← Flight Aerodynamics Next: Avian Respiration →

Module 2: Feather Biochemistry

Feathers are among the most structurally sophisticated biological materials known. Evolved from reptilian scales approximately 150 million years ago, they achieve extraordinary multifunctionality: aerodynamic lift, insulation, waterproofing, camouflage, and vivid structural coloration — all from a single protein, beta-keratin, elaborated into hierarchically organised microarchitectures.

1. Beta-Keratin: The Structural Protein of Feathers

Molecular Structure

Unlike the alpha-helical keratin of mammalian hair and nails, avian feather keratin adopts a beta-pleated sheet secondary structure. The polypeptide chain ($\sim 100$ amino acids, MW $\sim 10\,\text{kDa}$) forms extended antiparallel beta-strands stabilised by:

Interstrand H-bonds: N-H···O=C between adjacent strands ($d \approx 0.29\,\text{nm}$), spacing 0.47 nm between strands
Disulfide crosslinks: Cys-Cys bridges ($-S-S-$) between chains, $E_{bond} \approx 250\,\text{kJ/mol}$
Hydrophobic core: Val, Ile, Leu side chains pack into a low-entropy core

Mechanical Properties

Young's modulus (E)

~2.5 GPa

Comparable to nylon; 10× more flexible than bone (20 GPa)

Tensile strength (σ_max)

~200 MPa

Stronger than bone in tension; flexible enough to bend without fracture

Strain at failure (ε)

~10–15%

Substantial plastic deformation before fracture — critical for vane integrity

The stress-strain curve of feather rachis exhibits an initial linear elastic region (governed by $\sigma = E\epsilon$) followed by a yield plateau attributed to beta-sheet unfolding and H-bond rupture. This strain-stiffening/softening behaviour is analogous to spider silk and provides energy absorption during impact (landing, collision).

Cysteine Content and Crosslink Density

Flight feather beta-keratin contains ~7% cysteine residues — relatively high compared to mammalian hair (4%). Each disulfide bond crosslinks two polypeptide chains, creating a covalent network that resists creep and fatigue. Contour feathers near the body contain more cysteine (better water resistance) while flight feather barbules optimise stiffness-to-weight ratio instead.

2. Feather Morphology: A Hierarchical Velcro

Feather architecture spans six levels of hierarchy, from molecular to macroscopic:

The Hooklet-Groove Mechanism

Flight feather barbules interlock via a biological Velcro system: distal barbules carry hooklets (hamuli, 150–200 per barbule) that engage with the grooved edge of the adjacent proximal barbule. Each hooklet exerts $\sim 1\,\text{nN}$ when engaged. A flight feather has $\sim 10^6$ hooklet-groove contacts — total interlocking force ~1 mN, sufficient to maintain vane integrity under aerodynamic loads of$\sim 100\,\text{Pa}$ while being freely unzipable (preening).

3. Melanin Biosynthesis: The Raper-Mason Pathway

Melanin pigments are synthesised from the amino acid tyrosine via an oxidative cascade catalysed primarily by tyrosinase (TYR), with branches controlled by TYRP1 and DCT (dopachrome tautomerase).

L-Tyrosine

TYR (tyrosinase)

→

L-DOPA

TYR

→

Dopaquinone

→

Leucodopachrome

Spontaneous

→

Dopachrome

→ EUMELANIN branch (black/brown)

Dopachrome → DHI/DHICA (DCT) → 5,6-dihydroxyindole → eumelanin polymer

Absorption: broad visible spectrum, UV protection. Black plumage in crows, ravens

→ PHEOMELANIN branch (red/yellow)

Dopaquinone + cysteine → cysteinyldopa → benzothiazine → pheomelanin

Shorter-wavelength absorption. Red-yellow plumage in goldfinch, Robin (in part)

Key Enzymes

Tyrosinase (TYR)

Hydroxylates Tyr→DOPA (monophenolase) and oxidises DOPA→dopaquinone (diphenolase). Rate-limiting, Cu-containing active site (2 Cu²⁺ per enzyme)

TYRP1

DHICA oxidase; determines melanin polymer chain length and degree of oxidation; mutations cause dilute coloration phenotypes in birds

DCT (TYRP2)

Dopachrome tautomerase; shifts equilibrium toward DHICA (slower oxidation) vs DHI; controls eumelanin/pheomelanin ratio

