2.4 4.

Communication

Accountable Accounting: Carbon-

Based Management on Marginal
Lands

Tara L. DiRocco, Benjamin S. Ramage, Samuel G. Evans and Matthew D. Potts

Special Issue
Forest and Wood Vegetation Carbon Stores and Sequestration

https://doi.org/10.3390/f5040847
Forests 2014, 5, 847-861; doi:10.3390/f5040847
OPEN ACCESS

forests
ISSN 1999-4907
www.mdpi.com/journal/forests
Communication

Accountable Accounting: Carbon-Based Management on

Marginal Lands
Tara L. DiRocco 1,*, Benjamin S. Ramage 1, Samuel G. Evans 2 and Matthew D. Potts 1,2
1
Department of Environmental Science, Policy & Management, University of California,
Berkeley, CA 94720-3114, USA; E-Mails: bsramage@berkeley.edu (B.S.R.);
mdpotts@berkeley.edu (M.D.P.)
2
Energy Biosciences Institute, University of California, Berkeley, CA 94704, USA;
E-Mail: sgevans@berkeley.edu

* Author to whom correspondence should be addressed; E-Mail: tdirocco@cal.berkeley.edu;

Tel.: +1-510-642-5580; Fax: +1-510-643-5438.

Received: 20 January 2014; in revised form: 4 April 2014 / Accepted: 11 April 2014 /
Published: 23 April 2014

Abstract: Substantial discussion exists concerning the best land use options for mitigating
greenhouse gas (GHG) emissions on marginal land. Emissions-mitigating land use options
include displacement of fossil fuels via biofuel production and afforestation. Comparing C
recovery dynamics under these different options is crucial to assessing the efficacy of
offset programs. In this paper, we focus on forest recovery on marginal land, and show
that there is substantial inaccuracy and discrepancy in the literature concerning carbon
accumulation. We find that uncertainty in carbon accumulation occurs in estimations of
carbon stocks and models of carbon dynamics over time. We suggest that analyses to date
have been largely unsuccessful at determining reliable trends in site recovery due to
broad land use categories, a failure to consider the effect of current and post-restoration
management, and problems with meta-analysis. Understanding of C recovery could be
greatly improved with increased data collection on pre-restoration site quality, prior
land use history, and management practices as well as increased methodological
standardization. Finally, given the current and likely future uncertainty in C dynamics, we
recommend carbon mitigation potential should not be the only environmental service
driving land use decisions on marginal lands.
Forests 2014, 5 848

Keywords: forest carbon sequestration; carbon dynamics; land-use change; forest

recovery; afforestation

1. Introduction

There is a growing interest in using marginal land for greenhouse gas (GHG) mitigation. Recent
estimates have suggested that 1107–1141 million hectares of marginal land may exist, of which
385–472 million hectares is thought to be abandoned agricultural land [1,2]. GHG mitigation options
currently under consideration for marginal land (defined below) include on-site carbon (C)
sequestration via reforestation or afforestation, as well as fossil fuel displacement via the production of
bioenergy crops [3]. The relative efficacy of these different options is likely to be highly
context-dependent, and thus determining the optimal GHG mitigation strategy for any given location
requires a thorough understanding of carbon dynamics.
However, there are several key gaps and limitations in the existing literature on carbon dynamics
that prevents the precise evaluation of these mitigation strategies, particularly with regards to marginal
land. These include inadequacies in C pool quantification, broad categorization of land use types
and conversion histories, inadequate measurement and/or reporting of critical site-specific factors,
and the inevitable uncertainty associated with future predictions [4,5]. These inaccuracies limit
the efficacy of global and national policies aimed at reducing atmospheric carbon levels through
terrestrial mitigation strategies [6]. Without reliable input data to support site-specific C sequestration
potential, resources may be invested in sequestration or offset projects that are either (a) unlikely to
achieve their stated objectives or (b) sub-optimal compared to other land use options for C offsets (5).
The objective of this paper is to synthesize trends and challenges in quantifying C on recovering
marginal lands. We focus in particular on the C sequestration potential of recovering forests, either via
natural succession or afforestation. We do not directly consider C offsets of bioenergy crops as this has
been discussed in detail elsewhere (e.g., [7,8]). In the context of this paper, our use of the term
marginal land refers specifically to land that: (a) is not being used for any clear economic or
subsistence purposes; (b) has been substantially altered by humans from its original condition and is
currently of limited value in terms of biodiversity and ecosystem services; and (c) has low potential
yields based on soil, temperature, precipitation, elevation, and slope [9]. While the scope of our work
is limited to the ecological aspects of C sequestration, it is essential to recognize that many
social, legal, and ethical aspects of marginal land management still require detailed analysis
and consideration.
The paper is organized as follows. We first summarize the difficulties of quantifying forest C
stocks, noting key sources of inaccuracy and suggesting improvements in methodology and
standardization. We then discuss barriers to the effective prediction of C recovery dynamics on
marginal lands, and outline key data to be recorded and considered prior to any land management
decision. We conclude with recommendations for future research and suggestions for policies to
mitigate the potential pitfalls of carbon-based management schemes.
Forests 2014, 5 849

2. Quantification of Forest Carbon Stocks

A methodological understanding of C stock quantification is integral to investigating current

discrepancies in the C recovery literature [10]. Forest carbon stocks are typically divided into five
pools: aboveground biomass (AGB), belowground biomass (BGB), soil organic carbon (SOC), coarse
woody debris (CWD), and litter (Figure 1) [11]. The proportions of C in each pool are affected by
numerous factors, including forest age, which is highly relevant because different pools exhibit
different rates of recovery following natural succession or afforestation [12]. The AGB and BGB pools
increase during forest recovery, but the relative contribution of each pool in relation to regeneration
method remains poorly understood due to a paucity of BGB studies [13,14]. Within the soil pool, C
typically decreases shortly after restoration then increases to levels that equal [15–19] or exceed
pre-restoration C stocks [20–23], but see exceptions observed in early succession: [16,24,25]. The
initial decrease in soil C is due to a lag in leaf litter accumulation, which upon decay contributes C to
the soil pool [26,27].

