Pressure and Duration of Constriction in Boa Constrictor is Influenced by a Simulated Prey Heartbeat

Transcription

1 Dickinson College Dickinson Scholar Student Honors Theses By Year Student Honors Theses Pressure and Duration of Constriction in Boa Constrictor is Influenced by a Simulated Prey Heartbeat Allison Elizabeth Hall Dickinson College Follow this and additional works at: Part of the Biology Commons Recommended Citation Hall, Allison Elizabeth, "Pressure and Duration of Constriction in Boa Constrictor is Influenced by a Simulated Prey Heartbeat" (2010). Dickinson College Honors Theses. Paper 86. This Honors Thesis is brought to you for free and open access by Dickinson Scholar. It has been accepted for inclusion by an authorized administrator. For more information, please contact scholar@dickinson.edu.

2 Pressure and Duration of Constriction in Boa constrictor is Influenced by a Simulated Prey Heartbeat By Allison E. Hall With the collaboration of Amanda Hayes and Katelyn McCann Submitted in partial fulfillment of Honors Requirements for the Department of Biology Dr. Scott Boback, Supervisor Dr. Charles Zwemer, Supervisor Dr. David Kushner, Reader May 18, 2010

3 Abstract Constricting prey is energetically costly for snakes and therefore it would be beneficial to minimize this cost. However, the consequences of arresting a constriction event too soon could be deadly. Thus, the duration of constriction is bounded by competing demands to kill prey and conserve energy. Snakes possess mechanoreceptors within their ventral and dorsal skin that are used for detecting approaching predators and prey. This experiment sought to determine whether Boas (Boa constrictor) can sense a simulated heartbeat in their prey. It was predicted that if snakes possess this ability, those constricting rats with a simulated heart would constrict with greater pressure and increased duration than snakes constricting rats without a simulated heartbeat. We recorded constriction pressure from snakes constricting rats with and without a simulated heartbeat. Using a two-way unbalanced analysis of variance (ANOV A) we found that boas constricting rats with a simulated heart did so for longer and with greater total pressure relative to those constricting rats without a simulated heart. These data suggest that snakes may be capable of sensing the simulated heartbeat and will adjust constriction pressure and duration accordingly. Introduction Boa constrictor is a species of snake within the family Boidae with a geographic range from Central Mexico to Argentina (Stafford and Meyer, 2000). Boas are non venomous and use constriction to subdue and kill their prey prior to consumption (Cundall et al., 2000). In general, constricting snakes use their axial musculature to constrict their prey; specifically, the three main epaxial muscles (spinalis-sernispinalis, longissimus dorsi and iliocostalis) produce the force needed to constrict a prey item (Lourdais et al., 2005; Figure 1). The epaxial muscles are thought to play a major role in flexing the vertebral column 2

4 laterally, which is an important part of both constriction and locomotion (Moon, 2000). Hence, the performance of these muscles may vary among and within species of snakes according to the type of locomotion used and the type of prey consumed. Snakes use a variety of sensory reception including infrared (IR), vibration, and olfactory mechanisms to locate prey (Cundall et al., 2000). Boas possess temperature sensitive neurons located beneath scales of their upper jaws (labial scales), giving them the ability to detect IR radiation (Ebert et al., 2007; Von During, 1974). This provides them with the ability to detect endothermic prey without the use of visual cues. The IR receptors provide input to the tectal region of the brain of the snake, where visual and heat information can be integrated (Buning, 1983). Like other snakes, boas also have the ability to detect vibrations (Proske, 1969). Proske ( 1969) demonstrated that Pseudechis porthyriacus has vibration sensitive nerves in the ventral and dorsal skin. These receptors also were found in branches of the trigeminal nerve feeding the mandible and maxilla. Other work has corroborated these findings, indicating that the vibration sensitivity is detected by two distinct sensory systems. The first is composed of receptors in the skin of the snake (as indicated by Proske) whose signals are processed by the midbrain. The second set of signals can be detected by the eighth cranial nerve and are thus termed the 'VIII nerve system' (Hartline, 1971 ). These systems were found within three snake families: Boidae, Colubridae and Crotalidae (Hartline, 1971 ). The third recognized method of prey detection is olfactory reception. These sensory receptors are located in the vomeronasal organ and detect specific chemical cues from the snake's prey (Cundall et al., 2000). Once a potential prey item has been detected, the snakes initially strike, throw a coil around the prey, and proceed to constrict. In most strikes boas contact their prey with the 3

