THE TEGULA FUNEBRALIS SHORE-LEVEL SIZE GRADIENT: IMPLICATIONS OF SUSCEPTIBILITY TO PREDATION, AND SETTLEMENT R. Eben Polk June 7, 1998 Advisor: Jim Watanabe Biology 175H Hopkins Marine Station Permanent address Permission is granted to Stanford University to use the citation and abstract of this paper Abstract: Past studies have found that the black turban snail, Tegula funebralis, exhibits a shore-level size gradient, with mean shell size either increasing or decreasing in a down- shore direction, at different sites. Predation has been implicated as a driving factor. A shore-level size gradient was found at Monterey Bay, CA. Snails » 15 mm were more abundant than other size classes in the Low intertidal zone, while snails 9 -15 mm were the most abundant class in the Mid zone, and snails £9 mm were the most abundant class in the High zone. Mean size increased in a down-shore direction. The study site had virtually no Pisaster ochraceus, a known predator. Laboratory experiments revealed that small snails are significantly less susceptible to predation by Pisaster, compared with both medium and large snails. Experimental responses to direct contact with tube feet show some size-related differences in short-term behavior. However, these would tend to make small snails more susceptible than larger T. funebralis; mechanisms other than defensive behavior must keep Pisaster from preying on small snails. A qualitative search for juvenile T. funebralis « 2 mm was undertaken at two sites in Monterey Bay. Juveniles were abundant in the Mid zone at both sites, and rare or not present in the High and Low zone, indicating that settlement occurs in the Mid zone, with migration of T. funebralis becoming possible at greater sizes. Results suggest that other factors besides predation, such as settlement, are responsible for some components of a size gradient at a protected rocky shore in south Monterey Bay, including the general absence of small T. funebralis from the Low intertidal zone. INTRODUCTION Tegula funebralis, the black turban snail, is a common herbivorous intertidal gastropod that reaches a maximum size of approximately 22 to 24 mm in the Monterey Bay area. Its predators include Cancer crabs, octopus, and most notably, the seastar Pisaster ochraceus. It is one of several intertidal invertebrate species that exhibit shore- level size gradients (Doering et al. 1983), with mean size increasing or decreasing in a vertical direction along the intertidal zone, at a given geographic site. There are several biotic and abiotic factors which may create and maintain shore- level size gradients in T. funebralis. One such factor is environmental stress, including temperature extremes, desiccation (Marchetti and Geller 1987), and wave action. Another is habitat choice, on both the scale of microclimates and across the intertidal zone. Food availability and differential preferences could also contribute to a size gradient. Other factors include responses to light and gravity (Doering et al. 1983), settlement (Fawcett 1979), and potentially competition. The most well-studied factor that may affect Tegula size gradients is predation. Paine (1969) concluded that a size gradient in which mean shell size increases in a down- shore direction is caused by the presence of Pisaster ochraceus in the low zone. From observations made in Washington, Paine claimed that juvenile T. funebralis settled high in the intertidal. In this scheme, with increasing size over their life span, T. funebralis would migrate down into the low zone to take advantage of the greater food resources found in the form of macroalgae. Although 25-28% of adult snails would be eaten each year by Pisaster in areas of overlap, those that survived there would grow faster. Pisaster would maintain levels of predation in this zone that would continue to leave enough available resources to maintain a selective advantage to downward migration. In the absence of Pisaster, Paine felt that there would no longer be a strong advantage to migrating downward, due to increased intraspecific competition. Consequently, any size gradient would be less marked, or non-existent. Fawcett (1984) concluded that a size gradient with mean shell size increasing over decreasing intertidal heights would actually be observed in an areas with very little or no predation, including that of crabs and octopus. This model for size gradients operated under the assumption that T. funebralis settled as larvae in the mid-zone, the direction of migration of large snails would then determine the size gradient. Large snails, as they are more mobile, would either move down to take advantage of more abundant food resources or, in the case of intense predation, up beyond smaller snails. In areas with high predation, size gradients in which mean shell size increased in the upward direction were observed at several sites, while the opposite gradient was observed