Demography and Populations

Age At First Breeding; Intervals Between Breeding

Originally reported as 3 yr (Sykes 1979, Nichols et al. 1980). More recently as first year following fledging (Beissinger 1986, Snyder et al. 1989a). Approximately 22 and 17% of yearlings attempted nesting in 1979 (Snyder et al. 1989a) and 1992 (Bennetts and Kitchens 1992), respectively. Snyder et al. (1989a) suggested that most Snail Kites are potential breeders by second year.

Generally assumed that adults attempt to breed every year except during droughts (Nichols et al. 1980, Snyder et al. 1989a), and may attempt to breed more than once per year (Beissinger 1986, Beissinger and Snyder 1987, Snyder et al. 1989a, Bennetts and Kitchens 1992). Many birds may not breed during drought years (Sykes 1979, Nichols et al. 1980, Beissinger 1986, Snyder et al. 1989a).

Clutch

Range 1–5 since 1960s (see also Breeding: eggs). Mean clutch size in Florida 2.9, 2.5, 2.7, and 2.8 reported by Sykes (1987c), Bennetts et al. (1988), Snyder et al. (1989a), and J. A. Rodgers (n = 796 clutches, unpubl. data), respectively. Clutch sizes, as reported by Sykes (1987c), Bennetts et al. (1988), and Snyder et al. (1989a), respectively — 1-egg: 2, 6, and 1%; 2-egg: 17, 32, and 33%; 3-egg: 80, 61, and 65%; and 4-egg, all 1%. J. A. Rodgers (unpubl. data from Florida) found clutch sizes (1980–1993) as follows — 1 egg: 1.0% (n = 8); 2 eggs: 21.7% (n = 173); 3 eggs: 75.0% (n = 597); 4 eggs: 2.1% (n = 17); and 5 eggs: 0.1% (n = 1).

Larger clutches (3) tend to be in nests with higher male courtship feeding rates (Beissinger 1987a). Clutch size not affected by annual changes in water levels, seasons, or nesting in colony versus solitarily, but is larger on Lakes Okeechobee, Kissimmee, and Tohopekaliga than in Water Conservation Area (WCA) 3A (Snyder et al. 1989a).

Often renests following failed attempt as well as after successful attempt (Beissinger 1986, Snyder et al. 1989a, Bennetts and Kitchens 1992); actual number of clutches/season has not been documented, however. Snyder et al. (1989a) suggested an average of 2.7 nesting attempts/pair in 1978, but this estimate included nesting attempts in which eggs were not laid and may not represent actual number of clutches laid. Their estimate also required making several strong assumptions (e.g., that a statewide count represented an accurate census of the population) and was estimated during only 1 yr (1978) when overall productivity was extremely high (see below, Population Status: numbers). Consequently, these conditions may have influenced the validity of their estimate (Bennetts and Kitchens 1993).

Annual Reproductive Success

Hatching Success. Sykes (1987c) reported 57.4% (204 of 355 eggs laid) overall hatching success (survival of eggs to hatching) from 1968 to 1978. Bennetts et al. (1988) found hatchability (hatching success of eggs that have survived to time of hatching) to be 80.7% (352 of 436 eggs) for 1986 and 1987. Snyder et al. (1989a) reported hatchability of 80.9% (402 of 497 eggs) for 1966–1983. Overall hatching success was 58.0% (n = 609 nests); hatching success of eggs in nests that hatched at least 1 egg was 94.4% (JAR).

Nest Success. Unless otherwise stated, values for nest success (a nest in which at least 1 chick fledged) are herein based on nests in which at least 1 egg was laid. Sykes (1979) reported average nest success for 1968–1976 of 48%; range 17.1 (1974)–84.6 (1968). Later, Sykes (1987c) included 2 additional years (1977–1978) and revised average to 50% (includes some manipulated nests, i.e., nests placed in baskets on poles). Nests that fledged at least 1 bird: 41.7% (n = 938 nests; JAR). Bennetts et al. (1988) reported nest success of 30 and 46% for 1986 and 1987, respectively. Estimates of nest success derived using Mayfield’s method (Mayfield 1961, 1975) for those same years were 23 and 36%, respectively (approximately 20–25% lower than observed estimates). Snyder et al. (1989a) reported average nest success of 32% for the period 1968–1983, and 22.8% for all nests (including “pre-laying” nests).

