Module 3: Bar-Tailed Godwit — 11 000 km Non-Stop

The bar-tailed godwit is a large migratory shorebird in the Scolopacidae, breeding across the high Arctic tundra of Eurasia and western Alaska. Of the five recognised subspecies, the Alaskan-New Zealand flyway population belongs to L. l. baueri, a bird that averages 250 g lean mass and 37 cm body length, with a moderately long, slightly upturned bill. The sister subspecies L. l. menzbieri winters in northern Australia and stages in the Yellow Sea; the Eurasian L. l. lapponica winters in western Africa and European estuaries. None of these populations attempts the single-hop Pacific crossing: only baueri regularly executes a non-stop flight of this length.

Landis et al. (2019, Ibis) have shown that the closely related Limosa haemastica (Hudsonian godwit) of the Atlantic flyway performs a shorter but still demanding 6 000 km hop from Hudson Bay to the Pampas. The Atlantic route benefits from land bailouts along the Caribbean coast, but the birds largely fly non-stop anyway, suggesting that the long non-stop strategy predates the expansion of the baueri Pacific route and reflects a general phylogenetic predisposition in Limosa.

The Pacific flyway

In September, baueri departs the Alaskan estuarine staging grounds and heads south-southwest, crossing the entire Pacific without encountering land. The route passes east of Hawai‘i but well west of Tonga, avoiding both the central Pacific calms and the tropical-storm corridor. Recovery data from geolocators and satellite tags show a remarkably tight cluster of trajectories, indicating either a strong innate heading or learned route fidelity in adults.

Robert E. Gill Jr. and colleagues at the USGS Alaska Science Center deployed satellite transmitters on a cohort of baueri godwits in 2006–2008 and published the resulting trajectories in Proceedings of the Royal Society B (2009). The most famous individual, a female coded E7, carried a 26 g PTT (platform transmitter terminal) and was tracked from Alaska across the Pacific to the Firth of Thames, New Zealand: 11 680 km in 8.1 days, the longest uninterrupted migratory flight ever documented in a bird. The transmitter continued to function after arrival, showing only a brief fuelling interval before the northward leg via the Yellow Sea the following spring.

The Gill 2009 dataset showed several surprising features. Ground speed was remarkably steady, varying from 11 m/s during calms near the doldrums to almost 25 m/s in tailwind-assisted segments near the polar front. Flight altitude, measured indirectly through temperature sensors, was typically 1 000–2 000 m, well below the passerine nocturnal ceiling. And the flight was overwhelmingly diurnal and nocturnal combined, with no evidence of regular rest or roosting breaks.

Gill 2023 update: the B6 record

Gill et al. (2023, Avocetta) reported an updated record: the 5-month-old juvenile male B6 flew a great-circle route of 13 560 km non-stop from Alaska to Ansons Bay, Tasmania, completing the flight in 11 days and 1 hour. Because Tasmania lies further south-west than typical baueriwintering grounds, this trajectory is likely an anomaly driven by drift off the normal route, but the physiological implication is profound: a 5-month-old juvenile, never having flown before the start of migration, is capable of sustaining flight for nearly two weeks without refuelling.

Wind-aided routes

The timing of autumn departure correlates tightly with the position of the polar front jet, which sweeps south-eastward off Alaska in September as the Aleutian Low deepens. By departing during the passage of a low-pressure system, godwits gain 5–15 m/s of tailwind along the first 2–3 000 km, equivalent to a 30–100% boost in ground speed and a major saving of fuel. Shamoun-Baranes et al. (2006) and Felicísimo et al. (2008) have shown that such wind-riding is a statistically significant feature across many long-distance migrants.

Hyperphagia is the dramatic increase in food intake and body mass that precedes the migratory flight. In baueri godwits the process takes 4–6 weeks on the estuarine mudflats of the Yukon-Kuskokwim Delta and results in a body mass increase from \(\sim 250\) g lean to \(\sim 600\) g total — almost 150% of lean mass as pure fat. This is close to the upper limit observed in any migrant bird; some Arctic-breeding shorebirds and small passerines achieve similar multipliers.

The digestive machinery

To sustain the intake rate needed to deposit up to 14 g of fat per day, baueri re-shapes its entire digestive system. Gut mass, measured by Battley and Piersma on freshly killed birds, can treble during the staging phase, from ~20 g to ~55 g, to handle the marine-worm and bivalve diet of the mudflats. Liver mass also increases to handle lipogenesis; kidney mass declines because water balance is easier on a high-fat diet; and the crop is expanded to store undigested prey between tides.