4. Carotenoid Pigmentation: Dietary Colour

Unlike melanins, birds cannot synthesise carotenoids de novo — they must obtain them from their diet (insects, fruits, algae). Once ingested, carotenoids can be:

Deposited directly — dietary zeaxanthin, lutein deposited in feather follicles
Enzymatically modified — a ketolase enzyme (CYP2J19, identified 2016) converts yellow dietary carotenoids (zeaxanthin, lutein) into red ketocarotenoids (astaxanthin, canthaxanthin) in species like cardinals, flamingos

Chemical Basis of Colour

Carotenoids are polyene pigments with alternating C=C double bonds. The delocalized$\pi$-electron system allows photon absorption in the blue-green range (430–500 nm), reflecting red-yellow. Adding carbonyl groups (ketolation: C=O) extends the conjugation, red-shifting absorption further into the green (550–580 nm) — producing brilliant reds. The chain length determines colour: lycopene (11 C=C, red), beta-carotene (9 C=C, orange), zeaxanthin (9 C=C + hydroxyl, yellow).

Honest Signalling: Why Carotenoid Colour is Costly

Because carotenoids are obtained from food, plumage redness honestly signals foraging ability and parasite resistance. Additionally, carotenoids serve as antioxidants (quenching reactive oxygen species), creating a trade-off: bright birds invest carotenoids in display rather than immune protection. This is the "carotenoid allocation hypothesis" — experimentally supported in House Finches and Zebra Finches.

5. Structural Coloration: Physics of Iridescence

Fiery-throated Hummingbird showing iridescent structural coloration across its plumage

Fiery-throated Hummingbird (Panterpe insignis) — every colour on this bird's throat and breast is structural, not pigmentary. Thin-film interference in melanin-keratin nanostructures within the barbules produces angle-dependent iridescence spanning the entire visible spectrum.

Image: Joseph F. Pescatore

5.1 Thin-Film Interference

In species like the peacock, Common Kingfisher, and Hummingbirds, colour arises from interference of light waves reflected at multiple interfaces within barbule nanostructures. The barbule cortex is a stack of alternating layers: $\beta$-keratin ($n_k = 1.56$) and melanin granules ($n_m = 2.0$), each 100–400 nm thick, sandwiched between air ($n_0 = 1.0$). When white light enters, each interface partially reflects and partially transmits. The reflected waves from successive interfaces interfere — constructively for certain wavelengths, destructively for others — producing vivid colour.

Derivation: Optical Path Difference

Consider a ray incident at angle $\theta_i$ on a thin film of refractive index $n$and thickness $d$. By Snell's law the refracted angle satisfies:

\[ n_1 \sin\theta_i = n_2 \sin\theta_t \quad \Rightarrow \quad \theta_t = \arcsin\!\left(\frac{n_1}{n_2}\sin\theta_i\right) \]

The ray refracted into the film travels a geometric path of length $d / \cos\theta_t$ to reach the second interface, then reflects and traverses the same distance back. Meanwhile, the ray reflected at the first interface has already progressed along the surface. The difference in optical path is:

\[ \Delta = 2\,n\,\frac{d}{\cos\theta_t} - 2\,d\,\tan\theta_t \sin\theta_i \]

Using Snell's law ($\sin\theta_i = (n/n_1)\sin\theta_t$) and simplifying:

\[ \Delta = 2\,n\,d\,\cos\theta_t \]

Constructive interference occurs when $\Delta$ equals an integer number of wavelengths:

\[ 2 n d \cos\theta_t = m \lambda \quad (m = 1, 2, 3, \ldots) \]

Constructive interference at wavelength $\lambda$ when optical path difference equals an integer multiple of the wavelength. Blue-green colour: $d \approx 100\text{–}150\,\text{nm}$; green: $d \approx 150\text{–}200\,\text{nm}$; red: $d \approx 300\text{–}400\,\text{nm}$.

Derivation: Fresnel Reflectance

The amplitude reflection coefficient at the interface between media $i$ and $j$(for normal incidence, s-polarization) follows from the electromagnetic boundary conditions:

\[ r_{ij} = \frac{n_i - n_j}{n_i + n_j}, \qquad t_{ij} = \frac{2 n_i}{n_i + n_j} \]

For air/keratin: $r_{ak} = (1.0 - 1.56)/(1.0 + 1.56) = -0.219$; keratin/melanin: $r_{km} = (1.56 - 2.0)/(1.56 + 2.0) = -0.124$.

