Biochar application to soil is currently widely advocated for a variety of reasons related to sustainability. Typically, soil amelioration with biochar is presented as a multiple-‘win’ strategy, although it is also associated with potential risks such as environmental contamination. The most often claimed benefits of biochar (i.e. the ‘wins’) include (i) carbon sequestration; (ii) soil fertility enhancement; (iii) biofuel/bioenergy production; (iv) pollutant immobilization; and (v) waste disposal. However, the vast majority of studies ignore possible trade-offs between them. For example, there is an obvious trade-off between maximizing biofuel production and maximizing biochar production. Also, relatively little attention has been paid to mechanisms, as opposed to systems impacts, behind observed biochar effects, often leaving open the question as to whether they reflect truly unique properties of biochar as opposed to being simply the short-term consequences of a fertilization or liming effect. Here, we provide an outline for the future of soil biochar research. We first identify possible trade-offs between the potential benefits. Second, to be able to better understand and quantify these trade-offs, we propose guidelines for robust experimental design and selection of appropriate controls that allow both mechanistic and systems assessment of biochar effects and trade-offs between the wins. Third, we offer a conceptual framework to guide future experiments and suggest guidelines for the standardized reporting of biochar experiments to allow effective between-site comparisons to quantify trade-offs. Such a mechanistic and systems framework is required to allow effective comparisons between experiments, across scales and locations, to guide policy and recommendations concerning biochar application to soil.
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
Biochar is produced from the thermal degradation of organic material in the absence of oxygen. It differs from charcoal in that it is produced with the intention of application to soil rather than as fuel (Lehmann & Joseph 2009). Biochar is often promoted as having several potential benefits or ‘wins’, including carbon (C) sequestration, soil fertility enhancement, provision of biofuels, pollutant immobilization and disposal of organic wastes, among others. However, in many instances, it is not possible for all benefits to be simultaneously maximized and negative effects may also occur such as priming of soil organic matter (Cross & Sohi 2011; Zimmerman et al. 2011), or the introduction of contaminants into the soil (Chan & Xu 2009). Gains over one timeframe could also turn into losses over the longer term, when considering all aspects. The longevity of biochar implies that negative effects can endure in soil, potentially for thousands of years (Glaser et al. 2002). The type of feedstock and production conditions affect the properties of the resulting biochar and its associated effects (Jeffery et al. 2011; Zimmerman 2010). The question remains whether other effects of biochar application could be as long-lasting as the C storage; clarification of risks and testing across a broad spectrum of soil-crop combinations is vital before large-scale applications can be advocated.
Proponents of biochar deployment frequently draw parallels with historical practices of soil improvement conducted by Amerindians in the Amazon basin (creating the Amazonian Dark Earth or ‘terra preta de Indio’, generally referred to as ‘terra preta’) (Sombroek 1966). Terra preta appears to achieve two ‘wins’ through achieving high levels of fertility compared to surrounding unamended soils, while demonstrably sequestering C in the soil. Extrapolating to the current day, this suggests circumvention of the ‘carbon dilemma’ described by Janzen (2006). Janzen stated that a ‘paradox’ exists with regard to soil organic matter (SOM). SOM stocks should be conserved to sequester C, but at the same time decomposition of SOM is the driving force for increasing overall soil quality through activation of the soil food web and mineralization of nutrients. Terra preta appears to store large amounts of C at the same time as achieving relatively high levels of fertility for which the cycling of SOM is conventionally assumed necessary, thereby supposedly achieving the ‘win–win’ stated earlier.
For biochar precisely the same ‘win–win’ was proposed as was claimed for terra preta (Glaser et al. 2002). This led to the proposition that biochar application to soil outside of the Amazon basin has the same potential to sequester carbon and improve soil fertility in the manner described for terra preta. However, despite the link between biochar and terra preta being almost routinely made in the introductory sections of biochar reports, the justification for this association is often poorly outlined and there is as yet little to no evidence that adding biochar to soils will create terra preta like soils.
Recent work has identified further ‘wins’ that are often associated with biochar. Laird (2008) described biochar as a ‘win–win–win’ technology due to the production of biofuel during biomass pyrolysis. Other potential benefits or ‘wins’ include the suppression of greenhouse gas emission such as N2O and CH4 from soils (Karhu et al. 2011; Cayuela et al., 2013a, 2013b); the remediation of contaminated soils as biochar strongly binds to most organic pollutants (Cornelissen et al. 2005); and waste disposal (Cascarosa et al. 2013).
In this article, we consider the potential benefits that are associated with biochar before discussing why it is not possible for all such benefits to be simultaneously maximized and some negative effects may occur; trade-offs, therefore, are inevitable. We then provide a conceptual framework to allow identification of the potential benefits of biochar application in a given situation and hence the trade-offs between them. Finally, we make a call for standardized experimental techniques and reporting of results to allow robust and policy-relevant judgements to be made. Key to any advances is to acknowledge that biochars often have very different properties (Schimmelpfennig & Glaser, 2012), to the extent that the characterization of a material as ‘biochar’ can be insufficient and must be connected to a description of the biochar that has been studied or discussed. This critical point should be kept in mind when the term ‘biochar’ is used in the following sections.
