Introduction

Forestry ecosystems play a vital role in global carbon cycling, biodiversity conservation, and sustainable resource management (Miura et al. 2015; Watson et al. 2018; Migliavacca et al. 2021; Taye et al. 2021). Understanding nutrient dynamics within these ecosystems is crucial for effective management and conservation (Silver et al. 1996; Watson et al. 2018; Meier et al. 2019). Among essential nutrients, phosphorus (P) often limits plant growth, particularly in highly weathered and leached soils (Du et al. 2020; Hou et al. 2020), which are prevalent in southern China (Zhang et al. 2021; Yan et al. 2023). In response to P limitation, plants have evolved various adaptive strategies, such as the release of root exudates, especially carboxylates and phosphatases, to enhance P acquisition from P-impoverished soils (Lambers et al. 2015b; Güsewell and Schroth 2017; Lambers 2022). Investigating and quantifying root-exudation patterns in P-limited forestry ecosystems is therefore essential for understanding nutrient cycling processes and optimizing forest management practices. However, due to the challenge of directly accessing and measuring plant-soil nutrient dynamics in forests (Oburger and Jones 2018; Escolà Casas and Matamoros 2021), root exudation patterns across different forest ecosystems are largely unknown.

Leaf manganese (Mn) concentration ([Mn]) has been proposed as a proxy to estimate the extent of root exudation for P acquisition in a number of plant species (Lambers et al. 2015a, 2021; Pang et al. 2018; Zhou et al. 2022; Yu et al. 2023b). Some plants experiencing P limitation increase their root carboxylate exudation and, in turn, enhance Mn uptake from the soil (Shane and Lambers 2005; Lambers et al. 2015a). Plants absorb and transfer soil Mn through a combination of passive and active processes, lacking dedicated transporters for Mn (shared with iron and other elements), and the poor control of Mn uptake makes it possible to indicate its availability in rhizosphere soil (Page and Feller 2005; Millaleo et al. 2010). Manganese is transported to leaf cells through various transporters present in the plasma membrane, and subsequently accumulates in mature leaves. Manganese uptake rates may exceed the plant’s requirement by over 100-fold in some species, resulting in hyperaccumulation (Millaleo et al. 2010; Losfeld et al. 2015). Excessive Mn can be toxic, promoting plants to develop various mechanisms to accumulate Mn without adverse effects. These mechanisms include redistributing and compartmentalizing Mn within different tissues to maintain optimal [Mn] in essential tissues while sequestering excess Mn in less sensitive compartments such as cell walls (oxidized Mn and phenolic compounds) (Wissemeier and Horst 1992). In addition, plants can chelate excess Mn within cells in vacuoles by binding it to organic compounds such as citrate, oxalate, and malate (Memon and Yatazawa 1984; Millaleo et al. 2010). Due to the involvement of Mn in redox reactions and generation of reactive oxygen species (ROS), plants possess antioxidant defense systems to scavenge and neutralize ROS (Sytar et al. 2021; Kosakivska et al. 2021; Zhao et al. 2022). These findings suggest that leaf [Mn] has the potential to serve as a reliable indicator of rhizosphere carboxylate concentrations and P-acquisition strategies across various ecosystems (Lambers et al. 2015a; Huang et al. 2017).

Low soil P availability promotes leaf Mn accumulation in plants, and is often observed in P-deficient soils where plants adjust their nutrient-uptake and -allocation strategies to cope with the limited P availability (Millaleo et al. 2010; Lambers et al. 2015a, 2021). One of the mechanisms involves the exudation of carboxylates and phosphatases from their roots, enhancing the solubilization and mobilization of soil P. However, these carboxylates also influence the availability of other nutrients, including Mn (Shane and Lambers 2005). Under P-deficient conditions, increased secretion of carboxylates enhances the desorption of Mn from soil particles, making it more accessible to plant roots (Kochian et al. 2004). Subsequently, root uptake and shoot transport result in elevated [Mn] within mature leaves (Page and Feller 2005; Page et al. 2006; Losfeld et al. 2015).

