Migration

Long-distance migrant and one of the longest-distance migrants in the animal kingdom. Western Hemisphere–breeding groups travel from middle- and high-arctic breeding latitudes to e. and w. Atlantic Coasts between north and south temperate latitudes, including southernmost continental lands of South America. Knots are a classic “jump” migrant (Piersma 1987), undertaking long flights that may span thousands of kilometers without stops. At some stages of migration, very high proportions of entire populations may use a single migration staging site to prepare for long flights, giving rise to conservation concerns (see Conservation and management, below). Unknown numbers remain south of breeding grounds throughout the breeding season.

This section focuses on knots in the Western Hemisphere. For details elsewhere, see Piersma and Davidson 1992 .

C. c. islandica: Details of migration route(s) traveled between e. arctic Canada, Greenland, and the Old World are unclear (Morrison 1984). No major autumn-migration staging sites are known on the western side of the Atlantic, but R. I. G. Mor-rison (unpubl.) noted that islandica captured on Ellesmere I., Northwest Territories, near Alert during Aug 1996 had sufficient fat to fly 4,000 km without stops. This energy reserve would enable nonstop flight to staging areas on Waddenzee coasts of the Netherlands and Germany, where C. c. islandica gather by tens of thousands during Aug and Sep (Piersma et al. 1993a). Following a Prebasic molt on coasts of Waddenzee, most islandica shift westward to winter principally in bays and estuaries of the United Kingdom (Davidson and Wilson 1992), but small numbers also winter on coasts of France and perhaps Portugal.

Spring migration for many islandica involves transoceanic flights from the British Isles to staging sites on west coast of Iceland during mid- to late May (Davidson and Wilson 1992). Others (perhaps most) return to w. Waddenzee coast of the Netherlands during Mar and Apr (Piersma et al. 1994) and then briefly move to German Waddenzee coast before flying to w. Iceland and n. Norway staging sites in early May. From there, islandica fly directly to breeding grounds in Greenland and ne. Canada, arriving in late May–early Jun (Morrison and Wilson 1992).

C. c. rufa and C. c. roselaari: Because migration route(s) and wintering areas of roselaari are essentially undocumented, individuals that migrate and spend their winter in Western Hemisphere are grouped together in this account. Much of the following information comes from Morrison 1984 and references therein, from Morrison and Harrington 1992, and from Harrington 1996 and unpubl.

Atlantic Coast

Winter. Major rufa wintering zone is on coasts of s. Chile and Argentina (Morrison and Ross 1989), with extraordinary concentrations (>50,000) in n. Tierra del Fuego; lower numbers (about 10,000; Harrington et al. 1988), perhaps roselaari, winter on west coast of Florida, and probably also on Atlantic coast of ne. Brazil and Colombia.

Fall Migration. Southward migration of rufa begins mid-Jul. Northernmost staging sites are on coasts of Hudson and James Bays and on Atlantic Coast of U.S., especially at small number of sites in Massachusetts and New Jersey (Morrison and Harrington 1979). Here numbers of adult rufa increase steadily until early Aug and then decline rapidly between 10 and 20 Aug (International Shorebird Surveys [ISS] unpubl.). First juveniles arrive 20–30 Aug, but do not concentrate at traditional staging sites to the same degree as adults (BAH).

In Atlantic Provinces of Canada during autumn, highest numbers visit during same period as in Massachusetts (Hicklin 1987). At St. Pierre and Miquelon Is., near Newfoundland, knots occur only in autumn (R. Etcheberry pers. comm.), restricted to late parts of the migration period, suggesting that virtually all are juveniles.

Banding (Harrington and Leddy 1982) and other evidence show that adult rufa fly over the Atlantic from U.S. staging areas to South America. Species is uncommon in se. U.S. before Aug. Also uncommon in West Indies (Raffaele et al. 1998, ISS unpubl.).

