Factors influencing the distribution pattern of different ages of hauled out harbor seals were studied for a 5-week period in Spring 1991. Particular emphasis was placed on the impact of substrate characteristics on the observed age distribution. The seals in this area utilize irregularly-shaped offshore rocks for haul out. Aggressive behavior was almost completely absent in this area and seals were never observed battling over haul out space. Substrate features which were examined included: distance to an escape channel; % total and individual algal cover; size of hauling ground; and degree of incline to reach the hauling ground. These features at least partly resulted in the observed segregation of ages as adults were limited by the size of the hauling ground, while subadults and juveniles were limited by the % algal cover and degree of incline. Social organization may be an additional influence on the age-structured distributions, but the short term of this study prevented any extensive investigation into the social structure of this population. INTRODUCTION: The distribution of the harbor seal, Phoca vitulina, in the eastem Pacific extends from Baja Califomnia, Mexico to the northern coast of Alaska (Knudtson 1977). Throughout its range, the harbor seal hauls out on a variety of substrates which include: cobblestone and sandy beaches, sloping rock outcroppings and rock platforms, emergent offshore tidal rocks, rocky shorelines with large boulders and rubble, sandbars, mudflats, sheltered crevices or sea caves, and tidepools (Riedman 1990). In addition to hauling out in order to give birth and nurse its young, it has been postulated that hauling out is an important aspect of maintaining the harbor seal's daily energy budget as it spends 44-50% of its time hauled out (Riedman 1990, Sullivan 1982, Newby 1973). Several abiotic factors have been recorded as influencing haul out behavior of harbor seals, such as: tidal height (Scheffer and Slipp 1944, Venables and Venables 1955, Bishop 1968. Loughlin 1974, Schneider and Payne 1983, Osbomn 1985); wave intensity (Venables and Venables 1955, Bishop 1968, Schneider and Payne 1983, Slater and Markowitz 1983); disturbance (Scheffer and Slipp 1944, Venables and Venables 1955, Stewart 1984, Osbom 1985); and time of day (Stewart 1984. Osborn 1985, Calambokidis et al 1987). In terms of biotic factors influencing haul out, Bishop (1968) noted that the presence of other hauled out seals is an important stimulus to general haul out. Furthermore, Sullivan (1982) and Osborn (1985) reported that harbor seals have a greater tendency to haul out in areas where other seals are resting than areas which contain no resting seals even though the uninhabited areas are physically suited for haul out. Many past researchers of harbor seal behavior have addressed the issue of social organization among hauled out seals (Scheffer and Slipp 1944, Bishop 1968, Loughlin 1974, Knudtson 1977, Wilson 1978). These researchers either supported or rejected the presence of social organization based on whether individuals formed groups within the hauling ground area. More recently, Sullivan (1982) and Osbom (1985) have commented on social organization by noting that harbor seals of Humboldt Bay, California and Elkhom Slough, California respectively form groups based on age and sex. Sullivan studied the interactive and dominance behavior of the seals and provided a detailed analysis on the effectiveness of various behaviors in obtaining and retaining haul out space. Sullivan and Osborn both concluded that the observed age and sex groupings were the result of a complex hierarchy system headed by adult males. This paper addresses the age distribution of harbor seals hauling out on irregularly shaped offshore rocks and presents data on substrate characteristics which influence the age distribution of seals within the hauling grounds. MATERIALS AND METHODS: This study was conducted at Hopkins Marine Station (HMS), which is located in the city of Pacific Grove at the souther tip of Monterey Bay, California (see Map 1). This study was conducted from 4-22-91 to 5-27-91 for a total of 127 hours of observation. The dimensions of the entire study area are approximately 57 x 29 meters (see Map 2). Due to the close proximity of the study area to shore, observations were made with the naked eye and with the aid of binoculars. By using the natural boundaries formed by the irregular configuration of the rocks, the study area was divided into 21 distinct hauling grounds (see Maps 2-5). Only 2 of these hauling grounds are accessible at high tide, and thus observations were conducted for the full duration of the daytime ebbing tidal cycle. For each seal that hauled out, its age class, sex, time of haul out, interaction