The History of Farallon Island Marine Bird Populations, 1854-1972

THE HISTORY OF FARALLON ISLAND MARINE BIRD POPULATIONS, 1854-1972

DAVID G. AINLEY AND T. JAMES LEWIS

Point Reyes Bird Observatory Bolinas, California 94924

The marine bird populations of the Farallon THE FARALLON ISLANDS Islands, California, have long been of great The Farallones are a group of five granitic islands interest to ornithologists. In fact, the first and associated rocks that lie at the edge of the scientific information on the status of these continental shelf off Central California, -due west

populations dates back to a series of publica- of San Francisco (fig.I, 1). This naner and almost all previous reports concern South -Farallon Island tions that, in the l%Os, marked the very be- which, with its accompanying offshore rocks, is about ginnings of modern ornithology in western 44 ha in area and is the largest of the group. The North America: the Reports, of Explorations North Farallon Islands lie about 8 km NW of South and Surveys, to Ascertain the Most Practicable Farallon, and together are probably less than 1 ha in area. Hence, the number of birds they support and Economic Route for a Railroad from the is small. They are practicallv inaccessible to humans Mississippi River to the Pacific Ocean. Since because of their sheer cliffs. The remaining islets those Reports, more than 70 articles and books of the Farallon group are either continually or often have been published that describe the wild- awash by the sea. life, and mainly the birds, of the Farallones. The reasons for all the interest are readily evi- METHODS dent. On this group of rocky islands, situated The methods used to determine present-day popu- just 43 km out to sea from San Francisco, is lation sizes differed according to species. Storm- not just the largest concentration of breeding petrels were the most difficult to estimate because of their nocturnal activity and their habit of nesting marine birds in the United States outside of in cavities within talus slopes. We estimated Alaska and Hawaii, but it is a diverse avifauna the Ashy Storm-Petrel (Qceunodroma homochroa) as well. Represented are 12 of the 16 species breeding population by computing a Lincoln index of marine birds known to breed on the U.S. for birds mist-netted at night; where the population- _ Pacific Coast including some that are adapted (N) = Mnjm, and M is the number originally banded and released. n is the total marked and un- to pelagic, offshore, inshore, and intertidal marked birds caught during the recapture period, ways of life. The overall result of the many and m is the number of that total that were marked reports is a documentation of an avian com- (Mosby 1963:107). A night-vision scope (on loan munity over a time span probably unsurpassed from the U.S. Army) allowed us to determine petrel distributions on the island and to select good mist- in the Western Hemisphere. net locations. Petrel calls played on a tape recorder There has never been a complete review attracted birds to nets from a wide area. Immatures of the history of this community. We now were separated from breeding birds by differences present such a review with a time perspective in incubation patch characteristics, molt, and wing lengths Cunnubl. data). For the Lincoln index in of almost 120 years. That entire period has 1972 M ~was 921 birds, m was 145, and n was 394 been marked by intense disruption of the for the recapture period 15 July to 21 September. populations by human activities on or near During the recapture period, most immature, nonbreeding birds that were marked earlier were not the islands, but in 1972 protection of the recaptured. They apparently did not visit the island breeding grounds became nearly complete. during the late summer. This resulted in an over- For this reason, it now seems especially ap- estimation of the breeding population since about 15% of the birds banded in 1972 (M) were imma- propriate to review the populations during tures. Therefore we lowered the estimate from the those 120 years. In addition to discussing the Lincoln index by 15% to arrive at a less biased figure. effects that human activities have had, we To calculate the size of the breeding population present hypotheses on the probable roles two for the entire island we multiplied the modified estimate by two since the area in which we netted major oceanographic events have played in birds in 1972 represented about half the populations’ shaping the history of the populations, one distribution on the island. Although we trapped event being an extended period of unusual almost the same number of Ashy Storm-Petrels in ocean conditions in an early time and the 1971 (981) in the other half of their distribution, other being the complete disappearance of a our efforts during the recapture period were neither regular nor frequent enough to attempt calculation major fish population (a prey species) in a of an estimate. There was little movement by adults later time. between the 1971 and 1972 capture areas. t43.21 The Condor 76:432446, 1974 HISTORY OF FARALLON ISLAND MARINE BIRDS 433

I I A D-C CORMORANT 0 PEL. CORMORANT 120’ 1160 -Lo I , BRAN&S CORMORANT A WESTERN GULL 0 380 SAN FRANCISCO

MONTEREY BAY 200 KM

36’

POINT CONCEPTION 340. t _----< LOS CORONADOS’

FIGURE 2. A graphic summary of the history of

,SLA DE GUADALUPEf population changes in four marine bird species (non- alcid) on the Farallones. Symbols in parentheses esti- -260 ‘8 mates determined as explained in the methods. Other 124” 120’ 1160 . symbols were placed according to the information I I I presented in the tables. FIGURE 1. The Pacific coast of California and Baja California, showing islands and other geographic names mentioned in the text. The cold California each bird were blocked off visually into groups of 50, Current moves parallel to the coast. At Point Con- using natural reference points whenever possible. ception it begins to swing gradually westward flowing Most counts were made from the land using a along the western shores of the northern Channel spotting scope but a few were made from a boat Islands as it does so. using binoculars. Pigeon Guillemots (Cepphus cokmdm) were counted one evening in May just prior to the onset of egg-laying. At that time of We used the same estimation technique and mist- year both members of each pair sit by their nest net sites for Leachs’ Storm-Petrels (0. leuco~rhoa). entrances in the evenings. Tufted Puffins (Lunclu For the Lincoln index, M was 378, m was 47, and n cirrhatu) were counted as they sat at burrow was 169. We again reduced the result by 15% to entrances in the mornings just prior to egg-laying. compensate for nonbreeding immatures. In the 1971 Rhinoceros Auklets (Cerorhincu monocerutu) were capture area, we caught only 83 birds, indicating that counted in the waters near the island in the evenings. the species is more concentrated in the 1972 capture They tended to be diurnal ( !) in their nest visits, and area. Confirming this conclusion is the fact that few they were counted at that time, too. Cassins’ Auklet Leachs’ Storm-Petrels could be heard calling in the (Ptychorumphus uleuticus) numbers are quoted from 1971 area compared to the 1972 area. To arrive at the reports of recent detailed studies (Sibley and an estimate for the whole island we increased the Robert, unpubl. data; Manuwal 1972). 1972 modified result by 0.25 (roughly 83/378) to Finally, figures 2 and 3 are reconstructions of the include the population in the 1971 capture area. trends in populations as we have best been able to The remaining species of breeding birds were easier interpret them. The early reports only occasionally to estimate. Many of our counts were made through presented numerical records of populations but often binoculars or spotting scopes. Population sizes of presented rankings of species by abundance. Estima- Double-crested and Brandts’ Cormorants (Phalacro- tions of early population levels for the purpose of corur uuritus and P. penicillatus) were determined by these figures and this paper were done with the few counting active nests. This was done on foot in numerical data, the rankings, and the descriptions by September just after the young had fledged. In- Emerson ( 1904) of colony locations in various years. tensive observation of a colony from a blind provided After about 1900, most reports included numerical criteria for judging active nests. Pelagic Cormorants estimates of populations. Figure 2 depicts popula- (P. pelagicus) were counted pair-by-pair on several tion trends for nonalcids and figure 3 does the same evenings in May during peak egg-laying. At that for alcids. time all birds were present at their nests. Black Oystercatchers (Huemutopus bachmuni ) RESULTS were counted on their feeding and breeding territories and when flocked together during storms. The A SUMMARY OF HUMAN ACTIVITY ON Western Gull (Lams occident&) breeding popula- SOUTH FARALLON tion was estimated by dividing the island among Doughty (1971) has provided a useful sum- several observers who then simultaneously made mary of human activities. American and Rus- counts in their areas. This was done on 3 consecutive days in June at peak egg-hatching, with the observers sian sealers occupied South Farallon from taking different areas each day. The results were 1807 until the end of the 1830s. They ex- averaged and corrections were made to allow for terminated the elephant seal (Mi~ounga fluctuation in gull numbers relative to time of day angustirostris), the fur seal ( Arctocephalus (unpubl. data). Common Murres ( UTia uulge) were counted bird- philippi or Callorhinus ursinus), and the sea by-bird in their colonies. Colonies too large to count otter (Enhydra Zutris). The sealers ’ impact 434 DAVID G. AINLEY AND T. JAMES LEWIS