There is a lack of comprehensive studies examining the correlation between root exudation, leaf [Mn], and P limitation in forestry ecosystems of southern China. Very few studies have explored the relationship between root exudates and P availability among different plant species and ecosystems in south China. For example, research conducted in subtropical rainforests has shown that root exudation is a major root functional trait (Sun et al. 2021), and root exudation is largely affected by soil P availability and mycorrhizal type (Jiang et al. 2022). However, how prevalent the release of root exudates is among different species in P-limited forest ecosystems remains unexplored (Wang and Lambers 2020). Are species in any plant family consistently exhibiting high levels of root exudation in both P-rich and P-limited environments? Does root carboxylate exudation demonstrate specific trends among plant life forms? Responses to these questions are essential to bridge the knowledge gap and provide valuable insights into the nutrient dynamics of significant forest ecosystems.

In the context of forestry ecosystems in southern China, characterized by acidic soils with low P availability, investigating the correlation between root exudation, leaf [Mn], and P limitation holds particular promise. This region also encompasses vast areas of plantation forests dominated by fast-growing tree species widely utilized for timber production and ecological restoration (Zhou et al. 2017; Sun et al. 2021; Yan et al. 2023). Understanding the mechanisms underlying nutrient acquisition and utilization in these ecosystems is crucial for sustainable forest management and ecosystem functioning. Additionally, using leaf [Mn] as a proxy can allow us to assess whether plants have the capacity to be hyperaccumulators and remove excessive soil Mn for soil restoration (Li et al. 2007; Liu et al. 2020; Wildová et al. 2021).

This study aimed to investigate the potential of using leaf [Mn] to proxy root exudation in different P-limited forestry ecosystems in southern China as well as in P-rich systems in northern China. Based on previous measurements (Lai et al. 2024), we assumed that the forest ecosystems in southern China likely experience P limitation due to their acidic soils and significant erosion. We hypothesized: 1) The leaf [Mn] would be higher in the forest ecosystems of southern China (low-P soils) than those in northern China (relatively P-rich soils). Therefore, we expected plants in the southern forest ecosystems to exhibit faster root carboxylate-exudation rates, leading to increased Mn uptake and consequently higher leaf [Mn] than those in the northern forest ecosystems. According to previous studies showing that root exudation is widespread across different plant families and plant forms, we further hypothesized: 2) high leaf [Mn] in P-limited southern China would be more common compared to that in the relatively P-rich ecosystems in northern China.

To test the above hypotheses, we assessed multiple forest stands across different soil types with different levels of P availability. Leaf and soil samples were collected from selected stands encompassing a range of plant families and life forms, and leaf [Mn] was analyzed to proxy root carboxylate-exudation patterns. Furthermore, we quantified soil P availability and characterized soil properties to assess their impact on the dynamics of root exudation.

Methods and materials

Study sites and species selection

The study was conducted in 2020, assessing multiple forest stands across different soil types with different P availability in two distinct natural forest ecosystems (Fig. 1; this map was made using R package “ggmap”; the source is “stadia” with a type “stamen_terrain”). One ecosystem is located in the Maijishan Region (34°20′N, 106°01′E) of Gansu Province, northern China. Known for its rich biodiversity, the Maijishan ecosystem has a total of 2371 species of vascular plants, representing 218 families and 962 genera (Lu 2006). Located within the Qinling Mountains, it spans altitudes ranging from 1200 to 1600 m above sea level. It experiences a continental climate characterized by cool, humid, and semi-humid zones, with an average annual rainfall of 700 mm, primarily occurring between July and September. The mean temperature is 9 °C with a relative air humidity of approximately 73% (Yang et al. 2023). The dominant soils in this region are mountain brown loams (Tudi et al. 2022).

Fig. 1
figure 1

Map of the study sites in the relatively phosphorus-rich Maijishan region, Gansu, China and in the phosphorus-limited Damingshan region, Guangxi, China

Full size image

Another ecosystem studied is situated in the Damingshan Region (23°24′N, 108°31′E), within the south-central Guangxi Zhuang Autonomous Region of southern China. Due to its unique geographic location, Damingshan exhibits rich biodiversity with a total of 2374 vascular plant species belonging to 918 genera across 234 families (Wen et al. 1998). This ecosystem lies between the northern tropics and southern subtropics at an average altitude of 1200 m, experiencing a southern subtropical monsoon climate. It receives an average precipitation of 2630 mm and an average temperature of 15.1℃ (Zhu et al. 2011).