To conceptualize the trade-offs associated with different uses of biochar, we present a graphical framework that is focused on the most frequently reported potential benefits of biochar (Fig. 1). This framework will aid the identification of the best biochar to apply to a soil in a given situation. The identification of trade-offs could result in maximization of a particular benefit and/or minimization of trade-offs. In combination with life-cycle assessment and other information to determine how to weigh multiple biochar benefits and trade-offs, optimization of feedstock, pyrolysis technology and application modes should be possible (Roberts et al. 2010; Sparrevik et al. 2013). While we focus here on trade-offs between the most often reported benefits of biochar application, we acknowledge that there are other potential benefits that are not explicitly considered here (e.g., water retention, effects on soil biodiversity through increased refugia, decreasing greenhouse gas emission). Furthermore, currently the axes on Fig. 1 are qualitative owing to insufficient data to produce quantitative axes. There is currently no single metric by which trade-offs can be quantified. Our aim here was to present the concept of trade-offs and the figure is a way of conceptualizing them. Furthermore, the figure will help to direct future research as researchers can aim to produce the necessary data and analyses to allow conversion of the axes from qualitative to quantitative.
Here, we present three different scenarios, represented by Fig 1 a–c. Figure 1a shows an idealized biochar where all five of the most often reported benefits or ‘wins’ are maximized (solid black line). Figure 1b shows that when biochar is produced for maximum soil fertility effects, this trades-off against biofuel production for reasons discussed below. The residence time of such ‘agronomic’ biochars in the soil are relatively high and so they still count as a win for climate change mitigation. Figure 1c shows a biochar made from waste products (solid black line) that maximizes the waste disposal win but at a cost of not maximizing soil fertility effects as much as may be possible if a different feedstock was used. As stated above, these figures are only conceptual and there are currently insufficient data to make the axes quantitative. This will change as new data become available.
Climate change mitigation vs. soil fertility
A substantial proportion of biochar can remain un-decomposed in the soil over centuries to millennia (Spokas 2010; Zimmerman 2011). This implies that biochar has the potential to mitigate climate change through sequestering C for extended periods, in addition to other mechanisms that reduce greenhouse gas emissions (see below). However, the timeframe over which C remains sequestered in the soil is uncertain as biochar properties vary and depend on feedstock type and processing conditions (e.g., fast vs. slow pyrolysis, pyrolysis temperature) (Spokas 2010; Zimmerman 2010), local climatic conditions, soil type and native soil biota. The main factor affecting the turnover rate of uncharred C in soils is its interaction with the organo-mineral fraction of the soil that can lead to physical and physico-chemical stabilization (Liang et al. 2008; Schmidt et al. 2011; Dungait et al. 2012). However, charring confers a greater stability of residues. Direct oxidative ageing methods that compare biochar of various types with natural char in soil suggest that >60% of the C in fresh biochar will remain after 100 years (Cross & Sohi 2013). The chemical composition of the biochar must, therefore, play some role. It is yet to be determined whether the stabilization mechanisms proposed by Schmidt et al. (2011), such as physical disconnection and organo-mineral associations, apply to all instances of biochar and their relative importance.
In addition to C sequestration, other effects on the soil greenhouse gas balance have been reported (Sohi et al. 2009). Biochar addition resulted in reductions in methane (CH4) emission from the soil, possibly through increasing soil aeration (Rondon et al. 2005; Karhu et al. 2011). Furthermore, biochar decreased nitrous oxide (N2O) emissions from soils (e.g., Schouten et al. 2012) via a number of mechanisms (Singh et al. 2010; Cayuela et al., 2013a, 2013b). For example, biochar addition might change the microbial community including N2O-producing organisms, or alter soil structure and thereby the anaerobic volume of the soil in which denitrification takes place (Van Zwieten et al. 2009). It may also cause a shift of the product ratio N2O: N2 towards N2 after acid neutralization by alkaline biochar (Dörsch et al. 2012), probably alleviating the posttranscriptional inhibition of N2O reductase in acid soils (Liu et al. 2010). Biochar has also been suggested to function as an ‘electron shuttle’ that facilitates the transfer of electrons to denitrifying microorganisms (Cayuela et al., 2013a, 2013b). A last possible mechanism is that biochar could sorb N2O sufficiently to suppress emissions (Van Zwieten et al. 2009; Cornelissen et al. 2013a). However, it is important to note that the time frame over which these effects persist remains to be determined. Such information will be vital for inclusion in life-cycle assessments that aim to quantify trade-offs.
