Module 8

Climate Change & Biochemical Adaptation

Shifting metabolic zones, phenological mismatch, ocean deoxygenation, permafrost thaw, and evolutionary rescue

Climate change is fundamentally a biochemical perturbation. Rising temperature alters enzyme kinetics, shifts the balance between C3 and C4 photosynthesis, reduces ocean oxygen solubility, and accelerates decomposition of permafrost carbon. Organisms must adapt their biochemistry at rates unprecedented in evolutionary history. This module examines how climate change cascades from atmospheric chemistry through metabolic physiology to ecosystem reorganization, and asks whether evolutionary rescue can keep pace.

8.1 Shifting Metabolic Zones

The distribution of photosynthetic pathways across Earth’s surface is determined by temperature and aridity. C4 photosynthesis (using PEP carboxylase to concentrate CO₂ around RuBisCO) is advantageous at high temperatures because it suppresses photorespiration, which increases exponentially with temperature in C3 plants. The C3/C4 crossover temperature — where C4 becomes more efficient than C3 — is approximately 22–30$^\circ$C, depending on CO₂ concentration.

C3 vs C4 Quantum Yield Temperature Dependence

The quantum yield of CO₂ fixation for C3 plants depends on the specificity factor of RuBisCO ($\tau$) and the ratio of O₂ to CO₂ at the active site:

C3 Quantum Yield

\[ \Phi_{\text{C3}} = \frac{1}{4.5 + 10.5 \cdot \Gamma^*/C_i} \]

where $\Gamma^*$ is the CO₂ compensation point due to photorespiration, which increases with temperature: $\Gamma^* \propto \exp(E_a^{\text{oxy}}/RT)$.

For C4 plants, the CO₂-concentrating mechanism eliminates photorespiration, giving a temperature-independent quantum yield:

\[ \Phi_{\text{C4}} \approx \frac{1}{6.0} \approx 0.167 \quad \text{(constant with temperature)} \]

The crossover temperature $T^*$ where $\Phi_{\text{C3}} = \Phi_{\text{C4}}$shifts with atmospheric CO₂. Higher CO₂ raises $C_i$, reduces the photorespiration penalty, and shifts $T^*$ upward — paradoxically favoring C3 plants under rising CO₂ despite warming temperatures.

Elevational and Latitudinal Shifts

As isotherms migrate poleward and upslope, species must track their thermal niches. The rate of elevational shift can be derived from the environmental lapse rate:

Elevational Shift Rate

\[ v_{\text{shift}} = \frac{dT/dt}{dT/dz} \]

where $dT/dt \approx 0.02\text{--}0.04$ $^\circ$C/yr (warming rate) and $dT/dz \approx -6.5$ $^\circ$C/km (adiabatic lapse rate).

This gives $v_{\text{shift}} \approx 3\text{--}6$ m/yr upslope, or approximately 11 m per decade. Observed treeline advances range from 5–20 m per decade, broadly consistent with this estimate. For latitudinal shifts:

\[ v_{\text{lat}} = \frac{dT/dt}{dT/d\phi} \approx \frac{0.03}{0.005} \approx 6 \text{ km/yr poleward} \]

where $dT/d\phi \approx 0.5\text{--}0.7$ $^\circ$C per degree latitude ($\approx 0.005$ $^\circ$C/km). Many organisms cannot disperse at 6 km/yr, creating an adaptation deficit.

Thermal Niche Width

Each species occupies a thermal niche defined by its biochemical performance curve:

\[ P(T) = P_{\max} \cdot \exp\!\left(-\frac{(T - T_{\text{opt}})^2}{2\sigma_T^2}\right) \cdot \left(1 - \frac{1}{1 + e^{-k(T - T_{\text{crit}})}}\right) \]

The first term is a Gaussian centered on optimal temperature $T_{\text{opt}}$ with thermal breadth $\sigma_T$. The second term adds a sharp decline above the critical thermal maximum $T_{\text{crit}}$ due to protein denaturation. Tropical ectotherms have narrow $\sigma_T$ (~2–5$^\circ$C) and live near their $T_{\text{crit}}$, making them especially vulnerable.

8.2 Phenological Mismatch

Climate change is desynchronizing ecological interactions that evolved under stable seasonality. Spring phenological events — leaf-out, flowering, insect emergence — are advancing by 2–5 days per decade in response to warming. However, migratory birds use photoperiod (day length) as their primary cue for spring migration, which does not change with climate.

The Mismatch Problem

Consider a bird that must arrive at its breeding grounds when caterpillar biomass peaks to feed its chicks. The caterpillar peak depends on temperature (spring degree-days accumulation) while the bird’s arrival depends on photoperiod. As springs warm:

Caterpillar peak: Advances by $\Delta t_{\text{resource}} = \beta_T \cdot \Delta T$ days (typically $\beta_T \approx 3\text{--}5$ days/$^\circ$C)
Bird arrival: Advances by $\Delta t_{\text{bird}} \approx 0\text{--}1$ day/$^\circ$C (photoperiod-cued, weak temperature response)
Mismatch: $\Delta m = \Delta t_{\text{resource}} - \Delta t_{\text{bird}} = (\beta_T - \beta_P) \cdot \Delta T$

Fitness Cost of Mismatch

The fitness (reproductive success) of a consumer depends on the temporal match between its peak demand and the peak resource availability. Modeling both as Gaussian pulses:

Mismatch Fitness Model

\[ W = W_{\max} \cdot \exp\!\left(-\frac{(t_{\text{peak}} - t_{\text{resource}})^2}{2\sigma^2}\right) \]

where $t_{\text{peak}}$ is the consumer’s peak demand timing,$t_{\text{resource}}$ is the resource peak timing,$\sigma$ is the temporal width of the resource pulse, and$W_{\max}$ is maximum fitness at perfect synchrony.

The fitness cost of a $\Delta m$-day mismatch is:

\[ \frac{W}{W_{\max}} = \exp\!\left(-\frac{(\Delta m)^2}{2\sigma^2}\right) \]

For a caterpillar peak with $\sigma = 10$ days and a 15-day mismatch:$W/W_{\max} = \exp(-225/200) = 0.32$ — a 68% fitness reduction.

Empirical evidence: in the Netherlands, pied flycatchers (Ficedula hypoleuca) experienced population declines of up to 90% in forests with the strongest phenological mismatch (Sanz et al. 2003; Both et al. 2006). Great tits (Parus major), being resident, can partially track the advancing caterpillar peak and have fared better.