According to the literature (Wen et al. 1998; Suo et al. 2008; Zhu et al. 2011; Yang et al. 2023) and recommendations from local botanists, we established a 200 m × 200 m study site in each ecosystem in 2020. Each site was subdivided into eight to 10 subsites based on the landscape for the purpose of investigating and selecting plant species. Taking into consideration factors such as abundance, life form, and endemism, a total of 188 species were collected from the two ecosystems: 38 plant families, 81 species from relatively P-rich soils of Gansu; 56 plant families, 107 species from low-P soils of Guangxi (Supplementary Table 1).

Reference species selection

The leaf [Mn] can serve as an indicator of root carboxylate release, implying that higher concentrations suggest plants release greater amount of carboxylates that mobilize Mn and enhance its uptake from the soil. To assess the relationship between different leaf [Mn] levels and the extent of root carboxylate release, a “negative reference” is necessary (Lambers et al. 2021; Lambers 2023). Therefore, we selected the family of Dryopteridaceae (ferns) species as the negative reference for several reasons: 1) Dryopteridaceae are adapted to efficiently utilize limited Mn in their habitats (Grosjean et al. 2019; Schmitt et al. 2017) which may not provide sufficient amounts of Mn for these ferns to accumulate high concentrations in their leaves (Cornara et al. 2007); these adaptations often involve conservative strategies for Mn uptake and allocation within the plant, resulting in lower leaf [Mn] (Bai et al. 2020; Zhou et al. 2022); 2) Dryopteridaceae typically exhibit slow growth rates compared with many vascular plant species (Pinson et al. 2017; Rünk and Zobel 2007), leading to reduced demand for Mn and consequently lower concentrations in their leaves (Reimann et al. 2007); 3) Dryopteridaceae occurred in both studied ecosystems. It is worth noting that leaf [Mn] in different fern families may vary due to their specific growth microhabitats, suggesting a discernible difference in leaf [Mn] among different fern families. In this study, when referring to “high” leaf [Mn], it indicates a comparison with the negative reference group of Dryopteridaceae.

Leaf and soil sampling

All samples were collected between August and October 2020. Five plants per species (except for a few species, which included three to four plants) were selected, and mature leaves were utilized for chemical analyses. The leaves were placed in envelopes for subsequent determination of dry weight (at 60℃ for seven days). Bulk soil samples were obtained by combining soil from three points within each plot of every subsite. All soil samples were collected in plastic bags, air-dried for one week, sieved to a particle size of 2 mm, and then ground.

Leaf and soil analyses

Aliquots of 80 mg sample were digested using concentrated HNO3 and HClO4 (v/v 3:1) and diluted to 10 ml with Milli-Q water, and leaf [Mn] was analyzed by inductively coupled plasma optical emission spectroscopy (ICP-OES) (Perkin Elmer, PE Avio 500, Shelton, CT, USA). Leaf P concentration ([P]) was analyzed using the molybdenum blue-based method (Ames 1966) with a microplate reader (Thermo Fisher Scientific, Varioskan Flash, Finland).

Aliquots of 0.5 g soil sample were weighed in a digestion tube and digested using concentrated H2SO4 and HClO4 at 360℃ for one hour. The total nitrogen concentration ([N]) and total [P] were then determined using an automatic chemical analyzer (AMS, Smartchem 450, Rome, Italy) and the molybdenum blue-based method (Olsen and Sommers 1982), respectively. Soil nitrate and ammonium were extracted with 2 M KCl and measured using a continuous segmented flow autoanalyzer (AA3, SEAL Analytical, Norderstedt, Germany). Anion exchange membranes (AEMs, VWR Chemical, Leuven, Belgium) were used to measure soil resin P (‘plant-available P’) (Bentley et al. 1999). Soil pH (water based) was measured using a pH meter.

