Tansey Abstract The skeletal system in starfish greatly influences behavior and ecology. Locomotory behavior, especially aspects of flex¬ iblity, is dependent on the properties of the skeletal system. Anatomical and mechanical properties are here examined to determine the relationship of ossicle mass, material stiffness, and flexibility. Flexibility behavior is also observed to examine the range of degree and rate of bending among the3 subject species. It is concluded that a) The greater the volume fraction of ossicles, the stiffer the tissue and b) The greater the volume fraction of ossicles, the lesser the flex¬ ibility. It is also found that rate of flection varies more among the species than does the degree of flexibility. Introduction Much of starfish behavior is influenced by the skeletal system. This effect has been studied in feeding on bivalves (Eylers '76) and in burrowing behavior (Heddle '67). The effect of skeletal structure on flexibility, however, has not been examined. The ability to survive on a complex 3-dimen¬ sional rocky coastline with wave action, predators, tides and varying food sources is probably in part a function of flex¬ ibility. Sturdiness would benefit an intertidal starfish that deals with constant compressive wave forces while good flexibility allows movement in a greater variety of micro¬ habitats (crevices and open faces) giving greater access to a greater number of prey and hiding places. This study was undertaken to examine the effect of anatomical and Tansey and material properties of 3 species of intertidal starfish; 2ycnopodia helianthoides,Pisaster ochraceus, and Patiria miniata,on their flexibility. These species were chosen because of their differing habitats, morphologies and behaviors. Materials and Methods Ossicle Anatomy Starfish of the above species were collected from Monterey Bay, California and kept in seatable tanks at Hopkins Marine Station, Pacific Grove, Ca. Patiria and Pisaster of a typical size (130 and 170 g respectively) and small to medium size Pycnopodia (260 g) were used for a preliminary ossicle morphology examination. The animals were first relaxed in a solution of 50% sea water and 50% magnesium chloride solution (70 g Mgcl in 1 1. H,O) (Heddle '67) then cleaned of soft tissue in a solution of one-third household bleach (active ingredient 5.25% sodium hypochlorite) and two-thirds sea water for 2-4 hours. An examination was made of the individual ossicles and their organization. Flexibility Flexibility experiments were performed on instarfish of each species; the anus to arm tip lengths of which were recorded. The starfish were relaxed in the Mgcl solution for over 12 hours, then manually deformed around glass tubing with radii ranging from 1.5 mm to 5.6 cm. Measure¬ ments of the minimum radius to which each could bend were -2- Tansey taken for a lateral bend, an oral flection and an aboral flection. The radius of curvature was then normalized to the arm length. An obstacle tank (fig. 1) was constructed to observe natural bending behavior. 11 PVC pipes of radii ranging from 1 cm to 5.5 cm were attached to the bottom and sides of the concrete tank with marine epoxy. Natural bending around the pipes was observed in 12 starfish; 6 Patiria 4 Pisaster, and 2 Pycnopodia. Bending was also induced by manually setting the starfish on or against various pipes. The starfish were also manually inverted so that righting behavior could be observed. Material Properties The ratio of dry ossicle mass to total wet mass was determined for a single arm of 1 individual of each species. After the arms were removed from the animals and soaked in 50% NaoH (Eylers '76) to remove soft tissue, the ossicles were separated, dried with a heat lamp, and weighed. The stress-strain relationships were determined for patches of aboral skeleton (the more interspecifically varied of the 2 sybsystems) for each species. These stress¬ strain curves were determined from the data yielded by the starfish skeleton stretching machine (fig. 2) The test sample was clamped between a clamp mounted on an anchored double cantilever beam and a clamp mounted on a plate moved by a lead screw. Force was sensed by deflection of the double beam and was monitored by a linearly variable differ¬ -3 Tansey entiated transducer (LVDT) on the anchored plate. Deforma¬ tion of the sample was monitored by a LVDT mounted on the moving plate. The voltage output of each LVDT was recorded on a chart recorder. The machine could measure strain up to 40% and stress up to.85 N. The 18 aboral patches (6 from each species) were removed from starfish that had been relaxed in Mgcl and cut into hourglass shapes with a rectangular center sections. The volumes of the center sections were calculated from meas¬ ured length, width and thickness. For the sake of numerical convenience, the length was kept at a constant 5 mm for all tests. Volumes varied between 7 x 10"0 m2 and 3.6 x 10' m for the 18 patches. Assuming the sample to be isovolumetric, a new cross-sectional area was determined at each strain. Stress (0) and strain (8) could then be calculated. force 6 cross-sectional area n LAL where L is original length and LAL is the new stretched length. These stress-strain curves were determined at 2 different strain rates (strain rate is £ - 8/unit time): £ - 3.5 x 10 " (elapsed time = 96 sec) and £ = 2,8 x 10 sec sec (elapses time = 12 sec). 