Measuring soil organic carbon: which technique and where to from here?
Organic carbon (OC) in soils is contained in materials that have been derived from a variety of biological sources. Much stored OC will have come from plants (particularly roots) that have grown in the soil (Baldock 2007). Animal manures and remains are often part of the cycling of OC. Soil invertebrates such as ants, springtails, worms, termites, spiders, slaters and dung beetles (to name a few) also play a role in storing and cycling OC (Jonas et al. 2007; Velasquez et al. 2007; Ngo et al. 2012; Blouin et al. 2013). A further fraction will have been derived from a range of other soil organisms and their exudates and life stages (Kogel-Knabner 2002; Artursson et al. 2006; Kramer and Gleixner 2008). Anthropogenic activities (particularly fire) in different parts of the world have also added charred residues to the soil that range from elemental carbon structures to aromatic, aliphatic and carbohydrate compounds (Calvelo Pereira et al. 2014). These products may have also been added to soils via natural phenomena such as fires caused by lightning strike. Biomass-derived OC is generally liable to composting or humification processes (degradation and subsequent stabilisation), further increasing the range of organic materials in the soil matrix. Materials containing OC are also involved in aggregation processes, which can make them physically and/or chemically protected (Tisdall and Oades 1979, 1982). Confounding many methods in the analysis of total soil OC (TOC) is the presence of inorganic carbon compounds, which are generally carbonate minerals that may be unable to be separated easily from the soil matrix, and may even be bound to soluble OC within the soil (Schmidt et al. 2012). This gives a background of a wide variety of different organic chemical structures existing in the soil, with different levels of protection or chemical attraction, which goes part of the way to explaining why finding a single technique to analyse TOC that works in all cases has proven difficult.
Since the advent of TOC analysis there has been the drive to have methods that were more rapid (generally as the first consideration) and less expensive. When looking at the broad scale, OC is not evenly distributed over large areas, depths or soil types (Singh et al. 2012). This level of uncertainty increases the need for data that are accurate and relatively easy to obtain. The urgency of these requirements has only intensified with time as the need to accurately assess carbon stocks on a global scale has similarly intensified (Metz et al. 2007). Interest in ways to accurately measure TOC are part of an overall need to assess the global carbon cycle and stocks. Climate change scenarios and mitigation efforts will require improved understandings, with more data and better accuracy to allow appropriate measurement of what is occurring to carbon, and in what form it is stored or emitted (Lai et al. 2007). The world's oceans (including the ocean floor sediments) are the most significant sink of global carbon, but the soils of the world also contain a significant part of the terrestrial carbon at three- to fourfold the amount held in vegetation (and double of what is in the atmosphere; Ciais et al. 2013).
Sampling or not?
Clearly, a technique that does not require sample removal, is relatively low cost and can interrogate large areas with accuracy would be ideal in most circumstances. No current technique can yet claim to have fully achieved this ideal, but various techniques are moving towards this goal. The current standard, or at least the most common approach, remains removal of samples, with some pretreatment and laboratory-based analysis. However, the merits of taking physical samples of soil for measurement (either on-site or in a laboratory) versus techniques that do not remove the soil are not for specific discussion here. The present discussion focuses on the different measurement techniques, ranging from traditional sampling followed by laboratory-based analysis through to in situ or on-the-go analyses and remote sensing techniques. Samples removed from site and prepared in laboratories for TOC measurement generally undergo at least drying (air dried or at 105[degrees]C) and grinding to pass through a 2-mm sieve in order to enable representative subsampling. Some soils and techniques require finer grinding.
What is included in TOC determinations?
Carbonates and bicarbonates are considered to be inorganic components and should not be included in TOC analyses. Methods that include carbonates and TOC would be defined as total carbon (TC) analyses. These techniques are more prevalent and are often adapted to pretreat samples for inorganic carbon when analysing for TOC (Rayment and Higginson 1992). The biomass (living entities and their exudates) is generally included. For some researchers, living or not fully decomposed plant and animal tissue and the soil biomass have been excluded (MacCarthy et al. 1990). Humic substances (large and small organic molecules produced as a result of the decomposition and stabilisation of biomass) are considered to be integral and there appears no debate about their inclusion (MacCarthy et al. 1990). Some early methods sought to analyse soils specifically for humates as the organic proportion of soil (Degtjareff 1930). There has been debate about whether charcoal and charred materials should be excluded (Oades 1988). The argument for exclusion of charred materials could be on the basis that the material can be largely inert and does not participate in the carbon cycle within the soil to any great extent (Skjemstad et al. 1999). The opposing view would consider that the origin (parent matter) of the materials is organic, and hence chars and related materials should be included. Further, the chemical structures of these materials are not just stabilised elemental carbon structures, and composition depends greatly on the conditions under which the charring occurred (Calvelo Pereira et al. 2014). In fire-prone or intentionally human-modified soils, charred materials may constitute a significant proportion of the carbon present (Chia et al. 2010; Barrow 2012; Calvelo Pereira et al. 2014).
Conceptual models such as the Rothamsted Carbon Model (RothC; emanating from the Rothamsted Experimental site in the UK) describe the turnover of carbon in terms of decay (and resistance to decay) of organic inputs added on a monthly basis (Jenkinson 1990). RothC describes inputs (crop residues) that will have a proportion that is decomposable as well as a proportion resistant to fast decomposition, although both are assumed to decay to or become part of the same products, namely one of C[0.sub.2] (mineralisation), microbial biomass or humus (Jenkinson 1990). Aside from this, the soil is assumed to have a proportion of organic matter that is resistant to decomposition (i.e. inert organic matter (IOM)). The process of inputs decaying into specific products gives a possible analytical measurement direction. Models such as RothC have been developed using longitudinal data. The advantage of being able to directly measure the pools in the model would be the ability to start from any point in time (by simply measuring the amounts in each pool at that point in time) and being able to predict the future fate of OC (Skjemstad et al. 2004). Thus, this type of model gives a grouping of the types of constituents of TOC and assists in the understanding of what is present and how these pools interact in the soil environment.
Some researchers have attempted to link models to a method of direct measurement (Skjemstad et al. 2004; Zimmermann et al. 2007; Baldock et al. 2013b). In Australian soils, where charcoal contents can be a significant component of TOC, an approach using size fractionation as the basis has been used (Skjemstad et al. 2004). The [less than or equal to] 2-mm soil sample is divided into fractions consisting of 2000-53 [micro]m and <53 [micro]m, with an assumption that the charcoal and humic content resides only within the <53 [micro]m fraction. Skjemstad et al. (1999) used photooxidation followed by [sup.13]C nuclear magnetic resonance (NMR) to separate the charcoal from non-charcoal content of the <53 [micro]m fraction. This allowed measurement of the charcoal content as the IOM component of the model. When used along with measurement of the TOC and particulate OC (in the 2000-53 [micro]m fraction), the humic component was calculated as the TOC subtracting both the particulate OC and the IOM components. This approach enables the use of RothC for sites with no measurement history to have inputs modelled with some accuracy. The measurements undertaken to achieve this outcome are quite time consuming, but it nevertheless allows modelling to be conducted at these new sites. More recently, a similar but more extensive technique (analysing both 2000-50 and [less than or equal to] 50 [micro]m fractions for charcoal content) was used to allocate Australian soils to the RothC pools (Baldock et al. 2013b). Significantly, the authors found noteworthy amounts of charcoal content in the 2000-50 [micro]m fraction (Baldock et al. 2013b), unlike the earlier work (Skjemstad et al. 2004).
Although leading to a better understanding of the constituent OC components, in most cases TOC is sufficient because of the effort required to extract the finer detail involved with these procedures. The pragmatic issue of being able to analyse various portions of the OC while excluding others is clearly time consuming. Tables 1 and 2 provide method comparisons, including the TOC portion measured, calibration data and several opportunities for further research from methods discussed herein.
The reference method for TOC has been dry combustion, which measures all the forms of OC mentioned above with pretreatment necessary to exclude carbonates. For the remainder of this discussion, the term 'TOC' will include all carbon-containing materials in the soil, with the exception of inorganic carbon. In a practical sense, the measurements generally used to determine TOC do not discriminate against particular groups of compounds, wholly or with certainty in all cases. It is therefore important to be aware of the various factors that may affect analysis of particular proportions of OC compounds in each case. There is no one-size-fits-all approach; however, the different analyses discussed below increase our understanding of what each approach includes and what the results of these analyses may mean.
Oxidation techniques have been, and continue to be, the mainstay of measurement of carbon in soils. The dry combustion method for carbon was pioneered by Justus von Liebig and released in 1831, and although this method has received multiple improvements, dry combustion methods have remained as the standard or reference method ever since (Kosaka et al. 1959; Brock 2013). Wet acidic oxidations for determining TOC were appearing in the literature in the late 1800s and early 1900s (Ames and Gaither 1914). Oxidation methods are, by nature, indirect measurement of the carbon in soils because they entail conversion of the carbon in its various forms (in the soil) to carbon dioxide, which is then measured directly or indirectly. Both wet and dry oxidation account for the majority of soil carbon analyses still performed today. These methods have their limitations and interferences that must be dealt with appropriately to be as accurate as possible for the measurement of TOC. Many discoveries have been made in the intervening years to improve the accuracy of each and these are discussed in greater detail below. As with most developments, the thrust of the advances has not just been for the purpose of discovering and eliminating accuracy issues and interferences, but to increase the speed of analysis and decrease cost.
Dry oxidation methods
The German chemist Justus von Liebig is described by some as a father of agricultural chemistry and the chemical fertiliser industry because of his work on the plant nutrient value of nitrogen (Brock 2013). It was von Liebig's method of dry combustion using a five-bulb device called the Kaliapparat (or potash bulb) released in 1831 that became the reference method for TOC analysis for decades (Brock 2013). The Liebig method involved burning the organic sample with copper oxide (catalyst), the products of which were then passed through calcium chloride (to trap water) and a caustic solution (to trap carbon dioxide), with the amount of carbon dioxide captured determined gravimetrically (Brock 2013). The (Liebig) method markedly improved speed, allowing analysis of seven samples per day, whereas previously this number would have taken a week (Brock 2013). There were improvements to the method such as that developed by Dennstedt, whereby a double stream of oxygen was passed by the sample being heated (as described in Walkley 1935). Combustion products and interferences, including oxides of nitrogen and halogens, were removed using lead peroxide and water with calcium chloride, and the carbon dioxide was trapped using a two-level capture of soda lime (Walkley 1935). These methods used either burners or electrical heating of the sample bulb.
The modern dry combustion methods generally involve a furnace (either resistance or reduction) run at temperatures exceeding 1000[degrees]C using purified oxygen and perhaps other admixtures as catalysts in order to convert all carbon present into carbon dioxide. Heating organic materials leads to evolution of carbon monoxide and carbon dioxide and the ratios are highly temperature dependent, as described by the Boudouard reaction (Skjemstad and Baldock 2008; Lahijani et al. 2015). The production of carbon monoxide can be an issue that will lead to underestimation of TOC, as can the production of charcoal, which may not fully oxidise if the combustion does not continue for long enough (Skjemstad and Baldock 2008). Dry combustion methods do not discriminate for TOC because the carbonates in the sample are also decomposed at elevated temperatures (between 400[degrees]C and 1000[degrees]C) and produce carbon dioxide that will be measured as part of the analysis (Rayment and Higginson 1992; Skjemstad and Baldock 2008). So, unless some sample pretreatment occurs, these methods determine TC. Of course, not all soils contain significant levels of carbonate, so if the sample is of a soil that is essentially carbonate free, the TC will equal the TOC. Testing for the presence of carbonates is usually a simple fizz test, whereby drops of 1 M hydrochloric acid are placed on a separate finely ground portion of the soil. The presence of a fizz (evolution of carbon dioxide) denotes carbonates in the soil; conversely, lack of fizz denotes no significant carbonates are present. An issue with analysing TOC in calcareous soils using combustion methods is that some of the OC may be attached to carbonates, so pretreatments using an acid to remove carbonates can lead to the loss of OC (e.g. hydrochloric or hydrofluoric acid) or not removing all inorganic carbon (e.g. sulfurous acid) before analysis depending on the acid used (Heron et al. 1997; Schmidt et al. 2012).
In the 1960s, attempts at improved recovery of TOC using wet combustion methods were still being developed because the dry combustion methods were too slow (Allison 1960; Mebius 1960). By the early 1970s, some improvements had been made in terms of the speed of dry combustion by automation and increased temperatures (e.g. LECO Carbon Analyser St Joseph, MI, USA; Tabatabai and Bremner 1970). Using the new instrumentation, 0.2-0.3 g soil (along with ~1 g each of iron chips, tin and tin-coated copper as catalysts and to increase reaction temperature) were heated to 1650[degrees]C in an induction furnace with a stream of purified oxygen. The gases produced and the oxygen carrier were passed through various traps to remove non-carbon oxides, chlorides and water and through a heating tube to increase carbon dioxide conversion (Tabatabai and Bremner 1970). The gas stream was then collected in a cylinder at 45[degrees]C and carbon dioxide content measured using thermal conductivity with a thermistor-type cell in combination with an amplifier for the signal. This tipped the balance in favour of dry combustion for carbon analysis in soils (certainly in terms of time, although perhaps not in terms of cost), meaning around 150 samples or more could be analysed by a single operator in one day (Tabatabai and Bremner 1970).
