添加堆肥并且种植苜蓿和黑麦草对芘污染的土壤进行植物修复

Phytoremediation of pyrene contaminated soils amended with compost and planted with ryegrass and alfalfaM.C.Wang a ,Y.T.Chen a ,S.H.Chen b ,S.W.Chang Chien a ,⇑,S.V.Sunkara aa Department of Environmental Engineering and Management,Chaoyang University of Technology,Wufong District 41349,Taichung City,Taiwan,ROC bDepartment of Geography,Chinese Culture University,Yang Ming Shan,Taipei 11114,Taiwan,ROCa r t i c l e i n f o Article history:Received 7September 2011Received in revised form 21December 2011Accepted 22December 2011Available online 14January 2012Keywords:Phytoremediation Pyrene Ryegrass AlfalfaOrganic acids Dissipationa b s t r a c tRyegrass (Lolium perenne )and alfalfa (Medicago sativa )were planted in pots to remediate pyrene contam-inated quartz sand (as a control group),alluvial and red soils amended with and without compost.The pyrene degradation percentages in quartz sand,alluvial soil,and red soil amended with compost (5%,w/w)and planted with ryegrass and alfalfa for 90d growth were 98–99%and 97–99%,respectively,while those of pyrene in the corresponding treatments amended without compost but planted with ryegrass and alfalfa were 91–96%and 58–89%,respectively.Further,those of pyrene in the respective treatments amended with and without compost but unplanted were 54–77%and 51–63%,respectively.Pyrene con-tents in both roots and aboveground parts of ryegrass and alfalfa after 90d growth in quartz sand and the two soils amended with or without compost were trace amounts.Statistical analyses for the parameters of ryegrass planted in red and alluvial soils including the concentrations of total water-soluble volatile low molecular weight organic acids,microbial population,pyrene degradation percentage,and spiked pyrene concentration show significant correlations at 5%and mostly 1%probability levels,by the analysis of variance.It was thus suggested that the interactions among the consortia of plant root exudates,micro-organisms,and amended compost in rhizosphere soils could facilitate bioavailability of pyrene and sub-sequently enhance its dissipation.Ó2012Elsevier Ltd.All rights reserved.1.IntroductionPolycyclic aromatic hydrocarbons (PAHs)are produced by the incomplete combustion of organic material arising,partly,from natural combustion such as forest fire and volcanic eruption,but with the majority due to anthropogenic emission such as cigarette smoking,automobile exhaust,and processing production and spill-age of petroleum (Samanta et al.,2002;Ravindra et al.,2008).Pyr-ene is a PAH consisting of four condensed aromatic rings and is a recalcitrant contaminant in the environment (Cerniglia,1992).Pyr-ene contamination of soils occurs naturally as a result of incom-plete combustion of organic materials such as fossil fuels and vegetation.Although pyrene is not carcinogenic,it is one of 16PAHs on the USEPA list of priority contaminants that can pose a hu-man health risk because pyrene can transform to benzopyrene,which is toxic (White,2002).Therefore,there is a continuous need for the remediation of such organic compounds from contaminated soils.An option in certain situations,phytoremediation is gaining popularity for its low cost and regarded as an effective method in remediation of organic contaminants such as PAHs.This remedia-tion method approach is an in-situ ,not destructive and could rem-edy the soil structure and recover the biological environment (Tesar et al.,2002).Phytoremediation is the use of plants and their associated soil microorganisms,soil amendments,and agronomic techniques to remove or render harmless environmental contami-nants (Cunningham et al.,1996).Kirk et al.