Carbon Mineralization And Sequestration Response To Residue Return And Nitrogen Fertilization In Soils With Different Texture And Cropping Histories

Soil organic carbon (SOC) mineralization and external organic material decompositionare important processes in soil carbon cycling. Use of diversified farming systems that rely oncrop rotation and return of crop residues and/or manures is advocated as one way to improveagricultural sustainability and increase SOC level. Carbon turnover response differently toresidue return and inorganic fertilization in soils with various texture and cropping histories.Predicting carbon mineralization and sequestration response to residue and nitrogen fertilizeradditions by looking into soil basic properties is with both theoretical and practicalsignificance.In this study, we investigated carbon mineralization characteristics and the fate and rateof soil native and residue-derived C distributions into soil physical fractions by adding13C-labled wheat residue (residue return rate:8580kg ha-1) to clay-loam Mollisols from Iowa(IAsoil) and silt-loam Ultisols from Maryland (MDsoil). Both soils were collected fromlong-term trials comparing two year rotations amended with inorganic fertilizers (Conv), with3yr (Med) and longer (Long) rotations that were manure amended. Four treatments wereincluded: Control: soil with no residue or N fertilizer added. N: with N fertilization. R: withresidue application. RN: applied with both residue and N fertilizer. The CO2emission rateswere monitored for330days and cumulative CO2-C release data were fit into a doubleexponential model to estimate the size (Ca, Cs) and decay rates (ka, ks) of active and slow Cpools. Soil samples collected at the beginning, after180and330days, were fractionated intolight fraction (LF,<1.60g cm-3), occluded particulate organic matter (OPOM,0.053～2mm),and heavy fraction (HF,<53μm), in order to study the decomposition of soil native carbonand the distribution of residue-derived carbon in different SOM fractions. Results showed asfollows:(1) Soils with different texture and cropping histories showed different SOC contents andcarbon mineralization characteristics.The IAsoil containing high-activity smectite clays showed higher SOC level (28.0g kg-1) than MDsoil (14.1g kg-1) which contains semi-active kaolinitic clays, while25.5%of SOCexisted as particulate organic carbon (POM-C) in MDsoil and10%in IAsoil, suggestinghigher lability of SOC in MDsoil than in IAsoil. For IAsoil, the organic cropping histories(Med and Long) increased SOC by9.1%on average, increased POM-C by29.2%andOPOM-C by12.5%, these increases were insignificant; for MDsoil, organic croppinghistories significantly increased SOC, POM-C and OPOM-C by29.5%,50%and45.8%,respectively (p<0.05).After330days incubation, the cumulative amount of CO2-C release were similarbetween two soils (IAsoil=0.96～1.11g C·kg-1and MDsoil=0.92～1.22g C·kg-1), while thepercentage of SOC mineralized was higher in MDsoil than in IAsoil (7.9%vs3.7%). Thetwo-pool exponential model showed that the IAsoil maintained larger labile pools (Ca, Cs) andhigher decay rates (ka, ks) than MDsoil (p<0.05). Compared to Conv, Med and Long increasedthe cumulative amount of CO2-C release by33%and increased ka, ksby32%and46%respectively in MDsoil, while organic cropping histories did not show significant influence oncarbon mineralization in IAsoil.(2) Carbon mineralization response differently in two types of soils to residue return andN fertilization. Residue addition doubled the amount of cumulative CO2-C release in bothsoils, and prompted greater increases in Ca(340%vs230%) and Cs(38%vs21%) anddecreases in ka(58%vs9%) in IAsoil than MDsoil. Nitrogen fertilization dampened carbonmineralization in both soils by decreasing cumulative CO2-C release by5.7%in IAsoil and by8.8%in MDsoil. N fertilization increased ka by25%in IAsoil and ks by16%in MDsoil.The significant interactions were observed between residue addition and N fertilizationor cropping histories. In IAsoil, when no residue was added, Med and Long cropping historiesdecreased kaand increased kscomparing with Conv, and N fertilization increased kacomparedto no N-fertilized treatments, while the differences resulted from different cropping historiesor N fertilization were alleviated when residue added. In MDsoil, the Cashowed nosignificant difference between treatment Control and N (both were0.24g C kg-1) when noresidue added; residue addition decreased Caand made that the Caof treatments RN (0.76gC kg-1) was significantly lower than Caof treatment R (0.82g C kg-1), which suggested thatthe reduce of Cacaused by N fertilization could only be observed when with residue addition.