Soil magnetic susceptibility mapping as a pollution and provenance tool: an example from southern New Zealand

Soil magnetic susceptibility mapping as a pollution and provenance tool: an example from southern... Summary The presence or absence, degree and variation of heavy metal contamination in New Zealand soils is a matter of ongoing debate as it affects soil quality, agriculture and human health. In many instances, however, the soil heavy metal concentration data do not exist to answer these questions and the debate is ongoing. To address this, magnetic susceptibility (a common proxy for heavy metal contamination) values were measured in topsoil (0–30 cm) and subsoil (50–70 cm) at grid sites spaced at 8 km intervals across ca. 20 000 km2 of southern New Zealand. Samples were measured for both mass- and volume-specific magnetic susceptibility, with results being strongly, positively correlated. Three different methods of determining anomalies were applied to the data including the topsoil–subsoil difference method, Tukey boxplot method and geoaccumulation index method, with each method filtering out progressively more anomalies. Additional soil magnetic (hysteresis, isothermal remanence and thermomagnetic) measurements were made on a select subset of samples from anomalous sites. Magnetite is the dominant remanence carrying mineral, and magnetic susceptibility is governed by that minerals concentration in soils, rather than mineral type. All except two anomalous sites have a dominant geogenic source (cf. anthropogenic). By proxy, heavy metal contamination in southern New Zealand soils is minimal, making them relatively pristine. The provenance of the magnetic minerals in the anomalous sites can be traced back to likely sources in outcrops of igneous rocks within the same catchment, terrane or rock type: a distance of <100 km but frequently <1 km. Soil provenance is a key step when mapping element or isotopic distribution, vectoring to mineralization or studying soil for agricultural suitability, water quality or environmental regulation. Measuring soil magnetic susceptibility is a useful, quick and inexpensive tool that usefully supplements soil geochemical data. New Zealand, Environmental magnetism, Magnetic mineralogy and petrology, Rock and mineral magnetism, Spatial analysis INTRODUCTION Soil is the product of its source material, climate, age of formation and bioturbation, among other factors (Jenny 1941). Soil can act as an excellent tracer through the landscape, if an appropriate proxy is measured linking it to a probable source (Cullers et al.1988; Haughton et al.1991). Soil may be studied in numerous ways, including its physical properties, chemistry, stable or radiogenic isotopic composition or by its volumetric composition (e.g. organic matter, lithics, clays and minerals). This study focuses on the magnetic mineralogy of soil. Regional- and country-scale studies of magnetic susceptibility and frequency-dependent magnetic susceptibility in topsoil (0– ≤30 cm) have been conducted in England and Wales (Dearing et al.1996, 1997; Blundell et al.2009), Poland, Czech Republic and Germany (Magiera et al.2006) and Bosnia and Herzegovina (Hannam & Dearing 2008). In Austria, magnetic susceptibility measurements were made of both the topsoil (0–20 cm) and subsoil (50–70 cm; Hanesch et al.2007). In addition to magnetic susceptibility, Bian et al. (2014) measured anhysteretic remanence, isothermal remanence, temperature-dependent magnetic susceptibility and other parameters in topsoil (0–20 cm) from the Pearl River Delta area (ca. 41 000 km2), China. In a similar study, Jordanova et al. (2016) measured anhysteretic remanence, isothermal remanence, hysteresis parameters, among other parameters in Bulgarian topsoils. In the Free State of Saxony, Germany, Rachwal et al. (2017) measured magnetic susceptibility as well as the concentration of potentially toxic elements in the organic horizon and topsoil. The motivation for these studies includes understanding pollution, geogenic controls and magnetic mineralogy in soils to characterize anthropogenic affect, pollution sources and better detection of unexploded ordnance. The severity and distribution of heavy metal pollution on New Zealand soils is still a matter of debate, as is the proximal versus distal provenance of material in certain soil types. Using the regional- and national-scale surveys (above) as a model, soil samples from ca. 20 000 km2 of southern New Zealand were collected on an 8 km grid (Fig. 1a) and studied. This area is equivalent in size to Wales or West Virginia. Both topsoil (0–30 cm) and subsoil (50–70 cm) were measured for their volume- and mass-specific magnetic susceptibility by handheld and laboratory techniques. A subset of samples was chosen from anomalous magnetic susceptibility sites and from background sites, for additional measurement of their hysteresis, isothermal remanence and thermomagnetic properties. To our knowledge, this is the first time such a study has been conducted in the Southern Hemisphere. In New Zealand, small-scale investigations have used: magnetic susceptibility changes in loess to study climate change (Pillans & Wright 1990); magnetic fingerprinting to trace increased sedimentation (e.g. Nichol et al.2000) and; magnetic susceptibility for stratigraphic correlation in environmental studies (Turner 1997). This regional-scale study will identify magnetic anomalies, geogenic versus anthropogenic sources, discuss magnetic mineralogy and identify likely sources of magnetic mineralogy in soil. The results will help quantify heavy metal pollution in southern New Zealand, which can affect land and shallow groundwater quality, and help trace soil provenance. This paper is one of many recent and ongoing outputs from the geochemical baseline soil survey of southern New Zealand, and is a pilot study before a national survey of New Zealand is undertaken. Figure 1. View largeDownload slide The location of sample sites and the underlying soil and rock types. (a) Position of the survey within New Zealand. (b) The major New Zealand classified soil orders in southern New Zealand as taken from the Landcare Research Fundamental Soils Layer (lris.scinfo.org.nz; last accessed 2016 August 2). The international equivalents to the soil orders in this study are shown in Table 1. (c) The major lithostratigraphic units after Heron (2014) and Mortimer et al. (2014). The three townships (Dunedin, Milton and Invercargill) in the study area are marked for reference. The position of the Otago Schist metamorphic belt is shown by dashed lines. Figure 1. View largeDownload slide The location of sample sites and the underlying soil and rock types. (a) Position of the survey within New Zealand. (b) The major New Zealand classified soil orders in southern New Zealand as taken from the Landcare Research Fundamental Soils Layer (lris.scinfo.org.nz; last accessed 2016 August 2). The international equivalents to the soil orders in this study are shown in Table 1. (c) The major lithostratigraphic units after Heron (2014) and Mortimer et al. (2014). The three townships (Dunedin, Milton and Invercargill) in the study area are marked for reference. The position of the Otago Schist metamorphic belt is shown by dashed lines. REGIONAL SETTING Southern New Zealand has diverse geology and soil cover (Fig. 1). Five major soil orders found in New Zealand were part of this study (Fig. 1b; Hewitt 2010), with the world reference base soil classification equivalents shown in Table 1. Brown (n = 171) and pallic (n = 88) soils are particularly common in the study area, with recent, gley and podzol soils also well represented (Drewry et al.2000; Martin et al.2015). Of particular interest in this study are soils derived from igneous parent material, which include mafic brown and allophanic soils. The latter contains allophane minerals (Parfitt 1990) as well as ferrihydrite and imogolite (Molloy & Christie 1988) and typically occur as the weathering products of igneous rocks (Fieldes 1955; Parfitt et al.1983). Table 1. The New Zealand soil classification types, and their international equivalents, together with the Tukey boxplot (Fig. 5b) mass-specific magnetic susceptibility (10−8 m3 kg−1). Group  #  WRB  US soil taxonomy  Lithology  Min  Q1  Median  Q3  Max  Outlier  Far  Mean  Topsoil  Brown  158  Cambisols  Dystrochrepts  Schist; sediment  0.49  10.37  16.31  34.77  70.85  71.37  107.97  36.26  Pallic  83  Luvisols  Fragiudalfs, Fragiochrepts, Haplustalfs  Schist; sediment  3.51  8.23  12.67  21.55  34.63  41.53  61.51  21.13  Recent  30  Regosols  Fluvents, Orthents, Udepts  Schist; sediment  7.92  11.03  18.82  55.48  121.93  122.17  188.86  49.88  Gley  16  Gleysols  Aquepts  Sediment  4.89  11.23  29.65  61.15  115.78  136.04    46.25  Podzol  12  Podzols  Aquods, Orthods  Sediment  1.38  6.55  8.21  29.65  42.70  –  –  17.32  Basic  11  Mixed  Mixed  Basic igneous  3.67  14.02  42.16  120.09  249.18  279.20  –  92.91  Subsoil  Brown  161  Cambisols  Dystrochrepts  Schist; sediment  0.18  8.13  12.67  26.09  48.99  53.02  79.95  35.11  Pallic  85  Luvisols  Fragiudalfs, Fragiochrepts, Haplustalfs  Schist; sediment  1.80  6.57  9.09  18.15  35.24  35.53  52.90  15.56  Recent  29  Regosols  Fluvents, Orthents, Udepts  Schist; sediment  5.24  10.31  17.19  30.08  30.34  59.74  89.41  43.45  Gley  17  Gleysols  Aquepts  Sediment  2.79  5.61  17.06  33.94  68.22  76.43  118.93  33.07  Podzol  13  Podzols  Aquods, Orthods  Sediment  4.22  5.86  12.55  19.63  34.03  –  –  13.89  Basic  10  Mixed  Mixed  Basic igneous  5.81  12.97  17.30  149.18  350.62  –  –  85.07  Group  #  WRB  US soil taxonomy  Lithology  Min  Q1  Median  Q3  Max  Outlier  Far  Mean  Topsoil  Brown  158  Cambisols  Dystrochrepts  Schist; sediment  0.49  10.37  16.31  34.77  70.85  71.37  107.97  36.26  Pallic  83  Luvisols  Fragiudalfs, Fragiochrepts, Haplustalfs  Schist; sediment  3.51  8.23  12.67  21.55  34.63  41.53  61.51  21.13  Recent  30  Regosols  Fluvents, Orthents, Udepts  Schist; sediment  7.92  11.03  18.82  55.48  121.93  122.17  188.86  49.88  Gley  16  Gleysols  Aquepts  Sediment  4.89  11.23  29.65  61.15  115.78  136.04    46.25  Podzol  12  Podzols  Aquods, Orthods  Sediment  1.38  6.55  8.21  29.65  42.70  –  –  17.32  Basic  11  Mixed  Mixed  Basic igneous  3.67  14.02  42.16  120.09  249.18  279.20  –  92.91  Subsoil  Brown  161  Cambisols  Dystrochrepts  Schist; sediment  0.18  8.13  12.67  26.09  48.99  53.02  79.95  35.11  Pallic  85  Luvisols  Fragiudalfs, Fragiochrepts, Haplustalfs  Schist; sediment  1.80  6.57  9.09  18.15  35.24  35.53  52.90  15.56  Recent  29  Regosols  Fluvents, Orthents, Udepts  Schist; sediment  5.24  10.31  17.19  30.08  30.34  59.74  89.41  43.45  Gley  17  Gleysols  Aquepts  Sediment  2.79  5.61  17.06  33.94  68.22  76.43  118.93  33.07  Podzol  13  Podzols  Aquods, Orthods  Sediment  4.22  5.86  12.55  19.63  34.03  –  –  13.89  Basic  10  Mixed  Mixed  Basic igneous  5.81  12.97  17.30  149.18  350.62  –  –  85.07  Notes: Group: New Zealand soil type (brown, pallic, recent, gley and podzol) with a common acidic sedimentary lithology, or variable soil type with a common basic igneous lithology (basic); #: number of samples; WRB: world reference base soil classification; Min: minimum regular (whisker on boxplot); Q1: quartile one (bottom of box on boxplot); median (line through box on boxplot); Q3: quartile three (top of boxplot); Outlier (hollow circle on boxplot); Far: far outlier (hollow triangle on boxplot) and Mean (filled circle on boxplot). View Large The geology of New Zealand can be subdivided into relatively old basement rocks of the Austral Superprovince and relatively young cover rocks of the Zealandia Megasequence(Mortimer et al.2014; Fig. 1c). The former can be further subdivided into the Western and the Eastern Provinces, with soil samples used in this study collected above eight of the major basement terranes that comprise these provinces. The basement terranes are characterized by rocks of distinctive composition and age relating to the formation and evolution of Gondwana, and include a variety of continental-sourced quartzo-feldspathic metasedimentary rocks, calcalkaline volcanic rocks, island arc igneous and sedimentary rocks, ultramafic ophiolite rocks and I-, S- and A-type plutonic rocks (Heron 2014; Edbrooke et al.2015). The majority of sample locations in this study were above Eastern Province basement rocks, but one transect also sampled above a number of Western Province rocks (Fig. 1c). Rocks of the Caples and Rakaia terranes have been overprinted by, and form the protolith to, the Otago Schist—a regionally significant metamorphic belt (Fig. 1c). The Zealandia Megasequence is comprised largely of sediments derived by erosion of the basement rocks, and here has been subdivided into the PākihiSupergroup, equivalent to Quaternary aged rocks, and into other older cover, which includes four older sedimentary supergroups (Fig. 1c). This was deemed appropriate because of the similarity in results obtained for the other older cover. One other Zealandia Megasequence unit, the Dunedin Volcanic Group, has been highlighted (Fig. 1c), because its alkalic igneous chemistry and mineralogy is important to this study. The Dunedin Volcanic Group, including the Dunedin Volcano, is Oligocene to Miocene in age and comprised mainly of basanite lava flows, intrusions and pyroclastic deposits, with subordinate phonolite and trachyte rock compositions locally important (Benson 1968; Bishop & Turnbull 1996; Coombs et al.2008). Large catchments in the area, such as the Waiau and Oreti (Fig. 1c), drain southwards, eroding and redepositing sediment over distances of ≤100 km. The climate is cool temperate, with rainfall heaviest in the west (>4000 mm yr−1) and generally decreasing to the east, with the lowest rainfall being 250–500 mm yr−1. Of the ca. 220 000 people that live in the study area, 73 per cent live within two cities (Dunedin and Invercargill; Fig. 1c). The rest of the land area is sparsely populated and used for agriculture (beef, sheep, deer and dairy) or exotic forestry. METHODS Sample collection and preparation A regional geochemical baseline soil survey was undertaken in 2015 over southern New Zealand (Martin et al.2016, 2017; Rogers et al.2017; Fig. 1a). Soil samples were collected at 8 km intervals by hand auger at two depths. Topsoil samples were collected at 0–30 cm in the A-horizon and subsoil samples collected at 50–70 cm in the B-horizon, or rarely the C-horizon where soil cover was shallow (Rattenbury et al.2014). Samples from 323 sites (topsoil and subsoil) were analysed and form part of this study. Each sample was 5–10 kg in weight. A reference sample set is kept by GNS Science (the national geological survey organization) and forms the basis for this study. Samples were air-dried, sieved, split and the sub 2 mm portion analysed. Laboratory methods and procedures Handheld volume-specific magnetic susceptibility measurements (handheld magnetic susceptibility hereafter) were made using a commercially available Terraplus KT-10 unit on the dried and sieved samples in their open sample bags, with the device in contact with the soil (i.e. not through a sample bag). The sensitivity and the lower method detection limit of the volume magnetic susceptibility is 0.01 × 10−8 S.I. The handheld measurements were made at a GNS Science facility in Dunedin during a single session lasting a few hours, and measured both topsoil and subsoil (total of 323 localities). Laboratory mass-specific (cf. volume-specific) magnetic susceptibility measurements (laboratory magnetic susceptibility hereafter) were performed at the OtagoPalaeomagnetic Research Facility at the University of Otago, Dunedin, New Zealand using a single-frequency AGICO MFK-1A CS3 kappabridge. The MFK-1A measures samples in an applied field of between 2 and 700 A m−1 with a frequency of 976 Hz. Measurements were made on both topsoil and subsoil samples (total of 315 localities) at 200 A m−1. Eight localities did not have enough material available to complete a laboratory analysis. After anomalous magnetic susceptibility values had been determined (see Discussion section for an overview of how anomalous values were determined), an additional 21 sites were chosen for further study, these represented 