Most individuals depart e. U.S. and Canada during Aug, making landfall mid- to late Aug along South American coasts of the Guianas (Spaans 1978). Then move south and east (BAH), possibly to just east of Belem (Brazil), before flying overland across Brazil to Atlantic coasts of s. Brazil, Uruguay, and Argentina, where first arrivals are seen in early Oct (ISS unpubl.). Individuals then appear to move south along Atlantic coast to principal wintering grounds in s. Argentina and Tierra del Fuego (ISS). According to data from Jan surveys by Morrison and Ross (1989), a high proportion of rufa population spends the boreal winter season concentrated in a single bay, Bahìa Lomas in Chile, near eastern end of Strait of Magellan (52°30’S; 69°00’W). Frequently but uncommonly observed in Bermuda, where records range from 1 Aug to Nov (Amos 1991), and recorded in West Indies primarily Sep and Oct (Raffaele et al. 1998).

Few individuals use se. U.S. coasts before Aug, when numbers increase from low and middle hundreds to thousands (Marsh and Wilkinson 1991, B. Winn pers. comm.). On basis of migration timing, along with plumage and molt condition (BAH), it seems likely that these birds are roselaari destined to winter on sw. Florida coast (see Harrington et al. 1988). However, average measurements of captured individuals from Florida during winter (bill 36.25 mm ± 1.94 SD [n = 238]; natural wing-chord 162.3 mm ± 4.4 SD [n = 211]); from Argentina dur-ing Apr (bill 36.19 mm ± 1.92 SD [n = 279]; wing 162.2 mm ± 4.78 SD [n = 273]); from New Jersey during May (bill 35.60 mm ± 1.78 SD [n = 466]; wing 162.6 mm ± 4.06 SD [n = 439]); or from Massachusetts during Jul–Aug (bill 35.63 mm ± 1.94 SD [n = 498]; wing 162.2 mm ± 162.2 SD) are not significantly different (BAH).

Spring Migration. Northward migration of rufa on central Atlantic coast of Argentina includes a distinct and relatively protracted passage beginning in Feb, with numbers peaking mid-Mar–mid-Apr, then declining sharply by end of Apr (Blanco et al. 1992, Gonzalez et al. 1996). As was true in Chubut Province, Argentina (ISS unpubl.), counts during northward migration were substantially higher than during southward migration (Blanco et al. 1992).

Species is abundant on coasts of s. Brazil during late Apr–early May (Vooren and Chiaradia 1990). Most are in Alternate plumage, and in 1984 they departed during early May after attaining high fat levels (Harrington et al. 1986). Species rare on easternmost Brazilian coasts, although habitat there appears suitable. Severino Mendes de Azevedo Júnior and collaborators routinely counted shorebirds on sandy tidal flats of a river inlet near Recife for 3 yr and found only 1 knot (J. Azevedo in litt.; see also Júnior 1992).

Near Belem in n. Brazil, Rodriguez (1993) found hundreds of knots (unknown whether they were rufa or roselaari) between Jan and Apr 1992, but not during May 1991. Cassler and Lira (1979) describe a similar abundance pattern from the Maracaibo region of Venezuela. Counts from Suriname (Spaans 1978) and French Guiana (ISS unpubl.) show few individuals between Jan and end of Apr, then a brief, notable increase (hundreds) during early May; numbers do not reach the magnitude known during southward migration, however.

Species observed even less frequently in West Indies during northward migration than southward migration (Raffaele et al. 1998). Frequently but uncommonly observed in Bermuda, where records range from 12 May to 4 Jun (Amos 1991).

On se. U.S. Atlantic Coast, few wintering individuals (Am. Birds [Christmas Bird Counts, 1960–1989] and ISS unpubl.). Numbers increase during Mar and reach maxima early Apr–early May (ISS unpubl.; max 8,500 [P. Nugent pers. comm.]). An isolated record of 12,000 moving northward along Georgia coast on 22 May (Coolidge 1972) is extraordinary; typically species is scarce on southeast coast of U.S. after Apr. In Virginia, where relatively few individuals occur during fall, spring numbers peak around 24 May (ISS unpubl.).