Considerable variability in nest success among years, locations, and local nest environments (Sykes 1979, 1987c, Beissinger 1986, Bennetts et al. 1988, 1994, Snyder et al. 1989a). Sykes (1987c) attributed much of variability to water levels. Snyder et al. (1989a) found differences in success between years of high and low water (including a 1-yr lag time). These are driven by differences in nest site selection (between woody and herbaceous substrates) and water depth. Bennetts et al. (1988) found no affects for “within year” water depth during relatively high-water years. Differences reported in nesting success related to water levels tend to emphasize “between year” differences of relatively low- and high-water years (i.e., drought vs. nondrought years) rather than “within year” differences among sites. Drought or low-water conditions may influence nest success via depressed apple-snail populations (Kushlan 1975, Beissinger and Takekawa 1983), increased access by terrestrial predators (Beissinger 1986), or structural failure of nests placed in herbaceous vegetation (Sykes and Chandler 1974, Beissinger 1986, Snyder et al. 1989).

Other weather factors may influence nest success. Wind may cause structural collapse or tilting of nests, with loss of eggs or young (Chandler and Anderson 1974, Sykes and Chandler 1974, Beissinger 1986, Sykes 1987c, Bennetts et al. 1988). On 13 Mar 1993 (the “storm of the century”), 26 of 51 nests (51.0%) on Lake Okeechobee collapsed, and 11 of 51 nests (21.6%) were abandoned. Total of 72.6% of nests on Lake Okeechobee failed (JAR). Cold weather can stop nest-building (Beissinger 1986, Bennetts et al. 1988, Snyder et al. 1989a), probably owing to decreased availability of apple snails (Cary 1985) or mortality of young from exposure (Sykes 1987c). Heavy rain may cause direct mortality of young through exposure or decreased foraging ability of tending adults (Sykes 1987c), but nesting success may be higher in wet years (Beissinger 1995).

Substrate in which nest is placed may influence susceptibility to collapse. Snyder et al. (1989a) found higher success in woody versus herbaceous substrates. Bennetts et al. (1988) and Snyder et al. (1989a) found no difference among woody substrates. Snyder et al. (1989a) considered structural collapse (12.6–30.8%, 1978–1983 WCA 3A) the most important cause of nest failure. J. A. Rodgers (unpubl. data) found 16.3% of nests collapsed (n = 1,016 nests). Use of herbaceous substrates occurs primarily on lake habitats and may be influenced by water levels. This cause of failure is less frequent in Everglades and other nonlake habitats (Snyder et al. 1989, Bennetts et al. 1994a).

Predation is a frequent cause of nest failure after eggs have been laid (Sykes 1979, 1987b, c, Bennetts et al. 1988, Snyder et al. 1989a; see Behavior: predation). Susceptibility to predation may vary with locality, habitat, or combinations of habitats. Nest height could influence predation by cottonmouth moccasins, which are poor climbers (Bennetts et al. 1988).

Abandonment of nests before egg-laying is common, particularly during drought or passage of cold fronts (Beissinger 1986, Snyder et al. 1989a). Such abandonments are often temporary delays in courtship rather than termination of a nesting attempt (Bennetts et al. 1988, 1994a). Abandonment of eggs or young is rare, except during drought (Beissinger and Snyder 1987, Sykes 1987b, Bennetts et al. 1988).

Chandler and Anderson (1974) reported 31 young fledged from 82 eggs laid (37.8%) at Lake Okeechobee (most nests placed in artificial structures) from 1966 to 1973. During 1986 and 1987, 237 young fledged from 819 eggs (28.9%; REB).

Sykes (1979) reported 161 young fledged from 183 nests (0.88 young/nest) from 1968 through 1976 (includes 25 nests manipulated using nest baskets at Lake Okeechobee, and may be biased high). Considerable interyear and intersite variation, but overall combined fledging success (to nestling age of 6 wk) 0.83 fledgling/nest, or 0.29 fledgling/egg (n = 776 nests, 1980–1993, no nest manipulation; JAR). Bennetts et al. (1988) estimated 0.44 and 0.76 young/nest for 1986 and 1987 (no nest manipulation), respectively. All these reports were of nests containing at least 1 egg. Thus, these estimates are equivalent to young per clutch, but not per pair since the makeup of pairs can change because of mate desertion or renesting (see also Snyder et al. 1989a). J. A. Rodgers (unpubl. data) modeled Snail Kite nestling survivorship and found at 57–63 d of age (fledging age) survivorship per egg to be 0.27 (JAR).

