How do plants avoid hard soil?

Soil compaction affects global crop cultivation by reducing root penetration in both the upper and deeper soil layers (1). Modern agricultural practices have exacerbated soil compaction, largely because of intensification of operations leading to the deployment of heavier machinery and tillage practices (2, 3), severely degrading ~65 million hectares of land globally (4). Compaction increases soil bulk density and reduces soil porosity, limiting the availability and transport of water and nutrients (4, 5). The decrease in soil pore space, especially in large air-filled pores (Fig. 1, A to D, figs. S1 and S2, and movies S1 and S2), also restricts diffusion of gases between roots and the rhizosphere (6). To deal with compacted soils and penetrate cracks, roots are reported to undergo adaptive growth responses, including increased radial expansion of root tips (1). However, the predominant response of roots is cessation of growth, for which the mechanistic basis remains unclear. Here, we report that entrapped ethylene functions as a key signal regulating root growth in compacted soils.

Fig. 1 Soil compaction reduces the larger pores and triggers root growth responses mimicking ethylene treatment.

(A and B) CT images showing higher porosity (outlined in white) in uncompacted [1.1 g cm^–3 bulk density (BD)] (A) versus compacted soil (1.6 BD) (B). (C and D) Representative three-dimensional images of air-filled soil pores for a 100 × 100 × 100 voxel region from 1.1 BD (C) and 1.6 BD (D) soil cores. (E and F) Arabidopsis EIN3-GFP reporter exhibits elevated signal after covering root tip with high-vacuum silicone grease (+Gas Barrier) for 10 hours (F) relative to control (–Gas Barrier) (E). (G and H) Confocal images of radial cross sections of rice primary roots through meristem (MZ), elongation (EZ), and differentiation (DZ) zones grown in 1.1 BD (G) and 1.6 BD (H) soils. (I and J) Relative to control rice roots (I), roots treated with 10 ppm ethylene exhibit cortical cell expansion (J), mimicking the effect of compacted soil conditions (H). Scale bars, 1.25 mm [(A) and (B)], 100 μm [(E) to (J)].

Ethylene is produced by root tissues, and its level increases when roots are exposed to compacted soil (7, 8). Ethylene concentrations outside the root could increase as a result of the reduction in soil pore space in compacted soil, which affects gas diffusion from root tissues (Fig. 1, A to D, and figs. S1 and S2). To test this “restricted gas diffusion” model, we used the EIN3-GFP (green fluorescent protein) Arabidopsis ethylene response reporter (9) (fig. S3, A to C) and examined the effect of covering root tips with a gas-impermeable barrier. In agreement with model assumptions, restricting gas diffusion from root tip tissues triggered a rapid and sustained increase in EIN3-GFP in root elongation zone cell nuclei relative to controls (Fig. 1F versus Fig. 1E; fig. S3, D to G). This result is consistent with (i) limitation of ethylene release from root tip tissues and (ii) changes in gas diffusion rate between roots and the external environment, inducing ethylene accumulation and signaling. To rule out that changes in ethylene signaling were related to reduced oxygen levels in root tip tissues, we treated roots expressing the hypoxia markers pPCO1:GFP-GUS, pPCO2:GFP-GUS (10), and RAP2.12-GFP (11) with the gas-impermeable barrier. Hypoxia reporters were not induced by the gas barrier but were induced by submergence (figs. S4 to S6). We conclude that EIN3-GFP induction results from restricted ethylene diffusion rather than from hypoxic conditions (11).

Roots exposed to elevated levels of ethylene exhibited growth inhibition (Fig. 1, I and J), which phenocopied the impact of soil compaction (Fig. 1, G and H). We observed that rice roots grown in 1.1 g cm^–3 (uncompacted) versus 1.6 g cm^–3 (compacted) soil bulk densities exhibited reduced root length when exposed to compacted conditions (fig. S7, A and B). Root anatomical analysis revealed that compaction caused a factor of 3 decrease in epidermal cell length (fig. S7C), matched by a factor of 3 increase in cortical cell diameter (compare Fig. 1, G and H, and fig. S7D). Similarly, ethylene treatment reduced root length (fig. S8A) while increasing root width (Fig. 1, I and J) by decreasing epidermal cell length and increasing cortical cell diameter (fig. S8, B and C).

