c ○ 2000 Kluwer Academic Publishers. Manufactured in The Netherlands. Meaningful Alignment
- 格式:pdf
- 大小:638.81 KB
- 文档页数:17
104105Materials and methodsGrowth of plant materialChara corallina is a freshwater alga with single giant internodal cells which can be typically8cm long.The internodal cells used in the study were collected from established cultures growing in indoor tanks.Inter-nodal cells were cut from neighbouring cells,washed in distilled water and bathed in artificial pond wa-ter(APW:0.2m M K2SO4,1.0m M NaCl,1.0m M CaSO4,5.0m M Mes-NaOH,pH6.0)prior to experi-mental treatment.Internodes were transferred to APW solutions containing either0.5m M or1m M NH4+(as (NH4+)2SO4)for a minimum of two days before the measurements of intracellular NH4+.Chara sap analysisSingle Chara internodal cells were transferred from the nutrient solution to a glass microscope slide and one end was cut using a scalpel allowing the sap to exude onto the slide.A known volume(3–9µL)of vacuolar sap was collected using a Microcap(Drum-mond Scientific Co.,Broomall,PA)and immediately frozen in liquid nitrogen for storage.Stored samples were defrosted and analysed for NH4+using a Dionex Model DX500HPLC,fitted with an Ionpac Fast Cation1column(Dionex,Camberley,UK)and detected by non-suppressed conductivity.NH4+selective electrode measurementsDouble-barrelled micropipettes were prepared and sil-anized as described previously(Miller and Zhen, 1991).The composition of the ammonium sensor cocktail was10%(w/w)nonactin(ammonium iono-phore I,Fluka09877),1%potassium tetrakis(4-chlorophenyl)borate and89%dibutyl sebacate.Fi-nally,20%high molecular weight polyvinyl chloride was added to this cocktail and the mixture dissolved in5volumes of tetrahydrofuran(THF).All chem-icals were of analytical grade and were purchased from Fluka,except THF(Sigma).After a minimum of48h at room temperature in silica-dried air,the ammonium-selective membrane had gelled in the tip and each barrel was backfilled with electrolyte solu-tion.The ammonium-selective barrel was backfilled with100m M NH4Cl using a29-G needle attached to a1mL disposable syringe.The reference barrel was backfilled with200m M NaCl using a Microfil needle (World Precision Instruments(WPI),Sarasota,USA).The calibration and intracellular measurements were made using a high-impedance differential electrometer (model FD223,WPI)which was connected to a PC as described previously(Walker et al.,1995).The circuit was completed with a bath electrode(Flexref,WPI) which was placed in the APW bathing the Chara cells.The NH4+barrel was calibrated using solutions containing a constant background activity of K+(a K) of72m M and buffered at pH6.0.For intracellular recordings,Chara internodal cells were clamped in a 5mL perspex chamber and perfused with APW solu-tion.Microelectrodes were calibrated for NH4+before and after intracellular recordings and the combined calibration data were used to produce a curve from which the intracellular a NH4was calculated. ResultsIn the presence of background cytosolic a K(Walker et al.,1996),the electrodes can detect NH4+activ-ities in the m M range(Figure1A).Calibrating these microelectrodes in this background gave a mean slope ±sd of55.6±5.0mV(sample size,n=47)per ten fold change in concentration and a detection limit of9.4m M(IUPAC,1994).However,when several calib-ration points are obtained around the detection limit, the practical use of the electrodes can be extended to around1m M by using the non-linear portion of the calibration curve(see Figure1A).In Chara cells growing in APW containing0.5m M NH4+,the electrode measurements were almost at the limit of detection(Figure1B).The electrodes were more successful when cells were incubated in APW supplemented with1m M NH4+.Intracellular meas-urements of a NH4in Chara cells were in the range 6–39m M and these measurements could be separ-ated into two populations(Figure1B).However,the membrane potential measurements obtained with each a NH4value did not separate into two populations(data not shown).For cells placed in the higher NH4+con-centration,HPLC analysis of whole cell sap samples gave a mean value±sd for NH4+concentration of 25.1±1.3(n=4)m M.DiscussionWe have developed an ion-selective microelectrode, based on the sensor nonactin,that allows the direct measurement of intracellular a NH4in the millimolar106range.In the cytosol of plant cells,these electrodes can measure intracellular a NH4down to1m e of the electrodes on cells of Chara incubated in APW supplemented with1m M NH4+,gave two populations of measurements(Figure1B),with means of7.3and 30.8m M.It is probable that these represent the cytosol and vacuole,respectively.In Chara,the vacuole oc-cupies90–96%of the intracellular volume and the cytoplasm4–10%(see references in Miller and Zhen, 1991).This means that the whole-cell sap samples must be dominated by the vacuolar a NH4and thus the HPLC analysis will measure vacuolar paring the results of the sap sample analysis with the electrode measurements suggests that the higher population in Figure1B must represent vacu-olar impalements.Therefore,the cytosolic electrode measurements are the lower population with values around7m M.Further evidence for this assignment of these two populations comes from previous work measuring vacuolar concentrations of NH4in Chara growing in similar conditions which showed a mean vacuolar concentration of31m M(Ryan and Walker, 1994).Using14C-labelled methylammonium as an analogue for NH4+,these authors calculated the cyto-plasmic concentration to be18m M,but suggested that further compartmentation into acidic organelles may give2mm in the actual cytosol(Ryan and Walker, 1994).Another advantage of the electrode technique is that the measurement directly reports cytosolic a NH4 and so does not require such assumptions to be made.These are thefirst intracellular measurements of NH4+using NH4+-selective microelectrodes and they demonstrate the feasibility of the method.The dis-tribution of NH4+in Chara cells reveals a4-fold accumulation of ammonium in the vacuole compared to the cytosol.The future use of triple-barrelled mi-croelectrodes incorporating a pH-selective barrel will allow the unequivocal measurement of vacuolar and cytosolic NH4+activities.In these electrodes,the pH barrel allows identification of the compartment in which the electrode tip is located(Walker et al., 1995).Measurements of compartmental pH are also important as the equilibrium between NH3and NH4+ will be determined by this parameter.Finally,even though cytosolic K+activity is regulated in higher plants(Walker et al.,1996),it may be important to check with K+-selective microelectrodes that the cytosolic K+activity in Chara cells does not change when NH4+is added to the APW. AcknowledgementsWe wish to thank John Andralojc and David Carden for their help with the HPLC analysis.This work was funded by EU grant number BIO4-CT97-2310.IACR receives grant-aided support from the Biotechnology and Biological Sciences Research Council of the UK. ReferencesHenriksen G H,Bloom A J and Spanswick R M1990Measure-ment of netfluxes of ammonium and nitrate at the surface of barley roots using ion-selective microelectrodes.Plant Physiol.93,271–280.IUPAC1994Recommendations for the nomenclature of ion-selective electrodes(IUPAC recommendations1994).Pure& Appl.Chem.66,2528–2536.Kronzucker H J,Siddiqi M Y and Glass A D M1995Compartment-ation andflux characteristics of ammonium in spruce.Planta196, 691–698.Miller A J and Zhen R-G.1991Measurement of intracellular nitrate concentrations in Chara using nitrate-selective microelectrodes.Planta184,47–52.Ryan P R and Walker N A1994The regulation of ammonia uptake in Chara australis.J Exp.Bot.45,1057–1067.Scholer R P and Simon W1970Antibiotika-Membranelektrode zur selektiven Erfassung von Ammoniumionenaktivitäten.Chimia 24,372–374.Tyerman S,Whitehead L F and Day D A1995A channel-like trans-porter for NH4+on the symbiotic interface of N2-fixing plants.Nature378,629–632.Walker D J,Smith S J and Miller A J1995Simultaneous meas-urement of intracellular pH and K+or NO3−in barley root cells using triple-barreled,ion-selective microelectrodes.Plant Physiol.108,743–751.Walker D J,Leigh R A and Miller A J1996Potassium homeostasis in vacuolate plant A93,10510–10514.Wang M Y,Siddiqi M Y,Ruth T J and Glass A D M1993 Ammonium uptake by rice roots.I.Fluxes and subcellular distribution of13NH4+.Plant Physiol.103,1249–1258.。
外文期刊论文格式【篇一:外文期刊的论文要求格式】your papers title starts here: please centeruse helvetica (arial) 14full first author1, a, full second author2,b and last author3,c12full address of first author, including country3full address of second author, including country list all distinct addresses in the same wayaemail, bemail, cemailkeywords: list the keywords covered in your paper. these keywords will also be used by the publisher to produce a keyword index.for the rest of the paper, please use times roman (times new roman) 12abstract. this template explains and demonstrates how to prepare your camera-ready paper for trans tech publications. the best is to read these instructions and follow the outline of this text. please make the page settings of your word processor to a4 format (21 x 29,7 cm or 8 x 11 inches); with the margins: bottom 1.5 cm (0.59 in) and top 2.5 cm (0.98 in),right/left margins must be 2 cm (0.78 in). (we shall be able to publish your paper in electronic form on our web page , if the paper format and the margins are correct. if not, we will have to scan your paper which, when compared with an electronic version, results in very poor quality) your manuscript will be reduced by approximately 20% by the publisher. please keep this in mind when designing your figures and tables etc.introductionall manuscripts must be in english, also the table and figure texts, otherwise we cannot publish your paper.please keep a second copy of your manuscript in your office. when receiving the paper, we assume that the corresponding authors grant us the copyright to use the paper for the book or journal in question. should authors use tables or figures from other publications, they must ask the corresponding publishers to grant them the right to publish this material in their paper.use italic for emphasizing a word or phrase. do not use boldface typing or capital letters except for section headings (cf. remarks on section headings, below).organization of the textsection headings. the section headings are in boldface capital and lowercase letters. second level headings are typed as part of the succeeding paragraph (like the subsection heading of this paragraph). page numbers. do not number your paper:tables. tables (refer with: table 1, table 2, ...) should be presented as part of the text, but in such a way as to avoid confusion with the text. a descriptive title should be placed above each table. units in tables should be given in square brackets [mev]. if square brackets are not available, use curly {mev} or standard brackets (mev).macros. do not use any macros for the figures and tables. (we will not be able to convert such papers into our system)language. all text, figures and tables must be in english. figures. figures (refer with: fig. 1, fig. 2, ...) also should be presented as part of the text, leaving enough space so that the caption will not be confused with the text. the caption should be self-contained and placed below or beside the figure. generally, only original drawings or photographic reproductions are acceptable. only very good photocopies are acceptable. utmost care must be taken to insert the figures in correct alignment with the text. half-tone pictures should be in the form of glossy prints. if possible, please include your figures as graphic images in the electronic version. for best quality the pictures should have a resolution of 300 dpi(dots per inch).color figures are welcome for the online version of the journal. generally, these figures will be reduced to black and white for the print version. the author should indicate on the checklist if he wishes to have them printed in full color and make the necessary payments in advance.equations. equations (refer with: eq. 1, eq. 2, ...) should be indented 5 mm (0.2). there should be one line of space above the equation and one line of space below it before the text continues. the equations have to be numbered sequentially, and the number put in parentheses at the right-hand edge of the text. equations should be punctuated as if they were anordinary part of the text. punctuation appears after the equation but before the equation number, e.g.c2 = a2 + b2. (1) literature referencesreferences are cited in the text just by square brackets [1]. (if square brackets are not available, slashes may be used instead, e.g. /2/.) two or more references at a time may be putin one set of brackets [3,4]. the references are to be numbered in the order in which they are cited in the text and are to be listed at the end of the contribution under a heading references, see our example below. summaryif you follow the “checklist” your paper will conform to the requirements of the publisher and facilitate a problem-free publication process.references[1] dj.m. maric, p.f. meier and s.k. estreicher: mater. sci. forum vol. 83-87 (1992), p. 119[2] m.a. green: high efficiency silicon solar cells (trans tech publications, switzerland 1987).