0 0 Harbor Seal (Phoca vitulina richardi) Haul-Out Impact on and Behavior in the Rocky Midtidal Zone Jean Boal Boal Seal Haul-Out Impact and Behavior Introduction The common harbor seal Phoca vitulina has been studied both physiologically, and behaviorally in reference particu- larly to fishing impact and mother-pup relations (Anderson, 1969: Ridgway, 1972: Wilson, 1974). However, little work has been done on individual seal behavior with reference to haul-out site, and on the impact of seal haul-out on the intertidal forest community. This study on Phoca vitulina richardi, the subspecies found on western North American coasts, documents compositional and morphological changes in algae as well as the differential animal populations found on haul- out sites. A description of observed seal behaviors, par¬ ticularly as they relate to hauling out, is includéd. Seals first began hauling out here at Hopkins Marine Station Biological Preserve, Pacific Grove, California, only about five to ten years ago (Abbott, 1979: Baldridge, 1979). The local population appears to be stable (Harrold, 1977). The area surrounding the point is urban; however, due to the Preserve, island disturbance has been low. Study Sites Three principle and two auxiliary sites were used, all within the Biological Preserve. All sites were chosen for comparability of wave action, tidal height, and suitable rock availability. Boal Seal Haul-Out Impact and Behavior The first site, the southwest portion of a rocky island (Seal Island) about 40 meters east of shore, is the haul-out site of ten to fifty seals every low tide (see Fig. 1). This part of the island is the most sheltered from wave action, and the most heavily used by the seals. The second site was located on the southwest of Bird Island. It differed from Seal Island in having higher resident gull and cormorant populations, slightly greater wave exposure, and less easily accessible deep water suitable for escape (Loughlin, 1974) at low tide. The third study site chosen, a piece of the main land facing Seal Island, was comparable to Seal Island in all but its proximity to human activity and its absence of bird inhabitation. Finally, two auxiliary study sites were used, both on the north side of Bird Island. Wave exposure was high here. One site was accessible from the bay by deep nearby channels at all tides; this area was used for haul-out by from zero to eight seals (mean of three) at low tides. The other site, located adjacently, was only reachable by shallow channels at low tide. It was closer to the main land shore and no seals hauled out at this location. Behavioral observations were done primarily from the roof of three-story building A (see Fig. 1), and also from points B and C. In-water observations came from the coves off beaches M and N. Forty hours of observations over forty-five days were made. Boal Seal Haul-Out Impact and Behavior Effects on Algae In order to demonstrate possible differences in algal species distribution, quadrates were placed in the study sites, and the species under individual points were determined. Quad- rate locations were randomly determined within axes parallel and perpendicular to the coast. Individual points within the quadrate were determined by random coordinates within a grid system. Tidal heights from 15 to 190 cm. above MLL were in¬ cluded. A total of 180 points on 18 different rocks were deter- mined for each of the three principle study sites, as well as 50 points on five rocks in each of the two auxiliary sites. Although no significant differences were found in the three areas due to the wide height range included, possible trends in both Gigarina canaliculata and Rhodoglossum affine were suggested (see Fig. 2). Algal counts from the two auxiliary areas for these two species were compared (see Fig. 3). G. canaliculata was found more frequently on the haul-out sites as expected; however, R. affine was found significantly less frequently. This suggests that R. affine differences may not, in fact, be due to the hauling out. Per cent composition was then determined by ten 20 x 20 quadrates in each principle study area. Rocks heavily used by seals were sampled on Seal Island, and rocks as similar as possible in morphology and tidal height were chosen from Bird Island and the Main Land. In this comparison, G. canaliculata Boal Seal Haul-Out Impact and Behavior and R. affine cover on Seal