研究结果能否被重复是科学研究的基本要求。材料与方法部分描述研究是在何种物质条件下怎样展开的，是快速判定研究结果能否被重复的重要途径，其主要内容包括对材料进行描述和对方法进行描述两部分。

对材料进行描述指对研究所用材料进行描述，其内容包括对材料的概述，对材料的结构、主要成分、重要特性及设备的功能等的详述。对方法进行描述指按研究步骤的时间顺序或重要程度来描述研究所用的方法，其内容包括环境或条件，研究对象选择的方法，选用特定材料、设备或方法的理由、步骤或程序，所用的统计分析方法等。对于实验方法来说，主要阐述有关实验仪器、实验设备、实验条件和测试方法等事项，并介绍主要的实验过程，涉及实验对象，实验材料的名称、来源、性质、数量、选取及处理方法，实验目的，使用的仪器、设备（包括型号、名称、测量范围及精度等），实验及测定的方法和过程，出现的问题及采取的措施等。

材料与方法部分描述的内容通常较多，而且具体内容因研究领域、范围、主题等的不同会有很大的差别，不同出版物的要求也不尽相同。因此，这部分在写作上没有统一的规范，也很难统一要求。下面只阐述材料与方法部分规范写作的一些通用原则。

1.要清楚地指出研究对象的数量、来源和准备方法。给出诸如实验所用原料或材料的技术要求、性质、数量、来源，以及材料的选取、处理、制备方法等信息，有时甚至还要列出所用试剂的有关物理、化学性质。通常应使用通用、标准的名词、名称和术语，避免使用商业名称。

2.要结合各有关方法来描述。避免机械地按年、月的次序来描述，要使有能力的科技工作者按这部分提供的信息复用文中所述的实验与实验结果，达到对实验结果的“再现性”或“可重复性”的要求。

3.要具体真实地进行阐述。采用前人的方法时，只需注明引文出处；改进前人的方法时，须交代改进之处；提出自己不曾发表过的方法时，须详细说明，必要时可用示意图、方框图或照片图等配合表述，尽可能提供全部所需的细节。

材料与方法部分要描述得尽可能详细，此部分大多以过去时态来表达，并且通常都有副标题（Subheadings）。想要知道是否可在材料与方法部分使用副标题及使用哪一种形式的副标题，应该查看欲投稿期刊上的类似论文。要尽可能使材料与方法部分的副标题与结果部分中出现的内容前后一致。如果能使论文前后一致，在撰写论文的这两部分时都会容易不少，同时读者也能很快将论文中的某个方法与结果部分的相应内容联系起来。

常犯的错误

论文作者常犯的错误是把本应放在结论部分的内容放在材料与方法部分。衡量材料与方法部分优劣的唯一准则就是：该部分是否给出了足够的信息，使得具备一定能力的同行能够重复论文中的实验。

在语法和标点符号上犯错误。在引言部分和讨论部分，在描述一般性的概念时偶犯语法或标点符号上的错误也不会影响读者理解论文。但是在材料与方法部分，凡事都应力求准确与精确，所以在这一部分必须注意避免语法和标点错误。有时即使是漏掉一个逗号都会造成大错，如“Employing a straight platinum wire rabbit，sheep and human blood agar plates were inoculated...”，这个句子一开头就因使用虚悬分词而出现语法错误，但还不至于让人无法理解；但当作者在“wire”后面漏用逗号后，句子的整体意思就完全改变了。

很多写作书籍都建议写作时少用被动句式。但在科技论文的材料与方法部分，被动句式却很常用，因为这部分更注重的是做了什么，而不是谁做的。比如，这部分会常用类似“Mice were injected with...”这样的句式，而不用“I injected the mice with...”，“A technician injected the mice with...”或“A student injected the mice with...”这样的句式。这部分还常用“We injected...”这样的句式，即使有时只有一个人完成这项科研工作（尽管有人坚信期刊论文不应使用第一人称，但很多期刊还是允许出现“I”和“We”）。