[3] y. mishing, in: diffusion processes in advanced technological materials, edtied by d. guptanoyes publications/william andrew publising, norwich, ny (2004), in press.chemistry, edited by s.d. schwartz, volume 5 of progress in theoretical chemistry and physics, chapter, 10, kluwer academic publishers (2000).[5] r.j. ong, j.t. dawley and p.g. clem: submitted to journal of materials research (2003)从这里开始你的论文的标题:请中心使用helvetica字体(宋体)14全第一author1,整个第二author2,b和最后author31full地址的第一作者,包括国家第二作者2full地址,包括国家3list所有不同的地址,以同样的方式aemail,bemail,cemail关键词:列出你的论文中所涉及的关键字。
ASSESSMENT OF THE HEA VY METAL POLLUTION EFFECTS ON THE SOIL RESPIRATION IN THE BAIX LLOBREGAT(CATALONIA,NE SPAIN)DANUTA ZIMAKOWSKA-GNOI´NSKA1,JAUME BECH2and FRANCISCO J.TOBIAS21International Centre of Ecology PAS,Dziekanów Le´s ny n.Warsaw,05-092Łomianki,Poland 2Catedra de Pedologia,Facultat de Biologia,Universitat de Barcelona,Spain(Received21July1998;accepted13January1999)Abstract.A main goal of investigations is to determine could a soil respiration be an indicator of the soil pollution.In this case a measured level of the soil oxygen of its pollution.It also means that the pollution reduces biological processes soil samples were taken from polluted and non-polluted places in the Baix(Catalonia, NE Spain).Soil samples were taken from the top of soil(0–5cm)without a litter.Soil analysis were done,determining percentage shares of coarse fragments,coarse sand,fine sand,coarse silt,fine silt, clay,CaCO3,organic matter as well as water pH and conductivity CE(1:5[mS cm−1]).Also were determined(in mg kg−1)quantities of heavy metals,as Fe,Al,Mn,Zn,Cr,Ni,V,Cu,Cd,Pb.The soil respiration was investigated in temperatures15and30◦C and with controlled humidity.The respiration in30◦C is number of times greater then in15◦C both for polluted and non-polluted soils.Particularly high coefficients of correlation between the soil respiration and soil pollution in polluted soils were obtained for Pb:r=0.75in15◦C and r=0.98in30◦C;for Ba:0.90and0.57;for V:0.99and0.81.In non-polluted soils highest correlation coefficients are for Pb:r=0.70in15◦C; Fe:0.60and0.72;Al:0.68and0.64;Mn:0.51and0.66;Ba:0.63and0.61;Cr:0.94and0.70;Ni: 0.64and0.65;Cu:0.69and0.48;as well as V:0.62in15◦C;and Cd:0.69in15◦C.This way the soil respiration could be a good indicator of the soil pollution.Keywords:biological activity of soil,constant-pressure volumetric respirometer,heavy metals,respirom-etry methods,soil,soil degradation1.IntroductionThe aim of this study is to try correlating the effect of metal pollution and the respiration of soil in two areas–one polluted and a neighbour non-polluted area.An unfavourable influence of heavy metals on soil environment is well known. Many autors indicate their particularly destructive influence on soil microflora (Strojan,1975;Freedman and Hutchinson,1980).An activity of soil microbial population is also influenced by heavy metals as well as an intensity of biological processes in edaphon(Nordgren et al.,1983,1985,1986).Scientists are looking for relatively sensitive indicator of degradating processesin soil environment.An intensity of biological processes is generally estimated by quantity of ejected carbon dioxide(Witkamp,1969;Edwards,1974;Elmholt,1992;Environmental Monitoring and Assessment61:301–313,2000.©2000Kluwer Academic Publishers.Printed in the Netherlands.302 D.ZIMAKOWSKA-GNOI´NSKA ET AL.Gildon et al.,1993;Nakadai et al.,1993).This indicator covers both metabolisms –aerobic and anaerobic.Investigation of oxygen respiration could give more sen-sitive and explicit indication.Investigations of Fischer(1996)show that measured changes of the carbon dioxide ejection are less sensitive to changes of temperature and humidity of the soil as well as to changes of the soil pollution.Probably,the carbon dioxide ejection is a result of two simultanous processes–aerobic and anaerobic.The oxygen intake is connected only with aerobic process.In presented paper a respirometry method was applied to oxygen respiration measurements of polluted and non-polluted soils from the Barcelona area.Results have to show an influence of pollution on biological activity of soils. These results are also influenced by other physical and biological characteristics of investigated soil sample.The main objective of said study is to prove that res-piration of soil depends of soil pollution and moreover indicates the extent of the pollution harmfulness on soil biological processes.2.Area of Studies,Climatic Conditions and VegetationStudied areas are located in the Baix Llobregat,about20km SW of Barcelona and inside of the metropolitan area of the city.For purposes of this paper two areas were taken into account.Thefirst one,named‘non-polluted area’(Figure1),belongs to mountainous side of Gavá,Viladecans and Torelles de Llobregat.This area can be considered as non-polluted area because it is forest zone with relatively low anthropogenic impact.The second one,named‘polluted area’(Figure2),is located SW of Gaváand Viladecans and represents degradated part of metropolitan area.The climate of area is typically Mediterranean with mean annual temperature of 16◦C and mean annual rainfall(mostly in spring and in autum)of650mm.Spring and autumn precipitation are135and240mm,respectively;summer precipitation is only105mm.The summer is hot and dry with mean temperature over20◦C from June to September.The mean January temperature is9±10◦C.According to the Koppen climatic classification,it is‘Csa’;pedoclimic regimes are xeric(SMR) and thermic(STR).A main vegetation in the non-polluted area are low‘garriga’scrubs,Quercus coccifera,Pistacia lentiscus,Olea europea,Chamaerops humilis,Rhamnus ly-cioides,Rosmarinus officinalis,Erica multiflora,Ampelodesma mauritanica,Cer-atonia siliqua,Fumana laevipes,frequently accompanied by Pinus halepensis.In degradated area are herbes such as Hyparrhenia hirta,Sonchus tenerrimus, Brachypodium ramosum,Brachypodium phoenicoides,Foeniculum vulgare,Pso-ralea bituminosa,Inula viscosa etc.HEA VY METAL POLLUTION EFFECTS ON THE SOIL RESPIRATION303Figure1.The sampling points of the non polluted area are indicated.Number2shows the location of the polluted area.304 D.ZIMAKOWSKA-GNOI´NSKA ET AL.Figure2.Sampling points in the polluted area:enlarged view(A:composting plant,B:used cars dump).3.Material and Methods3.1.M ATERIALWere studied22topsoil samples(0–5cm,without litter).Eleven samples were taken from non-polluted area(Figure1)and other eleven from polluted area(Fig-ure2).Those from polluted area were taken from three transects–two near a composting plant and one near an used car dump.3.2.A NALYTICAL METHODSSoil samples were air-dried and passed through a2mm sieve to obtain a‘fine earth’(fraction less then2mm).Sieving has also to eliminate plant roots and other plant parts.In this‘fine earth’shares of coarse fractions(>0.2mm)andfine fractions (<0.2mm)were determined for sands and for silts.In the‘fine earth’the organic carbon content was determined using the Walkley-Black dichromate acid oxidation method.Soil pH values were determined for soil solution suspensions in distilled water and in the1N KCl.This determination was done potentiometrically with a glass electrode.Calcium carbonate contents were determined using the acid neutralization method.Particle-size distribution was determined using the pipette method and by sieving sand fractions.Conduc-HEA VY METAL POLLUTION EFFECTS ON THE SOIL RESPIRATION305 timetry was determined in1:5soil to water extract.Total nitrogen contents were determined using the Carlo Erba1500elemental analyzer.The‘fine earth’was grounded in a tungsten carbide mill,dried at105◦C and then used to determine metal contents.Soil samples(3g)were extracted with28ml of aqua regia(HNO3to HCl as1:3)during two hours at130◦C after a predigestion of16hr at a room temperature.Metal concentrations were measured infiltered extract,using the Polyscan61E spectrometer.3.3.S OIL RESPIRATIONSoil respirations were determined in laboratory because in this case experimental conditions are more stabile than in the case offield measurements.Soil samples were taken from the top of soil(0–5cm)without a litter,trans-ported to laboratory and stored there in a cooler in temperature3◦C.Soil samples were sieved to eliminate plant roots,as recommends Anderson(1982).Soil respirations were measured using constant pressure volumetric respirime-ters,modified by Klekowski(1975).Respirometers were placed in water ther-mostate with temperature controlled with±0.1◦C precision.Experiments were done in two temperatures:15and30◦C.In respirometric vessels were placed30g samples of soil.Samples were moist-ened up to60%of humidity adding sufficient quantity of boiled cold water.Moist-ened samples were adapted during12hr to the temperature of experiments.Mea-surements of oxygen consumption were done for the next3hr.After respiration and after drying in temperature55◦C a dry weight of sample was weighted.An humidity of sample was determined from the ratio of dry weight to wet weight of sample.A share of organic matter in the soil sample was determined by combustion in a muffle oven in temperature450◦C.Results obtained when using this method of determination are more applicable to biological processes in the investigated soil.The oxygen consumption for a sample30g of the wet soil is expressed in mm3O2·h−1and per gram of the dry soil or per gram of the organic matter in mm3O2·h−1·g−1.Mean values and standard deviations were calculated for all experiments.The aim of this preliminary study is tofind simple correlations between heavy metal pollution and soil respiration in two areas:polluted and non-polluted.4.Results and DiscussionThe increase of antropogenic heavy metals in soils is significant on outskirts of industrial zones like the Metropolitan Area of Barcelona.In Table I are compared physico-chemical characteristics of sampled soils.In Table II are shown shares of heavy metals in the same soil samples.In polluted area