3 trials were run for each species at each strain rate. Young's moduli were then calculated from the slope of a plot averaged for the 3 trials. Young's modulus is the stress at a natural strain divided by that natural strain. The strains observed in nature -4- Tansey were calculated using an average radius of curvature observed from the starfish obstacle tank experiment X -4L - where X is extension, ris the distance from the neutral axis (fig. 3) to the surface undergoing tension of compressive stress, and R is the radius to which the arm is bent (Wainwright et al '76) Results Ossicle Anatomy The starfish skeletal system has been illustrated elsewhere (Eylers '76, Heddle '67) for other species. Both the ambulacral and aboral skeletal systems for the 3 species examined here are pictured in figure 4. The ambulacral systems are similar in the species if the 2 rows of plate-like ossicles next to the ambulacral ossicles in Pycnopodia are considered part of the aboral skeleton. The aboral "lattice-work" structure in Patiria and Pisaster differs greatly from the scattered spines in Pycnopodia, but upon closer examination it is seen that even the 2 "lattice-work" structures are basically different. Patiria has a flat, 2-dimensional shingle-like form; the anchored ends of each ossicle being covered by the exposed portion of the ossicles immediately below it (more oral). Pisaster has a complex 3-dimensional lattice composed of small irregularly shaped ossicles anchored together like -5 Tansey a stone wall. Flexibility Pycnopodia is the most flexible of the 3 starfish examined (table 1). Lateral flexibility in this species is approximately half that of oral or aboral flixibility. Pisaster shows approximately equal values of flexibility in all directions. Patiria, the least flexible species, is half as flexible as Pisaster in the lateral and oral flections and one fourth as flexible in the aboral flection. To check the values from this experiment, behavioral observations were made of the 12 starfish in the obstacle tank. If placed on the pipes manually, all three species bend around the pipe to hold on with a majority of tube feet. It is noted that if the radius of pipe is smaller, the period of time of bending is longer. The best demon¬ strations of flexibility by all three species is obtained by inverting the starfish and observing righting behavior, The rate of flection seen in this experiment varies among the species more than the degree of flexibility. Pycnopodia was the fastest bender, and Patiria was the slowest. Material Properties The contribution of dry ossicle mass to total wet mass is 4% in Pycnopodia, 15% in Pisaster, and 16% in Patiria. Stress-strain curves are determined for patches of the aboral skeleton at 2 different strain rates (as described in material and methods) and are shown in figures 5 and 6. Each outgoing curve (extension) and returning curve -6 Tansey (relaxation) represents the average of 3 tests. At the higher strain rate, considerable plastic flow is observed. (20% in Pisaster, 40% in Patiria and 50% in Pycnopodia.) At the lower strain rate, plastic flow is less (20% in Pisaster, 0% in Patiria, and 0% in Pycnopodia). Naturally occurring strain is calculated (as described in materials and methods) to be 0.28. Young's modulus was calculated for each extension using this natural strain and the corresponding stress. Patiria is stiffest at the higher & (3 times as stiff as at the lower 6mp). Pisaster and Pycnopodia are stiffer in the slower test than in the faster test (Pisaster 5 times as much). Discussion The comparative analysis of the 3 species in this study revealed a relation between skeletal anatomy, mechani¬ cal properties and flexibility. Is there a mechanistic basis for this relationship? The differences in organization of ossicles among species can be used to explain the results obtained from the manual deformation of relaxed starfish. Pycnopodia is only half as flexible in a lateral bend as in either an oral or aboral direction. This fact can be explained by the presence of the 2 lateral rows of ossicles on either side of the ambulacral ossicles. When compressed, these ossicles jam into each other to form a solid rigid unit that resists further compression. Patiria's aboral flection -7- Tansey is half that of other bends due to the greater height of the arm (greater arm cross-sectional area) and the large sturdy ossicles found on the proximal aboral surface. The greater arm cross-sectional area would need a greater stress to show the same strain. The may be responsible for Patiria's general