有时作者在材料与方法部分提供的信息过分简短，以致一些必要的细节被省略。最常见的错误就是在应该明确指出动作及动作的实施者时，却忽略了后者。比如这个句子：“To determine its respiratory quotient，the organism was...”，初看好像是指出了动作的实施者“the organism”，但实际上这个句子有语法错误，因为“the organism”不可能具备做出一个决定的能力。还有这样一个句子：“Having completed the study，the bacteria were of no further interest.”。真令人难以相信细菌能“completed the study”；如果细菌在“completed the study”后不再有“further interest”，这些细菌就显得有些忘恩负义了。

“Blood samples were taken from 48 informed and consenting patients...the subjects ranged in age from 6 months to 22 years.”（Pediatr.Res.6∶26，1972），这句话倒没有什么语法错误，但读者看完后会觉得奇怪，6个月大的婴儿怎么能发表同意接受实验研究的声明。

例：

2 Materials and Methods

2.1.Plant material and site description

The experiment was conducted in an orchard planted with Conference pear trees on a Quince Adams rootstock，situated in Belgium，Sint-Truiden（50°.45′59.46″N，5°9′24.68″E）.Belgium is situated in a temperate climate zone with frequent rainfall events and a relatively low evapotranspiration during the growing season.Average rainfall in Belgium during the growing season from Aprilto August is 67 mm/month，average reference evapotranspiration（ETo）is 85 mm/month.However in 48%of the years between 1959 and 2012，rain deficits of 60 mm/month occurred.The trees were planted in 1996 with a planting distance of 3.5 m by 1.5 m.The average tree height was 3.5 m.The trees were trained in a free spindle system.The orchard was situated on a uniform silt loam textured soil.The organic carbon content in the upper soil layer（0-23cm）was 1.1%.Rainfall was recorded on site；Etowas calculated using the Penman-Montheith equation（Allen et al.，1998）based on data recorded in a regional weather station at 20 km from the site.In the orchard a drip irrigation system was installed with line drippers every 20 cm with a discharge rate of 2 L/h.Distance between the line drippers and the trunk was 35 cm.Management practices such as pruning，disease control，fertilization and mulching were carried out in the same way as in a commercial orchard.The EC of the irrigation water was 0.87 dSm-1at 25℃.

2.2.Soil water potential（Ψsoil）observations

Three plots（plot A，B and C）in the centre of the orchard were selected for the experiment.Every plot consisted of four trees within the centre one tree around which Watermark sensors were installed（Fig.1）.Sensors were installed on six positions perpendicular to the tree line.The numericalΨsoil calculations were executed in 2D in the plane XZ，with X being the horizontal coordinate perpendicular to the tree line and Z being the vertical coordinate.The calculation of Ψsoilin 2D is a simplification of the reality but was done to ease the computation time.Previously the calculation of water distribution after drip irrigation，with the drippers in line，has been calculated successfully in 2D in a plane perpendicular to the drip line（Skaggs et al.，2004；Zhou et al.，2007）.All sensors were installed at a depth of 30 cm in search of a gradient in Ψsoill independent from suction due to gravity.It is expected that root concentration is highest in the soil layers close to 30 cm depth.Installing more sensors in the root zone would possibly disturb the soil too much for a representative experiment.To supply information on water content in the deeper soil layers gravimetric soil moisture samples were taken at a depth of 30-60 cm，at reasonable distance from the sensors to prevent further soil disturbance.The Watermark sensors were connected to a data logger which recorded Ψsoilevery 4h.The standard manufacturer calibration was used to compute Ψsoil from the electrical resistance measured by the sensors.In every plot the sensors were brand new and used for the first time.Sensors were installed 1 day before the start of the observation period according to manufactory guidelines.In plot A Ψsoilwas recorded in 2009 while in plot B and C Ψsoilwas recorded in 2011.In the irrigated plots irrigation was scheduled using the Watermark sensors.Irrigation was initiated when Ψsoil decreased to-40 kPa，the irrigation dose ranged between 1 and 3 mm/day.