17 anomalous sites and 4 sites from the Otago Schist for a background comparison. A strong geogenic control on magnetic susceptibility (again see Discussion) meant the additional 21 samples were chosen from the subsoil. Hysteresis and isothermal remanent magnetization (IRM) analyses on crushed soil samples of between 0.15 and 0.1 g were measured on a Princeton Measurements Corporation Vibrating Sample Magnetometer (MicroMag 2900) to determine the magnetic mineralogy. Hysteresis analyses were conducted in saturating fields of 500 mT at 4mT increments and IRM analyses were conducted to 1 T. Thermomagnetic measurements were made on an AGICO MFK-1A CSkappabridge on ca. 0.25 cm3 samples in air with an applied field of 200 A m−1. The Curie temperature was estimated using the 1/k method in the AGICO Cureval software to identify the presence of haematite. A precise Curie temperature determination was not attempted. RESULTS The full data are included in the online extra supplementary material (ESM1). The handheld magnetic susceptibility results range between 1.1 × 10−8 and 424 × 10−8 S.I. (median: 13.8 × 10−8 S.I.) for the topsoil and between 1.2 and 594 × 10−8 S.I. (median: 13.0 × 10−8 S.I.) for the subsoil. The laboratory data results range between 0.5 × 10−8 and 534 × 10−8 m3 kg−1 (median: 15.9 × 10−8 m3 kg−1) for topsoil samples and 0.2 × 10−8 and 640 × 10−8 m3 kg−1 (median: 12.4 × 10−8 m3 kg−1) for subsoil samples (Fig. 2). A histogram shows the distribution of magnetic susceptibility values in the topsoil (Fig. 3a) and bivariate plots show strong relationships between laboratory magnetic susceptibility values measured in the topsoil versus subsoil (Fig. 3b; r2 = 0.87) and between laboratory measured magnetic susceptibility values and handheld magnetic susceptibility values in topsoil (not shown) and subsoil (Fig. 3c; r2 = 0.91). Figure 2. View largeDownload slide The range and pattern of mass-specific magnetic susceptibility values showing (a) topsoil and (b) subsoil. The points are size-coded, with larger dots equating to higher magnetic susceptibility. The coloured map shows an inverse distance weighted interpolation between samples using ArcGIS software to create 500 m cell size grids using a distance power of 1.5 and a fixed 9 km search radius. The interpolation has been portrayed using a 10-class quantile colour ramp. Grid ticks are New Zealand Transverse Mercator 2000 (NZTM2000) eastings and northings. Figure 2. View largeDownload slide The range and pattern of mass-specific magnetic susceptibility values showing (a) topsoil and (b) subsoil. The points are size-coded, with larger dots equating to higher magnetic susceptibility. The coloured map shows an inverse distance weighted interpolation between samples using ArcGIS software to create 500 m cell size grids using a distance power of 1.5 and a fixed 9 km search radius. The interpolation has been portrayed using a 10-class quantile colour ramp. Grid ticks are New Zealand Transverse Mercator 2000 (NZTM2000) eastings and northings. Figure 3. View largeDownload slide The magnetic susceptibility data shown on various plots. (a) A histogram of topsoil data showing mass-specific magnetic susceptibility. The black circles indicate values where additional magnetic mineralogy data were collected from subsoil samples. (b) Topsoil versus subsoil plot of mass-specific magnetic susceptibility data. (c) Topsoil mass-specific magnetic susceptibility (laboratory method) versus topsoil volume-specific magnetic susceptibility (handheld method). The strong correlation (R2 = 0.91) allowed a linear regression to be calculated as shown on the diagram. Figure 3. View largeDownload slide The magnetic susceptibility data shown on various plots. (a) A histogram of topsoil data showing mass-specific magnetic susceptibility. The black circles indicate values where additional magnetic mineralogy data were collected from subsoil samples. (b) Topsoil versus subsoil plot of mass-specific magnetic susceptibility data. (c) Topsoil mass-specific magnetic susceptibility (laboratory method) versus topsoil volume-specific magnetic susceptibility (handheld method). The strong correlation (R2 = 0.91) allowed a linear regression to be calculated as shown on the diagram. The IRM analyses resulted in a coercivity of remanence (Hcr) of between 25 and 77 mT (median: 41 mT) and a saturation remanent magnetization (Mrs) of between 0.12 and 131 mAm2 kg−1 (median: 17 mAm2 kg−1). Hysteresis analyses revealed a coercivity (Hc) of between 2.3 and 19 mT (median: 7.9 mT), saturation magnetization (Ms) of between 0.85 and 938 mAm2 kg−1 (median: 251 mAm2 kg−1) and remanent magnetization (Mr) of between 0.10 and 134 mAm2 kg−1 (median: 20 mAm2 kg−1). Hysteresis loops are narrow waisted with all samples saturating at low fields (Figs 4a and b). Thermomagnetic data for sample GB00221 (Fig. 4c) contains no detectable heating curve inflection, which may indicate only minor quantities of magnetite are present. However, the sample undergoes thermochemical alteration with the production of magnetite, which is likely derived from oxidation or dewatering to clay minerals (e.g. Hirt & Gehring 1991). Sample GB0064B (Fig. 4d) experienced significant thermochemical alteration during heating with magnetite being the dominant magnetic mineral. Figure 4. View largeDownload slide Plots of hysteresis and thermomagnetic data. Hysteresis data for (a) GB00221 and (b) GB00064. Samples GB00221 has a mild pot-bellied morphology indicating minimal contribution of super paramagnetic magnetite and a dominance of single-domain grains (Tauxe et al.1996). Thermomagnetic data for (c) GB00221 and (d) GB00064. Sample GB00221 has very weak magnetic susceptibility with no discernible inflections during heating (black arrows). The cooling curves (grey arrows) indicate magnetite has formed possibly from dehydration/alteration of clay minerals. Figure 4. View largeDownload slide Plots of hysteresis and thermomagnetic data. Hysteresis data for (a) GB00221 and (b) GB00064. Samples GB00221 has a mild pot-bellied morphology indicating minimal contribution of super paramagnetic magnetite and a dominance of single-domain grains (Tauxe et al.1996). Thermomagnetic data for (c) GB00221 and (d) GB00064. Sample GB00221 has very weak magnetic susceptibility with no discernible inflections during heating (black arrows). The cooling curves (grey arrows) indicate magnetite has formed possibly from dehydration/alteration of clay minerals. DISCUSSION Influence of rock type, soil type and depth on magnetic susceptibility The laboratory magnetic susceptibility subdivided by major stratigraphic units is shown in Fig. 5(a). The median value of magnetic susceptibility is significantly elevated in Dunedin Volcanic Group rocks, relative to other stratigraphic units, and the median value in CaplesTerrane rocks is low relative to other stratigraphic units. There are relatively small differences between topsoil and subsoil values, and a number of stratigraphic units (Brook Street, Dun-Mountain Maitai, Dunedin Volcanic Group and Western Province), are represented by <10 samples. A more representative subdivision of the magnetic susceptibility data was by soil type and lithology. Figure 5. View largeDownload slide Plots showing variation in mass-specific magnetic susceptibility between topsoil and subsoil. (a) Tukey boxplot of data from different stratigraphic units. P: Pākihi Supergroup; ZM: Zealandia Megasequence; R: Rakaia Terrane; C: Caples Terrane; M: Murihiku Terrane; DM: Dun Mountain-Maitai Terrane; BS: Brook Street Terrane; DVG: Dunedin Volcanic Group; WP: Western Province rocks. (b) Tukey boxplot of data from five major soil types occurring over rocks with a common, acidic sedimentary lithology (brown, pallic, recent, gley and podzol) and a sixth group of mixed soils occurring over basic igneous rock types (basic). (c) Median magnetic susceptibility normalized to the median value in the subsoil. Only podzol soil shows a relative depletion. Diagram drawn after Hanesch & Scholger (2005). Figure 5. View largeDownload slide Plots showing variation in mass-specific magnetic susceptibility between topsoil and subsoil. (a) Tukey boxplot of data from different stratigraphic units. P: Pākihi Supergroup; ZM: Zealandia Megasequence; R: Rakaia Terrane; C: Caples Terrane; M: Murihiku Terrane; DM: Dun Mountain-Maitai Terrane; BS: Brook Street Terrane; DVG: Dunedin Volcanic Group; WP: Western Province rocks. (b) Tukey boxplot of data from five major soil types occurring over rocks with a common, acidic sedimentary lithology (brown, pallic, recent, gley and podzol) and a sixth group of mixed soils occurring over basic igneous rock types (basic). (c) Median magnetic susceptibility normalized to the median value in the subsoil. Only podzol soil shows a relative depletion. Diagram drawn after Hanesch & Scholger (2005). The five major soil types examined in this study are sampled over rocks that share a common lithology of acidic sedimentary rocks. The brown, pallic and recent soils have formed over either Otago Schist rock with a Mesozoic age, or Zealandia Megasequence sedimentary rocks with a Cenozoic age, with a skew in the age range towards younger, Quaternary aged rocks. The gley and podzol soil types included in this study form over the latter only (Cenozoic-aged sedimentary rocks). These five groups show the five different soil types formed over a common, acidic sedimentary lithology and account for 97 per cent of the data. The other 3 per cent are the sixth group of sites collected over basic igneous rock types. The mean value of magnetic susceptibility in samples collected over basic igneous rock types is nearly double that seen in the other five groups (Table 1 and Fig. 5b). Soils formed from basic igneous protoliths are typically associated with high magnetic susceptibility values (e.g. Fialová et al.2006). There are small differences in the mean, median and upper whisker values of topsoil samples versus subsoil samples for these groupings (Fig. 5b). In Fig. 5(c), the median magnetic susceptibility value has been normalized to that value obtained in the subsoil. In this way, the topsoil values for the recent soils are very weakly enriched (1.1), brown and pallic soils (the majority of sample points) are weakly enriched (1.3 and 1.4), gley soils and the basic group are moderately enriched (1.7 and 2.4) and podzol soils are weakly depleted, relative to the subsoil. These enrichments can occur naturally through the activity of iron-reducing bacteria, especially for well-drained, brown and pallic soils (Verosub & Roberts 1995; Maher 1998; Hanesch & Petersen 1999). In Podzol soils, transport and solution processes can lead to iron-oxide depletion and hence lower magnetic susceptibility values (e.g. Hanesch & Scholger 2002). Handheld versus laboratory techniques There is a strong, positive relationship between mass-specific susceptibility measured in the laboratory and volume-specific susceptibility measured by handheld techniques (Fig. 3c). A linear regression equation (Fig. 3c) calculated for this data (including outliers) means the handheld data can be converted into mass-specific values (10−8 m3 kg−1). This is important for studies where budget, time or laboratory access are issues, such as for exploration personnel, science students or other researchers where magnetic susceptibility is a secondary concern. The associated costs in personnel and laboratory times are an order of magnitude less using a handheld magnetic susceptibility meter. The advantages to the laboratory technique are higher precision. A range of additional data about remanence, mineralogy, etc., can also be determined in the laboratory, assuming the time and equipment are available. Using a handheld meter in the field is thus acceptable for studies of soil magnetic susceptibility where an appropriate regression can be calculated or studies where high precision is a lower priority (e.g. Hanesch & Scholger 2002; Schmidt et al.2005). Anomaly detection Three different approaches were used to construct maps of anomalous magnetic susceptibility values, namely the topsoil–subsoil difference method, the Tukey boxplot method and the geoaccumulation index method. The topsoil–subsoil difference method assumes that magnetic material accumulates in the topsoil from anthropogenic input, meaning an enrichment in the topsoil relative to the subsoil indicates an anthropogenic anomaly (e.g. Kapička et al.2001). This can be displayed graphically (Fig. 6a), and the enrichments subdivided (following Hanesch & Scholger 2002) into possible geogenic or pedogenic anomalies if the susceptibility difference is below −20 × 10−8 m3 kg−1, and possible anthropogenic anomalies if >20 × 10−8 m3 kg−1. Approximately 7 per cent of the data have a susceptibility difference >20 × 10−8 m3 kg−1 and thus a potential anthropogenic source per this method. It has been noted in central European studies that one limitation of this method is where high susceptibility values exist, for example, >200 × 10−8 m3 kg−1, a difference of >20 × 10−8 m3 kg−1 can also be a geogenic anomaly (Fialová et al.2006; Hanesch et al.2007). Before considering these data further, the other anomaly detection methods will be presented. Figure 6. View largeDownload slide Maps showing anomalous mass-specific magnetic susceptibility data determined by various methods. (a) Topsoil–subsoil difference method. Values <−20 and >20 are anomalous by this method. Some workers interpret the <−20 data (square symbols) as representing a geogenic source and > 20 data representing an anthropogenic source, though this simple interpretation is complicated when the soil parent material comes from high magnetic susceptibility source rocks. (b) Tukey boxplot method on topsoil data. The outlier data are shown as a point source map with a four-part, natural breaks (Jenks 1967) subdivision of the data value range. The four rock units shown for comparison are the Median Batholith, Murihiku Terrane, Dun Mountain-Maitai Terrane (DMMT) and Dunedin Volcanic Group (DVG). The locality of the Longwood Range (LR) and Takitimu Mountains (TM) are also shown. The sites where additional magnetic measurements were made on subsoil samples is indicated by a sexagon symbol (solid line for anomalous sites and dashed line for background sites). (c) Geoaccumulation Index (Igeo) method on topsoil data. The Waiau and Oreti & Aparima catchments are shown, as well as various locations discussed in the text (italics) and two sample sites (GB00064 and GB00071). Some major rivers are shown as blue lines in the catchment areas. Figure 6. View largeDownload slide Maps showing anomalous mass-specific magnetic susceptibility data determined by various methods. (a) Topsoil–subsoil difference method. Values <−20 and >20 are anomalous by this method. Some workers interpret the <−20 data (square symbols) as representing a geogenic source and > 20 data representing an anthropogenic source, though this simple interpretation is complicated when the soil parent material comes from high magnetic susceptibility source rocks. (b) Tukey boxplot method on topsoil data. The outlier data are shown as a point source map with a four-part, natural breaks (Jenks 1967) subdivision of the data value range. The four rock units shown for comparison are the Median Batholith, Murihiku Terrane, Dun Mountain-Maitai Terrane (DMMT) and Dunedin Volcanic Group (DVG). The locality of the Longwood Range (LR) and Takitimu Mountains (TM) are also shown. The sites where additional magnetic measurements were made on subsoil samples is indicated by a sexagon symbol (solid line for anomalous sites and dashed line for background sites). (c) Geoaccumulation Index (Igeo) method on topsoil data. The Waiau and Oreti & Aparima catchments are shown, as well as various locations discussed in the text (italics) and two sample sites (GB00064 and GB00071). Some major rivers are shown as blue lines in the catchment areas. One of the most useful methods to determine the background element concentration in soil is the Tukey boxplot (Tukey 1977; Reimann 2005). It divides the ordered values of the data into equal parts by finding the interquartile range between the 25th and 75th percentiles. The background concentration can be defined as 1.5 times the interquartile range (e.g. Hanesch et al.2007). The