Increased soil fertility has been reported following biochar application with results being highly variable (Jeffery et al. 2011; Biederman & Harpole, 2013). Elucidation of several of the mechanisms behind these observed effects, and quantification of their longevity, is still required. Evidence suggests that the underlying mechanisms may include: increased cation exchange capacity (CEC) (Liang et al. 2006); increased plant-available water contents (Karhu et al. 2011), improved drainage of excess water (Ayodele et al. 2009); a liming effect in acidic soils (Yamato et al. 2006) or acid neutralization through the addition of organic anions in biochars produced at low temperatures (Yuan et al. 2011); the presence of available nutrients in the biochar (Angst & Sohi, 2013); and increased abundance of soil microbes (O'Neill et al. 2009; Liang et al. 2010), including mycorrhizal fungi (Warnock et al. 2007; Solaiman et al. 2010) and decomposers (Zackrisson et al. 1996).
The time range over which these effects operate varies considerably. Liming effects and direct nutrient addition effects are likely transient as nutrients are utilized or leached from the system. Such effects likely result from the addition of ash inclusions with the biochar (Mukherjee et al. 2011) rather than from the biochar C itself. Other effects may be longer lived, but slower to develop, such as increased CEC and associated nutrient-binding effects through surface oxidation of biochar particles (Liang et al. 2006) or increased plant-available water due to the high porosity of biochar particles increasing the water-holding capacity of soils (Karhu et al. 2011).There is evidence that some types of biochar have phytotoxic effects depending on the original feedstock and temperature of pyrolysis (Gell et al. 2011), but evidence also exists that charcoal can adsorb and inactivate phytotoxic compounds (Hille and Ouden 2005).
It is probable that there will be a trade-off between these two established ‘wins’. This could be the definitive case of ‘hoarding’ rather than ‘using’ in the context of Janzen's ‘Carbon Dilemma’ (Janzen 2006). As mentioned above, terra preta apparently achieves beneficial results with regard to both C sequestration and maintenance of enhanced levels of fertility compared to surrounding soils (Glaser et al. 2002). However, whether such concurrent benefits can also be achieved with biochar addition to soil, and at what level the benefits may trade-off against each other, remains to be determined, particularly in the long term and in temperate regions. For example, Quilliam et al. (2012) reported that double dosing and extra loading of biochar in a temperate field plot only provided transient effects on soil fertility over 4 years. However, Mao et al. (2012) found that the relatively high CEC of Mollisols in Iowa were due to the high black C contents of the soils. This suggests that biochar (which contains large proportions of black C, its defining chemical feature) has the potential to increase soil fertility through CEC effects, even in temperate soils. It should be noted, however, that mollisols have a high pH with relatively high Ca and P contents. Hence, it is possible that this is an exceptional case. A trade-off may occur between producing biochar that maximizes C sequestration potential vs. biochars with desired agronomic properties, due to the use of oxidation to ‘age’ biochars (i.e. accelerate the formation of biochar properties that develop over time in the soil) and increase their CEC. Such increased CECs will help the biochar to adsorb cations and reduce nutrient leaching, but the oxidation process leads to a loss of C, thereby reducing C sequestration potential. Evidence suggests that oxidation of biochars may concur with reduced recalcitrance of the biochar in soil, further reducing the biochar's C sequestration potential (Nguyen et al. 2010).
Biochar has been reported to cause priming of SOM (both positive and negative), specifically over short periods of time (Steinbeiss et al. 2009; Zimmerman et al. 2011; but see Cross & Sohi, 2011; Jones et al. 2012). This suggests another potential trade-off of biochar even when focusing on C sequestration. Potential short-term losses of native SOM are smaller than the C gain of biochar and might be negligible in many cases (Woolf & Lehmann 2012). However, a trade-off may also exist with faster cycling SOM potentially leading to reductions in the quality of C available in the soil for use by the biota (i.e. as SOM is reduced), even if the overall quantity of C actually increases (i.e. in the form of biochar).
Provision of biofuel vs. production of biochar
Growing crops for biofuel production have given rise to concern about the trade-off regarding the land that is needed to grow these crops, as this land could otherwise be used for food crops or conservation (e.g. Tilman et al. 2009). Pyrolysis of waste products to produce biochar while concurrently producing biofuel is thought to at least partially circumvent this problem. However, some trade-offs between stabilized and actively cycling SOM are likely unavoidable because plant material that is removed from the field and converted to biofuel is no longer available for decomposition. Although biochar C when returned to the soil would more than compensate for the C removed in plant material, most of which would decompose, there is an important temporal delay with respect to the soil C balance and soil fertility (Whitman et al. 2010, 2011). Furthermore, the quality of the C will differ as discussed previously, and allocation of biochar in different locations than those from where biomass was removed may create additional trade-offs.
