Temperature dependence of cardiac performance in the lobster Homarus americanus

The North American lobster Homarus americanus inhabits both the deep ocean waters and the coastal shallows and estuaries. Water temperature in its habitat varies over a range of 25°C(Lawton and Lavalli, 1995). Like other poikilotherms, lobsters are unable to regulate their own body temperature. However, they apparently detect changes in water temperature with great resolution, as they exhibit altered cardiac activity when temperature shifts by less than 0.5°C (Jury and Watson, 2000). Significant changes in environmental temperature pose an obvious physiological challenge to the lobster: how can its organs sustain appropriate performance, as each element of the system (including the ion channels, cell membranes, neural circuits and contractile machinery) is likely to depend on temperature in a different way? In the wild, lobsters undergo seasonal migrations (Campbell,1986; Campbell,1989; Campbell and Stasko,1986; Cooper and Uzmann,1971; Ellis, 2001; Ennis,1984; Estrella and Morrissey,1997; Watson et al.,1999). These migrations are thought to enable the lobsters to reach or remain at water temperatures favorable for growth, development and molting, all of which become compromised below 10°C (reviewed by Ennis, 1995; Waddy et al., 1995).

However, more recent data provide direct evidence that some lobster populations do not escape the extremes of winter. Temperature probes affixed to Gulf of Maine lobsters record extremely cold (0–6°C) water temperatures throughout six months of winter and early spring (Cowan, 2004). Under laboratory conditions, lobsters exposed to thermal gradients show a preference for significantly warmer water temperatures, in the range 13–20°C, and avoid water at higher temperatures(Crossin et al., 1998; Reynolds and Casterlin, 1979). In the wild, relatively warm temperatures are most likely to be found in shallow waters in summer months. Water temperature in shallow water is strongly affected by winds and tides. Temperature swings of greater than 12°C during the daily tides have been recorded at the level of the lobster traps (Fig. 1)(Manning, in press). Thus,lobsters face thermal shifts not only throughout the seasons but on shorter time scales, potentially even as they move in and out of thermoclines on the ocean floor.

In this study we explore how the cardiac system of the lobster responds to temperature changes in the physiological range. Because crustacean hearts are regulated by physiological inputs (neural, neurohormonal and chemosensory) as well as environmental influences (oxygen levels and salinity)(McMahon, 1999), it is also of interest to know whether temperature affects the response of the heart to other modulatory factors. We address this question by testing whether modulation of the cardiac performance by serotonin is temperature dependent. Serotonin is a neurohormone localized in and released from the pericardial organs in the lobster (Cooke,1966; Pulver and Marder,2002). It has been shown previously that serotonin modulates lobster hearts both in vivo and in vitro(Battelle and Kravitz, 1978; Guirguis and Wilkens, 1995; Wilkens and Kuramoto, 1998; Wilkens et al., 1996). Whether these modulatory effects might depend on temperature is unknown.

Here we characterize the effects of temperature on cardiac performance in the lobster hearts by measuring the strength and frequency of the heartbeat and the cardiac output of the heart in vitro. For comparison, we examine the temperature dependence of heart rates in intact animals. In addition, we test whether the modulation of in vitro cardiac activity by serotonin depends on temperature. The results demonstrate that multiple parameters of cardiac performance depend on temperature, and that these temperature dependencies interact to produce a relatively constant relationship between cardiac output and temperature. In addition, serotonin modulates the strength and frequency of the lobster heartbeat in a manner that is independent of temperature.

Lobsters (Homarus americanus Milne-Edwards 1837) were obtained from commercial sources and kept in artificial seawater at 5°C. To minimize the potentially confounding effects of seasonal variability, animals were maintained for a month or longer at 5°C before being used in experiments. Lobsters were anesthetized on ice, after which the heart was isolated for in vitro studies as described previously(Worden et al., 1995). Briefly, a small section of the dorsal thoracic carapace was removed with the heart attached and surrounding tissues were dissected away. The alary ligaments and muscles were left intact, providing stretch to the mechanoreceptors (Cooke, 2002). The preparation was pinned in a recording chamber ventral side up and submerged in 60 mls of lobster saline (462 mmol l^–1 NaCl, 16 mmol l^–1 KCl, 26 mmol l^–1 CaCl₂,8 mmol l^–1 MgCl₂, 11 mmol l^–1glucose, 10 mmol l^–1 Hepes, pH 7.4). The interior of the recording chamber was encircled by several loops of Tygon tubing through which refrigerated coolant (antifreeze:water in a 50:50 mix) was pumped by a refrigerated circulator (Isotemp model 1016P, Fisher Scientific, Pittsburgh,PA, USA) to control bath temperature in the range 2–25°C.

