Module 8: Migration & Endurance

The bar-tailed godwit flies 11,000 km non-stop in 8 days without eating, drinking, or sleeping. The Arctic tern completes an annual 70,000 km round-trip between poles. These feats push the biophysical limits of animal performance. This module derives the energetics, fluid balance, and navigational mechanisms that make long-distance migration possible.

1. Fat as Aviation Fuel

1.1 Why Fat?

The energy density of fuel determines how much mass must be carried per unit range — analogous to aircraft fuel efficiency. Comparing metabolic fuels:

Fuel	Energy density (kJ/g)	Metabolic water (g H2O/g)	Respiratory Quotient
Triglyceride (fat)	39.3	1.07	0.71
Carbohydrate (glycogen)	17.2	0.56	1.00
Protein (lean muscle)	18.0	0.40	0.80

Fat provides 2.3× more energy per gram than carbohydrate. Additionally, glycogen is stored hydrated (bound to 3–4 g water per gram glycogen), whereas fat is stored anhydrously in lipid droplets. The effective energy per unit mass of fat stores is therefore approximately 6–8× greater than glycogen stores on a wet-mass basis.

Fat oxidation also produces metabolic water: complete oxidation of a typical triglyceride (tripalmitin, C$_{51}$H$_{98}$O$_6$):

\[ \text{C}_{51}\text{H}_{98}\text{O}_6 + 72.5\ \text{O}_2 \rightarrow 51\ \text{CO}_2 + 49\ \text{H}_2\text{O} \]\[ \text{Metabolic water yield} = \frac{49 \times 18\ \text{g/mol}}{806\ \text{g/mol}} \approx 1.09\ \text{g H}_2\text{O/g fat} \]

During the bar-tailed godwit's 8-day flight, burning ∼55 g fat produces ∼59 g metabolic water — nearly enough to compensate for respiratory and cutaneous water loss, drastically reducing the need for external water.

1.2 Hyperphagia & Pre-Migration Fueling

In the weeks before departure, migratory birds enter hyperphagia: dramatically increased food intake (2–4× basal) regulated by seasonal changes in hypothalamic neuropeptides (NPY, AgRP up; POMC, CART down). This is accompanied by organ remodeling:

Organs that SHRINK

▼Gut (stomach, intestine): 25–50% mass reduction. Reduces protein catabolism during flight but also reduces digestive capacity (offset by post-migration re-growth).
▼Liver: 15–30% reduction. Less metabolic processing capacity needed during sustained flight.
▼Kidney: 20–35% reduction. Less filtration needed (low protein intake during flight).
▼Reproductive organs: essentially regress to zero outside breeding season.

Organs that HYPERTROPHY

▲Flight muscles (pectoralis major): 20–40% increase. Greater power output for sustained flight.
▲Heart: 20–30% increase. Higher cardiac output to supply enlarged flight muscles.
▲Brain: modest increase in certain regions (hippocampus for spatial memory in some species).
▲Subcutaneous + visceral fat deposits: can double body mass (e.g., godwit from 190 g to 380+ g).

2. Breguet Range Equation: Step-by-Step Derivation

The Breguet range equation (from fixed-wing aircraft theory, first derived by Louis Breguet in 1920) applies equally well to birds. We derive it from first principles.

Step 1: Power balance in level flight

For level flight at constant speed $V$, the bird must generate aerodynamic lift$L = W = mg$ (weight) and overcome drag $D$. The mechanical power required:

\[ P_{\text{mech}} = D \cdot V = \frac{W}{L/D} \cdot V \]

$L/D$ = lift-to-drag ratio, typically 10–20 for migrating birds.

Step 2: Metabolic power from fat

The metabolic power is related to mechanical power by the flight muscle efficiency $\eta$(∼0.23–0.27 for flight muscle):

\[ P_{\text{metabolic}} = \frac{P_{\text{mech}}}{\eta} \]

The rate of fuel (fat) consumption:

\[ \dot{m}_{\text{fuel}} = \frac{P_{\text{metabolic}}}{e_f} = \frac{W \cdot V}{\eta \cdot (L/D) \cdot e_f} \]

where $e_f = 39300$ J/g = 39.3 MJ/kg is the specific energy of fat.