Figure 1. Carbon (C) continuously cycles between the atmosphere, the dead and living
biomass, and soils of forest ecosystems [28]. For quantification and subsequent sampling
purposes, it is recommended that forest C should be divided into five different pools:
aboveground biomass, belowground biomass, soil, coarse woody debris, and litter [26].
The division of a forest ecosystem into C pools is shown below.
Forest Carbon Pools

Aboveground Biomass (AGB)

Coarse Woody Debris (CWD)

(standing)

(fallen)

Litter

Soil Organic Carbon

Belowground Biomass (BGB)

AGB: living vegetation above the soil including stems, branches, bark, foliage
BGB: live roots >2 mm
CWD: non-living biomass not contained in litter; either standing or fallen; includes
wood lying on surface, dead roots, and stumps >10 cm diameter
Litter: non-living biomass of size greater than SOM limit (~2mm) and less than CWD
limit (~10cm) lying dead above soil
SOC: organic carbon in mineral soils and in live or dead fine roots (<2mm)

Sources: Sierra et al. [28], Dierkes [29], Ngo et al. [6].

Forests 2014, 5 850

Despite general knowledge of coarse trends such as these, the dynamic processes by which C cycles
and accumulates remains poorly understood in forests [4,22,26], particularly regenerating forests [30],
and precise stock proportions for any given forest can rarely be estimated with confidence. Sources
of uncertainty in estimates of forest carbon sequestration occur at multiple temporal levels resulting
from the interplay of numerous factors (Figure 2). The two main sources of uncertainty with regard to
stock estimates are (1) failure to measure all C pools and (2) pool-specific methodological
problems [28,31]. Below we discuss in detail the implications of these shortcomings for carbon centric
land use decisions.

Figure 2. Sources of uncertainty in estimates of forest carbon sequestration. Uncertainty

in forest carbon estimates occurs at multiple temporal levels for various reasons outlined
below and discussed in the following sections.

2.1. Failure to Include All C Pools

Failure to include all five pools of C in studies gives an incomplete measurement of C accumulation
during land restoration and makes it difficult to compare C accumulation across studies [32]. Most
studies that quantify C stocks following land-use change solely include the aboveground biomass pool,
with the inherent assumption that this is indicative of total C recovery [13]. However, solely looking at
the AGB pool can lead to an overestimation of C stocks in earlier years of succession, failing to
account for C losses in soil and debris respiration. There is also evidence that the BGB pool could be
68% larger than previously thought, and that allocation of biomass above versus belowground differs
with regeneration method [33]. For example, Cuevas et al. [14] found that 44% of biomass production
in a regenerating secondary forest was belowground versus only 6% in a paired afforested
monoculture. While meta-analyses often ignore the BGB pool because of a paucity of data (e.g., [12]),
including the BGB pool could drastically change results especially with regards to the sequestration
potential of one regeneration method over another [13]. Similarly, studying changes in the SOC pool
without accompanied litter C data is misleading because litter C decays into SOC pools. Failure to
include the litter pool has led studies to conclude losses in SOC with regeneration, when in fact the net
ground C remained the same due to increases in litter C from tree growth [22].
Forests 2014, 5 851

2.2. Pool-Specific Methodological Errors

In addition to inaccuracies arising from pool inclusion, there are many inaccuracies when
quantifying the C in each individual pool. Estimations of C in the aboveground biomass (AGB) pool,
where the majority of C sequestration during restoration occurs, vary significantly depending on
methodology [10,34]. The most accurate method of calculating AGB C content is to apply site-specific
allometric equations (derived from local destructive sampling) to forest inventory data [35].
The choice of allometric biomass equation can significantly affect the estimation of AGB C stocks,
causing potentially large errors in AGB C quantification when inappropriate allometric equations are
used [4,34,36]. A further source of error is the assumption that all trees contain a C content that is 50%
its AGB, which is widely used despite little chemical verification [37]. Not only do most trees contain
less than 50% C by weight, but studies are also finding significant variation in C content between
species (e.g., Elias and Potvin [38] found a significant variation in tree C content of 32 tropical trees,
ranging from 44.4% to 49.4%). Accounting for the variation in species-specific C content can reduce
errors in estimates of AGB C sequestration by 10% in some cases [38], but this parameter is rarely
considered in C estimations [37]. Additionally, several studies have found a significant difference
in the C content of the heartwood versus sapwood of a tree, which should also be considered when
processing wood core samples and calculating AGB C [39].
Belowground, the spatial heterogeneity of root biomass and consequent sampling challenges make
the belowground biomass pool the most difficult to quantify [40]. When the pool is not ignored
altogether, many studies use inaccurate root: shoot ratios (estimating BGB from AGB data)
or inadequately sample their study sites due to time and financial constraints [41,42]. Inadequate
sampling methodologies have recently been calculated to underestimate the global belowground
biomass pool by 50%–68% [33,43]. Further errors occur from the loss of sample mass during
preparation and storage. These differences in mass alone can underestimate root content by 1/3 [42].
While more accurate methods of belowground biomass sampling use soil monoliths and trench
sampling, an adequate sample volume has yet to be determined [42,44]. Since SOC accumulates at
different rates within different soil depths during site regeneration [20], the most accurate sampling
methodology is to repetitively sample soil at 1–2 m depth at the same site. However, most studies
sample only to much shallower depths often due to sampling difficulties in the degraded soils
characteristic of marginal landscapes [4]. This data gap in deep soil C dynamics in conjunction with a
paucity of studies on later successional SOC limits our understanding of soil nutrient stocks and
dynamics [24].