in areas with low predation intensities. Despite the prevalence of theories for as a mechanism for establishing and maintaining a size gradient, data on Tegula's size-specific susceptibility to predation is lacking. Differences in susceptibility to predation, among snail sizes, when combined with intertidal height-specific predation, might possibly be an important component of a size gradient. In the present report, I present evidence for a size gradient at Hopkins Marine Station at Monterey Bay, CA, and for size-related differential susceptibility to predation by Pisaster. Evidence is presented for the nature of this size-related difference, gathered in short-term behavioral experiments. Because results pointed to the importance of the specific nature of settlement, and because the predation-based size gradient models depend implicitly on the location of juvenile settlement, I also made qualitative observations of juvenile distributions; these findings are also presented here. MATERIALS AND METHODS Vertical Size Gradient Over 28 days, 10 transects 10 m in length were selected by placing a measuring tape parallel to the water line (horizontal across the shore) in an area within the protected rocky habitat at Hopkins Marine Station, Monterey Bay, CA (36°36' N 121°54' W). I collected T. funebralis from transects during low tides, and ascertained quadrat heights relative to the predicted low-tide sea level with a surveyor's level and a stadia rod. I randomly selected five quadrats 25 cm in size within each transect. Each random distance along a 10 m length of measuring tape was the center of a quadrat. Surface characteristics and habitat type of each quadrat were also noted. I collected all Tegula spp. easily visible to the unaided eye, and measured their sizes (maximum basal diameter) in the lab. I partitioned the snails into groups based on the following intertidal height zones (Ricketts et al 1985): a High Zone (between 5 and 2.5 ft, or 1.52-.76 m, above MLLW), a Mid Zone (between 2.5 and 0 ft, or .76-0 m, above MLLW), and a Low Zone (between 0 and 1.8 ft, or 0-.55 m, below MLLW). Snails were also grouped into size classes: small Tegula (S9 mm), medium (9 - 15 mm), and large ( 15 mm). Because most snails collected were between 3 and 21 mm, arbitrary size classes spanning roughly 6 mm were appropriate. Mean snail sizes in each quadrat with 6 or more snails were tested for correlation with quadrat height (as per Fawcett 1984). Abundances of T. funebralis (per 25 cm' quadrat) were compared by a two-factor ANOVA with size class and height zone as orthogonal fixed factors. Prior to analysis I tested for heterogeneity of variances, using Cochran's test. Susceptibility to Predation by Pisaster In five outdoor tanks, measuring 49 cm x 52 cm x 32 cm deep, I conducted replicate susceptibility experiments. In each tank, I placed two Pisaster ochraceus, a known predator of Tegula spp. Three non-contiguous size classes of snails were used: small (S9 mm), medium (12 - 14 mm), and large Q 17 mm). Ratios of the numbers of snails in the three size classes were chosen with the goal of equalizing the number of random encounters between a seastar and snails of a size class. A simple 1:1:1 ratio of snails in each class would lead to relatively more seastar encounters with large snails, as small snails on an individual basis are less likely to be encountered. On the other extreme, equalizing the substrate area covered by snails in each size class (a 1:2:5 ratio, between Large, Medium, and Small snails, respectively) would lead to relatively more encounters with small snails, assuming they were not completely clustered. Consequently, an intermediate ratio of 1:2:3 was used. Sea stars were collected from the field and placed in holding tanks. I did not begin a given trial until seastars began to feed on Tegula, to ensure that all were capable of capturing and digesting prey. Aster one to three days, all of the seastars had successfully preyed on a few snails. The preliminary snails were removed, and 120 recently-collected snails were placed in each tank (20 Large, 40 Medium, and 60 Small). At least once a day, empty shells were collected and measured. I maintained a constant ratio of size classes in each tank by replacing snails that had been preyed upon. These replacements came directly from the field, when collection was possible; when replacing snails during high tides, I used Tegula that had been kept in an aquarium for one or two days. Each replicate experiment continued for five days. Because the natural reaction of T. funebralis to Pisaster is to climb vertical surfaces to heights above the water line, barriers to prevent this behavior from leading to escape were necessary. Lengths of thin galvanized steel edging, that had a right angle with one inch sides, were cut and placed along the edges of the tanks, about an inch above the water line. I then placed 1.5 inch strips