The total reflectance of a thin film (two-beam approximation) sums the first-surface and second-surface reflections with their phase difference $\delta = 4\pi n d \cos\theta_t / \lambda$:

\[ R = \frac{r_1^2 + r_2^2 + 2 r_1 r_2 \cos\delta}{1 + r_1^2 r_2^2 + 2 r_1 r_2 \cos\delta} \]

This is the Airy formula for a Fabry-Pérot cavity. For melanin layers, absorption must also be included via the extinction coefficient $k \approx 0.3$: the transmitted amplitude is attenuated by $e^{-\alpha d}$ where $\alpha = 4\pi k / \lambda$.

Iridescence: Angle-Dependent Colour

As the viewing angle increases (larger $\theta_i$), $\cos\theta_t$ increases (since$\theta_t < \theta_i$ in a denser medium), which means the product $2nd\cos\theta_t$ grows. To maintain constructive interference ($= m\lambda$), the peak wavelength $\lambda$ must decrease — a blue-shift. This is exactly what is observed in peacock feathers, hummingbirds, and starlings: they shimmer from green to blue as you change viewing angle.

\[ \lambda_{peak}(\theta) = \frac{2 n d}{m} \cos\!\left[\arcsin\!\left(\frac{\sin\theta_i}{n}\right)\right] \]

For a peacock barbule ($d = 150\,\text{nm}$, $n = 2.0$, $m = 1$):$\lambda_{peak}(0°) = 600\,\text{nm}$ (green) →$\lambda_{peak}(60°) \approx 520\,\text{nm}$ (blue-green).

5.2 Why Blue Birds Have No Blue Pigment

$How birds produce structural blue colour — wrinkled rachis surface diffracts blue light while smooth surface reflects all wavelengths equally$

Structural blue in feathers. Birds produce blue in barbs, barbules, and even the rachis. A wrinkled rachis surface acts as a diffraction grating, selectively scattering blue wavelengths while absorbing longer wavelengths via an underlying melanin layer. A smooth rachis surface reflects all wavelengths (white/specular reflection).

Inside each barb of a blue feather sits a spongy matrix of $\beta$-keratin threaded with air voids roughly 100–200 nm across. The voids are not randomly arranged — they form a quasi-ordered pattern with a characteristic spacing. When white light enters this medium, short-wavelength (blue) components scatter coherently: waves reflected from neighboring air–keratin interfaces interfere constructively at blue wavelengths and destructively at longer ones. A layer of melanin granules behind the spongy matrix absorbs whatever light slips through, so only the blue reflection escapes back to the viewer.

Not Rayleigh Scattering — Coherent Photonic Nanostructure

This mechanism was once lumped together with Rayleigh scattering (the classic “Tyndall blue” explanation for the sky), but electron microscopy and optical modelling from the 1990s onward — largely the work of Richard Prum and collaborators — showed the voids are too uniform and too closely packed for Rayleigh. They behave as a quasi-ordered photonic structure, and the peak wavelength is set by the Fourier transform of the void spatial distribution, not by a $1/\lambda^4$ law.

Derivation: Coherent Scattering from a Quasi-Ordered Array

For an array of scatterers (air voids) with a spatial correlation described by the radial distribution function $g(r)$, the scattered intensity is proportional to the structure factor:

\[ I(q) \propto |F(q)|^2 \cdot S(q), \qquad S(q) = 1 + \rho \int [g(r) - 1] e^{i\mathbf{q}\cdot\mathbf{r}} d\mathbf{r} \]

where $q = 4\pi n \sin(\theta/2)/\lambda$ is the scattering wavevector,$F(q)$ is the form factor of individual voids, and $S(q)$ encodes the spatial correlations. For quasi-ordered arrays with a characteristic spacing $a$,$S(q)$ peaks sharply near $q^* = 2\pi/a$.

The peak reflected wavelength in backscattering ($\theta = 180°$) is therefore:

\[ \lambda_{peak} = 2 n_{avg} \cdot a \]

where $n_{avg}$ is the volume-averaged refractive index of the spongy matrix and $a \approx 150\text{–}200\,\text{nm}$ is the characteristic void spacing. This gives $\lambda_{peak} \approx 400\text{–}500\,\text{nm}$ — blue.