2.3. The Importance of Standardized and Scientifically Sound Methodology

Before comparing C accumulation between sites, it is vital to understand the calculation of those C
values and their level of reliability and accuracy. Above, we have demonstrated the multiple levels of
uncertainty and variability in forest C estimations. It is not our intent to design extensive methods for
accurate forest C quantification, as has been done extensively elsewhere [45–47]. Instead we stress the
rationale behind these protocols, outlining the risks and errors incurred from falling short on the
recommended sampling protocols. In doing so, we emphasize the importance in the global application
Forests 2014, 5 852

of these standardized methods that is necessary to accurately determine trends in C recovery on

marginal land [12,48]. Understanding the sources of uncertainty in C estimations allows decision
makers to analyze whether or not a forest C study is accountable for a C-trading mechanism. Ideally,
previously collected data should be reanalyzed when possible using the types of methodology
presented by IPCC or the Center for Tropical Forest Science to provide a more complete global picture
of C sequestration from land-use change. Without such standardization, it is difficult to reliably
estimate C sequestration following land-use change [12].

3. Barriers to Effective Prediction of C Recovery Dynamics

Beyond the difficulties of quantifying existing C stocks at a particular point in time, further
uncertainty arises when assessing the sequestration potential of a parcel of marginal land. Factors
affecting C recovery (stocks and rates in all pools) include regeneration method, previous land use
history, climate, geomorphology, latitude, tree species, age, and distance to seed source, among
others [12,21,49–51]. However, there is little cohesion in the literature as to which parameter is most
indicative of site recovery. From a critical review of individual studies and meta-analyses, we have
determined that the literature is inconclusive about which biophysical parameters are most predictive
of site recovery for three main reasons: (1) the use of broad categories of land use history; (2) a failure
to consider the effect of current and post-restoration management; (3) problems with meta-analysis.

3.1. Broad Categories of Land Use History

Rather than collecting or finding detailed information on site quality, studies often consider land
use history as a proxy for regeneration and production capacity. Here, site quality is defined as
the ability of a piece of land to support vegetative growth and is influenced by a host of factors ranging
from climate to soil structure. Previous work has demonstrated that land use category i.e., cropland,
pasture, or mining, duration of land use, and the number of types of previous land use are all indicative
of pre-recovery site quality [12,17,52–56]. However, while similar land use histories often create
similar site characteristics, leading to similar biomass accumulation trends [25], there are several
problems with this approach.
First, most research fails to acknowledge that there is substantial variation of practices within all of
the land use categories typically specified in the literature, i.e., pasture, cleared, agriculture,
fire [12,13]. For instance, within the category of agriculture, there exist a variety of practices ranging
from swidden cultivation to intensively managed and fertilized cropland. Fertilized cropland usually
has less degradation and higher fertility than lands under swidden cultivation, leading to higher
biomass accumulation rates than on the ash-fertilized acidic soils following swidden
cultivation [12,57]. While both types of landscapes are grouped into the category of abandoned
agricultural land, it is clear that different management practices can lead to large differences in site
quality [25]. This is also the case for the pasture land use category, where the details of pastoral
activity (e.g., the duration of fallow periods, grazing intensity, and number of cycles) differentially
impact the ability of abandoned pasture to regenerate forest [58,59]. Whereas old pastures are typically
carbon-depleted and therefore experience greater rates of C sequestration into the soil pool than do
younger pastures [60], the rate of biomass accumulation is greater on younger pastures [58,59].
Forests 2014, 5 853

Grouping all pastoral landscapes into one category for analysis overlooks these land use details that
may be more indicative of regeneration potential. Letcher and Chazdon [53] warn of such ―crude‖
division of land use categories. It is also important to consider that the distribution of land uses may
reflect preferential site selection for desirable characteristics, confounding land use effects with actual
site quality effects [61].
Despite these issues, several general trends emerge from the literature that can inform
carbon-centric management practices, outlined in Table 1.

Table 1. Prior land use practices and their associated effect on forest recovery. These are
trends generally agreed upon in the literature concerning C sequestration on abandoned
land given certain site history characteristics.
Prior Land Use
Effect on Forest Recovery
Characteristics
Fertilized cropland will likely experience higher rates of biomass
Fertilized > Not Fertilized accumulation than unfertilized land, due to increased soil nutrient
availability.
Pasture lands tend to be more degraded than cropland, as low
fertility and soil compaction cause reduced aeration and soil
Cropland > Pasture biological activity. However, this trend could be due to the
preferential selection of less productive sites for pasture in the
first place.
Younger pastures are less degraded than older pastures since they
have not experienced as much soil compaction and nutrient loss
Young Pasture > Old Pasture
from pastoral activity. Therefore, there is faster biomass
accumulation on younger pastures than older pastures.
Although an ideal fallow period to maintain site fertility has yet to
be established in the literature, longer fallow periods allow the
Longer fallow > Shorter
replenishment of site nutrients. Increased site nutrients are
fallow
associated with increased biomass growth and subsequent C
sequestration.
Fewer cycles of cultivation means a less degraded landscape. A
Fewer cycles > More cycles less degraded landscape has better site quality with the nutrients
necessary for higher carbon accumulation rates.
Sources: Vesterdal and Rosenqvist [25], Kotto-same et al. [57], Hughes et al. [59], Milakovsky et al. [40],
Klanderud et al. [62], Silver et al. [12], Fearnside and Guimares [63], Feldpausch [58].