In ne. U.S. and e. Canada, highest numbers during spring occur on New Jersey and Delaware shores of Delaware Bay (Harrington 1986, Clark et al. 1993). At Cape May, NJ, migration occurs between third week of Apr and first week of Jun (Sibley 1997). In most years, steady buildup from mid- to late May, with maximum counts >40,000 (Clark et al. 1993). Few remain after 5 Jun. During spring, few individuals found in Massachusetts (Veit and Petersen 1993) or Atlantic Provinces of Canada (Hicklin 1987).

Interior

Negligible numbers occur at interior sites during both spring and fall, with slightly greater numbers during fall in some areas (e.g., Ohio). Records range from 10 May and 17 Jun, peaking 16–28 May, to 15 Jul–20 Oct (rarely later), peaking 20 Aug–15 Sep in Ohio, where most are observed along shores of Lake Erie (Peterjohn 1989). In Kansas, records range from 17 Apr through 1 Jun and 2 Aug through 26 Sep (rarely later), peaking in Sep, with most observed at Cheyenne Bottoms Wildlife Management Area in Barton Co. (Thompson and Ely 1992).

Pacific Coast

Winter. Low numbers on Pacific Coast of South America (Johnson 1965, Morrison and Ross 1989). Status poorly known in Mexico, where possibly hundreds or low thousands winter on e. Sea of Cortés (Morrison et al. 1992, Harrington 1993). Elsewhere few reported in Yucatán, Veracruz, Sonora, Baja California, and Baja California Sur (ISS unpubl.).

Spring Migration. Uncommon migrant along Pacific Coast of Panama Mar–Apr (Ridgely and Gwynne 1989). In Costa Rica, uncommon migrant along Pacific Coast mainly mid-Mar–late Apr (Stiles and Skutch 1989). Occurs regularly during northward migration on Salton Sea in California (Am. Birds 1978–1982), where counts sometimes reach 200–300 during late Apr–early May. Although unsubstantiated, it seems likely that these knots come from the small populations (perhaps roselaari) that winter on Sea of Cortés and Baja California Pacific Coasts of Mexico (see above).

Species (perhaps roselaari) uncommon on U.S. Pacific Coast. Numbers rarely exceed low thousands in Gray’s Harbor/Willapa Bay region of Washington between late Apr and mid-May (Paulson 1993). About a thousand counted in San Francisco Bay during Apr (G. Page in litt.). In general, migration occurs early Apr–late May, peaking during second half of Apr, in California (Small 1994); mid-Apr–third week of May, peaking during first half of May, in Oregon (Gilligan et al. 1994); and mid-Apr–early Jun, peaking late Apr and early May, in British Columbia, where recorded along entire coast (Campbell et al. 1990).

In Alaska, up to 100,000 knots reported by Isleib (1979) during second or third week of May at Copper/Bering River delta. Gill and Handel (1990) reported a similar number from Yukon-Kusko-kwim Delta on 21 May. These dates are before peak migration periods of rufa in e. U.S. (see above); possibly they are roselaari as suggested by Roselaar (1983), en route to Alaskan and/or Wrangel I. breeding areas. Few individuals occur as migrants in Alaska for remainder of year.

According to Roselaar (1983: 107), C. c. roselaari “do not reach their breeding areas from the south-west but from the east-south-east, with a last stop-over area at Copper and Bering River Deltas in southern Alaska, where over 40,000 birds have been seen (Kessel and Gibson 1978). Earlier spring stop-over sites of this population are at Gray’s Harbor (Bent 1927) and Willapa Bay (Gabrielson and Lincoln 1959) on the Washington coast of the U.S., where large numbers assemble in May. The other whereabouts of this population is a mystery: they probably winter on the Pacific coast somewhere between southern California and Chile . . . ”

During northward migration, individuals banded during fall and winter on west coast of Florida (perhaps roselaari) have been observed in low numbers at the Delaware Bay staging site, which is used mostly by the rufa that winter in austral South America (see above).

