A small (4 ½ -5 ¾ inches) bunting, the Savannah Sparrow is most easily identified by its mottled brown back, streaked breast and belly, and yellow eye-stripes. Numerous geographic races exist in this species’ wide range, some much paler or darker on the back and head than the typical race. Male and female Savannah Sparrows are similar to one another in all seasons. The Savannah Sparrow breeds across Alaska, Canada, and the northern half of the United States. In winter, most populations migrate south to the southern half of the United States, Mexico, and parts of Central America. Populations breeding in coastal California, Baja California, and central Mexico are non-migratory. Savannah Sparrows breed in open and semi-open habitats, including grassland, marshes, agricultural fields, and tundra. Similar habitat types are utilized in winter as in summer. Savannah Sparrows primarily eat insects in summer, switching over to seeds and grains during the winter. In appropriate habitat, Savannah Sparrows may be observed foraging for food on the ground below shrubs and grasses. Birdwatchers may also listen for this species’ song, a whistling “tsit-tsit-tsit tseee-tsaaay,” in order to separate it from other drab grassland birds. Savannah Sparrows are most active during the day, but, like many migratory songbirds, this species migrates at night.
- Wheelwright, N. T. and J. D. Rising. 2008. Savannah Sparrow (Passerculus sandwichensis), The Birds of North America Online (A. Poole, Ed.). Ithaca: Cornell Lab of Ornithology; Retrieved from the Birds of North America Online: http://bna.birds.cornell.edu/bna/species/045
- Peterson, Roger Tory. Birds of Eastern and Central North America. Boston: Houghton Mifflin, 1980. Print.
- eBird Range Map - Savannah Sparrow. eBird. Cornell Lab of Ornithology, N.d. Web. 20 July 2012. http://ebird.org/ebird/map/savspa.
- Passerculus sandwichensis. Xeno-canto. Xeno-canto Foundation, n.d. Web. 20 July 2012. http://xeno-canto.org/browse.php?query=Passerculus+sandwichensis.
- Savannah Sparrow (Passerculus sandwichensis). The Internet Bird Collection. Lynx Edicions, n.d. Web. 20 July 2012. http://ibc.lynxeds.com/species/savannah-sparrow-passerculus-sandwichensis.
occurs (regularly, as a native taxon) in multiple nations
Regularity: Regularly occurring
Type of Residency: Breeding
Regularity: Regularly occurring
Type of Residency: Year-round
Global Range: (>2,500,000 square km (greater than 1,000,000 square miles)) BREEDING: northern Alaska and northern Canada south of arctic islands east to northern Labrador and Newfoundland, south through the western U.S. and Mexico (locally) to southwestern Guatemala, and south to southern Iowa and New Jersey (AOU 1983). NON-BREEDING: southern British Columbia, southern Nevada, Gulf states, and Massachusetts to northern Honduras and West Indies (AOU 1983). RESIDENT: Pacific Coast from central California south to Baja California Sur and Gulf of California Coast from Sonora south to central Sinaloa (Wheelwright and Rising 1993). Accidental in Hawaii.
Length: 14 cm
Weight: 25 grams
Differs from song sparrow (MELOSPIZA MELODIA) by usually having a yellowish eyebrow stripe, whitish crown stripe, short notched tail (vs. long and rounded), and pinker legs (Peterson 1990). Lacks the white outer tail feathers of the vesper sparrow (POOECETES GRAMINEUS).
Comments: BREEDING: Prefers habitat with short to intermediate vegetation height, intermediate vegetation density, and a well developed litter layer. These preferred habitats cover a wide range of vegetation types, including alpine and arctic tundra, coastal salt marshes, sedge bogs, grassy meadows, and native prairie (Wheelwright and Rising 1993).