in the Llobregat delta plane306 D.ZIMAKOWSKA-GNOI´NSKA ET AL.TABLE IPhysico-chemical data of soil samplesNo.Coarse Coarse Fine Coarse Fine Clay Organic pH N CaCO3CE1:5 fragments sand sand silt silt matter H2O(%)(%)(%)(%)(%)(%)(%)(%)(%)(mS cm−1)1 3.6254.7122.53 4.407.5010.85 3.827.290.1025.580.23029.8922.0220.0215.6622.4819.83 1.887.340.0923.400.5843 3.5735.6010.0234.430.0019.97 3.597.120.2417.850.764419.6142.8611.719.8225.4410.17 2.907.230.1116.650.804565.1635.8815.677.0721.3120.078.037.260.7218.82 2.368619.7144.8812.267.1317.4018.32 4.167.540.2115.200.240722.2430.8216.2012.6319.8520.51 6.097.530.3118.820.303833.5746.4212.287.8919.2514.169.177.450.1529.680.347937.3634.7014.5412.8719.6218.27 4.277.270.1623.890.27710 6.3318.4110.758.1739.3823.29 6.557.300.2831.130.551 1119.8217.988.817.5640.3925.279.637.210.4329.440.653 1232.1617.55 6.5612.9338.7724.19 3.707.260.197.850.517 1352.0615.00 6.3214.1342.1122.44 3.137.360.17 2.500.103 1465.5522.52 6.3911.0339.5920.4714.537.200.60 3.450.213 1511.1340.3441.50 4.44 6.277.45 6.097.060.23 2.610.104 1646.5816.07 4.63 5.3137.6936.307.58 6.470.49 4.040.135 1715.7014.128.7210.7931.7334.64 4.967.180.1847.740.157 1813.667.8169.649.18 5.787.59 3.13 6.920.09 1.550.089 1913.527.4238.6516.4918.8118.62 6.557.370.2411.430.204 2019.6115.0245.407.9013.5918.09 6.217.480.1811.540.231 2140.2826.0916.2610.1124.5023.0412.827.070.38 5.590.319 2250.5514.94 5.447.4931.4340.718.157.300.22 6.880.183main sources of pollution are composting plant and the used car dump.The non-polluted area is nearby hilly forest area.Point16placed in this area is an anomaly –its soil contains natural high share of metals.These tables confirm that the polluted area is really a polluted one and concen-tration of heavy metals grows when approaching to the plant and car dump.In paper Bech et al.(1994)were presented correlations between physico-chemical characteristics and concentrations of heavy metals.For example,conductivity(CE 1:5)positively correlates with concentration of heavy metals in the contaminated area,but not in noncontaminated area.Mean conductivity in polluted area is0.65mS cm−1.In extremely polluted point No.5conductivity is even2.37mS cm−1.HEA VY METAL POLLUTION EFFECTS ON THE SOIL RESPIRATION307TABLE IIShares of heavy metals in soil samplesNo.%Fe%Al Mn Zn Cr Ni V Cu Cd Pb(ppm)1 1.56 1.35445.86197.0120.6616.1297.5871.930.71166.762206 2.24455.38122.1327.1920.69100.2833.040.4462.723 2.88 2.57575.72355.8050.0938.01128.49131.18 1.84241.304 3.12 2.34497.67898.4344.5628.79114.91196.33 4.46761.21513.64 2.961395.843770.36705.91130.17128.121093.7411.911970.656 1.98 1.78 4.38166.6324.0821.0298.9035.390.7465.067 3.08 2.24549.01537.1039.9335.21102.25112.85 1.30205.738 3.54 1.94610.201573.7155.3463.7383.16216.29 1.88458.669 2.32 2.36501.02391.9932.2024.13117.3858.310.76149.1710 2.22 2.48512.89199.0627.7622.89104.3745.200.4963.2911 2.15 2.48488.32247.5227.8922.96111.1629.140.9648.7212 4.11 3.10813.0877.6432.8437.7593.1025.200.3739.0813 4.60 3.68705.9484.1942.0544.21106.0327.270.3041.9714 3.47 2.79978.56214.3343.0638.25100.4033.090.5885.50 150.710.72103.4347.4511.608.7765.98 4.720.3028.04 1610.77 4.3313584.39111.4556.72111.15169.70129.82 1.2974.45 17 1.80 2.77426.8292.7127.9319.30117.8411.480.4740.94 180.870.98522.3323.2019.438.8879.07 3.980.1220.2619 1.79 2.20805.0652.0323.2519.9587.4015.550.3027.0020 1.68 1.851153.5742.6521.5915.1498.0310.020.3026.8621 3.64 1.74786.78128.3721.5036.3890.3358.600.6977.7122 4.31 4.16256.9178.3341.0442.74116.7225.17 1.0850.99In polluted area is observed considerable concentration of Fe,Mn,Zn,Be,Cr, Ni,V,Cu,Cd and Pb.In non-polluted area concentration of heavy metals is much lower and mean conductivity is only0.20mS cm−1.Share of organic matter decreases with increased concentration of heavy metals and increased salinization of soils.In the non-polluted area,natural Fe,Al,Zn,Cr,Ni,V,Cd and Pb show correla-tions with size fractions(negative correlations withfine sand and positive withfine silt and clay).In both areas metal contents generally correlate with each other:–in the non-polluted area these correlations are due to similar pedogonic processes and also to similar parent materials.308 D.ZIMAKOWSKA-GNOI´NSKA ET AL.Figure3.Oxygen consumptions of soil samples from contaminated areas:=15◦C, =30◦C.HEA VY METAL POLLUTION EFFECTS ON THE SOIL RESPIRATION309Figure4.Oxygen consumptions of soil samples from contaminated areas:=15◦C, =30◦C.310 D.ZIMAKOWSKA-GNOI´NSKA ET AL.–in the polluted area these correlations are due to contaminant processes which lead to a proportional increase of metal concentrations.In Figure3are presented oxygen consumptions of soil samples from contaminated areas and in Figure4oxygen consumptions of soil samples from noncontaminated areas.In both cases results are given for two temperatures of experiment:15and 30◦C.Was observed generally lower oxygen consumption in soil samples from con-taminated areas in comparison with oxygen consumption in soil samples from noncontaminated areas.In the extremely contaminated point No.5oxygen con-sumption has the lowest value.In this point concentration of heavy metals is con-siderably higher then in other points and also there is an extremal salinization (conductivity2.37mS cm−1compared with mean values:0.65in contaminated area and0.20in noncontaminated).In contaminated area a share of the organic matter is significantly lower,a mean value for all points in this area is only9.18%,when in noncontaminated area a mean value is12.77%.It could be a result of unfavourable influence of heavy metals and their salts on biological processes in edaphon,mostly on soil microfauna and soil microbial populations.In Table III are presented correlations between measured soil respirations and components of soil samples from contaminated area.In Table IV are presented correlations between measured soil respirations and components of soil samples from noncontaminated area.Particularly high coefficients of correlation between the soil respiration and soil pollution in polluted soils were obtained for Pb:r=0.75in15◦and r=0.98in 30◦C,for Ba:0.90and0.57;for V:0.99and0.81.In non-polluted soils highest correlation coefficients are for Pb:r=0.70in15◦C;Fe:0.60and0.72;Al:0.68 and0.64;Mn:0.51and0.66;Ba:0.63and0.61;Cr:0.94and0.70;Ni:0.64and 0.65;Cu:0.69and0.48;V:0.62in15◦C;and Cd:0.69in15◦C.Obtained results attest that measurements of soil respirations using volumetric respirometry could be used for estimations and comparations of soil ecological conditions as well as biological activity of soils.5.Conclusions1.Obtained results of experiments and analysis confirm that some investigatedsoils in the Baix Llobregat near Barcelona are contaminated by heavy metals.Concentration of pollutants increases in vinicity of a composting plant and a car dump.2.In contaminated soils conductivities are three times higher than in noncontam-inated soils.Soil conductivity depends in high degree on salinization of soils by dissociated heavy metal salts.HEA VY METAL POLLUTION EFFECTS ON THE SOIL RESPIRATION311TABLE IIICorrelations between measured soil respirationand components of this soil from contaminatedarea15◦C30◦CSoil humidity–0.326∗–0.471∗0.328∗∗0.144∗∗Organic matter–0.543∗–0.385∗0.084∗∗0.243∗∗CaCO3–0.544∗–0.334∗0.083∗∗0.315∗∗CE–0.493∗–0.528∗0.123∗∗0.095∗∗Fe–0.445∗–0.382∗0.170∗∗0.247∗∗Al–0.432∗–0.341∗0.185∗∗0.305∗∗Mn–0.532∗–0.442∗0.092∗∗0.173∗∗Sr–0.546∗–0.446∗0.082∗∗0.169∗∗Zn–0.476∗–0.304∗0.139∗∗0.364∗∗Ba0.044∗0.193∗0.897∗∗0.570∗∗Cr–0.418∗–0.399∗0.201∗∗0.224∗∗Ni–0.517∗–0.368∗0.104∗∗0.266∗∗V–0.003∗–0.084∗0.992∗∗0.806∗∗Cu–0.431∗–0.359∗0.186∗∗0.278∗∗Cd–0.352∗–0.295∗0.288∗∗0.379∗∗Pb–0.109∗0.010∗0.749∗∗0.976∗∗∗:r=Correlation Coefficient.∗∗:Significance level p<0.05.312 D.ZIMAKOWSKA-GNOI´NSKA ET AL.TABLE IVCorrelations between measured soil respirationand components of this soil from noncontami-nated area15◦C30◦CSoil humidity–0.325∗–0.569∗0.330∗∗0.068∗∗Organic matter–0.515∗–0.708∗0.105∗∗0.015∗∗CaCO3–0.655∗–0.515∗0.029∗∗0.105∗∗CE0.313∗–0.192∗0.349∗∗0.572∗∗Fe0.179∗–0.122∗0.599∗∗0.721∗∗Al–0.141∗–0.161∗0.679∗∗0.637∗∗Mn0.221∗–0.152∗0.513∗∗0.657∗∗Sr–0.595∗–0.606∗0.053∗∗0.048∗∗Zn–0.138∗–0.542∗0.685∗∗0.085∗∗Ba0.162∗–0.175∗0.633∗∗0.607∗∗Cr0.025∗–0.132∗0.941∗∗0.699∗∗Ni0.159∗–0.153∗0.641∗∗0.653∗∗V–0.170∗–0.337∗0.617∗∗0.311∗∗Cu0.135∗–0.238∗0.692∗∗0.482∗∗Cd–0.137∗–0.474∗0.690∗∗0.141∗∗Pb–0.133∗0.492∗0.697∗∗0.124∗∗∗:r=Correlation Coefficient.∗∗:Significance level p<0.05.HEA VY METAL POLLUTION EFFECTS ON THE SOIL RESPIRATION313 3.In contaminated area a share of the organic matter is significantly lower(9.18%)then in noncontaminated area(12.77%).It could be a result of unfavourable influence of heavy metals and their salts on biological processes in edaphon, mostly on soil microfauna and soil microbial populations.4.Oxygen consumptions are considerably lower in soil samples from contami-nated areas in comparation with noncontaminated areas.5.V olumetric respirometry could be used for estimations and comparations ofecological condition and biological activity of soils.ReferencesAnderson,J.P.E.:1982,‘Soil Respiration’,in:Page,A.L.,Miller,R.H.and Keeney,D.R.(eds.),Methods of Soil Analysis:2Chemical and Microbiological Properties,Am.Soc.Agron., Madison,Wisconsin,pp.467–476.Bech,J.,Tobias,F.J.and Zimakowska-Gnoi´n ska,D.:1994,‘Comparison of Heavy Metals in pol-luted and unpolluted soils at the NE edge of the Garraf Massif(Catalonia,Spain)’,Abstracts of the15th International Congress of Soil Science,Acapulco,Mexico,152–153.Edwards,N.T.:1974,‘A moving chamber design for measuring soil respiration rates’,OIKOS25, 97–101.Elmholt,S.:1992,‘Effect of Propiconazole on Substrate Amended Soil Respiration Following Laboratory and Field Respiration’,Pestic.Sci.34,139–146.Fischer,Z.:1996,‘Ecological equilibrium disturbances in the soil exposed to simulated acid rain.Part VI’,Environ.Monit.Assess.41,61–65.Freedman,B.and Hutchinson,T.C.:1980,‘Effect of smelter pollutants on forest litter decomposition near a nickel-copper smelter at Sudbury,Ontario’,Can.J.Bot.58,1722–1736.Freedman,B.and Hutchinson,T.C.:1980,‘Pollutant inputs from the atmosphere and accumulation in soils and vegetation near a nickel-copper smelter at Sudbury,Ontario’,Can.J.Bot.58,108–132.Gildon,A.and Rimmer,D.L.:1993,‘Soil respiration on reclaimed coal-mine spoil’,Biol.Fertile Soils16,41–44.Klekowski,R.Z.:1975,‘Constant Pressure V olumetric Microrespirometer for Terrestrial Inver-tebrates’in W.Grodzi´n ski,R.Z.Klekowski and A.Duncan(eds.),Methods for Ecological Bioenergetics,IBP Handbook Blackwell Sci.Publ.,Oxford-London-Edinburgh-Melbourne,24, pp.212–225.Nakadai,T.,Koizumi,H.,Usami,Y.,Satoh,M.and Oikawa,T.:1993,‘Examination of the method for measuring soil respiration in cultivated land:Effect of carbon dioxide concentration on soil respiration’,Ecological Research8,65–71.Nordgren,A.,Baath,E.and Soderstrom,B.:1983,‘Microfungi and microbial activity along a heavy metal gradient’,Appl.Environ.Microbiol.46,1829–1837.Nordgren,A.,Baath,E.and Soderstrom,B.:1985,‘Soil microfungi in an area polluted by heavy metals’,Can.J.Bot.63,448–455.Nordgren,A.,Kauri,T.,Baath,E.and Soderstrom,B.:1983,‘Soil microbial activity,mycelial lengths and physiological groups of bacteria in a heavy metal polluted area’,Environ.Poll.(Serie A)41, 89–100.Strojan,C.L.:1978,‘Forest litter decomposition in the vicinity of a zinc smelter’,Oecologia(Berl.) 32,203–212.Witkamp,M.:1969,‘Cycles of temperature and carbon dioxide evolution from litter and soil’, Ecology50,922–924.。