5 mandible alone initially, which then is followed by the maxilla, although this pattern can vary slightly between individual strikes (Cundall and Deufel, 1999). Although the initial strike pattern may vary amongst snakes, prey restraint behaviors vary little within boas and pythons (Cundall and Deufel, 1999). Boas strike the anterior portion of their prey and use a horizontal coil; the prey item is thus maintained horizontal to the substrate. The coil is a ventral-lateral coil, such that the first loop contacts the prey ventrally while the second loop contacts the prey laterally (Mehta and Burghardt, 2008; Heinrich and Klaasen, 1985). Additionally, boas constrict prey by placing their anatomical right against the prey significantly more often than their left (Heinrich and Klaasen, 1985). Constriction of prey is energetically costly for the snakes. For example, Canjani et al. (2003) measured the aerobic metabolism of Boa constrictor amarali before, during, and after constriction and found constriction times of up to an average of sixteen minutes and average oxygen consumption of up to ml 02 g-1 h-1, a six-fold increase from rest. Comparatively, Python molurus averaged 0.20 ml 02 g-1 h-1 during digestion, nearly a sixfold increase above standard metabolic rate (SMR). However, because digestion spanned an eight-day period this cost is likely greater than constriction (Secor and Diamond, 1997). Moon (2000) suggested that snakes use only as much force, and thus energy, as is necessary to subdue a prey item. Regardless, because constriction is costly, it is plausible that a mechanism exists for a snake to determine precisely when a prey item has expired and constriction is no longer necessary. This study aims to determine if that signal is the prey's heartbeat. There are many different cues that could indicate to a snake when the prey has expired. Movement, struggling, and breathing in addition to a beating heart would indicate a 4

6 living prey item (Moon, 2000). Snakes possess the ability to detect vibration, and movement in general, so they potentially have the ability to sense a beating heart in a prey item when it is being constricted (Hartline, 1971; Young and Morain, 2002). The cessation of the heartbeat would indicate that the prey item is dead and constriction is no longer necessary. Further, the mechanism by which prey are killed via constriction may be circulatory arrest (Hardy, 1994). If this is true, detecting the cessation of cardiac activity could be important in determining when it is safe to release. A similar study, conducted by Moon (2000) examined the constriction pressure produced by gopher snakes (Pituophis melanof eucus) and king snakes (Lampropeltis getula) in response to prey movement, simulated heartbeat, and respiratory movement. In preliminary trials, it was found that the snakes responded to the prey movement by increasing their constriction pressure, but the simulated ventilation and heartbeat did not prolong constriction times or increase constriction effort and were thus eliminated from the rest of the study (Moon, 2000). We hypothesized that boas have the ability to detect the heartbeat of their prey during a constriction event and alter their constriction pressure based on the presence or absence of a heartbeat. If the snakes have the ability to detect prey heartbeat, and utilize this to assess when prey have expired, we predict that snakes constricting prey with a sustained heartbeat would maintain a higher pressure for a longer duration relative to snakes constricting prey without a heartbeat. Methods Subjects The fourteen snakes of the species Boa constrictor used in this study were collected between 2002 and 2003 from the mainland and islands off of the coast of Belize and several 5