Step 3: Mass change differential equation

As fuel is burned, mass decreases. Let $m(t)$ be total bird mass at time $t$,$m_0$ departure mass, $m_f$ arrival mass. Since $W = mg$:

\[ \frac{dm}{dt} = -\dot{m}_{\text{fuel}} = -\frac{mg \cdot V}{\eta (L/D) e_f} \]\[ \Rightarrow \frac{dm}{m} = -\frac{g \cdot V}{\eta (L/D) e_f} dt = -\frac{g}{\eta (L/D) e_f} \cdot \frac{dR}{1} \]

Step 4: Integrate to get the Breguet range equation

Integrating from departure ($m_0$) to arrival ($m_f = m_0 - m_{\text{fuel}}$):

\[ \int_{m_0}^{m_f} \frac{dm}{m} = -\frac{g}{\eta (L/D) e_f} \int_0^R dR \]\[ \ln\frac{m_f}{m_0} = -\frac{g R}{\eta (L/D) e_f} \]\[ \boxed{R = \frac{\eta}{g} \cdot \frac{L}{D} \cdot e_f \cdot \ln\!\left(\frac{m_0}{m_f}\right)} \]

This is the Breguet range equation. The key factors:

●$\eta$: muscle efficiency (∼0.25) — hard to improve biologically
●$L/D$: aerodynamic efficiency (∼10–20) — improved by long pointed wings, V-formation
●$e_f$: specific energy (∼39.3 MJ/kg for fat) — why fat not glycogen
●$\ln(m_0/m_f)$: fuel ratio — doubling body mass gives $\ln(2) \approx 0.693$

Step 5: Numerical check for Bar-tailed Godwit

Parameters: $\eta = 0.25$, $L/D = 16$, $e_f = 39300$ J/g,$m_0 = 380$ g, $m_f = 180$ g (arrival), $g = 9.81$ m/s$^2$:

\[ R = \frac{0.25}{9.81} \times 16 \times 39300 \times \ln\!\left(\frac{380}{180}\right) \times 10^3\ \text{m/km} \]\[ R = \frac{0.25 \times 16 \times 39300}{9.81} \times \ln(2.11)\ \text{m} \approx 16067 \times 0.748 \approx 12018\ \text{km} \]

This theoretical maximum of ∼12,000 km is consistent with the observed 11,000 km Alaska–New Zealand flight — the bird flies near its theoretical range limit with some margin for winds and navigation errors. Favorable tailwinds from the Aleutian Low effectively boost $V$and hence range by 15–20%.

3. Remarkable Migration Feats

Bar-tailed Godwit (Limosa lapponica baueri)

11,000 km non-stop, 8 days (Alaska to New Zealand)

No feeding, drinking, or sleeping. Departs late September using Aleutian Low tailwinds over Pacific. Gut, liver, kidney atrophy to 25% normal mass. Heart + pectoralis hypertrophy. Satellite tracking confirmed by Gill et al. 2009 (Nature). Sleep: unihemispheric slow-wave sleep (USWS) during flight — one brain hemisphere sleeps while other monitors flight.

Arctic Tern (Sterna paradisaea)

~90,000 km round-trip (Arctic to Antarctica and back)

Longest migration of any animal. Follows two different routes: outbound via Africa, return via Americas. Exploits inter-tropical convergence zone (ITCZ) and trade winds for minimal energy expenditure. Tracked by geolocators (Egevang et al. 2010). Lifetime distance: >2.4 million km (equivalent to 3 lunar round trips).

Ruby-throated Hummingbird (Archilochus colubris)

800 km non-stop Gulf of Mexico crossing

Remarkable because hummingbirds have highest mass-specific metabolic rate of any bird (10-14x BMR in hovering flight). Doubles mass (from ~2.5 to ~5 g) before departure with nectar hyperphagia. Body mass ratio: 2.0 → Breguet range ~2800 km theoretical. Actual ~800 km represents efficient routing.

