Demography and Populations

Age At First Breeding; Intervals Between Breeding

Typically begins breeding at 3–5 yr (Schreiber et al. 1989), but some have bred, with lower success, at 1 and 2 yr (Blus and Keahey 1978, U.S. Fish Wildl. Serv. 1983, McNease et al. 1984). Females tend to mature sooner and begin breeding at a younger age than males do (U.S. Fish Wildl. Serv. 1983, Schreiber et al. 1989). Usually breeds annually, possibly every 8.5–10 mo below 20°N latitude (Harris 1969, Schreiber 1980a). Segments of Pacific coast populations may skip breeding during ENSO events (Tovar et al. 1987, Valle et al. 1987, Schreiber et al. 1989).

Clutch

Number of Eggs. Clutches from various North American populations in oological collections had mean of 2.95 eggs (95% confidence limits 2.68–3.22, n = 236), with no significant differences among populations (Anderson and Hickey 1970). Modal clutch size in most populations is 3 eggs. Maximum reported for Brown Pelican is 5 (n = 3 out of about 10,000 clutches in S. Carolina; Blus and Keahey 1978). Report of 18 eggs in 1 nest in Texas extraordinary; highly unlikely that all 18 laid by 1 female (Pearson 1921). Maximum reported for Peruvian Pelican is 8 (Coker 1920). On Cape Fear River, NC, mean clutch size 2.71 (n = 98) in 1982 (Parnell et al. 1984), 2.92 ± 0.3 SD (n = 370) in 1991–1993; clutch-size distribution (1991–1993): 1 egg, 0.3%; 2 eggs, 8.6%; 3 eggs, 90.0%; 4 eggs, 1.1% (MAS). Eight-year (1969–1976) mean at Tarpon Key, FL, 2.62 (n = 328); clutch-size distribution: 1 egg, 5.5%; 2 eggs, 26.5%; 3 eggs, 68.0% (Schreiber 1979). At Alafia Banks, FL, mean 2.4 ± 0.7 SD (n = 53) in 1989, 2.5 ± 0.6 SD (n = 71) in 1990 (Ploger 1992). Annual means in Texas (1975–1981) ranged from 2.0 (n = 18) to 3.25 (n = 16; King et al. 1985). On Isla Isabel, Mexico, mean 2.31 ± 0.8 SD (n = 49); clutch-size distribution: 1 egg, 18.4%; 2 eggs, 34.7%; 3 eggs, 44.9%; 4 eggs, 2.0% (Pinson and Drummond 1993). Annual means in Gulf of California (1974–1976) ranged from 2.78 (n = 73) to 2.95 (n = 123; Keith 1978). On Taboga I., Panama, mean 2.76 (n = 239; Montgomery and Martínez 1984). Means at Montalva Bay and Conejo Cay, Puerto Rico (1980–1982) ranged from 2.39 (n = 39) to 2.88 (n = 9); mean at Congo Cay and Dutch Cap, U.S. Virgin Is., 2.54 (n = 137) and 2.66 (n = 53), respectively (Collazo et al. 1998). Clutch size of immatures in S. Carolina (2.16; range 1–3, n = 37) significantly smaller than that of adults (2.85; range 2–4, n = 89; Blus and Keahey 1978). Seven-year (1987–1993) mean of primarily immature breeders in Maryland 2.3 (n = 113; D. Brinker pers. comm.). Mean of 3-yr-old breeders at Grand Terre, LA, 2.0 (n = 11; McNease et al. 1984).

Number of Clutches per Breeding Season. Normally one, but may replace lost clutch (see Breeding: eggs, above).

Annual And Lifetime Reproductive Success

Breeding success varies widely among years and colonies (Table 1), mainly in response to variation in food availability (Schreiber 1979, Anderson et al. 1982). Productivity averages about 1 fledgling/nest (Schreiber 1979), less when colonies subjected to human disturbance (Anderson and Keith 1980, Anderson 1988), ectoparasite infestation (King et al. 1977a, 1977c), and inclement weather (McNease et al. 1984). Breeders in immature plumage less productive than those in adult (Definitive) plumage (Blus and Keahey 1978). Effective reproductive life span 4–7 yr in Brown Pelican (Schreiber and Mock 1988), probably shorter in Peruvian Pelican because of ENSO-related mortality (Duffy 1980a). No direct measures of lifetime reproductive success available.

