Abstract- The wave-exposed, rocky outcrops of the intertidal zone at Hopkins Marine Station (Pacific Grove, CA) exhibit large numbers of Lottia gigantea grazing and homing on what appears to be shared territories. This contradicts previous studies in which Lottia were found to be solitary and territorial, aggressively repelling intruders and forming distinct territories of bare rock. This study observed 58 individually tagged limpets on four sites located on two vertical rock faces. By taking multiple observations of limpet displacements, recorded on predrawn maps of each site, limpet home scars and territory boundaries were identified. Limpet size and home scar height above mean lower water were also measured. Nineteen individuals were identified as non¬ territorial, 32 as territorial, and 7 as indistinguishable (no movement was recorded). Non-territorial limpets grazed an area larger than those of their territorial counterparts (mean = 351 cm2 vs. 199 cm2). The variation suggests that Lottia territorial behavior may be more complex than previously thought. Page Introduction: The Lottia gigantea is a large, 5-8 cm long limpet that inhabits the rocky shorelines of Western North America. They are found on outcrops exposed to the water, preferring vertical walls where wave action is heaviest. Lottia gigantea are found from the tip of Point Reyes, CA, down to Cerros Island, Mexico. In 1970, John Stimson studied the territorial behavior of Lottia gigantea on the coast of Santa Barbara County, CA. He described the territory of a limpet as a distinct, round or oval patch of bare rock, bordered by algal growth and sessile organisms such as mussels. The rock was slightly discolored with an algal film, and within each area existed one solitary Lottia that grazed on the film. The limpet showed homing behavior, as described by Fisher (1904), but also exhibited a defensive territorial behavior in which a resident limpet would aggressively shove other Lottia and individuals of other species off its territory. Much work has been done on Lottia since Stimson's original observations, and most of it has noted that these organisms are indeed aggressively territorial (Stimson, 1973; Wright, 1982; Wright and Shanks, 1992). However, Lottia observed on the coast of the Monterey Peninsula, CA, have exhibited a drastically different behavior. Although many individuals are observed to maintain territorial patches of rock as described by Stimson, others were found living together in close proximity on large, bare rock walls without territorial boundaries. This apparent lack of territories suggests possible variation in limpet behavior. Page 3 This study investigates the communal behavior of Lottia gigantea near Monterey. It attempts to determine whether limpets in the Monterey area are territorial or are more tolerant of each other and show no distinct territories, and to determine, if they show variation in behavior, what causes this variation. Methods and Materials: Experimental Location and Design The study was conducted at Point Cabrillo, Hopkins Marine Station, Pacific Grove, Ca. Two vertical granite rock faces were chosen as test locations because of the abundant Lottia present, and their ease of access and observation. Both WNW-orientated faces lie in the mid-to-high intertidal zone, and are directly exposed to incoming waves. The first rock face contained two test sites. The two are situated two meters apart, with the smaller Site 2 partially shielded from waves by a large boulder. Sites 3 and 4 are located adjacent to each other on the second rock face, divided by a crevice that prevents limpets from traversing from site to site. This face is located within a shallow cove; although it is exposed to incoming waves, the cove may reduce hydrodynamic forces imposed on the face and its inhabitants. The sites were photographed and measured. Maps of each site were carefully drawn onto graph paper, using the photographs to mark site boundaries and all physical features, including rock formations, algal patches, and mussel beds. Depending on the site, map scales were either 15 or 30 cm of actual substratum per inch of paper. At 30 cm intervals on the sites, 9 mm Page 4 circular markers painted fluorescent orange were epoxied onto the rock with Z- Spar Splashzone Compound marine epoxy. These markers and the cor- responding maps allowed accurate measurements of limpet location. Length and width of shell were initially measured for 58 limpets in the sites. Shell area was calculated by assuming shell shape to be an