(2005)reported that in petroleum contaminated soil,the plant species perennial ryegrass (Lolium perenne )and perennial ryegrass/alfalfa (Medicago sativa )mixture caused the greatest change in the rhizosphere bacterial community and these changes contributed to degradation of petro-leum hydrocarbons in contaminated soil.The microbial commu-nity in the rhizosphere is greatly influenced by roots and root exudates that vary depending on the conditions of plant growth.Root exudation varies both qualitatively and quantitatively be-tween plant species,as a function of plant age,mycorrhizal infec-tion and nutrient status (Jay and Jodi,1995;Günther et al.,1996;Grayston et al.,1996).It has been suggested that faster growing flora like grass species are effective plants for phytoremediation of PAHs in contaminated soils (Fang et al.,2001;Chiapusio et al.,2007;Wang et al.,2008).Grasses have fibrous root systems,resulting in large root length and surface area per unit volume of surface soil.The fibrous roots would provide a larger surface for colonization by soil microorgan-isms than a taproot (Anderson et al.,1993)and allow for greater0045-6535/$-see front matter Ó2012Elsevier Ltd.All rights reserved.doi:10.1016/j.chemosphere.2011.12.063Corresponding author.Tel.:+886423339373;fax:+886423339713.E-mail address:swcc@.tw (S.W.Chang Chien).interaction between the rhizosphere microbial community and the contaminant(Schwab and Banks,1994).Pyrene degradation was evaluated in contaminated soil with Bermuda grass and found that rhizosphere soil has a more diverse and active microbial commu-nity compared to nonvegetated soil.Consequently,the rhizosphere pyrene degrader population and pyrene degradation may be en-hanced compared to nonvegetated bulk soil(Krutz et al.,2005). Soleimani et al.(2010)investigated the effects of two grass species (Festuca arundinacea Schreb.and Festuca pratensis Huds.),infected and non-infected by endophytic fungi(Neotyphodium coenophi-alum and Neotyphodium uncinatum,respectively)on the degrada-tion of petroleum hydrocarbons in an aged petroleum contaminated soil.They reported that regardless of endophyte infection,pyrene removal in the rhizosphere of the two grass spe-cies were93–97%,whereas the removal in control was73%after 210d.The general subject of phytoremediation and particular its pro-cesses have been reviewed in many journal articles and book chap-ters.Literature review shows that many of the research articles contain some sections on rhizosphere aspects of phytoremediation. Moreover,the complexity and heterogeneity of contaminated‘‘real world’’soils still require the design of integrated approaches of rhi-zosphere management,e.g.,by combining those plants which have the potential to use in rhizo-degradation process and soil manage-ment.An improved understanding of the rhizosphere will help to translate the results of simplified pot experiments to the full com-plexity and heterogeneity offield applications(Wenzel,2009).The aim of the present study was thus to investigate the phytoremedi-ation of pyrene in quartz sand and two kinds of soils by using plant species ryegrass and alfalfa with their rhizosphere microbial com-munities and relation to organic matter amendments.In addition, different parameters were evaluated including pyrene uptake by plants,and microbial population and water-soluble volatile low molecular weight organic acids(LMWOAs)in rhizosphere soils. 2.Materials and methods2.1.ChemicalsPyrene(purity96%)used in this study was purchased from Supelco,USA.Pyrene stock solution(1g LÀ1)was prepared by dis-solving100mg of pyrene in100mL of high performance liquid chromatography(HPLC)-grade acetone.Acetonitrile used as the mobile phase in the HPLC analysis for pyrene was of HPLC grade. Double deionized water was used in all aqueous solutions and dilutions throughout the experiment.2.2.Collection of quartz sand,soils,and compost and their characterizationsTwo types of soils,alluvial soil and red soil,were collected from the surface soils(0–15cm in depth)of agriculturalfields in Taiwan. Sampling locations were identified by a Global positioning system. The collected soil samples were air-dried,ground,passed through a 2-mm sieve,and then stored in closed containers in a dry place at room temperature prior to use.Purchased quartz sand was passed through a2-mm sieve,and the collected quartz sand(with particle size less than2mm)was washed with0.1M HCl(1:10,w/v)to re-move trace amounts of adsorbed oxides,then washed several times with deionized