Abstract Previous reports on the behavior of Macclintockia scabra have implied that the intertidal limpet moves every time it is wetted by the incoming tide. In a three-week study on the potential effects of location, splash and light level on limpet movement, frequency of movement (number of instances individual limpet moved divided by the number of instances limpet was wet - averaged over all observed limpets) was calculated to be 12%. This is based on 2660 field observations on seventy-nine limpets, dispersed among three sites, differing in degree of wave exposure. I also made 600 observations on twenty-five limpets that had established home scars on rocks transported to the laboratory. Frequency of movement in the lab was 15%. A two-factor analysis of variance (ANOVA) was carried out on frequency of movement, with degree of splash and field site as the two factors. Data from during the day, and at night had to be analyzed separately because all possible combinations of "splash", "site" and "light level" did not occur during my field study. Frequency of movement did not differ significantly among sites. No movement was observed at low tide in either the lab or field. An increased frequency of movement was found to accompany a corresponding increase in splash level, both at night and during the day. However, lab results showed a decreased frequency of movement when limpets were submerged rather than sprinkled by a spray bar. A significantly greater amount of movement (25%) was seen at night than during the day (7%) - a finding that could be due to the higher nighttime tidal heights. Lab results showed no significant difference in day and night movement frequency. In a separate experiment, limpets were exposed to an artificial high tide, during a normally low tidal period. The limpets responded by producing a significantly lower-than-normal frequency of movement (8%). 10/15 tidal simulations resulted in a movement frequency of zero. My findings suggest that Macclintockia scabra may use a combination of splash level, light, and other as yet identified "tidal-timing" cues to induce movement. Location appears to play a less significant role. Introduction The movement behavior and foraging patterns of Macclintockia scabra have been observed and studied for over one hundred years. All of the classic reports present a similar story regarding the limpets' movement: The movements of the limpets are largely controlled by the tides. When the tide is out, they remain practically motionless on the rocks and present no visible sign of life. With the first dash of spray from the incoming tide they begin to move and are apparently active until the water recedes once more" (Wells, 1917). Macclintockia scabra find a very specific location on a rock and grow their shell to fit that location. They establish the site as their home scar and return to rest on the scar every time the tide is low. As described above, during the incoming high tide, the limpets lift up their shells and begin to forage around for food - once again returning to their home scars when the tide recedes. This movement pattern, closely correlated with the natural tidal cycle, has been found to be invariable and all-inclusive. In an observational study on the movement of the limpets, Willis Hewatt (1940) notes: "After the animals had been submerged several times by the oscillating water, they lifted their shells and started moving very slowly in various directions. The movements began at 8:45 a.m., and by 9:00 a.m. all of the marked limpets had moved away from their spots." In making preliminary observations of Macclintockia scabra, living in the high- intertidal region of the Hopkins Marine Life Refuge, it was quickly apparent that rather than foraging actively during the high tide, the majority of limpets remained motionless on their home scars. This low-level and inconsistency of movement was unexpected. After several daytime high tide observations of-0% movement frequency, I observed the limpets' behavior during a nighttime high tide, and was surprised to see a substantially greater number of limpets moving. This was another unexpected discovery, as previous studies have found no significant difference in the daytime and nighttime movement frequencies of Macclintockia scabra (Sutherland, 1969). Tembarked on a three-week observational study, with the purpose of determining an overall movement frequency for the limpets, as well as exploring what cues the limpets were using to induce them to move. Did the limpets move significantly less during the day than at night? If so, was this due to a difference in day/night tidal conditions or were the limpets responding to the amount of light exposure they were receiving? Did the movement frequency of the limpets