Tukey boxplot method has been applied using the soil-lithology subdivision outlined above (Table 1 and Fig. 5b). Outlier data are shown in Fig. 6(b), and subdivided based upon their concentration above background levels. The total number of anomalies by this method (n = 40) is slightly lower that the difference method used above (Fig. 6a) and the locality of the anomalies is different in some of the cases. One suggested approach to further refine the detection of anomalies is to calculate and plot the geoaccumulation index (Igeo; Müller 1979):   \begin{equation}{\rm{Igeo}} = {\rm{log}}2\left( {{\rm{Cn}}/1.5{\rm{Bn}}} \right)\end{equation} (1)where Cn is the value of the measured elemental concentration and Bn is the background value as determined by the Tukey boxplot method. The Igeo method attempts to compensate for natural background variation within groups. It reduces the total number of anomalies identified to 22 (Fig. 6c), but within similar areas of the survey to the Tukey boxplot method. The 22 anomalies identified by the Igeo method can be subdivided into four areas and two sites (Fig. 6c) within a common river catchment (Waiau catchment or Oreti and Aparima Catchment), terrane (Murihiku Terrane) or rock type (basic). Site GB00071 was collected in Milton township and site GB00064 was collected from farmland. These anomalous areas and sites are discussed further below. Neither the Tukey boxplot nor Igeo method distinguishes geogenic/pedogenic from anthropogenic anomalies. The Tukey boxplot method identifies all areas of anomalous results and is especially useful for groups of data numbering <50 (Reimann 2005). The Igeo method focuses on the upper end of the outlier data and suggests qualitative categories of contamination, for example, moderately to strongly contaminated for Igeo >2 (Müller 1979). Both methods appear useful for distinguishing anomalous magnetic susceptibility results from southern New Zealand. Nature of magnetic minerals Samples from anomalous sites identified by the Tukey boxplot and Igeo methods (Fig. 6b) were selected for further examination to understand the nature of magnetic mineralogy in them. Some samples from sites considered within normal background were also studied for comparison. The Hcr values are consistent with magnetite mineralogy where values can vary depending on grain size and composition (Fig. 7a; Day et al.1977) in agreement with thermomagnetic data (Figs 4c and d). Overall, the median coercivity of 41 mT indicates magnetite is the dominant remanence carrier at anomalous sites. A minor contribution from haematite may be present in some samples with higher Hcr values above 60 mT (Özdemir & Dunlop 2014). A day plot (Day et al.1977; Dunlop 2002a,b; Fig. 7b) indicates magnetic grains are mostly in the pseudo-single domain (PSD) and single domain (SD) grain size range. Anomalous sample sites from the Waiau catchment, Oreti and Aparima catchments and Murihiku Terrane partially overlap (Fig. 7b) at the centre of the PSD field with the exception of one Murihiku sample which falls in the multidomain (MD) field (labeled Longwood on Fig. 7b). The Dunedin Volcanic Group and Otago Schist rock samples plot slightly higher, with the Otago Schist samples lying at the PSD/SD-SP field transition, indicating a possible contribution of super paramagnetic (SP) grains from the Otago Schist or pedogenic processes above Otago Schist basement rock. All samples from the different groups show a degree of scatter, which is caused by the varying contribution of the soft component (SP or MD) in mixtures with SD grains (Dunlop 2002a). The two Dunedin Volcanic Group samples have different magnetic mineral concentrations (Mrs of 82 and 1.2 mAm2 kg−1) and coercivities (Hcr 40 and 25 mT) from one another, suggesting complexity in the magnetic mineralogy of the Dunedin Volcanic Group or that magnetic minerals have altered since erosion from their source. Figure 7. View largeDownload slide Subsoil magnetic data from 17 anomalous sites and four background sites. (a) An IRM Mrs versus Hcr plot indicating titanomagnetite is the main grain type. (b) A day plot indicating pseudo-single domain (PSD) magnetite is the dominant remanence carrier (Day et al.1977; Dunlop 2002a,b). Some analyses which approach the PSD-single domain (SD) mixing line may be biased by minor contributions of higher coercivity grains and a single sample from near the Longwood Range plots in the multidomain (MD) field and likely indicates a geogenic contribution of large titanomagnetite grains. (c) Magnetic susceptibility versus hysteresis and IRM determined magnetic concentration parameters (Ms and Mrs), which demonstrates a good correlation between magnetic mineral specific concentrations (Ms and Mrs) and magnetic susceptibility, which can be sensitive to contributions from paramagnetic and diamagnetic components in soil. Figure 7. View largeDownload slide Subsoil magnetic data from 17 anomalous sites and four background sites. (a) An IRM Mrs versus Hcr plot indicating titanomagnetite is the main grain type. (b) A day plot indicating pseudo-single domain (PSD) magnetite is the dominant remanence carrier (Day et al.1977; Dunlop 2002a,b). Some analyses which approach the PSD-single domain (SD) mixing line may be biased by minor contributions of higher coercivity grains and a single sample from near the Longwood Range plots in the multidomain (MD) field and likely indicates a geogenic contribution of large titanomagnetite grains. (c) Magnetic susceptibility versus hysteresis and IRM determined magnetic concentration parameters (Ms and Mrs), which demonstrates a good correlation between magnetic mineral specific concentrations (Ms and Mrs) and magnetic susceptibility, which can be sensitive to contributions from paramagnetic and diamagnetic components in soil. A biplot of magnetic susceptibility versus Ms and Mrs (Fig. 7c) reveals a close correlation between the two, indicating that magnetic susceptibility at most anomalous sites (e.g. Waiau and Oreti and Aparima catchments, Murihiku Terrane), is controlled by changes in mineral concentration rather than changes in mineral type, or in the proportions of ferromagnetic versus para/diamagnetic contributions. High coercivity minerals such as haematite, however, have saturation magnetization and magnetic susceptibility, which are orders of magnitude smaller than magnetite. Their relative contribution from sample to sample may therefore be underestimated from concentration-dependent measurements. Samples from the Waiau catchment have a higher average Hcr value of 61 which may indicate a greater quantity of high coercivity grains relative to other anomalous zones. Source of magnetic susceptibility anomalies Waiau catchment The Waiau catchment is an area of anomalous magnetic susceptibility identified by all three anomaly detection methods (Fig. 6). This area is sparsely populated, drained by the Waiau River and is bordered by Fiordland National Park. An obvious anthropogenic source for these anomalies is not present, but several, high magnetic susceptibility rock types occur within the catchment. For example, moderately to strongly magnetic rocks from the Median Batholith with I-type granites, gabbros and peridotites (Woodward & Hatherton 1975; Smale 1990) or Dun Mountain Ophiolite Belt ultramafic rocks (Hunt & Mumme 1978; Eccles et al.2005). Furthermore, the Longwood Range contains basic and ultrabasic rocks of the Longwood Suite that are known to contain titanomagnetite-bearing rock types (Price et al.2011). The day plot shows a site adjacent to the Longwood Range with particularly low Mrs/Ms and high Hcr/Hc relative to other sample sites in the Waiau catchment (Fig. 7b) indicating a dominance of large, MD grains that may be sourced from the Longwood Range (Ashley et al.2012). The Takitimu Mountains also contain likely titanomagnetite-bearing rocks (Houghton 1981) and anomalous soils sites adjacent to these mountains have formed mafic melanic and mafic brown soils and acidic allophanic brown soils. These soil types are the weathering products of igneous rocks and contain magnetite (Prasad & Ghildyal 1975), ferrihydrite and imogolite. Although a mineral that is uncommonly magnetic at ambient temperatures, ferrihydrite may carry a remanence (Johari Pannalal et al.2005). Soil samples from the Takitimu Mountains area also have higher coercivity values (Hc and Hcr) on average than samples from the rest of the studied area, which may indicate that the Takitimu Mountains are an unrecognized source of a high coercivity mineral such as haematite. Using the differences method of anomaly detection, two sites in the Waiau catchment suggested an anthropogenic source (triangle symbols in Fig. 6a), however, the strong magnetic susceptibility (>200 × 10−8 m3 kg−1) of probable source rocks in this catchment invalidates the criteria for assigning an anthropogenic source via this method. The preferred interpretation, then, is that there is a geogenic source to the magnetic susceptibility anomalies in the Waiau catchment. The provenance of magnetic minerals in soil in the catchment is suggested to be from igneous rocks of the Median Batholith, sourced from the upper part of the catchment in Fiordland National Park. There is likely local input from the Takitimu Mountains, and in the southern part of the catchment, input from the Longwood Suite in the Longwood Range may be the dominant source. Oreti and Aparima catchments Samples with anomalous magnetic susceptibility also occur along the flood plain of the Oreti and Aparima rivers (Fig. 6) above Late Pleistocene river gravels in brown or recent soil types. Six anomalous sites have been identified by the Igeo method (Fig. 7c). Again this is a sparsely populated region, dominated by dairy, sheep and beef farming, without an obvious anthropogenic source of contamination. The Oreti River does, however, drain the Dun Mountain Ophiolite Belt, and the Aparima River drains the Takitimu Mountains, all of which have high concentrations of magnetic minerals. It has been shown elsewhere that heavy minerals from the ophiolite belt have been eroded and redeposited along flood plains of the Oreti River (Martin et al.2016) and a similar explanation is envisioned to explain the Oreti and Aparima catchment anomalies. Namely, magnetite has been eroded and redeposited along the Oreti and Aparima river floodplains and captured during the study, a distance ≤100 km. Murihiku Terrane At least three anomalous sites have been identified in the Murihiku Terrane using each of the three methods of anomaly detection (Fig. 6). In absence of anthropogenic sources of contamination, a natural source is probable, which is consistent with the difference between the topsoil and subsoil being −20 × 10−8 mg3 kg−1 or less (square symbols Fig. 6a). The anomalous samples have formed adjacent to tuffaceous beds (e.g. Campbell et al.2003; Turnbull & Allibone 2003) and in magnetic, allophanic soils and brown soils. Some anomalous sites are formed in fragic pallic soils sourced from adjacent Dun Mountain-Maitai Terrane rocks, most probably Little Ben Sandstone rocks that were derived from a basaltic–andesitic volcanic arc. Little Ben Sandstone rocks are known to contain magnetite, haematite and pseudomorphs after haematite (Landi 1980). This suggests the provenance of magnetic minerals in anomalous sites in the Murihiku Terrane is proximal, certainly less than the terrane width (maximum 80 km), but most likely <1 km. Basic rocks The number of anomalies above Dunedin Volcanic Group rocks decreases from eight using the differences method, down to three using the Tukey boxplot method, to two using the Igeo method. The three anomalies above Dunedin Volcanic Group rocks (Fig. 7b) using the Tukey boxplot method need not invoke an anthropogenic source, and despite their relative proximity to Dunedin City, they occur in areas of farmland or forestry. The anomalous sites occur in mafic melanic soils and mafic brown soils, formed form igneous protoliths and, the alkalic igneous rocks that comprise Dunedin Volcanic Group are known to contain titanomagnetite (Wright 1967; Sherwood 1988), and are relatively magnetic (Woodward & Hatherton 1975). The provenance of magnetic minerals in these anomalous sites is proximal (<1 km). One sample site occurs over ultramafic to mafic rocks of the Dun Mountain Ophiolite Belt (Figs 6b and c) and also has a geogenic source. Anthropogenic sources Two anomalous sites (GB00064 and GB00071) are identified by all three anomaly methods (Fig. 7) but have no immediately obvious geogenic source. The topsoil–subsoil difference in magnetic susceptibility values at each site is consistent with anthropogenic contamination (Fig. 7a). At site GB00064, relatively magnetic Dunedin Volcanic Group rocks occur within 5 km of the sample area; however, this sample was also collected adjacent to a metalled road. In this region, aggregate from Dunedin Volcanic Group rocks is commonly used as roading aggregate. A probable explanation for the magnetic susceptibility anomaly for this location, then, is sample bias included road metal. Site GB00071 was collected in the township of Milton (population ca. 2000). The site is known to have relatively high concentrations of heavy metals, assigned to an anthropogenic source (Martin et al.2016). This suggests the magnetic susceptibility anomaly at this site is anthropogenic and proximal. SUMMARY AND CONCLUSIONS The volume-specific magnetic susceptibility (handheld method) and mass-specific magnetic susceptibility (laboratory method) of topsoil and subsoil were measured in a regional study of southern New Zealand. A strong correlation between the volume- and mass-specific methods suggests the faster and more economic handheld method could be justifiably used in future soil studies in New Zealand. A linear regression calculation on the data in this study could be used in the future to convert volume- to mass-specific magnetic susceptibility. Samples collected in this study can be divided into groups with shared soil type and/or lithology. All groups have similar median values of magnetic susceptibility, except for sites sampled over basic igneous rocks, which were significantly higher. Most groups were slightly too moderately enriched in the topsoil, relative to the subsoil, in a natural pedogenic process. Three different methods, the difference method, Tukey boxplot method and Igeo method, were used to identify sites with anomalous magnetic susceptibility in soil from southern New Zealand. The threshold for identifying anomalies is as follows: difference method < Tukey boxplot < Igeo, such that the difference method detects many more anomalies than the Igeo method, as has been found in other studies (Reimann 2005; Hanesch et al.2007). The magnetic mineralogy is dominated by magnetite, but certain areas around the Takitimu Mountains, Longwood Range and Dunedin Volcanic Group indicated variable magnetic mineralogy (e.g. minor contribution of haematite). In all except two anomalous sites, a geogenic source was the most probable explanation for magnetic susceptibility above natural background. This is consistent with the soil types at these anomalous sites that typically formed from igneous sources, the underlying bedrock geology that is a known source of magnetite or being in a catchment draining areas with rocks of known magnetic mineralogy. There is a minimal anthropogenic input detectable by magnetic susceptibility methods, and by proxy, a minimum of heavy metals, such as Cu, Pb and Zn, contaminating southern New Zealand soils. This is consistent with findings of low anthropogenic input of heavy metals in southern New Zealand, relative to elsewhere in New Zealand (Martin et al.2016). The provenance of magnetic mineralogy in anomalous sites is generally within the bounds of the rock type, catchment or terrane they were sampled within. This is a distance ≤100 km and frequently much closer (<1 km). In the Waiau catchment, the source of magnetic mineralogy is from Median Batholith rocks from within or near Fiordland National Park, with proximal sources from the Takitimu Mountains or Longwood Range becoming significant, or even dominant, in the southern areas of the catchment. In the Oreti and Aparima catchments, the dominant source of magnetic minerals is the Dun Mountain Ophiolite Belt or Takitimu Mountains. The study of soil magnetic susceptibility allows a more thorough understanding of soil distribution and is a powerful tool for sediment provenance studies. This will be important when trying to understand element or isotopic distribution, vectors to mineralization or variations in soil quality (for farming or for water quality). The low proportion of anomalous sites with anthropogenic contamination would be expected to increase when areas adjacent to larger urban centres on the North Island of New Zealand are studied. In future studies in New Zealand, it may be sufficient to study only the topsoil sample for magnetic susceptibility using a handheld method. This study should serve as a magnetic susceptibility reference library for similar soil types found in New Zealand, and guide future soil sampling studies. Acknowledgements This work was funded by the Government of New Zealand through a GNS Science strategic development fund. Phil Rieger, Amy Beatson and Chantelle Hillier assisted with sample collection and processing. We thank two anonymous reviewers and the editor for their constructive comments which substantially improved the manuscript. REFERENCES Ashley P., Craw D., Mackenzie D., Rombouts M., Reay A., 2012. Mafic and ultramafic rocks, and platinum mineralisation potential, in the Longwood Range, Southland, New Zealand, N. Z. J. Geol. Geophys. , 55( 1), 3– 19. https://doi.org/10.1080/00288306.2011.623302 Google Scholar CrossRef Search ADS   Benson W., 1968. Dunedin District. 