Island was found to be significantly higher on Seal Island (see Fig. 4). In order to determine whether the observed differences were due to general location differences, rock side algae composition was studied. Four points on the north, the south, the east, and the west sides of five rocks in each major area were counted for algal species with use of a 20 x 20 quadrate. Quadrate loca¬ tion and point determination were done as on the algae rock top study. No significant differences were found between Seal Island and Bird Island or the Main Land (see Fig. 3). Comparisons by height of the four principle alga, Endocladia muricata, G. canaliculata, Gigartina papillata, and R. affine showed range variations only in R. affine, which was found from 50-155 cm. above MLL on Seal Island, and from 45-130 and 80-130 cm. above MLL on the Main Land and Bird Island, respectively. More study is required to determine if the difference is actually attributable to the seals. Preliminary viewing had suggested possible morphological changes in Seal Island G. canaliculata. Five rock tops in each area were sampled for frond height and branching studies. Five fronds from dispersed holdfasts on the rock tops were chosen. Holdfasts were chosen to be representative of G. canaliculata on that rock, but frond selection was blind. In addition, one holdfast determined by coin toss location was picked and brought into the laboratory. Holdfast fronds were spread out and random counts determined two additional fronds which were measured for Boal Seal Haul-Out Impact and Behavior height (see Fig. 5). In the 25 sample study, significantly smaller plants were found on Seal Island. In the 10 frond sample, differences were not significant. The twenty-five fronds per area collected above were meas¬ ured for branching. A frond was considered to start from the holdfast itself and branches over 2 mm. in length were counted. The Seal Island fronds were not significantly less branched, but a trend for furthur study is suggested. Six holdfasts from three rocks in each major area were coded and three people not knowing the coding ranked the samples by color. All Seal Island samples were universally ranked as darker in color than all other samples. Finally, observations had indicated Phyllospadix was torn on Seal Island, so ten 20 x 20 cm. quadrates were placed on rocks predominantly covered by Phyllospadix, and visible roots counted (see Fig. 3). Clearly more were found on the Seal Island rocks than in the other two study areas. Effects on Animals For determination of faunal differences, 20 x 20 quadrates were randomly placed in the study sites and all animals easily visible to the unaided eye were counted. For rock top versus side comparisons, 50 cm. transects lines were run down the north, south, east, and west sides of the rock from the quadrate, and animals within 5 cm. of the lines also counted. Significant differneces in numbers of Littorina scutulata and Tegula fune- Boal Seal Haul-Out Impact and Behavior bralis were found on the rock tops, but not on the rock sides (see Fig. 6). Animals too small to be counted with the unaided eye were counted by bringing 18 random 2 x 2 cm. algae samples from each major site into the laboratory. Sample locations were determined by random coordinates beneath the 40 x 40 cm. quadrates used in algal studies. Barleeia haliotiphila, Lasea cistula, and Tri- colia pulloides were all found in far greater numbers on haul¬ out sites, and amphipods were found significantly less frequently (see Figs. 9 and8). (Amphipods were not counted on the auxil¬ iary sites). Bacterial Differences Bacteria on G. canaliculata was cultured on standard nutrient agar medium (Difco) to determine if different bacterial composi¬ tion might be a factor in the greater small animal populations of Seal Island. A total of fifteen 1.5 cm. high, comparable fronds from five different rocks were taken from each principle study area. One side of each frond was placed in contact with the agar surfact, and then removed. The reverse side of the frond was then touched to the agar surface of a second plate. The plates were left under continuous florescent light for three days, after which bacteria colonies were counted and described (see Table 1). Total bacterial counts were not significantly differ¬ ent between Seal Island and the Main Land. However, if the Boal Seal Haul-Out Impact and Behavior