Life-cycle assessment has demonstrated that, if potential soil effects are not included, greenhouse gas (GHG) reduction effects are similar when biomass is used to produce biochar as when it is subject to complete combustion for energy production (Roberts et al. 2010; Woolf et al. 2010). When potential effects such as increased plant growth, reduced N2O emissions, etc., are factored in, biochar production can be favourable compared to combustion of biomass (Hammond et al. 2011). However, there is an inherent trade-off in the pyrolysis process between production of energy and production of biochar; increasing biochar production will always decrease energy production within the same energy pathway (Gaunt & Lehmann 2008). There is thus a trade-off in policy objectives between bioenergy and energy security vs. C abatement (Sohi 2012).
Different pyrolysis conditions and feedstocks lead to the production of different proportions of biochar, condensable gas and bio-oil. This gas and bio-oil can then be collected and used as a biofuel (Mahinpey et al. 2009). Slow pyrolysis of feedstock has been calculated to be more energy efficient, in terms of energy input vs. energy output, than production of biofuel through fermentation of feedstock to produce ethanol. Gaunt & Lehmann (2008) reported that where slow pyrolysis technology is optimized to produce biochar for soil application, a reduction in energy output of ca. 30% occurs compared to fast pyrolysis optimized for biofuel production. Thus, maximizing the production of biofuel through fast pyrolysis reduces the amount of biochar that is produced (IEA 2006). If all biochars are similarly stable in the soil (stability is critical for C sequestration potential), this would suggest lower overall C abatement. Slow pyrolysis favours biochar production and by the same token could maximize the C sequestration potential, at the cost of diminished output of biofuel products. Recent work has demonstrated that biochar produced through pyrolysis has beneficial effects when compared to the solid residue (i.e. solid by-products) of bioethanol production in terms of CO2 and N2O emissions (Cayuela et al. 2013a). Also, a smaller quantity of biochar of high stability can have the same C abatement value as a larger quantity of less stable biochar (Crombie et al. 2012). Further work is needed to compare energy pathways to allow quantification of trade-offs that inevitably occur between energy and biochar production.
Changing pyrolysis conditions to maximize either biochar or biofuel production is likely to affect the C : N : P stoichiometry of the resulting biochar. However, while the majority of the beneficial effects of biochar on crop productivity stems from pH effects and the ability of biochars to retain N and P from other sources, evidence suggests that N and P can be available from some types of biochars (De la Rosa and Knicker, 2011; Wang et al. 2012; Chintala et al. 2013). Differential losses of C and N during pyrolysis occur as a function of temperature (Enders et al. 2012). Additional properties (such as those affecting soil pH) may impact the bioavailability of nutrients, particularly P, present in both biochar and bulk soil. This confounds prediction of the likely effects of a given biochar on the stoichiometry of the soil solution. However, it is evident that trade-offs will occur between maximizing biofuel production from pyrolysis and producing biochar with optimal C : N : P stoichiometry for a given soil/crop/climate combination (Gaunt & Lehmann 2008). Therefore, the effects of different pyrolysis conditions on the C : N : P stoichiometry of biochars and their effects on bioavailability in the soil is an important area of future research.
Feedstock selection vs. the use of wastes
One readily apparent trade-off regarding choice of feedstock for biochar production is that of stability of the resulting biochar vs. its nutrient content. For example, evidence suggests that biochars made from poultry litter support greater increases in crop productivity than those made from wood (Jeffery et al. 2011) which probably at least partly result from higher nutrient contents in this feedstock. However, biochars made from poultry litter are less stable in the soil than those made from wood (Singh et al. 2012).
The paucity of feedstocks in many parts of the world suggests that another trade-off may occur. In such areas, such as much of Africa, all of the aboveground crop biomass produced, and not just the grain or fruit, is utilized for a range of purposes such as animal feed, roofing materials, mulch to reduce water requirements or incorporation into the soil to improve organic matter content. Therefore, it may be difficult to reserve significant amounts of feedstock for biochar production, and environmental degradation may result when alternative feedstocks for biochar production are sought (e.g., through deforestation). As such, entry points for adoption of biochar technologies may occur through substitution of current technologies, such as the switch from burning fire wood to pyrolysis stoves for cooking (Torres-Rojas et al. 2011). However, the implementation of biochar technology in Africa can be a contentious issue for a variety of social reasons (for further discussion see Leach et al. 2012).
One method that potentially circumvents the problem of paucity of feedstocks is the use of waste materials to produce biochar. In theory, any C-based feedstock can be pyrolysed to produce biochar, and so biochar production has the potential to mitigate the increasing global problem of waste disposal. To date, a wide range of waste streams have been considered and tested, including biosolids (Chan & Xu, 2009), tannery wastes (Muralidhara, 1982), paper sludge (Rajkovich et al. 2011) and sewage and wastewater sludge (Bridle & Pritchard 2004; Hossain et al. 2010). The type of feedstock affects the properties of the resulting biochar (Kloss et al. 2012) in terms of crop yield effects (Jeffery et al. 2011) and recalcitrance in the soil (Zimmerman 2010; Singh et al. 2012). Furthermore, it is likely to affect whether the resulting biochar is classified as a waste product, with implications regarding its permissibility for soil application (Sohi et al. 2010). Legislative issues surrounding biochar application to soils produced from waste products, and the classification of such biochar in terms of policy, is vital before its large-scale application can be implemented. Such a discussion is beyond the scope of this article but has been covered in terms of European policy implications by Van Den Bergh (2009).