To directly test the functional importance of ethylene during soil compaction, we examined root growth responses of wild-type rice versus ethylene-insensitive mutants osein2 and oseil1 (12). OsEIN2 (ETHYLENE INSENSITIVE2) encodes a key ethylene signaling component (13). OsEIL1 (EIN3-like 1) encodes a critical transcription factor in the ethylene transduction pathway downstream of OsEIN2 (9). Mutations in rice OsEIN2 and OsEIL1 genes confer ethylene-insensitive root elongation phenotypes (12) (fig. S9, A and B). To analyze the impact of soil compaction on wild-type versus osein2 root growth, we grew rice lines in columns entirely filled with either uncompacted soil (1.1 g cm^–3) or highly compacted soil (1.6 g cm^–3, with a 1-cm top layer packed at 1.1 g cm^–3 to help establish seedling root growth). Penetrometer resistance analysis demonstrated that the root elongation rate was sensitive to increased soil strength (fig. S10).

To quantify the impact of soil compaction on root length of wild-type versus ethylene mutant lines, we used computed tomography (CT; Fig. 2, A to G). CT imaging revealed that unlike the wild type (Fig. 2B), both osein2 and oseil1 roots were able to penetrate highly compacted soil (Fig. 2, D and F; quantified in Fig. 2G). This result reveals that ethylene signaling is critical for triggering root growth responses upon soil compaction. Anatomical analysis of rice mutant roots further demonstrated that under compacted soil conditions, osein2 and oseil1 root epidermal cells continued to elongate normally, whereas cortical cells did not undergo radial expansion (figs. S11 and S12) relative to the wild type (fig. S13). Moreover, this growth response also occurred in other classes of roots: Primary and lateral root growth and cortical responses induced by soil compaction were blocked in the ethylene-insensitive Arabidopsis mutant etr1 (figs. S14 to S17). Similarly, ethylene-insensitive mutants in rice (osein2 and oseil1) and Arabidopsis (ein3eil1) accumulated significantly higher shoot and root biomass in compacted soil conditions relative to the wild type (figs. S18 and S19). Hence, our rice and Arabidopsis mutant analysis revealed that ethylene plays an inhibitory role in both monocot and eudicot root (and shoot) tissues when experiencing soil compaction.

Fig. 2 Disrupting ethylene response in rice confers root growth resistance to compacted soil.

(A to F) CT images of primary roots of wild-type (WT) [(A) and (B)], osein2 [(C) and (D)], and oseil1 [(E) and (F)] in 1.1 BD [(A), (C), and (E)] versus 1.6 BD [(B), (D), and (F)]. (G) Violin plots of primary root length in uncompacted (1.1 BD) versus compacted (1.6 BD) conditions for WT, osein2, and oseil1 rice seedlings. (H to K) Representative images showing root cap area in WT [(H) and (I)] and osein2 [(J) and (K)] in 1.1 BD [(H) and (J)] versus 1.6 BD [(I) and (K)] soils. (L and M) Ethylene treatment of WT roots showing reduction in root cap area [(M) versus (L)]. (N) Violin plots showing reduction of root cap area after ethylene treatment. (O) Violin plots showing reduction of root cap area of WT but not osein2 roots when grown in 1.6 BD versus 1.1 BD soils. Columella cells are marked in red [(L) and (M)]. *P ≤ 0.05, **P ≤ 0.001, ***P ≤ 0.0001 (Student’s t test). Scale bars, 10 mm [(A) to (F)], 100 μm [(H) to (M)].

Our results suggest that reduced root growth triggered by soil compaction does not arise from mechanical impedance, but instead represents a timely response controlled by ethylene, perhaps to avoid growth in compacted soils (14). To discriminate between the effects mediated by mechanical impedance and by ethylene, we compared their impact on root tip shape. Soil compaction caused wild-type rice roots to double in width and their root caps to develop a “flattened” shape (compare Fig. 2, H and I). Soil compaction–induced radial growth and root cap shape changes were blocked in osein2 (Fig. 2, J, K, and O). Hence, root tip shape changes induced by soil compaction appear to be controlled primarily by ethylene and not by mechanical impedance. Indeed, ethylene treatment alone was sufficient to trigger equivalent changes in root width (Fig. 1, I and J, and fig. S8, B and C) and cap shape (Fig. 2, L to N, and fig. S20) similar to roots exposed to soil compaction. Therefore, in plants, ethylene represents a critical signal controlling shape changes that underpin root compaction responses.

Given ethylene’s functional importance during root responses to compaction, we investigated whether soil mechanical impedance triggered increased ethylene signaling in root tissues. We used transgenic Arabidopsis and rice expressing either an ethylene biosensor featuring EIN3 (9) or OsEIL1 sequences fused with GFP (fig. S21). In uncompacted soil, 35S:EIN3-GFP or proOsEIL1:OsEIL1-GFP reporters in root nuclei were not detectable (Fig. 3, A and D). However, when reporter lines were grown in compacted soil, both ethylene reporters were detected in root elongation zone cells (Fig. 3, B, C, and E). To probe the role of ethylene in other soil types, we grew rice reporter lines in two contrasting soils. Compaction triggered a root ethylene response in clay soil (figs. S21 and S22) and sandy loam soil (Fig. 3E and fig. S23). Hence, the ethylene-based compaction mechanism appears to operate in different soil types.