Broad land use categories are not informative with regard to most of the factors discussed in
Table 1. Thus, it is extremely important that before any land use decisions are made, a comprehensive
assessment of biophysical site quality be undertaken [60]. Unfortunately, the majority of studies rarely
report site quality data, which partially explains why the literature has been inconclusive in providing
reliable predictors of site regeneration capacity [13]. On a final note, if forest regeneration is
to proceed naturally without planting, it is also important to consider the surrounding landscape
matrix and proximity to forest [64]. While the duration and intensity of previous land use influences
Forests 2014, 5 854

within-site propagule availability, the surrounding landscape determines the seed arrival that is
necessary for forest establishment [51].

3.2. Effect of Current and Post-Restoration Management

In the same way that site quality caused by land use history affects regeneration potential, land
management during and after restoration affects C sequestration potential [65,66]. The frequent
categorization of regenerating landscapes into the dichotomy of secondary forest or monoculture
plantation often fails to consider the specific details of regeneration schemes implemented on
marginal landscapes. Regeneration schemes vary widely in degrees of management intervention, such
as thinning, fertilization, the planting of N-fixing species, mycorrhizal inoculation, and environmental
restoration plantings [67–70]. These forms of assisted regeneration, with often extensive management
interventions, can lead to higher C sequestration rates than those of unassisted stands [5,17]. For
example, while it is generally believed that plantations sequester more C than secondary forests,
Bonner et al. [13] found that this result disappears when one compares unfertilized plantations to
secondary forests. Considering the effects of active management techniques on C accumulation
rates is therefore important to avoid the confounding of these actions with pre-restoration site
recovery potential.

3.3. Problems with Meta-Analysis

As noted in the introduction, much C policy today is being guided by the results of reviews
and meta-analyses. The failure of most primary studies to report details on site quality, land use
history, and management practices seriously limits the inferences that can be drawn from subsequent
syntheses of these studies. For instance, the recent meta-analysis by Bonner et al. [13] is limited by
a paucity of replicates for several important combinations of factors. While they found that tropical
AGB accumulation was not significantly affected by previous land use, annual precipitation, or soil
order, this is probably more a reflection of sample size limitations in meta-analyses than reality;
evidence for this assertion can be found in related studies with different methodologies that have found
these factors to significantly affect forest growth (e.g., [12,17,71]). Given the current state of the
literature, even if a meta-analysis were to determine accurate global averages, these global estimates
are unlikely to be applicable to any given individual site considering the variety of mechanisms that
affect C sequestration potential and realization [4].

4. Research, Policy, and Decision-Making Implications

4.1. Research Implications

A thorough understanding of site carbon sequestration potential requires pairing carbon

accumulation measurements with detailed knowledge of pre-restoration biophysical site
characteristics. Accurate projections of future dynamics on marginal land will depend on detailed
models that include most or all of the key factors discussed above. Thus, future studies should strive to
provide detailed site quality data [60], as well as detailed land use history information (although it
must be recognized that even extremely detailed land use history data is only coarsely indicative of
Forests 2014, 5 855

pre-restoration biophysical attributes). In addition, because pre-restoration site quality interacts with
subsequent management, these management actions should be investigated and reported. From our
review of the literature, we have compiled a table of parameters deemed significant to assessing site C
sequestration potential. This list combines the comparative parameters used in multiple studies and
meta-analyses into one complete table. Given sites with similar geography and climate, the following
data presented in Table 2 is necessary to assess the carbon sequestration potential of one site
over another.

Table 2. Data necessary to assess a site’s carbon sequestration potential. These data should
ideally be collected in all sequestration studies and at sites prior to the implementation of
any carbon land management scheme.
Data Category Relevant Variables
A range of physical (slope, aspect, altitude) and soil characteristics
Site Quality including pH, bulk density, N, Ca, P, K, micronutrient levels, soil
texture and structure, stone content (pre-restoration data)
Type, duration, number of cycles, number of types, fertilization, remnant
Land Use History
vegetation, other detailed practices (pre-restoration data)
Fertilizer applications during planting/natural succession, ploughing,
Management tilling, mycorrhizal inoculations, thinning, planting of N-fixing species,
any other strategies for assisted regeneration (data during restoration)
Sources: Feldpausch [58], Hughes et al. [59], Gamboa et al. [72], Silver et al. [12], Bonner et al. [13],
Kasel and Bennett [55], Peichl et al. [65], Paul et al. [17], Holl [64].

Ideally, site quality data should be collected directly (i.e., via soil testing), and management actions
should be tracked and recorded as they occur. However, this of course may not possible if research
begins after the recovery process has started. Potential ways to gather these important data include
increased use of archival data and interviews with local inhabitants, as well as increased use of
remotely sensed data [60]. However, it should be noted that while interviews with locals can fill
important gaps in knowledge, interviews can be time consuming and provide information of varying
quality [24]. In addition, while the use of aerial imaging will provide some aspects of land use history
(e.g., [19]), it also leaves out many important site quality details such as fertilization regimes and
intrinsic soil properties.