7 were born in captivity (Boback, 2006). Six males and eight females were used, nine were wild caught snakes and five were captive born snakes. Selection of prey In free-ranging boas, a typical prey item is between 20 and 40% of the snake body mass (Moon, 2000; Loop and Bailey, 1972). Therefore we used rats that were 20% (± 1 %) of the snake body mass for all constriction tests. The selected rats (Rodentpro.com, Inglefield, IN) were thawed from a frozen state to the room temperature of a snake holding room (28 C) overnight, prior to dry heating with a heating blanket. Instrumentation In order to monitor the pressure produced by the constricting snake, two pressure systems were used, one in the abdomen and one in the thoracic region of the rat. The thoracic system included both the simulated heart and a pressure-sensitive probe. Both were outfitted with a pressure transducer (Gould P.T.J. 4771, Holliston, MA, Figure 3). A 4.0 mm endotracheal (ET) tube (Sheridan, Loveland, CO) was used for the simulated heart. The 4.0 ET tube was cut along its length to create a pocket in which a 3.0 mm ET tube could reside. This allowed the thoracic system (simulated heart and pressure-sensitive probe) to be as compact as possible so that insertion into the rat was minimally destructive. The bulb of the 3.0 ET tube acted as the pressure probe. The bulb of the 3.0 ET tube was aligned with the bulb of the simulated heart to ensure proper placement of both within the thoracic cavity of the rat (Figure 4). The abdominal system consisted of one 4.0 mm ET tube. Both the thoracic and abdominal ET tubes were connected in series with intravenous tubing, three-way stopcocks 6

8 and individual pressure transducers (Figure 3). Air was removed from the systems by vacuum suction and the systems were back-filled with tap water. The prey were instrumented after thawing. The incisor teeth of the rats were removed prior to instrumentation to avoid puncture of the ET tube bulbs. The thoracic system was inserted into the esophagus of the rat using a trocar and positioned at the level of the heart. Proper placement was assessed via palpation. The abdominal system was inserted rectally and positioned caudal to the diaphragm. Both systems were secured to the muzzle and tail respectively using suture and several single square knots. Prey preparation To ensure snakes were presented with rats of appropriate temperature we determined how quickly the rats would cool to a non-physiologic temperature (37 C) after they were removed from heat, the rates of body cooling were determined for three rat sizes (Figure 2). The internal body temperature was monitored during these cooling tests and for all constriction tests using a k-type thermocouple within the liver of the rat. The temperature was recorded using a Physitemp BAT-10 (Physitemp, Clifton, NJ). The rats were warmed to 38 C (normal rodent body temperature Dilsaver et al., 1992) using a Harvard Homeothennic Blanket Control Unit (Harvard, Edenbridge, KY). Once the rats reached 38 C, the heat was removed and internal temperature was monitored and recorded every minute (Figure 2). It was determined that a maximum of five minutes between removal of the rat from heat and presentation to the snake should be achieved. This would allow the rat to cool only about 1 C. In all constriction tests the rats were maintained in the heat blanket at 38 C until immediately prior to testing. Calibration of the AID Recording System 7

9 The pressure transducers were secured at the level of the cage of the snake being tested and calibrated. A two-point calibration was performed using a mercury filled U-tube manometer at 0 and 100 mm Hg. The pressure transducers were connected in series to an analog to digital (AID) recording system (MPlOO Biopac Systems, Santa Barbara, CA). The AJD system recorded pressures using AcqKnowledge 3.9 software on Mac OS X. Constriction Test When the rats reached 38 C, they were removed from the heat and the implanted probes were connected to the pressure transducer and AID system. All air was removed from the system prior to testing. A Harvard Instruments Rodent Ventilator, Model 683 (Holliston, MA), was used to simulate a beating heart in a dead rat. We modified the ventilator by filling the system with water and hydraulically connecting the piston to the bulb of the ET tube implanted into the thoracic cavity of the rat, creating a complete, closed water column (Figure 3). The stroke volume of the ventilator was set to 2.5 ml for every trial. The final stroke volume achieved within the actual heart bulb was an average of 0.85 ml, comparable to the average stroke volume of a rat at 0.6 ml (Gunther, 1975). The average heart rate of a rat is beats/min (Harkness and Wagner, 1995). The ventilator was set to its maximum of 195 cycles per minute in tests with the simulated heart and turned off in tests without the simulated heart. This rate is on the low end of normal rat heart rates, but well within the typical mammalian prey of this mass consumed by Boa constrictor (Baudinette, 1978). To monitor the function of the ventilator the pressure produced was measured via a third pressure transducer. It was also ensured that the medical tubing was not in contact with the body of the snake during constriction. 8

