygen consumption (P<0.001 for both the medial and lateral PECT; N=56; ANCOVA; Fig. 9). The trajectories of increases in shivering intensity (expressed as a factorial increase) were similar among the animals of a given age. However, the rate of increases in EMG output differed significantly among the age groups (P<0.001 for both sites in PECT; ANCOVA); in general, the EMG output of the PECT increased more and at a significantly faster rate with a given increase in metabolic rate in older nestlings than in younger nestlings (compare the steepness of the curves for 8-day-old nestlings with those of 5-day-old and 3-day-old nestlings in Fig. 9). Interestingly, shivering intensity of the PECT muscles at a given metabolic rate was consistently higher when the bird was becoming hypothermic and therefore suffering a decrease in its rate of oxygen consumption (dashed lines in Fig. 9) than when the bird was normothermic at milder ambient temperatures (solid lines).
Spectral analysis
Fig. 10 presents the results of the spectral analyses of the EMGs during shivering and voluntary movements for an ontogenetic series of red-winged blackbirds of several ages.
Both the range and absolute value of the prominent frequencies in the PECT during shivering increased dramatically with age. The dominant frequencies in the EMG signal from this muscle in 3-day-old nestlings were in the range of approximately 50–150 Hz, whereas a substantial proportion of the corresponding ranges in 5-day-old (approximately 40–300 Hz) and 8-day-old (40–400 Hz) nestlings occurs at progressively higher frequencies. Further increases in these parameters occur after fledging as well, judging from the substantial differences between 8-day-old nestlings (i.e. just before fledging) and adults (40–925 Hz).
The frequency spectra of voluntary movements were remarkably similar to those of shivering for a nestling at a given age (Fig. 10), though the amplitudes of all elements of the spectra are usually higher during the former activity. Therefore, age rather than the type of activity is more important in determining the frequencies of contraction in the PECT muscles.
The characteristics of the corresponding spectral analyses of the GAST muscles (Fig. 11) differed in several respects. This muscle shivers at lower frequencies than the PECT muscles of the same animal. In addition, the GAST of both 8-day-old nestlings and adults shivers at frequencies below 350 Hz, and this does not change appreciably after fledging. Finally, unlike the PECT, the dominant frequencies during voluntary movements appear to be lower than those during shivering in a given GAST muscle (Fig. 11).
Hypothermia significantly affected the spectral characteristics of the EMG signal of the PECT muscle during shivering. A selective attenuation occurred in the higher-frequency components of the signal (Fig. 12).
Spatial heterogeneity within the PECT muscles
A significant spatial heterogeneity within the PECT muscles was not detected in any bird. The medial and lateral PECT were recruited simultaneously (i.e. had identical Tth values) and the overall patterns of changes in the intensity of shivering in the two sites with decreasing temperature were similar (see Fig. 13 for the highly significant linear relationship between factorial increases in shivering intensity for two sites of the PECT in each animal; each r2≥:0.862, P<0.005). Interestingly, however, the responses of the two sites differed quantitatively. The factorial increases in Umrv in the lateral PECT were consistently higher than the corresponding changes in the medial PECT in all of the nestlings. The rates of the increases in the lateral PECT averaged 2.3-fold (range 1.2-to 5.7-fold) higher than those in the medial PECT. The case for adult-2 was the opposite of that for the nestlings.
DISCUSSION
The ontogeny of heat production in the red-winged blackbird
The establishment of thermoregulatory ability in nestling birds is a complex process, dependent upon the functional maturation of the neural, muscular and endocrine systems. Thermoregulatory competence requires that these systems be sufficiently developed to support a rate of heat production that is both effective and sustainable. Measurements of the mass-specific and the mean rectified value of the EMG of the skeletal muscles during shivering provide two indices of thermogenesis. Simultaneous measurements of these two indices can elucidate whether shivering or non-shivering thermogenesis is the primary mode of heat production. In addition, measurements of EMG, in particular, are instantaneous and therefore are a very sensitive indicator of changes in the thermoregulatory behavior of the bird.