The modern take on the dry combustion method for carbon is a highly automated instrument that often analyses for other elements (e.g. nitrogen, sulfur and sometimes hydrogen and oxygen) at the same time, which can increase the usefulness of this kind of analysis (LECO 2013; Elementar 2014). The soil samples are introduced into the furnace, usually at a temperature between 950[degrees]C and 1400[degrees]C, and sometimes a tin container is used to assist combustion at higher temperatures (~1800[degrees]C) than the furnace temperature (Elementar 2014; Skalar 2014). Where multiple analytes are required, a gas chromatographic separator is generally used and detection is often via a thermal conductivity detector, infrared or sometimes mass spectrometer (LECO 2013; Elementar 2014; Skalar 2014). Where gas chromatography is not used for separation, a series of traps is used to separate interfering gases. Some modern instruments can also use sample mass in the order of 1-2 g instead of earlier models, which may use one-tenth or even less, reducing issues of the lack of representativeness of earlier models (Tam and Yao 1998).
Dry combustion using modern instruments as described above is reliable and represents the current standard for TC analysis. Sample presentation is straightforward and the instruments do not use harmful chemicals. The main disadvantages would be the high initial instrument cost, the cost of consumables (e.g. purified gases and tin or nickel boats, if required), the energy use of running a reaction chamber at elevated temperatures and pretreatments required when inorganic carbon is present. Some instruments offer automated treatment of samples (acidification) for determination of TOC by difference (Skalar 2014). Where inorganic carbon is not present or at very low levels, it would be the preferred current method for measurement of TOC.
Loss-on-ignition (LOI) methods are also based on the oxidation of the organic matter present in the soils. They differ from the combustion method in that rather than trapping and/or measuring the carbon dioxide evolved upon heating, they rely on gravimetric analysis that measures the sample before and after heating, determining the loss of the organic matter. There are LOI methods approximating TOC levels without a direct calibration with a standard method. These are temperature based only (e.g. mass loss between 105[degrees]C and 550[degrees]C usually expressed as a percentage of the oven dry temperature) and should be regarded as indicative only (Matthiessen et al. 2005; Rayment and Lyons 2011). The LOI method for TOC is more directly related when a calibration is constructed using a standard method (e.g. dry combustion) for measurement of TOC compared with the lost mass for set conditions of temperature and time of ashing. There are many possible contributions to mass loss from soil samples ranging between 30[degrees]C and 1000[degrees]C, not just OC. The lost mass will be inherent water from various minerals, oxides of nitrogen, sulfur and carbon from organic matter and other possible sources depending on the nature of the soil (Schulte and Hopkins 1996). Caution in using this method would be advised when samples contain significant possible interferences (e.g. gypsum, Epsom salts, calcium chloride and clays) that lose structural water at temperatures in the range of interest, especially when these interferences are not consistently distributed in the analysis set of samples (Schulte and Hopkins 1996).
The selection of an exact, one-size-fits-all temperature range for TOC analysis via LOI is not well defined in the literature and there are examples ranging from 360[degrees]C to 850[degrees]C, with the other major variables to be controlled being sample size, the time allowed for combustion and the temperature and time of drying (Ball 1964; Ben-Dor and Banin 1989; Schulte et al. 1991 ; Bisutti et al. 2004). Temperatures of drying and combusting will best be selected by understanding the relevant interferences in the samples of interest, which may mean increased levels of initial assaying. Given these limitations, there are successful examples of good calibrations in the literature. Generally, the conditions where the best calibrations are achieved relate to circumstances where the appropriate temperature range is chosen to account for the known likely interferences and the soils are relatively consistent in structure, geographical location and chemical composition. For example, Konen et al. (2002) in the north central US successfully constructed calibrations using dry combustion against LOI (360[degrees]C for 2h with 105[degrees]C pre-drying overnight) when using soil samples from the same geographical areas ([r.sup.2] = 0.94-0.98). However, a single calibration constructed using the entire dataset from five different regions combined gave an inferior calibration, highlighting that this method likely works best when using soils with similar properties (i.e. same type and location; Konen et al. 2002). It should be noted that this approach at 360[degrees]C will not fully combust the organic material present in the sample, which may take 24 h; however, the majority (>80%) is likely to be lost in the first 2 h (Ben-Dor and Banin 1989).
LOI determinations do not suffer from the specific measurement interferences for dry combustion methods and carbon dioxide detection because a specific species is not being measured. Rather, the amount of organic material lost from the sample under the specified conditions is related to a standard method for TOC (such as dry combustion). LOI techniques are simple, relatively cheap, use simple laboratory equipment and do not use expensive or toxic chemicals. Multiple analyses can be run simultaneously (depending on the size of the furnace). For example, the method could take as little as 2 h to analyse approximately 70 samples (aside from pretreatment), which is relatively good throughput (Konen et al. 2002). One of the main drawbacks of using LOI for TOC determinations is the need to use a standard TOC method for establishing the calibration and for quality control checks. The method requires energy for running a furnace at elevated temperatures (although these temperatures are lower than those used for dry combustion, they are required for longer periods and in much larger chambers). LOI is also less suited for analysis of samples across different soil types and locations unless multiple calibrations are used (Konen et al. 2002).
Wet oxidation methods
Early wet methods outlined chemical determination of TOC by three main substances: acidified or alkaline solutions of permanganate or dichromate or, alternatively, hydrogen peroxide in conjunction with chromic acid (Ames and Gaither 1914; Degtjareff 1930). Methods for the quantitative determination of OC using permanganate, which is a powerful oxidising agent, were being explored in the late 19th and early 20th centuries (Ames and Gaither 1914; Degtjareff 1930). Alkaline solutions of potassium permanganate were used to determine some of the constituents of coal in the 1930s (Randall et al. 1938) and later to measure the constituents of soil, such as humic acids, in the 1960s and early 1970s (Schnitzer and Desjardins 1964; Matsuda and Schnitzer 1972). However, early researchers were having doubts as to the extent to which permanganate could oxidise organic matter and were choosing dichromate instead for TOC analyses (Cameron and Breazeale 1904; Ames and Gaither 1914). As a result, the use of permanganate has been reserved for more specific purposes; for example, extraction of humic acids or more dilute solutions for biologically significant OC fraction analysis such as active (or permanganate-oxidisable) carbon (Matsuda and Schnitzer 1972; Blair et al. 1995; Weil et al. 2003; Culman et al. 2012). The method of Degtjareff using hydrogen peroxide for TOC was shown to be flawed by Walkley and Armstrong Black (1934), who outlined that the method introduced an error due to differences between soil and blank reaction stoichiometry. From the 1930s, aside from attempts at specific fraction analysis of OC in soil, the main efforts at deriving a rapid TOC method were focused on the dichromate method, which continued for decades until (and even after) the automation of dry combustion methods (Tabatabai and Bremner 1970).
The significant advantage of dichromate methods is that they do not measure inorganic carbon and therefore are perhaps the most distinctly selective TOC measurement. A rapid, approximate method for the determination of TOC in soils by the addition of potassium dichromate followed by concentrated sulfuric acid was proposed by Walkley and Armstrong Black (1934) along the lines of Schollenberger (1927). Schollenberger's (1927) rapid method involved a heating step and a pre-acidified solution of the dichromate for the redox reaction. The work by Walkley and Armstrong Black (1934) was proffered as a rapid method to be used in situations where exact determinations were not required. They stated recoveries between 60% and 86%, a standard error of [+ or -] 5.6%, with an average recovery of 76% for the 20 soils examined using the method and a correction factor of 100/76 or 1.32 for results obtained based on this average (Walkley and Armstrong Black 1934). Walkley (1935) followed up this work with a paper shortly after that compared the method with a variety of the analyses for TOC at the time. In that paper, dry combustion technique (Dennstedt) was compared with the Bangor method (a modified Kjeldahl method for simultaneous determination of carbon and nitrogen), the hypobromite method and the dichromate method. The recovery using the dichromate technique was the same as that reported in the 1934 study. The Bangor technique gave recoveries averaging approximately 90%, compared with approximately 65% with the hypobromite method, and the dry combustion technique was used as the reference (Walkley 1935).
It was the speed and ease of the analysis compared with all the other techniques that commended Walkley (and subsequently others) to use what has become known as the Walkley Black (WB) method above the other various attempts at improving the dichromate technique. Schollenberger (1945) had developed another version of the method with improved recoveries and repeatability, but it contained extra steps and hazards, such as heated sulfuric acid baths (210[degrees]C) to increase the temperature of reaction. Despite the lack of accuracy and recoveries that differed depending on soil type and locality, the WB method was relatively well used according to Walkley (1947). By this time, the method had evolved to be basically as it is used now: to a soil sample passing through a 0.5-mm screen (containing between 10 and 25 mg TOC), add 10mL of 0.167 M [K.sub.2][Cr.sub.2][0.sub.7] solution and 20 mL of [greater than or equal to] 96% sulfuric acid, shake and stand for 20-30 min; then, add 200-300 mL water, 10mL of 85% phosphoric acid and phenylamine indicator. Finally, titrate with ~1 M solution of FeS[0.sub.4] x 7[H.sub.2]0 to determine unreacted dichromate. This was a straightforward method and it is largely because of this simple approach that the method started to dominate in use. In that paper, Walkley (1947) also highlighted several areas pertaining to the dichromate method that have been pursued since. He discussed the possibility of using the dichromate method without correction as a measure of the amount of readily oxidisable OC in the soil; he increased the size of the glassware to improve titration endpoint determination used for the reaction (and therefore reduced the recovery of TOC because of reduced heat of reaction); and he discussed ambient temperature for its effect on recovery, the importance of the hydrogen ion concentration as the driver of the redox reaction (showing the relationship) and the importance of the concentration of potassium dichromate in terms of the amount of carbon in the sample that can be oxidised. The main interferences, such as the thermal decomposition of chromic acid, carbonates (not significant), chloride (methods to account for), soluble ferrous compounds and oxides of manganese, were also discussed (Walkley 1947).
Mebius (1960) put forward an alternative rapid dichromate method, similar to the WB method, that also included a boiling step for 30 min at around 150[degrees]C in a sand bath and used a reflux condenser. This tended to increase the recovery of TOC to levels similar to dry combustion. The technique was a vast improvement on the WB method in terms of accuracy when compared with dry combustion. There is a reliance on the blank determination for all subsequent determinations, which helps account for breakdown of the dichromate at elevated temperatures (Heanes 1984). An improvement on the Mebius method was made by researchers in Indiana (USA), who changed the reaction vessel to Folin-Wu tubes and included an aluminium heating block set at 150[degrees]C, giving good recoveries (Nelson and Sommers 1974).
Heanes (1984) made the further definitive step in improving the dichromate oxidation method for TOC taking into account most of the issues that arose with the technique experimenting on Australian soils. It appears that this method is the most accurate, repeatable and comparable with the modern dry oxidation techniques developed of the dichromate suite. It remains relatively rapid but with a slightly higher level of equipment required (compared with the WB method), which is neither expensive nor difficult to keep in a modern laboratory. In this adaptation, the main differences are finer grinding of the soil sample, 100mL borosilicate tubes for the reaction vessel fitting into a preheated aluminium block held at 135[degrees]C for 30 min and specified 98% concentrated sulfuric acid. It also uses an adaptation first proposed by Sims and Haby (1971), where the absorbance of the solution is read at 600 nm in a 1-cm path-length cell (reading the absorption of the [Cr.sup.3+] product rather than unused dichromate measured in the titrimetric determination). A calibration with sucrose as the standard OC source is used, whereby the standards go through the same process as the soil samples. The method addressed the issue of breakdown of dichromate at higher temperatures by limiting the temperature of reaction to 135[degrees]C and thereby limiting the breakdown of dichromate compared with the Mebius method at 150[degrees]C. This also takes out the uncertainties of the WB method, which can be affected by the size of the glassware, the temperature of the room and perhaps even factors like what type of bench or bench mat are used, which demonstrates the importance of the temperature of the reaction created with the rapid addition of sulfuric acid. The use of a spectrometer removes the subjective aspects of titration colour change. The Heanes method may also have some advantages over other techniques in calcareous soils where carbonate interferences are minimised (Schmidt et al. 2012). Heanes (1984) also discredited the use of silver sulfate to correct for chloride interference, instead recommending correction using the chloride concentration for saline soils of 1/12 the concentration of chloride detected, although saline soils may be more suited to combustion analysis for this reason.
Around the same time as Mebius, Allison (1960) proposed a dichromate method using various traps for interferences to allow the gravimetric analysis of carbon trapped as carbon dioxide in the final vessel of a glassware train and trap. This method removed the need for titration and in essence is comparable to the dry combustion method (TC) because it also measures carbonates in the sample. A similar adaptation was used by Snyder and Trofymow (1984) that uses a sodium hydroxide trap to measure carbon dioxide with addition of barium chloride and subsequent titration with hydrochloric acid, rather than a colourimetric or titrimetric determination of the chromate oxidation. These methods may be particularly suitable for circumstances where there are negligible amounts of inorganic carbon present, chloride is present or where dry combustion equipment is not readily available.