but continued with a crew of four keepers. In December 1972 the aids-to-navigation equipment was automated and the keepers were removed. Since the 1960s helicopters A)-- have been used to transfer Coast Guard personnel or equipment to the island when boat e COMMON MURRE transportation was not possible. Until 1972 0 PIGEON GUILLEMOT A TUFTED PUFF IN these aricraft usually flew at low altitude over A CASSIN’S AUKLET much of the island on each visit. Besides disturbance of nesting by human residents, a new problem beset the avian popu- n lations in the early 1900s: oil pollution. From 1900 into the 1940s oil tankers, on their way into one of the busiest ports in the world, routinely flushed their tanks close to the island before entering San Francisco Bay. Dawson (1911) reported island shores strewn with oiled bird carcasses, and the keepers ’ logs often reported oil slicks with oiled murres, puffins, and cormorants as being common. During this period two major oil spills also occurred in the area. Written accounts of the I I I I I 11 11 ’ 1 ’ 1860 I660 1900 1920 1940 1960 1937 incident, resulting from a wrecked tanker at the mouth of San Francisco Bay, reported FIGURE 3. A graphic summary of the history of population changes. in four alcid species on the thousands of alcids oiled and dead on main- Farallones. Placement of symbols is explained in land beaches. In the last two decades the bird figure 2. Note the discontinuity in the vertical scale populations have escaped major oil pollution to allow for the rather large changes in murre and incidents, but this has sometimes been miracu- auklet numbers. lous as in 1971 (Smail et al. 1972). In 1909, North and Middle Farallon (the on the bird populations is not known, but they large rocks located 8 and 4 km N of South did use birds and eggs for food. In 1848, eggs Farallon) were made a National Wildlife were taken commercially and by 1855 the Refuge; and in 1969 South Farallon was in- “Farallon Egg Company” was founded to cluded. In 1972 the Bureau of Sport Fisheries supply eggs to the growing human popula- and Wildlife contracted the Point Reyes Bird tions of Gold Rush California. The egging was Observatory (PRBO) to be caretaker of the carried out each year from April through June islands. In 1968 PRBO established a year by a dozen persons. In 1881 the company was round research station on South Farallon and disbanded; however, fishermen and lighthouse since then one or two PRBO biologists plus keepers continued the egging into the early assistants have been present at all times. In 1900s in spite of a prohibition instigated by addition, each year scientists from other insti- the A.O.U. Committee on Bird Protection and tutions have conducted their own research. declared in 1896 by the Lighthouse Board. Beginning in 1968, PRBO closed breeding The number of murre eggs taken in 45 years colonies (by fairly successful persuasion) to was well over 14 million. There was no regard the wanderings of resident personnel other for a sustained yield harvest, and no protec- than active researchers. Beginning in 1971, tion was extended to other species. Cor- West End and several other important morants fled their nests ahead of the eggers, breeding areas were closed even to research. leaving their eggs to scavenging gulls. Gull In 1972 PRBO removed the last domestic cat eggs and young were stepped on in an effort (there were five in 1967). In late 1972 a to reduce the gull population and thereby to helicopter flight pattern was worked out with reduce competition for murre eggs. the Coast Guard that succeeded in reducing In 1855 the first lighthouse also began aircraft disturbance almost to zero. Part of operation. Until 1965 it was attended by four the plan restricts helicopter use, except in keepers and their families (15-20 people). extreme circumstances, to the period Septem- The families kept dogs, cats, hogs, and a ber through March when no species are mule. During World War II, about 50 people breeding. were present to operate gun emplacements. In the late 1800s Old World rabbits In 1965 the U.S. Coast Guard removed families (Oryctolagus cuniculus) were introduced to HISTORY OF FARALLON ISLAND MARINE BIRDS 435

TABLE 1. The history of Leachs’ Stan-Petrel. TABLE 3. The history of Double-crested Cormorant.

Date status Source Date status SOUIW

1896 Present Loomis 1896 1859 Present; late egg-laying Heermann 1859 1897 Present Barlow 1897 1862 Present Gruber 1884 1904 Could not find Ray 1904 1886 Present Taylor 1887 1911 Present Dawson 1911 1886 Second most abundant cor- Bryant 1888 1959 100-200 birds Thoresen unpubl. data morant 1972 ca. 1400 birds This study 1896 Not as abundant as ex- Loomis 1896 pected 1903 At least three colonies gone Emerson 1904 1904 Only 20 nests; all on Main- Ray 1904 South Farallon. They compete with the larger top species of burrowing alcids for nesting cavi- 1911 35 pair Dawson 1911 ties; and their grazing results in a depauperate 1959 33 pair on Maintop Thoresen unpubl. data flora. For these reasons extermination mea- 1972 40 pair on Maintop This study sures are now underway, and hopefully will be completed by the end of 1973. House mice (Mus musculus) were introduced at some Both species concentrate their nesting to the unknown time, possibly in the feed for the southeast quarter of the island where there island mule. They have no effect on the are areas of few or no nesting gulls (fig. 4). breeding birds. Luckily, no rats are present. Up until 1965 many Ashy Storm-Petrels nested in stone walls and building foundations in that BIRD POPULATIONS area (E. James-Veitch, unpubl. map in litt.). The numbers of Leachs’ and Ashy Storm- Since then, the Coast Guard destroyed many Petrels have probably changed little over the of these structures. In 1971 and 1972 we re- years, but the difficulty in observing them built a few walls to encourage return of the makes this uncertain (tables 1 and 2). In fact, petrels to these areas. Coulter and Rise- it was not until 1896 that Leachs’ Storm-Petrels brough (1973) detected high pesticide levels were discovered on the Farallones (Loomis and egg-shell thinning in Farallon Ashy Storm- 1896). The petrels ’ pelagic feeding areas, their Petrels, but our breeding biology studies have nocturnal island activities, and the inaccessi- not yet disclosed any adverse effects of the bility of their nests-in narrow crevices and thinning. under rocks deep within stone walls and talus Three species of cormorants breed on the slopes-have undoubtedly been helpful in their island: Double-crested, Brandts,’ and Pelagic. avoidance of humans and domestic animals. Throughout the 1800s and into the early 19OOs, In 1959 Thoresen (unpubl. data) estimated observers found no cormorant eggs until late 3000-4000 Ashy Storm-Petrels and our estimate in 1972 is similarly about 4000 breeding birds. Thoresen estimated that 100-200 TABLE 4. The history of Brandts’ Cormorant. Leachs’ Storm-Petrels were breeding on the island in 1959; and this was our estimate in Date status SOUCtZ 1971. However, in 1972 we revised this esti- 1859 Present; no eggs or chicks in Heermann mate following discovery of a dense concentra- in June 1859 tion of this species high on Lighthouse Hill. 1862 Present Gruber 1884 Most numerous cormorant; Taylor 1887 The revised estimate is 1400 breeding birds. 1886 400 nests in largest colony 1886 Most nmnerous cormorant Bryant 1888 1896 No chicks and few entire Barlow 1897 TABLE 2. The history of Ashy Storm-Petrel. clutches in June 1896 In July: l-2 eggs per nest, Loomis 1896

Date status SOUlXX no chicks 1903 Two largest colonies gone; Emerson 1904 1862 Present Coues 1862 187 nests in largest colony 1886 Rare Taylor 1887 1904 Most numerous cormorant Ray 1904 1886 Seventh in abundance Bryant 1888 1911 Most numerous cormorant; Dawson 1911 1895 More than in 1885 Barlow 1897 600 birds on slope of Main- 1896 Abundant all parts of Loomis 1896 top island 1924 Hundreds vs. thousands of Chaney 1924 1904 Present Ray 1904 gulls 1911 Second in abundance Dawson 1911 1959 18,000-20,000 birds Thoresen un- 1959 3000-4000 birds Thoresen unpubl. data publ. data 1972 22,000 birds; lOOO+ nests in This study 1972 ca. 4000 birds This study two largest colonies 436 DAVID G. AINLEY AND T. JAMES LEWIS

TABLE 5. The history of Pelagic Cormorant. TABLE 6. The history of Black Oystercatcher.