Number Of Broods Normally Reared Per Season

One.

Proportion Of Total Females That Rear At Least One Brood To Independence

Unknown.

No valid estimates of age-specific survival rates available. Band recoveries suggest that only 30% survive first year, fewer than 2% survive beyond 10 yr; maximum life span 43 yr (Schreiber and Mock 1988). Peruvian Pelican probably experiences higher mortality rate and has shorter life span than Brown Pelican because of effects of periodic ENSO events (Duffy 1980a).

Diseases

Gastroenteritis and peritonitis, apparently caused by ingestion of toxins produced by the bacterium Clostridium perfringens, caused 72 deaths and illnesses in Florida in May–Jun 1983 after pelicans fed in waters contaminated by human sewage (Ankerberg 1984). Botulism caused 25 deaths in Puerto Rico in Nov–Dec 1982 (Collazo 1985). Recurring botulism outbreaks at Salton Sea have caused >1,500 deaths since 1996; more sickened pelicans captured, rehabilitated, and released; cholera caused 14 deaths in 2000 (U.S. Fish Wildl. Serv. data, available online at http: //pacific.fws.gov/salton/default.htm). Erysipelas, an infection caused by the bacterium Erysipelothrix rhusiopathiae, killed about 350 pelicans that had fed on fish scraps discarded by restaurants and commercial fishing industry in central California in Oct–Dec 1987 (Windingstad 1991); many additional deaths reported in central and s. California in early 1988 (F. Gress pers. comm.). Forty-three deaths in California in Sep–Oct 1991 (Fritz et al. 1992) and 150 deaths in Cabo San Lucas, Mexico, in winter 1995–1996 (Beltrán et al. 1997) attributed to domoic acid poisoning, which resulted from pelicans feeding on fish that had ingested toxin-producing diatoms (Pseudonitzschia spp.). Isolated deaths in Florida and S. Carolina caused by enteritis (n = 8), acute bacteremia (n = 1), pneumonia (n = 1), and other respiratory problems (n = 3; Blus et al. 1977, 1979b). Incidence of nonfatal cases of these diseases not known.

Body Parasites

Ectoparasites. Fowl ticks commonly inhabit nests and feed on adults and nestlings. Ornithodoros capensis recorded from colonies in Texas (King et al. 1977a), Georgia, S. Carolina, and N. Carolina (Keirans et al. 1992); O. denmarki from islands in Gulf of California (King et al. 1977c); and O. amblus from Peruvian Pelican colonies (Duffy 1983c). Tick infestations have caused nest desertions in N. and S. Carolina (Wilkinson et al. 1994), Texas (King et al. 1977a), Gulf of California (King et al. 1977c), Galápagos Is., and Peru (Duffy 1983c). Mean number of ticks/500-ml sample of nest material from nests not deserted 7.2 ± 1.7 SE on Isla Macabi and 42.2 ± 5.4 SE (n = 10) on Isla Mazorca, Peru (Duffy 1983c). Although ticks harbor viruses that may be transmitted to pelicans, no cases of illness or death due to tick-borne viral diseases known (King et al. 1977a, 1977c; Clifford et al. 1980; Keirans et al. 1992). Hippoboscid fly (Olfersia sordida) and epidermoptid mite (Myialges caulotoon) found on pelicans in Galápagos Is. (Madden and Harmon 1998). Five species of bird lice (Mallophaga) recovered from P. o. californicus, including Menopon titan, which infests inner surface of gular pouch as well as feathers (Kellogg 1896). Swarm of mosquitoes caused abandonment of 600 nests with young at Pelican I., FL (Nelson 1911).

