In situ Field Capacity in Brazilian Soils and a Derived Irrigation Management Practice Based on Water Suction

Field capacity (FC) is a fundamental parameter in soil and water engineering and hydrologic modeling. Despite its relevance, the in situ determination of this parameter is not standardized and its determination by indirect methods is dubious. This study presents a method of calculation of in situ FC and its corresponding water suction (hFC), using the van Genuchten equation for water retention and the pedotransfer function by Ottoni Filho et al. (2016) for standardized in situ determination of FC. The methodology was applied to HYBRAS, a database of hydrophysical data for Brazilian soils with 1,075 soil samples from 15 Brazilian states. FC and hFC were confirmed to depend on textural class and pedogenetic origin (weathered and unweathered soils). Our analysis justified why FC must not be determined based only on a single predetermined water suction value. A simplified method is proposed for the management of irrigated soils through the determination of water suction in the root zone and the mode and confidence interval values of hFC corresponding to soil groups formed from textural classes and pedological nature. Various statistical calculations of FC and hFC are presented for these groups.


Introduction
According to the Soil Taxonomy (1975), a soil is characterized by an arrangement of pores formed by inorganic and organic matter that serves as a support to plants in the field. It is an environment capable of having the air of its pores renewed with air from the atmosphere or filled with water. In irrigation engineering, soil is interpreted as being a reservoir of water and nutrients for crops. The main objective of its agricultural management is to provide optimum conditions for the development of plants with minimum impact to the environment. In this context, the concept of field capacity (FC) was introduced by Hendrickson (1931, 1949) as the soil moisture that corresponds to the maximum capacity of the soil to hold water available for use by plants, also characterized as the water content stabilized in the soil pores after the soil profile has been drained following an irrigation or rain event. This parameter has been largely used in hydrodynamic and hydrologic models involving soils (Kannan, White, Worral, &Whelan, 2007;Nasta & Romano, 2016;De Jong Van Lier, 2017), as well as in projects of irrigation and drainage systems and in water and soil management in general, including studies of groundwater recharge.
More specifically, Hendrickson (1931, 1949) defined FC as "the water content retained by the soil after an infiltration event and the drainage of the excess water, with a sharp decrease in the rate of downward water percolation". In their definition, these authors commented that this usually occurs two or three days after a rain or irrigation event. Similarly, the Glossary of Soil Science Terms (Soil Science Society of America, 2008) defines FC as "the water content retained in a uniform soil profile two or three days after it has been fully wetted and when free drainage in the root zone has become negligible". after a few days of drainage due to the fact that hydraulic equilibrium is not usually fully reached (Reichardt, 1988;Hillel, 1998;Romano & Santini, 2002;Reynolds, 2018). The main criticism falls on the lack of clarity and standardization of the explanation of hydraulic processes and field procedures related to the FC definition, which leads to significant differences between reported FC values, depending on the determination method used (Richards, 1960;Reichardt & Timm, 2004;Silva, Silva, Oliveira, Ferreira, & Serafim, 2014;Reynolds, 2018;Ribeiro, Costa, Silva, Franco, & Borges, 2018, Turek, Armido, Wendroth, & Santos, 2018. As a result, to maximize the standardization of a Veihmeyer and Hendrickson's reference method of determination of FC, Ottoni Filho et al. (2014) redefined the FC concept as being "the vertical distribution of the volumetric water content in the upper part of a soil profile that, in the course of ponded infiltration (of water from any source and with ponding depth smaller than 10 cm), becomes fully wetted at the end of infiltration and remains exposed to the subsequent process of drainage without evapotranspiration or rain for 48 h".
