Elsevier

Geoderma

Volume 363, 1 April 2020, 114139
Geoderma

Assessing the effects of deforestation and intensive agriculture on the soil quality through digital soil mapping

https://doi.org/10.1016/j.geoderma.2019.114139Get rights and content

Highlights

  • Different soil quality indices were applied to evaluate soil quality.

  • Different data set and scoring approaches used to calculate soil quality indices.

  • SQIn showed better estimation of soil quality comparing to SQIa and SQIw.

  • Deforestation led to a decrease in soil quality.

  • RF technique through digital soil mapping is a robust method to monitor soil quality.

Abstract

This study was designed to evaluate soil quality (SQ) in deforested and intensively cultured lands in Mazandaran Province, Iran. For this purpose, three soil quality indices (SQIs: additive soil quality index (SQIa), nemoro soil quality index (SQIn) and weighted additive soil quality index (SQIw)) and two datasets (the total data set (TDS) and minimum data set (MDS)) were determined. The linear (L) and non-linear (NL) scoring systems were also used to calculate each SQI. Eight soil properties, including pH, cation exchange capacity (CEC), electrical conductivity (EC), bulk density (BD), soil microbial respiration (SMR), total nitrogen (TN), soil organic carbon (SOC), and calcium carbonate equivalent (CCE), were measured in 108 locations (0–30 cm depth). The MDS was determined by the principal component analysis. A digital soil mapping (DSM) method, more specifically the random forest technique, was applied to produce the SQI maps. The maximum and minimum values of the SQIs were obtained in the natural forest (NF) and the dry farming (DR) land uses, respectively. The results indicated that both methods (i.e., TDS and MDS) properly could describe SQ in this area but MDS can be recommended as an appropriate method because it was able to classify the SQ using a lower number of the soil properties without losing information to SQ assessment. In addition, NF and the more-than-ten-year-old commercial garden (G10 + ) land uses had the highest proportions of SQ grades I and II, respectively. Conversely, DR and the less-than-ten-year-old commercial gardens (G10) land uses had the highest proportions of the SQ grade V (very low quality) among the six SQIs. The findings indicated that the L scoring system had higher agreement values for all SQIs compared to the NL scoring system, and also SQIn (R2 = 0.874) could provide a better estimation compared to the SQIa (R2 = 0.864) and SQIw (R2 = 0.865). The spatial distribution of the SQ grades using DSM indicated that the land use conversion could decrease SQIs, suggesting that close attention should be paid to the sustainable use of the agricultural lands to increase the SQ.

Introduction

Over the last 50 years, the major concerns in environmental studies have been soil degradation, climate change and, water and food security. Besides these negative aspects, land use change, land degradation, and deforestation can intensify the challenges of water and food security and climate change (Villarino et al., 2017). Therefore, understanding and enhancing soil quality (SQ) is vital and is a key point to address these global issues (Lal, 2011). Land use changes, specifically deforestation, has globally been identified as the most severe type of land use conversion (Smith et al., 2016), leading to the greatest impact on biodiversity, land degradation, and climate change (Foley et al., 2007). The sustainable use of land resources, restoration of degraded lands, management of land use, and maintenance of SQ require a profound understanding of the soil system (Ayoubi et al., 2018).

The soil quality of the deforested lands is meaningfully influenced by several critical environmental processes and management factors, including the balance between soil organic carbon (SOC) inputs and outputs (Korkanc et al., 2008), topographic properties (Ajami et al., 2016), and anthropogenic factors (Khaledian et al., 2017). Dorji et al. (2014) stated that deforestation and intensive agricultural activities could remarkably deplete SOC in the form of CO2 emission into the atmosphere. In a deforested area, Khormali et al. (2009) reported that the land use conversion led to a significant reduction in the soil organic matter compared to the natural forest (NF). Based on their results, SOC storage in the surface layers (0–60 cm of soil profile) decreased from 184.8 to 58.8 ton ha−1 when the NF converted to the cultivated lands, respectively. Similarly, cation exchange capacity (CEC) could play an important role in SQ; thus this soil property could have a great impact on the soil physicochemical properties and biological characteristics (Czarnecki and Düring, 2015). In previous studies, CEC was commonly used as an appropriate indicator of soil degradation (Khaledian et al., 2017).

Many methods have been proposed for SQ assessment, such as dynamic SQ models (Larson and Pierce, 1994), SQ cards (Ditzler and Tugel, 2002), geostatistical approaches (Sun et al., 2003), and soil quality indices (SQIs) (Andrews et al., 2002). Although these methods, each having their advantages and disadvantages, were used to monitor SQ changes, SQI methods were commonly used in the previous studies (Andrews et al., 2002) due to their simplicity, practicality, and quantitative flexibility (Qi et al., 2009). So far, several studies (Nabiollahi et al., 2018a, Nabiollahi et al., 2018b, Rahmanipour et al., 2014, Zhang et al., 2016) have used SQIs to monitor and assess SQ in various ecosystems under different land use management.