ubiquitous "white blobs" are excluded, the remaining bacterial counts are quite different. Bacteria distribution is likely too highly variable for this small sampling to be conclusive. Discussion That the algal and snail differences of Seal Island were only found on rock tops, and that the differences were also found in the auxiliary sites, clearly indicates that harbor seals have a definate impact on the midtidal biota where they haul out. This impact can be expected to fall into three categories: mechanical, chemical, and the resultant environmental. The action of crawling out and sliding in may cause structural stress to algae, includ¬ ing the tearing noted in Phyllospadix root counts. Less tightly attached animals are probably getting knocked off, as the low L. scutulata and T. funebralis numbers suggest. In addition, 200 - 100 lbs. of seal resting, on average, 10 hours a day (Loughlin, 1974) on the rocks may be related to the smaller G. can¬ aliculata plants, and may be squashing animal life such as amphi¬ pods. Chemically, on two instances feces were found on the rocks, and on one, some liquid discharge was found pooled in a rock crevice. This must also be affecting the plant and smaller animal life, even though the rocks are washed off by each high tide. Less quantifiable stress factors include decreased sunlight Boal Seal Haul-Out Impact and Behavior and warm, humid conditions maintained on the rocks during the time the rocks are normally the most exposed. Qualitative (tactile comparison) observations of rocks just vacated support this stress hypothesis. Possibly, the higher range of R. affine may be due to this seal-incubation phenomenon. The possibility that seals may occationally feed in the mid- tidal area places them in the position of top carnivore of the mid¬ tidal community. A comparison of fish found in seal stomachs (Bigg, 1969 and Newby, 1969-71) and those found in the Preserve intertidal (Nichols, 1979) suggest cabezon and sculpins as pos¬ sible prey. But, although as many as 32 seals have been sighted in the area at high tide, a more usual number would be two to four, and only upon disturbance were more than ten seals seen in the water at a time. If the seals are feeding at this time, this surely would affect the fishes and hence the community as a whole. Thus, it is clear that at and near the harbor seals' hauling out site, the seals have a definite and important impact on the midtidal forest community. This should not be overlooked in future studies of the midtidal region. Seal Behavior The harbor seal population at Seal Island ranged form seven to fifty, with a mean of 31, at low tide during full daylight. Observations were made for a total of about 40 hours in 45 days, Boal 10 Seal Haul-Out Impact and Behavior over a period of six weeks (April to June). Although too few counts were made to show significance, trends show a possibility of more seals hauling out in sun than in rain, and more at nega¬ tive tides than positive low tides (see Table 2). In general, at high tide from one up to six seals cound be seen in the water around and between points A and D (see Fig. 1). These most often were smaller seals. Because the haul-out site was so close to and in full view of the shore, and because of the smallness of the population, indi¬ viduals were able to be distinguished. Eleven different indivi- duals, relatively easy to site, were observed for rock preference and individual behavior. (For aide in future studies a sample score sheet for seal identification has been included. See Table 3). Rock preference was definately shown by the individuals followed. At least five different individuals were found to be returning consistantly to one to six particular rocks (some, more than 95% of observation sightings). With only one exception every identifiable seal seemed to have a preferred social set¬ ting: either in the thick of the crowd (there was generally a crowd of 10 + seals in the most sheltered part of the island for that tide), on the fringes of the crowd, or off by itself. Certain types of rocks appeared to be preferred. They were low, had flat top surfaces, were well sheltered from waves, and were near escape channels. In times of crowding, conflicts errupted over choice sites. Snorting, snarling, and head-thrusts Boal 11 Seal Haul-Out Impact and Behavior (Bishop, 1968, as cited in Loughlin, 1974) were all observed. In