In North American grasslands, occupies tallgrass prairie, idle and lightly grazed mixed-grass prairie, shortgrass, wet meadow zones surrounding prairie wetlands, alfalfa (MEDICAGO SATIVA)/brome (BROMUS spp.) hayfields, native and tame dense nesting cover (DNC), Conservation Reserve Program lands, weedy crop and stubble fields, retired cropland, and wheat fields (Stewart 1975; Salt and Salt 1976; Renken 1983; Dale 1993; Hartley 1994; Johnson and Igl 1995; Patterson and Best 1996; Prescott and Murphy 1995, 1996). Although occasionally breeds in cropland, is more abundant in idle native, DNC, and Conservation Reserve Program lands (Hartley 1994, Johnson and Igl 1995, Patterson and Best 1996). In Illinois, found in mixed and pure stands of hay, pastures, and idle grasslands, but reached highest densities in pastures and idle grasslands (Graber and Graber 1963). In Wisconsin and Ohio, abundance was positively correlated to percent herbaceous vegetation cover (Sample 1989, Swanson et al. 1997). In North Dakota and Saskatchewan, abundance was positively correlated with percent grass cover (Renken 1983, Renken and Dinsmore 1987, Sutter 1996). However, at a less arid site in Saskatchewan, abundance was negatively correlated to percent grass cover (Sutter 1996).
Most abundant on Conservation Reserve Program fields with high percent grass and low percent legume cover in the northern Great Plains (Johnson and Schwartz 1993). In Oregon and Nevada, Rotenberry and Wiens (1980) found a positive correlation between abundance and percent forb cover. In Wisconsin, abundance was negatively correlated to maximum vegetation height and vegetation height-density (Sample 1989). Wiens (1969, 1973) stated that low, dense vegetation was required for nest sites; grass-dominated habitats with little forb cover are preferred (Wiens 1969, 1973; Welsh 1975; Knight 1989; Vickery et al. 1992). In Wisconsin, avoided habitats with tall, dense vegetation and nested primarily in managed or disturbed habitats such as pastures and hayfields (Sample 1989). In Michigan, nested in hayfields of clover (MELILOTIS and TRIFOLIUM spp.), alfalfa, brome (BROMUS spp.), and timothy (PHLEUM PRATENSE), and in clumps of grass near cow pies in an overgrazed pasture (George 1952). In Maine, nesting birds in areas of mainly forb and shrub cover experienced lower reproductive success than those nesting in predominantly grass cover (Vickery et al. 1992). In Quebec, vegetation height did not differ between nest sites and random points, but successful nests were surrounded by taller vegetation than unsuccessful nests (Bedard and LaPointe 1984).
Avoids areas with extensive tree cover (Wheelwright and Rising 1993). Wiens (1969) noted that most breeding territories were located in the center of grassland habitats, away from cultivated fields and fence lines, and Sample (1989) found a negative correlation between abundance and percent shrub cover. In North Dakota, were found only on shrubless transects (Arnold and Higgins 1986). In Wyoming sagebrush steppe, were observed only on burned and herbicide treated areas with fewer shrubs and more grass and forb cover (Kerley and Anderson 1995). In West Virginia, nesting territory often included small trees, shrubs, and fence posts (Shields 1935). In Saskatchewan, nested in or near clumps of sparse western snowberry (SYMPHORICARPOS OCCIDENTALIS) shrubs (Lein 1968). Although total woody cover in habitats used for nesting was low (< 1 percent), Sample (1989) reported that the birds often used small trees and shrubs (< 2 m tall), fence posts and wire, and tall herbaceous stems as song perches.
NON-BREEDING: In migration, open fields, roadsides, dunes, coastal marshes, edges of ponds, and rarely in open woodlands (Wheelwright and Rising 1993). In winter, cultivated fields, pastures, golf courses, roadsides, dunes, and salt marshes (Wheelwright and Rising 1993).
Habitat and Ecology
Non-Migrant: Yes. At least some populations of this species do not make significant seasonal migrations. Juvenile dispersal is not considered a migration.