The definition above is considered more adequate because it keeps the original meaning of the FC concept expressed by the Glossary of Soil Science Terms at the same time that it establishes the drainage time and minimizes the inaccuracies and inconsistencies related to the expressions "uniform soil profile", "two or three days", " negligible drainage", "free drainage" that appear in the definition from the Glossary, which also omits the specification of the water application event in the soil profile, as well as the inexistence of rain or evapotranspiration after wetting. A consequence of the definition by Ottoni Filho et al. (2014) is that it leads to a greater standardization of the field test and of the method of determination of a reference FC. FC has been determined by the in situ direct method, that is, through experimental field infiltration and drainage or, generally, by the two following main subtypes of indirect methods: either pedotransfer functions (PTFs) or laboratory tests based mainly on the concept that FC is the soil water content associated with an arbitrary predetermined suction value.
To determine the FC value in situ, Embrapa (1979) advises the full wetting of the soil profile, that is, saturating it by inundation by applying a water depth to a 1mx1m soil area without vegetation. After wetting, the area must be covered with a piece of canvas or plastic to prevent further wetting by rain or water loss by evaporation, and until the FC is measured directly along the profile.
However, the indirect method most used to determine FC is the one that considers it to be the water content at a predetermined suction value, a moisture value frequently determined in the laboratory. Alternatively, this water content at a predetermined suction can also be obtained by PTFs. Usually, water suction values of 60, 100 or 330 cm associated with FC are adopted (Mello, Oliveira, Ferreira, & Lima, 2002;Reichardt & Timm, 2004;Ottoni Filho, Ottoni, Oliveira, Macedo, & Reichardt, 2014;Silva, Silva, Oliveira, Ferreira, & Serafim, 2014;De Jong Van Lier, 2017). In the laboratory, the method of equivalent moisture is also used. By this method, a centrifuge force one thousand times greater than the gravity force is applied for 30 min to previously saturated samples, which produces a water content called moisture equivalent that some researchers consider to be the FC (Cassel & Nielsen, 1986;Hillel, 1998;Ruiz, Ferreira, & Pereira, 2003).
Although the indirect method above, which equates FC to a water content associated with a predetermined suction value, is largely used by the scientific community, Reichardt (1988), Hillel (1998), Netto et al. (1999, Reichardt andTimm (2004), Ottoni Filho et al. (2016) and Reynolds (2018), among others, point out that these results are not representative of the actual FC of the soil profile and can at best be correlated to it. The reason is that the FC concept is derived from a specific water movement process by drainage through the soil profile and not from hydrostatic characteristics of the soil. The suction values of 60, 100 and 330 cm usually adopted in the determination of FC are considered arbitrary in the literature above and incompatible with the very definition of FC given by the Glossary of Soil Science Terms, not having any scientific grounds to be considered FC suction values. For example , Ottoni Filho et al. (2016) demonstrated that none of these three suction values in isolation represented in situ FC in their database adequately. Turek et al. (2018) analyzed various approaches of indirect determination of FC based on hydrodynamic and hydrostatic criteria, also adopting the suction values of 60, 100 and 330 cm. Their results showed that the FC values in general varied significantly depending on the determination criterion used.
The objective of the present study was to propose a method of calculation of a standardized in situ FC and its  To analyze the results, the 1,075 samples were separated into two groups -weathered and unweathered soils, as described above. In addition, the soils were separated again taking into account their textural classes. Three great textural class groups were considered, according to Cassel et al. (1983): fine texture, FT (silty clay loam, clay loam, silty clay, sandy clay and clay classes); mean texture, MT (sandy loam, loam, sandy clay loam, silt loam and silt classes); coarse texture, CT (sand and loamy sand classes). The textural classes in parenthesis are the same as in Figure 2.
Among other data, HYBRAS contains information on the five parameters (θs, θr, α, n, m) of the van Genuchten (VG) equation (Equation 1), which models the volumetric soil water content, θ (m³/m³), as a function of suction h (cm): Where, θ s and θ r are the saturation and the residual water contents, respectively, and α (cm -1 ), n (dimensionless) and m (dimensionless) are shape parameters of the water retention curve, θ(h).

