Fresh litter acts as a substantial phosphorus source of plant species appearing in primary succession on volcanic ash soil

Plants have difficulty absorbing phosphorus from volcanic ash soils owing to the adsorption of phosphorus by aluminum and iron in the soils. Thus, on volcanic ash soils, the phosphorus source for natural vegetation is expected to be organic matter, however, there is a lack of experimental evidence regarding this occurrence. Here, we studied the effect of organic matter on plant growth of some species that occur in primary successions of volcanic ash soil ecosystems, based on growth experiments and chemical analyses. We found that a large amount of inorganic phosphorus (but only a limited amount of inorganic nitrogen) is leached from fresh leaf litter of the pioneer spices Fallopia japonica at the initial stage of litter decomposition. Phosphorus from the fresh litter specifically activated the growth of subsequently invading nitrogen-fixing alder when immature volcanic soil was used for cultivation. In contrast, old organic matter in mature soil was merely a minor source of phosphorus. These results suggest that fresh litter of F. japonica is essential for growth of nitrogen-fixing alder because the litter supplies phosphorus. We consider that rapid phosphorus cycles in fresh litter-plant systems underlie the productivity of natural vegetation even in mature ecosystems established on volcanic ash soils.

www.nature.com/scientificreports/ and whether they contributed to vegetation development. We focused on organic matter and set up a working hypothesis that phosphorus leached from the organic matter promotes plant growth in volcanic ash soil. First, growth experiments and chemical analyses were conducted to determine whether phosphorus leached from fresh litter promoted plant growth in a typical primary succession plant species growing in volcanic soils-Fallopia japonica, Alnus inokumae, and Betula ermanii. They were grown in immature volcanic ash soil and their growth rate, with and without litter, were compared. Second, nutrient supply by old organic matter in mature volcanic ash soil was verified similarly as described above. These experiments show that fresh litter leaves are a major source of natural vegetation in volcanic ash soil and that old organic matter in mature soil is merely a minor phosphorus source. Finally, we discussed the possibility of a rapid phosphorus cycle in fresh litter-plant systems that underlie the productivity of natural vegetation established on volcanic ash soil.

Materials and methods
Volcanic ash soil. We used commercially available Kanuma soil as volcanic ash soil (fine-grained pumice, Akagi Engei Co., Ltd.) for growth experiment 1. Kanuma soil is a fully weathered pyroclastic fall from the eruption of Mt. Akagi 44,000 years ago 19 . The soil contains 30.8% aluminum (allophane and imogolite) and 1.4% iron (ferrihydrite) 20 . For growth experiment 2, we used three natural volcanic ash soil types-immature soil of pumice (Kanuma soil, C horizon), as well as mature soils of andosol (A-B horizon) and topsoil (the surface of andosol, P to A horizon)-collected from a riverbed in Kanuma City (36°35′ N, 139°44′ E; 200 m a.s.l.), central Japan, where the vegetation is a cypress forest. This place is managed by the Kanuma Civil Engineering Office. The topsoil was collected at a depth of approximately 0-10 cm from the soil surface after removing the fallen leaves on the soil surface. The andosol layer, typically distributed at a depth of approximately 10-75 cm, was collected from a depth of approximately 10-30 cm. Below the andosol layer, the Akadama soil layer is distributed; further below, the pumice Kanuma soil is distributed. The pumice was collected approximately 50 cm under the Akadama layer. Temporal information on soil formation was confirmed by direct radiocarbon dating of the soil samples. After removing soil carbonate with 1.0 M HCl, the total organic fraction was analyzed using an accelerator mass spectrometer (0.5MV compact AMS system, NEC) at the laboratory of radiocarbon dating, University of Tokyo. Conventional radiocarbon age after correction of isotopic fractionation with δ 13 C values was calibrated to a calendar date with the calibration dataset IntCal13 21 .