with other seals, and duration of haul out were recorded. Furthermore, its initial location and movements were recorded on the appropriate map. Four age classes were noted: adult (4+ years), subadult (2-4 years), juvenile (1-2 years), and pup (El year). Age class was primarily based on size using age-size data compiled by Fisher (1952) and Bishop (1968) (see Fig. 1), although pelt characteristics were used to distinguish some pups and juveniles. Seals were measured by comparing them to known-length areas on the rocks. Sex was indicated by the presence or absence of a penile opening and penile line, or the presence of teats visible on lactating females. However, most hauled out seals were not sexed because they did not position themselves on the rocks with their ventral side towards shore and sex-determination was almost impossible once the pelt dried. Tidal height, wave intensity, and disturbance were recorded each day Tidal height was measured in feet and was given by the tide schedule for Monterey Bay. The maximum and minimum tidal heights during which observations were made were +4.1 feet and +0.2 feet. Seals were not present in the study area at a tidal height less than -O.5 feet because the channels within the study area were too shallow for the seals to swim in them. Wave intensity was measured daily at each of the 21 hauling grounds. Multiple recordings were made throughout the observation period at one hour intervals. It was noted that wave impact remained constant throughout the study period until the tide level dropped below +2 feet. Once the tide height dropped below +2 feet, the wave intensity factor would decrease by one category. Wave intensity never changed by more than one category during any obervation period. There were five levels of wave intensity based on the percent of the hauling ground covered by wave splash: (1) hauling ground unaffected by wave splash; (2) 0-25% covered; (3) 25- 50% covered; (4) 50-75% covered; and (5) 75-100% covered. Because the entire hauling ground was not exposed at higher tide heights, these percentages were based on the area of the hauling ground exposed between wave sets. Only the available haul out area was considered, meaning that the portion of the hauling grounds which was underwater due to the tide was ignored. A "disturbance" was said to have occurred if any seal returned to the water via quick ungulations of its body which have been described by Sullivan (1982) as escape behavior. Time and type of disturbance were recorded as well as the seals' subsequent reaction to the disturbance. Physical characteristics of the rocks which were measured included 1) distance to escape channel; 2) % total algal cover; 3) % algal cover by Mastocarpus papillatus (narrow blade version), Mastocarpus papillatus (broad blade version), Endocladia muricata, and Iridaea flaccida; 4) size of hauling ground; and 5) degree of incline to reach hauling ground. Distance to an escape channel was measured in meters and given by the mean observed distance between a seal and the water's edge. Total and individual algal cover was assessed by using a 50-cm square grid subdivided into 5-cm squares. Relative percentages were recorded for each 50-cm square section within a hauling ground and the mean taken for the overall percent-cover. Due to the irregular shape of most hauling grounds, the size of the hauling ground was based on the number of large adult males which could be accommodated within the area. This number is not the maximum number of adult seals ever observed in the area. It is important to note that this value does not indicate the total number of individuals able to haul out in a given area because the area occupied by one large adult equals that of two subadults or three juveniles based on area comparisons in hauling grounds where all age classes hauled out together. Site 14 was the reference area used for these area calculations because the number of adult males within the site and the amount of space they occupied could be measured and then applied to other hauling grounds. Some of the smaller areas (i.e. sites 3, 6, 7, 13, 17, and 19) consisted of small, single rocks, and in these cases the central width and length were measured in meters and the mean taken if there was more than one rock. Although adults were observed hauling out in these areas on a few occasions, the same unit of measure was not used because: 1) adults very rarely hauled out in these areas and 2) the entire body of the adult did not fit on the rock - both its head and tail hung over the edges. Degree of inclination was measured using a protractor with an attached plumb. The angle of the side of the rock used for hauling out was measured and the mean taken if there were multiple haul out