The Effects of Temperature on the Jet Escape Response of the Sc quid Loligo opalescens Beth Riple William F. Gilly Laboratory Hopkins Marine Station Abstract Loligo opalescens migrates vertically, experiencing large temperature changes in short amounts of time. Loligo depends on jet-propulsion for escape behavior, and at 12°C shows an escape response that uses coordination between two motor systems employing giant and non-giant neural fibers. Little is known about escape behavior at colder temperatures, however, and experiments were done to determine how this behavior changed over the biologically relevant temperature range of 12'-4’C. Squid were restrained and recordings of stellar nerve activity and intra-mantle pressure transients were taken during escape behavior elicited by strobe flash stimuli. As temperature decreased, pressure showed an increase in amplitude and duration, as well as an introduction of a plateau followed by a slower increase in amplitude below a critical transition temperature at 8'-10'C. Neural recordings suggest that these findings may be due to both the introduction of two giant axon spikes at the transition temperature, as well as an increase in small fiber activity at lower temperatures. In free-swimming squid, these changes are manifested as a longer period of propulsive thrust which moves the squid a larger distance in a given time. These findings suggest that Loligo is able to maintain performance at colder temperatures through coordination of the giant and non-giant motor systems in a way that is similar to, yet distinct from, that at warmer temperatures. Introduction Escape behavior in Loligo utilizes a form of jet propulsion in which water is taken into the mantle cavity and is subsequently expelled through the funnel via synchronous contraction of circular muscles (Packard, 1968). This expulsion of water provides an explosive jet that propels the animal large distances in short times, providing an effective means of escape from predators. Escape behavior has been studied in vivo by restraining animals in order to permit recordings of stellar nerve motor activity and intra-mantle pressure transients (Ötis and Gilly, 1990). This work revealed the roles played by both the giant and non-giant axon systems in different types of escape responses, and showed that both types of motor axons have distinct functions that can either be employed separately or in concert during escape behavior The giant axon system of squid has been a well-studied subject for over half a century, and a wealth of knowledge exists concerning its functions. The all-or-nothing nature of giant axon excitability was initially studied using stellar nerve-mantle preparations, and the axon’s role in stereotyped escape jetting was inferred from these studies (Prosser, and Young, 1937; Young, 1938). More recent research done in vive targets the giant axon system as the controlling factor in escape responses elicited by a strobe flash (Ötis and Gilly, 1990). These flash-elicited responses involve the single firing of a giant axon which translates into a stereotyped intra-mantle pressure in terms of peak amplitude and duration. Non-giant axons run parallel to giant axons and, in contrast to the giant axons, must fire repetitively to generate graded mantle contractions (Wilson, 1960). Until recently, it was believed that these non-giant axons only mediated weak respiratory contractions and slow, constant swimming (Young, 1938). Current literature suggests that small non-giant axons also play an important role in more complex, delayed escape responses that can be elicited by an electrical shock applied to the arms of the squid (Otis and Gilly, 1990). Such complex, delayed escape responses are characterized by a longer latency of 2200ms compared to the more stereotyped flash responses (approx. 50ms) and show highly variable intra-mantle pressure transients which translate into jets of different strengths (Otis and Gilly, 1990). Thus, there appears to be an initial period of "decision-making" where an animal may or may not elicit a jet, presumably determined by an assessment of the stimulus. Additionally, the strength and number of jets executed is very plastic, and has been attributed to a complex central coordination of giant and non-giant motor systems. Small axons always fire in this type of response, and giant axons may or may not follow a burst of small axon activity (Otis & Gilly, 1990). Giant axons clearly provide a boost to the intra¬ mantle pressure, providing a more powerful jet, but activation of the giant axons is not necessary for strong escape behavior. These escape responses afford Loligo high performance levels at ambient (surface) sea water temperatures, but little is known about how these responses are affected by temperature changes. Many components of nervous function are highly temperature dependent, such as conduction characteristics and synaptic transmission but it is unclear how any individual component would affect whole animal performance. A better understanding