Determination of the in situ Field Capacity (FC) and Its Associated Suction (h FC )
The in situ field capacity (FC) and the associated suction (h FC ) are determined using Equation 1 and the PTF by Ottoni Filho et al. (2016), which is better described next . Ottoni Filho et al. (2014 standardized the process for obtaining in situ FC using the FC definition expressed in their study and following the recommendations of Embrapa (1979), as was mentioned in the Introduction. FC data were collected at different depths of 29 soils from the state of Rio de Janeiro (Brazil) (207 samples) based on the direct measurement of FC from the soil profile and always following the same laboratory methodological protocol in all tests, thus reducing experimental inconsistencies. The pedological classification of the 29 profiles, with varied pedogenesis, is given in Ottoni Filho et al. (2016). The profile lengths where FC was monitored varied from 30 cm to 70 cm. Using this database (N = 207), a PTF was developed and the estimation of the in situ FC was made using only the soil moisture for the 60 cm-suction value ( C -clay, SC -sandy clay, CL -clay loam, SCL -sandy clay loam, SL -sandy loam, LS -loamy sand, S -sand, L -loam, SiC -silty clay, SiCL -silty clay loam, SiL -silty loam, Si -silt jas.ccsenet.org Journal of Agricultural Science Vol. 14, No. 3; determine in situ FC based on the soil granulometry, bulk density, organic matter content and/or soil moisture at a given suction (Macedo, Menegueli, Ottoni Filho, & Souza Lima, 2002;Nemes, Pachepsky, & Timlin, 2011;Ottoni Filho, Ottoni, Oliveira, Macedo, & Reichardt, 2014;Ottoni Filho, Leal, Macedo, & Reis, 2016;Ribeiro, Costa, Silva, Franco, & Borges, 2018), however, in general, with smaller efficiency than that given by Equation 2 with the 207 samples ( Figure 3). FC = 0.560·θ 60 2 + 0.576·θ 60 + 0.0436 (2) Where, FC and θ 60 are given in m³/m³ and θ 60 is the water content for the 60-cm suction.
We show below how FC and h FC can be calculated using the five parameters of Equation 1 and the three constants of Equation 2. Using Equation 2 requires that θ 60 be calculated with Equation 1: Where, Where, S FC is the effective saturation at FC.
For greater reliability of the results of Equations 6 (which calculates FC) and 7 (which calculates h FC ), two criteria were used to select and reject HYBRAS soil samples. water contents encompassed large suction ranges, usually from 30 cm to 15,000 cm (with few data in the 0-30-cm range). The water content at zero suction was always characterized in HYBRAS.
Where, θ j calculated and θ j measured are the calculated and measured water content associated with the N values of suction measured in the sample; the p value represents the number of optimized parameters of the VG equation, which, in the case of HYBRAS, was p = 4 (5-1), since parameter m was not optimized, but obtained from parameter n (m = 1 -1/n), as is usual.
The second sample rejection criterion used aimed at eliminating all the samples with calculated h FC or with the 60-cm suction (used in Equation 2) out of the range of measured non-zero suctions in the sample, in order to avoid h FC being determined by extrapolation of the optimized water retention curve, θ(h).
The To evaluate the suitability of the FC calculation proposed here, the non-parametric Wilcoxon test for paired groups (Bradley, 1968;Zar, 1984) was used, as well as the RMSE value calculated from Equation 10: (10) Where, FC j calculated and FC j measured are the calculated and measured FC values of the N samples (N = 77).
Considering only samples not eliminated by the two criteria above, the distribution of the calculated values of h FC and FC in the various possible groups of soils in HYBRAS, mentioned in Section 2.1, was described statistically taking into account all the combinations of the four textural groups (all soils, fine texture (FT), mean texture (MT) and coarse texture (CT)), with the three groups of pedological classes (all soils, weathered and unweathered). Traditional statistical measures were used in the description: mean, coefficient of variation (CV), median, mode, minimum, maximum and confidence interval. The Mann-Whitney non-parametric test for unpaired groups (Field, 2005) was used to compare the median values of h FC or FC involving pairs of subgroups of soils above. The normality of the statistical distribution was verified with the Shapiro-Wilk test (Shapiro & Wilk, 1965).