water until it was ClÀfree,andfinally air dried.This quartz sand was used as the control group in this study. The pH of the quartz sand and soil samples were measured in dou-ble deionized water(1:1,w/v)using a model682Titroprocessor pH-meter(Metrohm,Swiss)(McLean,1982).The soil samples were digested with0.01M HCl(Tiessen et al.,1981)to remove inorganic carbon,and their total organic carbon(TOC)content was deter-mined using a Heraeus CHN–O-rapid elemental analyzer.The cat-ion exchange capacity(CEC)of the soils was determined according to the NaOAc method(Rhoades,1982).Particle size distributions were determined with a hydrometer method(Gee and Bauder, 1986).The compost which is commonly used by the farmers in Tai-wan was purchased from fertilizer retailer.The pH of the compost in double deionized water(1:5,w/v)was measured using a pH-meter model682Titroprocessor(Metrohm,Swiss)(McLean, 1982).The suspension,which was used for measuring compost pH,wasfiltered withfilter paper(Whatman42,pore size 2.5l m).The compost was digested with0.01M HCl(Tiessen et al.,1981),and then its organic C and total N contents were deter-mined by a Heraeus CHN–O-rapid elemental analyzer.After diges-tion of compost with concentrated HNO3and30%H2O2,it was analyzed for the contents of P,Ca,Mg,K,Na,B,Zn,Fe,and Mn by an inductively coupled plasma optical emission spectrometer (ICP-OES)(Jones and Case,1990).All of the measurements and determinations were of four replicates for the quartz sand,soil, and compost samples.2.3.Preparation of pyrene contaminated quartz sand and soilsBased on the obtained upper tolerant concentrations of pyrene in quartz sand,alluvial soil,and red soil for the growth of alfalfa and ryegrass in the preliminary experiment(data not shown), the amounts of200,400,and600mg of pyrene were weighed and separately dissolved in4L acetone in glass containers.These various concentrations of pyrene solution were prepared the re-quired sets.Then4kg of quartz sand,alluvial soil,or red soil was added to each of the three concentration levels of pyrene solutions. The resulted quartz sand or soil slurry was vigorously mixed by a stirrer for2h.The contaminated quartz sand and soil slurry sam-ples were equilibrated for2d to allow the sorption of contaminant onto the quartz sand or soils in a fume-hood cabinet.After2d,sol-vents were evaporated in a fume-hood cabinet at room tempera-ture.Thefinal concentration of pyrene(in the quartz sand and soil samples)was50,100,and150mg kgÀ1of quartz sand/soil, which were less than that of upper tolerant concentrations de-scribed above.The contaminated quartz sand,red soil,and alluvial soil were then placed on the shelves of a cabinet,and the doors were closed to maintain dark conditions for60d.During the aging period,the pyrene contaminated quartz sand,red soil,and alluvial soil were amended with the amount of double deionized water equivalent to theirfield capacity moisture tension(33.3kPa)at each3d interval by the weighing method.Moisture tension mea-surements of quartz sand(33.3kPa)and the two soils were ob-tained using a standard(14.5L)pressure cooker(Klute,1986). The aged pyrene contaminated quartz sand,red soil,or alluvial soil was divided into two sets.One set of each was amended with com-post(5%,w/w)while the other set was not.For the amendment of compost to the aged pyrene contaminated quartz sand,red soil,or alluvial soil,the required amount of air-dried and aged quartz sand or soil was ground to pass through a2-mm sieve.Similarly,the compost was pulverized to pass through a2-mm sieve.Then the required amount of sieved compost was mixed thoroughly with quartz sand,alluvial soil,or red soil.Each plastic pot with upper diameter12.1cm,lower diameter9.0cm,and height11.0cm and lined with clean plastic sheet from bottom to the upper rim of pot wasfilled with500g of pyrene contaminated quartz sand, red soil,or alluvial soil amended with or without compost.2.4.Planting of alfalfa and ryegrassIn order to let ryegrass and alfalfa grow to form as turf in quartz sand,red soil,or alluvial soil in a pot,100cultured seedlings of