Biochemical Basis of Phenological Cues

Plants use a combination of temperature accumulation (chilling hours + growing degree-days) and photoperiod (phytochrome signaling via PIF4/CO/FT pathway) to trigger spring development. The biochemical cascade: phytochrome B converts between Pr (red-absorbing, inactive) and Pfr (far-red-absorbing, active) forms. Pfr accumulation under long days activates CONSTANS (CO) transcription factor, which induces FLOWERING LOCUS T (FT) — the mobile “florigen” transported from leaves to the shoot apical meristem.

8.3 Ocean Deoxygenation

Ocean oxygen content has declined by approximately 2% globally since 1960 and is projected to decrease by 3–4% by 2100 under high-emission scenarios. Two synergistic mechanisms drive this: reduced solubility (warmer water holds less gas) and increased stratification (warm surface layer inhibits deep mixing that replenishes oxygen at depth).

Oxygen Solubility: Henry’s Law with Temperature

The saturation concentration of dissolved oxygen follows:

Oxygen Solubility vs. Temperature

\[ C_{\text{sat}} = \exp\!\left(A_1 + \frac{A_2}{T} + A_3 \ln T + A_4 T + A_5 S_{\text{sal}} + A_6 S_{\text{sal}}^2\right) \]

The Benson-Krause equation (UNESCO 1986), where $T$ is temperature in Kelvin and $S_{\text{sal}}$ is salinity. The coefficients $A_1$ through $A_6$are empirically determined.

A simpler approximation for the temperature dependence at standard salinity:

\[ C_{\text{sat}}(T) \approx 14.6 - 0.39T + 0.0077T^2 - 0.0000646T^3 \quad \text{(mg/L, } T \text{ in } ^\circ\text{C)} \]

At 5$^\circ$C: $C_{\text{sat}} \approx 12.8$ mg/L; at 25$^\circ$C: $\approx 8.3$ mg/L; at 35$^\circ$C: $\approx 7.0$ mg/L. Warming from 5 to 25$^\circ$C reduces oxygen capacity by 35%.

Oxygen Minimum Zone (OMZ) Expansion

Oxygen minimum zones occur at intermediate depths (200–1000 m) where microbial respiration of sinking organic matter consumes oxygen faster than ventilation replenishes it. The oxygen balance at depth $z$:

\[ \frac{\partial [\text{O}_2]}{\partial t} = K_z \frac{\partial^2 [\text{O}_2]}{\partial z^2} - R(z) \]

where $K_z$ is the vertical diffusion coefficient (reduced by stronger stratification) and $R(z)$ is the respiration rate (increased by warming via the Arrhenius factor).

Both changes drive $[\text{O}_2]$ lower: reduced $K_z$ means less oxygen supply from the surface, and increased $R$ means faster oxygen consumption. OMZs are expanding both vertically and horizontally, compressing the habitable volume for oxygen-dependent fish and invertebrates. Species that require $[\text{O}_2] > 3.5$ mg/L for aerobic metabolism are being squeezed into an ever-thinner surface layer — a process termed habitat compression.

8.4 Permafrost Thaw & Methane Feedback

Northern permafrost soils contain approximately 1,500 Gt C — roughly twice the atmospheric carbon pool. This organic carbon accumulated over millennia because freezing temperatures inhibited microbial decomposition. As the Arctic warms at 2–3$\times$ the global average rate, permafrost thaw exposes this carbon to microbial metabolism.

Arrhenius-Scaled Decomposition

The rate of microbial decomposition of soil organic carbon follows temperature-dependent Arrhenius kinetics:

Temperature-Dependent Decomposition

\[ \frac{d[C]}{dt} = -k_0 \cdot \exp\!\left(-\frac{E_a}{RT}\right) \cdot [C] \]

where $k_0$ is the pre-exponential factor, $E_a \approx 50\text{--}70$ kJ/mol for soil organic matter decomposition, $R = 8.314$ J/(mol$\cdot$K), and $[C]$ is the soil carbon stock.

The temperature sensitivity is characterized by $Q_{10}$:

\[ Q_{10} = \exp\!\left(\frac{10 E_a}{RT^2}\right) \approx 2\text{--}3 \text{ for typical soil temperatures} \]

A $Q_{10}$ of 2.5 means that a 10$^\circ$C warming increases decomposition rate by 2.5$\times$. At Arctic temperatures near 0$^\circ$C, $Q_{10}$ is higher (~3–5) because $Q_{10}$ itself increases at lower temperatures.

Aerobic vs. Anaerobic: CO₂ vs. CH₄

The product of decomposition depends critically on oxygen availability:

Aerobic (drained soils): C₆H₁₂O₆ + 6O₂ $\to$ 6CO₂ + 6H₂O. GWP of CO₂ = 1
Anaerobic (waterlogged, thermokarst): C₆H₁₂O₆ $\to$ 3CO₂ + 3CH₄. GWP of CH₄ = 28 (100-yr)

The warming potential of anaerobic decomposition is ~4$\times$ greater per unit carbon released because methane has 28$\times$ the global warming potential of CO₂. Thermokarst lake formation (ground collapse from ice lens melting) creates ideal anaerobic conditions for methanogenesis.

Positive Feedback Quantification

The permafrost carbon feedback creates a self-amplifying loop:

\[ \Delta T_{\text{extra}} = \lambda \cdot f \cdot C_{\text{pf}} \cdot \text{GWP}_{\text{eff}} \]

where $\lambda$ is climate sensitivity (~0.5 $^\circ$C per 100 Gt C),$f$ is the fraction of permafrost carbon released by 2100 (~5–15%),$C_{\text{pf}} = 1500$ Gt C, and GWP$_{\text{eff}}$ accounts for the CH₄/CO₂ mix. Estimates range from 0.1–0.5$^\circ$C additional warming by 2100, on top of anthropogenic forcing.

8.5 Evolutionary Rescue

Evolutionary rescue occurs when adaptation is rapid enough to prevent extinction in a deteriorating environment. The central question: can biochemical adaptation keep pace with climate change?

Critical Rate of Adaptation

Lynch & Lande (1993) derived the condition for population persistence when the environment changes at rate $d\theta/dt$ (where $\theta$ is the optimal phenotype, e.g., optimal heat tolerance):

Evolutionary Rescue Condition

\[ \frac{d\bar{z}}{dt} \geq \frac{d\theta}{dt} \]

The rate of change in mean character value ($\bar{z}$) must equal or exceed the rate of change in the optimum ($\theta$). If $d\bar{z}/dt < d\theta/dt$, the population accumulates a maladaptation load that reduces fitness below replacement.

