Introduction

Rare earth elements (REEs) include scandium (Sc), yttrium (Y), and 15 lanthanides, which have similar physical and chemical properties1. These elements typically exist as trivalent cations in the environment and can be categorized into two subgroups based on their atomic numbers: light REEs (LREEs) and heavy REEs (HREEs)2. In soils, LREEs such as cerium (Ce) tend to be enriched in organic matter, while HREEs such as Y are commonly associated with iron oxides3. REEs are widely used in high-technology industries, including catalysts, smartphones, batteries, solar panels, and magnets3, earning them the label of “technology-critical elements”4. However, the environmental risk associated with REEs is increasing due to discharge from REEs mining areas and the disposal of REE-containing wastes, leading to REEs being considered as emerging contaminants5. The release of REEs into the environment may affect crop yield and quality, hence influencing nutrient cycling and the food chain6.

In agriculture, REEs-based fertilizers have been applied in China due to their reported benefits on crop yield, nutrient uptake, seed germination, and chlorophyll concentration1. For example, the root weight of Triticum durum Desf. was significantly enhanced at lanthanum (La) concentrations of 0.01 and 0.1 mM compared to the control7. The concentrations of calcium (Ca), magnesium (Mg), iron (Fe), manganese (Mn), and molybdenum (Mo) significantly increased in Nymphoides peltata (Gmel.) O. Kuntze treated with 5 mg Y L− 1 by hydroponic experiments8. Chlorophyll a + b levels in the leaves of Triticum aestivum L. significantly increased at 0.5 and 1.0 µM of both La and Ce compared to the controls9.

Despite the beneficial effects of REEs on plant growth, high doses of REEs could be harmful to plant physiology. The respiration rate of Glycine max L. significantly decreased at 0.4 and 1.2 mM of La10. In addition, proline concentration in the leaves of Oryza sativa L. increased at a dose of 150 mg La kg− 1 dry soil, indicating the crop was experiencing La-induced stress11. Due to the similar ionic radii of REEs and Ca, REEs can compete with Ca for binding sites or interfere with Ca-dependent signaling systems in Amaranthus caudatus L12. In Phytolacca americana L., LREEs were found to utilize Ca ion channels for translocation, while HREEs share transport pathways with aluminum (Al)13. The dual beneficial and toxic effects of REEs are referred to as “hormetic effect”, where plant growth is promoted at low doses but inhibited at high doses14.

Previous studies on plant physiological responses to REEs have primarily focused on LREEs under hydroponic conditions7,9,10, which may not typically represent REEs behavior in the terrestrial environment. The bioavailability of soil REEs to plant is controlled by soil properties such as texture, pH, redox potential, and organic matter1,15. For example, REEs are poorly retained on particle surfaces in coarse-textured soils, leading to higher REEs concentrations in the soil solution and increased uptake by the roots of Boehmeria nivea L16. However, limited studies have investigated the physiological effects of soil LREEs and HREEs on crops, particularly vegetables. Taiwan, a humid tropical country with an annual rainfall of approximately 2000–2500 mm, is dominated by acidic soils, where REEs-containing processing industries are commonly located in many agricultural areas. According to soil REEs concentrations in Taiwan, the most abundant elements of LREEs and HREEs are Ce and Y, respectively17. Moreover, no control standard currently exists for soil REEs concentrations in many countries including Taiwan.

Although previous studies have investigated the hormetic effects of REEs14, the influence of soil texture on REEs bioavailability16, and the differential toxicity between LREEs and HREEs13, these factors have seldom been considered together. Thus, this study focused on the effects of Ce and Y on pak choi (Brassica rapa subsp. chinensis (L.) Hanelt), a vegetable of dietary importance in Asia and a potential REEs accumulator, grown in two soils with contrasting textures. We hypothesized that soil texture controls REEs concentrations in the soil solution, subsequently affecting the bioavailability of REEs and the physiological responses of pak choi to different subgroups of these elements such as Ce and Y. The objectives of this study were to (1) assess soil Ce and Y bioavailability to the tested crop in two acidic soils with contrasting textures, (2) determine their effects on crop biomass and nutrient concentrations, and (3) elucidate their physiological effects on the tested crop.