Locally Migrant: Yes. At least some populations of this species make local extended movements (generally less than 200 km) at particular times of the year (e.g., to breeding or wintering grounds, to hibernation sites).
Locally Migrant: Yes. At least some populations of this species make annual migrations of over 200 km.
Breeding populations throughout Canada and most of U.S. are long-distance migrants; arrive on nesting grounds in the northern and central U.S. and Canada in March-April, in far north late May-early June; departs far northern nesting areas usually by late August (Johnson and Herter 1989). Begins to arrive in the southeastern U.S. during the last week of September, with numbers increasing throughout the fall and reaching peak densities by mid-December; numbers are relatively stable until late March, when an abrupt decline begins; only a few scattered birds remain through late April and early May. Nonmigratory or local migrant in parts of west coast. Subspecies ROSTRATUS breeds along north and east coasts of the Gulf of California, many migrate north to winter in southwestern California (Zink et al. 1991). Subspecies PRINCEPS breeds on Sable Island, Nova Scotia, and winters on dune beaches of the Middle Atlantic states (McLaren, 1979 COSEWIC report).
Comments: During summer eats insects, spiders, and snails; breeding adults ate average of 18-21 g of fresh arthropods daily (Williams 1987). Feeds on seeds of primarily herbaceous plants at other times of year. Forages on ground, sometimes scratches. Adults feed arthropods to young (Meunier and Bedard 1984).
Number of Occurrences
Note: For many non-migratory species, occurrences are roughly equivalent to populations.
Estimated Number of Occurrences: 81 to >300
10,000 to >1,000,000 individuals
Territories are small, ranging from 0.05 to 1.25 hectares (George 1952, Lein 1968, Wiens 1969, Potter 1972, Welsh 1975, Piehler 1987, Wheelwright and Rising 1993).
Life History and Behavior
Lifespan, longevity, and ageing
Arrive on the breeding grounds between late March and early May, and begin nesting in May (George 1952, Baird 1968, Lein 1968, Maher 1973, Welsh 1975, Bedard and LaPointe 1984b, Wheelwright and Rising 1993). In North Dakota, breed from late May through late July, with peak breeding occurring from early June to mid-July (Stewart 1975). If the first nesting attempt fails, will renest, and many females produce a second clutch after a successful first nest (George 1952, Lein 1968, Taber 1968, Wiens 1969, Weatherhead 1979, Wheelwright and Rising 1993). In northern areas, may be limited to a single brood (Maher 1973, Weatherhead 1979).
In Saskatchewan, departed from breeding territories in early August, but remained in the area, foraging in weedy fields, along road edges, and along the margins of lakes and sloughs (Lein 1968). Fall migration occurs in mid- to late September (George 1952, Maher 1973).
In the Canadian Maritimes, male and female are philopatric (Bedard and LaPointe 1984b, Wheelwright and Rising 1993). Philopatry to breeding territories in Newfoundland is high; 95 percent of surviving males and 90 percent of surviving females return to within 40 meters of territories from a previous year (Wheelwright and Rising 1993). There is no information on philopatry in the Great Plains, and philopatry of grassland birds in the Great Plains may be low (Wiens 1974, Johnson 1996). Polygyny is common in populations residing in the Canadian Maritimes, but populations in other portions of the species range are generally monogamous (Welsh 1975, Wheelwright and Rising 1993).
Molecular Biology and Genetics
Barcode data: Passerculus sandwichensis
There are 22 barcode sequences available from BOLD and GenBank. Below is a sequence of the barcode region Cytochrome oxidase subunit 1 (COI or COX1) from a member of the species. See the BOLD taxonomy browser for more complete information about this specimen and other sequences.