Two Decisive Experiments

1.Crush test: Crush a blue feather between your fingers — it turns grayish-brown. You have destroyed the nanostructure; only the melanin backing remains. If the colour were pigmentary, crushing would not remove it.
2.Backlight test: Backlight a blue feather and it appears brown, because there is nothing blue to transmit — the colour exists only in reflection. Pigmentary colours (carotenoid reds, melanin browns) transmit their colour.

The Comparative Colour Palette of Birds

The comparative picture is worth keeping in mind:

Reds, oranges, yellows

Carotenoids — acquired through diet (insects, fruits, algae). Cannot be synthesized de novo.

Blacks, browns, grays

Melanins — synthesized endogenously in melanocytes from tyrosine via the melanogenesis pathway.

Greens

Usually structural blue layered over yellow carotenoid. True pigmentary green is rare — turacos are the notable exception (copper-based turacoverdin).

Blues

Always structural — never pigmentary in birds. The one colour birds have never solved chemically, so they solved it geometrically.

Simulation: Thin-Film Interference Spectra

Structural Colour: Thin-Film Interference and Melanin Biosynthesis

Python

script.py189 lines

import numpy as np
import matplotlib
matplotlib.use('Agg')
import matplotlib.pyplot as plt

# Thin-film interference: reflectance spectrum from melanin layers
# in barbule nanostructures (peacock/iridescent feathers)
# Model: air / keratin (n=1.56) / melanin (n=2.0) / keratin / air
# Two-beam approximation for thin melanin layer

wavelengths = np.linspace(350, 750, 800)  # nm
n_melanin = 2.0    # refractive index of eumelanin
n_keratin = 1.56   # refractive index of beta-keratin
n_air     = 1.0
theta_i   = 0.0    # normal incidence (degrees)

# Fresnel reflection coefficients at interfaces (normal incidence)
# r_12 at air/keratin, r_23 at keratin/melanin
r_ak = (n_air - n_keratin) / (n_air + n_keratin)
r_km = (n_keratin - n_melanin) / (n_keratin + n_melanin)

thicknesses = [100, 150, 200, 250, 300, 380, 450, 500]  # nm melanin layer

fig, axes = plt.subplots(2, 2, figsize=(14, 10))
fig.patch.set_facecolor('#030712')
axes = axes.flatten()

colors_plot = ['#22d3ee','#34d399','#a78bfa','#f0abfc','#fbbf24','#f87171','#fb923c','#60a5fa']

# Panel 1: Reflectance spectra for different thicknesses
ax1 = axes[0]
ax1.set_facecolor('#0c1a2e')

reflectances_at_peak = []
peak_wavelengths = []

for idx, d in enumerate(thicknesses):
    # Optical path difference: delta = 2 * n_melanin * d * cos(theta_t)
    # theta_t from Snell's law: sin(theta_t) = n_air * sin(theta_i) / n_melanin
    theta_r = np.radians(theta_i)
    theta_t = np.arcsin(n_air * np.sin(theta_r) / n_melanin)

# Phase difference
    delta = 4 * np.pi * n_melanin * d * np.cos(theta_t) / wavelengths  # radians

# Two-beam approximation: R = r1^2 + r2^2 + 2*r1*r2*cos(delta)
    # Account for absorption in melanin: add imaginary component
    k_mel = 0.3  # extinction coefficient
    alpha = 4 * np.pi * k_mel / wavelengths * d  # attenuation
    transmission = np.exp(-alpha)

r1 = abs(r_ak)
    r2 = abs(r_km) * transmission

R = (r1**2 + r2**2 + 2*r1*r2*np.cos(delta)) / (1 + r1**2 * r2**2 + 2*r1*r2*np.cos(delta))

ax1.plot(wavelengths, R * 100, color=colors_plot[idx], linewidth=1.8,
             label=f"d = {d} nm", alpha=0.9)

peak_idx = np.argmax(R)
    reflectances_at_peak.append(R[peak_idx] * 100)
    peak_wavelengths.append(wavelengths[peak_idx])

ax1.set_xlabel("Wavelength (nm)", color='white', fontsize=11)
ax1.set_ylabel("Reflectance (%)", color='white', fontsize=11)
ax1.set_title("Thin-Film Reflectance Spectra\n(melanin layer in keratin matrix)", color='#7dd3fc', fontsize=11)
ax1.tick_params(colors='white')
ax1.grid(True, alpha=0.2, color='#38bdf8')
for sp in ax1.spines.values(): sp.set_color('#38bdf8')
ax1.legend(fontsize=8, facecolor='#0c1a2e', edgecolor='#38bdf8', labelcolor='white', ncol=2)