Materials and methods

Soil collection and analysis

Two acidic soils derived from slate alluvium were collected from the topsoil (0–20 cm) in agricultural fields from northern Taiwan. These soils had similar pH and cation exchange capacity (CEC), low organic carbon (OC) content (< 2%), but different textures (high clay and low clay soils) (Table S1). The collection of soil samples has been obtained with permission from the landowners. The soil samples were air-dried, ground, and passed through a 2 mm sieve for further analyses in the laboratory. Soil pH was measured using a 1:1 (w/v) soil to deionized water mixture with a glass electrode (LAQUA F-71, HORIBA, Kyoto, Japan)18. The pipette method was used to determine the particle size distribution19. OC was determined using the Walkley-Black wet oxidation method20. Ammonium acetate leaching method was used to determine CEC21. Available nitrogen (N) was extracted with 2 M KCl (1:10, w/v) and determined through the Kjeldahl method22. Available phosphorus (P) was measured through Bray 1 method23. Available potassium (K) was determined by Mehlich 1 method24. Aqua regia (37% HCl + 65% HNO3, 3:1) was used to digest total Ce and Y in the experimental soils25, and the total concentrations of Ce and Y were determined by inductively coupled plasma optical emission spectroscopy (ICP-OES, Optima 2100DV, PerkinElmer, Waltham, MA, USA). The soil standard reference material (REE-1 from the Canadian Certified Reference Materials Project) was used to check the recoveries of Ce and Y, which were 103 and 86.7%, respectively.

Experimental setup

For the pot experiments, field soil samples were air-dried, ground, and passed through a 5 mm sieve to remove rock fragments and roots. To prepare REEs-spiked soils, the chloride forms of REEs were used, including CeCl3 and YCl3. Ce and Y were chosen as the representatives of LREEs and HREEs, respectively. Each of these REEs chlorides was individually spiked into the soils for the respective treatments. The target added doses for each treatment were as follows: Ce and Y at 0, 25, 50, and 100 mg kg−1. These doses were chosen to ensure that the hormetic effect of REEs on the tested crop could be observed based on previous studies9,11,14,15. The corresponding additional amounts of these REEs chlorides were calculated based on the target added doses. To ensure the homogenization of spiking, the REEs chlorides were dissolved in deionized water and gradually sprayed onto the soils while continuously stirring. After spiking, the REEs-spiked soils were immediately divided into four pots (15 cm in diameter and 14 cm in depth) for each treatment, with each pot containing 1.5 kg of soil. Soil moisture was maintained at 60% of field capacity, with weekly water supplementation. The pots were kept at room temperature and incubated for 50 days to ensure adsorption equilibrium before planting.

Plant harvest and soil solution analysis

The seeds of pak choi were obtained from the Known-You Seed Company, a professional seed company in Taiwan. Seedlings of the tested crop were generated by germinating mature seeds in potting soil for 10 days. Afterwards, three seedlings were transplanted into each pot and grown in a phytotron (25/20℃ day/night temperature) for 25 days. Each pot was supplied with 300 mg of N through (NH4)2SO4, along with 120 mg of P and 150 mg of K from KH2PO4. Each treatment was conducted in four replicates, and the soil moisture in the pot was maintained at 60% of field capacity to ensure a well-aerated and non-limiting moisture condition for plant growth.

At the end of the pot experiment, roots and shoots were harvested separately and oven-dried at 70℃ for 72 h to a constant weight for dry weight determination. The root and shoot samples were ground and passed through a 35-mesh sieve for digestion. Additionally, the third fully expanded leaf from one plant in each pot was collected for physiological analyses.

One day before the harvest, soil solution from each pot was collected using the Rhizon soil moisture sampler (Eijkelkam, Giesbeek, The Netherlands). The sampler consisted of a 10 cm porous polymer tube with a porous glass fiber core and was inserted into the pot. The soil solution was drawn through the tube by suction with an attached syringe. The concentrations of Ce and Y in the soil solutions were determined using inductively coupled plasma mass spectrometry (ICP-MS, Agilent 7700e, Agilent Technologies, Santa Clara, CA, USA), while the concentrations of Ca, Mg, Fe, and Al were determined using ICP-OES26. The plant material was used in accordance with all relevant institutional, national, and international guidelines and legislation27,28.