-- end --
Download FASTA File
Statistics of barcoding coverage: Passerculus sandwichensis
Public Records: 22
Specimens with Barcodes: 25
Species With Barcodes: 1
National NatureServe Conservation Status
Rounded National Status Rank: N5B - Secure
Rounded National Status Rank: N5B,N5N : N5B: Secure - Breeding, N5N: Secure - Nonbreeding
NatureServe Conservation Status
Rounded Global Status Rank: G5 - Secure
Reasons: Very large range, relatively common.
IUCN Red List Assessment
Red List Category
Red List Criteria
Comments: The frequency of brood parasitism by Brown-headed Cowbirds (MOLOTHRUS ATER) appears to be low across much of the range (Wheelwright and Rising 1993). However, high rates of brood parasitism have been reported. In Manitoba, 33 percent of nests were parasitized (Davis 1994, Davis and Sealy in press). In Saskatchewan, 44 percent of nests contained > 1 cowbird egg (SWCC 1997).
Restoration Potential: Plant fields with mixture of forbs and short and tall grasses (Jones and Vickery 1997).
Preserve Selection and Design Considerations: May occupy small (< 5 ha) areas of suitable habitat (Potter 1972). However, Jones and Vickery (1997) suggest that minimum grassland size is 20-40 acres and Herkert et al. (1993), based on data from Illinois, categorized the species as highly sensitive to habitat fragmentation. Species with a high area sensitivity are generally restricted to nesting in large, contiguous habitats and rarely nest in small habitat fragments. In New York, density was positively correlated with field size (Bollinger 1988, 1995), whereas in Saskatchewan, occurrence was negatively associated with area (SWCC 1997). In Illinois, Herkert (1991) found no sparrows on grassland tracts < 10 ha and noted that they were significantly more likely to occur on large than small grasslands. Incidence increased with area and reached 50 percent at about 10 ha in Maine (Vickery et al. 1994) and 40 ha in Illinois (Herkert 1994).
In Saskatchewan, 700-1,600 ha may be required to halve brood parasitism (SWCC 1997). In Minnesota tallgrass prairie, abundance was higher, and nest depredation and cowbird brood parasitism rates lower, as distance to wooded edges increased; nest depredation rates were lower on large (130-486 ha) than small (16-32 ha) grassland fragments (Johnson and Temple 1986, 1990).
Management Requirements: BURNING: In Minnesota, Wisconsin, and South Dakota, higher nest densities were found in unburned (for > 12 months) than burned grasslands (Tester and Marshall 1961, Martin 1967, Halvorsen and Anderson 1983, Huber and Steuter 1984). Halvorsen and Anderson (1983) attributed the difference in nesting densities between burned and unburned areas to the lack of litter on burned areas required for nesting. In Minnesota, nesting was significantly correlated with litter cover and litter depth, and may require > 2 years of litter accumulation postfire before using a grassland for nesting (Tester and Marshall 1961). In North Dakota, reached highest densities 1-5 years (Johnson 1997) and 6-7 years postfire (Madden 1996).
Common on an idle control (Johnson 1997), and, although they reached their highest density 6-7 years postfire, no significant response to fire (Madden 1996). Fire reduced abundance on native fescue (FESTUCA ALTAICA) for 3 years in Saskatchewan (Pylypec 1991). However, in Illinois, preferred recently burned grassland for nesting (Herkert 1994). Densities were highest in grasslands the first growing season postfire, lower on grasslands the second growing season postfire, and were not encountered > 3 growing seasons postfire (Herkert 1994).
MOWING/HAYING: In Alberta, nesting densities were highest in undisturbed (>3 growing seasons since last mowing) grasslands, but they also nested in grasslands that had been mowed the previous summer (Owens and Myres 1973). In Saskatchewan, more abundant in tame hayland idled for > 3 years than in annually hayed tame vegetation or in idle native prairie (Dale 1992, Dale et al. 1997). In Manitoba, most abundant and most productive in hayland compared to native grassland, tame grasslands, and woodlands; productivity was evaluated using a behavioral index that ranked behavior indicative of nestling or fledgling care (Jones 1994). Abundance was highest in recently (within 4 months of 1 May) mowed grasslands, which provided the low-to-medium vegetation height preferred by this species (Herkert 1991). In Michigan, abundance was reduced after hayfields were mowed, and mowing exposed nests and young to predators (George 1952). On seven rural Illinois airports, mowing destroyed 44 percent of all grassland bird nests, and Savannah Sparrow nest success was 23 percent (Kershner and Bollinger 1996).
