Determining soil quality indicators by factor analysis

https://doi.org/10.1016/j.still.2005.03.011Get rights and content

Abstract

Soil quality indicators (SQIs) can be used to evaluate sustainability of land use and soil management practices in agroecosystems. The objective of this study was to identify appropriate SQI from factor analysis (FA) of five treatments: no-till corn (Zee mays) without manure (NT), no-till corn with manure (NTM), no-till corn–soybean (Glycine max) rotation (NTR), conventional tillage corn (CT), and meadow (M) in Coshocton, Ohio. Soil properties were grouped into five factors (eigenvalues > 1) for the 0–10 cm depth as: (Factor 1) water transmission, (Factor 2) soil aeration, (Factor 3) soil pore connection 1, (Factor 4) soil texture and (Factor 5) moisture status. Factor 2 was the most dominant, with soil organic carbon (SOC) the most dominant measured soil attribute contributing to this factor. For the 10–20 cm depth, factors identified were: (Factor 6) soil aggregation, (Factor 7) soil pore connection 2, (Factor 8) soil macropore, and (Factor 9) plant production. At 10–20 cm depth, Factor 6 was most dominant with SOC the most dominant measured soil attribute. Management × sample and slope position × sample interactions were significant among some factors for both depths. Overall, SOC was the most dominant measured soil attribute as a SQI for both depths. Other key soil attributes were field water capacity, air-filled porosity, pH and soil bulk density for the 0–10 cm depth, and total N and mean weight diameter of aggregates for the 10–20 depth. Therefore, SOC could play an important role for monitoring soil quality.

Introduction

Soil quality essentially means “the capacity of a soil to function” (Larson and Pierce, 1991, Doran and Parkin, 1994, Karlen et al., 1997). The particular soil function of concern varies depending on the interest of observer. Thus, for land managers it can mean the soil's capacity to sustain or enhance productivity while maintaining soil resources for the future. For conservationists it can mean sustaining soil resources and protecting the environment. For consumers it can mean production of healthy and inexpensive food products. For environmentalists it can mean capacity to maintain or enhance biodiversity, water quality, nutrient cycling and biomass yield (Mausback and Seybold, 1998).

Soil quality influences basic soil functions, such as moderating and partitioning water and solute movement and their redistribution and supply to plants; storing and cycling nutrients; filtering, buffering, immobilizing and detoxifying organic and inorganic materials; promoting root growth; providing resistance to erosion (Karlen et al., 1997). The capacity of soil to function can be reflected by measured soil physical, chemical and biological properties, also known as soil quality indicators (SQIs). There is a need to develop SQI in such a way so that they: (i) integrate soil physical, chemical and/or biological properties and processes, (ii) apply under diverse field conditions, (iii) complement either existing databases or easily measurable data, and (iv) respond to land use, management practices, climate and human factors (Doran and Parkin, 1994). Depending upon their temporal variability, SQI can be classified as static or dynamic (Carter et al., 1997, Shukla et al., 2004a). Monitoring changes in the key SQI with time can determine if quality of a soil under a given land use and management system is improving, stable or declining (Lal, 1998, Shukla et al., 2004a). Soil quality concept is both advocated (Karlen et al., 2001) and criticized in the literature (Sojka and Upchurch, 1999).

There are two fundamental approaches employed for evaluating the sustainability of a management system: (i) comparative assessment and (ii) dynamic assessment (Larson and Pierce, 1994). In a comparative assessment, performance of a system is evaluated in relation to alternatives at a given time only. In a dynamic assessment, performance of a system is evaluated in relation to alternatives across time (Larson and Pierce, 1994). Several minimum data sets of SQI have been proposed (Arshad and Coen, 1992, Doran and Parkin, 1994, Kennedy and Papendick, 1995, Larson and Pierce, 1991, Larson and Pierce, 1994). Other alternate techniques for SQI identification and/or interpretation are: linear and multiple regression analysis (Li and Lindstorm, 2001, Mendham et al., 2002), pedotransfer functions (Bouma, 1989, Larson and Pierce, 1991, Larson and Pierce, 1994, Salchow et al., 1996), scoring functions (Karlen et al., 1994, Hussain et al., 1999), and factor analysis (FA) (Bachmann and Kinzel, 1992, Wander and Bollero, 1999; Brejda et al., 2000a,b; Andews et al., 2002, Shukla et al., 2004a). The FA technique groups soil properties into a few factors (James and McCulloch, 1990). These factors, although statistically constructed, can be assessed with respect to specific soil functions (Johnson and Wichern, 1992; Brejda et al., 2000b). Changes in the properties of soil attributes associated with a factor can be used to classify soil quality as aggrading, degrading or stable (Brejda et al., 2000a; Shukla et al., 2004a). The objectives of this study were to: (i) identify SQI using FA of measured soil properties, and (ii) evaluate the influence of land use treatments and landscape positions on SQI according to the land owners or environmentalists’ perspective.