As technology has become available, different quick methodologies have been attempted as adaptations of the WB method to include a heating step. For example, a domestic microwave oven rated at 850-kW output was used for the dichromate method with a 7-min reaction time in the oven (Tam and Yao 1998). Again, this would be a quick method, but there were no data on the reaction temperatures reached and, in fact, the method gave a stated recovery of 100 [+ or -] 4% for reference materials and 101 [+ or -] 6% for field samples. The researchers were working on marine sediments, so chloride concentrations were a factor, and they used silver sulfate for chloride removal, which may have been less accurate than chloride determination at higher chloride levels (Clark and Ogg 1942; Heanes 1984). So, lack of temperature control (subsequent breakdown of dichromate) and chloride concentrations may have played a part in the recovery for marine sediments. However, further investigations could prove interesting and such a method could be a quicker, simpler version of other heating-step methods (Mebius 1960; Nelson and Sommers 1974; Heanes 1984) using what is now very cheap and available technology.
Research has been conducted comparing the various dichromate techniques since their inception, usually against a standard dry combustion method (Ames and Gaither 1914; Walkley 1935; Clark and Ogg 1942; Kosaka et al. 1959; Grewal et al. 1991; Conyers et al. 2011). A recent example by Conyers et al. (2011) is a comparison between the WB and Heanes methods and a dry combustion method (LECO), measuring 26 different materials that could be included in TOC. Materials included organic reagents, fresh plants, animal manures, composted products, wood and burnt or carbonised material, as well as lime and dolomite, to test the selectivity of the methods for inorganic carbon. The Heanes and dry combustion methods gave very similar recoveries for all materials with the exception of the lime and dolomite, which were essentially unmeasured by the Heanes method. The WB method had a range of recoveries depending on the source of the material (e.g. -100% for organic reagents; -90% for fresh plants, compost and manures; -95% for wood; and between 10% and 55% for burnt or carbonised materials). There was no analysis for the effects of aggregates or protected carbon. Neither the WB nor Heanes methods measured inorganic substances, whereas the dry combustion method measured inorganic carbon. Conyers et al. (2011) suggested that the best use of the WB method was as a measure of readily oxidisable carbon (ROC) and using it without a correction factor. This had been suggested and dismissed quickly by Walkley (1935, 1947), perhaps not wanting to trivialise the method, but given these results and the varying nature of recoveries achieved in much of the research, there may be good reason to treat the WB method as a measure of ROC (Gillman et al. 1986; Lettens et al. 2007). It certainly appears that, given the success of the Heanes method, which is only slightly more onerous in terms of equipment, the case for using the WB method would be one of two scenarios: use as ROC or where only a quick approximate assessment of TOC is required. For the WB method to be used as an ROC method, it would be useful to standardise the glassware, temperature of reaction and final analysis step. The WB method persists as a widely used soil TOC measurement in commercial laboratories in Australia. The widespread historical use of the WB method means that understanding the issues and constraints of the method are important when attempting to calculate changes in carbon stocks based on WB results (Lettons et al. 2007).
The accuracy of all redox methods depends on the oxidation state of the OC in soil. A sample containing a larger number of molecules with low ratio of carbon to hydrogen can give elevated TOC levels when compared to samples containing a larger number of molecules having low ratios of carbon to oxygen. In a whole soil, these types of compounds tend to cancel each other out, although in some soils it could clearly be a source of inaccuracy (Skjemstad and Baldock 2008). One would certainly need to be careful if using the dichromate method to analyse a subset of the TOC because this fraction may have a skewed carbon to oxygen (or hydrogen) ratio compared with the whole soil and therefore give unreliable results. The main drawbacks of the dichromate methods are that the testing is destructive, relatively hands-on for an analyst compared with more automated techniques and the chemicals used are hazardous. Concentrated sulfuric acid is hazardous to handle, as is potassium dichromate (chromium VI is a carcinogen and strong oxidant), which also needs consideration for disposal (Kerven et al. 2000). A method that can help mitigate the issue of unreacted chromium VI is reducing it to chromium III with a carbon source after analysis and before disposal (Bisutti et al. 2004). Nevertheless, the safe disposal or treatment of chromium remains an issue.
Colour and other reflectance methods
Colour (visible reflectance)
The use of soil colour to estimate soil attributes has long been part of the way humans have evaluated soil. Trends in colour roughly define many soil components because of the mineral makeup of the soil. For example, increasing redness in soils is often correlated with increasing concentrations of iron-containing minerals, whereas increasing blackness generally indicates increasing levels of organic matter (Mouazen et al. 2007a). Colour is not necessarily a clear-cut approach for direct determination of constituents. Early research noted that soils with red colour from a particular region may have equal or higher organic matter content than darker soils and that the soil colour depends on the whole soil matrix (Cameron 1905). In order to use colour to measure concentrations, the colour must be defined. In 1905, Albert H. Munsell (an artist and academic) outlined the basis of the Munsell colour system that describes colour in three dimensions (hue, value and chroma) and was a more sophisticated and scientific approach to define colour than earlier attempts (Munsell 2013). The method relies on human eye matching to a colour chip, much like paint chips in a paint store, which can be somewhat subjective. This colour description system would later (1930s USDA) be adopted for use in eye-matching for soil colour and related to other qualitative soil attributes, such as organic matter content and iron content for field assessments; this system is still used today (Baumgardner et al. 1985; Mouazen et al. 2007a; Ibanez-Asensio et al. 2013). Munsell colour charts and the human eye were used to relatively reliably (90% accuracy) group soils into organic matter
contents broadly into <3%, 3%-5% and >5% groups (Steinhardt and Franzmeier 1979). Using Munsell colour (chroma and value), Franzmeier (1988) defined TOC with separate relationships depending on soil particle size. Around this time, non-transferability of calibrations from one geographical or soil type area to another and human eye determination of Munsell colour were noted by researchers as key issues with using colour for TOC (Shields et al. 1968; Steinhardt and Franzmeier 1979; Franzmeier 1988). In order to overcome human eye colour determination inconsistencies, a chroma meter was used to correlate TOC with percentage reflectance, Munsell value, Munsell chroma, sand percentage, clay percentage and particle size on moist and air-dried samples (Konen et al. 2003). However, a significant issue with using colour charts for a definitive measure of TOC is the fact that there are a discrete number of colour chips, whereas the colour range itself is continuous (Viscarra Rossel et al. 2008). Subsequent colour models have been developed as three-dimensional continuous colour spaces.
In 1931, the Commission Internationale de l'Eclairage (CIE) released the CIE XYZ 1931, which described a three-dimensional mathematical colour space, derived from earlier experiments by Wright (1929) and Smith and Guild (1932), separately, on the perception of the human eye to the colours (red, green and blue; Fairman et al. 1997). X, Y and Z are measures rather than colours; X and Z are components of the primary spectra and Y is the luminance, depicted in the x-, y- and z-planes. This model became the standard from which later colour models have drawn. Many transformations of these models have followed, including those accounting for nonlinear and cylindrical representations of the colour space (Viscarra Rossel et al. 2006a). Viscarra Rossel et al. (2006a) used various colour models to frame reflectance measurements made using a spectrometer and mathematical transformations to compare colour to TOC the best correlation being [r.sup.2] = 0.67 using CIE L*c*h* on a diverse set of soil samples.
Digital camera images with controlled lighting conditions were used to prepare calibrations to analyse a set of soil samples for both TOC and iron content (Viscarra Rossel et al. 2008). The 103 soil samples used were from across Brittany (France), with a two-thirds calibration to one-third validation approach. This research showed that the measured red, green and blue (RGB) channels from the camera images could be converted using software into three-dimensional model coordinates of various colour space models to predict TOC and Fe. The samples were also analysed for their diffuse reflectance spectra in the visible (Vis)-near infrared (NIR) using a field spectrometer to validate the soil colour measurement by the digital camera. The predictions for TOC were better than those for Fe in that study and comparable to the field spectrometer used, with calibration and validation correlations both being [r.sup.2] = 0.88 for a log linear transformation of the v* signal of the CIE Lu*v* model and [r.sup.2] = 0.92 and 0.91, respectively, for a full factorial regression analysis of the C1E L*a*b* model (Viscarra Rossel et al. 2008). This type of colour measurement is relatively rapid, inexpensive, non-destructive and the digital camera equipment is robust. The only extra requirements are for controlled lighting conditions and a computer with appropriate software for the required transformations. There are many approaches that could be tried along similar lines. Uptake of mobile phone technology in recent years (most of which include a camera of sufficient quality for this use) means that cameras could provide on-the-go analysis with minimal extra equipment needs (Gomez-Robledo et al. 2013). The opportunities for colour analysis of TOC appear to be in developing a simpler calibration method that is quick, relatively minimal in terms of number of calibration points required and readily adapted to different sites. Limitations may be that the technique is not sophisticated enough to account for TOC in all its forms. The effects of different states of OC measured using the colour approach appears to be an area of research still to be explored.
Reflectance measurements in the visible and infrared regions
Methods for spectrometer-based reflectance measurement in the Vis, NIR and mid infrared (MIR) parts of the electromagnetic (EM) spectrum have received considerable attention in the past few decades. There are numerous advantages in NIR and/or MIR reflectance spectroscopic techniques: testing can have relatively high throughput, may not require extensive sample pretreatment, can be conducted without expensive reagents or consumables and is non-destructive. NIR spectroscopy seems best suited to infield or on-the go measurement, whereas the MIR works best on dried and ground samples, and therefore in-laboratory analysis (Bellon-Maurel and McBratney 2011; Baldock et al. 2013a). The spectra recorded in the infrared (IR) regions indicate the types of compounds, including functional groups, that may be present because radiation is absorbed, resulting in changes to vibrational stretching and bending energies of chemical bonds (Guerrero et al. 2010; Genot et al. 2011). Broadly, the MIR spectra give information of fundamental bands of molecular vibrations, whereas NIR spectra show overtones and combinations of fundamental bands (Hunt 1977; Bellon-Maurel and McBratney 2011). Therefore, IR spectra from a sample may deliver information not only on compounds contributing to TOC, but also on many other components and soil parameters, increasing the value of this type of analysis (Janik et al. 1998; Islam et al. 2003; Viscarra Rossel and Webster 2012).
Measuring reflectance of soils in the Vis and NIR wavelengths, and linking parameters such as organic matter, particle size and moisture content, was beginning in the 1960s (Bowers and Hanks 1965). Research into the use of reflectance measurement in the Vis-NIR for soils was gaining momentum in the late 1970s to analyse for minerals and organic matter (Hunt 1977; Krishnan et al. 1980). Interest in the use of NIR reflectance for soils grew out of the use of similar methods for analysis of foods (e.g. grains and meat) for measures like proteins, moisture and lipids, and many industrial applications ranging from finishes on textiles to moisture levels in pharmaceuticals (Wetzel 1983; Bellon-Maurel et al. 2010; Bellon-Maurel and McBratney 2011). In these applications, the materials may be complex mixtures of compounds but are relatively consistent with very similar structures. However, soils are often not similar and may have a completely different matrix depending on the parent minerals they are derived from, the topography of the land from which they are taken or even inputs in agricultural systems (Genot et al. 2011). Herein lies the major difficulty with using the spectroscopic reflectance techniques for TOC analysis in soils: the soil matrix itself. This is why in much of the research calibrations with better accuracy are within soil type and locality, which reduces the amount of variation being dealt with (Genot et al. 2011). Development of calibrations that hold for different geographical and soil type locations, modifying (or spiking) to allow analysis from new unknown samples or the development of large spectral libraries are the main approaches used so far to improve the usability and increase the uptake of IR spectroscopy techniques for the measurement of TOC and other soil parameters (Brown et al. 2006; Sankey et al. 2008; Guerrero et al. 2010; Viscarra Rossel and Webster 2012; Baldock et al. 2013a). These approaches help with the difficulties of developing a calibration that will rely on relatively large numbers of samples to have both their IR reflectance spectra scanned (perhaps multiple times) and separate standard method analysis for the analyte(s) of interest. The success of these or similar approaches may determine whether reflectance techniques can replace more resource-intense methods.
The power of using spectral reflectance measurement in the Vis-NIR-MIR regions to analyse for several soil parameters (including TOC) has been investigated (Dalai and Henry 1986; Morra et al. 1991; Ben-Dor and Banin 1995; Janik and Skjemstad 1995; Islam et al. 2003). One paper reported positive correlations for 36 different soil parameters to an MIR prediction with TOC one of the more highly correlated (Janik et al. 1998). Islam et al. (2003) measured 13 different parameters in a wide range of soils ranging from pH and EC through to TOC, exchangeable cations and clay content in the ultraviolet (UV)-Vis-NIR range, of which seven (including TOC) were successful. The attraction of determining multiple analytes from a relatively rapid non-destructive test shows why these techniques are being investigated and why attempts are made to improve accuracy, transferability and usability. Viscarra Rossel et al. (20066) reviewed the use of UV-Vis-NIR- MIR spectra for quantitative analysis, providing the spectral region range, the calibration and error values and including the chemometric method used. The selection of the chemometric method (data analysis) can be important to the ability to extract meaningful information from the data derived from the scan of the sample. According to that review, several elements (Al, Ca, Cu, Fe, K, Mg, Mn, C (total), C (inorganic), TOC, N, P) and measures (acid (exchangeable), cation exchange capacity (CEC), EC, pH, lime requirement, organic matter, carbonates) had been measured using the spectroscopic techniques with varying levels of accuracy and precision (Viscarra Rossel et al. 20066). The MIR region showed the best results for TOC. The most common chemometric technique used for MIR analysis for TOC was partial least-squares regression (PLSR; Janik and Skjemstad 1995; McCarty et al. 2002; Viscarra Rossel et al. 20066).