The ontogenetic changes in both thermogenic indices indicate that the capacity for heat production in nestling red-winged blackbirds improves markedly with age. The metabolic rates of 3-day-old nestlings decrease during exposure to gradually falling Ta; at best, these nestlings are only able to maintain mass-specific at levels similar to or slightly above those at thermoneutral temperatures over a small decrease in Ta before metabolic rates decreased with further cooling. The 5-day-old nestlings increase mass-specific modestly above RMR when exposed to the same experimental conditions. In contrast, 8-day-old nestlings can increase metabolic rates approximately 2- to 2.5-fold over basal levels and sustain these elevated rates for a longer time before becoming hypothermic. The profound improvements in thermoregulatory capabilities with age are similar to the results obtained in more extensive ontogenetic studies of red-winged blackbird nestlings that measure metabolic rates using the same protocol (Olson, 1992) or in steady-state conditions (Hill and Beaver, 1982).
The time course of the ontogenetic improvements in shivering ability parallel those observed for mass-specific . As with metabolic rates, older nestlings and adults attain much larger factorial increases in the intensity of shivering during exposure to cold temperatures. Furthermore, the intensity of shivering at a given Ta below the TLC is positively correlated with age at every temperature, particularly in the PECT muscles, suggesting that the ability to adjust shivering to a level commensurate with the severity of the cold challenge also improves during postnatal development. Similar improvements in shivering ability have been observed in both precocial and semi-precocial (willow ptarmigan, Lagopus lagopus, Aulie, 1976; capercaillie, Tetrao urogallus, Hissa et al. 1983; western gull, Larus occidentalis, Eppley, 1987) and altricial (cattle egret, Bubulcus ibis, Hudson et al. 1974) nestlings.
The larger factorial increases in the intensity of shivering achieved by older birds may be due in part to concomitant increases in the number and perhaps the size of the motor units. All nestling and adult red-winged blackbirds regulate the intensity of shivering thermogenesis primarily through adjustments in the amplitude of the EMG and not through significant adjustments in frequency. Graded responses during shivering are accomplished in part through differential recruitment of motor units in skeletal muscle (Hemingway, 1963; Hohtola, 1982b). Growth of the effector tissues for shivering thermogenesis during this time undoubtedly contributes to these improvements. For example, the mass of the two PECT muscles in red-winged blackbirds represents 1.8±0.3% (mean ± 1 S.D.; N=6) of the total body mass at hatching, but increases disproportionately to body mass during the postnatal period to 5.1±0.7% (N=14) and 6.6±1.4% (N=17) of the mass of 8-and 10-day-old nestlings, respectively. Further substantial increases occur after fledging, so that the two PECT muscles represent 15.5±0.9% (N=12) of the total body mass of adults (Olson, 1990). Such an increase would allow larger factorial increases in the calorigenic output of the muscles, thereby improving the thermogenic capability of growing nestlings (O’Connor, 1975; Aulie, 1976). In addition, the morphology of the PECT muscles changes over this same period. Substantial increases in both the size and number of muscle fibers occur in the PECT muscles of the red-winged blackbird during the postnatal period, even though the relative proportion of fast-twitch fiber types is apparently relatively stable (Olson, 1990). More work is required to clarify the details of postnatal growth of skeletal muscle in this and other altricial species, particularly the changes in the patterns of innervation and the organization of motor units during myogenesis. However, it is clear that the growth and development of skeletal muscles are responsible for allowing older nestlings and adults to defend Tb levels over a wider range of cold challenges than can younger nestlings.
In addition to the quantitative improvements in the intensity of shivering, qualitative changes in the neuromuscular system apparently also occur, at least in the PECT muscle. The range and value of the dominant frequencies during both shivering and voluntary movements in this muscle increase throughout the postnatal period (see Fig. 10). This trend occurs despite an accompanying order-of-magnitude increase in body mass, an allometric relationship opposite to that found in interspecific comparisons of muscles from adult birds (Hohtola, 1982b). In this regard, it is noteworthy that the dominant frequencies obtained for the PECT muscles of adult pigeons (Columba livia), evening grosbeaks (Coccothraustes vespertinus) and common grackles (Quiscalus quiscula) (West et al. 1968; Hohtola, 1982a) are lower than those of the adult red-winged blackbird measured here. No other study has addressed the question of ontogenetic changes in the spectral characteristics of shivering.