Date status SOUrCe Date status Source

1859 Present; eggs, no chicks in Heermann 1859 1859 One breeding pair in June Heermann 1859 late June 1862 Not reported Gruber 1884 1862 Present Gruber 1884 1903 A few present, not breeding Kaeding 1903 1886 Present; not most numerous Taylor 1887 1944 No longer breeding Grinnell and cormorant Miller 1944 1886 Least numerous cormorant Bryant 1888 1956 Two seen in spring Coeswell and 1892 Least numerous cormorant Blankinship and ~tallcup 1956 Keeler 1892 1956 Eight seen in October Craig and Cogs- 1896 Not as numerous as ex- Loomis 1896 well 1956 pected 1959 16 present May to August Thoresen un- 1897 Least numerous cormorant Barlow 1897 publ. data 1903 At least one colony gone Emerson 1904 1972 52 present, including 20 This study 1904 Largest colony of 20 nests Ray 1904 breeding pair 1911 Present Dawson 1911 1959 700-800 birds Thoresen unpubl. data other cormorants, they were at their lowest 1972 2000 birds; largest colony This study of 100 nests numbers by the end of egging when at most only a few hundred birds were present. They then began to increase in numbers very slowly June. This is quite different from the present. so that by 1959, Thoresen counted 700-800 In 1971 and 1972 the Double-crested began birds. Since then, they have increased further laying in March, the Brandts’ in late April, and to 1000 individuals. In some areas they are the Pelagic early May. Late laying in the past still increasing. probably resulted from the intense distur- The Black Oystercatcher has had an in- bance. Cormorants are the most timid birds teresting history on the island (table 6). It on the island, and will flee their nests if a disappeared in the 1860s probably as a result human shows himself within 50-100 m. of too much disturbance from humans and During the egging years, human activity prob- domestic animals. From our experience, eggs ably kept them from their nests until after and newly hatched chicks are easy prey for the harvest ended. They may have been either gulls if adults are kept long from their nests. prevented from breeding early or the gulls Other than a few birds present in 1903, but may have eaten the early eggs. Early ob- not included in a list of breeding species, they servers also reported unusually small clutches; were next reported in 1956. The observers at perhaps as an effect of late breeding. Another that time (Cogswell and Stallcup 1956; Craig expected effect of late breeding would be and Cogswell 1956) apparently did not realize greater mortality of chicks and fledglings. In the significance of their records. The species any case, populations declined (fig. 2). increased to 16 birds by 1959, and they now Double-crested Cormorants once numbered number 52 with 20 breeding pairs. Unless in the thousands and were the second most present feeding territories can be further com- abundant cormorant species (table 3). Emer- pressed, the breeding population has perhaps son (1904) noted the disappearance of at reached maximum. least three colonies since his 1887 visit. Ray Western Gulls had also declined to their (1904) estimated a total of 70 birds all in lowest numbers at the turn of the century one colony on Maintop. The size and loca- (table 7). This was probably the result of tion of that colony has remained the same; and disturbance from humans and domestic ani- Western Gulls now nest sparsely in the areas mals but more so from the extermination mea- where other colonies once existed. Brandts’ sures of the eggers. In 1904 Emerson reported Cormorants have always been the most com- their disappearance from five areas since his mon cormorant, but they too experienced a 1887 visit. In 1903 they nested mostly on in- great population decrease (table 4). By the accessible cliff ledges, not the optimal habitat turn of the century, there were only a few for this species. In 1911 Dawson estimated thousand birds, and their largest colony was 6000 gulls present, which may have been an 187 nests. With the end of egging they began increase from a few years earlier. Since then, to recover, and by 1959 we estimated 22,000 their numbers have increased further. By breeding birds were present in several colonies 1959, 23,000 breeding gulls were present and each having over 1000 nests. Finally, Pelagic the nmnber of breeding birds has remained Cormorants were the least abundant cormo- at that level. They are perhaps more nu- rants prior to the egging (table 5). Like the merous now than ever before due to increased HISTORY OF FARALLON ISLAND MARINE BIRDS 437

TABLE 7. The history of Western Gull. TABLE 9. The history of Tufted Puffin.

Date stntus SOUlYX Date status SOUIYX

1859 Present Heermann 1859 1856 Numerous Hutchings 1856 1862 Second most numerous spe- Gruber 1884 1859 Present Heermann 1859 cies 1862 Present Gruber 1884 1886 Second most numerous Taylor 1887 1886 Rather common Taylor 1887 1886 Nest in colonies of 10 pair Bryant 1888 1886 Fifth most numerous spe- Bryant 1888 1895 Breed on flat at West End Barlow 1897 cies; 50 easily accessible 1896 Downy young and well-in- Loomis 1896 nests cubated eggs 8-16 July 1895 A veritable rookery on West Barlow 1897 1903 Gone from five colonies Emerson 1904 End 1904 Breed on West End and a Ray 1904 1896 Second most numerous di- Loomis 1896 few other places urnal bird 1911 6000 birds Dawson 1911 1903 Status unchanged from Emerson 1904 1959 22,000-23,000 birds Thoresen un- 1887; many West End publ. data and main island 1972 23,000 birds This study 1904 Many West End and main Ray 1904 _ island; saw 43 at one time 1911 Several thousand Dawson 1911 survival in the winter resulting from new food 1933 300 birds Smith 1934 sources in the form of garbage and fish offal, 1959 26 birds Thoresen un- and due to the availability of nesting habitat publ. data once occupied by fur seals. At present there 1972 50-60 birds entire island This study seems to be a growing surplus population, now about 1000 adult plumaged birds, that form several “clubs.” able increase in their numbers has resulted, In the 1850s Common Murres numbered with 20,500 birds currently present, and they about 400,000 birds (fig. 3; table 8). This continue to increase rapidly. Gress et al. estimate is based on the number of eggs (1971) discovered shell-thinning in murre taken then in a single year (Hutchings 1856), eggs on the Farallones, but our studies of their and the fact that a female can lay only two breeding biology have not yet revealed detri- (often fewer) replacement clutches (Tuck mental effects. 1960). By the end of egging just after 1900, In contrast to the murres, Tufted Puffins they had been reduced to 60,000 birds, and apparently declined little if any during the with the advent of oil pollution, they began egging years (table 9). They were protected to decline further. In 1959 there were only by their deep nest cavities and a reputation 6000 murres. Since then, they have escaped for giving painful bites. With the arrival of major oil slicks and have enjoyed increasing oil pollution, however, their numbers dropped protection from human disturbance. A notice- suddenly from an estimated several thousand in 1911 to 300 by 1933. In the last 15 years, at least, their numbers have been stable at TABLE 8. This history of Common Murre. 50-60 birds. The Pigeon Guillemot, another burrow-nesting alcid, also was apparently Date status SOUXW unaffected by the earlier disturbance and 1854 ca. 400,000 birds Nordhoff 1874 began to decline only after the coming of oil 1856 Myriads; on every peak Hutchings 1856 1859 Countless numbers Heermann 1859 1862 Most numerous species Gruber 1884 TABLE 10. The history of Pigeon Guillemot. 1886 Most numerous species Taylor 1887 1886 Outnumbers all others com- Bryant 1888 Date status SOUUX bined 1895 Most numerous species Barlow 1897 1859 Present Heermann 1859 1896 60,600 birds; only in inac- Loomis 1896 1862 Present Gruber 1884 cessible places 1886 Present Bryant 1888 1903 Gone from some areas; re- Emerson 1904 1895 Present Barlow 1897 duced elsewhere 1896 Common Loomis 1896 1904 Outnumbers all others com- Ray 1904 1903 Gone from one area; same Emerson 1904 bined elsewhere as in 1887 1911 20,000 birds Dawson 1924 1904 Present Ray 1904 1923 Counted by tens vs. gulls Chaney 1924 1911 about 200 birds Dawson 1911 by thousands 1959 750-1000 birds Thoresen un- 1959 6000-7000 birds Thoresen unpubl. data publ. data 1972 2000 birds; nest in same This study 1972 20,500 birds This study areas as in 1903 43% DAVID G. AINLEY AND T. JAMES LEWIS