4.2. Policy and Decision-Making Implications

While the quantification of existing forest C stocks is error prone due to insufficient methodology
and the failure to include all C pools, studies on future forest C recovery contain even more uncertainty
because they involve future projections. Ziegler et al.’s [73] meta-analysis of 250 studies in Southeast
Asia concludes that it is ―virtually impossible‖ with our current state of knowledge to predict how
land-use changes affect total ecosystem C stocks. Uncertainty at local scales and within individual
pools may have little consequence for global C estimates, provided there is no systematic bias.
However, even if this is true, these broad estimates have little relevance to any individual site, and it is
at this local scale that land use decisions are actualized. A wide range of factors influences local
C stocks and dynamics (e.g., climate, tree species, latitude, precipitation, proximity to forest), but these
Forests 2014, 5 856

relationships have been poorly quantified to date [4]. As such, reliable C recovery projections for a
particular parcel of land under a particular management regimen will likely remain elusive for some
time to come.
Given these limitations, it is important for policy and decision makers to acknowledge these
uncertainties when making decisions about the future use of marginal lands. Policy and decision
makers would be better served by making land use decision in a broader context that acknowledges
climate mitigation as one of the many environmental services forest landscapes may provide. For
example, there are many other services provided by forests such as watershed protection, pollination,
and biodiversity [58,74] and while the most carbon-positive land use transition may not maximize
these non-carbon benefits [73], focusing on non-carbon benefits as well as carbon benefits may lead to
more resilient systems. Illustrative examples include Thompson et al. [75] finding that monocultures
are less resilient than a system containing a greater species richness, making a less C rich forest more
favorable in the face of climate change and disturbance. Further, while afforested monocultures may
be credited with more C sequestration [12], they are often more susceptible to disease and therefore
result in less resilient ecosystems [76].

5. Conclusions

In conclusion, C recovery would be better understood with increased data collection on

pre-restoration site quality, prior land use history, and management practices as well as increased
methodological standardization. In addition, given the current and likely future uncertainty in C
dynamics, we recommend carbon mitigation potential should not be the only environmental service
driving land use decisions on marginal lands.

Acknowledgments

This work was supported by funds from the Energy Biosciences Institute.

Author Contributions

Tara L. DiRocco performed the research for the paper and wrote the first draft of the paper.
Matthew D. Potts and Benjamin S. Ramage conceived the research framework and questions.
Samuel G. Evans advised on the agricultural portion of the manuscript. All co-authors assisted with
writing and revising successive drafts.

Conflicts of Interest

The authors declare no conflict of interest.

References

1. Campbell, J.E.; Lobell, D.B.; Genova, R.C.; Field, C.B. The global potential of bioenergy on
abandoned agriculture lands. Environ. Sci. Technol. 2008, 42, 5791–5794.
2. Cai, X.; Zhang, X.; Wang, D. Land availability for biofuel production. Environ. Sci. Technol.
2011, 45, 334–339.
Forests 2014, 5 857

3. Righelato, R.; Spracklen, D.V. Environment. Carbon mitigation by biofuels or by saving and
restoring forests? Science 2007, 317, 902.
4. Kauffman, J.B.; Hughes, R.F.; Heider, C. Carbon pool and biomass dynamics associated with
deforestation, land use, and agricultural abandonment in the neotropics. Ecol. Appl. 2009, 19,
1211–1222.
5. Omeja, P.A.; Obua, J.; Rwetsiba, A.; Chapman, C.A. Biomass accumulation in tropical lands with
different disturbance histories: Contrasts within one landscape and across regions. For. Ecol.
Manag. 2012, 269, 293–300.
6. Ngo, K.M.; Turner, B.L.; Muller-Landau, H.C.; Davies, S.J.; Larjavaara, M.; Nik Hassan, N.F.B.;
Lum, S. Carbon stocks in primary and secondary tropical forests in Singapore. For. Ecol. Manag.
2013, 296, 81–89.
7. Scown, C.D.; Nazaroff, W.W.; Mishra, U.; Strogen, B.; Lobscheid, A.B.; Masanet, E.;
Santero, N.J.; Horvath, A.; Mckone, T.E. Lifecycle greenhouse gas implications of US national
scenarios for cellulosic ethanol production. Environ. Res. Lett. 2012, 7, 1–9.
8. Edwards, R.; Larive, J.F.; Rickeard, D.; Weindorf, W. Well-to-Tank Version 4 JEC Well-to-
Wheels Analysis; European Commission Joint Research Center: Brussels, Belgium, 2013; p. 134.
9. Fischer, G.; Hiznyik, E.; Priler, S.; Shana, M.; van Velthuizen, H. Global Agro-Ecological
Assessment for Agriculture in the 21st Century: Methodology and Results; IIASA and FAO:
Laxenburg, Austria; Rome, Italy, 2002.
10. Ramankutty, N.; Gibbs, H.K.; Achard, F.; Defries, R.; Foley, J.A.; Houghton, R.A. Challenges to
estimating carbon emissions from tropical deforestation. Glob. Chang. Biol. 2007, 13, 51–66.
11. Woodall, C.W.; Domke, G.M.; Riley, K.L.; Oswalt, C.M.; Crocker, S.J.; Yohe, G.W.
A framework for assessing global change risks to forest carbon stocks in the United States.
PLoS One 2013, 8, e73222.
12. Silver, W.L.; Ostertag, R.; Lugo, A.E. The Potential for carbon sequestration through
reforestation of abandoned tropical agricultural and pasture lands. Restor. Ecol. 2000, 8, 394–407.
13. Bonner, M.T.L.; Schmidt, S.; Shoo, L.P. A meta-analytical global comparison of aboveground
biomass accumulation between tropical secondary forests and monoculture plantations. For. Ecol.
Manag. 2013, 291, 73–86.
14. Cuevas, E.; Brown, S.; Lugo, A.E. Above- and belowground organic matter storage and
production in tropical pine plantation and a paired broadleaf secondary forest. Plant Soil 1991,
135, 257–268.
15. Davis, M.R.; Condron, L.M. Impact of grassland afforestation on soil carbon in New Zealand:
A review of paired-site studies. Aust. J. Soil Res. 2002, 40, 675–690.
16. Rhoades, C.C.; Eckert, G.E.; Coleman, D.C. Soil carbon differences among forest, agriculture,
and secondary vegetation in Lower Montane Ecuador. Ecol. Appl. 2000, 10, 497–505.
17. Paul, K.I.; Polglase, P.J.; Nyakuengama, J.G.; Khanna, P.K. Change in soil carbon following
afforestation. For. Ecol. Manag. 2002, 168, 241–257.
18. Powers, J.S. Changes in soil carbon and nitrogen after contrasting land-use transitions in
northeastern Costa Rica. Ecosystems 2004, 7, 134–146.
19. Schedlbauer, J.L.; Kavanagh, K.L. Soil carbon dynamics in a chronosequence of secondary
forests in northeastern Costa Rica. For. Ecol. Manag. 2008, 255, 1326–1335.
Forests 2014, 5 858