2.2.1.1.Plot A Ψsoilobserved in 2009 in an irrigated plot

In plot A soil was observed between 04/06/2009 and 15/08/2009.Sensors were only installed at positions 2，3，4 and 5 according to Fig.1.Total irrigation amount during this period was 77 mm，132 mm rainfall was recorded and total EToduring this period was 255 mm.

2.2.1.2.Plot B Ψsoilobserved in 2011 in an irrigated plot

In plot B　Ψsoilwas observed between 20/04/2011 and 15/07/2011.Sensors were installed at positions 1，2，3，4，5 and 6 according to Fig.1.Total irrigation amount during this period was 45 mm，112 mm rainfall was recorded and total ETo during this period was 300 mm.

2.2.1.3.Plot C Ψsoilobserved in 2011 in a non irrigated plot

Similar to plot B Ψsoilwas observed between 20/04/2011 and 15/07/2011.Sensors were installed at positions 1，2，3，4，5 and 6 according to Fig.1.In this period 112 mm rainfall was recorded and total ETo during this period was 300 mm.In plot C no irrigation was supplied to assure lower Ψsoil values in one of the three experimental plots.

Fig.1.Schematic top view with positions of the Watermark sensors which recorded Ψsoilin every plot on six positions on the axis perpendicular to the tree line at a depthof 30 cm.

2.3.Soil water content（θ）

In plot B and C soil water content（θ）was measured on the irrigated side and the non irrigated side（Fig.1）of the trees with gravimetric moisture samples.Samples were collected on 30/06/2011，05/07/2011 and 12/07/2011，towards the end of the observation period.Samples were taken with a gauge auger of 30 cm in the soil layers 0-30 cm and 30-60 cm.One sample consisted of minimal eight subsamples taken randomly within the treatment in the weed free strip beneath the canopy of the four trees within a plot.Gravimetric water content was measured by drying the samples at 105℃during 24 h.

2.4.Soil hydraulic properties

The retention points were measured on pressure plates at 0，-10 kPa，-20 kPa，-31.6 kPa，-70.8 kPa，-100 kPa，-200 kPa and-1600 kPa on soil samples taken in four replications at 30 cm depth and at 60 cm depth.Saturated hydraulic conductivity（Ksat）was measured in situ in four replications with the inversed auger hole method（Kessler and Oosterbaan，1974）and was 144±24 cm day1for the soil layer 0-70 cm and 20±0.1 cm day1for the soil layer 0-200 cm.

2.5.Root distribution

2.5.1.Contours of the root zone

To estimate the maximal contours of the root zone in the horizontal（X）and vertical（Z）directions，the central tree in plot A was excavated in January 2010 using low water pressure.The architecture of the coarse root system was measured in the lab with a compass，inclinometer and calliper and registered in the software ARCHIROOT（Dupuy.2003，www.archiroot.org.uk）which translates the measurements in a multi-scale tree graph（MTG）developed by Godin and Caraglio（1998）.The MTG is a multi-scale presentation of the tree，or root system and permits representation and analysis in a grid with the plant architectural model PlantGL（Pradal et al.，2009）.

2.5.2.Fine root distribution

To obtain fine root distributions cylindrical soil cores of 880 cm3 were sampled.Roots were washed from the soil using fresh water.All roots with a diameter＜2 mm were weighted with an accuracy of 0.001 g.Root length of these fine roots was determined on photographic scans of the roots with the ASSESS software（Lamari，2002）.Cores were taken at the six positions were the Watermark sensors were installed（Fig.1）to a depth of 90 cm for plot A and to a depth of 45 cm for plot B and C.The height of the soil cores was 15 cm.In plot A 36 soil cores were taken，in plot B and C 18 cores per plot.

2.6.Numerical calculations with HYDR US(www.daowen.com)

2.6.1.Boundary conditions

An automated mesh was generated with HYDRUS 2D with boundaries such as the atmospheric boundary，free drainageboundary，no flux boundary and the variable flux boundary to account for drip irrigation（Fig.2）.

