**4. Discussion**

Overall, there is a trend of converting pure plantations into mixed plantations for increasing stability and sustainability due to, in theory, complementary resource use, environmental benefits, and soil improvement in mixed forests [5]. However, the practical e ffect does not always match with our expectation. Until now, there have been few reports on the negative e ffect of mixture with broadleaved trees and coniferous trees [31].

### *4.1. Mixture E*ff*ect on Nutrient Concentrations in Rhizosphere and Bulk Soils*

Our results showed that organic C, total N, and total extractable P in both rhizosphere and bulk soils were higher in MP than PP. In other words, mixed Chinese fir plantations with broadleaved species increased nutrient concentrations in both rhizosphere and bulk soils. These results were in line with the hypothesis that the mixture with broadleaved trees improves soil P availability in Chinese fir plantations. Greater litter production or changes in the timing of litter inputs or di fferences of leaf nutrient concentration in mixed-species plantations could increase the soil nutrient supply, relative to monocultures, if decomposition rates were constant or faster [30]. Generally, the decomposition rate of litter and fine roots for coniferous species are slower than broadleaf species [32]. Additionally, some studies support that the decomposition rate of leaves and roost for *M. macclurei* and *S. superba* were faster than those of Chinese fir [33,34]. Moreover, compared with a pure Chinese fir plantation, the soil fertility and nutrient return were raised significantly in the mixed Chinese fir plantation with *M. macclurei*, due to substantial amount of litter and higher turnover rate of fine root [24]. Therefore, broadleaved trees with higher nutrient concentrations in the mixture could increase soil nutrient availability by altering the amount and quality of litter input as well as the amount and chemical composition of the root mass and exudates through fine root decomposition.

### *4.2. Mixture E*ff*ect Varies with Soil P Fraction*

The average concentrations of soil available P, slowly released P, and occluded P, but not soluble P, mineral P, and residual P, were significantly higher in MP than PP and three of these showed positive rhizosphere e ffects. Since P is an element of depositional cycle, P absorption by plants depends on its concentration gradient and di ffusivity in the soil near the roots. Both available P and soluble P are labile and are considered as the available fractions for plant growth [35]. Some studies have found that amending phosphates increases the immediate phosphorus availability and the rate of available P dissolution can be enhanced by the rhizosphere e ffect [36]. Moreover, the litter decomposition experiment of *S. superba* on a Chinese fir plantation shows that P concentration in the *S. superba* litter dropped rapidly, which could release 32% of the initial P into the soil during the first three months in Hunan Province [37].

Furthermore, the total extractable P was higher in MP than PP in both rhizosphere and bulk soils through the introduction of a broadleaf species in the mixed plantation. Recent studies indicated that the inactive P fractions could be converted into plant available forms with the help of the necessary manipulation of the rhizosphere environment [38]. The release of root exudates, such as organic ligands, is an activity of the root that can alter the concentration of P in the soil solution [12]. Some studies showed that cyclic dipeptides, which caused autoinhibition of Chinese fir, may be released into the soil through litter decomposition and root exudation. Moreover, root exudates provided more contributions to soil cyclic dipeptide levels than litter in Chinese fir plantations [39]. Thus, the introduction of broadleaf species to a pure Chinese fir plantation may relieve the autoinhibition of Chinese fir and alter the concentration of P in the soil solution. Furthermore, *S. superba* is a Mn-accumulating subtropical tree species and Mn hyperaccumulation is associated with Mn mobilization in the rhizosphere, most likely due to the release of protons. The carboxylates generated to produce the protons released into the rhizosphere are used internally in the plant to mobilize soil inorganic and organic P [40]. Therefore, the overall combination of changes in the P fractions in soil demonstrates that a mixture

effect may induce an increase in soil P availability through the solubility of the unavailable form of soil P and the rhizosphere effect. After all, the mixture effect on P availability in rhizosphere soil varies with the P fraction, and may depend on the litter decomposition and root exudates of introduced broadleaf species.

### *4.3. The Response of Chinese Fir to Mixture E*ff*ect Varies with Root Orders, Needle, and Twig Ages*

Our results showed that P and N concentrations in one-year old fresh needles and one-year old twigs were not significantly different between PP and MP in both study sites, although the mixture effect may induce an increase in soil P availability. Generally, evergreen trees tend to maintain a relatively favorable nutrient status in active young leaves for positive carbon (C) gain and high N use efficiency [41]. Trees can transport nutrients from old and senescing leaves to new organs to support new growth, which has been verified as a key mechanism of nutrient conservation and reuse in plants [42]. Thus, old leaves might be more sensitive to environmental variation, such as those mixed with a broadleaf species than young leaves. Generally, P and N concentrations in absorptive and transportive roots, two-year old fresh needles and two-year old twigs were significantly higher in PP than MP, although these concentrations in the other tissues were not different between both stand types. These results sugges<sup>t</sup> that the introduction of a broadleaf species to pure Chinese fir plantation might decrease the nutrient concentration of Chinese fir. In contrast, most of studies suggested that mixed species stands have higher nutrient availability than monoculture, even in the absence of N-fixing species [43], though few studies have contrasted nutrient concentrations for species in mixtures [21]. Therefore, foliar nutrient concentrations did not show a general trend in the mixtures studied so far. Moreover, since the Chinese fir biomass in the mixed plantation was lower than that in the pure Chinese fir stand in our study, we deduced that lower P concentration in Chinese fir tissue may be attributing to the stronger competition with the broadleaved tree. Obviously, our result did not support the hypothesis that the mixture can increase tree tissue nutrient availability but, rather, supported the hypothesis that the response of Chinese fir nutrients to the mixed with broadleaved tree varies with root orders, and needle and twig ages. These results may be caused by the lower competition of Chinese fir than broadleaved species for nutrient uptake.

### *4.4. Linkages between Soil, Root, and Needle Nutrients*

For the linkages among soil, root, and needle nutrients, we did not find any positive relationships between all P fractions in rhizosphere soil and P concentration in roots of different orders and in leaves of various ages. In contrast, slowly released P, occluded P, and residual P negatively correlated with the P concentrations in some root orders and fresh leaves of some ages, and the extractable P negatively correlated with P concentration in one-year old needle. Many studies sugges<sup>t</sup> that any shift in plant species composition, which is able to alter soil nutrient stoichiometry, can influence rhizosphere microbial and soil enzyme activities, which can further induce plant community species shifts and alter ecosystem function [44]. Two adjacent plants can simultaneously participate in competition and facilitation processes, and the direction and intensity of plant interactions are determined by the sum of the co-occurring negative and positive effects of one to another. These negative linkages may cause by the negative plant–plant interactions between Chinese fir and the broadleaf species. In general, plants take up most mineral nutrients through the rhizosphere where microorganisms interact with root exudates. A study reported that the seedling survivorship of *S. superba* was significantly inhibited by *eucalyptus* robusta Smith litter addition alone, meanwhile the seedling height of *S. superba* and *M. macclurei* was significantly suppressed when eucalyptus roots were present [45]. This negative nutrient feedback (homeostasis) supports plant species coexistence with lower proportional changes in consumer stoichiometry compared with resource stoichiometry [46]. Evaluating homeostatic relationships can provide valuable insight into assessing plant competition or plant coexistence [47]. Therefore, negative relationships between P fractions in rhizosphere soil and tree tissue nutrients of Chinese fir may be caused by plant competition between Chinese fir and the broadleaved species.
