Elsevier

Geoderma Regional

Volume 21, June 2020, e00277
Geoderma Regional

Estimating water retention and availability in cultivated soils of southern Brazil

https://doi.org/10.1016/j.geodrs.2020.e00277Get rights and content

Highlights

  • We generated pedofunctions to estimate water retention in subtropical, tilled soils.

  • Soil granulometry, consistency limit and cation exchange capacity affect soil water.

  • Subtropical soils need specific pedofunctions for water retention and availability.

  • Water retention pedofunctions are more accurate than available water pedofunctions.

Abstract

The objective of this study was to generate pedotransfer functions (PTFs) to estimate water retention and availability in conventionally-tilled soils and validate some PTFs available in the literature. Disturbed samples were collected from 34 subtropical soil profiles, corresponding to 85 horizons, to determine soil chemical and physical properties and PTFs were generated and validated. The correlations with physical and chemical soil properties were higher for soil water retention at specific tensions (r from 0.36 to 0.92) than for availability (r from 0.13 to 0.66). Water retained at 10, 33, 100 and 1500 kPa tensions may be accurately (i.e. R2 > 0.85) estimated by PTFs based on soil chemical and physical properties as predictive variables, but available water capacity may be estimated only with medium or lower accuracy (i.e. R2 < 0.85). Of those available in the literature, only PTFs of Giarola et al. (2002) at 10 kPa tension, of Masutti (1997) and of Oliveira et al. (2002) at 33 and 1500 kPa tensions for tropical soils, and of Reichert et al. (2009) at 10, 33 and 1500 kPa tensions for subtropical soils, estimated soil water retention with medium accuracy (i.e. R2 from 0.40 to 0.85). These PTFs had R2, root mean square error and mean error similar to PTFs that include only particle size distribution and soil organic matter. Further, as was the case for our study, these equations were all generated from a set of soils ranging from low (2:1 type clays) to high weathering (1:1 clays and oxides), thus demonstrating PTFs for hydraulic soil properties are region specific. Finally, none of the PTFs from temperate soils was accurate for estimating water retention in conventionally-tilled soils from subtropical environment, and thus should be replaced by the PTFs from this study.

Introduction

Soil water retention information is applied in agronomy, hydrology, ecology, and environmental issues. Agricultural crops are strongly influenced by water availability, which depends on soil moisture, granulometry, mineralogy, soil organic matter and structure. At low tensions (drier soil), water retention is related to soil pore size and organization, whereas at high tension (wet soil) this retention is associated with particle size and activity. Sandy soils, compared to clayey soils, have lower specific surface and surface charges and higher macro and total porosity (Brady and Weil, 2008), decreasing water retention along the water retention curve. Cation exchange capacity (CEC) and specific surface are surface phenomena whose magnitude depend on soil organic matter, granulometry and mineralogy, all of which affect water retention (Souza Júnior et al., 2007; Tessier et al., 1999).

Soil organic matter has high specific surface and surface charge density compared to most clay minerals and thus increases water retention (Bigorre et al., 2000), both as a reactive surface and soil aggregation factor, but relationship of soil water retention to organic matter is affected by soil granulometry (Rawls et al., 2003; Minasny and McBratney, 2018). Liquid and plastic limits are higher in soils with high contents of clay, organic matter and expansive clay minerals, and CEC (Baver and Gardner, 1972), properties that affect water holding by capillary retention and by adsorption, phenomena that are dependent on specific surface and surface charges.

Laboratory determination of soil water retention and availability to plants is time-consuming, and data are scarce (Ottoni et al., 2018). An alternative for the lack of data is to estimate these hydraulic properties, to optimize input data acquisition and reducing costs, through pedotransfer functions - PTFs (e.g McNeill et al., 2018), for specific tensions such as field capacity and permanent wilting point, or for available water capacity.

Few PTFs are available to estimate soil water retention for subtropical Brazilian soils (Bortolini, 2016; Costa et al., 2013; Giarola et al., 2002; Reichert et al., 2009). The soil properties usually selected in PTFs are the physical properties sand, silt and clay fractions and soil bulk density, and the chemical property soil organic matter (or carbon) whereas cation exchange capacity is rarely selected. Even when available, sometimes the PTFs were generated with a small number of soils, which reduces representativeness and accuracy. On the other hand, there is great demand for soil hydrological data (De Jong van Lier et al., 2015) for process modeling in subtropical environments.

The use of PTFs generated from temperate soils might not estimate water retention in tropical and subtropical soils with the necessary accuracy (Botula et al., 2012), since soils with low clay contents and high silt and cation exchange capacity predominate in temperate regions (Minasny and Hartemink, 2011). While region-specificity for PTFs for hydraulic soil properties has not definitely been shown, the subtropical soils of southern Brazil include soils both with tropical (weathered sandy-loam to clay soils with kaolinite and oxides in the clay fraction) and temperate characteristics (expansive clays and higher silt content), which differentiate the subtropical soils from both groups.