of this overall coordination is particularly relevant, given Loligo's frequent exposure to large temperature changes. Loligo is pelagic and vertically migrates through temperatures ranging from 13'-5°C. Loligo encounters predators throughout this temperature range, and it is presumably important that it be able to maintain escape performance at all temperatures. This study addresses the question as to how Loligo might compensate for these temperature changes by studying the temperature dependence of intra-mantle pressure transients and the patterns of stellar nerve motor activity during escape jetting. Materials and Methods Animals Loligo opalescens was collected from Monterey Bay and maintained in large holding tanks filled with free-flowing seawater (12-13'C). Animals were separated by sex in order to prolong health and were kept in the laboratory for up to two weeks after collection dates. Behavioral experiments Animals were restrained by gluing the dorsal mantle surface to a plastic holding platform with cyanoacrylate cement in order to obtain neural and pressure recordings. This holding platform was suspended in an aquarium with free-flowing, aerated seawater. Squid could be removed from this platform at the completion of an experiment, and were often returned to the holding tanks where they displayed normal behavior and sometimes survived for lengths of time comparable to squid not experimented on. Using a cooling unit and the arrangement schematized in Fig. 1, water temperatures in the aquarium were cooled from 13'-5°C at a rate of 1C per 5 minute interval, and could be warmed back to 13'C, then cooled again in the duration of one experiment. Free-flowing, ambient seawater (12°C) could be turned on or off by a valve. The aquarium was filled with this 12°C water, then the source was shut off and an alternate water source from the cooling unit pumped 3'C water into the aquarium. This water mixed with the 12°C water, and as water continued to circulate through the aquarium and back into the cooling unit, overall aquarium temperature was decreased. Aquarium temperature could be raised by shutting off the 3'C water line and turning back on the 12°C line, allowing the circulation of this water through the flow-through system to gradually warm the aquarium. Water was also warmed from 13°-17°C by adding warmed water into the tank, then cooled again using the 12’C water line. Temperature was constantly monitored with a thermistor probe in the aquarium and was recorded on a chart recorder. Escape jets were elicited using either a strobe flash or an electrical shock delivered to the base of the arms. Strobe flashes visually stimulated the animal directly, and produced a stereotyped response. Electric shocks produced more delayed and variable responses. Trials were videotaped using a Sony CCD-IRIS video camera. Additionally, movies were captured by a frame-grabber on a laboratory computer in 33ms frames which could later be resolved into 16ms frames and analyzed using NIH- image software. Intra-mantle pressure was recorded with an analog transducer that was attached to a hyperdermic needle inserted into the mantle cavity. Pressure transients were recorded onto both a chart recorder and a DC-coupled digital audio recorder (Sony DTC-670). Extra-cellular recordings of neural activity were made by cutting a hole over the stellate ganglion of a restrained animal and then exposing stellar nerves for en passant recordings. Recordings were made by a polyethylene suction electrode, usually placed on the second or third-hindmost stellar nerve just past the stellate ganglion, and were recorded onto Hi-8 tapes using a Sony (EVÖ 9700) 8mm video deck and also onto the digital audio tapes to be later resampled using a laboratory computer. Some neural recordings were made after cutting two 4cm slits in the ventral surface of the mantle, and this procedure minimized artifacts due to water being expelled through the hole above the stellate ganglion during strong jets which splashed the recording electrode. Free-swimming experiments Three squid were placed in a fifty gallon tank filled with seawater and escape jets were elicited using a strobe flash. Temperature was increased from 4'-12'C, and responses were recorded onto Hi-8 tapes using the Sony video camera. Trials were later resampled onto a laboratory computer and were analyzed using NIH image. Results Changes in flash-elicited escape jets with temperature Escape responses were elicited using a strobe flash, and Fig. 2Ai shows an example of the resulting pressure transient (upper trace) and stellar nerve discharge (lower trace) at 12.5°C. The complete pressure transient is displayed at a slower time base in Fig. 2Aii. Figures B-E display pressure transients and neural activity as temperature is sequentially lowered to 5.0'C. Pressure transients at 12°C confirm the stereotyped neural activity and pressure transients revealed in earlier experiments (Ötis and Gilly, 1990). However, the pressure transient shows a substantial increase in both