The organization of the lobster heart has been reviewed elsewhere(McMahon, 1995). Briefly,hemolymph enters the single chamber of the lobster heart through paired ostia and is pumped out through seven arteries. Isolated hearts continue to beat because of rhythmic neural output from the cardiac ganglion, located on the dorsal inner wall of the heart. Following isolation of the heart, the sternal artery was cannulated and perfused with temperature-controlled lobster saline at 3.5 ml min^–1 to maintain stretch. In some experiments the antennal arteries were tied off with 6.0 surgical silk and attached to a tension transducer (Grass Instruments, West Warwick, RI, USA; model FT-03) and amplifier (CyberAmp model 320; Axon Instruments, Union City, CA, USA) to record contractions of the heart during each heartbeat. Contractions were measured as force (g). In some experiments cardiac output out via the sternal artery was measured ultrasonically (in ml min^–1)using an inline flow transducer (Model T206, Transonic Systems, Ithaca, NY,USA). In these experiments the cannula was led back into the water at zero afterload and saline was perfused through the heart through one of the ostia. Temperature in the bath was continually monitored. All physiological signals were recorded on a video cassette recorder (VCR) and digitized by an analog to digital converter using pClamp software (Digidata1200 A-D converter and pClamp software from Axon Instruments-Molecular Devices, Union City, CA, USA).

In each experiment, the parameters of cardiac function were measured as the temperature was changed from 2 to 22°C at a rate of 0.9°C min^–1. In some experiments, the temperature was returned to 2°C and serotonin (1 μmol l^–1) was added to the saline perfusing the heart through the sternal artery cannula. After waiting for the serotonin effects to reach steady state (defined as a potentiation of contraction amplitude that varied by no more than 5% over a period of 5 min),temperature was again warmed from 2 to 22°C in the presence of serotonin.

To record cardiac activity in vivo two holes were drilled on the midline of the dorsal carapace for chronic implantation of electrodes. The holes were located 1 cm and 4 cm rostral to the posterior edge of the carapace. Analysis of the location of the holes by later dissection showed that they were in positions rostral and caudal to the perimeter of the underlying heart. Two wires insulated to within 0.5 cm of their tips were inserted into the holes, fixed in position with cryanoacrylate glue and duct tape, and connected to an impedance converter (UFI model 2991). The DC and AC outputs of the impedance converter were connected to a VCR as well as an analog to digital converter to record data with pClamp software. Animals were implanted at least 48 h before data were collected.

To record changes in cardiac activity as a function of the environmental temperature the animals were placed in a wire mesh cylinder surrounded by coils of tubing through which coolant was pumped by a refrigerated circulator. The cylinder was then submerged in an insulated chamber (23 cm×18 cm×16 cm) filled with 3.5 l of seawater (Crystal Sea Salt, Marine Enterprises International, Baltimore, MD, USA) until the dorsal carapace was 2–3 cm below the surface of the water. The temperature of the seawater bath was regulated by adjusting the refrigerated circulator. The bath temperature in the chamber was continually monitored by a temperature probe fixed in position over the dorsal carapace between the two recording electrodes. In all experiments the animals were acclimated to 2°C for at least 30 min before data were collected, during which time heart rate was monitored. During each experiment the temperature of the seawater was warmed from 2 to 22°C over a period of approximately 40–50 min.

To verify whether the temperature within the thoracic cavity of a living animal approximated the temperature of the surrounding water we repeated the temperature protocol while measuring the temperature within the pericardial sinus. The temperature of the pericardial sinus becomes isothermal with the external temperature within the first 5–10 min of the 30 min acclimation period at 2°C. Fig. 2 shows that the temperature within the pericardial sinus rises as external water temperature warms. In three experiments of this type the difference between internal and external temperature was 1.31±0.42°C (mean ±s.d.) over the temperature range from 2 to 22°C, confirming that the internal body temperature in the region of the heart closely approximates external water temperature.