The type of feedstock and pyrolysis conditions also affect the types and concentration of contaminants in the resulting biochar. For example, heavy metals, which are generally found in high concentrations in sewage sludge and biosolids, are increased in concentrations following pyrolysis (Chan & Xu 2009). To assess potential trade-offs, an assessment of whether heavy metals pose a greater risk in the sewage/biosolids stream or in the soil biochar stream is required. Such an assessment would necessarily examine whether such contaminants are more or less bioavailable in biochar compared to the original feedstock. Biochar can reduce the bioavailability of contaminants such as heavy metals that are already present in the soil (Park et al. 2011). In addition, while biochar can contain polycyclic aromatic hydrocarbons (PAHs; by-products of incomplete combustion) these amounts are low and hardly bioavailable such that biochars leach PAHs far below water quality criteria (Hale et al., 2012). However, as discussed previously biochars can vary considerably in their physical and chemical properties and as such bioavailability of PAHs (and other contaminants) should be monitored. Owing to the high cost of such monitoring this presents a potential hurdle to wide-scale implementation of biochar application to soil.
The use of waste products to produce biochar can lead to biochar with suboptimal properties. One example is the high sodium (Na) content that biochars produced from food wastes sometimes contain (Rajkovich et al. 2011). However, Na is mobile in soil and will leach out relatively quickly. Nevertheless, such biochars, which could reduce crop growth initially, need to be applied in smaller amounts, or they need to be applied well in advance of planting so that the Na has time to leach out. This leads to a further trade-off between applying biochar when possible (e.g., when labour is available, at an appropriate time to plough the field) compared to when it would function optimally (e.g., sufficiently before planting to allow time for Na to be leached).
The fact that any C-rich feedstock can be turned into biochar has also led to suggestions by the popular media that other waste materials, such as plastic, could be used for biochar production too (e.g. Harrabin 2009; Lovelock 2009; Black 2010). However, ‘plastic’ denotes a wide range of different compounds, and what is suitable for use as a feedstock needs further research both in terms of biofuel production and suitability of the produced biochar for soil application. While pyrolysing plastics may release compounds such as syngas that could be used for energy, from a climate change mitigation viewpoint, plastics are generally highly recalcitrant and it seems probable that subjecting them to pyrolysis will release more C into the atmosphere than would occur if they were buried in a landfill. If C sequestration is the goal, this is perhaps not a sensible option.
Contaminated-soil remediation vs. soil fertility
Addition of activated carbon to contaminated soil and sediment is sometimes used as a remediation strategy. Biochar also has the potential to perform this role as it has a high sorption capacity for persistent organic pollutants and pesticides (although generally lower than activated carbon) (Luthy et al. 1997; Cornelissen et al. 2005). While the use of activated carbon is an established practice for soil remediation, amending biochar to soil or sediment can also lower the bioavailable concentrations of pollutants and pesticides by one to two orders of magnitude. This is similar to what can be achieved through adding activated carbon, but is potentially less costly (Yang & Sheng 2003; Ghosh et al. 2011; Jakob et al. 2012).
Soil fertility increases resulting from biochar addition are often relatively modest in well-fertilized soils in temperature regions (Jeffery et al. 2011). As the effect of biochar on pollutant immobilization can be rather strong, the use of biochar in the temperate zone may be more relevant for alleviating soil and sediment contamination than for promoting crop growth. Life-cycle assessment (LCA) is required to examine the relative benefits of both uses. For example, one LCA study showed that activated biochar was greatly superior to an anthracite-based activated carbon for remediating a dioxin-contaminated fjord system in Norway, even though the biochar was slightly less chemically active in immobilizing dioxins, owing to its C sequestration potential (Sparrevik et al. 2011).
Further to the above-mentioned trade-offs between the most often reported ‘wins’ (also see Fig. 1), other trade-offs related to biochar use are evident. Here, we provide three further examples of such trade-offs.
