Module 4: Passerine Nocturnal Migration

The nocturnal strategy is used by the majority of small migratory land birds in the Holarctic. Kerlinger & Moore (1989, Current Ornithology) identified four main adaptive advantages:

• Reduced predation. Raptors such as sharp-shinned hawks and peregrines forage visually and are essentially absent from the night sky. Nocturnal migrants avoid the highest-risk predator guild for all but the take-off and landing minutes.
• Thermal environment. Night air is cooler and more humid, easing evaporative heat dissipation during sustained flapping flight.
• Smoother winds. Daytime convective mixing gives way to a stable nocturnal boundary layer where winds are laminar and often stronger at 500–1500 m (the low-level jet).
• Daylight for refuelling. Nocturnal flight leaves the whole day for foraging at the stopover site, essential for small birds that must refuel every 1–4 nights.

Small passerines generally migrate at altitudes of 500–2 000 m above ground level. Altitude is chosen dynamically as a function of wind, temperature, and cloud cover. Radar and thermal-imagery data show that in tailwind nights the bulk of birds climb to the level of maximum wind support—often 1 000–1 500 m above ground level in spring and 1 500–2 500 m in fall in North America—while in headwind nights they stay low to reduce ground-speed penalty.

Exceptional cases are well known. Geese, cranes and the Himalayan bar-headed goose reach altitudes above 6 000 m on mountain crossings. Some passerines on trans-ocean flights climb to 4 000 m to exploit the polar-front jet. But for the vast majority of nocturnal passerine migrants, the 500–2000 m stratum captures >90% of the biomass, as confirmed by Doppler radar VAD (Velocity-Azimuth Display) products.

Air density and power

Altitude raises two physical concerns: lower air density increases induced power, and lower partial pressure of oxygen reduces aerobic capacity. For the Pennycuick flight model at a 15 g bird, cruising at 1500 m costs about 4% more mechanical power than at sea level—a modest penalty, comfortably offset by a 2–4 m/s tailwind.

Nocturnal migration was first quantified through moon-watching, pioneered by George H. Lowery at Louisiana State University (Louisiana State University Museum of Zoology, 1951). Observers counted silhouettes passing across the face of a full moon through a telescope, recording direction, timing, and approximate altitude (from apparent size). The method is labour-intensive but gave the first quantitative estimates of migration volume; a good Louisiana night in April produced silhouette counts of 20–80 birds per hour per moon-face, corresponding to densities of 10⁴–10⁵ birds per km³.

Lowery & Newman (1966) extended the analysis to direction-of-flight statistics and showed that nocturnal migrants maintain tight bearing distributions with standard deviations of 10–20°, consistent with a well-calibrated compass. The technique laid the groundwork for later radar-based statistical analyses.

Ceilometer and thermal imaging

Fixed-beam ceilometer spotlights and more recently thermal imaging cameras provide automated analogues of moon-watching. A thermal camera can reliably detect a 12 g bird at 300 m altitude against a clear sky, and modern deep-learning classifiers distinguish passerine wingbeat patterns from bat, insect, and aircraft signatures.

The tool that transformed the field was radar ornithology: using weather-radar networks (principally the WSR-88D NEXRAD network in the US, operated by the National Weather Service since 1988, and the OPERA network in Europe) as continental-scale bird detectors. The same radar that tracks thunderstorms inadvertently sees the dawn and dusk bloom of migrating birds, with a distinctive biological signature:

• Low reflectivity (5–30 dBZ) organised in a thin vertical layer.
• Radial velocity strongly peaked along the migration azimuth.
• Temporal onset at sunset + 30–60 min and offset at sunrise - 30 min.
• Spectrum width narrow (~1 m/s), in contrast to wind-blown insects which show broad spectra.

Horton 2019

Kyle Horton and colleagues at Cornell published the landmark continent-scale analysis in Nature Ecology & Evolution (2019), estimating roughly 4  billion birds per year cross the continental USA at night, with a 25–30% decline since 1970 concordant with population declines reported by the North American Breeding Bird Survey (Rosenberg et al., 2019). The radar-derived flux density maps showed a clear Mississippi-flyway corridor with smaller Atlantic and Pacific flyways, reproducing with unprecedented resolution what flyway biologists had mapped from band recoveries decades earlier.

BirdCast nowcasts

The Cornell Lab of Ornithology and partners operate BirdCast, a real-time radar-ornithology service that produces 24-hour nowcasts of nocturnal migration intensity over the US with 3-hourly temporal resolution (Farnsworth & Russell, 2021). The service supports Lights Out programmes and conservation interventions.

Many nocturnal migrants give short flight calls—50–300 ms tones between 2 and 12 kHz—as they pass overhead. The species-specific call repertoire allows acoustic monitoring of migration with directional microphones on rooftops, feeding back complementary information to radar (which identifies overall biomass but not species).