For both in vivo and in vitro experiments, the frequency of the heart beat was measured over 60 s at each temperature, with at least five heartbeats recorded in each trial. In some in vivo experiments the heart occasionally stopped temporarily. If the heart failed to beat for 15 s or longer, the frequency data were analyzed by excluding the period of heart failure. The strength of the heartbeat was measured as the peak amplitude of the contraction with respect to baseline, at least five heartbeats were measured at each temperature. Acute values of Q₁₀ were calculated as Q₁₀=(fh₂/fh₁)exp[10/(t₂–t₁)]where fh₂ is the heart rate at temperature t₂ and fh₁ is the heart rate at temperature t₁(DeWachter and Wilkens, 1996). Stroke volume was calculated by dividing cardiac output by the heart rate.

Repeated measures models (Crowder and Hand, 1990) were used for the analyses of experiments involving multiple measurements on the same lobster as well as to investigate specific hypotheses concerning comparisons between groups or to specific temperatures. F tests were used to compare the values for cardiac parameters at different temperatures. Analyses were done using SAS software (The SAS Institute, Cary, NC, USA) version 9.1 module `PROC MIXED'.

To examine the temperature dependence of cardiac performance in the lobster, spontaneously beating isolated hearts were exposed in vitroto a steady temperature increase throughout the physiological range 2–22°C. As shown in the tension recordings in Fig. 3, temperature alters both the strength and frequency of the heartbeat. At the coldest temperatures the contractions of the heart are large in amplitude and heart rate is relatively low. Both parameters change as temperature warms, the heartbeats become smaller in amplitude and faster in frequency. In addition, the shape of the heartbeat changes, contraction and relaxation during each heartbeat proceed relatively slowly at cold temperatures but become faster at warmer temperatures. Overall, therefore, heartbeats are long in duration and large in amplitude at the cold extreme of the temperature range where heart rate is slow, and they become faster and weaker as temperature warms and heart rate increases. Finally, in this preparation the baseline tonus of the heart also changed, increasing with temperature up to 16°C and decreasing at higher temperatures. Similar changes in tonus were observed in three other experiments, but were variable and not analyzed further.

To describe these phenomena quantitatively, data from nine isolated hearts were pooled. Contraction amplitude decreases significantly as the temperature warms (Fig. 4). At temperatures of 16°C and higher the mean amplitude of the heartbeat is less than 35% of its initial value at 2°C. The temperature dependence of the shape of the heartbeat is described in Fig. 5. Contractions develop and relax relatively slowly at the coldest temperatures but become faster as temperature warms(Fig. 5A). This phenomenon is illustrated quantitatively by the phase plots(Fig. 5B,E) describing the relation of the rate of change in force to the contractile force recorded during the heartbeat. These plots show pronounced differences in shape as temperature changes because at warm temperatures the amplitude of the heartbeat is comparatively small while the rate of change in force is comparatively fast for the contraction and the relaxation phase of the heartbeat. From 2 to 10°C the rates of contraction and relaxation of each heartbeat more than double, before slowing as temperature becomes warmer(Fig. 5C). To examine the temperature dependence of contraction and relaxation without the confounding effect of temperature on heartbeat strength, data from Fig. 5C were replotted after normalizing for contraction amplitude (Fig. 5F). Normalization changes the shape of these plots without changing the conclusion of the analysis. Calculation of acute Q₁₀values for changes in rates of contraction and relaxation demonstrate these parameters are strongly temperature dependent in the 2–4°C range(acute Q₁₀ values >20 and >7 for raw and normalized data,respectively; Fig. 5D,G). At warmer temperatures rates of contraction and relaxation vary from weakly temperature dependent to temperature independent.