vary significantly in different locations? Materials and Methods 1 conducted a three-week observational study of Macclintockia scabra at three field sites, each characterized by a different degree of wave exposure. I used a variety of brightly colored nail polish to identify limpets and their home scar location at each field site (Table 1). I made observations of every limpet during daytime and nighttime high and low tides. For each low tide observation, I located each limpet and checked to see whether it had returned to its home scar or had established a new home scar location. For each high tide observation, I categorized the degree of splash (Table 2) each limpet was receiving and noted which limpets had left their home scars to forage for food. Limpets were considered wetted at splash 1 category or above. I calculated a movement frequency (E'times moved times wetted) for each limpet and used these to tabulate overall frequencies for each combination of splash level, site and degree of light exposure (day or night). These were compared using analysis of variance (ANOVA). No degree 3 splash occurred at either of the protected sites during the day throughout the observational period, making it impossible for me to conduct a three factor ANOVA of the data. Instead, 1 conducted a two factor ANÖVA using splash and site as the factors for day and night data separately, and a single factor ANOVA with degree of light exposure (day or night) as the single factor. In addition to my field observations, I brought granite rocks, along with twenty- five limpets from the intertidal zone into a 250 gallon, outdoor tank. I suspended a spray bar with running seawater over the rocks. The tank could be filled to submerge the rocks. By controlling the extent and duration of water sprayed onto the rocks, along with the degree to which the rocks were submerged, I could simulate artificial tides, consisting of either degree 2 (spray, but no submergence) or degree 3 (spray combined with submergence) splash conditions. Äfter giving the limpets a week to adjust to their surroundings and establish home scars, day and night observations were made on the individual limpets in a similar fashion to those in the field. For each observation, the splash level was recorded, along with which limpets were moving. I calculated movement frequencies for each limpet, and used these to calculate an average movement frequency for each degree of splash, both during the day and at night. I conducted a two- factor ANOVA on the lab data, using degree of splash and degree of light exposure (day or night) as the two factors. To test whether the timing of the natural tidal cycle was a cue the limpets were using to move, 1 exposed field limpets to an artificial high tide at an unnatural time, by spraying them with salt water during low tide. This technique produced a level of wetness equivalent to a degree 2 splash. To determine a movement frequency for each spraying session, 1 simply counted the limpets that were moving after a given period of spray and divided that number by the number I was spraying. A total of fifteen spraying sessions occurred, divided amongst daytime and nighttime hours. The movement frequencies of each session were averaged, and this number was compared to the calculated field movement frequency. Results Frequency of movement ( times limpet movec times limpet was "wet"), pooled over all 2660 day and night field observations was 12%. Frequency of movement, pooled over all 600 day and night lab observations, was 15%. Frequencies of movement did not differ notably from site to site (Fig. 1). It appears that overall differences in wave exposure do not affect limpets' tendency to move. For day and nighttime observations, frequency of movement in the field increased as the degree of splash increased (Fig. 2 and 3). No degree 3 splash occurred at either protected site during the day, so each site, splash combination was analyzed as a separate level of a single factor ANÖVA (Table 3). Although the overall ANOVA was highly significant (PS.001), no unambiguous pattern of significant differences emerged from the post hoc SNK test (Table 3). However, there was a consistent trend of more frequent movement at higher degrees of splash during the day (Fig. 2, Table 3). Maximum daytime movement was 29% at the exposed site with splash level 3 (Fig. 2). At night, movement was significantly greater at each increasing level of splash within each site (with one exception - see Table 4 and Fig. 3). The maximum frequency of movement 1 observed in the entire study was 41% at protected site 2, with splash level 3, at night (Fig. 3). In the lab, frequency of movement showed the opposite trend from the field data, with decreasing movement as the degree of