Thèse Gestion des Pâturages Pastoraux Échanges de Gaz à Effet de Serre et Potentiel de Séquestration du Carbone dans les Écosystèmes Arides de Djibouti une Approche Intégrée pour des Transitions D H/F - Doctorat.Gouv.Fr

Établissement : Université de Toulouse École doctorale : SDU2E - Sciences de l'Univers, de l'Environnement et de l'Espace Laboratoire de recherche : LAERO - Laboratoire d'Aérologie Direction de la thèse : Dominique SERCA ORCID 0000000186881440 Début de la thèse : 2027-09-01 Date limite de candidature : 2026-11-23T23:59:59 This PhD project falls under BEST's scientific axis: 'Quantifying and Analyzing the Societal Impact of Global Change.' It targets the arid pastoral ecosystems of Djibouti, which face challenges similar to those in the drylands of East Africa (Ethiopia, Horn of Africa), including soil degradation, invasive species, water scarcity, pressure on woody vegetation (a source of energy), and climate variability.

Objectives:
-Measure and model greenhouse gas (GHG) fluxes (CO, CH, NO) at high spatial resolution to refine national inventories (transitioning from a Tier 1 method-default IPCC emission factors-to Tier 2/3 methods with local data).
-Assess the carbon sequestration potential in soils and vegetation by comparing shrub-dominated areas with open grasslands.
-Propose nature-based solutions to enhance the climate resilience of pastoral communities, aligning with SDGs 2 (Zero Hunger), 13 (Climate Action), and 15 (Life on Land).

The project follows an integrated approach (WEFE Nexus) and adopts an interdisciplinary perspective (climate, ecology, animal science, social science) and a participatory framework (involving pastoralists, policymakers, and consultants) to analyze interdependencies between:
-Water: Management of watering points and watershed dynamics.
-Energy: Reducing dependence on woody biomass (a major driver of deforestation) through sustainable alternatives (e.g., biogas).
-Food: Food security linked to rangeland and livestock productivity.
-Ecosystems: Preservation of biodiversity and ecosystem services (e.g., soil fertility, pollination).

Methodological Innovations:
-Field data: GHG flux measurements, pastoralist surveys, transects to assess vegetation cover (LAI, diversity), soil characteristics (pH, texture, organic matter), and herd structure.
-Modeling: Simulation of carbon/water/nutrient dynamics (dynamic vegetation models and AI tools) to predict the impact of pastoral practices.
-Regional upscaling: Collaboration with Haramaya University (Ethiopia) for comparative analysis in the Horn of Africa's drylands.
-Participatory science: Involvement of pastoralists in data collection and co-design of solutions.
-Policy analysis: Evaluation of incentives for carbon sequestration.

Societal and Policy Impacts:
-Capacity building: Training pastoral communities in sustainable management and carbon sequestration mechanisms.
-Support for Djiboutian institutions (e.g., Ministry of Agriculture) in developing climate policies.

The project will rely on both academic partnerships (ILRI, Haramaya University) and non-academic partnerships (pastoral communities, Djibouti's Ministry of Agriculture, PASTOC consultancy).

1. Arid Ecosystems, Pastoralism, and Climate Change
Arid and semi-arid zones cover over 40% of global land and play an ambivalent role in the carbon cycle:
- GHG sources: Degraded soils and intensive pastoral practices (overgrazing, trampling) release CO, CH (enteric methane from ruminants), and NO (linked to manure management).
- Potential carbon sinks: Pasture soils and woody vegetation (shrubs, trees) can store organic carbon if management practices are sustainable.
- Scientific gaps: National GHG inventories (e.g., Djibouti's) use Tier 1 methods (IPCC default emission factors), which are imprecise for arid ecosystems due to:
o A lack of local data on pastoral activities (e.g., herd size, mobility).
o The absence of emission factors specific to arid contexts (e.g., emissions linked to Prosopis juliflora).
Climate change impacts in arid zones:
- Amplified stress: More frequent droughts, reduced precipitation, and rising temperatures exacerbate soil degradation and biodiversity loss.
- Feedback loop: Pasture degradation reduces their carbon sequestration capacity, thereby worsening climate change.
Pastoralism is the dominant livelihood in the arid zones of Djibouti and the Horn of Africa, providing food security (meat, milk) for millions. It demonstrates resilience to climate shocks through seasonal herd mobility. However, several factors increase the system's vulnerability:
- Demographic pressure increases demand for animal products and fuelwood.
- Invasion by Prosopis juliflora reduces pasture availability for livestock and alters GHG fluxes (e.g., increased NO emissions in soils).
- Soil degradation due to erosion, loss of organic matter, and increased salinity (linked to poorly managed irrigation).
- Water scarcity due to the depletion of water points and use conflicts (agriculture vs. pastoralism).
- Marginalization of herders and a lack of policies to reconcile ecosystem conservation with pastoral development.

