Earth-Science Reviews 155 (2016) 28–48 Contents lists available at ScienceDirect Earth-Science Reviews journal homepage: www.elsevier.com/locate/earscirev Magnetic mineral assemblages in soils and paleosols as the basis for paleoprecipitation proxies: A review of magnetic methods and challenges Daniel P. Maxbauer a,b,⁎, Joshua M. Feinberg a,b, David L. Fox a a Department of Earth Sciences, University of Minnesota, Minneapolis, MN, United States b Institute for Rock Magnetism, University of Minnesota, Minneapolis, MN, United States article info abstract Article history: Magnetic iron oxide minerals, principally magnetite, maghemite, hematite, and goethite are formed in well- Received 10 August 2015 drained soils in response to a suite of physical, chemical, and biological factors. Despite a wide range of complex- Received in revised form 11 January 2016 ity in the pedogenic processes that lead to magnetic mineral formation, dissolution, and transformation, there are Accepted 26 January 2016 well-documented empirical relationships between various magnetic mineral assemblages in soils with environ- Available online 27 January 2016 mental and climatic conditions. Recently there has been an increase in the number of quantitative magnetic Keywords: paleoprecipitation proxies that have been developed, and there is great potential for magnetic methods to be Paleoprecipitation used in the geologic record to develop reconstructions of past climates. Magnetic paleoprecipitation proxies Proxy have been widely utilized in Quaternary or younger loess–paleosol systems; however, they have yet to be utilized Soil magnetism in the pre-Quaternary fossil record. Future studies of magnetic mineralogy of soils and paleosols should aim to Paleosols explore non-loessic modern soils and pre-Quaternary paleosols with more focus on understanding the interac- Iron oxides tion between magnetic mineral assemblages and soil moisture. Applications of existing and novel magnetic Environmental magnetism paleoprecipitation proxies in the fossil record should prove to be a valuable resource for paleoclimatologists. © 2016 Elsevier B.V. All rights reserved. Contents 1. Introduction...............................................................29 2. Majorironoxidesinsoil..........................................................29 2.1. Magnetiteandmaghemite......................................................29 2.2. Hematiteandgoethite.......................................................30 2.3. Ferrihydrite............................................................31 3. Formationofironoxidesinsoils......................................................31 3.1. Ironoxideformationmodelsdescribingmagneticenhancement.....................................31 3.1.1. Redoxoscillationsandthefermentationmechanism......................................31 3.1.2. Agingpathwayofferrihydritetohematite..........................................32 3.2. Goethiteandhematiteformationanddistribution...........................................32 4. Characterizingironoxidemineralassemblages................................................33 4.1. Frequencydependenceofsusceptibility................................................33 4.2. HIRM...............................................................33 4.3. Unmixingmagneticmineralcomponents...............................................34 4.4. Determinationofgoethiteandhematiteconcentrations........................................35 5. Magneticproxiesforprecipitation.....................................................35 5.1. Relationshipsbetweenmagneticenhancementandprecipitationinloessicsoils.............................35 5.2. Relationshipsbetweenprecipitationandabundancesofgoethiteandhematite..............................37 5.3. Recognizingerrorinmagneticpaleoprecipitationproxies........................................38 6. Physical,chemical,andbiologicalcomplications...............................................38 6.1. Physical..............................................................38 ⁎ Corresponding author at: 310 Pillsbury Drive SE, Minneapolis, MN 55455, United States. E-mail address: [email protected] (D.P. Maxbauer). http://dx.doi.org/10.1016/j.earscirev.2016.01.014 0012-8252/© 2016 Elsevier B.V. All rights reserved. D.P. Maxbauer et al. / Earth-Science Reviews 155 (2016) 28–48 29 6.2. Chemical............................................................. 39 6.3. Biological............................................................. 39 7. Diageneticconcerns............................................................ 40 8. Challengesforfuturework......................................................... 41 Acknowledgments............................................................... 