As shown in the raw data of Fig. 3, the frequency of the heartbeat tends to increase as a function of temperature. Responses of ten individual isolated hearts to temperature change are shown in Fig. 6A. Heart rates increased in all hearts between 2 and 10°C, but there was considerable variability at warmer temperatures with some of the isolated hearts slowing or failing. Fig. 6B compares the mean heart rates recorded as a function of temperature from isolated hearts in vitro and from hearts in intact animals in vivo. In both isolated hearts and intact animals,increases in heart rate occurred gradually as temperature warmed with no observed periods of bradycardia or tachycardia. Both sets of data show significant increases in heart rate as temperature warms (see Fig. 6 legend). Over the temperature range from 2 to 22°C, the heart rates of isolated hearts increased approximately fourfold in vitro (closed squares) whereas heart rates measured in vivo increased approximately 2.5-fold (closed circles). Isolated hearts tend to beat more slowly than hearts in intact animals, these difference are statistically significant at 2°C and in the temperature range from 12 to 22°C. In isolated hearts, the maximum heart rate observed was 1.53 Hz (99 beats min^–1) at a temperature of 20°C. In intact animals the maximum heart rate observed was 1.85 Hz(111 beats min^–1) at 22°C, slightly faster than the previously reported upper limit of 100 beats min^–1 for lobsters walking on treadmills (O'Grady et al., 2001). Finally, to test whether heart rates in intact animals might thermoadapt over long periods at thermal equilibrium, lobsters were warmed from 2°C to a new steady state level of 12°C. Fig. 6C shows that heart rate increases as temperature warms from 2 to 12°C and remains constant for 45 min under steady state temperature conditions. Thus, we find no evidence that thermoadaptation alters heart rate. This observation is consistent with the fact that our data, recorded from lobsters under laboratory conditions, is in good agreement with heart rates measured in lobsters in the wild(Fig. 6B, open symbols).

Values for acute Q₁₀ calculated for heart rates in vivoand in vitro are shown in Fig. 7. For isolated hearts, the acute Q₁₀ value associated with the increase in heart rate between 2 to 4°C is exceptionally high(acute Q₁₀=60), owing to the steep increase in beat frequency as the heart speeds up from an exceptionally low heart rate at 2°C. Otherwise acute Q₁₀ values up to 20°C ranged from 3.5 to 1.0 and were similar in both datasets, suggesting the temperature dependence of the beating frequency in isolated hearts is similar to that in intact animals over most of the temperature range.

Because the spontaneously beating heart is driven by rhythmic output of the cardiac ganglion, the relationship between the frequency of the heartbeat and the strength of the heartbeat is of particular interest. The neurocardiac synapses on the heart muscle gate calcium influx into heart muscle through voltage-dependent channels, and these synapses facilitate strongly as a function of frequency of motoneuron firing(Anderson and Cooke, 1969; Mahadevan et al., 2004). Previously, Mahadevan et al. have observed a positive correlation between frequency and strength of the lobster heart beat that they attributed to enhanced synaptic depolarization at high burst frequencies(Mahadevan et al., 2004). However, in our experiments we observed that these parameters behave differently in response to temperature change: the strength of contractions decreases (Fig. 4) whereas the frequency of contractions increases (Fig. 6; see also raw data in Fig. 3).

To examine the potential coupling between strength and frequency of the heartbeat we plotted the relation between these parameters for each of seven isolated hearts that were exposed to temperatures warming from 2 to 22°C(Fig. 8). None of these plots shows a positive relationship between the amplitude and the frequency of contraction. In general, the contraction amplitude of the heartbeat appears to decrease as frequency increases. Almost all of the plots showed regions where similar amplitudes were recorded at different frequencies, or temperature ranges where contraction amplitude changed while heartbeat frequency remained constant. Therefore, measured as a function of changing temperature, the parameters of heart rate and strength of the heartbeat do not positively correlate.

The cardiac output depends both on the stroke volume and the heart rate. To test the possibility that the temperature-dependent decrease in heartbeat strength (Fig. 4) might be offset by the concurrent temperature-dependent increase in heart rate(Fig. 6) we measured cardiac output directly. Fig. 9A shows examples of cardiac flow recorded from the same isolated heart at two different temperatures. Measurements of this type in four isolated hearts were analyzed over the temperature range from 2 to 22°C. In agreement with earlier results (Fig. 6B),heart rate increases with temperature from 2°C up to 12°C to 14°C(Fig. 9B). In contrast, stroke volume is maximal at the coldest temperatures and decreases as temperature warms (Fig. 9C). Interestingly,cardiac output shows a temperature-dependent trend that is well described by a quadratic equation with a maximum at approximately 10°C(Fig. 9D). Between 2 and 20°C there are no statistically significant differences in cardiac output,thus cardiac output appears relatively constant over this temperature range. At 22°C cardiac output decreases significantly compared to the data at 2°C.