The elemental analysis of total phosphorus, nitrogen, and carbon in the soil samples was performed by Createrra Inc. Litter incubation experiment. Samples (1 g) of F. japonica litter leaves-collected upon leaf fall on an autumn day, dried at 80 °C for at least 48 h, and then crushed-were placed in cultivating tubes (n = 5). Then, 5 g of wet soil from the Nikko Botanical Garden (36°45′ N, 139°35′ E; 647 m a.s.l.) in Nikko, central Japan, was added to 500 mL of water and stirred (solution I). As inoculation, 0.1 mL of the supernatant of solution I was added to the tubes 24 . Considering that the amounts of phosphorus and nitrogen in the solution I were approximately 0.003 mg/L and 0.3 mg/L, respectively, they were determined to have not affected the initial value (t = 0). Next, 2 mL of water was added to the tubes, which were then kept at 30 °C. The tubes were left open to maintain an aerobic environment. The efflux of phosphorus and nitrogen from the leaves was measured every week for ten weeks. For these measurements, 5 mL of water was added and the tube was centrifuged for 10 min (solution II). The supernatant of solution II was then used for phosphorus and nitrogen measurements, and the residue was continuously kept at 30 °C.  www.nature.com/scientificreports/ Growth experiment 1: Comparative experiment on the growth of plant species with and without litter. The seedlings used for the experiment were from the species F. japonica, A. inokumae, and B. ermanii. A similar seedling size was used for each plant species. Seedlings of A. inokumae coexist with N-fixing actinomycetes. Six plants per species were collected before cultivation (t = 0) and dried in an oven at 80 °C for at least 48 h to measure the dry weight. There were four experimental groups for each species: a control (Con), a nitrogen addition (N: 10 mM NH 4 NO 3 ), a phosphorus addition (P: 10 mM NaH 2 PO 4 ), and a nitrogen and phosphorus addition (NP: 10 mM NH 4 NO 3 + 10 mM NaH 2 PO 4 ). Once a week, 50 mL of each nutrients was added to a 0.25-L garden pot. To verify whether the addition of litter (denoted by +) improved plant growth, litter leaves of F. japonica were placed on the soils. To verify if nutrients leached from litter sustained plant growth, we also combined nutrient and litter additions (Con+, N+, P+, NP+). When nutrients were added to the soil once a week, litter bag was removed before fertilizer application and returned after that.
To reproduce how litter is deposited and supplies nutrients on volcanic ash soil in primary succession, F. japonica litter was collected in Nikko in the autumn of 2018 and dried at 80 °C or 2 days or more (the same litter was used in incubation). Approximately 9 g of litter leaves was packed in a tea mesh bag 25 to prevent it from flying in the wind and placed on the soil surface of the garden pots. As indicated by the equation below, the amount of litter added to the 8 × 8 cm (0.0064 m 2 ) garden pot used in this experiment amounts to approximately three years of litter production when converted to the amount of leaf litter in a 15-year-old alder forest, i.e., about 430 g/m 2 per year 26 .
Six seedlings per group of A. inokumae and B. ermanii were cultivated for approximately 2 months (June 7-August 22, 2019) and 12 seedlings per group of F. japonica were cultivated for about 1 month (September 10-October 15, 2019). The experiment was stopped after 1 month for F. japonica as it grew rapidly in 2 nutrient conditions (NP, NP+) and the roots overflowed from the garden pot. At the end of the experiment, growth was evaluated by measuring dry weight after drying seedlings at 80 °C for at least 48 h. Subsequently, the total phosphorus and nitrogen content of the dried seedlings were also measured (chemical analysis).
The mass of phosphorus leached from litter during the cultivation period was calculated from the difference in the phosphorus contents of the litter before and after cultivation.
Growth experiment 2: Comparative experiment on plant growth with old organic matter. Eight F. japonica seedlings were cultivated in three different soil-types (pumice, andosol, and topsoil, as mentioned above) under three experimental conditions (Con, N, P, same nutrition as growth experiment 1) from May 29 to July 12, 2019. These plants were then harvested and oven-dried at 80 °C for at least 48 h to measure dry weight. Subsequently, the total phosphorus and nitrogen content of the seedlings were also measured (chemical analysis).