From the breeder’s equation, the maximal sustainable rate of evolution is:

\[ \frac{d\bar{z}}{dt} = \frac{h^2 \sigma_P^2}{\sigma_s^2} \cdot \frac{1}{\tau} \]

where $h^2$ is heritability, $\sigma_P^2$ is phenotypic variance,$\sigma_s^2$ is the width of the fitness function, and $\tau$ is generation time.

The critical rate of environmental change for persistence:

\[ \left(\frac{d\theta}{dt}\right)_{\text{crit}} = \frac{h^2 V_P}{\sigma_s^2 \tau} \]

For typical values ($h^2 = 0.3$, $V_P = 4$ $^\circ\text{C}^2$,$\sigma_s = 5$ $^\circ$C, $\tau = 5$ yr):$(d\theta/dt)_{\text{crit}} = 0.3 \times 4 / (25 \times 5) = 0.0096$ $^\circ$C/yr. Current warming of 0.02–0.04$^\circ$C/yr exceeds this for many species.

Molecular Targets for Rapid Adaptation

Heat shock proteins (HSPs): Promoter variation in Hsp70 and Hsp90 genes controls the temperature at which the heat shock response activates. Standing genetic variation in HSP promoter sequences provides raw material for rapid thermal adaptation.
CYP450 detoxification enzymes: Cytochrome P450 gene family expansion enables rapid evolution of detoxification capacity. Insects exposed to novel pesticides can evolve resistance within 10–50 generations through gene duplication and regulatory changes.
Membrane lipid desaturases: Enzymes that adjust membrane fluidity by inserting double bonds into fatty acid chains. Rapid changes in desaturase expression allow ectotherms to maintain membrane function across temperature changes (homeoviscous adaptation).

Epigenetic Plasticity as a Buffer

Before genetic adaptation can occur, epigenetic plasticity provides a rapid-response buffer. DNA methylation changes, histone modifications, and small RNA-mediated gene silencing can alter gene expression within a single generation. In coral bleaching, individuals that survived previous heat stress show altered DNA methylation patterns at heat shock genes and are more thermotolerant — a form of “epigenetic memory” that can be transmitted to offspring for 1–3 generations, buying time for genetic adaptation to catch up.

8.6 Climate Change Biochemical Cascade

The diagram below traces the biochemical cascade from rising atmospheric CO₂ through warming, ocean acidification, deoxygenation, and ultimately biodiversity loss, highlighting the positive feedback loops that amplify the initial perturbation.

8.7 Computational Simulations

Four simulations exploring climate change biochemistry: (1) shifting C3/C4 boundary under RCP scenarios, (2) phenological mismatch model, (3) ocean O₂ solubility projections, and (4) permafrost carbon release feedback.

Climate Change Biochemical Impacts: C3/C4 Shifts, Mismatch, O2, Permafrost

Python

script.py190 lines

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

np.random.seed(42)
fig, axes = plt.subplots(2, 2, figsize=(14, 11))

# ====== Panel 1: C3/C4 Crossover Temperature vs CO2 ======
ax = axes[0, 0]

# Simplified model: C3 quantum yield depends on Gamma*/Ci
# Gamma* increases with temperature (photorespiration)
# Ci ~ 0.7 * Ca for C3 plants
T_range = np.linspace(10, 45, 200)  # Celsius

CO2_levels = {
    'Pre-industrial (280 ppm)': {'co2': 280, 'color': '#3b82f6'},
    'Current (420 ppm)': {'co2': 420, 'color': '#2dd4bf'},
    'RCP 4.5 (550 ppm)': {'co2': 550, 'color': '#f59e0b'},
    'RCP 8.5 (900 ppm)': {'co2': 900, 'color': '#ef4444'},
}

# C4 quantum yield is approximately constant
phi_c4 = 1.0 / 6.0  # ~0.167

for label, params in CO2_levels.items():
    Ci = 0.7 * params['co2']  # intercellular CO2
    # Gamma* (CO2 compensation point) increases with T
    # Gamma* ~ 42.75 * exp(0.0635*(T-25)) at 21% O2
    gamma_star = 42.75 * np.exp(0.0635 * (T_range - 25))
    phi_c3 = 1.0 / (4.5 + 10.5 * gamma_star / Ci)
    ax.plot(T_range, phi_c3, color=params['color'], linewidth=2.5,
            label=label)

ax.axhline(y=phi_c4, color='#a78bfa', linewidth=2, linestyle='--',
           label='C4 quantum yield')
ax.set_xlabel('Temperature (\u00b0C)', fontsize=12, color='#99f6e4')
ax.set_ylabel('Quantum Yield (mol CO\u2082 / mol photons)', fontsize=12, color='#99f6e4')
ax.set_title('C3/C4 Crossover Temperature', fontsize=14, fontweight='bold', color='#5eead4')
ax.legend(fontsize=8.5, facecolor='#0a0a0a', edgecolor='#5eead4',
          labelcolor='#d1d5db', loc='lower left')
ax.set_facecolor('#0a0a0a')
ax.tick_params(colors='#94a3b8')
ax.spines['bottom'].set_color('#5eead4')
ax.spines['left'].set_color('#5eead4')
ax.spines['top'].set_visible(False)
ax.spines['right'].set_visible(False)
ax.grid(True, alpha=0.15, color='#5eead4')
ax.set_ylim(0, 0.12)

# ====== Panel 2: Phenological Mismatch Model ======
ax = axes[0, 1]

years = np.arange(1980, 2101)
# Resource peak advancing at 3 days/decade
t_resource_base = 120  # day of year (May 1)
resource_advance = 3.0 / 10  # days per year
t_resource = t_resource_base - resource_advance * (years - 1980)

# Bird arrival: slower advance (1 day/decade, photoperiod-limited)
bird_advance = 1.0 / 10
t_bird = t_resource_base - bird_advance * (years - 1980)

# Mismatch
mismatch = t_bird - t_resource

# Fitness
sigma = 10  # days, resource pulse width
fitness = np.exp(-mismatch**2 / (2 * sigma**2))

ax2 = ax.twinx()
ax.plot(years, mismatch, color='#ef4444', linewidth=2.5, label='Mismatch (days)')
ax2.plot(years, fitness, color='#2dd4bf', linewidth=2.5, label='Relative fitness')

ax.set_xlabel('Year', fontsize=12, color='#99f6e4')
ax.set_ylabel('Mismatch (days late)', fontsize=12, color='#ef4444')
ax2.set_ylabel('Relative Fitness (W/W_max)', fontsize=12, color='#2dd4bf')
ax.set_title('Phenological Mismatch', fontsize=14, fontweight='bold', color='#5eead4')

