Demography and Populations

Age At First Breeding; Intervals Between Breeding

Based on band recovery data, most individuals do not breed until at least 3 yr of age (in their fourth summer); a few may attempt to breed as 2 yr olds (in their third summer; Ludwig 1942, Gill and Mewaldt 1983). A small proportion of population is made up of nonbreeding adults (Ludwig 1942, Cuthbert 1985b); five to ten 1- to 2-yr old birds (<1%) were seen at colony sites annually in n. Lake Michigan. These individuals appeared mid-Jun–mid-Jul and spent several days at a site (FJC).

Usually breeds every year.

Clutch

Typically lays 1–3 eggs, rarely 4; 2 most frequent in South, 3 more common in North (Bent 1921). In Grays Harbor, WA, average clutch size 2.5 eggs ± 0.7 SD (mode 3; n = 2,315 nests; Penland 1981). In Great Lakes, mean clutch sizes from 2.4 (n = 2,170 nests; Shugart et al. 1978) to 2.81 (n = 225 nests; Ludwig 1965). On South Limestone I., Ontario, 2 eggs more common; 19781979 mean clutch size 1.88 eggs ± 0.033 SE (n = 303 nests; Quinn and Morris 1986). In San Francisco Bay area, CA, mean clutch size 1.86 (2 most common; n = 378 nests; Miller 1943); in San Diego Bay area, mean clutch size 2.08 (n = 351 nests; Kirven 1969); in s. California, mean clutch size 1.81 (n = 110 nests), and in central California 1.92 (range 1–3; n = 85 nests; Schew et al. 1994). In Texas, 2-egg clutches most common (Bent 1921, Mitchell and Custer 1986).

For late-nesting and renesting birds, clutch size generally smaller than in initial nesting attempt early in season (Bergman 1953, Soikkeli 1973a, Fetterolf and Blokpoel 1983); on Lake Michigan, average clutch size of renesters was 1.8 (n = 276 nests), compared to 2.5 for initial nesters (n = 686 nests; Shugart et al. 1978).

Supernormal clutches of 4–6 eggs reported in Pacific Northwest, California, and Michigan (Miller 1943, Conover 1983, Penland 1984, FJC). Rare in North America before 1940, but in U.S. increased in frequency after 1950. In Pacific Northwest, >4% of nests in 2 colonies contained 4 eggs; 0.8% contained 5–6 eggs (n = 561). Trapping and sexing all attendants at two 3-egg nests, two 4-egg nests, and one 5-egg nest revealed that supernormal clutches were incubated by female–female pairs, while 3-egg clutches were incubated by heterosexual pairs (Conover 1983). On Shoe I., MI, many supernormal clutches (4–6 eggs) were found adjacent to a gravel ridge following a major storm that washed out most nests on this ephemeral shoal. Adults reconstructed nests and rolled abandoned eggs into them from waterline of storm. Some of these eggs eventually hatched (FJC).

Annual And Lifetime Reproductive Success

One successful brood/season. Annual reproductive success varies widely (Table 1). Hatching success in 5 well-studied colonies across North America ranged from 70 to 85%; number hatched per nest ranged from 1.4 to 1.7. Fledging success (chicks fledged per eggs hatched) ranged from 52 to 66%, and adult reproductive success (chicks fledged per nest) ranged from 0.6 to 1.6 (see Table 1).

Reproductive success is heavily impacted by several factors, especially predation and human disturbance (see Conservation and management: effects of human activity, below). At Rice I., OR, where predation by Bald Eagles and gulls is high, fledging success is very low; proportion of nests with ≥1 chick raised to fledging is estimated at only about 5% (Roby et al. 1998). Reproductive success can vary greatly among years within a colony. In Great Lakes, at Hat I., Lake Michigan, number of chicks fledged per pair ranged from 0.5 to 1.2 between 1976 and 1978. In 1977, success very low (0.5 chick fledged/pair, n = 686 nests) because of researcher activity. In 1976 and 1978, success higher: 1.2 (n = 732 nests) and 0.9 (n = 124 nests), respectively; significant investigator disturbance did not occur during these years. Other factors affecting reproductive success between years include initial clutch size and storm-related nest failures (Shugart et al. 1978). Late and second nests produce fewer young and lower fledging rates than first nesting attempts (Fetterolf and Blokpoel 1983).

No information on lifetime reproductive success, but mean breeding-life expectancy of birds from Great Lakes and Pacific Coast estimated at 8.8 and 8.6 yr, respectively (Ludwig 1965, Gill and Mewaldt 1983). In Great Lakes, 3- to 15-yr-olds made up nuc-leus of breeding population (Ludwig 1965).