of galvanized mesh over the top of the edging Although snails could crawl an inch or less out of the water, they were forced to stop, or crawl upside-down on the underside of the edging, where the Pisaster could reach. I observed several snails moving up, away from the tank wall under the edging, and then falling to the bottom after reaching the edge; when snails were found that could not right themselves after falling, they were turned over with a clear length of plastic. In addition to the susceptibility experiments with three sizes classes, I conducted three more replicates using only 120 small snails. After five days, all snails were removed and counted. Single-factor fixed effects ANOVAs were used to determine differences in the susceptibility of the three size classes, based on the different proportions of each class that was preyed upon. Behavioral Responses to Physical Stimuli To ascertain if differential susceptibility is the result of behavioral differences among the size classes, I conducted the following experiments. In the laboratory, twelve snails of each size class (total n = 36) were subjected to three treatments: no stimulus, a tweezer point stimulus, and a tube foot stimulus. These experiments were carried out on a submerged horizontal glass surface in the lab. Underneath a glass pane, I placed a sheet of waterproof paper marked with concentric circles of radii 2 cm, 4 cm, and 6 cm. Radii were also marked at 30° intervals. Snails were collected immediately prior to the experiment to ensure that none had time to acclimate to laboratory conditions. I started each replicate run by placing a snail in the center of the circles. After the snail had reattached its foot to the substrate, oriented itself upright, and prepared to crawl, by feeling the surface with its cephalic tentacles, the stimulus (if any) was delivered to a single cephalic tentacle. The length of time taken for the snail to cross the 4 em circle was measured. The initial direction that the snail was facing when the stimulus was delivered was also noted. As the snail moved, I traced its path on a map of the circular "grid, noting the angle between the direction the snail faced initially and the direction in which it crossed the 4 cm circle. Each snail was given all three treatments consecutively, with no stimulus given first, followed by tweezers and then the tube foot. Following a snail's final run, the glass pane was cleaned to remove mucus and other traces, and the water in the tub was replaced. Two-factor ANOVAs were used to test for differences in both displacement time and the number of degrees between the initial direction and final location of a snail at 4 cm displacement. Stimulus type and size class were fixed factors. Distribution of juveniles To define the vertical distribution of recently-settled T. funebralis, I conducted a field search for juveniles. In Oregon, Moran (1997) found that most newly settled juveniles (£3 mm) were on the underside of cobbles. As cobble-sized rocks are found in all height zones, this micro-habitat was one of a few relevant to a determination of height- specific distribution and settlement. Juveniles raised in the lab grew to 2 mm in approximately 160 days (Moran 1997), so I assumed that snails under 2 mm were 5 months or less in age. I searched for juvenile T. funebralis less than 2 mm in maximum diameter at two sites: the protected area at Hopkins Marine Station (hereafter H.M.S.) described above, and a very protected area behind the Monterey Breakwater. The Breakwater site has a more even slope, and many cobble-size rocks, whereas most rocks at H.M.S. are large and fixed. I took two samples in each of the three height zones at each site; for each, a thorough inspection of a group of 8-12 cobbles lying adjacent to each other was made with a hand lens. Presence or absence of juvenile snails was noted. In each sample where snails were found, several were returned to the lab for identification under a microscope, using characteristics outlined by Moran (1997 and personal communication). RESULTS A Vertical Size Gradient Average snail size increased with decreasing intertidal height (Fig. 1), and the two were significantly correlated (p +.001). In sum, 1902 snails from 39 quadrats were included in the correlation analysis. A Cochran’s test revealed a heterogeneity of variances (Test value =.254, Critical value = 235), so data underwent a square root (x+1) transformation before further analysis. There was a significant interaction between intertidal height zone and size class, for T. funebralis abundances (Table 2). Small snails were the most prevalent size class in the High zone. In the Mid zone, medium snails were the most abundant, and Large snails dominated the Low zone (Fig. 2) While large and medium T. funebralis snails were found in all zones, small snails were significantly lower in number than large snails, in the Low zone (Fig. 3, Table 2). In this zone, many small and juvenile Tegula brunnea were collected. Large snails increased in abundance with decreasing intertidal height, while small and medium snails reached peak numbers in the Mid zone. There were significantly more small and medium snails in the Mid zone, compared to the Low zone. Overall, there were twice as many snails in the Mid zone than either of the other height zones (Table 1). Additional post-hoc comparisons showed there was also a significantly higher average total number of snails in each quadrat in the Mid zone, when compared with both the High and Low zones (Table 2). Predation by Pisaster The seastars ate significantly more large snails than medium snails, and more medium snails than small, despite the 1:2:3 ratio in favor of the small size class. In addition, proportions eaten (Fig. 4) were significantly different (Table 5). While roughly 4% of the small snails were taken, over one quarter of the medium snails and over one half of the large snails were taken (Table 4). Variances were not heterogenous (Cochran's Test). In experimental tanks with only small snails, an average of 8% were consumed over 5 days (Fig. 5). Behavioral Responses to Physical Stimuli Times for each snail to move 4 cm from the center point were compared across size classes and stimulus types (Fig. 6). Displacement times were shorter for all size classes when the stimulus was a tube foot, compared to the treatment without a stimulus. Displacement times in response to a tube foot were also lower than those in response to the tweezer-point stimulus. Larger snails were, on average, faster than small snails, for each treatment. Both stimulus type and size class had a significant effect on displacement time, but there was not a significant interaction between the two (Table 6). Although times decrease across stimuli, the relationship between size classes remains roughly the same (Fig. 6); thus, stimuli and size class effects seem to operate independently. Within terms were pooled with Interaction terms, and planned comparisons were made. Average displacement times of small snails were significantly greater than both large and medium snails. The second behavioral measure was absolute degrees difference between initial direction of the snail and the final direction at 4 cm displacement, again compared across size classes and stimulus types (Fig. 7). The interaction between size class and stimulus type was significant (Table 6). The relationship between size classes changes with stimulus type. Mean degrees difference increased for both medium and large snails, from the no-stimulus treatments to the tube foot treatments. In contrast, mean degrees difference was less for the small snails when stimulated by tube feet, compared with the difference when no stimulus was given. Planned comparisons showed that within the group of behavioral responses to a tube foot, small snails displaced themselves significantly fewer degrees, in comparison with medium snails. Other comparisons 10 between size classes for the tube foot stimulus were not significant. Representative snail paths are shown in Fig. 8. Intertidal Distribution of Juvenile T. funebralis At H.M.S., I found no juveniles under 2 mm in the High zone. While juveniles were abundant in the Mid zone, I found only a small number in the Low zone. At the Monterey Breakwater site, I was able to find a small number of juvenile funebralis in the High zone. Juveniles again were abundant in the Mid zone. I failed to find any in the Low zone. See Table 3. DISCUSSION The results of extensive intertidal sampling indicate the presence of a size gradient in the protected rocky habitat at H.M.S.: both mean size and relative abundances of different size classes change significantly among intertidal zones. Because snails increased in mean size in a down-shore direction and Pisaster seemed to be almost completely absent from this site, my findings are consistent with Fawcett's (1984) explanation of how predation (or its absence) can maintain a size gradient. Although the correlation of mean size with intertidal height explained only 11% of the variation in snail sizes, the slope was significantly negative. However, a U-shaped curve might better fit these data, as many of the average sizes in the highest quadrats increased over those in high mid-zone quadrats (Fig. 1). This possibility is intriguing, especially when considered in the context of the possible size-gradient factor of environmental stress. Marchetti and Geller (1987) demonstrated that small T. funebralis desiccated faster than large snails in laboratory conditions. If large snails, in addition to being more mobile, can better withstand the higher desiccation rates of the high intertidal, one might expect to see a slight increase in mean size within a quadrat in the uppermost limits of Tegula habitat, despite the fact that food availability would tend