Plant and Soil222:119–137,2000.©2000Kluwer Academic Publishers.Printed in the Netherlands.119Foliar free polyamine and inorganic ion content in relation to soil and soil solution chemistry in two fertilized forest stands at the Harvard Forest, MassachusettsRakesh Minocha1,∗,Stephanie Long1,Alison H.Magill2,John Aber2and William H. McDowell31USDA Forest Service,Northeastern Research Station,P.O.Box640,Durham,NH03824,USA;2Complex Systems Research Center,University of New Hampshire,Durham,NH03824,USA and3Department of Natural Resources, University of New Hampshire,Durham,NH03824,USAReceived24September1999.Accepted in revised form29February2000Key words:ammonium nitrate,calcium,Harvard Forest,magnesium,nitrate leaching,polyaminesAbstractPolyamines(putrescine,spermidine,and spermine)are low molecular weight,open-chained,organic polycations which are found in all organisms and have been linked with stress responses in plants.The objectives of our study were to investigate the effects of chronic N additions to pine and hardwood stands at Harvard Forest,Petersham, MA on foliar polyamine and inorganic ion contents as well as soil and soil solution chemistry.Four treatment plots were established within each stand in1988:control,low N(50kg N ha−1yr−1as NH4NO3),low N+sulfur(74kg S ha−1yr−1as Na2SO4),and high N(150kg N ha−1yr−1as NH4NO3).All samples were analyzed for inorganic elements;foliage samples were also analyzed for polyamines and total N.In the pine stand putrescine and total N levels in the foliage were significantly higher for all N treatments as compared to the control plot.Total N content was positively correlated with polyamines in the needles(P≤0.05).Both putrescine and N contents were also negatively correlated with most exchangeable cations and total elements in organic soil horizons and positively correlated with Ca and Mg in the soil solution(P≤0.05).In the hardwood stand,putrescine and total N levels in the foliage were significantly higher for the high N treatment only as compared to the control plot.Here also, total foliar N content was positively correlated with polyamines(P≤0.05).Unlike the case with the pine stand,in the hardwood stand foliar polyamines and N were significantly and negatively correlated with foliar total Ca,Mg, and Mn(P≤0.05).Additional significant(P≤0.05)relationships in hardwoods included:negative correlations between foliar polyamines and N content to exchangeable K and P and total P in the organic soil horizon;and positive correlations between foliar polyamines and N content to Mg in soil solution.With few exceptions,low N +S treatment had effects similar to the ones observed with low N alone for both stands.The changes observed in the pine stand for polyamine metabolism,N uptake,and element leaching from the soil into the soil solution in all treatment plots provide additional evidence that the pine stand is more nitrogen saturated than the hardwood stand. These results also indicate that the long-term addition of N to these stands has species specific and/or site specific effects that may in part be explained by the different land use histories of the two stands.Abbreviations:Perchloric acid(PCA),Polyamines(PAs),Zero tension lysimeter(ZTL),hardwoods(HW)IntroductionThere is increasing concern about the potential ad-verse effects of elevated rates of N deposition on water ∗FAX No:6038687604.E-mail:rminocha@ quality and the health of forest ecosystems(Aber et al.,1989,1998;Rasmussen and Wright,1998).This concern stems from the fact that in1990the United States(US)Clean Air Act targeted a50%reduction in S deposition but only a10%reduction in N depos-120ition(McNulty et al.,1996).Although most temperate forests are N limited,continuous deposition of N from the atmosphere can move them towards nitrogen sat-uration.Nitrogen saturation has been defined as the availability of ammonium and nitrate in excess of total combined plant and microbial nutritional demand (Aber et al.,1989).It is important to understand how chronic additions of N to ecosystems change the struc-ture and function of forest ecosystems(Asner et al., 1997;Jefferies and Maron,1997).Long-term elevated N deposition typically leads to an increase in the concentration of total foliar N, with or without similar changes in the important base elements such as Ca,Mg and K(Aber et al.,1995; Magill et al.,1997,1999;Rasmussen and Wright, 1998;van Dijk and Roelofs,1988).This increase in leaf N content also leads to significant shifts in the internal partitioning of N within the leaf.For example in conifers,N deposition significantly increases leaf N present in the form of free amino acids such as arginine (Ericsson et al.,1993,1995;Näsholm et al.,1997). Little is known about N partitioning for hardwoods under these wlor(1992)suggested that these changes in N partitioning are probably not re-lated to leaf function.This idea,however,has yet to be experimentally tested in terms of whether the alter-ations in N partitioning due to long-term N deposition actually have a positive or a negative effect on photo-synthetic capacity and biomass production.A possible decoupling of the relationship between foliar N and photosynthetic rate may occur under these conditions.The aliphatic polyamines(putrescine,spermidine, and spermine)play an important role in the growth and development of all living organisms.They are meta-bolically derived from the amino acids arginine and ornithine,and at cellular pH they carry a net positive charge(Cohen,1998).Abiotic stress conditions such as low pH,high SO2,high salinity,osmotic shock, nutrient stress,low temperature(Flores,1991and ref-erences therein)and high Al(Minocha et al.,1992, 1996)all result in an increase in cellular putrescine levels.Polyamines show a reverse proportionality to cellular elements such as Ca,Mg,Mn,and K in re-sponse to Al treatment in periwinkle(Catharanthus roseus)and red spruce(Picea rubens)cells(Minocha et al.1992,1996;Zhou et al.,1995).A key distinction between the polyamines(organic cations)and the inor-ganic elements is that,even if the latter(e.g.Ca2+and Mg2+)undergo recompartmentalization in response to external stimuli,their cellular levels are derived mainly from uptake across the biological membranes and this uptake ultimately depends upon their availab-ility in the soil or soil solution.In contrast,polyamines are synthesized within the cell,permitting adjustment of their cellular concentrations to meet physiological needs.Also,cellular polyamine levels can be regu-lated by conjugation,degradation,and sequestration via enzymatic means(Minocha et al.,1996).Thus polyamine synthesis may play an important role in the survival of plants under stress(Galston,1989).Recent work examining the concentration of polyamines in plant foliage has been aimed at using foliar polyam-ine concentrations as indicators of stress(Dohmen et al.,1990;Minocha et al.,1997;Santerre et al.,1990). In the case of mature red spruce trees,an increase in foliar putrescine concentration was associated with a decrease in foliar and soil Ca and Mg concentrations and an increase in the Oa horizon soil Al or Al:Ca ratios(Minocha et al.,1997).Structurally,polyamines are composed of carbon, hydrogen,and nitrogen.Therefore,we suspect that their levels may also change in response to chronic N additions,thus affecting the internal N partitioning within the leaf,a situation similar to that observed with free amino acids in conifers.The objective of this study was to determine the effects of chronic additions of N on:(1)foliar polyamines(a proposed stress in-dicator)and inorganic ions;(2)soil and soil solution inorganic ion chemistry,especially Al mobilization; and(3)the relationship between polyamines and soil chemistry in pine and hardwood stands.Materials and methodsStudy sitesThe study plots are located at Harvard Forest,Peter-sham,MA(42◦30 N,72◦10 W).This site is a part of the National Science Foundation’s Long-Term Eco-logical Research(LTER)program.These plots are a part of the ongoing study on chronic nitrogen addi-tions since1988(Aber et al.,1993;Magill et al.,1997, 1999).An even-aged red pine(P.resinosa)stand,74 yr old,and an adjacent mixed hardwood stand,approx. 55yr old,were chosen for this study.The hardwood stand is dominated by black oak(Quercus vetulina Lam.)and red oak(Q.Rubra Michx.f.,respectively) with significant amounts of black birch(Belutinaa lenta L.),red maple(Acer rubrum L.),and American beech(Fagus grandifolia Ehrh.).Most of the currently forested area at this site was in cultivation or pas-tureland during the mid-1800’s(Foster,1992).The121dominant soil types are stony sandy loams classified as Typic Dystrochrepts.Mean annual temperature ranges from19◦C in July to–12◦C in January and mean total annual precipitation is112cm.The estimated nitrogen deposition to the forest is about6kg ha−1yr−1wet and2kg ha−1yr−1dry(Aber et al.,1993;Magill et al.,1997;Ollinger et al.,1993).TreatmentsFour treatment plots were established within each stand:control,low N,low N+sulfur(N+S),and high N.Each plot measured30×30m(0.09ha)and was divided into36subplots(each5×5m).Fertilizer additions of NH4NO3and Na2SO4began in1988as six equal applications over the growing season.The fertilizer was weighed,mixed with20L of water (equivalent to0.002cm rainfall)and applied using a backpack sprayer.Two passes were made across each plot to ensure an even distribution of fertilizer.As described in Magill et al.(1997),a partial ap-plication was made in year1(1988).The total amount of fertilizer applied was38kg N ha−1yr−1to the low nitrogen treatment and the nitrogen portion of the N+S treatment,113kg N ha−1yr−1to the high nitrogen treatment,and74kg(S)ha−1yr−1to the N+S plots. Applications for all following years were at the rate of 50kg N ha−1yr−1to the low and N+S plots and150 kg N ha−1yr−1to the high N plots.Sulfur additions remained the same as used for year one.Collection and analyses of needle samplesFoliage samples were collected during thefirst week of August each year from mid to upper canopy branches of dominant or co-dominant trees using a shotgun.Early August was chosen as the sampling time because the trees were still physiologically very active at this time,compared to early or late summer. At each sampling time,current-year needle samples were collected from20different red pine trees in the pine stand and leaves from10different trees of black or red oak and red maple each were collected from the hardwood plots.A sub-sample was taken from each in-dividual tree collection for analyses of polyamine and exchangeable inorganic elements.The remaining pine needle samples from20different trees in each plot were pooled into5composite samples of4trees per sample for total inorganic elements and N analyses. Similarly,the remaining samples from10–12trees per species collected from each plot in the hardwood stand were pooled into4composite samples.Red and black oak were treated as a single species in all collections. Total elements and nitrogen analysesThe composite samples were dried at70◦C for48 h and ground using a Wiley mill with a1mm mesh screen.The ground samples were dried overnight at 70◦C and analyzed for N content using near-infrared (NIR)spectroscopy(Bolster et al.,1996;McLel-lan et al.,1991).These samples were also used for extraction of total inorganic ion content by a modi-fication of the method of Isaac and Johnson(1976) as described in Minocha and Shortle(1993).The ex-tracts were analyzed for total Ca,Mg,Mn,K,Al, and P content using a Beckman Spectrospan V ARL DCP(Direct Current Plasma Emission Spectrometer, Beckman Instruments,Inc.,Fullerton,CA)using the Environmental Protection Agency’s method number 66-AE0029(1986).Analysis of polyamines and exchangeable inorganic elementsThe fresh foliage samples were placed in individual pre-weighed microfuge tubes containing1ml of5% perchloric acid(PCA).The tubes were kept on ice dur-ing transportation to the laboratory and then stored at –20◦C until they were processed.The samples were weighed,frozen and thawed(3X),and centrifuged at 13,000rpm in a microfuge for10min.Details of the freeze-thawing extraction procedure are described in Minocha et al.