GRAZING: In short- and mixed-grass prairie, moderately and heavily grazed areas do not provide the dense ground cover required by nesting birds (Dale 1983). Several studies conducted in the Great Plains reported higher nesting densities on idle or lightly grazed grasslands compared to moderately or heavily grazed areas (Owens and Myres 1973; Maher 1974,1979; Kantrud 1981; Kantrud and Kologiski 1982; Anstey et al. 1995). For example, average breeding densities were 7.7 pairs per hectare in ungrazed versus 0.3 pair per hectare in grazed grasslands (Maher 1979). In Tennessee, no longer used a grassy field after it was heavily grazed (Knight 1989). In Idaho, used both grazed and ungrazed riparian habitats, but were more abundant in ungrazed areas (Medin and Clary 1990).
Although heavily grazed areas are unsuitable for nesting, grazing can be used to improve habitat. In North Dakota, densities of were highest in fields one year postgrazing, followed by dense nesting cover (DNC), grazed, and idle fields (Renken 1983). In Saskatchewan, foraged most frequently in grazed grasslands, but the number of nesting pairs was consistently lower in grazed than ungrazed areas (Dale 1984). Herkert et al. (1993) recommended light grazing, resulting in > 40 percent vegetative cover > 25 cm tall, to produce the intermediate vegetation height and density. In Alberta, frequency of occurrence did not significantly differ between four grazing treatments: early season tame (grazed from late April to mid-June), early season native (grazed in early summer), deferred grazed native (grazed after 15 July), and continuously grazed native (Prescott and Wagner 1996). In North Dakota, densities did not differ significantly between season-long (continuously grazed), short-duration, and twice-over rotation grazing treatments (Messmer 1990). Short-duration grazing involves a system of pastures rotated through a grazing schedule of approximately one week grazed and one month ungrazed, repeated throughout the season. Twice-over rotation involves grazing a number of pastures twice per season, with about a 2 month rest in between grazing. Season-long grazing involves leaving a herd on the same pasture all season. Higher densities of Brown-headed Cowbirds were observed in summer-grazed versus winter-grazed pastures in Colorado, and Knopf et al. (1988) cautioned that summer grazing could lower reproductive success of Savannah Sparrows and other grassland-nesting birds.
CULTIVATION/CROPLAND: In Saskatchewan, were more abundant in hayland and tame pasture than in native pasture, and were detected least frequently in cropland (Anstey et al. 1995); in Alberta, the species also occurred more frequently in tame than native pasture (Prescott and Murphy 1996). In native pasture, preferred low to moderate cover diversity and moderate to tall grass; in tame pasture, the species preferred high herbaceous biomass (Prescott and Murphy 1996). However, another Saskatchewan study found that Savannah Sparrow occurred more frequently on native pasture than on pastures seeded to pure crested wheatgrass (AGROPYRON CRISTATUM); no difference was found in frequency of occurrence when native pastures were compared to pastures of crested wheatgrass/grass (brome [BROMUS INERMUS] and Kentucky bluegrass (POA PRATENESIS) or crested wheatgrass/alfalfa (MEDICAGO SATIVA) (Davis and Duncan, in press).
In eastcentral Saskatchewan, were as common in fallow cropland as in two kinds of planted cover (Dale 1993). In Manitoba, Savannah Sparrows were as equally abundant in native DNC as in tame DNC and native grassland (Dhol et al. 1994). However, they were more productive in native DNC than in native grassland. Avian productivity was evaluated using a behavioral index that ranked behavior indicative of nestling or fledgling care. Most abundant in 1-3 years old DNC fields in Alberta, although they were present in all age classes studied (0-5 years since seeding) (Prescott and Murphy 1995).