Number of Young per Successful Pair. Average values reported for 1968–1978: 1.9 (Sykes 1979, 1987c); for 1966–1983: 1.96 (Snyder et al. 1989a). Beissinger (1986) reported 2.0 for 1979, 1982, and 1983, and Bennetts et al. (1988) reported 1.44 and 1.65 for 1986 and 1987, respectively.

Lifetime Reproductive Success

No information.

Number Of Broods Normally Reared Per Season

Snyder et al. (1989a) suggested that multiple brooding was widespread, with individuals potentially raising 4 broods/yr, but documented only 3 cases of raising 2 broods successfully in a given year and no cases of individuals raising > 2 broods. Similarly, Bennetts and Kitchens (1992) and REB found (using radio telemetry) that some individuals may successfully raise 2 broods, but found no evidence for successful rearing of > 2 broods.

Proportion Of Total Females That Rear At Least One Brood To Nest-Leaving Or Independence

No information.

In the wild, life span at least 9 yr (Sykes 1979), 11–13 yr (Beissinger 1986), 17 yr (Bennetts and Kitchens 1993). Beissinger (1986) suggested that average adult in Florida lived to be 5–8 yr old, owing to presumed high mortality during droughts. At least 22 additional individuals of known-age ≥ 13 yr old have been observed (REB), indicating that this species may commonly live to at least 13 yr.

Nichols et al. (1980) estimated that adult and juvenile survival of Snail Kites is 0.90 and 0.57, respectively. These were admittedly “best guesses” for the purpose of demographic modeling and were not based on empirical data. Snyder et al. (1989a) suggested average adult survival probably exceeds 90% under good conditions. During 1979, all of 13 juveniles equipped with radio transmitters survived at least 1.5 yr (Beissinger 1986, Snyder et al. 1989a, b). Bennetts and Kitchens (1993) reported preliminary survival estimates of 98.7% for adults (n = 45) and 88.6% for juveniles (n = 37) for 1992, based on radio telemetry; estimates probably biased high owing to some assumptions of censusing lost birds.

Survival may be lower during periods of drought (Beissinger 1986, Snyder et al. 1989a, Takekawa and Beissinger 1989), although few data exist for empirical estimates. Beissinger (1984, 1986, 1988) and Takekawa and Beissinger (1989) reported a population decline from 650 to 250 birds (38% survival) during a 1981 drought based on differences between the 1981 and 1982 annual counts; however, the validity of using the annual count for this purpose is questionable (see below, Population Status: numbers). Snyder et al. (1989a, b) reported survival of at least 7 of 8 (88%) radio-transmittered adult Snail Kites from the end of drought to the following spring. Beissinger (1986) reported that only 2 of 4 young survived 2 mo postfledging during the 1981 drought at Lake Okeechobee.

Information limited. Nestlings 5–8 wk old from Lake Okeechobee died of acute pneumonia of unknown etiology (Kerr 1977). Fledgling found dead at E. Lake Tohopekaliga infected with Aspergillus sp. (possibly A. fumigatus) but did not have aspergillosis (N. J. Thomas, R. A. Cole, D. J. Forrester and M. G. Spalding unpubl. data). Blood smears from 1 fledgling and 40 nestlings, 1973–1993, were all negative for parasitic protozoans, and 2 of 20 fecal samples were positive for oocysts of intestinal coccidia and probably represent a species of Eimeria, the significance of which is not known for Snail Kites (Sykes and Forrester 1983, D. J. Forrester and M G. Spalding unpubl. data). Four species of trematodes known from Snail Kites, but only Bothrigaster variolaris reported in Florida from bronchi and air sacs of juvenal bird (Travassos 1923, Dubois 1959, D. J. Forrester and M. G. Spalding unpubl. data).