rye-218M.C.Wang et al./Chemosphere87(2012)217–225grass or alfalfa were transplanted to each pot in greenhouse for 90d growth period because this period has been tested to attain the maximum pyrene degradation in quartz sand and the two soils in the preliminary experiment.All the treatments of experiment including the controls were carried out four replicates.During the growth period,the pyrene contaminated quartz sand,red soil, and alluvial soil in pots were amended with the amount of double deionized water equivalent to theirfield capacity moisture tension (33.3kPa)at each3d interval by the weighing method.2.5.Extraction of pyrene from contaminated soilsIn the present study,continuous batch extraction method(Mara et al.,2007)was selected due to its high efficiency,quick recovery, and lowest volume of solvent required for pyrene extraction from contaminated quartz sand and soils.In the batch extraction proce-dure,1g of aged pyrene contaminated quartz sand,alluvial soil,or red soil was placed in a20mL centrifuge glass tube,after which 10mL of acetonitrile was added and the tube was capped with a screw cap lined with a Teflon pad.The centrifuge tubes were then placed in a reciprocal mechanical shaker(5cm amplitude and 150rpm)and agitated at room temperature(25±1°C)for a pre-scribed time(2h)to attain equilibrium.At the end of the predeter-mined time intervals,the sample was removed,and the supernatant was separated from the quartz sand,alluvial soil,or red soil by centrifugation at2000g for10min.Each centrifugate was thenfiltered using a syringefilter equipped with a0.22l m membranefilter(CRITICAL).The pyrene concentrations of thefil-tered solutions were then determined by HPLC(Hitachi High Tech-nologies,Tokyo,Japan).The column used was Mightysil RP-18GP (250Â4.6mm in size).The sample injection volume was10l L, and the mobile phase was90%acetonitrile and10%double deion-ized water with aflow rate of1mL minÀ1.The detector wave-length was220nm.The background responses,determined for blank samples with similarly treated quartz sand and the two soils (without pyrene),were subtracted from the experimental re-sponses obtained for their corresponding systems(with pyrene), and the resulting differences were taken as the initial pyrene con-centrations.After90d growth period,the remaining pyrene in the composite samples of rhizosphere and bulk quartz sand,red soil, and alluvial soil were used to calculate the pyrene degradation percentages.2.6.Extraction of pyrene from plant tissuesAfter90d growth period,the average height of the ryegrass was around12.3cm while that of alfalfa was around15.2cm.The growth of the plants was well and showed good endurance to the pyrene contaminated quartz sand and the two soils.All har-vested ryegrass and alfalfa were cut into roots and aboveground parts,washed with tap water and then rinsed with deionized water two times and blotted with clean tissue paper in order to remove excess water.The fresh weights of roots and aboveground parts were weighed and then subjected to oven-dried at50°C in an oven for overnight and their oven-dried weights were determined.The pyrene in the aboveground parts and roots were extracted accord-ing to EPA Standard Method3540C(USEPA,1996a).All the samples were extracted with acetone and dichloromethane(1:1,v/v, 80mL)in Soxhlet apparatuses for18h.Florisil column was used for purifying the concentrated extract(EPA Standard Method 3620B)(USEPA,1996b).The eluant was evaporated to less than 2mL by using a rotary evaporator prior to analysis.Pyrene concen-trations in the aboveground parts and roots were analyzed by using Gas Chromatography/Mass Spectrometer(GC–MS),based on the EPA Standard Method8270C(USEPA,1996c).2.7.Measurement of microbial population in rhizosphere quartz sand and soils and observation of mycorrhizal growth on rootsThe quartz sand or soil particles which are easily shaken off from ryegrass or alfalfa roots were collected as non-rhizosphere quartz sand and soil,while thefine quartz sand or soil particles which are attached tightly to the roots were collected as rhizo-sphere quartz sand