NIR reflectance spectroscopy
NIR reflectance spectroscopy interrogates the spectra of sample reflected radiation in the range of approximately 2.5-25 [micro]m (4000 400 [cm.sup.-1]). In this part of the spectrum, radiation does not penetrate soil samples as deeply as the NIR, so it produces more specific and less diffuse spectra, but this is also a source of signal noise (Bellon-Maurel and McBratney 2011). In early research it was believed that it would be necessary to dilute soil samples using a non-absorbing substrate, such as potassium bromide, when analysing via absorption in the MIR because of the presence of strong absorption bands in the region (Farmer 1957). Although the pressed potassium bromide method produced excellent spectra, the technique involves more sample preparation and has some handling issues (the hygroscopic and reactive nature of potassium bromide and tedious preparation of samples), thus reflectance was favoured especially with the advent of Fourier transform instruments (Farmer 1957; Baes and Bloom 1989; Janik and Skjemstad 1995; Janik et al. 1998). The MIR reflectance spectra tend to give more accurate and reproducible values when samples are dried and ground to a fine, consistent size (and the time of grinding standardised) where laboratory-based research has shown the best results (Bellon-Maurel and McBratney 2011; Baldock et al. 2013a). MIR analysis has been shown in some studies to give more accurate results than NIR in laboratory-based analyses because of better specificity (McCarty et al. 2002; Viscarra Rossel et al. 20066). However, because more sample preparation is required, cost versus accuracy of analysis must be considered. Calibrations also need to be constructed using a reference method, which adds a further dimension of cost. A particular characteristic of the MIR region is the strong absorption band for carbonate that must be accounted for when analysing TOC (McCarty et al. 2002). Quartz and kaolin arc also possible interferences when analysing for TOC in the MIR (Viscarra Rossel et al. 20066).
NIR reflectance spectroscopy
NIR reflectance spectroscopy nominally interrogates the spectra of sample reflected radiation in the range 0.8-2.5 [micro]m (12 500-4000 [cm.sup.-1]), although this is dependent on the analytes of interest (Viscarra Rossel et al. 20066). Sometimes the visible part of the spectrum is included in scans down to around 0.4 [micro]m, because parts of the visible spectrum are useful in predicting TOC (Viscarra Rossel et al 2006b, 2008). Spectra in the NIR relate to overtones and combinations of the fundamental molecular vibrations (particularly of CH, OH and NH groups) and are therefore less specific (Ben-Dor and Banin 1995; Viscarra Rossel et al. 2006b). Radiation in the NIR part of the spectrum penetrates more deeply into the sample and the diffusion is more pronounced than in the MIR. These phenomena lead to the likely use of NIR spectroscopy for infield or less-processed samples because of its less specific nature, ability to handle coarser samples, rapid analysis and availability of light, portable instruments (Brown et al. 2006; Brunet et al. 2007; Chatterjee et al. 2009; Bellon-Maurel and McBratney 2011). Calibrations for coarser and less-processed soils are not as good as those for dried and ground samples; however, the trade-off between more determinations with less sample preparations will be important as to the method chosen (Brunet et al. 2007). The ability of available NIR instruments to scan relatively large sample areas in comparison with MIR also helps NIR analysis to be less encumbered by sample particle size (Reeves 2010). The review by Reeves (2010) gives an in-depth analysis of the advantages and disadvantages of MIR versus NIR reflectance for carbon measurement, describing specific instrumentation effects, characteristics of the spectra and various effects of interferences, that are beyond the scope of the present review.
Possible applications for in situ, on-the-go or remote sensing TOC analysis
In situ or on-the-go methods
The exact definition of these terms may be open to some debate, but the definition used herein essentially covers the types of techniques that are the main candidates for further work. In situ analysis describes that which occurs in the field, either on a freshly removed sample or in an undisturbed state. On-the-go analysis may be conducted either scanning an undisturbed soil or measuring in combination with an earth-moving implement (e.g. instrument in combination with a plough attached to a tractor). There is significant cost and time involved in removing samples and then the drying, weighing, crushing and sieving that are generally a part of laboratory-based methods for TOC determination as described above. It is for these reasons, combined with the need for more accurate understanding of OC stocks (i.e. more data over larger areas) that methods not involving sampling or those not involving sample removal from the site are attractive.
NIR reflectance spectroscopy
NIR reflectance is a significant candidate for in situ or on-the-go measurement of TOC for the reasons discussed in the previous section. Some approaches use an NIR sensor being towed behind a tractor (Mouazen et al. 20076; Christy 2008; Bricklemyer and Brown 2010). Others use various scanning heights (e.g. 1 m) above a ploughed field (Stevens et al. 2008). NIR analysis for TOC has also been conducted on sample cores removed in situ with moderate success.
Laser-induced breakdown spectroscopy
Laser-induced breakdown spectroscopy (LIBS) is a relatively new technique for the analysis of soils that has slowly increased in prominence over the past 15 years (Cremers et al. 2001). Like other modern soil carbon methods, it has been used in other applications before use on soils. LIBS has uses on gas, liquid, solid and mixed-phase materials for analysis pertaining to contaminants in air and waste water, as well as elements in ores and minerals or even toothpaste, with soils being a more recent use (Rusak et al. 1997; Condal et al. 2014). LIBS is an elemental analysis technique based on atomic emission spectroscopy (AES) methods. It uses a focused laser, which, when concentrated on a sample's surface, produces plasma representative of the elements (ions and molecules) from the sample in an electronically excited state. As the elements return to lower electronic states, they emit signature radiation, which, when appropriately detected, may give information as to their presence and concentrations (Bricklemyer et al. 2011; Martin et al. 2013). LIBS is a destructive testing procedure, leaving behind small craters in the surface of the sample, but is only focused on a relatively small volume (~0.1-0.5 mm wide and 0.01-0.2 mm deep depending on the number of pulses and instrumentation), so repeated testing is possible even from relatively small samples or surfaces (Harmon et al. 2006; Belkov et al. 2009).
Being an elemental technique, LIBS has been used to measure TC with some success, although the research shows an attempt at specifically measuring TOC (Martin et al. 2013). The technique is rapid (a single determination may be made in 1 min) and the type of sample presentation may range from dried and ground samples to recently removed cores to in situ, possibly in appropriate depth holes using a probe (Mosier-Boss et al. 2002; Harmon et al. 2006; da Silva et al. 2008; Belkov et al. 2009). The technique has good potential for portability (Harmon et al. 2006). However, early attempts at in situ measurement using probes or measuring directly from cores have found issues with soil matrix conditions, such as grain size, composition and porosity (because of the relatively small sampling volume) and water content (Mosier-Boss et al. 2002; Bricklemyer et al. 2011). For carbon analyses of the research reviewed, calibrations were far more accurate on dried (or cryogenically frozen) and ground samples or pellets than field moist cores (da Silva et al. 2008; Belkov et al. 2009; Bricklemyer et al. 2011). In some studies, it was shown that iron can be an interference in the measurement of carbon because of a strong emission peak near the carbon line at 247 nm, which has been addressed by measuring at the 193 nm line or combinations of both with satisfactory results (Ebinger et al. 2003; da Silva et al. 2008; Glumac et al. 2010). The technique usually uses a double pulse of the laser, which has been shown in some studies to increase the intensity and contrast of emission spectra compared with a single pulse, although various pulse techniques have been reported (Rozantsev et al. 1993; Belkov et al. 2009).
It appears that for the LIBS measurement of TOC, the samples or soils should contain insignificant levels of carbonates or be appropriately treated to remove them before analysis. As a laboratory technique alongside others requiring sample preparation (e.g. drying and grinding), LIBS has some favourable attributes given that the number of calibration points seems to be much less than those required for techniques such as 1R reflectance methods. The instrumentation is relatively robust, and although it is a destructive technique, there are no hazardous side effects except for the possibility of producing X-rays, which can be eliminated by correct set-up of the laser (da Silva et al. 2008). The relatively small volume of sampling for the analysis could lead to very localised and non-representative sampling that should be overcome by increased numbers of measurements taken and averaging spectra (da Silva et al. 2008). Given that the testing is destructive, the exact measurement cannot be repeated. The technique is in the early days of development as a soil carbon measurement, so opportunities to improve calibrations, and particularly as an in situ analysis technique with minimal sample preparation, appear to be the immediate priorities, as they are for any method with likely in-field capability. Perhaps the greatest issue presenting against its use as a portable technique will be overcoming sample heterogeneity issues.
Inelastic neutron scattering
Inelastic neutron scattering (INS) is a method under development for carbon measurements in surface soils, but similar methods have been used for decades to determine elemental concentrations in industrial applications, such as down boreholes of oil wells (Parsons et al. 2011). INS involves spectroscopy of gamma rays that emanate from the inelastic collisions of fast neutrons with the nuclei of elements present in the soil. These collisions result in emission of a characteristic radiation that may be read by an appropriate detector (or array of detectors; Wielopolski et al. 2008; Parsons et al. 2011). Other neutron interactions include elastic scattering leading to thermal neutrons that may also be measured because they are separated in time (from the inelastic collisions) when the neutron generator is pulsed (Wielopolski et al. 2008; Parsons et al. 2011). This allows an elemental analysis of a variety of soil constituents, including TC (Parsons et al. 2011). The time required for accurate measurements depends on the energy source and the concentration of the element of interest that relates to the number of counts at the appropriate point in the spectra compared with the background (Wielopolski et al. 2008; Falahat et al. 2012).
In research conducted with a customised instrument measuring a soil volume of 0.3 [m.sup.3] (~0.4 m depth), the time chosen for acquiring counts at each location was 1 h when operated statically (Wielopolski et al. 2010). The instrument can also be used in a scanning mode that is mobile and may average an area along a transect (Wielopolski et al. 2010, 2011). One hour for static measurement may seem quite long, but given that the volume sampled is so large, the method shows some promise. The system used by Wielopolski et al. (2011) uses electrical energy to produce the radiation and does not emit (gamma radiation) once turned off.
As with many of the other methods, INS does not discriminate between inorganic and organic carbon. The system proposed by Wielopolski et al. (2011) does not sample the soil by physical removal so measures TC. Measurement of TOC would need to be made either by taking separate soil samples and testing for carbonate presence, or using in locations with known carbonate status. From the research conducted thus far, the method looks more promising in terms of the principles of on-the-go analysis and especially at higher levels of soil carbon than other methods (Wielopolski et al. 2011). However, there is no discussion in the literature using INS on the soil surface of the possible effects on soil biota of bombardment with neutrons, particularly gamma radiation, which has been shown to cause DNA damage in humans (Olive 1998). It is stated that INS is a non-destructive technique, an assignment that may not take into account the effects on overall soil life, which could be important because the interaction volumes are quite large, especially if scanning along a transect (Wielopolski et al. 2011). If these issues are shown to be negligible, then the technique could be a leading contender in on-the-go analysis of surface soils for TC and TOC.
Airborne remote sensing
Remote sensing may have various definitions based on what it practically means. It could possibly be applied to any scanning or on-the-go measurements that do not physically remove samples. It could equally be applied to samples of removed soils that are not physically changed by the process (e.g. taking a photograph or spectrometer scan of a soil that does not physically alter it). Herein the discussion focuses on airborne remote sensing; measurement that is not in physical contact with the ground (i.e. aeroplane or satellite-based measurements). A more recent technology gaining more use in agricultural or farming applications such as weed control and mapping (among others) is unmanned flying equipment, such as drones or helicopters, that at some point could be used for remote sensing for TOC measurement. All the currently used techniques use principles of reflectance measurement based in the visible and 1R regions.
Chen et al. (2000) used colour aerial photographs from a plane, subsequently digitally scanned, to obtain information on the red, green and blue data transformed into a log linear relationship for TOC from a ploughed field. For that study, sampled areas at 2 x 2 m plots on the ground related to the pixel size resolution of the image information used, with very good agreement in the calibration ([r.sup.2] = 0.93) and validation ([r.sup.2] = 0.97) obtained. Ben-Dor et al. (2002) used airborne spectrometer-based data for soil parameter determination (including organic matter). That study was conducted over Israel, flying at a height of ~3000 m and measuring 72 spectral bands using the DAIS7915 scanner (GER Inc., USA) between 0.4 and 2.5 [micro]m. This provided an on-ground resolution of approximately 8-m square pixels and the ground testing was based on 16 pixels of information from the scanner. Understandably, the research suffered from many difficulties and interferences, giving perhaps qualitative information, but was an important first step in outlining the possibilities of airborne approaches. The work by Selige et al. (2006) in Germany used atmospheric and geographic correction for the data from the HyMap scanner (Integrated Spcctronics Pty Ltd, Baulkham Hills, NSW) that measures bandwidths of 15-20nm at 128 positions in the Vis-NIR. This approach gave good correlations for TOC with a relatively small dataset across 12 bare fields totalling ~700ha in a study area 200 [km.sup.2] in size. A more recent attempt in Luxembourg used an airplane and the AHS-160 sensor measuring 63 spectral bands between 430 and 2540 nm (Stevens et al. 2010). The image width was 1.96 km and the on-ground resolution was 2.6 x 2.6m with a study area of approximately 7 x 60 km. Stevens et al. (2010) used three different multivariate techniques and classified the dataset into region, soil classification and the five different hyperspectral images taken, as well as global data for analysis of TOC. Region and soil classification type-based datasets gave the strongest results. That study also used software to remove various anomalous information, such as bare road tracks and overly vegetated soils, before applying chemometrics (Stevens et al. 2010).