# Manual legend
from matplotlib.lines import Line2D
lines = [Line2D([0], [0], color='#ef4444', linewidth=2.5),
         Line2D([0], [0], color='#2dd4bf', linewidth=2.5)]
ax.legend(lines, ['Mismatch (days)', 'Relative fitness'],
          fontsize=10, facecolor='#0a0a0a', edgecolor='#5eead4',
          labelcolor='#d1d5db', loc='center left')

ax.set_facecolor('#0a0a0a')
ax.tick_params(axis='y', colors='#ef4444')
ax2.tick_params(axis='y', colors='#2dd4bf')
ax.tick_params(axis='x', colors='#94a3b8')
ax.spines['bottom'].set_color('#5eead4')
ax.spines['left'].set_color('#ef4444')
ax.spines['right'].set_color('#2dd4bf')
ax.spines['top'].set_visible(False)
ax2.spines['top'].set_visible(False)
ax.grid(True, alpha=0.15, color='#5eead4')
ax2.set_ylim(0, 1.1)

# ====== Panel 3: Ocean O2 Solubility Projections ======
ax = axes[1, 0]

T_ocean = np.linspace(0, 35, 200)
# Simplified O2 solubility: C_sat ~ 14.6 - 0.39T + 0.0077T^2 - 0.0000646T^3
C_sat = 14.6 - 0.39 * T_ocean + 0.0077 * T_ocean**2 - 0.0000646 * T_ocean**3

ax.fill_between(T_ocean, C_sat, alpha=0.3, color='#2dd4bf')
ax.plot(T_ocean, C_sat, color='#2dd4bf', linewidth=2.5, label='O\u2082 saturation')

# Hypoxia thresholds
ax.axhline(y=6.0, color='#f59e0b', linewidth=1.5, linestyle='--', alpha=0.7)
ax.text(30, 6.2, 'Mild hypoxia', color='#f59e0b', fontsize=10)
ax.axhline(y=3.5, color='#ef4444', linewidth=1.5, linestyle='--', alpha=0.7)
ax.text(30, 3.7, 'Severe hypoxia', color='#ef4444', fontsize=10)
ax.axhline(y=2.0, color='#dc2626', linewidth=2, linestyle='--', alpha=0.9)
ax.text(30, 2.2, 'Dead zone', color='#dc2626', fontsize=10, fontweight='bold')

# Mark current vs future temps
ax.axvline(x=15, color='#94a3b8', linewidth=1, linestyle=':', alpha=0.5)
ax.text(15.5, 13.5, 'Current\naverage', color='#94a3b8', fontsize=9)
ax.axvline(x=19, color='#f97316', linewidth=1, linestyle=':', alpha=0.5)
ax.text(19.5, 13.5, '+4\u00b0C\n(2100)', color='#f97316', fontsize=9)

ax.set_xlabel('Sea Surface Temperature (\u00b0C)', fontsize=12, color='#99f6e4')
ax.set_ylabel('O\u2082 Saturation (mg/L)', fontsize=12, color='#99f6e4')
ax.set_title('Ocean O\u2082 Solubility', fontsize=14, fontweight='bold', color='#5eead4')
ax.set_facecolor('#0a0a0a')
ax.tick_params(colors='#94a3b8')
ax.spines['bottom'].set_color('#5eead4')
ax.spines['left'].set_color('#5eead4')
ax.spines['top'].set_visible(False)
ax.spines['right'].set_visible(False)
ax.grid(True, alpha=0.15, color='#5eead4')

# ====== Panel 4: Permafrost Carbon Release ======
ax = axes[1, 1]

years_pf = np.arange(2000, 2301)
C_pf_total = 1500  # Gt C

# Temperature scenarios
scenarios_pf = {
    'RCP 2.6 (+1.5\u00b0C)': {'dT_per_yr': 0.005, 'T_max': 1.5, 'color': '#3b82f6'},
    'RCP 4.5 (+2.5\u00b0C)': {'dT_per_yr': 0.015, 'T_max': 2.5, 'color': '#f59e0b'},
    'RCP 8.5 (+4.5\u00b0C)': {'dT_per_yr': 0.035, 'T_max': 4.5, 'color': '#ef4444'},
}

for label, sc in scenarios_pf.items():
    # Arctic warming is 2x global
    T_arctic = np.minimum(sc['dT_per_yr'] * 2 * (years_pf - 2000), sc['T_max'] * 2)

# Arrhenius decomposition rate
    Ea = 55000  # J/mol
    R_gas = 8.314
    T_base = 273.15 - 5  # base permafrost temp
    T_actual = T_base + T_arctic

k = 0.001 * np.exp(-Ea / R_gas * (1/T_actual - 1/T_base))

# Integrate carbon loss
    C_remaining = np.zeros(len(years_pf))
    C_remaining[0] = C_pf_total
    for i in range(1, len(years_pf)):
        C_remaining[i] = C_remaining[i-1] * (1 - k[i])
        C_remaining[i] = max(C_remaining[i], 0)

C_released = C_pf_total - C_remaining
    ax.plot(years_pf, C_released, color=sc['color'], linewidth=2.5, label=label)

ax.set_xlabel('Year', fontsize=12, color='#99f6e4')
ax.set_ylabel('Cumulative C Released (Gt C)', fontsize=12, color='#99f6e4')
ax.set_title('Permafrost Carbon Feedback', fontsize=14, fontweight='bold', color='#5eead4')
ax.legend(fontsize=10, facecolor='#0a0a0a', edgecolor='#5eead4',
          labelcolor='#d1d5db', loc='upper left')
ax.set_facecolor('#0a0a0a')
ax.tick_params(colors='#94a3b8')
ax.spines['bottom'].set_color('#5eead4')
ax.spines['left'].set_color('#5eead4')
ax.spines['top'].set_visible(False)
ax.spines['right'].set_visible(False)
ax.grid(True, alpha=0.15, color='#5eead4')

fig.patch.set_facecolor('#0a0a0a')
plt.tight_layout()
plt.savefig('output.png', dpi=150, bbox_inches='tight',
            facecolor='#0a0a0a', edgecolor='none')
plt.close()
print("Climate change simulations complete: C3/C4 shifts, mismatch, O2 solubility, permafrost feedback.")

Click Run to execute the Python code

Code will be executed with Python 3 on the server

8.8 Carbon Dioxide Fertilization & Its Limits

Elevated atmospheric CO₂ directly stimulates C3 photosynthesis because RuBisCO is not CO₂-saturated at current ambient levels. The Farquhar-von Caemmerer-Berry (FvCB) model (1980) provides the mechanistic framework for understanding this response and, crucially, its limitations.