to drive the majority of large snails down in intertidal height. The predation experiments show clearly that small snails (£9 mm) are in much less danger of being preyed upon by Pisaster, in comparison with other size classes. If the decreased susceptibility of small snails in tanks with all three size classes was due simply and only to a preference of Pisaster for larger snails, then one would expect to see an increase in the proportion of small snails eaten, in tanks with only the small class. However, proportions and numbers of small snails eaten increased very little in tanks with only small snails. Together, these two sets of experiments indicate that the lower susceptibility of small snails may be due to an inability of seastars to detect, capture, or handle small snails. The next set of experiments were intended to examine at the nature of the differential susceptibility that was observed. Because large snails displaced themselves away from a tube foot stimulus faster than smaller snails, and because smaller snails did not seem to turn farther away from their initial direction when touched with a tube foot, small snails would in fact seem to be more susceptible to predation. Combined with the results of the predation experiments, this suggests that the decreased susceptibility is not due solely or perhaps much at all to predator choices based on preference. I conclude that some sort of physical limitation prevents small Tegula from being preyed upon. Perhaps Pisaster are less able to recognize small T. funebralis as prey, or perhaps they are too hard to handle. That small snails are in much less danger from predation by Pisaster suggests that a factor different than predation is responsible for their general absence from the Low zone of the intertidal. Settlement seems to be an obvious choice: Fawcett (1979) found that larvae settle mainly in the Mid intertidal region. The results of my search for juveniles under 2 mm, which are likely to be only about 150 days old (Moran 1997), support this. Because snails this small are virtually immobile relative to the whole intertidal zone, the height at which they are found is likely to be the height at which settlement occurred. It is unknown whether larvae actually settle in only the Mid zone, or whether they settle indiscriminately but suffer from high mortality in other height zones High mortality in other height zones, following indiscriminate settlement, would likely also be due to factors other than predation. Jensen (1981) found juvenile T. funebralis under 4 mm primarily in sandy holdfasts and on the undersides of rocks in pools, at H.M.S. If these microhabitats are the most favorable for small snails, and they are found primarily in the Mid zone, then microhabitat may play a role. In addition, Jensen found that 70% of the stomach contents (and hence of diet) of juvenile T. funebralis under 4 mm was detritus. How this apparently non-selective feeding behavior might influence distribution has not been determined. Interspecific competition may affect size classes differently, as well. Almost no small T. funebralis were found in the low zone, where juvenile T. brunnea were abundant in this zone. This zonation is quite clear, suggesting that T. brunnea may competitively displace juvenile T. funebralis in the low zone, although such an interspecific interaction does not seem to occur between adults of the species at this intertidal height (the transition from adult T. funebralis to adult T. brunnea occurs in the high subtidal). These other possible causative factors involving the Tegula funebralis size gradient invite and await future research. ACKNOWLEDGEMENTS I would like to thank my advisor, Jim Watanabe, for helping me in all stages of this research, from conception to completion, and for his generosity in giving time and sharing knowledge. I would also like to thank Isaac Kaplan, Natalie Lu, Sarah Girshick, and Tamara Jaron for making life in our house a ridiculously enjoyable experience. Thanks also to Amy Moran for providing helpful information on the location and identification of juveniles. LITERATURE CITED Doering, P.H., and D.W. Phillips. 1983. Maintenance of the shore-level size gradient in the marine snail Tegula funebralis (A. Adams): Importance of behavioral responses to light and sea star predators. J. Exp. Mar. Biol. Ecol. 67(2): 159-173. Fawcett, M.H. 1979. The consequences of latitudinal variation in predation for some marine intertidal herbivores. Dissertation. University of California, Santa Barbara, California, U.S.A. Fawcett, M.H. 1984. Local and latitudinal variation in predation on an herbivorous marine snail. Ecology. 65(4):1214-1230. Jensen, J.T. 1981. Distribution, activity, and food habits of juvenile Tegula funebralis and Littorina scutulata as they relate to resource partitioning. Veliger. 23(4):333-338. Marchetti, K.E., and J.B. Geller. 