In Alberta, abundance and productivity increased when winter and spring wheat were minimum tilled rather than conventional tilled (Dale and McKeating 1996). In Iowa, Savannah Sparrows nested at low densities in nontilled fields of corn and soybeans that were idle in fall and spring and contained year-round crop residue, rather than in tilled fields (Basore et al. 1986). In North Dakota, very few were found nesting in any type of cropland (conventional tillage, minimum-tillage, and organic fields in fallow, sunflowers, or wheat) (Lokemoen and Beiser 1997). Minimum-tillage and organic fields had more vegetation and attracted greater numbers and species of birds, but predation and mechanical activities resulted in low reproductive success (Lokemoen and Beiser 1997). Conservation tillage (reduced and no-tillage) may create ecological traps, as grassland nesting birds are attracted to the habitat provided, but have low nesting success (Best 1986).
Relevance to Humans and Ecosystems
Stewardship Overview: Promote management or enhancement activities that increase the amount of contiguous grassland habitat (Herkert 1991). Acquire large grassland tracts and minimize edge effects through reduction of woody vegetation along edges and within grasslands (Wray et al. 1982; Johnson and Temple 1986, 1990; Burger et al. 1994). In small grasslands, adjacent woody habitats may allow edge and woodland predators to penetrate interior grassland areas.
Aim grassland restoration at benefiting bird species most sensitive to habitat fragmentation; restoration projects should be > 50 ha, preferably > 100 ha (Herkert et al. 1993). Where grassland restoration > 30 ha is not possible, Herkert et al. (1993) recommended establishing several small grasslands, 6-8 ha minimum size, within 0.4 km of each other, and using adjacent grassland habitats (e.g., pasture, hayland, waterway) as corridors among tracts.
Avoid disturbing (e.g., burning, mowing, moderate or heavy grazing) suitable habitat during the breeding season, approximately 1 May to 1 August. Treatments in nesting habitat should be delayed until after 1 August to prevent destruction of fledglings and renesting females (Swanson 1996).
Burn grasslands managed for breeding habitat in early spring (March to April) or late fall (October to November). In grasslands > 50 ha, burn 25-30 percent annually (Herkert et al. 1993).
When possible (e.g., on federal lands or through cooperation with private landowners), delay mowing of hayfields until mid-July, which would allow many birds to raise at least one brood in years with normal breeding phenology; mowing should be delayed further if nesting is delayed by inclement spring weather (Dale et al. 1997). When mowing must be done during the breeding season, divide large fields, mowing only half each year, or mow individual fields every other year to provide refuge for fledglings (Dale et al. 1997).
On airports not large enough to provide habitat for nesting birds (e.g., where all of the grassland available must be mowed to meet Federal Aviation Administration standards), mow grass short enough (< 4 cm) to discourage nesting. This may cause birds to select alternative areas where nesting success would be higher (Kershner and Bollinger 1996).
Light grazing (leaving > 40 percent vegetation cover > 25 cm tall) can be used to create intermediate vegetation height and density (Herkert 1991).
The savannah sparrow (Passerculus sandwichensis) is a small American sparrow. It is the only widely accepted member of the genus Passerculus. Comparison of mtDNA NADH dehydrogenase subunit 2 and 3 sequences indicates that the Ipswich Sparrow, formerly usually considered a valid species (as Passerculus princeps), is a well-marked subspecies of the Savannah Sparrow, whereas the southwestern Large-billed Sparrow should be recognized as distinct species (Passerculus rostratus).
The common name comes from Savannah, Georgia, where one of the first specimens of this bird was collected.