Flight-muscle atrophy — counter-intuitive

A striking finding from Piersma, Bruinzeel, Drent & colleagues (reviewed in Piersma & van Gils 2011) is that the pectoralis major (the major flight muscle) does not hypertrophy during hyperphagia. In many species, including baueri, it can actually shrink slightly during peak fat loading. This appears paradoxical—more mass surely needs more muscle?—but the resolution lies in the flight-power scaling. Adding fat reduces the power demand per unit mass because induced power scales as \(W^2 / (\rho V b^2)\) and the bird can afford to fly at a slightly lower speed in the dense, fully fuelled state. At the moment of take-off, however, the muscle expands by 15–25% over the final week before departure to handle the maximum static load.

Fat is the universal migratory fuel. A gram of fat yields roughly \(E_{\text{fat}} \approx 37.7\) kJ, roughly nine times the yield of a gram of carbohydrate and almost twice that of lean protein. More importantly, complete oxidation of palmitate liberates about 1.07 g of water per gram of fat oxidised, giving the migrant a built-in water supply.

Daily consumption rate

At the sustained flight power of \(P_{\text{mech}} \approx 3.8\) Wand a biochemical efficiency \(\eta \approx 0.23\) (Pennycuick 2008), a baueri godwit burns chemical energy at \(P_{\text{chem}} \approx 16.5\) W, i.e. about 1.43 MJ/day, corresponding to \(\sim 14\) g of fat per day. Over 8 days of flight this totals \(\sim 115\) g, comfortably within the 350 g fat load taken on board.

Protein is not burned

Crucially, baueri does not draw significantly on its own muscle protein during the Pacific flight. Battley et al. (2012) used stable-isotope tracer methods to show that protein oxidation contributes less than 5% of flight energy, and the flight muscles lose less than 10% of their mass even after the longest flights. This is the metabolic ceiling of fat oxidation: the bird must carry enough oxygen delivery capacity and hepatic lipid-mobilisation machinery to avoid protein tapping, otherwise the respiratory quotient \(\text{RQ} = \text{CO}_2/\text{O}_2\) rises above the fat-pure value of 0.71 and flight muscle starts to degrade.

Dehydration limits

Even with metabolic water, the godwit loses water through respiration—at altitude and low humidity the evaporative loss can exceed metabolic gain. Klaassen (1996) modelled a water deficit of 2–4% body mass per day for migrating passerines, balanced mostly by metabolic water but tipping into net loss at altitude. The shorebird’s ability to reabsorb water from the cloacal excreta and to lower kidney filtration rates during flight is part of what makes the non-stop crossing physiologically feasible.

Colin Pennycuick’s book Modelling the Flying Bird (Academic Press, 2008) gives the standard decomposition of mechanical flight power into three terms:

Induced power dominates at low speed, parasite power at high speed, and the cruise speed that minimises total power—the ‘maximum range speed’—is found by differentiating total cost of transport. For a fully fuelled baueri, \(V_{\text{mr}} \approx 16\) m/s, consistent with the observed cruise airspeed.

Wing morphology

The bar-tailed godwit has wings optimised for long-distance flapping flight: long (span 76–82 cm), narrow (aspect ratio ~9), and with a moderately high wing loading of 70–90 N/m² at peak fuelling. These parameters place it in the efficient flapping-cruise regime, without the soaring specialisations of albatrosses or the short-haul elasticity of passerines.

Energy cost of transport

Bairlein’s wind tunnel measurements

Franz Bairlein’s wind tunnel at Wilhelmshaven has measured the metabolic rate of flying passerines and shorebirds. The results broadly confirm Pennycuick’s model at the level of 10–20% and identify a characteristic U-shaped power-speed curve for each species. For baueri, the minimum-power speed is about 12 m/s and the maximum-range speed about 16 m/s.

Theunis Piersma (NIOZ, University of Groningen) coined the term flexible phenotypeto describe the remarkable plasticity of shorebird body composition. Across the annual cycle, a knot or godwit may cycle through four or more distinct physiological configurations: breeding, post-breeding moult, pre-migration hyperphagia, migratory flight, and wintering. Each state involves a different balance of gut, muscle, fat, liver, kidney and heart mass. The changes are large—organ masses vary by factors of 2–3—and reversible within days to weeks.

The gut-flight trade-off

The most dramatic component is the gut-flight trade-off. A big digestive tract is essential for the refuelling phase—up to 55 g of intestine is needed to process the daily food intake—but during flight the same organ is dead weight, imposing an extra 3 W of parasite cost over the 8-day crossing. Piersma’s solution: the gut shrinks by two-thirds before take-off, while the flight muscles grow in anticipation. Upon arrival in New Zealand, the opposite restructuring begins within hours.