The development of a quantitative SQI method is based on three-step processes, that is, i) selection of an appropriate indicator, ii) determination of indicator scoring method, and iii) combination of all indicator scores within an SQI system (Karlen et al., 2003, Andrews et al., 2002, Andrews et al., 2004). Since SQ is very complex, there is no agreement on a definitive set of soil properties (e.g., soil physicochemical and biochemical attributes) concerning the SQ assessment (Liu et al., 2014). Nonetheless, the SQ indicators are the properties that are rapidly affected by the changes in the soil conditions (Marzaioli et al., 2010).

Generally, two methods including total data set (TDS) and minimum data set (MDS) have commonly been used in the SQ studies (Zhang et al., 2016, Rahmanipour et al., 2014, Cheng et al., 2016, Biswas et al., 2017, Nabiollahi et al., 2017, Nabiollahi et al., 2018a, Yu et al., 2018). The MDS method is often considered to reduce the required time and cost of the SQ assessment (Qi et al., 2009, Zhang et al., 2016). In fact, MDS could be obtained from TDS using various mathematical and statistical methods, such as multiple and linear regression analysis, discriminant analysis, scoring functions, and factor analysis (Yao et al., 2013). However, there is no standardized method to establish MDS because it largely depends on the practical aspects of research (Zhang et al., 2016). For example, Cheng et al. (2016) made use of CEC, total phosphorus, sand, clay, available phosphorus, iron, boron, and exchangeable Mg as the MDS to assess SQ in the farmlands of northeast China. Likewise, Nabiollahi et al. (2017) selected pH, soil electrical conductivity (EC), bulk density (BD), CEC, and exchangeable sodium percentage (ESP) to evaluate SQ in the agricultural lands of Kurdistan, Iran.

Among different soil properties derived from different numerical scales (as indicators), two scoring methods, including linear (L) and non-linear (NL) scoring systems, are frequently applied to normalize particular soil properties (Larson and Pierce, 1994). The additive soil quality index (SQIa) (Zhang et al., 2016, Askari and Holden, 2014), weighted additive soil quality index (SQIw) (Raiesi and Kabiri, 2016, Askari et al., 2014), and nemoro soil quality index (SQIn) (Rahmanipour et al., 2014) are three approaches most commonly used for integrating dimensionless indicators into SQIs (Nabiollahi et al., 2017, Nabiollahi et al., 2018a). Virtually, all these integrated approaches have successfully been applied for SQI assessment around the world (Qi et al., 2009, Askari et al., 2014, Nabiollahi et al., 2017, Rahmanipour et al., 2014, Raiesi and Kabiri, 2016, Cheng et al., 2016, Biswas et al., 2017).

Although numerous studies have been conducted on SQI assessment in natural recourses (Khaledian et al., 2017, Qi et al., 2009, Askari et al., 2014, Rahmanipour et al., 2014, Cheng et al., 2016, Biswas et al., 2017), a few studies have investigated the effects of deforestation on SQI and/or defined SQI classes under different land uses using robust mapping techniques, such as digital soil mapping (DSM) (Nabiollahi et al., 2018a, Nabiollahi et al., 2018b). DSM is a powerful technique to predict and map soil properties using auxiliary data (Minasny and Hartemink, 2011). Numerous studies have also applied DSM to predict soil properties using environmental variables (Forkuor et al., 2017, Kempen et al., 2009, Taghizadeh-Mehrjardi et al., 2014, Zeraatpisheh et al., 2017, Zeraatpisheh et al., 2019a).

Deforestation, forest conversion to agricultural lands and urbanization, occurring in the natural forests in the north of Iran (about 52% in less than half a century), have led to soil degradation and soil nutrient deterioration (Ajami et al., 2016). However, understanding the effects of changing the land use from NF to agricultural lands on SQ is required. Therefore, the present study was aimed to i) explore the effects of land use conversion on SQ in the NFs and the adjacent agricultural lands, ii) determine the best SQI, scoring method, and indicator selection technique, and iii) evaluate the spatial variations of SQI using DSM approach.

Section snippets

Study area description

The study area (approximately 276 ha) is located between 52° 57′ 50.9″ and 52° 59′ 12.37″ E longitudes and 36° 29′ 4.83″ and 36° 29′ 50.9″ N latitudes, Sari region, Mazandaran Province, north of Iran (Fig. 1). The mean annual temperature, precipitation, and elevation are 17.9 °C, 789 mm, and 35 m a.s.L, respectively. The main land uses in the study area were defined as follows: commercial gardens (160 ha, having different cultivars of Citrus sp., divided into two classes, i.e., the

The TDS indicator method

The results showed that the range of data were 0.86 to 1.55 (g cm−3) for BD, 5.00 to 26.50% for CCE, 5.11 to 8.45 for pH, 0.08 to 0.52 (ds m−1) for EC, 0.35 to 4.87% for SOC, 0.06 to 0.48% for TN, 11.90 to 20.57 (cmol+ kg−1) for CEC and 15.89 to 85.54 (mg C g−1 d-1) for SMR. The lowest and highest coefficient of variation (CV) was also observed in CEC and SOC, respectively (Table 2). According to the CVs proposed by Wilding (1985), CCE, SOC, EC, and TN indicated high variability, while BD, pH,