ellipse, and using the formula: A-nab. a=0.5(length) b=O.5(width) Limpet shells were cleaned and tagged with 1x3 cm color-coded markers glued onto the shells with marine epoxy. The markers were cloth strips painted with fluorescent paint using four colors in various combinations. The cloth allowed permanent adhesion of paint and prevented disintegration of the tags by constant exposure to the waves. Bright colors ensured ease of visibility from a distance in all light conditions. All edges of the tags had to be buried in the epoxy to prevent waves from ripping them out. Several limpets were retagged after waves dislodged the markers before the epoxy hardened. Thus, each limpet was identified by a color code and recorded by an identification number assigned to each tag. Data Collection Observations were taken from vantage points situated high enough above the waves to minimize the exposure of the observer and far enough away to minimize detection of the observer by the limpets. For sites 1 and 2, this was a tall boulder located approximately 4-5 meters in front of the sites. Sites 3 and 4 were observed through 10x binoculars from approximately 40 Page 5 meters on the opposite side of the cove when the tide was high, and by wading out into the cove when the tide was low. Observation sessions included recording limpet location onto the maps and looking for any overt behavior (e.g., confrontation between limpets). Time of day, sea and weather conditions were also noted for each session. A single map was used to record several sessions, using colored pens to differentiate between time points. The distribution of observation sessions was as follows: site 1 was measured 43 times, site 2, 20 times, site 3 and 4, 18 times each. Data Preparation After collecting all the data, each location recorded for each limpet in a given site was labeled on a set of "master" maps constructed on Adobe Photoshop 3.0 using a Macintosh computer. Connecting the outer-most points delineated grazing area boundaries for each limpet. Small areas showing multiple points indicated home scars for each limpet. These locations of high point density occurred because limpets were often recorded at their home scars. Because the limpets were plotted with different colors, the master maps clearly showed grazing territories of each individual. The area of each grazing range was measured by transferring the boundary lines onto grid paper, counting grid boxes, and multiplying by the scaled area of each box. Once home scars were identified, the height of each could be measured as a distance above mean lower water. Height measurements were taken with a surveyor's transit and stadia rod, using USGS Topographic Marker "Mussel Point +1" as the reference point. Each measure¬ Page 6 ment recorded for every limpet was then converted into a distance from that limpet's home scar by measuring the minimum distance between the point and home scar. Tidal heights at the times of observation were calculated using Bryan Wade's Tide Calculator tide prediction Fortran code program. For each time point, the tidal heights were interpolated using the heights at the hours before and after the session. Weather conditions, recorded once per day unless drastic changes occurred, included sky (sunny, cloudy, foggy), and air (windy or calm, temperature) conditions. The data were then compiled in two manners: per limpet and per time point. For each limpet size, home scar height, territory area, number of times when displaced from home scar, fraction of observations when displaced from home scar, and average distance of displacement were listed. At every time point time, tide height, weather condition, number of limpets displaced from home scar, fraction of limpets displaced from home scar, and average distance of displacement were listed. Next, the data taken per limpet were divided into two separate groups, territorial and nonterritorial limpets, that were used for comparisons. The limpets were classified as one or the other depending on two criteria: (1) whether they exhibited aggressive territorial behavior, and (2) whether their territories showed overlapping areas. Aggressive limpets with non-overlapping ranges were deemed territorial. Page 7 Results: Fifty-one of 58 limpets were observed displaced from their home scars more than twice. Two limpets did not move at all during the times observed, and the remaining five moved once or twice for a distance less than their shell length. These seven limpets could not be classified as territorial or nonterritorial; they