1:50,000. NZ Geological Survey Miscellaneous Series Map 1 . Wellington, New Zealand, Department of Scientific and Industrial Research. Bian Y., Ouyang T., Zhu Z., Huang N., Wan H., Li M., 2014. Magnetic properties of agricultural soil in the Pearl River Delta, South China — Spatial distribution and influencing factor analysis, J. appl.Geophysics , 107, 36– 44. https://doi.org/10.1016/j.jappgeo.2014.05.003 Google Scholar CrossRef Search ADS   Bishop D., Turnbull I., 1996. Geology of the Dunedin Area: Scale 1:250 000, pp. Geological Map 21. 52 p. + 21 Folded Map . Institute of Geological & Nuclear Sciences, Lower Hutt. Blundell A., Dearing J.A., Boyle J.F., Hannam J.A., 2009. Controlling factors for the spatial variability of soil magnetic susceptibility across England and Wales, Earth-Sci. Rev. , 95( 3–4), 158– 188. https://doi.org/10.1016/j.earscirev.2009.05.001 Google Scholar CrossRef Search ADS   Campbell H.J., Mortimer N., Turnbull I.M., 2003. Murihiku Supergroup, New Zealand: redefined, J. R. S. N. Z. , 33( 1), 85– 95. https://doi.org/10.1080/03014223.2003.9517722 Google Scholar CrossRef Search ADS   Coombs D.S., Adams C.J., Roser B.P., Reay A., 2008. Geochronology and geochemistry of the Dunedin Volcanic Group, eastern Otago, New Zealand, N. Z. J.Geol. Geophys. , 51( 3), 195– 218. https://doi.org/10.1080/00288300809509860 Google Scholar CrossRef Search ADS   Cullers R.L., Basu A., Suttner L.J., 1988. Geochemical signature of provenance in sand-size material in soils and stream sediments near the Tobacco Root batholith, Montana, U.S.A., Chem. Geol. , 70( 4), 335– 348. https://doi.org/10.1016/0009-2541(88)90123-4 Google Scholar CrossRef Search ADS   Day R., Fuller M., Schmidt V.A., 1977. Hysteresis properties of titanomagnetites: grain-size and compositional dependence, Phys. Earth planet. Inter. , 13( 4), 260– 267. https://doi.org/10.1016/0031-9201(77)90108-X Google Scholar CrossRef Search ADS   Dearing J.A., Bird P.M., Dann R.J.L., Benjamin S.F., 1997. Secondary ferrimagnetic minerals in Welsh soils: a comparison of mineral magnetic detection methods and implications for mineral formation, Geophys. J. Int. , 130( 3), 727– 736. https://doi.org/10.1111/j.1365-246X.1997.tb01867.x Google Scholar CrossRef Search ADS   Dearing J.A., Hay K.L., Baban S.M.J., Huddleston A.S., Wellington E.M.H., Loveland P.J., 1996. Magnetic susceptibility of soil: an evaluation of conflicting theories using a national data set, Geophys. J. Int. , 127( 3), 728– 734. https://doi.org/10.1111/j.1365-246X.1996.tb04051.x Google Scholar CrossRef Search ADS   Drewry J.J., Littlejohn R.P., Paton R.J., 2000. A survey of soil physical properties on sheep and dairy farms in southern New Zealand, N. Z. J. Agric. Res. , 43( 2), 251– 258. https://doi.org/10.1080/00288233.2000.9513425 Google Scholar CrossRef Search ADS   Dunlop D., 2002a. Theory and application of the Day plot (Mrs/Ms versus Hcr/Hc) 1. Theoretical curves and tests using titanomagnetite data, J. geophys. Res. , 107, EPM 4-1–EPM 4-22. Dunlop D., 2002b. Theory and application of the Day plot (Mrs/Ms versus Hcr/Hc) 2. Application to data for rocks, sediments, and soils, J. geophys. Res. , 107, EPM 5-1–EPM 5-15. Eccles J.D., Cassidy J., Locke C.A., Sporli K.B., 2005. Aeromagnetic imaging of the Dun Mountain Ophiolite Belt in northern New Zealand: insight into the fine structure of a major SW Pacific terrane suture, J. geol. Soc. , 162( 4), 723– 735. https://doi.org/10.1144/0016-764904-060 Google Scholar CrossRef Search ADS   Edbrooke S.W., Heron D.W., Forsyth P.J., Jongens R., 2015. Geological map of New Zealand 1:1 000 000. GNS Science Geological Map 2. 2 sheetsGNS Science , Lower Hutt, New Zealand. Fialová H., Maier G., Petrovský E., Kapička A., Boyko T., Scholger R., 2006. Magnetic properties of soils from sites with different geological and environmental settings, J. appl. Geophys. , 59( 4), 273– 283. https://doi.org/10.1016/j.jappgeo.2005.10.006 Google Scholar CrossRef Search ADS   Fieldes M., 1955. Clay mineralogy of New Zealand soils, N. Z. J. Sci. Technol. , B37, 336– 350. Hanesch M., Petersen N., 1999. Magnetic properties of a recent parabrown-earth from Southern Germany, Earth planet. Sci. Lett. , 169( 1–2), 85– 97. https://doi.org/10.1016/S0012-821X(99)00076-X Google Scholar CrossRef Search ADS   Hanesch M., Rantitsch G., Hemetsberger S., Scholger R., 2007. Lithological and pedological influences on the magnetic susceptibility of soil: their consideration in magnetic pollution mapping, Sci. Total Environ. , 382( 2–3), 351– 363. https://doi.org/10.1016/j.scitotenv.2007.04.007 Google Scholar CrossRef Search ADS PubMed  Hanesch M., Scholger R., 2002. Mapping of heavy metal loadings in soils by means of magnetic susceptibility measurements, Environ. Geol. , 42( 8), 857– 870. https://doi.org/10.1007/s00254-002-0604-1 Google Scholar CrossRef Search ADS   Hanesch M., Scholger R., 2005. The influence of soil type on the magnetic susceptibility measured throughout soil profiles, Geophys. J. Int. , 161( 1), 50– 56. https://doi.org/10.1111/j.1365-246X.2005.02577.x Google Scholar CrossRef Search ADS   Hannam J.A., Dearing J.A., 2008. Mapping soil magnetic properties in Bosnia and Herzegovina for landmine clearance operations, Earth planet. Sci. Lett. , 274( 3–4), 285– 294. https://doi.org/10.1016/j.epsl.2008.05.006 Google Scholar CrossRef Search ADS   Haughton P.D.W., Todd S.P., Morton A.C., 1991. Sedimentary provenance studies, Geol. Soc. Lond. Spec. Publ. , 57( 1), 1– 11. https://doi.org/10.1144/GSL.SP.1991.057.01.01 Google Scholar CrossRef Search ADS   Heron D.W., 2014. Geological Map of New Zealand 1:250,000 GNS Science. GNS Science Geological Map 1 , Lower Hutt. Hewitt A.E., 2010. New Zealand Soil Classification , third edn, Manaaki Whenua Press, Lincoln, New Zealand. Hirt A.M., Gehring A.U., 1991. Thermal alteration of the magnetic mineralogy in ferruginous rocks, J. geophys. Res. , 96( B6), 9947– 9953. https://doi.org/10.1029/91JB00573 Google Scholar CrossRef Search ADS   Houghton B., 1981. Lithostratigraphy of the Takitimu Group, central Takitimu Mountains, western Southland, New Zealand, N. Z. J. Geol. Geophys. , 24, 333– 348. Hunt T., Mumme T., 1978. Geophysical Reconnaissance of the Red Mountain Ophiolite Complex, Southern Westland, Geophysics Division Report, 128 . Department of Scientific and Industrial Research, Wellington. Jenks G., 1967. The data model concept in statistical mapping, Inter. Yearbook Cartogr. , 7, 186– 190. Jenny H., 1941. Factors of soil formation , McGraw-Hill Book Company. https://doi.org/10.1097/00010694-194111000-00009 Joharipannalal S., Crowe S., Cioppa M., Symons D., Sturm A., Fowle D., 2005. Room-temperature magnetic properties of ferrihydrite: a potential magnetic remanence carrier?, Earth planet. Sci. Lett. , 236( 3–4), 856– 870. https://doi.org/10.1016/j.epsl.2005.05.019 Google Scholar CrossRef Search ADS   Jordanova N., Jordanova D., Petrov P., 2016. Soil magnetic properties in Bulgaria at a national scale—Challenges and benefits, Global Planet. Change , 137, 107– 122. https://doi.org/10.1016/j.gloplacha.2015.12.015 Google Scholar CrossRef Search ADS   Kapička A., Petrovský E., Jordanova N., Podrázský V., 2001. Magnetic parameters of forest top soils in Krkonoše mountains, Czech Republic, Phys. Chem. Earth A: Solid Earth Geod. , 26( 11–12), 917– 922. https://doi.org/10.1016/S1464-1895(01)00142-9 Google Scholar CrossRef Search ADS   Landis C.A., 1980. Little Ben Sandstone, Maitai Group (Permian): nature and extent in the Hollyford-Eglinton region, South Island, New Zealand, N. Z. J. Geol. Geophys. , 23( 5–6), 551– 567. https://doi.org/10.1080/00288306.1980.10424127 Google Scholar CrossRef Search ADS   Magiera T., Strzyszcz Z., Kapicka A., Petrovsky E., 2006. Discrimination of lithogenic and anthropogenic influences on topsoil magnetic susceptibility in Central Europe, Geoderma , 130( 3–4), 299– 311. https://doi.org/10.1016/j.geoderma.2005.02.002 Google Scholar CrossRef Search ADS   Maher B.A., 1998. Magnetic properties of modern soils and Quaternary loessic paleosols: paleoclimatic implications, Palaeogeog. Palaeoclimat. Palaeoecol. , 137( 1–2), 25– 54. https://doi.org/10.1016/S0031-0182(97)00103-X Google Scholar CrossRef Search ADS   Martin A.P., Turnbull R.E., Rattenbury M.S., Baisden W.T., Christie A.B., Cohen D.R., Hoogewerff J.A., Rogers K.M., 2015. Geochemical Atlas of Southern New Zealand, p. 221, GNS Science Report 2015/26. vii , Lower Hutt, New Zealand. Martin A.P., Turnbull R.E., Rattenbury M.S., Cohen D.R., Hoogewerff J., Rogers K.M., Baisden W.T., Christie A.B., 2016. The regional geochemical baseline soil survey of southern New Zealand: design and initial interpretation, J. Geochem. Explor. , 167, 70– 82. https://doi.org/10.1016/j.gexplo.2016.05.009 Google Scholar CrossRef Search ADS   Martin A.P., Turnbull R.E., Rissmann C.W., Rieger P., 2017. Heavy metal and metalloid concentrations in soils under pasture of southern New Zealand, Geoderm. Region. , 11( Suppl. C), 18– 27. https://doi.org/10.1016/j.geodrs.2017.08.005 Google Scholar CrossRef Search ADS   Molloy L.F., Christie Q.R., 1988. Soils in the New Zealand Landscape: The Living Mantle , Mallinson Rendel in Association with the New Zealand Society of Soil Science, Wellington. Mortimer N. et al.  , 2014. High-level stratigraphic scheme for New Zealand rocks, N. Z. J. Geol. Geophys. , 57, 402– 419. https://doi.org/10.1080/00288306.2014.946062 Google Scholar CrossRef Search ADS   Müller G., 1979. Schwermetalle in den sedimenten des rheins—veränderungen seit 1971, Umsch. Wiss. Tech. , 79, 778– 783. Nichol S.L., Augustinus P.C., Gregory M.R., Creese R., Horrocks M., 2000. Geomorphic and sedimentary evidence of human impact on the New Zealand coastal landscape, Phys. Geogr. , 21, 109– 132. Özdemir Ö., Dunlop D.J., 2014. Hysteresis and coercivity of hematite, J. geophys. Res. , 119, 2582– 2594.. https://doi.org/10.1002/2013JB010739 Google Scholar CrossRef Search ADS   Parfitt R., 1990. Allophane in New Zealand—a review, Aust. J. Soil Res. , 28, 343– 360. https://doi.org/10.1071/SR9900343 Google Scholar CrossRef Search ADS   Parfitt R.L., Russell M., Orbell G.E., 1983. Weathering sequence of soils from volcanic ash involving allophane and halloysite, New Zealand, Geoderma , 29, 41– 57. https://doi.org/10.1016/0016-7061(83)90029-0 Google Scholar CrossRef Search ADS   Pillans B., Wright I., 1990. 500,000-Year paleomagnetic record from New Zealand loess, Quat. Res. , 33, 178– 187. https://doi.org/10.1016/0033-5894(90)90017-F Google Scholar CrossRef Search ADS   Prasad B., Ghildyal B.P., 1975. Magnetic susceptibility of lateritic soils and clays, Soil Sci ., 120, 219– 229. https://doi.org/10.1097/00010694-197509000-00010 Google Scholar CrossRef Search ADS   Price R., Spandler C., Arculus R., Reay A., 2011. The Longwood Igneous Complex, Southland, New Zealand: a Permo-Jurassic, intra-oceanic, subduction-related, I-type batholithic complex, Lithos , 126, 1– 21. https://doi.org/10.1016/j.lithos.2011.04.006 Google Scholar CrossRef Search ADS   Rachwał M., Kardel K., Magiera T., Bens O., 2017. Application of magnetic susceptibility in assessment of heavy metal contamination of Saxonian soil (Germany) caused by industrial dust deposition, Geoderma , 295, 10– 21. https://doi.org/10.1016/j.geoderma.2017.02.007 Google Scholar CrossRef Search ADS   Rattenbury M.S., Martin A.P., Turnbull R.E., Christie A.B., 2014. Sampling methodology for a regional multi-element geochemical baseline survey. GNS Science Report 2014/62GNS Science, Lower Hutt, New Zealand. Reimann C., 2005. Sub-continental-scale geochemical mapping: sampling, quality control and data analysis issues, Geochem. Explor. Environ. Anal. , 5, 311– 323. https://doi.org/10.1144/1467-7873/03-065 Google Scholar CrossRef Search ADS   Rogers K.M., Turnbull R.E., Martin A.P., Baisden W.T., Rattenbury M.S., 2017. Stable isotopes reveal human influences on southern New Zealand soils, Appl. Geochem. , 82, 15– 24. https://doi.org/10.1016/j.apgeochem.2017.05.006 Google Scholar CrossRef Search ADS   Schmidt A., Yarnold R., Hill M., Ashmore M., 2005. Magnetic susceptibility as proxy for heavy metal pollution: a site study, J. Geochem. Explor. , 85, 109– 117. https://doi.org/10.1016/j.gexplo.2004.12.001 Google Scholar CrossRef Search ADS   Sherwood G.J., 1988. Rock magnetic studies of Miocene volcanics in eastern Otago and Banks Peninsula, New Zealand: comparison between Curie temperature and low temperature susceptibility behaviour, N. Z. J. Geol. Geophys. , 31, 225– 235. https://doi.org/10.1080/00288306.1988.10417771 Google Scholar CrossRef Search ADS   Smale D., 1990. Distribution and provenance of heavy minerals in the South Island: a review, N. Z. J. Geol. Geophys. , 33, 557– 571. https://doi.org/10.1080/00288306.1990.10421374 Google Scholar CrossRef Search ADS   Tauxe L., Mullender T.A.T., Pick T., 1996. Potbellies, wasp-waists, and superparamagnetism in magnetic hysteresis, J. geophys. Res. , 101, 571– 583. https://doi.org/10.1029/95JB03041 Google Scholar CrossRef Search ADS   Tukey J.W., 1977. Exploratory Data Analysis , Addison-Wesley. Turnbull I., Allibone A., 2003. Geology of the Murihiku Area, 1:250 000 Geological Map 20 . 1 sheet and 74 p. Institute of Geological & Nuclear Sciences Limited, Lower Hutt, New Zealand. Turner G.M., 1997. Environmental magnetism and magnetic correlation of high resolution lake sediment records from Northern Hawke's Bay, New Zealand, N. Z. J. Geol. Geophys. , 40, 287– 298. https://doi.org/10.1080/00288306.1997.9514761 Google Scholar CrossRef Search ADS   Verosub K.L., Roberts A.P., 1995. Environmental magnetism: past, present, and future, J. geophys. Res. , 100, 2175– 2192. https://doi.org/10.1029/94JB02713 Google Scholar CrossRef Search ADS   Woodward D.J., Hatherton T., 1975. Magnetic anomalies over southern New Zealand, N. Z. J. Geol. Geophys. , 18, 65– 82. https://doi.org/10.1080/00288306.1975.10426347 Google Scholar CrossRef Search ADS   Wright J.B., 1967. The iron-titanium oxides of some Dunedin (New Zealand) Lavas, in relation to their palaeomagnetic and thermomagnetic character (with an appendix on associated Chromiferous Spinel), Mineral. Mag. , 36, 425– 435. https://doi.org/10.1180/minmag.1967.036.279.13 Google Scholar CrossRef Search ADS   SUPPORTING INFORMATION Supplementary data are available at GJI online. aMartin_etal_SoilMagSus_SupplementaryMaterials_R2.pdf Please note: Oxford University Press is not responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the paper. © The Author(s) 2017. Published by Oxford University Press on behalf of The Royal Astronomical Society. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Geophysical Journal International Oxford University Press

Soil magnetic susceptibility mapping as a pollution and provenance tool: an example from southern New Zealand

Loading next page...
