Wildfire, soil and vegetation: new directions
(Williams 2016) find that a positive trend in PET accounts for 78 % of the positive trend in burned area since 1984 in W USA. The strength of the observed interannual relationship between PET and forest-fire area and the co-occurring positive trends suggest that (1) continued warming will promote continued increases in western USA forest-fire area while fuels are not limiting, consistent with previous empirical evaluations, and (2) other processes besides increased PET have also contributed to the increase in the western US forest fire area and will continue to do so.
Climate change will likely increase fire severity and occurrence across the boreal biome (Flannigan 1998) and with these changes, there will be an increase in total annual C emissions (Turetsky 2011).
(Boby 2010) nicely estimates N and C emission in Alaska. They show have this can be done using mainly the adventitious-root-height method (ARH). The ARH method may, however, be ineffective for estimating pools in intense fires that leave trees uprooted, or in sites where trees were rooted within a layer of organic soil that was completely combusted, leaving no evidence of the rooting position of the prefire trees. They investigated 28 independent unburned stands and found that the average height of the highest adventitious roots on the stem corresponded to the surface of the moss layer; roots were, on average, 3.2 ± 0.43 cm below the surface of the green moss. This pattern was consistent across a range of understory moss communities, including feather mosses, Sphagnum spp., and other moss species
(Morgan 2014) We suggest that instead of collapsing many diverse, complex and interacting fire effects into a single severity index, the effects of fire should be directly measured and then integrated into severity index keys specifically designed for objective severity assessment. Using soil burn severity measures as examples, we highlight best practices for selecting imagery, designing an index, determining timing and deciding what to measure, emphasising continuous variables measureable in the field and from remote sensing.
(Pastick 2014) estimated soil organic layer thickness in Yukon.
(Amiro 2000) looks at NPP during the first 15 years. Most of the area have a NPP of carbon (C) around 150-300 g m-2 yr-1.Only the 1-3 since fire have no area with very high NPP (>350). All regions (more or less) show a positive correlation between NPP and time since fire. For each year NPP increases with roughly 10 g m-2 yr-1. One study area (Alberta boreal plains) has data up to 60 years after fire. Here, NPP increases up to 30 yr since fire and then slowly increases a bit more until 50-60 yrs since fire.
(Bond-Lamberty 2004)Total NPP was low (50–100 g C m-2 yr-1) immediately after fire, highest 12–20 years after fire (332 and 521gCm-2 yr-1 in the dry and wet stands, respectively) but 50 percent lower than this in the oldest stands. Tree NPP was highest 37 years after fire but 16–39 percent lower in older stands, and was dominated by deciduous seedlings in the young stands and by black spruce trees (85 percent) in the older stands. Bryophytes comprised a large percentage of aboveground NPP in the poorly drained stands, while belowground NPP was 0–40 percent of total NPP. Net ecosystem production (NEP), calculated using heterotrophic soil and woody debris respiration data from previous studies in this chronosequence, implied that the youngest stands were moderate C sources (roughly, 100 g C m-2 yr-1), the middle-aged stands relatively strong sinks (100–300g C m-2 yr-1), and the oldest stands about neutral with respect to the atmosphere. Source to sink occurred around 20-30 yrs after fire, both in wet and dry stands. Respiration of woody debri important up to 20-30 yrs after fire, at later stages its a very small C source.
(Mkhabela 2009) This study has shown that the fire sites had generally higher GEP, Re and ET than the harvested sites, which we believe is largely a result of greater species diversity at the fire sites coupled with higher soil water content. Regardless of disturbance history, NEP was generally negative for the younger sites, indicating that recently disturbed sites are C sources. C dynamics following fire may go through four phases compared to three phases for harvested sites: soon after fire, burned sites become C sources; then become C sinks; then become C sources again when the dead woody material starts decaying; and thereafter become C sinks or neutral. In contrast, harvested sites are C sources soon after harvest; C sinks at intermediate age; and then C neutral or a small sink at maturity. This hypothesised pattern is still very uncertain though.