# Add visible spectrum color bar
spectrum_x = np.linspace(380, 700, 100)
for i in range(len(spectrum_x)-1):
    wl = spectrum_x[i]
    if wl < 440: c = (0.5*(440-wl)/60, 0, (wl-380)/60)
    elif wl < 490: c = (0, (wl-440)/50, 1)
    elif wl < 510: c = (0, 1, (510-wl)/20)
    elif wl < 580: c = ((wl-510)/70, 1, 0)
    elif wl < 645: c = (1, (645-wl)/65, 0)
    else: c = (1, 0, 0)
    ax1.axvspan(spectrum_x[i], spectrum_x[i+1], ymin=0, ymax=0.06, color=c, alpha=0.6)

# Panel 2: Peak wavelength vs thickness
ax2 = axes[1]
ax2.set_facecolor('#0c1a2e')
ax2.scatter(thicknesses, peak_wavelengths, c=colors_plot[:len(thicknesses)],
            s=100, zorder=5, edgecolors='white', linewidth=0.8)

# Theoretical: lambda_peak = 2 * n * d (first order)
d_range = np.linspace(80, 550, 300)
lambda_theory_1 = 2 * n_melanin * d_range  # m=1
lambda_theory_2 = 2 * n_melanin * d_range / 2  # m=2

ax2.plot(d_range, lambda_theory_1, color='#38bdf8', linestyle='--', linewidth=1.5,
         label="2nd-order: 2nd*n*d = m*lambda (m=1)")
ax2.plot(d_range, lambda_theory_2, color='#f0abfc', linestyle='--', linewidth=1.5,
         label="2nd-order: m=2")

# Color band backgrounds
ax2.axhspan(380, 440, alpha=0.08, color='purple')
ax2.axhspan(440, 495, alpha=0.08, color='blue')
ax2.axhspan(495, 570, alpha=0.08, color='green')
ax2.axhspan(570, 620, alpha=0.08, color='yellow')
ax2.axhspan(620, 750, alpha=0.08, color='red')

ax2.text(410, 410, 'V', color='#a78bfa', fontsize=9)
ax2.text(410, 460, 'B', color='#60a5fa', fontsize=9)
ax2.text(410, 520, 'G', color='#34d399', fontsize=9)
ax2.text(410, 595, 'Y', color='#fde68a', fontsize=9)
ax2.text(410, 680, 'R', color='#f87171', fontsize=9)

ax2.set_xlim(80, 560); ax2.set_ylim(380, 750)
ax2.set_xlabel("Melanin Layer Thickness (nm)", color='white', fontsize=11)
ax2.set_ylabel("Peak Reflected Wavelength (nm)", color='white', fontsize=11)
ax2.set_title("Peak Color vs Layer Thickness\n(structural coloration map)", color='#7dd3fc', fontsize=11)
ax2.tick_params(colors='white')
ax2.grid(True, alpha=0.2, color='#38bdf8')
for sp in ax2.spines.values(): sp.set_color('#38bdf8')
ax2.legend(fontsize=8, facecolor='#0c1a2e', edgecolor='#38bdf8', labelcolor='white')

# Panel 3: Angle dependence (iridescence)
ax3 = axes[2]
ax3.set_facecolor('#0c1a2e')

d_peacock = 150  # nm - green peacock feather
angles = np.linspace(0, 70, 200)

for wl_probe in [450, 500, 550, 600, 650]:
    R_vs_angle = []
    for ang in angles:
        ang_r = np.radians(ang)
        theta_t2 = np.arcsin(n_air * np.sin(ang_r) / n_melanin)
        delta2 = 4 * np.pi * n_melanin * d_peacock * np.cos(theta_t2) / wl_probe
        alpha2 = 4 * np.pi * 0.3 / wl_probe * d_peacock
        t2 = np.exp(-alpha2)
        R_a = (r_ak**2 + (r_km*t2)**2 + 2*abs(r_ak)*abs(r_km)*t2*np.cos(delta2))
        R_vs_angle.append(min(1.0, abs(R_a)) * 100)