Fig. 3 Compacted soil reduces ethylene diffusion and enhances root ethylene signaling.

(A and B) Arabidopsis ethylene reporter EIN3-GFP exhibits no nuclear GFP signal when grown in uncompacted soil (1.1 BD) (A) but is clearly detected in root EZ (elongation zone) cells when grown in compacted soil (1.4 BD) (B). (C) Violin plot of GFP signal in 1.1 BD versus 1.4 BD soil in EZ of 35S:EIN3-GFP/ein3eil1. (D and E) Relative to 1.1 BD soil (D), a rice OsEIL1-GFP–based ethylene translational reporter exhibits elevated signal in compacted (1.6 BD) soil (E). (F and G) Schematic figures of ethylene diffusion (denoted by red circles) in uncompacted (F) versus compacted (G) soil, illustrating preferential accumulation of ethylene around and in root tissues. (H) Model simulation showing rate of bulk diffusion of ethylene in soil pores in uncompacted (green line) and compacted (red line) soil. (I) Graphical representation of quantification of ethylene across 1.1 BD and 1.6 BD soil layers (1 cm). After 20 ppm of ethylene was injected in the top chamber, we used gas chromatography–mass spectrometry to measure ethylene diffusion in the bottom chamber across empty, uncompacted (1.1 BD), and compacted (1.6 BD) soils. ***P ≤ 0.0001 (Student’s t test).

How does soil compaction induce elevated ethylene signaling in root tissues? Mechanical impedance could cause roots to up-regulate ethylene synthesis. Profiling of the ethylene precursor 1-aminocyclopropane-1-carboxylic acid (ACC) in excised rice root tips detected no change in levels after growth in compacted soil versus uncompacted controls (fig. S24). Alternatively, plant roots may sense soil compaction by monitoring ethylene levels. Mathematical modeling predicted slower outward ethylene diffusion rates under compacted soil conditions (Fig. 3H and fig. S25) due to the decreased volume of air-filled pores (1) (Fig. 1, A to D, and movies S1 and S2). This will result in a higher ethylene concentration close to roots (Fig. 3, F and G) and therefore in root cells, consistent with soil compaction triggering an ethylene response (Fig. 3, B, C, and E).

We directly tested whether soil compaction restricted gas diffusion by experimentally measuring ethylene’s ability to move through compacted versus uncompacted soil. A 1-cm-thick soil column (connecting two air-filled chambers) was either left empty (control) or filled with uncompacted soil (1.1 g cm^–3) or compacted soil (1.6 g cm^–3) (Fig. 3I and fig. S25B). Ethylene was injected into the upper chamber (an increase in pressure was avoided) and ethylene concentrations were subsequently measured over time in the lower chamber until an equilibrium was reached between the chambers. In agreement with gas diffusion simulations, ethylene levels rapidly reached an equilibrium with the lower chamber in control conditions without soil resistance (Fig. 3I). Ethylene was also able to diffuse through uncompacted soil, albeit more slowly than the empty control by a factor of 10 to 50; in contrast, ethylene was unable to diffuse through compacted soil and was still undetectable in the lower chamber at 20 days (Fig. 3I). This result demonstrates that soil compaction (and the associated increase in soil moisture due to decreased porosity) affects ethylene diffusion rates, consistent with our restricted gas diffusion model. This much slower ethylene diffusion in compacted soil results in an enhanced ethylene response in root cells. This entrapped ethylene gas provides a fast and reliable signal for plants to interact with their environment, because nearly all roots produce ethylene under normoxic conditions (15).

Our results reveal how roots regulate growth responses to soil compaction. First, the inhibition of root growth by compacted soils is triggered by ethylene signaling, rather than simply by mechanical forces. Second, rather than using a dedicated mechano-perception mechanism, roots appear to sense soil compaction through restricted diffusion of this gaseous signal from the plant cells to the soil, causing ethylene to accumulate in root expansion zone cells and inhibiting elongation growth. Third, compaction and soil moisture status appear to have an impact on root elongation, not only because they control soil strength, but also through regulating ethylene diffusion. Fourth, we propose that ethylene acts as an early warning signal for roots to avoid compacted soils (14); if so, this could provide a pathway for how breeders might select crops resilient to soil compaction.