Moss δ13C provides information about photosynthetic performance and relative growth rates, integrates moss photosynthetic activity throughout a growth period, and typically increases as a function of water availability (Rice 1996, Rice 2000). (Deane-Coe 2015) found that air warming also resulted in a consistent reduction in δ13C in all three mosses, and we found that δ13C was positively correlated with NPP in both years, as has been observed in other moss communities (Rice 2000). In contrast to vascular plants, where low δ13C values typically result from greater growth rates due to high water use efficiency, low δ13C values in mosses are most commonly observed when growth rates are low due to low tissue water content, when chloroplastic demand and diffusional resistance of CO2 are both low ((Rice 1996, Rice 2000). Collectively, the reduction in NPP and δ13C in two of the three dominant mosses at our site points to the possibility of drier conditions that reduced moss growth as a result of air warming.
(Alexander 2015) Global change models predict that high-latitude boreal forests will become increasingly susceptible to fire activity as climate warms, possibly causing a positive feedback to warming through fire-driven emissions of CO2 into the atmosphere. However, fire-climate feedbacks depend on forest regrowth and carbon (C) accumulation over the post-fire successional interval, which is influenced by nitrogen (N) availability. To improve our understanding of post-fire C and N accumulation patterns in boreal forests, we evaluated above- and belowground C and N pools within 70 stands throughout interior Alaska. Increased fire activity is also expected to volatilize nitrogen (N) (Harden and others 2002; Neff and others 2005; Boby and others 2010) in this already N-limited ecosystem (Van Cleve and Alexander 1981) because most N is stored in organic soils which are often greater than 50% consumed by fire (Kasischke and others 1995). Reduced N availability may limit forest productivity and regeneration and thus, C accumulation and storage during the post-fire interval (Harden and others 2000). Changes in the fire regime may also alter patterns of C and N accumulation by triggering patterns of forest regrowth that differ from the pre-fire stand. For example, increased fire frequency and extent can reduce seed availability (Romme and others 1998; Johnstone and Chapin 2006b), while increased fire severity can increase seed germination (Johnstone and Chapin 2006a). These changes in initial establishment parameters, combined with species-specific growth habits and competitive abilities, can alter stand composition and structure, and ultimately the rate and amount of C and N stored during post-fire succession (Weber and Flannigan 1997). Thus, the balance between fire-driven losses of C and N and their re-accumulation rates will determine whether increased fire activity creates a positive, neutral, or negative feedback to the climate system.
(Bona 2016)Tree canopy is intimately linked to site moisture, in part because of shading but also because black spruce trees are more productive (with less open canopies) when the water table is lower and a larger oxic rooting zone is provided. Therefore, our relationships between canopy and moss NPP are likely explained by soil moisture content differences between open and closed canopy systems. Water table depth, air humidity, or soil moisture information could improve prediction of moss NPP. However, since indicators of soil moisture or water table depth at a national scale a re lacking, this study suggests that tree canopy can be an effective proxy because canopy openness controls, as well as responds to, several light and moisture related variables. Results used in model. Although MOSS-C does not explicitly represent the paludification process, our results indicate the MOSS-C sub model’s potential ability to predict which sites are more likely to accumulate larger peat stocks over longer time frames, where high water tables can restrict tree roots and lower tree productivity, and to distinguish these from sites that tend to be dryer with more productive trees and smaller moss derived C pools. No significant relationships were found that could help describe potential sources of the unexplained finer scale variation in organic horizon C for these plots, suggesting that, to further improve the model, more detailed plot level data would be required. Also has fire in their model....