or soil.10g of rhizosphere quartz sand,red soil, or alluvial soil in an autoclaved test tube were added with100mL sterilized double deionized water and the suspension was shaken and then stirred to make the suspension homogeneously.After the sedimentation of quartz sand or soil particles,10mL superna-tant were taken and diluted with sterilized double deionized water to a total volume100mL making the concentration of diluted mul-tiple10À1.This stepwise dilution procedure was consecutively conducted for six times to obtain the concentrations of diluted multiples10À2to10À7for each rhizosphere quartz sand or soil sample.The prepared culture medium of agar plates was divided into the required sets and each set was evenly smeared with 0.1mL of sterilized double deionized water as control and with that of each concentration of diluted multiple solutions.Finally, all the plates for the treatments were placed in an incubator at 30°C for3d.The growth of microbial colonies on agar plates thus can be used to count the microbial population in rhizosphere quartz sand or soil by colony forming unit(CFU).In addition,the mycorrhizal growth on the intact roots of both ryegrass and alfalfa was observed by a Nikon MM-400/S measuring microscope sys-tem.The magnified multiples vary from50to1000times.2.8.Extraction of water soluble LMWOAs in rhizosphere soils and their determinationEach portion of rhizosphere red soil or alluvial soil washed from each pot described above was extracted with doubled deionized water(500mL)using an end-over-end shaker(amplitude of 2.5cm and200rpm)for8h at25±1°C.The extract was centri-fuged(18600g)for15min at25±1°C to separate the supernatant from the sediment.The supernatant was thenfiltered through a 0.45l m pore size of cellulose nitrate membranefilter(Whatman). Thefiltrate was transferred to a plastic vial containing an anion ex-change membrane(1cmÂ7cm)in hydroxyl form(IONAC MA-3475)and shaken in an end-over-end shaker(amplitude of 2.5cm and200rpm)for8h at25±1°C.The membrane was then transferred to another vial containing HCl(5mL of0.1M solution) and again shaken in an end-over-end shaker(amplitude of2.5cm and200rpm)for8h at25±1°C.All the acidified aqueous extracts were directly used for gas chromatography(GC)analysis of volatile LMWOAs(Szmigielska et al.,1996,1997).The separate column (glass column,2.1m in length and4mm of inner diameter)was filled with SP-1000(Supelco)as stationary phase and aflame ion-ization detector was utilized for GC analysis.The injector,column, and detector were kept at200,150,and200°C,respectively.Nitro-gen gas(purity99.995%)was used as carrier gas at aflow rate of 60mL minÀ1,while the nitrogen gas pressure was kept at 98.1kPa.The HAMILTON(SIX-701N)5l L syringes were applied to inject the samples(1l L each).The chromatograms were re-corded and peaks were integrated using Hewlett–Packard HP3395integrator.All the standard samples of LMWOAs were pur-chased from Supelco,USA.2.9.Statistical analysesAfter90d growth of ryegrass in red and alluvial soils amended with and without compost,the statistical correlations among the concentration of total LMWOAs,microbial population,percentage of pyrene degradation,and spiked pyrene concentration describedM.C.Wang et al./Chemosphere87(2012)217–225219above were investigated.The purchased software SPSS17.0(2008) was used to compute all correlation coefficients among the corre-sponding parameters.3.Results and discussion3.1.Characteristics of quartz sand,soils,and compostAlluvial soil was collected from the sampling location (120°4102.900E,24°1052.500N)while red soil was from the location (120°3509.000E,24°11022.400N).The pH(6.71)of alluvial soil was slightly acid while pH(6.06)of red soil was significantly lower than that of alluvial soil.Quartz sand is chemically inert inorganic com-ponent of the soils in environment,and hence its pH(6.84)was close to the pH of doubled deionized water.The TOC content (4.6mg kgÀ1)of alluvial soil was significantly larger than that (1.6mg kgÀ1)of red soil.The CEC(10.5cmol(+)kgÀ1)of alluvial soil was also significantly larger than