Hg 260 ADDITIONAL INFORMATION, IF ANY, CONCERNING AUTHORS, ADDRESS, TITLE, OR CITATION DATA PLEASE TYPE ABSTRACT DOUBLE SPACED BELON CarolMillard Biologisal Ahstract MILLARD, CAROL S. (Hopkins Marine Station of Stanford University, Pacific Grove, Calif., USA.) The clustering behavior of Acmaea digitalis. The Veliger Acmaea digitalis popula- tions often cluster on intertidal rocks at low tide. Clustering occurs in areas which receive some shade and some protection from direct wave action. Animals disperse on a rising tide. Reclustering occurs during tidal ebb while rocks are still wet, and takes place both day and night. A cluster varying in size from 22 to 44 individuals occupied an average of 15 sq. in. (range 8 - 20 sq. in.) each day over a 32 day period. Daily shifts in cluster position occurred; total cumulative area occupied by the cluster in 32 days was 77 sq. in. Only 17 members of the original cluster re¬ mained after 32 days, while a total of 22 new members entered. PLEASE DO NOT IYPE BELOW THIS LINE The Clustering Behavior of Acmaea digitalis (Mollusca: Gastropoda: Prosobranchia) by Carol Spencer Millard Hopkins Marine Station of Stanford University Pacific Grove, California (4 Text figures) The limpet Acmaea digitalis Eschscholtz, 1833, is a common inhabitant of the high intertidal rocky shores of the North American Pacific coast. Populations of A. digitalis are not randomly distributed but are often clustered into groups in limited areas with comparatively few solitary limpets on the surrounding regions. No specific studies of clustering behavior have been made on A. digitalis, although Frank (1966) mentions the occurrence of aggregations in de¬ pressions and other irregularities in the rock. Abe (1932) studied the behavior of clusters of the limpet Acmaea dorsuosa Gould in Japan. Abe's study, however, covered a period of two years and his results showed mainly seasonal variations. *footnote 1. 26 Carol S. Millard The present study, conducted at Hopkins Marine Station on Mussel Point, Pacific Grove, California was carried out to provide a better picture of the clustering behavior of A. digitalis. On April 25, 1966, fourteen clusters were placed under observation. Eleven of these persisted only two to three days and only three clusters remained intact with regard to the area occupied and members present for a period of a month. This paper documents the activity of the largest of the three remaining clusters. For present purposes a cluster is defined as an aggre¬ gation of three or more animals, each within one centimeter of its nearest neighbor. Animals farther than one centi¬ meter from the closest member of a cluster were not included in the cluster. The clusters studied were located on the east face of a large island of granite rock some eighty feet offshore. Waves broke on the opposite side of the island. rolled around and surged below the limpets in a narrow channel, splashing them thoroughly at high water. The par- ticular cluster studied was located six feet above 0.O water level. No macroscopic algae were present in the area and the microscopic algae seemed evenly distributed. The rock sur- face was fairly smooth with a fold or crack rising up the center. The upper end of this fold, where the cluster studied was located, presented a shallow depression on the 2ed Carol S. Millard rock face which became shaded between 1:30 and 3:00 P.M. The location offered some protection from the prevailing winds. To facilitate the plotting of cluster position on the rock a grid of red dots was painted at 1½ inch intervals on the rock surface. Within this grid the positions of indi¬ viduals could be designated with good accuracy. The limpets present on April 25, 1966 were marked with India ink to distinguish them from each other and from the original members. The limpets were observed daily at low water for a week and their positions mapped at each observation. Thereafter, they were observed at intervals of one to seven days. Several attempts were made to chart the move¬ ments of the limpets at rising and falling tides. For these observations pencil notes were made on x-ray film with the emulsion removed and surface roughened, facili¬ tating writing underwater. The water invariably became so rough that plotting the movements of the cluster throughout an entire tidal cycle was impossible. General observation of the cluster showed that the limpets move when they are being wetted by the splash of incoming waves on a rising or falling tide. Within a half hour after they have first been splashed, the majority of 26 Carol S. Millard the limpets begin to move upward and laterally on the rock. In this reaction, corroborated by Miller (1966), the lower limpets begin to move before the higher ones; the low ani¬ mals are more frequently and heavily splashed. Once moye¬ ment has started the animals crawl upwards and laterally for distances of two to six feet. Thus at high water the cluster is dispersed. The animals remain more or less active while the tide is in. As the water level recedes again, they tend to regroup themselves into a cluster, in approximately the position occupied by the previous cluster. The first question posed was: does the cluster