were, thus, discarded from all data analysis. Thirty-two limpets appeared to graze in an area in which no other limpet was found. The remaining 19 limpets were found to share 7-59% of their grazing area with at least one other individual and are, therefore, nonterritorial. Limpet numbers 2 and 3, 9 and 14, 7 and 15, 10 and 11, 34 and 41, 53 and 57 shared territory space, limpet 433, 437, and 438 shared, as did limpet numbers 26-29 (see figures 1 and 2). During the course of the study, limpet 41 and 44, #36 and #49 showed aggressive encounters in which they met, turned towards each other, pushed and shoved briefly, then turned away from each other and continued grazing along their respective territorial boundaries. Limpet #33 had an encounter with a slightly larger, unmarked limpet that shoved 433 until 433 rotated and escaped back up into its territory. The unmarked limpet promptly turned around and returned to its own territory. Limpet statistics Initially, the limpets from all four sites were tested as one large sample (n = 51). Limpet shell size varied from 1423.2 to 3165.1 mm2 with mean = 1638.8 Page 8 mme and standard deviation = 409.8 mm2. The distribution of shell sizes is shown in figure 3. Territory size showed mean = 255 cm2 and standard deviation = 230.7 cm2 (range = 14 to 1026 cm2). Home scar height above MLLW ranged from 24 to 152 cm. Mean height was 78 cm with standard deviation = 57.6 cm. The limpets were found displaced on the average 35% of the time, and mean displacement = 13.9 cm. Regressions indicated that larger limpets had larger territories, and were found lower on the rock faces. Territory size plotted against shell size (fig. 4), with both axes log transformed, resulted in a significant positive correlation (R = 0.489, P « 0.001, n = 51). Home scar height regressed against shell size (fig. 5) also showed a significant relationship (R = 0.373, p « 0.01, n = 51). The log transformation of both axes in the first regression analysis was performed in order to minimize the increase in variance in the dependent variable (territory size) as the independent variable (limpet size) increased. The average displacement observed at a given time regressed against the fraction of limpets that moved at that time (fig. 6) indicated a significant positive slope (R = 0.580, p « 0.001, n = 98). In other words, as more limpets moved, they traveled farther distances. The ratio of average displacement to territory size regressed against fraction of times displaced (fig. 7) showed a significant relationship (R = 0.57274,p « 0.001, n = 51). This suggested that the more often a limpet moved, less distance it traveled each time. Page 9 No significant relationships were found between average distance or fraction of times displaced as a function of tidal height or time. Comparisons between territorial and non-territorial limpets For territorial limpets, mean shell size was 1679.9 mm2, with a range from 956.1 to 3165.1 mm2, mean territory size = 199 cm2 (14 - 846 cm2), mean home scar height = 73 cm (24 - 132 cm). Nonterritorial limpets averaged 1569.7 mm? (1080.1 - 2048.9 mm2) in shell size, 351 cm2 (77-1023 cm2) in territory size, and 86 cm (47 - 152 cm) in home scar height. T-tests performed between the territorial and nonterritorial means found no significant differences in either limpet size or home scar height. However, the test indicated a significant difference in territory sizes between territorial and nonterritorial limpets. The nonterritorial limpets had larger grazing ranges. Log transformed territory sizes were again regressed to limpet size, also log transformed. Nonterritorial limpets (fig. 8) showed no correlation between limpet size and grazing range size (R = 0.439, p » 0.05, n = 19), whereas territorial limpets (fig. 9) showed a significant positive correlation between the two sizes (R = 0.613, p « 0.001, n = 32). Since the former had a significantly larger range size than the latter, the groups were retested for the significant correlation after removing differences in shell size between the groups. A test for equality of slopes indicated no significant difference between the slopes of the two regression lines (p = 0.629); analysis of covariance could thus be performed. Äfter removing shell size as a source of variance between the two groups, the test still found the mean range size of 351 cm* for nonterritorial Page 10 limpets to be significantly greater than the 199 cm2 mean for territorial limpets (p = 0.002). Regression analysis showed a significant negative