Field Capacity and Its Corresponding Suction in HYBRAS
To evaluate the quality of the FC results calculated with Equation 6, we compared the FC measured in situ with the calculated FC of the 77 samples from HYBRAS mentioned in the explanation of Equation 10. The results demonstrated that Equation 6 estimated FC well, since RMSE = 0.0255 cm³/cm³ (Equation 10) was not high and the Wilcoxon test showed that the measured FC was statistically indistinguishable from the calculated FC (p < 0.05), which indicates the inexistence of an estimation bias. The good performance of Equation 6, and, therefore, of Equation 2, in the calculation of in situ FC of the 77 samples is shown in Figure 4. Filtering the HYBRAS data using the two sample rejection criteria described in Section 2.2 allowed 842 samples for analysis of the results, approximately two thirds of which were weathered (554 samples) and one third unweathered (288 samples). These fractions roughly correspond to the total fractions of weathered and unweathered soils in the Brazilian territory (Embrapa-Spi, 2006). In the total, 474 samples had fine texture, 349 mean texture and 19 had coarse texture. Note. * One sample with h FC = 15,100 cm (h FC maximum) was eliminated from the calculation. ** Analysis not performed for CT soils in the weathered and unweathered soil classes due to insufficient data.
The h FC results considering all the textural classes (Table 1) gave a median of 157 cm, a mean of 188 cm and a CV of 64% for weathered soils (554 samples), in comparison to a median of 156 cm, a mean of 191 cm and a CV of 66% for unweathered soils (288 samples). The similarity between the statistics of weathered and unweathered soils indicates that the h FC distribution was little influenced by pedological class when texture was not considered. The same was observed for the FC results (Table 1). This fact is illustrated in Figure 5 by the great proximity of the plots of probability distribution of h FC for weathered and unweathered soils.  Figure 6a, which presents the clear influence of textural class on FC values, as expected. This textural influence is confirmed by the comparison of the medians with the Mann-Whitney test involving the three groups, corroborating (p < 0.001) a decreasing tendency of FC values from the FT group to the CT group. We can also see that the FC means practically coincided with the medians above in the three textural groups (Table 1 and Figure 6a). A relevant hydraulic parameter associated with FC is drainable porosity (DP) (also called field air capacity or specific yield), defined as the volumetric part of the total porosity of a fully saturated soil that percolates by drainage until the soil moisture reaches FC. That is, DP = total porosity − FC. The total porosity was considered here to be the water content at zero pressure in HYBRAS. Figure 6b depicts the DP median results for the three textural classes (along with the mean and coefficient of variation values). The results confirm that DP (0.147 m³/m³) was greater for CT soils than for MT soils (0.129 m³/m³), which is expected and confirmed by the Mann-Whitney test (p = 0.016). However, the DP median of FT soils (0.173 m³/m³) was greater than that of CT soils, which is unexpected, since in general sandy soils are thought to drain a greater volume of water than clayey soils (Davis & DeWiest, 1966;Hillel, 1998). This unexpected result was also confirmed by the Mann-Whitney test, as it did not show any statistical difference (p = 0.528) between the DP medians of subgroups CT and FT. It also confirmed (p < 0.001) that the DP values of the FT group tend to be greater than those of the MT group (0.173 m³/m³ vs. 0.129 m³/m³).