10 The AID system was set to run for 60 minutes and the thoracic and abdominal pressure transducers were set to zero by removing excess water and thus pressure from the system. The rats were held by their tails using large forceps and presented headfirst to the snakes. The snakes were allowed to constrict on the rats and the AID system was run until the end of constriction. The snake was allowed to constrict until the point when it removed its head from the rat; this indicated the point at which the snake would begin to locate the head of the prey to begin ingestion (Canjani et al., 2003). During the process of head removal, but while the coil was still intact, the beginning and end of the coil was marked on the snake using a Sharpie marker. The coil was defined as the portion of the snake touching the body of the prey. Following the constriction test, the snake was anesthetized using isoflurane and morphometric measurements were taken of mass, snout-vent length (SVL), tail length, coil length, coil height and width of epaxial muscles to the nearest millimeter. The coil height and epaxial muscle width were measured using calipers at the beginning, middle and end of the coil. The epaxial muscles were measured as described by Lourdais et al. (2005), using calipers and measuring muscles on only the anatomical right of the snake, because Boas tend to constrict using their right side (Heinrich and Klaasen, 1985), as well as both sets in a cross-sectional plane. Each snake was tested at least once constricting a rat with and a rat without the simulated heart. A test with the simulated heart was followed by a test without the simulated heart and then again with the simulated heart in an attempt to minimize any potential learning effects (Holtzman, 1998). After each constriction event the snakes were fed and a minimum 9

11 of 14 days elapsed before the next test, to ensure complete digestion (Secor and Diamond, 1997). Data Analysis Constriction tests were recorded using the AcqKnowledge software at 200 Hz, producing a data point every 5 milliseconds. Raw data was exported to Microsoft Excel and averaged every 200 rows to determine the pressure for every second of the constriction. The beginning of a constriction event was determined as 15 seconds prior to the maximum initial pressure recorded in the thoracic bulb. The cessation of constriction was defined as the point at which the pressure fell below 3% of the maximum pressure. The thoracic pressures were analyzed because in most cases the snakes utilized one coil in the thoracic region of the rat. The total pressure over the constriction event was quantified using an integral of the pressure versus time curve for each trial of each snake. The integral was calculated using the trapezoidal rule. An unbalanced two-way analysis of variance (ANOV A) was used to determine significance for both the integral and the duration data using the following formula: µijk=µ+a;+~j+ (a~) ij+ ijk. Where the individual observation, µijk, is the sum of the grand mean(µ), the effect of the simulated heart, a, at level i (i=o without simulated heart, i=l with the simulated heart), the effect of individual snake.b, at observationj (j=l, 2,... n), the interaction of simulated heart, a, and snake, ~' and the random error, e. Duration and integral were determined using S+ statistical software. The two-way unbalanced ANOV A was performed using the proc GLM procedure in SAS. Results Pressure versus time plots of an individual snake show differences in response to the presence or absence of a simulated heart. For example, one snake produced maximum 10