that(8.5cmol(+)kgÀ1)of red soil. This is attributed to higher TOC content of alluvial soil compared to higher oxide contents of highly weathered red soil.The texture of alluvial soil was coarse sandy loam and that of red soil was clay, while that of quartz sand was sandy.The pH(5.40)of compost was acidic with high contents of soluble salts.The compost had high organic C(434g kgÀ1)and total N(25.8g kgÀ1)contents with C/N of17.The amounts of macronutrients(P0.82,K0.29,Ca8.37, and Mg2.01g kgÀ1)and micronutrients(Fe414,Mn63.0,B14.7, and Zn16.4mg kgÀ1)of the compost were within the range of common composts.3.2.Biomasses of alfalfa and ryegrass planted in quartz sand and soilsAt the end of90d growth period,dry biomasses of roots and aboveground parts of alfalfa and ryegrass planted in quartz sand, red soil,and alluvial soil spiked with pyrene at the concentration of0,50,100,or150mg kgÀ1with the amendment of compost were mostly and significantly greater than the corresponding treatments without the amendment of compost(Table1).This is mainly because the amended compost supplies inorganic and organic nutrients for the growth of alfalfa and ryegrass.Both without and with the amendment of compost,the dry biomasses of roots of alfalfa and ryegrass planted in quartz sand,red soil,and alluvial soil spiked with pyrene at the concentrations of0,50,100,and 150mg kgÀ1were mostly and significantly greater than the corresponding aboveground parts(Table1).This implicates that both alfalfa and ryegrass are highly root system development plants suitable for phytoremediation of contaminated soils.In comparison,both dry biomasses of roots and aboveground parts of ryegrass were mostly and significantly greater than those of al-falfa planted in red soil and alluvial soil in the corresponding treat-ment.However,both without and with the amendment of compost,much higher concentrations of spiked pyrene in quartz sand,red soil,and alluvial soil did not contribute to significant ef-fect of toxicity on the dry biomasses of roots and aboveground parts of alfalfa and ryegrass(Table1).Tolerance of plant roots to the contaminants in soils is an essential criterion for the process of successful phytoremediation.The planted ryegrass and alfalfa in this study were thus identified as potential plants for phyto-remediation of pyrene contaminated soils.3.3.Pyrene contents of roots and aboveground parts of alfalfa and ryegrassAfter90d growth of alfalfa and ryegrass in quartz sand,red soil, and alluvial soil spiked with pyrene and amended with or without compost,the adsorption of pyrene at the root surfaces and the absorption of pyrene by roots of the two plants were not much pro-nounced.In addition,the amount of pyrene adsorbed at the root surfaces and absorbed by plant roots did not depend on plant spe-cies,type of soils,spiked pyrene concentrations of quartz sand and soils,and the amendment of compost(Table2).Nevertheless,some treatments showed that pyrene contents in the roots of alfalfa and ryegrass were significantly higher than those in the aboveground parts.This indicates that some pyrene was adsorbed at the root surfaces and absorbed by plant roots but not so much transported to the aboveground parts.Same trend of results have been reported by Gao et al.(2006).They studied the interactions of rice(Oryza sativa L.)and PAH-degrading bacteria(Acinetobacter sp.)on en-hanced dissipation of spiked pyrene in waterlogged soil and re-ported that pyrene concentrations in roots ranged from20to 90mg kgÀ1,while the concentrations in shoots were generally lower than0.2mg kgÀ1.Lu et al.