routes to the hauling ground. RESULTS: Factors influencing overall haul out: Because only 9.5% (2 of 21) of the hauling grounds were available at high tide, an inverse relationship was observed between tide height and total number of seals hauled out. Wave action had a varying effect on haul out in terms of the numbers of seals hauled out and their distribution. Waves always approached the study area from the northwest. Due to the position of the hauling sites, sites 1-13 and 19-21 always experienced wave actions of the same intensity factor. Sites 1-13 and 19-21 received direct wave action from the incoming ocean swells, yet these sites provided protection for sites 14-18. A positive relationship was observed between the wave intensity factor for sites 1-13 and 19-21 and the number of seals hauled out in the more protected sites 14-18. A factor of 5 produced an overall average increase of 104% in haul out at sites 14-18; a 4 produced an overall average increase of 64%; a 3 produced an average increase 36%; and factor of 2 had no significant effect on haul out in sites 14-18. These percent-increases are based on the average maximum number of seals hauled out on calm days when the entire study area experienced a wave intensity factor of 1. The maximum number of seals hauled out in sites 14-18 on days of heaviest wave action was still less than the maximum total number of seals that hauled out when all hauling grounds were accessible. Disturbances were recorded on 7 days. Disturbances were the result of kayaks on 3 days, a helicopter 1 day, and human presence on the shore and in the intertidal zone the remaining 3 days. After fleeing the hauling ground, some seals eventually rehauled in the same area while others relocated to a new hauling ground. On six occasions, seals within a hauling ground fled as a group but no apparent cause for flight was observed. However, on each of these occasions, the tide height was less than +1 feet. No rehaul was observed on any of these occasions. A disturbance often affected those hauling grounds nearest the disturbance and on 3 occasions, the disturbance did not result in site-abandonment by all seals. The seals which remained were always adult males and they were always the highest up within the hauling ground. Age-structured distribution: The seals in the study area did exhibit an age structured distribution in some areas, yet a mixing of age classes was the norm Figures 2-5 illustrate for each hauling ground the percent days that each age class was observed hauled out with respect to the total number of days any haul out occurred within the particular site. Upon inspection of these graphs, it is apparent that not all age classes were observed at each hauling ground, nor were they observed hauled out with the same frequency. Table 2 provides a summary of this data. Adults were never observed in sites 7, 17, and 19; subadults never in sites 12 and 14; juveniles never in sites 9, 12, 14, and 16; and mother/pup pairs never in sites 3, 7, 13, 14, 16, and 17. It is important to note that mothers were not counted as adults because it was the pup's ability to remain hauled out which determined where the mother hauled out. Especially in early May, it was not uncommon for a mother and pup to unsuccessfully haul out in several areas before finding a rock that the pup could climb up and not be washed off by a wave. Although most hauling grounds were used by more than one age class, this does not imply that different age classes always hauled out together in these areas. This was significant for sites 3, 7, 9, 15, 16, and 17 where only once did more than one age class use the hauling ground at the same time. All other hauling grounds, except for site 14, were occupied by 2 or more age classes concurrently. When the age classes mixed on the hauling ground, juveniles and subadults were at times observed occupying areas higher up on the rocks than adults. The distribution of the age classes was not constant throughout the daily haul out period, nor was it consistent from day to day. On 10 occasions a juvenile was observed hauling out later than an adult, yet the juvenile hauled out to a higher position on the rocks. Interactions among seals: Interaction among these seals was very limited. Most seals seen eliciting social behavior on the rocks were mother/pup pairs. Intra-pair behavior consisted of naso-naso contact, nursing, and scratching. Most agonistic behavior seen was performed by mother seals protecting their pups from other curious seals. These behaviors included growling and head thrusts with and without biting. The only social interaction observed which did not involve a mother/pup pair was a large adult which pushed a subadult off a rock so that it could haul out onto the