occupy the same amount of space and the same area on the rock consistently? To answer this question the daily positions and areas occupied by the cluster at low water were compared for successive days (see Figures 1 and 2). It can be seen that the cluster shifted its position from day to day. The area occupied averaged 15 sq. in. and varied from 8 to 20 sg. in. The cumulative total area of rock surface occupied by the cluster during the month of observation was 77 sg. in. (see Figure 1,u). There was always some overlap between the positions occupied by the cluster on successive days. This overlap ranged from 37% to 75% of the total area occupied at the time of observation (see Table 1). New area, not 276 Carol S. Millard previously inhabited during the period of observation, was added as the cluster changed daily in its shape and location on the rock. Another question to be answered concerns cluster member¬ ship (Figures 3 and 4). The membership of the cluster varied from day to day as animals entered the cluster, remain asso¬ ciated with it for a period, and then either left the area or remained nearby but not actually in the cluster. Later. some of these limpets rejoined the cluster. Thus, it could be observed that the turnover and variation in cluster mem¬ bership was something occurring in a relatively constant group. Only a few unmarked limpets entered the cluster area after the first week, during which eleven newcomers joined the cluster. In the month that followed, fewer than ten un- marked limpets joined. Two animals were observed to wander one to four feet from the cluster area. The closer of the two returned after five days; the other did not return. Observations of Acmaea digitalis in several different locations around the Monterey Peninsula indicate that, in general, clusters are found in more protectéd areas. They tend to occur in spots which receive more shade than the sur¬ rounding rock face due to a slight depression in the rock barely noticeable when the full sun is on the whole rock face, 22 Carol S. Millard On the rock island where most of this study was carried out, between 1:30 and 3:00 P.M. the cluster areas would be out¬ lined by a shade pattern which covered them but left the rest of the rock in full sunlight. Clustering also tends to occur on surfaces at right angles to the sea which receive some surge and splash from waves but do not bear the full brunt of the breakers. In areas directly exposed to wave action Acmaea digitalis have not been observed to cluster, although their populations may be very dense. Here they remain randomly dispersed, although small, temporary clusters of three to six animals may aggregate in a crack or depres¬ sion in the rock. The factors stimulating clustering are not fully known. Animals cluster when the tide is out and disperse when the tide is in, at night as well as during the day, thus light does not seem to be a direct and immediate causal factor. Tide level and wave action appear to play an important role. The limpets recluster when they are still being splashed on an ebbing tide and the area is still totally wet, as though the frequency of splash and the amount of water passing over them were being measured. The fact that the animals recluster when the rock is wet and may do so at night seems to indicate that clustering is not similar 29. Carol S. Millard to the phenomenon of aggregation in sow bugs which keep moving until they end up in the dampest and shadiest spot available (Allee 1926) It is also difficult to look at clustering as a purely random occurrence because it is seen that limpets return to a fairly stable cluster area. Some observations made suggest that the height of the sea at high water and the degree of wave action may influ¬ ence the height on the rocks where clustering occurs, Toward the end of the month that the cluster shown in Figure 1 was observed, the weather grew progressively more stormy, and wave action increased. During this period the cluster rose in its position on the rock (Fig. 1), gradually occupy¬ ing a higher area. Also, the cluster underwent progressive fragmentation with time. This behavior recalls the finding of Abe (1932) that clustering is seasonal, and that the clusters of A. dorsuosa break up shortly after the winter storms begin. The means by which animals are able to find their ways back to the cluster area are not clear. There is a possibil¬ ity that mucous trails are involved. Another animal which occurs slightly above A. digitalis in the intertidal region, Littorina planaxis Philippi, has been proven to follow mucous trails left on rocks by members of its own or other species 2. Carol S. Millard (Miyamoto, 1964; Peters, 1964). Whatever the mechanism, it may be similar to that involved in "homing" in limpets. During the month that the clusters were observed four of the animals were seen to return to the same exact location and orientation at successive low tides for period of two or three days. They then changed resting spots and repeated this behavior for several more days. Homing in some A. digitalis was also noticed by Miller (1966). SUMMARY Acmaea digitalis populations often cluster on inter- tidal rocks at low tide. Clustering occurs in areas which receive some shade and some protection from direct wave action. Animals disperse on a rising tide. Reclustering occurs during tidal ebb while rocks are still wet, and takes place both day and night. A cluster varying in size from 22 to 44 individuals occupied an average of 15 sq. in. (range 8 - 20 sq. in.) each day over a 32 day period. Daily shifts in cluster position occurred; total cumulative area occupied was 77 sq. in. Only 17 members of the original cluster remained after 32 days while a total of 22 new members entered. 