(2010)also reported that there was a distinct difference in pyrene distribution between above-ground parts and roots of Bidens maximowicziana and that the roots had enrichment function to pyrene in certain degree,and the most amount of pyrene was probably deposited in roots and not trans-ferred to the shoots.They further elucidated that pyrene was hard to be transferred to the stems and leaves from the roots,or pyrene was decomposed in the process of transferring from roots toTable1Dry biomasses of roots and aboveground parts(g potÀ1)of alfalfa and ryegrass planted in quartz sand and soils spiked with pyrene and amended without or with compost at the end of90d growth period.aTreatment Quartz sand or soil Spiked pyrene concentration(mg kgÀ1)050100150Without compostAlfalfa Quartz sand 1.07GHb0.21Cc 1.08HGb0.29Dc 1.71CDa0.11Fc 2.17EFGa0.28Fc Red soil 2.39FGHb0.81Cc 2.8CDEab0.6Dc 2.84CDab0.56EFc 3.80CDEFa0.42FcAlluvial soil 2.93FGb0.84Cc 3.75CDb0.96CDc 3.95BCb0.98DEFc 5.58ABCa 1.34DecRyegrass Quartz sand0.56Hb 1.55Ca0.60Gab 1.10CDab0.57Db0.95DEFab0.83Gab0.72DEFab Red soil 3.68EFbc 4.64Ba 2.67DEFabcd 4.10Bab 2.15CDbcd 1.82Dcd 1.67FGcd 1.48DdAlluvial soil 5.62DEa 1.88Cb 5.84Ba 2.46Cb 6.84Aa 2.73Cb7.49ABa 2.94CbWith compostAlfalfa Quartz sand 5.22Ea 1.48Cc 4.52BCab 1.15CDc 2.96CDabc0.95DEFc 2.59DEFGbc0.53Fc Red soil7.61CDa 1.66Cc 6.01Bab0.85CDc 5.73ABab0.76EFc 5.01BCDb0.55EFcAlluvial soil10.32Ba 1.52Cc9.12Aab 1.51CDc7.41Ab 1.47Dec 6.95ABb 1.34DEcRyegrass Quartz sand 1.87FGHa 1.47Cab 1.83EFGa 1.44CDab 1.09CDab 1.11DEab0.83Gb0.89DEFb Red soil8.44BCa7.41Aabc7.67Aab 4.99ABab 5.88ABabc 4.68Bab 4.26CDEc 4.56BabAlluvial soil14.85Aa7.06Abcd8.99Ab 6.11Acd8.02Abc 5.51Ad7.59Abcd 5.33Ada Thefirst and second data of a data set in a row are the means of masses of roots and aboveground parts,respectively,of alfalfa and ryegrass.Different lower case and upper case letter following the data in a row and in a column means significant differences among the data by least significant difference at5%level.220M.C.Wang et al./Chemosphere87(2012)217–225shoots.As a result,the pyrene in shoots was comparatively lower than in roots.We thus suggested that the huge molecule of pyrene with the structure of four fused benzene rings make it difficultly to be absorbed by the roots and subsequently transported to the aboveground parts of the plants.Ke et al.(2003)investigated the potential of two mangrove plant species,Kandelia candel and Bru-guiera gymnorrhiza in wetland systems to remove pyrene from sur-face-or bottom-contaminated sediments.At the end of6-month treatment,significant accumulation of pyrene in roots was only found in microcosms having bottom-contaminated sediments, and pyrene concentrations were3.05and4.50mg kgÀ1in roots of K.candel and B.gymnorrhiza,respectively.They concluded that the overall contribution of root accumulation and plant uptake to the removal of pyrene from contaminated sediments was insignif-icant.As seen in Table2,the obtained data were comparable to those reported by Ke et al.(2003).We thus suggested that the insignificant overall contribution of root accumulation and plant uptake to the removal of pyrene from contaminated quartz sand and the two soils be probably because of the macromolecular size of pyrene,which was difficultly absorbed by the roots of alfalfa and ryegrass.Gao et al.(2011)studied arbuscular mycorrhizal phyto-remediation for soils contaminated with phenanthrene and pyr-ene.They reported that mycorrizal colonization caused increased accumulation of PAHs in alfalfa roots but a decrease in shoot.How-ever,plant uptake contributed negligibly to PAH dissipation in arbuscular mycorrhizal phytoremediation,and alfalfa accumulated PAHs amounted to less than3.24%of total PAH degradation in mycorrhizal soils.Zhu and Zhang(2008)pointed out that organic pollutants in plants depend primarily on the rate of uptake,the metabolism of these chemicals in plant tissues,and plant growth. PAHs can be catabolized by plant enzymes,either mineralized completely to inorganic compounds(e.g.,CO2,H2O),or degraded partially to a stable intermediate that is stored in the plants.Con-sequently,in this study some stable intermediate derived from pyrene catabolized by plant enzymes could occur and store in the roots and/or aboveground parts of ryegrass and alfalfa.3.4.Microbial populations in rhizosphere quartz sand and soilsAs seen from Table3,the microbial population in rhizosphere alluvial soil in the pots amended with compost and planted with ryegrass at spiked pyrene concentration up to150mg kgÀ1,was significantly larger than the corresponding treatment amended without