Pipette Method : Errors Resulting From Aliquot Collection Depth in Soil Clay Quantification

Granulometry represents the relative proportions of the fractions that compose the soil, being an important agronomic tool to infer mean values of density, water availability and cation exchange capacity, besides being useful in soil classification. Among the methods employed to determine the fractions composing the soil, those which consider the separation by sedimentation for the clay fraction still have problems in the analytical protocol, which are directly responsible of errors in the results obtained. Given the above, this study aimed to evaluate the best pipette immersion depth to collect the aliquot containing only clay, to calculate and discuss the errors associated with collection of the aliquot containing clay fraction in soil granulometric analysis. Samples for granulometric analysis were collected in the superficial layer and top of the B horizon of an Argissolo Amarelo, corresponding to the textural classes sandy loam and sandy clay. Regardless of soil textural class, the depth h = 5 cm established in the calculation using the Stokes’s equation leads to overestimation and underestimation of clay and silt fractions in the soil. The collection should be performed with the pipette tip positioned at h/2 = 2.5 cm.


Introduction
Granulometry represents the relative proportions of the fractions that compose the soil.In most cases, it is a reference to the quantities of sand, silt and clay, fractions which are not changed by soil use and management, because they result from pedogenetic processes.It has great agronomic importance because, based on it, it is possible to infer mean values of density (Sivarajan et al., 2018), water availability (Rabot et al., 2018) and cation exchange capacity, besides being useful in soil classification (Klein et al., 2013).Regarding laboratory protocols, there is a great variety of techniques in soil granulometric analysis, and approximately four hundred methods are known nowadays (Almeida, 2008).Among the most widely known methods, two deserve special attention for being the most commonly used: the pipette method (Robinson, 1922) and the hydrometer method (Bouyoucos, 1927).
Soil granulometric analysis is normally carried out in three steps: application of pretreatments to remove flocculating and cementing agents, sample dispersion, and quantification of fractions (Ruiz, 2005).The first step aims to remove organic matter, iron oxides, carbonates and soluble salts (Ruiz, 2005).Soil sample dispersion is conducted by using a combination of chemical processes and mechanical disaggregation (Durner et al., 2017).Sieving is used for the sand fraction and sedimentation for the smaller particles-silt and clay (Cunha et al., 2014b).
For the sand fraction, separation by sieving leads to results with better precision and accuracy; however, separation by sedimentation for silt and clay fractions still has several problems in the various methods used in the analytical protocol, which are directly responsible of errors contained in the results obtained in the measurement of these granulometric fractions (Durner et al., 2017).Incomplete dispersion of soil particles, associated with the inefficiency of the procedures used, leads to errors in the quantification of the clay fraction and, consequently, of silt and sand fractions (Cunha et al., 2014b).These experimental errors, depending on the magnitude, may cause serious consequences for the utilization of the soil granulometric analysis results, from the simplest applications, directly related to soil use and management, to the more complex ones, which involve the modelling of soil physical, chemical, biological and mechanical processes (Corá et al., 2009).
Pipette method is also one of the main methods used to determine clay content.The method is frequently used due to the technical impossibility of manufacturing sieves with adequate diameters for the dimensions of these particles (Dias, 2004).It is worth highlighting that clay particles in the soil exhibit a laminar shape and the mesh opening in sieve manufacturing considers the equivalent diameter of spherical particles.Thus, contents of clay particles in the soil are usually determined using a method that is based on the sedimentation velocity of the particles.The principle of this method is the quantification of clay content based on the weighing of samples, collected at a certain depth in aqueous medium within a given time of sedimentation.Among its limitations, special attention should be paid to the collection depth established in the calculations using the Stokes's law, a factor that leads to errors in the determination of clay contents by this method (Dias, 2004;Durner et al., 2017).
In this context, the hypothesis tested in the present study is that the aliquot collected at the depth h established in the calculation by the Stokes's law, leads to overestimation of soil clay fraction because sit also contains silt.Therefore, the objectives were to: i) define the best depth for pipette immersion to collect the aliquot containing only clay; and ii) calculate and discuss the errors associated with the collection of the aliquot containing the clay fraction.

Experimental Site
To perform soil granulometric analysis, samples were collected at the Federal University of Ceará, Campus of Pici, in the upper layer (0-5 cm) and from the first 5 cm of the B horizon (in both situations, in the first five centimeters) of a Argissolo Amarelo (Embrapa, 2013), corresponding sandy loam and sandy clay textures, respectively (Santos et al., 2013).All analyses were carried out at the Soil Physics Laboratory of the Soil Sciences Department in the Federal University of Ceará.

Pipette Immersion Depths for Aliquot Collection in the Quantification of Clay Fraction
For each treatment 10 samples of oven-dried fine earth (ODFE) weighing 20 g were used.Each sample was put in a 250 mL beaker to which 100 mL of distilled water and 10 mL of o 1 mol L -1 sodium hydroxide, a chemical dispersant, were added.The mixture was allowed to rest for 12 to 16 hours for chemical dispersion of particles.Following that, physical dispersion was carried out in vertical shaker at 12.000 rpm for 5 min for the sandy loam material, and 10 min for the sandy clay one.Subsequently, sands were separated in a sieve with 0.053 mm mesh.The filtrate, composed of clay and silt, was collected in a 1 L sedimentation cylinder.The cylinder volume was topped up with distilled water, and the suspension was agitated for homogenization per 20 seconds (Gee & Or, 2002).
According to the Stokes's law, the 5 cm depth containing only the clay fraction in suspension was considered.After the time of sedimentation, three procedures were used to collect the 25 mL aliquot containing clay notably (i) by suction, immersing the pipette tip to 2.5 and 5 cm depths from the liquid surface, (ii) by suction, immersing the pipette tip to 2.5 and 5 cm depths from the liquid surface, and (iii) by siphoning of solution containing clay + dispersant using a drain installed on the side of the sedimentation cylinder at 5 cm from the 1 L mark.(25 mL were collected from the drained volume.An illustration of the apparatus individually used in each treatment is presented in Figure 1.

The siphon suction, w
The aliquo the weight aliquot col where, Vt clay in 2.5 (g); and W

Statisti
For each s was condu three proce depth from suspension from the 1

Results
The     contain silt a ng counted as t at the depth h rotocol with s reas the proto y textural clas s mentioned ab e silt fraction on between bo between the m y similar to tho hod statistical Figure 4a).Th h = 2.5 cm in tion collection es it, the inter eous solution column h, it c automatic pipe r.Since the de nt in the fluid and clay parti clay, the pipe h used to calcu suction at h = ocol with suc s, there were bove. Figur