IFOSSA - Distribution of interaction networks and functions in a spatially heterogeneous system, case of agroforestry

Objectives

Trees in agricultural fields create specific habitat mosaics through their above- and belowground perennial structures. They strongly influence the heterogeneity of microclimates, which vary seasonally depending on tree phenology and morphology. As a result, they modify associated biodiversity and the activity of other organisms. In addition, the presence of trees and, in some cases, associated grassy strips, leads to localized concentrations of organic resources from above- and belowground litter, an increase in soil organic carbon, and higher nutrient availability. Soil structure and aggregate stability can also be improved beneath tree rows, creating spatial heterogeneity in soil microhabitats. The overall aim of this project is to assess how ecological interaction networks and ecosystem functions respond to spatial reorganization of the environment following the establishment of trees and grassy strips in arable systems. The main hypothesis is that introducing trees alters the spatial structure of fields in ways that enhance biodiversity and multifunctionality in agricultural landscapes.

Operational objectives are to (i) characterize biodiversity changes within taxa and trophic groups; (ii) infer ecological interaction networks among studied organisms; (iii) establish relationships between ecological networks, soil physicochemical functioning, and plant growth, in order to understand multifunctionality in these systems.

Approaches

This study was carried out at the DIAMs trial (Mediterranean Agroforestry under Water Constraints), located South of Montpellier at the INRAE experimental station UE Diascope (Mauguio). The experimental design covers 5 hectares and follows a factorial layout with three blocks. Each block compares three treatments: forest plantations (black locust, Robinia pseudoacacia), an agroforestry system, and arable crop plots. Within the agroforestry system, two main habitats are distinguished: tree rows with associated grassy strips and inter-row cultivated alleys.

Crops (rotations of cereals and legumes) are managed under conventional low-input practices. The site is fully instrumented, enabling continuous monitoring of radiation at the soil surface, soil temperature and moisture at different depths, root growth, and soil solution nutrient concentrations. Four treatments are studied: arable control (C), forest control (F), and, within agroforestry, tree row (AF-LSA) and cultivated alley (AF-C).

In each treatment and block, five georeferenced sampling points were selected, providing five field replicates to capture variability. In total, sixty measurements per variable (biodiversity, function) were obtained.

Biodiversity was assessed through multiple approaches: environmental DNA metabarcoding (bacteria, fungi, protists), Baermann funnel extraction (nematodes), Macfadyen method (microarthropods), soil hand-sorting (macroinvertebrates), pitfall traps (surface-dwelling macroarthropods), pan traps (flying arthropods in the canopy), leaf dissection (foliar arthropods), camera trapping (mammals), photographic surveys (pollinators).

Multifunctionality was measured through weed seed predation assays (seed cards), soil suppressiveness against root-knot nematodes (Meloidogyne javanica) in laboratory assays, litterbag experiments assessing decomposition of black locust litter, soil aggregate stability (mean weight diameter of macroaggregates after wet sieving), monitoring of crop growth and yields, as well as black locust growth and resource allocation strategies.

Results

Biodiversity

Morphological and molecular identification characterized communities of soil engineers, surface-dwelling macroarthropods, free-living and parasitic nematodes (plant-feeding and entomopathogenic), fungi and bacteria (both free-living and nematode-associated), weeds, and aerial invertebrates (herbivores, predators, parasitoids). Several hundred taxa were identified, providing an almost exhaustive inventory of taxa and trophic groups present during the growing season (some protocols were repeated three times in spring). Key findings include:

• Herbivory by mammals. Rabbit activity-density was higher in agroforestry plots than in monocultures.
• Flying insects. Invertebrate diversity was higher in the presence of trees than in cultivated areas. However, more individuals were collected in pan traps from monocultures than from forest plantations, with agroforestry showing intermediate values. Trophic group composition shifted with land use and across the growing season.
• Surface macrofauna. Tree presence did not affect the abundance and diversity of surface-dwelling invertebrates, unlike findings from the literature where agroforestry often enhances soil organism diversity.
• Free-living nematodes. Monocultures showed higher diversity of herbivores and lower diversity of predators compared to other treatments.

Soil functions

Interaction networks were reconstructed from field observations, either directly (bipartite interactions) or inferred by combining co-occurrence data with trophic knowledge bases on soil organisms. Structure and composition of reconstructed networks were analyzed in relation to agroforestry spatial organization. Highlights include:

• Weed regulation. Seed predation was highest in arable controls (~70%) and lowest in forest controls (~30%). Late-winter granivore activity (mainly Carabidae) was greater in arable plots than in agroforestry or forest treatments, likely reflecting species selection.
• Suppression of phytoparasitic nematodes. Greater suppressive capacity was observed in samples from vegetated plots compared to monocultures.
• Soil structural stability. Water infiltration rates were significantly higher under tree rows compared to inter-row alleys, highlighting the positive role of tree rows in field hydrology. Soil structural stability and infiltration rates were strongly correlated with soil organic carbon, underlining the structuring role of organic matter status for other soil functions.

These results still require further integration, particularly to link biodiversity and functional patterns in agroecosystems. Some findings are already published or in press, while others will not be published but will serve as a basis for broader reflections on biodiversity assessment in diversified agroecosystems under Mediterranean climate constraints.

Participants

INRAE structures

• UMR Eco&Sols - Écologie fonctionnelle & biogeochimie des sols & des agro-ecosystemes - IRD / Institut Agro Montpellier / Cirad
• UE DIASCOPE - Unité expérimentale Diversité (des plantes et de la biodiversité en général) et de leurs capacités d’Adaptation à des environnements divers
• UMR DGIMI - Diversité, Génomes & Interactions Microorganismes-Insectes
• UMR PHIM - Plant Health Institute Montpellier - Cirad