Analogous response of interspecies NuREs to interannual climate variability
In contrast to deciduous plants on a global scale (Vergutz et al. 2012), the three oak species under study demonstrated higher N, P, and K resorption efficiencies, but lower C and Mg resorption efficiencies. The NuRE of N and P in this study was higher than that of deciduous plants in a subtropical karst region (Liu et al. 2014), and higher than that of Q. variabilis across China (Sun et al. 2015). This signified that plant growth in this region was more restricted by N and P. This may be because the plants grew at high altitudes, where due to the low temperature, they required higher levels of nutrient resorption to store nutrients for the following growing season. This was demonstrated by Du et al. (2017), who found that the resorption of N, P, and S, at least in Q. variabilis, was the greatest at high altitudes. Moreover, PRE in the three oaks was greater than NRE (except for 2018). This was in alignment with the research of Chai et al. (2015) and Du et al. (2017), who revealed that in the same or neighboring regions, the growth of oaks was commonly restrained by P compared with N.
As a conservation mechanism, nutrient resorption plays a key role in balancing plant nutrient demands and acquisition (Brant and Chen 2015; Killingbeck 2004). Our results revealed that NuRE varied significantly among the different years but did not differ between species. Interestingly, the three oak species displayed analogous NuRE responses to interannual climate variability, particularly for C, N, P, S, Mg, Fe, Mn, and Cu (Fig. 1). The similarity in NuRE between different plants indicated that the same plant types (only deciduous plants were investigated in the study) growing at the same sites might have parallel or convergent mechanisms in terms of nutrient conservation (Yeaman et al. 2016). This aligned well with a study by Bahamonde et al. (2019), who found that the NuRE did not vary between Nothofagus pumilio and N. Antarctica at the sites they investigated. Similar results were reported from a five year manipulative field study by Prieto and Querejeta (2020), and demonstrated by Lü et al. (2013), who found that with N inputs, plants exhibited a convergent response in the resportion of N and P. These results indicated that plants growing under the same conditions, or occupying the same niche, may experience analogous limitations for nutrient absorption and conservation processes (Bahamonde et al. 2019). Thus, our findings may have important implications for the study of localized plant adaptationsand evolution.
Effects of interannual climate, and leaf and soil nutrient status on NuRE
Significant fluctuations in nutrient resorption due to interannual variabilities in temperature and precipitation were reported in two prior studies (May and Killingbeck 1992; Killingbeck 1993). Which elements are more sensitive to interannual climate variability? Of the 14 elements examined, the NuRE of N, P, S, Ca, and Mg was closely associated with MAT and MAP (Figs. 3, 4). The results suggested that these elements were more sensitive to interannual climate variability compared to others, which due to their importance, translated to limitations in photosynthesis and growth in plants (Elser et al. 2000; Marschner and Marschner 2012). This finding signified that responses of resorption to climatic variability were contingent on element type (Vergutz et al. 2012; Sun et al. 2015; Du et al. 2017), where some elements directly related to growth (such as nucleic acid-proteins and photosynthetic elements) may be more sensitive to climate change.
The NRE decreased with higher temperatures across all species, which has been demonstrated in numerous previous studies (Yuan and Chen 2009; Sun et al. 2015; Prieto and Querejeta 2020). This is mainly because colder temperatures not only inhibited the metabolic activity of plants, but also restricted nutrient release and root N uptake. Therefore, plants need to extract additional N to enhance metabolic activity and growth rates, while reducing their dependence on available soil nutrients. This result also signified the interesting finding that temperature had a consistent influence on NRE regardless of spatial or temporal scales. However, precipitation had more complex influence on nutrient resorption. For instance, NRE and PRE initially decreased and then increased with MAP (Fig. 4). Comparatively, Yuan and Chen (2009) found that P resorption increased with higher precipitation, while Vergutz et al. (2012) reported that N and P resorption decreased with precipitation on a global scale. Moreover, Du et al. (2017) found that N and P resorption was enhanced with precipitation along an altitude gradient of the same area. Our study indicated that excessive or low precipitation may restrict the growth of plants, which resulted in high N and P resorption. The results indicated that precipitation may play a predominant role in the determination of nutrient demands and plant growth in local ecosystems.
In the drought year (2014), CRE was significantly enhanced in the plants (Fig. 1), which suggested that plants may store additional energy for reproduction under drought conditions. Correspondingly, the concentrations of C in the litter of the forest floor were reduced, as was demonstrated by Sardans et al. (2008). This could influence litter quality and decomposition, and even ecosystem C cycling downstream. Based on the predicted changes in local temperature and precipitation (0.9oC higher annual mean temperature and 6.4 mm lower total precipitation, see Zhao et al. (2012)), the resorption efficiency of C, N, and P will likely, to some extent, decrease over the next twenty years (Table S6), as demonstrated by Prieto and Querejeta (2020). This decrease likely reduced nutrient demands and plant growth and may have some influence on litter quality and decomposition (Prescott 2010; Prieto et al. 2019).