2. Carol S. Millard ACKNOWLEDGMENTS I would especially like to thank Dr. Donald P. Abbott of Hopkins Marine Station for his assistance and guidance during this study. I also wish to thank Galen Hilgard and Raymond Markel who helped me obtain equipment, and Tudy Balesteri for keeping an eye on me when field work was necessary at high tide and access to my offshore rock was difficult. My sin¬ cere appreciation goes to Susan E. Fitzpatrick for the let¬ tering on the text figures. This work was made possible by Grant GY 806 from the Undergraduate Research Participation Program of the National Science Foundation. 29 10 Carol S. Millard LITERATURE CITED Abe, Noboru 1932. The colony of the limpet (Acmaea dorsuoso Gould). Tohoku Imperial University Science Reports, 4th Ser. (Biol.) 7:169-187. Allee, Warder Clyde 1926. Studies in animal aggregations: causes and effects of bunching in land isopods. Journ. Exp. Zool. 45:255-77. Frank, Peter W. 1966. The biodemography of an intertidal snail popu¬ lation. Ecology 46(6):831-843. Miller, Alan Charles 1966. Orientation and movement of the limpet Acmaea digitalis (Gastropoda: Prosobranchia) on vertical rock surfaces. Veliger ---- Miyamoto, Alan C. 1964. Clustering in Littorina planaxis. Unpublished student research paper, Hopkins Marine Station of Stanford University, 20 p. Peters, Ronald S. 1964. Function of the cephalic tentacles in Littorina planaxis Philippi. Veliger 7(2):143-148. 2e e 11 Carol S. Millard FIGURE CAPTIONS Figure 1, A - U. Daily changes in size, shape, and number of individuals in a cluster of Acmaea digitalis on a vertical rock face, April 26 to May 27, 1966. Outlined clusters con¬ tain limpets one centimeter or less from nearest neighbors. The number on the cluster shows the number of limpets in it, Day and month are shown in arabic and roman numerals respectively; hours of observation are given in the lower right hand corner of each figure. Lines on the grid were 1.5 inches apart on the rock. The darkened square occupies the same space on the rock in each figure. Figure 1,u shows the cumulative total of all the area occupied by the cluster during the month of the study, Figure 2. Area occupied by the cluster on successive days, based on the data in Figure 1. Line A represents the total area covered by the cluster at any given observation time. Line B represents the area occupied by the cluster during a given observation which was not occupied at the previous observation. Line C represents the overlap of the cluster at any given reading with the area occupied at the previous observation. Line D represents the area which has not been occupied previously by the cluster at any time during the present study. 27 0 Carol S. Millard Figure 3. Turnover rate of limpets in cluster membership between successive observations. Dates and times are the same as those in Figure 1. Figure 4. Composition of the cluster at each observation period. Line A represents the total number of limpets pres¬ ent in the cluster during each observation. Line B represents the number of limpets present which were members of the clus¬ ter when the study began; some individuals left the cluster for a few days and later rejoined. Line C represents limpets present which were not charter members but joined the cluster after the study began. 27 1. Permanent address: FOOTNOTES Carer s. Mrlar 2 14 Carol S. Millard TABLE CAPTIONS Table I. Comparison of area occupied by the cluster with number of animals in the cluster and the percentage of the area occupied which overlaps that occupied the pre¬ vious observation. DATE+ HOUR 26 - IV - 66: 1050 27 - IV - 66: 0930 28 - IV - 66: 0930 29 - IV - 66: 1030 2 - V - 66: 1345 3 -V - 66: 1645 5 -V. - 66: 0700 6 -V - 66: 1530 6 -V - 66: 1100 6 - V - 66: 1500 6 - V - 66: 1850 12 - V - 66: 1200 14 - V - 66: 1500 16 - V. - 66: 1530 17 -V - 66: 1100 17 - V - 66: 1700 19 - V - 66: 0100 19 -V - 66: 0300 26 - V - 66: 1300 27 - V - 66: 1300 -5- TABLE I AREA OCCUPIED BY THE CLUSTER IN SO. INCHES 14.75 16 18.75 20.25 16.75 13.5 19.5 18.7 16.75 16.25 17.25 14.75 15.25 10.25 11.5 8.25 13 15.25 11.25 Carol S. Millard NUMBER OF ANIMALS 40 % OVERLAP WITH PREVIOUS DAY —--- 40 % 7% 57 % 64 % 46 % 75 56 % 43 % 41 % 40 % 2 tag 8 0 2 26w 1050 2 29.V 1030 2 5- 0700 10 4 0 27-V 0930 2 9 02 24 1345 5-V 1530 I 2 C 6 81 6-V uoc 2 4 128-1y 0920 3-V 1645 2 34 A — 004 6 0 141 16- 1630 6V 1850 M 13 75 4 A t 1 4-V 1500 e 6-V 1500 40. 17 S 2.V 1200 9 2 J 71 D a 17V —+ 2 2 D 2 10 19-V 300 4 267 1300 2 L 27V 1234 CUMULA AL 6 O FR-0. a 9 tatataa- 1 e 6. . 6 20 öddeda 29 L Jass Sou o- 1 8 98 8 8 2 Jssa OC o — - e 50. 46 44 42 40 38 39 32 24 54 6 2 2 20 4 ) 9. kkklt MAY APRIL DAS FIG. *.