rock. There was no fighting among these two seals the bigger one simply pushed the smaller one off into the water. Cuts and blood were often seen on these seals, but these marks were not the result of a aggressive battle on the hauling grounds. Rather, either the seals had the cuts at the time of haul out, indicating they were the result of an in-water confrontation, or the marks were made by the sharp barnacles on the rocks when the seal moved around. Substrate physical features and age class haul out: Adults were able to haul out at almost all hauling grounds unless the rocks were too small to accommodate their larger size. When a juvenile was observed at a higher position in the hauling ground than an adult, the juvenile was always resting on a rock too small for an adult. Adults were seen hauled out twice at sites 2 and 3, once at sites 13 and 21, and never at sites 7, 17, and 19. Table 2 indicates that, except for site 2, the space available at these sites ranged from 0.375 m2 to 1.8 m2. Although site 2 had large enough rocks, it was only exposed at a tidal height of +1.8 feet and less. Juveniles and subadults often arrived at this site before any adults, but on the two occasions that adults were observed at site 2 they were the first to arrive. No relation was observed between degree of incline, percent algal cover, or distance to escape channel and an adult's ability/preference to haul out at a particular site. Degree of incline and the percent algal cover did limit subadult and juvenile haul out. If the total algal cover was greater than 50%, then either 1) the incline had to be less than 20' or 2) % Mastocarpus (narrow blade) cover was less that 20% or 3) % Iridaea cover was less than 10% (see Table 3). Iridaea and Mastocarpus (narrow blade) were the most slippery of the four types of algae surveyed with Iridaea being the most slippery of all. On several days, both subadults and juveniles were seen attempting to haul out on rocks which had an algal cover of 80% or greater and an incline of 50' or greater. However, each of these attempts was unsuccessful because the seals were slipping down on the algae faster than they could climb up it. Although sites 3, 7, and 17 met the above conditions for haul out, subadults were only observed at these sites twice, once, and once respectively. Most often, juveniles arrived at these areas first and occupied all the available hauling area. Subadults were observed gliding by these sites, looking at the seals (juveniles) on the rocks, and proceeding to a different site to haul out. Site 1 did not meet the above conditions for haul out, vet both subadults and juveniles were seen at this hauling ground on four occasions. On each of these days, the seals did not haul out until the area had been exposed to the sun and wind for a minimum of three hours which made the algae less slippery. No relation was apparent between area of hauling ground or distance to escape channel and ability/preference of subadults and juveniles to haul out. Pup haul out was also influenced by degree of incline and percent algal cover. However, even more influential was the presence of other seals. An inverse relation was observed between the presence of juveniles, subadults, or male adults and the likelihood that a mother/pup pair would haul out at the same site as these other individuals. But, if the other seals were another mother/pup pair, haul out was quite likely. Once again, no relation was observed between distance to escape channel and mother/pup preference to haul out. DISCUSSION: Tide height, wave action and disturbance can be said to be the classic factors which have been studied in relation to overall haul out numbers. As with other studies, this study showed an inverse relationship between tide height, wave action, and disturbance with number of seals hauled out. However, with reference to disturbances, seals were observed leaving a hauling ground in mass although no apparent disturbance had occurred to produce such a reaction. Because the tide height was always less than +1 foot when this occurred and these seals always rehauled in a new area which had easy access to deep water, it is possible that the escape channel at the original site had reached a critical minimum depth and this forced the seals to leave. These were the only instances that distance to an escape route appeared to influence haul out, but the impact was on general haul out, not age-specific haul out. Sullivan's work with harbor seals in Humboldt Bay also involved a substrate of irregularly shaped offshore rocks. His project was centered around explaining why there existed a predictable spatial patter of the various age classes within the hauling grounds. The age-structured distribution that he repeatedly