(1994).The freeze-thawing method was also chosen for the extraction of exchangeable fraction of inorganic ions.This method extracted a consistent fraction of the total acid extractable inor-ganic ions from foliage of various species of mature trees and the quantity of this fraction varied for each ion type and tree species(Table1and Minocha et al.,1994).The moisture content data for individual exchangeable ion samples was not collected.There-fore,for comparison of total and exchangeable ions on weight basis,an average moisture content of53%, 62%,and61%has been used for this site from data collected in1992for pine(n=19),oak(n=33),and maple(n=45),respectively to convert fresh weight to dry weight.Briefly,the samples for both inorganic ions as well as for polyamines were frozen at–20◦C and thawed at room temperature,repeating the process two more times.Duration of the freezing step could vary from4122parison of effects of nitrogen treatments on total and exchangeable inorganic ion levels in the foliage of pine,oak and maple trees.Data has been averaged over three year period(1995–1997).Data for exchangeable ions are mean±se of n=60for pine stand and n=30for hardwoods.Data for total ions are mean±se of n=15for pine stand and n=12for hardwoods.The numbers in parenthesis indicate%of total ions extracted as exchangeable.The moisture content data for individual exchangeable ion samples was not collected.Therefore,for comparison of total and exchangeable ions on weight basis,an average moisture content of53%,62%,and61%has been used for this site from data collected in1992for pine(n=19),oak(n=33),and maple(n=45),respectively to convert FW to DWTreatment Ion Inorganic ions Pine(µmol g−1DW)Oak(µmol g−1DW)Maple(µmol g−1DW)Control Ca Total62.5±4.8121.5±5.9156.0±6.8Low N72.3±5.1113.5±8.1137.0±6.7High N60.0±5.571.8±6.2116.4±6.1Control Exchangeable11.4±0.6(18.3%)102.1±5.9(84.0%)79.2±4.7(50.8%)Low N11.9±0.7(16.5%)87.1±4.7(76.7%)69.7±2.3(50.9%)High N9.9±0.9(16.5%)57.3±5.0(79.9%)59.2±3.4(50.9%)Control Mg Total32.4±1.562.8±2.164.7±3.3Low N35.5±1.363.1±2.359.9±1.3High N35.4±1.354.1±3.355.7±2.9Control Exchangeable13.6±0.5(41.9%)56.1±3.4(89.3%)43.7±2.3(67.5%)Low N14.7±0.4(41.3%)58.0±3.4(91.9%)43.5±1.4(72.7%)High N13.5±0.4(38.1%)51.0±3.0(94.2%)38.5±1.9(69.1%)Control Mn Total16.7±1.141.6±2.337.4±2.3Low N22.4±2.127.3±2.023.6±0.9High N15.8±1.615.3±0.922.5±1.9Control Exchangeable 3.6±0.2(21.3%)50.8±3.5(122.1%)∗21.1±1.5(56.6%)Low N 3.9±0.3(17.3%)33.4±2.4(122.4%)∗14.5±1.0(61.6%)High N 3.2±0.4(20.5%)17.4±1.5(113.4%)∗11.1±0.9(49.5%)Control K Total120.1±6.7218.7±9.5184.6±11.8Low N125.5±7.2265.7±12.0171.6±10.5High N135.3±7.4202.2±9.4176.0±9.7Control Exchangeable72.8±3.8(60.6%)158.0±6.8(72.3%)132.2±6.0(71.6%)Low N79.6±3.6(63.4%)165.1±7.2(62.2%)138.8±5.9(80.9%)High N84.0±3.2(62.1%)157.5±6.7(77.9%)147.0±5.8(83.5%)∗The calculation of greater than100%extraction for Mn in the case of oak was possibly caused by slight overestimation of the mean moisture content for1995–1997.h to a few days.Samples were allowed to thaw com-pletely(approximate time1–1.5h)before refreezing. After freeze-thawing,the samples were centrifuged at 13,000×g.This supernatant was used directly for free polyamine analysis without further dilution and for inorganic-ion analysis after proper dilution with distilled,deionized water(final concentration of PCA 0.01or0.02N)by the procedures described below. The diluted fractions were analyzed for inorganic ion content with a Beckman Spectrospan V ARL DCP as described above.For quantitation of polyamines, heptanediamine was added as an internal standard to aliquots of the above extracts prior to dansylation. Fifty or one hundredµL of the extract were dansylated according to the procedure described in Minocha et al.(1990).Dansylated polyamines were separated by reversed phase HPLC(Perkin-Elmer Corp.,Norwalk,CT)using a gradient of acetonitrile and heptane-sulfonate,and quantified by afluorescence detector (Minocha et al.,1990).Collection and analyses of soil and soil solution samplesThree sets of two adjacent soil cores(<30cm apart) were taken to a depth of10cm in the mineral soil in each of the three designated subplots(nine samples per plot).Cores were split into organic(Oe+Oa)and min-eral horizons(top10cm)and placed in gas-permeable polyethylene bags.The soils were air-dried,sieved (<2mm size),and stored at room temperature prior to analyses.Before analyses,the mineral soil and organic soil samples were oven-dried at105◦C and70◦C, respectively.123Exchangeable inorganic elements were determined from a sub-fraction of the above samples using a modi-fication of the procedure of Taylor(1987).Briefly, either6g of mineral soil or1g of organic soil was added to30ml of extraction solution(0.05N HCl and 0.025N H2SO4)and placed on a gyratory shaker at90 rpm for15min.The extract wasfiltered with a glass fiber syringefilter(Gelman A/E,Gelman Sciences, Ann Abror,MI)and stored at4◦C until quantitation by DCP.Prior to digestion for total elemental analysis,a sub-sample of air-dried and sieved soil was processed for1min in a Shatter Box Laboratory Mill(Spex Industries,Inc.,Edison,NJ)to powder the sample. Microwave digestion(Hallett and Hornbeck,1997) was used for obtaining inorganic elements.Briefly, for mineral soils,0.1g sample was digested with5 ml of concentrated HNO3,2ml of concentrated HCl, 2ml offluoroboric acid(HBF4)and2ml of H2O2. For organic soils,0.1g sample was digested the same way with acids but without the presence of H2O2.The following microwave programs were used.Given in pairs are:Time(min)–Power(Watts).For mineral soil, 1–250,1–0,5–250,5–400,5–500,and5–600.For organic soil,1–250,1–0,5–250,5–400,1–0,5–500, 1–0,5–600,1–0,and2.5–650.Total running time was 27.5min.The details on installation of zero tension lysi-meters(ZTL)and soil solution sample collection are described in Currie et al.(1996)and McDowell et al. (1998).Briefly,5polyethylene ZTL’s were installed per plot except for N+S plots.Solutions were collected after major rain events and all5samples per plot were pooled prior to analyses.Over the course of three years (1995–1997),the collections were made approx.50 times from each plot.Samples were transported on ice to the laboratory andfiltered through pre-combusted Whatman GF/F glassfiberfilters(Whatman Inc., Clifton,NJ)within36h of collection before freezing. These solutions were analyzed for inorganic elements using DCP as described earlier.Statistical methodsLinear regression analyses were performed to es-tablish the strength and significance of relationships between two different variables(n=12except for soil samples where n=8)using Excel5.0for Windows(Mi-crosoft Corporation,Roselle,IL).Data for each vari-able(e.g.,foliar or soil Ca,Mg and Al)were analyzed as a series of one-way analysis of variance(ANOV A)to determine whether statistically significant differ-ences occurred between control and treatment plots for each individual variable.When F values for one-way ANOV A were significant,differences in treatment means were tested by Tukey’s multiple comparis-ons test.ANOV A and Tukey’s tests were performed with Systat for Windows,version7.01(SYSTAT Inc., Evanston,Il)and a probability level of0.05was used for tests unless specified otherwise.ResultsWith few exceptions,the low N+S treatment showed effects similar to the ones observed for low N treat-ment alone for both stands.For this reason,results for low N+S will not be discussed separately.Also,in foliage exchangeable ions always represented a con-sistent fraction of the total ions.Even though the quantity of this fraction varied for each ion type and tree species,nitrogen treatments had similar effects on both exchangeable and total Ca,Mg,Mn,and K levels in each case for all three species(Table1).Thus due to this similarity of trends between exchangeable and total ions,only total inorganic ion data are used for further comparison of foliar results with soil and soil solution data.Foliar polyamines and N contentPine:There was a significant increase in the level of putrescine in the needles of trees growing on all three treatment plots as compared to the control plot (Figure1A–D).A small but statistically significant increase was also observed in spermidine levels in response to high N treatment.Spermine,which was a relatively small proportion of free polyamines in red pine,increased significantly in response to low and high N additions.In spite of year to year vari-ations in the total amounts of polyamines(possibly due to different growth conditions resulting from vari-able weather patterns),similar trends were observed for each of the three years of data collection. Hardwoods:Putrescine levels in oak leaves were three-to four-fold higher in response to high N treat-ment for each of the three years of this study(Fig-ure2A–D).For other treatments no significant change was observed.The level of spermidine and spermine did not change significantly in response to low and high N treatments.While there were annual variations124Figure1.The effect of chronic N additions on foliar polyamines in pine.Data presented for each treatment are mean±SE of n=20for A–C. Figure1D presents cumulative data for three years(n=60).The asterisks denote the significant differences from control.in the total amount of polyamines,the trends were the same for each year.High N treatment caused a significant increase in putrescine levels in maple leaves(Figure3A–D)for two of the three years.Spermidine and spermine levels were also present in quantities comparable to putres-cine in maple leaves.For1995,a parallel increase in spermidine was also observed(Figure3A).No change in spermine content was observed for any of the three years.As with pine and oak,year to year variation in the three polyamines was also observed in maple.The total N content of pine needles and maple leaves increased in response to all N treatments(Fig-ure4A and4C).The total N content in oak foliage also rose significantly but only in response to the high N treatment(Figure4B).The changes in N were always positively and significantly correlated with changes in foliar polyamines for all three species(Figure4A–C). Foliar inorganic elementsPine:No significant changes in total foliar Ca,Mg, Mn,and K were observed in response to chronic N additions to the soil.However,there was a decrease in total Al concentration in response to all N treatments and an increase in total P with high N treatment only (Figure5).Hardwoods:Total Ca and Mn levels showed a sig-nificant decrease in response to high N treatment in both maple and oak leaves(Figure6).Changes in total Mn were also significant for low N treatment in both species.The only other statistically significant change observed was a decrease in P with all N treat-ments in maple.Mg showed a statistically insignificant decrease with high N addition for both species. Foliar Al:Ca and Mg:N ratiosA significant decrease in foliar total Mg:N and Al:Ca ratios was observed with all N treatments in case of pine(Figure7A–B).In contrast,an increase in the Al:Ca ratio(though statistically insignificant for maple)accompanied by a decrease in total Mg:N ra-tio was observed in response to high N treatment for maple and oak(Figure7C–F).Inorganic elements in the organic horizon of soil Pine:A significant decrease in exchangeable Ca, Mg,Mn,and K in the organic soil was observed in125 Figure2.The effect of chronic N additions on foliar