(Turetsky 2010) In Alaskan forests, moss abundance showed a unimodal distribution with time since fire, peaking 30–70 years post-fire. Mosses contributed 48% and 20% of wetland and upland productivity, respectively, but produced tissue that decomposed more slowly than both nonwoody and woody vascular tissues. The roles of moss traits in regulating key aspects of boreal performance (ecosystem N supply, C sequestration, permafrost stability, and fire severity) represent critical areas for understanding the resilience of Alaska’s boreal forest region under changing climate and disturbance regimes. REALLy good intro for highlightning mosses role. Eg. Mosses, one of the major groups of bryophytes, are ubiquitous and dominant components of ground-layer vegetation in both upland forests and peatlands across the boreal biome. These plants have received attention in several recent reviews for their importance in regulating soil hydroclimate and nutrient cycling in boreal ecosystems (van Breemen 1995; Turetsky 2003; Nilsson and Wardle 2005). Recent studies affiliated with the Bonanza Creek Long Term Ecological Research (BNZ-LTER) program also have documented relationships between moss composition and ecosystem parameters such as aboveground tree productivity and soil C storage (Hollingsworth et al. 2008) and have suggested that moss abundance plays a critical role in post-fire successional trajectories (Johnstone et al. 2010)
(Mkhabela 2009) Higher ET at the burned sites compared to harvested. The higher ET at the fire sites compared to the harvested sites was likely, in part, a result of the presence of both deciduous and coniferous trees. In addition, the fire sites had higher soil water content than the harvested sites, which might have enhanced ET.
(Kuglerová 2014) describes a hydrological model for boreal landscape using a high res DEM.
(Kuglerová 2015) Mapped groundwater discharge and found that it provided riparian-like habitat further away from the streams and also in upland-forest sites compared to the non-discharge counterparts. In addition, soil chemistry (C:N ratio, pH) and light availability were important predictors of vascular plant species richness. Mosses and liverworts responded to the availability of specific substrates (stones and topographic hollows), but were also affected by soil C:N. Overall, assemblages of mosses and vascular plants exhibited many similarities in how they responded to hydrological gradients, whereas the patterns of liverworts differed from the other two groups.
(Deluca 2012) Process models allow us to predict soil organic C losses but we presently lack a clear understanding of all the processes involved in forest soil C dynamics. Currently, this incomplete understanding of forest soil C dynamics and linkages to other cycles and responses to different disturbances all contribute to what we consider an over simplification of forest soil representation in models as a ‘C reservoir.’
Boreal forest ecosystems account for ∼50 per cent, or more, of world forest ecosystem C stocks. Soil C in boreal ecosystems has been reported to account for ∼85 per cent of the total biome C (Malhi et al., 1999) BUT taken from (Deluca 2012). Of the soil carbon, peatlands, although covering a small area (sweden 10-15%), account for probably more than 60 per cent while uplands account for the rest (my calcs). UPLANDS: Total ecosystem C storage in boreal forests varies by age, structure and stand history, all of which have a great influence on total C storage as reflected by changes in forest biomass, coarse woody debris, forest floor, and to a lesser extent, mineral soil. Mineral soil C (to 1.0 m depth) remains relatively stable and accounts for the majority of total ecosystem C across forest maturity and disturbance regimes. (Deluca 2012).
(Olsson 2009) that 60% of soil carbon is stored in the mineral soil for podzols. In general they found, the correlation coefficients for the linear relationship between SOC stock and site characteristics to be highest for N deposition, which explained up to 25% of variation, and latitude, which explained up to 20% of variation. Altitude had the lowest degree of explanation.
(Seedre 2011) Across much of the boreal, fire functions as a fundamental disturbance process that consumes the understory and moss bottom layer along with a portion of the humus pool and with a significant portion of the total C stored in the O horizon (Kasischke and Stocks, 2000) and partially resets the successional clock (Engelmark, 1999) with mineral soil C remaining greatly unchanged.
(Kelly 2016)used charcoal reconstructions of fire in Alaskan boreal forest to drive model simulations of carbon dynamics from AD 850–2006 and finds that fire was likely the dominant source of carbon-stock variability (accounting for 84 % of C stock variability) in boreal forests and that a recent increase in fire frequency since 1950 has led to large carbon losses.