correlation between limpet size and height of home scar for territorial limpets (R = 0,519, p « 0.005) (fig. 10), but no significant correlation between the two for nonterritorial limpets (R = 0.110, p » 0.50) (fig. 11). Again, a test for equality of slopes found no difference between the slopes of the two regressions(p = 0.480), and the corresponding analysis of covariance found no significant differences in home scar height between the two groups after removing limpets size as a source of variation (p = O.194). Lastly, regression analysis showed a significant inverse power relationship between the average displacement-range size ratio and the fraction of times displaced for territorial limpets (R = 0.595, p « 0.001) (fig. 12), but no similar correlation (R = 0.356, p » 0.10) for nonterritorial limpets (fig. 13). The slopes of the lines were not significantly different (p = 0.282), and the following analysis of covariance found no significant difference (p = 0.184) between the two groups. Nonterritorial limpets There was no significant correlation between the measured grazing area size and the amount of area shared (R = 0.213, p» 0.20, n = 19) (fig. 14). Each limpet was assigned an adjusted area, defined as a sum of the unshared parts of its measured grazing area and a fraction of the area shared with other limpets. The fraction equaled 1/2 if two limpets were sharing the area, 1/3 if Page 11 three were sharing, etc. The mean of these adjusted area sizes = 307 cm2. The adjusted area sizes were then plotted against the measured area sizes (fig. 15). and a highly significant positive relationship was found (R = 0.994, p « 0.001, n = 19). A final t-test comparing the means of the adjusted areas of nonterritorial limpets and the areas of territorial limpets showed no significant difference in territory size. In other words, all the limpets grazed areas that were approximately the same size, but those that shared grazing areas with others required more total range area than the limpets that grazed isolated ranges. Discussion: About Limpet Statistics... Mean grazing range sizes for both territorial or nonterritorial limpets (199 and 351 cm“ respectively) were found to be much smaller than grazing range sizes measured by other researchers. Stimson calculated territory sizes averaging 900 cm2 per limpet (Stimson, 1970). Various factors may contribute to this large discrepancy. First, Lottia grazing range size changes with algal abundance, shrinking when algae are abundant and enlarging when scarce (Stimson, 1973). Although no data were taken on algal abundance in this study, variation in algal density between Santa Barbara County, where Stimson took his measurements, and here in Monterey might contribute to the size differences. Page 12 Next, the large number of limpets in close proximity on the rock faces might prevent individuals from defending large territories. While Stimson did not taken population density measurements, preventing any comparison with this study, this might also contribute to the communal variation in their behavior. Limpets may lear that defending large territories with others around consumes a large proportion of energy and time that might not be readily available. Tolerance of others could have been learned from experience, just as territorial behavior might have been conditioned from previous experience (Wright and Shanks, 93). Regression analysis showed a correlation between limpet size and home scar height for territorial limpets, but not for nonterritorial limpets. The tolerance that nonterritorial limpets show of others and their ability to share grazing areas allow nonterritorial limpets more freedom of movement to locate areas of better food abundance. As a result, they may not be limited as to where they form a home scar. On the other hand, existence of home scars indicates that they do prefer a "home;" thus, they are not completely nomadic. On territoriality... The two criteria used to determine territoriality of the limpets appeared to be sufficient. During the course of observations, the limpets were seen to graze in distinct areas. For many of the limpets, the boundaries of these areas may be underestimated. Insufficient data prevented a more accurate presentation of these areas, especially those under 100 cm2. However, all limpets designated Page 13 "nonterritorial" were seen in areas shared by other limpets, and all the "territorial" limpets did show recognition of boundary lines, For example, limpet +22 and +23, whose home scars were only 20 cm apart, carefully