Influence of agroforestry plant species on the infiltration of S-Metolachlor in buffer soils

Highlights

• We evaluated how 4 plant species modulate water and S-Metolachlor fluxes in soils.

• The trees increased the soil organic matter content and sorption of S-Metolachlor.

• The trees increased the macroporal flow in buffer soils.

• The composition of plant species influences the efficiency of agroforestry buffers.

• Introducing trees in vegetated buffers help reduce herbicide diffuse pollutions.

Abstract

Agroforestry practices have been acknowledged for reducing pesticide losses while maintaining land productivity. Pesticide removal from overland flow results from great infiltration capacities of the buffer soils. This can in turn threaten the quality of groundwater in case of poor pesticide sorption and degradation in the root-influenced zone. These mechanisms and their balance are likely to be influenced by plant species. However, little is known about the role of agroforestry species in the infiltration of herbicides. The aim of this study was thereby to evaluate how popular agroforestry species modulate the infiltration of water and of a widely used herbicide. We established large buffer microcosms by planting Brome grass, Black walnut, Pin oak and Poplar trees in repacked soil columns. After a growth season of 4 months, we performed ponded infiltration experiments with bromide and S-Metolachlor. We used then the HYDRUS 1D model to compare the hydrodynamic properties and S-Metolachlor transport patterns between the microcosms. In addition, we compared the sorption properties of the rhizosphere and bulk soils. We found that the tree species increased the sorption of S-Metolachlor in soil with Kd being 3 times greater than in the un-vegetated and Brome grass microcosms. Poplar trees increased the hydraulic conductivity (Ksat) compared to the control and was associated to a low retardation of S-metolachlor, which increases the risk of groundwater contamination. With slightly reduced Ksat and retardation factor in the root zone, 1.6 to 1.8 times greater than in the control treatment, Black walnut appears as an optimal species for mitigating S-Metolachlor. The Brome grass and oak microcosms had the lowest Ksat of all treatments and S-Metolachlor retardation factors were equal and slightly increased compared to the control, respectively. These results show that agroforestry buffer's efficiency can be optimized by selecting appropriate species.

Introduction

Pesticides losses from croplands via runoff or drainage constitute up to 10% of pesticides sprayed to protect crops from pests and weeds (Louchart et al., 2001; Tang et al., 2012; Voltz and Louchart, 2001). This non-point source pollution (NPSP) threatens the quality and ecological health of both surface water bodies and groundwater thereby restricting specific usages such as drinking water supply and engendering significant remedial cost worldwide (Li and Jennings, 2018; Reichenberger et al., 2007; Schultz et al., 1995). Pesticides losses from croplands to surrounding surface and sub-surface water bodies is closely linked to landscape organization and land management (Louchart et al., 2001; Tang et al., 2012; Udeigwe et al., 2015; Voltz and Louchart, 2001). Non-cropped elements of landscapes such as vegetated ditches, hedgerows or vegetated riparian zones reduce overland runoff, filter shallow groundwater and mitigate pesticides thanks to an array of hydrological and biochemical processes (Dosskey, 2001, Dosskey et al., 2010; Lacas et al., 2005; Needelman et al., 2007; Reichenberger et al., 2007). However, despite the financial and technical support for buffer's implementation, farmers remain reluctant to convert croplands into non-productive buffer zones (Lovell and Sullivan, 2006). Agroforestry practices such as in-field agroforestry buffers or alley cropping, by combining intensively cropped areas with economically viable non-cropped buffer zones, may offer an optimal landscape setting for both maintaining land productivity and reducing pesticide losses (Jose, 2009; Nair, 2007). In alley cropping agroforestry systems, high value hardwood trees are planted in rows, between which, crops are established in allay ways (Kremer and Kussman, 2011). Agroforestry buffers refer to in-field, edge-of-field or riparian buffers planted with both grass and wooded species (Caron et al., 2010a, Caron et al., 2010b; Udawatta et al., 2011).

Evidence on efficiency of agroforestry systems in reducing NPSP from overland flow has been provided by several studies across North America and Europe. Pesticides abatements rates varying between 58 and 90%, have been measured between the inlet and outlet of edge-of-field and in-field agroforestry buffers (Borin et al., 2010; Caron et al., 2010a, Caron et al., 2010b; Lin et al., 2011). The pesticide abatements in those studies were attributed to significant runoff reduction and sediment trapping. Indeed, buffer vegetation provides roughness that slows overland flow and increases infiltration duration (Caron et al., 2010a; Dosskey, 2001; Dosskey et al., 2010). In addition, enhanced evapotranspiration by buffer vegetation with deeper root systems induces drier soil moisture conditions than for adjacent row crop soils thereby enhancing infiltration and water storage capacities in the buffers (Anderson et al., 2008; Dosskey et al., 2010; Lin et al., 2011; Sahin et al., 2016; Schultz et al., 1995). Moreover, roots and biological activity increase the soil macroporosity in vegetated buffers creating preferential flow pathways. While the pesticide reduction from overland flow in agroforestry buffers has been fairly well described, the fate of pesticide in buffer soils and the risk for groundwater contamination remains to be documented.

Preferential flow in soils leads to rapid leaching of pollutants and consequently to non-equilibrium sorption which increases the risk of groundwater contamination (Dages et al., 2015; Pot et al., 2005; Vanderborght et al., 2002). However, the vicinity of root channels, which are major conducting macropores, might be enriched in specific soil organic carbon and microbial biomass enhancing sorption and degradation potentials. Efficiency of agroforestry buffer zones depend on a balance of these mechanisms.

The magnitude of water and pesticide infiltration fluxes is assumed to be mainly dependent on vegetation type because of contrasted root biomass distribution along the soil profile (Lin et al., 2011; Schultz et al., 1995), diverse composition of soil organic carbon and microbial activities (Chu et al., 2013; Lin et al., 2004, Lin et al., 2010) and differing soil structure and moisture modification (Anderson et al., 2008; Sahin et al., 2016; Schultz et al., 1995; Seobi et al., 2005). Pesticide retention patterns would then be contrasted among plant species and between the rhizosphere and the soil matrix (Chu et al., 2013; Lin et al., 2010; Rodríguez-Cruz et al., 2012). Water and pesticide transport and retention in the root zone of a few herbaceous species have been evaluated and compared (Belden and Coats, 2009; Lin et al., 2004, Lin et al., 2011) but there is, to our knowledge, no study investigating the influence of temperate agroforestry species on those mechanisms. Accordingly, the objectives of this study were i) to compare the influence of popular agroforestry buffer trees and herbaceous species on the mechanisms of water and herbicide fluxes and ii) to compare soil textural, chemical and sorption properties between the bulk soil and rhizosphere of the tested species.

In order to isolate the effect of the plant species from pedo-climatic effects that can be locally highly variable and difficult to monitor precisely, we chose to implement a large microcosm approach. The water and herbicide infiltration parameters obtained from microcosm approaches might not be directly transferable to all field situations. However, the differential in water and herbicide infiltration parameters measured between microcosms planted with various species indicates if a given species has the ability to increase or decrease the hydraulic conductivity or herbicide sorption in the root influenced zone. This information is useful for designing agroforestry buffers with optimized herbicide mitigation capacity.