Charcoal is formed by the incomplete oxidation of organic matter heated to temperatures that drive off volatile elements such as N, S and O, resulting in increased carbon density of the remaining organic matter (Demirbas 2008).
(González-Pérez 2004) “identify the following main effects of fire on soil organic matter: (I) general removal of external oxygen groups that yields materials with comparatively reduced solubility; (ii) reduction of the chain length of alkyl compounds, such as alkanes, fatty acids, and alcohols; (iii) aromatisation of sugars and lipids; (iv) formation of heterocyclic N compounds; (v) macromo-lecular condensation of humic substanc; and (vi) production of an almost unalterable component, the so called black carbon.”
(Johnson 2001) Meta-analysis: “it is clear that fire need not necessarily lead to a loss of soil C or N and indeed may cause increases in soil C and N by incorporation of charcoal and hydrophobic organic matter or by the invasion of N-fixing vegetation.”
(Hart 2013) Charchoal amounts of 12.5-340 g C m-2 are reported for boreal ecosystems Stand-specific factors (fire intensity, vegetation type and burning efficiency) probably play a much larger role in determining charcoal levels than time since formation. Rosengren (2000) found higher charcoal quantities in Scandinavian forest stands subject to more intense fires, the result of more biomass on more productive sites, where more biomass was consumed, compared with stands of low-intensity ground fires, consuming smaller amounts of biomass.
(Deluca 2012) Approximately 10 per cent of the woody biomass consumed by fire is converted to charcoal, a uniquely stable form of C with mean residence times measured in thousands of years (Figure 4) as opposed to months for twigs and small stems (DeLuca 2008). This stable form of C is often not accounted for when evaluating the influence of fire on total C storage in soil ecosystems. Interestingly, charcoal commonly accounts for approximately 500–1000 kg C ha−1 in the O horizon of boreal forest soils with estimated mean ages of 600–2000 years old (Zackrisson 1996, DeLuca 2008).
(Deluca 2012) To date, modelling efforts have not effectively accounted for charcoal generation in long-term C accounting in boreal forest systems.
(Zackrisson 1996) estimated that charred materials ranged from 984 to 2,074 kg ha, quantities sufficient to exert important ecological effects. In particular, their sorptive abilities were very effective at reducing phytotoxicity from phenols produced by Empetrum hermaphroditum. Sorptive abilities, however, tended to decrease in time, and to disappear after a century. Reduction of the binding action of phenols released by the understorey of Vaccinium myrtillus imposed by freshly charred materials exerted important effects on the renovation of a burnt Betula pendula forest because of a greater uptake of N and other nutrients by seedlings(Wardle 1998).
(Turetsky 2011a) Depth of burn in ditched peatland 20 cm which may correspond to 450 yrs of peat accumulation. This study used C14 to date the peat.
(Boby 2010) Well done study (ARH measurements etc) looking at DOB and biomass consumption in Alaska. Carbon emission rates averaged 3.3 kg C/m2 which is similar to other studies in N Am.
Litter decomposition: Temperature is a dominant driver of organic matter decomposition rates in soil environments (Deluca 2012). In addition to cool temperatures, boreal litter forest litter tends to be composed of phenol rich substrates that are relatively resistant to decomposition (Nilsson et al., 2008). These factors reduce rates of litter decomposition and increase rates of soil organic matter accumulation. Most of the C accumulates as surface organic matter as a result of the acidic litter types and poorly drained forest conditions. Importantly, organic matter decay is influenced by temperature and oxygen contents including moisture, soil physical properties, substrate quality (recalcitrance of litter) and nutrient availability (Stevenson and Cole, 1999). (Deluca 2012) Boreal forest litter is generally of relatively poor quality and the associated soils are inherently C rich and exhibit low nutrient availability, especially N. Substrate quality is dictated by two primary factors: (1) The recalcitrance of organic compounds that make up the material (e.g. lignin or phenol contents); (2) The nutrient content, and in particular, the N and P content of the matter (Berg 2008).