Physiological measurement

Concentrations of chlorophyll a + b and carotenoids in leaves were determined according to the colorimetric method29. Fresh leaves (50 mg) were ground with 200 µL 95% ethanol in a homogenizer (SH-100, KUROBO, Japan). Then, each sample was thoroughly mixed with 800 µL 95% ethanol and stored in dark at 4℃ for 30 min. The samples were centrifuged at 12,000 rpm and the supernatants were collected for absorbance measurement at wavelengths of 664, 649, and 470 nm using a UV-visible spectrophotometer (SpectraMax ABS Plus, San Jose, CA, USA). The chlorophyll a + b and carotenoids concentrations were calculated based on the formula29.

Respiration activity in leaves was determined using 2, 3, 5 triphenyl tetrazolium chloride (TTC) reduction method30. Fresh leaves (50 mg) were placed in a test tube with 4 mL of TTC solution (0.8% w/v TTC in 50 mM phosphate buffer at pH 7.0). To ensure complete submersion, the test tubes were placed in a vacuum desiccator for 10 min. The samples were gently shaken in the dark for 24 h. After being carefully rinsed with deionized water, the leaves were ground with 5 mL of 50% (v/v) ethanol in a mortar, mixed with 5 mL of hexane, and centrifuged at 3000 rpm for 10 min. The solution was separated into two phases, and the upper phase was measured at wavelength of 545 nm using the spectrophotometer.

The proline concentration in the leaves was determined using the extraction method by acid ninhydrin31. Fresh leaves (100 mg) were ground with 10 mL of 3% sulfosalicylic acid and centrifuged at 2500 rpm for 10 min. A 2 mL aliquot of the extract was transferred to a test tube, mixed with 2 mL of ninhydrin reagent and 2 mL of glacial acetic acid. The test tube was heated at 100℃ for 60 min and then cooled rapidly. Subsequently, 4 mL of toluene was added, and the absorbance of the upper phase was measured at 520 nm using the spectrophotometer.

Plant elemental analysis

To determine the concentrations of Ce, Y, Ca, Mg, Fe, and Al in the roots and shoots, samples were digested using a microwave-assisted digestion method32. A 0.2 g sample was mixed with 6 mL of HNO3 in a Teflon vessel for 14 h. Then, the sample was pre-digested with heating to 100℃ in 10 min in a microwave oven (Speedwave Entry, Berghof, Eningen, Germany). Subsequently, 2 mL of 35% H2O2 was added, and the oven temperature was gradually increased to 120℃ in 8 min and held at 120℃ for 2 min. The temperature kept increasing to 160℃ in 5 min and held at 160℃ for 5 min. For the final stage, the temperature was raised to 180℃ in 5 min and maintained at 180℃ for 15 min. After cooling, the digested solution was diluted to 10 mL with 2% HNO3 and filtered through a 0.45 μm Millipore. The concentrations of Ce and Y in the filtrate were measured using the ICP-MS, while the concentrations of Ca, Mg, Fe, and Al were determined using ICP-OES.

For quality assurance and control, two standard reference materials, GBW 10,052 (green tea) from the National Institute of Metrology (Beijing, China) and SRM 1573a (tomato leaves) from the National Institute of Standards and Technology (Gaithersburg, MD, USA), were digested using the same digestion method. The recoveries of Ce and Y from GBW 10,052 were 82.9% and 105%, respectively. The recoveries of Ca, Mg, Fe, and Al from SRM 1573a were 96.0%, 101%, 99.2%, and 96.9%, respectively. For every ten samples in each batch, a standard solution was determined to ensure the relative error was within ± 5%.