The advent of the Landsat satellites (1972-) and information from the on-board Multi-Spectral Scanner (MSS) systems increased interest in airborne reflectance measurement, including the ability to gamer important agricultural information (Irons 2014). Researchers have used these satellite resources with success in several areas; however, to date, measurement of TOC has not achieved satisfactory results. Laboratory-based reflectance measurements started to be undertaken in order to determine the feasibility of whether the data from satellites and planes could be of any use in the determination of TOC (Henderson et al. 1992). The MSS systems on Landsat scan the Earth's surface, measuring reflected sunlight in several bands in the Vis and NIR spectra (USGS 2013). Sensors aboard these satellites initially recorded data in relatively wide (~100nm) and few (six) bands. In addition, the resolution of earlier Landsat-based scanners was 80 m at ground level, which was reduced to 30 m with more recent satellites in the series (USGS 2013). This lack of specificity has hindered analysis of the data to accurately predict properties such as TOC because of the many variations that may occur, including very heterogeneous surfaces within a 30-m wide area. In more recent times, the possibility of multispectral scanning information from satellites with resolution into <2-m increments has occurred (e.g. GeoEye-1 or Pleiades-1A; see www. satimagingcorp.com/, accessed 25 July 2014). This level of resolution fits with the work conducted by Chen et al. (2000), who obtained excellent results from photographs taken from an airplane and using 2 x 2 m pixel resolution.
Meteorological conditions need to be accounted for when using airborne data because cloud, sun position and other localised phenomena affect the reflected light. Therefore, airborne data often need ground-based observations to help correct for interferences (Ben-Dor et al. 2009; Croft et al. 2012). There have been relatively few attempts to measure TOC using satellite data, and calibrations have not been as successful as ground or laboratory-based data, despite early optimism. An issue for satellite spectral IR information remains a low signal to noise ratio (Gomez et al. 2008). Satellite imagery from Landsat was used to attempt monitoring of agricultural practice that may lead to carbon sequestration through crop residue biomass (Bricklemyer et al. 2007). Gomez et al. (2008) used data from the Hyperion hyperspectral instrument aboard the NASA Earth Observation-1 satellite (Ichoku and Przyborski 2014). Gomez et al. (2008) showed that predictions using the spectra obtained from Hyperion data were less accurate than a field-based core-scanned analysis and only gave qualitative results with prediction [r.sup.2] = 0.51. The researchers put this difference down to the signal to noise ratio and the 30 m surface resolution of the Hyperion data being too coarse. Correction for anomalies, such as those used by Stevens et al. (2010), was not applied in the study of Gomez et al. (2008).
The approach of airborne remote sensing continues to be a subject of research, and seems capable of producing quantitative results with the appropriate amount of data preprocessing. The ability to more accurately predict TOC levels using this type of approach would allow the best type of broad-scale analysis for surface TOC because of its non-invasive, wide-area data capture. This could lead to decreased costs of this type of analysis if it were able to be applied for scans that are often kilometres wide, improving the economy of scale. This type of approach is clearly limited to surface soils and best applied to surfaces with vegetation removed (e.g. ploughed fields, which, ironically, are part of practices that likely lead to loss of OC; Dalai and Mayer 1986; Cannell and Hawes 1994; Olson et al. 2005). An attempt at measuring TOC with some vegetative cover has been made (Bartholomeus et al. 2011). That study showed that vegetative cover (early maize crop) significantly affected the prediction of TOC, but the researchers were also able to make a correction to somewhat account for the partial ground cover by using a filter derived from a model of simulated leaf reflectance. Significantly, because the ground cover was only partial, the readiness of an approach to account for well-vegetated soils is not evident (Bartholomeus et al. 2011). Drones or other unmanned flying equipment could be particularly useful for more localised images and scanning closer to the soil surface, which is not possible with satellite or aeroplane-based platforms. Significantly, drones may be able to zoom in on small, slightly exposed soil surfaces among otherwise vegetated areas, which could increase the land area that this type of approach could be used for. Surface soils contain significant levels of OC, with concentrations generally decreasing with depth, so this approach still warrants attention. For these reasons alone it should continue to attract research and approaches to improve accuracy.
All the TOC methods described herein are encumbered by at least one of interference, toxic side effect or a lack of accuracy desired for wider-scale use. The technique chosen depends on the scale and type of data required and consideration of the resources available for sampling and laboratory analysis versus the various in-field or remote approaches. The current state-of-the-art approaches for samples analysed in the laboratory remain dry combustion for non-calcareous soils, with the Heanes method a better option in highly calcareous soils. Of the modern contenders, MIR analysis appears to be the approach most likely to overtake laboratory-based measurements with appropriate access to spectral libraries and techniques to analyse unknown origin samples. The race is more wide open for infield, remote or on-the go techniques, with NIR more developed. Evolving techniques, such as LIBS, INS or airborne remote sensing, require approaches to improve precision and accuracy so the results obtained can be confidently used.
Allison LE (1960) Wet-combustion apparatus and procedure for organic and inorganic carbon in soil. Soil Science Society of America Journal 24, 36-40. doi: 10.2136/sssaj 1960.03615995002400010018x
Ames JW, Gaither EW (1914) Determination of carbon in soils and soil extracts. Journal of Industrial and Engineering Chemistry 6, 561-564. doi: 10.1021/ie50067a010
Artursson V, Finlay RD, Jansson JK (2006) Interactions between arbuscular mycorrhizal fungi and bacteria and their potential for stimulating plant growth. Environmental Microbiology 8, 1 10. doi: 10.1111/j. 14622920.2005.00942.x
Baes AU, Bloom PR (1989) Diffuse reflectance and transmission Fourier transform infrared (DRIFT) spectroscopy of humic and fulvic acids. Soil Science Society of America Journal 53, 695-700. doi: 10.2136/sssaj 1989.03615995005300030008x
Baldock JA (2007) 'Nutrient cycling in terrestrial ecosystems.' (Springer-Verlag: Berlin, Heidelberg)
Baldock JA, Hawke B, Sanderman J, Macdonald LM (2013a) Predicting contents of carbon and its component fractions in Australian soils from diffuse reflectance mid-infrared spectra. Soil Research 51, 577-595. doi: 10.1071 /SR 13077
Baldock JA, Sanderman J, Macdonald LM, Puccini A, Hawke B, Szarvas S, McGowan J (2013b) Quantifying the allocation of soil organic carbon to biologically significant fractions. Soil Research 51, 561-576. doi:10.1071/SR12374
Ball DF (1964) Loss-on-ignition as an estimate of organic matter and organic carbon in non-calcareous soils. Journal of Soil Science 15, 84-92. doi: 10.1111/j. 1365-2389.1964.tb00247.x
Barrow CJ (2012) Biochar: potential for countering land degradation and for improving agriculture. Applied Geography 34, 21-28. doi:10.1016/j.apgeog.2011.09.008
Bartholomeus H, Kooistra L, Stevens A, van Leeuwen M, van Wesemael B, Ben-Dor E, Tychon B (2011) Soil organic carbon mapping of partially vegetated agricultural fields with imaging spectroscopy. International Journal of Applied Earth Observation and Geoinformation 13, 81-88. doi: 10.1016/j.jag.2010.06.009
Baumgardner MF, Silva LF, Biehl LL, Stoner ER (1985) Reflectance properties of soils. In 'Advances in agronomy'. Vol. 38. (Ed. NC Brady) pp. 1-14. (Academic Press: Orlando, FL)
Belkov MV, Burakov VS, De Giacomo A, Kiris VV, Raikov SN, Tarasenko NV (2009) Comparison of two laser-induced breakdown spectroscopy techniques for total carbon measurement in soils. Spectrochemica Acta Part B 64, 899-904. doi:10.1016/j.sab.2009.07.019
Bellon-Maurel V, McBratney A (2011) Near-infrared (NIR) and midinfrared (MIR) spectroscopic techniques for assessing the amount of carbon stock in soils: critical review and research perspectives. Soil Biology & Biochemistry 43, 1398-1410. doi:10.1016/j.soilbio.2011.02. 019
Bellon-Maurel V, Fernandez-Ahumada E, Palagos B, Roger J-M, McBratney A (2010) Critical review of chemometric indicators commonly used for assessing the quality of the prediction of soil attributes by NIR spectroscopy. Trends in Analytical Chemistry 29, 1073-1081. doi: 10.1016/j.trac.2010.05.006
Ben-Dor E, Banin A (1989) Determination of organic matter content in aridzone soils using a simple 'loss-on-ignition' method. Communications in Soil Science and Plant Analysis 20, 1675-1695. doi: 10.1080/0010 3628909368175
Ben-Dor E, Banin A (1995) Near-infrared analysis as a rapid method to simultaneously evaluate several soil properties. Soil Science Society of America Journal 59, 364-372. doi:10.2136/sssaj1995.0361599500 59000200I4x
Ben-Dor E, Patkin K, Banin A, Kamieli A (2002) Mapping of several soil properties using DAIS-7915 hyperspectral scanner data: a case study over clayey soils in Israel. International Journal of Remote Sensing 23, 1043-1062. doi:10.1080/01431160010006962
Ben-Dor E, Chabrillat S, Dematte JAM, Taylor GR, Hill J, Whiting ML, Sommer S (2009) Using imaging spectroscopy to study soil properties. Remote Sensing of Environment 113(Suppl. 1), S38-S55. doi: 10.1016/j. rse.2008.09.019
Bisutti I, Hilke I, Raessler M (2004) Determination of total organic carbon: an overview of current methods. Trends in Analytical Chemistry 23, 716-726. doi: 10.1016/j.trac.2004.09.003
Blair GJ, Lefroy RDB, Lisle L (1995) Soil carbon fractions based on their degree of oxidation and the development of a carbon management index for agricultural systems. Australian Journal of Agricultural Research 46, 1459-1466. doi:10.1071/AR9951459
Blouin M, Hodson ME, Delgado EA, Baker G, Brussaard L, Butt KR, Dai J, Dendooven L, Peres G, Tondoh JE, Cluzeau D, Brun JJ (2013) A review of earthworm impact on soil function and ecosystem services. European Journal of Soil Science 64, 161-182. doi: 10.1111/ejss. 12025
Bowers SA, Hanks RJ (1965) Reflection of radiant energy from soils. Soil Science 100, 130-138. doi: 10.1097/00010694-196508000-00009
Bricklemyer RS, Brown DJ (2010) On-the-go VisNIR: potential and limitations for mapping soil clay and organic carbon. Computers and Electronics in Agriculture 70, 209-216. doi:10.1016/j.compag. 2009.10.006
Bricklemyer RS, Lawrence RL, Miller RM, Battogtokh N (2007) Monitoring and verifying agricultural practices related to soil carbon sequestration with satellite imagery. Agriculture, Ecosystems & Environment 118, 201-210. doi: 10.1016/j.agee.2006.05.017
Bricklemyer RS, Brown DJ, Barefield JE, Clegg SM (2011) Intact soil core total, inorganic, and organic carbon measurement using laser-induced breakdown spectroscopy. Soil Science Society of America Journal 75, 1006-1018. doi: 10.2136/sssaj2009.0244
Brock WH (2013) Justus, baron von Liebig. In 'Encyclopaedia Britanica'. Available at: www.britannica.com/biography/Justus-Freiherr-von-Liebig (accessed 4 April 2014).