2. Regional Context: Djibouti and the Horn of Africa
Djibouti and neighboring regions in Ethiopia and Somalia (Horn of Africa) host semi-arid ecosystems facing environmental challenges such as:
- Biological invasions (e.g., Prosopis juliflora).
- Water management dependent on over-exploited aquifers.
- Conflicts between pastoralists, farmers, and urban areas.
- Pressure on woody resources (90% of energy needs), leading to accelerated deforestation.
The ecosystems of the Horn of Africa share similarities with other arid regions, such as:
- Pasture degradation due to overgrazing and drought.
- Dependence on pastoralism as the dominant economic activity.
- Transboundary dynamics (transhumance).
- Data gaps on GHG fluxes and carbon sequestration.
Many results could be transposed to other arid regions (Sahel, Mediterranean).

3. Scientific Context: Integrated Approaches
Several gaps have been identified in the Horn of Africa region, summarized in the table below. By combining field data, modeling, AI, and participatory science, the project aims to address these gaps.

This project is set against a global backdrop of climate crisis and arid ecosystem degradation, with specific challenges in the Horn of Africa (biological invasions, resource pressure, data scarcity). It proposes an innovative approach combining:
1. Precise measurements of GHG fluxes and carbon stocks.
2. Advanced modeling (Tier 2/3 + AI) to predict the impact of pastoral practices.
3. Integration of the WEFE Nexus for sustainable resource management.
4. Involvement of local stakeholders (herders, institutions) for actionable research.
Project originality:
- First study of this scale in the Horn of Africa.
- Systemic approach (ecology + climate + social + policy).
- Potential for transposition to other arid regions.
Expected impacts:
- Scientific: Advances in understanding carbon-GHG dynamics in arid environments.
- Societal: Concrete solutions for herders and decision-makers (e.g., carbon sequestration policies).
- Policy: Contribution to national and international climate goals (SDGs, Paris Agreement).


Objective 1: Quantify and Model Greenhouse Gas (GHG) Fluxes in Arid Pastoral Ecosystems of Djibouti
Context:
National GHG inventories in Djibouti currently rely on Tier 1 methods (IPCC default emission factors), which are poorly suited to local specificities due to a lack of activity data and country-specific emission factors. This project aims to improve estimation accuracy by transitioning to Tier 2 (national emission factors) or Tier 3 (local, high-resolution data) methods.
Sub-objectives:
- Measure GHG fluxes (CO, CH, NO) at high spatial and temporal resolution in pilot sites (grazing lands, livestock enclosures, water points, grazing gradients) using flux chambers along transects to assess spatial variability.
- Identify carbon sources and sinks: Compare emissions between areas dominated by invasive shrubs (Prosopis juliflora) and open grasslands, and analyze the impact of pastoral practices (seasonal mobility, enclosure management) on fluxes.
- Model carbon/water dynamics: Use dynamic vegetation models and AI tools (neural networks) to predict GHG flux evolution based on practices, and upscale results to the regional level (Horn of Africa) in collaboration with Haramaya University (Ethiopia).

Objective 2: Assess Carbon Sequestration Potential in Soils and Vegetation
Context:
Arid ecosystems in Djibouti lack baseline data on carbon stocks (soils and vegetation). Yet, pastures could play a key role in climate change mitigation through sequestration.
Sub-objectives:
- Characterize carbon stocks: Measure soil organic matter (SOM), soil organic carbon (SOC), and plant biomass across landscape types (shrub-dominated vs. grassland areas), and evaluate the impact of soil degradation (erosion, fertility loss) and shrub encroachment on sequestration. This work will be conducted in partnership with ILRI (International Livestock Research Institute) in Addis Ababa and Nairobi, specialists in such measurements in East Africa.
- Identify leverage points for enhanced sequestration: Study the role of sustainable management practices (e.g., rotational grazing, restoration of degraded areas) and analyze the influence of native vs. invasive plant species on carbon storage.
- Propose restoration strategies: Develop nature-based solutions (e.g., agro-pastoral forestry, pasture rehabilitation) to improve ecosystem resilience.

Objective 3: Adopt an Integrated (WEFE Nexus) Approach to Analyze Interdependencies
Context:
The project uses the Water-Energy-Food-Ecosystems (WEFE) Nexus framework to address challenges holistically, accounting for links between resources and ecosystem services.
Sub-objectives:
- Water: Optimize the management of water points and watersheds to reduce resource pressure, and study the impact of irrigation and urban water supply on pastoral ecosystems.
- Energy: Reduce dependence on fuelwood (a major driver of deforestation) through sustainable alternatives (e.g., biogas from animal waste).
- Food: Ensure food security for pastoral communities by improving pasture and livestock productivity.
- Ecosystems: Preserve biodiversity (native plant species, pollinators) and ecosystem services (soil fertility, nutrient cycling).