 
/lp/ou_press/soil-magnetic-susceptibility-mapping-as-a-pollution-and-provenance-emtqQKouBN
Publisher
Oxford University Press
Copyright
© The Author(s) 2017. Published by Oxford University Press on behalf of The Royal Astronomical Society.
ISSN
0956-540X
eISSN
1365-246X
D.O.I.
10.1093/gji/ggx484
Publisher site
See Article on Publisher Site

Abstract

Summary The presence or absence, degree and variation of heavy metal contamination in New Zealand soils is a matter of ongoing debate as it affects soil quality, agriculture and human health. In many instances, however, the soil heavy metal concentration data do not exist to answer these questions and the debate is ongoing. To address this, magnetic susceptibility (a common proxy for heavy metal contamination) values were measured in topsoil (0–30 cm) and subsoil (50–70 cm) at grid sites spaced at 8 km intervals across ca. 20 000 km2 of southern New Zealand. Samples were measured for both mass- and volume-specific magnetic susceptibility, with results being strongly, positively correlated. Three different methods of determining anomalies were applied to the data including the topsoil–subsoil difference method, Tukey boxplot method and geoaccumulation index method, with each method filtering out progressively more anomalies. Additional soil magnetic (hysteresis, isothermal remanence and thermomagnetic) measurements were made on a select subset of samples from anomalous sites. Magnetite is the dominant remanence carrying mineral, and magnetic susceptibility is governed by that minerals concentration in soils, rather than mineral type. All except two anomalous sites have a dominant geogenic source (cf. anthropogenic). By proxy, heavy metal contamination in southern New Zealand soils is minimal, making them relatively pristine. The provenance of the magnetic minerals in the anomalous sites can be traced back to likely sources in outcrops of igneous rocks within the same catchment, terrane or rock type: a distance of <100 km but frequently <1 km. Soil provenance is a key step when mapping element or isotopic distribution, vectoring to mineralization or studying soil for agricultural suitability, water quality or environmental regulation. Measuring soil magnetic susceptibility is a useful, quick and inexpensive tool that usefully supplements soil geochemical data. New Zealand, Environmental magnetism, Magnetic mineralogy and petrology, Rock and mineral magnetism, Spatial analysis INTRODUCTION Soil is the product of its source material, climate, age of formation and bioturbation, among other factors (Jenny 1941). Soil can act as an excellent tracer through the landscape, if an appropriate proxy is measured linking it to a probable source (Cullers et al.1988; Haughton et al.1991). Soil may be studied in numerous ways, including its physical properties, chemistry, stable or radiogenic isotopic composition or by its volumetric composition (e.g. organic matter, lithics, clays and minerals). This study focuses on the magnetic mineralogy of soil. Regional- and country-scale studies of magnetic susceptibility and frequency-dependent magnetic susceptibility in topsoil (0– ≤30 cm) have been conducted in England and Wales (Dearing et al.1996, 1997; Blundell et al.2009), Poland, Czech Republic and Germany (Magiera et al.2006) and Bosnia and Herzegovina (Hannam & Dearing 2008). In Austria, magnetic susceptibility measurements were made of both the topsoil (0–20 cm) and subsoil (50–70 cm; Hanesch et al.2007). In addition to magnetic susceptibility, Bian et al. (2014) measured anhysteretic remanence, isothermal remanence, temperature-dependent magnetic susceptibility and other parameters in topsoil (0–20 cm) from the Pearl River Delta area (ca. 41 000 km2), China. In a similar study, Jordanova et al. (2016) measured anhysteretic remanence, isothermal remanence, hysteresis parameters, among other parameters in Bulgarian topsoils. In the Free State of Saxony, Germany, Rachwal et al. (2017) measured magnetic susceptibility as well as the concentration of potentially toxic elements in the organic horizon and topsoil. The motivation for these studies includes understanding pollution, geogenic controls and magnetic mineralogy in soils to characterize anthropogenic affect, pollution sources and better detection of unexploded ordnance. The severity and distribution of heavy metal pollution on New Zealand soils is still a matter of debate, as is the proximal versus distal provenance of material in certain soil types. Using the regional- and national-scale surveys (above) as a model, soil samples from ca. 20 000 km2 of southern New Zealand were collected on an 8 km grid (Fig. 1a) and studied. This area is equivalent in size to Wales or West Virginia. Both topsoil (0–30 cm) and subsoil (50–70 cm) were measured for their volume- and mass-specific magnetic susceptibility by handheld and laboratory techniques. A subset of samples was chosen from anomalous magnetic susceptibility sites and from background sites, for additional measurement of their hysteresis, isothermal remanence and thermomagnetic properties. To our knowledge, this is the first time such a study has been conducted in the Southern Hemisphere. In New Zealand, small-scale investigations have used: magnetic susceptibility changes in loess to study climate change (Pillans & Wright 1990); magnetic fingerprinting to trace increased sedimentation (e.g. Nichol et al.2000) and; magnetic susceptibility for stratigraphic correlation in environmental studies (Turner 1997). This regional-scale study will identify magnetic anomalies, geogenic versus anthropogenic sources, discuss magnetic mineralogy and identify likely sources of magnetic mineralogy in soil. The results will help quantify heavy metal pollution in southern New Zealand, which can affect land and shallow groundwater quality, and help trace soil provenance. This paper is one of many recent and ongoing outputs from the geochemical baseline soil survey of southern New Zealand, and is a pilot study before a national survey of New Zealand is undertaken. Figure 1. View largeDownload slide The location of sample sites and the underlying soil and rock types. (a) Position of the survey within New Zealand. (b) The major New Zealand classified soil orders in southern New Zealand as taken from the Landcare Research Fundamental Soils Layer (lris.scinfo.org.nz; last accessed 2016 August 2). The international equivalents to the soil orders in this study are shown in Table 1. (c) The major lithostratigraphic units after Heron (2014) and Mortimer et al. (2014). The three townships (Dunedin, Milton and Invercargill) in the study area are marked for reference. The position of the Otago Schist metamorphic belt is shown by dashed lines. Figure 1. View largeDownload slide The location of sample sites and the underlying soil and rock types. (a) Position of the survey within New Zealand. (b) The major New Zealand classified soil orders in southern New Zealand as taken from the Landcare Research Fundamental Soils Layer (lris.scinfo.org.nz; last accessed 2016 August 2). The international equivalents to the soil orders in this study are shown in Table 1. (c) The major lithostratigraphic units after Heron (2014) and Mortimer et al. (2014). The three townships (Dunedin, Milton and Invercargill) in the study area are marked for reference. The position of the Otago Schist metamorphic belt is shown by dashed lines. REGIONAL SETTING Southern New Zealand has diverse geology and soil cover (Fig. 1). Five major soil orders found in New Zealand were part of this study (Fig. 1b; Hewitt 2010), with the world reference base soil classification equivalents shown in Table 1. Brown (n = 171) and pallic (n = 88) soils are particularly common in the study area, with recent, gley and podzol soils also well represented (Drewry et al.2000; Martin et al.2015). Of particular interest in this study are soils derived from igneous parent material, which include mafic brown and allophanic soils. The latter contains allophane minerals (Parfitt 1990) as well as ferrihydrite and imogolite (Molloy & Christie 1988) and typically occur as the weathering products of igneous rocks (Fieldes 1955; Parfitt et al.1983). Table 1. The New Zealand soil classification types, and their international equivalents, together with the Tukey boxplot (Fig. 5b) mass-specific magnetic susceptibility (10−8 m3 kg−1). Group  #  WRB  US soil taxonomy  Lithology  Min  Q1  Median  Q3  Max  Outlier  Far  Mean  Topsoil  Brown  158  Cambisols  Dystrochrepts  Schist; sediment  0.49  10.37  16.31  34.77  70.85  71.37  107.97  36.26  Pallic  83  Luvisols  Fragiudalfs, Fragiochrepts, Haplustalfs  Schist; sediment  3.51  8.23  12.67  21.55  34.63  41.53  61.51  21.13  Recent  30  Regosols  Fluvents, Orthents, Udepts  Schist; sediment  7.92  11.03  18.82  55.48  121.93  122.17  188.86  49.88  Gley  16  Gleysols  Aquepts  Sediment  4.89  11.23  29.65  61.15  115.78  136.04    46.25  Podzol  12  Podzols  Aquods, Orthods  Sediment  1.38  6.55  8.21  29.65  42.70  –  –  17.32  Basic  11  Mixed  Mixed  Basic igneous  3.67  14.02  42.16  120.09  249.18  279.20  –  92.91  Subsoil  Brown  161  Cambisols  Dystrochrepts  Schist; sediment  0.18  8.13  12.67  26.09  48.99  53.02  79.95  35.11  Pallic  85  Luvisols  Fragiudalfs, Fragiochrepts, Haplustalfs  Schist; sediment  1.80  6.57  9.09  18.15  35.24  35.53  52.90  15.56  Recent  29  Regosols  Fluvents, Orthents, Udepts  Schist; sediment  5.24  10.31  17.19  30.08  30.34  59.74  89.41  43.45  Gley  17  Gleysols  Aquepts  Sediment  2.79  5.61  17.06  33.94  68.22  76.43  118.93  33.07  Podzol  13  Podzols  Aquods, Orthods  Sediment  4.22  5.86  12.55  19.63  34.03  –  –  13.89  Basic  10  Mixed  Mixed  Basic igneous  5.81  12.97  17.30  149.18  350.62  –  –  85.07  Group  #  WRB  US soil taxonomy  Lithology  Min  Q1  Median  Q3  Max  Outlier  Far  Mean  Topsoil  Brown  158  Cambisols  Dystrochrepts  Schist; sediment  0.49  10.37  16.31  34.77  70.85  71.37  107.97  36.26  Pallic  83  Luvisols  Fragiudalfs, Fragiochrepts, Haplustalfs  Schist; sediment  3.51  8.23  12.67  21.55  34.63  41.53  61.51  21.13  Recent  30  Regosols  Fluvents, Orthents, Udepts  Schist; sediment  7.92  11.03  18.82  55.48  121.93  122.17  188.86  49.88  Gley  16  Gleysols  Aquepts  Sediment  4.89  11.23  29.65  61.15  115.78  136.04    46.25  Podzol  12  Podzols  Aquods, Orthods  Sediment  1.38  6.55  8.21  29.65  42.70  –  –  17.32  Basic  11  Mixed  Mixed  Basic igneous  3.67  14.02  42.16  120.09  249.18  279.20  –  92.91  Subsoil  Brown  161  Cambisols  Dystrochrepts  Schist; sediment  0.18  8.13  12.67  26.09  48.99  53.02  79.95  35.11  Pallic  85  Luvisols  Fragiudalfs, Fragiochrepts, Haplustalfs  Schist; sediment  1.80  6.57  9.09  18.15  35.24  35.53  52.90  15.56  Recent  29  Regosols  Fluvents, Orthents, Udepts  Schist; sediment  5.24  10.31  17.19  30.08  30.34  59.74  89.41  43.45  Gley  17  Gleysols  Aquepts  Sediment  2.79  5.61  17.06  33.94  68.22  76.43  118.93  33.07  Podzol  13  Podzols  Aquods, Orthods  Sediment  4.22  5.86  12.55  19.63  34.03  –  –  13.89  Basic  10  Mixed  Mixed  Basic igneous  5.81  12.97  17.30  149.18  350.62  –  –  85.07  Notes: Group: New Zealand soil type (brown, pallic, recent, gley and podzol) with a common acidic sedimentary lithology, or variable soil type with a common basic igneous lithology (basic); #: number of samples; WRB: world reference base soil classification; Min: minimum regular (whisker on boxplot); Q1: quartile one (bottom of box on boxplot); median (line through box on boxplot); Q3: quartile three (top of boxplot); Outlier (hollow circle on boxplot); Far: far outlier (hollow triangle on boxplot) and Mean (filled circle on boxplot). View Large The geology of New Zealand can be subdivided into relatively old basement rocks of the Austral Superprovince and relatively young cover rocks of the Zealandia Megasequence(Mortimer et al.2014; Fig. 1c). The former can be further subdivided into the Western and the Eastern Provinces, with soil samples used in this study collected above eight of the major basement terranes that comprise these provinces. The basement terranes are characterized by rocks of distinctive composition and age relating to the formation and evolution of Gondwana, and include a variety of continental-sourced quartzo-feldspathic metasedimentary rocks, calcalkaline volcanic rocks, island arc igneous and sedimentary rocks, ultramafic ophiolite rocks and I-, S- and A-type plutonic rocks (Heron 2014; Edbrooke et al.2015). The majority of sample locations in this study were above Eastern Province basement rocks, but one transect also sampled above a number of Western Province rocks (Fig. 1c). Rocks of the Caples and Rakaia terranes have been overprinted by, and form the protolith to, the Otago Schist—a regionally significant metamorphic belt (Fig. 1c). The Zealandia Megasequence is comprised largely of sediments derived by erosion of the basement rocks, and here has been subdivided into the PākihiSupergroup, equivalent to Quaternary aged rocks, and into other older cover, which includes four older sedimentary supergroups (Fig. 1c). This was deemed appropriate because of the similarity in results obtained for the other older cover. One other Zealandia Megasequence unit, the Dunedin Volcanic Group, has been highlighted (Fig. 1c), because its alkalic igneous chemistry and mineralogy is important to this study. The Dunedin Volcanic Group, including the Dunedin Volcano, is Oligocene to Miocene in age and comprised mainly of basanite lava flows, intrusions and pyroclastic deposits, with subordinate phonolite and trachyte rock compositions locally important (Benson 1968; Bishop & Turnbull 1996; Coombs et al.2008). Large catchments in the area, such as the Waiau and Oreti (Fig. 1c), drain southwards, eroding and redepositing sediment over distances of ≤100 km. The climate is cool temperate, with rainfall heaviest in the west (>4000 mm yr−1) and generally decreasing to the east, with the lowest rainfall being 250–500 mm yr−1. Of the ca. 220 000 people that live in the study area, 73 per cent live within two cities (Dunedin and Invercargill; Fig. 1c). The rest of the land area is sparsely populated and used for agriculture (beef, sheep, deer and dairy) or exotic forestry. METHODS Sample collection and preparation A regional geochemical baseline soil survey was undertaken in 2015 over southern New Zealand (Martin et al.2016, 2017; Rogers et al.2017; Fig. 1a). Soil samples were collected at 8 km intervals by hand auger at two depths. Topsoil samples were collected at 0–30 cm in the A-horizon and subsoil samples collected at 50–70 cm in the B-horizon, or rarely the C-horizon where soil cover was shallow (Rattenbury et al.2014). Samples from 323 sites (topsoil and subsoil) were analysed and form part of this study. Each sample was 5–10 kg in weight. A reference sample set is kept by GNS Science (the national geological survey organization) and forms the basis for this study. Samples were air-dried, sieved, split and the sub 2 mm portion analysed. Laboratory methods and procedures Handheld volume-specific magnetic susceptibility measurements (handheld magnetic susceptibility hereafter) were made using a commercially available Terraplus KT-10 unit on the dried and sieved samples in their open sample bags, with the device in contact with the soil (i.e. not through a sample bag). The sensitivity and the lower method detection limit of the volume magnetic susceptibility is 0.01 × 10−8 S.I. The handheld measurements were made at a GNS Science facility in Dunedin during a single session lasting a few hours, and measured both topsoil and subsoil (total of 323 localities). Laboratory mass-specific (cf. volume-specific) magnetic susceptibility measurements (laboratory magnetic susceptibility hereafter) were performed at the OtagoPalaeomagnetic Research Facility at the University of Otago, Dunedin, New Zealand using a single-frequency AGICO MFK-1A CS3 kappabridge. The MFK-1A measures samples in an applied field of between 2 and 700 A m−1 with a frequency of 976 Hz. Measurements were made on both topsoil and subsoil samples (total of 315 localities) at 200 A m−1. Eight localities did not have enough material available to complete a laboratory analysis. After anomalous magnetic susceptibility values had been determined (see Discussion section