To clarify the two unexpected results above, we hypothesized that the nature of the clay fraction in weathered Brazilian soils influences DP, since drainability greatly depends on the saturated hydraulic conductivity and water retention of the soil, which are acknowledged to differ between weathered tropical soils and temperate climate soils, as demonstrated by Tomasella et al. (2000) and Ottoni et al. (2018Ottoni et al. ( , 2019. According to these authors, the distinct nature of weathered clays from tropical climate in relation to temperate climate clays results in a peculiar granular aggregation of clayey weathered soils, forming a pore structure somewhat similar to that of coarse textured soils. This characterizes the so-called hybrid behavior of weathered tropical clayey soils, which gives them aeration, permeability and water availability characteristics similar to those of sandy soils. This would justify the fact that FT soils in HYBRAS tend to have DP values similar or even higher than the DP values of CT and MT soils, as confirmed in the previous paragraph. To test the hypothesis made in this paragraph, Table 2 gives the DP statistics for a subdivision of the FT subgroup of weathered soils considering only the clay textural class (311 samples). In Table 2, DP5 and DP95 represent the drainable porosity which is exceeded at 95% and 5% of probability, respectively. That is, according to Table 2 there is just a small chance of 5% of DP values being smaller than 0.10 m³/m³ in weathered clays from HYBRAS, which would be a surprising result in a database of clays from temperate climate. Table 2 also shows that both the median and the mean of DP of these clays are around 0.20 m³/m³, a value typical of sandy materials reported in the literature on temperate climate soils (Davis & DeWiest, 1966;Hillel, 1998). All this indicates that the nature (weathered or unweathered) of Brazilian clays influences soil drainability, that is to say, FC values.  Note. * DP5 and DP95 represent the drainable porosity exceeded at 95% and 5% of probability, respectively.
Evaluating the statistics in Table 1 and consistently with that pointed out in the previous paragraph, we observe that the median of h FC for FT samples of unweathered soils (219 cm) is greater than the median of weathered soils (173 cm), which also applied to the means (251 cm and 209 cm, respectively). The same was observed for FC in the FT group, which had a greater median (0.388 m³/m³) and mean (0.411 m³/m³) in unweathered soils than in weathered soils (0.348 m³/m³ and 0.347 m³/m³, respectively). The h FC and FC medians and means of MT soils were also greater in unweathered soils than in weathered soils (Table 1), which demonstrates the influence of pedogenetic origin on the h FC and FC values in soils of similar texture. The Mann-Whitney test of comparison of medians also confirmed that the h FC and FC values tended to be greater (p < 0.011) in unweathered soils in relation to the corresponding weathered soil subgroups in the comparisons above.
As to FC, while the mean and median values were close in HYBRAS (Table 1), this did not happen for h FC , which had means about 20% greater than the medians, except in the CT soil group. This implies an asymmetrical statistical distribution of h FC values, with the distribution tail tending to high values, as shown in Figure 7. However, the proximity of the mean and median values of FC did not imply a normal statistical distribution of FC in HYBRAS, according to the Shapiro-Wilk test (N = 842, p < 0.001). The h FC values were rather variable in all soil subgroups of HYBRAS, with coefficients of variation around 60% (Table 1), except in the CT class (CV = 21%). This makes h FC a highly variable property (Warrick, 1998) and points to great errors of estimation of FC when the indirect method of estimation with a single arbitrary and predetermined value of h FC is used. The FC values in general had a smaller variation in HYBRAS than h FC , with a CV around 25% in all the subgroups (Table 1). The minimum FC value in HYBRAS was 0.119 m³/m³ (a sand from the Podzol class, with a high DP value of 0.315 m³/m³), and a maximum of 0.628 m³/m³ (a clay from the Gleysol class, with a low DP value of 0.064 m³/m³). The minimum value of h FC was 60 cm (for the same sand above) and the maximum was 15,100 cm, as previously mentioned. Further information on h FC in HYBRAS is given in Section 3.2.