Parasitic arthropods known to infest Snail Kites in Florida include mites (Ornithonyssus bursa and unidentified mites), chewing lice (Colpocephalum turbinatum, Falcolipeurus guadriguttatus, and Craspedorrhynchus obscurus), and mosquitoes (Anopheles crucians, A. walkerii) (Malcomson 1960, Sykes 1983c, Sykes and Forrester 1983, Bennetts et al. 1988, D. J. Forrester and M G. Spalding unpubl. data). When mites and mosquitoes occur in large numbers, they harass nestlings (can cause mortality), but lice are probably of little significance (D. J. Forrester and M G. Spalding unpubl. data). Larval dermestid beetles (Dermestes nidum) feed on some nestlings, creating abdominal lesions, which later heal; most affected have survived (Snyder et al. 1984). Botflies (species unknown) occasionally parasitize Snail Kite nestlings in Venezuela (S. R. Beissinger pers. comm.).

Exposure

No information on adults (but see above, Measures of Breeding Activity: annual reproductive success).

Predation

Predation is a primary cause of death for nestlings (see Behavior: predation). Although overall rates are low, predation may be a major cause of adult mortality; e.g., Great Horned Owls (Bennetts and Kitchens 1993, REB).

Competition With Other Species

Snyder and Snyder (1969) studied comparative foraging behavior of 2 potential competitors. Their results were inconclusive regarding the degree of interspecific competition among Limpkins, Boat-tailed Grackles, and Snail Kites. They report some antagonistic interactions among these species, generally initiated by Boat-tailed Grackles, and probably a form of interference competition. Limpkins, more dependent on apple snails than Boat-tailed Grackles, generally feed in different habitat (denser vegetation) than Snail Kites.

Emaciation

Probably a common cause of death for newly independent or dispersing juveniles (Bennetts and Kitchens 1992). Drought (leading to starvation) is likely the root cause of adult mortality (Beissinger 1986, 1988).

Disease And Parasites

Disease or parasites may occasionally contribute to, or cause, death of nestlings (see above, Disease and Body Parasites).

Shooting

Occurs occasionally, particularly during waterfowl hunting season (Nov–Jan) (Sprunt 1945, Stieglitz and Thompson 1967, Sykes 1978, 1979, REB).

Initial Dispersal From Natal Site

Little information. At least 7 of 37 juveniles (19%) dispersed from their natal site during first 4 mo after fledging, whereas at least 11 (30%) remained during this period (Bennetts and Kitchens 1992). One juvenile was recovered 225 km northwest of its natal site within 1 mo after fledging (Bennetts and Kitchens 1992).

Fidelity To Breeding Site And Winter Home Range

Little information. Some individuals show fidelity to breeding location, but exceptions are common (REB). Also may attempt to breed at >1 location within a given year (Beissinger 1986, Bennetts and Kitchens 1993).

No information about fidelity to wintering areas, but birds often move southward in Florida peninsula during cold winters (Sykes 1983a).

Dispersal From Breeding Site

Some movement of Snail Kites following nesting; varies with year and locale (Sykes 1978, Beissinger et al. 1983, Bennetts and Kitchens 1993).

Home Range

In Florida, semi-nomadic (Sykes 1978, 1979, Beissinger and Takekawa 1983, Bennetts 1993) and may move throughout range within state over lifetime (Bennetts and Kitchens 1992, 1993, Bennetts 1993; also see Fig. 1). Thus, a long-lived bird’s home range could potentially be roughly the geographic range of species in Florida.

Beissinger et al. (1983) hypothesized that Snail Kites may move between Florida and Cuba; this is based on the apparent “disappearance” of birds from Florida during late summer, distances moved by banded birds in Florida, large decrease in numbers observed directly after a drought, and Amadon’s (1975) revision of subspecies. Although possible, there has been no evidence to support this hypothesis. During a 1-wk search of Zapata Swamp, Cuba, 55 Snail Kites were observed but no banded birds were seen (Beissinger et al. 1983). Approximately 350 Snail Kites banded in Florida during 5 yr preceding their expedition (Beissinger et al. 1983). Bennetts and Kitchens (1993) also found that movements of Snail Kites in Florida during this period of “disappearance” often were to habitats in the state peripheral to major breeding locations.

Numbers

Annual count (ground; mainly with airboat) conducted in Florida since 1969 (Sykes 1979, 1982, Rodgers et al. 1988, Bennetts and Maier 1991, Bennetts et al. 1992, 1993, 1994a). This is a quasi-systematic (dense vegetation often precludes a completely systematic search) count which includes combination of transect and roost counts (Sykes 1982, Rodgers et al. 1988).