addition, the seals displayed a tapping of the front flippers on the other seal. Both seals might continue this tapping at eachother for 5-10 minutes or more. Finally, if one seal did not move over sufficiently, the larger seal would push the smaller seal off the rock. This pushing was observed on at least eight separate occasions, the small seal losing in every case. Con¬ flicts were observed greater than thirty times, and snarling and flipper-taps included in approximately 90% of the displays. Smaller seals were noted for choosing rocks for consistant hauling out that were not preferred sites. These were largely high rocks (bare or partly bare of algae), and in some cases were pointed or steeply sloped. This choice is understandable in light of the observed large seals' dominence of preferred rock¬ types. Using size as an indication of age, the population of Seal Island is young. Large seals accountfor from O to 7 of the hauled out seals sighted, 7 to 20 were quite small relative to these, and remaining 20-, medium sized. Behavior differences were noted between large seals and others. In case of disturbance, the smaller were quicker to look around. The larger seals never tol- erated a close human approach and were never observed to swim up and watch people on the shore or follow the small boats departing from the beaches, as were smaller seals. It is possible that smaller seals are more accustomed to people and avoid them less, which would help account for why this island population, so close to Marine Station activity, is predominantly made up of small Boal 12 Seal Haul-Out Impact and Behavior seals. Play in the water off the beaches was observed about 60% of observation days by at least one and up to nine seals at a time. Water slapping was definately not correlated with warning, but was displayed frequently by single seals. From observation to obser¬ vation time, consistent in-water individual seal identification was not generally possible. However, in any given observation period, seals were observed to slap the water with only the right or the left flipper. This preliminary observation indicates handedness, which although not mentioned in previous harbor seal studies, has been a noted characteristic of other marine mammals. Only one mother and pup were observed in the Preserve this spring. The pup was the first ever sighted here. It may not have been born here as it was only sighted beginning the last week of May, and in the second week of observations, was not observed to be permitted to nurse (although it was still quite small). This lack of pupping here is probably due to the amount of human activ¬ ity nearby. Two seals were recognizable in the water. A medium sized seal and a small seal, they were sighted swimming together on more than twenty different occasions throughout the study. This suggests some lasting pair-bonding may occur, even at non-mating times of the year. Boal 13 Seal Haul-Out Impact and Behavior Summary and Conclusions Seal impact on haul-out sites was evidenced by; 1. Greater G. canaliculata and R. affine cover. 2. Shorter and darker G. canaliculata plants. 3. Plant tearing shown by Phyllospadix root visibility. 4. Increased B. haliotiphila, L. cistula, and T. pulloides populations. 5. Decreased amphipod populations. 6. Fewer L. scutulata and T. funebralis. 7. Possible alteration of bacterial composition. Seal haul-out behavior noted include: 1. Low tide population of 7 to 52 with a mean of 31. 2. Possible preference to haul out under conditions of sun and negative tides. 3. Preferred rocks low, sheltered, near deep water, inacces¬ sible to land animals, and with as little nearby human activity as possible. 4. Particular rock preferences shown by individuals. 5. Rock conflicts characterized by snorting, snarling, head¬ thrusts, flipper tapping, and pushing, with larger seals' success. 6. Sustained pair bonding outside of mating season. Boal 14 Seal Haul-Out Impact and Behavior Acknowledgenents My greatest thanks for the great help provided particularly by Robin Burnett, and also by Alan Baldridge, Donald P. Abbott, and Judson Vandevere. References Abbott, Donald P. 1979. Personal communication. Anderson, Harald T., ed. 1969. The Biology of Marine Animals. Academic Press, N.Y. Baldridge, Alan. 1979. Personal communication. Bishop, Richard H. 1968. Reproduction, age determination, and behavior of the Harbor Seal, Phoca vitulina L., in the Gulf of Alaska. M.S. thesis, Univ. of Alaska, College, Alaska. 