TABLE 11. The history of Rhinoceros Auklet. TABLE 12. The history of Cassins’ Auklet. _ Date status Source Date status Source

1859 Present and breeding Heermann 1859 1859 Present in winter but not Heermann 1859 1862 Breeding; rarest species Gruber 1884 summer 1927 No longer breeding; disap- Grinnell 1926 1862 Rare; present in winter Gruber 1884 peared in 1860s 1886 Present and breeding Taylor 1887 1971 Four to five birds present This study 1895 Common; on all parts of the Barlow 1897 spring and summer; pos- island sible breeder 1896 Abundant Loomis 1896 1972 Two to three pair; prob- This study 1903 Same as in 1887 Emerson 1904 able breeder 1904 Abundant; thousands Ray 1904 1911 Most abundant species; Dawson 1911 lOO,OOO-200,000 birds pollution (table 10). Emerson (1904) noted 1971 105,000-120,000 birds Manuwal 1972 Sibley and Rob- the disappearance of only one small colony ert unpubl. since his 1887 visit and added that the status data of this bird had changed little. Pigeon Guil- lemots were common and, as mapped, nested in the same areas where they nest today (see it can prevent sleep. The birds occupy their below). By 1911, however, Dawson counted burrows throughout the year. They are now only 200 birds, indicating that a sudden de- so ubiquitous that their history as reported is cline in numbers had occurred during the almost unbelievable. In the 1850s and 1860s first decade of the 1900s. After the oil pollu- they were rare and present only in the winter tion lessened, they began to increase. By according to the same persons who reported 1959 the population had reached 1000 birds. nocturnal Rhinoceros Auklets and storm- Since then, they have increased further to petrels. By the 1880s they were common and number about 2000 birds at present, and they breeding, and by the 1890s they were abun- again nest in the areas described by Emerson dant. In 1911 Dawson estimated lOO,OOO- for 1887 and 1903. Their population is prob- 200,000 birds which includes the range of ably close to maximum since most suitable present estimates. At the least, a great change crevices are now occupied except in the im- in numbers occurred during the 1800s but mediate vicinity of the human quarters. The its exact extent is not known. The probable population of rabbits introduced in the late cause for the change is discussed later. 1800s will hopefully soon be removed with one intended result being to release a few CHANGES IN NESTING DISTRIBUTIONS more nesting cavities for guillemots and puf- Coincident with changes in population sizes fins. in the past 90 years, changes have occurred in A fourth alcid, the Rhinoceros Auklet, dis- colony distributions on South Farallon. Emer- appeared in the early 1860s (table 11). Only son (1904) published a map comparing dis- a few pairs were present then, and after tributions in 1887 with those in 1903. He several birds were collected over 3 years, few mainly noted many disappearances of whole if any remained (Grinnell 1926). In 1971 colonies and reductions in others. The map these auklets were present and if they did not later published by Bowman (1961) showed breed, they at least scouted the island a great that many of the lost colonies had returned by deal. In 1972 at least two burrows and pos- 1958. In 1972 we again mapped breeding sibly a third were used by pairs daily from colonies, and shown in figure 4 are the nesting h4ay through July. They entered and departed distributions for three species compared to from their burrows, to our surprise, during 1958. all daylight hours. We did not check for Common Murres have been displacing gulls eggs for fear of disturbing thousands of murres and Brandts’ Cormorants, and in their turn and cormorants nesting near by. Brandts’ Cormorants have been displacing The last species to be discussed, Cassins’ gulls. Murres have displaced the much larger Auklet, is easily the most abundant breeding species through (1) their rapid expansion in species with present day estimates (table 12) numbers; (2) the expansion of their dense ranging from 105,000 (Manuwal 1972) to 120,000 birds (Sibley and Robert, unpubl. colonies outward from the centers (they nest data). Many extensive areas of the island shoulder-to-shoulder) ; and (3) their arrival surface are densely pocketed with their bur- to claim territories a full month before the rows. At night the noise of the combined gulls and 5 months before the cormorants singing of the total population is so loud that (unpubl. data). The cormorants displace gulls HISTORY OF FARALLON ISLAND MARINE BIRDS 439

BREEDING COLONIES

NEW SINCE 1958

Brandt’s Cormorant tz%%i pJgJ Western Gull

Common Murre

NO CHANGE 1958 TO 1972

Brandt’s Cormorant E3 m Western Gull

Common Murre

FIGURE 4. The changes in nesting distributions between I958 and 1972 for three species of marine birds on South Farallon. The distributions for 1958 were drawn after Bowman ( 1961). Also shown are the areas on the island mentioned in the text and in the tables. through ( 1) their denser nesting colonies (i.e., DISCUSSION cormorants are spaced no more than pecking During the last 120 years, the populations of distance apart but gulls are usually 2-3 m five marine bird species on the Farallones apart); and (2) their sheer size advantage: have changed in the same way: decline in many gulIs nesting adjacent to cormorant numbers during the last half of the 1800s or colonies stand aside squawking as they lose during the first decades of the 1900s and then their nests to cormorants searching for nest recovery or approach toward recovery in more material. Earlier in the season, cormorants recent years (figs. 2 and 3). In extreme cases bodily displace gulls that attempt to claim of this pattern, two other species completely territories in cormorant nesting areas. In con- disappeared and later returned. The factors trast to these trends, murres have not displaced shaping these patterns have been evident: the smaller Pelagic Cormorants, the other spe- prolonged low reproductive success due to cies requiring cliff ledges for nest sites. Pe- disruption of nesting by human activities or lagic Cormorants are resident the year round and, on cliffs where they and murres nest, they prolonged high mortality from oil pollution have already occupied their sites when murres as the causes of the declines; and then signifi- arrive. The displaced gulls have been re-estab- cant lessening of these stresses allowing for lishing nests in areas close to human dwellings, recovery. Not including the storm petrels and are the only birds willing to do this. Some whose populations have changed little, three have been forced to nest once again on cliff species have exhibited patterns that differ ledges. from the one just summarized. In one case, 440 DAVID G. AINLEY AND T. JAMES LEWIS