(Deluca 2012) Combining of climatic variables with litter N content and indices of litter recalcitrance may be the most effective means of predicting litter decomposition rates; however, further work is necessary to better capture litter quality estimates.
(Deluca 2012) Modelling studies seem to show that controls are system specific. For example, unlike more temperate ecosystems, one of the key mechanisms involved in C preservation in boreal soils is the cooling of subsurface soil layers as soil depth increases rather than increasing recalcitrance in subsurface soils (Carrasco 2006). (Carrasco 2006) Results indicate that total C accumulation is controlled by the rate of carbon input, decomposition rates, and the presence of historical permafrost. However, unlike more temperate ecosystems, one of the key mechanisms involved in C preservation in boreal soils examined here is the cooling of subsurface soil layers as soil depth increases rather than increasing recalcitrance in subsurface soils.
Leachates from both burned upland and peatland soils had increased aromaticity and causing decreased biodegradation and increased UV-mediated. Soil C solubility in spruce forests decreased following wildfire and that DOC/TN decreased following fire in both upland and peatland leachates, indicating that N became relatively more mobile. For peatland soils, the effects of wildfire persisted 6 years after fire in hollows where vegetation had not regenerated, but we found little evidence of a fire effect on regenerated hummocks (Olefeldt 2013).
Selective removal of highly aromatic DOC in lakes through UV-mediated processes implies that organic sources that are considered stabile in terrestrial ecosystems can be readily mineralized once entering aquatic ecosystems. Together, our results suggest that regional characteristics (climate, surface geology and lake morphometry) can prevent wildfire from causing pulse perturbations to the linkages between terrestrial and aquatic C cycling and also regulate the processes that dominate within-lake removal of terrestrial DOC (Olefeldt 2013a).
Hypothesis: fire severity affects formation of pyrogenic coal (charcoal) - the higher severity, the more charcoal. A high burn severity will, however, results in a greater C loss during fire. At a longer time scale, the higher pyrogenic carbon content found in severely burned areas will facilitate soil (and above ground) carbon accumulation.
- Charchoal increases productivity (higher pH, promote decay). Peatlands are very nutrient poor and such nutrient “kick” should increase Sphagnum growth.
Nitrogen is the primary limiting nutrient in boreal forest ecosystems (Tamm 1991).
(González-Pérez 2004) The C/N ratios of soil after burning are usually lower than in the original soils (Table 3), a phenomenon frequently cited in several types of post-fire soils. Also, after a fire there is an increase of available nutrients in soil, mainly in the form of water-soluble components of ash that became available to living organisms. Part of this effect derives from an increase in soil pH frequently observed after a fire which is associated to an increase in exchangeable cations in soil Viro, 1974 and Raison, 1979 resembling the effect of liming the soil.
(Certini 2005)The immediate response of soil organic N to heating is a decrement because of some loss through volatilisation (Fisher and Binkley 2000). However, a substantial portion of soil organic N survives low intensity fires, maybe changing its form.Moderate to high intensity fires convert most soil organic nitrogen to inorganic forms. Amonium is a direct product of the combustion, while nitrate forms from ammonium some weeks or months after fire as a result of biochemical reactions called nitrification (Covington and Sackett 1992). Both NH4+–N and NO3–N are available to the biota, but if not promptly uptaken, they follow quite different destinies: nitrate is soon leached downwards, while ammonium is adsorbed onto the negatively charged surfaces of minerals and organics and, thus, is held by the soil (Mroz et al.1980).
Grogan et al. (2000) assessed that the NH4+ pulse generated by a severe wildfire in a mature forest of P. muricata was dissipated by the end of the second growing season.