12 pressure of 225 mmhg when constricting a rat with the simulated heart versus a maximum pressure of 175 mmhg when constricting a rat without a simulated heart (Figure 5). The snake constricted for a total duration of just over 2500 seconds (41.7 minutes) when constricting a rat with a simulated heart versus just over 1000 seconds (16. 7 minutes) when constricting a rat without a simulated heart. Additionally, when constricting a rat with a simulated heart the snake produces several peaks in pressure while constricting a rat without the simulated heart it produces one peak that gradually decreases with time. This difference between both pressure and duration was fairly consistent among all tests. The average total time of constriction for snakes constricting rats with a simulated heart was significantly longer ( seconds, min) than snakes constricting rats without a simulated heart ( seconds, min; Table 1; Figure 6). The average total pressure over total time was taken for all of the constriction events; this was the integral of the pressure versus time graphs. The average total pressure over the entire constriction event for snakes constricting rats with a simulated heart was again significantly larger than snakes constricting rats without a simulated heart (Table 1; Figure 7). Discussion These results are consistent with our predictions and support our hypothesis in that snakes constricting prey with a simulated heart constrict for significantly longer (Table 1, Figure 6) and with significantly greater pressure (Table 1, Figure 7) over the entire constriction event as compared to snakes constricting prey without a simulated heart. Additionally, we found a significant effect of individual snake on integral, suggesting there is a pattern to the total amount of pressure generated by snakes. These findings suggest that 11

13 Boas have the ability to detect the presence of a simulated heart in their prey and use this stimulus as feedback for the necessary duration and total pressure of constriction. Although the total time that snakes constricted rats with a simulated heart was significantly greater than snakes constricting rats without a simulated heart, there was no significance impact of the individual snake. This means that snakes responded to the presence of the simulated heart by constricting for a greater period of time in the majority of cases regardless of the specific snake. The integral for snakes constricting rats with a simulated heart was also significantly greater than snakes constricting rats without the simulated heart. However, there was also a significant interaction of individual snakes on the total pressure produced over the constriction event. This result may be due to the varying size of the snakes. The snakes used in this study varied in size from 668 g to 2745 g. A larger snake may produce a greater force over the same time period than a smaller snake would, and could be one reason for the significance in this case. Additionally, there is greater variation seen amongst the snakes constricting rats with the simulated heart versus snakes constricting rats without the simulated heart in both the duration and the integral. Although the reason for this is unknown, it is possible that the overall function of the ventilator throughout the constriction event, coil placement, or potentially snake size may have contributed to this result. Having the ability to detect a heartbeat in prey would be beneficial to snakes for several reasons. As previously stated, constriction is metabolically costly (as measured by oxygen consumption) for snakes and thus it would be beneficial to have a means of determining when expending more energy is no longer necessary. During constriction, some snakes increase their oxygen consumption by a factor of six (Canjani et al., 2003). A human 12

14 going from rest to maximal exertion increases oxygen consumption by around a factor of eight (Schmidt-Nielsen and Duke, 1990). This exertion, however, would not be maintained for an extended period of time, whereas the snakes displaying a six-fold increase in oxygen consumption was a mean over 16 minutes (Canjani et al., 2003). In our study we observed constriction events that lasted an average of 22 minutes and some as long as 45 minutes. Because the simulated heart was active throughout the constriction this suggests that the proximate stimulus for the snakes releasing may have been exhaustion. Canjani et al. (2003) also showed that snakes constricting for the longest period of time (16 minutes) also spent the most time inspecting their prey post-constriction (an average of 10 minutes). They suggested that this provided a means of recovery for the snake after constriction because the duration of constriction and amount of time spent inspecting the prey were directly related (Canjani et al., 2003). In their natural environment constricting to the point of exhaustion would be detrimental to snakes. They would potentially lose their ability to effectively defend themselves. Possessing a means to determine when constriction is no longer necessary would benefit the snakes greatly. In addition to constricting for an unnecessarily long time, constricting for too little time could be a problem as well. Because snakes lack limbs they must use their trunk to subdue their prey. If a snake releases its prey before that animal has died the snake opens itself up to potential injury. Snakes constrict with greater pressure when they sense movement in their prey, but simply detecting movement may not be enough. Boas feed on a range of vertebrates, including ectotherms such as iguanas and other lizards (Stafford and Meyer, 2000; Greene, 1983). Many ectotherms possess the ability to withstand relatively high levels of hypoxia, as compared to many endotherms, due to lower mass specific 13

