Decoupled Soil Labile Carbon and Nitrogen Dynamics Triggered by Urban Vegetation


 Purpose Studies about soil carbon (C) and nitrogen (N) dynamics with land use change are urgently needed for urban ecosystems. We used fractionation of soils combined with stable isotopic analysis to examine soil C and N cycles after decadal forest and lawn planting in the Pearl River Delta, China. Methods Soil samples from bare soil (CK) and four land use treatments (55 and 20 years of forest plantation, F-55 and F-20; 55 and 20 years of lawn plantation, L-55 and L-20) were split into different chemical fractions. Then we analyzed the C and N contents, C/N ratio, δ13C and δ15N, C and N recalcitrant indices (RIC, RIN), and a C pool management index (CPMI).Results Forest vegetation substantially enhanced soil organic carbon (SOC) caused by the recalcitrant (RC) and labile C (LC) pools, while the larger soil organic nitrogen (SON) was ascribed to the increased recalcitrant N (RN). Enhanced LC but minor changes in labile N (LN) suggested a decoupled C and N in labile fractions of the forest soils. In contrast, the larger LN, and the enhanced decomposition of SOC, indicated that the lawns may have inhibited N mineralization of labile pools, also suggesting a decoupled C and N turnover and leading to low RIN values. Conclusions Urban forest and lawn plantations significantly changed the soil C and N dynamics, and emphasized the inconsistency between C and N processes, especially in labile pools, which would eventually lead to minor changes in N and limit the ecosystem C sequestration.


37
The rapid increase of urbanization has become one of the most important global issues in the          were rinsed using 20 ml of deionized water. Hydrolyzates were described as active pools (1).

184
Then the residues were performed a stoving at 65 °C. Remainings were hydrolyzed using 2 185 ml of 13 mol/L sulfuric acid at approximately 25°C overnight with continuously shakings.

186
The next step is to dilute the sulphuric acid to 1 mol/L, then hydrolyzing the samples at

C and N contents and isotopes in SOM
Specifically, X represents C or N, h refers to the heavy isotopes, l is the light isotopes. The 207 isotopic ratio of C ( 13 C) reflects the comparative value of PeeDee Belemnite Standard (δ 13 C = 208 0.0112372‰), while N stable isotopic ratio ( 15 N) is presented as the comparative value of 209 atmosphere (δ 15 N = 0.0‰). Standard samples are measured for each 10 samplings; and the 210 accuracy of the measurement is ± 0.13 ‰ for δ 13 C and ± 0.21‰ for δ 15 N, respectively.

211
For various types of land use plots (i.e. forests, lawns), the δ 13 C values were used to 212 estimate a percentage of fresh C (ffresh, that is, the plant-derived fresh C residue) and of aged   in root than in leaf and litter under forests. By and large, the averaged C/N ratios of the plant 266 materials were greater in forests compared with lawns (Table 1).

267
Moreover, no significant changes in soil C/N ratios occurred among the forests, lawns 268 and CK except for that in the L-55 soils, which had the lowest C/N ratios compared with 269 other land uses ( Table 2). The F-20 and L-20 increased but the F-55 decreased the soil BD,    (Table   281 3). No remarkable differences were found in δ 13 C values of SOM pools between the age of 20 282 and 55 years in both soil layers under each type of land cover (i.e. forest and/or lawn) except 283 for the δ 13 C values of RP at the 0-20 cm depth in lawn soils (Table 3). By comparison, more  (Table 3). were greatly increased in forest soils than those in bare soils, while there was no obvious 300 change in labile N (LN) storage between forest soils and CK (Fig. 1). The enhanced soil C 301 and N storage of SOMP, RP and LP were found in L-55 soils at the both depths, whereas no 302 significant increases in RC and RN storage were found between L-20 soils and CK (Fig. 1).

303
No notable discrepancies in soil C storage were found between F-55 and F-20 soils, while the 304 the top depth (Fig. 1). The C and N storage in recalcitrant pools was higher in F-55 (3347.36 306 g C/m 2 and 253.24 g N/m 2 , respectively) compared with F-20 soils (2366.82 g C/m 2 and 307 147.37 g N /m 2 , respectively), while LC storage was less in F-55 (1479.18 g C/m 2 ) than in 308 F-20 soils (2359.86 g C/m 2 ) in upper soil layer (P < 0.05; Fig. 1).

309
In general, The SOC, RC and RN content decreased in the following order: F-55 > F-20 >  Remarkable discrepancies were found in RIC ratios (P < 0.05) for the top layer, as well 337 as in RIN ratios and CPMI (P < 0.001) for the both soil layers among all land covers (Fig. 2).

338
Old forests such as F-55 markedly increased the RIC ratios (68.75%), while other land uses 339 had no significantly impact on RIC ratios at the topsoil compared with CK (Fig. 2). In general, and lawn systems compared with the control (P < 0.05; Table 4 and Fig.1 found in forests but enriched δ 13 C occurred in lawns in comparison with the control (Table 3), 356 possibly due to that a large amount of the C3 plant materials inputs into forest soil while 357 massively mixed C3 and C4 residues inputs into lawn soils (Table 1) (Table 4). Generally, the subsoil usually stores more stable C (i.e. physically or 369 chemically protected C) and has been suggested as a considerable C sink (Chen et al. 2013).

370
A faster SOC decay rate in deep layer could interpret the less SOC and SON content relative to 371 topsoil in lawns (Table 5), while a massive plant biomass mainly governed the residue return 372 into SOM stocks at the topsoil in forests and hence led to a higher content at the 0-20 cm 373 layer (Table 4).

374
Chemical fractionation analysis showed that the conversion from uncultivated area to 375 forest substantially enhanced the SOC due to the increased RC and LC pools (Fig. 1), while 376 an improved SON storage was ascribed to the increased RN stores in forests (Table 4). In 377 addition, we found that higher RC contents and stocks occurred in forests and L-55 plots 378 compared with CK (Fig.1). Generally, the recalcitrant pool has been considered as a stable

393
Furthermore, larger RC, RN as well as organic C storages were found in forest soils 394 compared with lawn soils (Fig.1), which proved that a better capacity for forest soil C and N 395 sequestration primarily owing to more plant residue inputs with a higher C/N ratio and a 396 slower soil C decomposition in forests, compared to lawns (Tables 1 and 5 litter and roots) inputs in forest soils than lawn soils (Table 4). In contrast, SON stocks did 401 not significantly differ between forest and lawn in the same planting ages, probably because 402 that a larger recalcitrant N storage was offset by its smaller labile N storage in forest soils 403 than those in lawn soils (Fig. 1) forest soils (Fig. 1). In addition, our results showed that soil LC storage was greater in forest 416 soils compared with CK (Fig. 1) (Table 5), but it may inhibit N mineralization of 424 labile pools (i.e. great labile N stores; Fig. 1), also stressing a decoupled C and N turnover 425 level and leading to a low RIN values (Fig. 2). Besides, herbaceous species such as Cynodon     Land use × Depth *** *** *** *** Table 4 Mean values (n = 9) for the C, N content, C/N ratios of SOM pools (0-40 cm) under different land covers. The abbr for land covers and years are the same as shown in Table 1. Means ± s.d. with different letters for a variable represent significant differences (P < 0.05). Note: n.s. = not significant; *P < 0.05; **P < 0.01; ***P < 0.001.