Common Swift (Apus apus)

10 months airborne (longest unbroken flight)

Tracked by miniature geolocators. Swifts eat, sleep, and mate on the wing. Sleep accomplished via USWS. Only land during breeding. Some individuals spend entire non-breeding period (Aug-May) airborne. ~200,000 km flight in 10 months.

4. Navigation, Compasses & Migratory Restlessness

4.1 Star Compass (Stellar Rotation)

Migratory birds learn north as the center of stellar rotation during a critical imprinting window (∼10–65 days post-hatch). Emlen (1969) exposed Indigo buntings (Passerina cyanea) to a planetarium sky with an artificial "north star" — birds oriented relative to the false center. The star compass is learned from the rotation axis, not innate.

Since Polaris is offset from true celestial pole by ∼0.7°, and will drift further over millennia (26,000-year precession cycle), a learned star compass is more robust than a fixed neural constellation map.

4.2 Magnetic Inclination Compass

Migratory birds use an inclination compass: they detect the angle between the magnetic field vector and gravity, not field polarity. This means "poleward" = "toward the shallower field inclination" (toward the equator = increasing inclination).

Wiltschko & Wiltschko (1972) showed European robins (Erithacus rubecula) orient correctly when field strength is halved or doubled, but become disoriented when field lines are made horizontal (no inclination) — confirming inclination sensitivity.

Calibration: the magnetic compass is recalibrated at sunset/sunrise by polarized light sky patterns. Birds can use celestial information to resolve magnetic ambiguities at magnetic anomalies.

4.3 Circadian Clock & Seasonal Timing

Migration timing is controlled by the hypothalamo-pituitary-gonadal (HPG) axis, photoperiodically regulated. The signal cascade:

1.Increasing day length (>critical photoperiod) detected by deep brain photoreceptors in hypothalamus (not retina) — light penetrates skull and directly stimulates TSH (thyroid stimulating hormone) release from pars tuberalis.
2.TSH activates type 2 deiodinase (DIO2) in ependymal cells, converting T4 → T3 locally in the hypothalamus — T3 then stimulates GnRH neurons.
3.GnRH stimulates pituitary FSH/LH release; gonads develop (breeding season) or regress (fall migration). GnIH (gonadotropin-inhibiting hormone) from dorsomedial hypothalamus acts in opposition.
4.Prolactin, corticosterone, and melatonin mediate body weight regulation, hyperphagia, and Zugunruhe timing.

The circadian clock (SCN suprachiasmatic nucleus, period genes Per1, Per2, Cry1, Cry2; clock gene Bmal1) provides the 24-hour base frequency on which photoperiodic measurement is based. BMAL1:CLOCK heterodimers activate E-box promoters of Per and Cry; PER:CRY complexes feed back to inhibit BMAL1:CLOCK. This molecular oscillator runs with a ∼24-hour period (exact period $\tau$ varies 23.5–24.7 h by species).

4.4 Zugunruhe: Quantifying Migratory Restlessness

Zugunruhe (German: "migration restlessness") is the spontaneous locomotor activity shown by caged migratory birds at night during the migratory season. Measured by the Emlen funnel (directional activity detected by ink on paper lining).

The total amount of Zugunruhe predicts migratory distance: blackcap warblers (Sylvia atricapilla) from populations with different migration distances show proportional Zugunruhe duration. Cross-breeding experiments (German blackcaps × African blackcaps) produce hybrids with intermediate Zugunruhe — demonstrating genetic control of migration distance and direction.

Candidate "migratory vector" genes: ADCYAP1 (PACAP neuropeptide precursor), VIP, and alleles in the VIP/ADCYAP1R1 axis correlate with migration distance in several warbler species. Clock gene polymorphisms (especially in Clock poly-Q repeat region) correlate with breeding latitude.

Figure 1: Schematic Migration Routes

5. Python: Breguet Range & Godwit Flight Simulation

Computing maximum flight range vs body mass using the Breguet equation for birds of different sizes, and simulating the mass and fuel budget of the bar-tailed godwit's 11,000 km Alaska–New Zealand flight.