Greatest mortality occurs during first 6 mo of life (Ludwig 1942, Soikkeli 1970). In Great Lakes, on basis of 280 band recoveries, approximately 62% of birds died before third year of life or first breeding year; postfledging mortality highest in Aug and Sep (Ludwig 1965). On Pacific Coast, annual survival rates were estimated on basis of band recoveries of living and dead birds; estimates for second-year, third-year, and after-third-year birds were 0.79, 0.87, and 0.89, respectively (i.e., 79% of second-year birds entered third year, etc.). Hatch-year birds had estimated survival rate of 0.82 for 4-mo period between fledging and end of first calendar year, and annual survival rate of 0.55 (55% lived to second year); 57% of fledglings reached fourth year or adulthood (n = 397 band recoveries; Gill and Mewaldt 1983).

Once a bird reaches maturity, it will likely survive a long time; in Great Lakes, on basis of 260 band recoveries, birds attaining adulthood had average life span of 11.9 yr; 16 lived >15 yr (Ludwig 1965). Maximum age recorded is 26 yr 2 mo; 4 birds banded in Great Lakes all lived minimum of 26 yr (Bergstrom 1952, Ludwig 1965, Clapp et al. 1983).

External parasites found on birds from e. U.S. include 4 species of lice: Actornithophilus funebre, Degeeriella praestans, Menophon sp., and Philopterus melanocephalus (Peters 1936). Internal parasites recorded in birds from Great Lakes include the cestodes Dibothriocephalus oblongatum (used lake herring as final intermediate host; Thomas 1947 in Ludwig 1965), Schistocephalus solidus, and Paricterotaenia sp.; the nematode Cosmocephalus sp. (possibly oblevatus); and the trematodes Diplostomum sp., Cotylurus sp., Ornithobilharzia sp. (possibly lari), Clinostomum sp., and Stephanoprora sp. (Ludwig 1965).

Egg Mortality

Causes include embryonic death, predation, abandonment (see Conservation and management: effects of human activity, below), and strong winds and high tides washing eggs from or flooding nests. At a dike in San Francisco Bay, CA, DeGroot (1931) found >50 eggs rolled to water edge following strong winds. In n. Lake Michigan, 12% of nests monitored (n = 2,170) were destroyed by storm-driven waves (Shugart et al. 1978).

Chick Mortality

Factors include attacks by adult terns, predation, and exposure to cold temperatures and drifting sand, which result when parent birds are flushed from nests and leave young birds exposed. At a subcolony on South Limestone I., Ontario, with >70 nests, only 1 chick fledged; most chicks were preyed on when ≥7 d old by Herring Gulls from a colony bordering the Caspian Tern colony (J. Quinn pers. comm.). Chicks 1–3 wk old are most susceptible to death as result of entering another gull or tern’s territory during a disturbance (Penland 1976). High tides and floods also wash chicks from nests. Starvation can be a major mortality factor among chicks, especially among younger chicks in broods of 2–3 (Soikkeli 1973a, Quinn 1980, Pierce 1984).

Juvenile And Adult Mortality

Ludwig (1965) used band recovery data to determine cause of death for 369 terns from Great Lakes during 1922–1953. Of these birds, nearly 39% were adults (>3 yr old); 11% were subadults (18 mo–3 yr); 22% were immatures (6–18 mo); and 29% were juveniles (0–6 mo). Fifty-five percent were found dead of unknown cause, 22% were shot or trapped, 13% were found injured or sick, and 5% were collected for scientific purposes. Additional causes (4%) included weather, entanglement, collisions, and predation. At Rice I., OR, Bald Eagle predation has been a significant source of mortality for adults (Roby et al. 1998).

Competition With Other Species

In Grays Harbor, WA, Western Gulls (Larus occidentalis), Glaucous-winged Gulls (L. glaucescens), and Western x Glaucous-winged gull hybrids compete with Caspian Terns for nesting sites. Gulls also eat tern eggs. Gulls arrive and establish nest territories a month or more before terns (Penland 1976, 1982). In Great Lakes, Herring and Ring-billed gulls also arrive at nesting sites earlier than Caspians, and limit available nest sites; Herring Gulls usually win aggressive encounters with Caspians, while Caspians can usually displace Ring-billed Gulls (Shugart et al. 1978). Caspian Terns and Herring Gulls kill each other’s young if young wander in colonies (Ludwig 1942). In Atlantic Canada, growing gull populations have displaced terns from some of the best breeding habitat and increased predation on young and eggs (Lock 1993).