observed was a vertical stratification of age classes in which adult males occupied the highest, most protected sites while younger seals were forced to haul out at progressively less favorable areas which were more susceptible to wave splash. The result was a pyramid shaped distribution with adult males at the top. Sullivan's group of seals were very social as they aggressively competed for haul out space on the rocks. Sullivan recorded the frequency and success rate of eight different aggressive behaviors for each age class. In the end, he concluded that the age-structured distributions were the result of different success rates of aggressive behavior for each age class. In other words, the adult males were at the top of the pyramid because they were the most successful in battles for space Likewise, the younger animals which occupied less favorable areas were less and less successful. In this study, vertical stratification topped by adult males was observed in two manners, between sites and within sites. An example of "between-site' vertical stratification is that site 14 (adults only) and site 17 (primarily subadults and juveniles) are in the same proximity with site 14 being higher than 17. Within-site" stratification means that when all age classes were present at one hauling ground, at times adults were the highest and juveniles and pups were the lowest. However, within-site vertical stratification was not stable and changed from day to day as subadults and juveniles were also observed as being the highest age class. Therefore, the spatial age patterns which Sullivan observed were present in the study area but they did not occur with the same predictable regularity. The complex social interactions that Sullivan had observed and concluded were the cause of the age-structured distribution were not seen in this study. One possible explanation for this lack of aggressive behavior is that space is not a limited resource in this area. This is likely since even on the day with the largest total count, there was still space left for haul out but no seals tried to obtain this space. Furthermore, there were several occasions where a seal approached a hauling ground which was occupied by a smaller age class and after surveying this hauling ground, the bigger seal proceeded to haul out at a different location rather than forcing the smaller seal off the rock. Perhaps this was because space was not limited and it was more costly to engage in conflict than to simply find another area to haul out. Another possible reason for limited aggressive behavior involves the social organization of the population. Perhaps the population Sullivan observed had a very high turnover rate and thus the social hierarchy had to be reestablished daily. The seals of this study may have been part of a more stable population where the social organization is more or less permanently established. An observation which gives some weight to this argument is that at site 14 where only large adult males were seen hauled out, not once did another age class even attempt haul out at this site even though the physical features were such that younger age classes should be able to haul out there. Site 14 was the most protected site and there was always space available for more seals to haul out. But, this area may be known within the population as being only for adult males and no other seals want to challenge that fact. Although this study did not attempt to define the social structure of the population, this does not mean that some form of social organization fails to exist, nor does it imply that social organization did not in some way influence the observed age-structured haul out. Since the same age distributions were observed in this study as Sullivan's study, even in the absence of aggressive behavior, this suggests that there is a factor other than aggression which influences the age-structured patterns. Sullivan did not investigate other possibilities for the patterns he observed. This study explored the possibility that the physical features of the hauling grounds influenced the age distribution. Size of hauling ground, degree of incline, and percent algal cover were observed to have different relations with the haul out success of different age classes. Degree incline and percent algal cover had no significant effect on adult haul out. Two possible explanations for this are that 1) adult bodies are long enough to span over the slippery sections of algae, and/or 2 adults possess better coordinated haul out manuevers such that a steep incline does not pose a problem. However, smaller age classes do not possess these assets and thus are hindered by a steep incline and high percentage of algal cover. There is no reason to assume that the