(Boby 2010) estimated that nitrogen emission rates averaged 0.09 kg N/m2 in Alaska. There are few published values to compare to our estimate, but the magnitude alone indicates that fire is an important pathway of N loss from these ecosystems. If a linear N input of 0.0002 kg N·m−2·yr−1 is assumed, then the fires investigated, on average, emitted 450 years of N accumulation.
Read More: http://www.esajournals.org/doi/full/10.1890/08-2295.1
From (DeLuca 2006) Our results corroborate the largely untested hypothesis that frequent fire in ponderosa pine forests increases inorganic N availability in the long term and emphasize the need to study natural, unmanaged sites in far greater detail.
(Certini 2005) Little loss of P through leaching and volatilisation - so enrichment of soil P, but it doesnt last long. Concentrations of cations, such as Ca2+,Mg2+, and K+,and the anion SO42increase considerably in the soil solution immediately following burning (Khanna and Raison 1986). The behaviour of micronutrients, such as Fe, Mn, Cu, Zn, B, and Mo, with respect to fire is not well know but seems to move down (in the soil) slowly.
(Wirth 2002) Following Vitoussek and Howarth (1991) the ecosystem N pool in fire ecosystems should decrease over time because fire events usually cause losses of N mainly in the form of Nox (oxidation), N2 (pyro-denitrification), organic compounds (volatilisation), ash particles (convection) and nitrate (post-fireleaching) (e.g., Grogan et al., 2000; Kuhlbusch et al., 1991; Lorbert and Warnatz, 1993; Neary et al., 1999). Consequently, one would expect to measure lowest ecosystem N pools in the heavily burned stands of the lichen type. Interestingly, this was not the case and surface fires seemed to only redistribute soil N from the organic into the mineral layer. The total ecosystem N pool was 7.4±1.5 mol N m−2 on average of which only 25 percent were stored in biomass or coarse woody debris. Total ecosystem N was independent of stand age, surface fire regime and site type. No correlation was found between total ecosystem C and N pools. Average total ecosystem C:N ratio was 114±35 mol C mol N−1.
Areas which have higher concentrations of stored S from past acid precipitation or have large areas of peatlands in the watershed may have aggravated losses of S and H+ after drought and fire (Bayley 1992).
(DeLuca 2002)In northern Eurasia, Scots pine forests are more likely to survive fire events but also experience establishment of various birch species as an early successional invader. In North America and in Eurasia, feather mosses re-establish during the 20–50 years following fire and bring with them the capacity to accumulate N through their cyanobacterial associates. Nitrogen fixation in feather moss communities after fires is an important ingredient in maintaining forest in fire-maintained ecosystems (DeLuca 2002) and is therefore essential for the re-accumulation of C. Recurrent fire on an excessively short interval has the potential to result in a net loss of N from the ecosystem and may inhibit ecosystem recovery after fire (DeLuca 2002).
As previously noted, feather mosses and sphagnum peat mosses cover the forest floor of most boreal ecosystems. Both moss types harbour N2 fixing cyanobacteria creating a niche for N2 fixation in an otherwise hostile (temporally dry or excessively wet) environment (DeLuca 2002). Although N2 fixation in individual moss shoots is relatively low, collectively, the mosses provide the vast majority of fixed N in secondary successional boreal forests (Zackrisson 2004).
(Holden 2016) Soil basal respiration exhibited a similar response to fire severity as microbial biomass, with greater reductions in higher severity sites. These findings are not in agreement with hypothesized post-fire increases in microbial decomposition from classic ecosystem theory of secondary succession (Chapin et al 2011, Harmon et al 2011). Nevertheless, they are consistent with recent studies reporting that microbial respiration decreases or shows no change following boreal forest fires (Goulden et al 2011). Post-fire microbial respiration may have been limited by low microbial biomass, low water availability, or low soil C quality (Dooley and Treseder 2012).