for an overview of how anomalous values were determined), an additional 21 sites were chosen for further study, these represented 17 anomalous sites and 4 sites from the Otago Schist for a background comparison. A strong geogenic control on magnetic susceptibility (again see Discussion) meant the additional 21 samples were chosen from the subsoil. Hysteresis and isothermal remanent magnetization (IRM) analyses on crushed soil samples of between 0.15 and 0.1 g were measured on a Princeton Measurements Corporation Vibrating Sample Magnetometer (MicroMag 2900) to determine the magnetic mineralogy. Hysteresis analyses were conducted in saturating fields of 500 mT at 4mT increments and IRM analyses were conducted to 1 T. Thermomagnetic measurements were made on an AGICO MFK-1A CSkappabridge on ca. 0.25 cm3 samples in air with an applied field of 200 A m−1. The Curie temperature was estimated using the 1/k method in the AGICO Cureval software to identify the presence of haematite. A precise Curie temperature determination was not attempted. RESULTS The full data are included in the online extra supplementary material (ESM1). The handheld magnetic susceptibility results range between 1.1 × 10−8 and 424 × 10−8 S.I. (median: 13.8 × 10−8 S.I.) for the topsoil and between 1.2 and 594 × 10−8 S.I. (median: 13.0 × 10−8 S.I.) for the subsoil. The laboratory data results range between 0.5 × 10−8 and 534 × 10−8 m3 kg−1 (median: 15.9 × 10−8 m3 kg−1) for topsoil samples and 0.2 × 10−8 and 640 × 10−8 m3 kg−1 (median: 12.4 × 10−8 m3 kg−1) for subsoil samples (Fig. 2). A histogram shows the distribution of magnetic susceptibility values in the topsoil (Fig. 3a) and bivariate plots show strong relationships between laboratory magnetic susceptibility values measured in the topsoil versus subsoil (Fig. 3b; r2 = 0.87) and between laboratory measured magnetic susceptibility values and handheld magnetic susceptibility values in topsoil (not shown) and subsoil (Fig. 3c; r2 = 0.91). Figure 2. View largeDownload slide The range and pattern of mass-specific magnetic susceptibility values showing (a) topsoil and (b) subsoil. The points are size-coded, with larger dots equating to higher magnetic susceptibility. The coloured map shows an inverse distance weighted interpolation between samples using ArcGIS software to create 500 m cell size grids using a distance power of 1.5 and a fixed 9 km search radius. The interpolation has been portrayed using a 10-class quantile colour ramp. Grid ticks are New Zealand Transverse Mercator 2000 (NZTM2000) eastings and northings. Figure 2. View largeDownload slide The range and pattern of mass-specific magnetic susceptibility values showing (a) topsoil and (b) subsoil. The points are size-coded, with larger dots equating to higher magnetic susceptibility. The coloured map shows an inverse distance weighted interpolation between samples using ArcGIS software to create 500 m cell size grids using a distance power of 1.5 and a fixed 9 km search radius. The interpolation has been portrayed using a 10-class quantile colour ramp. Grid ticks are New Zealand Transverse Mercator 2000 (NZTM2000) eastings and northings. Figure 3. View largeDownload slide The magnetic susceptibility data shown on various plots. (a) A histogram of topsoil data showing mass-specific magnetic susceptibility. The black circles indicate values where additional magnetic mineralogy data were collected from subsoil samples. (b) Topsoil versus subsoil plot of mass-specific magnetic susceptibility data. (c) Topsoil mass-specific magnetic susceptibility (laboratory method) versus topsoil volume-specific magnetic susceptibility (handheld method). The strong correlation (R2 = 0.91) allowed a linear regression to be calculated as shown on the diagram. Figure 3. View largeDownload slide The magnetic susceptibility data shown on various plots. (a) A histogram of topsoil data showing mass-specific magnetic susceptibility. The black circles indicate values where additional magnetic mineralogy data were collected from subsoil samples. (b) Topsoil versus subsoil plot of mass-specific magnetic susceptibility data. (c) Topsoil mass-specific magnetic susceptibility (laboratory method) versus topsoil volume-specific magnetic susceptibility (handheld method). The strong correlation (R2 = 0.91) allowed a linear regression to be calculated as shown on the diagram. The IRM analyses resulted in a coercivity of remanence (Hcr) of between 25 and 77 mT (median: 41 mT) and a saturation remanent magnetization (Mrs) of between 0.12 and 131 mAm2 kg−1 (median: 17 mAm2 kg−1). Hysteresis analyses revealed a coercivity (Hc) of between 2.3 and 19 mT (median: 7.9 mT), saturation magnetization (Ms) of between 0.85 and 938 mAm2 kg−1 (median: 251 mAm2 kg−1) and remanent magnetization (Mr) of between 0.10 and 134 mAm2 kg−1 (median: 20 mAm2 kg−1). Hysteresis loops are narrow waisted with all samples saturating at low fields (Figs 4a and b). Thermomagnetic data for sample GB00221 (Fig. 4c) contains no detectable heating curve inflection, which may indicate only minor quantities of magnetite are present. However, the sample undergoes thermochemical alteration with the production of magnetite, which is likely derived from oxidation or dewatering to clay minerals (e.g. Hirt & Gehring 1991). Sample GB0064B (Fig. 4d) experienced significant thermochemical alteration during heating with magnetite being the dominant magnetic mineral. Figure 4. View largeDownload slide Plots of hysteresis and thermomagnetic data. Hysteresis data for (a) GB00221 and (b) GB00064. Samples GB00221 has a mild pot-bellied morphology indicating minimal contribution of super paramagnetic magnetite and a dominance of single-domain grains (Tauxe et al.1996). Thermomagnetic data for (c) GB00221 and (d) GB00064. Sample GB00221 has very weak magnetic susceptibility with no discernible inflections during heating (black arrows). The cooling curves (grey arrows) indicate magnetite has formed possibly from dehydration/alteration of clay minerals. Figure 4. View largeDownload slide Plots of hysteresis and thermomagnetic data. Hysteresis data for (a) GB00221 and (b) GB00064. Samples GB00221 has a mild pot-bellied morphology indicating minimal contribution of super paramagnetic magnetite and a dominance of single-domain grains (Tauxe et al.1996). Thermomagnetic data for (c) GB00221 and (d) GB00064. Sample GB00221 has very weak magnetic susceptibility with no discernible inflections during heating (black arrows). The cooling curves (grey arrows) indicate magnetite has formed possibly from dehydration/alteration of clay minerals. DISCUSSION Influence of rock type, soil type and depth on magnetic susceptibility The laboratory magnetic susceptibility subdivided by major stratigraphic units is shown in Fig. 5(a). The median value of magnetic susceptibility is significantly elevated in Dunedin Volcanic Group rocks, relative to other stratigraphic units, and the median value in CaplesTerrane rocks is low relative to other stratigraphic units. There are relatively small differences between topsoil and subsoil values, and a number of stratigraphic units (Brook Street, Dun-Mountain Maitai, Dunedin Volcanic Group and Western Province), are represented by <10 samples. A more representative subdivision of the magnetic susceptibility data was by soil type and lithology. Figure 5. View largeDownload slide Plots showing variation in mass-specific magnetic susceptibility between topsoil and subsoil. (a) Tukey boxplot of data from different stratigraphic units. P: Pākihi Supergroup; ZM: Zealandia Megasequence; R: Rakaia Terrane; C: Caples Terrane; M: Murihiku Terrane; DM: Dun Mountain-Maitai Terrane; BS: Brook Street Terrane; DVG: Dunedin Volcanic Group; WP: Western Province rocks. (b) Tukey boxplot of data from five major soil types occurring over rocks with a common, acidic sedimentary lithology (brown, pallic, recent, gley and podzol) and a sixth group of mixed soils occurring over basic igneous rock types (basic). (c) Median magnetic susceptibility normalized to the median value in the subsoil. Only podzol soil shows a relative depletion. Diagram drawn after Hanesch & Scholger (2005). Figure 5. View largeDownload slide Plots showing variation in mass-specific magnetic susceptibility between topsoil and subsoil. (a) Tukey boxplot of data from different stratigraphic units. P: Pākihi Supergroup; ZM: Zealandia Megasequence; R: Rakaia Terrane; C: Caples Terrane; M: Murihiku Terrane; DM: Dun Mountain-Maitai Terrane; BS: Brook Street Terrane; DVG: Dunedin Volcanic Group; WP: Western Province rocks. (b) Tukey boxplot of data from five major soil types occurring over rocks with a common, acidic sedimentary lithology (brown, pallic, recent, gley and podzol) and a sixth group of mixed soils occurring over basic igneous rock types (basic). (c) Median magnetic susceptibility normalized to the median value in the subsoil. Only podzol soil shows a relative depletion. Diagram drawn after Hanesch & Scholger (2005). The five major soil types examined in this study are sampled over rocks that share a common lithology of acidic sedimentary rocks. The brown, pallic and recent soils have formed over either Otago Schist rock with a Mesozoic age, or Zealandia Megasequence sedimentary rocks with a Cenozoic age, with a skew in the age range towards younger, Quaternary aged rocks. The gley and podzol soil types included in this study form over the latter only (Cenozoic-aged sedimentary rocks). These five groups show the five different soil types formed over a common, acidic sedimentary lithology and account for 97 per cent of the data. The other 3 per cent are the sixth group of sites collected over basic igneous rock types. The mean value of magnetic susceptibility in samples collected over basic igneous rock types is nearly double that seen in the other five groups (Table 1 and Fig. 5b). Soils formed from basic igneous protoliths are typically associated with high magnetic susceptibility values (e.g. Fialová et al.2006). There are small differences in the mean, median and upper whisker values of topsoil samples versus subsoil samples for these groupings (Fig. 5b). In Fig. 5(c), the median magnetic susceptibility value has been normalized to that value obtained in the subsoil. In this way, the topsoil values for the recent soils are very weakly enriched (1.1), brown and pallic soils (the majority of sample points) are weakly enriched (1.3 and 1.4), gley soils and the basic group are moderately enriched (1.7 and 2.4) and podzol soils are weakly depleted, relative to the subsoil. These enrichments can occur naturally through the activity of iron-reducing bacteria, especially for well-drained, brown and pallic soils (Verosub & Roberts 1995; Maher 1998; Hanesch & Petersen 1999). In Podzol soils, transport and solution processes can lead to iron-oxide depletion and hence lower magnetic susceptibility values (e.g. Hanesch & Scholger 2002). Handheld versus laboratory techniques There is a strong, positive relationship between mass-specific susceptibility measured in the laboratory and volume-specific susceptibility measured by handheld techniques (Fig. 3c). A linear regression equation (Fig. 3c) calculated for this data (including outliers) means the handheld data can be converted into mass-specific values (10−8 m3 kg−1). This is important for studies where budget, time or laboratory access are issues, such as for exploration personnel, science students or other researchers where magnetic susceptibility is a secondary concern. The associated costs in personnel and laboratory times are an order of magnitude less using a handheld magnetic susceptibility meter. The advantages to the laboratory technique are higher precision. A range of additional data about remanence, mineralogy, etc., can also be determined in the laboratory, assuming the time and equipment are available. Using a handheld meter in the field is thus acceptable for studies of soil magnetic susceptibility where an appropriate regression can be calculated or studies where high precision is a lower priority (e.g. Hanesch & Scholger 2002; Schmidt et al.2005). Anomaly detection Three different approaches were used to construct maps of anomalous magnetic susceptibility values, namely the topsoil–subsoil difference method, the Tukey boxplot method and the geoaccumulation index method. The topsoil–subsoil difference method assumes that magnetic material accumulates in the topsoil from anthropogenic input, meaning an enrichment in the topsoil relative to the subsoil indicates an anthropogenic anomaly (e.g. Kapička et al.2001). This can be displayed graphically (Fig. 6a), and the enrichments subdivided (following Hanesch & Scholger 2002) into possible geogenic or pedogenic anomalies if the susceptibility difference is below −20 × 10−8 m3 kg−1, and possible anthropogenic anomalies if >20 × 10−8 m3 kg−1. Approximately 7 per cent of the data have a susceptibility difference >20 × 10−8 m3 kg−1 and thus a potential anthropogenic source per this method. It has been noted in central European studies that one limitation of this method is where high susceptibility values exist, for example, >200 × 10−8 m3 kg−1, a difference of >20 × 10−8 m3 kg−1 can also be a geogenic anomaly (Fialová et al.2006; Hanesch et al.2007). Before considering these data further, the other anomaly detection methods will be presented. Figure 6. View largeDownload slide Maps showing anomalous mass-specific magnetic susceptibility data determined by various methods. (a) Topsoil–subsoil difference method. Values <−20 and >20 are anomalous by this method. Some workers interpret the <−20 data (square symbols) as representing a geogenic source and > 20 data representing an anthropogenic source, though this simple interpretation is complicated when the soil parent material comes from high magnetic susceptibility source rocks. (b) Tukey boxplot method on topsoil data. The outlier data are shown as a point source map with a four-part, natural breaks (Jenks 1967) subdivision of the data value range. The four rock units shown for comparison are the Median Batholith, Murihiku Terrane, Dun Mountain-Maitai Terrane (DMMT) and Dunedin Volcanic Group (DVG). The locality of the Longwood Range (LR) and Takitimu Mountains (TM) are also shown. The sites where additional magnetic measurements were made on subsoil samples is indicated by a sexagon symbol (solid line for anomalous sites and dashed line for background sites). (c) Geoaccumulation Index (Igeo) method on topsoil data. The Waiau and Oreti & Aparima catchments are shown, as well as various locations discussed in the text (italics) and two sample sites (GB00064 and GB00071). Some major rivers are shown as blue lines in the catchment areas. Figure 6. View largeDownload slide Maps showing anomalous mass-specific magnetic susceptibility data determined by various methods. (a) Topsoil–subsoil difference method. Values <−20 and >20 are anomalous by this method. Some workers interpret the <−20 data (square symbols) as representing a geogenic source and > 20 data representing an anthropogenic source, though this simple interpretation is complicated when the soil parent material comes from high magnetic susceptibility source rocks. (b) Tukey boxplot method on topsoil data. The outlier data are shown as a point source map with a four-part, natural breaks (Jenks 1967) subdivision of the data value range. The four rock units shown for comparison are the Median Batholith, Murihiku Terrane, Dun Mountain-Maitai Terrane (DMMT) and Dunedin Volcanic Group (DVG). The locality of the Longwood Range (LR) and Takitimu Mountains (TM) are also shown. The sites where additional magnetic measurements were made on subsoil samples is indicated by a sexagon symbol (solid line for anomalous sites and dashed line for background sites). (c) Geoaccumulation Index (Igeo) method on topsoil data. The Waiau and Oreti & Aparima catchments are shown, as well as various locations discussed in the text (italics) and two sample sites (GB00064 and GB00071). Some major rivers are shown as blue lines in the catchment areas. One of the most useful methods to determine the background element concentration in soil is the Tukey boxplot (Tukey 1977; Reimann 2005). It divides the ordered values of the data into equal parts by finding the interquartile range between the 25th and 75th