Endoparasites. Host to 31 species of helminths in Florida and Louisiana (Courtney and Forrester 1974). The most prevalent and abundant are the trematodes Phagicola longus, Mesostephanus appendiculatoides, Galactostomum darbyi, and Stephanoprora denticulata, which inhabit the small intestine; the kidney fluke Renicola thapari; and the nematodes Contracaecum multipapillatum and C. spiculigerum, which live in the proventriculus. Mean of 7,134 helminths/bird (Humphrey et al. 1978). Three species of diplostomes, Bolbophorus confusus, Bursacetabulus pelecanus, and Bursacetabulus [Bursatintinnabulus] macrobursus (Tehrany et al. 1999), found in small intestines of individuals in Texas (Dronen et al. 1999). No helminth-caused diseases or deaths known (Greve et al. 1987). Three species of Hypoderidae mites, Phalacrodectes punctatissimus, P. pelecani, and Pelecanectes apunctatus, found in subcutaneous tissues of trachea and neck of pelicans from Florida and Louisiana; no pathology observed (Pence and Courtney 1973). Coccidian sporozoa Eimeria pelecani and Eimeria sp. recovered from fecal samples of nestlings from Florida (Courtney and Ernst 1975).

Exposure

Cold temperatures cause mortality through hypothermia, frostbite damage to foot webs and gular pouch, and starvation resulting from reduced availability of fishes in surface waters (Schreiber and Mock 1988, McNease et al. 1992). Cold spells at Pelican I., FL, resulted in ≥500 deaths of nestlings in 1908 (Chapman 1908) and caused desertion of eggs in 1979 (Schreiber 1980a). In N. Carolina, blizzard in Mar 1980 killed many adults; extensive mortality of adults and immatures also reported during winter of 1981–1982, when starving birds frequented piers, marinas, and fish houses, where they received handouts (Parnell and Soots 1982). Periods of subfreezing temperatures in winter 1989–1990 killed nearly 1,000 pelicans, mostly immatures, in N. Carolina (W. Golder pers. comm.) and 169 adults and immatures in Louisiana (McNease et al. 1992). Sensitivity to cold limits winter distribution mainly to areas south of 38°N (see Distribution, above) and restricts breeding range to regions with >4.5 consecutive months of above-freezing temperatures (Schreiber 1980a). Hurricanes killed many individuals in S. Carolina in 1894 (Wayne 1894) and 1989 (Wilkinson et al. 1994). Tidal flooding destroys many nests on low islands along Atlantic and Gulf coasts of U.S. As many as 400 nestlings and large numbers of eggs lost to single flood in Louisiana (McNease et al. 1992). Hundreds of eggs washed from nests by storm tide on Pelican I., FL (Chapman 1908). In S. Carolina, tidal flooding destroyed 29 of 125 sample nests (Blus and Keahey 1978). In N. Carolina, 16 of 76 sample nests with eggs destroyed by flooding after storm dropped 15 cm of rainfall in 3 h on 20 Apr 1991 (MAS).

Predation

See Behavior: predation, above.

Competition With Other Species

No reports.

Starvation

Reduced prey availability during severe ENSO events results in mass starvation. During 1982–1983 ENSO, large numbers of dead and dying Brown Pelicans observed in Galápagos Is. and Panama (Ainley et al. 1986); 70,000 adult and 210,000 nestling Peruvian Pelicans starved to death in Peru (Tovar et al. 1987). Nestlings starved in Galápagos Is. during weaker 1972 ENSO, but no adult mortality reported (Boersma 1978). Annual reproductive success in S. California Bight correlated with anchovy abundance; in years when anchovies scarce, many adults abandon their nests and nestlings starve (Anderson et al. 1982). Several hundred downy young apparently died of starvation in S. Carolina in 1976; reason for lack of food unknown (Blus et al. 1979b). At Alafia Banks, FL, starvation accounted for 10% of 109 nestling deaths whose cause was known (Ploger 1992).

Migration

Most recoveries of birds banded as nestlings in N. and S. Carolina occur during Nov–Dec and Feb–Mar; deaths apparently caused by stress of migration and cold weather (Schreiber and Mock 1988). First-year birds dispersing from colonies in Gulf of California usually die after strong winds push them inland to Sonoran Desert (Anderson et al. 1977a).