Imaging elemental distribution in the crop leaves

To demonstrate the effect of different treatments on the distribution of Ce, Y, Ca, Mg, Fe, and Al in leaves, first leaves from the tested crop grown in low clay soil under five treatments were collected: 0 (control), 25 mg Ce kg− 1, 100 mg Ce kg− 1, 25 mg Y kg− 1, and 100 mg Y kg− 1. The method used for analysis was described in a previous study33. In summary, a laser ablation inductively coupled plasma mass spectrometer (LA-ICP-MS) was used, consisting of a 266 nm Nd: YAG laser (NWR 266, ESI, UK) for ablation and a quadrupole-based ICP-MS detector (Agilent 7800, Agilent Technologies, USA) for analysis. The laser operated at a fluence of 6.5 J cm− 2, with a frequency of 50 Hz, using a 30 μm beam and a scan speed of 469 μm s− 1. Helium (He) was used as the carrier gas at a flow rate of 800 mL min− 1 to transfer the ablated material to the ICP-MS for measurement. To correct for differences in tissue density, 13C+ was used as an internal standard to normalize the results and improve accuracy. The samples were mounted on glass slides and analyzed by LA-ICP-MS, measuring elements like 13C+, 24Mg+, 27Al+, 44Ca+, 56Fe+, 140Ce+, and 89Y+. For two-dimensional (2D) imaging, all sample sections were mounted on double sided tape (Scotch, 3 M 666, USA) in glass slides. The ICP-MS data were exported into Excel files and then transformed into 2D element distribution images using Iolite computing software.

Statistical analysis

To compare the differences in biomass, concentrations of Ce, Y, Ca, Mg, Fe, and Al, and physiological responses among the tested crop exposed to different doses of Ce and Y in two soils. Two-way analysis of variance (ANOVA) was conducted, followed by least significant difference (LSD) post hoc tests with different letters denoting a significant difference at p < 0.05. All statistical analyses were performed using R version 4.0.034.

Results

Concentrations of REEs, Ca, Mg, Fe, and al in the soil solution

The concentrations of Ce, Y, Ca, Mg, Fe, and Al in the soil solution with the control treatment represent their concentrations in the initial status, which can be compared with those treated with Ce and Y doses, and determine the changes caused by Ce and Y application (Table 1). Concentrations of Ce in the soil solution under the control treatment were 0.63 and 0.58 µg L−1 in the high clay soil and low clay soil, respectively (Table 1). As Ce doses increased, Ce in the soil solution increased in both soils. Notably, for all Ce treatments, concentrations of Ce in the soil solution were higher in the low clay soil than those in the high clay soil. In terms of Ca, its concentration in the soil solution was not significantly (p < 0.05) different across different Ce doses in both soils. Mg concentration showed significantly higher in the low clay soil than the high clay soil at any given Ce dose. No significant difference of Fe and Al concentrations were observed between different Ce doses in the high clay soil, but their concentrations increased significantly at 50 mg Ce kg−1 in the low clay soil.

Table 1 The concentrations of REEs, Ca, Mg, Fe, and al in the soil solution with Ce and Y treatments in the high clay and low clay soils at plant harvest. Values represent the mean ± standard deviation (n = 4). Columns followed by different letters are significantly different at 5% level according to the least significant difference post hoc test.
Full size table

Regarding Y treatments, concentration of Y in the soil solution increased with increasing Y doses in both soils (Table 1). Similar to Ce treatment, concentration of Y was consistently higher in the low clay soil compared to those in the high clay soil at all Y doses. While Ca concentration in the high clay soil remained unaffected by Y doses, significant increases were observed at 50 and 100 mg Y kg−1 in the low clay soil. Mg concentrations exhibited a significant rise at 100 mg Y kg−1 in the low clay soil. Regarding Fe, its concentration in the high clay soil showed no significant change, while a significant increase was observed at 100 mg Y kg−1 in the low clay soil. Similar to Fe, Al concentration showed no significant difference in the high clay soil, while it significantly increased at 100 mg Y kg−1 in the low clay soil.

Physiological responses to REEs exposure

The plant root biomass showed a significant (p < 0.05) increase at 25 mg Ce kg− 1 in the high clay soil, while in the low clay soil, it declined at all Ce doses compared to the control (Fig. 1a). However, the root biomass was significantly reduced at 50 and 100 mg Y kg− 1 in the high clay soil. In the low clay soil, the root biomass also significantly decreased at all Y doses compared to the control (Fig. 1b). Regarding the plant shoot biomass, no significant differences were observed across all Ce treatments in both studied soils (Fig. 1c). In contrast, at 25 mg Y kg− 1, the shoot biomass significantly increased in the high clay soil, while it remained unchanged in the low clay soil at all Y treatments (Fig. 1d).

Fig. 1
Fig. 1
Full size image

Root dry biomass of the crop with Ce treatment (a) and Y treatment (b), and shoot dry biomass of the crop with Ce treatment (c) and Y treatment (d) in the high clay and the low clay soils. Values represent the mean ± standard deviation (n = 4). Different letters are significantly different at 5% level according to least significant difference post hoc test.

