Robustness versus performance

Two agricultures put to the test by crises and the living world

Two agricultural regimes: inhabiting an environment or optimizing a system

Contemporary agriculture operates between two poles. The first is based on the logic of being part of the living world: soils structured by biological activity (earthworms, micro-organisms), hedgerows regulating water and climate flows, crop diversity, short circuits and territorial roots.

INRAE research shows that functional biodiversity - particularly soil biodiversity - plays a central role in fertility, pest regulation and the resilience of agricultural systems (INRAE, 2021).

By way of example, the DEPHY (Ecophyto) programs document farms that have significantly reduced pesticide use while maintaining economic viability, by mobilizing long rotations, associated crops and biological regulation (Ministry of Agriculture, 2022).

The second pole is based on a logic of production optimization: heavy mechanization, chemical inputs, crop specialization and integration into globalized markets. Here, agriculture becomes a system driven by indicators (yield, productivity), reinforced today by precision agriculture (sensors, GPS, predictive models).

These two regimes reflect two relationships with the environment:

an ecological and relational relationship,
an instrumental and calculative one.

Robustness and performance: logics that cannot be superimposed

Robustness refers to a system's ability to maintain its functions despite disturbances. It is based on diversity, redundancy and ecological interactions. The French Environment and Energy Management Agency (ADEME) points out that agroecological systems, by promoting soil organic matter and biodiversity, improve water retention capacity and yield stability (ADEME, 2020).

A significant example is conservation agriculture systems: by increasing soil organic matter levels, they store more water and are more drought-resistant (Pellerin et al., 2019).

Conversely, productive performance is measured in yield. In France, Agence Bio data show persistent gaps between organic and conventional farming, particularly for field crops (Agence Bio, 2023).

However, this performance is based on structural dependencies:

nitrogen fertilizers from fossil fuels,
pesticides,
machinery,
international markets.

Analyses by the French General Council for Food, Agriculture and Rural Areas highlight this systemic vulnerability: productivity is high, but conditioned by unstable external flows (CGAAER, 2021).

In this way, robustness and performance are not in direct opposition to each other: they are the result of different dependency regimes.

Crises as indicators: water, energy, health, geopolitics

Contemporary crises act as tests of resistance.

Energy costs and the geopolitical situation

Rising fertilizer prices in 2022 and 2026, linked in particular to international tensions, have strongly affected intensive systems. Fertilizer-independent farms were less exposed (FAO, 2022).

Water stress

Droughts are becoming more frequent in France. BRGM (Bureau de Recherches Géologiques et Minières) reports a general decline in water tables in several regions (BRGM, 2023). Mega-basins" illustrate a technical response aimed at securing irrigation, but they crystallize conflicts of use: appropriation of the resource, evaporation, impacts on ecosystems.

At the same time, digital infrastructures are becoming water players. Data centers require large quantities of water for cooling, introducing new competition between agriculture and digital technology (ADEME, 2022).

Health effects

ANSES research highlights the links between agricultural intensification, pesticide use and risks to human and environmental health (ANSES, 2021).

Climate change

Specialized systems are more vulnerable to climatic hazards. Conversely, diversification (mixed farming, agroforestry) helps to spread risks (Pellerin et al., 2019).

Land, water, data: towards a recomposition of power relations

The growing interest in agricultural land by major economic players is part of this context of systemic tensions. Bill Gates, who has become a major agricultural landowner, illustrates this dynamic.

Land is becoming a strategic asset for several reasons:

food security,
storage of value, scarcity
energy production,
agricultural data capture.

Precision farming, supported by artificial intelligence, is also transforming practices: optimizing inputs, anticipating yields, soil modeling. But these developments are increasing dependence on digital and energy infrastructures.

At the same time, the issue of water is becoming central. Irrigated agriculture, domestic uses and industrial needs (notably digital) are all competing for water. The work of France Stratégie underlines that this tension could become structuring in the coming decades (France Stratégie, 2022).

From system robustness to the ability to learn together

Analysis of the two agricultural models reveals a profound tension between two forms of intelligence:

an intelligence of forecasting and optimization,
an intelligence of relationships and adaptation.

But this opposition is now reaching its limits. Systemic crises show that neither performance alone, nor robustness in isolation, is enough.

What is decisive is the ability of players to learn together.

Agroecological transitions in France rely heavily on farmer collectives (CIVAM networks, DEPHY groups, GIEEs) who experiment, share and adjust their practices. Knowledge is not simply transmitted, but co-produced through experience.

Similarly, agricultural technological innovations mobilize ecosystems of players: researchers, engineers, farmers, institutions.

So, robustness lies not just in soils or techniques, but in collective learning dynamics. This implies :