Statistical analyses

All statistical analyses were performed using the R software platform (version 4.0.2, R Core Team 2020). Shapiro-Wilk normality was tested using ‘rstatix’ (version 0.7.2, Kassambara 2023), and homogeneity was tested using ‘stats’ (R Core Team 2020). Considering the varying sample sizes, such as three to five replicates for each species and one to 10 species from different families, the utilization of Welch t-test is more suitable for this study (Stewart-Oaten et al 1992; Ruxton 2006). The Welch t-test was used to compare the difference of mean concentrations between the “negative reference” and target families, and the difference between two locations using ‘stats’ (R Core Team 2020). All plant species from two locations were compared for significance (Figs. 2, 3 and 5), excluding the top 50% of species with the highest leaf [Mn], as depicted in Fig. 4. Appropriate generalised least square models (GLS) were selected based on the lowest Akaike's Information Criterion corrected (AICc) values and the Tukey's HSD post hoc test (Pinheiro & Bates 2000), to compare the significance among life forms at the same location (p < 0.05) in Fig. 5.

Fig. 2
figure 2

Leaf phosphorus (P) (a) and manganese (Mn) (b) concentration in different locations of relatively P-rich soils in Gansu and low-P soils in Guangxi. Each point indicates an individual plant, with a total of 327 and 395 individuals in Gansu and Guangxi, respectively. Asterisks **** indicate significant difference between values in Gansu and Guangxi using the Welch t-test, p < 0.0001

Full size image
Fig. 3
figure 3

Leaf manganese (Mn) concentration of 38 families in relatively P-rich soils of Gansu (a) and 56 families in low-P soils of Guangxi (b). Asterisks * indicates significant difference compared with the negative reference (Dryopteridaceae, marked as R) using a Welch t-test, p < 0.05

Full size image
Fig. 4
figure 4

Top 50% plant species with the highest leaf manganese (Mn) concentration in relatively P-rich soils of Gansu (a) and low-P soils of Guangxi (b)

Full size image
Fig. 5
figure 5

Leaf manganese (Mn) concentration of different plant life forms in relatively P-rich soils Gansu (a) and low-P soils of Guangxi (b). Different letters indicate significant difference among life forms within the same location using appropriate generalized least square models based Tukey's HSD post hoc test, p < 0.05

Full size image

Results

Soil characteristic in different ecosystems

The soil in Guangxi exhibited significantly lower pH than that in Gansu (Table 1), accompanied by lower concentrations of total P and resin P (less than 20% of the levels observed in Gansu). Conversely, the soil total N concentration was higher in Guangxi, although not statistically significant; the concentration of NH4+-N in Gansu was 70% higher than that in Guangxi (p < 0.05), whereas the soil in Guangxi exhibited a higher level of NO3−-N (p < 0.05). Although we did not measure soil [Mn], literature data indicates a 1.9-fold higher level of soil [Mn] in Gansu compared with that in Guangxi (Wang et al. 2012, 2016).

Table 1 Bulk soil characteristics of Maijishan in Gansu and Damingshan in Guangxi
Full size table

Leaf P and Mn concentration in different ecosystems, families and species

In the ecosystem of low-P Guangxi, plants exhibited 79% lower leaf [P] (0.9 mg g−1, p < 0.0001, Fig. 2a), but 51% higher leaf [Mn] (230 mg kg−1; p < 0.0001, Fig. 2b) compared with those in the ecosystem of relatively P-rich soils of Gansu (1.6 mg g−1 leaf [P], and 113 mg kg−1 leaf [Mn]; Fig. 2).

We examined 38 and 56 families in relatively P-rich soils of Maijishan (Gansu) and low-P soils of Damingshan (Guangxi), respectively (Fig. 3). Approximately half of the families from relatively P-rich soils (16 out of 38) exhibited significantly higher leaf [Mn] than the reference group, while a substantial majority of the families from low-P soils (46 out of 56) showed significantly elevated leaf [Mn] relative to the negative reference. Leaf [Mn] was generally higher in low-P soils than in relatively P-rich soils (Fig. 3). Notably, in low-P soils, the families Altingiaceae, Verbenaceae, Clethraceae, Chloranthaceae, Araliaceae, Melastomataceae, and Selaginellaceae all displayed average leaf [Mn] values exceeding 500 mg kg−1, while in relatively P-rich soils, only the Fagaceae family exhibited an average leaf [Mn] above this threshold. When compared with the negative reference using a Welch t-test, the p-values were marginal for the Sapindaceae and Lauraceae families (0.052 and 0.051, respectively) in relatively P-rich soils, and for the Anacardiaceae, Juglandaceae, and Lindsaeeacea families (0.051, 0.056, and 0.05, respectively) in low-P soils.