This passerine bird breeds in Alaska, Canada, northern, central and Pacific coastal USA, Mexico and Guatemala. The Pacific and Mexican breeders are resident, but other populations are migratory, wintering from the southern United States across Central America and the Caribbean to northern South America. It is a very rare vagrant to western Europe.
This species has a typically sparrow-like dark-streaked brown back, and whitish underparts with brown or blackish breast and flank streaking. It has whitish crown and supercilium stripes, sometimes with some yellow (more often near the beak). The cheeks are brown and the throat white. The flight feathers are blackish-brown with light brown or white border. The eyes are dark. The feet and legs are horn-colored, as is the lower part of the bill, with the upper part being dark grey.
The Savannah Sparrow is a very variable species, with numerous subspecies, several of which have been split as separate species at various times. The different forms vary principally in the darkness of the plumage. The variation generally follows Gloger's Rule, with Alaskan and interior races the palest, and southwestern coastal forms the darkest. There are some exceptions, though, most conspicuously in some island populations that presumably were strongly affected by founder effects. The general pattern of variation has a fairly clear divide, southwest of which the birds become notably darker; this agrees quite well with the limit between P. sandwichensis and P. (s.) rostratus. Savannah Sparrows show some variation in size across subspecies. The total length can range from 11 to 17 cm (4.3 to 6.7 in), wingspan ranges from 18 to 25 cm (7.1 to 9.8 in) and body mass from 15 to 29 g (0.53 to 1.02 oz). In the nominate subspecies, the body weight averages 20.1 g (0.71 oz).
The Savannah Sparrows proper (see below) are very similar, and migrant birds can not usually be related to a breeding population with certainty. The resident or partially migratory subspecies are well distinguishable by size and, particularly between groups, coloration.
|Problems playing this file? See media help.|
These birds forage on the ground or in low bushes; particularly in winter they are also found in grazed low-growth grassland. They mainly eat seeds, but also eat insects in the breeding season. They are typically encountered as pairs or family groups in the breeding season, and assemble in flocks for the winter migration. The flight call is a thin seep. Sensu lato, the Savannah Sparrow is considered a threatened species by the IUCN. The song is mixture of chirps and trills.
Seventeen subspecies (including the Large-billed Sparrows) are currently recognized, though many are only described from wintering birds and much of the variation seems to be clinal. Four additional subspecies are no longer generally accepted. The complex is usually divided into several groups:
Savannah Sparrows proper
All are migratory; wintering ranges overlap widely.
- P. s. labradorius, breeds in Newfoundland, Labrador, and N Quebec
- P. s. oblitus, breeds in N Ontario and Manitoba
- P. s. savanna (Eastern Savannah Sparrow), breeds in the NE USA and adjacent Canada (includes P. s. mediogriseus)
- P. s. sandwichensis (Aleutian Savannah Sparrow), breeds on the Aleutian Islands and W Alaskan Peninsula
- P. s. anthinus, breeds in the remainder of Alaska, south and east to central British Columbia and north of the Great Plains to Manitoba
- P. s. brooksi (Dwarf Savannah Sparrow), breeds in southernmost British Columbia to northernmost California
- P. s. alaudinus, breeds in coastal northern and central California
- P. s. nevadensis, breeds in the N Great Plains and the Great Basin
- P. s. brunnescens, breeds from central Mexico south to Guatemala (includes P. s. rufofuscus)
Some post-breeding dispersal. Formerly considered a distinct species.
- P. s. princeps, breeds almost exclusively on Sable Island
- The Ipswich Sparrow is somewhat larger and paler in colour than other eastern Savannah Sparrows. The breast streaks are narrower and pale brown. Some birds overwinter on the island; others migrate south along the Atlantic coast, usually departing later and returning sooner than mainland birds. Some birds interbreed with P. s. savanna in Nova Scotia. These birds frequently raise three broods in a year. This bird was first observed in winter on the dunes near the town of Ipswich, Massachusetts.