splash increased for both day and nighttime observations (Fig. 4 and 5). The frequency of movement of field limpets was 8% when exposed to an artificial degree 2 splash. Of fifteen spraying sessions, ten resulted in no movement at all (Fig. 6). Frequency of movement in the field was significantly greater at night than during the day, pooled over all levels of splash and sites (Table 6). Although no statistical comparisons could be made, frequency of movement for each degree of splash was consistently higher for the nighttime field observations than the day (Fig. 7). Frequency of movement was also greater at each field site during the night than during the day (Fig, 8). In the lab, limpets also moved more frequently at night than during the day, but these differences were not significant (Table 5). Although frequency of movement did not differ among sites (wave exposures), a combination of splash and time of day appear to be the primary stimuli that induce high¬ intertidal limpets to move. Discussion Based on previous studies, I expected to see -100% movement frequency of Macclintockia scabra at high tide, rather than the observed 12%. Very little literature actually addresses the movement and feeding habits of Macclintockia scabra, but rather focuses on the species' characteristic return to a specified home scar at low tide; perhaps my expectation of 100% movement frequency was unreasonable. However, it is unclear why I saw such a low movement frequency. Enright's (1977) study of vertical migration by zooplankton suggested that under certain conditions of food availability, it might be metabolically advantageous for herbivores to feed intermittently rather than continuously. Little (1989) proposed a similar mechanism for benthic grazers in the rocky intertidal. If algal films are sufficiently scarce in the high-intertidal zone, perhaps it is metabolically beneficial for Macclintockia scabra to forage only when necessary, rather than every tidal cycle. Foraging 12% of the time could be sufficient to meet their energetic requirements. It would be insightful to compare the movement frequencies of the high and mid-intertidal Macclintockia scabra and combine this information with whether there is a significant difference in the amount of micro-algae on the rocks in both regions. The consistency of movement at each field site suggests that location is not one of the major factors affecting the limpets' observed movement frequency. In comparing the movement frequencies for each degree of splash, I found significant differences in the majority of the comparisons, regardless of site (Tables 3 and 4). Though not all differences were significant, the general trend was to see more movement as the splash level increased. The difference in lab results is perplexing (Fig. 4 and 5), as both daytime and nighttime movement frequencies were lower for a splash 3 than a splash 2 high tide. Perhaps lab splash simulations did not mimic sufficiently the field splash 2 and 3 conditions. For example, the lab splash 3 was a gentle, gradual submergence of the limpets and rocks, while a field splash 3 consisted of more wavy, variable conditions. If I could replicate and categorize the lab and field degrees of splash more closely, perhaps I would obtain more consistent results. It would not be valid to conclude that the limpets possess an internal tidal timing factor (endogenous clock?), based on the spraying data alone. A series of more convincing isolation studies would need to occur in order to legitimately approach this subject. However, because the limpets responded to the spraying with an appreciably lower frequency of movement than that seen under their natural conditions, it appears as if they are not simply responding to the degree of splash they are being exposed to, but the timing of the splash as well. This topic bears further investigation. Because the field movement frequency, observed during the day, was significantly less than the nighttime field movement frequency (Table 6), it appears as if limpets use the amount of light they are exposed to as a movement cue. This difference was significant when pooled over all splash level and site combinations. Once again, it is perplexing why there was no significant difference between day and night movement in the lab (Table 5). Perhaps the artificial conditions caused the lab limpets to behave in an unnatural manner. Because no nighttime low high tide and daytime high high tide occurred during my observational period, it is difficult to infer whether the limpets are responding to a reduced amount of light, or an increased amount of splash by moving more at night. IfI were to continue this study, and given the chance to compare the movement frequencies at a nighttime low high tide