Cassins’ Auklets increased rapidly in number waters for the unusually extended period of during the late 1800s and have since remained 1853, or earlier, to the 1870s. Exactly when in high numbers (fig. 3); and in the other the period began is not known because few case, Double-crested Cormorant (fig. 2) and data exist prior to 1853, when the Pacific Rail- Tufted Puffin (fig. 3) populations declined road Surveys began. Hubbs ’ data consisted of due to disturbance and oil pollution, but un- temperature records and records of warm- like the other species, they have not recovered water animals much farther north than re- nor have they even begun to do so. Rather, corded since. Many of these organisms were they have remained stable at low numbers for notably sessile and thus required an extended many years. In the following two sections we period of warm water to move north. The present hypotheses to explain the last two existence of the warm-water period has since patterns, and in the concluding section we been confirmed by Robinson (1965). comment on the pattern exhibited by the The occurrence of warm waters as far north storm petrels. as Monterey Bay (36.7”N) was strikingly apparent during the mid-1800s but strong in- CASSIN’S AUKLET POPULATION CHANGES trusion to the San Francisco-Farallones area The most plausible explanation for the in- (37.7”N; only 150 km farther) was not so creased auklet population in the mid-1800s obvious because data both for and against relates the change to an extended period of this was in evidence. Since Hubbs ’ paper ap- more than two decades when warm tropical peared, however, additional fauna1 records waters moved unusually far north along the have been discovered to indicate that warm California coast. However, some discussion waters did affect the ocean as far as 37.7”N. of warm-water periods and the oceanographic First, Orr (1963) reported the skull of a tropi- conditions of central California waters is re- cal dolphin (Tursiops @ii) from San Fran- quired before a link to auklet population cisco Bay, and he estimated that the individual changes is suggested. probably died 50-100 years earlier. This The annual pattern of currents off Cali- probably places it we11 within the mid-1800s fornia is reviewed by Sverdrup et al. (1942). warm-water period and suggests the presence In brief, the cold, nutrient-rich waters of the of warm water in that area. The only other California Current move south along the coast verifiable record of that species north of the in a sluggish flow 560 km wide. The current San Diego area is one for Monterey Bay in displaces the warm, subtropical, nutrient-de- 1873, during the warm-water period (Banks pleted oceanic waters that would otherwise be and Brownell 1969). Second, the skull of a present. The colder waters along the coast frigatebird (Fregata magnificans) was found usually reach a little beyond Point Conception on the Farallones in 1861 (Bryant 1888). Nor- (34”N), south of which they swing westward mally restricted to tropical waters, this species and the warming effects of the tropical waters has been recorded to 37.7”N or beyond on become increasingly stronger (see fig. 1 for the West Coast in only 3 other years: 1888 geographic locations). During the spring and and 1905 (Grinnell and Miller 1944), both early summer, strong prevailing northwest warm-water years (Robinson 1965), and in winds help to move the current south at full 1972 (DeSante et al. 1972), another warm- strength, and at the same time produce signifi- water year (see below). Usually, this species cant upwelling of cold, nutrient-rich waters in reaches only as far north as the San Diego area, several places along the edge of the continent. and thus the 1861 frigatebird record for the The cold temperatures of the current are thus FaraIIones indicates an unusually strong north- maintained. In the late summer and fall, the ward movement of warm water. Third, in northwest winds cease or subside, causing a 1859 during the warm-water period, the reduction in the strength of the current and Pelagic Red Crab (Pkuroncodes planipes), a the upwelling. As a result, warm oceanic species normally found south of 28”N (central water moves coastward and a warm, north- Baja California; Boyd 1967), was reported, ward-flowing counter-current appears. From and first described, in Monterey Bay. This time to time since the early 1800s (and earlier: organism was next recorded there during the Hubbs 1948), circulation in the California years 1957-60 (Glynn 1961) when warm Current has become altered with the result waters moved well north of the Farallones that especially warm waters moved unusually (Radovich 1961); and its third and latest far north (to 37”N and beyond) for extended occurrence in Monterey Bay was during the periods, i.e., Ionger than just the fall. winter of 1972-73 (Baldridge, pers. comm.), Hubbs (1948) presented convincing data when warm water again affected the ocean as for an intensive northward intrusion of warm far as 37.7”N (see below). Longhurst ( 1967) HISTORY OF FARALLON ISLAND MARINE BIRDS 441

related that Pburoncodes may not reach ma1 production of young auklets should have Monterey Bay with the warm waters per se resulted in a reduced population. The auklets but through the unusual current conditions began to increase in the 1870s the time when that also bring significant intrusion of warm the cold waters returned to the region, and water. Finally, as referred to above, 1972 was by similar timing of these two events, this indeed a year of significant warm waters in further links low auklet numbers in the mid- California. Dozens of warm-water fishes- 1800s to warm-water conditions. The increase especially scombrids and jacks-showed up in in auklet numbers after the 1870s may have southern California during 1972 and 1973, but been a return to population levels existing it is not known how much farther north many prior to the warm-water period. This Banks actually moved (Fitch, pers. comm.). Some, and Brownell (1969) hypothesized for changes to be sure, did move at least as far as Point in dolphin distributions in the same ocean Reyes (35 km N of the Farallones), Pacific region during the return of the colder waters. Bonito (Sarda chiliensis) being one example They also suggested that the changes resulted (Whitt and Ainley, pers. catch). Also indi- from alterations in the distribution of the cating warm-water conditions in 1972 to be dolphins ’ prey (fish). However, for Cassins’ of strength similar to the mid-1800s were the Auklets a change in availability of food re- pelagic crab and frigatebird records men- sources seems more likely since they breed tioned above. on islands in warmer waters off Baja Cali- How do similarities among three warm- fornia (AOU 1957) and thus suitable prey water periods in central California relate to species must be present in warmer waters. Farallon Cassins’ Auklet populations? During How other Farallon marine birds were af- the two most recent warm-water periods, fected by the warm waters of the 1800s is auklet breeding productivity on the Farallones not known. During the warm-water years of was lower than during years of colder sea 1972 and 1973, most species were affected in temperatures. In 1959, the third year of a various ways, generally to result in reduced warm-water period, Thoresen (1964) recorded productivity. Some were able to adapt some- 0.27 chicks fledged per breeding pair (75 what by altering their patterns of resource nests studied); and in 1972, another warm- exploitation, but that is an involved subject water year, we recorded 0.58 chicks per pair to be discussed in a future report. At the (69 nests). In contrast, during 2 years of least, the added environmental stress of warm- colder sea temperatures, 1970 and 1971, Manu- water conditions in the mid-1800s may have wal ( 1972) recorded 0.71 and 0.62 chicks per rendered the birds particularly susceptible to breeding pair (234 and 201 nests). The dif- the disturbance occurring at that time. ferences are significant (Sokal and Rohlf 1969: Another explanation for the auklet popula- 608) between breeding success in 1959 and tion increase relates the change to increased each of the other years (P < O.OOl), and be- food availability resulting from the decrease tween 1972 and 1970 (P < 0.05); but are not in other Farallon wildlife. This can be ruled significant between 1972 and 1971 (P > 0.05) out since no other Farallon animals compete nor between 1970 and 1971 (P = 0.05). with the auklets for food to an appreciable The observed decrease in auklet productiv- degree (unpubl. data). Increased availability ity could result from reduced food resources of nest sites due to a decrease in other species during warm-water conditions. Auklets of the can also be ruled out, because abundant rock Farallones feed on zooplankton, and mostly cavities exist that larger cavity nesters (guil- euphausiids ( Manuwal 1972). Warm tropical lemots and puffins) cannot use, and in any waters in the southern part of the California case, the auklets became abundant 20 years Current have 100 times less the biomass of before potential nest-site competitors began zooplankton, and particularly euphausiids, to decline in number. than do the subarctic waters of the northern part (Aron 1960). This gradient occurs from RECOVERY FAILURE IN DOUBLE-CRESTED Vancouver (48”N) to San Diego (32”N) as CORMORANTS AND TUFTED PUFFINS the influence of tropical waters increases. Contrasting with most other Farallon marine Additionally, the reduction in northwest winds bird populations, Double-crested Cormorants during warm-water years results in reduced and Tufted Puffins have failed to recover from upwelling of cold, nutrient-rich waters and a decimation. The pattern exhibited by these further decline in productivity (Sverdrup et al. two species (i.e., decline in numbers and then 1942). Thus an extended period of warm failure to repopulate) is not confined to the water and decreased food resources in the Farallon populations, however, and in addi- mid-1800s with resultant prolonged subopti- tion, it is exhibited by a marine mammal of 442 DAVID G. AINLEY AND T. JAMES LEWIS