inflexibility. The aboral ossicles are large enough to effect aboral flection in the same way that the lateral ossicles in Pycnopodia effect lateral bend in that species. A rough prediction of flexibility can be computed from the ratios of ossicle mass to total mass. Using the relationship of density, mass and volume, these mass ratios can be expressed as volume fraction ratios. If the density of soft tissue is approximately 14 and the density of ossicles is 2.6 g cm (Jones '70 as cited by Koehl 182), the ratio of ossicle volume to total volume (the ossicle volume fraction, Vpo) is Vo OM 1 2.6 q cm Vro - Vr 2.6 g om total massl m This equation yields values of ossicle volume fraction of .016 for Pycnopodia, .064 for Pisaster, and .068 for Patiria. (These values normalize to a 1:4:4.25 ratio for the 3 species). The Young's modulus of ossicle Caco, is approxi¬ mately 137 GN m (Wainwright et al '76) making it very inextensible in comparison to the soft tissues likely to be found in the arm, such as ligament, muscle, connective tissue, and epidermis, which have Young's moduli between 10* and 102 N m (Wainwright '76). Thus as the volume -8 Tansey fraction of ossicles increases the extensibility of the tissue decreases. If the volume fraction of ossicles is 1, the total extensibility (Kay will be essentially O. If, however, the volume fraction of ossicles is O, the will represent the extensibility of the soft tissues max (k). This linear function is illustrated in figure 7, and can be expressed as - -k Vpo t k max The extension experienced by the bending arm as a whole is X-2L - (fig. 3) L where X is extension of the aboral tissue, r is the distance from the neutral axis, and R is the radius to which the arm is bent (Wainwright et al '76). Therefore, Vro tk --kyptkor where 1/R is equal to flexibility. This shows that the greater the volume fraction of ossicles, the larger the radius of bending and the smaller the flexibility. If the k values are the same among the 3 species, the radii of bending will have the same 1:4:4.25 ratio as the ossicle volume fractions. This ratio is not experimentally observed (table 3) implying that the k values differ. It should, however, be noted that the experimental results are qualita¬ tively correct: the greater the ossicle volume fraction, the less the flexibility. Using an average r value of 2.8 cm and the observed radius of curvature (oral flection) the k values can be computed using the formula -9 Tansey kRIIo as 14.2 for Bycnopodia, 2.3 for Pisaster, and 1.4 for Patiria, The results of relaxed starfish flexibility tests conflict somewhat with observations of natural bending in the obstacle tank. The degree of flexibility seems to vary less among species at natural rates of flection. The importance of such observations would be greatly strength¬ ened if more exhaustive statistically relevant examination were made. Nevertheless, these results encouraged analysis of material properties to explain the relation of material stiffness and strain rate. Stiffness varies among species and between rates of strain for each species. This variation of stiffness can be related to ossicle mass (table 3). For a composite material arranged serially (roughly the situation in a starfish arm) the inverse of Young's modulus is equal to the sum of the ratios of volume fraction to stiffness of each component in the material (Wainwright et al '76). . 1 - Vro + Vro FST where Eg is the total stiffness, Eg is soft tissue stiffness, E, is ossicle stiffness and Vy is ossicle volume fraction. Again the ossicle stiffness is so much greater than the stiffness of the soft tissue that the ratio of ossicle volume fraction to ossicle stiffness is negligible. Therefore, Esr E 1-V -10 Tansey The greater the ossicle volume fraction, the greater the Young's modulus. If the Eg values are the same among the species, the total stiffness will be directly proportional to the ossicle volume fraction. An exact ratio is not observed in the data implying different Eg values for the species. It should again be noted that while the data didn't perfectly conform to expectations, it still reflected the general trend that the greater ossicle volume fraction, the greater the stiffness. Using E values from the faster strain test, approximate Eg values can be calculated using the formula Esr“ Er(1 - Vpo) as 6.1 x 10° N m"4 for Pycnopodia, 1.6 x 10 Nm2 for Pisaster, and 1.3 x 10° N m4 for Patiria. Any relation between stiffness and flexibility in these experiments is not clear. The major factor of confusion is the change of stiffness with strain rate. Pycnopodia is approximately equally stiff at both strain rates. Patiria is 3 times stiffer at the faster strain rate; Pisaster is 5 times stiffer at the slower strain rate, Four possible explanations can be set forth. 1) Individual differences among members of the same species may be reponsible. Different individual starfish were used in Pisaster and Patiria for the 2 different strain rates. 