and a daytime high high tide, I could more readily separate the day v. night and degree of splash effects on the limpets. However, because the limpets did respond with a greater frequency of movement at night than during the day over the given observational period, the possibility remains that degree of light is indeed a movement cue utilized by the limpets. Acknowledgements Thank you to Freya Sommer for assisting me with my lab setup and Chris Patton for helping me with my computer "woes A heart-felt thank you to my advisor, Professor Jim Watanabe, for the time, wisdom and patience he gave to me throughout this project, and for the dedication he devotes to undergraduate education on a daily basis. I am privileged to have had this chance to work with him. Literature Cited Enright, J.T. 1977. Diurnal vertical migration: adaptive significance and timing. Part 1. Selective advantage: a metabolic model. Limnology and Oceanography. 22: 856-872. Hewatt, Willis G. 1940. Observations on the homing limpet, Acmaea scabra Gould. American Midland Naturalist. 24.1: 205-208. Little, Colin. 1989. Factors governing patterns of foraging activity in littoral marine herbivorous molluscs. Journal of Molluscan Studies. 55: 273-284. Sutherland, John P. 1969. Dynamics of high and low populations of the limpet, Acmaea Scabra (Gould). Ecological Monographs. 40.2: 169-188. Wells, Morris M. 1917. The behavior of limpets with particular reference to the homing instinct. Journal of Animal Behavior. 7.6: 387-395. Tables Table 1. The number of limpets observed at each field and lab site. Limpets Site Marked Exposed Protected Protected Lab Table 2. Definition of Splash Categories Degree Wetness of Lab of Splash Splash Condition: Rock none dry inconsistent damp consistent, moderate wet times of excessive submergence — Table 3. Analysis of Variance of Daytime Field Observations Due to missing cells, combinations of site and splash were analyzed as separate levels of a single factor ANOVA. Factors: Site, Splash Transformation: Log (percentage + 1) Weighted Average Sample Size: 27.3 Sum-of-Squares df Source Mean-Square F-ratio Treatment 103.094 17.182 9.310 —0.000 Within 354.338 192 1.846 SNK Test KEY Ex: Exposed Site P1: Protected 1 Site P2: Protected 2 Site 1: Degree 1 Splash : Degree 2 Splash 3: Degree 3 Splash Ex-2 Ex-3 P2-2 P1-2 P2-1 Ex-1 P1-1 Table 4. Analysis of Variance of Nighttime Field Observations Factors: Site, Splash Transformation: Log (percentage + 1) Weighted Average Sample Size: 24.5 Source Sum-of-Squares df Mean-Square F-rat 14.635 7.318 2.691 LASH 130.696 24.034 65.348 TE'SPLASH 24.584 6.146 2.260 Withii 2.719 595.465 219 SNK Test (See Table 3 Key) EXPOSED: PROTECTEDI: PROTECTED2 SPLASH 3: P2 SPLASH 2: SPLASH 1: PI Ex 0.070 0.000 0.064 Table 5. Analysis of Variance of Lab Observations Factors: Degree of Light Exposure (Day or Night), Splash Sum-of-Squares Source Mean-Square F-Ratio DayNight 0.012 0.279 0.012 0.595 Splash 0.595 13.706 DayNight“ 0.134 0.134 3.085 Splash Within 90 3.910 0.043 Table 6. Analysis of Variance of Field Observations Factor: Day or Night Source Sum-of-Squares df Mean-Square F-ratio 36.840 12.212 0.001 LIGHTDARK 36.840 Withir 3.017 1282.085 425 0.599 0.000 0.082 Figure Legend Fig. 1. Movement frequencies calculated at each field and lab site. Error bars are standard errors of the mean. Fig. 2. Daytime movement frequencies, calculated at each field site for each degree of splash. Error bars are standard errors of the mean. Fig. 3. Nighttime movement frequencies, calculated at each field site for each degree of splash. Error bars are standard errors of the mean. Fig. 4. Movement frequencies, based on daytime lab observations, calculated for each degree of splash. Error bars are standard errors of the mean. Fig. 5. Movement frequencies, based on nighttime lab observations, calculated for each degree of splash. Error bars are standard errors of the mean. Fig. 6. Number of instances a given movement frequency was observed, when limpets were exposed to an artificial degree 2 splash high tide, at an unnatural time. Fig. 7. A comparison of daytime and nighttime movement frequencies, calculated at each degree of splash pooled over all field sites. Error bars are standard errors of the mean. Fig. 8. A comparison of daytime and nighttime movement frequencies, calculated at each site, pooled over all levels of splash. Error bars are standard errors of the mean. Figure 1. Figure 2. 0.6 E 0.0 0O T Splash SITE exposed protected1 protected2 Figure 3. Figure 4. 08 0.4 50.2 Splash 0.7 0.4 0.1 O.O Splash SITE exposed protected1 protected2 Figure 5. Figure 6. 86—0 —3 Splash 04 0.5 Movement Frequency Figure 7. Figure 8. 0.2 2 0.1 OO Splash 20. 0.0 exposed lab protectedlprotected2 Site DAYNIGHT day night DAYNIGHT night