Chemical analysis. Phosphorus. We used the dry destruction method to pretreat total phosphorus measurements in plant tissue 27 . A sample of the plant (0.05 g) was burned at 550 °C for 1 h. The plant ash was dissolved in 10 mL of 2 M H 2 SO 4 and shaken for over 16 h; then, the solution was filtered. The filtrate was diluted at a 1:10 ratio with tris(hydroxymethyl)aminomethane (pH 8.0).
The soil for available phosphorus were pretreated by Truog' s method 28 . The soil (0.05 g) was dissolved in 10 mL of 0.002 M H 2 SO 4 , shaken for 30 min, and the solution was filtered. The filtrate was diluted at a 1:10 ratio with water.
The amount of phosphorus in the sample solution was measured by the molybdenum blue colorimetric method 29 .
Nitrogen. The total nitrogen in plant tissue was measured using an elemental analyzer (EA; Vario Macro cube, Elementar, Germany). A few milligrams of the dried plant sample were placed in a tin capsule for EA combustion. EA carried out sample combustion and N 2 separation/detection from the combusted gases and provided us with nitrogen contents.
The soil sample preparation for available nitrogen measurements was based on the incubation methodology 30 . Half of the sampled soils were analyzed fresh, and the other half incubated for four weeks at 30 °C before analysis. 2 M KCl (20 mL) was added to 2 g of the soil sample; the solution was shaken for 1 h and filtered. The filtrate was collected, and the volume of nitrogen was measured by indophenol blue absorptiometry after reducing all to ammonia using Pack Test WAK-TNi (Kyoritsu Chemical-Check Lab., Corp, Tokyo, Japan). Available nitrogen was taken as the difference in the concentration of inorganic nitrogen (NO 3 -N, NO 2 -N and NH 4 -N) between incubated and fresh soil.

Statistical analysis. All statistical analyses were performed with EZR 31 (Saitama Medical Center, Jichi
Medical University, Saitama, Japan), which is a graphical user interface for R (The R Foundation for Statistical Computing, Vienna, Austria). More precisely, it is a modified version of R commander designed to add statistical functions frequently used in biostatistics. The figure's values are mean ± SE. Intergroup differences for nutrition conditions in soil, and soil-types were evaluated using non-parametric Kruskal-Wallis with posthoc Steel-Dwass tests. In addition, comparisons between with or without litter were evaluated using two-tailed Mann-Whitney U-test. p values are * p < 0.05, ** p < 0.01, *** p < 0.001. www.nature.com/scientificreports/

Results
Litter incubation experiment. Inorganic phosphorus and nitrogen liberated from the litter are shown in Fig. 1. About 20% of total phosphorus in litter leaves was inorganic phosphorus at t = 0 (80% was organic and remaining in litter), and about 40% of the total phosphorus was leached during the initial three weeks (Fig. 1c).
In contrast, as compared to phosphorus, only small relative changes were observed in leached nitrogen.
Growth experiment 1. The growth of A. inokumae was significantly increased with litter compared to that without litter (A. inokumae: Con vs. Con+, p = 0.009) (Fig. 2a, Supplementary Figure S1). Compared to the control group, the growth of F. japonica and B. ermanii significantly increased in the groups treated with nitrogen and phosphorus (F. japonica: Con vs. NP, p = 0.0003; B. ermanii: Con vs. NP, p = 0.02), as shown in Fig. 2b and Supplementary Figure S1. The growth of A. inokumae significantly increased in the groups supplemented with phosphorus (A. inokumae: Con vs. P, p = 0.04; Con vs. NP, p = 0.03). Therefore, the limiting  To verify the effects of litter addition for growth, we cultivated plant species with litter and nutrients. The growth of F. japonica and B. ermanii was significantly increased by the addition of nitrogen and litter (F. japonica: Con+ vs. N+, p = 0.0002; B. ermanii: Con+ vs. N+, p = 0.03), and there was almost no difference in A. inokumae by addition of nutrients (Fig. 2c, Supplementary Figure S1). These results suggest that all plants would obtain a sufficient amount of phosphorus from litter.