In both soils, no significant difference in chlorophyll a + b was observed among all Ce treatments (Fig. 2a). However, concentration of chlorophyll a + b significantly decreased with increasing Y doses in the high clay soil, whereas in the low clay soil, the dose of 50 mg Y kg−1 caused a significant reduction in chlorophyll concentration (Fig. 2b).

Fig. 2
Fig. 2
Full size image

Concentration of chlorophyll a + b in leaves of the crop with Ce treatment (a) and Y treatment (b) in the high clay and low clay soils. Values represent the mean ± standard deviation (n = 4). Different letters are significantly different at 5% level according to least significant difference post hoc test.

Carotenoids showed no significant difference in the high clay soil across Ce treatments, while concentration of carotenoids was significantly reduced at 100 mg Ce kg−1 in the low clay soil (Fig. 3a). Regarding Y treatment, concentration of carotenoids was significantly reduced at 100 mg Y kg−1 in the high clay soil, while a significant decrease was detected at 50 mg Y kg−1 in the low clay soil (Fig. 3b).

Fig. 3
Fig. 3
Full size image

Concentration of carotenoids in leaves of the crop with Ce treatment (a) and Y treatment (b) in the high clay and low clay soils. Values represent the mean ± standard deviation (n = 4). Different letters are significantly different at 5% level according to least significant difference post hoc test.

In the TTC reduction test, the absorbance values significantly decreased in the two soils in all Ce treatments, compared to the control (Fig. 4a). Nevertheless, there was no significant difference among Ce doses of 25, 50, and 100 mg kg−1 in both soils. The addition of Y led to a significant reduction in the absorbance value at all Y doses in both soils (Fig. 4b).

Fig. 4
Fig. 4
Full size image

Absorbance at 545 nm (A545) of TTC reduction test in leaves of the crop with Ce treatment (a) and Y treatment (b) in the high clay and low clay soils. Values represent the mean ± standard deviation (n = 4). Different letters are significantly different at 5% level according to least significant difference post hoc test.

In the high clay soil, proline concentration decreased with increasing Ce dose (Fig. 5a). In the low clay soil, all Ce doses caused a significant decrease of proline. A significant difference was exhibited at 25 mg Y kg−1 in the high clay soil, however, in the low clay soil, significant decreases of proline levels were observed at all Y treatments (Fig. 5b).

Fig. 5
Fig. 5
Full size image

Concentration of proline in leaves of the crop with Ce treatment (a) and Y treatment (b) in the high clay and low clay soils. Values represent the mean ± standard deviation (n = 4). Different letters are significantly different at 5% level according to least significant difference post hoc test.

Concentrations of REEs, Ca, Mg, Fe, and al in the plant

At the dose of 100 mg Ce kg−1, Ce concentration in the roots were significantly (p < 0.05) higher in both studied soils (Fig. 6a). However, no significant differences of Ca concentration were observed in the two soils (Fig. 6b). In the high clay soil, Mg, Fe, and Al concentrations significantly increased in the roots supplied with 25 mg Ce kg−1 (Figs. 6c–e). In contrast, in the low clay soil, the concentrations of these elements exhibited a decreasing trend as Ce doses increased. In the Y treatments, no significant change of Y concentration in the roots was observed in the high clay soil, while it significantly increased at 50 and 100 mg Y kg−1 in the low clay soil (Fig. 6f). In the high clay soil, Ca concentration in the roots was not significantly affected, whereas it significantly increased at 100 mg Y kg−1 in the low clay soil (Fig. 6g). However, Mg concentration was significantly reduced in roots of plants grown with 25 mg Y kg−1 in the both soils (Fig. 6h). Fe and Al concentrations in the roots decreased at 25 mg Y kg−1 in the high clay soil, while both elements showed significant reduction at all Y doses in the low clay soil (Fig. 6i and j).

Fig. 6
Fig. 6
Full size image

Concentrations of Ce (a), Ca (b), Mg (c), Fe (d), and Al (e) in roots of the crop with Ce treatment, and concentrations of Y (f), Ca (g), Mg (h), Fe (i), and Al (j) in roots of the crop with Y treatment in the high clay and low clay soils. Values represent the mean ± standard deviation (n = 4). Different letters are significantly different at 5% level according to least significant difference post hoc test.