avoided the other’s territory so strictly that when one came to the boundary, it would halt, pause, rotate around until it was positioned parallel to the boundary, then continue grazing along the boundaries. Limpet #1 and 44, ended a confrontation when each stopped pushing, rotated in opposite directions, and continued grazing along the border of their adjacent territories in opposite directions. These limpets were also only 7% different in size (1561.4 vs. 1664.2 mm2). Also, limpet 433, designated "nonterritorial," entered foreign territory, only to be repelled and briefly chased by the slightly larger territorial limpet, which left its own ground to chase #33 into the latter's territory. Äfter this encounter, the unmarked territorial limpet retreated directly back into its territory, stopping and turning only when it reached its own territory. This indicates two things. First, all of the limpets seem to recognize borders and, thus, territories, as seen by the regular grazing of each limpet in a certain area. This implies a possible chemosensitive (olfactory), visual, or even tactile behavioral response not yet identified. Since Lottia have been commonly observed responding defensively to contact with Pisaster tube feet (Bullock, 1953), it may be possible that they also detect the presence or grazing areas of other limpets. Next, not all limpets respond to, although they may detect, the presence of other limpets' territories. One obviously notes the Page 14 shared grazing areas of nonterritorial limpets, but also some individuals, like f33, disregard the boundaries of even territorial limpets. Interestingly, no limpet, territorial or otherwise, was observed intruding into the area of another's home scar. Some limpets such as numbers 1, 26, 33 and 37 graze past other home scars, but never have they been seen grazing over a scar. Distance between home scars has no relationship to territoriality of neighbor limpets; some territorial limpets home at locations just 10 cm apart while others are 40 cm apart, and nonterritorial limpets and combinations of both show similar numbers in home scar distances. Previous papers (Stimson 1970, 1973) describe limpet territories as circular or oval areas. Although this may be true for solitary, territorial limpets, this did not seem to be the case for communal limpets. Again, insufficient data may be the culprit, but the tested individuals often showed long, narrow territories that border on large rock formations or algal patches. Individuals grazing on such territories include limpet numbers 15, 21, 29, 32, 33, and 53. Rock formations are difficult to bulldoze, so limpets have to accommodate to such obstructions One such formation seems to prevent limpet #32 and #33 from grazing on each other's area. On the other hand, limpets are known to bulldoze algal clumps when forming and maintaining territories, but the limpets in this study were observed grazing around patched of Endocladia in their territories. Some limpets homed at one location, but then traverse great distances to graze. Limpet 410, 411, 418, show this pattern. Page 15 „.And the nonterritorial limpets The amount of shared grazing area may depend on limpet density, local resource abundance, and location. At site 2, four of five limpets shared substantial proportions of their territories (30%-59%). These limpets also homed at close proximity to each other; limpet +27 and +28 homed the closest at 6 cm apart, limpet +25 was 15 cm from +27, and +29 was 30 cm from +28 (fig. Apart from site 2, in which close proximity of individuals and large overlap among ranges showed otherwise, no two limpets were seen in shared areas simultaneously. When one limpet was within the shared area, the other was observed grazing elsewhere. This may be due to pure coincidence, distinct grazing patterns that placed limpets apart at a given time, or a sense of temporal territoriality. Although verification of this speculation will require a completely new study, perhaps the limpets can sense the presence of others such that they will actively avoid the other when grazing in a shared area. This again suggests possible chemoreceptive behaviors in Lottia. Conclusion: The results of this study indicated that variation does exist in the territorial behavior of Lottia gigantea. Nineteen of 51 limpets were designated as nonterritorial limpets because they grazed on shared territories, and exhibited no aggressive behavior towards others. The other 32 limpets were considered territorial since they appeared to maintain and