Counts since 1969 have ranged from 65 in 1972 (Sykes 1983b) to 996 in 1994 (Bennetts et al. 1994a; see Fig. 7). However, Bennetts et al. (1993, 1994a) caution that 1993 and 1994 counts were conducted when numerous radio-transmittered birds (about 100) were being monitored, which may have influenced totals (radio-transmittered birds led to previously unknown roost sites).

Beissinger (1995) states that percentage change from year to year in Snail Kite numbers on the counts is strongly positively related to mean and minimum spring and summer water levels, positively correlated with total annual and winter rainfall, strongly correlated with length of breeding season, and in close agreement with predicted population sizes from demographic analyses using parameters that were partly derived from the counts. Year-to-year change in numbers is weakly related to water levels and rainfall at time of the counts and nesting success. Together, length of breeding season and nesting success account for 74% of year-to-year variation in percentage change in numbers.

Annual count has been widely cited and interpreted in a variety of contexts (e.g., Sykes 1979, Beissinger 1984, 1986, 1995, Bennetts et al. 1988, 1994a, Rodgers et al. 1988, Snyder et al. 1989a, Takekawa and Beissinger 1989)—including being used as an index to the use of surveyed wetlands (e.g., Rodgers et al. 1988) and as a population census (e.g., Sykes 1979, Snyder et al. 1989a)—but it has limitations. Rodgers et al. (1988) point out that annual changes in the count can result from changes in survival, dispersal (often into areas not counted), or recruitment, and that we are unable to distinguish these effects. Additional problems include observer effects, detectability in different habitats, weather, and lack of confidence intervals.

Trends

Reports of Snail Kite numbers in Florida during late 1800s and early 1900s (reviewed by Sykes 1984) consistently indicated that species was locally abundant in a few localities (Scott 1881, Wayne 1895, Howell 1932). Beginning in 1930s and continuing through mid-1960s, virtually all reports indicated severe declines in numbers and distribution (Howell 1932, Bent 1937, Sprunt 1945, 1954, Wachenfeld 1956, Stieglitz 1965). Sprunt (1945) estimated that in early 1940s between 50 and 100 Snail Kites remained in Florida, later revised to 50–75. Wachenfeld (1956) and later Stieglitz (1965) suggested that ≤ 20 birds remained. Although these estimates were undoubtedly influenced by the difficulty of access to major habitats, they consistently indicated low numbers.

The annual count since 1969 has indicated a generally increasing trend (Sykes 1979, Rodgers et al. 1988, Bennetts et al. 1994a). Deterministic life-table approach by Nichols et al. (1980) with low estimated finite rate of increase now out of date as current evidence indicates high productivity and survivorship. There have been sharp declines in numbers in some years since 1967 (e.g., 1981, 1985, 1987), but it is not known to what extent this reflects actual changes in population (see also Numbers, above). Declines generally occur during dry years, but have occurred in some wet years as well.

Hydrology (water level) is probably directly or indirectly the primary environmental influence on Snail Kite populations in Florida (Stieglitz 1965, Stieglitz and Thompson 1967, Sykes 1979, 1983a, Beissinger 1986, 1995, Bennetts et al. 1988, 1994a, Snyder et al. 1989a). Influences may include direct loss of habitat via drainage (Stieglitz and Thompson 1967, Sykes 1978, Beissinger 1988), effects on apple-snail populations (Kushlan 1975), effects on reproduction (Nichols et al. 1980, Snyder et al. 1989a), effects on survival (Beissinger 1986, Takekawa and Beissinger 1989), and changes in habitat structure (Bennetts et al. 1988, 1994a).

Long-term trends in numbers and distribution generally coincide with hydrologic regimes. Reports of declining numbers and range reductions coincide with periods of major drainage in Florida (Howell 1932, Bent 1937, Sprunt 1945, 1954, Wachenfeld 1956, Stieglitz 1965). Distribution in recent decades also coincides with areas of relatively long hydroperiod (Stieglitz and Thompson 1967, Bennetts et al. 1988).

Recently, Beissinger (1995) suggested factors that regulate Snail Kite population size: water levels (i.e., average, minimum), frequency and strength of dry-out or drought events (Snail Kite population viability is likely to be highly influenced by interdrought intervals); and duration of nesting season.