Python

script.py178 lines

import numpy as np
import matplotlib.pyplot as plt
import matplotlib.gridspec as gridspec

fig = plt.figure(figsize=(14, 11))
gs = gridspec.GridSpec(2, 2, figure=fig, hspace=0.44, wspace=0.38)

# ============================================================
# Physical constants & model parameters
# ============================================================
g_accel   = 9.81       # m/s^2
e_fat     = 39.3e3     # J/g = 39.3 MJ/kg (fat energy density)
eta       = 0.25       # flight muscle efficiency
rho_air   = 1.2        # kg/m^3 at sea level
rho_high  = 0.9        # typical cruise altitude ~3000m

# Allometric relationships (Pennycuick 1989 / Alerstam 1990):
# Wing span: b = 1.17 * m^0.39 (m in kg, b in m)
# Wing area: S = 0.16 * m^0.72 (m in kg, S in m^2)
# L/D peaks near optimal speed; typical L/D = 12-18 for migratory shorebirds

def breguet_range(m0_g, fuel_fraction, LoverD=15, eta_val=eta):
    m0 = m0_g / 1000.0   # convert to kg
    mf = m0 * (1 - fuel_fraction)
    R  = (eta_val / g_accel) * LoverD * e_fat * 1000 * np.log(m0 / mf)  # meters
    return R / 1000  # km

# ============================================================
# Panel A: Range vs body mass for different fuel fractions
# ============================================================
ax1 = fig.add_subplot(gs[0, :])

m_range_g = np.logspace(0.5, 4, 200)   # 3 g to 10 kg

fuel_fracs = [0.3, 0.4, 0.5, 0.55]
colors_fuel = ['#60a5fa', '#38bdf8', '#f59e0b', '#f97316']
labels_fuel = ['30% fuel (modest hyperphagia)', '40% fuel', '50% fuel (heavy loading)',
               '55% fuel (max, bar-tailed godwit)']

for ff, col, lab in zip(fuel_fracs, colors_fuel, labels_fuel):
    R_arr = [breguet_range(m, ff) for m in m_range_g]
    ax1.plot(m_range_g, R_arr, color=col, linewidth=2.2, label=lab)

# Mark specific species
species_marks = [
    ('Ruby-throated\nHummingbird', 3.5,  0.50,  '#fb923c'),
    ('Garden Warbler',             20,   0.50,  '#a78bfa'),
    ('Dunlin',                     55,   0.40,  '#34d399'),
    ('Red Knot',                   200,  0.50,  '#38bdf8'),
    ('Bar-tailed Godwit',          380,  0.55,  '#f59e0b'),
    ('Whooper Swan',               10000, 0.30, '#f472b6'),
]

for (name, m_g, ff, col) in species_marks:
    R_val = breguet_range(m_g, ff)
    ax1.scatter([m_g], [R_val], color=col, zorder=6, s=90, edgecolors='white', linewidth=0.8)
    y_off = 300 if name == 'Bar-tailed Godwit' else 150
    ax1.annotate(name, xy=(m_g, R_val), xytext=(m_g * 1.5, R_val + y_off),
                fontsize=7.5, color=col,
                arrowprops=dict(arrowstyle='->', color=col, lw=0.8))

# Godwit actual record
ax1.axhline(11000, color='#f59e0b', linewidth=1.5, linestyle='--', alpha=0.6,
            label='Godwit actual record: 11,000 km')

ax1.set_xscale('log')
ax1.set_xlabel('Body mass (g) at departure', fontsize=11)
ax1.set_ylabel('Maximum range (km)', fontsize=11)
ax1.set_title('Breguet Range Equation: Maximum Flight Range vs Body Mass\n'
              r'$R = rac{eta}{g}cdotrac{L}{D}cdot e_f cdot ln(m_0/m_f)$   (L/D=15)', fontsize=11)
ax1.legend(fontsize=8.5, loc='upper left')
ax1.grid(True, alpha=0.3, which='both')
ax1.set_xlim(3, 12000)
ax1.set_ylim(0, 18000)

# ============================================================
# Panel B: Godwit 11,000 km flight mass budget
# ============================================================
ax2 = fig.add_subplot(gs[1, 0])

distance_km = np.linspace(0, 11000, 1000)

# Initial masses (g)
m0_total        = 380.0
m0_fat          = 200.0    # ~53% of body mass
m0_protein_lean = 80.0     # flight muscle (some catabolized late)
m0_structure    = 100.0    # bones, other tissues (constant)