percentiles. The background concentration can be defined as 1.5 times the interquartile range (e.g. Hanesch et al.2007). The Tukey boxplot method has been applied using the soil-lithology subdivision outlined above (Table 1 and Fig. 5b). Outlier data are shown in Fig. 6(b), and subdivided based upon their concentration above background levels. The total number of anomalies by this method (n = 40) is slightly lower that the difference method used above (Fig. 6a) and the locality of the anomalies is different in some of the cases. One suggested approach to further refine the detection of anomalies is to calculate and plot the geoaccumulation index (Igeo; Müller 1979):   \begin{equation}{\rm{Igeo}} = {\rm{log}}2\left( {{\rm{Cn}}/1.5{\rm{Bn}}} \right)\end{equation} (1)where Cn is the value of the measured elemental concentration and Bn is the background value as determined by the Tukey boxplot method. The Igeo method attempts to compensate for natural background variation within groups. It reduces the total number of anomalies identified to 22 (Fig. 6c), but within similar areas of the survey to the Tukey boxplot method. The 22 anomalies identified by the Igeo method can be subdivided into four areas and two sites (Fig. 6c) within a common river catchment (Waiau catchment or Oreti and Aparima Catchment), terrane (Murihiku Terrane) or rock type (basic). Site GB00071 was collected in Milton township and site GB00064 was collected from farmland. These anomalous areas and sites are discussed further below. Neither the Tukey boxplot nor Igeo method distinguishes geogenic/pedogenic from anthropogenic anomalies. The Tukey boxplot method identifies all areas of anomalous results and is especially useful for groups of data numbering <50 (Reimann 2005). The Igeo method focuses on the upper end of the outlier data and suggests qualitative categories of contamination, for example, moderately to strongly contaminated for Igeo >2 (Müller 1979). Both methods appear useful for distinguishing anomalous magnetic susceptibility results from southern New Zealand. Nature of magnetic minerals Samples from anomalous sites identified by the Tukey boxplot and Igeo methods (Fig. 6b) were selected for further examination to understand the nature of magnetic mineralogy in them. Some samples from sites considered within normal background were also studied for comparison. The Hcr values are consistent with magnetite mineralogy where values can vary depending on grain size and composition (Fig. 7a; Day et al.1977) in agreement with thermomagnetic data (Figs 4c and d). Overall, the median coercivity of 41 mT indicates magnetite is the dominant remanence carrier at anomalous sites. A minor contribution from haematite may be present in some samples with higher Hcr values above 60 mT (Özdemir & Dunlop 2014). A day plot (Day et al.1977; Dunlop 2002a,b; Fig. 7b) indicates magnetic grains are mostly in the pseudo-single domain (PSD) and single domain (SD) grain size range. Anomalous sample sites from the Waiau catchment, Oreti and Aparima catchments and Murihiku Terrane partially overlap (Fig. 7b) at the centre of the PSD field with the exception of one Murihiku sample which falls in the multidomain (MD) field (labeled Longwood on Fig. 7b). The Dunedin Volcanic Group and Otago Schist rock samples plot slightly higher, with the Otago Schist samples lying at the PSD/SD-SP field transition, indicating a possible contribution of super paramagnetic (SP) grains from the Otago Schist or pedogenic processes above Otago Schist basement rock. All samples from the different groups show a degree of scatter, which is caused by the varying contribution of the soft component (SP or MD) in mixtures with SD grains (Dunlop 2002a). The two Dunedin Volcanic Group samples have different magnetic mineral concentrations (Mrs of 82 and 1.2 mAm2 kg−1) and coercivities (Hcr 40 and 25 mT) from one another, suggesting complexity in the magnetic mineralogy of the Dunedin Volcanic Group or that magnetic minerals have altered since erosion from their source. Figure 7. View largeDownload slide Subsoil magnetic data from 17 anomalous sites and four background sites. (a) An IRM Mrs versus Hcr plot indicating titanomagnetite is the main grain type. (b) A day plot indicating pseudo-single domain (PSD) magnetite is the dominant remanence carrier (Day et al.1977; Dunlop 2002a,b). Some analyses which approach the PSD-single domain (SD) mixing line may be biased by minor contributions of higher coercivity grains and a single sample from near the Longwood Range plots in the multidomain (MD) field and likely indicates a geogenic contribution of large titanomagnetite grains. (c) Magnetic susceptibility versus hysteresis and IRM determined magnetic concentration parameters (Ms and Mrs), which demonstrates a good correlation between magnetic mineral specific concentrations (Ms and Mrs) and magnetic susceptibility, which can be sensitive to contributions from paramagnetic and diamagnetic components in soil. Figure 7. View largeDownload slide Subsoil magnetic data from 17 anomalous sites and four background sites. (a) An IRM Mrs versus Hcr plot indicating titanomagnetite is the main grain type. (b) A day plot indicating pseudo-single domain (PSD) magnetite is the dominant remanence carrier (Day et al.1977; Dunlop 2002a,b). Some analyses which approach the PSD-single domain (SD) mixing line may be biased by minor contributions of higher coercivity grains and a single sample from near the Longwood Range plots in the multidomain (MD) field and likely indicates a geogenic contribution of large titanomagnetite grains. (c) Magnetic susceptibility versus hysteresis and IRM determined magnetic concentration parameters (Ms and Mrs), which demonstrates a good correlation between magnetic mineral specific concentrations (Ms and Mrs) and magnetic susceptibility, which can be sensitive to contributions from paramagnetic and diamagnetic components in soil. A biplot of magnetic susceptibility versus Ms and Mrs (Fig. 7c) reveals a close correlation between the two, indicating that magnetic susceptibility at most anomalous sites (e.g. Waiau and Oreti and Aparima catchments, Murihiku Terrane), is controlled by changes in mineral concentration rather than changes in mineral type, or in the proportions of ferromagnetic versus para/diamagnetic contributions. High coercivity minerals such as haematite, however, have saturation magnetization and magnetic susceptibility, which are orders of magnitude smaller than magnetite. Their relative contribution from sample to sample may therefore be underestimated from concentration-dependent measurements. Samples from the Waiau catchment have a higher average Hcr value of 61 which may indicate a greater quantity of high coercivity grains relative to other anomalous zones. Source of magnetic susceptibility anomalies Waiau catchment The Waiau catchment is an area of anomalous magnetic susceptibility identified by all three anomaly detection methods (Fig. 6). This area is sparsely populated, drained by the Waiau River and is bordered by Fiordland National Park. An obvious anthropogenic source for these anomalies is not present, but several, high magnetic susceptibility rock types occur within the catchment. For example, moderately to strongly magnetic rocks from the Median Batholith with I-type granites, gabbros and peridotites (Woodward & Hatherton 1975; Smale 1990) or Dun Mountain Ophiolite Belt ultramafic rocks (Hunt & Mumme 1978; Eccles et al.2005). Furthermore, the Longwood Range contains basic and ultrabasic rocks of the Longwood Suite that are known to contain titanomagnetite-bearing rock types (Price et al.2011). The day plot shows a site adjacent to the Longwood Range with particularly low Mrs/Ms and high Hcr/Hc relative to other sample sites in the Waiau catchment (Fig. 7b) indicating a dominance of large, MD grains that may be sourced from the Longwood Range (Ashley et al.2012). The Takitimu Mountains also contain likely titanomagnetite-bearing rocks (Houghton 1981) and anomalous soils sites adjacent to these mountains have formed mafic melanic and mafic brown soils and acidic allophanic brown soils. These soil types are the weathering products of igneous rocks and contain magnetite (Prasad & Ghildyal 1975), ferrihydrite and imogolite. Although a mineral that is uncommonly magnetic at ambient temperatures, ferrihydrite may carry a remanence (Johari Pannalal et al.2005). Soil samples from the Takitimu Mountains area also have higher coercivity values (Hc and Hcr) on average than samples from the rest of the studied area, which may indicate that the Takitimu Mountains are an unrecognized source of a high coercivity mineral such as haematite. Using the differences method of anomaly detection, two sites in the Waiau catchment suggested an anthropogenic source (triangle symbols in Fig. 6a), however, the strong magnetic susceptibility (>200 × 10−8 m3 kg−1) of probable source rocks in this catchment invalidates the criteria for assigning an anthropogenic source via this method. The preferred interpretation, then, is that there is a geogenic source to the magnetic susceptibility anomalies in the Waiau catchment. The provenance of magnetic minerals in soil in the catchment is suggested to be from igneous rocks of the Median Batholith, sourced from the upper part of the catchment in Fiordland National Park. There is likely local input from the Takitimu Mountains, and in the southern part of the catchment, input from the Longwood Suite in the Longwood Range may be the dominant source. Oreti and Aparima catchments Samples with anomalous magnetic susceptibility also occur along the flood plain of the Oreti and Aparima rivers (Fig. 6) above Late Pleistocene river gravels in brown or recent soil types. Six anomalous sites have been identified by the Igeo method (Fig. 7c). Again this is a sparsely populated region, dominated by dairy, sheep and beef farming, without an obvious anthropogenic source of contamination. The Oreti River does, however, drain the Dun Mountain Ophiolite Belt, and the Aparima River drains the Takitimu Mountains, all of which have high concentrations of magnetic minerals. It has been shown elsewhere that heavy minerals from the ophiolite belt have been eroded and redeposited along flood plains of the Oreti River (Martin et al.2016) and a similar explanation is envisioned to explain the Oreti and Aparima catchment anomalies. Namely, magnetite has been eroded and redeposited along the Oreti and Aparima river floodplains and captured during the study, a distance ≤100 km. Murihiku Terrane At least three anomalous sites have been identified in the Murihiku Terrane using each of the three methods of anomaly detection (Fig. 6). In absence of anthropogenic sources of contamination, a natural source is probable, which is consistent with the difference between the topsoil and subsoil being −20 × 10−8 mg3 kg−1 or less (square symbols Fig. 6a). The anomalous samples have formed adjacent to tuffaceous beds (e.g. Campbell et al.2003; Turnbull & Allibone 2003) and in magnetic, allophanic soils and brown soils. Some anomalous sites are formed in fragic pallic soils sourced from adjacent Dun Mountain-Maitai Terrane rocks, most probably Little Ben Sandstone rocks that were derived from a basaltic–andesitic volcanic arc. Little Ben Sandstone rocks are known to contain magnetite, haematite and pseudomorphs after haematite (Landi 1980). This suggests the provenance of magnetic minerals in anomalous sites in the Murihiku Terrane is proximal, certainly less than the terrane width (maximum 80 km), but most likely <1 km. Basic rocks The number of anomalies above Dunedin Volcanic Group rocks decreases from eight using the differences method, down to three using the Tukey boxplot method, to two using the Igeo method. The three anomalies above Dunedin Volcanic Group rocks (Fig. 7b) using the Tukey boxplot method need not invoke an anthropogenic source, and despite their relative proximity to Dunedin City, they occur in areas of farmland or forestry. The anomalous sites occur in mafic melanic soils and mafic brown soils, formed form igneous protoliths and, the alkalic igneous rocks that comprise Dunedin Volcanic Group are known to contain titanomagnetite (Wright 1967; Sherwood 1988), and are relatively magnetic (Woodward & Hatherton 1975). The provenance of magnetic minerals in these anomalous sites is proximal (<1 km). One sample site occurs over ultramafic to mafic rocks of the Dun Mountain Ophiolite Belt (Figs 6b and c) and also has a geogenic source. Anthropogenic sources Two anomalous sites (GB00064 and GB00071) are identified by all three anomaly methods (Fig. 7) but have no immediately obvious geogenic source. The topsoil–subsoil difference in magnetic susceptibility values at each site is consistent with anthropogenic contamination (Fig. 7a). At site GB00064, relatively magnetic Dunedin Volcanic Group rocks occur within 5 km of the sample area; however, this sample was also collected adjacent to a metalled road. In this region, aggregate from Dunedin Volcanic Group rocks is commonly used as roading aggregate. A probable explanation for the magnetic susceptibility anomaly for this location, then, is sample bias included road metal. Site GB00071 was collected in the township of Milton (population ca. 2000). The site is known to have relatively high concentrations of heavy metals, assigned to an anthropogenic source (Martin et al.2016). This suggests the magnetic susceptibility anomaly at this site is anthropogenic and proximal. SUMMARY AND CONCLUSIONS The volume-specific magnetic susceptibility (handheld method) and mass-specific magnetic susceptibility (laboratory method) of topsoil and subsoil were measured in a regional study of southern New Zealand. A strong correlation between the volume- and mass-specific methods suggests the faster and more economic handheld method could be justifiably used in future soil studies in New Zealand. A linear regression calculation on the data in this study could be used in the future to convert volume- to mass-specific magnetic susceptibility. Samples collected in this study can be divided into groups with shared soil type and/or lithology. All groups have similar median values of magnetic susceptibility, except for sites sampled over basic igneous rocks, which were significantly higher. Most groups were slightly too moderately enriched in the topsoil, relative to the subsoil, in a natural pedogenic process. Three different methods, the difference method, Tukey boxplot method and Igeo method, were used to identify sites with anomalous magnetic susceptibility in soil from southern New Zealand. The threshold for identifying anomalies is as follows: difference method < Tukey boxplot < Igeo, such that the difference method detects many more anomalies than the Igeo method, as has been found in other studies (Reimann 2005; Hanesch et al.2007). The magnetic mineralogy is dominated by magnetite, but certain areas around the Takitimu Mountains, Longwood Range and Dunedin Volcanic Group indicated variable magnetic mineralogy (e.g. minor contribution of haematite). In all except two anomalous sites, a geogenic source was the most probable explanation for magnetic susceptibility above natural background. This is consistent with the soil types at these anomalous sites that typically formed from igneous sources, the underlying bedrock geology that is a known source of magnetite or being in a catchment draining areas with rocks of known magnetic mineralogy. There is a minimal anthropogenic input detectable by magnetic susceptibility methods, and by proxy, a minimum of heavy metals, such as Cu, Pb and Zn, contaminating southern New Zealand soils. This is consistent with findings of low anthropogenic input of heavy metals in southern New Zealand, relative to elsewhere in New Zealand (Martin et al.2016). The provenance of magnetic mineralogy in anomalous sites is generally within the bounds of the rock type, catchment or terrane they were sampled within. This is a distance ≤100 km and frequently much closer (<1 km). In the Waiau catchment, the source of magnetic mineralogy is from Median Batholith rocks from within or near Fiordland National Park, with proximal sources from the Takitimu Mountains or Longwood Range becoming significant, or even dominant, in the southern areas of the catchment. In the Oreti and Aparima catchments, the dominant source of magnetic minerals is the Dun Mountain Ophiolite Belt or Takitimu Mountains. The study of soil magnetic susceptibility allows a more thorough understanding of soil distribution and is a powerful tool for sediment provenance studies. This will be important when trying to understand element or isotopic distribution, vectors to mineralization or variations in soil quality (for farming or for water quality). The low proportion of anomalous sites with anthropogenic contamination would be expected to increase when areas adjacent to larger urban centres on the North Island of New Zealand are studied. In future studies in New Zealand, it may be sufficient to study only the topsoil sample for magnetic susceptibility using a handheld method. This study should serve as a magnetic susceptibility reference library for similar soil types found in New Zealand, and guide future soil sampling studies. Acknowledgements This work was funded by the Government of New Zealand through a GNS Science strategic development fund. Phil Rieger, Amy Beatson and Chantelle Hillier assisted with sample collection and processing. We thank two anonymous reviewers and the editor for their constructive comments which substantially improved the manuscript. REFERENCES Ashley P., Craw D., Mackenzie D., Rombouts M., Reay A., 2012. Mafic and ultramafic rocks, and platinum mineralisation potential, in the Longwood Range, Southland, New Zealand, N. Z. J. Geol. Geophys. , 55( 1), 3– 19. https://doi.org/10.1080/00288306.2011.623302 Google Scholar CrossRef Search ADS   Benson W., 1968. Dunedin District. 