if wl_probe == 450: col = '#60a5fa'
    elif wl_probe == 500: col = '#34d399'
    elif wl_probe == 550: col = '#a3e635'
    elif wl_probe == 600: col = '#fbbf24'
    else: col = '#f87171'
    ax3.plot(angles, R_vs_angle, color=col, linewidth=2, label=f"{wl_probe} nm")

ax3.set_xlabel("Viewing Angle (degrees from normal)", color='white', fontsize=11)
ax3.set_ylabel("Reflectance (%)", color='white', fontsize=11)
ax3.set_title("Iridescence: Reflectance vs Viewing Angle\n(d = 150 nm, peacock-like)", color='#7dd3fc', fontsize=11)
ax3.tick_params(colors='white')
ax3.grid(True, alpha=0.2, color='#38bdf8')
for sp in ax3.spines.values(): sp.set_color('#38bdf8')
ax3.legend(fontsize=9, facecolor='#0c1a2e', edgecolor='#38bdf8', labelcolor='white')

# Panel 4: Melanin biosynthesis pathway (bar chart of enzyme activities)
ax4 = axes[3]
ax4.set_facecolor('#0c1a2e')

steps = ['Tyrosine', 'DOPA', 'DOPAquinone', 'Eumelanin\
(+ DHI)', 'Pheomelanin\
(+ cysteine)']
eumelanin_activity = [100, 85, 70, 60, 10]
pheomelanin_activity = [100, 85, 70, 15, 65]

x_pos = np.arange(len(steps))
width = 0.35
ax4.bar(x_pos - width/2, eumelanin_activity, width, color='#374151', edgecolor='#22d3ee',
        linewidth=1.5, label='Eumelanin pathway (brown/black)', alpha=0.9)
ax4.bar(x_pos + width/2, pheomelanin_activity, width, color='#78350f', edgecolor='#fbbf24',
        linewidth=1.5, label='Pheomelanin pathway (red/yellow)', alpha=0.9)

ax4.set_xticks(x_pos)
ax4.set_xticklabels(steps, fontsize=9, color='white')
ax4.set_ylabel("Relative Pathway Flux (%)", color='white', fontsize=11)
ax4.set_title("Melanin Biosynthesis Pathway\n(Relative enzyme flux)", color='#7dd3fc', fontsize=11)
ax4.tick_params(colors='white', labelsize=9)
ax4.grid(True, alpha=0.2, color='#38bdf8', axis='y')
for sp in ax4.spines.values(): sp.set_color('#38bdf8')
ax4.legend(fontsize=9, facecolor='#0c1a2e', edgecolor='#38bdf8', labelcolor='white')
ax4.set_ylim(0, 115)

for i, (e, p) in enumerate(zip(eumelanin_activity, pheomelanin_activity)):
    ax4.text(i - width/2, e+2, str(e), ha='center', color='#22d3ee', fontsize=8)
    ax4.text(i + width/2, p+2, str(p), ha='center', color='#fbbf24', fontsize=8)

plt.tight_layout()
plt.savefig('output.png', dpi=150, bbox_inches='tight', facecolor='#030712')

Click Run to execute the Python code

Code will be executed with Python 3 on the server

6. Preen Oil and the Uropygial Gland

The uropygial gland (preen gland) is a bilobed sebaceous gland at the base of the tail, present in most birds. It secretes a complex mixture of wax esters, free fatty acids, triglycerides, and squalene.

Waterproofing Mechanism

Wax esters (C16–C22 fatty acids esterified with C16–C22 fatty alcohols) spread over the feather barbule surface, creating a hydrophobic monolayer. Contact angle of water on treated feathers: $\theta \approx 150°$ (superhydrophobic). Critical for aquatic birds — without preening, plumage becomes waterlogged ($\Delta\rho \approx 0.3\,\text{g/cm}^3$).