1987. The effects of aggregation and microhabitat on desiccation and body temperature of the black turban snail, Tegula funebralis (A. Adams, 1855). Veliger. 30(2):127-133. Moran, A.L. 1997. Spawning and larval development of the black turban snail Tegula funebralis (Prosobranchia: Trochidae). Mar. Bio. 128:107-114. Paine, R.T. 1969. The Pisaster-Tegula interaction: Prey patches, predator food preference, and intertidal community structure. Ecology. 50:950-961. Ricketts E.F., J. Calvin, J. Hedgpeth, and D.W. Phillips. 1985. Between Pacific Tides Stanford University Press, Stanford, CA. Table 1. Size gradient data for abundance of each size class at each intertidal height zone, including total abundance and average abundance per quadrat. Small Total Medium Large 179 High Intertidal Abundance 403 132 92 (14 Quadrats) Average Abundances 12.79 9.43 6.57 9.60 Per Quadrat Abundance 869 Mid Intertidal 301 374 194 (14 Quadrats) 26.71 Average Abundances 21.50 13.86 20.69 Per Quadrat Low intertidal Abundance 118 338 464 (14 Quadrats) Average Abundances 0.57 8.43 24.14 11.05 Per Quadrat Table 2. Size gradient ANOVA and post-hoc comparison data, for the two-factor ANOVA comparing abundances of size classes in different height zones. Source of MS P-value Variation Height Zones 0.001 27.70 7.60 3.23 Size Class 0.89 0.414 5.001 Iinteraction 20.47 5.62 117 364 Within Fotal 125 Significant Tukeys Tests IP value Total 4 Mid Zone » Total &4 Low Zone p-.002 Total ≈ Mid Zone » Total 4 High Zone 1p5.004 Bonferroni Adjustment (18/36 comparisons) Pvalue Small Class, Low « Mid p-.001 Medium Class, Low « Mid P=.041 Low Zone, Large » Smal 1p5.001 ble 3. Qualitative search results for juveniles under 2 mm, at two sites. Table 4. Relative susceptibility: total numbers eaten and mean proportions. Relative Total Numbers Average Susceptibility Proportion Eaten Eaten (5 Tanks ISmall Class 0.04 Medium Class 0.27 Large Class 93 0.53 Table 5. Single factor ANOVA and planned comparisons for susceptibility experiments. Source of F oritical value MS P-value Variation Between Size 0.2956 62.50 4.52E-07 3.89 Classes Within Groups 0.0047 Total Significant Planned Comparisons E value F(025, 1. 12) Crtical value Proportion Small Eaten « Medium Eaten 27.48 6.55 Proportion Medium Eaten « Large Eaten 35.18 Table 6. ANOVA and planned comparisons for analysis of 4 cm displacement times. Source of MS P-value F ontical Variation value 78966 5.94E-06 Stimulus 13.62 3.09 Size Class 41993 7.24 0.001 3.09 Interaction 1875 0.86 2.46 0.32 99 5797 Within 107 Total Significant Planned Comparisons for Displacement Times F1o25 1 103) critical value 5.17 Small Class « Large Class 11.03 Small Class « Medium Class 11.29 Table 7. ANOVA and planned comparisons for analysis of degrees difference between initial direction a snail faced and its final degree location. Source of P-value MS Foritical Variation value 3.09 Stimulus 13277 5.74 0.004 3.09 Size Class 0.39 0.94 2176 7478 3.23 0.015 2.46 interaction Within 2312 Total 107 Significant Planned Comparison for Degrees Difference Etou. 1. 9) crftical value Small Class « Medium Class 5.89 8.46 FIGURE LEGEND Figure 1. Plot of mean snail size in each quadrat vs Intertidal height, fit with a correlation line. Standard error (+-SE) bars. Figure 2. Graph of relative proportions of size classes in each intertidal height zone. Figure 3. Graph of abundances of each size class in each intertidal height zone. Figure 4. Relative susceptibility: mean proportions of each size class eaten in susceptibility experiments, (+/- SE). Figure 5. Proportions of small size class eaten, with and without other sizes present in experimental tanks (+/- SE). Figure 6. Mean displacement times for each size class at the three stimuli (+/- SE). Figure 7. Mean degrees difference for each size class at the three stimuli (+/- SE). Figure 8. Representative snail paths. Shown are three snails, one in each size class. For each snail, no stimulus treatment and tube foot treatment are shown. Note times of displacement, snail paths, and initial direction marked as e 9 25 20 y=-1.2x+18.4 r2-11. P5001 — I I I 28 2 1 kkkv- -2 -1 2 4 HEIGHT (ft) Figure 1 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 Small E Medium HLarge High Intertidal Mid Intertidal Low Intertidal (5-2.5ft) (2.5-Oft) (0--1.8ft) Height Zone Relative to MLLW Figure 2 400 350 300 250 200 150 100 50 Small (x « or = 9 mm) Medium (9 «X « or - 15) ILarge (15 mm «x) Mid Low High Intertidal Intertidal Intertidal (5 -2.5ft) (0--1.8ft) (2.5 - Oft) Height Zone Relative to MLLW Figure 3 0.6 0.5 EO.4 20.3 0.2 0.1 Small (9 mm or less) Medium (12 14 mm) HLarge (17 mm or more Size Class Figure 4 0.8 0.7 0.6 10.5 50.4 20.3 Q. 0.2 0.1 Small, Medium, and Large Size Classes HSmall Size Class Only Experimental Type Figure 5 250 200 None Tweezers Stimulus Type Figure 6 Tube Foot ESmall E Medium ELarge 160 140 120 None Tweezers Physical Stimulus Figure 7 Tube Foot Small Medium HLarge 8 0 9 Tant.