Fig.2.Mesh generated in HYDRUS with the selected boundary conditions.

Table 1 Soil properties used in HYDRUS simulation.

θris residual water content considered at h=-16，000 cm=-1600 kPa，θs is saturated water content measured at h=0 cm=0 kPa and Ksatis saturated hydraulic conductivity.

αd reflects the inverse of the air-entry value during drying of the soil，αw during wetting of the soil，n pore size distribution and I pore-connectivity（Van Genuchten，1980）.

2.6.2.Hydraulic soil properties

For establishing the K（h）relationship expressed in Eq.（1）the soil hydraulic functions described by Van Genuchten（1980）were used as implemented in HYDRUS:

where，Seis the effective water content，θr residual water content considered at h=-16000 cm=-1600 kPa，θssaturated water content measured at h=0 cm=0 kPa and Ksat（cm/day）is saturated hydraulic conductivity.The parameters α and n according to VanGenuchten（1980），necessary for the HYDRUS calculation，were fitted through measured water retention points（Table 1；Fig.3a）.In the simulations hysteresis was considered and the wetting curve differed from the drying curve，αwfrom the wetting curve equaled two times αd from the drying curve after Kool and Parker（1987）.

2.6.3.Evapotranspiration

Evaporation and transpiration was combined in the calculation as evapotranspiration（ET）.Evapotranspiration（ETc）of the tree was estimated with the crop specific Kc factor and Eto（Allen et al.，1998）.

The Kc factor was assumed to be 1.06 times higher than that measured by Girona et al.（2004）（Fig.3b）who obtained a‘Conference’pear tree Kc factor in a lysimeter in an orchard planted where distance between the trees in the row was 1.06 lower than the orchard in Sint-Truiden.This assumption was made due to the lack of actual measurements of light interception.The water stress response reduction function，α（h）in Eq.（2），used in the study is described by Feddes et al.（1978）as implemented in HYDRUS:

where，h1，h2，h3 and h4 are four critical pressure heads for root water uptake.Eq.（6）is displayed in Fig.3c.h1is water content at saturation of the soil ath=0 cm=0 kPa，h4 is wilting point at h=-16，000 cm=-1600 kPa.In this study the threshold for water stress h3 was set to-400 cm=-40 kPa independently from transpiration.The threshold for waterlogging h2 was set to-10 cm=-1 kPa.

2.6.4.Initial conditions

Initial conditions were calculated with HYDRUS for a simulation period prior to the Ψsoilobservation period.Soil was calculated between 01/04/2009 and 03/06/2009 prior to the Ψsoilobservationsin plot A.Ψsoilwas calculated between 01/04/2011 and 19/04/2011 prior to the Ψsoilobservations in plot B and C.On April 1st the soil was assumed to be at Field Capacity，-10 kPa，over the entire flow domain.April 1st can be considered as the end of winter in Belgium and the beginning of‘Conference’pear transpiration.

2.6.5.Root distribution

The normalized root distributionβ（x，z）can in HYDRUS be described by the following function proposed by Vrugt et al.（2001 a）：

where，xm（m），zm（m）maximum rooting depths in the X-and Z-direction，x and z are distances from the origin in the X-and Z-direction.px，pz，x*and z*are empirical parameters.

Maximum rooting depths（xm and zm）were for all plots derived from the coarse root excavation and the RLD observations to-90 cm in plot A.Maximal rooting length in X-direction（xm）was assumed 2 m and maximal rooting depth in Z（zm）direction was assumed 0.9 m.For each plot the empirical parameters px，pz，x*and z*were first parameterized based on the observations of root length density of the fine roots（cm/cm3）（RLD）.Next the function was parameterized based on the observations of root weight density（g/cm3）（RWD）.Thirdly the function was parameterized based on two root distributions found in literature.Root observations in literature for pear tree are scarce but root distribution of apple was sampled by various authors.Gong et al.（2006）documented RLD observations for a 7 year old apple tree on a loam soil.Besharat et al.（2010）documented RLD observations for a 6 year old apple tree on a clay loam soil.Both root distributions were used to parameterize root density in the present Ψsoilcalculations.