Parasitic Jaegers (Stercorarius parasiticus) regularly steal food from Caspian Terns in Northwest Territories at Great Slave Lake (J. Sirois pers. comm.), as do Western x Glaucous-winged gull hybrids at Rice I., OR (Roby et al. 1998), and Herring Gulls in Great Lakes (FJC). Caspian Terns opportunistically steal food from other fish-eating birds and humans (King 1987)—e.g., from Red-breasted Mergansers (Mergus serrator) fishing adjacent to a colony in n. Lake Michigan (FJC).

Initial Dispersal From Natal Site

Returns to general area of hatching (e.g., Great Lakes, Pacific Coast), but no strong tendency to breed at natal colony. For example, in Pacific Coast population, 58% of adult birds were found at nonnatal colonies during breeding season; only 42% were found at natal colony (n = 79; Gill and Mewaldt 1983). In Great Lakes during 1960–1967, 57% of adults returned to nest at colonies other than their natal ones (n = 233; Ludwig 1968). In Lakes Huron and Michigan during 1962–1986, 69.9% (n = 336) and 58.6% (n = 285), respectively, of adults were recovered outside their natal blocks (L’Arrivee and Blokpoel 1988). Because Caspian Terns do not breed until 3 yr of age, fidelity to natal colony is measured most accurately by study of colony use in this age group. Data suggest that first-time breeders tend to nest at colonies other than natal colonies. In Great Lakes, only 10% of color-banded 3-yr-olds nested at colony of hatching; 90% were recorded breeding at other locations (n = 20; Cuthbert 1981).

Fidelity To Breeding Site

Adults show strong fidelity to colony where they bred the previous year. In ne. Lake Michigan, 69% (n = 55) of terns that were followed for 2 consecutive breeding seasons nested at site of previous year’s breeding; 29% bred at one of 3 additional colony sites in same area. Previous year’s reproductive success appears to influence tenacity to colony site. Majority of birds whose nests failed the previous year nested at different colony sites the following year; >75% of the birds that nested successfully the previous year returned to previous year’s colony site. In addition, when initial nesting attempts failed, most birds abandoned the colony, and of those that renested in same season (n = 15), 87% relocated to different colony site (Cuthbert 1988).

Entire colonies readily abandon breeding sites if harassed by humans (see Conservation and management: effects of human activity, below) or predators. Colonies at Grays Harbor, WA, and on mainland site at Hamilton Harbour, Ontario, are believed to have abandoned these sites because of disturbance by Bald Eagle and red fox, respectively (D. Roby pers. comm., Quinn and Sirdevan in press).

Fidelity To Winter Home Range

No information.

Dispersal From Breeding Site Or Colony

After young fledge, family groups move to feeding areas near breeding colonies (see Migration, above, and Breeding: fledgling and immature stages, above). In Great Lakes, adults and fledglings spend 1–2 wk along mainland coastal shoreline or large inland lakes (15–70 km from colony), where they roost and forage before slowly moving south during fall migration (FJC).

Home Range

Unknown, but nesting birds may fly long distances on fishing flights. In n. Lake Michigan, marked individuals were observed foraging along mainland shoreline up to 50 km from breeding colony (FJC); on Pacific Coast, nesting individuals foraged up to 62 km from breeding colony (Gill 1976). Individuals fitted with transmitters at Lake Ontario colonies almost always foraged at least 2.5 km offshore, and large areas were believed to be necessary to support the dietary needs of this species (Sirdevan and Quinn 1997). Long-distance fishing flights also recorded for Baltic Sea colonies (Bergman 1953, Soikkeli 1973b).

Numbers

Although disjunct, North American population is largest of continental populations. Counts have varied greatly in method and extent, but estimated 33,000–35,000 pairs nested at locations across North America in 1980s and 1990s. In late 1970s–early 1980s, U.S. breeding population was estimated at about 9,500 pairs (Spendelow and Patton 1988); estimates in 1980s and 1990s show that this population has doubled.

Number of birds that winter in North America not available, but highest abundances are on eastern coast of Florida south of Cocoa Beach, and on Gulf Coast of Texas north of Houston to south of Corpus Christi (Root 1988).

Baltic population (Finland, Sweden, and Estonia) was estimated at 1,850–1,950 pairs in 1984 (Hario et al. 1987). In Afrotropical region (mainly w. Africa), a few thousand pairs; in s. Africa, about 500 pairs (Cooper et al. 1992). In New Zealand, 3,500 to <5,000 pairs; in Australia, “many thousands” (Challies 1985, del Hoyo et al. 1996).

Trends

Species most abundant on Pacific Coast, and in Manitoba on Lakes Winnipeg and Winnipegosis. Numbers appear to have increased substantially in all major breeding areas across range (see Appendix 2). Breeding Bird Survey results show significant increase (85.9%, p < 0.10) in Caspian Terns during 1966–1993, and 55.5% increase (not significant) during 1984–1993 (Price et al. 1995).

