How to Bend the Curve of Biodiversity Loss?

From Conceptual Debates to Concrete Pathways

CIRAD (Agricultural Research for Sustainable Development)

Damien Beillouin (UR Hortsys) et Bruno Rapidel (UR Absys)

Context: A Scientific Consensus

  • Biodiversity as a Planetary Boundary
    • The biosphere integrity is transgressed globally, threatening ecosystem functioning and resilience (Steffen et al., 2015; Rockström et al., 2009).
    • Loss of species and functional diversity compromises ecosystem services critical to human well-being.
    • Biodiversity decline is systemic and multi-scalar, not confined to specific regions or taxa.
  • Agriculture as the Dominant Driver
    • Land-use change, habitat fragmentation, monocultures, and intensification are primary pressures.
    • Global syntheses identify agriculture as the principal driver of terrestrial biodiversity loss (IPBES, 2019; FAO, 2022).

Act I — Land Sparing vs. Land Sharing

  • Land Sparing

  • High-yield production on a smaller area

  • Freeing land for conservation

  • Works best where: Intensification is possible without major environmental trade-offs, Wild species depend on large, undisturbed habitats, Governance can secure protected areas,…


  • Land Sharing

  • Lower-intensity, wildlife-friendly farming

  • Larger agricultural area

  • Works best where: Landscapes are already highly mosaic, Smallholders dominate, Cultural landscapes have intrinsic ecological value,…

Act II — Why Land Use Matters

Bending the biodiversity curve requires addressing three concrete questions:

  • Where should we produce food?
  • How should we produce food?
  • How should land be allocated between production and conservation?

This is where the sparing–sharing framework becomes operationally useful.

Two Contrasting Views on Agricultural Productivity

1️⃣ High-Yield / Pragmatic Focus (Balmford et al.)

  • Core idea: Increase crop yields to reduce the amount of land needed for farming.
  • Other solutions like reducing food waste or changing diets are less impactful.
  • Requires access to fertilizers, improved varieties (including GMOs), markets, and good agronomic advice.
  • Jevons paradox explained: Improving efficiency can sometimes lead to more consumption, but in agriculture, these “rebound effects” are rare.
  • Approach: Maintain productivity where necessary while being mindful of environmental outcomes.

Implications for policy and practice:
- Intensify production on suitable land to spare natural habitats elsewhere.
- Focused interventions are often more effective than spreading low-intensity farming everywhere.

2️⃣ Critical / Holistic View (Beillouin et al.)

  • Core idea: Focusing only on yield has failed to deliver sustainable food systems.
    ““Decades of yield-centric agriculture, divorced from ecological and social concerns, have proven inadequate” Beillouin, 2025
  • Ignoring environmental and social consequences can lead to:
    • Biodiversity loss
    • Ecosystem degradation
    • Social inequities
  • Hidden costs of the global food system are estimated at $10 trillion (PPP, 2020).
  • Rebound / Jevons effects and spatial displacement are potential risks if intensification is not carefully managed.

Implications for policy and practice:
- Yield alone is insufficient.
- Policies must also consider ecological health, social equity, and long-term sustainability.

Two Contrasting Views on Biodiversity conservation

1️⃣ Land Sparing / Spatial Protection Focus (Balmford et al., Jan 2025)

  • Core idea: Protecting species by reserving land, reducing conversion from natural habitats.
  • Most trade-offs are assumed concave → protecting high-biodiversity areas while intensifying agriculture elsewhere is effective.
  • Focus: Spatial protection (i.e. agroecology) framed as a trade-off against yield.

Implications for policy and practice:
- Prioritize strict protected areas in landscapes with high biodiversity.
- Use high-yielding farmland to reduce pressure on remaining natural habitats.
- Spatial planning is central, but may overlook on-farm biodiversity in productive areas.

2️⃣ Critical / Holistic / Landscape-Scale View (Beillouin et al., Aug 2025)

  • Core idea: Effectiveness of spared land is overestimated; designation alone is insufficient.
  • “Effective protection requires more than spatial designation” — policies must ensure real conservation outcomes.
  • Risk of agrobiodiversity loss if only large areas are spared and productive lands are simplified.
  • GBF targets 4 and 10 highlight the need to include on-farm biodiversity in conservation strategies.
  • Win-win perspective: In some contexts, high yields and on-farm biodiversity can coexist (Beillouin et al., 2021; Jones et al., 2023)

Implications for policy and practice:
- Integrate biodiversity into productive landscapes (trees, hedgerows, semi-natural patches).
- Landscape-scale, multifunctional management is necessary.
- Conservation policies must go beyond land sparing, addressing both spared and farmed areas.