Biochar and conservation tillage
A seldom acknowledged trade-off that may occur when applying biochar to soil is that it is necessary to bury the biochar in order to prevent it from being eroded by wind or water. This is usually done by mixing the biochar into the topsoil, either mechanically or by hand. Such mixing requires disturbance and cultivation of the soil, which promotes native SOM loss as well as potentially other side effects in those situations where incorporation is not part of on-going management. Therefore, wide-scale biochar application may trade-off against the benefits that no-till farming brings. Direct surface application (e.g., added to slurry or in muck spreading) is also possible, although such application techniques run the risk of the biochar being eroded by wind or rain (Rumpel et al., 2006). A possible response to mitigate this potential trade-off is to combine biochar with minimum tillage conservation farming practices, and apply biochar, for example, in the hoe basins or rip lines where cultivation takes place (Hobbs et al. 2008; Giller et al. 2009). This would also reduce the amount of biochar needed for fertility effects (Cornelissen et al., 2013b) and so helps circumvent some of the trade-offs that occur due to competition for feedstocks. However, such a strategy is labour intensive and unlikely to be implemented in large-scale arable systems, unless it can be combined with application of manures which may already be part of on-going soil management.
Biochar production and human health
Another potential trade-off exists between soil fertility and human health effects. For example, a biochar produced from maize cobs proved to be very effective for soil fertility in Zambia, increasing harvests by up to a factor four (Cornelissen et al., 2013a). However, the only realistic way for these farmers to produce biochar in the near term is through traditional kilns that emit particles <10 μm (PM10), CH4 and carbon monoxide (CO), as they cannot afford cleaner retort pyrolysis technologies. Charcoal particles (i.e. soot) have adverse effects on human health such as causing the lung disease pneumoconiosis, in addition to contributing to global warming (Baveye 2007). Thus, while beneficial to the farmers in terms of crop yield, a complete LCA using technology that was available to farmers showed that biochar implementation may have adverse health effects and was only slightly beneficial for climate change mitigation (since the emitted CH4 is a strong GHG, offsetting C sequestration) (Sparrevik et al. 2013). The areas that are most likely to experience these problems are also those where the potential benefit of biochar to crop production appears to be the highest: smallholder farms in developing countries (Sparrevik et al. 2013). The level of this trade-off is likely to vary and could be reduced or eliminated with the use of lower emission pyrolysis cook stoves.
Increased resistance to pests vs. decreased pesticide effectiveness
Elad et al. (2010) and Meller Harel et al. (2012) have reported that biochar can induce systemic resistance in some plant species (peppers, tomatoes and strawberries) to some fungal pathogens and other pests such as the broad mite (Polyphagotarsonemus latus). Should this effect also be found for field-grown crops, biochar may reduce the need for fungicides and potentially other pesticides. However, a trade-off likely exists as some evidence also suggests that pesticides are less effective when applied following biochar application (Yu et al. 2009; Graber et al. 2012). The issue of whether increased resistance to pests is sufficient to counter the reduced efficacy of pesticides and the wider-scale implications of biochar application to soil on pest populations requires further work.
There is now a large body of pertinent and robust literature on the effects of biochar on soil properties and processes. These include crop yields from multi-year field experiments (e.g. Kimetu et al. 2008; Gaskin et al. 2010; Major et al. 2010; Jones et al. 2012; Güereña et al., 2013), possible priming effects on soil organic matter (e.g. Zimmerman et al. 2011) and disease resistance (Meller Harel et al. 2012). However, most research is still phenomenological and descriptive; there is comparatively little research studying mechanisms behind observed effects (Sohi et al. 2010; Güereña et al., 2013). Such a mechanistic understanding is imperative to allow the assessment of potential trade-offs that is needed before large-scale application of biochar should be promoted. For this reason, we provide recommendations for experimental design to guide progress towards on the one hand (i) a mechanistic (i.e. reductionist) understanding and on the other hand; (ii) a systems (i.e. holistic) understanding.
Use of experimental controls
In order to allow elucidation of underlying mechanisms in comparison to the life-cycle effects of implementing a biochar system, quite different experimental designs are required. Traditionally, biochar effects have been assessed through comparisons with negative controls (i.e. with no addition). However, considerable scope also exists for comparing biochar addition with positive controls to address systems-level questions. The choice of controls (negative and/or positive) should depend on the situation and the hypothesis being tested. One way forward in this field could involve the use of multiple positive controls; having only one control allows the testing of only one hypothesis, whereas multiple controls allow the testing of several competing hypotheses. In addition, resolving questions about the systems impact of implementing biochar may require different controls than resolving questions about soil and plant processes. Therefore, in many cases, comparisons with positive controls that contain the uncharred feedstocks from which the biochar was produced are desirable. For example, many of the beneficial effects of adding biochar produced from poultry manure to a soil may also occur when adding the un-pyrolysed poultry litter (Chan et al. 2008). Inclusion of positive controls in experimental designs is vital to allow for the effects of biochar per se to be quantified and to move towards a systems understanding of the effects of biochar application to soil.