Osteology of long flight

Even the skeleton reflects flight specialisation. The furcula (wishbone) and sternum carry the pectoralis attachment; the coracoid transmits thrust. In baueri, the sternum keel is deep and the coracoid is heavily buttressed. The humerus is pneumatised to reduce inertia during the wingbeat; the hand bones are fused to increase stiffness at the wingtip. Despite these skeletal optima, total skeleton mass remains only 6–7% of body mass, a remarkable compromise.

For a small bird flying 8–10 days non-stop, the question of sleep is inescapable. Direct EEG measurements on a freely flying godwit are not yet feasible, but work on frigatebirds by Rattenborg, Voirin, Cruz and colleagues (Nature Communications, 2016) has established that some seabirds perform brief bouts of unihemispheric slow-wave sleep (USWS) during flight—one hemisphere of the brain sleeps at a time, while the other stays vigilant. Total sleep duration in frigatebirds during multi-day flights was remarkably low (42 min/day vs 12 h on land), suggesting extreme tolerance of sleep loss.

Whether godwits use the same strategy is not yet established. Behavioural observations during the Alaskan-New Zealand flight reveal no obvious hesitation or drift; the tagged birds maintain heading and altitude through the entire 8-day crossing without interruption. The working assumption is that some form of USWS is also present, probably coupled to automatic central-pattern generation in the spinal cord that maintains the wingbeat without active cortical input.

Phil Battley (Massey University) and colleagues have carried out the most detailed physiological studies of baueri at the New Zealand arrival site, measuring body composition, metabolic rate and tissue water content on birds captured shortly after landfall. The 2012 paper in Oecologia used doubly labelled water to estimate the field metabolic rate during the flight, yielding a value of 0.60 W/g lean mass, consistent with the Pennycuick prediction at the observed ground speed.

Battley also measured kidney mass on arrival: it was reduced by ~30% relative to pre-migration staging birds, as predicted by the flexible-phenotype framework, and recovered within five days. This is consistent with the hypothesis that during the flight the bird dials down renal water-and-solute filtration to conserve body water, relying on the high metabolic-water output of fat oxidation.

A useful comparison is the Atlantic flyway of Calidris canutus rufa, the Red Knot subspecies breeding in the Canadian Arctic and wintering in Tierra del Fuego. Landis and colleagues have tracked rufa across the Atlantic with geolocators and satellite transmitters and found that, despite an overall journey of 30 000 km round-trip, individual flight segments are shorter than baueri—typically 2 000 to 5 000 km—with a critical refuelling stop at Delaware Bay to exploit the horseshoe crab egg bloom. The Atlantic route thus trades route length for predictable staging grounds, a strategy made feasible by the continental geography of the Americas.

The contrast highlights that baueri’s single-hop strategy is not a generic optimum but a specific solution to the Pacific’s lack of landfall. Over the open ocean, there is simply nowhere to stop; the bird must carry the fuel for the entire journey. Over a continental flyway, distributed staging is both possible and energetically cheaper, provided the staging resources remain reliable.

At a cruise altitude of 1 500 m, the air temperature in the North Pacific in September is 0–10 °C, colder than at sea level. The godwit’s metabolic heat production during flight (16 W chemical, of which 3.8 W is mechanical and 12 W is heat) is more than adequate to keep its 600 g body at 40 °C; the challenge is actually to shed heat, not to generate it. Feather re-arrangement, leg extension into the slipstream, and occasional climbs to cooler air layers are the available strategies.

On arrival in the warmer New Zealand summer, the opposite problem dominates: the bird must unload heat after 10 days of forced high production. Observations at the Firth of Thames report that arriving birds often stand at the water’s edge with wings slightly open for several hours before resuming normal behaviour.

The figure below sketches the Alaska-to-New Zealand great-circle route followed by E7 and B6, overlaid with the approximate September 500 hPa wind-vector pattern inferred from reanalysis. The polar front jet provides a strong tailwind in the first 3 000 km, which then transitions to the equatorial calms and the south-eastern trade winds.

The Alaska-New Zealand population has been declining since the mid-2000s, largely because of loss of tidal-flat staging habitat around the Yellow Sea on the northward leg (Piersma et al., 2016). Even though the Pacific southward flight is non-stop and therefore insulated from intermediate habitat loss, the birds return via a series of tidal-flat stopovers on the coast of China and the Korean Peninsula, where industrial reclamation has destroyed >60% of the historic mudflat area since 1980. The Yellow Sea is the bottleneck of the East Asian–Australasian flyway and currently limits the whole population.

Climate-driven shifts in the jet-stream position and in the timing of the Alaskan staging bloom add further uncertainty. The Gill 2023 update noted that departure dates have shifted ~1 week later since 2008, possibly tracking the later onset of the autumn polar-front circulation. Whether these small shifts can be tolerated indefinitely by a species that is pushing against physiological limits at every fuelling and flight is one of the open questions motivating continued satellite tracking.