In agreement with earlier reports, serotonin increases both the strength and the frequency of the heartbeat (Fig. 10). At 2°C the effects of serotonin on seven isolated hearts reached steady state within 12.2±8.5 min (mean ± s.d.; range 2.3–21 min). In these experiments serotonin increased contraction amplitude of the heart by an average of 128% (range 13–497%) and increased the frequency of the heartbeat by an average of 64% (range 18–171%). Fig. 11 shows the effects of serotonin on contraction amplitude and frequency over the entire temperature range, expressed as a ratio between the value measured in the presence of serotonin as compared to the control value at that temperature. In some experiments performed in summer months serotonin was particularly potent in increasing contraction amplitude at the warm temperatures of 16–18°C (raw data not shown). However, results averaged from all seven experiments showed no statistically significant temperature dependence of the effect of serotonin on contraction amplitude(Fig. 11A). Serotonin's effect in increasing the heart rate was also temperature independent, serotonin increased the frequency of the heartbeat approximately two fold at all temperatures (Fig. 11B). The shapes of the phase plots for heartbeats in the presence and absence of serotonin are similar (Fig. 11C), indicating that the effects of serotonin on contraction and relaxation rates are proportional to its effect in increasing contraction amplitude.

Temperature is a critically important parameter both for lobsters in the wild and for physiologically excitable cell membranes. Here we focus on the temperature dependence of a crustacean heart, a complex physiological system paced by the rhythmic output of the cardiac ganglion motoneurons and modulated by neural inputs and circulating hormones. However, unlike neurohormones and transmitters, which are targeted to cells expressing the appropriate receptors, temperature will affect all cells by altering the activity of ion channels and membrane pumps both in the plasma membranes and in the intracellular membranes. Temperature, therefore, is a ubiquitous modulator,potentially affecting cardiac physiology at multiple levels. In this study we describe the temperature dependencies of several parameters of cardiac performance in the lobster Homarus americanus. Interestingly,although contraction amplitude, heart rate, stroke volume and heartbeat kinetics have different patterns of temperature dependence, cardiac output remains relatively constant over a wide temperature range.

A major finding in this study is that heart rates increase with temperature both in vitro and in vivo. These results are in agreement with several previous reports on Homarus(Hokkanen and DeMont, 1997; McMahon, 1999; Schreiber et al., 2000; Schreiber et al., 1998) as well as other species of lobster (Nakamura et al., 1994; Zainal et al.,1992). We observe faster heart rates in intact lobsters compared to isolated hearts, especially in the range 12–20°C. Similar observations have been made previously, both in lobsters(Schreiber et al., 1998) and crabs (Wilkens, 1994), and attributed to the overall cardioexcitatory effects of the cardioregulatory nerves and endogenous hormones in intact animals. It is also possible that the different heart rates observed in isolated and intact hearts reflect differences in filling pressure, vascular resistance and stretch of the heart muscle.

Our demonstration that the acute Q₁₀ values for heart rate are similar in isolated heart and in intact animals (except at the extreme of 2°C) suggests that the temperature dependence of the lobster heart rate arises mainly from the intrinsic properties of the cardiac ganglion, rather than from activity in the cardioregulatory nerves or from hormonal signals arising from other regions of the CNS. Interestingly, Jury and Watson(Jury and Watson, 2000)reached a different conclusion. Based on their observations that severing the cardioregulatory nerves in intact animals abolished the response of the heart to changes in temperature, they suggested that the cardioregulatory nerves mediate temperature dependence of the heart rate(Jury and Watson, 2000). It is not entirely clear how to reconcile their results with ours. However, we note that their experiments were performed in a relatively narrow temperature range(starting at 15°C and warming or cooling by 1.5°C) and are in agreement with our data showing a small but significant difference in temperature dependence of intact and isolated hearts at 14°C (see Fig. 7). Nevertheless, over the relatively large temperature range of 4–22°C our observation that the temperature dependence of heart rate is similar in vitro and in vivo supports the hypothesis that temperature dependence of the heartbeat arises mainly from temperature effects on the cardiac ganglion neurons. A minor contribution by cardioregulatory nerves and/or hormones to the temperature dependence of the heart rate is suggested by the difference in the slopes of the plots describing the temperature dependence of the heart rate in vivo and in vitro, as shown in Fig. 6B.