When soil fertility effects are the main subject of investigation, a positive control might involve the addition of the un-pyrolysed and/or ashed feedstock (i.e. a positive control). This would allow quantification of the effects of pyrolysis on increasing the availability of nutrients relative to fresh organic matter, and controlling the provision of soluble nutrients present in ash. It would reflect direct soil incorporation and combustion as two alternative uses for the same biomass. To establish trade-offs in feedstock use, rates of addition should probably be based on equivalent mass of the starting material. This would allow quantification of the effects of reduced residue incorporation on soil processes that are likely to occur if such residue is removed for biochar production. Such approaches make sense in systems and life-cycle studies, whereas process-based investigations may benefit from other sets of controls. The intent and purpose of the research is therefore important for driving decisions on choice of controls.
In addition to a systems understanding, a mechanistic understanding is also critical. It has been reported repeatedly that biochar application reduces CH4 and N2O emissions from the soil (Singh et al. 2010; Liu et al. 2011; Cayuela et al., 2013a, 2013b). However, without understanding the mechanisms behind these results it is not possible to robustly extrapolate to other environments, soil types or soil and crop combinations. As mentioned above, possible mechanisms include biochar reducing soil N availability directly, or indirectly through increasing P availability and thereby plant N uptake and immobilization. Such hypothesized mechanisms could be tested experimentally through comparison of biochar addition with a carefully selected positive control (i.e. with addition) as well as a negative control (i.e. without addition). For example, if the reduced N2O emissions occur as a result of reduced N availability due to changes in the C : N ratio, comparisons with positive control treatments containing high C : N ratio uncharred residues such as wood or straw may be informative. This would be pertinent if N-limited crops such as cereals were grown. Alternatively, increasing P availability in the positive controls and monitoring N2O emissions would allow testing of the hypothesis that changes in P availability increases plant N uptake, reducing N surplus within the soil thereby reducing N2O emissions.
Types of positive control
Biochar often leads to acid neutralization or a liming effect in acidic soils, although the strength of the liming effect will vary between biochar types. Soil pH plays an important role in the availability of nutrients and other ions in the soil, some of which can increase to toxic levels in acidic soils (e.g. Al), (Liang et al. 2006). Therefore, a control treatment in which the pH is adjusted to bring the soil pH in line with the biochar treatments would often be useful (Hass et al. 2012) to allow distinguishing those biochar effects that occur beyond pH effects. Other types of positive controls may include the use of plastic chips of the same size as biochar particles, or perlite to increase the soil WHC. Again, the choice of positive control will depend on the hypothesis being tested.
Finally, additions of nutrients to positive controls at the same rate as those present in leachable ash would allow differentiation between effects resulting from changes in nutrient availability relative to direct or indirect effects of biochar on the soil microbial community, water retention, etc. Quantification of nutrients could take place as a pilot experiment in which the amount of available nutrients such as N, P and K following biochar addition are analysed, and then this amount applied in the main experiment as a control. While the rate of release may be different, it would at least allow some control of nutrient effects and so aid identification of biochar effects without the influence of nutrients. While use of such a control has several potential weaknesses that would need to be overcome (such as different release rates of nutrients from biochar compared to fertilizer) efforts in this direction are necessary to allow identification of biochar effects beyond nutrient effects. Inclusion of positive controls in an experimental design should not (and indeed, must not in most instances) preclude the inclusion of negative controls as well.
Positive controls seek to match the projected, initial effects of a biochar addition. Over time the residual effects of a biochar and positive control are likely to diverge. Thus, an alternative approach is to remove a property of the biochar such as the potential nutrient content. This could be achieved by understanding the pattern of nutrient release from biochar and repeated leaching (Angst & Sohi 2013) or, for the effects of time on CEC, biochar ‘ageing’ (Cross & Sohi 2013).
A tiered approach to choice of controls
To achieve a mechanistic understanding of the effects of biochar, we suggest that moving forward with biochar research may benefit from a three-tiered approach, a framework for which is provided in Fig. 2. We suggest that, when the question at hand warrants it, a useful approach for experiments looking at the effects of biochar application could involve the use of both positive (i.e. addition of un-pyrolysed feedstock) and negative (i.e. no addition) controls (Level One controls; Fig. 2). It may also be useful to include ‘subtractive’ controls (Fig. 2). This would consist of controls in which the amount of un-pyrolysed feedstock added is reduced by an equivalent amount to that which is removed to produce the biochar. This would allow investigation of the effects of reduced C inputs into the soil due to removal of crop residues for biochar production.
However, to achieve a full mechanistic understanding of biochar effects it is necessary to distinguish between the effects of the biochar itself and effects such as nutrients added through ash, which are likely to be short-lived. For such experiments, we suggest that further controls would likely be necessary; examples of which are given in Level Two of Fig. 2. Inclusion of such controls has the added advantage of overcoming some of the shortcomings of relatively short-term experiments. For example, nutrient or liming effects associated with biochar are likely to be relatively short-lived as nutrients are utilized or leached from the soil. Addition of nutrients to controls at the quantities as present in biochar will allow for effects beyond nutrient additions to be quantified. While the rate of release between biochar and fertilizer (including slow-release fertilizer) is likely different, this may give an indication of what benefits remain once such nutrients are no longer present or available.