polyamines in oak.Data presented for each treatment are mean±SE of n=10for A–B and n=5for C.Figure2D presents cumulative data for three years(n=25).The asterisks denote the significant differences from control.response to all N treatments.There was an increase in exchangeable P in response to the low N treatment only(Figure8:Pine).However,there was no signi-ficant effect of these treatments on exchangeable Al levels.Whereas total Ca decreased in response to high N treatment only,total Mn and P decreased in response to all N treatments with no significant change in K in the organic soil.A slight but statistically insignific-ant decrease in the level of total Mg and Al with N treatments was also observed(Figure9:Pine). Hardwoods:N additions alone had no significant effect on exchangeable or total Ca levels of the or-ganic horizon.However,exchangeable Mg,K,and P concentrations were significantly lower in high N plots with Mn being low in all N treatment plots(Fig-ure8:HW).Total Mg,Mn,and P levels decreased in response to all N treatments in this soil horizon (Figure9:HW).Inorganic elements in the mineral horizon of soil Pine:Whereas changes in the exchangeable Ca,Mg, Mn,and K did not show any specific and statistically significant patterns,P increased significantly in rela-tion to all N additions in the mineral soil.Both low and high N treatments decreased the exchangeable Al con-centrations significantly in this soil layer(Figure10: Pine).Hardwoods:The exchangeable ion chemistry of the mineral horizon showed decreases in the levels of some inorganic elements,but the changes were not statistically significant except for Al in the high N plot (Figure10:HW).Inorganic elements in zero tension lysimeter(ZTL) solutionPine:A significant increase was observed in the levels of all inorganic elements tested in response to high N addition.Levels of P and Al were high even for low N treatment(Figure11:Pine). Hardwoods:There was typically a small increase in the levels of most elements in the ZTL solution samples in response to N treatments,but the only sig-nificant increase that was observed was in Al levels in response to high N treatment(Figure11:HW).126Figure3.The effect of chronic N additions on foliar polyamines in maple.Data presented for each treatment are mean±SE of n=10for A–B with one exception and n=4for C.Figure3D presents cumulative data for three years(n=24).The asterisks denote the significant differences from control.Relationship between foliar and soil chemistry Pine:Putrescine levels did not correlate with most inorganic elements in the pine needles.However,pu-trescine was negatively correlated with all but P of the exchangeable and all of the total inorganic ions in the organic soil horizon(Table2).There was no correlation between putrescine or total polyamines and exchangeable elements in the mineral soil except for P and Al.In contrast to the results for organic soil hori-zon,there was a positive correlation between Ca,Mg, and Al in soil solution and foliar putrescine as well as total polyamines(Table2).N content in the pine needles was positively correl-ated with foliar Mg and P only.There was also a strong negative correlation between foliar N and exchange-able Ca,Mg,Mn,and K in the organic soil horizon. Similar to the situation with total polyamines,foliar N showed a significant negative correlation only with total Mn and P in the organic soil horizon.N content, however,was positively correlated with all elements in the soil solution(Table2).Hardwoods:Unlike the situation with pine,putres-cine and total PAs in oak and maple leaves showed a strong negative correlation with foliar Ca,Mg,Mn, and P.In the organic soil horizon,putrescine,total polyamines as well as N were significantly and in-versely related with exchangeable K and total P in both oak and maple.In the mineral soil,putrescine in oak was negatively correlated only with Al but in maple it had a negative correlation with several elements.Fo-liar putrescine and N were also positively correlated with soil solution Mg and/or Al.For more details on individual correlations refer to Table2.DiscussionBiochemical changes that occur in response to expos-ure to a particular stress can be measured before any visible symptoms appear at the level of the organism and may even be used as early indicators of change(s) in the vitality of trees within a stand(Baur et al.,1998; Minocha et al.,1997).Among the physiological and molecular responses of plants to high N deposition, is the increase in cellular content of leaf N.Nitrogen has been shown to be stored in the leaf in the form of nitrate and/or specific free amino acids such as ar-ginine(Aber et al.,1995;Ericsson et al.,1993,1995;。
政治学主要英文期刊简介1、刊名称American Political Science Review, with PS: Political Science & Politics.ISSN: 0003-0554创刊年:1906出版期次4/yr. 开本:21.5x15.5 页数:300 目录价:154/USD出版机构American Political Science Association,USA 出版地:美国译名简介《美国政治科学评论》附《政治科学与政治学》美国政治科学学会会刊。
刊载政治理论、政治学、比较政治学、国际政治学等方面的文章、札记和书评。
附刊《政治科学与政治学》,报道该学会的研究动态和会议消息,并刊载学位论文。
SSCI来源刊,影响因子2.4482、报刊名称Comparative Politics. 创刊年:1968ISSN:0010-4159出版期次4/yr. 开本:21.5x15.5 页数:90 目录价:69/USD出版机构City University of New York, Political Science Program,USA 出版地:美国译名简介《比较政治学》刊载美国和其它国家的政治、社会、经济与宗教的研究和比较研究方面的文章与评论。
SSCI来源刊,影响因子1.0833、报刊名称Capitalism, Nature, Socialism. 创刊年:1989ISSN:1045-5752出版期次4/yr. 开本:21.5x15.5 页数:144 目录价:99/USD出版机构Guilford Publications Inc.,USA 出版地:美国译名简介《资本主义、自然、社会主义》由加利福尼亚大学社会学与经济学教授James O’Connor担任主编的探讨和论述社会主义生态学的国际性刊物。
内容涉及政治学、经济学、社会学和环境科学等。
PROQUEST有全文4、报刊名称Public Administration. 创刊年:1922ISSN:0033-3298出版期次4/yr. 开本:21.5x15.5 页数:124 目录价:324/GBP出版机构Blackwell Publishers Ltd.,UK 出版地:英国译名简介《公共行政》刊载公共行政管理和政策制定以及国际行政管理和比较研究等方面的文章。
一、中文数学相关网▪中国数学资源网▪陈省身数学研究所▪学问社区▪中国数学会▪中国科学院▪中国科学院数学与系统科学研究院▪中国科学院数学研究所▪中国科学院应用数学研究所▪中国科学院系统科学研究所▪中国科学院计算数学与科学工程计算研究所▪中国科学院数学机械化研究中心▪中国科学院数学与系统科学研究院图书馆▪数学中国▪中国数学在线▪晨兴数学中心▪北京大学数学科学学院▪复旦大学数学科学学院▪浙江大学数学系主页▪浙江大学数学中心▪香港中文大学数学研究所▪南开大学数学科学学院▪南开大学组合数学中心▪中山大学数学与计算科学学院▪MathWorld▪博士数学论坛▪中国数学建模网▪中国运筹学论坛▪数学教研网▪中国运筹学会▪美国数学协会▪加拿大数学协会▪匈牙利科学院数学研究所▪MathSciNet▪Zentralblatt MATH▪EMIS欧洲数学信息服务▪中国国家图书馆▪维基百科,自由的百科全书▪丘成桐中学数学奖▪丘成桐中学数学奖(海外)▪丘成桐中学数学奖(台湾)▪美国高中数学建模竞赛网二、外文数学相关网站▪Analysis - Math Forum 数学论坛——分析数学,链接到一些最好的因特网分析数学资源,如网络站点、软件、因特网项目、出版物和可供讨论的公众论坛。
▪David Marius Bressoud David Marius Bressoud的主页。
David Marius编写了一系列的大学课本,在这些课本中,他使用物理和数学的历史来激发学生对多变量微积分学,数论以及实际分析的学习兴趣。
三、数学词表▪1991 Mathematics Subject Classification (MSC) - Chris Eilbeck; Heriot-Watt University, Edinburgh 爱丁堡Heriot-Watt大学Chris Eilbeck编著的1991数学主题分类表。
这是1991数学主题分类表(Mathematics Subject Classification)的超文本版本。
Keywords:alignment,Gestalt,Helmholtz,principle,probabi 1.IntroductionMost theories of image analysis tend tofind in a given image geometric structures(regions,contours,lines, convex sets,junctions,etc.).These theories generally assume that the images contain such structures and thentry to compute their best description.The variational framework is quite well adapted to such a viewpoint(for a complete review,see e.g.Morel and Solimini, 1995).The general idea is to minimize a functional ofthe kindF(u,u8Desolneux,Moisan and Moreland the resulting segmentation depends a lot upon the value of these constants.The other point is that they will always deliver a minimum for their functional and so they assume that any image may be segmented(even a white noise).Indeed,they do not yield any criterion to decide whether segmentation is relevant or not.Of course,the probabilistic framework leading to varia-tional methods should in principle give a way to es-timate the parameters of the segmentation functional. In the deterministic framework,these parameters can sometimes be estimated as Lagrange multipliers when (e.g.)a noise model is at hand,like in the Rudin-Osher-Fatemi method(see Rudin et al.,1992).It is nonetheless easy to check that,first,most variational methods propose a very rough and inaccurate model for the image,second,their parameters are generally not correctly estimated anyway,yielding to supervised methods.Actually,we should not be fair if we claimed that what we propose immediately yields a more re-liable segmentation method.In fact,we only intend to point out the possibility of checking any proposed segmentation,by any segmentation method,from the point of view of meaningfulness.So far,this check will only be analysed in detail for straight boundaries: given a segmentation performed by any other method, we can,with the method proposed here,a posteriori decide about the meaninfulness of straight parts of the proposed boundaries.Another drawback of most segmentation meth-ods is their locality.Despite the Gestaltists theories, they look rather for local structure.Let us mention some nonlocal theories of image analysis:the Hough Transform(see Maitre,1985),the detection of glob-ally salient structures by Sha’Ashua and Ullman(see Sha’Ashua and Ullman,1988),the Extension Field of Guy and Medioni(see Guy and Medioni,1992)and the Parent and Zucker curve detector(see Parent and Zucker,1989).These methods have the same drawback as the variational models of segmentation described above.The main point is that they a priori suppose that what they want tofind(lines,circles,curves,...)is in the image.They mayfind too many or too little such structures in the image and do not yield an existence proof for the found structures.As a main example,let us describe the Hough transform.We assume that the image under analysis is made of dots which may cre-ate aligned patterns or not.We then compute for each straight line in the image,the number of dots lying on the line.In fact,the Hough transform describes a fast algorithm to do so.The result of the Hough trans-form is then a map associating with each line a number of dots.Then,“peaks”of the Hough transform may be computed:they indicate the lines which have more dots.Which peaks are significant?Clearly,a threshold must be used.For the today technology,this threshold generally is given by a user or learned.The Hough transform is nothing but a particular kind of“group-ing”.According to Gestalt theory,“grouping”is the law of visual perception(see Kanizsa,1997).Its main idea is that whenever points(or previously formed vi-sual objects)have a characteristic in common,they get grouped and form a new,larger visual object, a“Gestalt”.Some of the main grouping characteris-tics are colour constancy,“good continuation”,align-ment,parallelism,common orientation,convexity and closedness(for a curve),...In addition,the group-ing principle is recursive.For example,if points have been grouped into lines,then these lines may again be grouped according(e.g.)to parallelism.Our purpose is not to propose a new segmentation method.We rather propose a computational method to decide whether a given Gestalt(obtained by any seg-mentation or grouping method)is sure or not.Although most of what we write here can be generalized to other geometric structures,we shall focus on alignments,one of the most basic Gestalt(see Wertheimer,1923).In this paper,we push the study to the end for the detection of alignments,but we willfirst give a general definition of what we will call“a meaningful event”. Many of our statements will apply to other Gestalt as well.Our main idea is that a meaningful event is an event that,according to probabilistic estimates,should not happen in an image and therefore is significant.In that sense,we shall say that it is a“proven event”.The above informal definition immediately raises an objec-tion:if we do probabilistic estimates in an image,this means that we have an a priori model.We are there-fore losing any generality in the approach,unless the probabilistic model could be proven to be“the right one”for any image.In fact,we shall do statistical