In Great Lakes, numbers have increased steadily since 1960s, nearly tripling since 1963; greatest increases on Lake Ontario (Neuman and Blokpoel 1997). Pacific Coast numbers have more than doubled since 1980, when population was estimated at 6,000 pairs (Gill and Mewaldt 1983), and population has expanded north into British Columbia and Alaska (Campbell et al. 1990, Gibson and Kessel 1992). Largest colony in North America (>10,000 pairs) is on dredge-spoil island at Rice I., OR, established in 1986; since that time, has increased by >600%. Rapid buildup in 1980s–1990s is due largely to shifting of breeding pairs from Grays Harbor, Willapa Bay, Puget Sound, WA, and East Sand I., OR, near Columbia River mouth (Roby et al. 1998). In central Canada, numbers in Manitoba have more than tripled since 1970 (see Appendix 2), and in Saskatchewan and s. Alberta new colonies were discovered in late 1980s and early 1990s (Bennett 1995, Smith 1996, M. Preston pers. comm.). On Gulf Coast, substantial increases in Louisiana, Alabama, and Florida since 1960s and 1970s (see Appendix 2). In Texas, Christmas Bird Counts show positive trend during past 35 yr, which may indicate that more terns remain during mild winters (Chaney et al. 1996). On the Atlantic Coast, substantial declines have occurred since the late nineteenth century. In Quebec, “traditional” egg harvesting nearly extirpated largest known colony along lower North Shore (Nettleship and Lock 1973, Chapeldaine 1996, M. Robert pers. comm.) On the Virginia coast, egging and harvest for millinery trade severely impacted the species (Weske et al. 1977). Numbers have not recovered in either locale. Elsewhere on Atlantic Coast, breeding colonies are small (≤100 pairs), but stable or increasing. In west-central interior, numbers may be declining in Wyoming and Utah, owing to fluctuating water levels; in Utah, human disturbance and predation by California Gulls may also impact terns (F. Howe pers. comm., A. Cerovski pers. comm.).

Reasons for increases are multifaceted and in some areas unknown (see Distribution: historical changes, above). In Great Lakes, legislation has protected pristine habitat. Most colonies are on remote, inaccessible islands, and few have been impacted by contaminants (see Conservation and management: effects of human activity, below). From late 1930s–1950s dramatic declines occurred in large predatory fish populations (Weseloh and Collier 1995), and by the late 1950s smaller fish species, such as alewife and smelt became an abundant and steady food source (Ludwig 1965). New colonies recently established on artificial sites at Hamilton Harbour (Ontario) and Saginaw Bay (MI; Blokpoel and Scharf 1991). On Pacific and Gulf Coasts, artificial habitat provides quality breeding sites; many colonies now found on dredge-spoil islands and salt dikes (Stadtlander et al. 1993, Parkin 1998, Roby et al. 1998). On Lake Winnipegosis, Manitoba, increases may be related to increases in forage fish abundance, resulting from excessive exploitation of large predatory fish (Hobson et al. 1989).

Outside of North America, significant declines in Baltic in Sweden and Finland, and in Denmark, West Germany, East Germany, Romania, and Tunisia (Cramp 1985, Hario et al. 1987). Decline in Baltic population may be due to higher mortality on African wintering grounds, which experienced severe drought conditions in 1970s and 1980s (Hario et al. 1987). Heavy egg-collecting in 1800s greatly diminished colonies in Denmark, Romania, and Danube River delta (Cott 1953). In s. Africa, classified as “rare” (Brooke 1984), but temporal trends hard to determine; populations in this area may move between breeding localities, or in some years large proportions of adult population may not breed (Cooper et al. 1992).

The primary factor limiting populations appears to be availability of high-quality nest sites protected from storms and free of mammalian and avian predators (Penland 1976, Shugart et al. 1978, Cuthbert 1981). Caspian Terns require nest sites on beaches high enough to avoid flooding and washouts during storms that generate large waves. Earlier nesting by gulls sometimes forces terns into lower, less safe areas, and limits the number of terns that can nest successfully, or prevents some birds from nesting at all (Penland 1976, Shugart et al. 1978). Human-created habitat providing high-quality nest sites is associated with population increases throughout North America (see Population status, above). Although many colony sites are isolated from predator populations, at some locations Caspians have experienced high mortality due to avian predators (Bald Eagles and Great Horned Owls) and mammals, especially canids. Scarce food resources during chick-rearing periods (Soikkeli 1973a, Quinn 1980, Pierce 1984) and drought conditions at wintering sites result in higher-than-usual mortality rates and may limit populations (Hario et al. 1987).