The Farquhar Model of C3 Photosynthesis

Net photosynthesis is the minimum of two rates minus dark respiration:

Farquhar Model

\[ A = \min(A_c, A_j) - R_d \]

The RuBisCO-limited rate ($A_c$) depends on the maximum carboxylation velocity and the competition between CO₂ and O₂ at the RuBisCO active site:

RuBisCO-Limited Assimilation

\[ A_c = V_{c\max} \cdot \frac{C_i - \Gamma^*}{C_i + K_c\left(1 + \frac{O}{K_o}\right)} \]

where $V_{c\max}$ is the maximum carboxylation rate (~50–120 $\mu$mol m$^{-2}$ s$^{-1}$), $C_i$ is the intercellular CO₂concentration, $\Gamma^*$ is the CO₂ compensation point in the absence of dark respiration, $K_c$ is the Michaelis constant for CO₂ (~270 $\mu$mol/mol),$O$ is the O₂ concentration, and $K_o$ is the Michaelis constant for O₂ (~165 mmol/mol).

The RuBP-regeneration-limited rate ($A_j$) depends on electron transport:

\[ A_j = J \cdot \frac{C_i - \Gamma^*}{4C_i + 8\Gamma^*} \]

where $J$ is the electron transport rate, limited by light availability and the maximum electron transport rate $J_{\max}$.

At current CO₂ (~420 ppm), most C3 plants are RuBisCO-limited ($A_c < A_j$). Raising CO₂ to 550 ppm increases $C_i$, directly increasing $A_c$. However, above ~600–800 ppm, photosynthesis becomes RuBP-regeneration-limited ($A_j < A_c$), and further CO₂ increases yield diminishing returns.

FACE Experiments: The Reality Check

Free-Air CO₂ Enrichment (FACE) experiments elevate CO₂ over intact ecosystems. Results from over 15 FACE sites worldwide reveal:

Initial boost: +15–25% net primary productivity (NPP) at 550 ppm CO₂
Photosynthetic acclimation: After 3–5 years, $V_{c\max}$ downregulates by 10–20% (less RuBisCO protein produced as N is reallocated)
Diminishing returns: NPP enhancement declines to +5–10% after a decade
Nutrient limitation: The response is strongly constrained by N and P availability

Progressive Nitrogen Limitation (PNL)

The progressive nitrogen limitation hypothesis (Luo et al. 2004) explains why the CO₂ fertilization effect weakens over time. The mechanism is a biogeochemical feedback loop:

Step 1: Elevated CO₂ increases photosynthesis and biomass production
Step 2: More biomass requires more N, drawn from the available soil N pool
Step 3: More litter production with higher C:N ratio (N-diluted tissue)
Step 4: High-C:N litter decomposes slowly (microbial N immobilization), sequestering N in soil organic matter
Step 5: Soil available N declines progressively, limiting further growth response to CO₂

The net result: the terrestrial carbon sink provided by CO₂ fertilization is self-limiting. Models that omit PNL overestimate future land carbon uptake by 50–100%, with profound implications for climate projections.

8.9 Tipping Points & Biogeochemical Feedbacks

Earth system tipping points are thresholds beyond which positive feedbacks drive the system irreversibly to a new state. Unlike gradual degradation, tipping points involve nonlinear, self-amplifying dynamics where small additional forcing can trigger catastrophic change.

Amazon Dieback: Rainfall-Forest Feedback

The Amazon rainforest generates ~30–50% of its own rainfall through evapotranspiration. This creates a positive feedback loop that can operate in reverse:

Reduced rainfall (from climate change + deforestation) $\to$
Forest drought stress $\to$ tree mortality, increased fire susceptibility
Forest fires $\to$ massive CO₂ release + canopy loss
Reduced evapotranspiration $\to$ even less rainfall recycling
Savannification: forest converts irreversibly to grassland/savanna

The critical precipitation threshold can be derived from the ecosystem water balance:

Critical Precipitation Threshold

\[ P_{\text{crit}} = \text{ET}(\text{LAI}) + R \]

where $P$ is precipitation, $\text{ET}$ is evapotranspiration (which depends on leaf area index LAI), and $R$ is runoff. The feedback enters because LAI depends on $P$: $\text{LAI} = f(P)$, creating a self-referential equation. Below $P_{\text{crit}} \approx 1,500$ mm/yr, the forest cannot sustain itself and transitions to savanna.

Stommel Two-Box Model: Atlantic Thermohaline Circulation

The Atlantic Meridional Overturning Circulation (AMOC) is driven by density differences between cold, salty North Atlantic water (dense, sinks) and warm, fresh tropical water. Stommel (1961) showed this system has two stable states using a two-box model:

Stommel Two-Box Model

\[ \frac{d(\Delta T)}{dt} = \eta_1 - |\Psi| \cdot \Delta T \]

\[ \frac{d(\Delta S)}{dt} = \eta_2 - |\Psi| \cdot \Delta S \]

\[ \Psi = k(\alpha \cdot \Delta T - \beta \cdot \Delta S) \]

where $\Delta T$ and $\Delta S$ are the temperature and salinity differences between the boxes, $\Psi$ is the overturning flow rate, $\alpha$ and$\beta$ are the thermal and haline expansion coefficients, and $\eta_1, \eta_2$represent atmospheric forcing. The competition between temperature (driving circulation) and salinity (opposing it) creates bistability.

Freshwater input from Greenland ice sheet melting increases $\Delta S$, weakening$\Psi$. If freshwater flux exceeds a critical threshold (~0.1–0.5 Sv), the AMOC collapses to a “shutdown” state with dramatic consequences: cooling of Northern Europe by 5–10$^\circ$C, southward shift of the Intertropical Convergence Zone, and disruption of global precipitation patterns. AMOC is currently at its weakest in ~1,000 years.

8.10 Nature-Based Solutions

Nature-based solutions (NbS) harness natural biochemical processes for climate change mitigation. Three approaches with the highest potential are blue carbon ecosystems, biochar, and enhanced rock weathering.

