Isotope Analysis Reveals Differential Impacts of Articial and Natural Afforestation On Soil Organic Carbon Dynamics in Abandoned Farmland

: Backgrounds A multitude of studies have applied different methods to study the 18 dynamics of soil organic carbon (SOC), but the differential impact of artificial and natural afforestation on SOC dynamic are still poorly understood. Methods and aims We investigated the SOC dynamics following artificial and natural afforestation in Loess Plateau of China, characterizing soil structure and stoichiometry using stable isotope carbon and radiocarbon models. We aim to 23 compare SOC dynamics, clarify SOC source under different afforestation, examine 24 comparability of the study areas and find how soil aggregate size classes control SOC 25 dynamics, finally to evaluate effect of reforestation project. 26 Results The 0-10cm and 10-20 cm SOC stocks were significant higher than other 27 two land-use system. At other depths, there is no significant difference among the 28 three land-use system. Total top soil SOC stocks, C:N and C:P of differently sized soil 29 aggregates significantly increased following afforestation. 13 results and 30 Radiocarbon models indicated that the SOC decomposition rate and new SOC input 31 rate were lower under natural afforestation than artificial afforestation. 32 Conclusions Afforestation can accumulate SOC in top soils mainly resulting 33 from in topsoil changing. SOC resource is mainly from macroaggregate formation 34 provided by fresh plant residues. SOC loss from soil respiration was derived from 35 microaggregates during afforestation. The “ space-for- time substitution” method is 36 suitable for comparability of the study areas. 37

stoichiometry using stable isotope carbon and radiocarbon models. We aim to 23 compare SOC dynamics, clarify SOC source under different afforestation, examine 24 comparability of the study areas and find how soil aggregate size classes control SOC 25 dynamics, finally to evaluate effect of reforestation project. 26 Results The 0-10cm and 10-20 cm SOC stocks were significant higher than other 27 two land-use system. At other depths, there is no significant difference among the 28 three land-use system. Total top soil SOC stocks, C:N and C:P of differently sized soil 29 aggregates significantly increased following afforestation. 13 C results and 30 Radiocarbon models indicated that the SOC decomposition rate and new SOC input 31 rate were lower under natural afforestation than artificial afforestation. 32 Conclusions Afforestation can accumulate SOC in top soils mainly resulting 33 from in topsoil changing. SOC resource is mainly from macroaggregate formation 34 provided by fresh plant residues. SOC loss from soil respiration was derived from 35 microaggregates during afforestation. The"space-for-time substitution" method is 36 suitable for comparability of the study areas.  Human ecosystem interventions such as deforestation and afforestation, can 64 cause rapid and persistent changes in vegetation and soil. Wei et al. (2013) found SOC 4 stocks decreased most rapidly during the first 4 years of cropland cultivation after 66 deforesting, but Rytter (2016) reported that SOC pools were generally unchanged 67 after five years of Salicaceae growth during afforestation of former agricultural land.  In this study, we combined (1) a time dependent steady-state box model based on 81 radiocarbon ( 14 C) to estimate SOC decomposition rates in different land-use systems; 82 (2) a natural abundance stable carbon isotope ( 13 C) study to quantify old and new     103 We obtained estimates of the recovery periods of different communities from   Plots were spaced approximately 50m apart in each land-use system. Meanwhile, each 121 plot was at least 40m from land-use system boundary to minimize edge effects. 122 We then dug five 20cm-deep pits for topsoil sampling at the four corners and the 123 center in each plot, and took three soil bulk density (BD) samples. BD was measured 124 at 0-20 cm in each subplot using a stainless steel cutting ring (5×5 cm). The soil cores 125 were dried at 105℃ for 24h. Five soil samples (0-20 cm) were collected from the five 126 pits in each plot, and the collected samples were mixed to form one homogeneous 127 sample for analysis of δ 13 C, 14 C, and aggregate size distribution. Three aggregate-size    Where Rsample is the 13 C/ 12 C ratio of the sample and Rstandard is the 13 C/ 12 C ratio in 170 the PDB standard.

171
The soil samples were pretreated for 14 C analyses as standardized by Zhou et al.  Where D is the thinness (cm) of the soil layer, BD is the bulk density (g cm -3 ), and OC is the soil organic carbon concentrate (g kg -1 ) at 0-20 cm.

198
Stocks of soil OC in each soil aggregate size class were calculated as: Where C0 is the initial SOC stock (SOC stock in the reference sites), Ct is initial SOC 214 stock remaining (old C stock) at time t (year) since ecosystem change.