Furthermore, soils for grain production are cultivated under conventional tillage, which represent a significant practice in southern Brazil (18.6% of the area) and even more when in the whole country (32.2% - Telles, 2015). Since structure affects the representativeness of the PTFs (Nguyen et al., 2014), the accuracy of PTFs will be higher when using samples with similar soil structure (undisturbed or disturbed), especially for low tensions when soil structure exerts a significant effect on water retention (Vereecken et al., 1989).We hypothesized that PTFs developed for temperate and tropical soils do not accurately estimate water retention for conventionally-tilled subtropical soils, suggesting these pedotransfer functions are region specific. The objectives were to (i) generate PTFs to estimate water retention at different tensions in conventionally-tilled subtropical soils with a broad range of chemical, physical, and mineralogical properties, and (ii) validate some PTFs available in the literature for tropical, subtropical and temperate environments.

Section snippets

Soils and laboratory analyses

Soil samples with disturbed structure were collected from 34 profiles in southern Brazil (Rio Grande do Sul state), in 85 surface and subsurface soil horizons. The soils were classified according to the World Reference Base system (IUSS, 2015; Appendix) as Acrisols (10), Arenosols (1), Cambisols (2), Ferralsols (5), Leptosols (2), Luvisols (3), Nitisols (5), Phaeozems (1), Planosols (3), and Vertisols (2).

The soil samples were air-dried; manually, gently disrupted/disaggregated along natural

Correlation between water retention and availability with soil properties

The soils used to generate the PTFs had broad variability in soil chemical and physical properties (Table 2), including several textural classes except the silt textural class (Fig. 1). Water retention was highly variable among soils and horizons; for example, gravimetric moisture (θg) at the tension of 10 kPa ranged from 53 to 774 g kg−1 (Table 2). Most physical properties showed correlation with water retention to a higher or lesser degree. The correlation coefficients (r) with physical soil

Water retention and availability as affected by soil physical and chemical properties

Water retention was negatively correlated with coarse sand, total sand, and soil bulk density (Table 3). Sand has an indirect effect on water retention due to its low specific surface and surface charges (Brady and Weil, 2008). Moreover, the negative correlations between water retention and bulk density are also partially due to the indirect effect of particle size distribution, since sandier soils had the greatest densities, with a correlation coefficient ranging from −0.71 to −0.81 between

Conclusions

  • 1.

    Water retention in subtropical conventionally-tilled soils, is mainly related to soil granulometry, liquid and plastic limits, and cation exchange capacity, thus including physical and chemical properties related to surface activity of soil particles.

  • 2.

    Water retained by subtropical conventionally-tilled soils at different tensions, varying from 10 to 1500 kPa, is accurately estimated by means of pedotransfer functions generated by using chemical and physical properties as prediction variables,

Acknowledgments

The authors thank CNPq, Capes – finance code 001, and Fapergs for fellowships/assistantships and financial resources, and to students who assisted with data organization.

References (54)

  • M. van den Berg et al.

    Pedotransfer functions for estimation of moisture retention characteristics of Ferralsols and related soils

    Geoderma

    (1997)
  • M.H. Ali et al.

    Retention and release of soil water as related to mineralogy of the soil clays

    J Indian Soc Soil Sci

    (1971)
  • F. Arruda et al.

    Parâmetros de solo para o cálculo da água disponível com base na textura de solos

    Rev. Bras. Cienc. Solo

    (1987)
  • M.L.L. Assad et al.

    Relação entre água retida e conteúdo de areia total em solos brasileiros

    R Bras Agrometeorologia

    (2001)
  • A.C. Auler et al.

    Influence of physical attributes and pedotransfer function for predicting water retention in management systems

    R Bras Eng Agric Ambiental

    (2017)
  • L.D. Baver et al.

    Soil Physics

    (1972)
  • M.A. Bell et al.

    Soil pedotransfer functions for four Mexican soils

    Soil Sci Soc Am J

    (1995)
  • T.D. Biswas et al.

    Retention and availability of soil water as influenced by soil organic carbon

    Indian Agric Res Inst

    (1969)
  • D. Bortolini

    Estimativa da retenção e disponibilidade de água em solos de Santa Catarina

    (2016)
  • N.C. Brady et al.

    The Nature and Properties of Soils

    (2008)
  • A.S. Brito et al.

    Estimativa da capacidade de campo pela curva de retenção e pela densidade de fluxo da água

    Rev Bras Cienc Solo

    (2011)
  • A. Costa et al.

    Pedotransfer functions to estimate retention and availability of water in soils of the state of Santa Catarina

    Brazil Rev Bras Cienc Solo

    (2013)
  • N.F.B. Giarola et al.

    Relações entre propriedades físicas e características de solos da região sul do Brasil

    Rev. Bras. Cienc. Solo

    (2002)
  • P.I. Gubiani et al.

    Assessing errors and accuracy in dew-point potentiometer and pressure plate extractor measurements

    Soil Sci Soc Am J

    (2013)
  • S.C. Gupta et al.

    Estimating soil water retention characteristics from particle size distribution, organic matter percent and bulk density

    Water Res Res.

    (1979)
  • IUSS Working Group WRB

    World reference base for soil resources 2014, update 2015

  • N. Kämpf et al.

    Estrutura básica de argilomineral 2:1 com hidroxi-Al entrecamadas em Latossolo Bruno no Rio Grande do Sul

    Rev. Bras. Cienc. Solo

    (1995)
  • Cited by (37)

    View all citing articles on Scopus
    View full text