Fall Migration. Status unclear on Pacific Coast during fall migration. Where species is sometimes abundant during spring, it is less numerous during fall. Only low numbers (dozens) reported from Alaska. In British Columbia, observed mainly along extreme southwestern coast late Jun through Nov, peaking anywhere between mid-Jul and late Sep; individuals observed more often than flocks (Campbell et al. 1990). In Oregon, usually rare, occasionally uncommon transient during autumn along coast (Gilligan et al. 1994). In California, migration occurs late Jul–early Oct (Small 1994). Up to a thousand counted at San Francisco Bay (G. Page in litt.). Low numbers in Costa Rica (Smith and Stiles 1979), and no substantive migration reported from Pacific coast of South America.

Information in this section comes from Piersma et al. 1990a, 1990b, 1997; see also Tulp et al. 1994 and Battley 1997 . Each kind of shorebird has its own range of optimal flying conditions (speed, altitude, flock configuration, etc.), so it is not surprising that migrating shorebirds generally fly in single-species flocks. Flock configurations may vary (in order of prevalence at Banc d’Arguin, Mauritania) from V-formations to echelons, clusters, or bunches. Migration departure times of most shorebirds, including Red Knots, tend to be on sunny days in the few hours before twilight. Flock size tends to differ among species. Red Knots tend to have larger flocks than most other shorebirds; mean size of 64 flocks departing north from Mauritania was 29 ± 28 SD, with the “average bird” being in a flock of 56 ± 31 SD birds. Mean size of flocks departing the Dutch Waddenzee during spring was 92 ± 100 SD (the average bird was in a flock of 197 ± 110 SD); during fall, comparable values on the Waddenzee were 113 ± 121 SD and 236 ± 108 SD, respectively. Directions taken by departing flocks seem to be on a near-direct course toward their destination, with little allowance for wind drift, possibly excepting conditions when departures take place with tailwinds. Departing flocks observed in Mauritania gained altitude at rate of 0.91 m/s, a relatively higher value than for 7 other species measured (see also Piersma et al. 1997).

At Delaware Bay in the U.S., northward-departing Red Knot flocks typically start out as clusters, sometimes of mixed species, but with increasing altitude change to V-formations and echelons, with species tending to segregate (BAH). Generally fly on northwesterly track, but often deviate—especially toward southwest or south—as long as flocks are in sight.

The amazing physiological preparations that shorebirds complete before long-distance migration flights were first suggested by Cooke (1910), who described shorebirds flying nonstop over the ocean from e. Canada to South America, and from Alaska to the Hawaiian Is.; he noted that some birds arrive at their landfalls still retaining fat, whereas others arrive in emaciated condition. Harrington et al. (1986) and Gudmundsson et al. (1991) showed that knots about to depart on long migration flights have mean body masses 50–55% greater than estimated fat-free mass, with a mean daily increase of 3.1 g, or about 2.3% of the lean body mass. Rates of mass change may differ substantially during various stages of migration stopovers (T. Piersma pers. comm.).

Recent studies, many involving Red Knots, have shown that physiological preparations for migration are far more extensive than simply laying on of fat. Prior to migration, knots undergo pronounced changes in basal metabolic rates, organ sizes, muscle masses, and fat mass along with overall increase of mass (Piersma et al. 1995a). The increase of mass, once thought entirely attributable to increased fat, includes shifts in the relative sizes of different muscle masses (e.g., leg muscle mass decreases, flight muscle mass increases) and organs (e.g., the stomach and proventricular gizzard masses decrease), as well as increased fat mass (Piersma et al. 1996).