This way four Ψsoilcalculations per plot where executed：（1）Ψsoilcalculated withβ（x，z）based on observed RLD，（2）Ψsoilcalculated withβ（x，z）based on observed RWD，（3）Ψsoilcalculated withβ（x，z）based on RLD observations of Gong et al.（2006）and（4）Ψsoilcalculated withβ（x，z）based on root RLD observations of Besharat et al.（2010）.Purpose of the four simulations was to evaluate to what extent an in situ observation of root distribution contributes to a good calculation of Ψsoilwhich was one of the objectives of the study.

2.6.6.Comparison between observation and calculation

Each Ψsoilcalculation in the flow domain was registered with six observation nodes placed at the location of the WatermarkΨsoil sensors.The average daily output of the Watermark sensor was compared with the average daily Ψsoilcalculated on the corresponding observation node.The coefficient of determination（R2）and the root mean square error（RMSE）were used to quantify the quality of the simulation.

Besides Ψsoilalso average θ in the flow domain was calculated in the time steps when θ was measured.Average θ and observed θ was compared for the soil layers 0-30 cm and 30-60 cm.

Fig.3.The fitted drying and wetting curve using Van Genuchten（1980）for the soil layer（0-30 cm（）a），the crop facto（rKc）derived from Girona et al.（2004）which relates reference evapotranspiratio（nETo）to maximal crop evapotranspiration（ET（c）b）and the water stress response function as used by Feddes et al（.1978）and as used in the calculation ofΨ soi（lc）.

2.7.Plant water status

Plant water status was recorded in plot B and C on three trees per plot by measurements of sap flow and stem water potential（Ψstem）.Objective was to observe possible water stress in the non irrigated plot C and to see whether it was reflected in the HYDRUS calculation of root water uptake.

2.7.1.Sap flow

Sap flow was monitored with thermal dissipation（TD）probes.Two needles of 2 mm diameter and 20 mm long were inserted in the trunk 10 cm apart.The upper probe was heated with a constant power of 0.2 W.Based on the temperature difference between the two needles sap flux density（Jp，m3m-2s-1）was calculated according to Granier（1985）who derived an empirical relationship between Jp and a dimensionless flow index K.

where T is temperature difference between the two needles To is the temperature difference under zero flow conditions which was taken as the temperature difference at night.

Sap flow observations using the TD technique cannot be considered as an absolute estimate of sap flow or sap flux density（Gonzalez-Altozano et al.，2008；Steppe et al.，2008）.The major drawbacks of the technique are that Eq.（9）is an empirical relationship which can differ between tree species.Furthermore the basic assumptions using this technique are debatable:uniform sap flow in the entire conducting sap wood area，zero sap flow at night and no vertical temperature gradient.Consequently water stress can only be detected by comparing a well irrigated plot，in this case plot B，with less irrigated plot，in this case plot C.According to Fernandez et al.（2008）this approach leads to satisfactory water stress observations.

Fig.4.Rainfall，irrigation and Ψsoilobservations with Watermark sensors at a depth of 30 cm in plot A（a），plot B（b）and plot C（c）.Position of the sensors is outlined in Fig.1.

Fig.5.Schematic of coarse roots of the tree in plot A obtained after excavation of the tree.Root architecture was measured in the lab with a compass，inclinometer and calliper and registered in the software ARCHIROOT（Dupuy，2003）.

2.7.2.Stem water potential（Ψstem）

On three trees in plot B and C　Ψstem measurements were carried out on 06/05/2011，24/05/2011，30/06/20111 and 08/07/2011 on sunny days without rainfall.For each measurement three leaves per tree were selected from the inner part of the canopy.While still being attached，these leaves were enclosed in plastic bags covered with aluminium foil.After 60 min，the leaves were detached and theΨstem was determined immediately using a pressure chamber（Scholander et al.，1965）.The Ψstem was only recorded on sunny days without rainfall.Measurements were performed between 13.00 h and 15.00 h.