# Power consumption: ~8x BMR = 8 * 0.01 * m^0.67 W for a godwit
# Approximate: total fuel per km from Breguet
LoverD = 16.0
# From dR = (eta * L/D * e_fat / g) * dm/m  ->  dm = m * g * dR / (eta * L/D * e_fat)
# Solve numerically
m_total = np.zeros(len(distance_km))
m_fat   = np.zeros(len(distance_km))
m_prot  = np.zeros(len(distance_km))
m_total[0] = m0_total
m_fat[0]   = m0_fat
m_prot[0]  = m0_protein_lean

dR = (distance_km[1] - distance_km[0]) * 1000  # meters per step

for i in range(1, len(distance_km)):
    m_prev = m_total[i-1]
    # Breguet: dm = m * g * dR / (eta * LoverD * e_fat * 1000)
    dm_fuel = m_prev * g_accel * dR / (eta * LoverD * e_fat * 1000)
    # Late in flight (fat < 20% remaining) also catabolize some protein
    fat_frac = m_fat[i-1] / m_prev
    if fat_frac > 0.15:
        dm_fat  = dm_fuel
        dm_prot = 0.0
    else:
        dm_fat  = dm_fuel * 0.75
        dm_prot = dm_fuel * 0.25
    m_fat[i]   = max(0, m_fat[i-1]   - dm_fat)
    m_prot[i]  = max(0, m_prot[i-1]  - dm_prot)
    m_total[i] = m0_structure + m_fat[i] + m_prot[i]

# Metabolic water produced (1.07 g per g fat burned)
water_produced = (m0_fat - m_fat) * 1.07

ax2.plot(distance_km, m_total, color='#a78bfa', linewidth=2.5, label='Total body mass')
ax2.fill_between(distance_km, m0_structure, m_total, alpha=0.25, color='#a78bfa')
ax2.plot(distance_km, m_fat + m0_structure, color='#f59e0b', linewidth=2, label='Structure + fat')
ax2.fill_between(distance_km, m0_structure, m_fat + m0_structure, alpha=0.2, color='#f59e0b', label='Fat reserves')
ax2.axhline(m0_structure, color='#94a3b8', linewidth=1, linestyle=':', label='Structural mass (constant)')
ax2.axvline(11000, color='#4ade80', linewidth=1.5, linestyle='--', alpha=0.7, label='New Zealand (11,000 km)')

ax2.set_xlabel('Distance flown (km)', fontsize=10)
ax2.set_ylabel('Mass (g)', fontsize=10)
ax2.set_title('Godwit Mass Budget\nAlaska to New Zealand (non-stop)', fontsize=10)
ax2.legend(fontsize=7.5)
ax2.grid(True, alpha=0.3)
ax2.set_xlim(0, 11000)
ax2.annotate(f'Arrival mass: {m_total[-1]:.0f} g\n({100*(m0_total-m_total[-1])/m0_total:.0f}% mass loss)',
             xy=(10800, m_total[-1]), xytext=(7000, 220),
             fontsize=8.5, color='#a78bfa',
             arrowprops=dict(arrowstyle='->', color='#a78bfa'))

# ============================================================
# Panel C: Metabolic water and hydration balance
# ============================================================
ax3 = fig.add_subplot(gs[1, 1])

# Water loss: respiratory (~0.4 mg/km/g initial mass for birds in flight)
water_loss_resp = 0.40 * m0_total * distance_km / 1000  # mg -> g/1000
# Cutaneous water loss: ~0.3 mg/km/g
water_loss_skin = 0.30 * m0_total * distance_km / 1000
water_loss_total = water_loss_resp + water_loss_skin