Siblicide And Infanticide

First-hatched nestling often kills ≥1 sibling either directly by pecking or indirectly by driving it from nest and preventing it from feeding (see Breeding: young birds, above). At Alafia Banks, FL, siblicide accounted for 30% of 109 nestling deaths whose cause could be determined; another 28% of deaths attributed to siblicide and/or starvation (Ploger 1992). On Isla Isabel, Mexico, only 15% of successful nests in which ≥1 egg hatched (n = 27) fledged >1 young; first-hatched was sole survivor in 70% of broods. Over half of 58 dead chicks examined bore peck wounds, and 43% were found outside the nest, suggesting cause of death was siblicide (Pinson and Drummond 1993). Marauding fledglings and immatures killed 11 nestlings at Alafia Banks, FL (Ploger 1992). Adults and nonbreeding immatures at ground colony in N. Carolina pecked and killed several young that fled their nests in response to human disturbance (MAS). Four instances of cannibalism observed at Pelican I., FL, also facilitated by human disturbance, which caused brooding parents to leave nestlings unattended (Chapman 1908). Adult Peruvian Pelicans sometimes reach into other nests and toss nestlings out, which may then die of exposure (Coker 1920).

Initial Dispersal From Natal Site

Sightings of marked individuals indicate high degree of natal philopatry, but movement among established colonies not uncommon. In S. Carolina, all 26 wing-tagged or color-banded breeders observed on Marsh I. in the early 1980s had hatched there; 34 marked breeders observed on Bird-Key Stono either hatched there or on nearby Deveaux Bank, which had washed away in 1980, forcing colony relocation (Wilkinson 1984). Most prefledglings transplanted from Florida to Queen Bess and North Is., LA, began nesting near their release sites within 3 yr, although a few returned to Florida (McNease et al. 1984). Similarly, fledglings from Queen Bess I. transplanted to Last I., LA, began breeding there within 3 yr (McNease et al. 1992). New colonies often founded by immatures (Williams and Joanen 1974, Schreiber and Schreiber 1982). Band re-turns suggest that birds dispersing from St. Andrews Bay, FL, colony may have contributed to growth of Mobile Bay, AL, colony 210 km to the west, and probably established colony on Lanark I., FL, 110 km to the east (Wood et al. 1995).

Fidelity To/Dispersal From Breeding Site And Winter Home Range

Stable, undisturbed colony sites tend to be occupied year after year, some for several decades or longer. Mean annual colony turnover rate 5% in Louisiana (Visser and Peterson 1994) and 8% in N. Carolina (Parnell et al. 1997). However, low turnover rate in Louisiana related to lack of suitable alternative sites rather than stability of used sites (Visser and Peterson 1994). Breeders displaced from Lanark I., FL, by Hurricane Allison in Jun 1995 apparently started new colony 19 km to the east on Palmetto I. in Oyster Bay (Wood et al. 1995). Exchange among colonies in California and w. Mexico common (Anderson 1983); colony locations regularly shift, apparently in response to changes in prey abundance and distribution, human disturbance, and tick infestation (Everett and Anderson 1991). Little information on fidelity to winter range.

Home Range

Forages mainly within 20 km of nesting islands (Briggs et al. 1981). Spends most time on or near nest when not foraging (Schreiber 1977a). Range expands during nonbreeding season; regularly observed up to 175 km from s. California mainland and 75 km from nearest island, during fall and early winter (Briggs et al. 1981). Little information on space use by marked individuals.