Section snippets

Experimental site

This study was conducted at the North Appalachian Experimental Watersheds (NAEW) near Coshocton, Ohio (40° 22′N, 81° 48′W; elevation 300–600 m), located in the Western Allegheny Plateau of the Appalachian Mountains. The dominant soil is Coshocton silt loam (fine-loamy, mixed, mesic Aquultic Hapludalfs), developed mostly on upper and mid slopes. Soil on the lower slopes is Clarksburg silt loam (fine-loamy, mixed, superactive, mesic Oxyaquic Fragiudalfs). The climate is continental with mean

Grouping of soil properties

Correlation analysis of the 20 soil attributes representing soil physical and chemical properties and biomass yield resulted in significant correlation (P < 0.05) in 136 of the 210 soil attribute pairs (Table 1). Highest positive correlations were obtained for ic versus I, i5 versus I, and SOC versus SN (r > 0.90). Highest negative correlations were obtained for sand versus silt concentration (r > 0.94) and EC versus pH (r > 0.74). Biomass yield was positively correlated with SOC concentration (r = 0.59)

Discussion

Factor analysis grouped 20 measured attributes into five factors for the 0–10 cm depth. All five factors related to one or more soil functions (e.g., water and nutrient retention and transport, soil structure, aeration, etc.) to influence soil pore structure and the capacity of soil to accept, store and release water and nutrients. Factor 2 (Soil Aeration) was the most dominant for 0–10 cm and Factor 6 (Soil Aggregation) for 10–20 cm depth. Measured attributes constituting these factors influence

Conclusions

The dominant factors in assessing soil quality varied with soil depth. Soil aeration was the most discriminating factor for the 0–10 cm depth and soil aggregation was the most significant factor for the 10–20 cm depth. For each factor, the dominant measured soil attribute was SOC. If only one soil attribute were to be used for monitoring soil quality changes every 3–5 years, SOC should be selected.

Acknowledgements

The authors gratefully acknowledge the Los Alamos National Laboratory, New Mexico, for funding the project. Authors also thank Dr. Lloyd Owns and other staff members of North Appalachian Experimental Watersheds (NAEW), USDA, Coshocton for their help and support during soil sampling.

References (45)

  • Bouwer, H., 1986. Intake rate: cylinder infiltrometer. In: Klute, A. (Ed.), Methods of Soil Analysis, Part 1, 2nd ed.,...
  • J.I. Brejda et al.

    Identification of regional soil quality factors and indicators. II. Northern Mississippi loess hills and palouse prairie

    Soil Sci. Soc. Am. J.

    (2000)
  • J.I. Brejda et al.

    Identification of regional soil quality factors and indicators. I. Central and southern high plains

    Soil Sci. Soc. Am. J.

    (2000)
  • Doran, J.W., Parkin, T.B., 1994. Defining and assessing soil quality. In: Doran, J.W., et al. (Eds.), Defining Soil...
  • Gee, G.W., Bauder, J.W., 1986. Particle size analysis. In: Klute, A. (Ed.), Methods of Soil Analysis. Part 1, 2nd ed.,...
  • F.C. James et al.

    Multivariate analysis in ecology and systematics: panacea or pandora's box?

    Annu. Rev. Ecol. Syst.

    (1990)
  • R.A. Johnson et al.

    Applied multivariate statistical analysis

    (1992)
  • D.L. Karlen et al.

    Soil quality: a concept, definition, and framework for evaluation

    Soil Sci. Soc. Am. J.

    (1997)
  • Kelley, G.E., Edwards, W.E., Harrold, L.L., 1975. Soils of the North Appalachian Experimental Watersheds, Miscellaneous...
  • A.C. Kennedy et al.

    Microbial characteristic of soil quality

    J. Soil Water Conserv.

    (1995)
  • Klute, A. (Ed.), 1986. Water retention: laboratory methods. Methods of Soil Analysis, Part 1, 2nd ed. Agron. Monograph...
  • Klute, A., Dirkson, C., 1986. Hydraulic conductivity and diffusivity: laboratory methods. In: Klute, A. (Ed.), Methods...
  • Cited by (433)

    View all citing articles on Scopus
    View full text