McMahon has emphasized the importance of not relying solely on the heart rate as an indicator of cardiac performance(McMahon, 1999; McMahon, 2001) and our observations strongly support this view. Whereas the lobster heart rate increases linearly as a function of temperature over most of the range from 2 to 22°C two other parameters of cardiac performance, the strength of the heartbeat and the stroke volume, both decrease. Similar observations have been reported for crab (Cancer magister) hearts(DeWachter and McMahon, 1996; DeWachter and Wilkens, 1996). Moreover, cardiac output in the lobster displays a pattern of temperature dependence that probably results from the opposing temperature-dependent effects on heart rate and stroke volume. Cardiac output was maximal at 10°C, but became compromised as the temperature warmed above 20°C,with a significant decrease in cardiac output in hearts that continued to beat and an increase in the number of hearts failing. It is possible that cardiac performance in intact animals might be protected somewhat from thermal stress by endogenous neural and hormonal signals, since nearly all the hearts in intact animals (20 out of 21) continued to beat at 22°C. Although we did not make measurements at temperatures warmer than 22°C, it is likely that cardiac performance will degrade as animals approach their lethal temperatures of 25 to 30°C (McLeese,1956).

It is important to note that the rate of temperature change in our experiments (approximately 1°C min^–1) is considerably faster than temperature changes occurring in the wild during a tide or a season. (For example, the fastest rate of temperature change during the course of tide is approximately 3°C h^–1 in the record shown in Fig. 1.) In the ocean, it is possible that the animals might be able to acclimate to warmer temperatures if temperature changes sufficiently slowly, thereby improving cardiac performance at warm temperatures. However, it is also likely that animals moving through thermoclines on the ocean floor will experience relatively rapid and extreme changes in temperature. It is interesting to note that our measurements of heart rate in lobsters exposed to acute changes in temperature in the lab are in good agreement with those obtained from ocean animals experiencing seasonal temperature changes over the course of a year (see Fig. 6B). These results suggest that our relatively fast thermal challenges in the laboratory stimulate cardiac responses very similar to those in lobsters experiencing seasonal thermal shifts in their natural habitat. Future experiments will be directed toward examining whether the temperature dependence of cardiac performance differs for animals acclimated to different seasonal temperatures or animals exposed to temperature change at different rates.

Finally, the temperature dependence of cardiac performance in the lobster differs in several respects from that reported for crab C. magister(DeWachter and Wilkens, 1996),a Pacific coast species that inhabits a similar thermal range. In isolated crab hearts the temperature dependence of the heart rate is only one third that measured in intact animals and cardiac output is maximal at 2°C and decreases as a function of temperature until heart failure occurs at 20°C. In contrast, isolated lobster hearts appear to be more robust in the face of warming temperature. Most are viable up to 22°C, show temperature-dependent increases in heart rate that are similar to those observed in hearts in intact animals, and exhibit an increase in cardiac output as temperature warms above 2°C that reaches a maximum at 10°C.

A previous study of isolated lobster hearts concluded that the strength of the heartbeat and the frequency of the heartbeat are coupled(Mahadevan et al., 2004),because these parameters show a positive correlation in spontaneously beating hearts and in hearts paced by external stimuli. This interpretation seems reasonable, given that the same study showed that the magnitude of synaptic depolarization at neurocardiac synapses is linearly related to burst frequency. However, the experiments were performed at constant temperature(fixed between 10 and 14°C), thereby constraining the range of possible physiological responses. Our results, collected over a broad temperature range, show no evidence for coupling between the strength and frequency of the heartbeat. We favor the viewpoint that strength and frequency of the heartbeat are independent parameters and independently regulated. At constant temperatures heartbeat frequency will be a critical determinant of contraction amplitude, but changes in temperature will affect both parameters independently, and alter any relationship between them. The `uncoupling' of heartbeat frequency and amplitude in the lobster has been observed previously in response to hypoxia (McMahon,1999) and during adaptation to stimulation by nitrous oxide(Mahadevan et al., 2004).