Objective 4: Propose Solutions for a Climate-Smart Pastoralism Transition
Context:
Pastoral communities in Djibouti are vulnerable to climate shocks (droughts, precipitation variability) and anthropogenic pressures (overgrazing, resource overexploitation). The project aims to co-develop adaptive solutions with local stakeholders.
Sub-objectives:
- Analyze current pastoral practices: Conduct surveys on seasonal mobility, herd management, and resource use (water, pastures, fuelwood); assess the impact of these practices on ecosystem health (e.g., pressure on herbaceous layers) and GHG emissions.
- Co-design solutions with herders: Reduce emissions by optimizing enclosure management (CH reduction) and recycling livestock effluents; improve resilience by securing seasonal herd mobility and promoting alternatives to fuelwood (e.g., biogas, renewable energy).
- Strengthen food security: Link livestock productivity to sustainable pasture management, and integrate nutritional needs of communities into management strategies.

Objective 5: Generate Societal and Policy Impact
Context:
The project aims to strengthen local capacities and inform public policies for the sustainable management of arid ecosystems.
Sub-objectives:
- Capacity building: Train pastoral communities in sustainable management techniques (e.g., rotational grazing, carbon sequestration) and raise awareness of climate finance mechanisms (e.g., carbon credits).
- Institutional support: Collaborate with Djibouti's Ministry of Agriculture to develop tailored climate policies (e.g., incentives for carbon sequestration) and produce policy briefs for decision-makers.

Cross-Cutting Objective: Methodological Innovation
Key Approaches:
- Participatory science: Involve herders in data collection (e.g., vegetation cover monitoring) and co-design of solutions.
- Interdisciplinarity: Combine climatology, ecology, animal science, and social sciences for a comprehensive analysis.
- Technological tools: Use dynamic modeling (e.g., vegetation models + AI to predict carbon fluxes) and remote sensing (satellites, drones) for large-scale ecosystem monitoring.

Summary:
This PhD combines fundamental research (measurements, modeling) and applied research (community solutions) to address the climatic, ecological, and socio-economic challenges of Djibouti's arid ecosystems, while adopting a participatory and interdisciplinary approach.
The methodology relies on a mixed approach combining field research, modeling, AI tools, and participatory science. It is organized into 4 phases: data collection, analysis, modeling, and validation/co-design.
The project begins with in-situ field data collection to obtain precise local data on:
- GHG fluxes.
- Carbon stocks.
- Pastoral practices.
This data will be supplemented with secondary data (pastoralism statistics, satellite data on vegetation cover, NDVI, etc.).
Data Analysis
- Descriptive analysis: Means, standard deviations, normality tests (Shapiro-Wilk), and analysis of variance (ANOVA).
- Multivariate analysis: Principal Component Analysis (PCA) to identify key variables influencing GHG fluxes.
- Multiple regressions: To model relationships between:
o Pastoral practices and GHG fluxes.
o Soil characteristics (e.g., SOM) and carbon stocks.
- Spatial interpolation: Kriging for mapping GHG fluxes and carbon stocks.
- Layer overlay: Vegetation, soils, and pastoral practices to identify critical areas (e.g., emission hotspots).
Dynamic Modeling
- Vegetation dynamics will be modeled using LPJ-GUESS and/or DayCent.
- Once calibrated, models will test scenarios for:
o Pasture management practices.
o Rehabilitation of degraded areas (e.g., Prosopis removal).
o Climate scenarios to assess the impact of climate change on GHG fluxes and sequestration.
AI and Upscaling
- Machine learning (e.g., Random Forest) will be used to:
o Predict GHG fluxes from climatic, vegetation cover, and pastoral practice data.
o Extrapolate results to the Horn of Africa scale.
Participatory Science
- Herders will be involved in:
o Data collection (e.g., seasonal herd movements).
o Co-design of sustainable practices (e.g., rotational grazing, water management).
- Preliminary results will be presented to herders for feedback and adjustment.
Stakeholder Engagement
- Semi-structured interviews with decision-makers (Ministry of Agriculture, NGOs) to conduct a SWOT analysis (Strengths, Weaknesses, Opportunities, Threats) of existing policies (e.g., subsidies for carbon sequestration).
- Policy recommendations for Djiboutian institutions in the form of policy briefs.

All or part of the following skills will be appreciated:

-Scientific skills: knowledge of the carbon cycle, ecology of arid ecosystems, natural resource management, pastoralism, agronomy;

-Technical or methodological skills: field data collection and analysis (measurement of fluxes, ecosystem characteristics, surveys, and interviews), remote sensing measurements, modeling, AI, statistics, GIS, visualization tools (Python libraries);

-Cross-cutting skills: interdisciplinary approach, interest in participatory science, working with local communities, collaboration with institutions (policymakers), project management, and international collaboration (partnership coordination, planning, and monitoring), communication and valorization (scientific writing, dissemination of results, scientific outreach);

-Personal skills: adaptability, problem-solving mindset, critical thinking, leadership and autonomy, scientific ethics and integrity, intercultural sensitivity.

Application link : https://edd-projets.utoulouse.fr/

• Python
• Analyse de la variance (ANOVA)
• Machine learning