Brown DJ, Shepherd KD, Walsh MG, Dewayne Mays M, Reinsch TG (2006) Global soil characterization with VNIR diffuse reflectance spectroscopy. Geoderma 132, 273-290. doi:10.1016/j.geoderma.2005. 04.025
Brunet D, Barthes BG, Chotte J-L, Feller C (2007) Determination of carbon and nitrogen contents in Alfisols, Oxisols and Ultisols from Africa and Brazil using NIRS analysis: effects of sample grinding and set heterogeneity. Geoderma 139, 106-117. doi: 10.1016/j. geoderma. 2007.01.007
Calvelo Pereira R, Camps Arbestain M, Kaal J, Vasquez Suero M, Sevilla M, Hindmarsh J (2014) Detailed carbon chemistry in charcoals from preEuropean Maori gardens of New Zealand as a tool for understanding biochar stability in soils. European Journal of Soil Science 65, 83-95. doi: 10.1111/ejss. 12096
Cameron FK (1905) A comparison of the organic matter in different soil types. Journal of the American Chemical Society 27, 256-258. doi: 10.1021/ja01981a009
Cameron FK, Breazeale JF (1904) The organic matter in soils and subsoils. Journal of the American Chemical Society 26, 29-45. doi: 10.1021/ja01991a003
Cannell RQ, Hawes JD (1994) Trends in tillage practices in relation to sustainable crop production with special reference to temperate climates. Soil & Tillage Research 30, 245-282. doi: 10.1016/0167-1987(94) 90007-8
Chatterjee A, Lai R, Wielopolski L, Martin MZ, Ebinger MH (2009) Evaluation of different soil carbon determination methods. Critical Reviews in Plant Sciences 28, 164-178. doi: 10.1080/0735268090 2776556
Chen F, Kissel DE, West LT, Adkins W (2000) Field-scale mapping of surface soil organic carbon using remotely sensed imagery. Soil Science Society of America Journal 64, 746-753. doi:10.2136/sssaj 2000.642746x
Chia CH, Munroe P, Joseph S, Lin Y (2010) Microscopic characterisation of synthetic Terra Preta. Australian Journal of Soil Research 48, 593-605. doi: 10.1071/SR10012
Christy CD (2008) Real-time measurement of soil attributes using on-the-go near infrared reflectance spectroscopy. Computers and Electronics in Agriculture 61, 10-19. doi:10.1016/j.compag.2007.02.010
Ciais P, Sabine C, Bala G, Bopp L, Brovkin V, Canadell J, Chhabra A, DeFries R, Galloway J, Heimann M, Jones C, Le Quere C, Myneni RB, Piao S, Thornton P (2013) Carbon and other biogeochemical cycles. In 'Climate Change 2013: The Physical Science Basis. Contribution of Working Group 1 to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change'. (Eds TF Stocker, D Qin, G-K Planner, M Tignor, SK Allen, J Boschung, A Nauels, Y Xia, V Bex, PM Midgle) pp. 465-570. (Cambridge University Press: Cambridge, UK and New York, USA)
Clark NA, Ogg CL (1942) A wet combustion method for determining total carbon in soils. Soil Science 53, 27-36. doi:10.1097/00010694194201000-00003
Conyers MK, Poile GJ, Oates AA, Waters D, Chan KY (2011) Comparison of three carbon determination methods on naturally occurring substrates and the implication for the quantification of 'soil carbon'. Soil Research 49, 27-33. doi: 10.1071/SR10103
Cremers DA, Ebinger MH, Breshears DD, Unkefer PJ (2001) Measuring total soil carbon with laser-induced breakdown spectroscopy (LIBS). Journal of Environmental Quality 30, 2202-2206. doi: 10.2134/jeq2001.2202
Croft H, Kuhn NJ, Anderson K (2012) On the use of remote sensing techniques for monitoring spatio-temporal soil organic carbon dynamics in agricultural systems. Catena 94, 64-74. doi: 10.1016/j. catena.2012.01.001
Culman SW, Snapp SS, Freeman MA, Schipanski ME, Beniston J, Lai R, Drinkwater LE, Franzluebbers AJ, Glover JD, Grandy AS, Lee J, Six J, Maul JE, Mirsky SB, Spargo JT, Wander MM (2012) Permanganate oxidizable carbon reflects a processed soil fraction that is sensitive to management. Soil Science Society of America Journal 76, 494-504. doi: 10.2136/sssaj2011.0286
da Silva RM, Milori DMBP, Ferreira EC, Krug FJ, Martin-Neto L (2008) Total carbon measurement in tropical soil sample. Spectrochemica Acta Part B 63, 1221-1224. doi:10.1016/j.sab.2008.09.003
Dalal RC, Henry RJ (1986) Simultaneous determination of moisture, organic carbon and total nitrogen by near infrared reflectance spectrophotometry. Soil Science Society of America Journal 50, 120-123. doi: 10.2136/sssaj1986.03615995005000010023x
Dalal RC, Mayer RJ (1986) Long term trends in fertility of soils under continuous cultivation and cereal cropping in southern Queensland. 1. Overall changes in soil properties and trends in winter cereal yields. Australian Journal of Soil Research 24, 265-279. doi: 10.1071/SR9860265
Degtjareff WT (1930) Determining soil organic matter by means of hydrogen peroxide and chromic acid. Soil Science 29, 239-246. doi: 10.1097/00010694-193003000-00007
Ebinger MH, Norfleet ML, Breshears DD, Cremers DA, Ferris MJ, Unkefer PJ, Lamb MS, Goddard KL, Meyer CW (2003) Extending the applicability of laser-induced breakdown spectroscopy for total soil carbon measurement. Soil Science Society of America Journal 67, 1616-1619. doi:10.2136/sssaj2003.1616
Elementar (2014) Vario MACRO cube. Available at: www.elementar.de/en/ products/elementar-products/vario-macro-cube.html (accessed 9 July 2014). Fairman HS, Brill MH, Hemmindger H (1997) How the CIE 1931 colormatching functions were derived from Wright-Guild data. Color Research and Application 22, 11-23. doi:10.1002/(SICI)1520-6378 (199702)22:1<1 l::AID-COL4>3.0.CO;2-7
Falahat S. Koble T, Schumann O, Waring C, Watt G (2012) Development of a surface scanning soil analysis instrument. Applied Radiation and Isotopes 70, 1107-1109. doi:10.1016/j.apradiso,2011.12.024
Farmer VC (1957) Effects of grinding during the preparation of alkali halide disks. Spectrochemica Acta 8, 374-389. doi: 10.1016/0371-1951 (57) 80197-0
Fidencio PH, Poppi RJ, de Andrade JC (2002) Determination of organic matter in soils using radial basis function networks and near infrared spectroscopy. Analytica Chimica Acta 453, 125-134. doi:10.1016/ S0003-2670(01)01506-9
Franzmeier DP (1988) Relation of organic matter content to texture and colour of Indiana soils. Proceedings of the Indiana Academy of Sciences 98, 463-471.
Genot V, Colinet G, Bock L, Vanvyve D, Reusen Y, Dardenne P (2011) Near infrared reflectance spectroscopy for estimating soil characteristics valuable in the diagnosis of soil fertility. Journal of Near Infrared Spectroscopy 19, 117-138. doi: 10.1255/jnirs.923
Gillman GP, Sinclair DF, Beech TA (1986) Recovery of organic carbon by the Walkley and Black procedure in highly weathered soils. Communications in Soil Science and Plant Analysis 17, 885-892. doi: 10.1080/00103628609367759
Glumac NG, Dong WK, Jarrell M (2010) Quantitative analysis of soil organic carbon using laser-induced breakdown spectroscopy: an improved method. Soil Science Society of America Journal 74, 1922-1928. doi: 10.2136/sssaj2010.0100
Gomez C, Viscarra Rossel RA, McBratney A (2008) Soil organic carbon prediction by hyperspectral remote sensing and field vis-NIR spectroscopy: an Australian case study. Geoderma 146, 403-411. doi: 10.1016/j.geoderma.2008.06.011
Gomez-Robledo L, Lopez-Ruiz N, Melgosa M, Palma AJ, Fermin Capitan-Vallvey L, Sanchez-Maranon M (2013) Using the mobile phone as Munsell soil-colour sensor: an experiment under controlled illumination conditions. Computers and Electronics in Agriculture 99, 200-208. doi: 10.1016/j.compag.2013.10.002
Gondal MA, Maganda YW, Dastageer MA, Al Adel FF, Naqvi AA, Qahtan TF (2014) Detection of the level of fluoride in the commercially available toothpaste using laser induced breakdown spectroscopy with the marker atomic transition line of neutral fluorine at 731.1 nm. Optics & Laser Technology 57, 32-38. doi:10.1016/j.optlastec.2013.09.035
Grewal KS, Buchan GD, Sherlock RR (1991) A comparison of three methods of organic carbon determination in some New Zealand soils. Journal of Soil Science 42, 251-257. doi:10.1111/j. 1365-2389.1991. tb00406.x
Guerrero C, Zomoza R, Gomez I, Mataix-Beneyto J (2010) Spiking of NIR regional models using samples from target sites: effect of model size on prediction accuracy. Geoderma 158, 66-77. doi:10.1016/j.geoderma. 2009.12.021
Harmon RS, DeLucia FC, McManus CE, McMillan NJ, Jenkins TF, Walsh ME, Miziolek A (2006) Laser-induced breakdown spectroscopy: an emerging chemical sensor technology for real-time field-portable, geochemical, mineralogical, and environmental applications. Applied Geochemistry 21, 730-747. doi:10.10!6/j.apgeochem.2006.02.003
Heanes DL (1984) Determination of total organic-C in soils by an improved chromic acid digestion and spectrophotometric procedure. Communications in Soil Science and Plant Analysis 15, 1191-1213. doi: 10.1080/00103628409367551
Henderson TL, Baumgardner MF, Franzmeier DP, Stott DE, Coster DC (1992) High dimensional reflectance analysis of soil organic matter. Soil Science Society of America Journal 56, 865-872. doi: 10.2136/ sssaj1992.03615995005600030031x
Heron G, Barcelona MJ, Andersen ML, Christensen TH (1997) Determination of nonvolatile organic carbon in aquifer solids after carbonate removal by sulfurous acid. Ground Water 35, 6-11. doi: 10.1111/j. 1745-6584.1997.tb00053.x
Hunt GR (1977) Spectral signatures of particulate minerals in the visible and near infrared. Geophysics 42, 501-513. doi: 10.1190/1.1440721
Ibanez-Asensio S, Marques-Mateu A, Moreno-Ramon H, Balasch S (2013) Statistical relationships between soil colour and soil attributes in semiarid areas. Biosystems Engineering 116, 120-129. doi: 10.1016/j. biosystemseng.2013.07.013
Ichoku C, Przyborski P (2014) Hyperion. Available at: http:// earthobservatory.nasa.gov/Features/E01Tenth/page3.php (accessed 25 July 2014).
Irons JR (2014) Landsat Science: The Multispectral Scanner System. Available at: http://landsat.gsfc.nasa.gov/?p=3227 (accessed 23 May 2014).
Islam K, Singh B, McBratney A (2003) Simultaneous estimation of several soil properties by ultra-violet, visible, and near-infrared reflectance spectroscopy. Australian Journal of Soil Research 41, 1101-1114. doi: 10.1071/SR02137
Janik LJ, Skjemstad JO (1995) Characterisation and analysis of soils using mid-infrared partial least squares: II. Correlations with some laboratory data. Australian Journal of Soil Research 33, 637-650. doi: 10.1071/SR9950637
Janik LJ, Merry RH, Skjemstad JO (1998) Can mid infrared diffuse reflectance analysis relace soil extractions? Australian Journal of Experimental Agriculture 38, 681-696. doi:10.1071/EA97144
Jenkinson DS (1990) The turnover of organic carbon and nitrogen in soil. Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences 329, 361-368. doi: 10.1098/rstb. 1990.0177
Jonas JL, Wilson GWT, White PM, Joem A (2007) Consumption of mycorrhizal and saprophytic fungi by Collembola in grassland soils. Soil Biology & Biochemistry 39,2594-2602. doi: 10.1016/j.soilbio.2007. 05.004
Kerven GL, Menzies NW, Geyer MD (2000) Analytical methods and quality assurance. Communications in Soil Science and Plant Analysis 31, 1935-1939. doi: 10.1080/00103620009370551
Kogel-Knabner I (2002) The macromolecular organic composition of plant and microbial residues as inputs to soil organic matter. Soil Biology dc Biochemistry 34, 139-162. doi:10.1016/S0038-0717(01)00158-4
Konen ME, Jacobs PM, Burras CL, Talaga BJ, Mason JA (2002) Equations for predicting soil organic carbon using loss-on-ignition for north central U.S. soils. Soil Science Society of America Journal 66, 1878-1881. doi: 10.2136/sssaj2002.1878
Konen ME, Burras CL, Sandor JA (2003) Organic carbon, texture, and quantitative color measurement relationship for cultivated soils in north central Iowa. Soil Science Society of America Journal 67, 1823-1829. doi: 10.2136/sssaj2003.1823
Kosaka J, Honda C, Iseki A (1959) A new rapid and accurate method for the determination of carbon in soils. Soil Science and Plant Nutrition 5, 77-83. doi: 10.1080/00380768.1959.10430898
Kramer C, Gleixner G (2008) Soil organic matter in soil depth profiles: distinct carbon preferences of microbial groups during carbon transformation. Soil Biology & Biochemistry 40, 425-433. doi: 10.1016/j.soilbio.2007.09.016
Krishnan P, Alexander JD, Butler BJ, Hummel JW (1980) Reflectance technique for predicting soil organic matter. Soil Science Society of America Journal 44, 1282-1285. doi:10.2136/sssajl980.0361599500 4400060030x
Lahijani P, Zainal ZA, Mohammadi M, Mohamed AR (2015) Conversion of the greenhouse gas C[0.sub.2] to the fuel gas CO via the Boudouard reaction: a review. Renewable & Sustainable Energy Reviews 41, 615-632. doi: 10.1016/j.rser.2014.08.034
Lal R, Follett RF, Stewart BA, Kimble JM (2007) Soil carbon sequestration to mitigate climate change and advance food security. Soil Science 172, 943-956. doi: 10.1097/ss.0b013e31815cc498
LECO (2013) TruMac Series. Available at: www.leco.com/products/analytical-sciences/ carbon-nitrogen-protein-sulfur/trumacseries#features (accessed 10 July 2014).