Socially Just Approaches to Agricultural Transitions

1️⃣ Yield-Centric / Top-Down Focus (Balmford et al., Jan 2025)

  • Core idea: Land sparing framed primarily as a conservation tool, focusing on yields.
  • Downstream solutions like diet change or waste reduction are considered insufficient.
  • Social equity is treated as secondary; limited attention to farmer autonomy or local contexts.

Implications for policy and practice:
- Policies risk ignoring local social dynamics and smallholder needs.
- Conservation and productivity goals may not align with social justice if interventions are top-down.

2️⃣ Critical / Socio-Ecological View (Beillouin et al., Aug 2025)

  • Core idea: Agricultural transitions are complex socio-ecological processes, shaped by power dynamics, access to resources, and social, institutional, and cultural factors.
  • Risks reinforcing corporate dominance and marginalizing smallholders:
    • “Technology-based intensification is capital-intensive, which smallholders struggle to compete with, leading to displacement or dependence on large agribusinesses.”
  • Adoption of ecologically friendly practices depends on multiple social, economic, and cultural determinants.

Implications for policy and practice:
- Design transitions that consider smallholder capacity, equity, and local governance.
- Multi-criteria approaches are needed to ensure that environmental improvements do not exacerbate social inequalities.

Act III — Why the Debate Stalls (and What Actually Matters)

The “which strategy is better?”

The real constraints are: - Socio-economic (farm size, tenure, labour) - Landscape configuration (fragmented vs. intact) - Governance (zoning enforceability) - Market incentives (price premiums, certification) - Climate constraints (water, soil fertility)

Most regions cannot implement a pure sparing or pure sharing model. The most effective solution is context-dependent!!

Empirical Evidence
- Sparing & sharing overlap; consider continuum approach (Sidemo-Holm, 2021)
- No universal solution; local context matters (Augustiny, 2025)
- Sparing & sharing can be complementary, not exclusive (Valente, 2022)

Beyond Land Sparing vs Sharing: Scientific & Practical Bottlenecks

1. Conceptual & epistemic limitations
- Simplified yield-biodiversity models fail to capture multi-dimensional socio-ecological realities.
- Evidence is fragmented: site-specific, short-term, or taxonomically narrow.
- Weak integration of ecological theory, agronomy, and socio-economic contexts.


2. Scientific debates unresolved
- Trade-offs between yield, biodiversity, and social outcomes remain poorly quantified.
- Leakage, rebound effects, and global land-use consequences are incompletely understood.
- Lack of multi-criteria metrics integrating biodiversity, ecosystem services, and social equity.

What We Know About Agriculture → Biodiversity

(200 meta-analyses, 9,000 studies)

  • Evidence is uneven: mostly USA–China–Brazil–EU; tropics under-studied.
  • What we study: mainly field practices (fertilizers, pesticides, tillage, diversification).
    → System-level & landscape drivers = rare.
  • Which taxa: strong bias to insects; plants & vertebrates largely missing.
  • Which metrics: mostly abundance; diversity <25%; functional metrics almost absent.
  • Knowledge gaps: 47/130 practice × taxon combos never tested
    (e.g., agroforestry × fauna, landscape × soil biota, water management).
  • Overall: effects = mixed, context-dependent, often based on small samples.

Source: Bonfanti et al., 2025)

Global Evidence on Agriculture & Biodiversity

1. Sources of evidence
- Second-order meta-analyses (Beillouin et al., 200+ MAs, 9,000+ studies): quantify average effects of practices on biodiversity.
- PREDICTS database: maps biodiversity responses across space, taxa, and land-use intensities.
- Other global syntheses: e.g., FAO, IPBES, FABLE modelling.

2. Core insight
- Biodiversity recovery requires bundles of complementary practices, not single measures.
- Integrating field- to landscape-level evidence from MAs + global databases (PREDICTS) supports quantitative pathways for sustainable agriculture.

Source: EAT Lancet et al., 2025; Hudson et al., 2017 (PREDICTS)

Limitations for Land Sparing / Sharing Assessment

Key methodological constraints

  • Coarse land-use categories:
    GLOBIO (that uses PREDICTS) aggregates agriculture into broad classes (e.g., “extensive” vs “intensive”), making it difficult to distinguish agroecological practices such as intercropping, agroforestry, or hedgerows.

  • Fixed MSA values per category:
    Mean Species Abundance (MSA) is static (e.g., intensive agriculture = 10%), failing to reflect variability within practices or improvements from biodiversity-friendly farming.

  • No spatial configuration:
    Effects of land sparing vs land sharing depend on landscape layout, corridors, and patch size, which is poorly considered.

  • Limited representation of ecosystem services:
    Pollination, pest control, soil fertility, water retention benefits from diverse farming systems are ignored, underestimating potential win-win outcomes.