Accumulation of phosphorus and nitrogen in plants with or without litter was shown in Fig. 2d. For A. inokumae and B. ermanii, the total phosphorus and total nitrogen in the plant were measured at the start of the growth experiment (t = 0) and after cultivation (t = 2 months). Since the mass F. japonica at t = 0 was insufficient for elemental analysis, F. japonica was not included in Fig. 2d. The addition of litter (Con+) increased the total phosphorus and nitrogen contents (Fig. 2d). Especially, total phosphorus contents of A. inokumae significantly increased by litter additions (A. inokumae: t = 0 vs. Con+, p = 0.01; Con vs. Con+, p = 0.02). During the cultivation period, phosphorus leached from litter was more than that absorbed by plants (Table 2). Therefore, litter is considered the major phosphorus source of the litter addition experiments, especially for A. inokumae.
Growth experiment 2. Growth experiment 1 was conducted to verify the effects of litter, the new organic matter. However, old organic matter is also present in mature natural vegetation volcanic ash soil; thus, we tried to verify the effects of old organic matter, using mature Kanuma soil collected from a cypress plantation. Table 3 shows the age of organic matter in each soil. The organic fraction dates were 420 to 540 cal AD for the topsoil and 385 to 535 cal AD for the andosol (95.4% probability interval). The pumice could not be measured because of quite low carbon content.
Phosphorus significantly promoted F. japonica growth in andosol and topsoil (andosol: Con vs. P, p = 0.01; topsoil: Con vs. P, p = 0.003) ( Fig. 3a and Supplementary Figure S2). Figure 3b shows the total phosphorus and nitrogen contents of the plants after cultivation under control condition (Con), suggesting phosphorus deficiency in andosol and topsoil ( Fig. 3b and Supplementary Table S3). Further, the plants grown on andosol and topsoil had considerably higher nitrogen contents than those grown on pumice, suggesting the existence of available nitrogen in andosol and topsoil (Table 3). These results indicate that the plants are able to absorb sufficient nitrogen from the mature soil containing old organic matter.

Phosphorus leaching from litter promotes growth of N2-fixing A. inokumae, a succeeding species in the middle stage of primary succession. In the incubation experiment, approximately 40%
of the total phosphorus of litter was in an inorganic form within three weeks after litter decomposition started (Fig. 1c). In contrast, almost all nitrogen persisted in an organic form incorporated in litter. These results suggest that a large amount of phosphorus can be supplied to plants from newly fallen litter, whereas little nitrogen is available. Tateno (1988) showed a similar tendency for nitrogen mineralization in an incubation experiment where nitrogen was not mineralized from fresh leaves of F. japonica in a 250 d incubation period, supporting our present result regarding nitrogen behavior in litter 24 .
In the growth experiment using immature pumice, the growth of N2-fixing A. inokumae was promoted by adding phosphorus or litter (Fig. 2a,b,c), while non-N2-fixing F. japonica and B. ermanii needed nitrogen in addition to phosphorus or litter. This is probably because the Alnus species coexist with N2-fixing actinomycetes 32 . In fact, A. inokumae had higher nitrogen content than that exhibited by B. ermanii (Fig. 2d) when only litter was placed on the pumice. In the same experiment, the addition of litter increased the amount of phosphorus absorbed by plants (Fig. 2d). This answers our research question, and shows that litter acts as a substantial phosphorus source for plants, at least in the case of volcanic ash soils, although mineral soil may be a phosphorus source in some cases 33 .
Similarly, the addition of litter increased the amount of nitrogen absorbed by plants (Fig. 2d). However, there was no improvement in the growth of non-N2-fixing B. ermanii by the addition of both phosphorus and litter. It is probable that, since plant tissues require more nitrogen than phosphorus, the amount of nitrogen leached from fresh litter is not sufficient for plant growth at the stage of primary succession of volcanic ash soil with low litter deposition.
It is notable that even N2-fixing A. inokumae hardly grew on pumice without additional phosphorus or litter. In the middle stage of primary succession in Japanese volcanic ash soil, A. inokumae invades after F. japonica 23 . Therefore, F. japonica litter is essential for expressing the N2-fixing ability of A. inokumae because the litter supplies phosphorus to A. inokumae (Fig. 4).