With increasing Ce doses, Ce concentration in the shoots showed an increasing trend in both studied soils (Fig. 7a). However, Ca and Mg concentrations were not significantly different with different Ce doses in both soils (Fig. 7b and c). As for Fe and Al, their concentrations in the shoots were consistent in the high clay soil (Fig. 7d and e). However, the concentrations of Fe and Al significantly increased in shoots of plants treated with 100 mg Ce kg−1 in the low clay soil. In Y treatments, Y concentration in the shoots also followed an increasing trend in both soils (Fig. 7f). In the plants grown on soil containing 50 mg Y kg−1, Ca and Mg concentrations were significantly higher in the high clay soil (Fig. 7g and h). In contrast, in the low clay soil, Ca and Mg concentrations did not differ significantly among Y doses. Fe and Al concentrations were not significantly different among Y doses in the high clay soil, whereas in the low clay soil, concentrations of both elements significantly increased at 50 mg Y kg−1 compared to plants grown on the corresponding control soils (Fig. 7i and j).

Fig. 7
Fig. 7
Full size image

Concentrations of Ce (a), Ca (b), Mg (c), Fe (d), and Al (e) in shoots of the crop with Ce treatment, and concentrations of Y (f), Ca (g), Mg (h), Fe (i), and Al (j) in shoots of the crop with Y treatment in the high clay and low clay soils. Values represent the mean ± standard deviation (n = 4). Different letters are significantly different at 5% level according to least significant difference post hoc test.

Elemental distribution in leaves

The distributions of Ce, Y, Ca, Mg, Fe, and Al in leaves in the low clay soil with different Ce and Y treatments were visualized using LA-ICP-MS. In the control treatment, Ce was faintly detected in the leaf, while Ca, Mg, Fe, and Al were abundant and distributed throughout the leaves (Fig. 8a). At 25 mg Ce kg− 1, Ce ion intensity increased, while Ca ion intensity decreased compared to the control (Fig. 8b). In contrast, there were no clear changes in the intensities of Mg, Fe, and Al ions at the dose of 25 mg Ce kg− 1. At 100 mg Ce kg− 1, the distribution of Ce in the leaf appeared as concentrated points (Fig. 8c). Ca and Mg ion intensities remained largely unchanged, whereas the distribution patterns of Fe and Al were similar to that of Ce at 100 mg Ce kg− 1.

Fig. 8
Fig. 8
Full size image

Laser ablation images of ion intensities in first leaves of the crop in the low clay soil with 0 (a), 25 (b), and 100 (c) mg Ce kg− 1 doses. The measured ions are Ce, Ca, Mg, Fe, and Al. Images in the leftmost column are the surface appearance of the leaves.

Low ion intensity of Y was observed in the leaf in the control treatment, while Ca, Mg, Fe, and Al had higher ion intensities and distributed throughout the leaves (Fig. 9a). At 25 mg Y kg− 1, Y was accumulated in specific areas of the leaf, appearing as a light blue color (Fig. 9b). Ca and Mg levels did not show obvious changes compared to the control, while Fe and Al formed concentrated points with similar distribution patterns. At 100 mg Y kg− 1, Y ion intensity further increased and was broadly accumulated throughout the entire leaf (Fig. 9c). Ca and Mg remained unchanged, while Fe and Al ion intensities increased to higher levels.

Fig. 9
Fig. 9
Full size image

Laser ablation images of ion intensities in first leaves of the crop in the low clay soil with 0 (a), 25 (b), and 100 (c) mg Y kg− 1 doses. The measured ions are Y, Ca, Mg, Fe, and Al. Images in the leftmost column are the surface appearance of the leaves.