20. Lima, A.M.N.; Silva, I.R.; Neves, J.C.L.; Novais, R.F.; Barros, N.F.; Mendonça, E.S.; Smyth,
T.J.; Moreira, M.S.; Leite, F.P. Soil organic carbon dynamics following afforestation of degraded
pastures with eucalyptus in southeastern Brazil. For. Ecol. Manag. 2006, 235, 219–231.
21. Sauer, T.J.; James, D.E.; Cambardella, C.A.; Hernandez-Ramirez, G. Soil properties following
reforestation or afforestation of marginal cropland. Plant Soil 2012, 360, 375–390.
22. Li, D.; Niu, S.; Luo, Y. Global patterns of the dynamics of soil carbon and nitrogen stocks
following afforestation: A meta-analysis. New Phytol. 2012, 195, 172–181.
23. Don, A.; Schumacher, J.; Freibauer, A. Impact of tropical land-use change on soil organic carbon
stocks-a meta-analysis. Glob. Chang. Biol. 2011, 17, 1658–1670.
24. Neumann-Cosel, L.; Zimmermann, B.; Hall, J.S.; van Breugel, M.; Elsenbeer, H. Soil carbon
dynamics under young tropical secondary forests on former pastures—A case study from Panama.
For. Ecol. Manag. 2011, 261, 1625–1633.
25. Vesterdal, L.; Rosenqvist, L. Carbon sequestration in soil and biomass following afforestation:
Experiences from oak and Norway spruce chronosequences in Denmark, Sweden and the
Netherlands. Eff. Afforestation 2007, 1, 19–51.
26. Wang, J.; Epstein, H.E. Estimating carbon source-sink transition during secondary succession in a
Virginia valley. Plant Soil 2012, 362, 135–147.
27. Sang, P.M.; Lamb, D.; Bonner, M.; Schmidt, S. Carbon sequestration and soil fertility of tropical
tree plantations and secondary forest established on degraded land. Plant Soil 2012, 362,
187–200.
28. Sierra, C.A.; del Valle, J.I.; Orrego, S.A.; Moreno, F.H.; Harmon, M.E.; Zapata, M.; Coorado,
G.J.; Herrera, M.A.; Lara, W.; Restrepo, D.E.; et al. Total carbon stocks in a tropical forest
landscape of the Porce region, Colombia. For. Ecol. Manag. 2007, 243, 299–309.
29. Dierkes, C. Accounting for carbon in Great Lakes Forests. OSU Climate Change Webinar Series;
OSU Climate Change: Columbus, OH, USA, 2011; pp. 1–3.
30. Hooker, T.D.; Compton, J.E. Forest ecosystem carbon and nitrogen accumulation during the first
century after agricultural abandonment. Ecol. Appl. 2003, 13, 299–313.
31. Jaramillo, V.J.; Ahedo-Hernndez, R.; Kauffman, J.B. Root biomass and carbon in a tropical
evergreen forest of Mexico: Changes with secondary succession and forest conversion to pasture.
J. Trop. Ecol. 2003, 19, 457–464.
32. Nunes, L.; Patricio, M.; Tomé, J.; Tomé, M. Carbon and nutrients stocks in even-aged maritime
pine stands from Portugal. For. Syst. 2010, 19, 434–448.
33. Robinson, D. Implications of a large global root biomass for carbon sink estimates and for soil
carbon dynamics. Proc. R. Soc. B: Biol. Sci. 2007, 274, 2753–2759.
34. Preece, N.D.; Crowley, G.M.; Lawes, M.J.; van Oosterzee, P. Comparing above-ground biomass
among forest types in the Wet Tropics: Small stems and plantation types matter in carbon
accounting. For. Ecol. Manag. 2012, 264, 228–237.
35. Brown, S. Measuring carbon in forests: Current status and future challenges. Environ. Pollut.
2002, 116, 363–372.
36. Chave, J.; Andalo, C.; Brown, S.; Cairns, M.A.; Chambers, J.Q.; Eamus, D.; Folster, H.;
Fromard, F.; Higuchi, N.; Kira, T.; et al. Tree allometry and improved estimation and balance in
tropical forests of carbon stocks. Oecologia 2005, 145, 87–99.
Forests 2014, 5 859