duration and amplitude as temperature is lowered. Duration increases from around 300ms at 12°C to 700ms at 5°C, and peak amplitude increases substantially. Changes in the overall shape of the pressure transient are also evident, although the initial rate of rise and the maximum rate of relaxation do not appear to be markedly affected. At a critical transition range of 8°-10°C, a plateau appears (Fig 2Dii), which extends the time of peak amplitude before relaxation. At temperatures lower than this range, an additional slower rise in pressure amplitude follows the plateau period (Fig 2Eii). Neural activity accompanying escape jets at different temperatures Neural activity shows two important changes over the temperature range. First a second giant axon spike in Fig. 2Di appears quite reliably in the 8°-10’C range, and sometimes one or two spikes are present at essentially the same temperature (Fig. 20i and Fig. 2Di). However, this second giant axon spike can also appear at temperatures lower than this range and rarely (once) does not occur at all upon cooling. Additionally, an increase occurs at lower temperatures in both the intensity and duration of the burst of non-giant axon firing after the giant spike. This increase in intensity of non-giant firing appears to be a coordinated and synchronous discharge manifested as a slow wave immediately following each giant spike in Fig 2. An attempt to quantify the increase in non-giant activity is presented in Fig 3. The rms value of each 7.5ms segment of neural activity was determined using data sampled on a laboratory computer, and served as a measure of total deviation of neural activity from the baseline defined by background activity. More activity exists at 6'C than at 12°0. This increase in non-giant activity thus appears to be important. In contrast to repetitive giant axon activity, which appears at a certain temperature and remains unchanged thereafter, the increase in intensity and duration of small axon activity appears to be graded over the entire temperature range. Escape jets in non-restrained squid at different temperatures Free-swimming animals were also studied in order to establish how temperature changes affected whole-animal escape performance. Fig. 4A shows the distance traveled at both 5.5°C and 12°C (average of 3 animals) as a function of time elapsed after the stimulus. Results show that animals at lower temperatures consistently traveled further distances than animals at warmer temperatures, with intermediate distances traveled by animals in the middle of this temperature range. At 5.5°C an escape jet propelled the squid twice as far as at 12°0. Fig. 4B shows the average, instantaneous velocity traveled by the same three animals, as derived from data in Fig. 4A. Animals at colder temperatures reach a slightly higher maximum velocity and clearly maintain elevated velocity for a longer period of time. However, animals at colder temperatures show a longer delay from time of stimulus to peak velocity than animals at warmer temperatures, a trend that can also be seen in Fig. 4A. Escape jets above 12°0 Preliminary work shows that upon warming from 12'-17°C, several changes occur in flash-elicited escape responses. At a critical temperature range between 14'- 16°C, intra-mantle pressures decrease suddenly to an amplitude of only about 10% of that at 12°C (Table 1). Neural recordings indicate that the disappearance of the giant axon spike corresponds to this dramatic decrease in pressure. Only small fiber activity was seen at temperatures above this range, and no summation or longer duration of firing relative to activity at 12'C was witnessed. Shock-elicited escape responses Preliminary work with delayed-type escape responses elicited by electric shocks indicates that the peak amplitude of delayed jets decreases as temperature decreases, reflecting a general trend of compromised performance at colder temperatures (Table 1). Attempts to detect coordinated firing of the giant and non¬ giant axons, regularly seen at 12'C or above (Otis & Gilly, 1990), were unsuccessful at temperatures below 10°C. However, it is difficult to determine whether this reflects the actual absence of giant fiber activity or was instead due to a shortcoming in the recordings. Discussion Flash-elicited responses As temperature was decreased during flash-elicited escape responses, pressure transients showed an increase in amplitude and duration, as well as an introduction of a plateau followed by a subsequent slower increase in amplitude after a transition range of 8°-10°C. Neural recordings showed both the appearance of a second giant axon spike in a temperature range of 8'-10'C and an increase in small axon activity marked by both a greater intensity and duration of firing and an increasing summation of activity. Previous work done by Young on the repetitive firing of giant axons suggests that this repetitive firing causes an increase in the duration of a pressure transient by extending the duration of the peak amplitude before the pressure