15 metabolic rate (Boutilier, 2001). Early during constriction, prey are likely under hypoxic and hypercapnic stress, so prey that have the ability to function under higher levels of hypoxia could survive the constriction event. Under hypoxic conditions these ectotherms would greatly reduce their movement. There are reports of ectotherms that have been constricted to a point where no signs of life (including respiration) were obvious and the lizard was unresponsive for 2 L hours, but still managed to fully recover after an extended period of time (Reed et al., 2006). Because these ectotherms can withstand tissue hypoxia they are not as easily killed by constriction (Hardy, 1994). Thus, relying solely on the detection of movement in prey would not allow snakes to fully determine when prey have expired. Having the ability to detect a beating heart, however, would allow the snake to constrict until the prey item is dead and no longer a risk. The means by which prey are killed by constriction is not definitively known. It long has been thought that prey die by suffocation or asphyxiation due to the crushing force placed on the lungs during constriction (Hardy, 1994). Recently, it has been suggested that prey may be dying as a result of circulatory failure and not actually suffocation. It has been shown that prey tend to die much more quickly than would be expected if they were killed by suffocation alone (Hardy, 1994). Circulatory failure could be attained in three ways; an increase in thoracic pressure, which does not allow for blood to flow back to the heart, compression of the heart which prevents the heart from filling during diastole, or increased pressure following contraction to the point that the heart muscles cannot relax to contract again (Hardy, 1994). This thought is consistent with our findings: if prey die due to circulatory failure, the snake's ability to sense a beating heart would be an important indication that constriction was no longer necessary. 14

16 To test our hypothesis our experiment sought to simulate a prey heartbeat in a manner as close to the normal physiology of a rat as possible. Moon (2000) examined how prey movement, simulated heartbeat and respiratory movement of prey affected constriction in gopher snakes (Pituophis melanoleucus) and king snakes (Lampropeltis getula). It was determined that the simulated heartbeat and ventilation did not have a significant effect on constriction time or pressure and were removed from the remainder of the study. To simulate the heartbeat Moon pulsed an implanted bulb of an unknown size within a mouse; this was done by hand and may not have approximated the volume, pressure, or rate of typical cardiac activity. Using a ventilator to simulate the heart, our experiment simulated a heartbeat in a more controlled, constant manner, similar to that of a living prey item. Snakes appear to be much more sensitive to substrate vibrations than air-borne sound and can detect these vibrations in both their head and along their body (Hartline, 197 I). The pit viper (Trimeresurus jlavoviridis) has both touch and vibrotactile neurons. The touch neurons responded to von Frey hairs at 5-10 mg of stimulation (Terashima and Liang, 1994). Our study did not analyze the amount of pressure produced by the simulated heart on the body wall of the rat, but this could be evaluated in the future. It must be taken into consideration that because the detection of vibration is so sensitive, other factors may come into play when snakes encounter rats with the simulated heart. The ventilator produces vibration that could potentially be detected by the snakes. We discount this as a possible explanation for the differences in snake response to the presence or absence of the simulated heart because we ensured that there was no physical contact between the snake cages and the ventilator. Additionally, after each strike we verified that the external tubing was not in contact with the snake, if it was it was relocated while the snake was forming its coil. 15