physical features examined in this study are the only features which may influence haul out behavior. Those features chosen for any study will be highly dependent on the type of substrate being used for haul out. For example, distance from an escape channel was not shown to have an impact on haul out for any age class in this study. However, other investigators which have studied seal haul out behavior on sandbars and mudflats have repeatedly noted a preference by all age classes for area in close proximity to an escape channel. Scheffer and Slipp (1944) studied seals on mudflats in Washington State and observed that those seals which hauled out did not allow themselves to be farther than 5 feet from the water's edge and would abandon a site if the escape channel depth was below 1.5 feet. Davis and Renouf (1987) studied seals on sandbars at Miquelon and also observed that the seals would only haul out in close proximity to an escape channel. In fact, grouping was thought to be the result of a limited number of sites with suitable escape routes. Opinions on the presence of age structure has also varied with the type of substrate being used. Pitcher and McAllister (1981) and Pitcher and Calkins (1983) studied seals on gravel beaches in the Gulf of Alaska. Neither study observed age-structured grouping. Bishop (1968) studied harbor seal behavior on sandy beaches in the Gulf of Alaska. He reported "strong gregariousness among seals yet saw little social structure as the sexes and ages were often intermixed. He did note that dominance was related to size and that there was social pressure towards a spatially static group. On the other hand, Scheffer and Slipp (1944), on mudflats, concluded that the seals were loosely gregarious. Davis and Renouf (1987), observed a predictable age-structured arrangement on sandbars, yet the seals failed to interact socially to a significant degree. This study does not provide specific data on the relative influence of substrate physical features and social organization on age-structured distributions. Instead, it suggests that haul out behavior may first be guided by the physical requirements of haul out. Then, working within these boundaries, the social organization further dictates where each age class hauls out. This appears to be especially true when the substrate being used for haul out is irregularly shaped rocks, yet substrate physical features may not be as important a factor for other substrates (i.e. beaches, sandbars, mudflats). This study does not discredit the presence nor the importance of social structure within harbor seal populations. Rather, it illustrates the need to consider the physical features of potential hauling grounds before drawing conclusions on the significance of social organization within a population. i e 52 5 . - 2 — — 180 - 160 140 - 120 100 - FIG 1: Standard length in relation to age based on a compilation of data by Fisher (1952) and Bishop (1968). k — Age (years) FIG 2: Percent days that adults hauled out in each site as a function of total number days (N) any haul out occurred in each site. N=23 15 14 15 14 17 10 11 19 14 8 20 15 12 6 20 2 100.0% - 90.0% - 80.0% 70.0% 60.0% 50.0% 40.0% 30.0% 20.0% - 10.0% 0.0% 8 9 10 11 12 14 15 16 17 18 19 20 21 SITE NUMBER FIG 3: Percent days that subadults hauled out in each site as a function of total number days (N) any haul out occurred in each site. N=23 15 14 15 14 17 10 11 19 14 8 20 21 20 20 15 12 6 100.0% 90.0% 80.0% 70.0% 60.0% - 50.0% 40.0% 30.0% 20.0% 10.0% 0.0% 8 9 10 11 12 13 14 15 16 17 18 19 20 21 SITE NUMBER FIG 4: Percent days that juveniles hauled out in each site as a function of total number days (N) any haul out occurred in each site. N=23 15 14 15 14 17 10 11 19 14 8 2 7 20 21 21 20 20 15 12 6 100.0% 90.0% - 80.0% 70.0% 60.0% 50.0% 40.0% 30.0% 20.0% 10.0% 0.0% 8 9 10 11 12 13 14 15 16 17 18 19 20 21 SITE NUMBER FIG 5: Percent days that mother/pup pairs hauled out in each site as a function of total number days (N) any haul out occurred in each site. N= 23 15 14 15 14 17 10 11 19 14 8 7 20 21 21 20 20 15 12 6 100.0% 90.0% 80.0% 70.0% 60.0% 50.0% 40.0% 30.0% - 20.0% 10.0% 0.0% 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 SITE NUMBER TABLE 1: Mean maximum numbers of seals hauled out in sites 14-18 as a function of the wave intensity factor of sites 1-13 and 19-21. % comparisons were based on the total number of 25 seals hauled out when a factor of 1 (no wave action) was present. WAVE INTENSITY OF SITES 1-13, 19-21. BASE VALUE 14 14 11 17 8 8 15 2 3 2 E 16 2 3 3 1 2 4 17 2 17 23 15 18 12 12 11 27 34 41 51 25 TOTALS: 25 TABLE 2: Age classes observed within each hauling ground. (+ present, - absent) JUVENILE MOTHER/PUP ADULT SUBADULT SITE + 2 4 + + + + + + + + + + + + + + + + + + + 10 11 12 13 14 15 16 17 18 19 20 21 TABLE 3: Measurement of specified substrate characteristics for each hauling ground. 