The increasing importance of higher-frequency components in the EMGs of older nestlings and adults suggests one or more of the following possibilities. First, higher-threshold, fast-twitch fibers may contribute more to the production of heat during shivering in the muscles of older nestlings and adults. Alternatively, ontogenetic changes in the electrical conductivity of the membranes of the motoneurons or muscles themselves, in the efficiency of the contractile machinery (e.g. in the production of ATP or the cycling of Ca2+) of the muscle and/or in the development of the γ-afferent motoneurons (fusimotor system) are perhaps responsible (see, for example, Hohtola, 1982b). Such changes are not totally unprecedented in birds: the electrical properties of the membranes of several skeletal muscles of the chick change during postnatal development (e.g. see Karzel, 1968). Furthermore, the concurrent development of the innervation patterns and the morphology of the muscle apparently follow different trajectories, at least in the precocial chick. For instance, the caudal latissimus dorsi muscle in embryonic chicks contracts more slowly than in the adult, despite the presence of the complement of fast-type myosin light chains typical of the adult muscle (Pette et al. 1979). The greater contribution of higher-frequency signals may indicate a decreased synchronization of motor unit firing and a concomitantly lower tremor amplitude in older nestlings and adults. This trend would be adaptive because it would reduce convective heat losses (Hohtola and Stevens, 1986). Further research is required to resolve the ontogenetic changes both in the muscles themselves and in the interaction between the muscles and the α- and γ-motoneurons innervating them.
Unexpectedly, a significant shivering response was observed in the PECT of one 3-day-old nestling (3d-1), suggesting that at least some 3-day-old nestlings are capable of muscular thermogenesis. Such incipient endothermic capabilities are modest, but clearly indicate that active thermogenesis is important in nestlings several days before the primary transition to endothermy (stage 3 in Olson, 1992). Such active thermogenesis undoubtedly contributes to the stability of over the extended plateau phase during gradual cooling in this nestling and in several 3-and 4-day-old nestlings in the larger study of Olson (1992). These capabilities suggest that the extended plateau in these birds is not entirely due to the inertial advantages of the increasing body mass during this period (the ‘inertial phase’ of Hill and Beaver, 1982), but rather is due to metabolic heat production. The heat produced through shivering apparently suffices to offset the van’t Hoff effect for a short period as ambient temperatures decrease. However, further cooling renders shivering by these nestlings inadequate, owing to their relatively high surface-to-volume ratios and consequent high rates of heat loss.
Primary role of the pectoralis muscles in shivering thermogenesis
Among the muscles studied here, the PECT muscles are the principal sites of shivering thermogenesis in all nestling and adult red-winged blackbirds. Shivering in these muscles is a ‘first line defense’ against cold in red-winged blackbirds. The Tth for shivering (determined to the nearest 0.5°C from raw EMGs) in both sites of the PECT muscles coincides with the TLC in the rate of oxygen consumption, and the subsequent increases in EMG activity with further cooling correlate well with the corresponding increases in mass-specific (Figs 4–7, 9). Such a correlation exists despite the fact that other muscle groups also contribute to overall heat production. The PECT muscles also have both the highest EMG activities and much larger factorial increases in electrical output with decreasing Ta. Furthermore, the PECT muscles in red-winged blackbirds can apparently shiver before any significant use of them is made in voluntary movements. A primary role of the PECT muscles has also been observed in precocial nestlings (Hissa et al. 1983; Barré et al. 1985) and in adult birds (e.g. Steen and Enger, 1957; West, 1965; Hohtola and Stevens, 1986).
In contrast, shivering in the leg muscles was less important in both nestlings and adults. Even though the GAST muscles are functional and are used for postural support soon after hatching, shivering was not detected in the GAST of either 3-day-old and did not contribute significantly to heat production until nestlings were more than 5 days of age (8 days in this study). Even when shivering was detected, it was of relatively low intensity (Figs 5–8), and the threshold temperature for shivering in this muscle was always lower than that of the PECT muscles and therefore lower than the TLC.