Our results shows that litter acts as an important phosphorus source for plants. The results also give a cue for understanding N2 fixation in terrestrial ecosystems. N2-fixing species do not appear in early stage nor late stage of mature vegetation 22,23,34 . N2-fixing species cannot sufficiently absorb phosphorus without litter from pioneer species. Therefore, N2-fixing species cannot invade the early stage. Considering that N2 fixation is a costly process 35 , the advantage of N2-fixing ability disappears in the late stage of primary succession where N2-non-fixing species utilize sufficient nitrogen accumulated as a part of SOM.
Rapid phosphorus cycle in ecosystems on volcanic ash soil. A question arises from our growth experiment: What is the nitrogen source in mature ecosystems in volcanic ash soil? Fig. 3a shows that the addition of phosphorus promotes the growth of F. japonica in andosol and topsoil of mature Kanuma soils, suggesting that phosphorus is the limiting nutrient in the soils and that plant species can absorb nitrogen in the soils. In fact, F. japonica grown on andosol and topsoil had considerably increased nitrogen contents compared to those www.nature.com/scientificreports/ grown on pumice (Fig. 3b). The topsoil and andosol contained large amounts of old organic matter (organic matter age: about 400 AD to 500 AD, Table 3). Therefore, we suggest that nitrogen can be supplied from the mature soil, while the major source of phosphorus is newly fallen litter, even in mature ecosystems established on volcanic ash soils. However, it is possible that the andosol and topsoil contained a small amount of new organic matter that could not be detected by radiocarbon dating, the age of the nitrogen source is still unknown. These results suggest that old organic matter may act as a nitrogen source for plants. However, there is a problem remaining. Old organic matter in andosol and topsoil is presumed to be persistent humus formed from plant litter 36 . It is probable that the organic nitrogen component of humus 37 decomposes over a long period of time. The difference in nitrogen mineralization rate between litter and humus may be explained by the CN ratio 38,39 . However, the CN ratio of F. japonica litter (Supplementary Table S4) was about 25 that was relatively similar to that of the humus in topsoil and andosol (Table 3). A similar tendency has previously been reported, which demonstrates that nitrogen hardly mineralizes after 250 d of litter and fresh leaves of F. japonica incubation, although the CN ratio of fresh leaves has been estimated to decrease to about 12 by the end of the incubation Figure 2. Effect of litter on the growth of Fallopia japonica (n = 11-12), Alnus inokumae (n = 5-6), and Betula ermanii (n = 5-6). (a) Differences in growth with and without litter. (b) Differences in growth with nutrition. (c) Verification of growth-promoting nutrition supplied by litter. (d) Differences in total phosphorus (n = 5-6) and nitrogen (n = 3) of A. inokumae and B. ermanii before and after cultivation, and with or without litter addition. The raw data are available in Supplementary Table S2. The figure's value are mean ± SE. Mann-Whitney U-test was used for (a). Kruskal-Wallis with post-hoc Steel-Dwass tests were used for (b), (c) and (d). p values are *p < 0.05, **p < 0.01, ***p < 0.001. Growth of A. inokumae was promoted by phosphorus (P) or litter leaves (+). Growth of F. japonica and B. ermanii were promoted by both nitrogen and phosphorus (NP) and both nitrogen and litter leaves (N+). Litter leaves (+) increased phosphorus, especially in A. inokumae, and increased nitrogen levels in plants.  www.nature.com/scientificreports/  www.nature.com/scientificreports/ experiment 24 . Further research is required to investigate the mechanism of nitrogen mineralization in organic matter for nature of primary succession. This study suggests that phosphorus is rapidly leached from litter and contributes to the vegetation development on volcanic ash soil. Model analysis of strongly weathered tropical soils deficient in phosphorus revealed that organic phosphorus was the major source of the available phosphorus 40 . The high productivity of tropical rainforests may be supported by the rapid leaching of phosphorus from fresh litter as shown in this study.

Data availability
All data generated or analyzed during this study are included in this published article and its supplementary information files.