Although we have not investigated the physiological mechanisms underlying temperature dependence in this system, our results suggest several possible mechanisms. As heart rate is determined by the rhythmic neural output of the cardiac ganglion, it is possible that temperature regulates the frequency and patterns of bursting of the Homarus cardiac ganglion motoneurons and pacemaker cells. Temperature-dependent decreases in the strength of the heartbeat might arise either from the properties of the Homaruscardiac muscle fibers, or of the cardiac ganglion motoneuron synapses that innervate them. For example, temperature may affect both the size of the neurocardiac synaptic potentials and the extent to which they facilitate, as shown in the myocardium of another lobster species Panulirus japonicus (Kuramoto,1994). In addition, it is also probable that the calcium dynamics within the cardiac muscle fibers might be temperature dependent, thereby regulating the availability of calcium in the cytoplasm and the rate at which it activates the contractile machinery. Further experiments will be required to investigate these possibilities.

In agreement with earlier reports, we find that serotonin increases both the rate and the strength of the heartbeat. The onset of serotonin effects on the heart at 2°C is comparatively slow, reaching steady state in approximately 12 min, on average. By comparison, other studies performed at 12–13°C have demonstrated that potentiation of lobster heart rate by serotonin reaches steady state within 5 min(Guirguis and Wilkens, 1995; Kuramoto et al., 1995). In our experiments modulation of the heart by serotonin was effective over the whole temperature range tested. Although there was some variability between preparations, we did not observe a statistically significant temperature dependence of the modulatory effects overall. In contrast, serotonin elicits several temperature-dependent effects on a lobster skeletal muscle neuromuscular system. Serotonin increases muscle tonus most strongly at the coldest temperatures (2°C), increases nerve-evoked contractions most strongly at warm temperatures (16–18°C) and enlarges the temperature range over which the inhibitory motoneuron is effective in relaxing the muscle(M.K.W., unpublished observations).

There are likely to be multiple sites of action for serotonin in the lobster cardiac system. Serotonin increases the heartbeat of intact animals even after the cardioregulatory nerves are severed(Guirguis and Wilkens, 1995),suggesting that serotonin acts directly on the cardiac ganglion. This was subsequently verified by Berlind's observation that serotonin increases and prolongs the bursting output of the cardiac ganglion in vitro(Berlind, 1998). Such effects would be consistent with serotonin increasing heartbeat frequency and contraction amplitude. In Homarus serotonin also decreases output through the sternal arterial valve, thereby redirecting arterial flow(Wilkens et al., 1996). However, it is also probable that serotonin acts on the neurocardiac synapses as well as the muscle fibers to increase contractility. Cooke reported preliminary evidence that serotonin potentiates synaptic transmission at synapses on the lobster heart (Cooke,1966). Serotonin potentiates both synaptic transmission and muscle contractility in lobster dactyl opener neuromuscular systems(Kravitz et al., 1980) as well as other invertebrate neuromuscular systems(Worden, 1998).

Finally, it is important to note that the temperature can be a confounding variable in physiology experiments, because multiple physiological processes underlie cellular function and each can depend differently on temperature. For example, a recent study of excitation–contraction coupling in lobster cardiac muscle at constant temperature concluded with the prediction that an increase in heart rate will enhance calcium loading of the sarcoplasmic reticulum, thereby accelerating relaxation of the heartbeat, increasing diastolic filling and contraction amplitude(Shinozaki et al., 2002). Further, these authors predicted that heartbeat frequency-dependent increases in contraction amplitude will cause stroke volume and cardiac output to increase. However, using an experimental protocol that changes temperature we observe different relationships between the parameters underlying cardiac performance. As the lobster heart rate increases both contraction amplitude and stroke volume decrease, while cardiac output itself remains relatively constant with a maximum at 10°C and a significant decrease at temperatures warmer than 20°C. Changes in temperature can therefore alter the relationships between parameters of cardiac function to produce physiological effects that cannot be predicted from observations at a constant fixed temperature.

The authors are grateful to Claire Edwards and Joseph Camacho for help with data collection, Deforest Mellon and Lynne Fieber for comments on the manuscript, and Brian Duling for helpful discussions. This work was supported by funds from the University of Virginia School of Medicine.