water_balance = water_produced - water_loss_total  # positive = net gain

ax3.plot(distance_km, water_produced,   color='#38bdf8', linewidth=2.5, label='Metabolic water produced (g)')
ax3.plot(distance_km, water_loss_total, color='#f97316', linewidth=2.5, linestyle='--', label='Total water loss (g)')
ax3.fill_between(distance_km, water_loss_total, water_produced,
                 where=(water_produced > water_loss_total),
                 alpha=0.2, color='#38bdf8', label='Net water surplus')
ax3.fill_between(distance_km, water_loss_total, water_produced,
                 where=(water_produced <= water_loss_total),
                 alpha=0.2, color='#f97316', label='Net water deficit')
ax3.axvline(11000, color='#4ade80', linewidth=1.5, linestyle='--', alpha=0.7)
ax3.set_xlabel('Distance flown (km)', fontsize=10)
ax3.set_ylabel('Cumulative water (g)', fontsize=10)
ax3.set_title('Hydration Balance During Godwit Flight\n(1.07 g H2O per g fat oxidized)', fontsize=10)
ax3.legend(fontsize=8)
ax3.grid(True, alpha=0.3)
ax3.set_xlim(0, 11000)

final_balance = water_balance[-1]
ax3.annotate(f'Net balance at 11,000 km:\n{final_balance:.1f} g ({"+".join([""] if final_balance>=0 else [])}{final_balance:.1f} g)',
             xy=(11000, water_produced[-1]),
             xytext=(6500, water_produced[-1] * 0.65),
             fontsize=8.5, color='#38bdf8',
             arrowprops=dict(arrowstyle='->', color='#38bdf8'))

fig.suptitle('Avian Long-Distance Migration: Breguet Range & Energy Budget', fontsize=13, fontweight='bold')
plt.savefig('output.png', dpi=130, bbox_inches='tight', facecolor='#0f172a')

Click Run to execute the Python code

Code will be executed with Python 3 on the server

6. Sleep During Flight & Unihemispheric Processing

A major physiological puzzle of multi-day non-stop migration was: how do birds manage sleep deprivation? Frigate birds monitored by EEG during soaring flight (Rattenborg et al. 2016) showed they sleep during flight, but only for ~0.69 hours/day— versus ∼12 hours at rest. They use two mechanisms:

Unihemispheric Slow-Wave Sleep (USWS)

One cerebral hemisphere enters slow-wave sleep (SWS) while the other remains awake. The sleeping hemisphere's contralateral eye closes; the awake hemisphere's eye stays open for visual monitoring. This allows rest of half the brain while maintaining flight control and environmental awareness. Delta waves (0.5–4 Hz, SWS signature) recorded from sleeping hemisphere simultaneously with wake patterns in the other.

Rapid Eye Movement (REM) in Flight

Very brief REM sleep episodes (<3 seconds) were also recorded in flying frigate birds. During REM, muscle atonia normally occurs — but flying birds suppress this to maintain wing beats. These microsleeps may serve synaptic consolidation functions. Interestingly, REM sleep during soaring (using thermals, minimal flapping) was more common than during powered flight.

Sleep pressure (homeostatic sleep drive, adenosine-mediated) accumulates more slowly in flight than at rest — possibly because the aerobic exercise itself has restorative effects (exercise-induced BDNF production supports synaptic maintenance), and because preflight sleep banking (sleeping more in the days before migration begins) partially pre-compensates for in-flight sleep loss.

References

Hedenström, A. (2010). Extreme endurance migration: what is the limit to non-stop flight? PLoS Biology, 8, e1000362.
Gill, R. E. et al. (2009). Extreme endurance flights by landbirds crossing the Pacific Ocean. Proceedings of the Royal Society B, 276, 447–457.
Alerstam, T. (2011). Optimal bird migration revisited. Journal of Ornithology, 152, 5–23.
Piersma, T. & van Gils, J. A. (2011). The Flexible Phenotype: A Body-Centred Integration of Ecology, Physiology, and Behaviour. Oxford University Press.
Pennycuick, C. J. (2008). Modelling the Flying Bird. Academic Press.
Weber, J.-M. (2009). The physiology of long-distance migration: extending the limits of endurance metabolism. Journal of Experimental Biology, 212, 593–597.
Battley, P. F. et al. (2012). Contrasting extreme long-distance migration patterns in bar-tailed godwits. Journal of Avian Biology, 43, 21–32.
Gill, F. B. (2007). Ornithology, 3rd ed. W. H. Freeman.

←Song & Acoustics Course Overview→