California. Furthermore, an explanation of 1963). The Double-crested Cormorant breeds this pattern must account for the opposite well into the north (northwest Alaska), but it trend in populations of some ecologically simi- also breeds farther south than these other lar species. An hypothesis we wish to present species (AOU 1957). relates the patterns of change in these Cali- In contrast to the above three species, in fornia populations of several marine birds and the same period and region, two other pin- mammals to the demise of the Pacific Sardine niped species and a bird have been increasing (Sardinops caerulea) in the California Cur- as recoveries from former decimations. The rent. The loss of the immense sardine popu- California Sea Lion (Zulophus californianus) lations was a major event in the offshore breeding population of the Channel Islands waters of California and Baja California. Such increased logarithmically from 700 animals in a loss should have had much broader implica- the 1930s to almost 20,000 by the late 1950s tions than simply the disappearance of the (Bartholomew and Boolootian 1960) and have largest human fishery in the Western Hemi- since stabilized at the last figure (Ode11 1971; sphere from the 1920s through the 1940s. R. DeLong, pers. comm.). These islands (San First, the Double-crested Cormorant col- Miguel) are the northernmost breeding area onies south of the Farallones on the Channel of this species. Similarly, the Northern Ele- Islands and on the islands off the west coast phant Seal (Mirounga ungustirostris) has of northern Baja California have all declined been re-establishing its former breeding col- since the 1920s and 1930s (Gress et al. 1973). onies since the 1940s. The island-by-island On Los Coronados, Huey (in Gress et al. 1973) recolonization has moved northward from a attributed population reductions during that remnant population of Isla de Guadalupe to time to activities of tourists; but no explana- the Farallones (Le Boeuf et al. 1974). The tion has been offered for the decline in the total population is now over 30,000 animals immense populations farther south on Isla San (Peterson and Le Boeuf 1969). Finally, the Martin, and so far as known, yet farther south Rhinoceros Auklet, which recently re-estab- no apparent changes have occurred. Since lished itself on the Farallones, has in the last 1945, populations on Los Coronados and the 15 years established populations on other Channel Islands have been affected by DDT islands from British Columbia down to north- contamination (Gress et al. 1973), but not SO ern California (Scott et al., unpubl. data). on the Farallones (PRBO unpubl.) nor on San This species does not breed as far north as Martin (Gress et al. 1973). The overall pat- the puffin and cormorant (AOU 1957; Udvardy tern, then, for this species from the Farallones 1963), but it is increasing where these others to Isla San Martin (northern Baja) is one of are not. initial decimation, due in most cases to human In summary, since the period 1920-40 in the activity, and then failure to recover. region from central California to northern Second, the Tufted Puffin, a species not Baja California, the Double-crested Cormo- affected much by direct human disturbance, rant, Tufted Puffin, and Stellers’ Sea Lion was a common resident during the early 1900s populations have declined or have failed to in the northern Channel Islands (Wheelock recover from former decimation, while the 1904; Howell 1917). At least one of these Rhinoceros Auklet, California Sea Lion, and islands is now barren of breeding puffins Elephant Seal populations have increased or (Banks 1966)) and if they still breed on the recovered rapidly. These changes, with the others, then their numbers are few (Small possible exception of the cormorant, suggest 1960). Tufted Puffins have also disappeared a long-term amelioration in marine climate from Point Reyes ( Ainley, pers. obs.). Thus, along the Pacific Coast. However, data on the pattern shown by the Farallon popula- marine climate do not support this (Hubbs tion is the same in the entire southern part of 1960; Robinson 1965), and neither does the the species ’ range. recent colonization of San Miguel Island by a Third, in the northern Channel Islands the fur seal (Callorhinus ursinus) otherwise re- population of Stellers’ Sea Lion (Eumatopius stricted to the Bering Sea (Peterson et al. jubutus) began a sharp decline in the 1940s 1968). In fact, since the early 1940s ocean that has continued to the present: from 2000 temperatures in the area being discussed have animals in 1938 to about 35 now (Bartholomew been colder than the few decades preceding and Boolootian 1960; Ode11 1971; R. DeLong, the 1940s with the exception of the warm- pers. comm.). These islands are at the south- water periods mentioned previously. In the ern extreme of this species ’ range, and, in background of these fauna1 changes is the fact, the breeding ranges of this sea lion and loss during the 1940s of the sardine population the Tufted Puffin are identical (Udvardy formerly centered in the same geographic HISTORY OF FARALLON ISLAND MARINE BIRDS 443 region. It would be surprising if such a loss cormorants ( Pearson 1971)) puffins ( including passed without other fauna1 repercussions, Rhinoceros Auklet; Bedard 1969)) and otariids particularly to predator populations in the (sea lions and fur seals; Spalding 1964; Jones extremes of their respective species ’ ranges 1974) do divide up food resources partly in (cormorant excepted) where a species should this way. It so happens that in the region be most sensitive to the ecological balance. under discussion, the Double-crested Cormo- During the 1930s and 1940s Pacific Sardines rant is the largest cormorant (Palmer 1962; supported the largest human fishery in the Ainley pers. obs.), the Tufted Puffin is the Western Hemisphere, with its center in cen- larger puffin (Bedard 1969), and Stellers’ Sea tral California from about San Francisco to Lion is the largest otariid, while the Northern Monterey. The total fishery extended from Fur Seal is the smallest (Scheffer 1969). the Pacific Northwest to central Baja Cali- These larger predators should have been able fornia, and at its peak in the early 1940s the to exploit the larger sardine to a greater degree catch ranged up to 800,000 tons or more. A than could the smaller predators. Second, in- spectacular decline then began in the sardine formation on diets show, in fact, a greater use stocks, with a simultaneous southward con- of sardines in the past by the larger animals. traction of the population. The decline was Double-crested Cormorants fed heavily on sar- probably a result of overfishing at a time of dines in northern Baja (Wright 1913), and additional environmental stress-an extended thus probably also in central California. The period of cold water. By 1945, the sardine other cormorants did not feed much on sar- had all but disappeared from the Pacific dines in earlier days (Loomis 1895; Bent Northwest and had become a vestige of its 1922), and today they feed mostly on deep- former abundance in central California; by water and benthic fish but they also take some 1951, it was virtually gone north of Point Con- anchovies (Hubbs et al. 1970; Ainley and ception; and now the population is centered Fitch, unpubl. data). Rhinoceros Auklets fed off southern and central Baja California. As on small sardines to an unknown degree when might be expected, an ecologically equivalent wintering in southern California (Linton species, the Northern Anchovy (Engraulis 1908); however, we know nothing of their mordax), responded to the resources vacated food habits in central California today. In by the sardine. Anchovies had always been Washington, they prefer Sand Launce (Am- more abundant than sardines in the region, modytes; Richardson 1961), a species of but in the last 20 years they have staged a anchovy size not common in central California, population explosion. This has occurred, how- and they also feed heavily on anchovies and ever, almost entirely in the region south of smelt (Hypomesus; Cody 1973), another an- Point Conception. Both species are mid-water chovy-sized fish. The Stellers’ Sea Lion in schooling fishes having similar life histories, early years fed on sardines to an unknown ecologies, and distributions. An important dif- degree (Pike 1958), but other than being ference is that the mature sardine is almost noted once in association with a school of twice the size of the anchovy, adult lengths sardines, California Sea Lions are not known being about 38 cm and 20 cm, respectively. to have fed on them. More importantly, the It was mostly the older, larger sardines that California Sea Lion, in contrast to the Stellers,’ occurred off central California and farther feeds heavily on cephalopods (reviewed by north. This review of the recent history of Jones 1974), and thus the loss of sardines sardines and anchovies was summarized from would not have been as important to them as the following reports: Marine Research Com- to the Stellers.’ The Elephant Seal is special- mittee 1952, 1953; Clark and Marr 1955; Marr ized as a benthic or epibenthic feeder (More- 1960; Sette 1960; Ahlstrom 1966; Frey 1971; john and Blatz 1970), and the Northern Fur Vrooman and Smith 1971. Seal in central California feeds mostly on an- How might the loss of the sardine affect the chovies, hake ( LMerluccius) , and cephalopods populations of marine birds and mammals (North Pacific Fur Seal Commission 1969), under discussion? Some evidence suggests making the loss of sardines of little direct that those species declining or not recovering importance to these two predators. fed on sardines to a greater degree than did Two lines of evidence, a predator-size/ those that are increasing. First, a large body prey-size relationship and information on diet, of literature shows that larger predators feed suggest that in the region from central Cali- on larger prey than their smaller competitors, fornia south to northern Baja California, the and in that way among others, resources are once abundant sardine was a more important partitioned (see reviews by Selander 1966; food to certain larger members of some preda- MacArthur and Levins 1969). Specifically, tor groups-cormorants, puffins and otariids- 444 DAVID G. AINLEY AND T. JAMES LEWIS