Lettens S, de Vos B, Quataert P, van Wesemael B, Muys B, van Orshoven J (2007) Variable carbon recovery of Walkley-Black analysis and implications for national soil organic carbon accounting. European Journal of Soil Science 58, 1244-1253. doi:10.1111/j. 1365-2389. 2007.00916.x
MacCarthy P, Malcolm RL, Clapp CE, Bloom PR (1990) An introduction to soil humic substances. In 'Humic substances in soil and crop sciences: Selected readings'. (Ed. P MacCarthy) pp. 1-12. (American Society of Agronomy and Soil Science Society of America: Madison, WI)
McCarty GW, Reeves JB III, Reeves VB, Follett RF, Kimble JM (2002) Mid-infrared and near-infrared diffuse reflectance spectroscopy for soil carbon measurements. Soil Science Society of America Journal 66, 640-646. doi: 10.2136/sssaj2002.6400
Martin MZ, Mayes MA, Heal KR, Brice DJ, Wullschleger SD (2013) Investigation of laser-induced breakdown spectroscopy and multivariate analysis for differentiating inorganic and organic C in a variety of soils. Spectrochemica Acta Part B 87, 100-107. doi: 10.1016/j.sab.2013.05.026
Matsuda K, Schnitzer M (1972) The permanganate oxidation of humic acids extracted from acid soils. Soil Science 114, 185-193. doi: 10.1097/00010694-197209000-00005
Matthiessen MK, Larney FJ, Selinger LB, Olson AF (2005) Influence of loss-on-ignition temperature and heating time on ash content of compost and manure. Communications in Soil Science and Plant Analysis 36, 2561-2573. doi: 10.1080/00103620500257242
Mebius LJ (1960) A rapid method for the determination of organic carbon in soil. Analytica Chimica Acta 22, 120-124. doi: 10.1016/S0003-2670(00) 88254-9
Metz B, Davidson OR, Bosch PR, Dave R, Meyer LA (2007) Climate Change 2007: Working Group III: Mitigation of Climate Change (IPCC). Available at: www.ipcc.ch/publications_and_data/ar4/wg3/en/tssts-ts-83-mitigation-technologies.html (accessed 23 July 2014).
Morra MJ, Hall MH, Freeborn LL (1991) Carbon and nitrogen analysis of soil fractions using near-infrared reflectance spectroscopy. Soil Science Society of America Journal 55, 288-291. doi: 10.2136/sssaj 1991. 03615995005500010051x
Mosier-Boss PA, Liebermann SH, Theriault GA (2002) Field demonstrations of a direct push FO-LIBS metal sensor. Environmental Science <6 Technology 36, 3968-3976. doi:10.1021/es020528z
Mouazen AM, Karoui R, Deckers J, De Baerdemaeker J, Ramon H (2007a) Potential of visible and near-infrared spectroscopy to derive colour groups utilising the Munsell soil colour charts. Biosystems Engineering 97, 131-143. doi:10.1016/j.biosystemseng.2007.03.023
Mouazen AM, Maleki MR, De Baerdemaeker J, Ramon H (20076) On-line measurement of some selected soil properties using a VIS-NIR sensor. Soil & Tillage Research 93, 13-27. doi:10.1016/j.still.2006.03.009
Munsell (2013) Official Site of Munsell Color. Available at: http://munsell. com/about-munsell-color/how-color-notation-works/ (accessed 23 July 2014).
Nelson DW, Sommers LE (1974) A rapid and accurate procedure for estimation of organic carbon in soils. Proceedings of the Indiana Academy of Science 84, 456 462.
Ngo P-T, Rumpel C, Doan T-T, Jouquet P (2012) The effect of earthworms on carbon storage and soil organic matter composition in tropical soil amended with compost and vermicompost. Soil Biology & Biochemistry 50, 214-220. doi: 10.1016/j.soilbio.2012.02.037
Oades JM (1988) The retention of organic matter in soils. Biogeochemistry 5, 35-70. doi: 10.1007/BF02180317
Olive PL (1998) The role of DNA single and double strand breaks in cell killing by ionizing radiation. Radiation Research 150, S42-S52. doi: 10.2307/3579807
Olson K.R, Lang JM, Ebelhar SA (2005) Soil organic carbon changes after 12 years of no-tillage and tillage of Grantsburg soils in southern Illinois. Soil & Tillage Research 81, 217-225. doi: 10.1016/j.still.2004.09.009
Parsons A, Bodnarik J, Evans L, Floyd S, Lim L, McCanahan T, Namkung M, Nowicki S, Schweitzer J, Starr R, Trombka J (2011) Active neutron and gamma-ray instrumentation for in situ planetary science applications. Nuclear Instruments & Methods in Physics Research. Section A, Accelerators, Spectrometers, Detectors and Associated Equipment 652, 674-679. doi: 10.1016/j.nima.2010.09.157
Randall RB, Benger M, Groocock CM (1938) The alkaline permanganate oxidation of organic substances selected for their bearing upon the chemical constitution of coal. Proceedings of the Royal Society of London. Series A 165, 432-452. doi:10.1098/rspa,1938.0068
Rayment GE, Higginson FR (1992) 'Australian laboratory handbook of soil and water chemical methods.' (Inkata Press: Melbourne)
Rayment GE, Lyons DJ (2011) 'Soil chemical methods - Australasia.' (CSIRO Publishing: Melbourne)
Reeves JB III (2010) Near- versus mid-infrared diffuse reflectance spectroscopy for soil analysis emphasizing carbon and laboratory versus on-site analysis: where are we and what needs to be done? Geoderma 158, 3-14. doi:10.1016/j.geoderma.2009.04.005
Reeves JB III, McCarty GW (2001) Quantitative analysis of agricultural soils using near infrared reflectance spectroscopy and a fibre-optic probe. Journal of Near Infrared Spectroscopy 9, 25 34. doi: 10.1255/Jnirs.291
Rozantsev VA, Shirokanov AD, Yankovskii AA (1993) Effect of the time interval between single laser pulses on the character of the laser plasma spectrum. Journal of Applied Spectroscopy 59, 797-800. doi: 10.1007/BF00664918
Rusak DA, Castle BC, Smith BW, Winefordner JD (1997) Fundamentals and applications of laser-induced breakdown spectroscopy. Critical Reviews in Analytical Chemistry 27, 257-290. doi: 10.1080/1040 8349708050587
Sankey JB, Brown DJ, Bernard ML, Lawrence RL (2008) Comparing local vs. global visible and near-infrared (VisNIR) diffuse reflectance spectroscopy (DRS) calibrations for the prediction of soil clay, organic C and inorganic C. Geoderma 148, 149-158. doi:10.1016/j.geoderma. 2008.09.019
Schmidt A, Smemik RJ, McBeath TM (2012) Measuring organic carbon in Calcarosols: understanding the pitfalls and complications. Soil Research 50, 397-405. doi: 10.1071 /SR 12134
Schnitzer M, Desjardins JG (1964) Further investigations on the alkaline permanganate oxidation of organic matter extracted from a podzol Bh horizon. Canadian Journal of Soil Science 44, 272-279. doi: 10.4141/ cjss64-041
Schollenberger CJ (1927) A rapid approximate method for determining soil organic matter. Soil Science 24, 65-68. doi :10.1097/00010694192707000-00008
Schollenberger CJ (1945) Determination of soil organic matter. Soil Science 59, 53-56. doi: 10.1097/00010694-194501000-00008
Schulte EE, Hopkins BG (1996) Estimation of soil organic matter by weight loss-on-ignition. In 'Soil organic matter: analysis and interpretation'. (Eds FR Magdoff, MA Tabatabai, EA Hanlon) pp. 21-31. (Soil Science Society of America: Madison, WI)
Schulte EE, Kaufmann C, Peter JB (1991) The influence of sample size and heating time on soil weight loss-on-ignition. Communications in Soil Science and Plant Analysis 22, 159-168. doi: 10.1080/0010362910 9368402
Selige T, Bohner J, Schmidhalter U (2006) High resolution topsoil mapping using hyperspectral image and field data in multivariate regression modeling procedures. Geoderma 136, 235-244. doi: 10.1016/j. geoderma.2006.03.050
Shields JA, Paul EA, St Arnaud RJ, Head WK (1968) Spectrophotometry measurement of soil color and its relationship to moisture and organic matter. Canadian Journal of Soil Science 48, 271 280. doi: 10.4141/cjss68-037
Sims JR, Haby VA (1971) Simplified colorimetric determination of soil organic matter. Soil Science 112, 137-141. doi:10.1097/00010694197108000-00007
Singh K, Murphy BW, Marchant BP (2012) Towards cost-effective estimation of soil carbon stocks at the field scale. Soil Research 50, 672-684. doi: 10.1071/SR12119
Skalar (2014) Primacs series total organic carbon and total carbon analyzers. Available at: www.skalar.com/analyzers/total-organic-carbon-toc-andtotal-carbon-tc-analyzers/ (accessed 9 July 2014).
Skjemstad JO, Baldock JA (2008) Soil sampling and methods of analysis. In 'Total and organic carbon'. 2nd edn. (Eds MR Carter, EG Gregorich) pp. 225-237. (Canadian Society of Soil Science)
Skjemstad JO, Taylor JA, Smemik RJ (1999) Estimation of charcoal (char) in soils. Communications in Soil Science and Plant Analysis 30, 2283-2298. doi : 10.1080/00103629909370372
Skjemstad JO, Spouncer LR, Cowie B, Swift RS (2004) Calibration of the Rothamsted organic carbon turnover model (RothC ver. 26.3), using measurable soil organic carbon pools. Soil Research 42, 79-88. doi: 10.1071 /SR03013
Smith T, Guild J (1932) The C.I.E. colorimetric standards and their use. Transactions of the Optical Society 33, 73-134.
Snyder JD, Trofymow JA (1984) A rapid accurate wet oxidation diffusion procedure for determining organic and inorganic carbon in plant and soil samples. Communications in Soil Science and Plant Analysis 15, 587-597. doi: 10.1080/00103628409367499
Soon YK, Abboud S (1991) A comparison of some methods for soil organic carbon determination. Communications in Soil Science and Plant Analysis 22, 943-954. doi: 10.1080/00103629109368465
Steinhardt GC, Franzmeier DP (1979) Comparison of organic matter content with soil color for silt loam soils of Indiana. Communications in Soil Science and Plant Analysis 10, 1271-1277. doi:10.1080/0010362790 9366981
Stevens A, van Wesemael B, Bartholomeus H, Rosillon D, Tychon B, BenDor E (2008) Laboratory, field and airborne spectroscopy for monitoring organic carbon content in agricultural soils. Geoderma 144, 395-404. doi: 10.1016/j.geoderma.2007.12.009
Stevens A, Udelhoven T, Denis A, Tychon B, Lioy R, Hoffman L, van Wesemael B (2010) Measuring soil organic carbon in croplands at regional scale using airborne imaging spectroscopy. Geoderma 158, 32-45. doi:10.1016/j.geoderma.2009.11.032
Tabatabai MA, Bremner JM (1970) Use of the LECO automatic 70-second carbon analyzer for total carbon analysis in soils. Soil Science Society of America Journal 34, 608-610. doi:10.2136/sssaj1970.0361599500 3400040020x
Tam NFY, Yao MWY (1998) An accurate, simple and novel analytical method for the determination of total organic carbon in sediment. International Journal of Environmental Analytical Chemistry 72, 137-150. doi: 10.1080/03067319808035885
Tisdall JM, Oades JM (1979) Stabilization of soil aggregates by the root systems of ryegrass. Australian Journal of Soil Research 17, 429-441. doi: 10.1071/SR9790429
Tisdall JM, Oades JM (1982) Organic matter and water-stable aggregates in soils. Journal of Soil Science 33, 141-163. doi: 10.1111/j. 1365-2389. 1982.tb01755.x
USGS (2013) Landsat Missions Timeline. Available at: http://landsat.usgs. gov/about_mission_history.php (accessed 23 July 2014).
Velasquez E, Pelosi C, Brunet D, Grimaldi M, Martins M, Rendeiro AC, Barrios E, Lavelle P (2007) This ped is my ped: visual separation and near infrared spectra allow determination of the origins of soil macroaggregates. Pedobiologia 51, 75-87. doi:10.1016/j.pedobi.2007. 01.002
Viscarra Rossel RA, Webster R (2012) Predicting soil properties from the Australian soil visible-near infrared spectroscopic database. European
Journal of Soil Science 63, 848-860. doi: 10.1111/j. 1365-2389.2012. 01495.x
Viscarra Rossel RA, Minasny B, Roudier P, McBratney AB (2006a) Color space models for soil science. Geoderma 133, 320-337. doi: 10.1016/j.geoderma.2005.07.017
Viscarra Rossel RA, Walvoort DJJ, McBratney A, Janik LJ, Skjemstad JO (20066) Visible, near infrared, mid infrared or combined diffuse reflectance spectroscopy for simultaneous assessment of various soil properties. Geoderma 131, 59-75. doi: 10.1016/j.geoderma.2005.03.007
Viscarra Rossel RA, Fouad Y, Walter C (2008) Using a digital camera to measure soil organic carbon and iron contents. Biosystems Engineering 100, 149-159. doi:10.1016/j.biosystemseng.2008.02.007
Walkley A (1935) An examination of methods for determining organic carbon and nitrogen in soils. The Journal of Agricultural Science 25, 598-609. doi: 10.1017/S0021859600019687
Walkley A (1947) A critical examination of a rapid method for determining organic carbon in soils: effect of variations in digestion conditions and of inorganic soil constituents. Soil Science 63, 251-264. doi:10.1097/00010694-194704000-00001
Walkley A, Armstrong Black I (1934) An examination of the Degtjareff method for determining soil organic matter, and a proposed modification of the chromic acid titration. Soil Science 37, 29-38. doi: 10.1097/00010694-193401000-00003
Weil RR, Islam KR, Stine MA, Gruver JB, Samson-Liebig SE (2003) Estimating active carbon for soil quality assessment: a simplified method for laboratory and field use. American Journal of Alternative Agriculture 18, 3-17. doi: 10.1079/AJAA2003003
Wetzel DL (1983) Near-infrared reflectance analysis. Analytical Chemistry 55, 1165A-1176A. doi:10.1021/ac00262a718
Wielopolski L, Hendrey G, Johnsen KH, Mitra S (2008) Nondestructive system for analyzing carbon in the soil. Soil Science Society of America Journal 12, 1269-1277. doi:10.2136/sssaj2007.0177
Wielopolski L, Yanai RD, Levine CR, Mitra S, Vadeboncoeur A (2010) Rapid, non-destructive carbon analysis of forest soils using neutron-induced gamma-ray spectroscopy. Forest Ecology and Management 260, 1132-1137. doi: 10.1016/j.foreco,2010.06.039
Wielopolski L, Chatteijee A, Mitra S, Lai R (2011) In situ determination of soil carbon pool by inelastic neutron scattering: comparison with dry combustion. Geoderma 160,394-399. doi: 10.1016/j.geoderma.2010.10.009
Wright WD (1929) A re-determination of the trichromatic coefficients of the spectral colours. Transactions of the Optical Society 30, 141-164. doi: 10.1088/1475-4878/30/4/301
Zimmermann M, Liefeld J, Schmidt MWI, Smith P, Fuhrer J (2007) Measured soil organic matter fractions can be related to pools in the RothC model. European Journal of Soil Science 58, 658-667. doi: 10.1111/j.1365-2389.2006.00855.x
Timothy J. Johns (A,B), Michael J. Angove (A), and Sabine Wilkens (A)
(A) La Trobe University, School of Pharmacy and Applied Science, PO Box 199, Bendigo, Vic. 3552, Australia.