es-timates,but related not to a model of the images but to a general model of perception.We shall apply the so called Helmholtz principle.This principle attempts to describe when perception decides to group objects according to some quality(colour,alignment,etc.).It can be stated in the following way.Assume that ob-jects O1,O2,...,O n are present in an image.Assume that k of them,say O1,...,O k have a common feature, say,same colour,same orientation,etc.We are then facing the dilemna:is this common feature happen-ing by chance or is it significant?In order to answerMeaningful Alignments9this question,we make the following mental experi-ment:we assume that the considered quality has beenrandomly and uniformly distributed on all objects,i.e.O1,...O n.Notice that this quality may be spatial(like position,orientation);then we(mentally)assume thatthe observed position of objects in the image is a ran-dom realization of this uniform process.Then,we mayask the question:is the observed repartition probableor not?The Helmholtz principle states that if the expectationin the image of the observed configuration O1,...,O k is very small,then the grouping of these object makes sense,is a Gestalt.Definition1(ε-meaningful event).We say that an event of type“such configuration of points has such property”isε-meaningful,if the expectation in a im-age of the number of occurences of this event is less thanε.Whenε 1,we talk about meaningful events.This seems to contradict our notion of a parameter-less the-ory.Now,it does not,since theε-dependency of meaningfulness will be low(it will be in fact a log ε-dependency).The probability that a meaningful event is observed by accident will be very small.In such a case,our perception is liable to see the event,no matter whether it is“true”or not.Our termε-meaningful is related to the classical p-significance in statistics;as we shall see further on,we must use expectations in our estimates and not probabilities.The program we state here has been proposed several times in Computer Vision.We know of at least two in-stances:Lowe(1985)and Witkin-Tenenbaum(1983). Let us quote extensively David Lowe’s program,whose mathematical consequences we shall try to develop in this paper:“we need to determine the probabil-ity that each relation in the image could have arisen by accident,P(a).Naturally,the smaller that this value is,the more likely the relation is to have a causal interpretation.If we had completely accurate image measurements,the probability of accidental occurence could become vanishingly small.For example,the probability of two image lines being exactly parallel by accident of viewpoint and position is zero.However, in real images there are many factors contributing to limit the accuracy of measurements.Even more impor-tant is the fact that we do not want to limit ourselves to perfect instances of each relation in the scene—we want to be able to use the information available from even approximate instances of a relation.Given an im-age relation that holds within some degree of accuracy, we wish to calculate the probability that it could have arisen by accident to within that level of accuracy.This can only be done in the context of some assumption re-garding the surrounding distribution of objects,which serves as the null hypothesis against which we judge significance.One of the most general and obvious assumptions we can make is to assume that a back-ground of independently positioned objects in three-space,which in turn implies independently positioned projections of the objects in the image.This null hy-pothesis has much to recommend it.(...)Given the assumption of independence in three-space position and orientation,it is easy to calculate the probabil-ity that a relation would have arisen to within a given degree of accuracy by accident.For example if two straight lines are parallel to within5degrees,we can calculate that the chance is only5/180=1/36that the relation would have arisen by accident from two independent objects.”Some main points of the pro-gram we shall mathematically develop are contained in the preceding quotation:particularly the idea that significant geometric objects are the ones with small probability and the idea that this probability is anyway never zero because of the necessary lack of accuracy of observations in an image.Now,the preceding program is not accurate enough to give the right principles for computing Gestalt.The above mentionned example is e.g.not complete enough to be convincing.Indeed, we simply cannotfix a priori an event such as“these two lines are parallel”without merging it into the set of all events of the same kind,that is,all parallelisms. The space of straight lines in an image depends on the accuracy of the observations,but also on the size of the image itself.The fact that the mentionned prob-ability be“low”(1/36)does not imply that few such events will occur in the image:we have to look for the number of possible pairs of parallel lines.If this number is large,then we will in fact detect many non-significant pairs of parallel lines.Only if the expected number of such pairs is much below1,can we de-cide that the observed parallelism makes sense.Be-fore proceeding to the mathematical theory,let us give some other toy example and discuss our definition of “ε-meaningfulness”.Example and Discussion:Let us consider an image of size100×100pixels.We assume that the grey-level at each pixel is0or1,which means that we work on a binary image.Our main asumption is that if two points10Desolneux,Moisan and Moreldo not belong to the same object,then their grey-levels are independent(and equally distributed if the image is equalized).Now,imagine that we observe the following event:a black10×10square.The expectation of the number of10×10black squares in the image is simply the number of10×10squares in the100×100image times the probability that each pixel of a10×10square is black.And so the expectation is90·90·12100,which is much less than1.We conclude that this event is meaningful.Remarks:1)Subsquares(large enough)are also mean-ingful,and so are also candidates to be“Gestalt”.2) Interaction of Gestalts:if we take into account that we observe a10×10black square on a30×30white background,then the expectation of the number of oc-curences of this square-on-background event is70·70·12100·12800,and so we get a“much more meaningful”event.This is rather a toy example,but it shows immediatly which kind of difficulties and apories are associated with “meaningfulness”:1.Too many meaningful events:by the same argumentas above,all large enough parts of the black square are meaningful.If(e.g.)we take all parts of this square with cardinality larger than50,they are all meaningful and their number is larger than250!We will see how to solve the problem of having too many meaningful events by defining the notion of “maximal meaningful event”.2.Problem of the a priori/a posteriori definition of theevent:if we take an arbitrary10×10pattern in a 100×100random binary image,then the expec-tation of the number of occurences of this event is90·90·(12)100which is much less than1.The an-swer is that we need an a priori geometric definition of the event,as done in Gestaltism.The event can-not be defined from the observed image itself!3.Moreover,we can remark that the definition of thegeometric event changes its“meaningfulness”.For example if we consider our10×10black square asa convex set with area100,then the expectation be-comes(12)100times the number of convex sets witharea100.And so the event may loose its meaning-fulness.4.Abstract geometrical character of the information,lack of localization.ex.1:if we observe a meaningful black patch,allwhat we can say is:“there is a black patch andthe indicated dots may belong to it”.We do notknow which points belong“for sure”to the event.ex.2:if we observe a meaningful alignment ofpoints,then we can say“on that line,there arealigned points”but we are not able to define theendpoints.5.How many Gestalt?If we make a list of“pregnant”Gestalt,following Gestalt theory,the longer the list,the higher the expectation offinding“false gestalt”.Thus,perception,and also computer vision will atsome time meet the following problem:tofind thebest trade off between number of Gestalt(whichmight be a priori as high as possible)and the falsedetection rate.For the time being,we shall notadress this problem;it will be adressed only whenwe are in a position to do a correct theory for manyGestalt!Our plan is as follows.In Section2,we explain ourdefinition of meaningful alignments.Section3is de-voted to the structure properties of the“number of falsealarms”.In Section4,we give asymptotic(as l→∞)and non-asymptotic estimates about the meaningful-ness of the following observation:“k well-alignedpoints in a segment of length l”.Section5introducesmaximal meaningfulness as a mean to reduce the num-ber of events and localize them.Section6gives strongarguments in favour of our main conjecture:two maxi-mal meaningful segments on the same line are disjoints.In the experimental Section7,we compute mean-ingful and maximal meaningful alignments in severalimages.2.Definition of Meaningful Segments2.1.Very Local Computation of the Directionof the Level LinesLet us consider a gray image of size N(that is N2pixels).At each point,we compute a direction,whichis the direction of the level line passing by the pointcalculated on a q×q pixels neighbourhood(generallyMeaningful Alignments11 q=2).No previous smoothing on the image will beperformed and no restoration:such processes wouldloose the a priori independence of directions which isrequired for the detection method.The computation of the gradient direction is basedon an interpolation(we have q=2).We define thedirection at pixel(i,j)by rotating byπthe directionof the gradient of the order2interpolation at the centerof the2×2window made of pixels(i,j),(i+1,j),(i,j+1)and(i+1,j+1).We getdir(i,j)=1DD where D=−[u(i,j+1)+u(i+1,j+1)]+[u(i,j)+u(i+1,j)][u(i+1,j)+u(i+1,j+1)]−[u(i,j)+u(i,j+1)].Then we say that two points X and Y have the same direction with precision1nifAngle(dir(X),dir(Y))≤2πn.(4)In agreement with psychopysics and numerical exper-imentation,we consider that n should not exceed16.2.2.Probabilistic ModelAccording to the Helmholtz principle,our main as-sumption is following:we assume that the direction atall points in an image is a uniformly distributed randomvariable.In the following,we assume that n>2andwe set p=1<1;p is the accuracy of the direction. We interpret p as the probability that two independentpoints have the“same”direction with the given ac-curacy p.In a structureless image,when two pixelsare such that their distance is more than2,the direc-tions computed at the two considered pixels should beindependent random variables.We assume that everydeviation from this randomness assumption will leadto the detection of a structure(Gestalt)in the image.Alignments provide a more concrete way to under-stand Helmholtz principle.We know(by experience)that images have contours and therefore meaningfulalignments.This is mainly due to the smoothness ofcontours of solid objects and the generation of geo-metric structure by most physical and biological laws.Now,it can be assumed that in afirst approximation,the relative positions of objects are independent.Thismeans that whenever two points x and y belong to thesame contour,their directions are likely to be highlycorrelated,while if they belong to two different objects,their directions should be independent(see the above quoted Lowe’s program).From now on,the computations will be performed on any image presenting at each pixel a direction which is uniformly distributed,two points at a distance larger than q=2having independent directions.Let A be a segment in the image made of l independent pixels (it means that the distance between two consecutive points of A is2and so,the real length of A is2l).We are interested in the number of points of A which have the property of having their direction aligned with the direction of A.Such points of A will simply be called aligned points of A.The question is to know what is the minimal number k(l)of aligned points that we must observe on a length l segment so that this event becomes meaningful when it is observed in a real image.2.3.Definition of MeaningLet A be a straight segment with length l and x1, x2,...,x l be the l(independent)points of A.Let X i be the random variable whose value is1when the direc-tion at pixel x i is aligned with the direction of A,and 0otherwise.We then have the following distribution for X i:P[X i=1]=p and P[X i=0]=1−p.