begins to subside (Young, 1938). This can explain both the increase in duration seen in the pressure transients in Fig. 2 A-E as well as the introduction of a plateau period. However, Young's work has also shown that additional firing of giants cannot increase the peak amplitude of the pressure transient and therefore cannot explain the additional slow rise in amplitude seen in Fig. 2Eii. This suggests that the slow rise is not due to the repetitive firing of the giant axon. Repetitive stimulation of small axons can cause graded increases in pressure amplitude and duration (Wilson, 1959). Preliminary results showing increased small axon activity both in terms of intensity and duration suggest that the additional increase in amplitude is attributable to this. Further, it is hypothesized that the critically timed summation of small axon activity immediately following the giant spike might also lead to a more powerful circular muscle contraction. This suggests that as temperature is decreased, the parallel giant and small fiber systems become increasingly active both separately and in coordination. Contrary to warmer temperatures at which the contribution by the giant axon system is predominant, colder temperatures are conducive to summation by the two motor systems. Although it remains to be determined exactly how these two systems work together, it is clear that a giant axon always fires first and is then followed by a burst of small axon activity. This burst becomes more prominent at lower temperatures probably because of both increased synchronization, resulting in a summed, slow wave (Fig. 2), and an increase in the total period of discharge (Fig. 3). Both mechanisms could lead to a greater contribution of the small axon system to the jet. It also seems likely that the double firing of the giant axon allows an additional synchronous discharge of small axons which provides a second period of summation during the response and effectively supplies a second "dose" of summed small fiber activity. The additional firing of the giant axon itself appears to have a minimal effect at a given temperature and serves to hold the pressure at a steady level which may allow the small axon system to fire for a longer time and thereby provide the subsequent boost added to this maintained level. Reasons for the occasional single firing of a giant axon at colder temperatures where double firing is usually seen are not known. Preliminary research looking at warming above 12’C may serve to confirm the temperature dependence of coordination between the two systems. Just as cooling serves to increase the amount of summation by the two systems, warming seems to increasingly separate the systems to a point where the giant axons no longer respond and the small fiber system fires for only a very short time, producing only a weak response. This means that pressure transients with the largest peak amplitude are found at cold temperatures, with increasingly smaller peak amplitudes as temperatures are warmed. Shock-elicited responses As temperature is decreased, the peak amplitude of the pressure transient during a delayed-type escape response is decreased, and the giant axon system does not appear to be recruited. Therefore, shock-elicited escape responses seem to become greatly compromised as temperatures are lowered. It is not yet known why this may be, but it seems that the two systems lose their ability to coordinate together in complex ways, and therefore are not able to elicit as strong a pressure transient as is possible at warmer temperatures where the two systems work in concert. As temperature is increased above 12’C, delayed-type escape responses do not seem to be significantly compromised, and show peak pressure amplitudes comparable to those seen at 12°C. Though it has not been confirmed in the present experiments whether giant axons actually fire at elevated temperatures, the maintained level of performance seen relative to 12'C suggests that the giant and non¬ giant fiber systems are still working in coordination. All of ÖOtis and Gilly's (1990) work was actually carried out above 12°0. Coordination of giant and non-giant fiber systems As temperature is decreased, the stereotyped escape response elicited by a flash shows increasing coordination between parallel giant and non-giant fiber systems, providing not only maintenance of performance, but an actual increase in performance when compared to warmer temperatures. Conversely, when temperature is decreased the shock-elicited response shows decreasing ability to coordinate giant and small fiber systems, showing decreased performance at colder temperatures (Table 1). This study shows that these two types of escape responses coordinate the two motor systems in different ways. In a flash-elicited response the firing of the giant axon is followed by small axon activity, whereas in a shock-elicited response the firing of small axon activity is followed by giant axon firing if the giants are recruited. These different styles of coordination