NuRE was also associated with the elemental concentrations of fresh leaves. Surprisingly, there were positive relationships between leaf element concentrations and NuREs (except for P) (Fig. S3). Our results were unlike most previous studies, in which the high nutrient content of fresh leaves typically led to low nutrient resorption (Kobe et al. 2005; Ratnam et al. 2008; Vergutz et al. 2012), or there was no significant relationship between them (Aerts 1996; Chapin and Moilanen 1991). For our study, the relationships between the elemental concentrations of fresh leaves and climatic factors were also analyzed, which revealed that they were quite consistent with that which existed between NuRE and climatic factors. This revealed that both the elemental concentrations of fresh leaves and the resorption of nutrients were impacted by climatic factors, which may form coupling relationships. When plants possessed high nutrient concentrations during the growing season, increased nutrient resorption was likely to occur at the end of the growing season to maintain a source-sink balance. Furthermore, we found that soil had no influence on NuRE between years (Table S2), which agreed with the findings of Aerts (1996) and Aerts and Chapin (2000). This was likely caused by the low fluctuation of available soil nutrients between years and low sensitivity to changes in the soil due to long-term adaptation.
Regulation pattern of nutrient resorption of multiple elements
The elemental CVs between different years generally decreased with higher NuREs in all three oak species (Fig. 6). This was because high concentrations with low variability benefitted plants by maintaining their homeostasis, which was in alignment with the stability of limiting elements hypothesis (Han et al. 2011). Meanwhile, this was also supported by the study of Karimi and Folt (2006), who found a strong regulation of macronutrients, and weak regulation of micronutrients and non-essential elements in organisms. The exception was Al that appeared to have both low variation and resorption, which was likely due to the avoidance of phytotoxic effects by plants (Brunner and Sperisen 2013). These results suggested that nutrient resorption, akin to leaf stoichiometry, had a similar regulation pattern in homeostasis, i.e. higher nutrient concentrations or resorption efficiencies, with lower variability in plants.
Our results exhibited that nucleic acid-protein and photosynthesis-enzyme elements had higher resorption efficiency comparted to structural and toxic elements (Figs. 1, 5). This is because nucleic acid-protein elements are generally limited in supply (Elser et al., 2007; Du et al. 2020), and therefore have a high resorption efficiency. Further, although photosynthesis-enzyme elements are required in low quantities, they are essential for new growth (Marschner and Marschner 2012) and may therefore, to some extent, re-absorbed (Son et al. 2000). The negative resorption efficiencies of structural and toxic nutrients have also been previously reported (Killingbeck 1993; Liu et al. 2014; Du et al. 2017). This is likely due to these nutrients are freely loaded into xylem but highly restricted in phloem (Bukovac and Wittwer 1957; Hill 1980), which limits their inputs into perennial tissues. Therefore, they attain high levels at end of the growing season, in contrast to nucleic acid-protein nutrient elements (Saur et al. 2000; Andivia et al. 2010). The accretion of these nutrients from leaf shedding may assist with amplifying the reserves of other required nutrients, such as nucleic acid-protein and photosynthesis-enzyme elements, while concurrently playing a critical role in detoxification. This was also reflected through seasonal variations in the elements of Q. variabilis at the same sites, which revealed low structural and toxic nutrients and high nucleic acid-protein and photosynthesis-enzyme nutrient elements at the onset of the growing season (Du et al. 2021). These results revealed that plants engaged in resorption trade-offs as the end of the growing season approached. Generally, nucleic acid-protein and photosynthesis-enzyme elements were positively re-absorbed, with structural and toxic nutrients expelled in a trade-off manner between multiple elements at the conclusion of the growing season (Fig. 7). To facilitate this trade-off process, interannual climate and fresh leaf nutrient status may be referenced by plants.
Furthermore, Ca served not only as a structural element (Marschner and Marschner 2012) as was typically thought, but also the key regulator of plant growth and development (Hepler 2005). We found that Ca resorption was significantly correlated with multiple elements (Fig. S2). For instance, Ca could negatively regulate the resorption of C and N in Q. glandulifera and Q. variabilis, and this process could be coupled by N to influence the resorption of P and S (Fig. S2). Since leaf senescence is a genetically programmed process, Ca could act as a second messenger and involve a specific Ca2+-dependent protein kinase to regulate the process (Gepstein et al. 2003; Wang et al. 2019). Poovaiah and Leopold (1973) also demonstrated that due to the addition of Ca, the senescence of corn leaves was deferred. Therefore, we inferred that Ca might play a critical role in regulating the NuRE of plants, which needs to be further investigated.