than to those species ’ smaller ecological com- provide for the increased oil tanker traffic petitors. This loss, then, might have been resulting from the exploitation of the Alaska important enough to have, at least in part, oil fields. The existence of the populations of lessened the suitability of the region to the large alcids, including the Rhinoceros Auklet, larger predators from what it was prior to the on the Farallones depends upon the extent of sardines’ disappearance, or the loss might the expected coincident increase in oil pol- have given relatively greater competitive ad- lution. vantage to the smaller predators. Thus in Only those Farallon species “pre-adapted” marked contrast to the smaller species, the to avoid contact with humans have not thus Double-crested Cormorant (also a victim of far been adversely affected by them. These pesticides in part of the region during recent species are the most pelagic, are nocturnal in years), the Tufted Puffin, and the Stellers’ Sea their island activities, and nest in inaccessible Lion have declined or have failed to recover burrows; namely, Leachs’ and Ashy Storm- from decimations incurred just prior to the Petrels and Cassins’ Auklets. We hypothesized loss of sardines. The pattern of population that an extended mid-1800 warm-water period change in Double-crested Cormorants and caused a prolonged reduction in reproductive Tufted Puffins on the Farallones is part of this success in the auklets. This in turn resulted in situation. population decline, and the recorded population increase from 1870 to 1890 was a response SUMMARY AND CONCLUSIONS to the return of colder, more productive With satisfaction, we report increasing or waters. stable populations of marine birds on the If the continued existence of marine birds Farallon Islands, an uncommon story in many is desirable near areas of intense human ac- parts of todays’ world and even for the islands tivity, as is the case for the Farallones, the themselves during the 1800s and first half of need for protecting them from direct or in- the 1900s. Disruptive human activities from direct contact with that activity is well il- 1850 to well into the 19OOs,resulting from an lustrated in the history of the Farallon popu- egging industry and operation of a four-family lations. That full protection is the means to lighthouse station, caused sharp population this end is amply demonstrated by the results loss in breeding Double-crested, Brandts,’ and of protection given in recent years to the Pelagic Cormorants, Black Oystercatchers, Farallones. Western Gulls, and Common Murres. With the end of egging, the Brandts’ and Pelagic ACKNOWLEDGMENTS Cormorants, the oystercatchers, and the gulls This paper is dedicated to the many persons and began to recover; and with full protection in private foundations who, by their generous donations, recent years they now have done so fully. have made possible the existence of the Observatorys’ Farallon Research Station and the protection of the Double-crested Cormorants have never re- marine bird-breeding colonies on the Farallones. We covered but have remained stable, and this, particularly wish to-thank the Dean Witter Founda- we hypothesize, is due at least in part to the tion, the Charles E. Merrill Trust, and the National loss of an important prey fish, the sardine. Science Foundation, and in this regard, we acknowl- Collections of specimens around 1860 elimi- edge the coordinating efforts of J. Smail. The Fish and Wildlife Service, and especially R. D. Bauer, has nated Rhinoceros Auklets from the island, but aided in many ways. Logistic support and assistance in I972 they showed signs of recolonization. have been given by the U.S. Coast Guard and the Oil pollution caused sharp declines in the Farallon Patrol of the Oceanic Society. Unpublished populations of Tufted Puffins and Pigeon observations were kindlv offered bv A. Baldridge and R. L. DeLong; J. E. Fitch helped in the anal&is--of Guillemots, both of which were not affected bird-feeding habits; and M. C. Coulter coordinated by human activities, beginning just after 1900, gull censuses. Discussions with R. L. Brownell and and it caused further declines in the murre R. E. Jones on marine mammals were instructive, population. With a lessening of the pollution and the comments on various drafts of the oaoer by in the last 20 years and with full protection at D. W. Anderson, M. C. Baker, R. L. Brow&l< J. E. Fitch. H. W. Frev. R. E. Tones. G. W. Pane. and the Farallones, the murres and guillemots have H. C: Strong helped to improve the final result: We been increasing. Tufted Puffins have never greatly appreciate the help given by these persons recovered their former numbers, and we relate and groups. this to the loss of the sardines, too. The future, however, is tenuous for all alcid populations on LITERATURE CITED the Farallones. The islands are in the midst AHLSTROM, E. H. 1966. Distribution and abundance of one of the worlds busiest shipping lanes, of sardine and anchovy larvae in the California Current region off California, 1951-64: a sum- and in the region where large refineries are mary. U.S. Fish Wildl. Serv., Spec. Sci. Rept. located, preparations are now under way to Fish. 534:1-71. HISTORY OF FARALLON ISLAND MARINE BIRDS 445