215
For the long-term ecosystem, the decomposition rate constants k2 (yr -1 ) was 216 obtained through the bomb-14 C model (Cherkinsky and Brovkin 1993; Torn et al. where C is the organic carbon inventory of a soil sample (g C m -2 ), 14 C is the pMC of  Where ∆tn = tn+1-tn, which is the number of days between each field measurement 267 within the year; R is total soil CO2 emitted in the measurement period, and Fm,n is the   the OC stocks of macroaggregates and microaggregates significantly increased, but 305 the OC stocks of silt & clay were nearly 0 (far lower than for any other fraction), and could thus be considered a negligible part of total OC stocks. The OC stock of 307 macroaggregates was significant higher than that of microaggregate in AR, but the 308 OC stock of macroaggregate and microaggregate were nearly the same in NQ.

309
Meanwhile, The OC stock of macroaggregate and microaggregate in NQ were 310 respectively significant lower and higher than in AR (Fig. 2a). were significantly higher than AR, except for C:N ratio of macroaggregates in NQ 322 comparing with AR (Fig 3).  324 Over the afforestation period, the SOC decomposition rate (k1), and new SOC 325 input rate (k2) under artificial restoration were higher than natural restoration (Fig 4).

329
The amount of soil CO2 emissions across ecosystems was calculated from 330 continuous measurements over 4 years. Both natural and artificial restoration 331 significantly increased the amount of soil CO2 emissions, and the increased of amount 332 of soil CO2 emissions was higher under natural restoration (NQ) than artificial 333 restoration (AR) (Fig 6).  (Fig 7 a, b). In addition, F1 significantly increased and F2 slightly 340 decreased for microaggregate-OC (Fig 7 c, d). The aggregate-model (F1 & F2) results 341 indicated that increases in SOC stocks in macroaggregates and microaggregates under 342 artificial and natural afforestation were mainly due to increases in SOC concentration 343 rather than mass (Fig 7).  352 Restoration effect is the most important for evaluating reforestation project.

353
Actually this is very difficult to find primeval forest as undisturbed site for direct 354 comparison, because in this region there is no primary forest due to more than

436
In our study, the aggregate-model (F1 & F2) results indicated that increases in SOC stocks were mainly due to increases in SOC concentration rather than mass in 438 macroaggregates and microaggregates (Fig 7), and could conclude that for natural afforestation (Fig 6). Meanwhile, the analysis of soil stoichiometry in the 479 study suggested that C/N and C/P ratios in microaggregates were consistently higher 480 than in other fractions under afforestation (Fig 3). And Spohn and Chodak (2015) found that microorganisms increase their respiration rate with an increase in soil C/P 482 ratio and C concentration. Therefore, combining our results with previous studies, it 483 was concluded that SOC loss from soil respiration was mainly originated from 484 microaggregate resulting from higher C/P ratio after afforestation, supporting the   488 The comparability of the study areas is the central question in this study.

489
Previous studies in this area also demonstrate topography greatly influence soil

526
In conclusion, afforestation in study area is effective for SOC accumulation, and 527 natural afforestation is more significant than artificial, which is mainly resulting from 528 in topsoil changing. Carbon isotope and soil aggregates models analyzed that SOC 529 resource is mainly from macroaggregate formation provided by fresh plant residues.

530
The SOC concentration of soil aggregates played a dominant role in determining the 531 dynamics of SOC accumulation during afforestation period. Based on the analysis of 532 soil aggregates, soil respiration, and aggregate stoichiometry, we concluded that SOC 533 loss from soil respiration mainly originated from microaggregates during afforestation. 534 13 C and 14 C models proved effective tools that showed that recovery time is a key 535 factor determining the accumulation of SOC following afforestation. Especially, the 536 study confirm that"space-for-time substitution" method is suitable for comparability 537 of the study areas.                The different aggregate size OC stocks (a) and size class distributions (b) in the three land-use systems.

Figure 3
The C:N and C:P of different soil aggregate size classes in the three land-use systems. Macroaggregates were >0.25 mm; microaggregatew were between 0.25 and 0.053 mm; and silt & clay were <0.053 mm. Error bars represent the standard error of the mean. Signi cant differences are indicated by the different letter (p < 0.05); AF: abandoned farmland; AR: arti cial afforestation (plantation of Robinia pseudoacacia L); NQ: natural afforestation (natural forest -Quercus liaotungensis Koidz).

Figure 4
SOC decomposition rate constants (k1) and new SOC input rate (kg m-2 yr-1) calculated with a 13C model under (a) arti cial and (b) natural afforestation. Error bars represent the standard error of the mean. Signi cant differences are indicated by the asterisk symbol (*p < 0.05).  The mean annual soil CO2 emission over 4 years in the three land-use systems. Error bars represent the standard error of the mean. Signi cant differences are indicated by the different letter (p < 0.05); AR: arti cial afforestation (plantation of Robinia pseudoacacia L); NQ: natural afforestation (natural forest -Quercus liaotungensis Koidz).