We previously mentioned that the statistical measures calculated for FC and h FC in CT soils (Table 1) must be taken with caution due to the small number of samples (N = 19) in this group. In order to evaluate the quality of these statistical data in HYBRAS, all the methodology described here for the determination of FC and h FC was repeated in CT soils from HYPRES (Hydraulic Properties of European Soils) database (Wösten, Lilly, Nemes, & Le Bas, 1999; personal communication with Dr. Allan Lilly, one of the authors of the paper above, who kindly granted us access to HYPRES data). As its name suggests, HYPRES is a database with only European soils. Its CT soils, such as those in HYBRAS, also have N > 4, where N is the number of experimental data pairs in the water retention curve. Because the clay content of CT soils is rather low (lower than 15%), these soils are expected to present a pore structure simpler and less dependent on pedological conditions than those of other textural classes. Thus, we hypothesized that the pedogenetic conditions of CT soils from HYPRES do not exert a significant influence on the efficacy of determination of FC using the methodology adopted in this study, based on an equation (Equation 2) developed for Brazilian soils. This would justify using Equations 6 and 7 to determine FC and h FC in all CT soils from HYPRES, and would enable the comparison of statistics of CT soils from HYBRAS and HYPRES.
When the same sample selection criteria described in Section 2.2 were applied to HYPRES CT soils with the VG equation parameters of HYPRES, it was possible to calculate the FC and h FC values of 52 samples, a much larger number than the 19 CT HYBRAS soil samples. Table 3 compares the FC and h FC statistics of CT soils from HYBRAS and HYPRES. We can see that the median, mean and CV values of the two databases are very close. The Mann-Whitney test confirmed that the FC and h FC medians of CT soils from HYBRAS and HYPRES are statistically equal (p > 0.682). This lends further support to the statistical values of CT soils presented in Table 1.

Proposal for the Management of Irrigated Soils
For the use of FC and h FC in mathematical modeling or soil management, it is recommended that the FC value be measured preferably through in situ testing employing a standard experimental procedure, such as that suggested by Ottoni Filho et al. (2014, so that more appropriate and consistent values of FC and h FC can be obtained. If a direct determination of FC and h FC is not viable due the complexity of in situ testing, a second option is presented here. Undisturbed soil samples from the soil horizons of interest can be taken to the laboratory for analysis and their water contents can be determined at different suction values, but preferably in the 50-1,000-cm range, the predominant range where h FC was determined in this study. Next, the parameters of an appropriate E equation of representation of the water retention curve (such as Equation 1) of each sample must be provided. After that, the methodology of this study can be applied to determine the values of FC and h FC in the horizons wanted. Due to the great variability of h FC , as demonstrated in Section 3.1, we strengthen the inconvenience of adopting the usual method with a single arbitrary value of h FC in the estimation of FC, a method also largely criticized in the literature, as already mentioned.
If both options of the previous paragraph are unviable, a simplified method of irrigated soil management is proposed using only the in situ determination of water suction (using a tensiometer, for example) at a depth of relevant water extraction in the root zone, and some h FC statistics from HYBRAS. Our intention is to propose a simplified soil management tool that is easy to use, since specific plant and climate conditions are not taken into account, but rather, as we shall see, only textural and pedological characteristics of the soil.
The method is based on the 90%-probability confidence interval of h FC and on the most probable h FC value (mode), for eight subgroups of soils from HYBRAS already analyzed, organized according to their textural and pedogenetic characteristics (Figure 8). According to a previous analysis (former paragraph where Figure 5 was mentioned), in Figure 8 it is not necessary to differentiate soils in the weathered and unweathered subgroups when textural class information is unknown. Pedogenesis subdivisions are not presented for coarse soil texture either, due to a lack of data. jas.ccsenet.

Figure capacit
Note. wever, e, the m in n the n the suction value is greater than 290 cm (Figure 8), regardless of the pedological class. A field suction lower than 70 cm for two or more days after irrigation (weathered soils), 100 cm (unweathered soils) or 75 cm (pedological nature unknown) indicates that drainage is insufficient. As proposed, suctions of 100 cm (weathered soils) or 120 cm (unweathered soils) must be the guideline values to be attained in the field resulting from the irrigation of MT soils.