The relatively large size and proximity of the PECT muscles to the heart and other vital organs probably account for the central role of this muscle in heat production. Both pectoralis muscles constitute 15.5% of the total body mass of the adult red-winged blackbird (Olson, 1990). In addition, the lack of significant involvement of the PECT muscles in postural support when the bird is at rest probably contributes to their importance as a site of shivering thermogenesis. In contrast, the muscles of the leg are important in postural support and, therefore, vigorous shivering in these muscles would compromise postural stability during perching and could increase convective heat loss by increasing the overt tremor of the whole body. In order to limit overt tremor during intense shivering in the PECT muscles, either the primary antagonistic muscles must contract simultaneously or some other mechanism must prevent significant movement of the wing. In this regard, it is interesting that the antagonistic muscle to the PECT, the supracoracoideus, is so small. The role of this latter muscle during shivering is unclear, but it was active in the adult birds monitored in this study (data not shown). Although the small size of this muscle is obviously sufficient for effecting the upstroke (recovery phase) of the wing during flight, it could severely limit the potential heat production of the pectoralis if it alone were responsible for offsetting the contraction of the PECT during intense shivering thermogenesis. Alternatively, another mechanism, for instance an adjustment of the position of the wing, might obviate the need for an equal output from the supracoracoideus and PECT muscles to prevent large overt tremor amplitudes during vigorous shivering. Perhaps birds are able to ‘lock’ their wings mechanically during shivering, thereby allowing a higher heat output by the PECT muscles without an equivalent simultaneous contraction of the supracoracoideus.
The role of shivering as a first line defense against cold in red-winged blackbirds contrasts with the case in mammals with documented regulatory NST, in which the Tth of all skeletal muscles is below the TLC (e.g. Heldmaier et al. 1985). Such a discrepancy between TLC and Tth in mammals has been taken to indicate that hormonally mediated heat production is the first line defense in these animals (see Hemingway, 1963, for a review). However, a supplementary role of regulatory NST at lower temperatures cannot necessarily be ruled out in red-winged blackbirds. If present, however, the mechanisms of such NST remain unclear. The adult birds studied to date lack brown adipose tissue (Freeman, 1967; Johnston, 1971; Olson et al. 1988; Saarela et al. 1991), the specialized tissue and principal site of NST in mammals. Moreover, in contrast to the profound calorigenic effect of norepinephrine, the agonist in the brown adipose tissue of mammals, the corresponding effect in adult birds is at best equivocal. Several investigations failed to find a significant thermogenic effect of this monoamine (e.g. Chaffee et al. 1963; Hissa and Palokangas, 1970; Hissa and Rautenberg, 1974; Marley and Stephenson, 1975; Hissa et al. 1975, 1980a; Koban and Feist, 1982). The results of some studies have been taken to indicate that catecholamines are calorigenic in birds (e.g. El-Halawani et al. 1970; Arieli et al. 1978). This conclusion, however, is often inferred from the thermolytic effects of the administration of the β-blocker propranolol, which in turn has been shown to also inhibit shivering thermogenesis in birds (Hissa et al. 1980b). Regulatory NST may be present only in young birds, particularly in chicks of precocial species, as has been implicated in response to the administration of catecholamines in the chick of the domestic fowl (Freeman, 1966; Wekstein and Zolman, 1968, but see Palokangas and Hissa, 1971) and glucagon in king penguin chicks (Barré and Rouanet, 1983) and Muscovy ducklings (Barré et al. 1987). Furthermore, such NST may be inducible in nestling birds, for example through experimental cold-acclimation (e.g. in Muscovy ducklings, Barré et al. 1985). In another study, Barré and his colleagues (Barré et al. 1986) identified an uncoupling effect of FFA in mitochondria of skeletal muscle in Muscovy ducklings, suggesting the presence of NST in this tissue. Further research is required to determine whether regulatory NST plays a significant role in the production of heat in either nestling or adult birds.