Blue Carbon: Coastal Ecosystem Sequestration

Blue carbon ecosystems — mangroves, seagrass meadows, and salt marshes — sequester carbon at rates 2–10$\times$ faster than terrestrial forests on a per-area basis:

Mangroves: ~6–8 t CO₂ ha$^{-1}$ yr$^{-1}$. Deep peat soils (up to 10 m) store ~1,000 t C/ha. Tannin-rich, anoxic sediments inhibit decomposition
Seagrass: ~3–5 t CO₂ ha$^{-1}$ yr$^{-1}$. Root mats trap sediments; lignocellulosic tissue resists decay. Posidonia oceanica meadows store C for millennia
Salt marshes: ~5–9 t CO₂ ha$^{-1}$ yr$^{-1}$. Sulfate-reducing conditions produce sulfide minerals that complex with organic C, further slowing decomposition

The net climate benefit must account for lifecycle emissions:

Net CO₂ Removal Accounting

\[ \text{Net removal} = C_{\text{sequestered}} - C_{\text{construction}} - C_{\text{opportunity}} \]

where $C_{\text{construction}}$ includes emissions from restoration activities and$C_{\text{opportunity}}$ accounts for alternative land use emissions. For mangrove restoration, the payback period is typically 3–5 years.

Biochar: Pyrolytic Carbon Stabilization

Biochar is produced by pyrolysis (thermal decomposition at 350–700$^\circ$C in the absence of oxygen) of biomass. The process converts labile organic carbon (cellulose, hemicellulose) into recalcitrant aromatic carbon with a mean residence time exceeding 1,000 years in soils.

The carbon efficiency of biochar depends on pyrolysis temperature: higher temperatures produce more stable char but with lower yield. The optimal balance (~500$^\circ$C) retains ~50% of biomass C as biochar with $>80$% aromatic carbon content. Co-benefits include improved soil water retention, cation exchange capacity, and reduced N₂O emissions from soil.

Enhanced Rock Weathering

Natural silicate weathering removes ~0.3 Gt CO₂/yr from the atmosphere over geological timescales. Enhanced weathering accelerates this by spreading crushed silicate rocks (e.g., basalt) on agricultural soils:

Silicate Weathering Reaction

\[ \text{CaSiO}_3 + 2\text{CO}_2 + 3\text{H}_2\text{O} \to \text{Ca}^{2+} + 2\text{HCO}_3^- + \text{H}_4\text{SiO}_4 \]

Each mole of wollastonite (CaSiO₃) permanently removes 2 moles of CO₂ by converting it to dissolved bicarbonate (HCO₃$^-$), which is transported to the ocean where it remains stable for $>10^5$ years. The reaction also raises soil pH and provides Ca$^{2+}$ and Si nutrients, benefiting crop growth.

Global potential: spreading basalt on ~50% of cropland could remove 0.5–4 Gt CO₂/yr, depending on particle size, climate, and soil conditions. The reaction rate increases with temperature, creating a natural negative feedback: as climate warms, weathering accelerates and removes more CO₂. Costs are estimated at $80–180/t CO₂, competitive with direct air capture.

8.11 Advanced Simulations: Fertilization, Tipping Points & Blue Carbon

Three simulations: (1) Farquhar model CO₂ response curves comparing current vs elevated CO₂ with the transition from RuBisCO-limited to RuBP-limited photosynthesis, (2) Amazon tipping point dynamics showing the rainfall-forest positive feedback, and (3) blue carbon sequestration comparison across coastal ecosystems.

CO2 Fertilization, Amazon Tipping Point & Blue Carbon Sequestration

Python

script.py228 lines

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

np.random.seed(42)
fig, axes = plt.subplots(1, 3, figsize=(18, 5.5))

# ====== Panel 1: Farquhar Model CO2 Response Curves ======
ax = axes[0]

Ci_range = np.linspace(50, 1200, 500)  # intercellular CO2 (umol/mol)

# Parameters (typical C3 plant at 25C)
Vcmax = 80    # umol m-2 s-1
Jmax = 160    # umol m-2 s-1
Gamma_star = 43  # umol/mol (CO2 compensation point)
Kc = 270      # umol/mol (Michaelis for CO2)
Ko = 165000   # umol/mol (Michaelis for O2)
O = 210000    # umol/mol (atmospheric O2)
Rd = 1.5      # umol m-2 s-1 (dark respiration)

# RuBisCO-limited rate
Ac = Vcmax * (Ci_range - Gamma_star) / (Ci_range + Kc * (1 + O / Ko))

# Light for J calculation
J = Jmax * 0.9  # assuming near-saturating light
Aj = J * (Ci_range - Gamma_star) / (4 * Ci_range + 8 * Gamma_star)

# Net assimilation
A = np.minimum(Ac, Aj) - Rd

ax.plot(Ci_range, Ac, color='#ef4444', linewidth=2, linestyle='--', label='Ac (RuBisCO-limited)')
ax.plot(Ci_range, Aj, color='#3b82f6', linewidth=2, linestyle='--', label='Aj (RuBP-limited)')
ax.plot(Ci_range, A, color='#2dd4bf', linewidth=3, label='A = min(Ac, Aj) - Rd')

# Mark current and elevated CO2
Ci_current = 0.7 * 420  # ~294
Ci_elevated = 0.7 * 550  # ~385
Ci_high = 0.7 * 800     # ~560

for ci, label_text, col in [(Ci_current, '420 ppm', '#f59e0b'),
                             (Ci_elevated, '550 ppm', '#a78bfa'),
                             (Ci_high, '800 ppm', '#ef4444')]:
    A_val = min(Vcmax * (ci - Gamma_star) / (ci + Kc * (1 + O / Ko)),
                J * (ci - Gamma_star) / (4 * ci + 8 * Gamma_star)) - Rd
    ax.plot(ci, A_val, 'o', color=col, markersize=10, zorder=10, markeredgecolor='white', markeredgewidth=1.5)
    ax.annotate(f'Ci at {label_text}\nA={A_val:.1f}', xy=(ci, A_val),
                xytext=(ci + 80, A_val - 5), fontsize=8, color=col,
                arrowprops=dict(arrowstyle='->', color=col, lw=1.5))

# Transition line
trans_idx = np.argmin(np.abs(Ac - Aj))
ax.axvline(x=Ci_range[trans_idx], color='#94a3b8', linewidth=1, linestyle=':', alpha=0.5)
ax.text(Ci_range[trans_idx] + 10, 5, 'Limitation\ntransition', color='#94a3b8', fontsize=8)

ax.set_xlabel('Intercellular CO\u2082, Ci (\u00b5mol/mol)', fontsize=11, color='#99f6e4')
ax.set_ylabel('Assimilation rate (\u00b5mol m\u207b\u00b2 s\u207b\u00b9)', fontsize=11, color='#99f6e4')
ax.set_title('Farquhar Model: CO\u2082 Response', fontsize=12, fontweight='bold', color='#5eead4')
ax.legend(fontsize=8, facecolor='#0a0a0a', edgecolor='#5eead4', labelcolor='#d1d5db',
          loc='lower right')
ax.set_facecolor('#0a0a0a')
ax.tick_params(colors='#94a3b8')
ax.spines['bottom'].set_color('#5eead4')
ax.spines['left'].set_color('#5eead4')
ax.spines['top'].set_visible(False)
ax.spines['right'].set_visible(False)
ax.grid(True, alpha=0.15, color='#5eead4')
ax.set_ylim(-5, 45)

