WATERSEED - The role of water in the spatial structuring of plant biodiversity in wine-growing agrosystems

In order to limit the negative effects of agricultural intensification, it is necessary to rethink the interactions between ditches, inter-rows of perennial crops, headlands and agricultural plots. These elements host a spontaneous vegetation that provides numerous ecosystem services in terms of regulation and supply, and facilitates water circulation in the landscape. WATERSEED will analyse how water structures the distribution patterns of plant communities, the dispersal of seeds by water and the spatial structure of species according to soil moisture gradients.

Based on the hypothesis that water is a key driver in structuring plant communities in Mediterranean regions, the WATERSEED project aimed to:

• better understand the determinants of potential plant biodiversity structuring in hydrographic networks,
• analyze the factors of seed dispersal by hydrochory, and
• integrate the acquired knowledge into a model simulating dispersal and growth of spontaneous vegetation cover in order to test scenarios of interstitial space redesign for maximizing ecosystem functions of plant cover.

Approaches

Drawing on experimental plots and domains (Domaine de la Jasse and the Roujan Observatory) of LISAH, we highlighted increased plant diversity and a greater importance of hydrochorous and zoochorous dispersal in ditch networks compared to vineyard plots. From a field experiment, we demonstrated the variability of seed dispersal factors, depending on intrinsic traits (shape, buoyancy) and extrinsic conditions (micro-topography, vegetation cover density). Finally, the model built from these observations and experiments justified the importance of cover management practices to improve the resilience of vineyard systems facing hydroclimatic extremes.

This project re-examined the role of water in agroecosystems as a driver of plant biodiversity structuring. It advocates for a better understanding of the functional properties of vegetation in relation to water flows, in order to design solutions that maximize both water resources and biodiversity.

Results

Seed bank analysis

It highlighted the influence of intra-plot heterogeneity (variability in grass cover practices between rows and alternation between plots, headlands and ditches) on the composition and species richness of the seed bank, with an increase in these indicators in neighbouring uncultivated areas (headlands and ditches). These results suggest the importance of non-productive areas within the landscape in the composition and species richness of the soil seed bank.

Hydrochory experimentation

Seed morphology plays a role in their dispersal potential during secondary dispersal events, with more spherical seeds being more susceptible to being displaced by runoff. Seed morphologies with appendages (hooks, hairs, etc.) were more sensitive to soil surface properties that facilitated their trapping (roughness, vegetation cover).

Secondary dispersal therefore plays an important role in maintaining plant biodiversity within agroecosystems, and even increasing it without direct human intervention.

A striking result is the virtual absence of seed movement within the most grassy inter-rows.

Mechanistic modelling

By combining several models (multi-species plant growth under water stress, simulation of friction between vegetation cover and runoff water, and seed transfer), it has been possible to begin studying the influence of climatic variables (temperature, rainfall, ET0) on the friction generated by vegetation cover on water circulation.

The use of IPCC scenarios to simulate climate change in the model input seems to show an increase in friction generated by vegetation cover in the coming years, probably as a result of faster and more significant vegetation cover development during periods of heavy rainfall.

The prototype created will enable more practical questions to be answered, for example: How can the model be used as a virtual laboratory to test new spatial and temporal arrangements of landscape elements in order to maximise water resources and plant biodiversity?

However, this modelling work has also highlighted gaps in our knowledge. In order to make progress, it will be necessary to prioritise other seed dispersal factors and the importance of plant community determinants.

Valorisation for seedling recognition of collected data

An original dataset compiling photos of seedling development forms the basis of a research collaboration with the Pl@ntNet team to offer a service dedicated to the early recognition of wild species.

Participants

INRAE units involved

• UMR G-EAU - Gestion de l’Eau, Acteurs, Usages
• UMR LISAH - Laboratoire d'étude des Interactions Sol - Agrosystème - Hydrosystème
• UMR ABSYS - Agrosystèmes biodiversifiés

Partners

• UMR CITERES - Cités, Territoires, Environnement et Sociétés - Université de Tours