Evans & O’Brien (2002) catalogued nocturnal flight calls for most North American passerines. More recently, the Old Bird project (William Evans) and the BirdVox machine- learning pipeline (Lostanlen et al., 2018) have automated species identification of NFCs at scale. A typical autumn night at a Northeast US station logs 1 000–5 000 calls, with a strong Swainson’s thrush/American redstart/common yellowthroat dominance.

Why call at all?

The adaptive function of NFCs remains debated. Candidate hypotheses include flock cohesion (birds flying alone in darkness cannot see each other), altitude signalling (calls may encode flight altitude through Doppler shift), and social information exchange about weather or landing sites. The most-cited evidence comes from experimental playbacks that concentrate NFC-giving species relative to silent species, supporting some form of collective behaviour emergence.

Nocturnal migrants fly for 8–12 hours per night, then refuel or roost during the day. Whether they sleep aloft is a recurring question. Rattenborg, Voirin, Cruz and colleagues (Nature Communications, 2016) demonstrated unihemispheric slow-wave sleep (USWS) in great frigatebirds (Fregata minor) during multi-day flights, with total sleep of only 42 min/day aloft vs 12 h/day on land. For small passerines on single-night flights, the working assumption is that sleep is deferred to the diurnal stopover rather than taken in flight, but definitive EEG data are not yet available.

Some long-distance passerines, however, do sustain multi-day flights: the great reed warbler (Acrocephalus arundinaceus) crosses the Mediterranean and the Sahara in a single 3-day non-stop bout (Bairlein et al., 2012). For such cases, USWS in flight is highly plausible even if not yet measured.

Night-to-night variation in migration volume is enormous, as every radar station observes. On a “peak night” in early May over Texas, BirdCast may report 500 million birds aloft; on a headwind night two days later, only 10 million. The difference is driven almost entirely by the synoptic weather pattern. Migrants depart preferentially on:

• Nights with tailwind components of the seasonal bearing (northerly in fall, southerly in spring in the Americas).
• Clear skies (stars visible for celestial calibration).
• Falling pressure ahead of a warm front (spring) or rising pressure behind a cold front (fall).
• No precipitation expected along the flight path.

Fallout events

The converse is the “fallout”: an unexpected storm or headwind forces aloft migrants to descend en masse at the first landfall. Gulf Coast fallout events are famous among birders: a cold front stalling over the Gulf of Mexico in late April causes thousands of neotropical migrants—warblers, tanagers, buntings—to drop into coastal chenier woodlands. These events are ecologically important because they concentrate migrants into small areas where they are vulnerable to cat predation and habitat loss, but also biologically revealing because they reflect the boundary of migratory energetic capacity.

Stopover ecology is the study of what happens between flights. A 12 g Swainson’s thrush flying from Argentina to Yukon completes the journey in 4–5 weeks of actual flight distributed over 60–80 days, with 60–70% of the time spent at stopover sites. The critical insight from Moore, Simons & others in the 1990s is that the stopover sites are not interchangeable: some have high fuel-deposition rates because of abundant insect or fruit resources, others are effectively deserts for the bird.

Gulf Coast funnels

The Gulf Coast chenier woodlands, the Delaware Bay horseshoe-crab beaches, and the Pacific Northwest riparian corridors are well-documented high-value stopover sites. At each, the refueling rate can reach 5–10% body mass per day for the best individuals, enough to support a 300–600 km night of flight for every day of stopover.

Alerstam-Lindstrom optimisation

Alerstam & Lindström (1990) framed the stopover problem as an optimal-foraging trade-off: carrying more fuel extends range but increases flight cost, because induced power scales with \(m^{3/2}\) (and hence the marginal return to each extra gram of fat decreases). Weber & Houston (1997) solved the discrete-stopover optimisation to show that the time-minimising strategy is to carry approximately enough fuel for one hop and no more—a pattern confirmed in field data for many passerines.

Nocturnal migrants are uniquely vulnerable to collisions with illuminated structures. Loss, Will, Loss & Marra (Condor, 2014) synthesised the evidence for bird-building collisions in the USA, arriving at an estimate of 365–988 million birds killed per year by collisions with windows and towers. This is comparable to mortality from feral cats and constitutes a major anthropogenic source of decline.

Light pollution

Artificial lights at night disorient migrants, attracting them into urban airspace and confusing celestial-compass calibration. Gauthreaux & Belser (1999) documented mass attraction of passerines to the 9/11 World Trade Center memorial beams; van Doren, Horton, Dokter, Klinck, Elbin & Farnsworth (2017) quantified the phenomenon across US cities and found radar-detectable deviations from flyway axis near major urban lights.