Bååth et al. (1995) to demonstrate that in burnt coniferous forests fungi were reduced more than bacteria. Soil-dwelling invertebrates: very rapid recolonization and only marginal effects reported in literature. Fungivores decrease though. In a coniferous stand, up to 12 years were necessary for microbial biomass to return to pre-fire levels (Fritze et al. 1993) In a P. abies forest, Pietikainen and Fritze (1995) found that soil basal respiration diminishes after a low-intensity prescribed fire but not proportionally with the reduction in microbial biomass C, evidently because the specific respiration rate (CO2–C evolved per unit of microbial C) is greater in burnt areas than in the control. The impact of fire on biological properties of soil depends strictly on soil moisture. The highest decline was observed at the moistest condition, maybe as a result of faster heat transmission than in drier soils.
Read’s hypothesis that arbuscular mycorrhizal fungi should dominate ecosystems with low accumulation of surface litter, and ectomycorrhizal fungi should proliferate where organic horizons are well-developed. This pattern is expected because ectomycorrhizal fungi display a greater capacity to mineralize organic compounds than do arbuscular mycorrhizal fungi.
(Treseder 2004) Fire did not reduce the abundance of arbuscular mycorrhizal fungi. In contrast, ectomycorrhizal colonization required up to 15 years to return to pre-fire levels. As a result, dominant mycorrhizal groups shifted from arbuscular to ectomycorrhizal fungi as succession progressed. Altogether, microbes that can mineralize organic compounds (i.e., ectomycorrhizae and bacteria) recovered more slowly than those that cannot (i.e., arbuscular mycorrhizae). Potential net N mineralization and standing pools of ammonium-N were relatively low in the youngest site. In addition, glomalin stocks were positively correlated with arbuscular mycorrhizal hyphal length, peaking early in the chronosequence. Our results indicate that microbial succession may influence soil carbon and nitrogen dynamics in the first several years following fire, by augmenting carbon storage in glomalin while inhibiting mineralization of organic compounds.
(Deluca 2012) The main motivators for the elaboration of forest soil carbon (C) models are: (1) the scientific need for understanding C-related processes in forest soils and linkages across scales (soil ↔ ecosystem ↔ physical environment) and (2) quantifying ecosystem C stores. In case 1, models permit us to account for the multiple processes across scales of time and space, furthering our understanding of the system. The complexity of the processes involved in soil C accumulation make it difficult to develop accurate models, however, our need to predict process outcomes requires that we create models based on our current understanding of processes. In case 1, models permit us to account for the multiple processes across scales of time and space, furthering our understanding of the system. The complexity of the processes involved in soil C accumulation make it difficult to develop accurate models, however, our need to predict process outcomes requires that we create models based on our current understanding of processes. Process models: Models based on our understanding of biogeochemical and ecological exchanges in forest soils or on interactions among forest vegetation, soils, the underlying geology and geomorphology and the atmosphere are referred to as process models. Two models emerge as the most widely used and published forest soil models, YASSO (Tuomi et al., 2009) and CENTURY (Parton et al., 1993). Among the most complex published ecosystem model is Ecosys (Grant et al., 2006). NO PEATLANDS IN THEM I THINK!! Global C models that attempt to model vegetation dynamics, atmosphere–biosphere interaction or global C budgets increasingly incorporate soil representation (e.g. LPJ, DyN). Have maybe peatlands?? Empirical models: Models are rarely completely process based nor are they fully empirical. There are models at the empirical end of this continuum that use statistical methods to expand field measurements over a large land base. They do not, however, generally advance our understanding of the forest system the way process models do and they are not appropriate for use under changing environmental conditions. For temporal projections or for spatial projections outside the data range, process representation provides valuable constraints to the modelling system, while fitting the model (or parameters) to actual data grounds the projections in reality. Fortunately, linkages at other scales can also expand our potential for better modelling and may provide cross-scale C estimates. CENTURY, which already compared relatively well with eddy-flux measurements (Kirschbaum et al., 2007), has been linked to satellite data and forest inventory systems (Potter et al., 2008), providing a potential for wide application.