Numbers

Breeding. Appendix 2 . Total Brown Peli-can population estimate 101,300–104,500 breeding pairs: 50,000–51,000 P. o. californicus, 44,000–45,000 P. o. carolinensis, 4,800–5,000 P. o. occidentalis, ±2,000–3,000 P. o. urinator, 500+ P. o. murphyi . Nearly two-thirds of P. o. californicus population breeds in Gulf of California; about 17,000 pairs nest on Islas Salsipuedes, Las Animas, and San Lorenzo during non-ENSO years (Velarde and Anderson 1994). Largest P. o. californicus colony in U.S. on West Anacapa I., CA; 5,300 nesting attempts in 1999 (F. Gress pers. comm.). Nearly 40,000 pairs of P. o. carolinensis bred in U.S. in 1999, about 60% on Gulf coast. Largest colonies were on Gaillard I. (Mobile Co.), AL (5,225 nests; R. Clay pers. comm.), Last (Raccoon) I. (Terrebonne Parish; 3,700 nests) and Curlew/Stakes Is. (Plaquemines Parish, LA; 3,600 nesting birds; Louisiana Dep. Wildl. Fish., Nat. Heritage Prog. data), and Little Egg I. Bar (Glynn Co., GA; 3,372 nests; Georgia Dep. Nat. Resour., Nongame Wildl. Nat. Heritage Section data). In e. Mexico, largest known colony (300 pairs in 1986 and 700–1,000 pairs in 1996) on Isla Contoy, Quintana Roo; 4 other sites contained 128 nesting pairs when last surveyed in 1986 (Blankinship 1987, D. Blankinship pers. comm.). Fewer than 100 pairs estimated to breed off Belize and Caribbean coast of Honduras. Perhaps as many as 1,000 pairs breed along Pacific coast of Honduras and Costa Rica, but no recent census data available. The 2 largest colonies in Panama ( Islas Taboga and Uravá) contained 1,042 nests in 1997 and 1,455 nest in 1998; 7 other colonies had about 1,250 nests when last censused in the 1970s (G. Angehr pers. comm.). About 1,500 pairs of P. o. occidentalis breed in the West Indies (Collazo et al. 2000). No recent estimate of number in Venezuela, but > 3,300 nests counted in 1983 (Guzman and Schreiber 1987). Size of P. o. urinator population in Galápagos Is. roughly estimated to be “several thousand” (Valle et al. 1987) and “a few thousand” (Cepeda and Cruz 1994) breeding pairs. Only known P. o. murphyi colony consisted of ≥500 pairs in 1993 (Ridgely and Greenfield 2001). Size of Peruvian Pelican population in Peru 400,000 individuals in 1996 (Jahncke 1998). Total breeding population in Chile unknown, but 2,699 pairs bred on Isla Pájaro Niño in central Chile in 1995–1996; only 1,032 pairs bred in 1996–1997, none during 1997–1998 ENSO (Simeone and Bernal 2000).

Winter. Resident in most of breeding range, but numbers north of about 30°N decrease with increasing latitude; begins breeding in winter south of about 30°N (see Breeding: phenology, above). Rarely seen later than mid-Dec in Maryland portion of Chesapeake Bay, but small numbers (<20) regularly winter in mouth of bay in Virginia (Iliff 1999a). Several thousand winter as far north as the Carolinas (Parnell and Soots 1982, Wilkinson 1984). Postbreeding migration out of Gulf of California brings large numbers of P. o. californicus to Pacific coast of California, Oregon, and Washington in autumn. Numbers decline by early winter as birds withdraw to south. Between 1975 and 1978, peak autumn numbers in s. California 65,000–94.000; peak winter numbers 1,800–5,000 (Briggs et al. 1981). Numbers in central and n. California in 1980–1983 peaked at 23,000–34,000 in autumn and <400 in winter (Briggs et al. 1983). Numbers along Oregon and Washington coasts in mid-Sep increased annually from 4,194 in 1987 to 10,119 in 1991 (Jaques 1994). Fewer than 100 remain into winter (Tweit et al. 1999). Aerial survey of Mexican Gulf coast during Dec 1979–Jan 1980 yielded a total of 4,438 birds; 2,270 counted during survey of Gulf and Caribbean coasts of Mexico in Mar 1986 (Blankinship 1987). During Feb 1988, 6,031 pelicans observed in Gulf of Panama; in Jan 1993, 582 counted along Caribbean coast and 3,017 along Pacific coast of Panama (Butler et al. 1998).