The influence of the pedological nature shown in Figure 8 can be better observed in FT soils, which have greater clay contents, and, in turn, most influence the soil structure. Considering this, the recommended range of irrigation management, corresponding to the confidence interval of h FC , varied from 95 cm to 510 cm when the pedological nature is unknown, and from 100 cm to 410 cm in weathered soils or 70 cm to 800 cm in unweathered soils. In the latter case, if tensiometers are used, it is recommended to irrigate unweathered FT soils only when the instrument reading is nearly at its practical highest limit of use, which is around 800 cm. Likewise, the mode of h FC in this textural class changed from 140 cm in weathered soils to 220 cm in unweathered soils; these must be the guideline suction values for wetting the soil in the field in both cases. These results justify the differentiation in subgroups according to the pedological nature.
The analysis above reveals that the pedological class has a greater influence on the h FC mode and confidence interval in the textural group with a greater clay content (FT soils) than in the group with a smaller clay content (MT soils), which is expected. Therefore, it is reasonable to hypothesize that the influence of pedological class on CT soils, which have very low clay contents, is secondary with respect to the irrigation management method proposed, especially considering that the confidence interval of h FC has a reduced amplitude (from 60 cm to 120 cm) in CT soils when the pedogenetic origin is unknown.

Conclusions
When field capacity (FC) is not determined in situ it has been calculated using various methodologies usually inconsistent with each other, which is unfortunate. The most popular method is that which estimates FC from a predetermined arbitrary value of suction without any theoretical or experimental basis. In the present work we have calculated an standardized in situ FC and its corresponding suction (h FC ) for 842 soils from HYBRAS, a broad database of hydrophysical data of Brazilian soils, using the van Genuchten water retention equation (Equation 1) and the PTF of Equation 2. This PTF was calibrated for 207 samples of Brazilian soils with varied pedogenesis in order to estimate in situ FC values measured by a standard field test of water application and drainage. The methodology proposed in our work calculated in situ FC successfully for 77 samples from HYBRAS with small errors and without bias. As the PTF of Equation 2 was developed for Brazilian soils, for caution, it is recommended to apply this methodology to determine FC and h FC only in these soils. However, as the PTF above is based only on θ 60 = θ (h = 60 cm), a variable dependent mainly on the soil pore structure, we can argue that this methodology also applies to other pedological environments, which must be evaluated experimentally.
The statistical analysis of the distribution of FC and h FC values in various soil groups of HYBRAS taking three textural classes (fine, mean and coarse texture) and two pedological groups (weathered and unweathered soils) into account leads to the conclusion that both soil granulometry and pedological nature influenced the FC and h FC values. The FC and h FC means ranged from 0.19 m³/m³ and 79 cm for the coarse texture group to 0.41 m³/m³ and 251 cm for the unweathered fine texture group. The h FC varied greatly within groups, with coefficients of variation around 60% in general, which confirms the lack of consistency of the method used to determine FC from an arbitrary predetermined h value. The most probable value (mode) of h FC also varied among groups, from 80 cm (coarse texture) to 220 cm (unweathered fine texture).
When the methodology proposed for the determination of FC and h FC cannot be applied, we suggest a simplified method of management of irrigated soils through the determination of suction in the field (using tensiometers, for example) and the knowledge of the group of the soil in hand. This simplified management is based on the confidence interval of h FC (at 90% probability) and on the mode value of h FC , corresponding to the specific soil group (Figure 8). The high limit of these confidence intervals ranged from 120 cm in the coarse texture group to 800 cm in the unweathered soils of fine texture. The low limit ranged from 60 cm (coarse textured soils) to 100 cm (weathered fine textured soils or unweathered mean textured soils). When the suction determined in the field is out of the range of the confidence interval, it is indicative that irrigation is needed or the soil profile drainage is deficient. As proposed, the mode value serve as a guideline suction to be attained for the adequate wetting of the soil profile through irrigation.

Copyrights
Copyright for this article is retained by the author(s), with first publication rights granted to the journal.
This is an open-access article distributed under the terms and conditions of the Creative Commons Attribution license (http://creativecommons.org/licenses/by/4.0/).