1:50,000. NZ Geological Survey Miscellaneous Series Map 1 . Wellington, New Zealand, Department of Scientific and Industrial Research. Bian Y., Ouyang T., Zhu Z., Huang N., Wan H., Li M., 2014. Magnetic properties of agricultural soil in the Pearl River Delta, South China — Spatial distribution and influencing factor analysis, J. appl.Geophysics , 107, 36– 44. https://doi.org/10.1016/j.jappgeo.2014.05.003 Google Scholar CrossRef Search ADS   Bishop D., Turnbull I., 1996. Geology of the Dunedin Area: Scale 1:250 000, pp. Geological Map 21. 52 p. + 21 Folded Map . Institute of Geological & Nuclear Sciences, Lower Hutt. Blundell A., Dearing J.A., Boyle J.F., Hannam J.A., 2009. Controlling factors for the spatial variability of soil magnetic susceptibility across England and Wales, Earth-Sci. Rev. , 95( 3–4), 158– 188. https://doi.org/10.1016/j.earscirev.2009.05.001 Google Scholar CrossRef Search ADS   Campbell H.J., Mortimer N., Turnbull I.M., 2003. Murihiku Supergroup, New Zealand: redefined, J. R. S. N. Z. , 33( 1), 85– 95. https://doi.org/10.1080/03014223.2003.9517722 Google Scholar CrossRef Search ADS   Coombs D.S., Adams C.J., Roser B.P., Reay A., 2008. Geochronology and geochemistry of the Dunedin Volcanic Group, eastern Otago, New Zealand, N. Z. J.Geol. Geophys. , 51( 3), 195– 218. https://doi.org/10.1080/00288300809509860 Google Scholar CrossRef Search ADS   Cullers R.L., Basu A., Suttner L.J., 1988. Geochemical signature of provenance in sand-size material in soils and stream sediments near the Tobacco Root batholith, Montana, U.S.A., Chem. Geol. , 70( 4), 335– 348. https://doi.org/10.1016/0009-2541(88)90123-4 Google Scholar CrossRef Search ADS   Day R., Fuller M., Schmidt V.A., 1977. Hysteresis properties of titanomagnetites: grain-size and compositional dependence, Phys. Earth planet. Inter. , 13( 4), 260– 267. https://doi.org/10.1016/0031-9201(77)90108-X Google Scholar CrossRef Search ADS   Dearing J.A., Bird P.M., Dann R.J.L., Benjamin S.F., 1997. Secondary ferrimagnetic minerals in Welsh soils: a comparison of mineral magnetic detection methods and implications for mineral formation, Geophys. J. Int. , 130( 3), 727– 736. https://doi.org/10.1111/j.1365-246X.1997.tb01867.x Google Scholar CrossRef Search ADS   Dearing J.A., Hay K.L., Baban S.M.J., Huddleston A.S., Wellington E.M.H., Loveland P.J., 1996. Magnetic susceptibility of soil: an evaluation of conflicting theories using a national data set, Geophys. J. Int. , 127( 3), 728– 734. https://doi.org/10.1111/j.1365-246X.1996.tb04051.x Google Scholar CrossRef Search ADS   Drewry J.J., Littlejohn R.P., Paton R.J., 2000. A survey of soil physical properties on sheep and dairy farms in southern New Zealand, N. Z. J. Agric. Res. , 43( 2), 251– 258. https://doi.org/10.1080/00288233.2000.9513425 Google Scholar CrossRef Search ADS   Dunlop D., 2002a. Theory and application of the Day plot (Mrs/Ms versus Hcr/Hc) 1. Theoretical curves and tests using titanomagnetite data, J. geophys. Res. , 107, EPM 4-1–EPM 4-22. Dunlop D., 2002b. Theory and application of the Day plot (Mrs/Ms versus Hcr/Hc) 2. Application to data for rocks, sediments, and soils, J. geophys. Res. , 107, EPM 5-1–EPM 5-15. Eccles J.D., Cassidy J., Locke C.A., Sporli K.B., 2005. Aeromagnetic imaging of the Dun Mountain Ophiolite Belt in northern New Zealand: insight into the fine structure of a major SW Pacific terrane suture, J. geol. Soc. , 162( 4), 723– 735. https://doi.org/10.1144/0016-764904-060 Google Scholar CrossRef Search ADS   Edbrooke S.W., Heron D.W., Forsyth P.J., Jongens R., 2015. Geological map of New Zealand 1:1 000 000. GNS Science Geological Map 2. 2 sheetsGNS Science , Lower Hutt, New Zealand. Fialová H., Maier G., Petrovský E., Kapička A., Boyko T., Scholger R., 2006. Magnetic properties of soils from sites with different geological and environmental settings, J. appl. Geophys. , 59( 4), 273– 283. https://doi.org/10.1016/j.jappgeo.2005.10.006 Google Scholar CrossRef Search ADS   Fieldes M., 1955. Clay mineralogy of New Zealand soils, N. Z. J. Sci. Technol. , B37, 336– 350. Hanesch M., Petersen N., 1999. Magnetic properties of a recent parabrown-earth from Southern Germany, Earth planet. Sci. Lett. , 169( 1–2), 85– 97. https://doi.org/10.1016/S0012-821X(99)00076-X Google Scholar CrossRef Search ADS   Hanesch M., Rantitsch G., Hemetsberger S., Scholger R., 2007. Lithological and pedological influences on the magnetic susceptibility of soil: their consideration in magnetic pollution mapping, Sci. Total Environ. , 382( 2–3), 351– 363. https://doi.org/10.1016/j.scitotenv.2007.04.007 Google Scholar CrossRef Search ADS PubMed  Hanesch M., Scholger R., 2002. Mapping of heavy metal loadings in soils by means of magnetic susceptibility measurements, Environ. Geol. , 42( 8), 857– 870. https://doi.org/10.1007/s00254-002-0604-1 Google Scholar CrossRef Search ADS   Hanesch M., Scholger R., 2005. The influence of soil type on the magnetic susceptibility measured throughout soil profiles, Geophys. J. Int. , 161( 1), 50– 56. https://doi.org/10.1111/j.1365-246X.2005.02577.x Google Scholar CrossRef Search ADS   Hannam J.A., Dearing J.A., 2008. Mapping soil magnetic properties in Bosnia and Herzegovina for landmine clearance operations, Earth planet. Sci. Lett. , 274( 3–4), 285– 294. https://doi.org/10.1016/j.epsl.2008.05.006 Google Scholar CrossRef Search ADS   Haughton P.D.W., Todd S.P., Morton A.C., 1991. Sedimentary provenance studies, Geol. Soc. Lond. Spec. Publ. , 57( 1), 1– 11. https://doi.org/10.1144/GSL.SP.1991.057.01.01 Google Scholar CrossRef Search ADS   Heron D.W., 2014. Geological Map of New Zealand 1:250,000 GNS Science. GNS Science Geological Map 1 , Lower Hutt. Hewitt A.E., 2010. New Zealand Soil Classification , third edn, Manaaki Whenua Press, Lincoln, New Zealand. Hirt A.M., Gehring A.U., 1991. Thermal alteration of the magnetic mineralogy in ferruginous rocks, J. geophys. Res. , 96( B6), 9947– 9953. https://doi.org/10.1029/91JB00573 Google Scholar CrossRef Search ADS   Houghton B., 1981. Lithostratigraphy of the Takitimu Group, central Takitimu Mountains, western Southland, New Zealand, N. Z. J. Geol. Geophys. , 24, 333– 348. Hunt T., Mumme T., 1978. Geophysical Reconnaissance of the Red Mountain Ophiolite Complex, Southern Westland, Geophysics Division Report, 128 . Department of Scientific and Industrial Research, Wellington. Jenks G., 1967. The data model concept in statistical mapping, Inter. Yearbook Cartogr. , 7, 186– 190. Jenny H., 1941. Factors of soil formation , McGraw-Hill Book Company. https://doi.org/10.1097/00010694-194111000-00009 Joharipannalal S., Crowe S., Cioppa M., Symons D., Sturm A., Fowle D., 2005. Room-temperature magnetic properties of ferrihydrite: a potential magnetic remanence carrier?, Earth planet. Sci. Lett. , 236( 3–4), 856– 870. https://doi.org/10.1016/j.epsl.2005.05.019 Google Scholar CrossRef Search ADS   Jordanova N., Jordanova D., Petrov P., 2016. Soil magnetic properties in Bulgaria at a national scale—Challenges and benefits, Global Planet. Change , 137, 107– 122. https://doi.org/10.1016/j.gloplacha.2015.12.015 Google Scholar CrossRef Search ADS   Kapička A., Petrovský E., Jordanova N., Podrázský V., 2001. Magnetic parameters of forest top soils in Krkonoše mountains, Czech Republic, Phys. Chem. Earth A: Solid Earth Geod. , 26( 11–12), 917– 922. https://doi.org/10.1016/S1464-1895(01)00142-9 Google Scholar CrossRef Search ADS   Landis C.A., 1980. Little Ben Sandstone, Maitai Group (Permian): nature and extent in the Hollyford-Eglinton region, South Island, New Zealand, N. Z. J. Geol. Geophys. , 23( 5–6), 551– 567. https://doi.org/10.1080/00288306.1980.10424127 Google Scholar CrossRef Search ADS   Magiera T., Strzyszcz Z., Kapicka A., Petrovsky E., 2006. Discrimination of lithogenic and anthropogenic influences on topsoil magnetic susceptibility in Central Europe, Geoderma , 130( 3–4), 299– 311. https://doi.org/10.1016/j.geoderma.2005.02.002 Google Scholar CrossRef Search ADS   Maher B.A., 1998. Magnetic properties of modern soils and Quaternary loessic paleosols: paleoclimatic implications, Palaeogeog. Palaeoclimat. Palaeoecol. , 137( 1–2), 25– 54. https://doi.org/10.1016/S0031-0182(97)00103-X Google Scholar CrossRef Search ADS   Martin A.P., Turnbull R.E., Rattenbury M.S., Baisden W.T., Christie A.B., Cohen D.R., Hoogewerff J.A., Rogers K.M., 2015. Geochemical Atlas of Southern New Zealand, p. 221, GNS Science Report 2015/26. vii , Lower Hutt, New Zealand. Martin A.P., Turnbull R.E., Rattenbury M.S., Cohen D.R., Hoogewerff J., Rogers K.M., Baisden W.T., Christie A.B., 2016. The regional geochemical baseline soil survey of southern New Zealand: design and initial interpretation, J. Geochem. Explor. , 167, 70– 82. https://doi.org/10.1016/j.gexplo.2016.05.009 Google Scholar CrossRef Search ADS   Martin A.P., Turnbull R.E., Rissmann C.W., Rieger P., 2017. Heavy metal and metalloid concentrations in soils under pasture of southern New Zealand, Geoderm. Region. , 11( Suppl. C), 18– 27. https://doi.org/10.1016/j.geodrs.2017.08.005 Google Scholar CrossRef Search ADS   Molloy L.F., Christie Q.R., 1988. Soils in the New Zealand Landscape: The Living Mantle , Mallinson Rendel in Association with the New Zealand Society of Soil Science, Wellington. Mortimer N. et al.  , 2014. High-level stratigraphic scheme for New Zealand rocks, N. Z. J. Geol. Geophys. , 57, 402– 419. https://doi.org/10.1080/00288306.2014.946062 Google Scholar CrossRef Search ADS   Müller G., 1979. Schwermetalle in den sedimenten des rheins—veränderungen seit 1971, Umsch. Wiss. Tech. , 79, 778– 783. Nichol S.L., Augustinus P.C., Gregory M.R., Creese R., Horrocks M., 2000. Geomorphic and sedimentary evidence of human impact on the New Zealand coastal landscape, Phys. Geogr. , 21, 109– 132. Özdemir Ö., Dunlop D.J., 2014. Hysteresis and coercivity of hematite, J. geophys. Res. , 119, 2582– 2594.. https://doi.org/10.1002/2013JB010739 Google Scholar CrossRef Search ADS   Parfitt R., 1990. Allophane in New Zealand—a review, Aust. J. Soil Res. , 28, 343– 360. https://doi.org/10.1071/SR9900343 Google Scholar CrossRef Search ADS   Parfitt R.L., Russell M., Orbell G.E., 1983. Weathering sequence of soils from volcanic ash involving allophane and halloysite, New Zealand, Geoderma , 29, 41– 57. https://doi.org/10.1016/0016-7061(83)90029-0 Google Scholar CrossRef Search ADS   Pillans B., Wright I., 1990. 500,000-Year paleomagnetic record from New Zealand loess, Quat. Res. , 33, 178– 187. https://doi.org/10.1016/0033-5894(90)90017-F Google Scholar CrossRef Search ADS   Prasad B., Ghildyal B.P., 1975. Magnetic susceptibility of lateritic soils and clays, Soil Sci ., 120, 219– 229. https://doi.org/10.1097/00010694-197509000-00010 Google Scholar CrossRef Search ADS   Price R., Spandler C., Arculus R., Reay A., 2011. The Longwood Igneous Complex, Southland, New Zealand: a Permo-Jurassic, intra-oceanic, subduction-related, I-type batholithic complex, Lithos , 126, 1– 21. https://doi.org/10.1016/j.lithos.2011.04.006 Google Scholar CrossRef Search ADS   Rachwał M., Kardel K., Magiera T., Bens O., 2017. Application of magnetic susceptibility in assessment of heavy metal contamination of Saxonian soil (Germany) caused by industrial dust deposition, Geoderma , 295, 10– 21. https://doi.org/10.1016/j.geoderma.2017.02.007 Google Scholar CrossRef Search ADS   Rattenbury M.S., Martin A.P., Turnbull R.E., Christie A.B., 2014. Sampling methodology for a regional multi-element geochemical baseline survey. GNS Science Report 2014/62GNS Science, Lower Hutt, New Zealand. Reimann C., 2005. Sub-continental-scale geochemical mapping: sampling, quality control and data analysis issues, Geochem. Explor. Environ. Anal. , 5, 311– 323. https://doi.org/10.1144/1467-7873/03-065 Google Scholar CrossRef Search ADS   Rogers K.M., Turnbull R.E., Martin A.P., Baisden W.T., Rattenbury M.S., 2017. Stable isotopes reveal human influences on southern New Zealand soils, Appl. Geochem. , 82, 15– 24. https://doi.org/10.1016/j.apgeochem.2017.05.006 Google Scholar CrossRef Search ADS   Schmidt A., Yarnold R., Hill M., Ashmore M., 2005. Magnetic susceptibility as proxy for heavy metal pollution: a site study, J. Geochem. Explor. , 85, 109– 117. https://doi.org/10.1016/j.gexplo.2004.12.001 Google Scholar CrossRef Search ADS   Sherwood G.J., 1988. Rock magnetic studies of Miocene volcanics in eastern Otago and Banks Peninsula, New Zealand: comparison between Curie temperature and low temperature susceptibility behaviour, N. Z. J. Geol. Geophys. , 31, 225– 235. https://doi.org/10.1080/00288306.1988.10417771 Google Scholar CrossRef Search ADS   Smale D., 1990. Distribution and provenance of heavy minerals in the South Island: a review, N. Z. J. Geol. Geophys. , 33, 557– 571. https://doi.org/10.1080/00288306.1990.10421374 Google Scholar CrossRef Search ADS   Tauxe L., Mullender T.A.T., Pick T., 1996. Potbellies, wasp-waists, and superparamagnetism in magnetic hysteresis, J. geophys. Res. , 101, 571– 583. https://doi.org/10.1029/95JB03041 Google Scholar CrossRef Search ADS   Tukey J.W., 1977. Exploratory Data Analysis , Addison-Wesley. Turnbull I., Allibone A., 2003. Geology of the Murihiku Area, 1:250 000 Geological Map 20 . 1 sheet and 74 p. Institute of Geological & Nuclear Sciences Limited, Lower Hutt, New Zealand. Turner G.M., 1997. Environmental magnetism and magnetic correlation of high resolution lake sediment records from Northern Hawke's Bay, New Zealand, N. Z. J. Geol. Geophys. , 40, 287– 298. https://doi.org/10.1080/00288306.1997.9514761 Google Scholar CrossRef Search ADS   Verosub K.L., Roberts A.P., 1995. Environmental magnetism: past, present, and future, J. geophys. Res. , 100, 2175– 2192. https://doi.org/10.1029/94JB02713 Google Scholar CrossRef Search ADS   Woodward D.J., Hatherton T., 1975. Magnetic anomalies over southern New Zealand, N. Z. J. Geol. Geophys. , 18, 65– 82. https://doi.org/10.1080/00288306.1975.10426347 Google Scholar CrossRef Search ADS   Wright J.B., 1967. The iron-titanium oxides of some Dunedin (New Zealand) Lavas, in relation to their palaeomagnetic and thermomagnetic character (with an appendix on associated Chromiferous Spinel), Mineral. Mag. , 36, 425– 435. https://doi.org/10.1180/minmag.1967.036.279.13 Google Scholar CrossRef Search ADS   SUPPORTING INFORMATION Supplementary data are available at GJI online. aMartin_etal_SoilMagSus_SupplementaryMaterials_R2.pdf Please note: Oxford University Press is not responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the paper. © The Author(s) 2017. Published by Oxford University Press on behalf of The Royal Astronomical Society.

Journal

Geophysical Journal InternationalOxford University Press

Published: Feb 1, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off