AMERICAN ORNITHOLOGISTS' UNION. 1957. Check-list stranding of Pelagic Red Crabs, Pleuroncodes of North American birds. Fifth ed. A.O.U., plunipes, at Monterey Bay, California, since 1859, Baltimore. with notes on their biology. Calif. Fish and Game ARON, W. 1962. The distribution of animals in the 47:97-102. eastern North Pacific and its relationship to CRESS, F., R. W. RISEBROUGH, AND F. C. SIBLEY. physical and chemical conditions. J. Fish. Res. 1971. Shell thinning in eggs of the Common Bd. Canada 19:271-314. Murre, Uris a&e, from the Farallon Islands, BANKS, R. C. 1966. Terrestrial vertebrates of Anacapa California. Condor 73 : 368-69. Island, California. Trans. San Diego Sot. Nat. GRESS, F., R. W. RISEBROUGH, D. W. ANDERSON, L. Hist. 14: 173-88. F. KIFF, AND J. R. JEHL, JR. 1973. Reproductive BANKS, R. C., AND R. L. BROWNELL. 1969. Taxonomy failures of Double-crested- Cormorants in southern of the Common Dolphins of the eastern Pacific California and Baia, California. Wilson Bull. 85: Ocean. J. Mammal. 50:262-71. 197-208. BARLOW, C. 1897. The story of the Farallones. GRINNEL, J. 1926. The evidence as to the former H. R. Taylor, Alameda. nesting of the Rhinoceros Auklet in California. BARTHOLOMEW, G. A., AND R. A. BOOLOOTIAN. 1960. Condor 28: 3740. Numbers and population structure of the pin- GRINXEL, J., AND A. H. MILLER. 1944. The distribu- nipeds on the California Channel Islands. J. tion of the birds of California. Pacific Coast Mammal. 41:366-75. Avifauna No. 27. BEDARD, J. 1969. Adaptive radiation in Alcidae. Ibis GRUBER, F. VON. 1884. Die Seevogel der Farallone- lli:i89-98. - Inseln. Z. Gesamte Ornithol. 1:167-72. BENT, A. C. 1922. Life histories of North American HEER~IANN, A. L. 1859. Report upon the birds col- netrels and oelicans and their allies. U.S. Natl. lected on survey. U.S. Pacific Railroad Surv., Mus., Bull. 121. Repts. 10:29-80. BLANKINSHIP, J. W., AND C. A. KEELER. 1892. On HOWELL,’ A. B. 1917. Birds of the islands off the the natural history of the Farallon Islands. Zoe coast of Southern California. Pacific Coast Avi- 3: 144-86. fauna No. 12. BOWX+N, R. I. 1961. Late spring observations on HUBBS, C. L. 1948. Changes in the fish fauna of birds of South Farallon Island, California. Con- western North America correlated with changes dor 63:410416. in ocean temperatures. J. Mar. Res. 7:459-82. BOND, C. M. 1967. The benthic and pelagic habitats HUBBS, C. L. 1960. Quarternary paleoclimatology of of the red crab, Plewoncodes planipes. Pacific the Pacific Coast of North America. Calif. Coop. Sci. 21:394-403. Ocean. Fish. Investig., Repts. 7: 105-12. BRYANT, W. E. 1888. Birds and eggs from the HIJ~BS, C. L., A. L. KELLY, AND C. LIMBAUGH. 1970. Farallon Islands. Proc. Calif. Acad. Sci., 2nd Diversity in feeding by Brandts’ Cormorant near Ser. 125-50. San Diego. Calif. Fish and Game 56:156-65. CHANEY, R. W. 1924. Breeding condition of the HUTCHINGS, J. M. 1856. The Farallone Islands. murres on the Farallones in ”Tune. 1923. Condor Hutchings ’ Calif. Mag. 1:49-57. 26:30. JONES, R. E. 1974. A review of food habits of some CLARK, R. N., ANT) J. C. MARR. 1955. Population small marine mammals of central California. dvnamics of the Pacific Sardine. Calif. Coon. M.S. Thesis. California State Univ., Humboldt, dcean. Fish. Investig., Prog. Rept. 4:11-48. - Arcaia. CODY, M. L. 1973. Coexistence, coevolution and KAEDING, H. B. 1903. Bird life on the Farallone convergent evolution in seabird communities. Islands. Condor 5: 121-27. Ecology 54:31-44. LE BOEUF, B. J., D. G. AINLEY, AND T. J. LEWIS. COGSWELL, H. L., AND L. STALLCUP. 1956. Offshore 1974. Elephant Seals on the Farallones: popu- boat trips encounter rough water but some good lation structure of an incipient breeding colony. birds. Gull 38:30-31. COUES, E. 1862. A critical review of the family J. Mammal. 55 (in press). Procellariidae: Pt. I, embracing the Procellarieae, LINTON, C. B. 1908. Notes from Santa Cruz Island. or stormv netrels. Proc. Acad. Nat. Sci. Phila. Condor 10: 124. 16:76-78. - LONGHURST, A. R. 1967. The pelagic phase of COULTER, M. C., AND R. W. RISEBROUGH. 1973. Egg- Pleuroncodes planipes Stimpson (Crustacea, Gal- shell thinning in Ashy Petrels. Condor 75 :254- atheidae) in the California Current. Calif. Coop. 55. Ocean. Fish. Investig., Repts. 11:142-54. CRAIG, A. M., AND H. L. COGSWELL. 1956. October LOOMIS, L. M. 1895. California water birds. No. I. boat trip explores ocean beyond the Farallones. Monterey and vicinity from the middle of June Gull 38 : 48-49. to the end of August. Proc. Calif. Acad. Sci., 2nd DAWSON, W. L. 1911. Another fortnight on the Ser. 5: 177-224. Farallones. Condor 13: 171-83. LOO~IIS, L. M. 1896. California water birds. No. III. DESANTE, D., R. LEVALLEY, AND R. STALLCUP. 1972. South Farallon Island in July. Proc. Calif. Acad. The fall migration: middle Pacific Coast region. Sci., 2nd Ser. 6:353-66. Amer. Birds 26:112-17. MACARTHUR, R. H., AND R. LEVINS. 1964. Competi- DOUGHTY, R. W. 1971. San Franciscos’ nineteenth- tion, habitat selection and character displacement century egg basket: the Farallones. Geogr. Rev. in a patchy environment. Proc. Natl. Acad. Sci. 61:554-72. 51: 1207-10. EMERSON, W. 0. 1904. The Farallone Islands re- MANUWAL, D. A. 1972. The population ecology of visited, 1887-1903. Condor 6:6I-68. the Cassins’ Auklet on Southeast Farallon Island, FREY, H. W. 1971. Californias’ living marine re- California. Ph.D. Dissertation. Univ. California, sources and their utilization. Dept. Fish and Los Angeles. Game, Calif. Resources Agency. MARINE RESEARCH COMMITTEE. 1952. The sardine GLYNN, P. W. 1961. The first recorded mass and its environment, yesterday and today. Calif. 446 DAVID G. AINLEY AND T. JAMES LEWIS

Coop. Sardine Res. Program, Prog. Rept. 2:9- (BT) data with two types of historic and modem 44. sea surface data. Calif. Coop. Ocean. Fish. MARINE RESEARCH COMMITTEE. 1953. Sardines and Investig., Repts. 10:141-52. man. Calif. Coop. Ocean. Fish. Investig., Prog. SCHEFFER, V. B. 1969. Seals, sea lions and walruses: Rept. 3:25-30. a review of the pinnipedia. Stanford Univ. MARFI, J. C. 1960. Discussion, p. 192-94. In 0. E. Press, Palo Alto. Sette [ed.], The long term historical record of SELANDER, R. K. 1966. Sexual dimorphism and dif- meteorological, oceanographic and biological data. ferential niche utilization in birds. Condor 68: Calif. Coop. Ocean. Fish. Investig., Repts. 7. 113-52. MOREJOHN, G. V., AND D. M. BLATZ. 1970. Contents SETTE, 0. E. 1960. The long term historical record of the stomach of an Elephant Seal. J. Mammal. of meteorological. oceanoerauhic and biological 51: 173-74. data. Calif. coo,. Ocean.-Fish. Investig., Repts. Mosuy, H. S. 1963. Wildlife investigational tech- 7: 181-94. niques. Second ed. (revised). Wildlife Society, &AIL, J., D. G. AINLEY, AND H. STRONG. 1972. Washington, D.C. Notes on birds killed in the 1971 San Francisco NORDHOFF, C. 1874. The Farallon Islands. Harpers’ oil spill. Calif. Birds 3:25-32. New Monthly Mag. 48:617-25. SMALL, k. 1960. The nesting season: southern NORTH PACIFIC FUR SEAL COMMISSION. 1969. Re- Pacific Coast region. Audubon Field Notes 14: Dart on investigations from 1964 to 1966. Wash- 477. ington, D.C. _ S~IITH, C. F. 1934. Bird notes from the Farallon ODELL, D. K. 1971. Censuses of pinnipeds breeding Islands. Condor 36: 170-72. on the California Channel Islands. J. Mammal. SOKAL, R. R., AND F. J. ROHLF. 1969. Biometry. W. 52: 187-90. H. Freeman Co., San Francisco. ORR, R. T. 1963. A northern record for the Pacific SPALDING, D. J. 1964. Comparative feeding habits of Bottlenose Dolnhin. 1. Mammal. 44:424. the fur seal, sealion, and Harbour Seal on the PALMER, R. S. [eh.]. 1962. Handbook of North British Columbia coast. Fish. Res. Bd. Canada, American birds, Vol. 1. Yale Univ. Press, New Bull. 146: l-52. Haven. SVERDRUP, H. V., M. W. JOHNSON, AND R. H. PEARSON, T. H. 1968. The feeding biology of sea- FLEMING. 1942. The oceans: their physics. bird species breeding on the Farne Islands, chemistry, and general biology. Prentice-Hall; Northumberland. I. Anim. Ecol. 37:521-51. New York. PETERSON, R. S., AND h. J. LE BOEUF. 1969. Popu- TAYLOR, H. R. 1887. A trip to the Farallone Islands. lation study of seals and sea lions. Trans. 34th Ornithol. 001. 12:41-43. N.A. Wildl. Nat. Resour. Conf., Wildl. Mgmt. THORESEN, A. C. 1964. The breeding behavior of the Inst., Washington, D.C. Cassin Auklet. Condor 66:456-76. PETERSON, R. S., B. J. LE BOEUF, AND R. L. DELONG. TUCK, L. M. 1960. The murres: their distribution, 1968: Fur. seals from the Bering Sea breeding populations, and biology. Can. Wildl. Serv., in California. Nature 219:899-901. Rept. Ser., No. 1. Ottawa. PIKE, G. C. 1958. Food of the Northern Sealion. Fish. UDVAR