# ====== Panel 2: Amazon Tipping Point ======
ax = axes[1]

# Rainfall-forest feedback model
# P = P_external + alpha * LAI (forest recycles rainfall)
# LAI = f(P) = LAI_max * P / (P + K_P)  (LAI depends on rainfall)
# Stable states where P_total = P_ext + alpha * LAI_max * P_total / (P_total + K_P)

P_ext_range = np.linspace(500, 2500, 200)  # external precipitation mm/yr
alpha = 0.4    # recycling coefficient
LAI_max = 8.0  # maximum LAI
K_P = 800      # half-saturation for LAI response
P_crit = 1500  # critical precipitation

# For each P_ext, find total precipitation (solving self-consistent equation)
# P_total = P_ext + alpha * LAI_max * P_total / (P_total + K_P)
# This is a quadratic: P_total^2 + (K_P - P_ext - alpha*LAI_max)*P_total - K_P*P_ext = 0

forest_cover = np.zeros(len(P_ext_range))
P_total_arr = np.zeros(len(P_ext_range))

for i, P_ext in enumerate(P_ext_range):
    # Quadratic coefficients: P^2 + b*P + c = 0
    b = K_P - P_ext - alpha * LAI_max * K_P
    c = -K_P * P_ext
    discriminant = b**2 - 4*c
    if discriminant >= 0:
        P_total = (-b + np.sqrt(discriminant)) / 2
    else:
        P_total = P_ext
    P_total_arr[i] = P_total
    LAI = LAI_max * P_total / (P_total + K_P)
    forest_cover[i] = LAI / LAI_max * 100

# Simulate time trajectory: declining P_ext
years = np.arange(2000, 2151)
P_ext_time = np.maximum(2000 - 8 * (years - 2000), 600)  # declining from 2000 to 600

LAI_time = np.zeros(len(years))
LAI_time[0] = 7.5  # start near maximum forest

for i in range(1, len(years)):
    P_ext_t = P_ext_time[i]
    # P_total with current LAI feedback
    P_total_t = P_ext_t + alpha * LAI_time[i-1] * K_P
    # LAI responds to total precipitation
    LAI_eq = LAI_max * P_total_t / (P_total_t + K_P)
    # LAI adjusts slowly (inertia)
    tau = 10  # years response time
    LAI_time[i] = LAI_time[i-1] + (LAI_eq - LAI_time[i-1]) / tau
    LAI_time[i] = max(0.5, LAI_time[i])

ax2 = ax.twinx()
ax.plot(years, LAI_time / LAI_max * 100, color='#2dd4bf', linewidth=3, label='Forest cover (%)')
ax2.plot(years, P_ext_time, color='#3b82f6', linewidth=2, linestyle='--', label='External rainfall')

# Mark tipping point
tip_idx = np.argmax(np.diff(LAI_time) < -0.15)
if tip_idx > 0:
    ax.axvline(x=years[tip_idx], color='#ef4444', linewidth=2, linestyle=':', alpha=0.7)
    ax.text(years[tip_idx] + 2, 85, 'Tipping\npoint', color='#ef4444', fontsize=10, fontweight='bold')

ax.axhline(y=30, color='#f59e0b', linewidth=1, linestyle='--', alpha=0.5)
ax.text(2005, 32, 'Savanna threshold', color='#f59e0b', fontsize=9)

ax.set_xlabel('Year', fontsize=11, color='#99f6e4')
ax.set_ylabel('Forest cover (%)', fontsize=11, color='#2dd4bf')
ax2.set_ylabel('Rainfall (mm/yr)', fontsize=11, color='#3b82f6')
ax.set_title('Amazon Dieback Tipping Point', fontsize=12, fontweight='bold', color='#5eead4')

from matplotlib.lines import Line2D
lines = [Line2D([0], [0], color='#2dd4bf', linewidth=3),
         Line2D([0], [0], color='#3b82f6', linewidth=2, linestyle='--')]
ax.legend(lines, ['Forest cover', 'External rainfall'],
          fontsize=9, facecolor='#0a0a0a', edgecolor='#5eead4', labelcolor='#d1d5db')

ax.set_facecolor('#0a0a0a')
ax.tick_params(axis='y', colors='#2dd4bf')
ax2.tick_params(axis='y', colors='#3b82f6')
ax.tick_params(axis='x', colors='#94a3b8')
ax.spines['bottom'].set_color('#5eead4')
ax.spines['left'].set_color('#2dd4bf')
ax.spines['right'].set_color('#3b82f6')
ax.spines['top'].set_visible(False)
ax2.spines['top'].set_visible(False)
ax.grid(True, alpha=0.15, color='#5eead4')
ax.set_ylim(0, 105)

# ====== Panel 3: Blue Carbon Sequestration ======
ax = axes[2]

ecosystems = ['Mangrove', 'Seagrass', 'Salt marsh', 'Tropical\nforest', 'Temperate\nforest', 'Boreal\nforest']

# Sequestration rates (t CO2/ha/yr)
seq_rates = [7.0, 4.0, 6.5, 3.5, 2.0, 1.0]

# Soil C stocks (t C/ha, top 1m)
soil_stocks = [950, 500, 650, 150, 120, 200]

colors_eco = ['#2dd4bf', '#3b82f6', '#a78bfa', '#f59e0b', '#f97316', '#94a3b8']

x = np.arange(len(ecosystems))
width = 0.35

# Left bars: sequestration rate
bars1 = ax.bar(x - width/2, seq_rates, width, color=colors_eco, edgecolor='#0a0a0a',
               linewidth=1.5, label='Sequestration rate')

# Right bars: soil C stock (scaled)
ax2 = ax.twinx()
bars2 = ax2.bar(x + width/2, soil_stocks, width, color=colors_eco, edgecolor='#0a0a0a',
                linewidth=1.5, alpha=0.4, label='Soil C stock')