Finally, extrapolation of results from biochars produced from the same feedstocks but under different pyrolysis conditions should be undertaken with caution, despite being a potentially appealing short-cut in maximizing our understanding of biochar and its effects within the soil and wider environment. Biochars produced under different conditions likely show differences in terms of ash content, and therefore differences in liming effects, nutrient concentrations and ratios. It is vital that studies reporting the effects of biochar application to soil also include as much information as possible about the biochar characteristics and manufacturing or preparation conditions.
Reporting of Results
Comprehensive reviews and meta-analyses are important for aggregating effects and allowing more robust extrapolation, and hence guidance of policy and directing of future research. They are particularly pertinent for areas in which experimental results at a relatively small scale need to be up-scaled due to proposed large-scale implementation, as is the case for biochar (similar to other areas such as compost research which suffers from many of the same problems as biochar research). It would therefore be useful if data presented in primary research manuscripts comply with the guidelines described by the Cochrane Collaboration (Higgins and Green 2009) concerning the production of systematic reviews and meta-analyses and as first suggested for biochar research by Verheijen et al. (2010).
To maximize the utility of studies for meta-analyses, as many auxiliary variables as possible should be reported. These should include both biochar and soil properties; e.g. CEC and pH of soil before and after application, nutrient content of biochar, pyrolysis conditions (especially temperature), feedstock, soil properties, crop or plant type, and fertilizer used (type and dose), as well as climatic information, particle size, and mode and depth of incorporation where possible. Supporting such advances, approaches to provide standards or specification have emerged, for example, by the International Biochar Initiative (IBI 2012). Use of a reference biochar, produced from the same feedstock under the same conditions, will aid cross-site comparisons and therefore quantification of differences in interactions of a given biochar with contrasting soil types.
Furthermore, it is desirable that such information is reported in a standardized way. In many experiments, particularly those involving pots or mesocosms, application rate has been variably reported as% w/w (e.g. Meller Harel et al. 2012), as t ha−1 of C equivalence (e.g. Woolf et al. 2010) or as t ha−1 of biochar mass equivalence (e.g. Karhu et al. 2011). This complicates comparison of between-site results and standardization of units across papers as much as is possible is therefore useful. We recommend that application rates should be reported as t ha−1 mass equivalents, because not all researchers have the necessary equipment to analyse their biochar in units of t C ha−1. This recommendation is in addition to study-specific reporting (such as % w/w in the case of microcosm experiments) where different units may be warranted such as volume if targeting the physical impact on bulk soil, or concentration if targeting chemical effects.
Future Research and Conclusions
Owing to the extensive range of combinations of biochar, soils and plants, much research still needs to be undertaken to understand the large variety of resulting interactions and their effects. As research progresses, it will be possible to make extrapolations with increasing robustness as, for example, the database upon which meta-analysis can be performed grows. Such information is vital to guide the development of certification schemes such as that proposed by the International Biochar Initiative, and The European Biochar Certificate, which is already implemented in part of Europe, as well as to guide policy. However, as discussed above, trade-offs will almost inevitably occur between the potential ‘wins’ following biochar application to soil; such trade-offs are generally not yet quantified, or even identified. Experimental designs that consider such trade-offs between the wins should therefore be a priority.
To effectively guide future research, a representativity analysis (i.e. an analysis of which soil, biochar, crop combinations have been studied and whether it is representative of combinations ultimately deployed) is urgently needed to allow identification of the gaps in current research. For example, the meta-analysis of Jeffery et al. (2011) indicates that relatively few biochar experiments have taken place in temperate regions, or have focused on major crops such as potato. A representativity analysis would be a very useful tool for researchers as well as providing guidance to policy makers as to where to direct research funding.
In conclusion, the large and growing body of research reported in the literature demonstrates the potential of biochar application to soil to provide a range of benefits. However, such benefits are unlikely to be maximized in all situations and trade-offs will inevitably occur between them. Furthermore, there are currently insufficient data in the literature to draw conclusions concerning biochar production and application to soil in all situations. Published long-term experiments in particular are lacking and are vital to assess the long-term implications of biochar application.
To quantify and predict such trade-offs it is necessary to move towards a mechanistic understanding of the effects of biochar application. One way to move towards such an understanding is through judicious choice of controls. Standardized reporting of results will aid cross-site comparisons and aid quantitative reviews. Such steps will allow biochar research to move forwards while remaining firmly grounded in robust science, and will allow policy to be effectively developed to maximize the potential benefits of biochar while concurrently avoiding or minimizing negative effects.
We gratefully acknowledge funding from the Graduate School Production Ecology and Resource Conservation in Wageningen, the Netherlands for financial support and we thank the participants of the biochar workshop ‘Biochar: The Soil is the Limit’ in Wageningen, May 2012 for their part in the interesting discussions that have led to the production of this manuscript.