17 To pursue this study further it would be beneficial to examine the response of snakes to the simulated heart at different points during the constriction event. Turning the simulated heart on further into a constriction, or turning it off once constriction has already begun would provide insight into how snakes use this detected information. It also would indicate the ability of the snakes to adapt to changes during constriction. Examining the relationship between the function of the simulated heart and the length and overall pressure of constriction would be beneficial water it did not maintain identical as well. Because the ventilator is not meant to function with performance throughout every constriction test. It would be interesting to examine whether that played any part in indicating when the snake removed its head from the prey. Overall, Boas seem to have the ability to detect a simulated heart in their prey and alter their constriction of that prey based on the presence or absence of the simulated heart. Boas seem to use the heartbeat of their prey to deterrn ine the necessary pressure and time needed to effectively subdue their prey. Acknowledgements I would like to recognize my honors committee, Professor Scott Boback, Professor Charles Zwemer and Professor David Kushner for all of their help and support throughout this project. I would especially like to thank Professor Boback for the amount of time and energy that he dedicated to this project. Special thanks to Professor Zwemer, without whom I would not have been involved in research in the first place. I would also like to acknowledge that this was a collaboration with both Amanda Hayes and Katie McCann, whose help was invaluable in completing this project. Thank you to Professor Forrester for 16

18 all of his help with statistics. I would like to thank Mike Potthoff for his technical support as well. References 1. Baudinette, R. V Scaling of heart rate during locomotion in mammals. J. Comp. Physiol. 127: Boback, S. M A morphometric comparison of island and mainland boas (Boa constrictor) in Belize. Copeia 2: Boutilier, R. G Mechanisms of cell survival in hypoxia and hypothermia. J. Exp. Biol. 204: Buning, T. C Thresholds of infrared sensitive tectal neurons in Python recticulatus, Boa constrictor and Agkistrodon rhodostoma. J Comp Physiol 151: Canjani, C., Andrade, D. V., Cruz-Neto, A. P., and Abe, A. S Aerobic metabolism during predation by a boid snake. Comp. Biochem. Physiol. A 133: Cundall, D. and Deufel, A Striking patterns in booid snakes. Copeia 4: Cundall, D., Greene, H. W. and Schwenk, K Feeding in snakes, p In: Form, function, and evolution in tetrapod vertebrates. Academic Press, San Diego, CA. 17

19 8. Dilsaver, S. C., Overstreet, D. H. and Peck, J. A Measurement of temperature in the rat by rectal probe and telemetry yields compatible results. Pharmacol Biochem Behav 42: Ebert, J., Mtiller, S. and Westhoff, G Behavioural examination of the infrared sensitivity of ball pythons. J. Zoo!. 272: Greene, H. W Boa constrictor (Boa, Bequer, Boa Constrictor). P In: D. H. Janzen, ed. Costa Rican Natural History. The University of Chicago Press, Chicago. 11. Gunther, B Dimensional analysis and theory of biological similarity. Physiol. Rev. 55: Hardy, D. L A re-evalutation of suffocation as the cause of death during constriction by snakes. Herpetol. Rev. 25: Harkness, J.E. and Wagner, J.E The Biology and Medicine of Rabbits and Rodents. Williams and Wilkins, Media, PA. 14. Hartline, P.H Physiological basis for the detection of sound and vibration in snakes. J Exp. Biol. 54: Heinrich, M. L. and Klaassen, H. E Side dominance in constricting snakes. J Herpetol. 19: Holtzman, D. A From slither to hither: orientation and spatial learning in snakes. Integr. Biol Loop, M. S. and Bailey, L. G The effect of relative prey size on the ingestion behavior of rodent-eating snakes. Psychon. Sci. 28:

20 18. Lourdais, 0., Brischoux, F. and Barantin, L How to assess musculature and performance in a constricting snake? A case study in the Colombian rainbow boa (Epicrates cenchria maurus). J. Zoo!. 265: Mehta, R. S. and Burghardt, G. M Contextual flexibility: reassessing the effects of prey size and status on prey restraint behaviour of macrostomate snakes. Ethof. 114: Moon, B. R The mechanics and muscular control of constriction in gopher snakes (Pituophis melanoleucus) and a king snake tlampropeltis getula). J Zoo!. 252: Proske, U Vibration-sensitive mechanorecptors in snake skin. Exp Neurol 23: Reed, R. N., Green, S., Boback, S. M. and Montgomery, C. E Ctenosaura melanosterna predation. Herpetol. Rev. 37: Schmidt-Nielsen, K. and Duke, J.B Animal Physiology. Cambridge University Press, New York, NY. 24. Secor, S. M. and Diamond, J Determinants of the postfeeding metabolic response ofburmese pythons, Python molurus. Physiol. Zoo!. 2: Stafford, P. J. and Meyer, J. R Boas, p In: A Guide to the Reptiles of Belize. Academic Press, San Diego, CA. 26. Terashima, S. and Liang, Y Touch and vibrotactile neurons in a crotalin snake's trigeminal ganglia. Somatosens. Motor Res. 11: Young, B. A. and Morain, M The use of ground-borne vibrations for prey localization in the Saharan sand vipers (Certastes). J. Exp. Biol. 205:

21 28. Von During, M The radiant heat receptor in the Scales of the upper jaw of Boa constrictor. Z. Ant. Entwickl 145: Dorsal Spinal is-semispinalis Longissimus dorsi Iliocostalis Ventral Figure 1. Cross sectional diagram of a snake displaying the three main epaxial muscles. The vertebra is depicted in gray. 20

22 (a) ~ u 38 QI ~ ~ 36 ~ 35 :l K 33 E ~ Time(min) {b) u 37 QI QI g, 36 QI e as QI ~ 34,. ~ 33 Q. ~ 32 ~ 31 ~ ~ 30' o Time(mln) (c) 39 ~ 38 u QI 37 QI ~ 36 c ~ 35 QI...,. l; 34 ~ 33 c. E 32 QI ~ o so 60 Time (min) Figure 2. Cooling curves for three different size rats. Displays the time to cool for (a) a 113.l g rat, (b) a 202.8g rat, and (c) a 350.2g rat. n=l in all cases. 21

23 Rat~~~- -- HMH -- Thoracic Abdominal Stop cock line line line Figure 3. Constriction testing setup. The ventilator was connected in series with intravenous medical tubing, two stopcocks and the probe inserted into the rat. The simulated heart, the thoracic pressure bulb and the abdominal pressure bulb all were in series with individual pressure transducers (circles), P 1, P2, and P3 respectively, and the AID recording system. 22

24 A B Figure 4. (A) Simulated heart and pressure probe setup, the pressure probe (white bulb) is within the probe used for the simulated heart (clear bulb). (B) Magnification of the simulated heart probe and pressure probe with suture to secure. 23

25 A 250, Ol I E 150 E '-" Q) '-- ::::i 100 Cl) Cl) Q) ' B Time (s) 250, Ol I E 150 E '-" Q) '-- ::::i 100 en Cl) Q) ' Time (s) Figure 5. Representative pressure vs. time graphs for an individual snake constricting a rat (a) with and (b) without the simulated heart. 24

26 Table 1. Results of an unbalanced two-way analysis of variance. Table displaysµ, the average total duration or average integral of constriction. The factor represents the effect of a, the simulated heart, ~' the effect of individual snakes, and a~, the interaction between individual snake and the simulated heart. The table displays the degrees of freedom ( dt), type III sum of squares, F-value and p-value. µ Factor df Type III SS F-value p-value Duration a al Integral a < al

27 0 0 - in 0J 0 0 0,.-.._ 0J ir: '--' (!) 0 a 0 io f:: l{) No Heart Simulated Heart Figure 6. Comparison of the total constriction duration for snakes constricting rats with and without the simulated heart. Along the x-axis the data is grouped into no heart and simulated heart. Within each one of those categories snakes are arranged 1-14 from left to right, multiple measures having the identical x-value. The horizontal lines represent the average duration of constriction. Duration was defined as the point from the initial strike to the point at which pressure dropped below 3% of maximum. (n = 53) 26

28 No Heart Simulated Heart Figure 7. Comparison of the area under the curve (integral) for pressure versus time graphs of snakes constricting rats with and without the simulated heart. Figure displays the integral for both abdominal and thoracic traces. X-axis labels as in Figure 6. Horizontal line represents the average integral of constriction. (n = 53) 27