Plant species with the highest leaf [Mn] in two ecosystems were presented in Fig. 4. In Maijishan, which had a higher soil P concentration, the top 10 species with the highest leaf [Mn] were Quercus spinosa (881 mg kg−1, Fagaceae), Taxus wallichiana var. chinensis (702 mg kg−1, Taxaceae), Quercus mongolica (516 mg kg−1, Fagaceae), Elaeagnus umbellate (442 mg kg−1, Elaeagnaceae), Symplocos tanakana (344 mg kg−1, Symplocaceae), Quercus aliena var. acutiserrata (310 mg kg−1, Fagaceae), Schisandra chinensis (303 mg kg−1, Schisandraceae), Deutzia grandiflora (247 mg kg−1, Hydrangeaceae), Litsea pungens (236 mg kg−1, Lauraceae) and Elaeagnus pungens (178 mg kg−1, Elaeagnaceae). In Damingshan where soil P concentration was lower than that in Majishan, the top 10 species with the highest leaf [Mn] were Clerodendrum kwangtungense (1506 mg kg−1, Verbenaceae), Altingia chinensis (1108 mg kg−1, Altingiaceae), Dendropanax dentiger (897 mg kg−1, Araliaceae), Selaginella doederleinii (852 mg kg−1, Selaginellaceae), Clethra delavayi (812 mg kg−1, Clethraceae), Blastus pauciflorus (756 mg kg−1, Melastomataceae), Betula austrosinensis (629 mg kg−1, Betulaceae), Elaeocarpus decipiens (628 mg kg−1, Elaeocarpaceae), Sarcandra glabra (625 mg kg−1, Chloranthaceae) and Lophatherum gracile (582 mg kg−1, Poaceae).

Compared with the top 10 species with the highest leaf [Mn] in relatively P-rich soils of Gansu (with an average of 416 mg Mn kg−1) and low-P soils of Guangxi (840 mg Mn kg−1), both the maximum and minimum leaf [Mn] in low-P soils were significantly higher than those observed on relatively P-rich soils, respectively. Amongst those top 10 species with highest leaf [Mn], Fagaceae (three species) and Elaeagnaceae (two species) were families most represented in relatively P-rich soils; in contrast, the top 10 highest leaf [Mn] species belonged to 10 different families in low-P soils.

In low-P soils of Guangxi, within the Verbenaceae family, two species demonstrated significant differences in leaf [Mn]: Callicarpa bodinieri exhibited 170 mg Mn kg−1, while Clerodendrum kwangtungense showed a value of 1500 mg Mn kg−1. Similarly, among Araliaceae, three species displayed high distinct levels of leaf [Mn], namely Dendropanax dentiger (894 mg Mn kg −1), Aralia armata (410 mg Mn kg−1), and Heptapleurum delavayi (454 mg Mn kg−1). In the Altingiaceae family, only one species was included: Altingia chinensis, and all five replicates exhibited high leaf [Mn], although there was considerable variation, ranging from 826 to 1813 mg Mn kg−1. Among the seven Poaceae there was considerable variation; Lophatherum gracile and Chimonobambusa damingshanensis had a leaf [Mn] of 582 mg Mn kg−1and 360 mg Mn kg−1, respectively. In contrast, Miscanthus floridulus and Setaria palmifolia exhibited low leaf [Mn], 12 mg Mn kg−1 and 24 mg Mn kg−1, respectively. However, all five Poaceae in relatively P-rich soils of Gansu exhibited relatively low leaf [Mn], with an average concentration of 38 ± 7 (standard error of the average) mg Mn kg−1.