37. Martin, A.R.; Thomas, S.C. A reassessment of carbon content in tropical trees. PLoS One 2011,
6, e23533.
38. Elias, M.; Potvin, C. Assessing inter- and intra-specific variation in trunk carbon concentration for
32 neotropical tree species. Can. J. For. Res. 2003, 33, 1039–1045.
39. Gifford, R.M. Carbon Contents of Above-Ground Tissues of Forest and Woodland Tress; National
Carbon Accounting System, Technical Report No 22; Australian Greenhouse Office: Canberra,
Australia, September 2000.
40. Milakovsky, B.; Frey, B.; James, T. Carbon dynamics in the boreal forest. In Managing Forest
Carbon in a Changing Climate; Ashton, M.S., Tyrrell, M.L., Spalding, D., Gentry, B., Eds.;
Springer: New York, NY, USA, 2012; p. 115.
41. Robinson, D. Scaling the depths: Below-ground allocation in plants, forests and biomes.
Funct. Ecol. 2004, 18, 290–295.
42. Taylor, B.N.; Beidler, K.V.; Cooper, E.R.; Strand, A.E.; Pritchard, S.G. Sampling volume in root
studies: The pitfalls of under-sampling exposed using accumulation curves. Ecol. Lett. 2013, 16,
862–869.
43. Mokany, K.; Raison, R.J.; Prokushkin, A.S. Critical analysis of root: Shoot ratios in terrestrial
biomes. Glob. Chang. Biol. 2006, 12, 84–96.
44. Ping, X.; Zhou, G.; Zhuang, Q.; Wang, Y.; Zuo, W.; Shi, G.; Lin, X.; Wang, Y. Effects of sample
size and position from monolith and core methods on the estimation of total root biomass in a
temperate grassland ecosystem in Inner Mongolia. Geoderma 2010, 155, 262–268.
45. Aalde, H.; Gonzalez, P.; Gytarsky, M.; Krug, T.; Kurz, W.A.; Ogle, S.; Somogyi, Z. Forest land.
In 2006 IPCC Guidelines for National Greenhouse Gas Inventories (Vol. 4: Agriculture, Forestry,
and Other Land Use); Eggleston, H.S., Buendia, L., Miwa, K., Ngara, T., Tanabe, K., Eds.;
Institute for Global Environmental Strategies: Hayama, Japan, 2006.
46. MacDicken, K. A Guide to Monitoring Carbon Storage in Forestry and Agroforestry Projects;
Winrock International Institute for Agricultural Development: Little Rock, AR, USA, 1997.
47. Center for Tropical Forest Science. Soil Carbon Sampling Protocol; Center for Tropical Forest
Science, Smithsonian Tropical Research Institute: Washington, DC, USA, 2010.
48. Fahey, T.J.; Woodbury, P.B.; Battles, J.J.; Goodale, C.L.; Hamburg, S.; Ollinger, S.; Woodall,
C.W. Forest carbon storage: Ecology, management, and policy. Front. Ecol. Environ. 2010, 8,
245–252.
49. Lemenih, M.; Olsson, M.; Karltun, E. Comparison of soil attributes under Cupressus lusitanica
and Eucalyptus saligna established on abandoned farmlands with continuously cropped farmlands
and natural forest in Ethiopia. For. Ecol. Manag. 2004, 195, 57–67.
50. Anderson, K.J.; Allen, A.P.; Gillooly, J.F.; Brown, J.H. Temperature-dependence of biomass
accumulation rates during secondary succession. Ecol. Lett. 2006, 9, 673–682.
51. Holl, K.D.; Aide, T.M. When and where to actively restore ecosystems? For. Ecol. Manag.
2011, 261, 1558–1563.
52. Niu, X.; Duiker, S.W. Carbon sequestration potential by afforestation of marginal agricultural
land in the Midwestern U.S. For. Ecol. Manag. 2006, 223, 415–427.
Forests 2014, 5 860