Vitamin D₃ Precursors

Preen oil contains 7-dehydrocholesterol — a vitamin D₃ precursor. When spread on feathers and exposed to UV-B (290–315 nm), photolysis produces pre-vitamin D₃:$7\text{-DHC} \xrightarrow{h\nu} \text{pre-D}_3 \xrightarrow{\Delta} \text{Vitamin D}_3$. Birds ingest this during preening — a unique exogenous photosynthesis strategy.

Chemical Ecology of Preen Oil

Preen oil composition is species-, sex-, and season-specific. Many birds use it for chemical signalling (semiochemicals) detected by mates via olfaction — challenging the historical view that birds are anosmic. The hoopoe (Upupa epops) adds bacteria to its oil that produce dimethyl sulfide and other volatiles to deter ectoparasites.

Module Summary

Beta-Keratin

Beta-pleated sheet; E ≈ 2.5 GPa; disulfide crosslinks from Cys residues; unique to sauropsids

Feather Hierarchy

6 levels: rachis → barb → barbule → hooklet → beta-keratin fibril → amino acid

Hooklet Mechanism

~10⁶ hooklet-groove contacts per feather; ~1 nN per contact; unzippable by preening

Melanin Biosynthesis

Tyrosine → DOPA → dopaquinone via TYR; branch to eumelanin (black) or pheomelanin (red)

Carotenoids

Dietary origin; CYP2J19 ketolase converts yellow → red; honest signal of individual quality

Thin-Film Interference

2nd·cosθ = mλ; d=100–150 nm → blue, d=200 nm → green; iridescence with viewing angle

Structural Blue

No blue pigment; coherent scattering from quasi-ordered melanin/air nanostructure; non-iridescent

Preen Oil

Uropygial gland; wax esters for waterproofing; 7-DHC → vitamin D₃ precursor; chemical signalling

References

Prum, R. O. (2006). Anatomy, physics, and evolution of structural colors. In Bird Coloration, Vol. 1: Mechanisms and Measurements (ed. G. E. Hill & K. J. McGraw), pp. 295–353. Harvard University Press.
Prum, R. O., Torres, R. H., Williamson, S. & Dyck, J. (1998). Coherent light scattering by nanostructured collagen arrays in the caruncles of the Malagasy asity. Journal of Experimental Biology, 201, 763–773.
Prum, R. O. & Torres, R. H. (2003). Structural colouration of avian skin: convergent evolution of coherently scattering dermal collagen arrays. Journal of Experimental Biology, 206, 2409–2429.
Zi, J. et al. (2003). Coloration strategies in peacock feathers. Proceedings of the National Academy of Sciences, 100, 12576–12578.
Stavenga, D. G., Leertouwer, H. L., Marshall, N. J. & Osorio, D. (2011). Dramatic colour changes in a bird of paradise caused by uniquely structured breast feather barbules. Proceedings of the Royal Society B, 278, 2098–2104.
Eliason, C. M. & Shawkey, M. D. (2012). A photonic heterostructure produces diverse iridescent colours in duck wing patches. Journal of the Royal Society Interface, 9, 2279–2289.
D'Alba, L. & Shawkey, M. D. (2019). Melanosomes: Biogenesis, Properties, and Evolution of an Ancient Organelle. Physiological Reviews, 99, 1–19.
McGraw, K. J. (2006). Mechanics of carotenoid-based coloration. In Bird Coloration, Vol. 1 (ed. G. E. Hill & K. J. McGraw), pp. 177–242. Harvard University Press.
Lopes, R. J. et al. (2016). Genetic basis for red coloration in birds. Current Biology, 26, 1427–1434. (CYP2J19 ketolase discovery)
Dove, C. J. & Agreda, A. (2007). Differences in plumulaceous feather characters of dabbling and diving ducks. The Condor, 109, 192–199.
Galván, I. & Solano, F. (2016). Bird Integumentary Melanins: Biosynthesis, Forms, Function and Evolution. International Journal of Molecular Sciences, 17, 520.
Jacob, J. & Ziswiler, V. (1982). The uropygial gland. In Avian Biology, Vol. 6 (ed. D. S. Farner, J. R. King & K. C. Parkes), pp. 199–324. Academic Press.
Gill, F. B. (2007). Ornithology, 3rd ed. W. H. Freeman & Co.
Proctor, N. S. & Lynch, P. J. (1993). Manual of Ornithology: Avian Structure and Function. Yale University Press.

← Flight Aerodynamics Next: Module 3 — Avian Respiration →