Leaf Mn concentration as dependent on life form

When plant species were classified into different life forms (Fig. 5), in relatively P-rich soils of Gansu, trees exhibited the highest leaf [Mn], while herbs and ferns presented the lowest leaf [Mn]; in contrast, in low-P soils of Guangxi, shrubs demonstrated the highest leaf [Mn], whereas lianas had the lowest leaf [Mn]. Notably, relatively P-rich soils of Gansu showed more pronounced variation among plant life forms compared with relatively smaller differences in low-P soils of Guangxi.

Discussion

The present findings provide robust support for our hypotheses that leaf [Mn] in P-limited forests of southern China is higher than that in the forests of northern China with higher soil P, despite Gansu having a higher soil [Mn]. Additionally, we observed a more widespread distribution of high leaf [Mn] compared with the negative reference across various plant families in low-P soils than in relatively P-rich soils. This suggests a root carboxylate-release strategy commonly exhibited by plants in response to P limitation in Guangxi forests.

Soil P limitation promotes leaf Mn accumulation

Although soil [Mn] was higher in relatively P-rich soils of Gansu than in low-P soils of Guangxi (Table 1), Mn availability can vary. Soils that are more acidic generally exhibit higher Mn availability, while alkaline or highly weathered soils may have a very low Mn availability (Sims 1986). Additionally, a high availability of soil P can reduce the uptake and accumulation of Mn in leaves (Pedas et al. 2011). This is attributed to high levels of soil P that can induce alterations in soil pH and affect the availability of elements, including Mn. Furthermore, it is well-documented that P and Mn often exhibit antagonistic interactions within the soil environment, leading to mutual inhibition of their respective uptake (Barben et al. 2010). To mitigate the confounding effect of soil pH in different locations, the negative control is recommended to be used to standardize leaf [Mn] in response to soil P (Lambers et al. 2021). Leaf [Mn] was significantly higher (Fig. 2) across various plant families in P-limited Guangxi (Fig. 3), indicating enhanced mobilization and absorption of soil Mn by plants. The soil total P and resin P concentration both indicated a more severe P-deficiency in Guangxi than in Gansu (Table 1) which is consistent with previous studies (Zhang et al. 2005; Li et al. 2015; Liu et al. 2022). The combination of a lower soil pH, karst land, and higher precipitation in Guangxi exacerbates the severity of P deficiency when compared with that in relatively P-rich soils of Gansu (Zhang et al. 2021; Li et al. 2023).

Mycorrhizal fungi play a pivotal role in enhancing nutrient acquisition in plants through symbiotic interactions. However, the soil microbial diversity in low-P soils of Guangxi forests is comparatively lower than that at other locations (Zhao et al. 2020). Furthermore, the richness of arbuscular mycorrhizal fungi (AMF) exhibits a significant positive correlation with plant-available P and soil pH (Xiao et al. 2019). This symbiotic association and nutrient-uptake ability can be suppressed under severe P limitation (Abbott et al. 1984; Bolan et al. 1984; Treseder and Allen 2002; Albornoz et al. 2017), despite mycorrhizal symbiosis still playing a vital role in pathogen defense (Guillemin et al. 1994; Branzanti et al. 1999; Gille et al. 2024). In such scenarios, plants may rely on alternative strategies, such as root exudates, for P acquisition. This would partially explain the higher leaf [Mn] widely observed across various plant families in low-P soils of Guangxi.

A high leaf Mn concentration in mature leaves is a proxy for significant root exudation

Manganese accumulation in plants requires more study, as some species, such as Banksia (Proteaceae) and some Eucalyptus species (Myrtaceae) show stronger Mn enrichment characteristics than others in P-deficient conditions (Shane and Lambers 2005; Lambers et al. 2021; Zhou et al. 2022). Several families, including Fagaceae, Symplocaceae, Smilacaceae, Araliaceae, Pinaceae, Cupressaceae exhibited higher leaf [Mn] than the negative references in both relatively P-rich soils and low-P soils. Moreover, many of these species are known to release carboxylates from their roots (Table 2). For instance, Quercus species (Fagaceae) in temperate and tropical forests release significant amounts of oxalate from their fine roots (Sun et al. 2017; Wang et al. 2021; Nottingham et al. 2022); while Panax species (Araliaceae) release glycolate, propionate, glycerate and glutarate (Luo et al. 2022). Similarly, Picea species (Pinaceae) exude root carboxylates like oxalate and lactate into the rhizosphere (Sandnes et al. 2005). Monocarboxylates like lactate and glycolate do not chelate Mn, but they may affect Mn availability in soil when associated with a decrease in rhizosphere pH (Lambers et al. 2021).

