Performance of Soil Condition Indicators Across Taxonomic Groups and Land Uses

soil condition to those characteristics suitable (fit) for a particular land use. Also implicit in this brief definition Information on soil conditions in New Zealand is needed to assess is the capacity for the soil to maintain its fitness into soil quality at a national scale. We tested a standard set of 16 primary indicators at 29 sites (0–10 cm depth) across nine soil great groups the future. with matched examples of indigenous forest, plantation forest, pasThere have been many studies on specific aspects of tures, and crops. Soils under indigenous forest were acidic (pH 5.5– soil condition under different land uses, and investiga5.7), low in Olsen P (5–14 mg cm23), with high microbial C (814–1228 tors have used many different indicators and differing mg cm23), respiration (1.1–1.4 mg C cm23 h21), total C (31.8–52.9 sampling strategies (e.g., Reganold et al., 1993; Schacht mg cm23), macroporosity (9.6–11.7% v/v), and total available water et al., 1996; Werner, 1997; Boehm and Anderson, 1997; (29.2–31.5% v/v). Plantation forest soils had generally similar characRobertson et al., 1997; Staben et al., 1997). These studies teristics. Pasture soils were less acidic (pH 5.3–6.9) than forest soils, provide good information for those particular soils and but with more available P (5.5–43.0 mg cm23), higher total C (30.7– land-management practices, but comparison between 141.5 mg cm23), total N (2.7–9.0 mg cm23), and mineralizable N (68– these studies and others is difficult because of the variety 175 mg cm23). The physical condition was similar to forest soils. Cropped soil had low total C (20–34 mg cm23), microbial C (160–956 of indicators and methodologies used. A standardized mg cm23), respiration (0.29–1.33 mg C cm23 h21), and total available methodology may to be too broad when applied across water (6.7–30.1% v/v), but high pH (5.8–7.2), Olsen P (11.2–199 mg contrasting soils and land uses. However, it is not practicm23), and bulk density (0.96–1.3 g cm23). Principal component analycal to optimize sampling and analytical techniques for sis identified outlier sites and grouped land uses independently of each soil and land use for extensive sampling on a nasoil great groups. Some indicators were less useful because of high tional scale. Standardized methods are common in largevariability (unsaturated hydraulic conductivity), correlation to other scale soil fertility studies where soils are usually sampled indicators (microbial C) or interpretation difficulties (respiration). with a fixed-depth coring device and the same analyses Overall, the standardized approach provided useful information about and extraction methods are applied to all samples (e.g., soil conditions on a national scale. the Olsen P test). Although the methods are standardized, the results are interpreted differently, depending on the soil group and land use (Saunders et al., 1987). I of soil quality are required for environNew Zealand has diverse soils and multiple land uses mental reporting; they help us to assess human and (Molloy, 1988). The responsibility for soil condition renatural impacts on soils and to identify sustainable landporting lies with 12 Regional Councils who desired stanmanagement practices (Doran and Parkin, 1994; Sims dardized methods for measurements. It was important et al., 1997). Numerous soil chemical, biological, and that the methods could be applied to the diversity of physical characteristics have been suggested as suitable soils and land uses in the different regions, were afindicators for these purposes (Garlynd et al., 1994; Harfordable, and internationally acceptable (Hortensius ris and Bezdicek, 1994; Jordan et al., 1995). Land adminand Welling, 1996). We selected a set of soil condition istrators involved in regional and national reporting indicators based on current international literature (e.g., would prefer a standard set of indicators so that trends Doran et al., 1994; Doran and Jones, 1996) and applied in soil condition across regions, soils, time, and land use these indicators across 9 different soil great groups and can be readily discerned and analyzed. Various minimultiple land uses. Our objective was to test whether a mum data sets have been proposed (Harris and Bezdistandard set of indicators could discriminate between cek, 1994; Doran and Parkin, 1994; Pankhurst et al., land-management practices and provide soil condition 1994) but have been the subject of ongoing discussion. information at a national scale. Currently, there is no consensus on a definitive data set for soil-quality monitoring, nor consensus on how the MATERIALS AND METHODS indicators should be interpreted. Part of this lack of consensus about soil-quality monitoring arises because Soils and Land Use different soil conditions are desirable, depending on the The selection of sites and land uses was made after conland use. There are various lengthy definitions of soil sulting Regional Council staff responsible for soil monitoring quality (e.g., see Doran and Parkin, 1994; Carter et al., and reporting. Excluding contaminated or eroding soils, prin1997). To interpret soil condition in terms of soil quality cipal soil condition concerns raised by New Zealand land we have used the concise fitness for use criteria sugmanagers and regulators were those resulting from intensificagested by Larson and Pierce (1994). In this definition, tion or change in land use. Examples were impacts of longterm market gardening and continuous arable cropping, consoil quality is defined in the context of matching the version of sheep pasture to intensive dairying, irrigation onto peat, and establishment of pine plantations on former indigeLandcare Research, Private Bag 3127, Hamilton, New Zealand. Received 10 June 1998. *Corresponding author (sparlingg@ Abbreviations: ANOVA, analysis of variance; CEC, cation-exchange landcare.cri.nz). capacity; CV, coefficient of variance; K240, unsaturated conductivity at 240 kPa; PCA, principal components analysis. Published in Soil Sci. Soc. Am. J. 64:300–311 (2000).