defend territorial spaces. All Page 16 limpets, however, lived communally and in close proximity on the rock faces, Resource abundance and space limitations may play a role in determining the behavior of the limpets, as might possible chemoreceptive mechanisms or a sense of temporal territoriality. Since the study did not include such measurements, this is not conclusive. As indicated, the traditional view of the aggressive Lottia and its territory is not entirely correct; the territorial behavior in the limpets may be more complicated than previously deemed. Acknowledgments: Much gratitude to Mark Denny for his guidance, input, and advise in collecting data and writing this paper. Also, thanks to Jim Watanabe for his help with the statistical analysis, John Lee for his help with the camera set up, and to Bryan Sun and Steve Randle for their help in collecting data. Literature Cited: Bullock, T. (1953). Predator recognition and escape responses of some intertidal gastropods in the presence of starfish. Behavior 5: 130-140. Fisher, W. (1904). Anatomy of Lottia gigantea. Zool. Jahrb. Anat. 20(1): 1-66. Stimson, J. (1970). Territorial behavior of the owl limpet Lottia gigantea. Ecology 51: 113-118. Stimson, J. (1973). The role of the territory in the ecology of the intertidal limpet Lottia gigantea. Ecology 54: 1020-1030. Wright, G. W. and A. L. Shanks (1993). Previous experience determines territorial behavior in an archaeogastropod limpet. Exp. Mar. Biol. Ecol. 166: 217-229. je 19 Figure Legend: Fig. 1) The maps of sites 1 and 2, showing all locations observed of all limpets. Each limpet is color coded, and the lines delineate boundaries of grazing ranges. The areas within each range that show a high density of points indicates home scars. Fig. 2) The maps of sites 3 and 4, showing all locations observed of all limpets. Each limpet is color coded, and the lines delineate boundaries of grazing ranges. The areas within each range that show a high density of points indicates home scars. Fig. 3) Histogram distribution of limpet shell sizes. mean = 1638.8 mm2, standard deviation = 409.8 mm2 range = 956.1 - 3165.8 mm2 n =51 Fig. 4) Territory size (log transformed) regressed against limpet size (log transformed). R =0.489, p « 0.001, n = 51 Territory size shows a significant positive correlation with limpet size. Territory size increases as limpets get larger. Fig. 5) Limpet size regressed against home scar height. R = 0.373, p « 0.01, n =51 Limpet size is negatively correlated with height of home scar. Larger limpets form home scars lower on the rock, closed to the water. Fig. 6) Average displacement of limpets recorded at any given observation session regressed against the percentage of the limpets that moved at that session. R = 0.580, p « 0.001, n = 98 There was a positive correlation between average displacement of limpets and the fraction of limpets displaced at a given observation session. Under favorable environmental conditions, when more limpets move, they move greater distances. Fig. 7) The ratio of average displacement to territory size regressed against the fraction of times a limpet was found displaced. R =0.573, p « 0.001, n = 51 Since the average displacement for a limpet depended on territory size, the ratio of the two provided a normalized measure that could be compared to the frequency of limpet displacements. A significant negative relationship was found. A limpet that moved more frequently covered less distance in its territory each time compared to a limpet that moved infrequently, but traveled greater distances when it did move. Page 20 Fig. 8) Territory size (log transformed) regressed against shell size (log transformed) of nonterritorial limpets. R =0.439, p» 0.05, n = 19 In nonterritorial limpets, there was no significant correlation between territory size and shell size. Size of territory did not depend on size of limpet. Fig. 9) Territory size (log transformed) regressed against shell size (log transformed) of nonterritorial limpets. R =0.613, p « 0.001, n = 32 In territorial limpets, there was a significant positive correlation between territory size and shell size. Territory size did depend, to a great degree, on size of the limpet. Fig. 10) Territorial limpet size regressed against home scar height. R =0.516, p + 0.005, n =32 In territorial limpets, limpet size was negatively correlated with home scar height. Larger limpets were found lower on the rock, closer to the water. Fig. 11) Nonterritorial limpet size regressed against home scar height. R=0.110, p» 0.50, n = 19 In nonterritorial limpets, limpet size did not determine the height of its home scar. These limpets were found homing at all heights. Fig 12) The ratio of average displacement to territory size regressed against the fraction of times a territorial limpet was found displaced. R =0.595, p + 0.001, n =32 Since the average displacement for a limpet depended on territory size, the ratio of the two provided a normalized measure that could be compared to the frequency of limpet displacements. A significant negative relationship was found. A territorial limpet that moved more frequently covered less distance in its territory each time compared to a limpet that moved infrequently, but traveled greater distances when it did move. Fig. 13) The ratio of average displacement to territory size regressed against the fraction of times a nonterritorial limpet was found displaced. R = 0.356, p» 0.10, n = 19 No significant correlation was found between the average displacement to territory size ratio and fraction of times a nonterritorial limpets was displaced. The trend indicates that these limpets behave similarly to territorial limpets, but insufficient data on limpets that moyed less than 10% of times prevented a statistically conclusive correlation. Page 21 Fig. 14) The fraction of shared territory regressed against the measured territory size of nonterritorial limpets. R=0.213, p» 0.20, n = 19 No significant correlation was found. The fraction of territory shared between limpets did not depend on the observed measured size of the territories. Fig. 15) The adjusted territory size regressed against the measured territory size. R =0.994, p + 0.001, n = 19 There was a significant positive correlation between adjusted territory and measured territory sizes. Regardless of the proportion of territory that was shared, the adjusted territory size was directly proportional to measured territory size. Page Page 5 1 2 2*) 4. a 15 cm 30 cm Map! Map 2 Figure 1 268. 29 Page 23 . 97 5 8 3 5 5 2 Page 24 2 3* Figure 3 Shell Size Histogram — -- ssssseseeeee sss — essess---- T 500 1000 1500 2000 2500 3000 3500 Shell Size (mm2) Page 25 Figure 4 Territory Size in Relation to Limpet Size (Log Trans.) E (. — O ) — 2.6 . QOH90 2.2 O O 1.8 ..C 1.4 E 2.9 3 3.1 3.2 3.3 3.4 3.5 3.6 Log Limpet Size (mm»2) Page 26 Limpet Size in Relation to Home Scar Hgt. 3500 — 3000 0 2500 2000 O O OO 1500 0 O O 1000 50E 20 40 60 80 100 120 140 160 Home Scar Hgt. (cm) Figure 5 Page 27 Average Displacement Recorded at any Given Time Point in Relation to the Percentage of Limpets that Moved 40 — 30 O 20 Oc 00 2 O seseseeeseset 10 O 2) eessesesseses -0.2 0.2 0.4 0.6 0.8 0 1.2 % of Limpets that Moved Figure 6 Page 28 Figure 7 0.7 0.6 0.5 0.4 0.3 0.2 0.1 Average Displacement/Grazing Range Size in Relation to % of Times Displaced O... -- ö Q 80o8 do —O L P 0.2 0.4 0.8 0.6 % Times Displaced for a given Limpet Page 29 Figure 8 Territory Size in Relation to Nonterritorial-Limpet Size (Log Transformed.) 9.. 0 2.6 o . Meeeseeesee — o 2.2 1.8 E 1.4 E 2.9 3 3.1 3.2 3.3 3.4 3.5 3.6 Log Limpet Size (mm2) Page 30 Figure 9 Territory Size in Relation to Territorial-Limpet Size (Log Transformed) E — 4 — 2.6 E GO Q OO 2.2 C 1.8 C . 1.4 — — E 2.9 3 3.1 3.2 3.3 3.4 3.5 3.6 Log Shell Size (mm2) Page 31 Territorial-Limpet Size in Relation to Home Scar Hgt. — 3500 — 0 sssces 3000 O 2500 — 2000 O 88 0 000 1500 O — ( 1000 50E 20 40 60 80 100 120 140 160 Home Scar Hgt. (cm) Figure 10 Page 32 Nonterritorial-Limpet Size in Relation to Home Scar Height. 3500 3000 . 2500 2000 C 5 1500 O . 0 1000 50 ETIET 20 40 60 80 100 120 140 160 Home Scar Height. (cm) Figure 11 Page 33 Ave. Displacement/Grazing Range Size in Relation to % Times Displaced of Territorial Limpets — 0.7 — 0.6 Leseeeeeeeeessssssssse 0.5 E 0.4 O --¬ 0.3 3 0.2 8 S0 0.1 3B0o 0 0.4 0.6 0.8 0.2 % Times Displaced Figure 12 Page 34 Ave. Displacement/Grazing Range Size in Relation to % Times Displaced of Nonterritorial Limpets 0.7 —— —------- 0.6 8 0.5 -e- 0.4 c 0.3 — essseeseesssssesesee 0.2 . 0.1 0 0 P 0.4 0.6 0.8 0.2 % Times Displaced Figure 13 Page 35 Nonterritorial Limpet Grazing Area Size and Corresponding Percentage of Shared Area — esssesesssessseesessee 0.8 —.. Seesseeeeeesesessssese 0.6 —. Messseeseeeeeseeessese 0.4 . O 0.2 00 E 200 400 600 0 800 1000 1200 Grazing Area Size Figure 14 Page 36 Measured Grazing Area Size Compared to Adjusted Grazing Area Size 1000 — ... leeeeeessessssssesssesee 800 600 — 400 80 -n----------- 200 88 200 400 600 800 1000 1200 0 Actual Grazing Area Size (cm?) Figure 15 Page 37