2) There may exist an optimal strain rate for each species where the value of stiffness is lowest. If so, the stiffnesses recorded here represent only 2 points on -11 Tansey a curve, the minimum value of which is yet to be determined. 3) The amount of tissue damage done at the 2 strain rates may have been different. The plastic flow, a measurement of material destruction, is much greater at the faster rate (up to 50%) than at the slower one (only 20%). Patiria may have stretched at the fast rate while the other 2 species may have just torn. Both the problems of individual dif¬ ference and breaking/stretching can only be fully examined by statistically valid repetition. The last explanation (4) may not only clarify the differ¬ ences in the stiffness at different strain rates but also explain the lack of a direct parallel between ossicle mass ratios, degree of flexibility and stiffness. Even if it is assumed that the soft tissue components of each species skeletal system are the same, the relative volume of any 1 type of soft tissue (e.g. ligament, muscle) to ossicle volume and the organization of the soft tissue around the ossicles differs greatly. An extensionin one species may be lim¬ ited by ligament while in another species muscle may be the limiting factor. These differences may account for the variety of k and Eg values calculated in previous equations. Further experiments should be done to clarify this situation. Both tensile and shear stress-strain curves need to be determined at many more strain rates and percent strains for a larger number of individuals. Further ana¬ tomical and histological examinations would be invaluable. Nevertheless, these experiments do show that flexibility and stiffness are related to ossicle mass. It is also -12 Tansey found that rate of flection varies more among the species than does degree of flexibility. Perhaps these results and theoretical predictions will be considered in further ex¬ planations of starfish flexibility and locomotory behavior. Acknowledgements I am very grateful to M. Denny for his patient and encouraging direction, consultation and inspiration. His contribution of time and energy was limitless. I would like to also thank S. Denny and K. Denny for their under- standing. Thanks also go to M. Diab and Wm. Gilly for critical discussion, K. Calhoun for showing pity and providing typing skills, and F. Sommer for technical assistance and humorous inspiration. -13 Tansey Literature Cited 1) Eylers, John P. (1976) Aspects of Skeletal Mechanics of the Starfish Asterias forbesii. J. Morph., 149: 353-368. 2) Heddle, Duncan (1967) Versatility of Movement and the Origin of the Asteroids. Symp. Zool. Soc. Lond., No. 20, 125-141. 3) Koehl, M.A.R. (1982)- Mechanical Design of Spicule¬ Reinforced Connective Tissue: Stiffness. J. Exp. Biol. 98: 239-267. 4) Wainwright, S.A., Biggs, W.D., Currey, J.D., Gosline, J.M. (1976) Mechanical Design in Organisms. London: Edward Arnold. Tansey Figure Legend Figure 1. Starfish obstacle tank. Eleven PVC pipes with radii ranging from 1 cm to 5 cm. Used for behavioral observations. Figure 2. Starfish skeleton stretching machine. Force and deformation were measured for aboral skeletal patches to determine stress-strain curves, Figure 3. Beam theory extension. Engineering model to relate surface extension to radius of curvature. Figure 4. Ossicle morphology. Ambulacral and aboral skeletal systems are illustrated. P-proximal; D= distal; AB=aboral; O-oral. Figure 5. Stress-strain curves. Low plastic flow at low strain rate. Pathway represents extension and relaxation. Figure 6. Stress-strain curves. High plastic flow (502) at high strain rate. Pathway represents extension and relaxation. Figure 7. Mathematical relation of extension and ossicle volume fraction. -15 Tansey STARFISH ARM FLEXABILITY (MINIMUM BENDING RADIUS /ARM LENGTHJX IO SPECIES ABORAL LATERAL ORAL FLECTION FLECTION BEND PYCNOPODIA .55 .3 .23 PISASTER 2.4 2.4 .9 PATIRIA 4.3 8.9 4.3 Table 1 ABORAL SKELETAL STIFFNESS YOUNGIS MODULUSX ION SPECIES &2.8XIO 2SEC £3.5XO 9SEC 1.2 PYCNOPODIA 62 7.8 PISASTER .7 4.9 4 PATIRIA CALCULATED AT &=.28 Table 2 C X 0 0 I — o u LL O — E I — 0 O o — — -17- Tansey O 1 — L — STARFISH ÖBSTACLE TANK Figure 1 Tansey STARFISH SKELETON STRETCHING MACHINE LVDT ANCHORED LEVER PLATE MOVING PLATE LVDT Figure 2 -19 DOUBLE CANTILEVER O CLAMP ......... G 0 CLAMPS Tansey BEAM THEORY EXTENSION — X = r/R Figure: 3 -20- Tansey COMPRESION NEUTRAL AXIS TENSION Tansey OSSICLE MORPHOLOGY ABORAL PYCNOPODIA A Da 4 o O LATERAL AMBULACRAL PISASTER. ABORAL AMBULACRAL PATIRIA AB O E S ABORAL AMBULACRAL Figure 4 -21 C 50 45 40 35 X N/m 25. 20 15 O. Tansey STRESS-STRAIN CURVES EPYCNOPODIA APISASTER O PATIRIA - 3.5XO SSEC TIME 96 SEC 8 8 O L .3 .O5 15 .2 .25 Figure 5 -22- .35 50- 45 40 35 0 X1O N/m 25 20 15 0 STRESS-STRAIN CURVES H PYCNOPODIA A PISASTER O PATIRIA -2.8XO 2SEC TIME2 SEC — + . 5 .25 2 .O5 6 Figure 6 -23 Tansey —2 — 3 .35 Tansey EXTENSION-OSSICLE VOLUME FRACTION E QUATION 0 X-.-kVrok Figure 7 -24-