Trends

Figure 4 . Numbers in U.S. have increased dramatically since late 1950s to early 1970s, when pesticides reduced populations to their lowest levels. Total population size now exceeds historical levels. Number of nests (colonies) of P. o. carolinensis along Atlantic coast steadily increased from 2,796 (8) in 1970 (Wilkinson et al. 1994) to 15,670 (32) in 1999 (Appendix 2). Largest absolute growth occurred in Georgia, N. Carolina, and Virginia. In S. Carolina, 59% drop in nests between 1989 and 1995 due to loss of key nesting islands caused by erosion; individuals apparently relocated to neighboring states (Wilkinson 1997). Number breeding along Florida Atlantic coast relatively stable since 1989. Along U.S. Gulf coast, numbers of nests (colonies) fluctuated between 4,030 and 6,686 (17–30) from 1970 to 1983 (Wilkinson et al. 1994), then rapidly rose to 24,195–27,869 (39–50) in late 1990s. Greatest increases recorded in Louisiana, Alabama, and Texas. Number nesting on Florida Gulf coast in 1999 lowest since 1983. Statewide, number of nesting pairs averaged 7,658 between 1968 and 1990 and 9,750 between 1991 and 1998 (Florida Fish Wild. Conserv. Comm. data). Numbers seen in interior Florida have increased since 1986 (McNair 2000). As many as 60–80 pelicans observed on Lake Okeechobee in late winter to summer 1988–1992; 3 nests discovered in 1991 and at least 14 nests built in 1992 (Smith and Goguen 1993); no subsequent reports of breeding. In California, breeding population of P. o. californicus increased dramatically during 1980s and has remained fairly stable since early 1990s (about 6,000 nesting attempts/yr during non-ENSO years). First reported breeding at Salton Sea in 1996 when 3 nests with downy young discovered in Jun; another nest started in Dec 1996 and 6 started in early 1997 failed (Sturm 1998). Small numbers also nested unsuccessfully in 1998; no breeding in 1999 (D. Anderson pers. comm.).

Breeding population of P. o. californicus in Gulf of California, less impacted by pesticides, apparently has remained stable since studies began in 1970; possible decline in Sinaloa mainland population due to abandonment of colony site in Ensenada del Pabellon in 1998 (D. Anderson pers. comm.). Numbers along Pacific coast of Baja apparently stable (Everett and Anderson 1991). Known breeding population of P. o. carolinensis in e. Mexico doubled between 1986 (428 pairs) and 1996 (700–1,000+ pairs; D. Blankinship pers. comm.). In Panama, numbers at 2 largest colonies relatively stable since at least 1979; no recent censuses of other Panamanian colonies (G. Angerh pers. comm.). Data insufficient to assess trends elsewhere in range of P. o. carolinensis . Size of P. o. occidentalis population down about 20% from previous estimate of 6,200 pairs, which was considered unrealistically low (van Halewyn and Norton 1984). No known reason for apparent decline, but human disturbance and habitat loss possible (Collazo et al. 2000). Mean number wintering in Puerto Rico declined by 74% between 1980–1982 (2,289 individuals) and 1992–1995 (593; Collazo et al. 1998). No information on P. o. murphyi or P. o. urinator trends. Peruvian Pelican population fluctuates greatly in response to ENSO conditions. Population of 420,000 adults in 1981–1982 declined to 110,000 during the 1982–1983 ENSO event, then grew to 620,000 by 1985–1986 (Tovar et al. 1987). Numbers well below highest levels attained during 1950s and 1960s, prior to collapse of anchoveta fishery (Duffy 1983b, 1994).

Populations regulated mainly by density-independent climatic and oceanographic phenomena that influence prey availability (Duffy 1980a, Schreiber and Schreiber 1983, Anderson and Gress 1984). Fledging rate most responsive demographic variable; food shortage during breeding results in nest abandonment and nestling starvation (Schreiber 1979, Anderson et al. 1982), but usually little mortality of adults (Anderson and Gress 1984), which can immigrate to areas with greater prey densities. Pacific populations experience low productivity during periodic ENSO events, when warming of surface waters reduces prey availability; adult survivorship reduced during severe conditions (Ainley et al. 1986). Little known about causes or extent of variation in prey availability in Atlantic, Gulf of Mexico, and Caribbean. Because Brown Pelicans are long-lived, sporadic breeding failures have little effect on long-term population stability, but recurring breeding failures result in population decline (Schreiber 1980b). Decreased adult survival has greater influence on population stability than decreased productivity. Peruvian Pelican population regulated by catastrophic breeding failure and mass adult mortality associated with periodic severe ENSO conditions; high productivity in years when prey abundant allows rapid recovery (Duffy 1980a, 1983b).