Discussion

REEs accumulation in the plant

In both studied soils, the concentration of Ce in the soil solution was significantly (p < 0.05) lower than that of Y at a given dose (Table 1). Because the soil solid/liquid partition coefficient (Kd) of Ce is higher than that of Y, Ce tends to be adsorbed by clay colloids, resulting in a lower concentration of Ce in the soil solution35. Regarding soil texture, soils with higher clay content have a greater capacity to retain cations36, leading to lower cation concentrations in the soil solution. Thus, in all Ce and Y treatments, the concentrations of Ce and Y in the soil solution were lower in the high clay soil compared to the low clay soil. This, in turn, reduced the bioavailability of Ce and Y and their concentrations in the roots in the high clay soil, compared to the low clay soil with all treatments (Fig. 6a and f). Additionally, concentrations of Ce and Y were higher in the roots (Fig. 6a and f) than in the shoots (Fig. 7a and f), indicating that the tested crop minimized the translocation of Ce and Y from root to shoot to mitigate their toxicity to photosynthesis in leaves15. With increasing Ce and Y doses, the concentrations of Ca, Mg, Fe, and Al generally increased in the soil solution, especially in the low clay soil (Table 1). Since REEs have similar ionic radii to Ca and Mg and ionic charge to Fe and Al, Ce and Y can compete for the sorption sites in soil with Ca, Mg, Fe, and Al, causing the desorption of these elements into soil solution1.

Shi et al.37 investigated Ce and Y in the shoots of eight vegetables, including Brassica species from Mongolia, Shandong, Guangdong, and Jiangxi in China, and the concentrations of Ce and Y were lower than 0.05 mg kg−1. However, the concentrations of Ce and Y in the shoots in the control were 0.66 and 0.24 mg kg−1 in the high clay soil, and 0.70 and 0.23 mg kg−1 in the low clay soil (Fig. 7a and f). Moreover, their concentrations at the Ce and Y doses of 25, 50, and 100 mg kg−1 were much higher than those in the control, suggesting the tested crop was a REEs accumulator in this study. In all cases of the present study, the concentration of Ce in the shoots (Fig. 7a) was lower than that of Y (Fig. 7f), indicating that Y formed stronger metal-organic complexes than Ce to facilitate long-distance transport and accumulate in shoots, as demonstrated in P. americana by Yuan et al.13. Moreover, such enrichment of HREEs in shoots was also observed in Phytolacca species38, B. nivea39, and Camellia sinensis (L.) O. Kuntze40. To mitigate metal toxicity, trichomes, external structures in plant leaves, serve as a sequestration sink for potentially toxic metals41. For example, Brassica juncea (L.) Czern. accumulated excessive cadmium (Cd) in leaf trichomes to mitigate Cd toxicity42. In this study, at the dose of 100 mg Ce kg−1, the presence of dotted hotspots of Ce in the leaves reflected the accumulation of excessive Ce in trichomes to reduce cellular toxicity (Fig. 8c).

Physiological responses to REEs

The hormetic effect of Ce and Y on root and shoot biomass in this study (Fig. 1a and d) was explained by the fact that low doses of Ce and Y caused cytoplasmic expansion and relaxation, thus leading to increased protoplasmic volume for biomass accumulation43. However, at higher doses of REEs, excessive radicals were induced to damage cellular ultrastructure, thus inhibiting plant growth, as demonstrated in Arabidopis thaliana (L.) Heynh44. and O. sativa45. In the roots, the concentrations of Mg, Fe, and Al exhibited the hormetic effects induced by Ce, where their concentrations significantly (p < 0.05) increased at 25 mg Ce kg− 1 in the high clay soil (Figs. 6c–e). The enhanced uptake of Mg, Fe, and Al might be attributed to the activation of transport proteins on the root cell membrane, which increased the membrane permeability and stimulated the root uptake of these elements in soil46. Moreover, the accumulation of Mg and Fe in the roots also contributed to the significant increase in root biomass at 25 mg Ce kg− 1 in this study (Fig. 1a). Regarding the elemental dynamics in the shoots, the concentration of Ca, Mg, Fe, and Al demonstrated the hormetic effects with Y doses (Figs. 7g–j). Similar to Ce, Y can overexpress the transporter activity for the root absorption of Ca, Mg, Fe, and Al, while Y is easily complexed with ligands such as citrate and transported from root to shoot in the xylem compared with Ce46. Furthermore, Y has been suggested to share the same translocation pathway with Ca, Mg, Fe, and Al due to their similarities in ion radius and ion charge13,47,48, hence leading to the significant increase in the concentration of Ca, Mg, Fe, and Al in the shoots at 50 mg Y kg− 1 in this study (Figs. 7g–j). Moreover, REEs, Fe, and Al were co-localized in the ion mappings of leaves through LA-ICP-MS with Ce treatment (Fig. 8b and c) and Y treatment (Fig. 9b and c), implying that the tested crop utilized a similar detoxification strategy to sequester Ce, Y, Fe, and Al in trichomes49.