Table 2 Plants with relatively high leaf manganese concentration in this study reported to release root carboxylates or other exudates that change the soil nutrient dynamics
Full size table

Some plant families occurred at both locations but only showed significantly higher leaf [Mn] than the negative reference at the low-P location, such as Lauraceae, Asteraceae, Sapindaceae, Polypodiaceae, Fabaceae, Thelypteridaceae, Poaceae and Dennstaedtiaceae. Though previous studies found fern families in general have low leaf [Mn] (Grosjean et al. 2019; Schmitt et al. 2017), the present study observed that certain fern families (i.e. Thelypteridaceae and Dennstaedtiaceae) also responded to the soil P levels with the values at low-P soils of Guangxi being significantly higher than in relatively P-rich soils of Gansu (Fig. 5). This may be attributed to specific growth conditions, although such occurrences are not widespread (Reimann et al. 2007), which requires further investigation. The observation of high leaf [Mn] in some species aligns with the previous reports on the exudation of carboxylates from roots. For example, species belonging to Fabaceae and Poaceae are found to release root carboxylates such as citrate, oxalate and malate (Table 2). Additionally, species from Lauraceae release citrate (Aoki et al. 2012) and other root exudates (Sun et al. 2021); Asteraceae species release oxalate (Olivares et al. 2002). Interestingly, Cyperaceae showed higher leaf [Mn] than the negative reference in relatively P-rich soils of Gansu, but not in P-limited Guangxi (p = 0.07); many species of Cyperaceae produce dauciform roots and release root carboxylates and phosphatases (Playsted et al. 2006; Shane et al. 2006). The variation among Cyperaceae in P-limited Guangxi might be accounted for by cations other than protons accompanying the release of carboxylates (Roelofs et al. 2001; Zhu et al. 2005). Exudation of cations like K+ and Na+, rather than H+, contributes to maintaining the charge balance during the release of carboxylate anions. However, this affects the rhizosphere pH, and therefore the Mn availability, and subsequent leaf Mn accumulation (Fig. 3b) (Lambers et al. 2021). However, further investigation is required to fully comprehend these effects.

We observed significant variation in leaf [Mn] among certain families, such as Fagaceae, Taxaceae and Elaeagnaceae in relatively P-rich soils of Gansu. Moreover, a greater number of plant families exhibited significant variation in P-limited Guangxi, including Altingiaceae, Verbenaceae, Araliaceae, Melastomataceae, Selaginellaceae, Betulaceae, Zingiberaceae, Elaeocarpaceae, Polypodiaceae and Poaceae. These findings suggest that the capacity to accumulate leaf [Mn] varies within plant families. Additionally, some plants do not respond to low soil P by releasing root exudates and consequently maintain low leaf [Mn] (Lambers et al 2022).

Enhancing the perspectives on leaf manganese and root exudate research demands meticulous attention and focus

Investigating leaf [Mn] and root exudation of carboxylates is crucial for comprehending plant nutrient acquisition, rhizosphere dynamics, and plant-microbe interactions. Given the importance of further leaf Mn studies, it is important to note a number of factors affecting soil Mn availability and plant absorption and accumulation. It requires careful selection of the study sites. Climate conditions, soil moisture, and drainage can influence the availability and uptake of Mn by plants. Waterlogged or poorly-drained soils can enhance both Fe and Mn availability; however, plants down-regulate Fe uptake due to the tightly controlled Fe-acquisition mechanisms (Lambers et al. 2021), which consequently reduces Mn acquisition as Mn and Fe share the same transporter (Baxter et al. 2008). Conversely, dry soils may restrict root uptake and result in reduced leaf Mn accumulation. Soil characteristics such as pH, organic matter content, redox conditions, and [Mn] impact the availability of Mn for plant uptake.