53. Letcher, S.G.; Chazdon, R.L. Rapid recovery of biomass, species richness, and species
composition in a forest chronosequence in northeastern Costa Rica. Biotropica 2009, 41,
608–617.
54. Van Rooyen, M.W.; van Rooyen, N.; Stoffberg, G.H. Carbon sequestration potential of
post-mining reforestation activities on the KwaZulu-Natal coast, South Africa. Forestry 2012, 86,
211–223.
55. Kasel, S.; Bennett, L.T. Land-use history, forest conversion, and soil organic carbon in pine
plantations and native forests of south eastern Australia. Geoderma 2007, 137, 401–413.
56. Guo, L.B.; Gifford, R.M. Soil carbon stocks and land use change: A meta analysis. Glob.
Chang. Biol. 2002, 8, 345–360.
57. Kotto-same, J.; Woomer, P.L.; Appolinaire, M.; Louis, Z. Carbon dynamics in slash-and-bum
agriculture and land use alternatives of the humid forest zone in Cameroon. Agric. Ecosyst.
Environ. 1997, 809, 245–256.
58. Feldpausch, T.R.; Rondon, M.A.; Fernandes, E.C.M.; Riha, S.J.; Wandelli, E. Carbon and nutrient
accumulation in secondary forests regenerating on pastures in central Amazonia. Ecol. Appl.
2004, 14, S164–S176.
59. Hughes, R.F.; Kauffmann, J.B.; Jaramillo, V.J. Biomass, carbon, and nutrient dynamics of
secondary forests in a humid tropical region of Mexico. Ecology 1999, 80, 1892–1907.
60. De Koning, G.H.J.; Veldkamp, E.; López-Ulloa, M. Quantification of carbon sequestration in
soils following pasture to forest conversion in northwestern Ecuador. Glob. Biogeochem. Cycles
2003, 17, doi:10.1029/2003GB002099.
61. Powers, J.S.; Veldkamp, E. Regional variation in soil carbon and d13C in forests and pastures of
northeastern Costa Rica. Biogeochemistry 2005, 72, 315–336.
62. Klanderud, K.; Mbolatiana, H.Z.H.; Vololomboahangy, M.N.; Radimbison, M.A.; Roger, E.;
Totland, Ø.; Rajeriarison, C. Recovery of plant species richness and composition after
slash-and-burn agriculture in a tropical rainforest in Madagascar. Biodivers. Conserv. 2009, 19,
187–204.
63. Fearnside, P.M.; Guimaraes, W.M. Carbon uptake by secondary forests in Brazilian Amazonia.
For. Ecol. Manag. 1996, 80, 35–46.
64. Holl, K.D. Old field vegetation succession in the Neotropics. In Old Fields; Cramer, V.A.,
Hobbs, R.J., Eds.; Island Press: Washington, DC, USA, 2007; pp. 93–117.
65. Peichl, M.; Arain, M.A.; Brodeur, J.J. Age effects on carbon fluxes in temperate pine forests.
Agric. For. Meteorol. 2010, 150, 1090–1101.
66. Fonseca, W.; Rey Benayas, J.M.; Alice, F.E. Carbon accumulation in the biomass and soil of
different aged secondary forests in the humid tropics of Costa Rica. For. Ecol. Manag. 2011, 262,
1400–1408.
67. Black, K.; Byrne, K.A.; Mencuccini, M.; Tobin, B.; Nieuwenhuis, M.; Reidy, B.; Bolger, T.;
Saiz, G.; Green, C.; Farrell, E.T.; et al. Carbon stock and stock changes across a Sitka spruce
chronosequence on surface-water gley soils. Forestry 2009, 82, 255–272.
68. Lasco, R.D.; Pulhin, F.B. Carbon budgets of forest ecosystems in the Philippines. J. Eviron.
Sci. Manag. 2009, 12, 1–13.
Forests 2014, 5 861

69. Macedo, M.O.; Resende, A.S.; Garcia, P.C.; Boddey, R.M.; Jantalia, C.P.; Urquiaga, S.;
Campello, E.F.C.; Franco, A.A. Changes in soil C and N stocks and nutrient dynamics 13 years
after recovery of degraded land using leguminous nitrogen-fixing trees. For. Ecol. Manag. 2008,
255, 1516–1524.
70. Kanowski, J.; Catterall, C.P. Carbon stocks in above-ground biomass of monoculture plantations,
mixed species plantations and environmental restoration plantings in north-east Australia.
Ecol. Manag. Restor. 2010, 11, 119–126.
71. Mekuria, W.; Veldkamp, E.; Corre, M.D. Restoration of ecosystem carbon stocks following
exclosure establishment in communal. For. Range Wildland Soils 2011, 75, 246–256.
72. Gamboa, A.M.; Hidalgo, C.; de León, F.; Etchevers, J.D.; Gallardo, J.F.; Campo, J. Nutrient
addition differentially affects soil carbon sequestration in secondary tropical dry forests:
Early- versus Late-Succession Stages. Restor. Ecol. 2010, 18, 252–260.
73. Ziegler, A.D.; Phelps, J.; Yuen, J.Q.; Webb, E.L.; Lawrence, D.; Fox, J.M.; Bruun, T.B.; Leisz,
S.J.; Ryan, C.; Dressler, W.; et al. Carbon outcomes of major land-cover transitions in SE Asia:
Great uncertainties and REDD+ policy implications. Glob. Chang. Biol. 2012, 18, 3087–3099.
74. Hall, J.M.; Holt, T.; Daniels, A.E.; Balthazar, V.; Lambin, E.F. Trade-offs between tree cover,
carbon storage and floristic biodiversity in reforesting landscapes. Landsc. Ecol. 2012, 27,
1135–1147.
75. Thompson, I.; Mackey, B.; McNulty, S.; Mosseler, A. Forest resilience, biodiversity, and climate
change. A synthesis of the biodiversity/resilience/stability relationship in forest ecosystems.
In CBD Technical Series; Secretariat of the Convention on Biological Diversity: Montreal, QC,
Canada, 2009.
76. Hobbs, R.J.; Higgs, E.; Harris, J.A. Novel ecosystems: Implications for conservation and
restoration. Trends Ecol. Evol. 2009, 24, 599–605.

© 2014 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access article
distributed under the terms and conditions of the Creative Commons Attribution license
(http://creativecommons.org/licenses/by/3.0/).