Lights Out programmes

The operational response is the ‘Lights Out’ campaign, encouraging high-rise buildings to turn off or shield upward-facing lights during peak migration nights. New York, Chicago, Toronto and Washington DC now run Lights Out programmes coordinated with BirdCast nowcasts. Evidence from Van Doren et al. (2021, PNAS) suggests a halving of collision rates during Lights Out periods, though enforcement is voluntary and uneven.

Stephen Emlen’s planetarium experiments with indigo buntings (Emlen 1970, 1975) established that nocturnal passerines learn their star compass as nestlings by observing the rotational centre of the night sky. Birds hand-reared under a planetarium sky rotating around Betelgeuse (rather than the true Polaris) later oriented away from Betelgeuse as their “south” during migratory restlessness. The experiment cleanly disproved innate-map hypotheses for the star compass and showed that the axis of celestial rotation is learned, not hard-wired.

For module 4 the Emlen result matters because it establishes a key dependency of nocturnal migrants on a clear, observable night sky. Light pollution and cloudy nights degrade the star compass, forcing reliance on the magnetic or olfactory back-ups. The dual-redundancy of the compass system is a recurring theme of this course.

The figure below sketches the geometry of a WSR-88D radar observing nocturnal migration. The antenna sweeps in azimuth at a shallow elevation angle (0.5°), sampling the atmospheric volume within ~150 km of the station. Migrating birds cross the beam as diffuse biological scatter with characteristic narrow spectrum width, from which the reflectivity factor, vertical profile, and along-flyway flux can be derived.

Captive migrants held in cages on their normal photoperiod and temperature cycle display a characteristic nocturnal restlessness behaviour known by the German ornithological term Zugunruhe: hopping, wing-whirring, and oriented perching, confined precisely to the hours they would otherwise be in flight. This behaviour was first systematised by Kramer (1949) and then quantified by Emlen’s funnel cage (Emlen & Emlen 1966), which traps ink footprints on sloping paper to record directional preference.

Zugunruhe onset coincides with the physiological programme of migration: pre-migratory hyperphagia, fat deposition to 30–50% body mass, elevated hematocrit, expanded pectoral muscle, and a clock-gated release of corticosterone. The intensity and duration of Zugunruhe measured in captive birds correlates with the migration distance of wild conspecifics—long-distance migrants (e.g. garden warbler, blackpoll warbler) show Zugunruhe for 40–60 nights, whereas short-distance migrants (blackcaps, robins) show only 10–20 nights. Berthold’s (1991) crossing experiments between European blackcap populations demonstrated that the duration is heritable, with a specific genetic locus mapped in 2011 to chromosome 1A.

Photoperiodic triggering

The trigger for Zugunruhe is the change in photoperiod. As autumn shortens day length below a species-specific threshold, the pineal release of melatonin lengthens, and the hypothalamic-pituitary-gonadal axis switches from breeding mode to migratory mode. Artificial shortening of photoperiod in September triggers Zugunruhe in captive birds out of season, providing a powerful laboratory model for the migratory programme.

The transition from stopover to flight is one of the most stressed decisions a small migrant makes. Body condition (fat score), weather ahead, and time pressure all feed into the go-no-go choice. Sjoberg et al. (2015) fitted autumnal reed warblers (Acrocephalus scirpaceus) with micro-dataloggers at Baltic stopover sites and recorded departure altitude profiles. Birds climbed to 2 000–3 000 m in the first hour of flight, then settled to a cruise altitude chosen for maximal tailwind support.

Landing is the mirror image: as dawn approaches, migrants descend through layered stopover-detection mechanisms. They listen for daylight chorus (conspecific signals); they use the earliest sunlight gradients to identify habitat (forest vs open field); they evaluate fuel reserves to decide on a short or long stopover. Newton (2008, The Migration Ecology of Birds) reviews the extensive literature on arrival and stopover decision-making, highlighting the extraordinary sensitivity of small migrants to local weather conditions and habitat quality.

Passerine nocturnal migration is the largest single recurrent movement of vertebrate biomass on the planet. Roughly 4 × 10⁹ birds cross the US every year; comparable numbers cross Europe, the Sahara, and the East Asian flyway. The biophysical apparatus required—celestial compass calibrated on star rotation, magnetic-compass back-up from cryptochrome CRY4 (Module 2), optimal fat loading and stopover (Weber 1997), and a central-pattern-generated wingbeat sustainable for 8 hours of continuous flight—is a major integrative achievement of selection.

Modern radar ornithology has transformed the field from painstaking moon-watching to continent-scale nowcasts. The same tools are uncovering rapid declines in migrant populations and are increasingly used for operational interventions (Lights Out, turbine-curtailment) to protect nocturnal migrants. Module 5 will take up the parallel case of an insect migrant (the monarch butterfly) whose diurnal, multi-generational migration differs entirely in mechanism but faces many of the same threats.