(B) Corresponding author. Email: T.Johns@latrobe.edu.au
Table 1. Calibration and recovery data for key research on various total soil organic carbon (TOC) methods CIE, Commission Internationale de l'Eclairage; CV, coefficient of variation; DC, dry combustion; INS, inelastic neutron scattering; IR, infrared; LIBS, laser-induced breakdown spectroscopy; LOI, loss-on-ignition; MIR, mid infrared; N1R, near infrared; OM, organic matter; PLSR, partial least-squares regression; RMSE, root mean square error; Std ref, standard reference; WB, Walkley-Black method Method n Correlation coefficient ([r.sup.2]) or available data DC Tabatabai and Bremner 10 s.d. <0.03% C (except Glencoe (1970) soils/peat) Soon and Abboud (1991) 39 4.9% CV Gillman et al. (1986) 49 2.9% CV LOI (Konen et al. 2002) 68 0.98 LOI (Ben-Dor and Banin 1989) 93 0.945 Dichromate oxidation 20 60%-86% recovery, [+ or -] (Walkley and Armstrong 5.6% s.e. 0 Dichromate oxidation (Tam 15 Std ref materials 100 and Yao 1998) [+ or -] 4% 30 Marine sediments 101 [+ or -] 6% Dichromate oxidation Heanes (1984) 12 100 [+ or -] 2% recovery Conyers et al. (2011) 26 98.7% recovery, s.e. 0.019 Colour (Munsell; Steinhardt 262 Characterised into <3%, and Franzmeier 1979) 3%-5% and >5% OM; ~90% accuracy Colour (digital camera; 69 0.92, RMSE = 0.42, Viscarra Rossel calibration et al. 2008) 34 0.91, RMSE = 0.48, validation MIR (Baldock et al. 2013a) 13 100 0.93, RMSE = 0.42, calibration 6626 0.93, RMSE = 0.42, validation MIR (McCarty et al. 2002) 177 0.96, calibration 60 0.94, validation NIR (Reeves and McCarty 180 0.94, validation 2001) NIR (Fidencio et al. 2002) 60 0.92 LIBS (Belkov et al. 2009) 11 0.97, single laser pulse 8 0.97, double laser pulse LIBS (Martin et al. 2013) 29 0.98, RMSE = 3.9% 29 0.92, RMSE = 8.4% INS (Wielopolski 5 [+ or -] 4.5% to -9.4%, et al. 2011) field C >10% 5 [+ or -] 52% to -21%, field C <5% Airborne remote sensing 31 0.98 (Chen et al. 2000) Airborne remote sensing 12 0.89 (Selige et al. 2006) Method Standard Associated information method DC Tabatabai and Bremner DC with NaOH DC widely accepted as (1970) trap the reference method for carbon determinations Soon and Abboud (1991) DC LECO CR-12 with IR detector Gillman et al. (1986) DC LECO CR-12 LOI (Konen et al. 2002) DC (LECO) Correlations between 0.94 and 0.98 for different regions LOI (Ben-Dor and Banin 1989) Modified WB OM (not TOC) Dichromate oxidation DC -- (Walkley and Armstrong 0 Dichromate oxidation (Tam DC Microwave oven used for and Yao 1998) heating Dichromate oxidation Heanes (1984) DC Soils ground to <0.15 mm. Conyers et al. (2011) DC Heanes method used Colour (Munsell; Steinhardt Various OM (not TOC) and Franzmeier 1979) dichromate Colour (digital camera; DC CIE L * a * b * Viscarra Rossel model with full et al. 2008) factorial regression MIR (Baldock et al. 2013a) DC PLSR with nine factors, square root transformation; diverse soils set MIR (McCarty et al. 2002) DC PLSR with 17 factors NIR (Reeves and McCarty DC PLSR with eight factors 2001) NIR (Fidencio et al. 2002) Heated 150 Neural network with [degrees]C radial bias function dichromate LIBS (Belkov et al. 2009) DC Calibration data LIBS (Martin et al. 2013) DC Calibration data INS (Wielopolski DC Calibration data et al. 2011) Airborne remote sensing DC Aerial photograph (Chen et al. 2000) converted to digital; colour-based measurement Airborne remote sensing DC Aircraft-based (Selige et al. 2006) spectrometer measurement (HyMap) Table 2. Summary table of total soil organic carbon (TOC) measurement considerations and opportunities DC, dry combustion; INS, inelastic neutron scattering; IR, infrared; LIBS, laser-induced breakdown spectroscopy; LOI, loss-on-ignition; MIR, mid infrared; NIR, near infrared; OC, organic carbon; ROC, readily oxidisable carbon; TC, total carbon; WB, Walkley-Black method Method & TOC portion measured Interferences section DC Measures TC Carbonates (4.1.1) Pretreatment required for Depending on detection measurement of TOC method (e.g. water for IR) OC can be lost in treatment for carbonates Equipment is expensive LOI Dependent on other Carbonates at (4.1.2) standard method and temperatures quality of calibration >400[degrees]C Water from clay and mineral structures WB Variable amounts of OC Chlorides (4.2.1) depending on origin Sparingly measures burnt Oxides of Mn or carbonised materials Minimal measurement of [Fe.sup.2+] inorganic carbon Heanes Measurement of all forms Chlorides (4.2.1) of TOC, including charcoals; shows good agreement with DC (non-calcareous soils) Oxides of Mn [Fe.sup.2+] Colour Dependent on other Carbonates (white) (Munsell) standard method used (5.1) for calibration Unlikely to measure less Black minerals humified or fresher carbon due to colour Colour-- Dependent on other Carbonates (white) digital standard method used camera for calibration (5.1) Unlikely to measure less Black minerals humified or fresher carbon due to colour Non-humified OC Charred materials MIR Dependent on other Cal. Bias. (5.2.1) standard method and quality of calibration Quartz Kaolin Carbonates NIR Dependent on other Non-TOC soil substrates (5.2.2) standard method and absorbing in same quality of calibration regions LIBS Measures TC Carbonates (6.1.2) Some attempts made at Iron differentiation Dependent on other Water standard method for calibration INS Measures TC Carbonates (6.1.3) Airborne Surface TOC, as per Meteorological conditions remote colour, NIR, MIR sensing (6.2) As per colour, NIR Method & Issues Research opportunities section DC Requires pretreatment Carbonate removal (4.1.1) for carbonates pretreatments that minimise loss of OC Saline soils Resource intensive Destructive testing LOI Definition of best As per issues (4.1.2) conditions may be soil dependent Incomplete oxidation Defining best approaches of OC in terms of all the variables for different soil types No. possible interferences and uncertainties Destructive testing Local calibrations Time taken for drying/ ashing steps WB Use of toxic chemicals Further exploration of the (4.2.1) microwave heating step as per Tam and Yao (1998) Disposal of wastes Destructive testing Incomplete oxidation Definition of conditions for a readily oxidisable carbon measure based on WB method Sparing and variable measurement of charred materials Heat of reaction reached dependent on ambient conditions Heanes Use of toxic chemicals None identified (4.2.1) Disposal of wastes Destructive testing Minimal measurement of inorganic carbon Colour Non-continuous colour None identified (Munsell) chips (5.1) Modem colour models more suitable or compatible with instmmental techniques Colour-- Unlikely to measure less Developing simple digital humified OC calibrations for local camera conditions (5.1) Depends on another Phone camera applications standard method to derive calibration Local or soil type Developing more complex calibrations are calibrations to account required for different source OC contributions to soil MIR Extent of calibration Calibrations or techniques (5.2.1) samples required to to better handle set up analysis unknown origin samples Inability to deal with unknown origin samples NIR Extent of calibration Improvement of (5.2.2) samples required to laboratory-based set up analysis calibrations and measurements Inability to deal with Infield analyses (e.g. unknown origin aboard drone) samples LIBS Small soil volume Development as on-the-go (6.1.2) analysed, so non- or in situ measurement homogeneous soils will because of rapid data require multiple collection possibilities emission spectra Destructive technique Improved calibrations to overcome on-the-go or in situ interferences Possible production of X-rays depending on laser set-up INS Results thus far better Understanding of effects (6.1.3) on high-carbon soils on soil biota (10%-40% C) Gamma radiation Proving of the technique interaction with soil in different conditions life on broad-scale use Requires separate analysis for TOC determination Airborne Correct pixel Finding best sensors for remote resolution at most accurate analysis sensing ground level, avoid (6.2) multiple surfaces Specificity of equipment Increased ability to use (satellite) over large areas with sparse vegetation Only able to measure Use of drones for close-to- surface soils ground analysis Requires soil surfaces to be bare of vegetation, limiting widespread use Method & Best current applications References section DC For large numbers Tabatabai and Bremner (4.1.1) of samples (1970) Non-calcareous soils Schmidt et al. (2012); or soils without Skjemstad and added limes Baldock (2008) Schmidt et al. (2012) LOI Where equipment levels Konen et al. (2002) (4.1.2) are low but access to furnace at 400[degrees]C is possible Saline soils Ben-Dor and Banin (1989) Repeat analyses on the Schulte and Hopkins same soils where strong (1996) calibration relationship can be established WB Use as a measure of soil Walkley and Armstrong (4.2.1) ROC without correction Black (1934) Walkley (1947) Conyers et al. (2011) Where a quick Tam and Yao (1998) approximate assessment of soil TOC is required Where labour is cheap Where equipment budget is low or less toxic alternatives are not available Heanes Highly calcareous soils Heanes (1984) (4.2.1) To track changes in TOC Conyers et al. (2011) at same site Where equipment budget Walkley and Armstrong is low or less toxic Black (1934) alternatives not available Colour Quick field assessment Franzmeier (1988) (Munsell) (5.1) Where accuracy not as Steinhardt and important Franzmeier (1979) Where access to suitable Ibanez-Asensio et al. equipment is lacking (2013) Where non-destructive Baumgardner et al. testing is important (1985); Tabatabai and Bremner (1970) Colour-- Requires developing as a Konen et al. (2002); digital technique Viscarra Rossel camera et al. (2008) (5.1) Shows some promise for rapid determinations Where budget is low and accuracy not supremely important Where non-destructive testing is important MIR Where sample numbers Bellon-Maurel and (5.2.1) are high and sample McBratney (2011) preparation (grinding) is not a critical issue For ongoing analysis on Baldock el al. (2013a) similar soil types Viscarra Rossel et ai. (2006b) NIR Where sample numbers Reeves (2010) (5.2.2) are high and the effort to derive a calibration using several known samples from other techniques is warranted or suitable calibration library is available Where accuracy may not Brunet et al. (2007) be as important as through-put For on-going analysis on Ben-Dor and similar soil types as Banin (1995) calibration For development as an on- Viscarra Rossel et al. the-go analysis (2006b) LIBS Non-calcareous soils Belkov et al. (2009) (6.1.2) For rapid analysis of dried, Bricklemyer ground samples et al. (2011) Where equipment da Silva et al. (2008) available Martin et al. (2013) INS On-the-go analysis of total Wielopolski (6.1.3) carbon at field scale et al. (2008) Wielopolski et al. (2010) Wielopolski et al. (2011) Parsons et al. (2011) Falahat et al. (2012) Airborne Broad-scale applications Chen et al. (2000) remote on bare soils (e.g. sensing recently cultivated (6.2) surfaces) Gomez et al. (2008) Selige et al. (2006) Stevens et al. (2010) Stevens et al. (2008)
|Printer friendly Cite/link Email Feedback|
|Author:||Johns, Timothy J.; Angove, Michael J.; Wilkens, Sabine|
|Date:||Oct 1, 2015|
|Previous Article:||Some insights into the health of Dermosols around Tasmania, Australia.|
|Next Article:||Assessment of the accuracy of profile available water and potential rooting depth estimates held within New Zealand's fundamental soil layers...|