(3) The distribution of SOM fractions varied in two soils.The percentage of LF in whole soil in terms of fraction weight (relative weight of LF)were twice high in MDsoil as in IAsoil, the relative weight of HF was17.0%higher, andOPOM was33.8%lower in MDsoil than in IAsoil, these results were due to the higher siltcontent in MDsoil and the higher sand content in IAsoil. Either organic system or RN treatments improved the relative weight of LF in MDsoil (p<0.05), while showed nosignificant effect on other fractions of MDsoil or any fraction of IAsoil. The weight of LFgenerally decreased along incubation and no changes were observed for OPOM or HF.(4) The13C labeling technic was applied to investigate the turnover of soil native carbon(Csoil) and residue-derived carbon (Cnew).By the end of incubation (330days), the relative loss rate of Csoilin each SOM fraction(Ratio of the amount of Csoilloss from each fraction to the SOC content of whole soil) showedas HF (4.42%～6.22%)>OPOM (2.48%～1.46%)>LF (0.25%～0.94%) in both soils,indicating that HF was the main source for carbon loss. The Cnewentered into LF right afteraddition, then transported into OPOM and HF. The residual rate of Cnewshowed as OPOM(16.2%～17.3%)>HF (14.3%～12.8%)>LF (1.83%～7.35%), meaning that more Cnewwerestabilized in OPOM, followed by HF.Compared to no residue-added treatments, residue addition increased the relative lossrates of Csoilin HF from2.89%to5.97%(p<0.05) in IAsoil, suggesting positive priming; anddecreased the relative loss rates of Csoilin OPOM from2.09%to0.83%(p<0.05) in MDsoil,suggesting negative priming.The residual rates of Cnewin LF were higher in MDsoil (7.35%) than IAsoil (1.83%)(p<0.05) and no significant differences residual rates of Cnewin OPOM or HF were observedbetween two soils, resulting in higher total Cnewstabilization in MDsoil (37.5%) than in IAsoil(32.4%).Cropping systems did not significantly change Csoilor Cnew distributions in IAsoil.While in MDsoil, compared to Conv, organic cropping histories increased Csoilcontents of LFand OPOM by44.1%and40.0%, respectively (p<0.05); showed no significant effect on Csoilcontent of HF; and increased residual rates of Cnewin OPOM by20.4%, showed no significanteffect on residual rates of Cnewin LF or HF.The decomposition of Csoiland stabilization of Cnewresponsed differently to Nfertilization between the two types of soils. In MDsoil, N fertilization reduced the relative lossrate of Csoilin OPOM by39.6%(p<0.05), from2.60%to1.57%, when no residue was added;and increased the residual rate of Cnewin LF by119%(p<0.05), from4.6%to10.1%, whenresidue was added. In IAsoil, N fertilization increased the residual rate of Cnewin HF by6.95%(p<0.05) and enhanced the positive priming on decomposition of HF-Csoil, resulting inthe faster decomposition of HF-Csoil, and did not change total SOC content of IAsoil ingeneral.(5) The results of correlation analysis showed that the resistant carbon pool size (Cr=SOC-[Ca+Cs]) was significantly correlated to HF-C pool, the correlation coefficientswere0.976(IAsoil) and0.862(MDsoil), respectively (p<0.001). The Crwas quantitativelylarger than HF, thus Crcontained C pool of HF and partially of LF and OPOM. In MDsoil, theLF-C pool size determined the potentially mineralizable C pool size (Ca+Cs). In IAsoil, soilparticle components had no significant correlation with Caor Cs. For both soils, OPOM andHF showed the greater contribution to C mineralization (35.8%and64.2%respectively) thanLF when no residue was added; and C mineralization were mainly derived from HF (85.1%),followed by LF (27.0%) when residue was added.Conclusively, the carbon turnover of high-SOC IAsoil could not be easily affected bycropping systems or fertilizations and newly-added carbon was more easily mineralized inIAsoil and did not alter SOC level. While for low-SOC MDsoil, organic cropping or residueaddition both could prmote the stabilization of newly-added carbon thus increased SOCcontent, especially the SOC content in LF and OPOM fractions.