Chlorophyll a + b concentration serves as an indicator of photosynthetic activity, and Mg, as the central atom in chlorophylls, also plays a crucial role in photosynthetic efficiency50. In this study, no significant change in concentration of Mg in the shoots was observed in any of the Ce treatments in the studied soils (Fig. 7c). Chlorophyll a + b concentration in the leaves was also consistent in all Ce doses (Fig. 2a), meaning that photosynthetic activity was not affected by Ce. In comparison, increasing doses of Y reduced chlorophyll a + b concentration, indicating that Y was more toxic to photosynthetic pigments than Ce due to the higher binding affinity of Y to organic compounds3. As reported by Maksimovic et al., Y might enter the chloroplast and deform chlorophyll molecules, reducing the chlorophyll a + b concentration in Zea mays L. at Y doses higher than 8.9 mg L− 151. Moreover, the damage to chloroplast structures leads to reduced electron transport efficiency, ribulose 1,5-bisphosphate carboxylase (RuBPcase) activity, and carbon assimilation52.

The TTC reduction test can determine succinate dehydrogenase activity in mitochondria, and succinate dehydrogenase is a key enzyme for the electron transport system, ATP synthesis, and thus respiratory activity53. In this study, the TTC reduction test showed that all Ce and Y doses (25, 50, and 100 mg kg− 1) in both studied soils led to the reduction of absorbance at 545 nm, indicating that even low Ce and Y doses inhibited respiratory activity in the leaves (Fig. 4a and b). The suppression of plant respiration might be attributed to mitochondrial swelling, disrupted membrane potential, and damaged mitochondrial ultrastructure caused by REEs54. Moreover, mitochondria can accumulate a large amount of REEs, serving as a self-defense mechanism of plants against the damage caused by REEs55. Compared with the critical doses for inhibiting photosynthetic pigments (none for Ce and 50 mg kg− 1 for Y), leaf respiratory activity was significantly inhibited at the lowest doses (25 mg kg− 1 of Ce and Y), implying Ce and Y were more toxic to respiratory activity than photosynthesis.

REEs-induced stress is generally expected to stimulate proline production, since proline functions as a scavenger of reactive oxygen species (ROS) generated with REEs exposure, such as La and Ce56. However, in the present study, proline concentration was decreased across all Ce and Y treatments in both the high clay soil and the low clay soil (Fig. 5a and b). This finding is consistent with the result that proline concentration in the roots and shoots of Z. mays significantly decreased at 8.9 and 89 mg Y L− 151. The decrease in proline with increasing Ce and Y doses might result from inhibited proline biosynthesis and lower ROS levels in this study. Since glutamate is the primary precursor and mainly produced by the mitochondria through the conversion of α-ketoglutarate57, the inhibition of mitochondrial respiration induced by Ce and Y in this study could reduce glutamate production (Fig. 4a and b), thus limiting proline synthesis (Fig. 5a and b). Moreover, ROS is generated as a byproduct during energy production by mitochondria58; thus, reduced ROS levels due to inhibited mitochondrial respiration also decrease the synthesis of proline as a ROS scavenger. In this study, we demonstrated that Ce and Y application interfered with nutrient concentrations in the roots and shoots of the tested crop and exerted a greater toxic effect on the respiratory system than on photosynthesis, ultimately affecting proline production.

Conclusions

This study demonstrates that the soil bioavailability and phytotoxicity of Ce and Y in the tested crop are regulated by REEs type and soil texture. The hormetic effects of biomass and nutrient accumulation were induced by both Ce and Y, while Y exerted more pronounced inhibition on photosynthetic pigment levels than Ce. Moreover, the leaf respiratory activity was sensitive to both Ce and Y, especially in soils with low clay content, contributing to reduced proline production. These findings highlighted the importance of considering REEs type and soil texture when assessing the risks of REEs contamination in agricultural regions. Given that this study was conducted in pot experiments, the observed responses might differ from those in field conditions. Future work should integrate field trials, various crops, and different REEs to better understand the risks and sustainable use of these technology-critical elements.