Different plant species exhibit varying capacities for leaf [Mn]. Some plant species possess inherent adaptations for higher levels of [Mn], such as Schima superba (Theaceae), a Mn hyperaccumulator that can reach 10,000 mg Mn kg−1 in leaf in a pot experiment (Yang et al. 2008), Larix decidua (Pinaceae), Betula pendula (Betulaceae) and Vaccinium myrtillus (Ericaceae) accumulate 7,000 to 10,000 mg Mn kg−1 in leaf when growing in air-polluted mountain areas (Wildová et al. 2021). Certain plants form specialized roots such as cluster root in Proteaceae and many actinorhizal species and dauciform root in some Cyperaceae which enable them to release root carboxylates in an exudative burst thus efficiently mobilizing soil P to cope with P-impoverished environments in Western Australia, South Africa and South America (Neumann and Martinoia 2002; Shane et al. 2006; Lambers et al. 2015b). Additionally, an increasing number of species without specialized root systems have been reported to release root exudates for soil P acquisition (Table 2); however, it remains largely unknown how widespread this strategy is, globally.

Ericaceae exhibit a symbiosis with ericoid mycorrhizal fungi which is expected to enhance their P-acquisition (Smith et al. 2015). Additionally, the mycorrhizal hyphae can intercept mobilized Mn and prevent its accumulation in leaves (Hashem 1995). Interestingly, we collected 10 Ericaceae species in P-limited Guangxi, eight of them exhibiting relatively high leaf [Mn] ranging from 200 to 360 mg Mn kg−1 compared with the negative references. Similarly, other Ericaceae species have also been observed to accumulate high leaf Mn, such as Vaccinium myrtillus (Wildová et al. 2021), and Leucopogon verticillatus (Zhou et al. 2022). Notably, previous research has shown that Vaccinium species possess the ability to release root carboxylates like oxalate and citrate (Millaleo et al. 2020). Therefore, our findings suggest that symbiotic fungi did not intercept Mn and that they did not occupy a dominant position in the low-P soils, indicating that the symbiosis was possibly ineffective in facilitating plant P uptake. Instead, we propose that these Ericaceae species in Guangxi, which are limited by soil P availability, instead, rely on root carboxylate exudation for their P acquisition, similar to how some Eucalyptus plants that associate with AM and ECM fungi release root carboxylates to acquire P in severely P-depleted habitats in southwestern Australia (Zhou et al. 2022). These findings suggest that further research on leaf Mn dynamics along with investigations on mycorrhizal symbioses and root exudation is warranted, globally.

By integrating field observations (Lambers et al. 2021; Zhou et al. 2022), controlled-environment experiments (Pang et al. 2018; Yu et al. 2023b), and molecular analyses, we can gain deeper insights into the impact of environmental factors on leaf [Mn] and its correlation with root exudates. Long-term field studies offer valuable perspectives on the temporal and spatial patterns associated with these processes (Wildová et al. 2021). Molecular techniques such as transcriptomics, proteomics, and metabolomics can unveil the genetic and molecular mechanisms underlying leaf [Mn] and root carboxylate exudation (Sharma and Jha 2023; Yu et al. 2023a). Implementing these strategies will advance our understanding of leaf [Mn] and root carboxylate exudation to acquire deeper insights into plant nutrient-acquisition strategies and their ecological significance.

Conclusion

This study investigated leaf [Mn] in two forest ecosystems characterized by contrasting soil P availability. We observed that plants growing in low-P soils of Guangxi exhibited significantly higher average leaf [Mn] compared with those in relatively P-rich soils of Gansu, despite the fact that Gansu had higher soil [Mn]. By utilizing the same plant family, i.e. Dryopteridaceae, as a negative reference, the results indicated a greater number of species with high leaf [Mn] in low-P soils than in relatively P-rich soils. This trend was further confirmed across different plant families and life forms, suggesting that high leaf [Mn] is more prevalent in low-P soils. Furthermore, we synthesized literature data to demonstrate that a substantial proportion of plants with high leaf [Mn] also release significant amounts of root carboxylates, thereby establishing mature leaf [Mn] as a reliable proxy for root carboxylate release under field conditions. The utilization of root carboxylates for soil P acquisition represents a common strategy exhibited by plants inhabiting P-limited environments.