(5)The random variable representing the number of x i having the“good”direction isS l=X1+X2+···+X l.(6) Because of the independence of the X i,the law of S l is given by the binomial distributionP[S l=k]=lkp k(1−p)l−k.(7)When we consider a length l segment,we want to know whether it isε-meaningful or not among all the seg-ments of the image(and not only among the segments having the same length l).Let m(l)be the number of oriented segments of length l in a N×N image.We define the total number of oriented segments in a N×N12Desolneux,Moisan and Morelimage as the number of pairs(x,y)of points in the im-age(an oriented segment is given by its starting point and its ending point)and so we havel maxl=1m(l)=N2(N2−1) N4.(8)The estimate N4is accurate enough,taking into account that what matters here will be its logarithm.Definition2(ε-meaningful segment).A length l seg-ment isε-meaningful in a N×N image if it contains at least k(l)points having their direction aligned with the one of the segment,where k(l)is given byk(l)=mink∈N,P[S l≥k]≤εN4.(9)Let us develop and explain this definition.For1≤i≤N4,let e i be the following event:“the i-th seg-ment isε-meaningful”andχei denote the characteristicfunction of the event e i.We havePχei=1=PS li≥k(l i)where l i is the length of the i-th segment.Notice thatif l i is small we may have P[S li ≥k(l i)]=0.Let Rbe the random variable representing the exact number of e i occuring simultaneously in a trial.Since R=χe1+χe2+···+χeN4,the expectation of R isE(R)=Eχe1+Eχe2+···+EχeN4=l maxl=0m(l)P[S l≥k(l)].(10)We compute here the expectation of R but not its law because it depends a lot upon the relations of depen-dence between the e i.The main point is that segments may intersect and overlap,so that the e i events are not independent,and may even be strongly dependent. By definition we haveP[S l≥k(l)]≤εN4,so that E(R)≤εN4·N4≤ε.This means that the expectation of the number of ε-meaningful segments in an image is less thanε. This notion ofε-meaningful segments has to be related to the classical“α-significance”in statistics,whereαis simplyε/N4.The difference which leadsus to have a slightly different terminology is following:we are not in a position to assume that the segmentdetected asε-meaningful are independent in any-way.Indeed,if(e.g.)a segment is meaningful it maybe contained in many larger segments,which also are ε-meaningful.Thus,it will be convenient to compare the number of detected segments to the expectation ofthis number.This is not exactly the same situation as infailure detection,where the failures are somehow dis-joint events.See Remark(*)below.This means thatis an absolute parameter,not depending upon the sizeof the image,but only on the number of false detectionswhich the user allows.Of course,if the image is larger,it may be expected that an increasing number of falsedetections should be allowed.However,byfixing al-ways smaller than one,we decided not to take this op-portunity.Our proposed definition of meaningfulnessis also related to the statistical analysis of functionalmedical images(fMRI,PET)by Statistical ParameterMap(SPM),with two main differences,however.Thefirst one is this:in the recent work of Stuart Clare(FMRIB center,Oxford,see Clare(1997)),and in theworks of Friston et al.(1991)and Forman et al.(1995),an hypothesis testing method against white noise isperformed in time series.As in the present work,thebinomial law appears and a careful account of the ef-fect offiltering on the number of effective degrees offreedom:this leads e.g.S.Clare to divide this numberby three after a small gaussianfiltering and is relatedto our decision of considering only nets of points ata distance larger than2.S.Clare does as we do;hep-tests against the white noise assumption and admitsa p-value of0.005by patient.Here is the main dif-ference:the number of patients,and the length of thedata are not taken into account in the test.In particular,the time length of the test is of course just enough toperform a significant test and the p-value is a threshold“per patient”.In our case,we have two factors:thefirst one is that the number of“patients”is huge.Thus,with a p-test,the expectation of false detections wouldbe much above1,which is what we avoid by imposing much smaller than1and by entering into the com-putation the number of segments N4.This is why we compute an expectation and not a probability:we have too many and not independent trials.The reason for introducing expectation here is the non independence (contrarily to patients)and the huge number of trials, increasing with the image size.Meaningful Alignments13 Remark.We could have defined aε-meaningfullength l segment as a segmentε-meaningful onlyamong the set of the length l segments.It would havebeen a segment with at least k (l)points having the“good”direction where k (l)is defined by m(l)·[S l≥k (l)]≤ε.Notice that m(l) N3because there areapproximately N2possible discrete straight lines in aN×N image and on each discrete line,about N choicesfor the starting point of the segment.But we did notkeep this definition because when looking for align-ments we cannot a priori know the length of the seg-ment we look for.In the same way,we never considerevents like:“a segment has exactly k aligned points”,but rather“a segment has at least k aligned points”,andk must be given,as we do,by a detectability criterionand not a priorifixed.3.Number of False Alarms3.1.DefinitionDefinition3(Number of false alarms).Let A be asegment of length l0with at least k0points having theirdirection aligned with the direction of A.We definethe number of false alarms of A asNF(k0,l0)=N4·PS l≥k0=N4·l0k=k0l0kp k(1−p)l0−k.(11)Interpretation of this definition:the number NF(k0,l0) of false alarms of the segment A represents an upper-bound of the expectation in an image of the number of segments of probability less than the one of the consid-ered segment.Remark(*)(relative notion).Let A be a segment and NF(k0,l0)its number of false alarms.Then A is ε-meaningful if and only if NF(k0,l0)≤ε,but it is worth noticing that we could have compared NF(k0,l0) not toεbut to the real number of segments with prob-ability less than the one of A,observed in the image. For example,if we observe100segments of proba-bility less thanα,and if the expected value R of the number of segments of probability less thanαwas 10,we are able to say that this100-segments event could happen with probability less than1/10,since 10=E(R)≥100·P[R=100].Now,each of these 100segments only is10-meaningful!3.2.Properties of the Number of False AlarmsProposition1.The number of false alarms NF(k0, l0)has the following properties:1.NF(0,l0)=N4,which proves that the event for asegment to have more than zero aligned points is never meaningful!2.NF(l0,l0)=N4·p l0,which shows that a segmentsuch that all of its points have the“good”direction is ε-meaningful if its length is larger than(−4ln N+ lnε)/ln p.3.NF(k0+1,l0)<NF(k0,l0).This can be interpretedby saying that if two segments have the same length l0,the“more meaningful”is the one which has the more“aligned”points.4.NF(k0,l0)<NF(k0,l0+1).This property canbe illustrated by the followingfigure of a segment (where a•represents a misaligned point,and a→represents an aligned point):→→•→→••→→→→→•If we remove the last point(on the right),which is misaligned,the new segment is less probable and therefore more meaningful than the considered one.5.NF(k0+1,l0+1)<NF(k0,l0).Again,we canillustrate this property:→→•→→••→→→→→→If we remove the last point(on the right),which is aligned,the new segment is more probable and therefore less meaningful than the considered one.This proposition is a consequence of the defini-tion and properties of the binomial distribution(see Feller,1968).If we consider a length l segment(made of l inde-pendent pixels),then the expectation of the number of points of the segment having the same direction as the one of the segment is simply the expectation of the random variable S l,that isE(S l)=li=1E(X i)=li=1P[X i=1]=p·l.(12)We are interested inε-meaningful segments,which are the segments such that their number of false alarms is less thanε.These segments have a small probability14Desolneux,Moisan and Morel(less thanε/N4),and since they represent alignments (deviation from randomness),they should contain more aligned points than the expected number computed above.That is the main point of the following propo-sition.Proposition2.Let A be a segment of length l0≥1, containing at least k0points having the same direction as the one of A.If NF(k0,l0)≤p·N4,(which is the case when A is meaningful),thenk0≥pl0+(1−p).(13) This is a“sanity check”for the model.4.ThresholdsIn the following,εand p arefixed numbers smaller than1,and we use the notationP(k,l)=P[S l≥k]=li=klip i(1−p)l−i.(14)We recall that a segment of length l isε-meaningful as soon as it contains at least k(l)points having the“right”direction,where k(l)is defined byk(l)=mink∈N,P[S l≥k]≤εN4.(15)Thefirst simple necessary condition we can get is a threshold on the length l.For anε-meaningful segment, we havep l≤P[S l≥k(l)]≤εN4,(16)so thatl≥−4ln N+lnεln p.(17)Let us give a numerical example:if the size of the im-age is N=512,and if p=1/16(which corresponds to16possible directions),the minimal length of a 1-meaningful segment is l min=9.We can also give estimates of the thresholds k(l). The mathematical theorems are given in the Appendix. They roughly say thatk(l) pl+C·l·lnN,(18)Figure1.Estimates for the threshold of meaningfulness k(l).Themiddle(stepcase)curve represents the exact value of the minimalnumber of aligned points k(l)to be observed on a1-meaningful seg-ment of length l in an image of size512,for a direction precision of1/16.The upper and lower curves represent estimates of this thresh-old obtained by Proposition5and Proposition7(see Appendix).where2p(1−p)≤C≤1/2.Some of these resultsare illustrated by Fig.1.These estimates are not nec-essary for the algorithm(because P[S l≥k]is easyto compute)but they provide an interesting order ofmagnitude for k(l).5.Maximal Meaningful Segments5.1.DefinitionSuppose that on a straight line we have found a mean-ingful segment S with a very small number of falsealarms(i.e.NF(S) 1).Then if we add some“spuri-ous”points at the end of the segment we obtain anothersegment with probability higher than the one of S andhaving still a number of false alarms less than1,whichmeans that this new segment is still meaningful(seefigure).→→→→→→→→→→→→→→→→→→••••In the same way,it is likely to happen in general thatmany subsegments of S having a probability higherthan the one of S will still be meaningful(see experi-mental Section,where this problem obviously occursfor the DNA image).These remarks justify the intro-duction of the following notion of“maximal segment”.。