121 p. Cooper, John, and Mark Wieland; Anson Hines. 1977. Subtidal Abalone Populations in an Area Inhabited by Sea Otters. The Veliger, 20 (2): 164. Harrold, C. 1977. Data on file at Hopkins Marine Station Library. Loughlin, T. R. 1974. The Distribution and Ecology of the Harbor Seal in Humbolt Bay, California. M.A. thesis, Humbolt State University. Newby, Terrell C. 1969-71. Pacific Harbor Seal. Canada Fisheries Research Board Bulletin, Nos. 168-175. Boal 15 Seal Haul-Out Impact and Behavior Nichols, Andrew. 1979. Predation by Tidepool and Nearshore Fishes and its Impact on the Rocky Intertidal Zone Community. Biology 175H paper on file at Hopkins Marine Station Library. Ridgway, Sam H., ed. 1972. Mammals of the Sea, Biology and Medi¬ cine. Charles C. Thomas, Publisher. Springfield, Il1. Wilson, S. C. 1974. Eliciting Play: A Comparative Study. American Zoologist, 14: 341-370. Boal 16 Seal Haul-Out Impact and Behavior Figures Fig. 1 Map of Hopkins Marine Station Biological Preserve (Cooper, 1977). Principle study sites on Seal Island (SI), Bird Island (BI), and the Main Land (ML), with auxiliary stidy sites on Bird Island - North (BI-N). Seal in-water behavior observed between points (B) and (D). Observations from (A), (B), and (C). Fig. 2 Plant composition counts for 180 random sites at Bird Island, Main Land, and Seal Island study areas. A - G. canaliculata and R. affine counts from Fig. 2. Fig. 3 B - Haul-out and non-haul-out algal counts. C - Algal counts for rock sides. Studentst test done with data transformed to the.2 power. Significance not found. Fig. 4 Per cent cover of heavily used haul-out rocks (SI) and rocks similar in tidal height and morphology (BI and ML) show significant differences. G. canaliculata frond height comparisons. Fig. Fig. 6 G. canaliculata frond branching measured. Root visibility counts as an indication of plant tearing Fig. are higher on Seal Island. A - Fewer snails are found on Seal Island when rock tops Fig. 8 are studied, but no differences found on rock sides. B - Tegula recounts on rock tops confirm absence seen in (A). 0 Boal 17 Seal Haul-Out Impact and Behavior Fig. More B. haliotiphila, L. cistula, T. pulloides and fewer amphipods found on Seal Island rock tops. Fig. 10 Fig. 9 conclusions supported at auxiliary study sites. Table 1 Bacteria from G. canaliculata fronds from the three principle study locations show significant differences, Seal Island from Bird Island, and Seal Island from Main Land, excluding "white blobs"wwhich are ubiquitous. Table 2 Numbers of seals counted under various conditions. No differences significant due to high variability and few observations. Possible trends for future study are suggested, however. Table 3 Seal identification sample data sheet. r De 8 18 35. 9 g pn 0 2 A , R J X Hilden¬ brandia Irradaea Pelvetia Petrocelis Phyllo- spadix Pryanitis Rhodo¬ glossum affine Lepto¬ rhynchos No Algae 4 a 8 NUMBER OF PLANTS Cladophora columbiana Corallines 18 (articulated Corallines (crustose) Endocladia muricata Gastroclonium coulteri Gelidium Gigartina canaliculata Gigartina papillata Green Algae 0 e 19 8 8 8 2 USE BI Zm □SI &a 10 1 NHO 30 20 10 20 10 G. CANALICULATA 26 16 R. AEEINE 21 ROCK TOPS 0 28 ROCK SIDES 20 40 5 20 10 G. CANALICULATA R. AEEINE 24 0 00 o1 15 14 0 BI N= 10 BI +MLX SI M (= 4.8 § =7.96 I = 2.53 P ( .02 S1 21 BI E m LS1 ML X SI ML X SI N = 25 - 6.58 001 - 5.11 .001 4.2 N = 10 I = 1.29 NS = 0.07 NS 4.7 20 BI M 51 7.2 4.0 N = 25 23 5 PHYLLOSPADIX ROOT COUNTS ML BI SI 10 10 10 0.30 2.40 0.20 3.07 0.63 0.67 BIX SI ML X SI I=2.22 I - 2.11 P K.05 P(.05 24 A. N - 18 L SCUTULAIA L FUNEBRALIS LI SCUTULATA Li FUNEBRALIS B. N = 10 L EUNEBRALIS SNAIL COUNTS BL ML 81 80 31 19 345 119 139 83 41 33 SI 118 RXC IOP 6 = 20.12 P (.005 RXC G = 0.1 SIDES NS 25 — R S hoн sооаноNI 1о aa O 250 L o10 210 42 19 .. LR K IRICOLIA BARLEEIA LASEA BIRD ISLAND - NORTH AUXILLARY SITES RXC G = 129.82 P K .0 HAUL-OUT NON-HAUL-OUT 8 27 e WHITE BLOB WHITE WHITE RADIAL YELLOW CLUMPS YELLOW BLOB PINK EGGS CLEAR GRAINY CLEAR EGGS WHITE CLUMP TOTALS EXCLUDING WHITE BLOBS BI 6 0 22 BACTERIA ML 43 2 0 0 78 20 SI 19 60 X(-9 P (.001 28 e SEAL ISLAND TOTAL SUNNY OVERCAST RAINING NEG. LOW TIDES POS, LOW TIDES BIRD ISLA TOTAL SEAL COUNTS Hn M 25 32.5 34.6 11 29.0 25.5 41.5 30.3 18 3.0 11 815 2.02 3.76 5.57 1.50 6.81 10.42 0.54 29