Conclusions

The need for efficient and accurate methods, including DSM, is severely felt to monitor changes in SQ in various landscapes under different land uses. DSM can contribute to the development of advanced SQ maps using cost-effective and accessible spatial information. The regions, such as forests and rangelands, where human activities have degraded soils and converted farming land uses, require accurate and quantitative sources of information to maintain their sustainable development. In this

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

The field survey and laboratory analysis were financially supported by the Genetics and Agricultural Biotechnology Institute of Tabarestan (GABIT) and Sari Agricultural Sciences and Natural Resources University (SANRU) (under grant no. T214-96). The authors would like to thank the GABIT for providing the services and facilities during this research. We would like to thank Mr. Majid Vardan for his valuable help during the field survey and sampling collection.

References (71)

  • B. Kempen et al.

    Updating the 1: 50,000 Dutch soil map using legacy soil data: A multinomial logistic regression approach

    Geoderma

    (2009)
  • F. Khormali et al.

    Role of deforestation and hillslope position on soil quality attributes of loess-derived soils in Golestan province, Iran

    Agric. Ecosyst. Environ.

    (2009)
  • R. Lal

    Sequestering carbon in soils of agro-ecosystems

    Food Policy

    (2011)
  • Z. Liu et al.

    Soil quality assessment of Albic soils with different productivities for eastern China

    Soil Tillage Res.

    (2014)
  • R. Marzaioli et al.

    Soil quality in a Mediterranean area of Southern Italy as related to different land use types

    Appl. Soil Ecol.

    (2010)
  • A.B. McBratney et al.

    On digital soil mapping

    Geoderma

    (2003)
  • B. Minasny et al.

    Predicting soil properties in the tropics

    Earth Sci. Rev.

    (2011)
  • K. Nabiollahi et al.

    Assessing the effects of slope gradient and land use change on soil quality degradation through digital mapping of soil quality indices and soil loss rate

    Geoderma

    (2018)
  • K. Nabiollahi et al.

    Assessment of soil quality indices for salt-affected agricultural land in Kurdistan Province, Iran

    Ecol. Indic.

    (2017)
  • M.R. Pahlavan-Rad et al.

    Spatial variability of soil texture fractions and pH in a flood plain (case study from eastern Iran)

    Catena

    (2018)
  • J. Peters et al.

    Random forests as a tool for ecohydrological distribution modelling

    Ecol. Model.

    (2007)
  • Y. Qi et al.

    Evaluating soil quality indices in an agricultural region of Jiangsu Province, China

    Geoderma

    (2009)
  • F. Rahmanipour et al.

    Assessment of soil quality indices in agricultural lands of Qazvin Province, Iran

    Ecol. Indic.

    (2014)
  • F. Raiesi et al.

    Identification of soil quality indicators for assessing the effect of different tillage practices through a soil quality index in a semi-arid environment

    Ecol. Ind.

    (2016)
  • M. Shukla et al.

    Determining soil quality indicators by factor analysis

    Soil Tillage Res.

    (2006)
  • S.M.J. Sione et al.

    Evaluation of soil degradation produced by rice crop systems in a Vertisol, using a soil quality index

    Catena

    (2017)
  • B. Sun et al.

    Evaluation of spatial and temporal changes of soil quality based on geostatistical analysis in the hill region of subtropical China

    Geoderma

    (2003)
  • R. Taghizadeh-Mehrjardi et al.

    Digital mapping of soil salinity in Ardakan region, central Iran

    Geoderma

    (2014)
  • S.H. Villarino et al.

    Deforestation impacts on soil organic carbon stocks in the Semiarid Chaco Region, Argentina

    Sci. Total Environ.

    (2017)
  • R. Yao et al.

    Determining minimum data set for soil quality assessment of typical salt-affected farmland in the coastal reclamation area

    Soil Tillage Res.

    (2013)
  • P. Yu et al.

    Selecting the minimum data set and quantitative soil quality indexing of alkaline soils under different land uses in northeastern China

    Sci. Total Environ.

    (2018)
  • M. Zeraatpisheh et al.

    Comparing the efficiency of digital and conventional soil mapping to predict soil types in a semi-arid region in Iran

    Geomorphology

    (2017)
  • M. Zeraatpisheh et al.

    Digital mapping of soil properties using multiple machine learning in a semi-arid region, central Iran

    Geoderma

    (2019)
  • G. Zhang et al.

    Soil quality assessment of coastal wetlands in the Yellow River Delta of China based on the minimum data set

    Ecol. Ind.

    (2016)
  • J. Zhi et al.

    Predicting mattic epipedons in the northeastern Qinghai-Tibetan Plateau using Random Forest

    Geoderma Reg.

    (2017)
  • Cited by (103)

    • Impact of organic farming on soil health and nutritional quality of crops

      2023, Journal of the Saudi Society of Agricultural Sciences
    View all citing articles on Scopus
    View full text