Thermosensitivity of shivering
Both the intensity and spectral characteristics of shivering are profoundly affected by hypothermia in red-winged blackbird nestlings. Although the exact time course of the development of hypothermia in the experiments reported here is not known, the decrease in the temperatures of nestlings that did become hypothermic was probably relatively uniform, beginning at the time when mass-specific decreased or stabilized with a further drop in Ta. EMG activity and mass-specific decline roughly in parallel (see Figs 6 and 7 for 5-and 8-day-old nestlings, respectively). In addition, the higher-frequency components are attenuated more by hypothermia than are the lower ones (Fig. 12). Whether this reflects a differential deactivation in the effector tissues themselves (e.g. the fast-twitch muscle fibers or a change in the size-dependent recruitment of motor units; Hohtola, 1982b) or implicates a general effect of low body temperatures on the central nervous system (including the afferent and efferent pathways involved in the control of shivering thermogenesis), or both, is unknown. Cold temperatures may hyperpolarize membranes by affecting the ionic channels and/or enzyme-mediated pumps in the membranes or synapses (especially the neuromuscular junction). Alternatively, the profound shift in the dominant frequencies in the EMG signal may reflect a particular thermosensitivity of the thermosensors in the spinal cord and periphery and/or the motoneurons of the spinal cord. Unlike the case in mammals, the hypothalamus apparently does not play as important a role as either an afferent or central processor in the control of shivering in birds (Snapp et al. 1977; Simon-Oppermann et al. 1978). Rather, extrahypothalamic centers in the spinal cord (Rautenberg, 1969; Simon, 1974; Necker and Rautenberg, 1975; Görke and Pierau, 1979) and thermosensors in the periphery (e.g. feathered regions in the back and wing, Necker, 1977) appear to be more important. Whatever is the case, the results of this study indicate the importance of the homeothermy of nestlings in the field (Olson, 1991) and suggest that external sources of heat would be required to restore normothermic Tb values if young nestlings are cooled substantially.
Pattern of shivering
The pattern of shivering in red-winged blackbirds is nearly always continuous, varying little throughout the postnatal period. Variations in intensity (see West et al. 1968) are observed in red-winged blackbirds of several ages and, more rarely, a true bursting pattern of shivering occurs. Hohtola and Stevens (1986) have argued that the pattern of shivering is dependent upon the nature of the supply of ATP in the muscle; muscles with a large aerobic capacity (and therefore a more sustained supply of ATP) will shiver continuously, whereas those muscles more dependent upon anaerobic metabolism to supply ATP will exhibit a bursting pattern. In birds dependent upon sustained flight, the high aerobic capacity of the pectoralis muscles might ‘pre-adapt’ them for sustained shivering thermogenesis, resulting in a continuous pattern of shivering. Interspecific comparisons of shivering in the pectoralis muscles of columbiforms and passeriforms (with relatively high aerobic capacities; Calder and King, 1974) with that of galliforms (with relatively low aerobic capacities) offer initial support for this hypothesis (Hohtola and Stevens, 1986). However, it is noteworthy that shivering in red-winged blackbirds is usually continuous throughout the postnatal period, despite profound increases in the aerobic capacities (Olson, 1990), dominant frequencies during shivering (this study), and changes in the morphology (Olson, 1990) of the PECT muscle over this same period. This suggests that factors other than the biochemical profile of the muscles are important determinants of the pattern of shivering.
ACKNOWLEDGEMENTS
I thank Dr and Mrs Gosline for allowing me access to their pond to collect nestlings and to Dr Richard Dolbeer for helping me make arrangements to trap adult red-winged blackbirds in Huron, Ohio. Drs William Dawson, Carl Gans and Richard Marsh provided the space and equipment essential to the completion of this study. Chris Young assisted me in the Fast Fourier Transforms. Drs William Dawson, Mary McKitrick and Bryon Doneen offered useful comments on an earlier version of the manuscript. The research was partially funded by a predoctoral fellowship through the Rackham Graduate School at The University of Michigan.