41 AppendixA. Primeronmineralandenvironmentalmagnetism......................................... 41 A.1. Magneticsusceptibility................................................... 41 A.2. Magneticgrainsize.....................................................42 41 A.3. Magneticremanenceandhysteresis.............................................42 41 A.4. S-RatioandL-Ratio.....................................................45 41 A.5. Someusefulhighandlowtemperaturemeasurements....................................45 41 References..................................................................45 41 1. Introduction In the first part of this review we provide an overview of the major iron oxide minerals found in soils (Section 2). This is followed by a dis- Magnetism in well-drained soil is controlled by the abundance, grain cussion of the pedogenic processes that lead to the formation and trans- size, and chemical composition of various iron oxide and oxyhydroxide formation of magnetic iron oxides in soils (Section 3). We then discuss minerals (hereafter referred to simply as ‘oxides’). In soils, the most the relevant magnetic methods used to identify and quantify the abun- abundant (by volume) iron oxides are goethite (α-FeOOH) and hema- dance of magnetic minerals in soils (Section 4). The available magnetic tite (α-Fe2O3), which are antiferromagnetic and produce weak perma- paleoprecipitation proxies are reviewed in Section 5.InSection 6 we ad- nent magnetizations. Magnetite (Fe3O4) and maghemite (γ-Fe2O3), dress natural mechanisms that complicate and limit the applicability of both ferrimagnetic with strong magnetizations, are far less abundant different magnetic paleoprecipitation proxies. Further, we explore the in soils but tend to dominate bulk magnetic properties. Magnetic min- potential pathways for iron oxide mineral destruction or transformation erals form in soil in response to a suite of complex pedogenic processes due to diagenetic processes that occur during the transition from soil to that are sensitive to physical, chemical, and biological conditions. De- paleosol (Section 7). We conclude this review with a number of chal- spite these complexities, empirical relationships between soil iron ox- lenges and research themes that we hope will guide future research ides and climate have been observed for decades (e.g., Kampf and (Section 8). Schwertmann, 1983) and environmental magnetic studies of soils and We direct readers that are relatively new to the field of environmen- sediments routinely make qualitative climatic interpretations (see re- tal magnetism to the appendix (Appendix A) where we include a brief views by Maher, 1998, 2007, 2011; Liu et al., 2012). primer on many common magnetic properties. This review draws Quantitative reconstructions of past environmental conditions, such from a broad body of previously published work. For further details on as mean annual precipitation and temperature, are of fundamental in- specific topics, readers are referred to the following resources: for a terest to paleoclimatologists. For example, methods to reconstruct full review of iron oxide minerals see Cornell and Schwertmann paleoprecipitation in pre-Quaternary terrestrial systems (N2.6 Ma) (2003); for previous reviews on magnetism in soils see Mullins have been developed using leaf physiognomic approaches (Peppe (1977) and Maher (1998); for more encompassing reviews of environ- et al., 2011; Royer, 2012), bulk geochemical weathering indices of mental magnetism in general see Thompson and Oldfield (1986), Evans paleosols (Sheldon et al., 2002; Sheldon and Tabor, 2009), the depth and Heller (2003), Maher (2007), Maher (2011),andLiu et al. (2012). to the carbonate horizon of paleosols (Retallack, 2005), and the eco- physiology of mammalian fauna (e.g., Eronen et al., 2010a; Eronen 2. Major iron oxides in soil et al., 2010b). A growing number of studies have proposed methods to link magnetic minerals within a soil quantitatively to the mean annual We describe here the magnetic minerals that display correlation be- precipitation (MAP) under which the soil developed (e.g., Maher and tween mineral abundance and precipitation (i.e., goethite, hematite, Thompson, 1995; Geiss et al., 2008; Balsam et al., 2011; Orgeira et al., magnetite, maghemite). In addition, we have included some related in- 2011; Hyland et al., 2015). Historically this work has focused on magne- formation about ferrihydrite because it is a common soil constituent and tite/maghemite variations in loess-derived soils developed under a lim- often is involved as a precursor phase in pedogenic processes that lead − ited range of MAP (~200–1000 mm yr 1), although some recent studies to the formation of the more stable magnetic iron oxides. Lepidocrite have expanded their scope to