The Large-billed Sparrows proper are 2–3[verification needed] dark, large and strong-billed subspecies:
- Passerculus rostratus/sandwichensis rostratus, which breed on the Gulf Coast of NE Baja California and NW Sonora (some post-breeding dispersal).
- P. r./s. atratus, resident on the coast of central Sonora to central Sinaloa (resident)
- P. r./s. beldingi, resident on the Pacific Coast from Morro Bay, California, to El Rosario, Baja California (includes P. r./s. bryanti)
- P. r./s. anulus, resident around Sebastián Vizcaíno Bay, Baja California
- P. r./s. guttatus, resident around San Ignacio Lagoon
- P. r./s. magdalenae, resident around Magdalena Bay
- P. r./s. sanctorum
- This is a large-bodied and large-billed subspecies, similar to rostratus. They utilize different habitat and their breeding season does not seem to coincide[notes 1] with that of Belding's Sparrows. However, their bill size is due to convergent evolution and their habitat choice simply to the lack of alternatives on their barren island home; altogether, it appears to be a fairly recent offshoot from the Belding's Sparrows group. It appears as distinct evolutionarily from these as does the Ipswich Sparrow from the Savannah Sparrow proper group, only that there seems to have been more gene flow and/or a larger founder population in the case of the latter.
- For late April 1906, Thayer & Bangs (1907) report nestlings, young birds molting into adult plumage, but no eggs anymore. Rising in Zink et al. (2005) found no breeding activity in late April 1999. Consequently, breeding seems to take place in the winter months, with the last young fledging in April/May.
- BirdLife International (2012). "Passerculus sandwichensis". IUCN Red List of Threatened Species. Version 2013.2. International Union for Conservation of Nature. Retrieved 26 November 2013.
- Zink, Robert M.; Rising, James D.; Mockford, Steve; Horn, Andrew G.; Wright, Jonathan M.; Leonard, Marty & Westberg, M.C. (2005). "Mitochondrial DNA variation, species limits, and rapid evolution of plumage coloration and size in the Savannah Sparrow". Condor 107 (1): 21–28. doi:10.1650/7550. JSTOR 3247750.
- Byers, Clive; Olsson, Urban & Curson, Jon (1995): Sparrows and Buntings: A Guide to the Sparrows and Buntings of North America and the World. Houghton Mifflin, Boston. ISBN 0-395-73873-3
- "Savannah Sparrow, Life History, All About Birds – Cornell Lab of Ornithology". Allaboutbirds.org. Retrieved 2013-03-30.
- Sproule, Ken. "Sparrows". Toronto Wildlife. Retrieved 2013-03-30.
- CRC Handbook of Avian Body Masses by John B. Dunning Jr. (Editor). CRC Press (1992), ISBN 978-0-8493-4258-5.
- Olson, Storrs L.; James, Helen F. & Meister, Charles A. (1981). "Winter field notes and specimen weights of Cayman Island Birds". Bulletin of the British Ornithologists' Club 101 (3): 339–346.
- Thayer, John E. & Bangs, Outram (1907). "Birds Collected by W. W. Brown, Jr., on Cerros [sic], San Benito and Natividad Islands in the Spring of 1906, with Notes on the Biota of the Islands". Condor 9 (3): 77–81. doi:10.2307/1361136.
Names and Taxonomy
Comments: Some authors have recognized the various groups as separate species: princeps (Ipswich Sparrow); sandwichensis (Savannah Sparrow); beldingi (Belding's Sparrow); and rostratus (Large-billed Sparrow) (AOU 1983, 1998). Morphological evidence and the level and pattern of mtDNA variation support species status for rostratus (of southwestern California and the north and east coasts of the Gulf of California) (Zink et al. 1991).
Rising (2007) recommended that six subspecies (including alaudinus) be synonymized under P. s. sandwichensis (this was accepted by ITIS). Zink et al. (2005) recognizes P. rostratus as distinct, as well as questions the validity of certain subspecies of sandwichensis.