88 85 SITE 35 93 30 50 5 15 4 Adults 32 2 15 85 25 65 2 8. 2 Adults 18. 09 12 10 5 85 0 10 O5x15 21 95 —30 10 60 0 1 Adults 0 5 23 55 60 20 —20 0 8 Adults 15 6 015 30 —10 65 2 LOX10 5 15 15 15 0 015 8 10 O5x25 45 1 Adults 30 95 45 10 0 80 45 35 50 0 5 3 Adults. 53 91 10 015 50 10 35 5 2 Adults 15 11 90 UM 05 UM UM 4 Adults. 15 15 90 12 05 UIM UM 17 UM 25 4 Adults 13 05. 90 45 40 15 8 0 O5x20 14 0 0 50 0 0 0 10 12 Adults 15 20 0. 0 0 0. 45 0 3 Adults 50 40 16 L0 10 0 1Adult 0 62 17 30 0 0 20 0.75 80 0 O5x015 18 20 85 20 60 —20 0 8 Adults. 35 99 19 10 10 80 0 10 9 12X 10 20 15 60 90 4 Adults 11 5 5 0 21 2.5 90 UM UM UM 25 L5x12 20 UM = Twas unable to measure the exact algal composition of these rocks because the conditions were never such that it could be safely measured. % total algae and % Iridaea are estimates based on photographs - both are very visible at a distance. ACKNOWLEDGEMENTS: 1 would like to extend my deepest thanks and appreciation to our T. A., Terri Shaw. Her encouragement and willingness to discuss my project at all stages was very beneficial and helped me to remain focused throughout the quarter. I am also very thankful for Terri's frankness in pointing out and discussing potential weakness, not only during the actual study but also during the process of writing this paper. I would also like to thank my fellow classmates Teri Nicholson and Kirsten Hackney for their help in identifying individuals and discussing seal behavior at the neighboring Seal Rock area. Finally, I would like to thank Alan Baldridge for always being interested in my research and for sharing his insight and knowledge of the harbor seals in this particular location. LITERATURE CITED BISHOP, RICHARD H. 1968. Reproduction, Age Determination, and Behavior of the Harbor Seal, Phoca vitulina, in the Gulf of Alaska. M.S. Thesis for University of Alaska. CALAMBÖKIDIS, J. et.al. 1987. Distribution and Haulout Behavior of Harbor Seals in Glacier Bay, Alaska. Canadian Journal of Zoology. 65:1391. 1396. DAVIS, BENJAMIN M. and DEANE RENOUF. 1987. Social Behavior of Harbor Seals, Phoca vitulina, on the Hauling Out Grounds at Miquelon. Canadian Field Naturalist. 101(1):1-5. FISHER, H.D. 1952. The Status of the Harbor Seal in British Columbia with Particular Reference to the Skeena River. Canada Fisheries Research Board Bulletin. 93:1-58. KNUDTON, P.M. 1977. Observations on the Breeding Behavior of the Harbor Seal in Humboldt Bay, California. California Fish and Game. 63:66-70. LOUGHLIN, THOMAS. 1974. The Distribution and Ecology of the Harbor Seal in Humboldt Bay, Califomia. M.S. Thesis for Humboldt State University. 71 pp. NEWBY, T.C. 1973. Observations on the Breeding Behavior of the Harbor Seal in the State of Washington. Journal of Mammology. 54:540-543. ÖSBORN, LESLIE S. 1985. Population Dynamics, Behavior, and Effect of Disturbance on Haulout Patterns of the Harbor Seal, Phoca vitulina richardsi. M.S. Thesis for University of California, Santa Cruz. 74 pp. PITCHER, KENNETH W. and DONALD G. CALKINS. 1979. Biology of the Harbor Seal, Phoca vitulina richardsi, in the Gulf of Alaska in Environmental Assessment of the Alaskan Continental Shelf. 1983. 19: 231-302. PITCHER, KENNETH W. and DENNIS C. MCALLISTER. 1981. Movements and Haulout Behavior of Radio-tagged Harbor Seals, Phoca vitulina.. Canadian Field Naturalist. 95(3):292-297. RENOUF, DEANE (ed). 1991. Behavior of Pinnipeds. Chapman and Hall Ltd. 410 pp. RIEDMAN, MARIANNE. 1990. The Pinnipeds: Seals, Sea Lions, and Walruses. University of California Press. 352 pp. SCHEFFER, VICTOR B. and JOHN W. SLIPP. 1944. The Harbor Seal in Washington State. American Midland Naturalist. 32:373-416. SCHNEIDER, DAVID and P. MICHAEL PAYNE. 1983. Factors Affecting Haulout of Harbor Seals at a Site in Southeast Massachusetts. Journal of Mammology. 64(3):518-520. SLATER, L.M. and HAL MARKOWITZ. 1983. Spring Population Trends in Phoca vitulina richardi in two Central California Coast Areas. California Fish and Game. 69(4):217-226. STEWART, BRENT S. 1984. Diurnal Hauling Patterns of Harbor Seals at San Miguel Island, California. Journal of Wildlife Management. 48(4): 1459- 1461. SULLIVAN, RÖBERT M. 1982. Agonistic Behavior and Dominance Relationships in the Harbor Seal, Phoca vitulina. Journal of Mammology. 63(4):554-569. VENABLES, U.M. and L.S. VENABLES. 1955. Observations on a Breeding Colony of the Seal, Phoca vitulina, in Shetland. Proceedings of the Zoological Society of London. 125:521-532. WILSON, SUSAN 1978. Social Organization and Behavior of Harbor Seals, Phoca vitulina concolor, in Maine. Final Report to the Marine Mammal Commission. Report number MMC-76/10.