# Add value labels
for bar, val in zip(bars1, seq_rates):
    ax.text(bar.get_x() + bar.get_width()/2, val + 0.15,
            f'{val}', ha='center', va='bottom', color='#99f6e4', fontsize=9, fontweight='bold')

for bar, val in zip(bars2, soil_stocks):
    ax2.text(bar.get_x() + bar.get_width()/2, val + 20,
             f'{val}', ha='center', va='bottom', color='#94a3b8', fontsize=8)

# Bracket for blue carbon
ax.annotate('', xy=(-0.5, 8.5), xytext=(2.5, 8.5),
            arrowprops=dict(arrowstyle='-', color='#2dd4bf', lw=2))
ax.text(1, 9.0, 'BLUE CARBON', color='#2dd4bf', fontsize=11, fontweight='bold', ha='center')
ax.text(1, 8.6, '(2-10\u00d7 faster per area)', color='#2dd4bf', fontsize=8, ha='center')

ax.set_xticks(x)
ax.set_xticklabels(ecosystems, fontsize=9)
ax.set_ylabel('Sequestration (t CO\u2082/ha/yr)', fontsize=11, color='#99f6e4')
ax2.set_ylabel('Soil C stock (t C/ha, top 1m)', fontsize=11, color='#94a3b8')
ax.set_title('Blue Carbon vs Terrestrial', fontsize=12, fontweight='bold', color='#5eead4')

lines_leg = [Line2D([0], [0], color='#2dd4bf', linewidth=8, alpha=1.0),
             Line2D([0], [0], color='#2dd4bf', linewidth=8, alpha=0.4)]
ax.legend(lines_leg, ['Seq. rate', 'Soil C stock'],
          fontsize=9, facecolor='#0a0a0a', edgecolor='#5eead4', labelcolor='#d1d5db',
          loc='upper right')

ax.set_facecolor('#0a0a0a')
ax.tick_params(axis='y', colors='#99f6e4')
ax2.tick_params(axis='y', colors='#94a3b8')
ax.tick_params(axis='x', colors='#94a3b8')
ax.spines['bottom'].set_color('#5eead4')
ax.spines['left'].set_color('#5eead4')
ax.spines['right'].set_color('#94a3b8')
ax.spines['top'].set_visible(False)
ax2.spines['top'].set_visible(False)
ax.grid(True, alpha=0.15, color='#5eead4', axis='y')

fig.patch.set_facecolor('#0a0a0a')
plt.tight_layout()
plt.savefig('output.png', dpi=150, bbox_inches='tight',
            facecolor='#0a0a0a', edgecolor='none')
plt.close()
print("CO2 fertilization, Amazon tipping point, and blue carbon simulations complete.")

Click Run to execute the Python code

Code will be executed with Python 3 on the server

References

Ehleringer, J.R., Cerling, T.E. & Helliker, B.R. (1997). C4 photosynthesis, atmospheric CO₂, and climate. Oecologia, 112(3), 285–299.
Collatz, G.J., Berry, J.A. & Clark, J.S. (1998). Effects of climate and atmospheric CO₂ partial pressure on the global distribution of C4 grasses. Oecologia, 114(4), 441–454.
Lenoir, J., Gégout, J.C., Marquet, P.A., de Ruffray, P. & Brisse, H. (2008). A significant upward shift in plant species optimum elevation during the 20th century. Science, 320(5884), 1768–1771.
Both, C., Bouwhuis, S., Lessells, C.M. & Visser, M.E. (2006). Climate change and population declines in a long-distance migratory bird. Nature, 441(7089), 81–83.
Visser, M.E. & Both, C. (2005). Shifts in phenology due to global climate change: the need for a yardstick. Proceedings of the Royal Society B, 272(1581), 2561–2569.
Keeling, R.F., Körtzinger, A. & Gruber, N. (2010). Ocean deoxygenation in a warming world. Annual Review of Marine Science, 2, 199–229.
Breitburg, D., Levin, L.A., Oschlies, A., Grégoire, M., Chavez, F.P., Conley, D.J., ... & Zhang, J. (2018). Declining oxygen in the global ocean and coastal waters. Science, 359(6371), eaam7240.
Schuur, E.A.G., McGuire, A.D., Schädel, C., Grosse, G., Harden, J.W., Hayes, D.J., ... & Vonk, J.E. (2015). Climate change and the permafrost carbon feedback. Nature, 520(7546), 171–179.
Turetsky, M.R., Abbott, B.W., Jones, M.C., Anthony, K.W., Olefeldt, D., Schuur, E.A.G., ... & McGuire, A.D. (2020). Carbon release through abrupt permafrost thaw. Nature Geoscience, 13(2), 138–143.
Bell, G. & Gonzalez, A. (2009). Evolutionary rescue can prevent extinction following environmental change. Ecology Letters, 12(9), 942–948.
Lynch, M. & Lande, R. (1993). Evolution and extinction in response to environmental change. In Biotic Interactions and Global Change (pp. 234–250). Sinauer Associates.
Hoffmann, A.A. & Sgrò, C.M. (2011). Climate change and evolutionary adaptation. Nature, 470(7335), 479–485.
Farquhar, G.D., von Caemmerer, S. & Berry, J.A. (1980). A biochemical model of photosynthetic CO₂ assimilation in leaves of C3 species. Planta, 149(1), 78–90.
Ainsworth, E.A. & Long, S.P. (2005). What have we learned from 15 years of free-air CO₂ enrichment (FACE)? A meta-analytic review. New Phytologist, 165(2), 351–372.
Luo, Y., Su, B., Currie, W.S., Dukes, J.S., Finzi, A., Hartwig, U., ... & Field, C.B. (2004). Progressive nitrogen limitation of ecosystem responses to rising atmospheric carbon dioxide. BioScience, 54(8), 731–739.
Lovejoy, T.E. & Nobre, C. (2018). Amazon tipping point. Science Advances, 4(2), eaat2340.
Stommel, H. (1961). Thermohaline convection with two stable regimes of flow. Tellus, 13(2), 224–230.
Mcleod, E., Chmura, G.L., Bouillon, S., Salm, R., Björk, M., Duarte, C.M., ... & Silliman, B.R. (2011). A blueprint for blue carbon. Frontiers in Ecology and the Environment, 9(10), 552–560.
Beerling, D.J., Kantzas, E.P., Lomas, M.R., Wade, P., Eufrasio, R.M., Renforth, P., ... & Banwart, S.A. (2020). Potential for large-scale CO₂ removal via enhanced rock weathering with croplands. Nature, 583(7815), 242–248.

Share:X Reddit LinkedIn

M7. Conservation Biochemistry Course Overview