found in the two types of escape responses must play a biologically relevant role for squid traveling through large temperature ranges in relatively short periods of time. Changes in the way the two systems work in concert may take advantage of or compensate for different properties of the squid’s physiology that may be affected by temperature changes. For instance, the shock-elicited response seems to be more adaptive at warm temperatures, where its more delayed and variable jets may allow an animal to choose not only whether or not it responds to a stimulus, but also to what degree it responds. Such a highly flexible escape response may also act to conserve energy, in contrast to the stereotyped flash response where there is a reaction of fixed amount every time. However, if colder temperatures somehow compromise the central nervous system’s ability to provide rapid and accurate assessment of a threatening situation, it might be more advantageous to employ a more simple but powerful startle response that guarantees a strong and reliable jet. Therefore, although the more flexible, delayed-type response is compromised by colder temperatures, the startle response becomes stronger and more reliable. In this way, differing methods of coordinating parallel giant and non-giant motor axons may allow Loligo incredible flexibility that allows it to travel freely through the changing temperature conditions it experiences daily. Acknowledgements: Special thanks to William Gilly, Heike Neumeister, Thomas Preuss and the entire Gilly idance, as well as their patien te and positive words. Thanks Lab for their help and also to the Somero Lab and the Thompson Lab for the use of their equipment. References 1. Prosser, C.L. & Young, J.Z. 1937 Responses of muscles of the squid to repetitive stimulation of the giant nerve fibres. Biol. Bull. 7 3, 237-241 2. Young, J.Z. 1938 The functioning of the giant nerve fibres of the squid. J. Exp. Biol. 1 5, 170-185 3. Wilson, D.M. 1960 Nervous control of movement in cephalopods. J. Exp. Biol. 37,57-72 4. Packard, A. 1968 Jet propulsion and the giant fibre response of Loligo . Nature 221, 875-877 5. Otis, T.S. & Gilly, W.F. 1990 Jet-propelled escape in the squid Loligo opalescens: Concerted control by giant and non-giant motor axon pathways. Proc. Natl. Acad. Sci. USA 8 7, 2911-2915 Figure Legend Figure 1. Water temperature system. 12°C water is used to fill the experimental aquarium, then is shut off and 3°C water is pumped in. This 3°C water mixes with the 12•C water and gradually lowers the temperature while being recycled through the chiller in a closed water system. To warm the water, the 3°C water line is closed and 12°C water is pumped into the aquarium. Cold water is allowed to leave the system. Figure 2. Temperature-dependence of the stellar nerve discharge (lower traces) anc the corresponding intra-mantle pressure transient (upper traces) during flash-elicited startle responses in a restrained squid. Experiment was started at 12.5°C (A), and temperature was lowered to 5°C (E). Giant axon events are indicated by arrows. Complete pressure transients are displayed on the right. Figure 3. Temperature dependence of small fiber activity. Values were determined by calculating the rms values for 7.5 ms segments of neural activity. The rms value served as a measure of the amount of deviation of neural activity from a baseline defined by background activity (level between stimulus artifact and the large peak due to giant axon activity). Figure 4. Temperature dependence of escape-jetting performance in free-swimming squid. Values are the mean of three squid studied simultaneously in the same tank. (A.) Cumulative distance traveled following a flash stimulus. (B) Velocity computed from the data in (A). Squid were placed in a fifty gallon tank filled with 5.5°C water. and water was warmed up to 12°C by partially draining the tank and adding ambient¬ temperature seawater in increments. Fig. 1 outflow Aquarium inflow 12°C sea water valve valve 3'C sea water Water chiller circulates backto coler H vabe leaves circulating systeml valve Fig. 2 12.5° Ai — 100 ms Bi 10.5° 8.60 Ci 8.30C Di 5.0° — — t 10 kPa 200 uV 100m Fig. 3 20 Giant spike Giant spike Flash Boo Sggd — 100 Time (ms) -6•C —0—12° Boto 200 Table 1. Temperature-dependence of the peak amplitudes of pressure transients Peak amplitude: Peak amplitude: flash response shock response Above ambient (16°0) 1.8 kPa 13.4 kPa Ambient (12°C) 11.3 kPa 15,4 kPa Below ambient (5°0) 15.6 kPa 7.8 kPa Table 1. Temperature-dependence of peak amplitude of pressure transients from a representative experiment. As temperature is cooled, peak amplitudes elicited by a flash response increase, becoming larger than peak amplitudes elicited by shock responses. Peak amplitudes elicited by shock responses exceed flash-elicited, peak amplitudes both at and above ambient levels, but are much smaller below ambient temperature. Fig. 4 6 0 30 100 5 0 — —e—5.5°0 —0—12 C — 500 1000 Time after stimulus (ms) —e—5.5°C —0—12•C Mrdee 500 1000