compost.Significant difference in microbial population was also found for the same treatment but planted with alfalfa at spiked pyrene concentration50–150mg kgÀ1.The treatment amended with compost and simultaneously planted with alfalfa or ryegrass in alluvial soil showed that microbial population in-creased with increasing spiked pyrene concentration.This may be attributed to inorganic and organic nutrients provided by the decomposition of compost,promoting the growth of alfalfa and ryegrass(Table1),which their root exudates may thus stimulate the growth of microorganisms in rhizosphere alluvial soil(Table 3).Günther et al.(1996)studied the effect of growing ryegrass on the biodegradation of hydrocarbons in laboratory scale soil col-umns.They pointed out that in the rhizosphere soil system,ali-phatic hydrocarbons disappeared faster than in unvegetated columns.Elimination of contaminants was accompanied by an in-crease in microbial numbers and activities and stimulated by plant roots.In the review paper on rhizosphere carbonflow in trees, Grayston et al.(1996)concluded that a key factor governing micro-bial growth and activity in soils is carbon availability.In our exper-iment,the decomposition of amended compost and plant roots exudates mostly contributed to the available carbon in alluvial soil. The root exudates cause rhizosphere-inhabiting microbial popula-tions to increase well beyond those of the bulk soil(Jones,1998), attracting motile bacteria and fungal hyphae that stimulate an ar-ray of positive,neutral,or negative interactions with plants(Gerh-ardson,2002).In addition,the organic and inorganic nutrients of amended compost associated with root exudates in rhizosphere alluvial soil may also be utilized by microorganisms to stimulate microbial propagation and activity.Binet et al.(2000)investigated the dissipation of a mixture of eight PAHs,ranging from three to six rings,in the rhizosphere of ryegrass.They concluded that the in-creased PAH dissipation in rhizospheric soil was associated with an enhancement of PAH degraders.In our experiment,we found that microbial population in rhizosphere alluvial soil planted with ryegrass and amended with compost were significantly higher than that in the corresponding treatment but unplanted with rye-grass at the four spiked pyrene concentrations(Table3).Both with-out and with compost amendment,the treatments of unplanted and planted with alfalfa and ryegrass in the pots of quartz sand, red soil,and alluvial soil spiked with pyrene at the concentrations of50–150mg kgÀ1,the microbial population in rhizosphere quartzTable2Pyrene contents(mg kgÀ1roots or aboveground parts)of alfalfa and ryegrass planted in quartz sand and soils spiked with pyrene and amended without or with compost.a.Treatment Spiked pyrene concentration(mg kgÀ1quartz sand or soil)050100150AlfalfaQuartz sand ND b ND ND ND0.69BCa0.48Da0.36Da ND ND ND ND ND0.48Ca ND0.50Da0.43BaRed soil ND ND 4.17Aa 1.99Ab 1.50Abc0.97BCcd 1.19BCc0.43Bd ND ND0.80Cab0.57BCDab0.84ABCab0.60CDb 1.48ABa 1.38AbAlluvial soil ND ND 1.02Ca0.94Ba 1.29ABa0.99Ba0.81CDa0.77Ba ND ND 1.90Ba ND 1.44Aa 1.37Aa 1.83Aa0.64BbRyegrassQuartz sand ND ND0.44Cc0.38Dc0.48Cc0.41Dc0.79CDa0.69Bab ND ND0.46Ca ND0.34Cb0.38Dab0.38Dab0.31BbRed soil ND ND0.49Ca0.39CDa0.52Ca0.36Db0.54Da0.52Ba ND ND0.65Cab0.64BCab0.85ABCab0.60CDb0.91BCDa0.73BabAlluvial soil ND ND0.55Cc ND0.97ABCab0.74BCDbc 1.34ABCa0.43Bd ND ND 1.08Ca ND0.88ABCab0.33Dd 1.14BCa0.47Bbca Within the same row of thefirst and second data of each data set are the means of pyrene contents in roots and aboveground parts,respectively.Within the same column of thefirst and second data of each data set are the means of pyrene contents under the treatment amended without or with compost,respectively.Different lower case and upper case letters following the data in a row and in a column,respectively,means significant differences among the data by least significant difference at5%level.b ND indicates that pyrene content in digestion solution was lower than the detection limit of instrument.M.C.Wang et al./Chemosphere87(2012)217–225221。